From 3384c7c7641b44987e35eadbc9df6c16a0520159 Mon Sep 17 00:00:00 2001 From: Vadim Fedorenko Date: Fri, 10 Sep 2021 01:04:09 +0300 Subject: selftests/bpf: Test new __sk_buff field hwtstamp Analogous to the gso_segs selftests introduced in commit d9ff286a0f59 ("bpf: allow BPF programs access skb_shared_info->gso_segs field"). Signed-off-by: Vadim Fedorenko Signed-off-by: Daniel Borkmann Acked-by: Martin KaFai Lau Link: https://lore.kernel.org/bpf/20210909220409.8804-3-vfedorenko@novek.ru --- lib/test_bpf.c | 1 + 1 file changed, 1 insertion(+) (limited to 'lib') diff --git a/lib/test_bpf.c b/lib/test_bpf.c index 830a18ecffc8..0018d51b93b0 100644 --- a/lib/test_bpf.c +++ b/lib/test_bpf.c @@ -8800,6 +8800,7 @@ static __init struct sk_buff *build_test_skb(void) skb_shinfo(skb[0])->gso_type |= SKB_GSO_DODGY; skb_shinfo(skb[0])->gso_segs = 0; skb_shinfo(skb[0])->frag_list = skb[1]; + skb_shinfo(skb[0])->hwtstamps.hwtstamp = 1000; /* adjust skb[0]'s len */ skb[0]->len += skb[1]->len; -- cgit v1.2.3 From 8988bacd6045adf39719e5057e43170f83bd1709 Mon Sep 17 00:00:00 2001 From: Qu Wenruo Date: Tue, 31 Aug 2021 17:30:44 +0800 Subject: kobject: unexport kobject_create() in kobject.h The function kobject_create() is only used by one caller, kobject_create_and_add(), no other driver uses it, nor is exported to other modules. However it's still exported in kobject.h, and can sometimes confuse users of kobject.h. Since all users should call kobject_create_and_add(), or if extra attributes are needed, should alloc the memory manually then call kobject_init_and_add(). Signed-off-by: Qu Wenruo Link: https://lore.kernel.org/r/20210831093044.110729-1-wqu@suse.com Signed-off-by: Greg Kroah-Hartman --- include/linux/kobject.h | 1 - lib/kobject.c | 2 +- 2 files changed, 1 insertion(+), 2 deletions(-) (limited to 'lib') diff --git a/include/linux/kobject.h b/include/linux/kobject.h index ea30529fba08..efd56f990a46 100644 --- a/include/linux/kobject.h +++ b/include/linux/kobject.h @@ -101,7 +101,6 @@ int kobject_init_and_add(struct kobject *kobj, extern void kobject_del(struct kobject *kobj); -extern struct kobject * __must_check kobject_create(void); extern struct kobject * __must_check kobject_create_and_add(const char *name, struct kobject *parent); diff --git a/lib/kobject.c b/lib/kobject.c index ea53b30cf483..4a56f519139d 100644 --- a/lib/kobject.c +++ b/lib/kobject.c @@ -777,7 +777,7 @@ static struct kobj_type dynamic_kobj_ktype = { * call to kobject_put() and not kfree(), as kobject_init() has * already been called on this structure. */ -struct kobject *kobject_create(void) +static struct kobject *kobject_create(void) { struct kobject *kobj; -- cgit v1.2.3 From 4a7e1e5fc294687a8941fa3eeb4a7e8539ca5e2f Mon Sep 17 00:00:00 2001 From: Nathan Chancellor Date: Wed, 25 Aug 2021 13:38:59 -0700 Subject: crypto: sm4 - Do not change section of ck and sbox When building with clang and GNU as, there is a warning about ignored changed section attributes: /tmp/sm4-c916c8.s: Assembler messages: /tmp/sm4-c916c8.s:677: Warning: ignoring changed section attributes for .data..cacheline_aligned "static const" places the data in .rodata but __cacheline_aligned has the section attribute to place it in .data..cacheline_aligned, in addition to the aligned attribute. To keep the alignment but avoid attempting to change sections, use the ____cacheline_aligned attribute, which is just the aligned attribute. Fixes: 2b31277af577 ("crypto: sm4 - create SM4 library based on sm4 generic code") Link: https://github.com/ClangBuiltLinux/linux/issues/1441 Signed-off-by: Nathan Chancellor Reviewed-by: Tianjia Zhang Signed-off-by: Herbert Xu --- lib/crypto/sm4.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'lib') diff --git a/lib/crypto/sm4.c b/lib/crypto/sm4.c index 633b59fed9db..284e62576d0c 100644 --- a/lib/crypto/sm4.c +++ b/lib/crypto/sm4.c @@ -15,7 +15,7 @@ static const u32 fk[4] = { 0xa3b1bac6, 0x56aa3350, 0x677d9197, 0xb27022dc }; -static const u32 __cacheline_aligned ck[32] = { +static const u32 ____cacheline_aligned ck[32] = { 0x00070e15, 0x1c232a31, 0x383f464d, 0x545b6269, 0x70777e85, 0x8c939aa1, 0xa8afb6bd, 0xc4cbd2d9, 0xe0e7eef5, 0xfc030a11, 0x181f262d, 0x343b4249, @@ -26,7 +26,7 @@ static const u32 __cacheline_aligned ck[32] = { 0x10171e25, 0x2c333a41, 0x484f565d, 0x646b7279 }; -static const u8 __cacheline_aligned sbox[256] = { +static const u8 ____cacheline_aligned sbox[256] = { 0xd6, 0x90, 0xe9, 0xfe, 0xcc, 0xe1, 0x3d, 0xb7, 0x16, 0xb6, 0x14, 0xc2, 0x28, 0xfb, 0x2c, 0x05, 0x2b, 0x67, 0x9a, 0x76, 0x2a, 0xbe, 0x04, 0xc3, -- cgit v1.2.3 From db2b0c5d7b6f19b3c2cab08c531b65342eb5252b Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Thu, 24 Jun 2021 11:41:23 +0200 Subject: objtool: Support pv_opsindirect calls for noinstr Normally objtool will now follow indirect calls; there is no need. However, this becomes a problem with noinstr validation; if there's an indirect call from noinstr code, we very much need to know it is to another noinstr function. Luckily there aren't many indirect calls in entry code with the obvious exception of paravirt. As such, noinstr validation didn't work with paravirt kernels. In order to track pv_ops[] call targets, objtool reads the static pv_ops[] tables as well as direct assignments to the pv_ops[] array, provided the compiler makes them a single instruction like: bf87: 48 c7 05 00 00 00 00 00 00 00 00 movq $0x0,0x0(%rip) bf92 bf8a: R_X86_64_PC32 pv_ops+0x268 There are, as of yet, no warnings for when this goes wrong :/ Using the functions found with the above means, all pv_ops[] calls are now subject to noinstr validation. Signed-off-by: Peter Zijlstra (Intel) Link: https://lore.kernel.org/r/20210624095149.118815755@infradead.org --- lib/Kconfig.debug | 2 +- tools/objtool/arch/x86/decode.c | 34 ++++++- tools/objtool/check.c | 151 +++++++++++++++++++++++++++++--- tools/objtool/include/objtool/arch.h | 2 +- tools/objtool/include/objtool/elf.h | 1 + tools/objtool/include/objtool/objtool.h | 9 ++ tools/objtool/objtool.c | 22 +++++ 7 files changed, 208 insertions(+), 13 deletions(-) (limited to 'lib') diff --git a/lib/Kconfig.debug b/lib/Kconfig.debug index ed4a31e34098..63a4735cef66 100644 --- a/lib/Kconfig.debug +++ b/lib/Kconfig.debug @@ -458,7 +458,7 @@ config STACK_VALIDATION config VMLINUX_VALIDATION bool - depends on STACK_VALIDATION && DEBUG_ENTRY && !PARAVIRT + depends on STACK_VALIDATION && DEBUG_ENTRY default y config VMLINUX_MAP diff --git a/tools/objtool/arch/x86/decode.c b/tools/objtool/arch/x86/decode.c index 340a3dce94a0..3172983bf808 100644 --- a/tools/objtool/arch/x86/decode.c +++ b/tools/objtool/arch/x86/decode.c @@ -20,6 +20,7 @@ #include #include #include +#include #include static int is_x86_64(const struct elf *elf) @@ -102,12 +103,13 @@ unsigned long arch_jump_destination(struct instruction *insn) #define rm_is_mem(reg) (mod_is_mem() && !is_RIP() && rm_is(reg)) #define rm_is_reg(reg) (mod_is_reg() && modrm_rm == (reg)) -int arch_decode_instruction(const struct elf *elf, const struct section *sec, +int arch_decode_instruction(struct objtool_file *file, const struct section *sec, unsigned long offset, unsigned int maxlen, unsigned int *len, enum insn_type *type, unsigned long *immediate, struct list_head *ops_list) { + const struct elf *elf = file->elf; struct insn insn; int x86_64, ret; unsigned char op1, op2, @@ -544,6 +546,36 @@ int arch_decode_instruction(const struct elf *elf, const struct section *sec, *type = INSN_RETURN; break; + case 0xc7: /* mov imm, r/m */ + if (!noinstr) + break; + + if (insn.length == 3+4+4 && !strncmp(sec->name, ".init.text", 10)) { + struct reloc *immr, *disp; + struct symbol *func; + int idx; + + immr = find_reloc_by_dest(elf, (void *)sec, offset+3); + disp = find_reloc_by_dest(elf, (void *)sec, offset+7); + + if (!immr || strcmp(immr->sym->name, "pv_ops")) + break; + + idx = (immr->addend + 8) / sizeof(void *); + + func = disp->sym; + if (disp->sym->type == STT_SECTION) + func = find_symbol_by_offset(disp->sym->sec, disp->addend); + if (!func) { + WARN("no func for pv_ops[]"); + return -1; + } + + objtool_pv_add(file, idx, func); + } + + break; + case 0xcf: /* iret */ /* * Handle sync_core(), which has an IRET to self. diff --git a/tools/objtool/check.c b/tools/objtool/check.c index c6f206fee8ff..84e59a97bab6 100644 --- a/tools/objtool/check.c +++ b/tools/objtool/check.c @@ -382,7 +382,7 @@ static int decode_instructions(struct objtool_file *file) insn->sec = sec; insn->offset = offset; - ret = arch_decode_instruction(file->elf, sec, offset, + ret = arch_decode_instruction(file, sec, offset, sec->len - offset, &insn->len, &insn->type, &insn->immediate, @@ -420,6 +420,82 @@ err: return ret; } +/* + * Read the pv_ops[] .data table to find the static initialized values. + */ +static int add_pv_ops(struct objtool_file *file, const char *symname) +{ + struct symbol *sym, *func; + unsigned long off, end; + struct reloc *rel; + int idx; + + sym = find_symbol_by_name(file->elf, symname); + if (!sym) + return 0; + + off = sym->offset; + end = off + sym->len; + for (;;) { + rel = find_reloc_by_dest_range(file->elf, sym->sec, off, end - off); + if (!rel) + break; + + func = rel->sym; + if (func->type == STT_SECTION) + func = find_symbol_by_offset(rel->sym->sec, rel->addend); + + idx = (rel->offset - sym->offset) / sizeof(unsigned long); + + objtool_pv_add(file, idx, func); + + off = rel->offset + 1; + if (off > end) + break; + } + + return 0; +} + +/* + * Allocate and initialize file->pv_ops[]. + */ +static int init_pv_ops(struct objtool_file *file) +{ + static const char *pv_ops_tables[] = { + "pv_ops", + "xen_cpu_ops", + "xen_irq_ops", + "xen_mmu_ops", + NULL, + }; + const char *pv_ops; + struct symbol *sym; + int idx, nr; + + if (!noinstr) + return 0; + + file->pv_ops = NULL; + + sym = find_symbol_by_name(file->elf, "pv_ops"); + if (!sym) + return 0; + + nr = sym->len / sizeof(unsigned long); + file->pv_ops = calloc(sizeof(struct pv_state), nr); + if (!file->pv_ops) + return -1; + + for (idx = 0; idx < nr; idx++) + INIT_LIST_HEAD(&file->pv_ops[idx].targets); + + for (idx = 0; (pv_ops = pv_ops_tables[idx]); idx++) + add_pv_ops(file, pv_ops); + + return 0; +} + static struct instruction *find_last_insn(struct objtool_file *file, struct section *sec) { @@ -893,6 +969,9 @@ static struct reloc *insn_reloc(struct objtool_file *file, struct instruction *i return NULL; if (!insn->reloc) { + if (!file) + return NULL; + insn->reloc = find_reloc_by_dest_range(file->elf, insn->sec, insn->offset, insn->len); if (!insn->reloc) { @@ -1882,6 +1961,10 @@ static int decode_sections(struct objtool_file *file) mark_rodata(file); + ret = init_pv_ops(file); + if (ret) + return ret; + ret = decode_instructions(file); if (ret) return ret; @@ -2663,20 +2746,64 @@ static inline bool func_uaccess_safe(struct symbol *func) static inline const char *call_dest_name(struct instruction *insn) { + static char pvname[16]; + struct reloc *rel; + int idx; + if (insn->call_dest) return insn->call_dest->name; + rel = insn_reloc(NULL, insn); + if (rel && !strcmp(rel->sym->name, "pv_ops")) { + idx = (rel->addend / sizeof(void *)); + snprintf(pvname, sizeof(pvname), "pv_ops[%d]", idx); + return pvname; + } + return "{dynamic}"; } -static inline bool noinstr_call_dest(struct symbol *func) +static bool pv_call_dest(struct objtool_file *file, struct instruction *insn) +{ + struct symbol *target; + struct reloc *rel; + int idx; + + rel = insn_reloc(file, insn); + if (!rel || strcmp(rel->sym->name, "pv_ops")) + return false; + + idx = (arch_dest_reloc_offset(rel->addend) / sizeof(void *)); + + if (file->pv_ops[idx].clean) + return true; + + file->pv_ops[idx].clean = true; + + list_for_each_entry(target, &file->pv_ops[idx].targets, pv_target) { + if (!target->sec->noinstr) { + WARN("pv_ops[%d]: %s", idx, target->name); + file->pv_ops[idx].clean = false; + } + } + + return file->pv_ops[idx].clean; +} + +static inline bool noinstr_call_dest(struct objtool_file *file, + struct instruction *insn, + struct symbol *func) { /* * We can't deal with indirect function calls at present; * assume they're instrumented. */ - if (!func) + if (!func) { + if (file->pv_ops) + return pv_call_dest(file, insn); + return false; + } /* * If the symbol is from a noinstr section; we good. @@ -2695,10 +2822,12 @@ static inline bool noinstr_call_dest(struct symbol *func) return false; } -static int validate_call(struct instruction *insn, struct insn_state *state) +static int validate_call(struct objtool_file *file, + struct instruction *insn, + struct insn_state *state) { if (state->noinstr && state->instr <= 0 && - !noinstr_call_dest(insn->call_dest)) { + !noinstr_call_dest(file, insn, insn->call_dest)) { WARN_FUNC("call to %s() leaves .noinstr.text section", insn->sec, insn->offset, call_dest_name(insn)); return 1; @@ -2719,7 +2848,9 @@ static int validate_call(struct instruction *insn, struct insn_state *state) return 0; } -static int validate_sibling_call(struct instruction *insn, struct insn_state *state) +static int validate_sibling_call(struct objtool_file *file, + struct instruction *insn, + struct insn_state *state) { if (has_modified_stack_frame(insn, state)) { WARN_FUNC("sibling call from callable instruction with modified stack frame", @@ -2727,7 +2858,7 @@ static int validate_sibling_call(struct instruction *insn, struct insn_state *st return 1; } - return validate_call(insn, state); + return validate_call(file, insn, state); } static int validate_return(struct symbol *func, struct instruction *insn, struct insn_state *state) @@ -2880,7 +3011,7 @@ static int validate_branch(struct objtool_file *file, struct symbol *func, case INSN_CALL: case INSN_CALL_DYNAMIC: - ret = validate_call(insn, &state); + ret = validate_call(file, insn, &state); if (ret) return ret; @@ -2899,7 +3030,7 @@ static int validate_branch(struct objtool_file *file, struct symbol *func, case INSN_JUMP_CONDITIONAL: case INSN_JUMP_UNCONDITIONAL: if (is_sibling_call(insn)) { - ret = validate_sibling_call(insn, &state); + ret = validate_sibling_call(file, insn, &state); if (ret) return ret; @@ -2921,7 +3052,7 @@ static int validate_branch(struct objtool_file *file, struct symbol *func, case INSN_JUMP_DYNAMIC: case INSN_JUMP_DYNAMIC_CONDITIONAL: if (is_sibling_call(insn)) { - ret = validate_sibling_call(insn, &state); + ret = validate_sibling_call(file, insn, &state); if (ret) return ret; } diff --git a/tools/objtool/include/objtool/arch.h b/tools/objtool/include/objtool/arch.h index 6f482ae2d7d8..589ff58426ab 100644 --- a/tools/objtool/include/objtool/arch.h +++ b/tools/objtool/include/objtool/arch.h @@ -69,7 +69,7 @@ struct instruction; void arch_initial_func_cfi_state(struct cfi_init_state *state); -int arch_decode_instruction(const struct elf *elf, const struct section *sec, +int arch_decode_instruction(struct objtool_file *file, const struct section *sec, unsigned long offset, unsigned int maxlen, unsigned int *len, enum insn_type *type, unsigned long *immediate, diff --git a/tools/objtool/include/objtool/elf.h b/tools/objtool/include/objtool/elf.h index e34395047530..c3857fadee7a 100644 --- a/tools/objtool/include/objtool/elf.h +++ b/tools/objtool/include/objtool/elf.h @@ -57,6 +57,7 @@ struct symbol { struct symbol *pfunc, *cfunc, *alias; bool uaccess_safe; bool static_call_tramp; + struct list_head pv_target; }; struct reloc { diff --git a/tools/objtool/include/objtool/objtool.h b/tools/objtool/include/objtool/objtool.h index 24fa83634de4..f99fbc6078d5 100644 --- a/tools/objtool/include/objtool/objtool.h +++ b/tools/objtool/include/objtool/objtool.h @@ -14,6 +14,11 @@ #define __weak __attribute__((weak)) +struct pv_state { + bool clean; + struct list_head targets; +}; + struct objtool_file { struct elf *elf; struct list_head insn_list; @@ -25,10 +30,14 @@ struct objtool_file { unsigned long jl_short, jl_long; unsigned long jl_nop_short, jl_nop_long; + + struct pv_state *pv_ops; }; struct objtool_file *objtool_open_read(const char *_objname); +void objtool_pv_add(struct objtool_file *file, int idx, struct symbol *func); + int check(struct objtool_file *file); int orc_dump(const char *objname); int orc_create(struct objtool_file *file); diff --git a/tools/objtool/objtool.c b/tools/objtool/objtool.c index e21db8bce493..c90c7084e45a 100644 --- a/tools/objtool/objtool.c +++ b/tools/objtool/objtool.c @@ -135,6 +135,28 @@ struct objtool_file *objtool_open_read(const char *_objname) return &file; } +void objtool_pv_add(struct objtool_file *f, int idx, struct symbol *func) +{ + if (!noinstr) + return; + + if (!f->pv_ops) { + WARN("paravirt confusion"); + return; + } + + /* + * These functions will be patched into native code, + * see paravirt_patch(). + */ + if (!strcmp(func->name, "_paravirt_nop") || + !strcmp(func->name, "_paravirt_ident_64")) + return; + + list_add(&func->pv_target, &f->pv_ops[idx].targets); + f->pv_ops[idx].clean = false; +} + static void cmd_usage(void) { unsigned int i, longest = 0; -- cgit v1.2.3 From 12235da8c80a1f9909008e4ca6036d5772b81192 Mon Sep 17 00:00:00 2001 From: Maarten Lankhorst Date: Thu, 9 Sep 2021 11:32:18 +0200 Subject: kernel/locking: Add context to ww_mutex_trylock() i915 will soon gain an eviction path that trylock a whole lot of locks for eviction, getting dmesg failures like below: BUG: MAX_LOCK_DEPTH too low! turning off the locking correctness validator. depth: 48 max: 48! 48 locks held by i915_selftest/5776: #0: ffff888101a79240 (&dev->mutex){....}-{3:3}, at: __driver_attach+0x88/0x160 #1: ffffc900009778c0 (reservation_ww_class_acquire){+.+.}-{0:0}, at: i915_vma_pin.constprop.63+0x39/0x1b0 [i915] #2: ffff88800cf74de8 (reservation_ww_class_mutex){+.+.}-{3:3}, at: i915_vma_pin.constprop.63+0x5f/0x1b0 [i915] #3: ffff88810c7f9e38 (&vm->mutex/1){+.+.}-{3:3}, at: i915_vma_pin_ww+0x1c4/0x9d0 [i915] #4: ffff88810bad5768 (reservation_ww_class_mutex){+.+.}-{3:3}, at: i915_gem_evict_something+0x110/0x860 [i915] #5: ffff88810bad60e8 (reservation_ww_class_mutex){+.+.}-{3:3}, at: i915_gem_evict_something+0x110/0x860 [i915] ... #46: ffff88811964d768 (reservation_ww_class_mutex){+.+.}-{3:3}, at: i915_gem_evict_something+0x110/0x860 [i915] #47: ffff88811964e0e8 (reservation_ww_class_mutex){+.+.}-{3:3}, at: i915_gem_evict_something+0x110/0x860 [i915] INFO: lockdep is turned off. Fixing eviction to nest into ww_class_acquire is a high priority, but it requires a rework of the entire driver, which can only be done one step at a time. As an intermediate solution, add an acquire context to ww_mutex_trylock, which allows us to do proper nesting annotations on the trylocks, making the above lockdep splat disappear. This is also useful in regulator_lock_nested, which may avoid dropping regulator_nesting_mutex in the uncontended path, so use it there. TTM may be another user for this, where we could lock a buffer in a fastpath with list locks held, without dropping all locks we hold. [peterz: rework actual ww_mutex_trylock() implementations] Signed-off-by: Maarten Lankhorst Signed-off-by: Peter Zijlstra (Intel) Link: https://lkml.kernel.org/r/YUBGPdDDjKlxAuXJ@hirez.programming.kicks-ass.net --- drivers/gpu/drm/drm_modeset_lock.c | 2 +- drivers/regulator/core.c | 2 +- include/linux/dma-resv.h | 2 +- include/linux/ww_mutex.h | 15 +------ kernel/locking/mutex.c | 41 ++++++++++++++++++ kernel/locking/test-ww_mutex.c | 86 ++++++++++++++++++++++++++++---------- kernel/locking/ww_rt_mutex.c | 25 +++++++++++ lib/locking-selftest.c | 2 +- 8 files changed, 137 insertions(+), 38 deletions(-) (limited to 'lib') diff --git a/drivers/gpu/drm/drm_modeset_lock.c b/drivers/gpu/drm/drm_modeset_lock.c index fcfe1a03c4a1..bf8a6e823a15 100644 --- a/drivers/gpu/drm/drm_modeset_lock.c +++ b/drivers/gpu/drm/drm_modeset_lock.c @@ -248,7 +248,7 @@ static inline int modeset_lock(struct drm_modeset_lock *lock, if (ctx->trylock_only) { lockdep_assert_held(&ctx->ww_ctx); - if (!ww_mutex_trylock(&lock->mutex)) + if (!ww_mutex_trylock(&lock->mutex, NULL)) return -EBUSY; else return 0; diff --git a/drivers/regulator/core.c b/drivers/regulator/core.c index ca6caba8a191..f4d441b1a8bf 100644 --- a/drivers/regulator/core.c +++ b/drivers/regulator/core.c @@ -145,7 +145,7 @@ static inline int regulator_lock_nested(struct regulator_dev *rdev, mutex_lock(®ulator_nesting_mutex); - if (ww_ctx || !ww_mutex_trylock(&rdev->mutex)) { + if (!ww_mutex_trylock(&rdev->mutex, ww_ctx)) { if (rdev->mutex_owner == current) rdev->ref_cnt++; else diff --git a/include/linux/dma-resv.h b/include/linux/dma-resv.h index e1ca2080a1ff..39fefb86780b 100644 --- a/include/linux/dma-resv.h +++ b/include/linux/dma-resv.h @@ -173,7 +173,7 @@ static inline int dma_resv_lock_slow_interruptible(struct dma_resv *obj, */ static inline bool __must_check dma_resv_trylock(struct dma_resv *obj) { - return ww_mutex_trylock(&obj->lock); + return ww_mutex_trylock(&obj->lock, NULL); } /** diff --git a/include/linux/ww_mutex.h b/include/linux/ww_mutex.h index 29db736af86d..bb763085479a 100644 --- a/include/linux/ww_mutex.h +++ b/include/linux/ww_mutex.h @@ -28,12 +28,10 @@ #ifndef CONFIG_PREEMPT_RT #define WW_MUTEX_BASE mutex #define ww_mutex_base_init(l,n,k) __mutex_init(l,n,k) -#define ww_mutex_base_trylock(l) mutex_trylock(l) #define ww_mutex_base_is_locked(b) mutex_is_locked((b)) #else #define WW_MUTEX_BASE rt_mutex #define ww_mutex_base_init(l,n,k) __rt_mutex_init(l,n,k) -#define ww_mutex_base_trylock(l) rt_mutex_trylock(l) #define ww_mutex_base_is_locked(b) rt_mutex_base_is_locked(&(b)->rtmutex) #endif @@ -339,17 +337,8 @@ ww_mutex_lock_slow_interruptible(struct ww_mutex *lock, extern void ww_mutex_unlock(struct ww_mutex *lock); -/** - * ww_mutex_trylock - tries to acquire the w/w mutex without acquire context - * @lock: mutex to lock - * - * Trylocks a mutex without acquire context, so no deadlock detection is - * possible. Returns 1 if the mutex has been acquired successfully, 0 otherwise. - */ -static inline int __must_check ww_mutex_trylock(struct ww_mutex *lock) -{ - return ww_mutex_base_trylock(&lock->base); -} +extern int __must_check ww_mutex_trylock(struct ww_mutex *lock, + struct ww_acquire_ctx *ctx); /*** * ww_mutex_destroy - mark a w/w mutex unusable diff --git a/kernel/locking/mutex.c b/kernel/locking/mutex.c index d456579d0952..2fede72b6af5 100644 --- a/kernel/locking/mutex.c +++ b/kernel/locking/mutex.c @@ -94,6 +94,9 @@ static inline unsigned long __owner_flags(unsigned long owner) return owner & MUTEX_FLAGS; } +/* + * Returns: __mutex_owner(lock) on failure or NULL on success. + */ static inline struct task_struct *__mutex_trylock_common(struct mutex *lock, bool handoff) { unsigned long owner, curr = (unsigned long)current; @@ -736,6 +739,44 @@ __ww_mutex_lock(struct mutex *lock, unsigned int state, unsigned int subclass, return __mutex_lock_common(lock, state, subclass, NULL, ip, ww_ctx, true); } +/** + * ww_mutex_trylock - tries to acquire the w/w mutex with optional acquire context + * @ww: mutex to lock + * @ww_ctx: optional w/w acquire context + * + * Trylocks a mutex with the optional acquire context; no deadlock detection is + * possible. Returns 1 if the mutex has been acquired successfully, 0 otherwise. + * + * Unlike ww_mutex_lock, no deadlock handling is performed. However, if a @ctx is + * specified, -EALREADY handling may happen in calls to ww_mutex_trylock. + * + * A mutex acquired with this function must be released with ww_mutex_unlock. + */ +int ww_mutex_trylock(struct ww_mutex *ww, struct ww_acquire_ctx *ww_ctx) +{ + if (!ww_ctx) + return mutex_trylock(&ww->base); + + MUTEX_WARN_ON(ww->base.magic != &ww->base); + + /* + * Reset the wounded flag after a kill. No other process can + * race and wound us here, since they can't have a valid owner + * pointer if we don't have any locks held. + */ + if (ww_ctx->acquired == 0) + ww_ctx->wounded = 0; + + if (__mutex_trylock(&ww->base)) { + ww_mutex_set_context_fastpath(ww, ww_ctx); + mutex_acquire_nest(&ww->base.dep_map, 0, 1, &ww_ctx->dep_map, _RET_IP_); + return 1; + } + + return 0; +} +EXPORT_SYMBOL(ww_mutex_trylock); + #ifdef CONFIG_DEBUG_LOCK_ALLOC void __sched mutex_lock_nested(struct mutex *lock, unsigned int subclass) diff --git a/kernel/locking/test-ww_mutex.c b/kernel/locking/test-ww_mutex.c index 3e82f449b4ff..d63ac411f367 100644 --- a/kernel/locking/test-ww_mutex.c +++ b/kernel/locking/test-ww_mutex.c @@ -16,6 +16,15 @@ static DEFINE_WD_CLASS(ww_class); struct workqueue_struct *wq; +#ifdef CONFIG_DEBUG_WW_MUTEX_SLOWPATH +#define ww_acquire_init_noinject(a, b) do { \ + ww_acquire_init((a), (b)); \ + (a)->deadlock_inject_countdown = ~0U; \ + } while (0) +#else +#define ww_acquire_init_noinject(a, b) ww_acquire_init((a), (b)) +#endif + struct test_mutex { struct work_struct work; struct ww_mutex mutex; @@ -36,7 +45,7 @@ static void test_mutex_work(struct work_struct *work) wait_for_completion(&mtx->go); if (mtx->flags & TEST_MTX_TRY) { - while (!ww_mutex_trylock(&mtx->mutex)) + while (!ww_mutex_trylock(&mtx->mutex, NULL)) cond_resched(); } else { ww_mutex_lock(&mtx->mutex, NULL); @@ -109,19 +118,38 @@ static int test_mutex(void) return 0; } -static int test_aa(void) +static int test_aa(bool trylock) { struct ww_mutex mutex; struct ww_acquire_ctx ctx; int ret; + const char *from = trylock ? "trylock" : "lock"; ww_mutex_init(&mutex, &ww_class); ww_acquire_init(&ctx, &ww_class); - ww_mutex_lock(&mutex, &ctx); + if (!trylock) { + ret = ww_mutex_lock(&mutex, &ctx); + if (ret) { + pr_err("%s: initial lock failed!\n", __func__); + goto out; + } + } else { + if (!ww_mutex_trylock(&mutex, &ctx)) { + pr_err("%s: initial trylock failed!\n", __func__); + goto out; + } + } - if (ww_mutex_trylock(&mutex)) { - pr_err("%s: trylocked itself!\n", __func__); + if (ww_mutex_trylock(&mutex, NULL)) { + pr_err("%s: trylocked itself without context from %s!\n", __func__, from); + ww_mutex_unlock(&mutex); + ret = -EINVAL; + goto out; + } + + if (ww_mutex_trylock(&mutex, &ctx)) { + pr_err("%s: trylocked itself with context from %s!\n", __func__, from); ww_mutex_unlock(&mutex); ret = -EINVAL; goto out; @@ -129,17 +157,17 @@ static int test_aa(void) ret = ww_mutex_lock(&mutex, &ctx); if (ret != -EALREADY) { - pr_err("%s: missed deadlock for recursing, ret=%d\n", - __func__, ret); + pr_err("%s: missed deadlock for recursing, ret=%d from %s\n", + __func__, ret, from); if (!ret) ww_mutex_unlock(&mutex); ret = -EINVAL; goto out; } + ww_mutex_unlock(&mutex); ret = 0; out: - ww_mutex_unlock(&mutex); ww_acquire_fini(&ctx); return ret; } @@ -150,7 +178,7 @@ struct test_abba { struct ww_mutex b_mutex; struct completion a_ready; struct completion b_ready; - bool resolve; + bool resolve, trylock; int result; }; @@ -160,8 +188,13 @@ static void test_abba_work(struct work_struct *work) struct ww_acquire_ctx ctx; int err; - ww_acquire_init(&ctx, &ww_class); - ww_mutex_lock(&abba->b_mutex, &ctx); + ww_acquire_init_noinject(&ctx, &ww_class); + if (!abba->trylock) + ww_mutex_lock(&abba->b_mutex, &ctx); + else + WARN_ON(!ww_mutex_trylock(&abba->b_mutex, &ctx)); + + WARN_ON(READ_ONCE(abba->b_mutex.ctx) != &ctx); complete(&abba->b_ready); wait_for_completion(&abba->a_ready); @@ -181,7 +214,7 @@ static void test_abba_work(struct work_struct *work) abba->result = err; } -static int test_abba(bool resolve) +static int test_abba(bool trylock, bool resolve) { struct test_abba abba; struct ww_acquire_ctx ctx; @@ -192,12 +225,18 @@ static int test_abba(bool resolve) INIT_WORK_ONSTACK(&abba.work, test_abba_work); init_completion(&abba.a_ready); init_completion(&abba.b_ready); + abba.trylock = trylock; abba.resolve = resolve; schedule_work(&abba.work); - ww_acquire_init(&ctx, &ww_class); - ww_mutex_lock(&abba.a_mutex, &ctx); + ww_acquire_init_noinject(&ctx, &ww_class); + if (!trylock) + ww_mutex_lock(&abba.a_mutex, &ctx); + else + WARN_ON(!ww_mutex_trylock(&abba.a_mutex, &ctx)); + + WARN_ON(READ_ONCE(abba.a_mutex.ctx) != &ctx); complete(&abba.a_ready); wait_for_completion(&abba.b_ready); @@ -249,7 +288,7 @@ static void test_cycle_work(struct work_struct *work) struct ww_acquire_ctx ctx; int err, erra = 0; - ww_acquire_init(&ctx, &ww_class); + ww_acquire_init_noinject(&ctx, &ww_class); ww_mutex_lock(&cycle->a_mutex, &ctx); complete(cycle->a_signal); @@ -581,7 +620,9 @@ static int stress(int nlocks, int nthreads, unsigned int flags) static int __init test_ww_mutex_init(void) { int ncpus = num_online_cpus(); - int ret; + int ret, i; + + printk(KERN_INFO "Beginning ww mutex selftests\n"); wq = alloc_workqueue("test-ww_mutex", WQ_UNBOUND, 0); if (!wq) @@ -591,17 +632,19 @@ static int __init test_ww_mutex_init(void) if (ret) return ret; - ret = test_aa(); + ret = test_aa(false); if (ret) return ret; - ret = test_abba(false); + ret = test_aa(true); if (ret) return ret; - ret = test_abba(true); - if (ret) - return ret; + for (i = 0; i < 4; i++) { + ret = test_abba(i & 1, i & 2); + if (ret) + return ret; + } ret = test_cycle(ncpus); if (ret) @@ -619,6 +662,7 @@ static int __init test_ww_mutex_init(void) if (ret) return ret; + printk(KERN_INFO "All ww mutex selftests passed\n"); return 0; } diff --git a/kernel/locking/ww_rt_mutex.c b/kernel/locking/ww_rt_mutex.c index 3f1fff7d2780..0e00205cf467 100644 --- a/kernel/locking/ww_rt_mutex.c +++ b/kernel/locking/ww_rt_mutex.c @@ -9,6 +9,31 @@ #define WW_RT #include "rtmutex.c" +int ww_mutex_trylock(struct ww_mutex *lock, struct ww_acquire_ctx *ww_ctx) +{ + struct rt_mutex *rtm = &lock->base; + + if (!ww_ctx) + return rt_mutex_trylock(rtm); + + /* + * Reset the wounded flag after a kill. No other process can + * race and wound us here, since they can't have a valid owner + * pointer if we don't have any locks held. + */ + if (ww_ctx->acquired == 0) + ww_ctx->wounded = 0; + + if (__rt_mutex_trylock(&rtm->rtmutex)) { + ww_mutex_set_context_fastpath(lock, ww_ctx); + mutex_acquire_nest(&rtm->dep_map, 0, 1, ww_ctx->dep_map, _RET_IP_); + return 1; + } + + return 0; +} +EXPORT_SYMBOL(ww_mutex_trylock); + static int __sched __ww_rt_mutex_lock(struct ww_mutex *lock, struct ww_acquire_ctx *ww_ctx, unsigned int state, unsigned long ip) diff --git a/lib/locking-selftest.c b/lib/locking-selftest.c index 161108e5d2fe..71652e1c397c 100644 --- a/lib/locking-selftest.c +++ b/lib/locking-selftest.c @@ -258,7 +258,7 @@ static void init_shared_classes(void) #define WWAF(x) ww_acquire_fini(x) #define WWL(x, c) ww_mutex_lock(x, c) -#define WWT(x) ww_mutex_trylock(x) +#define WWT(x) ww_mutex_trylock(x, NULL) #define WWL1(x) ww_mutex_lock(x, NULL) #define WWU(x) ww_mutex_unlock(x) -- cgit v1.2.3 From 04e85bbf71c9072dcf0ad9a7150495d72461105c Mon Sep 17 00:00:00 2001 From: Alexey Dobriyan Date: Mon, 2 Aug 2021 23:43:15 +0300 Subject: isystem: delete global -isystem compile option Further isolate kernel from userspace, prevent accidental inclusion of undesireable headers, mainly float.h and stdatomic.h. nds32 keeps -isystem globally due to intrinsics used in entrenched header. -isystem is selectively reenabled for some files, again, for intrinsics. Compile tested on: hexagon-defconfig hexagon-allmodconfig alpha-allmodconfig alpha-allnoconfig alpha-defconfig arm64-allmodconfig arm64-allnoconfig arm64-defconfig arm-am200epdkit arm-aspeed_g4 arm-aspeed_g5 arm-assabet arm-at91_dt arm-axm55xx arm-badge4 arm-bcm2835 arm-cerfcube arm-clps711x arm-cm_x300 arm-cns3420vb arm-colibri_pxa270 arm-colibri_pxa300 arm-collie arm-corgi arm-davinci_all arm-dove arm-ep93xx arm-eseries_pxa arm-exynos arm-ezx arm-footbridge arm-gemini arm-h3600 arm-h5000 arm-hackkit arm-hisi arm-imote2 arm-imx_v4_v5 arm-imx_v6_v7 arm-integrator arm-iop32x arm-ixp4xx arm-jornada720 arm-keystone arm-lart arm-lpc18xx arm-lpc32xx arm-lpd270 arm-lubbock arm-magician arm-mainstone arm-milbeaut_m10v arm-mini2440 arm-mmp2 arm-moxart arm-mps2 arm-multi_v4t arm-multi_v5 arm-multi_v7 arm-mv78xx0 arm-mvebu_v5 arm-mvebu_v7 arm-mxs arm-neponset arm-netwinder arm-nhk8815 arm-omap1 arm-omap2plus arm-orion5x arm-oxnas_v6 arm-palmz72 arm-pcm027 arm-pleb arm-pxa arm-pxa168 arm-pxa255-idp arm-pxa3xx arm-pxa910 arm-qcom arm-realview arm-rpc arm-s3c2410 arm-s3c6400 arm-s5pv210 arm-sama5 arm-shannon arm-shmobile arm-simpad arm-socfpga arm-spear13xx arm-spear3xx arm-spear6xx arm-spitz arm-stm32 arm-sunxi arm-tct_hammer arm-tegra arm-trizeps4 arm-u8500 arm-versatile arm-vexpress arm-vf610m4 arm-viper arm-vt8500_v6_v7 arm-xcep arm-zeus csky-allmodconfig csky-allnoconfig csky-defconfig h8300-edosk2674 h8300-h8300h-sim h8300-h8s-sim i386-allmodconfig i386-allnoconfig i386-defconfig ia64-allmodconfig ia64-allnoconfig ia64-bigsur ia64-generic ia64-gensparse ia64-tiger ia64-zx1 m68k-amcore m68k-amiga m68k-apollo m68k-atari m68k-bvme6000 m68k-hp300 m68k-m5208evb m68k-m5249evb m68k-m5272c3 m68k-m5275evb m68k-m5307c3 m68k-m5407c3 m68k-m5475evb m68k-mac m68k-multi m68k-mvme147 m68k-mvme16x m68k-q40 m68k-stmark2 m68k-sun3 m68k-sun3x microblaze-allmodconfig microblaze-allnoconfig microblaze-mmu mips-ar7 mips-ath25 mips-ath79 mips-bcm47xx mips-bcm63xx mips-bigsur mips-bmips_be mips-bmips_stb mips-capcella mips-cavium_octeon mips-ci20 mips-cobalt mips-cu1000-neo mips-cu1830-neo mips-db1xxx mips-decstation mips-decstation_64 mips-decstation_r4k mips-e55 mips-fuloong2e mips-gcw0 mips-generic mips-gpr mips-ip22 mips-ip27 mips-ip28 mips-ip32 mips-jazz mips-jmr3927 mips-lemote2f mips-loongson1b mips-loongson1c mips-loongson2k mips-loongson3 mips-malta mips-maltaaprp mips-malta_kvm mips-malta_qemu_32r6 mips-maltasmvp mips-maltasmvp_eva mips-maltaup mips-maltaup_xpa mips-mpc30x mips-mtx1 mips-nlm_xlp mips-nlm_xlr mips-omega2p mips-pic32mzda mips-pistachio mips-qi_lb60 mips-rb532 mips-rbtx49xx mips-rm200 mips-rs90 mips-rt305x mips-sb1250_swarm mips-tb0219 mips-tb0226 mips-tb0287 mips-vocore2 mips-workpad mips-xway nds32-allmodconfig nds32-allnoconfig nds32-defconfig nios2-10m50 nios2-3c120 nios2-allmodconfig nios2-allnoconfig openrisc-allmodconfig openrisc-allnoconfig openrisc-or1klitex openrisc-or1ksim openrisc-simple_smp parisc-allnoconfig parisc-generic-32bit parisc-generic-64bit powerpc-acadia powerpc-adder875 powerpc-akebono powerpc-amigaone powerpc-arches powerpc-asp8347 powerpc-bamboo powerpc-bluestone powerpc-canyonlands powerpc-cell powerpc-chrp32 powerpc-cm5200 powerpc-currituck powerpc-ebony powerpc-eiger powerpc-ep8248e powerpc-ep88xc powerpc-fsp2 powerpc-g5 powerpc-gamecube powerpc-ge_imp3a powerpc-holly powerpc-icon powerpc-iss476-smp powerpc-katmai powerpc-kilauea powerpc-klondike powerpc-kmeter1 powerpc-ksi8560 powerpc-linkstation powerpc-lite5200b powerpc-makalu powerpc-maple powerpc-mgcoge powerpc-microwatt powerpc-motionpro powerpc-mpc512x powerpc-mpc5200 powerpc-mpc7448_hpc2 powerpc-mpc8272_ads powerpc-mpc8313_rdb powerpc-mpc8315_rdb powerpc-mpc832x_mds powerpc-mpc832x_rdb powerpc-mpc834x_itx powerpc-mpc834x_itxgp powerpc-mpc834x_mds powerpc-mpc836x_mds powerpc-mpc836x_rdk powerpc-mpc837x_mds powerpc-mpc837x_rdb powerpc-mpc83xx powerpc-mpc8540_ads powerpc-mpc8560_ads powerpc-mpc85xx_cds powerpc-mpc866_ads powerpc-mpc885_ads powerpc-mvme5100 powerpc-obs600 powerpc-pasemi powerpc-pcm030 powerpc-pmac32 powerpc-powernv powerpc-ppa8548 powerpc-ppc40x powerpc-ppc44x powerpc-ppc64 powerpc-ppc64e powerpc-ppc6xx powerpc-pq2fads powerpc-ps3 powerpc-pseries powerpc-rainier powerpc-redwood powerpc-sam440ep powerpc-sbc8548 powerpc-sequoia powerpc-skiroot powerpc-socrates powerpc-storcenter powerpc-stx_gp3 powerpc-taishan powerpc-tqm5200 powerpc-tqm8540 powerpc-tqm8541 powerpc-tqm8548 powerpc-tqm8555 powerpc-tqm8560 powerpc-tqm8xx powerpc-walnut powerpc-warp powerpc-wii powerpc-xes_mpc85xx riscv-allmodconfig riscv-allnoconfig riscv-nommu_k210 riscv-nommu_k210_sdcard riscv-nommu_virt riscv-rv32 s390-allmodconfig s390-allnoconfig s390-debug s390-zfcpdump sh-ap325rxa sh-apsh4a3a sh-apsh4ad0a sh-dreamcast sh-ecovec24 sh-ecovec24-romimage sh-edosk7705 sh-edosk7760 sh-espt sh-hp6xx sh-j2 sh-kfr2r09 sh-kfr2r09-romimage sh-landisk sh-lboxre2 sh-magicpanelr2 sh-microdev sh-migor sh-polaris sh-r7780mp sh-r7785rp sh-rsk7201 sh-rsk7203 sh-rsk7264 sh-rsk7269 sh-rts7751r2d1 sh-rts7751r2dplus sh-sdk7780 sh-sdk7786 sh-se7206 sh-se7343 sh-se7619 sh-se7705 sh-se7712 sh-se7721 sh-se7722 sh-se7724 sh-se7750 sh-se7751 sh-se7780 sh-secureedge5410 sh-sh03 sh-sh2007 sh-sh7710voipgw sh-sh7724_generic sh-sh7757lcr sh-sh7763rdp sh-sh7770_generic sh-sh7785lcr sh-sh7785lcr_32bit sh-shmin sh-shx3 sh-titan sh-ul2 sh-urquell sparc-allmodconfig sparc-allnoconfig sparc-sparc32 sparc-sparc64 um-i386-allmodconfig um-i386-allnoconfig um-i386-defconfig um-x86_64-allmodconfig um-x86_64-allnoconfig x86_64-allmodconfig x86_64-allnoconfig x86_64-defconfig xtensa-allmodconfig xtensa-allnoconfig xtensa-audio_kc705 xtensa-cadence_csp xtensa-common xtensa-generic_kc705 xtensa-iss xtensa-nommu_kc705 xtensa-smp_lx200 xtensa-virt xtensa-xip_kc705 Tested-by: Nathan Chancellor # build (hexagon) Signed-off-by: Alexey Dobriyan Acked-by: Ard Biesheuvel Signed-off-by: Masahiro Yamada --- Makefile | 2 +- arch/arm64/lib/Makefile | 2 ++ arch/nds32/Makefile | 2 ++ arch/powerpc/lib/Makefile | 2 ++ crypto/Makefile | 2 ++ lib/raid6/Makefile | 4 ++++ 6 files changed, 13 insertions(+), 1 deletion(-) (limited to 'lib') diff --git a/Makefile b/Makefile index 5e7c1d854441..ee5896261d2f 100644 --- a/Makefile +++ b/Makefile @@ -981,7 +981,7 @@ KBUILD_CFLAGS += -falign-functions=64 endif # arch Makefile may override CC so keep this after arch Makefile is included -NOSTDINC_FLAGS += -nostdinc -isystem $(shell $(CC) -print-file-name=include) +NOSTDINC_FLAGS += -nostdinc # warn about C99 declaration after statement KBUILD_CFLAGS += -Wdeclaration-after-statement diff --git a/arch/arm64/lib/Makefile b/arch/arm64/lib/Makefile index 0941180a86d3..29490be2546b 100644 --- a/arch/arm64/lib/Makefile +++ b/arch/arm64/lib/Makefile @@ -9,6 +9,8 @@ ifeq ($(CONFIG_KERNEL_MODE_NEON), y) obj-$(CONFIG_XOR_BLOCKS) += xor-neon.o CFLAGS_REMOVE_xor-neon.o += -mgeneral-regs-only CFLAGS_xor-neon.o += -ffreestanding +# Enable +CFLAGS_xor-neon.o += -isystem $(shell $(CC) -print-file-name=include) endif lib-$(CONFIG_ARCH_HAS_UACCESS_FLUSHCACHE) += uaccess_flushcache.o diff --git a/arch/nds32/Makefile b/arch/nds32/Makefile index ccdca7142020..c17fc3a755c3 100644 --- a/arch/nds32/Makefile +++ b/arch/nds32/Makefile @@ -9,6 +9,8 @@ endif # Avoid generating FPU instructions arch-y += -mno-ext-fpu-sp -mno-ext-fpu-dp -mfloat-abi=soft +# Enable +KBUILD_CFLAGS += -isystem $(shell $(CC) -print-file-name=include) KBUILD_CFLAGS += $(call cc-option, -mno-sched-prolog-epilog) KBUILD_CFLAGS += -mcmodel=large diff --git a/arch/powerpc/lib/Makefile b/arch/powerpc/lib/Makefile index 99a7c9132422..9e5d0f413b71 100644 --- a/arch/powerpc/lib/Makefile +++ b/arch/powerpc/lib/Makefile @@ -65,5 +65,7 @@ obj-$(CONFIG_FTR_FIXUP_SELFTEST) += feature-fixups-test.o obj-$(CONFIG_ALTIVEC) += xor_vmx.o xor_vmx_glue.o CFLAGS_xor_vmx.o += -maltivec $(call cc-option,-mabi=altivec) +# Enable +CFLAGS_xor_vmx.o += -isystem $(shell $(CC) -print-file-name=include) obj-$(CONFIG_PPC64) += $(obj64-y) diff --git a/crypto/Makefile b/crypto/Makefile index c633f15a0481..429c4d57458c 100644 --- a/crypto/Makefile +++ b/crypto/Makefile @@ -119,6 +119,8 @@ CFLAGS_aegis128-neon-inner.o += $(aegis128-cflags-y) CFLAGS_REMOVE_aegis128-neon-inner.o += -mgeneral-regs-only aegis128-$(CONFIG_CRYPTO_AEGIS128_SIMD) += aegis128-neon.o aegis128-neon-inner.o endif +# Enable +CFLAGS_aegis128-neon-inner.o += -isystem $(shell $(CC) -print-file-name=include) obj-$(CONFIG_CRYPTO_PCRYPT) += pcrypt.o obj-$(CONFIG_CRYPTO_CRYPTD) += cryptd.o diff --git a/lib/raid6/Makefile b/lib/raid6/Makefile index c770570bfe4f..45e17619422b 100644 --- a/lib/raid6/Makefile +++ b/lib/raid6/Makefile @@ -14,6 +14,8 @@ hostprogs += mktables ifeq ($(CONFIG_ALTIVEC),y) altivec_flags := -maltivec $(call cc-option,-mabi=altivec) +# Enable +altivec_flags += -isystem $(shell $(CC) -print-file-name=include) ifdef CONFIG_CC_IS_CLANG # clang ppc port does not yet support -maltivec when -msoft-float is @@ -34,6 +36,8 @@ endif # ARM/NEON intrinsics in a non C99-compliant environment (such as the kernel) ifeq ($(CONFIG_KERNEL_MODE_NEON),y) NEON_FLAGS := -ffreestanding +# Enable +NEON_FLAGS += -isystem $(shell $(CC) -print-file-name=include) ifeq ($(ARCH),arm) NEON_FLAGS += -march=armv7-a -mfloat-abi=softfp -mfpu=neon endif -- cgit v1.2.3 From 3229b906fb35b63515f0c703b917357c83e1ea22 Mon Sep 17 00:00:00 2001 From: Thomas Zimmermann Date: Thu, 16 Sep 2021 20:15:57 +0200 Subject: lib: devres: Add managed arch_phys_wc_add() Add devm_arch_phys_wc_add() as managed wrapper around arch_phys_wc_add(). Useful for several graphics drivers that set framebuffer memory to write combining. v2: * fix typo in commit description Signed-off-by: Thomas Zimmermann Reviewed-by: Hans de Goede Link: https://patchwork.freedesktop.org/patch/msgid/20210916181601.9146-2-tzimmermann@suse.de --- include/linux/io.h | 2 ++ lib/devres.c | 36 ++++++++++++++++++++++++++++++++++++ 2 files changed, 38 insertions(+) (limited to 'lib') diff --git a/include/linux/io.h b/include/linux/io.h index 9595151d800d..fcd8ea79c5df 100644 --- a/include/linux/io.h +++ b/include/linux/io.h @@ -132,6 +132,8 @@ static inline int arch_phys_wc_index(int handle) #endif #endif +int devm_arch_phys_wc_add(struct device *dev, unsigned long base, unsigned long size); + enum { /* See memremap() kernel-doc for usage description... */ MEMREMAP_WB = 1 << 0, diff --git a/lib/devres.c b/lib/devres.c index b0e1c6702c71..24d4d849ff67 100644 --- a/lib/devres.c +++ b/lib/devres.c @@ -528,3 +528,39 @@ void pcim_iounmap_regions(struct pci_dev *pdev, int mask) } EXPORT_SYMBOL(pcim_iounmap_regions); #endif /* CONFIG_PCI */ + +static void devm_arch_phys_ac_add_release(struct device *dev, void *res) +{ + arch_phys_wc_del(*((int *)res)); +} + +/** + * devm_arch_phys_wc_add - Managed arch_phys_wc_add() + * @dev: Managed device + * @base: Memory base address + * @size: Size of memory range + * + * Adds a WC MTRR using arch_phys_wc_add() and sets up a release callback. + * See arch_phys_wc_add() for more information. + */ +int devm_arch_phys_wc_add(struct device *dev, unsigned long base, unsigned long size) +{ + int *mtrr; + int ret; + + mtrr = devres_alloc(devm_arch_phys_ac_add_release, sizeof(*mtrr), GFP_KERNEL); + if (!mtrr) + return -ENOMEM; + + ret = arch_phys_wc_add(base, size); + if (ret < 0) { + devres_free(mtrr); + return ret; + } + + *mtrr = ret; + devres_add(dev, mtrr); + + return ret; +} +EXPORT_SYMBOL(devm_arch_phys_wc_add); -- cgit v1.2.3 From c822310725ee41af663de2448094155d442ff871 Mon Sep 17 00:00:00 2001 From: Thomas Zimmermann Date: Thu, 16 Sep 2021 20:15:58 +0200 Subject: lib: devres: Add managed arch_io_reserve_memtype_wc() Add devm_arch_io_reserve_memtype_wc() as managed wrapper around arch_io_reserve_memtype_wc(). Useful for several graphics drivers that set framebuffer memory to write combining. v2: * fix typo in commit description Signed-off-by: Thomas Zimmermann Reviewed-by: Hans de Goede Link: https://patchwork.freedesktop.org/patch/msgid/20210916181601.9146-3-tzimmermann@suse.de --- include/linux/io.h | 3 +++ lib/devres.c | 46 ++++++++++++++++++++++++++++++++++++++++++++++ 2 files changed, 49 insertions(+) (limited to 'lib') diff --git a/include/linux/io.h b/include/linux/io.h index fcd8ea79c5df..5fc800390fe4 100644 --- a/include/linux/io.h +++ b/include/linux/io.h @@ -168,4 +168,7 @@ static inline void arch_io_free_memtype_wc(resource_size_t base, } #endif +int devm_arch_io_reserve_memtype_wc(struct device *dev, resource_size_t start, + resource_size_t size); + #endif /* _LINUX_IO_H */ diff --git a/lib/devres.c b/lib/devres.c index 24d4d849ff67..14664bbb4875 100644 --- a/lib/devres.c +++ b/lib/devres.c @@ -564,3 +564,49 @@ int devm_arch_phys_wc_add(struct device *dev, unsigned long base, unsigned long return ret; } EXPORT_SYMBOL(devm_arch_phys_wc_add); + +struct arch_io_reserve_memtype_wc_devres { + resource_size_t start; + resource_size_t size; +}; + +static void devm_arch_io_free_memtype_wc_release(struct device *dev, void *res) +{ + const struct arch_io_reserve_memtype_wc_devres *this = res; + + arch_io_free_memtype_wc(this->start, this->size); +} + +/** + * devm_arch_io_reserve_memtype_wc - Managed arch_io_reserve_memtype_wc() + * @dev: Managed device + * @start: Memory base address + * @size: Size of memory range + * + * Reserves a memory range with WC caching using arch_io_reserve_memtype_wc() + * and sets up a release callback See arch_io_reserve_memtype_wc() for more + * information. + */ +int devm_arch_io_reserve_memtype_wc(struct device *dev, resource_size_t start, + resource_size_t size) +{ + struct arch_io_reserve_memtype_wc_devres *dr; + int ret; + + dr = devres_alloc(devm_arch_io_free_memtype_wc_release, sizeof(*dr), GFP_KERNEL); + if (!dr) + return -ENOMEM; + + ret = arch_io_reserve_memtype_wc(start, size); + if (ret < 0) { + devres_free(dr); + return ret; + } + + dr->start = start; + dr->size = size; + devres_add(dev, dr); + + return ret; +} +EXPORT_SYMBOL(devm_arch_io_reserve_memtype_wc); -- cgit v1.2.3 From cfecea6ead5f15880fc1fb31fc655f8be5cf7424 Mon Sep 17 00:00:00 2001 From: Kees Cook Date: Fri, 18 Jun 2021 10:57:38 -0700 Subject: lib/string: Move helper functions out of string.c The core functions of string.c are those that may be implemented by per-architecture functions, or overloaded by FORTIFY_SOURCE. As a result, it needs to be built with __NO_FORTIFY. Without this, macros will collide with function declarations. This was accidentally working due to -ffreestanding (on some architectures). Make this deterministic by explicitly setting __NO_FORTIFY and move all the helper functions into string_helpers.c so that they gain the fortification coverage they had been missing. Cc: Andrew Morton Cc: Nick Desaulniers Cc: Andy Lavr Cc: Nathan Chancellor Cc: Alexey Dobriyan Cc: Stephen Rothwell Cc: Bartosz Golaszewski Acked-by: Andy Shevchenko Signed-off-by: Kees Cook --- arch/arm/boot/compressed/string.c | 1 + arch/s390/lib/string.c | 3 + arch/x86/boot/compressed/misc.h | 2 + arch/x86/boot/compressed/pgtable_64.c | 2 + arch/x86/lib/string_32.c | 1 + lib/string.c | 210 +--------------------------------- lib/string_helpers.c | 193 +++++++++++++++++++++++++++++++ 7 files changed, 208 insertions(+), 204 deletions(-) (limited to 'lib') diff --git a/arch/arm/boot/compressed/string.c b/arch/arm/boot/compressed/string.c index 8c0fa276d994..fcc678fce045 100644 --- a/arch/arm/boot/compressed/string.c +++ b/arch/arm/boot/compressed/string.c @@ -5,6 +5,7 @@ * Small subset of simple string routines */ +#define __NO_FORTIFY #include /* diff --git a/arch/s390/lib/string.c b/arch/s390/lib/string.c index cfcdf76d6a95..392fb9f4f4db 100644 --- a/arch/s390/lib/string.c +++ b/arch/s390/lib/string.c @@ -8,6 +8,9 @@ */ #define IN_ARCH_STRING_C 1 +#ifndef __NO_FORTIFY +# define __NO_FORTIFY +#endif #include #include diff --git a/arch/x86/boot/compressed/misc.h b/arch/x86/boot/compressed/misc.h index 31139256859f..49bde196da9b 100644 --- a/arch/x86/boot/compressed/misc.h +++ b/arch/x86/boot/compressed/misc.h @@ -14,6 +14,8 @@ #undef CONFIG_KASAN #undef CONFIG_KASAN_GENERIC +#define __NO_FORTIFY + /* cpu_feature_enabled() cannot be used this early */ #define USE_EARLY_PGTABLE_L5 diff --git a/arch/x86/boot/compressed/pgtable_64.c b/arch/x86/boot/compressed/pgtable_64.c index 2a78746f5a4c..a1733319a22a 100644 --- a/arch/x86/boot/compressed/pgtable_64.c +++ b/arch/x86/boot/compressed/pgtable_64.c @@ -1,3 +1,5 @@ +// SPDX-License-Identifier: GPL-2.0 +#include "misc.h" #include #include #include diff --git a/arch/x86/lib/string_32.c b/arch/x86/lib/string_32.c index d15fdae9656e..53b3f202267c 100644 --- a/arch/x86/lib/string_32.c +++ b/arch/x86/lib/string_32.c @@ -11,6 +11,7 @@ * strings. */ +#define __NO_FORTIFY #include #include diff --git a/lib/string.c b/lib/string.c index b2de45a581f4..485777c9da83 100644 --- a/lib/string.c +++ b/lib/string.c @@ -6,20 +6,15 @@ */ /* - * stupid library routines.. The optimized versions should generally be found - * as inline code in + * This file should be used only for "library" routines that may have + * alternative implementations on specific architectures (generally + * found in ), or get overloaded by FORTIFY_SOURCE. + * (Specifically, this file is built with __NO_FORTIFY.) * - * These are buggy as well.. - * - * * Fri Jun 25 1999, Ingo Oeser - * - Added strsep() which will replace strtok() soon (because strsep() is - * reentrant and should be faster). Use only strsep() in new code, please. - * - * * Sat Feb 09 2002, Jason Thomas , - * Matthew Hawkins - * - Kissed strtok() goodbye + * Other helper functions should live in string_helpers.c. */ +#define __NO_FORTIFY #include #include #include @@ -238,40 +233,6 @@ ssize_t strscpy(char *dest, const char *src, size_t count) EXPORT_SYMBOL(strscpy); #endif -/** - * strscpy_pad() - Copy a C-string into a sized buffer - * @dest: Where to copy the string to - * @src: Where to copy the string from - * @count: Size of destination buffer - * - * Copy the string, or as much of it as fits, into the dest buffer. The - * behavior is undefined if the string buffers overlap. The destination - * buffer is always %NUL terminated, unless it's zero-sized. - * - * If the source string is shorter than the destination buffer, zeros - * the tail of the destination buffer. - * - * For full explanation of why you may want to consider using the - * 'strscpy' functions please see the function docstring for strscpy(). - * - * Returns: - * * The number of characters copied (not including the trailing %NUL) - * * -E2BIG if count is 0 or @src was truncated. - */ -ssize_t strscpy_pad(char *dest, const char *src, size_t count) -{ - ssize_t written; - - written = strscpy(dest, src, count); - if (written < 0 || written == count - 1) - return written; - - memset(dest + written + 1, 0, count - written - 1); - - return written; -} -EXPORT_SYMBOL(strscpy_pad); - /** * stpcpy - copy a string from src to dest returning a pointer to the new end * of dest, including src's %NUL-terminator. May overrun dest. @@ -514,46 +475,6 @@ char *strnchr(const char *s, size_t count, int c) EXPORT_SYMBOL(strnchr); #endif -/** - * skip_spaces - Removes leading whitespace from @str. - * @str: The string to be stripped. - * - * Returns a pointer to the first non-whitespace character in @str. - */ -char *skip_spaces(const char *str) -{ - while (isspace(*str)) - ++str; - return (char *)str; -} -EXPORT_SYMBOL(skip_spaces); - -/** - * strim - Removes leading and trailing whitespace from @s. - * @s: The string to be stripped. - * - * Note that the first trailing whitespace is replaced with a %NUL-terminator - * in the given string @s. Returns a pointer to the first non-whitespace - * character in @s. - */ -char *strim(char *s) -{ - size_t size; - char *end; - - size = strlen(s); - if (!size) - return s; - - end = s + size - 1; - while (end >= s && isspace(*end)) - end--; - *(end + 1) = '\0'; - - return skip_spaces(s); -} -EXPORT_SYMBOL(strim); - #ifndef __HAVE_ARCH_STRLEN /** * strlen - Find the length of a string @@ -688,101 +609,6 @@ char *strsep(char **s, const char *ct) EXPORT_SYMBOL(strsep); #endif -/** - * sysfs_streq - return true if strings are equal, modulo trailing newline - * @s1: one string - * @s2: another string - * - * This routine returns true iff two strings are equal, treating both - * NUL and newline-then-NUL as equivalent string terminations. It's - * geared for use with sysfs input strings, which generally terminate - * with newlines but are compared against values without newlines. - */ -bool sysfs_streq(const char *s1, const char *s2) -{ - while (*s1 && *s1 == *s2) { - s1++; - s2++; - } - - if (*s1 == *s2) - return true; - if (!*s1 && *s2 == '\n' && !s2[1]) - return true; - if (*s1 == '\n' && !s1[1] && !*s2) - return true; - return false; -} -EXPORT_SYMBOL(sysfs_streq); - -/** - * match_string - matches given string in an array - * @array: array of strings - * @n: number of strings in the array or -1 for NULL terminated arrays - * @string: string to match with - * - * This routine will look for a string in an array of strings up to the - * n-th element in the array or until the first NULL element. - * - * Historically the value of -1 for @n, was used to search in arrays that - * are NULL terminated. However, the function does not make a distinction - * when finishing the search: either @n elements have been compared OR - * the first NULL element was found. - * - * Return: - * index of a @string in the @array if matches, or %-EINVAL otherwise. - */ -int match_string(const char * const *array, size_t n, const char *string) -{ - int index; - const char *item; - - for (index = 0; index < n; index++) { - item = array[index]; - if (!item) - break; - if (!strcmp(item, string)) - return index; - } - - return -EINVAL; -} -EXPORT_SYMBOL(match_string); - -/** - * __sysfs_match_string - matches given string in an array - * @array: array of strings - * @n: number of strings in the array or -1 for NULL terminated arrays - * @str: string to match with - * - * Returns index of @str in the @array or -EINVAL, just like match_string(). - * Uses sysfs_streq instead of strcmp for matching. - * - * This routine will look for a string in an array of strings up to the - * n-th element in the array or until the first NULL element. - * - * Historically the value of -1 for @n, was used to search in arrays that - * are NULL terminated. However, the function does not make a distinction - * when finishing the search: either @n elements have been compared OR - * the first NULL element was found. - */ -int __sysfs_match_string(const char * const *array, size_t n, const char *str) -{ - const char *item; - int index; - - for (index = 0; index < n; index++) { - item = array[index]; - if (!item) - break; - if (sysfs_streq(item, str)) - return index; - } - - return -EINVAL; -} -EXPORT_SYMBOL(__sysfs_match_string); - #ifndef __HAVE_ARCH_MEMSET /** * memset - Fill a region of memory with the given value @@ -1141,27 +967,3 @@ void *memchr_inv(const void *start, int c, size_t bytes) return check_bytes8(start, value, bytes % 8); } EXPORT_SYMBOL(memchr_inv); - -/** - * strreplace - Replace all occurrences of character in string. - * @s: The string to operate on. - * @old: The character being replaced. - * @new: The character @old is replaced with. - * - * Returns pointer to the nul byte at the end of @s. - */ -char *strreplace(char *s, char old, char new) -{ - for (; *s; ++s) - if (*s == old) - *s = new; - return s; -} -EXPORT_SYMBOL(strreplace); - -void fortify_panic(const char *name) -{ - pr_emerg("detected buffer overflow in %s\n", name); - BUG(); -} -EXPORT_SYMBOL(fortify_panic); diff --git a/lib/string_helpers.c b/lib/string_helpers.c index 3806a52ce697..bde13612c25d 100644 --- a/lib/string_helpers.c +++ b/lib/string_helpers.c @@ -696,3 +696,196 @@ void kfree_strarray(char **array, size_t n) kfree(array); } EXPORT_SYMBOL_GPL(kfree_strarray); + +/** + * strscpy_pad() - Copy a C-string into a sized buffer + * @dest: Where to copy the string to + * @src: Where to copy the string from + * @count: Size of destination buffer + * + * Copy the string, or as much of it as fits, into the dest buffer. The + * behavior is undefined if the string buffers overlap. The destination + * buffer is always %NUL terminated, unless it's zero-sized. + * + * If the source string is shorter than the destination buffer, zeros + * the tail of the destination buffer. + * + * For full explanation of why you may want to consider using the + * 'strscpy' functions please see the function docstring for strscpy(). + * + * Returns: + * * The number of characters copied (not including the trailing %NUL) + * * -E2BIG if count is 0 or @src was truncated. + */ +ssize_t strscpy_pad(char *dest, const char *src, size_t count) +{ + ssize_t written; + + written = strscpy(dest, src, count); + if (written < 0 || written == count - 1) + return written; + + memset(dest + written + 1, 0, count - written - 1); + + return written; +} +EXPORT_SYMBOL(strscpy_pad); + +/** + * skip_spaces - Removes leading whitespace from @str. + * @str: The string to be stripped. + * + * Returns a pointer to the first non-whitespace character in @str. + */ +char *skip_spaces(const char *str) +{ + while (isspace(*str)) + ++str; + return (char *)str; +} +EXPORT_SYMBOL(skip_spaces); + +/** + * strim - Removes leading and trailing whitespace from @s. + * @s: The string to be stripped. + * + * Note that the first trailing whitespace is replaced with a %NUL-terminator + * in the given string @s. Returns a pointer to the first non-whitespace + * character in @s. + */ +char *strim(char *s) +{ + size_t size; + char *end; + + size = strlen(s); + if (!size) + return s; + + end = s + size - 1; + while (end >= s && isspace(*end)) + end--; + *(end + 1) = '\0'; + + return skip_spaces(s); +} +EXPORT_SYMBOL(strim); + +/** + * sysfs_streq - return true if strings are equal, modulo trailing newline + * @s1: one string + * @s2: another string + * + * This routine returns true iff two strings are equal, treating both + * NUL and newline-then-NUL as equivalent string terminations. It's + * geared for use with sysfs input strings, which generally terminate + * with newlines but are compared against values without newlines. + */ +bool sysfs_streq(const char *s1, const char *s2) +{ + while (*s1 && *s1 == *s2) { + s1++; + s2++; + } + + if (*s1 == *s2) + return true; + if (!*s1 && *s2 == '\n' && !s2[1]) + return true; + if (*s1 == '\n' && !s1[1] && !*s2) + return true; + return false; +} +EXPORT_SYMBOL(sysfs_streq); + +/** + * match_string - matches given string in an array + * @array: array of strings + * @n: number of strings in the array or -1 for NULL terminated arrays + * @string: string to match with + * + * This routine will look for a string in an array of strings up to the + * n-th element in the array or until the first NULL element. + * + * Historically the value of -1 for @n, was used to search in arrays that + * are NULL terminated. However, the function does not make a distinction + * when finishing the search: either @n elements have been compared OR + * the first NULL element was found. + * + * Return: + * index of a @string in the @array if matches, or %-EINVAL otherwise. + */ +int match_string(const char * const *array, size_t n, const char *string) +{ + int index; + const char *item; + + for (index = 0; index < n; index++) { + item = array[index]; + if (!item) + break; + if (!strcmp(item, string)) + return index; + } + + return -EINVAL; +} +EXPORT_SYMBOL(match_string); + +/** + * __sysfs_match_string - matches given string in an array + * @array: array of strings + * @n: number of strings in the array or -1 for NULL terminated arrays + * @str: string to match with + * + * Returns index of @str in the @array or -EINVAL, just like match_string(). + * Uses sysfs_streq instead of strcmp for matching. + * + * This routine will look for a string in an array of strings up to the + * n-th element in the array or until the first NULL element. + * + * Historically the value of -1 for @n, was used to search in arrays that + * are NULL terminated. However, the function does not make a distinction + * when finishing the search: either @n elements have been compared OR + * the first NULL element was found. + */ +int __sysfs_match_string(const char * const *array, size_t n, const char *str) +{ + const char *item; + int index; + + for (index = 0; index < n; index++) { + item = array[index]; + if (!item) + break; + if (sysfs_streq(item, str)) + return index; + } + + return -EINVAL; +} +EXPORT_SYMBOL(__sysfs_match_string); + +/** + * strreplace - Replace all occurrences of character in string. + * @s: The string to operate on. + * @old: The character being replaced. + * @new: The character @old is replaced with. + * + * Returns pointer to the nul byte at the end of @s. + */ +char *strreplace(char *s, char old, char new) +{ + for (; *s; ++s) + if (*s == old) + *s = new; + return s; +} +EXPORT_SYMBOL(strreplace); + +void fortify_panic(const char *name) +{ + pr_emerg("detected buffer overflow in %s\n", name); + BUG(); +} +EXPORT_SYMBOL(fortify_panic); -- cgit v1.2.3 From c430f60036af44079170ff71a461b9d7cf5ee431 Mon Sep 17 00:00:00 2001 From: Kees Cook Date: Wed, 14 Apr 2021 15:45:39 -0700 Subject: fortify: Move remaining fortify helpers into fortify-string.h When commit a28a6e860c6c ("string.h: move fortified functions definitions in a dedicated header.") moved the fortify-specific code, some helpers were left behind. Move the remaining fortify-specific helpers into fortify-string.h so they're together where they're used. This requires that any FORTIFY helper function prototypes be conditionally built to avoid "no prototype" warnings. Additionally removes unused helpers. Cc: Andrew Morton Cc: Daniel Axtens Cc: Vincenzo Frascino Cc: Andrey Konovalov Cc: Dan Williams Acked-by: Francis Laniel Reviewed-by: Nick Desaulniers Signed-off-by: Kees Cook --- include/linux/fortify-string.h | 7 +++++++ include/linux/string.h | 9 --------- lib/string_helpers.c | 2 ++ 3 files changed, 9 insertions(+), 9 deletions(-) (limited to 'lib') diff --git a/include/linux/fortify-string.h b/include/linux/fortify-string.h index c1be37437e77..7e67d02764db 100644 --- a/include/linux/fortify-string.h +++ b/include/linux/fortify-string.h @@ -2,6 +2,13 @@ #ifndef _LINUX_FORTIFY_STRING_H_ #define _LINUX_FORTIFY_STRING_H_ +#define __FORTIFY_INLINE extern __always_inline __attribute__((gnu_inline)) +#define __RENAME(x) __asm__(#x) + +void fortify_panic(const char *name) __noreturn __cold; +void __read_overflow(void) __compiletime_error("detected read beyond size of object (1st parameter)"); +void __read_overflow2(void) __compiletime_error("detected read beyond size of object (2nd parameter)"); +void __write_overflow(void) __compiletime_error("detected write beyond size of object (1st parameter)"); #if defined(CONFIG_KASAN_GENERIC) || defined(CONFIG_KASAN_SW_TAGS) extern void *__underlying_memchr(const void *p, int c, __kernel_size_t size) __RENAME(memchr); diff --git a/include/linux/string.h b/include/linux/string.h index 5e96d656be7a..ac1c769a5a80 100644 --- a/include/linux/string.h +++ b/include/linux/string.h @@ -249,15 +249,6 @@ static inline const char *kbasename(const char *path) return tail ? tail + 1 : path; } -#define __FORTIFY_INLINE extern __always_inline __attribute__((gnu_inline)) -#define __RENAME(x) __asm__(#x) - -void fortify_panic(const char *name) __noreturn __cold; -void __read_overflow(void) __compiletime_error("detected read beyond size of object passed as 1st parameter"); -void __read_overflow2(void) __compiletime_error("detected read beyond size of object passed as 2nd parameter"); -void __read_overflow3(void) __compiletime_error("detected read beyond size of object passed as 3rd parameter"); -void __write_overflow(void) __compiletime_error("detected write beyond size of object passed as 1st parameter"); - #if !defined(__NO_FORTIFY) && defined(__OPTIMIZE__) && defined(CONFIG_FORTIFY_SOURCE) #include #endif diff --git a/lib/string_helpers.c b/lib/string_helpers.c index bde13612c25d..faa9d8e4e2c5 100644 --- a/lib/string_helpers.c +++ b/lib/string_helpers.c @@ -883,9 +883,11 @@ char *strreplace(char *s, char old, char new) } EXPORT_SYMBOL(strreplace); +#ifdef CONFIG_FORTIFY_SOURCE void fortify_panic(const char *name) { pr_emerg("detected buffer overflow in %s\n", name); BUG(); } EXPORT_SYMBOL(fortify_panic); +#endif /* CONFIG_FORTIFY_SOURCE */ -- cgit v1.2.3 From c2a228d69cef802cf6bfd773c84f8419d2e2acf9 Mon Sep 17 00:00:00 2001 From: Johan Almbladh Date: Tue, 14 Sep 2021 11:18:29 +0200 Subject: bpf/tests: Allow different number of runs per test case This patch allows a test cast to specify the number of runs to use. For compatibility with existing test case definitions, the default value 0 is interpreted as MAX_TESTRUNS. A reduced number of runs is useful for complex test programs where 1000 runs may take a very long time. Instead of reducing what is tested, one can instead reduce the number of times the test is run. Signed-off-by: Johan Almbladh Signed-off-by: Daniel Borkmann Link: https://lore.kernel.org/bpf/20210914091842.4186267-2-johan.almbladh@anyfinetworks.com --- lib/test_bpf.c | 4 ++++ 1 file changed, 4 insertions(+) (limited to 'lib') diff --git a/lib/test_bpf.c b/lib/test_bpf.c index 0018d51b93b0..a406cbb2e34d 100644 --- a/lib/test_bpf.c +++ b/lib/test_bpf.c @@ -80,6 +80,7 @@ struct bpf_test { int expected_errcode; /* used when FLAG_EXPECTED_FAIL is set in the aux */ __u8 frag_data[MAX_DATA]; int stack_depth; /* for eBPF only, since tests don't call verifier */ + int nr_testruns; /* Custom run count, defaults to MAX_TESTRUNS if 0 */ }; /* Large test cases need separate allocation and fill handler. */ @@ -8631,6 +8632,9 @@ static int run_one(const struct bpf_prog *fp, struct bpf_test *test) { int err_cnt = 0, i, runs = MAX_TESTRUNS; + if (test->nr_testruns) + runs = min(test->nr_testruns, MAX_TESTRUNS); + for (i = 0; i < MAX_SUBTESTS; i++) { void *data; u64 duration; -- cgit v1.2.3 From 4bc354138d553bc48dc1fb1e184d50a524e6c20f Mon Sep 17 00:00:00 2001 From: Johan Almbladh Date: Tue, 14 Sep 2021 11:18:30 +0200 Subject: bpf/tests: Reduce memory footprint of test suite The test suite used to call any fill_helper callbacks to generate eBPF program data for all test cases at once. This caused ballooning memory requirements as more extensive test cases were added. Now the each fill_helper is called before the test is run and the allocated memory released afterwards, before the next test case is processed. Signed-off-by: Johan Almbladh Signed-off-by: Daniel Borkmann Link: https://lore.kernel.org/bpf/20210914091842.4186267-3-johan.almbladh@anyfinetworks.com --- lib/test_bpf.c | 26 ++++++++++++-------------- 1 file changed, 12 insertions(+), 14 deletions(-) (limited to 'lib') diff --git a/lib/test_bpf.c b/lib/test_bpf.c index a406cbb2e34d..c865fd584adb 100644 --- a/lib/test_bpf.c +++ b/lib/test_bpf.c @@ -8694,8 +8694,6 @@ static __init int find_test_index(const char *test_name) static __init int prepare_bpf_tests(void) { - int i; - if (test_id >= 0) { /* * if a test_id was specified, use test_range to @@ -8739,23 +8737,11 @@ static __init int prepare_bpf_tests(void) } } - for (i = 0; i < ARRAY_SIZE(tests); i++) { - if (tests[i].fill_helper && - tests[i].fill_helper(&tests[i]) < 0) - return -ENOMEM; - } - return 0; } static __init void destroy_bpf_tests(void) { - int i; - - for (i = 0; i < ARRAY_SIZE(tests); i++) { - if (tests[i].fill_helper) - kfree(tests[i].u.ptr.insns); - } } static bool exclude_test(int test_id) @@ -8960,7 +8946,19 @@ static __init int test_bpf(void) pr_info("#%d %s ", i, tests[i].descr); + if (tests[i].fill_helper && + tests[i].fill_helper(&tests[i]) < 0) { + pr_cont("FAIL to prog_fill\n"); + continue; + } + fp = generate_filter(i, &err); + + if (tests[i].fill_helper) { + kfree(tests[i].u.ptr.insns); + tests[i].u.ptr.insns = NULL; + } + if (fp == NULL) { if (err == 0) { pass_cnt++; -- cgit v1.2.3 From 68c956fe741757b760aa00fca8725c5651f5f77a Mon Sep 17 00:00:00 2001 From: Johan Almbladh Date: Tue, 14 Sep 2021 11:18:31 +0200 Subject: bpf/tests: Add exhaustive tests of ALU shift values This patch adds a set of tests for ALU64 and ALU32 shift operations to verify correctness for all possible values of the shift value. Mainly intended for JIT testing. Signed-off-by: Johan Almbladh Signed-off-by: Daniel Borkmann Link: https://lore.kernel.org/bpf/20210914091842.4186267-4-johan.almbladh@anyfinetworks.com --- lib/test_bpf.c | 260 +++++++++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 260 insertions(+) (limited to 'lib') diff --git a/lib/test_bpf.c b/lib/test_bpf.c index c865fd584adb..d599c2a8c860 100644 --- a/lib/test_bpf.c +++ b/lib/test_bpf.c @@ -497,6 +497,168 @@ static int bpf_fill_long_jmp(struct bpf_test *self) return 0; } +static int __bpf_ld_imm64(struct bpf_insn insns[2], u8 reg, s64 imm64) +{ + struct bpf_insn tmp[] = {BPF_LD_IMM64(reg, imm64)}; + + memcpy(insns, tmp, sizeof(tmp)); + return 2; +} + +/* Test an ALU shift operation for all valid shift values */ +static int __bpf_fill_alu_shift(struct bpf_test *self, u8 op, + u8 mode, bool alu32) +{ + static const s64 regs[] = { + 0x0123456789abcdefLL, /* dword > 0, word < 0 */ + 0xfedcba9876543210LL, /* dowrd < 0, word > 0 */ + 0xfedcba0198765432LL, /* dowrd < 0, word < 0 */ + 0x0123458967abcdefLL, /* dword > 0, word > 0 */ + }; + int bits = alu32 ? 32 : 64; + int len = (2 + 7 * bits) * ARRAY_SIZE(regs) + 3; + struct bpf_insn *insn; + int imm, k; + int i = 0; + + insn = kmalloc_array(len, sizeof(*insn), GFP_KERNEL); + if (!insn) + return -ENOMEM; + + insn[i++] = BPF_ALU64_IMM(BPF_MOV, R0, 0); + + for (k = 0; k < ARRAY_SIZE(regs); k++) { + s64 reg = regs[k]; + + i += __bpf_ld_imm64(&insn[i], R3, reg); + + for (imm = 0; imm < bits; imm++) { + u64 val; + + /* Perform operation */ + insn[i++] = BPF_ALU64_REG(BPF_MOV, R1, R3); + insn[i++] = BPF_ALU64_IMM(BPF_MOV, R2, imm); + if (alu32) { + if (mode == BPF_K) + insn[i++] = BPF_ALU32_IMM(op, R1, imm); + else + insn[i++] = BPF_ALU32_REG(op, R1, R2); + switch (op) { + case BPF_LSH: + val = (u32)reg << imm; + break; + case BPF_RSH: + val = (u32)reg >> imm; + break; + case BPF_ARSH: + val = (u32)reg >> imm; + if (imm > 0 && (reg & 0x80000000)) + val |= ~(u32)0 << (32 - imm); + break; + } + } else { + if (mode == BPF_K) + insn[i++] = BPF_ALU64_IMM(op, R1, imm); + else + insn[i++] = BPF_ALU64_REG(op, R1, R2); + switch (op) { + case BPF_LSH: + val = (u64)reg << imm; + break; + case BPF_RSH: + val = (u64)reg >> imm; + break; + case BPF_ARSH: + val = (u64)reg >> imm; + if (imm > 0 && reg < 0) + val |= ~(u64)0 << (64 - imm); + break; + } + } + + /* + * When debugging a JIT that fails this test, one + * can write the immediate value to R0 here to find + * out which operand values that fail. + */ + + /* Load reference and check the result */ + i += __bpf_ld_imm64(&insn[i], R4, val); + insn[i++] = BPF_JMP_REG(BPF_JEQ, R1, R4, 1); + insn[i++] = BPF_EXIT_INSN(); + } + } + + insn[i++] = BPF_ALU64_IMM(BPF_MOV, R0, 1); + insn[i++] = BPF_EXIT_INSN(); + + self->u.ptr.insns = insn; + self->u.ptr.len = len; + BUG_ON(i > len); + + return 0; +} + +static int bpf_fill_alu_lsh_imm(struct bpf_test *self) +{ + return __bpf_fill_alu_shift(self, BPF_LSH, BPF_K, false); +} + +static int bpf_fill_alu_rsh_imm(struct bpf_test *self) +{ + return __bpf_fill_alu_shift(self, BPF_RSH, BPF_K, false); +} + +static int bpf_fill_alu_arsh_imm(struct bpf_test *self) +{ + return __bpf_fill_alu_shift(self, BPF_ARSH, BPF_K, false); +} + +static int bpf_fill_alu_lsh_reg(struct bpf_test *self) +{ + return __bpf_fill_alu_shift(self, BPF_LSH, BPF_X, false); +} + +static int bpf_fill_alu_rsh_reg(struct bpf_test *self) +{ + return __bpf_fill_alu_shift(self, BPF_RSH, BPF_X, false); +} + +static int bpf_fill_alu_arsh_reg(struct bpf_test *self) +{ + return __bpf_fill_alu_shift(self, BPF_ARSH, BPF_X, false); +} + +static int bpf_fill_alu32_lsh_imm(struct bpf_test *self) +{ + return __bpf_fill_alu_shift(self, BPF_LSH, BPF_K, true); +} + +static int bpf_fill_alu32_rsh_imm(struct bpf_test *self) +{ + return __bpf_fill_alu_shift(self, BPF_RSH, BPF_K, true); +} + +static int bpf_fill_alu32_arsh_imm(struct bpf_test *self) +{ + return __bpf_fill_alu_shift(self, BPF_ARSH, BPF_K, true); +} + +static int bpf_fill_alu32_lsh_reg(struct bpf_test *self) +{ + return __bpf_fill_alu_shift(self, BPF_LSH, BPF_X, true); +} + +static int bpf_fill_alu32_rsh_reg(struct bpf_test *self) +{ + return __bpf_fill_alu_shift(self, BPF_RSH, BPF_X, true); +} + +static int bpf_fill_alu32_arsh_reg(struct bpf_test *self) +{ + return __bpf_fill_alu_shift(self, BPF_ARSH, BPF_X, true); +} + static struct bpf_test tests[] = { { "TAX", @@ -8414,6 +8576,104 @@ static struct bpf_test tests[] = { {}, { { 0, 2 } }, }, + /* Exhaustive test of ALU64 shift operations */ + { + "ALU64_LSH_K: all shift values", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu_lsh_imm, + }, + { + "ALU64_RSH_K: all shift values", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu_rsh_imm, + }, + { + "ALU64_ARSH_K: all shift values", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu_arsh_imm, + }, + { + "ALU64_LSH_X: all shift values", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu_lsh_reg, + }, + { + "ALU64_RSH_X: all shift values", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu_rsh_reg, + }, + { + "ALU64_ARSH_X: all shift values", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu_arsh_reg, + }, + /* Exhaustive test of ALU32 shift operations */ + { + "ALU32_LSH_K: all shift values", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu32_lsh_imm, + }, + { + "ALU32_RSH_K: all shift values", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu32_rsh_imm, + }, + { + "ALU32_ARSH_K: all shift values", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu32_arsh_imm, + }, + { + "ALU32_LSH_X: all shift values", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu32_lsh_reg, + }, + { + "ALU32_RSH_X: all shift values", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu32_rsh_reg, + }, + { + "ALU32_ARSH_X: all shift values", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu32_arsh_reg, + }, }; static struct net_device dev; -- cgit v1.2.3 From 9298e63eafea1ebe235919dbbbfc20c1c25000f8 Mon Sep 17 00:00:00 2001 From: Johan Almbladh Date: Tue, 14 Sep 2021 11:18:32 +0200 Subject: bpf/tests: Add exhaustive tests of ALU operand magnitudes This patch adds a set of tests for ALU64 and ALU32 arithmetic and bitwise logical operations to verify correctness for all possible magnitudes of the register and immediate operands. Mainly intended for JIT testing. The patch introduces a pattern generator that can be used to drive extensive tests of different kinds of operations. It is parameterized to allow tuning of the operand combinations to test. Signed-off-by: Johan Almbladh Signed-off-by: Daniel Borkmann Link: https://lore.kernel.org/bpf/20210914091842.4186267-5-johan.almbladh@anyfinetworks.com --- lib/test_bpf.c | 772 +++++++++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 772 insertions(+) (limited to 'lib') diff --git a/lib/test_bpf.c b/lib/test_bpf.c index d599c2a8c860..228e681e84b7 100644 --- a/lib/test_bpf.c +++ b/lib/test_bpf.c @@ -659,6 +659,451 @@ static int bpf_fill_alu32_arsh_reg(struct bpf_test *self) return __bpf_fill_alu_shift(self, BPF_ARSH, BPF_X, true); } +/* + * Common operand pattern generator for exhaustive power-of-two magnitudes + * tests. The block size parameters can be adjusted to increase/reduce the + * number of combinatons tested and thereby execution speed and memory + * footprint. + */ + +static inline s64 value(int msb, int delta, int sign) +{ + return sign * (1LL << msb) + delta; +} + +static int __bpf_fill_pattern(struct bpf_test *self, void *arg, + int dbits, int sbits, int block1, int block2, + int (*emit)(struct bpf_test*, void*, + struct bpf_insn*, s64, s64)) +{ + static const int sgn[][2] = {{1, 1}, {1, -1}, {-1, 1}, {-1, -1}}; + struct bpf_insn *insns; + int di, si, bt, db, sb; + int count, len, k; + int extra = 1 + 2; + int i = 0; + + /* Total number of iterations for the two pattern */ + count = (dbits - 1) * (sbits - 1) * block1 * block1 * ARRAY_SIZE(sgn); + count += (max(dbits, sbits) - 1) * block2 * block2 * ARRAY_SIZE(sgn); + + /* Compute the maximum number of insns and allocate the buffer */ + len = extra + count * (*emit)(self, arg, NULL, 0, 0); + insns = kmalloc_array(len, sizeof(*insns), GFP_KERNEL); + if (!insns) + return -ENOMEM; + + /* Add head instruction(s) */ + insns[i++] = BPF_ALU64_IMM(BPF_MOV, R0, 0); + + /* + * Pattern 1: all combinations of power-of-two magnitudes and sign, + * and with a block of contiguous values around each magnitude. + */ + for (di = 0; di < dbits - 1; di++) /* Dst magnitudes */ + for (si = 0; si < sbits - 1; si++) /* Src magnitudes */ + for (k = 0; k < ARRAY_SIZE(sgn); k++) /* Sign combos */ + for (db = -(block1 / 2); + db < (block1 + 1) / 2; db++) + for (sb = -(block1 / 2); + sb < (block1 + 1) / 2; sb++) { + s64 dst, src; + + dst = value(di, db, sgn[k][0]); + src = value(si, sb, sgn[k][1]); + i += (*emit)(self, arg, + &insns[i], + dst, src); + } + /* + * Pattern 2: all combinations for a larger block of values + * for each power-of-two magnitude and sign, where the magnitude is + * the same for both operands. + */ + for (bt = 0; bt < max(dbits, sbits) - 1; bt++) /* Magnitude */ + for (k = 0; k < ARRAY_SIZE(sgn); k++) /* Sign combos */ + for (db = -(block2 / 2); db < (block2 + 1) / 2; db++) + for (sb = -(block2 / 2); + sb < (block2 + 1) / 2; sb++) { + s64 dst, src; + + dst = value(bt % dbits, db, sgn[k][0]); + src = value(bt % sbits, sb, sgn[k][1]); + i += (*emit)(self, arg, &insns[i], + dst, src); + } + + /* Append tail instructions */ + insns[i++] = BPF_ALU64_IMM(BPF_MOV, R0, 1); + insns[i++] = BPF_EXIT_INSN(); + BUG_ON(i > len); + + self->u.ptr.insns = insns; + self->u.ptr.len = i; + + return 0; +} + +/* + * Block size parameters used in pattern tests below. une as needed to + * increase/reduce the number combinations tested, see following examples. + * block values per operand MSB + * ---------------------------------------- + * 0 none + * 1 (1 << MSB) + * 2 (1 << MSB) + [-1, 0] + * 3 (1 << MSB) + [-1, 0, 1] + */ +#define PATTERN_BLOCK1 1 +#define PATTERN_BLOCK2 5 + +/* Number of test runs for a pattern test */ +#define NR_PATTERN_RUNS 1 + +/* + * Exhaustive tests of ALU operations for all combinations of power-of-two + * magnitudes of the operands, both for positive and negative values. The + * test is designed to verify e.g. the JMP and JMP32 operations for JITs that + * emit different code depending on the magnitude of the immediate value. + */ + +static bool __bpf_alu_result(u64 *res, u64 v1, u64 v2, u8 op) +{ + *res = 0; + switch (op) { + case BPF_MOV: + *res = v2; + break; + case BPF_AND: + *res = v1 & v2; + break; + case BPF_OR: + *res = v1 | v2; + break; + case BPF_XOR: + *res = v1 ^ v2; + break; + case BPF_ADD: + *res = v1 + v2; + break; + case BPF_SUB: + *res = v1 - v2; + break; + case BPF_MUL: + *res = v1 * v2; + break; + case BPF_DIV: + if (v2 == 0) + return false; + *res = div64_u64(v1, v2); + break; + case BPF_MOD: + if (v2 == 0) + return false; + div64_u64_rem(v1, v2, res); + break; + } + return true; +} + +static int __bpf_emit_alu64_imm(struct bpf_test *self, void *arg, + struct bpf_insn *insns, s64 dst, s64 imm) +{ + int op = *(int *)arg; + int i = 0; + u64 res; + + if (!insns) + return 7; + + if (__bpf_alu_result(&res, dst, (s32)imm, op)) { + i += __bpf_ld_imm64(&insns[i], R1, dst); + i += __bpf_ld_imm64(&insns[i], R3, res); + insns[i++] = BPF_ALU64_IMM(op, R1, imm); + insns[i++] = BPF_JMP_REG(BPF_JEQ, R1, R3, 1); + insns[i++] = BPF_EXIT_INSN(); + } + + return i; +} + +static int __bpf_emit_alu32_imm(struct bpf_test *self, void *arg, + struct bpf_insn *insns, s64 dst, s64 imm) +{ + int op = *(int *)arg; + int i = 0; + u64 res; + + if (!insns) + return 7; + + if (__bpf_alu_result(&res, (u32)dst, (u32)imm, op)) { + i += __bpf_ld_imm64(&insns[i], R1, dst); + i += __bpf_ld_imm64(&insns[i], R3, (u32)res); + insns[i++] = BPF_ALU32_IMM(op, R1, imm); + insns[i++] = BPF_JMP_REG(BPF_JEQ, R1, R3, 1); + insns[i++] = BPF_EXIT_INSN(); + } + + return i; +} + +static int __bpf_emit_alu64_reg(struct bpf_test *self, void *arg, + struct bpf_insn *insns, s64 dst, s64 src) +{ + int op = *(int *)arg; + int i = 0; + u64 res; + + if (!insns) + return 9; + + if (__bpf_alu_result(&res, dst, src, op)) { + i += __bpf_ld_imm64(&insns[i], R1, dst); + i += __bpf_ld_imm64(&insns[i], R2, src); + i += __bpf_ld_imm64(&insns[i], R3, res); + insns[i++] = BPF_ALU64_REG(op, R1, R2); + insns[i++] = BPF_JMP_REG(BPF_JEQ, R1, R3, 1); + insns[i++] = BPF_EXIT_INSN(); + } + + return i; +} + +static int __bpf_emit_alu32_reg(struct bpf_test *self, void *arg, + struct bpf_insn *insns, s64 dst, s64 src) +{ + int op = *(int *)arg; + int i = 0; + u64 res; + + if (!insns) + return 9; + + if (__bpf_alu_result(&res, (u32)dst, (u32)src, op)) { + i += __bpf_ld_imm64(&insns[i], R1, dst); + i += __bpf_ld_imm64(&insns[i], R2, src); + i += __bpf_ld_imm64(&insns[i], R3, (u32)res); + insns[i++] = BPF_ALU32_REG(op, R1, R2); + insns[i++] = BPF_JMP_REG(BPF_JEQ, R1, R3, 1); + insns[i++] = BPF_EXIT_INSN(); + } + + return i; +} + +static int __bpf_fill_alu64_imm(struct bpf_test *self, int op) +{ + return __bpf_fill_pattern(self, &op, 64, 32, + PATTERN_BLOCK1, PATTERN_BLOCK2, + &__bpf_emit_alu64_imm); +} + +static int __bpf_fill_alu32_imm(struct bpf_test *self, int op) +{ + return __bpf_fill_pattern(self, &op, 64, 32, + PATTERN_BLOCK1, PATTERN_BLOCK2, + &__bpf_emit_alu32_imm); +} + +static int __bpf_fill_alu64_reg(struct bpf_test *self, int op) +{ + return __bpf_fill_pattern(self, &op, 64, 64, + PATTERN_BLOCK1, PATTERN_BLOCK2, + &__bpf_emit_alu64_reg); +} + +static int __bpf_fill_alu32_reg(struct bpf_test *self, int op) +{ + return __bpf_fill_pattern(self, &op, 64, 64, + PATTERN_BLOCK1, PATTERN_BLOCK2, + &__bpf_emit_alu32_reg); +} + +/* ALU64 immediate operations */ +static int bpf_fill_alu64_mov_imm(struct bpf_test *self) +{ + return __bpf_fill_alu64_imm(self, BPF_MOV); +} + +static int bpf_fill_alu64_and_imm(struct bpf_test *self) +{ + return __bpf_fill_alu64_imm(self, BPF_AND); +} + +static int bpf_fill_alu64_or_imm(struct bpf_test *self) +{ + return __bpf_fill_alu64_imm(self, BPF_OR); +} + +static int bpf_fill_alu64_xor_imm(struct bpf_test *self) +{ + return __bpf_fill_alu64_imm(self, BPF_XOR); +} + +static int bpf_fill_alu64_add_imm(struct bpf_test *self) +{ + return __bpf_fill_alu64_imm(self, BPF_ADD); +} + +static int bpf_fill_alu64_sub_imm(struct bpf_test *self) +{ + return __bpf_fill_alu64_imm(self, BPF_SUB); +} + +static int bpf_fill_alu64_mul_imm(struct bpf_test *self) +{ + return __bpf_fill_alu64_imm(self, BPF_MUL); +} + +static int bpf_fill_alu64_div_imm(struct bpf_test *self) +{ + return __bpf_fill_alu64_imm(self, BPF_DIV); +} + +static int bpf_fill_alu64_mod_imm(struct bpf_test *self) +{ + return __bpf_fill_alu64_imm(self, BPF_MOD); +} + +/* ALU32 immediate operations */ +static int bpf_fill_alu32_mov_imm(struct bpf_test *self) +{ + return __bpf_fill_alu32_imm(self, BPF_MOV); +} + +static int bpf_fill_alu32_and_imm(struct bpf_test *self) +{ + return __bpf_fill_alu32_imm(self, BPF_AND); +} + +static int bpf_fill_alu32_or_imm(struct bpf_test *self) +{ + return __bpf_fill_alu32_imm(self, BPF_OR); +} + +static int bpf_fill_alu32_xor_imm(struct bpf_test *self) +{ + return __bpf_fill_alu32_imm(self, BPF_XOR); +} + +static int bpf_fill_alu32_add_imm(struct bpf_test *self) +{ + return __bpf_fill_alu32_imm(self, BPF_ADD); +} + +static int bpf_fill_alu32_sub_imm(struct bpf_test *self) +{ + return __bpf_fill_alu32_imm(self, BPF_SUB); +} + +static int bpf_fill_alu32_mul_imm(struct bpf_test *self) +{ + return __bpf_fill_alu32_imm(self, BPF_MUL); +} + +static int bpf_fill_alu32_div_imm(struct bpf_test *self) +{ + return __bpf_fill_alu32_imm(self, BPF_DIV); +} + +static int bpf_fill_alu32_mod_imm(struct bpf_test *self) +{ + return __bpf_fill_alu32_imm(self, BPF_MOD); +} + +/* ALU64 register operations */ +static int bpf_fill_alu64_mov_reg(struct bpf_test *self) +{ + return __bpf_fill_alu64_reg(self, BPF_MOV); +} + +static int bpf_fill_alu64_and_reg(struct bpf_test *self) +{ + return __bpf_fill_alu64_reg(self, BPF_AND); +} + +static int bpf_fill_alu64_or_reg(struct bpf_test *self) +{ + return __bpf_fill_alu64_reg(self, BPF_OR); +} + +static int bpf_fill_alu64_xor_reg(struct bpf_test *self) +{ + return __bpf_fill_alu64_reg(self, BPF_XOR); +} + +static int bpf_fill_alu64_add_reg(struct bpf_test *self) +{ + return __bpf_fill_alu64_reg(self, BPF_ADD); +} + +static int bpf_fill_alu64_sub_reg(struct bpf_test *self) +{ + return __bpf_fill_alu64_reg(self, BPF_SUB); +} + +static int bpf_fill_alu64_mul_reg(struct bpf_test *self) +{ + return __bpf_fill_alu64_reg(self, BPF_MUL); +} + +static int bpf_fill_alu64_div_reg(struct bpf_test *self) +{ + return __bpf_fill_alu64_reg(self, BPF_DIV); +} + +static int bpf_fill_alu64_mod_reg(struct bpf_test *self) +{ + return __bpf_fill_alu64_reg(self, BPF_MOD); +} + +/* ALU32 register operations */ +static int bpf_fill_alu32_mov_reg(struct bpf_test *self) +{ + return __bpf_fill_alu32_reg(self, BPF_MOV); +} + +static int bpf_fill_alu32_and_reg(struct bpf_test *self) +{ + return __bpf_fill_alu32_reg(self, BPF_AND); +} + +static int bpf_fill_alu32_or_reg(struct bpf_test *self) +{ + return __bpf_fill_alu32_reg(self, BPF_OR); +} + +static int bpf_fill_alu32_xor_reg(struct bpf_test *self) +{ + return __bpf_fill_alu32_reg(self, BPF_XOR); +} + +static int bpf_fill_alu32_add_reg(struct bpf_test *self) +{ + return __bpf_fill_alu32_reg(self, BPF_ADD); +} + +static int bpf_fill_alu32_sub_reg(struct bpf_test *self) +{ + return __bpf_fill_alu32_reg(self, BPF_SUB); +} + +static int bpf_fill_alu32_mul_reg(struct bpf_test *self) +{ + return __bpf_fill_alu32_reg(self, BPF_MUL); +} + +static int bpf_fill_alu32_div_reg(struct bpf_test *self) +{ + return __bpf_fill_alu32_reg(self, BPF_DIV); +} + +static int bpf_fill_alu32_mod_reg(struct bpf_test *self) +{ + return __bpf_fill_alu32_reg(self, BPF_MOD); +} + static struct bpf_test tests[] = { { "TAX", @@ -8674,6 +9119,333 @@ static struct bpf_test tests[] = { { { 0, 1 } }, .fill_helper = bpf_fill_alu32_arsh_reg, }, + /* ALU64 immediate magnitudes */ + { + "ALU64_MOV_K: all immediate value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu64_mov_imm, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "ALU64_AND_K: all immediate value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu64_and_imm, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "ALU64_OR_K: all immediate value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu64_or_imm, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "ALU64_XOR_K: all immediate value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu64_xor_imm, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "ALU64_ADD_K: all immediate value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu64_add_imm, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "ALU64_SUB_K: all immediate value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu64_sub_imm, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "ALU64_MUL_K: all immediate value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu64_mul_imm, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "ALU64_DIV_K: all immediate value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu64_div_imm, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "ALU64_MOD_K: all immediate value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu64_mod_imm, + .nr_testruns = NR_PATTERN_RUNS, + }, + /* ALU32 immediate magnitudes */ + { + "ALU32_MOV_K: all immediate value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu32_mov_imm, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "ALU32_AND_K: all immediate value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu32_and_imm, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "ALU32_OR_K: all immediate value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu32_or_imm, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "ALU32_XOR_K: all immediate value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu32_xor_imm, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "ALU32_ADD_K: all immediate value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu32_add_imm, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "ALU32_SUB_K: all immediate value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu32_sub_imm, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "ALU32_MUL_K: all immediate value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu32_mul_imm, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "ALU32_DIV_K: all immediate value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu32_div_imm, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "ALU32_MOD_K: all immediate value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu32_mod_imm, + }, + /* ALU64 register magnitudes */ + { + "ALU64_MOV_X: all register value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu64_mov_reg, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "ALU64_AND_X: all register value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu64_and_reg, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "ALU64_OR_X: all register value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu64_or_reg, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "ALU64_XOR_X: all register value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu64_xor_reg, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "ALU64_ADD_X: all register value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu64_add_reg, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "ALU64_SUB_X: all register value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu64_sub_reg, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "ALU64_MUL_X: all register value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu64_mul_reg, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "ALU64_DIV_X: all register value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu64_div_reg, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "ALU64_MOD_X: all register value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu64_mod_reg, + .nr_testruns = NR_PATTERN_RUNS, + }, + /* ALU32 register magnitudes */ + { + "ALU32_MOV_X: all register value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu32_mov_reg, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "ALU32_AND_X: all register value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu32_and_reg, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "ALU32_OR_X: all register value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu32_or_reg, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "ALU32_XOR_X: all register value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu32_xor_reg, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "ALU32_ADD_X: all register value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu32_add_reg, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "ALU32_SUB_X: all register value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu32_sub_reg, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "ALU32_MUL_X: all register value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu32_mul_reg, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "ALU32_DIV_X: all register value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu32_div_reg, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "ALU32_MOD_X: all register value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu32_mod_reg, + .nr_testruns = NR_PATTERN_RUNS, + }, }; static struct net_device dev; -- cgit v1.2.3 From a5a36544de38057b8e8de8fb6b2bcd9c102640f4 Mon Sep 17 00:00:00 2001 From: Johan Almbladh Date: Tue, 14 Sep 2021 11:18:33 +0200 Subject: bpf/tests: Add exhaustive tests of JMP operand magnitudes This patch adds a set of tests for conditional JMP and JMP32 operations to verify correctness for all possible magnitudes of the immediate and register operands. Mainly intended for JIT testing. Signed-off-by: Johan Almbladh Signed-off-by: Daniel Borkmann Link: https://lore.kernel.org/bpf/20210914091842.4186267-6-johan.almbladh@anyfinetworks.com --- lib/test_bpf.c | 779 +++++++++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 779 insertions(+) (limited to 'lib') diff --git a/lib/test_bpf.c b/lib/test_bpf.c index 228e681e84b7..fb27a25e2c0c 100644 --- a/lib/test_bpf.c +++ b/lib/test_bpf.c @@ -1104,6 +1104,384 @@ static int bpf_fill_alu32_mod_reg(struct bpf_test *self) return __bpf_fill_alu32_reg(self, BPF_MOD); } + +/* + * Exhaustive tests of JMP operations for all combinations of power-of-two + * magnitudes of the operands, both for positive and negative values. The + * test is designed to verify e.g. the JMP and JMP32 operations for JITs that + * emit different code depending on the magnitude of the immediate value. + */ + +static bool __bpf_match_jmp_cond(s64 v1, s64 v2, u8 op) +{ + switch (op) { + case BPF_JSET: + return !!(v1 & v2); + case BPF_JEQ: + return v1 == v2; + case BPF_JNE: + return v1 != v2; + case BPF_JGT: + return (u64)v1 > (u64)v2; + case BPF_JGE: + return (u64)v1 >= (u64)v2; + case BPF_JLT: + return (u64)v1 < (u64)v2; + case BPF_JLE: + return (u64)v1 <= (u64)v2; + case BPF_JSGT: + return v1 > v2; + case BPF_JSGE: + return v1 >= v2; + case BPF_JSLT: + return v1 < v2; + case BPF_JSLE: + return v1 <= v2; + } + return false; +} + +static int __bpf_emit_jmp_imm(struct bpf_test *self, void *arg, + struct bpf_insn *insns, s64 dst, s64 imm) +{ + int op = *(int *)arg; + + if (insns) { + bool match = __bpf_match_jmp_cond(dst, (s32)imm, op); + int i = 0; + + insns[i++] = BPF_ALU32_IMM(BPF_MOV, R0, match); + + i += __bpf_ld_imm64(&insns[i], R1, dst); + insns[i++] = BPF_JMP_IMM(op, R1, imm, 1); + if (!match) + insns[i++] = BPF_JMP_IMM(BPF_JA, 0, 0, 1); + insns[i++] = BPF_EXIT_INSN(); + + return i; + } + + return 5 + 1; +} + +static int __bpf_emit_jmp32_imm(struct bpf_test *self, void *arg, + struct bpf_insn *insns, s64 dst, s64 imm) +{ + int op = *(int *)arg; + + if (insns) { + bool match = __bpf_match_jmp_cond((s32)dst, (s32)imm, op); + int i = 0; + + i += __bpf_ld_imm64(&insns[i], R1, dst); + insns[i++] = BPF_JMP32_IMM(op, R1, imm, 1); + if (!match) + insns[i++] = BPF_JMP_IMM(BPF_JA, 0, 0, 1); + insns[i++] = BPF_EXIT_INSN(); + + return i; + } + + return 5; +} + +static int __bpf_emit_jmp_reg(struct bpf_test *self, void *arg, + struct bpf_insn *insns, s64 dst, s64 src) +{ + int op = *(int *)arg; + + if (insns) { + bool match = __bpf_match_jmp_cond(dst, src, op); + int i = 0; + + i += __bpf_ld_imm64(&insns[i], R1, dst); + i += __bpf_ld_imm64(&insns[i], R2, src); + insns[i++] = BPF_JMP_REG(op, R1, R2, 1); + if (!match) + insns[i++] = BPF_JMP_IMM(BPF_JA, 0, 0, 1); + insns[i++] = BPF_EXIT_INSN(); + + return i; + } + + return 7; +} + +static int __bpf_emit_jmp32_reg(struct bpf_test *self, void *arg, + struct bpf_insn *insns, s64 dst, s64 src) +{ + int op = *(int *)arg; + + if (insns) { + bool match = __bpf_match_jmp_cond((s32)dst, (s32)src, op); + int i = 0; + + i += __bpf_ld_imm64(&insns[i], R1, dst); + i += __bpf_ld_imm64(&insns[i], R2, src); + insns[i++] = BPF_JMP32_REG(op, R1, R2, 1); + if (!match) + insns[i++] = BPF_JMP_IMM(BPF_JA, 0, 0, 1); + insns[i++] = BPF_EXIT_INSN(); + + return i; + } + + return 7; +} + +static int __bpf_fill_jmp_imm(struct bpf_test *self, int op) +{ + return __bpf_fill_pattern(self, &op, 64, 32, + PATTERN_BLOCK1, PATTERN_BLOCK2, + &__bpf_emit_jmp_imm); +} + +static int __bpf_fill_jmp32_imm(struct bpf_test *self, int op) +{ + return __bpf_fill_pattern(self, &op, 64, 32, + PATTERN_BLOCK1, PATTERN_BLOCK2, + &__bpf_emit_jmp32_imm); +} + +static int __bpf_fill_jmp_reg(struct bpf_test *self, int op) +{ + return __bpf_fill_pattern(self, &op, 64, 64, + PATTERN_BLOCK1, PATTERN_BLOCK2, + &__bpf_emit_jmp_reg); +} + +static int __bpf_fill_jmp32_reg(struct bpf_test *self, int op) +{ + return __bpf_fill_pattern(self, &op, 64, 64, + PATTERN_BLOCK1, PATTERN_BLOCK2, + &__bpf_emit_jmp32_reg); +} + +/* JMP immediate tests */ +static int bpf_fill_jmp_jset_imm(struct bpf_test *self) +{ + return __bpf_fill_jmp_imm(self, BPF_JSET); +} + +static int bpf_fill_jmp_jeq_imm(struct bpf_test *self) +{ + return __bpf_fill_jmp_imm(self, BPF_JEQ); +} + +static int bpf_fill_jmp_jne_imm(struct bpf_test *self) +{ + return __bpf_fill_jmp_imm(self, BPF_JNE); +} + +static int bpf_fill_jmp_jgt_imm(struct bpf_test *self) +{ + return __bpf_fill_jmp_imm(self, BPF_JGT); +} + +static int bpf_fill_jmp_jge_imm(struct bpf_test *self) +{ + return __bpf_fill_jmp_imm(self, BPF_JGE); +} + +static int bpf_fill_jmp_jlt_imm(struct bpf_test *self) +{ + return __bpf_fill_jmp_imm(self, BPF_JLT); +} + +static int bpf_fill_jmp_jle_imm(struct bpf_test *self) +{ + return __bpf_fill_jmp_imm(self, BPF_JLE); +} + +static int bpf_fill_jmp_jsgt_imm(struct bpf_test *self) +{ + return __bpf_fill_jmp_imm(self, BPF_JSGT); +} + +static int bpf_fill_jmp_jsge_imm(struct bpf_test *self) +{ + return __bpf_fill_jmp_imm(self, BPF_JSGE); +} + +static int bpf_fill_jmp_jslt_imm(struct bpf_test *self) +{ + return __bpf_fill_jmp_imm(self, BPF_JSLT); +} + +static int bpf_fill_jmp_jsle_imm(struct bpf_test *self) +{ + return __bpf_fill_jmp_imm(self, BPF_JSLE); +} + +/* JMP32 immediate tests */ +static int bpf_fill_jmp32_jset_imm(struct bpf_test *self) +{ + return __bpf_fill_jmp32_imm(self, BPF_JSET); +} + +static int bpf_fill_jmp32_jeq_imm(struct bpf_test *self) +{ + return __bpf_fill_jmp32_imm(self, BPF_JEQ); +} + +static int bpf_fill_jmp32_jne_imm(struct bpf_test *self) +{ + return __bpf_fill_jmp32_imm(self, BPF_JNE); +} + +static int bpf_fill_jmp32_jgt_imm(struct bpf_test *self) +{ + return __bpf_fill_jmp32_imm(self, BPF_JGT); +} + +static int bpf_fill_jmp32_jge_imm(struct bpf_test *self) +{ + return __bpf_fill_jmp32_imm(self, BPF_JGE); +} + +static int bpf_fill_jmp32_jlt_imm(struct bpf_test *self) +{ + return __bpf_fill_jmp32_imm(self, BPF_JLT); +} + +static int bpf_fill_jmp32_jle_imm(struct bpf_test *self) +{ + return __bpf_fill_jmp32_imm(self, BPF_JLE); +} + +static int bpf_fill_jmp32_jsgt_imm(struct bpf_test *self) +{ + return __bpf_fill_jmp32_imm(self, BPF_JSGT); +} + +static int bpf_fill_jmp32_jsge_imm(struct bpf_test *self) +{ + return __bpf_fill_jmp32_imm(self, BPF_JSGE); +} + +static int bpf_fill_jmp32_jslt_imm(struct bpf_test *self) +{ + return __bpf_fill_jmp32_imm(self, BPF_JSLT); +} + +static int bpf_fill_jmp32_jsle_imm(struct bpf_test *self) +{ + return __bpf_fill_jmp32_imm(self, BPF_JSLE); +} + +/* JMP register tests */ +static int bpf_fill_jmp_jset_reg(struct bpf_test *self) +{ + return __bpf_fill_jmp_reg(self, BPF_JSET); +} + +static int bpf_fill_jmp_jeq_reg(struct bpf_test *self) +{ + return __bpf_fill_jmp_reg(self, BPF_JEQ); +} + +static int bpf_fill_jmp_jne_reg(struct bpf_test *self) +{ + return __bpf_fill_jmp_reg(self, BPF_JNE); +} + +static int bpf_fill_jmp_jgt_reg(struct bpf_test *self) +{ + return __bpf_fill_jmp_reg(self, BPF_JGT); +} + +static int bpf_fill_jmp_jge_reg(struct bpf_test *self) +{ + return __bpf_fill_jmp_reg(self, BPF_JGE); +} + +static int bpf_fill_jmp_jlt_reg(struct bpf_test *self) +{ + return __bpf_fill_jmp_reg(self, BPF_JLT); +} + +static int bpf_fill_jmp_jle_reg(struct bpf_test *self) +{ + return __bpf_fill_jmp_reg(self, BPF_JLE); +} + +static int bpf_fill_jmp_jsgt_reg(struct bpf_test *self) +{ + return __bpf_fill_jmp_reg(self, BPF_JSGT); +} + +static int bpf_fill_jmp_jsge_reg(struct bpf_test *self) +{ + return __bpf_fill_jmp_reg(self, BPF_JSGE); +} + +static int bpf_fill_jmp_jslt_reg(struct bpf_test *self) +{ + return __bpf_fill_jmp_reg(self, BPF_JSLT); +} + +static int bpf_fill_jmp_jsle_reg(struct bpf_test *self) +{ + return __bpf_fill_jmp_reg(self, BPF_JSLE); +} + +/* JMP32 register tests */ +static int bpf_fill_jmp32_jset_reg(struct bpf_test *self) +{ + return __bpf_fill_jmp32_reg(self, BPF_JSET); +} + +static int bpf_fill_jmp32_jeq_reg(struct bpf_test *self) +{ + return __bpf_fill_jmp32_reg(self, BPF_JEQ); +} + +static int bpf_fill_jmp32_jne_reg(struct bpf_test *self) +{ + return __bpf_fill_jmp32_reg(self, BPF_JNE); +} + +static int bpf_fill_jmp32_jgt_reg(struct bpf_test *self) +{ + return __bpf_fill_jmp32_reg(self, BPF_JGT); +} + +static int bpf_fill_jmp32_jge_reg(struct bpf_test *self) +{ + return __bpf_fill_jmp32_reg(self, BPF_JGE); +} + +static int bpf_fill_jmp32_jlt_reg(struct bpf_test *self) +{ + return __bpf_fill_jmp32_reg(self, BPF_JLT); +} + +static int bpf_fill_jmp32_jle_reg(struct bpf_test *self) +{ + return __bpf_fill_jmp32_reg(self, BPF_JLE); +} + +static int bpf_fill_jmp32_jsgt_reg(struct bpf_test *self) +{ + return __bpf_fill_jmp32_reg(self, BPF_JSGT); +} + +static int bpf_fill_jmp32_jsge_reg(struct bpf_test *self) +{ + return __bpf_fill_jmp32_reg(self, BPF_JSGE); +} + +static int bpf_fill_jmp32_jslt_reg(struct bpf_test *self) +{ + return __bpf_fill_jmp32_reg(self, BPF_JSLT); +} + +static int bpf_fill_jmp32_jsle_reg(struct bpf_test *self) +{ + return __bpf_fill_jmp32_reg(self, BPF_JSLE); +} + + static struct bpf_test tests[] = { { "TAX", @@ -9281,6 +9659,7 @@ static struct bpf_test tests[] = { { }, { { 0, 1 } }, .fill_helper = bpf_fill_alu32_mod_imm, + .nr_testruns = NR_PATTERN_RUNS, }, /* ALU64 register magnitudes */ { @@ -9446,6 +9825,406 @@ static struct bpf_test tests[] = { .fill_helper = bpf_fill_alu32_mod_reg, .nr_testruns = NR_PATTERN_RUNS, }, + /* JMP immediate magnitudes */ + { + "JMP_JSET_K: all immediate value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_jmp_jset_imm, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "JMP_JEQ_K: all immediate value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_jmp_jeq_imm, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "JMP_JNE_K: all immediate value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_jmp_jne_imm, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "JMP_JGT_K: all immediate value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_jmp_jgt_imm, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "JMP_JGE_K: all immediate value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_jmp_jge_imm, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "JMP_JLT_K: all immediate value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_jmp_jlt_imm, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "JMP_JLE_K: all immediate value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_jmp_jle_imm, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "JMP_JSGT_K: all immediate value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_jmp_jsgt_imm, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "JMP_JSGE_K: all immediate value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_jmp_jsge_imm, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "JMP_JSLT_K: all immediate value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_jmp_jslt_imm, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "JMP_JSLE_K: all immediate value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_jmp_jsle_imm, + .nr_testruns = NR_PATTERN_RUNS, + }, + /* JMP register magnitudes */ + { + "JMP_JSET_X: all register value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_jmp_jset_reg, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "JMP_JEQ_X: all register value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_jmp_jeq_reg, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "JMP_JNE_X: all register value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_jmp_jne_reg, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "JMP_JGT_X: all register value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_jmp_jgt_reg, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "JMP_JGE_X: all register value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_jmp_jge_reg, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "JMP_JLT_X: all register value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_jmp_jlt_reg, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "JMP_JLE_X: all register value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_jmp_jle_reg, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "JMP_JSGT_X: all register value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_jmp_jsgt_reg, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "JMP_JSGE_X: all register value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_jmp_jsge_reg, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "JMP_JSLT_X: all register value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_jmp_jslt_reg, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "JMP_JSLE_X: all register value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_jmp_jsle_reg, + .nr_testruns = NR_PATTERN_RUNS, + }, + /* JMP32 immediate magnitudes */ + { + "JMP32_JSET_K: all immediate value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_jmp32_jset_imm, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "JMP32_JEQ_K: all immediate value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_jmp32_jeq_imm, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "JMP32_JNE_K: all immediate value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_jmp32_jne_imm, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "JMP32_JGT_K: all immediate value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_jmp32_jgt_imm, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "JMP32_JGE_K: all immediate value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_jmp32_jge_imm, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "JMP32_JLT_K: all immediate value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_jmp32_jlt_imm, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "JMP32_JLE_K: all immediate value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_jmp32_jle_imm, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "JMP32_JSGT_K: all immediate value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_jmp32_jsgt_imm, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "JMP32_JSGE_K: all immediate value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_jmp32_jsge_imm, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "JMP32_JSLT_K: all immediate value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_jmp32_jslt_imm, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "JMP32_JSLE_K: all immediate value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_jmp32_jsle_imm, + .nr_testruns = NR_PATTERN_RUNS, + }, + /* JMP32 register magnitudes */ + { + "JMP32_JSET_X: all register value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_jmp32_jset_reg, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "JMP32_JEQ_X: all register value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_jmp32_jeq_reg, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "JMP32_JNE_X: all register value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_jmp32_jne_reg, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "JMP32_JGT_X: all register value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_jmp32_jgt_reg, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "JMP32_JGE_X: all register value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_jmp32_jge_reg, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "JMP32_JLT_X: all register value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_jmp32_jlt_reg, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "JMP32_JLE_X: all register value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_jmp32_jle_reg, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "JMP32_JSGT_X: all register value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_jmp32_jsgt_reg, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "JMP32_JSGE_X: all register value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_jmp32_jsge_reg, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "JMP32_JSLT_X: all register value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_jmp32_jslt_reg, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "JMP32_JSLE_X: all register value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_jmp32_jsle_reg, + .nr_testruns = NR_PATTERN_RUNS, + }, }; static struct net_device dev; -- cgit v1.2.3 From a7d2e752e52050fcdf0c50cf343488891a8efd5b Mon Sep 17 00:00:00 2001 From: Johan Almbladh Date: Tue, 14 Sep 2021 11:18:34 +0200 Subject: bpf/tests: Add staggered JMP and JMP32 tests This patch adds a new type of jump test where the program jumps forwards and backwards with increasing offset. It mainly tests JITs where a relative jump may generate different JITed code depending on the offset size, read MIPS. Signed-off-by: Johan Almbladh Signed-off-by: Daniel Borkmann Link: https://lore.kernel.org/bpf/20210914091842.4186267-7-johan.almbladh@anyfinetworks.com --- lib/test_bpf.c | 829 +++++++++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 829 insertions(+) (limited to 'lib') diff --git a/lib/test_bpf.c b/lib/test_bpf.c index fb27a25e2c0c..7b3d801275e8 100644 --- a/lib/test_bpf.c +++ b/lib/test_bpf.c @@ -1481,6 +1481,426 @@ static int bpf_fill_jmp32_jsle_reg(struct bpf_test *self) return __bpf_fill_jmp32_reg(self, BPF_JSLE); } +/* + * Set up a sequence of staggered jumps, forwards and backwards with + * increasing offset. This tests the conversion of relative jumps to + * JITed native jumps. On some architectures, for example MIPS, a large + * PC-relative jump offset may overflow the immediate field of the native + * conditional branch instruction, triggering a conversion to use an + * absolute jump instead. Since this changes the jump offsets, another + * offset computation pass is necessary, and that may in turn trigger + * another branch conversion. This jump sequence is particularly nasty + * in that regard. + * + * The sequence generation is parameterized by size and jump type. + * The size must be even, and the expected result is always size + 1. + * Below is an example with size=8 and result=9. + * + * ________________________Start + * R0 = 0 + * R1 = r1 + * R2 = r2 + * ,------- JMP +4 * 3______________Preamble: 4 insns + * ,----------|-ind 0- if R0 != 7 JMP 8 * 3 + 1 <--------------------. + * | | R0 = 8 | + * | | JMP +7 * 3 ------------------------. + * | ,--------|-----1- if R0 != 5 JMP 7 * 3 + 1 <--------------. | | + * | | | R0 = 6 | | | + * | | | JMP +5 * 3 ------------------. | | + * | | ,------|-----2- if R0 != 3 JMP 6 * 3 + 1 <--------. | | | | + * | | | | R0 = 4 | | | | | + * | | | | JMP +3 * 3 ------------. | | | | + * | | | ,----|-----3- if R0 != 1 JMP 5 * 3 + 1 <--. | | | | | | + * | | | | | R0 = 2 | | | | | | | + * | | | | | JMP +1 * 3 ------. | | | | | | + * | | | | ,--t=====4> if R0 != 0 JMP 4 * 3 + 1 1 2 3 4 5 6 7 8 loc + * | | | | | R0 = 1 -1 +2 -3 +4 -5 +6 -7 +8 off + * | | | | | JMP -2 * 3 ---' | | | | | | | + * | | | | | ,------5- if R0 != 2 JMP 3 * 3 + 1 <-----' | | | | | | + * | | | | | | R0 = 3 | | | | | | + * | | | | | | JMP -4 * 3 ---------' | | | | | + * | | | | | | ,----6- if R0 != 4 JMP 2 * 3 + 1 <-----------' | | | | + * | | | | | | | R0 = 5 | | | | + * | | | | | | | JMP -6 * 3 ---------------' | | | + * | | | | | | | ,--7- if R0 != 6 JMP 1 * 3 + 1 <-----------------' | | + * | | | | | | | | R0 = 7 | | + * | | Error | | | JMP -8 * 3 ---------------------' | + * | | paths | | | ,8- if R0 != 8 JMP 0 * 3 + 1 <-----------------------' + * | | | | | | | | | R0 = 9__________________Sequence: 3 * size - 1 insns + * `-+-+-+-+-+-+-+-+-> EXIT____________________Return: 1 insn + * + */ + +/* The maximum size parameter */ +#define MAX_STAGGERED_JMP_SIZE ((0x7fff / 3) & ~1) + +/* We use a reduced number of iterations to get a reasonable execution time */ +#define NR_STAGGERED_JMP_RUNS 10 + +static int __bpf_fill_staggered_jumps(struct bpf_test *self, + const struct bpf_insn *jmp, + u64 r1, u64 r2) +{ + int size = self->test[0].result - 1; + int len = 4 + 3 * (size + 1); + struct bpf_insn *insns; + int off, ind; + + insns = kmalloc_array(len, sizeof(*insns), GFP_KERNEL); + if (!insns) + return -ENOMEM; + + /* Preamble */ + insns[0] = BPF_ALU64_IMM(BPF_MOV, R0, 0); + insns[1] = BPF_ALU64_IMM(BPF_MOV, R1, r1); + insns[2] = BPF_ALU64_IMM(BPF_MOV, R2, r2); + insns[3] = BPF_JMP_IMM(BPF_JA, 0, 0, 3 * size / 2); + + /* Sequence */ + for (ind = 0, off = size; ind <= size; ind++, off -= 2) { + struct bpf_insn *ins = &insns[4 + 3 * ind]; + int loc; + + if (off == 0) + off--; + + loc = abs(off); + ins[0] = BPF_JMP_IMM(BPF_JNE, R0, loc - 1, + 3 * (size - ind) + 1); + ins[1] = BPF_ALU64_IMM(BPF_MOV, R0, loc); + ins[2] = *jmp; + ins[2].off = 3 * (off - 1); + } + + /* Return */ + insns[len - 1] = BPF_EXIT_INSN(); + + self->u.ptr.insns = insns; + self->u.ptr.len = len; + + return 0; +} + +/* 64-bit unconditional jump */ +static int bpf_fill_staggered_ja(struct bpf_test *self) +{ + struct bpf_insn jmp = BPF_JMP_IMM(BPF_JA, 0, 0, 0); + + return __bpf_fill_staggered_jumps(self, &jmp, 0, 0); +} + +/* 64-bit immediate jumps */ +static int bpf_fill_staggered_jeq_imm(struct bpf_test *self) +{ + struct bpf_insn jmp = BPF_JMP_IMM(BPF_JEQ, R1, 1234, 0); + + return __bpf_fill_staggered_jumps(self, &jmp, 1234, 0); +} + +static int bpf_fill_staggered_jne_imm(struct bpf_test *self) +{ + struct bpf_insn jmp = BPF_JMP_IMM(BPF_JNE, R1, 1234, 0); + + return __bpf_fill_staggered_jumps(self, &jmp, 4321, 0); +} + +static int bpf_fill_staggered_jset_imm(struct bpf_test *self) +{ + struct bpf_insn jmp = BPF_JMP_IMM(BPF_JSET, R1, 0x82, 0); + + return __bpf_fill_staggered_jumps(self, &jmp, 0x86, 0); +} + +static int bpf_fill_staggered_jgt_imm(struct bpf_test *self) +{ + struct bpf_insn jmp = BPF_JMP_IMM(BPF_JGT, R1, 1234, 0); + + return __bpf_fill_staggered_jumps(self, &jmp, 0x80000000, 0); +} + +static int bpf_fill_staggered_jge_imm(struct bpf_test *self) +{ + struct bpf_insn jmp = BPF_JMP_IMM(BPF_JGE, R1, 1234, 0); + + return __bpf_fill_staggered_jumps(self, &jmp, 1234, 0); +} + +static int bpf_fill_staggered_jlt_imm(struct bpf_test *self) +{ + struct bpf_insn jmp = BPF_JMP_IMM(BPF_JLT, R1, 0x80000000, 0); + + return __bpf_fill_staggered_jumps(self, &jmp, 1234, 0); +} + +static int bpf_fill_staggered_jle_imm(struct bpf_test *self) +{ + struct bpf_insn jmp = BPF_JMP_IMM(BPF_JLE, R1, 1234, 0); + + return __bpf_fill_staggered_jumps(self, &jmp, 1234, 0); +} + +static int bpf_fill_staggered_jsgt_imm(struct bpf_test *self) +{ + struct bpf_insn jmp = BPF_JMP_IMM(BPF_JSGT, R1, -2, 0); + + return __bpf_fill_staggered_jumps(self, &jmp, -1, 0); +} + +static int bpf_fill_staggered_jsge_imm(struct bpf_test *self) +{ + struct bpf_insn jmp = BPF_JMP_IMM(BPF_JSGE, R1, -2, 0); + + return __bpf_fill_staggered_jumps(self, &jmp, -2, 0); +} + +static int bpf_fill_staggered_jslt_imm(struct bpf_test *self) +{ + struct bpf_insn jmp = BPF_JMP_IMM(BPF_JSLT, R1, -1, 0); + + return __bpf_fill_staggered_jumps(self, &jmp, -2, 0); +} + +static int bpf_fill_staggered_jsle_imm(struct bpf_test *self) +{ + struct bpf_insn jmp = BPF_JMP_IMM(BPF_JSLE, R1, -1, 0); + + return __bpf_fill_staggered_jumps(self, &jmp, -1, 0); +} + +/* 64-bit register jumps */ +static int bpf_fill_staggered_jeq_reg(struct bpf_test *self) +{ + struct bpf_insn jmp = BPF_JMP_REG(BPF_JEQ, R1, R2, 0); + + return __bpf_fill_staggered_jumps(self, &jmp, 1234, 1234); +} + +static int bpf_fill_staggered_jne_reg(struct bpf_test *self) +{ + struct bpf_insn jmp = BPF_JMP_REG(BPF_JNE, R1, R2, 0); + + return __bpf_fill_staggered_jumps(self, &jmp, 4321, 1234); +} + +static int bpf_fill_staggered_jset_reg(struct bpf_test *self) +{ + struct bpf_insn jmp = BPF_JMP_REG(BPF_JSET, R1, R2, 0); + + return __bpf_fill_staggered_jumps(self, &jmp, 0x86, 0x82); +} + +static int bpf_fill_staggered_jgt_reg(struct bpf_test *self) +{ + struct bpf_insn jmp = BPF_JMP_REG(BPF_JGT, R1, R2, 0); + + return __bpf_fill_staggered_jumps(self, &jmp, 0x80000000, 1234); +} + +static int bpf_fill_staggered_jge_reg(struct bpf_test *self) +{ + struct bpf_insn jmp = BPF_JMP_REG(BPF_JGE, R1, R2, 0); + + return __bpf_fill_staggered_jumps(self, &jmp, 1234, 1234); +} + +static int bpf_fill_staggered_jlt_reg(struct bpf_test *self) +{ + struct bpf_insn jmp = BPF_JMP_REG(BPF_JLT, R1, R2, 0); + + return __bpf_fill_staggered_jumps(self, &jmp, 1234, 0x80000000); +} + +static int bpf_fill_staggered_jle_reg(struct bpf_test *self) +{ + struct bpf_insn jmp = BPF_JMP_REG(BPF_JLE, R1, R2, 0); + + return __bpf_fill_staggered_jumps(self, &jmp, 1234, 1234); +} + +static int bpf_fill_staggered_jsgt_reg(struct bpf_test *self) +{ + struct bpf_insn jmp = BPF_JMP_REG(BPF_JSGT, R1, R2, 0); + + return __bpf_fill_staggered_jumps(self, &jmp, -1, -2); +} + +static int bpf_fill_staggered_jsge_reg(struct bpf_test *self) +{ + struct bpf_insn jmp = BPF_JMP_REG(BPF_JSGE, R1, R2, 0); + + return __bpf_fill_staggered_jumps(self, &jmp, -2, -2); +} + +static int bpf_fill_staggered_jslt_reg(struct bpf_test *self) +{ + struct bpf_insn jmp = BPF_JMP_REG(BPF_JSLT, R1, R2, 0); + + return __bpf_fill_staggered_jumps(self, &jmp, -2, -1); +} + +static int bpf_fill_staggered_jsle_reg(struct bpf_test *self) +{ + struct bpf_insn jmp = BPF_JMP_REG(BPF_JSLE, R1, R2, 0); + + return __bpf_fill_staggered_jumps(self, &jmp, -1, -1); +} + +/* 32-bit immediate jumps */ +static int bpf_fill_staggered_jeq32_imm(struct bpf_test *self) +{ + struct bpf_insn jmp = BPF_JMP32_IMM(BPF_JEQ, R1, 1234, 0); + + return __bpf_fill_staggered_jumps(self, &jmp, 1234, 0); +} + +static int bpf_fill_staggered_jne32_imm(struct bpf_test *self) +{ + struct bpf_insn jmp = BPF_JMP32_IMM(BPF_JNE, R1, 1234, 0); + + return __bpf_fill_staggered_jumps(self, &jmp, 4321, 0); +} + +static int bpf_fill_staggered_jset32_imm(struct bpf_test *self) +{ + struct bpf_insn jmp = BPF_JMP32_IMM(BPF_JSET, R1, 0x82, 0); + + return __bpf_fill_staggered_jumps(self, &jmp, 0x86, 0); +} + +static int bpf_fill_staggered_jgt32_imm(struct bpf_test *self) +{ + struct bpf_insn jmp = BPF_JMP32_IMM(BPF_JGT, R1, 1234, 0); + + return __bpf_fill_staggered_jumps(self, &jmp, 0x80000000, 0); +} + +static int bpf_fill_staggered_jge32_imm(struct bpf_test *self) +{ + struct bpf_insn jmp = BPF_JMP32_IMM(BPF_JGE, R1, 1234, 0); + + return __bpf_fill_staggered_jumps(self, &jmp, 1234, 0); +} + +static int bpf_fill_staggered_jlt32_imm(struct bpf_test *self) +{ + struct bpf_insn jmp = BPF_JMP32_IMM(BPF_JLT, R1, 0x80000000, 0); + + return __bpf_fill_staggered_jumps(self, &jmp, 1234, 0); +} + +static int bpf_fill_staggered_jle32_imm(struct bpf_test *self) +{ + struct bpf_insn jmp = BPF_JMP32_IMM(BPF_JLE, R1, 1234, 0); + + return __bpf_fill_staggered_jumps(self, &jmp, 1234, 0); +} + +static int bpf_fill_staggered_jsgt32_imm(struct bpf_test *self) +{ + struct bpf_insn jmp = BPF_JMP32_IMM(BPF_JSGT, R1, -2, 0); + + return __bpf_fill_staggered_jumps(self, &jmp, -1, 0); +} + +static int bpf_fill_staggered_jsge32_imm(struct bpf_test *self) +{ + struct bpf_insn jmp = BPF_JMP32_IMM(BPF_JSGE, R1, -2, 0); + + return __bpf_fill_staggered_jumps(self, &jmp, -2, 0); +} + +static int bpf_fill_staggered_jslt32_imm(struct bpf_test *self) +{ + struct bpf_insn jmp = BPF_JMP32_IMM(BPF_JSLT, R1, -1, 0); + + return __bpf_fill_staggered_jumps(self, &jmp, -2, 0); +} + +static int bpf_fill_staggered_jsle32_imm(struct bpf_test *self) +{ + struct bpf_insn jmp = BPF_JMP32_IMM(BPF_JSLE, R1, -1, 0); + + return __bpf_fill_staggered_jumps(self, &jmp, -1, 0); +} + +/* 32-bit register jumps */ +static int bpf_fill_staggered_jeq32_reg(struct bpf_test *self) +{ + struct bpf_insn jmp = BPF_JMP32_REG(BPF_JEQ, R1, R2, 0); + + return __bpf_fill_staggered_jumps(self, &jmp, 1234, 1234); +} + +static int bpf_fill_staggered_jne32_reg(struct bpf_test *self) +{ + struct bpf_insn jmp = BPF_JMP32_REG(BPF_JNE, R1, R2, 0); + + return __bpf_fill_staggered_jumps(self, &jmp, 4321, 1234); +} + +static int bpf_fill_staggered_jset32_reg(struct bpf_test *self) +{ + struct bpf_insn jmp = BPF_JMP32_REG(BPF_JSET, R1, R2, 0); + + return __bpf_fill_staggered_jumps(self, &jmp, 0x86, 0x82); +} + +static int bpf_fill_staggered_jgt32_reg(struct bpf_test *self) +{ + struct bpf_insn jmp = BPF_JMP32_REG(BPF_JGT, R1, R2, 0); + + return __bpf_fill_staggered_jumps(self, &jmp, 0x80000000, 1234); +} + +static int bpf_fill_staggered_jge32_reg(struct bpf_test *self) +{ + struct bpf_insn jmp = BPF_JMP32_REG(BPF_JGE, R1, R2, 0); + + return __bpf_fill_staggered_jumps(self, &jmp, 1234, 1234); +} + +static int bpf_fill_staggered_jlt32_reg(struct bpf_test *self) +{ + struct bpf_insn jmp = BPF_JMP32_REG(BPF_JLT, R1, R2, 0); + + return __bpf_fill_staggered_jumps(self, &jmp, 1234, 0x80000000); +} + +static int bpf_fill_staggered_jle32_reg(struct bpf_test *self) +{ + struct bpf_insn jmp = BPF_JMP32_REG(BPF_JLE, R1, R2, 0); + + return __bpf_fill_staggered_jumps(self, &jmp, 1234, 1234); +} + +static int bpf_fill_staggered_jsgt32_reg(struct bpf_test *self) +{ + struct bpf_insn jmp = BPF_JMP32_REG(BPF_JSGT, R1, R2, 0); + + return __bpf_fill_staggered_jumps(self, &jmp, -1, -2); +} + +static int bpf_fill_staggered_jsge32_reg(struct bpf_test *self) +{ + struct bpf_insn jmp = BPF_JMP32_REG(BPF_JSGE, R1, R2, 0); + + return __bpf_fill_staggered_jumps(self, &jmp, -2, -2); +} + +static int bpf_fill_staggered_jslt32_reg(struct bpf_test *self) +{ + struct bpf_insn jmp = BPF_JMP32_REG(BPF_JSLT, R1, R2, 0); + + return __bpf_fill_staggered_jumps(self, &jmp, -2, -1); +} + +static int bpf_fill_staggered_jsle32_reg(struct bpf_test *self) +{ + struct bpf_insn jmp = BPF_JMP32_REG(BPF_JSLE, R1, R2, 0); + + return __bpf_fill_staggered_jumps(self, &jmp, -1, -1); +} + static struct bpf_test tests[] = { { @@ -10225,6 +10645,415 @@ static struct bpf_test tests[] = { .fill_helper = bpf_fill_jmp32_jsle_reg, .nr_testruns = NR_PATTERN_RUNS, }, + /* Staggered jump sequences, immediate */ + { + "Staggered jumps: JMP_JA", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, MAX_STAGGERED_JMP_SIZE + 1 } }, + .fill_helper = bpf_fill_staggered_ja, + .nr_testruns = NR_STAGGERED_JMP_RUNS, + }, + { + "Staggered jumps: JMP_JEQ_K", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, MAX_STAGGERED_JMP_SIZE + 1 } }, + .fill_helper = bpf_fill_staggered_jeq_imm, + .nr_testruns = NR_STAGGERED_JMP_RUNS, + }, + { + "Staggered jumps: JMP_JNE_K", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, MAX_STAGGERED_JMP_SIZE + 1 } }, + .fill_helper = bpf_fill_staggered_jne_imm, + .nr_testruns = NR_STAGGERED_JMP_RUNS, + }, + { + "Staggered jumps: JMP_JSET_K", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, MAX_STAGGERED_JMP_SIZE + 1 } }, + .fill_helper = bpf_fill_staggered_jset_imm, + .nr_testruns = NR_STAGGERED_JMP_RUNS, + }, + { + "Staggered jumps: JMP_JGT_K", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, MAX_STAGGERED_JMP_SIZE + 1 } }, + .fill_helper = bpf_fill_staggered_jgt_imm, + .nr_testruns = NR_STAGGERED_JMP_RUNS, + }, + { + "Staggered jumps: JMP_JGE_K", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, MAX_STAGGERED_JMP_SIZE + 1 } }, + .fill_helper = bpf_fill_staggered_jge_imm, + .nr_testruns = NR_STAGGERED_JMP_RUNS, + }, + { + "Staggered jumps: JMP_JLT_K", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, MAX_STAGGERED_JMP_SIZE + 1 } }, + .fill_helper = bpf_fill_staggered_jlt_imm, + .nr_testruns = NR_STAGGERED_JMP_RUNS, + }, + { + "Staggered jumps: JMP_JLE_K", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, MAX_STAGGERED_JMP_SIZE + 1 } }, + .fill_helper = bpf_fill_staggered_jle_imm, + .nr_testruns = NR_STAGGERED_JMP_RUNS, + }, + { + "Staggered jumps: JMP_JSGT_K", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, MAX_STAGGERED_JMP_SIZE + 1 } }, + .fill_helper = bpf_fill_staggered_jsgt_imm, + .nr_testruns = NR_STAGGERED_JMP_RUNS, + }, + { + "Staggered jumps: JMP_JSGE_K", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, MAX_STAGGERED_JMP_SIZE + 1 } }, + .fill_helper = bpf_fill_staggered_jsge_imm, + .nr_testruns = NR_STAGGERED_JMP_RUNS, + }, + { + "Staggered jumps: JMP_JSLT_K", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, MAX_STAGGERED_JMP_SIZE + 1 } }, + .fill_helper = bpf_fill_staggered_jslt_imm, + .nr_testruns = NR_STAGGERED_JMP_RUNS, + }, + { + "Staggered jumps: JMP_JSLE_K", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, MAX_STAGGERED_JMP_SIZE + 1 } }, + .fill_helper = bpf_fill_staggered_jsle_imm, + .nr_testruns = NR_STAGGERED_JMP_RUNS, + }, + /* Staggered jump sequences, register */ + { + "Staggered jumps: JMP_JEQ_X", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, MAX_STAGGERED_JMP_SIZE + 1 } }, + .fill_helper = bpf_fill_staggered_jeq_reg, + .nr_testruns = NR_STAGGERED_JMP_RUNS, + }, + { + "Staggered jumps: JMP_JNE_X", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, MAX_STAGGERED_JMP_SIZE + 1 } }, + .fill_helper = bpf_fill_staggered_jne_reg, + .nr_testruns = NR_STAGGERED_JMP_RUNS, + }, + { + "Staggered jumps: JMP_JSET_X", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, MAX_STAGGERED_JMP_SIZE + 1 } }, + .fill_helper = bpf_fill_staggered_jset_reg, + .nr_testruns = NR_STAGGERED_JMP_RUNS, + }, + { + "Staggered jumps: JMP_JGT_X", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, MAX_STAGGERED_JMP_SIZE + 1 } }, + .fill_helper = bpf_fill_staggered_jgt_reg, + .nr_testruns = NR_STAGGERED_JMP_RUNS, + }, + { + "Staggered jumps: JMP_JGE_X", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, MAX_STAGGERED_JMP_SIZE + 1 } }, + .fill_helper = bpf_fill_staggered_jge_reg, + .nr_testruns = NR_STAGGERED_JMP_RUNS, + }, + { + "Staggered jumps: JMP_JLT_X", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, MAX_STAGGERED_JMP_SIZE + 1 } }, + .fill_helper = bpf_fill_staggered_jlt_reg, + .nr_testruns = NR_STAGGERED_JMP_RUNS, + }, + { + "Staggered jumps: JMP_JLE_X", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, MAX_STAGGERED_JMP_SIZE + 1 } }, + .fill_helper = bpf_fill_staggered_jle_reg, + .nr_testruns = NR_STAGGERED_JMP_RUNS, + }, + { + "Staggered jumps: JMP_JSGT_X", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, MAX_STAGGERED_JMP_SIZE + 1 } }, + .fill_helper = bpf_fill_staggered_jsgt_reg, + .nr_testruns = NR_STAGGERED_JMP_RUNS, + }, + { + "Staggered jumps: JMP_JSGE_X", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, MAX_STAGGERED_JMP_SIZE + 1 } }, + .fill_helper = bpf_fill_staggered_jsge_reg, + .nr_testruns = NR_STAGGERED_JMP_RUNS, + }, + { + "Staggered jumps: JMP_JSLT_X", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, MAX_STAGGERED_JMP_SIZE + 1 } }, + .fill_helper = bpf_fill_staggered_jslt_reg, + .nr_testruns = NR_STAGGERED_JMP_RUNS, + }, + { + "Staggered jumps: JMP_JSLE_X", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, MAX_STAGGERED_JMP_SIZE + 1 } }, + .fill_helper = bpf_fill_staggered_jsle_reg, + .nr_testruns = NR_STAGGERED_JMP_RUNS, + }, + /* Staggered jump sequences, JMP32 immediate */ + { + "Staggered jumps: JMP32_JEQ_K", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, MAX_STAGGERED_JMP_SIZE + 1 } }, + .fill_helper = bpf_fill_staggered_jeq32_imm, + .nr_testruns = NR_STAGGERED_JMP_RUNS, + }, + { + "Staggered jumps: JMP32_JNE_K", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, MAX_STAGGERED_JMP_SIZE + 1 } }, + .fill_helper = bpf_fill_staggered_jne32_imm, + .nr_testruns = NR_STAGGERED_JMP_RUNS, + }, + { + "Staggered jumps: JMP32_JSET_K", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, MAX_STAGGERED_JMP_SIZE + 1 } }, + .fill_helper = bpf_fill_staggered_jset32_imm, + .nr_testruns = NR_STAGGERED_JMP_RUNS, + }, + { + "Staggered jumps: JMP32_JGT_K", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, MAX_STAGGERED_JMP_SIZE + 1 } }, + .fill_helper = bpf_fill_staggered_jgt32_imm, + .nr_testruns = NR_STAGGERED_JMP_RUNS, + }, + { + "Staggered jumps: JMP32_JGE_K", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, MAX_STAGGERED_JMP_SIZE + 1 } }, + .fill_helper = bpf_fill_staggered_jge32_imm, + .nr_testruns = NR_STAGGERED_JMP_RUNS, + }, + { + "Staggered jumps: JMP32_JLT_K", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, MAX_STAGGERED_JMP_SIZE + 1 } }, + .fill_helper = bpf_fill_staggered_jlt32_imm, + .nr_testruns = NR_STAGGERED_JMP_RUNS, + }, + { + "Staggered jumps: JMP32_JLE_K", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, MAX_STAGGERED_JMP_SIZE + 1 } }, + .fill_helper = bpf_fill_staggered_jle32_imm, + .nr_testruns = NR_STAGGERED_JMP_RUNS, + }, + { + "Staggered jumps: JMP32_JSGT_K", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, MAX_STAGGERED_JMP_SIZE + 1 } }, + .fill_helper = bpf_fill_staggered_jsgt32_imm, + .nr_testruns = NR_STAGGERED_JMP_RUNS, + }, + { + "Staggered jumps: JMP32_JSGE_K", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, MAX_STAGGERED_JMP_SIZE + 1 } }, + .fill_helper = bpf_fill_staggered_jsge32_imm, + .nr_testruns = NR_STAGGERED_JMP_RUNS, + }, + { + "Staggered jumps: JMP32_JSLT_K", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, MAX_STAGGERED_JMP_SIZE + 1 } }, + .fill_helper = bpf_fill_staggered_jslt32_imm, + .nr_testruns = NR_STAGGERED_JMP_RUNS, + }, + { + "Staggered jumps: JMP32_JSLE_K", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, MAX_STAGGERED_JMP_SIZE + 1 } }, + .fill_helper = bpf_fill_staggered_jsle32_imm, + .nr_testruns = NR_STAGGERED_JMP_RUNS, + }, + /* Staggered jump sequences, JMP32 register */ + { + "Staggered jumps: JMP32_JEQ_X", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, MAX_STAGGERED_JMP_SIZE + 1 } }, + .fill_helper = bpf_fill_staggered_jeq32_reg, + .nr_testruns = NR_STAGGERED_JMP_RUNS, + }, + { + "Staggered jumps: JMP32_JNE_X", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, MAX_STAGGERED_JMP_SIZE + 1 } }, + .fill_helper = bpf_fill_staggered_jne32_reg, + .nr_testruns = NR_STAGGERED_JMP_RUNS, + }, + { + "Staggered jumps: JMP32_JSET_X", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, MAX_STAGGERED_JMP_SIZE + 1 } }, + .fill_helper = bpf_fill_staggered_jset32_reg, + .nr_testruns = NR_STAGGERED_JMP_RUNS, + }, + { + "Staggered jumps: JMP32_JGT_X", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, MAX_STAGGERED_JMP_SIZE + 1 } }, + .fill_helper = bpf_fill_staggered_jgt32_reg, + .nr_testruns = NR_STAGGERED_JMP_RUNS, + }, + { + "Staggered jumps: JMP32_JGE_X", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, MAX_STAGGERED_JMP_SIZE + 1 } }, + .fill_helper = bpf_fill_staggered_jge32_reg, + .nr_testruns = NR_STAGGERED_JMP_RUNS, + }, + { + "Staggered jumps: JMP32_JLT_X", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, MAX_STAGGERED_JMP_SIZE + 1 } }, + .fill_helper = bpf_fill_staggered_jlt32_reg, + .nr_testruns = NR_STAGGERED_JMP_RUNS, + }, + { + "Staggered jumps: JMP32_JLE_X", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, MAX_STAGGERED_JMP_SIZE + 1 } }, + .fill_helper = bpf_fill_staggered_jle32_reg, + .nr_testruns = NR_STAGGERED_JMP_RUNS, + }, + { + "Staggered jumps: JMP32_JSGT_X", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, MAX_STAGGERED_JMP_SIZE + 1 } }, + .fill_helper = bpf_fill_staggered_jsgt32_reg, + .nr_testruns = NR_STAGGERED_JMP_RUNS, + }, + { + "Staggered jumps: JMP32_JSGE_X", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, MAX_STAGGERED_JMP_SIZE + 1 } }, + .fill_helper = bpf_fill_staggered_jsge32_reg, + .nr_testruns = NR_STAGGERED_JMP_RUNS, + }, + { + "Staggered jumps: JMP32_JSLT_X", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, MAX_STAGGERED_JMP_SIZE + 1 } }, + .fill_helper = bpf_fill_staggered_jslt32_reg, + .nr_testruns = NR_STAGGERED_JMP_RUNS, + }, + { + "Staggered jumps: JMP32_JSLE_X", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, MAX_STAGGERED_JMP_SIZE + 1 } }, + .fill_helper = bpf_fill_staggered_jsle32_reg, + .nr_testruns = NR_STAGGERED_JMP_RUNS, + }, }; static struct net_device dev; -- cgit v1.2.3 From 2e807611945c2d36e25d10bc6f932e5f9943deea Mon Sep 17 00:00:00 2001 From: Johan Almbladh Date: Tue, 14 Sep 2021 11:18:35 +0200 Subject: bpf/tests: Add exhaustive test of LD_IMM64 immediate magnitudes This patch adds a test for the 64-bit immediate load, a two-instruction operation, to verify correctness for all possible magnitudes of the immediate operand. Mainly intended for JIT testing. Signed-off-by: Johan Almbladh Signed-off-by: Daniel Borkmann Link: https://lore.kernel.org/bpf/20210914091842.4186267-8-johan.almbladh@anyfinetworks.com --- lib/test_bpf.c | 63 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 63 insertions(+) (limited to 'lib') diff --git a/lib/test_bpf.c b/lib/test_bpf.c index 7b3d801275e8..8aef42cfb5fe 100644 --- a/lib/test_bpf.c +++ b/lib/test_bpf.c @@ -1104,6 +1104,60 @@ static int bpf_fill_alu32_mod_reg(struct bpf_test *self) return __bpf_fill_alu32_reg(self, BPF_MOD); } +/* + * Test the two-instruction 64-bit immediate load operation for all + * power-of-two magnitudes of the immediate operand. For each MSB, a block + * of immediate values centered around the power-of-two MSB are tested, + * both for positive and negative values. The test is designed to verify + * the operation for JITs that emit different code depending on the magnitude + * of the immediate value. This is often the case if the native instruction + * immediate field width is narrower than 32 bits. + */ +static int bpf_fill_ld_imm64(struct bpf_test *self) +{ + int block = 64; /* Increase for more tests per MSB position */ + int len = 3 + 8 * 63 * block * 2; + struct bpf_insn *insn; + int bit, adj, sign; + int i = 0; + + insn = kmalloc_array(len, sizeof(*insn), GFP_KERNEL); + if (!insn) + return -ENOMEM; + + insn[i++] = BPF_ALU64_IMM(BPF_MOV, R0, 0); + + for (bit = 0; bit <= 62; bit++) { + for (adj = -block / 2; adj < block / 2; adj++) { + for (sign = -1; sign <= 1; sign += 2) { + s64 imm = sign * ((1LL << bit) + adj); + + /* Perform operation */ + i += __bpf_ld_imm64(&insn[i], R1, imm); + + /* Load reference */ + insn[i++] = BPF_ALU32_IMM(BPF_MOV, R2, imm); + insn[i++] = BPF_ALU32_IMM(BPF_MOV, R3, + (u32)(imm >> 32)); + insn[i++] = BPF_ALU64_IMM(BPF_LSH, R3, 32); + insn[i++] = BPF_ALU64_REG(BPF_OR, R2, R3); + + /* Check result */ + insn[i++] = BPF_JMP_REG(BPF_JEQ, R1, R2, 1); + insn[i++] = BPF_EXIT_INSN(); + } + } + } + + insn[i++] = BPF_ALU64_IMM(BPF_MOV, R0, 1); + insn[i++] = BPF_EXIT_INSN(); + + self->u.ptr.insns = insn; + self->u.ptr.len = len; + BUG_ON(i != len); + + return 0; +} /* * Exhaustive tests of JMP operations for all combinations of power-of-two @@ -10245,6 +10299,15 @@ static struct bpf_test tests[] = { .fill_helper = bpf_fill_alu32_mod_reg, .nr_testruns = NR_PATTERN_RUNS, }, + /* LD_IMM64 immediate magnitudes */ + { + "LD_IMM64: all immediate value magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_ld_imm64, + }, /* JMP immediate magnitudes */ { "JMP_JSET_K: all immediate value magnitudes", -- cgit v1.2.3 From 27cc6dac6ec816cc31be9031edbee3e519234471 Mon Sep 17 00:00:00 2001 From: Johan Almbladh Date: Tue, 14 Sep 2021 11:18:36 +0200 Subject: bpf/tests: Add test case flag for verifier zero-extension This patch adds a new flag to indicate that the verified did insert zero-extensions, even though the verifier is not being run for any of the tests. Signed-off-by: Johan Almbladh Signed-off-by: Daniel Borkmann Link: https://lore.kernel.org/bpf/20210914091842.4186267-9-johan.almbladh@anyfinetworks.com --- lib/test_bpf.c | 3 +++ 1 file changed, 3 insertions(+) (limited to 'lib') diff --git a/lib/test_bpf.c b/lib/test_bpf.c index 8aef42cfb5fe..0c1c8cccf254 100644 --- a/lib/test_bpf.c +++ b/lib/test_bpf.c @@ -52,6 +52,7 @@ #define FLAG_NO_DATA BIT(0) #define FLAG_EXPECTED_FAIL BIT(1) #define FLAG_SKB_FRAG BIT(2) +#define FLAG_VERIFIER_ZEXT BIT(3) enum { CLASSIC = BIT(6), /* Old BPF instructions only. */ @@ -11280,6 +11281,8 @@ static struct bpf_prog *generate_filter(int which, int *err) fp->type = BPF_PROG_TYPE_SOCKET_FILTER; memcpy(fp->insnsi, fptr, fp->len * sizeof(struct bpf_insn)); fp->aux->stack_depth = tests[which].stack_depth; + fp->aux->verifier_zext = !!(tests[which].aux & + FLAG_VERIFIER_ZEXT); /* We cannot error here as we don't need type compatibility * checks. -- cgit v1.2.3 From d4ff9ee2dc0bbbdba204e215c8b6bf58f5773994 Mon Sep 17 00:00:00 2001 From: Johan Almbladh Date: Tue, 14 Sep 2021 11:18:37 +0200 Subject: bpf/tests: Add JMP tests with small offsets This patch adds a set of tests for JMP to verify that the JITed jump offset is calculated correctly. We pretend that the verifier has inserted any zero extensions to make the jump-over operations JIT to one instruction each, in order to control the exact JITed jump offset. Signed-off-by: Johan Almbladh Signed-off-by: Daniel Borkmann Link: https://lore.kernel.org/bpf/20210914091842.4186267-10-johan.almbladh@anyfinetworks.com --- lib/test_bpf.c | 71 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 71 insertions(+) (limited to 'lib') diff --git a/lib/test_bpf.c b/lib/test_bpf.c index 0c1c8cccf254..2eb1d0e4aff3 100644 --- a/lib/test_bpf.c +++ b/lib/test_bpf.c @@ -10709,6 +10709,77 @@ static struct bpf_test tests[] = { .fill_helper = bpf_fill_jmp32_jsle_reg, .nr_testruns = NR_PATTERN_RUNS, }, + /* Short relative jumps */ + { + "Short relative jump: offset=0", + .u.insns_int = { + BPF_ALU64_IMM(BPF_MOV, R0, 0), + BPF_JMP_IMM(BPF_JEQ, R0, 0, 0), + BPF_EXIT_INSN(), + BPF_ALU32_IMM(BPF_MOV, R0, -1), + }, + INTERNAL | FLAG_NO_DATA | FLAG_VERIFIER_ZEXT, + { }, + { { 0, 0 } }, + }, + { + "Short relative jump: offset=1", + .u.insns_int = { + BPF_ALU64_IMM(BPF_MOV, R0, 0), + BPF_JMP_IMM(BPF_JEQ, R0, 0, 1), + BPF_ALU32_IMM(BPF_ADD, R0, 1), + BPF_EXIT_INSN(), + BPF_ALU32_IMM(BPF_MOV, R0, -1), + }, + INTERNAL | FLAG_NO_DATA | FLAG_VERIFIER_ZEXT, + { }, + { { 0, 0 } }, + }, + { + "Short relative jump: offset=2", + .u.insns_int = { + BPF_ALU64_IMM(BPF_MOV, R0, 0), + BPF_JMP_IMM(BPF_JEQ, R0, 0, 2), + BPF_ALU32_IMM(BPF_ADD, R0, 1), + BPF_ALU32_IMM(BPF_ADD, R0, 1), + BPF_EXIT_INSN(), + BPF_ALU32_IMM(BPF_MOV, R0, -1), + }, + INTERNAL | FLAG_NO_DATA | FLAG_VERIFIER_ZEXT, + { }, + { { 0, 0 } }, + }, + { + "Short relative jump: offset=3", + .u.insns_int = { + BPF_ALU64_IMM(BPF_MOV, R0, 0), + BPF_JMP_IMM(BPF_JEQ, R0, 0, 3), + BPF_ALU32_IMM(BPF_ADD, R0, 1), + BPF_ALU32_IMM(BPF_ADD, R0, 1), + BPF_ALU32_IMM(BPF_ADD, R0, 1), + BPF_EXIT_INSN(), + BPF_ALU32_IMM(BPF_MOV, R0, -1), + }, + INTERNAL | FLAG_NO_DATA | FLAG_VERIFIER_ZEXT, + { }, + { { 0, 0 } }, + }, + { + "Short relative jump: offset=4", + .u.insns_int = { + BPF_ALU64_IMM(BPF_MOV, R0, 0), + BPF_JMP_IMM(BPF_JEQ, R0, 0, 4), + BPF_ALU32_IMM(BPF_ADD, R0, 1), + BPF_ALU32_IMM(BPF_ADD, R0, 1), + BPF_ALU32_IMM(BPF_ADD, R0, 1), + BPF_ALU32_IMM(BPF_ADD, R0, 1), + BPF_EXIT_INSN(), + BPF_ALU32_IMM(BPF_MOV, R0, -1), + }, + INTERNAL | FLAG_NO_DATA | FLAG_VERIFIER_ZEXT, + { }, + { { 0, 0 } }, + }, /* Staggered jump sequences, immediate */ { "Staggered jumps: JMP_JA", -- cgit v1.2.3 From c4df4559db8447cdae15254a713f7fd5d4cee3ab Mon Sep 17 00:00:00 2001 From: Johan Almbladh Date: Tue, 14 Sep 2021 11:18:38 +0200 Subject: bpf/tests: Add JMP tests with degenerate conditional This patch adds a set of tests for JMP and JMP32 operations where the branch decision is know at JIT time. Mainly testing JIT behaviour. Signed-off-by: Johan Almbladh Signed-off-by: Daniel Borkmann Link: https://lore.kernel.org/bpf/20210914091842.4186267-11-johan.almbladh@anyfinetworks.com --- lib/test_bpf.c | 229 +++++++++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 229 insertions(+) (limited to 'lib') diff --git a/lib/test_bpf.c b/lib/test_bpf.c index 2eb1d0e4aff3..2aa1d0d3c685 100644 --- a/lib/test_bpf.c +++ b/lib/test_bpf.c @@ -10709,6 +10709,235 @@ static struct bpf_test tests[] = { .fill_helper = bpf_fill_jmp32_jsle_reg, .nr_testruns = NR_PATTERN_RUNS, }, + /* Conditional jumps with constant decision */ + { + "JMP_JSET_K: imm = 0 -> never taken", + .u.insns_int = { + BPF_ALU64_IMM(BPF_MOV, R0, 1), + BPF_JMP_IMM(BPF_JSET, R1, 0, 1), + BPF_ALU64_IMM(BPF_MOV, R0, 0), + BPF_EXIT_INSN(), + }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 0 } }, + }, + { + "JMP_JLT_K: imm = 0 -> never taken", + .u.insns_int = { + BPF_ALU64_IMM(BPF_MOV, R0, 1), + BPF_JMP_IMM(BPF_JLT, R1, 0, 1), + BPF_ALU64_IMM(BPF_MOV, R0, 0), + BPF_EXIT_INSN(), + }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 0 } }, + }, + { + "JMP_JGE_K: imm = 0 -> always taken", + .u.insns_int = { + BPF_ALU64_IMM(BPF_MOV, R0, 1), + BPF_JMP_IMM(BPF_JGE, R1, 0, 1), + BPF_ALU64_IMM(BPF_MOV, R0, 0), + BPF_EXIT_INSN(), + }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + }, + { + "JMP_JGT_K: imm = 0xffffffff -> never taken", + .u.insns_int = { + BPF_ALU64_IMM(BPF_MOV, R0, 1), + BPF_JMP_IMM(BPF_JGT, R1, U32_MAX, 1), + BPF_ALU64_IMM(BPF_MOV, R0, 0), + BPF_EXIT_INSN(), + }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 0 } }, + }, + { + "JMP_JLE_K: imm = 0xffffffff -> always taken", + .u.insns_int = { + BPF_ALU64_IMM(BPF_MOV, R0, 1), + BPF_JMP_IMM(BPF_JLE, R1, U32_MAX, 1), + BPF_ALU64_IMM(BPF_MOV, R0, 0), + BPF_EXIT_INSN(), + }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + }, + { + "JMP32_JSGT_K: imm = 0x7fffffff -> never taken", + .u.insns_int = { + BPF_ALU64_IMM(BPF_MOV, R0, 1), + BPF_JMP32_IMM(BPF_JSGT, R1, S32_MAX, 1), + BPF_ALU64_IMM(BPF_MOV, R0, 0), + BPF_EXIT_INSN(), + }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 0 } }, + }, + { + "JMP32_JSGE_K: imm = -0x80000000 -> always taken", + .u.insns_int = { + BPF_ALU64_IMM(BPF_MOV, R0, 1), + BPF_JMP32_IMM(BPF_JSGE, R1, S32_MIN, 1), + BPF_ALU64_IMM(BPF_MOV, R0, 0), + BPF_EXIT_INSN(), + }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + }, + { + "JMP32_JSLT_K: imm = -0x80000000 -> never taken", + .u.insns_int = { + BPF_ALU64_IMM(BPF_MOV, R0, 1), + BPF_JMP32_IMM(BPF_JSLT, R1, S32_MIN, 1), + BPF_ALU64_IMM(BPF_MOV, R0, 0), + BPF_EXIT_INSN(), + }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 0 } }, + }, + { + "JMP32_JSLE_K: imm = 0x7fffffff -> always taken", + .u.insns_int = { + BPF_ALU64_IMM(BPF_MOV, R0, 1), + BPF_JMP32_IMM(BPF_JSLE, R1, S32_MAX, 1), + BPF_ALU64_IMM(BPF_MOV, R0, 0), + BPF_EXIT_INSN(), + }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + }, + { + "JMP_JEQ_X: dst = src -> always taken", + .u.insns_int = { + BPF_ALU64_IMM(BPF_MOV, R0, 1), + BPF_JMP_REG(BPF_JEQ, R1, R1, 1), + BPF_ALU64_IMM(BPF_MOV, R0, 0), + BPF_EXIT_INSN(), + }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + }, + { + "JMP_JGE_X: dst = src -> always taken", + .u.insns_int = { + BPF_ALU64_IMM(BPF_MOV, R0, 1), + BPF_JMP_REG(BPF_JGE, R1, R1, 1), + BPF_ALU64_IMM(BPF_MOV, R0, 0), + BPF_EXIT_INSN(), + }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + }, + { + "JMP_JLE_X: dst = src -> always taken", + .u.insns_int = { + BPF_ALU64_IMM(BPF_MOV, R0, 1), + BPF_JMP_REG(BPF_JLE, R1, R1, 1), + BPF_ALU64_IMM(BPF_MOV, R0, 0), + BPF_EXIT_INSN(), + }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + }, + { + "JMP_JSGE_X: dst = src -> always taken", + .u.insns_int = { + BPF_ALU64_IMM(BPF_MOV, R0, 1), + BPF_JMP_REG(BPF_JSGE, R1, R1, 1), + BPF_ALU64_IMM(BPF_MOV, R0, 0), + BPF_EXIT_INSN(), + }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + }, + { + "JMP_JSLE_X: dst = src -> always taken", + .u.insns_int = { + BPF_ALU64_IMM(BPF_MOV, R0, 1), + BPF_JMP_REG(BPF_JSLE, R1, R1, 1), + BPF_ALU64_IMM(BPF_MOV, R0, 0), + BPF_EXIT_INSN(), + }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + }, + { + "JMP_JNE_X: dst = src -> never taken", + .u.insns_int = { + BPF_ALU64_IMM(BPF_MOV, R0, 1), + BPF_JMP_REG(BPF_JNE, R1, R1, 1), + BPF_ALU64_IMM(BPF_MOV, R0, 0), + BPF_EXIT_INSN(), + }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 0 } }, + }, + { + "JMP_JGT_X: dst = src -> never taken", + .u.insns_int = { + BPF_ALU64_IMM(BPF_MOV, R0, 1), + BPF_JMP_REG(BPF_JGT, R1, R1, 1), + BPF_ALU64_IMM(BPF_MOV, R0, 0), + BPF_EXIT_INSN(), + }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 0 } }, + }, + { + "JMP_JLT_X: dst = src -> never taken", + .u.insns_int = { + BPF_ALU64_IMM(BPF_MOV, R0, 1), + BPF_JMP_REG(BPF_JLT, R1, R1, 1), + BPF_ALU64_IMM(BPF_MOV, R0, 0), + BPF_EXIT_INSN(), + }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 0 } }, + }, + { + "JMP_JSGT_X: dst = src -> never taken", + .u.insns_int = { + BPF_ALU64_IMM(BPF_MOV, R0, 1), + BPF_JMP_REG(BPF_JSGT, R1, R1, 1), + BPF_ALU64_IMM(BPF_MOV, R0, 0), + BPF_EXIT_INSN(), + }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 0 } }, + }, + { + "JMP_JSLT_X: dst = src -> never taken", + .u.insns_int = { + BPF_ALU64_IMM(BPF_MOV, R0, 1), + BPF_JMP_REG(BPF_JSLT, R1, R1, 1), + BPF_ALU64_IMM(BPF_MOV, R0, 0), + BPF_EXIT_INSN(), + }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 0 } }, + }, /* Short relative jumps */ { "Short relative jump: offset=0", -- cgit v1.2.3 From f1517eb790f97c1326016eb164a33a64d4d4fb7a Mon Sep 17 00:00:00 2001 From: Johan Almbladh Date: Tue, 14 Sep 2021 11:18:39 +0200 Subject: bpf/tests: Expand branch conversion JIT test This patch expands the branch conversion test introduced by 66e5eb84 ("bpf, tests: Add branch conversion JIT test"). The test now includes a JMP with maximum eBPF offset. This triggers branch conversion for the 64-bit MIPS JIT. Additional variants are also added for cases when the branch is taken or not taken. Signed-off-by: Johan Almbladh Signed-off-by: Daniel Borkmann Link: https://lore.kernel.org/bpf/20210914091842.4186267-12-johan.almbladh@anyfinetworks.com --- lib/test_bpf.c | 125 +++++++++++++++++++++++++++++++++++++++++---------------- 1 file changed, 91 insertions(+), 34 deletions(-) (limited to 'lib') diff --git a/lib/test_bpf.c b/lib/test_bpf.c index 2aa1d0d3c685..dd6bb0044097 100644 --- a/lib/test_bpf.c +++ b/lib/test_bpf.c @@ -463,47 +463,79 @@ static int bpf_fill_stxdw(struct bpf_test *self) return __bpf_fill_stxdw(self, BPF_DW); } -static int bpf_fill_long_jmp(struct bpf_test *self) +static int __bpf_ld_imm64(struct bpf_insn insns[2], u8 reg, s64 imm64) { - unsigned int len = BPF_MAXINSNS; - struct bpf_insn *insn; + struct bpf_insn tmp[] = {BPF_LD_IMM64(reg, imm64)}; + + memcpy(insns, tmp, sizeof(tmp)); + return 2; +} + +/* + * Branch conversion tests. Complex operations can expand to a lot + * of instructions when JITed. This in turn may cause jump offsets + * to overflow the field size of the native instruction, triggering + * a branch conversion mechanism in some JITs. + */ +static int __bpf_fill_max_jmp(struct bpf_test *self, int jmp, int imm) +{ + struct bpf_insn *insns; + int len = S16_MAX + 5; int i; - insn = kmalloc_array(len, sizeof(*insn), GFP_KERNEL); - if (!insn) + insns = kmalloc_array(len, sizeof(*insns), GFP_KERNEL); + if (!insns) return -ENOMEM; - insn[0] = BPF_ALU64_IMM(BPF_MOV, R0, 1); - insn[1] = BPF_JMP_IMM(BPF_JEQ, R0, 1, len - 2 - 1); + i = __bpf_ld_imm64(insns, R1, 0x0123456789abcdefULL); + insns[i++] = BPF_ALU64_IMM(BPF_MOV, R0, 1); + insns[i++] = BPF_JMP_IMM(jmp, R0, imm, S16_MAX); + insns[i++] = BPF_ALU64_IMM(BPF_MOV, R0, 2); + insns[i++] = BPF_EXIT_INSN(); - /* - * Fill with a complex 64-bit operation that expands to a lot of - * instructions on 32-bit JITs. The large jump offset can then - * overflow the conditional branch field size, triggering a branch - * conversion mechanism in some JITs. - * - * Note: BPF_MAXINSNS of ALU64 MUL is enough to trigger such branch - * conversion on the 32-bit MIPS JIT. For other JITs, the instruction - * count and/or operation may need to be modified to trigger the - * branch conversion. - */ - for (i = 2; i < len - 1; i++) - insn[i] = BPF_ALU64_IMM(BPF_MUL, R0, (i << 16) + i); + while (i < len - 1) { + static const int ops[] = { + BPF_LSH, BPF_RSH, BPF_ARSH, BPF_ADD, + BPF_SUB, BPF_MUL, BPF_DIV, BPF_MOD, + }; + int op = ops[(i >> 1) % ARRAY_SIZE(ops)]; - insn[len - 1] = BPF_EXIT_INSN(); + if (i & 1) + insns[i++] = BPF_ALU32_REG(op, R0, R1); + else + insns[i++] = BPF_ALU64_REG(op, R0, R1); + } - self->u.ptr.insns = insn; + insns[i++] = BPF_EXIT_INSN(); + self->u.ptr.insns = insns; self->u.ptr.len = len; + BUG_ON(i != len); return 0; } -static int __bpf_ld_imm64(struct bpf_insn insns[2], u8 reg, s64 imm64) +/* Branch taken by runtime decision */ +static int bpf_fill_max_jmp_taken(struct bpf_test *self) { - struct bpf_insn tmp[] = {BPF_LD_IMM64(reg, imm64)}; + return __bpf_fill_max_jmp(self, BPF_JEQ, 1); +} - memcpy(insns, tmp, sizeof(tmp)); - return 2; +/* Branch not taken by runtime decision */ +static int bpf_fill_max_jmp_not_taken(struct bpf_test *self) +{ + return __bpf_fill_max_jmp(self, BPF_JEQ, 0); +} + +/* Branch always taken, known at JIT time */ +static int bpf_fill_max_jmp_always_taken(struct bpf_test *self) +{ + return __bpf_fill_max_jmp(self, BPF_JGE, 0); +} + +/* Branch never taken, known at JIT time */ +static int bpf_fill_max_jmp_never_taken(struct bpf_test *self) +{ + return __bpf_fill_max_jmp(self, BPF_JLT, 0); } /* Test an ALU shift operation for all valid shift values */ @@ -8653,14 +8685,6 @@ static struct bpf_test tests[] = { { }, { { 0, 1 } }, }, - { /* Mainly checking JIT here. */ - "BPF_MAXINSNS: Very long conditional jump", - { }, - INTERNAL | FLAG_NO_DATA, - { }, - { { 0, 1 } }, - .fill_helper = bpf_fill_long_jmp, - }, { "JMP_JA: Jump, gap, jump, ...", { }, @@ -11009,6 +11033,39 @@ static struct bpf_test tests[] = { { }, { { 0, 0 } }, }, + /* Conditional branch conversions */ + { + "Long conditional jump: taken at runtime", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_max_jmp_taken, + }, + { + "Long conditional jump: not taken at runtime", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 2 } }, + .fill_helper = bpf_fill_max_jmp_not_taken, + }, + { + "Long conditional jump: always taken, known at JIT time", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_max_jmp_always_taken, + }, + { + "Long conditional jump: never taken, known at JIT time", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 2 } }, + .fill_helper = bpf_fill_max_jmp_never_taken, + }, /* Staggered jump sequences, immediate */ { "Staggered jumps: JMP_JA", -- cgit v1.2.3 From f536a7c80675e4875e50df9182881d7678e27651 Mon Sep 17 00:00:00 2001 From: Johan Almbladh Date: Tue, 14 Sep 2021 11:18:40 +0200 Subject: bpf/tests: Add more BPF_END byte order conversion tests This patch adds tests of the high 32 bits of 64-bit BPF_END conversions. It also adds a mirrored set of tests where the source bytes are reversed. The MSB of each byte is now set on the high word instead, possibly affecting sign-extension during conversion in a different way. Mainly for JIT testing. Signed-off-by: Johan Almbladh Signed-off-by: Daniel Borkmann Link: https://lore.kernel.org/bpf/20210914091842.4186267-13-johan.almbladh@anyfinetworks.com --- lib/test_bpf.c | 122 +++++++++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 122 insertions(+) (limited to 'lib') diff --git a/lib/test_bpf.c b/lib/test_bpf.c index dd6bb0044097..7f8ad4fa4add 100644 --- a/lib/test_bpf.c +++ b/lib/test_bpf.c @@ -6748,6 +6748,67 @@ static struct bpf_test tests[] = { { }, { { 0, (u32) cpu_to_be64(0x0123456789abcdefLL) } }, }, + { + "ALU_END_FROM_BE 64: 0x0123456789abcdef >> 32 -> 0x01234567", + .u.insns_int = { + BPF_LD_IMM64(R0, 0x0123456789abcdefLL), + BPF_ENDIAN(BPF_FROM_BE, R0, 64), + BPF_ALU64_IMM(BPF_RSH, R0, 32), + BPF_EXIT_INSN(), + }, + INTERNAL, + { }, + { { 0, (u32) (cpu_to_be64(0x0123456789abcdefLL) >> 32) } }, + }, + /* BPF_ALU | BPF_END | BPF_FROM_BE, reversed */ + { + "ALU_END_FROM_BE 16: 0xfedcba9876543210 -> 0x3210", + .u.insns_int = { + BPF_LD_IMM64(R0, 0xfedcba9876543210ULL), + BPF_ENDIAN(BPF_FROM_BE, R0, 16), + BPF_EXIT_INSN(), + }, + INTERNAL, + { }, + { { 0, cpu_to_be16(0x3210) } }, + }, + { + "ALU_END_FROM_BE 32: 0xfedcba9876543210 -> 0x76543210", + .u.insns_int = { + BPF_LD_IMM64(R0, 0xfedcba9876543210ULL), + BPF_ENDIAN(BPF_FROM_BE, R0, 32), + BPF_ALU64_REG(BPF_MOV, R1, R0), + BPF_ALU64_IMM(BPF_RSH, R1, 32), + BPF_ALU32_REG(BPF_ADD, R0, R1), /* R1 = 0 */ + BPF_EXIT_INSN(), + }, + INTERNAL, + { }, + { { 0, cpu_to_be32(0x76543210) } }, + }, + { + "ALU_END_FROM_BE 64: 0xfedcba9876543210 -> 0x76543210", + .u.insns_int = { + BPF_LD_IMM64(R0, 0xfedcba9876543210ULL), + BPF_ENDIAN(BPF_FROM_BE, R0, 64), + BPF_EXIT_INSN(), + }, + INTERNAL, + { }, + { { 0, (u32) cpu_to_be64(0xfedcba9876543210ULL) } }, + }, + { + "ALU_END_FROM_BE 64: 0xfedcba9876543210 >> 32 -> 0xfedcba98", + .u.insns_int = { + BPF_LD_IMM64(R0, 0xfedcba9876543210ULL), + BPF_ENDIAN(BPF_FROM_BE, R0, 64), + BPF_ALU64_IMM(BPF_RSH, R0, 32), + BPF_EXIT_INSN(), + }, + INTERNAL, + { }, + { { 0, (u32) (cpu_to_be64(0xfedcba9876543210ULL) >> 32) } }, + }, /* BPF_ALU | BPF_END | BPF_FROM_LE */ { "ALU_END_FROM_LE 16: 0x0123456789abcdef -> 0xefcd", @@ -6785,6 +6846,67 @@ static struct bpf_test tests[] = { { }, { { 0, (u32) cpu_to_le64(0x0123456789abcdefLL) } }, }, + { + "ALU_END_FROM_LE 64: 0x0123456789abcdef >> 32 -> 0xefcdab89", + .u.insns_int = { + BPF_LD_IMM64(R0, 0x0123456789abcdefLL), + BPF_ENDIAN(BPF_FROM_LE, R0, 64), + BPF_ALU64_IMM(BPF_RSH, R0, 32), + BPF_EXIT_INSN(), + }, + INTERNAL, + { }, + { { 0, (u32) (cpu_to_le64(0x0123456789abcdefLL) >> 32) } }, + }, + /* BPF_ALU | BPF_END | BPF_FROM_LE, reversed */ + { + "ALU_END_FROM_LE 16: 0xfedcba9876543210 -> 0x1032", + .u.insns_int = { + BPF_LD_IMM64(R0, 0xfedcba9876543210ULL), + BPF_ENDIAN(BPF_FROM_LE, R0, 16), + BPF_EXIT_INSN(), + }, + INTERNAL, + { }, + { { 0, cpu_to_le16(0x3210) } }, + }, + { + "ALU_END_FROM_LE 32: 0xfedcba9876543210 -> 0x10325476", + .u.insns_int = { + BPF_LD_IMM64(R0, 0xfedcba9876543210ULL), + BPF_ENDIAN(BPF_FROM_LE, R0, 32), + BPF_ALU64_REG(BPF_MOV, R1, R0), + BPF_ALU64_IMM(BPF_RSH, R1, 32), + BPF_ALU32_REG(BPF_ADD, R0, R1), /* R1 = 0 */ + BPF_EXIT_INSN(), + }, + INTERNAL, + { }, + { { 0, cpu_to_le32(0x76543210) } }, + }, + { + "ALU_END_FROM_LE 64: 0xfedcba9876543210 -> 0x10325476", + .u.insns_int = { + BPF_LD_IMM64(R0, 0xfedcba9876543210ULL), + BPF_ENDIAN(BPF_FROM_LE, R0, 64), + BPF_EXIT_INSN(), + }, + INTERNAL, + { }, + { { 0, (u32) cpu_to_le64(0xfedcba9876543210ULL) } }, + }, + { + "ALU_END_FROM_LE 64: 0xfedcba9876543210 >> 32 -> 0x98badcfe", + .u.insns_int = { + BPF_LD_IMM64(R0, 0xfedcba9876543210ULL), + BPF_ENDIAN(BPF_FROM_LE, R0, 64), + BPF_ALU64_IMM(BPF_RSH, R0, 32), + BPF_EXIT_INSN(), + }, + INTERNAL, + { }, + { { 0, (u32) (cpu_to_le64(0xfedcba9876543210ULL) >> 32) } }, + }, /* BPF_ST(X) | BPF_MEM | BPF_B/H/W/DW */ { "ST_MEM_B: Store/Load byte: max negative", -- cgit v1.2.3 From 18935a72eb25525b655262579e1652362a3b29bb Mon Sep 17 00:00:00 2001 From: Johan Almbladh Date: Tue, 14 Sep 2021 11:18:41 +0200 Subject: bpf/tests: Fix error in tail call limit tests This patch fixes an error in the tail call limit test that caused the test to fail on for x86-64 JIT. Previously, the register R0 was used to report the total number of tail calls made. However, after a tail call fall-through, the value of the R0 register is undefined. Now, all tail call error path tests instead use context state to store the count. Fixes: 874be05f525e ("bpf, tests: Add tail call test suite") Reported-by: Paul Chaignon Reported-by: Tiezhu Yang Signed-off-by: Johan Almbladh Signed-off-by: Daniel Borkmann Tested-by: Tiezhu Yang Link: https://lore.kernel.org/bpf/20210914091842.4186267-14-johan.almbladh@anyfinetworks.com --- lib/test_bpf.c | 37 +++++++++++++++++++++++++++---------- 1 file changed, 27 insertions(+), 10 deletions(-) (limited to 'lib') diff --git a/lib/test_bpf.c b/lib/test_bpf.c index 7f8ad4fa4add..a94ab634f947 100644 --- a/lib/test_bpf.c +++ b/lib/test_bpf.c @@ -12180,10 +12180,15 @@ static __init int test_bpf(void) struct tail_call_test { const char *descr; struct bpf_insn insns[MAX_INSNS]; + int flags; int result; int stack_depth; }; +/* Flags that can be passed to tail call test cases */ +#define FLAG_NEED_STATE BIT(0) +#define FLAG_RESULT_IN_STATE BIT(1) + /* * Magic marker used in test snippets for tail calls below. * BPF_LD/MOV to R2 and R2 with this immediate value is replaced @@ -12253,32 +12258,38 @@ static struct tail_call_test tail_call_tests[] = { { "Tail call error path, max count reached", .insns = { - BPF_ALU64_IMM(BPF_ADD, R1, 1), - BPF_ALU64_REG(BPF_MOV, R0, R1), + BPF_LDX_MEM(BPF_W, R2, R1, 0), + BPF_ALU64_IMM(BPF_ADD, R2, 1), + BPF_STX_MEM(BPF_W, R1, R2, 0), TAIL_CALL(0), BPF_EXIT_INSN(), }, - .result = MAX_TAIL_CALL_CNT + 1, + .flags = FLAG_NEED_STATE | FLAG_RESULT_IN_STATE, + .result = (MAX_TAIL_CALL_CNT + 1 + 1) * MAX_TESTRUNS, }, { "Tail call error path, NULL target", .insns = { - BPF_ALU64_IMM(BPF_MOV, R0, -1), + BPF_LDX_MEM(BPF_W, R2, R1, 0), + BPF_ALU64_IMM(BPF_ADD, R2, 1), + BPF_STX_MEM(BPF_W, R1, R2, 0), TAIL_CALL(TAIL_CALL_NULL), - BPF_ALU64_IMM(BPF_MOV, R0, 1), BPF_EXIT_INSN(), }, - .result = 1, + .flags = FLAG_NEED_STATE | FLAG_RESULT_IN_STATE, + .result = MAX_TESTRUNS, }, { "Tail call error path, index out of range", .insns = { - BPF_ALU64_IMM(BPF_MOV, R0, -1), + BPF_LDX_MEM(BPF_W, R2, R1, 0), + BPF_ALU64_IMM(BPF_ADD, R2, 1), + BPF_STX_MEM(BPF_W, R1, R2, 0), TAIL_CALL(TAIL_CALL_INVALID), - BPF_ALU64_IMM(BPF_MOV, R0, 1), BPF_EXIT_INSN(), }, - .result = 1, + .flags = FLAG_NEED_STATE | FLAG_RESULT_IN_STATE, + .result = MAX_TESTRUNS, }, }; @@ -12384,6 +12395,8 @@ static __init int test_tail_calls(struct bpf_array *progs) for (i = 0; i < ARRAY_SIZE(tail_call_tests); i++) { struct tail_call_test *test = &tail_call_tests[i]; struct bpf_prog *fp = progs->ptrs[i]; + int *data = NULL; + int state = 0; u64 duration; int ret; @@ -12400,7 +12413,11 @@ static __init int test_tail_calls(struct bpf_array *progs) if (fp->jited) jit_cnt++; - ret = __run_one(fp, NULL, MAX_TESTRUNS, &duration); + if (test->flags & FLAG_NEED_STATE) + data = &state; + ret = __run_one(fp, data, MAX_TESTRUNS, &duration); + if (test->flags & FLAG_RESULT_IN_STATE) + ret = state; if (ret == test->result) { pr_cont("%lld PASS", duration); pass_cnt++; -- cgit v1.2.3 From 29eef85be2f60b1027214b4bfc4b1a9d592830a7 Mon Sep 17 00:00:00 2001 From: Johan Almbladh Date: Tue, 14 Sep 2021 11:18:42 +0200 Subject: bpf/tests: Add tail call limit test with external function call This patch adds a tail call limit test where the program also emits a BPF_CALL to an external function prior to the tail call. Mainly testing that JITed programs preserve its internal register state, for example tail call count, across such external calls. Signed-off-by: Johan Almbladh Signed-off-by: Daniel Borkmann Link: https://lore.kernel.org/bpf/20210914091842.4186267-15-johan.almbladh@anyfinetworks.com --- lib/test_bpf.c | 86 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++-- 1 file changed, 83 insertions(+), 3 deletions(-) (limited to 'lib') diff --git a/lib/test_bpf.c b/lib/test_bpf.c index a94ab634f947..08f438e6fe9e 100644 --- a/lib/test_bpf.c +++ b/lib/test_bpf.c @@ -12208,6 +12208,30 @@ struct tail_call_test { offset, TAIL_CALL_MARKER), \ BPF_JMP_IMM(BPF_TAIL_CALL, 0, 0, 0) +/* + * A test function to be called from a BPF program, clobbering a lot of + * CPU registers in the process. A JITed BPF program calling this function + * must save and restore any caller-saved registers it uses for internal + * state, for example the current tail call count. + */ +BPF_CALL_1(bpf_test_func, u64, arg) +{ + char buf[64]; + long a = 0; + long b = 1; + long c = 2; + long d = 3; + long e = 4; + long f = 5; + long g = 6; + long h = 7; + + return snprintf(buf, sizeof(buf), + "%ld %lu %lx %ld %lu %lx %ld %lu %x", + a, b, c, d, e, f, g, h, (int)arg); +} +#define BPF_FUNC_test_func __BPF_FUNC_MAX_ID + /* * Tail call tests. Each test case may call any other test in the table, * including itself, specified as a relative index offset from the calling @@ -12267,6 +12291,28 @@ static struct tail_call_test tail_call_tests[] = { .flags = FLAG_NEED_STATE | FLAG_RESULT_IN_STATE, .result = (MAX_TAIL_CALL_CNT + 1 + 1) * MAX_TESTRUNS, }, + { + "Tail call count preserved across function calls", + .insns = { + BPF_LDX_MEM(BPF_W, R2, R1, 0), + BPF_ALU64_IMM(BPF_ADD, R2, 1), + BPF_STX_MEM(BPF_W, R1, R2, 0), + BPF_STX_MEM(BPF_DW, R10, R1, -8), + BPF_CALL_REL(BPF_FUNC_get_numa_node_id), + BPF_CALL_REL(BPF_FUNC_ktime_get_ns), + BPF_CALL_REL(BPF_FUNC_ktime_get_boot_ns), + BPF_CALL_REL(BPF_FUNC_ktime_get_coarse_ns), + BPF_CALL_REL(BPF_FUNC_jiffies64), + BPF_CALL_REL(BPF_FUNC_test_func), + BPF_LDX_MEM(BPF_DW, R1, R10, -8), + BPF_ALU32_REG(BPF_MOV, R0, R1), + TAIL_CALL(0), + BPF_EXIT_INSN(), + }, + .stack_depth = 8, + .flags = FLAG_NEED_STATE | FLAG_RESULT_IN_STATE, + .result = (MAX_TAIL_CALL_CNT + 1 + 1) * MAX_TESTRUNS, + }, { "Tail call error path, NULL target", .insns = { @@ -12345,17 +12391,19 @@ static __init int prepare_tail_call_tests(struct bpf_array **pprogs) /* Relocate runtime tail call offsets and addresses */ for (i = 0; i < len; i++) { struct bpf_insn *insn = &fp->insnsi[i]; - - if (insn->imm != TAIL_CALL_MARKER) - continue; + long addr = 0; switch (insn->code) { case BPF_LD | BPF_DW | BPF_IMM: + if (insn->imm != TAIL_CALL_MARKER) + break; insn[0].imm = (u32)(long)progs; insn[1].imm = ((u64)(long)progs) >> 32; break; case BPF_ALU | BPF_MOV | BPF_K: + if (insn->imm != TAIL_CALL_MARKER) + break; if (insn->off == TAIL_CALL_NULL) insn->imm = ntests; else if (insn->off == TAIL_CALL_INVALID) @@ -12363,6 +12411,38 @@ static __init int prepare_tail_call_tests(struct bpf_array **pprogs) else insn->imm = which + insn->off; insn->off = 0; + break; + + case BPF_JMP | BPF_CALL: + if (insn->src_reg != BPF_PSEUDO_CALL) + break; + switch (insn->imm) { + case BPF_FUNC_get_numa_node_id: + addr = (long)&numa_node_id; + break; + case BPF_FUNC_ktime_get_ns: + addr = (long)&ktime_get_ns; + break; + case BPF_FUNC_ktime_get_boot_ns: + addr = (long)&ktime_get_boot_fast_ns; + break; + case BPF_FUNC_ktime_get_coarse_ns: + addr = (long)&ktime_get_coarse_ns; + break; + case BPF_FUNC_jiffies64: + addr = (long)&get_jiffies_64; + break; + case BPF_FUNC_test_func: + addr = (long)&bpf_test_func; + break; + default: + err = -EFAULT; + goto out_err; + } + *insn = BPF_EMIT_CALL(BPF_CAST_CALL(addr)); + if ((long)__bpf_call_base + insn->imm != addr) + *insn = BPF_JMP_A(0); /* Skip: NOP */ + break; } } -- cgit v1.2.3 From 3d717fad5081b8e3bda76d86907fad95398cbde8 Mon Sep 17 00:00:00 2001 From: Kees Cook Date: Tue, 28 Sep 2021 16:09:45 -0700 Subject: bpf: Replace "want address" users of BPF_CAST_CALL with BPF_CALL_IMM In order to keep ahead of cases in the kernel where Control Flow Integrity (CFI) may trip over function call casts, enabling -Wcast-function-type is helpful. To that end, BPF_CAST_CALL causes various warnings and is one of the last places in the kernel triggering this warning. Most places using BPF_CAST_CALL actually just want a void * to perform math on. It's not actually performing a call, so just use a different helper to get the void *, by way of the new BPF_CALL_IMM() helper, which can clean up a common copy/paste idiom as well. This change results in no object code difference. Signed-off-by: Kees Cook Signed-off-by: Alexei Starovoitov Reviewed-by: Gustavo A. R. Silva Acked-by: Andrii Nakryiko Link: https://github.com/KSPP/linux/issues/20 Link: https://lore.kernel.org/lkml/CAEf4Bzb46=-J5Fxc3mMZ8JQPtK1uoE0q6+g6WPz53Cvx=CBEhw@mail.gmail.com Link: https://lore.kernel.org/bpf/20210928230946.4062144-2-keescook@chromium.org --- include/linux/filter.h | 6 +++++- kernel/bpf/hashtab.c | 6 +++--- kernel/bpf/verifier.c | 26 +++++++++----------------- lib/test_bpf.c | 2 +- 4 files changed, 18 insertions(+), 22 deletions(-) (limited to 'lib') diff --git a/include/linux/filter.h b/include/linux/filter.h index 4a93c12543ee..6c247663d4ce 100644 --- a/include/linux/filter.h +++ b/include/linux/filter.h @@ -365,13 +365,17 @@ static inline bool insn_is_zext(const struct bpf_insn *insn) #define BPF_CAST_CALL(x) \ ((u64 (*)(u64, u64, u64, u64, u64))(x)) +/* Convert function address to BPF immediate */ + +#define BPF_CALL_IMM(x) ((void *)(x) - (void *)__bpf_call_base) + #define BPF_EMIT_CALL(FUNC) \ ((struct bpf_insn) { \ .code = BPF_JMP | BPF_CALL, \ .dst_reg = 0, \ .src_reg = 0, \ .off = 0, \ - .imm = ((FUNC) - __bpf_call_base) }) + .imm = BPF_CALL_IMM(FUNC) }) /* Raw code statement block */ diff --git a/kernel/bpf/hashtab.c b/kernel/bpf/hashtab.c index 32471ba02708..3d8f9d6997d5 100644 --- a/kernel/bpf/hashtab.c +++ b/kernel/bpf/hashtab.c @@ -668,7 +668,7 @@ static int htab_map_gen_lookup(struct bpf_map *map, struct bpf_insn *insn_buf) BUILD_BUG_ON(!__same_type(&__htab_map_lookup_elem, (void *(*)(struct bpf_map *map, void *key))NULL)); - *insn++ = BPF_EMIT_CALL(BPF_CAST_CALL(__htab_map_lookup_elem)); + *insn++ = BPF_EMIT_CALL(__htab_map_lookup_elem); *insn++ = BPF_JMP_IMM(BPF_JEQ, ret, 0, 1); *insn++ = BPF_ALU64_IMM(BPF_ADD, ret, offsetof(struct htab_elem, key) + @@ -709,7 +709,7 @@ static int htab_lru_map_gen_lookup(struct bpf_map *map, BUILD_BUG_ON(!__same_type(&__htab_map_lookup_elem, (void *(*)(struct bpf_map *map, void *key))NULL)); - *insn++ = BPF_EMIT_CALL(BPF_CAST_CALL(__htab_map_lookup_elem)); + *insn++ = BPF_EMIT_CALL(__htab_map_lookup_elem); *insn++ = BPF_JMP_IMM(BPF_JEQ, ret, 0, 4); *insn++ = BPF_LDX_MEM(BPF_B, ref_reg, ret, offsetof(struct htab_elem, lru_node) + @@ -2397,7 +2397,7 @@ static int htab_of_map_gen_lookup(struct bpf_map *map, BUILD_BUG_ON(!__same_type(&__htab_map_lookup_elem, (void *(*)(struct bpf_map *map, void *key))NULL)); - *insn++ = BPF_EMIT_CALL(BPF_CAST_CALL(__htab_map_lookup_elem)); + *insn++ = BPF_EMIT_CALL(__htab_map_lookup_elem); *insn++ = BPF_JMP_IMM(BPF_JEQ, ret, 0, 2); *insn++ = BPF_ALU64_IMM(BPF_ADD, ret, offsetof(struct htab_elem, key) + diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index 7a8351604f67..1433752db740 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -1744,7 +1744,7 @@ static int add_kfunc_call(struct bpf_verifier_env *env, u32 func_id) desc = &tab->descs[tab->nr_descs++]; desc->func_id = func_id; - desc->imm = BPF_CAST_CALL(addr) - __bpf_call_base; + desc->imm = BPF_CALL_IMM(addr); err = btf_distill_func_proto(&env->log, btf_vmlinux, func_proto, func_name, &desc->func_model); @@ -12514,8 +12514,7 @@ static int jit_subprogs(struct bpf_verifier_env *env) if (!bpf_pseudo_call(insn)) continue; subprog = insn->off; - insn->imm = BPF_CAST_CALL(func[subprog]->bpf_func) - - __bpf_call_base; + insn->imm = BPF_CALL_IMM(func[subprog]->bpf_func); } /* we use the aux data to keep a list of the start addresses @@ -12995,32 +12994,25 @@ static int do_misc_fixups(struct bpf_verifier_env *env) patch_map_ops_generic: switch (insn->imm) { case BPF_FUNC_map_lookup_elem: - insn->imm = BPF_CAST_CALL(ops->map_lookup_elem) - - __bpf_call_base; + insn->imm = BPF_CALL_IMM(ops->map_lookup_elem); continue; case BPF_FUNC_map_update_elem: - insn->imm = BPF_CAST_CALL(ops->map_update_elem) - - __bpf_call_base; + insn->imm = BPF_CALL_IMM(ops->map_update_elem); continue; case BPF_FUNC_map_delete_elem: - insn->imm = BPF_CAST_CALL(ops->map_delete_elem) - - __bpf_call_base; + insn->imm = BPF_CALL_IMM(ops->map_delete_elem); continue; case BPF_FUNC_map_push_elem: - insn->imm = BPF_CAST_CALL(ops->map_push_elem) - - __bpf_call_base; + insn->imm = BPF_CALL_IMM(ops->map_push_elem); continue; case BPF_FUNC_map_pop_elem: - insn->imm = BPF_CAST_CALL(ops->map_pop_elem) - - __bpf_call_base; + insn->imm = BPF_CALL_IMM(ops->map_pop_elem); continue; case BPF_FUNC_map_peek_elem: - insn->imm = BPF_CAST_CALL(ops->map_peek_elem) - - __bpf_call_base; + insn->imm = BPF_CALL_IMM(ops->map_peek_elem); continue; case BPF_FUNC_redirect_map: - insn->imm = BPF_CAST_CALL(ops->map_redirect) - - __bpf_call_base; + insn->imm = BPF_CALL_IMM(ops->map_redirect); continue; } diff --git a/lib/test_bpf.c b/lib/test_bpf.c index 08f438e6fe9e..21ea1ab253a1 100644 --- a/lib/test_bpf.c +++ b/lib/test_bpf.c @@ -12439,7 +12439,7 @@ static __init int prepare_tail_call_tests(struct bpf_array **pprogs) err = -EFAULT; goto out_err; } - *insn = BPF_EMIT_CALL(BPF_CAST_CALL(addr)); + *insn = BPF_EMIT_CALL(addr); if ((long)__bpf_call_base + insn->imm != addr) *insn = BPF_JMP_A(0); /* Skip: NOP */ break; -- cgit v1.2.3 From f2ec8d9a3b8c0f22cd6a2b4f5a2d9aee5206e3b7 Mon Sep 17 00:00:00 2001 From: Masami Hiramatsu Date: Tue, 14 Sep 2021 23:40:36 +0900 Subject: kprobes: treewide: Replace arch_deref_entry_point() with dereference_symbol_descriptor() ~15 years ago kprobes grew the 'arch_deref_entry_point()' __weak function: 3d7e33825d87: ("jprobes: make jprobes a little safer for users") But this is just open-coded dereference_symbol_descriptor() in essence, and its obscure nature was causing bugs. Just use the real thing and remove arch_deref_entry_point(). Link: https://lkml.kernel.org/r/163163043630.489837.7924988885652708696.stgit@devnote2 Signed-off-by: Masami Hiramatsu Tested-by: Andrii Nakryiko Signed-off-by: Steven Rostedt (VMware) --- arch/ia64/kernel/kprobes.c | 5 ----- arch/powerpc/kernel/kprobes.c | 11 ----------- include/linux/kprobes.h | 1 - kernel/kprobes.c | 7 +------ lib/error-inject.c | 3 ++- 5 files changed, 3 insertions(+), 24 deletions(-) (limited to 'lib') diff --git a/arch/ia64/kernel/kprobes.c b/arch/ia64/kernel/kprobes.c index d4048518a1d7..0f8573bbf520 100644 --- a/arch/ia64/kernel/kprobes.c +++ b/arch/ia64/kernel/kprobes.c @@ -891,11 +891,6 @@ int __kprobes kprobe_exceptions_notify(struct notifier_block *self, return ret; } -unsigned long arch_deref_entry_point(void *entry) -{ - return ((struct fnptr *)entry)->ip; -} - static struct kprobe trampoline_p = { .pre_handler = trampoline_probe_handler }; diff --git a/arch/powerpc/kernel/kprobes.c b/arch/powerpc/kernel/kprobes.c index 7a7cd6bda53e..d422e297978b 100644 --- a/arch/powerpc/kernel/kprobes.c +++ b/arch/powerpc/kernel/kprobes.c @@ -542,17 +542,6 @@ int kprobe_fault_handler(struct pt_regs *regs, int trapnr) } NOKPROBE_SYMBOL(kprobe_fault_handler); -unsigned long arch_deref_entry_point(void *entry) -{ -#ifdef PPC64_ELF_ABI_v1 - if (!kernel_text_address((unsigned long)entry)) - return ppc_global_function_entry(entry); - else -#endif - return (unsigned long)entry; -} -NOKPROBE_SYMBOL(arch_deref_entry_point); - static struct kprobe trampoline_p = { .addr = (kprobe_opcode_t *) &kretprobe_trampoline, .pre_handler = trampoline_probe_handler diff --git a/include/linux/kprobes.h b/include/linux/kprobes.h index 0ba3f9e316d4..2ed61fcbc89c 100644 --- a/include/linux/kprobes.h +++ b/include/linux/kprobes.h @@ -381,7 +381,6 @@ int register_kprobe(struct kprobe *p); void unregister_kprobe(struct kprobe *p); int register_kprobes(struct kprobe **kps, int num); void unregister_kprobes(struct kprobe **kps, int num); -unsigned long arch_deref_entry_point(void *); int register_kretprobe(struct kretprobe *rp); void unregister_kretprobe(struct kretprobe *rp); diff --git a/kernel/kprobes.c b/kernel/kprobes.c index 8021bccb7770..550042d9a6ef 100644 --- a/kernel/kprobes.c +++ b/kernel/kprobes.c @@ -1861,11 +1861,6 @@ static struct notifier_block kprobe_exceptions_nb = { .priority = 0x7fffffff /* we need to be notified first */ }; -unsigned long __weak arch_deref_entry_point(void *entry) -{ - return (unsigned long)entry; -} - #ifdef CONFIG_KRETPROBES unsigned long __kretprobe_trampoline_handler(struct pt_regs *regs, @@ -2327,7 +2322,7 @@ static int __init populate_kprobe_blacklist(unsigned long *start, int ret; for (iter = start; iter < end; iter++) { - entry = arch_deref_entry_point((void *)*iter); + entry = (unsigned long)dereference_symbol_descriptor((void *)*iter); ret = kprobe_add_ksym_blacklist(entry); if (ret == -EINVAL) continue; diff --git a/lib/error-inject.c b/lib/error-inject.c index c73651b15b76..2ff5ef689d72 100644 --- a/lib/error-inject.c +++ b/lib/error-inject.c @@ -8,6 +8,7 @@ #include #include #include +#include /* Whitelist of symbols that can be overridden for error injection. */ static LIST_HEAD(error_injection_list); @@ -64,7 +65,7 @@ static void populate_error_injection_list(struct error_injection_entry *start, mutex_lock(&ei_mutex); for (iter = start; iter < end; iter++) { - entry = arch_deref_entry_point((void *)iter->addr); + entry = (unsigned long)dereference_symbol_descriptor((void *)iter->addr); if (!kernel_text_address(entry) || !kallsyms_lookup_size_offset(entry, &size, &offset)) { -- cgit v1.2.3 From caaaa1667bf198c54cc3141ad92ca6ce853e99cd Mon Sep 17 00:00:00 2001 From: Johan Almbladh Date: Fri, 1 Oct 2021 15:03:39 +0200 Subject: bpf/tests: Add tests of BPF_LDX and BPF_STX with small sizes This patch adds a series of tests to verify the behavior of BPF_LDX and BPF_STX with BPF_B//W sizes in isolation. In particular, it checks that BPF_LDX zero-extendeds the result, and that BPF_STX does not overwrite adjacent bytes in memory. BPF_ST and operations on BPF_DW size are deemed to be sufficiently tested by existing tests. Signed-off-by: Johan Almbladh Signed-off-by: Daniel Borkmann Link: https://lore.kernel.org/bpf/20211001130348.3670534-2-johan.almbladh@anyfinetworks.com --- lib/test_bpf.c | 254 +++++++++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 254 insertions(+) (limited to 'lib') diff --git a/lib/test_bpf.c b/lib/test_bpf.c index 21ea1ab253a1..a838a6179ca4 100644 --- a/lib/test_bpf.c +++ b/lib/test_bpf.c @@ -6907,6 +6907,260 @@ static struct bpf_test tests[] = { { }, { { 0, (u32) (cpu_to_le64(0xfedcba9876543210ULL) >> 32) } }, }, + /* BPF_LDX_MEM B/H/W/DW */ + { + "BPF_LDX_MEM | BPF_B", + .u.insns_int = { + BPF_LD_IMM64(R1, 0x0102030405060708ULL), + BPF_LD_IMM64(R2, 0x0000000000000008ULL), + BPF_STX_MEM(BPF_DW, R10, R1, -8), +#ifdef __BIG_ENDIAN + BPF_LDX_MEM(BPF_B, R0, R10, -1), +#else + BPF_LDX_MEM(BPF_B, R0, R10, -8), +#endif + BPF_JMP_REG(BPF_JNE, R0, R2, 1), + BPF_ALU64_IMM(BPF_MOV, R0, 0), + BPF_EXIT_INSN(), + }, + INTERNAL, + { }, + { { 0, 0 } }, + .stack_depth = 8, + }, + { + "BPF_LDX_MEM | BPF_B, MSB set", + .u.insns_int = { + BPF_LD_IMM64(R1, 0x8182838485868788ULL), + BPF_LD_IMM64(R2, 0x0000000000000088ULL), + BPF_STX_MEM(BPF_DW, R10, R1, -8), +#ifdef __BIG_ENDIAN + BPF_LDX_MEM(BPF_B, R0, R10, -1), +#else + BPF_LDX_MEM(BPF_B, R0, R10, -8), +#endif + BPF_JMP_REG(BPF_JNE, R0, R2, 1), + BPF_ALU64_IMM(BPF_MOV, R0, 0), + BPF_EXIT_INSN(), + }, + INTERNAL, + { }, + { { 0, 0 } }, + .stack_depth = 8, + }, + { + "BPF_LDX_MEM | BPF_H", + .u.insns_int = { + BPF_LD_IMM64(R1, 0x0102030405060708ULL), + BPF_LD_IMM64(R2, 0x0000000000000708ULL), + BPF_STX_MEM(BPF_DW, R10, R1, -8), +#ifdef __BIG_ENDIAN + BPF_LDX_MEM(BPF_H, R0, R10, -2), +#else + BPF_LDX_MEM(BPF_H, R0, R10, -8), +#endif + BPF_JMP_REG(BPF_JNE, R0, R2, 1), + BPF_ALU64_IMM(BPF_MOV, R0, 0), + BPF_EXIT_INSN(), + }, + INTERNAL, + { }, + { { 0, 0 } }, + .stack_depth = 8, + }, + { + "BPF_LDX_MEM | BPF_H, MSB set", + .u.insns_int = { + BPF_LD_IMM64(R1, 0x8182838485868788ULL), + BPF_LD_IMM64(R2, 0x0000000000008788ULL), + BPF_STX_MEM(BPF_DW, R10, R1, -8), +#ifdef __BIG_ENDIAN + BPF_LDX_MEM(BPF_H, R0, R10, -2), +#else + BPF_LDX_MEM(BPF_H, R0, R10, -8), +#endif + BPF_JMP_REG(BPF_JNE, R0, R2, 1), + BPF_ALU64_IMM(BPF_MOV, R0, 0), + BPF_EXIT_INSN(), + }, + INTERNAL, + { }, + { { 0, 0 } }, + .stack_depth = 8, + }, + { + "BPF_LDX_MEM | BPF_W", + .u.insns_int = { + BPF_LD_IMM64(R1, 0x0102030405060708ULL), + BPF_LD_IMM64(R2, 0x0000000005060708ULL), + BPF_STX_MEM(BPF_DW, R10, R1, -8), +#ifdef __BIG_ENDIAN + BPF_LDX_MEM(BPF_W, R0, R10, -4), +#else + BPF_LDX_MEM(BPF_W, R0, R10, -8), +#endif + BPF_JMP_REG(BPF_JNE, R0, R2, 1), + BPF_ALU64_IMM(BPF_MOV, R0, 0), + BPF_EXIT_INSN(), + }, + INTERNAL, + { }, + { { 0, 0 } }, + .stack_depth = 8, + }, + { + "BPF_LDX_MEM | BPF_W, MSB set", + .u.insns_int = { + BPF_LD_IMM64(R1, 0x8182838485868788ULL), + BPF_LD_IMM64(R2, 0x0000000085868788ULL), + BPF_STX_MEM(BPF_DW, R10, R1, -8), +#ifdef __BIG_ENDIAN + BPF_LDX_MEM(BPF_W, R0, R10, -4), +#else + BPF_LDX_MEM(BPF_W, R0, R10, -8), +#endif + BPF_JMP_REG(BPF_JNE, R0, R2, 1), + BPF_ALU64_IMM(BPF_MOV, R0, 0), + BPF_EXIT_INSN(), + }, + INTERNAL, + { }, + { { 0, 0 } }, + .stack_depth = 8, + }, + /* BPF_STX_MEM B/H/W/DW */ + { + "BPF_STX_MEM | BPF_B", + .u.insns_int = { + BPF_LD_IMM64(R1, 0x8090a0b0c0d0e0f0ULL), + BPF_LD_IMM64(R2, 0x0102030405060708ULL), + BPF_LD_IMM64(R3, 0x8090a0b0c0d0e008ULL), + BPF_STX_MEM(BPF_DW, R10, R1, -8), +#ifdef __BIG_ENDIAN + BPF_STX_MEM(BPF_B, R10, R2, -1), +#else + BPF_STX_MEM(BPF_B, R10, R2, -8), +#endif + BPF_LDX_MEM(BPF_DW, R0, R10, -8), + BPF_JMP_REG(BPF_JNE, R0, R3, 1), + BPF_ALU64_IMM(BPF_MOV, R0, 0), + BPF_EXIT_INSN(), + }, + INTERNAL, + { }, + { { 0, 0 } }, + .stack_depth = 8, + }, + { + "BPF_STX_MEM | BPF_B, MSB set", + .u.insns_int = { + BPF_LD_IMM64(R1, 0x8090a0b0c0d0e0f0ULL), + BPF_LD_IMM64(R2, 0x8182838485868788ULL), + BPF_LD_IMM64(R3, 0x8090a0b0c0d0e088ULL), + BPF_STX_MEM(BPF_DW, R10, R1, -8), +#ifdef __BIG_ENDIAN + BPF_STX_MEM(BPF_B, R10, R2, -1), +#else + BPF_STX_MEM(BPF_B, R10, R2, -8), +#endif + BPF_LDX_MEM(BPF_DW, R0, R10, -8), + BPF_JMP_REG(BPF_JNE, R0, R3, 1), + BPF_ALU64_IMM(BPF_MOV, R0, 0), + BPF_EXIT_INSN(), + }, + INTERNAL, + { }, + { { 0, 0 } }, + .stack_depth = 8, + }, + { + "BPF_STX_MEM | BPF_H", + .u.insns_int = { + BPF_LD_IMM64(R1, 0x8090a0b0c0d0e0f0ULL), + BPF_LD_IMM64(R2, 0x0102030405060708ULL), + BPF_LD_IMM64(R3, 0x8090a0b0c0d00708ULL), + BPF_STX_MEM(BPF_DW, R10, R1, -8), +#ifdef __BIG_ENDIAN + BPF_STX_MEM(BPF_H, R10, R2, -2), +#else + BPF_STX_MEM(BPF_H, R10, R2, -8), +#endif + BPF_LDX_MEM(BPF_DW, R0, R10, -8), + BPF_JMP_REG(BPF_JNE, R0, R3, 1), + BPF_ALU64_IMM(BPF_MOV, R0, 0), + BPF_EXIT_INSN(), + }, + INTERNAL, + { }, + { { 0, 0 } }, + .stack_depth = 8, + }, + { + "BPF_STX_MEM | BPF_H, MSB set", + .u.insns_int = { + BPF_LD_IMM64(R1, 0x8090a0b0c0d0e0f0ULL), + BPF_LD_IMM64(R2, 0x8182838485868788ULL), + BPF_LD_IMM64(R3, 0x8090a0b0c0d08788ULL), + BPF_STX_MEM(BPF_DW, R10, R1, -8), +#ifdef __BIG_ENDIAN + BPF_STX_MEM(BPF_H, R10, R2, -2), +#else + BPF_STX_MEM(BPF_H, R10, R2, -8), +#endif + BPF_LDX_MEM(BPF_DW, R0, R10, -8), + BPF_JMP_REG(BPF_JNE, R0, R3, 1), + BPF_ALU64_IMM(BPF_MOV, R0, 0), + BPF_EXIT_INSN(), + }, + INTERNAL, + { }, + { { 0, 0 } }, + .stack_depth = 8, + }, + { + "BPF_STX_MEM | BPF_W", + .u.insns_int = { + BPF_LD_IMM64(R1, 0x8090a0b0c0d0e0f0ULL), + BPF_LD_IMM64(R2, 0x0102030405060708ULL), + BPF_LD_IMM64(R3, 0x8090a0b005060708ULL), + BPF_STX_MEM(BPF_DW, R10, R1, -8), +#ifdef __BIG_ENDIAN + BPF_STX_MEM(BPF_W, R10, R2, -4), +#else + BPF_STX_MEM(BPF_W, R10, R2, -8), +#endif + BPF_LDX_MEM(BPF_DW, R0, R10, -8), + BPF_JMP_REG(BPF_JNE, R0, R3, 1), + BPF_ALU64_IMM(BPF_MOV, R0, 0), + BPF_EXIT_INSN(), + }, + INTERNAL, + { }, + { { 0, 0 } }, + .stack_depth = 8, + }, + { + "BPF_STX_MEM | BPF_W, MSB set", + .u.insns_int = { + BPF_LD_IMM64(R1, 0x8090a0b0c0d0e0f0ULL), + BPF_LD_IMM64(R2, 0x8182838485868788ULL), + BPF_LD_IMM64(R3, 0x8090a0b085868788ULL), + BPF_STX_MEM(BPF_DW, R10, R1, -8), +#ifdef __BIG_ENDIAN + BPF_STX_MEM(BPF_W, R10, R2, -4), +#else + BPF_STX_MEM(BPF_W, R10, R2, -8), +#endif + BPF_LDX_MEM(BPF_DW, R0, R10, -8), + BPF_JMP_REG(BPF_JNE, R0, R3, 1), + BPF_ALU64_IMM(BPF_MOV, R0, 0), + BPF_EXIT_INSN(), + }, + INTERNAL, + { }, + { { 0, 0 } }, + .stack_depth = 8, + }, /* BPF_ST(X) | BPF_MEM | BPF_B/H/W/DW */ { "ST_MEM_B: Store/Load byte: max negative", -- cgit v1.2.3 From 89b63462765cc0370f22ebec53d3e83cbbb17613 Mon Sep 17 00:00:00 2001 From: Johan Almbladh Date: Fri, 1 Oct 2021 15:03:40 +0200 Subject: bpf/tests: Add zero-extension checks in BPF_ATOMIC tests This patch updates the existing tests of BPF_ATOMIC operations to verify that a 32-bit register operand is properly zero-extended. In particular, it checks the operation on archs that require 32-bit operands to be properly zero-/sign-extended or the result is undefined, e.g. MIPS64. Signed-off-by: Johan Almbladh Signed-off-by: Daniel Borkmann Link: https://lore.kernel.org/bpf/20211001130348.3670534-3-johan.almbladh@anyfinetworks.com --- lib/test_bpf.c | 22 +++++++++++++++++----- 1 file changed, 17 insertions(+), 5 deletions(-) (limited to 'lib') diff --git a/lib/test_bpf.c b/lib/test_bpf.c index a838a6179ca4..f6983ad7b981 100644 --- a/lib/test_bpf.c +++ b/lib/test_bpf.c @@ -7398,15 +7398,20 @@ static struct bpf_test tests[] = { * Individual tests are expanded from template macros for all * combinations of ALU operation, word size and fetching. */ +#define BPF_ATOMIC_POISON(width) ((width) == BPF_W ? (0xbaadf00dULL << 32) : 0) + #define BPF_ATOMIC_OP_TEST1(width, op, logic, old, update, result) \ { \ "BPF_ATOMIC | " #width ", " #op ": Test: " \ #old " " #logic " " #update " = " #result, \ .u.insns_int = { \ - BPF_ALU32_IMM(BPF_MOV, R5, update), \ + BPF_LD_IMM64(R5, (update) | BPF_ATOMIC_POISON(width)), \ BPF_ST_MEM(width, R10, -40, old), \ BPF_ATOMIC_OP(width, op, R10, R5, -40), \ BPF_LDX_MEM(width, R0, R10, -40), \ + BPF_ALU64_REG(BPF_MOV, R1, R0), \ + BPF_ALU64_IMM(BPF_RSH, R1, 32), \ + BPF_ALU64_REG(BPF_OR, R0, R1), \ BPF_EXIT_INSN(), \ }, \ INTERNAL, \ @@ -7420,11 +7425,14 @@ static struct bpf_test tests[] = { #old " " #logic " " #update " = " #result, \ .u.insns_int = { \ BPF_ALU64_REG(BPF_MOV, R1, R10), \ - BPF_ALU32_IMM(BPF_MOV, R0, update), \ + BPF_LD_IMM64(R0, (update) | BPF_ATOMIC_POISON(width)), \ BPF_ST_MEM(BPF_W, R10, -40, old), \ BPF_ATOMIC_OP(width, op, R10, R0, -40), \ BPF_ALU64_REG(BPF_MOV, R0, R10), \ BPF_ALU64_REG(BPF_SUB, R0, R1), \ + BPF_ALU64_REG(BPF_MOV, R1, R0), \ + BPF_ALU64_IMM(BPF_RSH, R1, 32), \ + BPF_ALU64_REG(BPF_OR, R0, R1), \ BPF_EXIT_INSN(), \ }, \ INTERNAL, \ @@ -7438,10 +7446,13 @@ static struct bpf_test tests[] = { #old " " #logic " " #update " = " #result, \ .u.insns_int = { \ BPF_ALU64_REG(BPF_MOV, R0, R10), \ - BPF_ALU32_IMM(BPF_MOV, R1, update), \ + BPF_LD_IMM64(R1, (update) | BPF_ATOMIC_POISON(width)), \ BPF_ST_MEM(width, R10, -40, old), \ BPF_ATOMIC_OP(width, op, R10, R1, -40), \ BPF_ALU64_REG(BPF_SUB, R0, R10), \ + BPF_ALU64_REG(BPF_MOV, R1, R0), \ + BPF_ALU64_IMM(BPF_RSH, R1, 32), \ + BPF_ALU64_REG(BPF_OR, R0, R1), \ BPF_EXIT_INSN(), \ }, \ INTERNAL, \ @@ -7454,10 +7465,10 @@ static struct bpf_test tests[] = { "BPF_ATOMIC | " #width ", " #op ": Test fetch: " \ #old " " #logic " " #update " = " #result, \ .u.insns_int = { \ - BPF_ALU32_IMM(BPF_MOV, R3, update), \ + BPF_LD_IMM64(R3, (update) | BPF_ATOMIC_POISON(width)), \ BPF_ST_MEM(width, R10, -40, old), \ BPF_ATOMIC_OP(width, op, R10, R3, -40), \ - BPF_ALU64_REG(BPF_MOV, R0, R3), \ + BPF_ALU32_REG(BPF_MOV, R0, R3), \ BPF_EXIT_INSN(), \ }, \ INTERNAL, \ @@ -7555,6 +7566,7 @@ static struct bpf_test tests[] = { BPF_ATOMIC_OP_TEST2(BPF_DW, BPF_XCHG, xchg, 0x12, 0xab, 0xab), BPF_ATOMIC_OP_TEST3(BPF_DW, BPF_XCHG, xchg, 0x12, 0xab, 0xab), BPF_ATOMIC_OP_TEST4(BPF_DW, BPF_XCHG, xchg, 0x12, 0xab, 0xab), +#undef BPF_ATOMIC_POISON #undef BPF_ATOMIC_OP_TEST1 #undef BPF_ATOMIC_OP_TEST2 #undef BPF_ATOMIC_OP_TEST3 -- cgit v1.2.3 From f68e8efd7fa506928432b8cd41b8c7d91d804e02 Mon Sep 17 00:00:00 2001 From: Johan Almbladh Date: Fri, 1 Oct 2021 15:03:41 +0200 Subject: bpf/tests: Add exhaustive tests of BPF_ATOMIC magnitudes This patch adds a series of test to verify the operation of BPF_ATOMIC with BPF_DW and BPF_W sizes, for all power-of-two magnitudes of the register value operand. Also fixes a confusing typo in the comment for a related test. Signed-off-by: Johan Almbladh Signed-off-by: Daniel Borkmann Link: https://lore.kernel.org/bpf/20211001130348.3670534-4-johan.almbladh@anyfinetworks.com --- lib/test_bpf.c | 504 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++- 1 file changed, 503 insertions(+), 1 deletion(-) (limited to 'lib') diff --git a/lib/test_bpf.c b/lib/test_bpf.c index f6983ad7b981..84efb23e09d0 100644 --- a/lib/test_bpf.c +++ b/lib/test_bpf.c @@ -796,7 +796,7 @@ static int __bpf_fill_pattern(struct bpf_test *self, void *arg, /* * Exhaustive tests of ALU operations for all combinations of power-of-two * magnitudes of the operands, both for positive and negative values. The - * test is designed to verify e.g. the JMP and JMP32 operations for JITs that + * test is designed to verify e.g. the ALU and ALU64 operations for JITs that * emit different code depending on the magnitude of the immediate value. */ @@ -1137,6 +1137,306 @@ static int bpf_fill_alu32_mod_reg(struct bpf_test *self) return __bpf_fill_alu32_reg(self, BPF_MOD); } +/* + * Exhaustive tests of atomic operations for all power-of-two operand + * magnitudes, both for positive and negative values. + */ + +static int __bpf_emit_atomic64(struct bpf_test *self, void *arg, + struct bpf_insn *insns, s64 dst, s64 src) +{ + int op = *(int *)arg; + u64 keep, fetch, res; + int i = 0; + + if (!insns) + return 21; + + switch (op) { + case BPF_XCHG: + res = src; + break; + default: + __bpf_alu_result(&res, dst, src, BPF_OP(op)); + } + + keep = 0x0123456789abcdefULL; + if (op & BPF_FETCH) + fetch = dst; + else + fetch = src; + + i += __bpf_ld_imm64(&insns[i], R0, keep); + i += __bpf_ld_imm64(&insns[i], R1, dst); + i += __bpf_ld_imm64(&insns[i], R2, src); + i += __bpf_ld_imm64(&insns[i], R3, res); + i += __bpf_ld_imm64(&insns[i], R4, fetch); + i += __bpf_ld_imm64(&insns[i], R5, keep); + + insns[i++] = BPF_STX_MEM(BPF_DW, R10, R1, -8); + insns[i++] = BPF_ATOMIC_OP(BPF_DW, op, R10, R2, -8); + insns[i++] = BPF_LDX_MEM(BPF_DW, R1, R10, -8); + + insns[i++] = BPF_JMP_REG(BPF_JEQ, R1, R3, 1); + insns[i++] = BPF_EXIT_INSN(); + + insns[i++] = BPF_JMP_REG(BPF_JEQ, R2, R4, 1); + insns[i++] = BPF_EXIT_INSN(); + + insns[i++] = BPF_JMP_REG(BPF_JEQ, R0, R5, 1); + insns[i++] = BPF_EXIT_INSN(); + + return i; +} + +static int __bpf_emit_atomic32(struct bpf_test *self, void *arg, + struct bpf_insn *insns, s64 dst, s64 src) +{ + int op = *(int *)arg; + u64 keep, fetch, res; + int i = 0; + + if (!insns) + return 21; + + switch (op) { + case BPF_XCHG: + res = src; + break; + default: + __bpf_alu_result(&res, (u32)dst, (u32)src, BPF_OP(op)); + } + + keep = 0x0123456789abcdefULL; + if (op & BPF_FETCH) + fetch = (u32)dst; + else + fetch = src; + + i += __bpf_ld_imm64(&insns[i], R0, keep); + i += __bpf_ld_imm64(&insns[i], R1, (u32)dst); + i += __bpf_ld_imm64(&insns[i], R2, src); + i += __bpf_ld_imm64(&insns[i], R3, (u32)res); + i += __bpf_ld_imm64(&insns[i], R4, fetch); + i += __bpf_ld_imm64(&insns[i], R5, keep); + + insns[i++] = BPF_STX_MEM(BPF_W, R10, R1, -4); + insns[i++] = BPF_ATOMIC_OP(BPF_W, op, R10, R2, -4); + insns[i++] = BPF_LDX_MEM(BPF_W, R1, R10, -4); + + insns[i++] = BPF_JMP_REG(BPF_JEQ, R1, R3, 1); + insns[i++] = BPF_EXIT_INSN(); + + insns[i++] = BPF_JMP_REG(BPF_JEQ, R2, R4, 1); + insns[i++] = BPF_EXIT_INSN(); + + insns[i++] = BPF_JMP_REG(BPF_JEQ, R0, R5, 1); + insns[i++] = BPF_EXIT_INSN(); + + return i; +} + +static int __bpf_emit_cmpxchg64(struct bpf_test *self, void *arg, + struct bpf_insn *insns, s64 dst, s64 src) +{ + int i = 0; + + if (!insns) + return 23; + + i += __bpf_ld_imm64(&insns[i], R0, ~dst); + i += __bpf_ld_imm64(&insns[i], R1, dst); + i += __bpf_ld_imm64(&insns[i], R2, src); + + /* Result unsuccessful */ + insns[i++] = BPF_STX_MEM(BPF_DW, R10, R1, -8); + insns[i++] = BPF_ATOMIC_OP(BPF_DW, BPF_CMPXCHG, R10, R2, -8); + insns[i++] = BPF_LDX_MEM(BPF_DW, R3, R10, -8); + + insns[i++] = BPF_JMP_REG(BPF_JEQ, R1, R3, 2); + insns[i++] = BPF_MOV64_IMM(R0, __LINE__); + insns[i++] = BPF_EXIT_INSN(); + + insns[i++] = BPF_JMP_REG(BPF_JEQ, R0, R3, 2); + insns[i++] = BPF_MOV64_IMM(R0, __LINE__); + insns[i++] = BPF_EXIT_INSN(); + + /* Result successful */ + insns[i++] = BPF_ATOMIC_OP(BPF_DW, BPF_CMPXCHG, R10, R2, -8); + insns[i++] = BPF_LDX_MEM(BPF_DW, R3, R10, -8); + + insns[i++] = BPF_JMP_REG(BPF_JEQ, R2, R3, 2); + insns[i++] = BPF_MOV64_IMM(R0, __LINE__); + insns[i++] = BPF_EXIT_INSN(); + + insns[i++] = BPF_JMP_REG(BPF_JEQ, R0, R1, 2); + insns[i++] = BPF_MOV64_IMM(R0, __LINE__); + insns[i++] = BPF_EXIT_INSN(); + + return i; +} + +static int __bpf_emit_cmpxchg32(struct bpf_test *self, void *arg, + struct bpf_insn *insns, s64 dst, s64 src) +{ + int i = 0; + + if (!insns) + return 27; + + i += __bpf_ld_imm64(&insns[i], R0, ~dst); + i += __bpf_ld_imm64(&insns[i], R1, (u32)dst); + i += __bpf_ld_imm64(&insns[i], R2, src); + + /* Result unsuccessful */ + insns[i++] = BPF_STX_MEM(BPF_W, R10, R1, -4); + insns[i++] = BPF_ATOMIC_OP(BPF_W, BPF_CMPXCHG, R10, R2, -4); + insns[i++] = BPF_ZEXT_REG(R0), /* Zext always inserted by verifier */ + insns[i++] = BPF_LDX_MEM(BPF_W, R3, R10, -4); + + insns[i++] = BPF_JMP32_REG(BPF_JEQ, R1, R3, 2); + insns[i++] = BPF_MOV32_IMM(R0, __LINE__); + insns[i++] = BPF_EXIT_INSN(); + + insns[i++] = BPF_JMP_REG(BPF_JEQ, R0, R3, 2); + insns[i++] = BPF_MOV32_IMM(R0, __LINE__); + insns[i++] = BPF_EXIT_INSN(); + + /* Result successful */ + i += __bpf_ld_imm64(&insns[i], R0, dst); + insns[i++] = BPF_ATOMIC_OP(BPF_W, BPF_CMPXCHG, R10, R2, -4); + insns[i++] = BPF_ZEXT_REG(R0), /* Zext always inserted by verifier */ + insns[i++] = BPF_LDX_MEM(BPF_W, R3, R10, -4); + + insns[i++] = BPF_JMP32_REG(BPF_JEQ, R2, R3, 2); + insns[i++] = BPF_MOV32_IMM(R0, __LINE__); + insns[i++] = BPF_EXIT_INSN(); + + insns[i++] = BPF_JMP_REG(BPF_JEQ, R0, R1, 2); + insns[i++] = BPF_MOV32_IMM(R0, __LINE__); + insns[i++] = BPF_EXIT_INSN(); + + return i; +} + +static int __bpf_fill_atomic64(struct bpf_test *self, int op) +{ + return __bpf_fill_pattern(self, &op, 64, 64, + 0, PATTERN_BLOCK2, + &__bpf_emit_atomic64); +} + +static int __bpf_fill_atomic32(struct bpf_test *self, int op) +{ + return __bpf_fill_pattern(self, &op, 64, 64, + 0, PATTERN_BLOCK2, + &__bpf_emit_atomic32); +} + +/* 64-bit atomic operations */ +static int bpf_fill_atomic64_add(struct bpf_test *self) +{ + return __bpf_fill_atomic64(self, BPF_ADD); +} + +static int bpf_fill_atomic64_and(struct bpf_test *self) +{ + return __bpf_fill_atomic64(self, BPF_AND); +} + +static int bpf_fill_atomic64_or(struct bpf_test *self) +{ + return __bpf_fill_atomic64(self, BPF_OR); +} + +static int bpf_fill_atomic64_xor(struct bpf_test *self) +{ + return __bpf_fill_atomic64(self, BPF_XOR); +} + +static int bpf_fill_atomic64_add_fetch(struct bpf_test *self) +{ + return __bpf_fill_atomic64(self, BPF_ADD | BPF_FETCH); +} + +static int bpf_fill_atomic64_and_fetch(struct bpf_test *self) +{ + return __bpf_fill_atomic64(self, BPF_AND | BPF_FETCH); +} + +static int bpf_fill_atomic64_or_fetch(struct bpf_test *self) +{ + return __bpf_fill_atomic64(self, BPF_OR | BPF_FETCH); +} + +static int bpf_fill_atomic64_xor_fetch(struct bpf_test *self) +{ + return __bpf_fill_atomic64(self, BPF_XOR | BPF_FETCH); +} + +static int bpf_fill_atomic64_xchg(struct bpf_test *self) +{ + return __bpf_fill_atomic64(self, BPF_XCHG); +} + +static int bpf_fill_cmpxchg64(struct bpf_test *self) +{ + return __bpf_fill_pattern(self, NULL, 64, 64, 0, PATTERN_BLOCK2, + &__bpf_emit_cmpxchg64); +} + +/* 32-bit atomic operations */ +static int bpf_fill_atomic32_add(struct bpf_test *self) +{ + return __bpf_fill_atomic32(self, BPF_ADD); +} + +static int bpf_fill_atomic32_and(struct bpf_test *self) +{ + return __bpf_fill_atomic32(self, BPF_AND); +} + +static int bpf_fill_atomic32_or(struct bpf_test *self) +{ + return __bpf_fill_atomic32(self, BPF_OR); +} + +static int bpf_fill_atomic32_xor(struct bpf_test *self) +{ + return __bpf_fill_atomic32(self, BPF_XOR); +} + +static int bpf_fill_atomic32_add_fetch(struct bpf_test *self) +{ + return __bpf_fill_atomic32(self, BPF_ADD | BPF_FETCH); +} + +static int bpf_fill_atomic32_and_fetch(struct bpf_test *self) +{ + return __bpf_fill_atomic32(self, BPF_AND | BPF_FETCH); +} + +static int bpf_fill_atomic32_or_fetch(struct bpf_test *self) +{ + return __bpf_fill_atomic32(self, BPF_OR | BPF_FETCH); +} + +static int bpf_fill_atomic32_xor_fetch(struct bpf_test *self) +{ + return __bpf_fill_atomic32(self, BPF_XOR | BPF_FETCH); +} + +static int bpf_fill_atomic32_xchg(struct bpf_test *self) +{ + return __bpf_fill_atomic32(self, BPF_XCHG); +} + +static int bpf_fill_cmpxchg32(struct bpf_test *self) +{ + return __bpf_fill_pattern(self, NULL, 64, 64, 0, PATTERN_BLOCK2, + &__bpf_emit_cmpxchg32); +} + /* * Test the two-instruction 64-bit immediate load operation for all * power-of-two magnitudes of the immediate operand. For each MSB, a block @@ -10721,6 +11021,208 @@ static struct bpf_test tests[] = { { { 0, 1 } }, .fill_helper = bpf_fill_ld_imm64, }, + /* 64-bit ATOMIC magnitudes */ + { + "ATOMIC_DW_ADD: all operand magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_atomic64_add, + .stack_depth = 8, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "ATOMIC_DW_AND: all operand magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_atomic64_and, + .stack_depth = 8, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "ATOMIC_DW_OR: all operand magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_atomic64_or, + .stack_depth = 8, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "ATOMIC_DW_XOR: all operand magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_atomic64_xor, + .stack_depth = 8, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "ATOMIC_DW_ADD_FETCH: all operand magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_atomic64_add_fetch, + .stack_depth = 8, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "ATOMIC_DW_AND_FETCH: all operand magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_atomic64_and_fetch, + .stack_depth = 8, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "ATOMIC_DW_OR_FETCH: all operand magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_atomic64_or_fetch, + .stack_depth = 8, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "ATOMIC_DW_XOR_FETCH: all operand magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_atomic64_xor_fetch, + .stack_depth = 8, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "ATOMIC_DW_XCHG: all operand magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_atomic64_xchg, + .stack_depth = 8, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "ATOMIC_DW_CMPXCHG: all operand magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_cmpxchg64, + .stack_depth = 8, + .nr_testruns = NR_PATTERN_RUNS, + }, + /* 64-bit atomic magnitudes */ + { + "ATOMIC_W_ADD: all operand magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_atomic32_add, + .stack_depth = 8, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "ATOMIC_W_AND: all operand magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_atomic32_and, + .stack_depth = 8, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "ATOMIC_W_OR: all operand magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_atomic32_or, + .stack_depth = 8, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "ATOMIC_W_XOR: all operand magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_atomic32_xor, + .stack_depth = 8, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "ATOMIC_W_ADD_FETCH: all operand magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_atomic32_add_fetch, + .stack_depth = 8, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "ATOMIC_W_AND_FETCH: all operand magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_atomic32_and_fetch, + .stack_depth = 8, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "ATOMIC_W_OR_FETCH: all operand magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_atomic32_or_fetch, + .stack_depth = 8, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "ATOMIC_W_XOR_FETCH: all operand magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_atomic32_xor_fetch, + .stack_depth = 8, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "ATOMIC_W_XCHG: all operand magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_atomic32_xchg, + .stack_depth = 8, + .nr_testruns = NR_PATTERN_RUNS, + }, + { + "ATOMIC_W_CMPXCHG: all operand magnitudes", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_cmpxchg32, + .stack_depth = 8, + .nr_testruns = NR_PATTERN_RUNS, + }, /* JMP immediate magnitudes */ { "JMP_JSET_K: all immediate value magnitudes", -- cgit v1.2.3 From 0bbaa02b481682004cf812dbeca68272752a5e8a Mon Sep 17 00:00:00 2001 From: Johan Almbladh Date: Fri, 1 Oct 2021 15:03:42 +0200 Subject: bpf/tests: Add tests to check source register zero-extension This patch adds tests to check that the source register is preserved when zero-extending a 32-bit value. In particular, it checks that the source operand is not zero-extended in-place. Signed-off-by: Johan Almbladh Signed-off-by: Daniel Borkmann Link: https://lore.kernel.org/bpf/20211001130348.3670534-5-johan.almbladh@anyfinetworks.com --- lib/test_bpf.c | 143 +++++++++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 143 insertions(+) (limited to 'lib') diff --git a/lib/test_bpf.c b/lib/test_bpf.c index 84efb23e09d0..c7db90112ef0 100644 --- a/lib/test_bpf.c +++ b/lib/test_bpf.c @@ -10586,6 +10586,149 @@ static struct bpf_test tests[] = { {}, { { 0, 2 } }, }, + /* Checking that ALU32 src is not zero extended in place */ +#define BPF_ALU32_SRC_ZEXT(op) \ + { \ + "ALU32_" #op "_X: src preserved in zext", \ + .u.insns_int = { \ + BPF_LD_IMM64(R1, 0x0123456789acbdefULL),\ + BPF_LD_IMM64(R2, 0xfedcba9876543210ULL),\ + BPF_ALU64_REG(BPF_MOV, R0, R1), \ + BPF_ALU32_REG(BPF_##op, R2, R1), \ + BPF_ALU64_REG(BPF_SUB, R0, R1), \ + BPF_ALU64_REG(BPF_MOV, R1, R0), \ + BPF_ALU64_IMM(BPF_RSH, R1, 32), \ + BPF_ALU64_REG(BPF_OR, R0, R1), \ + BPF_EXIT_INSN(), \ + }, \ + INTERNAL, \ + { }, \ + { { 0, 0 } }, \ + } + BPF_ALU32_SRC_ZEXT(MOV), + BPF_ALU32_SRC_ZEXT(AND), + BPF_ALU32_SRC_ZEXT(OR), + BPF_ALU32_SRC_ZEXT(XOR), + BPF_ALU32_SRC_ZEXT(ADD), + BPF_ALU32_SRC_ZEXT(SUB), + BPF_ALU32_SRC_ZEXT(MUL), + BPF_ALU32_SRC_ZEXT(DIV), + BPF_ALU32_SRC_ZEXT(MOD), +#undef BPF_ALU32_SRC_ZEXT + /* Checking that ATOMIC32 src is not zero extended in place */ +#define BPF_ATOMIC32_SRC_ZEXT(op) \ + { \ + "ATOMIC_W_" #op ": src preserved in zext", \ + .u.insns_int = { \ + BPF_LD_IMM64(R0, 0x0123456789acbdefULL), \ + BPF_ALU64_REG(BPF_MOV, R1, R0), \ + BPF_ST_MEM(BPF_W, R10, -4, 0), \ + BPF_ATOMIC_OP(BPF_W, BPF_##op, R10, R1, -4), \ + BPF_ALU64_REG(BPF_SUB, R0, R1), \ + BPF_ALU64_REG(BPF_MOV, R1, R0), \ + BPF_ALU64_IMM(BPF_RSH, R1, 32), \ + BPF_ALU64_REG(BPF_OR, R0, R1), \ + BPF_EXIT_INSN(), \ + }, \ + INTERNAL, \ + { }, \ + { { 0, 0 } }, \ + .stack_depth = 8, \ + } + BPF_ATOMIC32_SRC_ZEXT(ADD), + BPF_ATOMIC32_SRC_ZEXT(AND), + BPF_ATOMIC32_SRC_ZEXT(OR), + BPF_ATOMIC32_SRC_ZEXT(XOR), +#undef BPF_ATOMIC32_SRC_ZEXT + /* Checking that CMPXCHG32 src is not zero extended in place */ + { + "ATOMIC_W_CMPXCHG: src preserved in zext", + .u.insns_int = { + BPF_LD_IMM64(R1, 0x0123456789acbdefULL), + BPF_ALU64_REG(BPF_MOV, R2, R1), + BPF_ALU64_REG(BPF_MOV, R0, 0), + BPF_ST_MEM(BPF_W, R10, -4, 0), + BPF_ATOMIC_OP(BPF_W, BPF_CMPXCHG, R10, R1, -4), + BPF_ALU64_REG(BPF_SUB, R1, R2), + BPF_ALU64_REG(BPF_MOV, R2, R1), + BPF_ALU64_IMM(BPF_RSH, R2, 32), + BPF_ALU64_REG(BPF_OR, R1, R2), + BPF_ALU64_REG(BPF_MOV, R0, R1), + BPF_EXIT_INSN(), + }, + INTERNAL, + { }, + { { 0, 0 } }, + .stack_depth = 8, + }, + /* Checking that JMP32 immediate src is not zero extended in place */ +#define BPF_JMP32_IMM_ZEXT(op) \ + { \ + "JMP32_" #op "_K: operand preserved in zext", \ + .u.insns_int = { \ + BPF_LD_IMM64(R0, 0x0123456789acbdefULL),\ + BPF_ALU64_REG(BPF_MOV, R1, R0), \ + BPF_JMP32_IMM(BPF_##op, R0, 1234, 1), \ + BPF_JMP_A(0), /* Nop */ \ + BPF_ALU64_REG(BPF_SUB, R0, R1), \ + BPF_ALU64_REG(BPF_MOV, R1, R0), \ + BPF_ALU64_IMM(BPF_RSH, R1, 32), \ + BPF_ALU64_REG(BPF_OR, R0, R1), \ + BPF_EXIT_INSN(), \ + }, \ + INTERNAL, \ + { }, \ + { { 0, 0 } }, \ + } + BPF_JMP32_IMM_ZEXT(JEQ), + BPF_JMP32_IMM_ZEXT(JNE), + BPF_JMP32_IMM_ZEXT(JSET), + BPF_JMP32_IMM_ZEXT(JGT), + BPF_JMP32_IMM_ZEXT(JGE), + BPF_JMP32_IMM_ZEXT(JLT), + BPF_JMP32_IMM_ZEXT(JLE), + BPF_JMP32_IMM_ZEXT(JSGT), + BPF_JMP32_IMM_ZEXT(JSGE), + BPF_JMP32_IMM_ZEXT(JSGT), + BPF_JMP32_IMM_ZEXT(JSLT), + BPF_JMP32_IMM_ZEXT(JSLE), +#undef BPF_JMP2_IMM_ZEXT + /* Checking that JMP32 dst & src are not zero extended in place */ +#define BPF_JMP32_REG_ZEXT(op) \ + { \ + "JMP32_" #op "_X: operands preserved in zext", \ + .u.insns_int = { \ + BPF_LD_IMM64(R0, 0x0123456789acbdefULL),\ + BPF_LD_IMM64(R1, 0xfedcba9876543210ULL),\ + BPF_ALU64_REG(BPF_MOV, R2, R0), \ + BPF_ALU64_REG(BPF_MOV, R3, R1), \ + BPF_JMP32_IMM(BPF_##op, R0, R1, 1), \ + BPF_JMP_A(0), /* Nop */ \ + BPF_ALU64_REG(BPF_SUB, R0, R2), \ + BPF_ALU64_REG(BPF_SUB, R1, R3), \ + BPF_ALU64_REG(BPF_OR, R0, R1), \ + BPF_ALU64_REG(BPF_MOV, R1, R0), \ + BPF_ALU64_IMM(BPF_RSH, R1, 32), \ + BPF_ALU64_REG(BPF_OR, R0, R1), \ + BPF_EXIT_INSN(), \ + }, \ + INTERNAL, \ + { }, \ + { { 0, 0 } }, \ + } + BPF_JMP32_REG_ZEXT(JEQ), + BPF_JMP32_REG_ZEXT(JNE), + BPF_JMP32_REG_ZEXT(JSET), + BPF_JMP32_REG_ZEXT(JGT), + BPF_JMP32_REG_ZEXT(JGE), + BPF_JMP32_REG_ZEXT(JLT), + BPF_JMP32_REG_ZEXT(JLE), + BPF_JMP32_REG_ZEXT(JSGT), + BPF_JMP32_REG_ZEXT(JSGE), + BPF_JMP32_REG_ZEXT(JSGT), + BPF_JMP32_REG_ZEXT(JSLT), + BPF_JMP32_REG_ZEXT(JSLE), +#undef BPF_JMP2_REG_ZEXT /* Exhaustive test of ALU64 shift operations */ { "ALU64_LSH_K: all shift values", -- cgit v1.2.3 From e2f9797b3c7396065ca3bc9f223225ca63c1e2bd Mon Sep 17 00:00:00 2001 From: Johan Almbladh Date: Fri, 1 Oct 2021 15:03:43 +0200 Subject: bpf/tests: Add more tests for ALU and ATOMIC register clobbering This patch expands the register-clobbering-during-function-call tests to cover more all ALU32/64 MUL, DIV and MOD operations and all ATOMIC operations. In short, if a JIT implements a complex operation with a call to an external function, it must make sure to save and restore all its caller-saved registers that may be clobbered by the call. Signed-off-by: Johan Almbladh Signed-off-by: Daniel Borkmann Link: https://lore.kernel.org/bpf/20211001130348.3670534-6-johan.almbladh@anyfinetworks.com --- lib/test_bpf.c | 267 ++++++++++++++++++++++++++++++++++++++++++--------------- 1 file changed, 197 insertions(+), 70 deletions(-) (limited to 'lib') diff --git a/lib/test_bpf.c b/lib/test_bpf.c index c7db90112ef0..201f34060eef 100644 --- a/lib/test_bpf.c +++ b/lib/test_bpf.c @@ -3744,76 +3744,6 @@ static struct bpf_test tests[] = { { }, { { 0, -1 } } }, - { - /* - * Register (non-)clobbering test, in the case where a 32-bit - * JIT implements complex ALU64 operations via function calls. - * If so, the function call must be invisible in the eBPF - * registers. The JIT must then save and restore relevant - * registers during the call. The following tests check that - * the eBPF registers retain their values after such a call. - */ - "INT: Register clobbering, R1 updated", - .u.insns_int = { - BPF_ALU32_IMM(BPF_MOV, R0, 0), - BPF_ALU32_IMM(BPF_MOV, R1, 123456789), - BPF_ALU32_IMM(BPF_MOV, R2, 2), - BPF_ALU32_IMM(BPF_MOV, R3, 3), - BPF_ALU32_IMM(BPF_MOV, R4, 4), - BPF_ALU32_IMM(BPF_MOV, R5, 5), - BPF_ALU32_IMM(BPF_MOV, R6, 6), - BPF_ALU32_IMM(BPF_MOV, R7, 7), - BPF_ALU32_IMM(BPF_MOV, R8, 8), - BPF_ALU32_IMM(BPF_MOV, R9, 9), - BPF_ALU64_IMM(BPF_DIV, R1, 123456789), - BPF_JMP_IMM(BPF_JNE, R0, 0, 10), - BPF_JMP_IMM(BPF_JNE, R1, 1, 9), - BPF_JMP_IMM(BPF_JNE, R2, 2, 8), - BPF_JMP_IMM(BPF_JNE, R3, 3, 7), - BPF_JMP_IMM(BPF_JNE, R4, 4, 6), - BPF_JMP_IMM(BPF_JNE, R5, 5, 5), - BPF_JMP_IMM(BPF_JNE, R6, 6, 4), - BPF_JMP_IMM(BPF_JNE, R7, 7, 3), - BPF_JMP_IMM(BPF_JNE, R8, 8, 2), - BPF_JMP_IMM(BPF_JNE, R9, 9, 1), - BPF_ALU32_IMM(BPF_MOV, R0, 1), - BPF_EXIT_INSN(), - }, - INTERNAL, - { }, - { { 0, 1 } } - }, - { - "INT: Register clobbering, R2 updated", - .u.insns_int = { - BPF_ALU32_IMM(BPF_MOV, R0, 0), - BPF_ALU32_IMM(BPF_MOV, R1, 1), - BPF_ALU32_IMM(BPF_MOV, R2, 2 * 123456789), - BPF_ALU32_IMM(BPF_MOV, R3, 3), - BPF_ALU32_IMM(BPF_MOV, R4, 4), - BPF_ALU32_IMM(BPF_MOV, R5, 5), - BPF_ALU32_IMM(BPF_MOV, R6, 6), - BPF_ALU32_IMM(BPF_MOV, R7, 7), - BPF_ALU32_IMM(BPF_MOV, R8, 8), - BPF_ALU32_IMM(BPF_MOV, R9, 9), - BPF_ALU64_IMM(BPF_DIV, R2, 123456789), - BPF_JMP_IMM(BPF_JNE, R0, 0, 10), - BPF_JMP_IMM(BPF_JNE, R1, 1, 9), - BPF_JMP_IMM(BPF_JNE, R2, 2, 8), - BPF_JMP_IMM(BPF_JNE, R3, 3, 7), - BPF_JMP_IMM(BPF_JNE, R4, 4, 6), - BPF_JMP_IMM(BPF_JNE, R5, 5, 5), - BPF_JMP_IMM(BPF_JNE, R6, 6, 4), - BPF_JMP_IMM(BPF_JNE, R7, 7, 3), - BPF_JMP_IMM(BPF_JNE, R8, 8, 2), - BPF_JMP_IMM(BPF_JNE, R9, 9, 1), - BPF_ALU32_IMM(BPF_MOV, R0, 1), - BPF_EXIT_INSN(), - }, - INTERNAL, - { }, - { { 0, 1 } } - }, { /* * Test 32-bit JITs that implement complex ALU64 operations as @@ -10586,6 +10516,203 @@ static struct bpf_test tests[] = { {}, { { 0, 2 } }, }, + /* + * Register (non-)clobbering tests for the case where a JIT implements + * complex ALU or ATOMIC operations via function calls. If so, the + * function call must be transparent to the eBPF registers. The JIT + * must therefore save and restore relevant registers across the call. + * The following tests check that the eBPF registers retain their + * values after such an operation. Mainly intended for complex ALU + * and atomic operation, but we run it for all. You never know... + * + * Note that each operations should be tested twice with different + * destinations, to check preservation for all registers. + */ +#define BPF_TEST_CLOBBER_ALU(alu, op, dst, src) \ + { \ + #alu "_" #op " to " #dst ": no clobbering", \ + .u.insns_int = { \ + BPF_ALU64_IMM(BPF_MOV, R0, R0), \ + BPF_ALU64_IMM(BPF_MOV, R1, R1), \ + BPF_ALU64_IMM(BPF_MOV, R2, R2), \ + BPF_ALU64_IMM(BPF_MOV, R3, R3), \ + BPF_ALU64_IMM(BPF_MOV, R4, R4), \ + BPF_ALU64_IMM(BPF_MOV, R5, R5), \ + BPF_ALU64_IMM(BPF_MOV, R6, R6), \ + BPF_ALU64_IMM(BPF_MOV, R7, R7), \ + BPF_ALU64_IMM(BPF_MOV, R8, R8), \ + BPF_ALU64_IMM(BPF_MOV, R9, R9), \ + BPF_##alu(BPF_ ##op, dst, src), \ + BPF_ALU32_IMM(BPF_MOV, dst, dst), \ + BPF_JMP_IMM(BPF_JNE, R0, R0, 10), \ + BPF_JMP_IMM(BPF_JNE, R1, R1, 9), \ + BPF_JMP_IMM(BPF_JNE, R2, R2, 8), \ + BPF_JMP_IMM(BPF_JNE, R3, R3, 7), \ + BPF_JMP_IMM(BPF_JNE, R4, R4, 6), \ + BPF_JMP_IMM(BPF_JNE, R5, R5, 5), \ + BPF_JMP_IMM(BPF_JNE, R6, R6, 4), \ + BPF_JMP_IMM(BPF_JNE, R7, R7, 3), \ + BPF_JMP_IMM(BPF_JNE, R8, R8, 2), \ + BPF_JMP_IMM(BPF_JNE, R9, R9, 1), \ + BPF_ALU64_IMM(BPF_MOV, R0, 1), \ + BPF_EXIT_INSN(), \ + }, \ + INTERNAL, \ + { }, \ + { { 0, 1 } } \ + } + /* ALU64 operations, register clobbering */ + BPF_TEST_CLOBBER_ALU(ALU64_IMM, AND, R8, 123456789), + BPF_TEST_CLOBBER_ALU(ALU64_IMM, AND, R9, 123456789), + BPF_TEST_CLOBBER_ALU(ALU64_IMM, OR, R8, 123456789), + BPF_TEST_CLOBBER_ALU(ALU64_IMM, OR, R9, 123456789), + BPF_TEST_CLOBBER_ALU(ALU64_IMM, XOR, R8, 123456789), + BPF_TEST_CLOBBER_ALU(ALU64_IMM, XOR, R9, 123456789), + BPF_TEST_CLOBBER_ALU(ALU64_IMM, LSH, R8, 12), + BPF_TEST_CLOBBER_ALU(ALU64_IMM, LSH, R9, 12), + BPF_TEST_CLOBBER_ALU(ALU64_IMM, RSH, R8, 12), + BPF_TEST_CLOBBER_ALU(ALU64_IMM, RSH, R9, 12), + BPF_TEST_CLOBBER_ALU(ALU64_IMM, ARSH, R8, 12), + BPF_TEST_CLOBBER_ALU(ALU64_IMM, ARSH, R9, 12), + BPF_TEST_CLOBBER_ALU(ALU64_IMM, ADD, R8, 123456789), + BPF_TEST_CLOBBER_ALU(ALU64_IMM, ADD, R9, 123456789), + BPF_TEST_CLOBBER_ALU(ALU64_IMM, SUB, R8, 123456789), + BPF_TEST_CLOBBER_ALU(ALU64_IMM, SUB, R9, 123456789), + BPF_TEST_CLOBBER_ALU(ALU64_IMM, MUL, R8, 123456789), + BPF_TEST_CLOBBER_ALU(ALU64_IMM, MUL, R9, 123456789), + BPF_TEST_CLOBBER_ALU(ALU64_IMM, DIV, R8, 123456789), + BPF_TEST_CLOBBER_ALU(ALU64_IMM, DIV, R9, 123456789), + BPF_TEST_CLOBBER_ALU(ALU64_IMM, MOD, R8, 123456789), + BPF_TEST_CLOBBER_ALU(ALU64_IMM, MOD, R9, 123456789), + /* ALU32 immediate operations, register clobbering */ + BPF_TEST_CLOBBER_ALU(ALU32_IMM, AND, R8, 123456789), + BPF_TEST_CLOBBER_ALU(ALU32_IMM, AND, R9, 123456789), + BPF_TEST_CLOBBER_ALU(ALU32_IMM, OR, R8, 123456789), + BPF_TEST_CLOBBER_ALU(ALU32_IMM, OR, R9, 123456789), + BPF_TEST_CLOBBER_ALU(ALU32_IMM, XOR, R8, 123456789), + BPF_TEST_CLOBBER_ALU(ALU32_IMM, XOR, R9, 123456789), + BPF_TEST_CLOBBER_ALU(ALU32_IMM, LSH, R8, 12), + BPF_TEST_CLOBBER_ALU(ALU32_IMM, LSH, R9, 12), + BPF_TEST_CLOBBER_ALU(ALU32_IMM, RSH, R8, 12), + BPF_TEST_CLOBBER_ALU(ALU32_IMM, RSH, R9, 12), + BPF_TEST_CLOBBER_ALU(ALU32_IMM, ARSH, R8, 12), + BPF_TEST_CLOBBER_ALU(ALU32_IMM, ARSH, R9, 12), + BPF_TEST_CLOBBER_ALU(ALU32_IMM, ADD, R8, 123456789), + BPF_TEST_CLOBBER_ALU(ALU32_IMM, ADD, R9, 123456789), + BPF_TEST_CLOBBER_ALU(ALU32_IMM, SUB, R8, 123456789), + BPF_TEST_CLOBBER_ALU(ALU32_IMM, SUB, R9, 123456789), + BPF_TEST_CLOBBER_ALU(ALU32_IMM, MUL, R8, 123456789), + BPF_TEST_CLOBBER_ALU(ALU32_IMM, MUL, R9, 123456789), + BPF_TEST_CLOBBER_ALU(ALU32_IMM, DIV, R8, 123456789), + BPF_TEST_CLOBBER_ALU(ALU32_IMM, DIV, R9, 123456789), + BPF_TEST_CLOBBER_ALU(ALU32_IMM, MOD, R8, 123456789), + BPF_TEST_CLOBBER_ALU(ALU32_IMM, MOD, R9, 123456789), + /* ALU64 register operations, register clobbering */ + BPF_TEST_CLOBBER_ALU(ALU64_REG, AND, R8, R1), + BPF_TEST_CLOBBER_ALU(ALU64_REG, AND, R9, R1), + BPF_TEST_CLOBBER_ALU(ALU64_REG, OR, R8, R1), + BPF_TEST_CLOBBER_ALU(ALU64_REG, OR, R9, R1), + BPF_TEST_CLOBBER_ALU(ALU64_REG, XOR, R8, R1), + BPF_TEST_CLOBBER_ALU(ALU64_REG, XOR, R9, R1), + BPF_TEST_CLOBBER_ALU(ALU64_REG, LSH, R8, R1), + BPF_TEST_CLOBBER_ALU(ALU64_REG, LSH, R9, R1), + BPF_TEST_CLOBBER_ALU(ALU64_REG, RSH, R8, R1), + BPF_TEST_CLOBBER_ALU(ALU64_REG, RSH, R9, R1), + BPF_TEST_CLOBBER_ALU(ALU64_REG, ARSH, R8, R1), + BPF_TEST_CLOBBER_ALU(ALU64_REG, ARSH, R9, R1), + BPF_TEST_CLOBBER_ALU(ALU64_REG, ADD, R8, R1), + BPF_TEST_CLOBBER_ALU(ALU64_REG, ADD, R9, R1), + BPF_TEST_CLOBBER_ALU(ALU64_REG, SUB, R8, R1), + BPF_TEST_CLOBBER_ALU(ALU64_REG, SUB, R9, R1), + BPF_TEST_CLOBBER_ALU(ALU64_REG, MUL, R8, R1), + BPF_TEST_CLOBBER_ALU(ALU64_REG, MUL, R9, R1), + BPF_TEST_CLOBBER_ALU(ALU64_REG, DIV, R8, R1), + BPF_TEST_CLOBBER_ALU(ALU64_REG, DIV, R9, R1), + BPF_TEST_CLOBBER_ALU(ALU64_REG, MOD, R8, R1), + BPF_TEST_CLOBBER_ALU(ALU64_REG, MOD, R9, R1), + /* ALU32 register operations, register clobbering */ + BPF_TEST_CLOBBER_ALU(ALU32_REG, AND, R8, R1), + BPF_TEST_CLOBBER_ALU(ALU32_REG, AND, R9, R1), + BPF_TEST_CLOBBER_ALU(ALU32_REG, OR, R8, R1), + BPF_TEST_CLOBBER_ALU(ALU32_REG, OR, R9, R1), + BPF_TEST_CLOBBER_ALU(ALU32_REG, XOR, R8, R1), + BPF_TEST_CLOBBER_ALU(ALU32_REG, XOR, R9, R1), + BPF_TEST_CLOBBER_ALU(ALU32_REG, LSH, R8, R1), + BPF_TEST_CLOBBER_ALU(ALU32_REG, LSH, R9, R1), + BPF_TEST_CLOBBER_ALU(ALU32_REG, RSH, R8, R1), + BPF_TEST_CLOBBER_ALU(ALU32_REG, RSH, R9, R1), + BPF_TEST_CLOBBER_ALU(ALU32_REG, ARSH, R8, R1), + BPF_TEST_CLOBBER_ALU(ALU32_REG, ARSH, R9, R1), + BPF_TEST_CLOBBER_ALU(ALU32_REG, ADD, R8, R1), + BPF_TEST_CLOBBER_ALU(ALU32_REG, ADD, R9, R1), + BPF_TEST_CLOBBER_ALU(ALU32_REG, SUB, R8, R1), + BPF_TEST_CLOBBER_ALU(ALU32_REG, SUB, R9, R1), + BPF_TEST_CLOBBER_ALU(ALU32_REG, MUL, R8, R1), + BPF_TEST_CLOBBER_ALU(ALU32_REG, MUL, R9, R1), + BPF_TEST_CLOBBER_ALU(ALU32_REG, DIV, R8, R1), + BPF_TEST_CLOBBER_ALU(ALU32_REG, DIV, R9, R1), + BPF_TEST_CLOBBER_ALU(ALU32_REG, MOD, R8, R1), + BPF_TEST_CLOBBER_ALU(ALU32_REG, MOD, R9, R1), +#undef BPF_TEST_CLOBBER_ALU +#define BPF_TEST_CLOBBER_ATOMIC(width, op) \ + { \ + "Atomic_" #width " " #op ": no clobbering", \ + .u.insns_int = { \ + BPF_ALU64_IMM(BPF_MOV, R0, 0), \ + BPF_ALU64_IMM(BPF_MOV, R1, 1), \ + BPF_ALU64_IMM(BPF_MOV, R2, 2), \ + BPF_ALU64_IMM(BPF_MOV, R3, 3), \ + BPF_ALU64_IMM(BPF_MOV, R4, 4), \ + BPF_ALU64_IMM(BPF_MOV, R5, 5), \ + BPF_ALU64_IMM(BPF_MOV, R6, 6), \ + BPF_ALU64_IMM(BPF_MOV, R7, 7), \ + BPF_ALU64_IMM(BPF_MOV, R8, 8), \ + BPF_ALU64_IMM(BPF_MOV, R9, 9), \ + BPF_ST_MEM(width, R10, -8, \ + (op) == BPF_CMPXCHG ? 0 : \ + (op) & BPF_FETCH ? 1 : 0), \ + BPF_ATOMIC_OP(width, op, R10, R1, -8), \ + BPF_JMP_IMM(BPF_JNE, R0, 0, 10), \ + BPF_JMP_IMM(BPF_JNE, R1, 1, 9), \ + BPF_JMP_IMM(BPF_JNE, R2, 2, 8), \ + BPF_JMP_IMM(BPF_JNE, R3, 3, 7), \ + BPF_JMP_IMM(BPF_JNE, R4, 4, 6), \ + BPF_JMP_IMM(BPF_JNE, R5, 5, 5), \ + BPF_JMP_IMM(BPF_JNE, R6, 6, 4), \ + BPF_JMP_IMM(BPF_JNE, R7, 7, 3), \ + BPF_JMP_IMM(BPF_JNE, R8, 8, 2), \ + BPF_JMP_IMM(BPF_JNE, R9, 9, 1), \ + BPF_ALU64_IMM(BPF_MOV, R0, 1), \ + BPF_EXIT_INSN(), \ + }, \ + INTERNAL, \ + { }, \ + { { 0, 1 } }, \ + .stack_depth = 8, \ + } + /* 64-bit atomic operations, register clobbering */ + BPF_TEST_CLOBBER_ATOMIC(BPF_DW, BPF_ADD), + BPF_TEST_CLOBBER_ATOMIC(BPF_DW, BPF_AND), + BPF_TEST_CLOBBER_ATOMIC(BPF_DW, BPF_OR), + BPF_TEST_CLOBBER_ATOMIC(BPF_DW, BPF_XOR), + BPF_TEST_CLOBBER_ATOMIC(BPF_DW, BPF_ADD | BPF_FETCH), + BPF_TEST_CLOBBER_ATOMIC(BPF_DW, BPF_AND | BPF_FETCH), + BPF_TEST_CLOBBER_ATOMIC(BPF_DW, BPF_OR | BPF_FETCH), + BPF_TEST_CLOBBER_ATOMIC(BPF_DW, BPF_XOR | BPF_FETCH), + BPF_TEST_CLOBBER_ATOMIC(BPF_DW, BPF_XCHG), + BPF_TEST_CLOBBER_ATOMIC(BPF_DW, BPF_CMPXCHG), + /* 32-bit atomic operations, register clobbering */ + BPF_TEST_CLOBBER_ATOMIC(BPF_W, BPF_ADD), + BPF_TEST_CLOBBER_ATOMIC(BPF_W, BPF_AND), + BPF_TEST_CLOBBER_ATOMIC(BPF_W, BPF_OR), + BPF_TEST_CLOBBER_ATOMIC(BPF_W, BPF_XOR), + BPF_TEST_CLOBBER_ATOMIC(BPF_W, BPF_ADD | BPF_FETCH), + BPF_TEST_CLOBBER_ATOMIC(BPF_W, BPF_AND | BPF_FETCH), + BPF_TEST_CLOBBER_ATOMIC(BPF_W, BPF_OR | BPF_FETCH), + BPF_TEST_CLOBBER_ATOMIC(BPF_W, BPF_XOR | BPF_FETCH), + BPF_TEST_CLOBBER_ATOMIC(BPF_W, BPF_XCHG), + BPF_TEST_CLOBBER_ATOMIC(BPF_W, BPF_CMPXCHG), +#undef BPF_TEST_CLOBBER_ATOMIC /* Checking that ALU32 src is not zero extended in place */ #define BPF_ALU32_SRC_ZEXT(op) \ { \ -- cgit v1.2.3 From e42fc3c2c40e0fb9d371c146dc8c0a70bee88a3c Mon Sep 17 00:00:00 2001 From: Johan Almbladh Date: Fri, 1 Oct 2021 15:03:44 +0200 Subject: bpf/tests: Minor restructuring of ALU tests This patch moves the ALU LSH/RSH/ARSH reference computations into the common reference value function. Also fix typo in constants so they now have the intended values. Signed-off-by: Johan Almbladh Signed-off-by: Daniel Borkmann Link: https://lore.kernel.org/bpf/20211001130348.3670534-7-johan.almbladh@anyfinetworks.com --- lib/test_bpf.c | 137 +++++++++++++++++++++++++++------------------------------ 1 file changed, 65 insertions(+), 72 deletions(-) (limited to 'lib') diff --git a/lib/test_bpf.c b/lib/test_bpf.c index 201f34060eef..919323a3b69f 100644 --- a/lib/test_bpf.c +++ b/lib/test_bpf.c @@ -538,6 +538,57 @@ static int bpf_fill_max_jmp_never_taken(struct bpf_test *self) return __bpf_fill_max_jmp(self, BPF_JLT, 0); } +/* ALU result computation used in tests */ +static bool __bpf_alu_result(u64 *res, u64 v1, u64 v2, u8 op) +{ + *res = 0; + switch (op) { + case BPF_MOV: + *res = v2; + break; + case BPF_AND: + *res = v1 & v2; + break; + case BPF_OR: + *res = v1 | v2; + break; + case BPF_XOR: + *res = v1 ^ v2; + break; + case BPF_LSH: + *res = v1 << v2; + break; + case BPF_RSH: + *res = v1 >> v2; + break; + case BPF_ARSH: + *res = v1 >> v2; + if (v2 > 0 && v1 > S64_MAX) + *res |= ~0ULL << (64 - v2); + break; + case BPF_ADD: + *res = v1 + v2; + break; + case BPF_SUB: + *res = v1 - v2; + break; + case BPF_MUL: + *res = v1 * v2; + break; + case BPF_DIV: + if (v2 == 0) + return false; + *res = div64_u64(v1, v2); + break; + case BPF_MOD: + if (v2 == 0) + return false; + div64_u64_rem(v1, v2, res); + break; + } + return true; +} + /* Test an ALU shift operation for all valid shift values */ static int __bpf_fill_alu_shift(struct bpf_test *self, u8 op, u8 mode, bool alu32) @@ -576,37 +627,19 @@ static int __bpf_fill_alu_shift(struct bpf_test *self, u8 op, insn[i++] = BPF_ALU32_IMM(op, R1, imm); else insn[i++] = BPF_ALU32_REG(op, R1, R2); - switch (op) { - case BPF_LSH: - val = (u32)reg << imm; - break; - case BPF_RSH: - val = (u32)reg >> imm; - break; - case BPF_ARSH: - val = (u32)reg >> imm; - if (imm > 0 && (reg & 0x80000000)) - val |= ~(u32)0 << (32 - imm); - break; - } + + if (op == BPF_ARSH) + reg = (s32)reg; + else + reg = (u32)reg; + __bpf_alu_result(&val, reg, imm, op); + val = (u32)val; } else { if (mode == BPF_K) insn[i++] = BPF_ALU64_IMM(op, R1, imm); else insn[i++] = BPF_ALU64_REG(op, R1, R2); - switch (op) { - case BPF_LSH: - val = (u64)reg << imm; - break; - case BPF_RSH: - val = (u64)reg >> imm; - break; - case BPF_ARSH: - val = (u64)reg >> imm; - if (imm > 0 && reg < 0) - val |= ~(u64)0 << (64 - imm); - break; - } + __bpf_alu_result(&val, reg, imm, op); } /* @@ -799,46 +832,6 @@ static int __bpf_fill_pattern(struct bpf_test *self, void *arg, * test is designed to verify e.g. the ALU and ALU64 operations for JITs that * emit different code depending on the magnitude of the immediate value. */ - -static bool __bpf_alu_result(u64 *res, u64 v1, u64 v2, u8 op) -{ - *res = 0; - switch (op) { - case BPF_MOV: - *res = v2; - break; - case BPF_AND: - *res = v1 & v2; - break; - case BPF_OR: - *res = v1 | v2; - break; - case BPF_XOR: - *res = v1 ^ v2; - break; - case BPF_ADD: - *res = v1 + v2; - break; - case BPF_SUB: - *res = v1 - v2; - break; - case BPF_MUL: - *res = v1 * v2; - break; - case BPF_DIV: - if (v2 == 0) - return false; - *res = div64_u64(v1, v2); - break; - case BPF_MOD: - if (v2 == 0) - return false; - div64_u64_rem(v1, v2, res); - break; - } - return true; -} - static int __bpf_emit_alu64_imm(struct bpf_test *self, void *arg, struct bpf_insn *insns, s64 dst, s64 imm) { @@ -7881,7 +7874,7 @@ static struct bpf_test tests[] = { "BPF_ATOMIC | BPF_DW, BPF_CMPXCHG: Test successful return", .u.insns_int = { BPF_LD_IMM64(R1, 0x0123456789abcdefULL), - BPF_LD_IMM64(R2, 0xfecdba9876543210ULL), + BPF_LD_IMM64(R2, 0xfedcba9876543210ULL), BPF_ALU64_REG(BPF_MOV, R0, R1), BPF_STX_MEM(BPF_DW, R10, R1, -40), BPF_ATOMIC_OP(BPF_DW, BPF_CMPXCHG, R10, R2, -40), @@ -7898,7 +7891,7 @@ static struct bpf_test tests[] = { "BPF_ATOMIC | BPF_DW, BPF_CMPXCHG: Test successful store", .u.insns_int = { BPF_LD_IMM64(R1, 0x0123456789abcdefULL), - BPF_LD_IMM64(R2, 0xfecdba9876543210ULL), + BPF_LD_IMM64(R2, 0xfedcba9876543210ULL), BPF_ALU64_REG(BPF_MOV, R0, R1), BPF_STX_MEM(BPF_DW, R10, R0, -40), BPF_ATOMIC_OP(BPF_DW, BPF_CMPXCHG, R10, R2, -40), @@ -7916,7 +7909,7 @@ static struct bpf_test tests[] = { "BPF_ATOMIC | BPF_DW, BPF_CMPXCHG: Test failure return", .u.insns_int = { BPF_LD_IMM64(R1, 0x0123456789abcdefULL), - BPF_LD_IMM64(R2, 0xfecdba9876543210ULL), + BPF_LD_IMM64(R2, 0xfedcba9876543210ULL), BPF_ALU64_REG(BPF_MOV, R0, R1), BPF_ALU64_IMM(BPF_ADD, R0, 1), BPF_STX_MEM(BPF_DW, R10, R1, -40), @@ -7934,7 +7927,7 @@ static struct bpf_test tests[] = { "BPF_ATOMIC | BPF_DW, BPF_CMPXCHG: Test failure store", .u.insns_int = { BPF_LD_IMM64(R1, 0x0123456789abcdefULL), - BPF_LD_IMM64(R2, 0xfecdba9876543210ULL), + BPF_LD_IMM64(R2, 0xfedcba9876543210ULL), BPF_ALU64_REG(BPF_MOV, R0, R1), BPF_ALU64_IMM(BPF_ADD, R0, 1), BPF_STX_MEM(BPF_DW, R10, R1, -40), @@ -7953,11 +7946,11 @@ static struct bpf_test tests[] = { "BPF_ATOMIC | BPF_DW, BPF_CMPXCHG: Test side effects", .u.insns_int = { BPF_LD_IMM64(R1, 0x0123456789abcdefULL), - BPF_LD_IMM64(R2, 0xfecdba9876543210ULL), + BPF_LD_IMM64(R2, 0xfedcba9876543210ULL), BPF_ALU64_REG(BPF_MOV, R0, R1), BPF_STX_MEM(BPF_DW, R10, R1, -40), BPF_ATOMIC_OP(BPF_DW, BPF_CMPXCHG, R10, R2, -40), - BPF_LD_IMM64(R0, 0xfecdba9876543210ULL), + BPF_LD_IMM64(R0, 0xfedcba9876543210ULL), BPF_JMP_REG(BPF_JNE, R0, R2, 1), BPF_ALU64_REG(BPF_SUB, R0, R2), BPF_EXIT_INSN(), -- cgit v1.2.3 From daed6083f4fbcbb57da26d80f15365219ae793de Mon Sep 17 00:00:00 2001 From: Johan Almbladh Date: Fri, 1 Oct 2021 15:03:45 +0200 Subject: bpf/tests: Add exhaustive tests of ALU register combinations This patch replaces the current register combination test with new exhaustive tests. Before, only a subset of register combinations was tested for ALU64 DIV. Now, all combinatons of operand registers are tested, including the case when they are the same, and for all ALU32 and ALU64 operations. Signed-off-by: Johan Almbladh Signed-off-by: Daniel Borkmann Link: https://lore.kernel.org/bpf/20211001130348.3670534-8-johan.almbladh@anyfinetworks.com --- lib/test_bpf.c | 834 ++++++++++++++++++++++++++++++++++++++++++++++++++++----- 1 file changed, 763 insertions(+), 71 deletions(-) (limited to 'lib') diff --git a/lib/test_bpf.c b/lib/test_bpf.c index 919323a3b69f..924bf4c9783c 100644 --- a/lib/test_bpf.c +++ b/lib/test_bpf.c @@ -1130,6 +1130,381 @@ static int bpf_fill_alu32_mod_reg(struct bpf_test *self) return __bpf_fill_alu32_reg(self, BPF_MOD); } +/* + * Test JITs that implement complex ALU operations as function + * calls, and must re-arrange operands for argument passing. + */ +static int __bpf_fill_alu_imm_regs(struct bpf_test *self, u8 op, bool alu32) +{ + int len = 2 + 10 * 10; + struct bpf_insn *insns; + u64 dst, res; + int i = 0; + u32 imm; + int rd; + + insns = kmalloc_array(len, sizeof(*insns), GFP_KERNEL); + if (!insns) + return -ENOMEM; + + /* Operand and result values according to operation */ + if (alu32) + dst = 0x76543210U; + else + dst = 0x7edcba9876543210ULL; + imm = 0x01234567U; + + if (op == BPF_LSH || op == BPF_RSH || op == BPF_ARSH) + imm &= 31; + + __bpf_alu_result(&res, dst, imm, op); + + if (alu32) + res = (u32)res; + + /* Check all operand registers */ + for (rd = R0; rd <= R9; rd++) { + i += __bpf_ld_imm64(&insns[i], rd, dst); + + if (alu32) + insns[i++] = BPF_ALU32_IMM(op, rd, imm); + else + insns[i++] = BPF_ALU64_IMM(op, rd, imm); + + insns[i++] = BPF_JMP32_IMM(BPF_JEQ, rd, res, 2); + insns[i++] = BPF_MOV64_IMM(R0, __LINE__); + insns[i++] = BPF_EXIT_INSN(); + + insns[i++] = BPF_ALU64_IMM(BPF_RSH, rd, 32); + insns[i++] = BPF_JMP32_IMM(BPF_JEQ, rd, res >> 32, 2); + insns[i++] = BPF_MOV64_IMM(R0, __LINE__); + insns[i++] = BPF_EXIT_INSN(); + } + + insns[i++] = BPF_MOV64_IMM(R0, 1); + insns[i++] = BPF_EXIT_INSN(); + + self->u.ptr.insns = insns; + self->u.ptr.len = len; + BUG_ON(i != len); + + return 0; +} + +/* ALU64 K registers */ +static int bpf_fill_alu64_mov_imm_regs(struct bpf_test *self) +{ + return __bpf_fill_alu_imm_regs(self, BPF_MOV, false); +} + +static int bpf_fill_alu64_and_imm_regs(struct bpf_test *self) +{ + return __bpf_fill_alu_imm_regs(self, BPF_AND, false); +} + +static int bpf_fill_alu64_or_imm_regs(struct bpf_test *self) +{ + return __bpf_fill_alu_imm_regs(self, BPF_OR, false); +} + +static int bpf_fill_alu64_xor_imm_regs(struct bpf_test *self) +{ + return __bpf_fill_alu_imm_regs(self, BPF_XOR, false); +} + +static int bpf_fill_alu64_lsh_imm_regs(struct bpf_test *self) +{ + return __bpf_fill_alu_imm_regs(self, BPF_LSH, false); +} + +static int bpf_fill_alu64_rsh_imm_regs(struct bpf_test *self) +{ + return __bpf_fill_alu_imm_regs(self, BPF_RSH, false); +} + +static int bpf_fill_alu64_arsh_imm_regs(struct bpf_test *self) +{ + return __bpf_fill_alu_imm_regs(self, BPF_ARSH, false); +} + +static int bpf_fill_alu64_add_imm_regs(struct bpf_test *self) +{ + return __bpf_fill_alu_imm_regs(self, BPF_ADD, false); +} + +static int bpf_fill_alu64_sub_imm_regs(struct bpf_test *self) +{ + return __bpf_fill_alu_imm_regs(self, BPF_SUB, false); +} + +static int bpf_fill_alu64_mul_imm_regs(struct bpf_test *self) +{ + return __bpf_fill_alu_imm_regs(self, BPF_MUL, false); +} + +static int bpf_fill_alu64_div_imm_regs(struct bpf_test *self) +{ + return __bpf_fill_alu_imm_regs(self, BPF_DIV, false); +} + +static int bpf_fill_alu64_mod_imm_regs(struct bpf_test *self) +{ + return __bpf_fill_alu_imm_regs(self, BPF_MOD, false); +} + +/* ALU32 K registers */ +static int bpf_fill_alu32_mov_imm_regs(struct bpf_test *self) +{ + return __bpf_fill_alu_imm_regs(self, BPF_MOV, true); +} + +static int bpf_fill_alu32_and_imm_regs(struct bpf_test *self) +{ + return __bpf_fill_alu_imm_regs(self, BPF_AND, true); +} + +static int bpf_fill_alu32_or_imm_regs(struct bpf_test *self) +{ + return __bpf_fill_alu_imm_regs(self, BPF_OR, true); +} + +static int bpf_fill_alu32_xor_imm_regs(struct bpf_test *self) +{ + return __bpf_fill_alu_imm_regs(self, BPF_XOR, true); +} + +static int bpf_fill_alu32_lsh_imm_regs(struct bpf_test *self) +{ + return __bpf_fill_alu_imm_regs(self, BPF_LSH, true); +} + +static int bpf_fill_alu32_rsh_imm_regs(struct bpf_test *self) +{ + return __bpf_fill_alu_imm_regs(self, BPF_RSH, true); +} + +static int bpf_fill_alu32_arsh_imm_regs(struct bpf_test *self) +{ + return __bpf_fill_alu_imm_regs(self, BPF_ARSH, true); +} + +static int bpf_fill_alu32_add_imm_regs(struct bpf_test *self) +{ + return __bpf_fill_alu_imm_regs(self, BPF_ADD, true); +} + +static int bpf_fill_alu32_sub_imm_regs(struct bpf_test *self) +{ + return __bpf_fill_alu_imm_regs(self, BPF_SUB, true); +} + +static int bpf_fill_alu32_mul_imm_regs(struct bpf_test *self) +{ + return __bpf_fill_alu_imm_regs(self, BPF_MUL, true); +} + +static int bpf_fill_alu32_div_imm_regs(struct bpf_test *self) +{ + return __bpf_fill_alu_imm_regs(self, BPF_DIV, true); +} + +static int bpf_fill_alu32_mod_imm_regs(struct bpf_test *self) +{ + return __bpf_fill_alu_imm_regs(self, BPF_MOD, true); +} + +/* + * Test JITs that implement complex ALU operations as function + * calls, and must re-arrange operands for argument passing. + */ +static int __bpf_fill_alu_reg_pairs(struct bpf_test *self, u8 op, bool alu32) +{ + int len = 2 + 10 * 10 * 12; + u64 dst, src, res, same; + struct bpf_insn *insns; + int rd, rs; + int i = 0; + + insns = kmalloc_array(len, sizeof(*insns), GFP_KERNEL); + if (!insns) + return -ENOMEM; + + /* Operand and result values according to operation */ + if (alu32) { + dst = 0x76543210U; + src = 0x01234567U; + } else { + dst = 0x7edcba9876543210ULL; + src = 0x0123456789abcdefULL; + } + + if (op == BPF_LSH || op == BPF_RSH || op == BPF_ARSH) + src &= 31; + + __bpf_alu_result(&res, dst, src, op); + __bpf_alu_result(&same, src, src, op); + + if (alu32) { + res = (u32)res; + same = (u32)same; + } + + /* Check all combinations of operand registers */ + for (rd = R0; rd <= R9; rd++) { + for (rs = R0; rs <= R9; rs++) { + u64 val = rd == rs ? same : res; + + i += __bpf_ld_imm64(&insns[i], rd, dst); + i += __bpf_ld_imm64(&insns[i], rs, src); + + if (alu32) + insns[i++] = BPF_ALU32_REG(op, rd, rs); + else + insns[i++] = BPF_ALU64_REG(op, rd, rs); + + insns[i++] = BPF_JMP32_IMM(BPF_JEQ, rd, val, 2); + insns[i++] = BPF_MOV64_IMM(R0, __LINE__); + insns[i++] = BPF_EXIT_INSN(); + + insns[i++] = BPF_ALU64_IMM(BPF_RSH, rd, 32); + insns[i++] = BPF_JMP32_IMM(BPF_JEQ, rd, val >> 32, 2); + insns[i++] = BPF_MOV64_IMM(R0, __LINE__); + insns[i++] = BPF_EXIT_INSN(); + } + } + + insns[i++] = BPF_MOV64_IMM(R0, 1); + insns[i++] = BPF_EXIT_INSN(); + + self->u.ptr.insns = insns; + self->u.ptr.len = len; + BUG_ON(i != len); + + return 0; +} + +/* ALU64 X register combinations */ +static int bpf_fill_alu64_mov_reg_pairs(struct bpf_test *self) +{ + return __bpf_fill_alu_reg_pairs(self, BPF_MOV, false); +} + +static int bpf_fill_alu64_and_reg_pairs(struct bpf_test *self) +{ + return __bpf_fill_alu_reg_pairs(self, BPF_AND, false); +} + +static int bpf_fill_alu64_or_reg_pairs(struct bpf_test *self) +{ + return __bpf_fill_alu_reg_pairs(self, BPF_OR, false); +} + +static int bpf_fill_alu64_xor_reg_pairs(struct bpf_test *self) +{ + return __bpf_fill_alu_reg_pairs(self, BPF_XOR, false); +} + +static int bpf_fill_alu64_lsh_reg_pairs(struct bpf_test *self) +{ + return __bpf_fill_alu_reg_pairs(self, BPF_LSH, false); +} + +static int bpf_fill_alu64_rsh_reg_pairs(struct bpf_test *self) +{ + return __bpf_fill_alu_reg_pairs(self, BPF_RSH, false); +} + +static int bpf_fill_alu64_arsh_reg_pairs(struct bpf_test *self) +{ + return __bpf_fill_alu_reg_pairs(self, BPF_ARSH, false); +} + +static int bpf_fill_alu64_add_reg_pairs(struct bpf_test *self) +{ + return __bpf_fill_alu_reg_pairs(self, BPF_ADD, false); +} + +static int bpf_fill_alu64_sub_reg_pairs(struct bpf_test *self) +{ + return __bpf_fill_alu_reg_pairs(self, BPF_SUB, false); +} + +static int bpf_fill_alu64_mul_reg_pairs(struct bpf_test *self) +{ + return __bpf_fill_alu_reg_pairs(self, BPF_MUL, false); +} + +static int bpf_fill_alu64_div_reg_pairs(struct bpf_test *self) +{ + return __bpf_fill_alu_reg_pairs(self, BPF_DIV, false); +} + +static int bpf_fill_alu64_mod_reg_pairs(struct bpf_test *self) +{ + return __bpf_fill_alu_reg_pairs(self, BPF_MOD, false); +} + +/* ALU32 X register combinations */ +static int bpf_fill_alu32_mov_reg_pairs(struct bpf_test *self) +{ + return __bpf_fill_alu_reg_pairs(self, BPF_MOV, true); +} + +static int bpf_fill_alu32_and_reg_pairs(struct bpf_test *self) +{ + return __bpf_fill_alu_reg_pairs(self, BPF_AND, true); +} + +static int bpf_fill_alu32_or_reg_pairs(struct bpf_test *self) +{ + return __bpf_fill_alu_reg_pairs(self, BPF_OR, true); +} + +static int bpf_fill_alu32_xor_reg_pairs(struct bpf_test *self) +{ + return __bpf_fill_alu_reg_pairs(self, BPF_XOR, true); +} + +static int bpf_fill_alu32_lsh_reg_pairs(struct bpf_test *self) +{ + return __bpf_fill_alu_reg_pairs(self, BPF_LSH, true); +} + +static int bpf_fill_alu32_rsh_reg_pairs(struct bpf_test *self) +{ + return __bpf_fill_alu_reg_pairs(self, BPF_RSH, true); +} + +static int bpf_fill_alu32_arsh_reg_pairs(struct bpf_test *self) +{ + return __bpf_fill_alu_reg_pairs(self, BPF_ARSH, true); +} + +static int bpf_fill_alu32_add_reg_pairs(struct bpf_test *self) +{ + return __bpf_fill_alu_reg_pairs(self, BPF_ADD, true); +} + +static int bpf_fill_alu32_sub_reg_pairs(struct bpf_test *self) +{ + return __bpf_fill_alu_reg_pairs(self, BPF_SUB, true); +} + +static int bpf_fill_alu32_mul_reg_pairs(struct bpf_test *self) +{ + return __bpf_fill_alu_reg_pairs(self, BPF_MUL, true); +} + +static int bpf_fill_alu32_div_reg_pairs(struct bpf_test *self) +{ + return __bpf_fill_alu_reg_pairs(self, BPF_DIV, true); +} + +static int bpf_fill_alu32_mod_reg_pairs(struct bpf_test *self) +{ + return __bpf_fill_alu_reg_pairs(self, BPF_MOD, true); +} + /* * Exhaustive tests of atomic operations for all power-of-two operand * magnitudes, both for positive and negative values. @@ -3737,77 +4112,6 @@ static struct bpf_test tests[] = { { }, { { 0, -1 } } }, - { - /* - * Test 32-bit JITs that implement complex ALU64 operations as - * function calls R0 = f(R1, R2), and must re-arrange operands. - */ -#define NUMER 0xfedcba9876543210ULL -#define DENOM 0x0123456789abcdefULL - "ALU64_DIV X: Operand register permutations", - .u.insns_int = { - /* R0 / R2 */ - BPF_LD_IMM64(R0, NUMER), - BPF_LD_IMM64(R2, DENOM), - BPF_ALU64_REG(BPF_DIV, R0, R2), - BPF_JMP_IMM(BPF_JEQ, R0, NUMER / DENOM, 1), - BPF_EXIT_INSN(), - /* R1 / R0 */ - BPF_LD_IMM64(R1, NUMER), - BPF_LD_IMM64(R0, DENOM), - BPF_ALU64_REG(BPF_DIV, R1, R0), - BPF_JMP_IMM(BPF_JEQ, R1, NUMER / DENOM, 1), - BPF_EXIT_INSN(), - /* R0 / R1 */ - BPF_LD_IMM64(R0, NUMER), - BPF_LD_IMM64(R1, DENOM), - BPF_ALU64_REG(BPF_DIV, R0, R1), - BPF_JMP_IMM(BPF_JEQ, R0, NUMER / DENOM, 1), - BPF_EXIT_INSN(), - /* R2 / R0 */ - BPF_LD_IMM64(R2, NUMER), - BPF_LD_IMM64(R0, DENOM), - BPF_ALU64_REG(BPF_DIV, R2, R0), - BPF_JMP_IMM(BPF_JEQ, R2, NUMER / DENOM, 1), - BPF_EXIT_INSN(), - /* R2 / R1 */ - BPF_LD_IMM64(R2, NUMER), - BPF_LD_IMM64(R1, DENOM), - BPF_ALU64_REG(BPF_DIV, R2, R1), - BPF_JMP_IMM(BPF_JEQ, R2, NUMER / DENOM, 1), - BPF_EXIT_INSN(), - /* R1 / R2 */ - BPF_LD_IMM64(R1, NUMER), - BPF_LD_IMM64(R2, DENOM), - BPF_ALU64_REG(BPF_DIV, R1, R2), - BPF_JMP_IMM(BPF_JEQ, R1, NUMER / DENOM, 1), - BPF_EXIT_INSN(), - /* R1 / R1 */ - BPF_LD_IMM64(R1, NUMER), - BPF_ALU64_REG(BPF_DIV, R1, R1), - BPF_JMP_IMM(BPF_JEQ, R1, 1, 1), - BPF_EXIT_INSN(), - /* R2 / R2 */ - BPF_LD_IMM64(R2, DENOM), - BPF_ALU64_REG(BPF_DIV, R2, R2), - BPF_JMP_IMM(BPF_JEQ, R2, 1, 1), - BPF_EXIT_INSN(), - /* R3 / R4 */ - BPF_LD_IMM64(R3, NUMER), - BPF_LD_IMM64(R4, DENOM), - BPF_ALU64_REG(BPF_DIV, R3, R4), - BPF_JMP_IMM(BPF_JEQ, R3, NUMER / DENOM, 1), - BPF_EXIT_INSN(), - /* Successful return */ - BPF_LD_IMM64(R0, 1), - BPF_EXIT_INSN(), - }, - INTERNAL, - { }, - { { 0, 1 } }, -#undef NUMER -#undef DENOM - }, #ifdef CONFIG_32BIT { "INT: 32-bit context pointer word order and zero-extension", @@ -10849,6 +11153,394 @@ static struct bpf_test tests[] = { BPF_JMP32_REG_ZEXT(JSLT), BPF_JMP32_REG_ZEXT(JSLE), #undef BPF_JMP2_REG_ZEXT + /* ALU64 K register combinations */ + { + "ALU64_MOV_K: registers", + { }, + INTERNAL, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu64_mov_imm_regs, + }, + { + "ALU64_AND_K: registers", + { }, + INTERNAL, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu64_and_imm_regs, + }, + { + "ALU64_OR_K: registers", + { }, + INTERNAL, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu64_or_imm_regs, + }, + { + "ALU64_XOR_K: registers", + { }, + INTERNAL, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu64_xor_imm_regs, + }, + { + "ALU64_LSH_K: registers", + { }, + INTERNAL, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu64_lsh_imm_regs, + }, + { + "ALU64_RSH_K: registers", + { }, + INTERNAL, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu64_rsh_imm_regs, + }, + { + "ALU64_ARSH_K: registers", + { }, + INTERNAL, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu64_arsh_imm_regs, + }, + { + "ALU64_ADD_K: registers", + { }, + INTERNAL, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu64_add_imm_regs, + }, + { + "ALU64_SUB_K: registers", + { }, + INTERNAL, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu64_sub_imm_regs, + }, + { + "ALU64_MUL_K: registers", + { }, + INTERNAL, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu64_mul_imm_regs, + }, + { + "ALU64_DIV_K: registers", + { }, + INTERNAL, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu64_div_imm_regs, + }, + { + "ALU64_MOD_K: registers", + { }, + INTERNAL, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu64_mod_imm_regs, + }, + /* ALU32 K registers */ + { + "ALU32_MOV_K: registers", + { }, + INTERNAL, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu32_mov_imm_regs, + }, + { + "ALU32_AND_K: registers", + { }, + INTERNAL, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu32_and_imm_regs, + }, + { + "ALU32_OR_K: registers", + { }, + INTERNAL, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu32_or_imm_regs, + }, + { + "ALU32_XOR_K: registers", + { }, + INTERNAL, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu32_xor_imm_regs, + }, + { + "ALU32_LSH_K: registers", + { }, + INTERNAL, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu32_lsh_imm_regs, + }, + { + "ALU32_RSH_K: registers", + { }, + INTERNAL, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu32_rsh_imm_regs, + }, + { + "ALU32_ARSH_K: registers", + { }, + INTERNAL, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu32_arsh_imm_regs, + }, + { + "ALU32_ADD_K: registers", + { }, + INTERNAL, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu32_add_imm_regs, + }, + { + "ALU32_SUB_K: registers", + { }, + INTERNAL, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu32_sub_imm_regs, + }, + { + "ALU32_MUL_K: registers", + { }, + INTERNAL, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu32_mul_imm_regs, + }, + { + "ALU32_DIV_K: registers", + { }, + INTERNAL, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu32_div_imm_regs, + }, + { + "ALU32_MOD_K: registers", + { }, + INTERNAL, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu32_mod_imm_regs, + }, + /* ALU64 X register combinations */ + { + "ALU64_MOV_X: register combinations", + { }, + INTERNAL, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu64_mov_reg_pairs, + }, + { + "ALU64_AND_X: register combinations", + { }, + INTERNAL, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu64_and_reg_pairs, + }, + { + "ALU64_OR_X: register combinations", + { }, + INTERNAL, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu64_or_reg_pairs, + }, + { + "ALU64_XOR_X: register combinations", + { }, + INTERNAL, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu64_xor_reg_pairs, + }, + { + "ALU64_LSH_X: register combinations", + { }, + INTERNAL, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu64_lsh_reg_pairs, + }, + { + "ALU64_RSH_X: register combinations", + { }, + INTERNAL, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu64_rsh_reg_pairs, + }, + { + "ALU64_ARSH_X: register combinations", + { }, + INTERNAL, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu64_arsh_reg_pairs, + }, + { + "ALU64_ADD_X: register combinations", + { }, + INTERNAL, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu64_add_reg_pairs, + }, + { + "ALU64_SUB_X: register combinations", + { }, + INTERNAL, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu64_sub_reg_pairs, + }, + { + "ALU64_MUL_X: register combinations", + { }, + INTERNAL, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu64_mul_reg_pairs, + }, + { + "ALU64_DIV_X: register combinations", + { }, + INTERNAL, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu64_div_reg_pairs, + }, + { + "ALU64_MOD_X: register combinations", + { }, + INTERNAL, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu64_mod_reg_pairs, + }, + /* ALU32 X register combinations */ + { + "ALU32_MOV_X: register combinations", + { }, + INTERNAL, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu32_mov_reg_pairs, + }, + { + "ALU32_AND_X: register combinations", + { }, + INTERNAL, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu32_and_reg_pairs, + }, + { + "ALU32_OR_X: register combinations", + { }, + INTERNAL, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu32_or_reg_pairs, + }, + { + "ALU32_XOR_X: register combinations", + { }, + INTERNAL, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu32_xor_reg_pairs, + }, + { + "ALU32_LSH_X: register combinations", + { }, + INTERNAL, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu32_lsh_reg_pairs, + }, + { + "ALU32_RSH_X: register combinations", + { }, + INTERNAL, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu32_rsh_reg_pairs, + }, + { + "ALU32_ARSH_X: register combinations", + { }, + INTERNAL, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu32_arsh_reg_pairs, + }, + { + "ALU32_ADD_X: register combinations", + { }, + INTERNAL, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu32_add_reg_pairs, + }, + { + "ALU32_SUB_X: register combinations", + { }, + INTERNAL, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu32_sub_reg_pairs, + }, + { + "ALU32_MUL_X: register combinations", + { }, + INTERNAL, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu32_mul_reg_pairs, + }, + { + "ALU32_DIV_X: register combinations", + { }, + INTERNAL, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu32_div_reg_pairs, + }, + { + "ALU32_MOD_X register combinations", + { }, + INTERNAL, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu32_mod_reg_pairs, + }, /* Exhaustive test of ALU64 shift operations */ { "ALU64_LSH_K: all shift values", -- cgit v1.2.3 From 6fae2e8a1d9ee09e25aee6514b7544e059a8ee68 Mon Sep 17 00:00:00 2001 From: Johan Almbladh Date: Fri, 1 Oct 2021 15:03:46 +0200 Subject: bpf/tests: Add exhaustive tests of BPF_ATOMIC register combinations This patch adds tests of all register combinations for BPF_ATOMIC operations on both BPF_W and BPF_DW sizes. Signed-off-by: Johan Almbladh Signed-off-by: Daniel Borkmann Link: https://lore.kernel.org/bpf/20211001130348.3670534-9-johan.almbladh@anyfinetworks.com --- lib/test_bpf.c | 422 +++++++++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 422 insertions(+) (limited to 'lib') diff --git a/lib/test_bpf.c b/lib/test_bpf.c index 924bf4c9783c..40db4cee4f51 100644 --- a/lib/test_bpf.c +++ b/lib/test_bpf.c @@ -1805,6 +1805,246 @@ static int bpf_fill_cmpxchg32(struct bpf_test *self) &__bpf_emit_cmpxchg32); } +/* + * Test JITs that implement ATOMIC operations as function calls or + * other primitives, and must re-arrange operands for argument passing. + */ +static int __bpf_fill_atomic_reg_pairs(struct bpf_test *self, u8 width, u8 op) +{ + struct bpf_insn *insn; + int len = 2 + 34 * 10 * 10; + u64 mem, upd, res; + int rd, rs, i = 0; + + insn = kmalloc_array(len, sizeof(*insn), GFP_KERNEL); + if (!insn) + return -ENOMEM; + + /* Operand and memory values */ + if (width == BPF_DW) { + mem = 0x0123456789abcdefULL; + upd = 0xfedcba9876543210ULL; + } else { /* BPF_W */ + mem = 0x01234567U; + upd = 0x76543210U; + } + + /* Memory updated according to operation */ + switch (op) { + case BPF_XCHG: + res = upd; + break; + case BPF_CMPXCHG: + res = mem; + break; + default: + __bpf_alu_result(&res, mem, upd, BPF_OP(op)); + } + + /* Test all operand registers */ + for (rd = R0; rd <= R9; rd++) { + for (rs = R0; rs <= R9; rs++) { + u64 cmp, src; + + /* Initialize value in memory */ + i += __bpf_ld_imm64(&insn[i], R0, mem); + insn[i++] = BPF_STX_MEM(width, R10, R0, -8); + + /* Initialize registers in order */ + i += __bpf_ld_imm64(&insn[i], R0, ~mem); + i += __bpf_ld_imm64(&insn[i], rs, upd); + insn[i++] = BPF_MOV64_REG(rd, R10); + + /* Perform atomic operation */ + insn[i++] = BPF_ATOMIC_OP(width, op, rd, rs, -8); + if (op == BPF_CMPXCHG && width == BPF_W) + insn[i++] = BPF_ZEXT_REG(R0); + + /* Check R0 register value */ + if (op == BPF_CMPXCHG) + cmp = mem; /* Expect value from memory */ + else if (R0 == rd || R0 == rs) + cmp = 0; /* Aliased, checked below */ + else + cmp = ~mem; /* Expect value to be preserved */ + if (cmp) { + insn[i++] = BPF_JMP32_IMM(BPF_JEQ, R0, + (u32)cmp, 2); + insn[i++] = BPF_MOV32_IMM(R0, __LINE__); + insn[i++] = BPF_EXIT_INSN(); + insn[i++] = BPF_ALU64_IMM(BPF_RSH, R0, 32); + insn[i++] = BPF_JMP32_IMM(BPF_JEQ, R0, + cmp >> 32, 2); + insn[i++] = BPF_MOV32_IMM(R0, __LINE__); + insn[i++] = BPF_EXIT_INSN(); + } + + /* Check source register value */ + if (rs == R0 && op == BPF_CMPXCHG) + src = 0; /* Aliased with R0, checked above */ + else if (rs == rd && (op == BPF_CMPXCHG || + !(op & BPF_FETCH))) + src = 0; /* Aliased with rd, checked below */ + else if (op == BPF_CMPXCHG) + src = upd; /* Expect value to be preserved */ + else if (op & BPF_FETCH) + src = mem; /* Expect fetched value from mem */ + else /* no fetch */ + src = upd; /* Expect value to be preserved */ + if (src) { + insn[i++] = BPF_JMP32_IMM(BPF_JEQ, rs, + (u32)src, 2); + insn[i++] = BPF_MOV32_IMM(R0, __LINE__); + insn[i++] = BPF_EXIT_INSN(); + insn[i++] = BPF_ALU64_IMM(BPF_RSH, rs, 32); + insn[i++] = BPF_JMP32_IMM(BPF_JEQ, rs, + src >> 32, 2); + insn[i++] = BPF_MOV32_IMM(R0, __LINE__); + insn[i++] = BPF_EXIT_INSN(); + } + + /* Check destination register value */ + if (!(rd == R0 && op == BPF_CMPXCHG) && + !(rd == rs && (op & BPF_FETCH))) { + insn[i++] = BPF_JMP_REG(BPF_JEQ, rd, R10, 2); + insn[i++] = BPF_MOV32_IMM(R0, __LINE__); + insn[i++] = BPF_EXIT_INSN(); + } + + /* Check value in memory */ + if (rs != rd) { /* No aliasing */ + i += __bpf_ld_imm64(&insn[i], R1, res); + } else if (op == BPF_XCHG) { /* Aliased, XCHG */ + insn[i++] = BPF_MOV64_REG(R1, R10); + } else if (op == BPF_CMPXCHG) { /* Aliased, CMPXCHG */ + i += __bpf_ld_imm64(&insn[i], R1, mem); + } else { /* Aliased, ALU oper */ + i += __bpf_ld_imm64(&insn[i], R1, mem); + insn[i++] = BPF_ALU64_REG(BPF_OP(op), R1, R10); + } + + insn[i++] = BPF_LDX_MEM(width, R0, R10, -8); + if (width == BPF_DW) + insn[i++] = BPF_JMP_REG(BPF_JEQ, R0, R1, 2); + else /* width == BPF_W */ + insn[i++] = BPF_JMP32_REG(BPF_JEQ, R0, R1, 2); + insn[i++] = BPF_MOV32_IMM(R0, __LINE__); + insn[i++] = BPF_EXIT_INSN(); + } + } + + insn[i++] = BPF_MOV64_IMM(R0, 1); + insn[i++] = BPF_EXIT_INSN(); + + self->u.ptr.insns = insn; + self->u.ptr.len = i; + BUG_ON(i > len); + + return 0; +} + +/* 64-bit atomic register tests */ +static int bpf_fill_atomic64_add_reg_pairs(struct bpf_test *self) +{ + return __bpf_fill_atomic_reg_pairs(self, BPF_DW, BPF_ADD); +} + +static int bpf_fill_atomic64_and_reg_pairs(struct bpf_test *self) +{ + return __bpf_fill_atomic_reg_pairs(self, BPF_DW, BPF_AND); +} + +static int bpf_fill_atomic64_or_reg_pairs(struct bpf_test *self) +{ + return __bpf_fill_atomic_reg_pairs(self, BPF_DW, BPF_OR); +} + +static int bpf_fill_atomic64_xor_reg_pairs(struct bpf_test *self) +{ + return __bpf_fill_atomic_reg_pairs(self, BPF_DW, BPF_XOR); +} + +static int bpf_fill_atomic64_add_fetch_reg_pairs(struct bpf_test *self) +{ + return __bpf_fill_atomic_reg_pairs(self, BPF_DW, BPF_ADD | BPF_FETCH); +} + +static int bpf_fill_atomic64_and_fetch_reg_pairs(struct bpf_test *self) +{ + return __bpf_fill_atomic_reg_pairs(self, BPF_DW, BPF_AND | BPF_FETCH); +} + +static int bpf_fill_atomic64_or_fetch_reg_pairs(struct bpf_test *self) +{ + return __bpf_fill_atomic_reg_pairs(self, BPF_DW, BPF_OR | BPF_FETCH); +} + +static int bpf_fill_atomic64_xor_fetch_reg_pairs(struct bpf_test *self) +{ + return __bpf_fill_atomic_reg_pairs(self, BPF_DW, BPF_XOR | BPF_FETCH); +} + +static int bpf_fill_atomic64_xchg_reg_pairs(struct bpf_test *self) +{ + return __bpf_fill_atomic_reg_pairs(self, BPF_DW, BPF_XCHG); +} + +static int bpf_fill_atomic64_cmpxchg_reg_pairs(struct bpf_test *self) +{ + return __bpf_fill_atomic_reg_pairs(self, BPF_DW, BPF_CMPXCHG); +} + +/* 32-bit atomic register tests */ +static int bpf_fill_atomic32_add_reg_pairs(struct bpf_test *self) +{ + return __bpf_fill_atomic_reg_pairs(self, BPF_W, BPF_ADD); +} + +static int bpf_fill_atomic32_and_reg_pairs(struct bpf_test *self) +{ + return __bpf_fill_atomic_reg_pairs(self, BPF_W, BPF_AND); +} + +static int bpf_fill_atomic32_or_reg_pairs(struct bpf_test *self) +{ + return __bpf_fill_atomic_reg_pairs(self, BPF_W, BPF_OR); +} + +static int bpf_fill_atomic32_xor_reg_pairs(struct bpf_test *self) +{ + return __bpf_fill_atomic_reg_pairs(self, BPF_W, BPF_XOR); +} + +static int bpf_fill_atomic32_add_fetch_reg_pairs(struct bpf_test *self) +{ + return __bpf_fill_atomic_reg_pairs(self, BPF_W, BPF_ADD | BPF_FETCH); +} + +static int bpf_fill_atomic32_and_fetch_reg_pairs(struct bpf_test *self) +{ + return __bpf_fill_atomic_reg_pairs(self, BPF_W, BPF_AND | BPF_FETCH); +} + +static int bpf_fill_atomic32_or_fetch_reg_pairs(struct bpf_test *self) +{ + return __bpf_fill_atomic_reg_pairs(self, BPF_W, BPF_OR | BPF_FETCH); +} + +static int bpf_fill_atomic32_xor_fetch_reg_pairs(struct bpf_test *self) +{ + return __bpf_fill_atomic_reg_pairs(self, BPF_W, BPF_XOR | BPF_FETCH); +} + +static int bpf_fill_atomic32_xchg_reg_pairs(struct bpf_test *self) +{ + return __bpf_fill_atomic_reg_pairs(self, BPF_W, BPF_XCHG); +} + +static int bpf_fill_atomic32_cmpxchg_reg_pairs(struct bpf_test *self) +{ + return __bpf_fill_atomic_reg_pairs(self, BPF_W, BPF_CMPXCHG); +} + /* * Test the two-instruction 64-bit immediate load operation for all * power-of-two magnitudes of the immediate operand. For each MSB, a block @@ -11976,6 +12216,188 @@ static struct bpf_test tests[] = { { { 0, 1 } }, .fill_helper = bpf_fill_ld_imm64, }, + /* 64-bit ATOMIC register combinations */ + { + "ATOMIC_DW_ADD: register combinations", + { }, + INTERNAL, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_atomic64_add_reg_pairs, + .stack_depth = 8, + }, + { + "ATOMIC_DW_AND: register combinations", + { }, + INTERNAL, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_atomic64_and_reg_pairs, + .stack_depth = 8, + }, + { + "ATOMIC_DW_OR: register combinations", + { }, + INTERNAL, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_atomic64_or_reg_pairs, + .stack_depth = 8, + }, + { + "ATOMIC_DW_XOR: register combinations", + { }, + INTERNAL, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_atomic64_xor_reg_pairs, + .stack_depth = 8, + }, + { + "ATOMIC_DW_ADD_FETCH: register combinations", + { }, + INTERNAL, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_atomic64_add_fetch_reg_pairs, + .stack_depth = 8, + }, + { + "ATOMIC_DW_AND_FETCH: register combinations", + { }, + INTERNAL, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_atomic64_and_fetch_reg_pairs, + .stack_depth = 8, + }, + { + "ATOMIC_DW_OR_FETCH: register combinations", + { }, + INTERNAL, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_atomic64_or_fetch_reg_pairs, + .stack_depth = 8, + }, + { + "ATOMIC_DW_XOR_FETCH: register combinations", + { }, + INTERNAL, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_atomic64_xor_fetch_reg_pairs, + .stack_depth = 8, + }, + { + "ATOMIC_DW_XCHG: register combinations", + { }, + INTERNAL, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_atomic64_xchg_reg_pairs, + .stack_depth = 8, + }, + { + "ATOMIC_DW_CMPXCHG: register combinations", + { }, + INTERNAL, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_atomic64_cmpxchg_reg_pairs, + .stack_depth = 8, + }, + /* 32-bit ATOMIC register combinations */ + { + "ATOMIC_W_ADD: register combinations", + { }, + INTERNAL, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_atomic32_add_reg_pairs, + .stack_depth = 8, + }, + { + "ATOMIC_W_AND: register combinations", + { }, + INTERNAL, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_atomic32_and_reg_pairs, + .stack_depth = 8, + }, + { + "ATOMIC_W_OR: register combinations", + { }, + INTERNAL, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_atomic32_or_reg_pairs, + .stack_depth = 8, + }, + { + "ATOMIC_W_XOR: register combinations", + { }, + INTERNAL, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_atomic32_xor_reg_pairs, + .stack_depth = 8, + }, + { + "ATOMIC_W_ADD_FETCH: register combinations", + { }, + INTERNAL, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_atomic32_add_fetch_reg_pairs, + .stack_depth = 8, + }, + { + "ATOMIC_W_AND_FETCH: register combinations", + { }, + INTERNAL, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_atomic32_and_fetch_reg_pairs, + .stack_depth = 8, + }, + { + "ATOMIC_W_OR_FETCH: register combinations", + { }, + INTERNAL, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_atomic32_or_fetch_reg_pairs, + .stack_depth = 8, + }, + { + "ATOMIC_W_XOR_FETCH: register combinations", + { }, + INTERNAL, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_atomic32_xor_fetch_reg_pairs, + .stack_depth = 8, + }, + { + "ATOMIC_W_XCHG: register combinations", + { }, + INTERNAL, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_atomic32_xchg_reg_pairs, + .stack_depth = 8, + }, + { + "ATOMIC_W_CMPXCHG: register combinations", + { }, + INTERNAL, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_atomic32_cmpxchg_reg_pairs, + .stack_depth = 8, + }, /* 64-bit ATOMIC magnitudes */ { "ATOMIC_DW_ADD: all operand magnitudes", -- cgit v1.2.3 From 68813605dea69ca0af26af7dd00384c78c1df05d Mon Sep 17 00:00:00 2001 From: Johan Almbladh Date: Fri, 1 Oct 2021 15:03:47 +0200 Subject: bpf/tests: Add test of ALU shifts with operand register aliasing This patch adds a tests of ALU32 and ALU64 LSH/RSH/ARSH operations for the case when the two operands are the same register. Mainly intended to test JITs that implement ALU64 shifts with 32-bit CPU instructions. Also renamed related helper functions for consistency with the new tests. Signed-off-by: Johan Almbladh Signed-off-by: Daniel Borkmann Link: https://lore.kernel.org/bpf/20211001130348.3670534-10-johan.almbladh@anyfinetworks.com --- lib/test_bpf.c | 162 ++++++++++++++++++++++++++++++++++++++++++++++++++++----- 1 file changed, 149 insertions(+), 13 deletions(-) (limited to 'lib') diff --git a/lib/test_bpf.c b/lib/test_bpf.c index 40db4cee4f51..dfcbdff714b6 100644 --- a/lib/test_bpf.c +++ b/lib/test_bpf.c @@ -660,37 +660,37 @@ static int __bpf_fill_alu_shift(struct bpf_test *self, u8 op, self->u.ptr.insns = insn; self->u.ptr.len = len; - BUG_ON(i > len); + BUG_ON(i != len); return 0; } -static int bpf_fill_alu_lsh_imm(struct bpf_test *self) +static int bpf_fill_alu64_lsh_imm(struct bpf_test *self) { return __bpf_fill_alu_shift(self, BPF_LSH, BPF_K, false); } -static int bpf_fill_alu_rsh_imm(struct bpf_test *self) +static int bpf_fill_alu64_rsh_imm(struct bpf_test *self) { return __bpf_fill_alu_shift(self, BPF_RSH, BPF_K, false); } -static int bpf_fill_alu_arsh_imm(struct bpf_test *self) +static int bpf_fill_alu64_arsh_imm(struct bpf_test *self) { return __bpf_fill_alu_shift(self, BPF_ARSH, BPF_K, false); } -static int bpf_fill_alu_lsh_reg(struct bpf_test *self) +static int bpf_fill_alu64_lsh_reg(struct bpf_test *self) { return __bpf_fill_alu_shift(self, BPF_LSH, BPF_X, false); } -static int bpf_fill_alu_rsh_reg(struct bpf_test *self) +static int bpf_fill_alu64_rsh_reg(struct bpf_test *self) { return __bpf_fill_alu_shift(self, BPF_RSH, BPF_X, false); } -static int bpf_fill_alu_arsh_reg(struct bpf_test *self) +static int bpf_fill_alu64_arsh_reg(struct bpf_test *self) { return __bpf_fill_alu_shift(self, BPF_ARSH, BPF_X, false); } @@ -725,6 +725,86 @@ static int bpf_fill_alu32_arsh_reg(struct bpf_test *self) return __bpf_fill_alu_shift(self, BPF_ARSH, BPF_X, true); } +/* + * Test an ALU register shift operation for all valid shift values + * for the case when the source and destination are the same. + */ +static int __bpf_fill_alu_shift_same_reg(struct bpf_test *self, u8 op, + bool alu32) +{ + int bits = alu32 ? 32 : 64; + int len = 3 + 6 * bits; + struct bpf_insn *insn; + int i = 0; + u64 val; + + insn = kmalloc_array(len, sizeof(*insn), GFP_KERNEL); + if (!insn) + return -ENOMEM; + + insn[i++] = BPF_ALU64_IMM(BPF_MOV, R0, 0); + + for (val = 0; val < bits; val++) { + u64 res; + + /* Perform operation */ + insn[i++] = BPF_ALU64_IMM(BPF_MOV, R1, val); + if (alu32) + insn[i++] = BPF_ALU32_REG(op, R1, R1); + else + insn[i++] = BPF_ALU64_REG(op, R1, R1); + + /* Compute the reference result */ + __bpf_alu_result(&res, val, val, op); + if (alu32) + res = (u32)res; + i += __bpf_ld_imm64(&insn[i], R2, res); + + /* Check the actual result */ + insn[i++] = BPF_JMP_REG(BPF_JEQ, R1, R2, 1); + insn[i++] = BPF_EXIT_INSN(); + } + + insn[i++] = BPF_ALU64_IMM(BPF_MOV, R0, 1); + insn[i++] = BPF_EXIT_INSN(); + + self->u.ptr.insns = insn; + self->u.ptr.len = len; + BUG_ON(i != len); + + return 0; +} + +static int bpf_fill_alu64_lsh_same_reg(struct bpf_test *self) +{ + return __bpf_fill_alu_shift_same_reg(self, BPF_LSH, false); +} + +static int bpf_fill_alu64_rsh_same_reg(struct bpf_test *self) +{ + return __bpf_fill_alu_shift_same_reg(self, BPF_RSH, false); +} + +static int bpf_fill_alu64_arsh_same_reg(struct bpf_test *self) +{ + return __bpf_fill_alu_shift_same_reg(self, BPF_ARSH, false); +} + +static int bpf_fill_alu32_lsh_same_reg(struct bpf_test *self) +{ + return __bpf_fill_alu_shift_same_reg(self, BPF_LSH, true); +} + +static int bpf_fill_alu32_rsh_same_reg(struct bpf_test *self) +{ + return __bpf_fill_alu_shift_same_reg(self, BPF_RSH, true); +} + +static int bpf_fill_alu32_arsh_same_reg(struct bpf_test *self) +{ + return __bpf_fill_alu_shift_same_reg(self, BPF_ARSH, true); +} + /* * Common operand pattern generator for exhaustive power-of-two magnitudes * tests. The block size parameters can be adjusted to increase/reduce the @@ -11788,7 +11868,7 @@ static struct bpf_test tests[] = { INTERNAL | FLAG_NO_DATA, { }, { { 0, 1 } }, - .fill_helper = bpf_fill_alu_lsh_imm, + .fill_helper = bpf_fill_alu64_lsh_imm, }, { "ALU64_RSH_K: all shift values", @@ -11796,7 +11876,7 @@ static struct bpf_test tests[] = { INTERNAL | FLAG_NO_DATA, { }, { { 0, 1 } }, - .fill_helper = bpf_fill_alu_rsh_imm, + .fill_helper = bpf_fill_alu64_rsh_imm, }, { "ALU64_ARSH_K: all shift values", @@ -11804,7 +11884,7 @@ static struct bpf_test tests[] = { INTERNAL | FLAG_NO_DATA, { }, { { 0, 1 } }, - .fill_helper = bpf_fill_alu_arsh_imm, + .fill_helper = bpf_fill_alu64_arsh_imm, }, { "ALU64_LSH_X: all shift values", @@ -11812,7 +11892,7 @@ static struct bpf_test tests[] = { INTERNAL | FLAG_NO_DATA, { }, { { 0, 1 } }, - .fill_helper = bpf_fill_alu_lsh_reg, + .fill_helper = bpf_fill_alu64_lsh_reg, }, { "ALU64_RSH_X: all shift values", @@ -11820,7 +11900,7 @@ static struct bpf_test tests[] = { INTERNAL | FLAG_NO_DATA, { }, { { 0, 1 } }, - .fill_helper = bpf_fill_alu_rsh_reg, + .fill_helper = bpf_fill_alu64_rsh_reg, }, { "ALU64_ARSH_X: all shift values", @@ -11828,7 +11908,7 @@ static struct bpf_test tests[] = { INTERNAL | FLAG_NO_DATA, { }, { { 0, 1 } }, - .fill_helper = bpf_fill_alu_arsh_reg, + .fill_helper = bpf_fill_alu64_arsh_reg, }, /* Exhaustive test of ALU32 shift operations */ { @@ -11879,6 +11959,62 @@ static struct bpf_test tests[] = { { { 0, 1 } }, .fill_helper = bpf_fill_alu32_arsh_reg, }, + /* + * Exhaustive test of ALU64 shift operations when + * source and destination register are the same. + */ + { + "ALU64_LSH_X: all shift values with the same register", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu64_lsh_same_reg, + }, + { + "ALU64_RSH_X: all shift values with the same register", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu64_rsh_same_reg, + }, + { + "ALU64_ARSH_X: all shift values with the same register", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu64_arsh_same_reg, + }, + /* + * Exhaustive test of ALU32 shift operations when + * source and destination register are the same. + */ + { + "ALU32_LSH_X: all shift values with the same register", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu32_lsh_same_reg, + }, + { + "ALU32_RSH_X: all shift values with the same register", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu32_rsh_same_reg, + }, + { + "ALU32_ARSH_X: all shift values with the same register", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_alu32_arsh_same_reg, + }, /* ALU64 immediate magnitudes */ { "ALU64_MOV_K: all immediate value magnitudes", -- cgit v1.2.3 From 7bceeb95726b105bd4241c9635acc0836df675d4 Mon Sep 17 00:00:00 2001 From: Johan Almbladh Date: Fri, 1 Oct 2021 15:03:48 +0200 Subject: bpf/tests: Add test of LDX_MEM with operand aliasing This patch adds a set of tests of BPF_LDX_MEM where both operand registers are the same register. Mainly testing 32-bit JITs that may load a 64-bit value in two 32-bit loads, and must not overwrite the address register. Signed-off-by: Johan Almbladh Signed-off-by: Daniel Borkmann Link: https://lore.kernel.org/bpf/20211001130348.3670534-11-johan.almbladh@anyfinetworks.com --- lib/test_bpf.c | 58 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 58 insertions(+) (limited to 'lib') diff --git a/lib/test_bpf.c b/lib/test_bpf.c index dfcbdff714b6..b9fc330fc83b 100644 --- a/lib/test_bpf.c +++ b/lib/test_bpf.c @@ -11133,6 +11133,64 @@ static struct bpf_test tests[] = { {}, { { 0, 2 } }, }, + /* BPF_LDX_MEM with operand aliasing */ + { + "LDX_MEM_B: operand register aliasing", + .u.insns_int = { + BPF_ST_MEM(BPF_B, R10, -8, 123), + BPF_MOV64_REG(R0, R10), + BPF_LDX_MEM(BPF_B, R0, R0, -8), + BPF_EXIT_INSN(), + }, + INTERNAL, + { }, + { { 0, 123 } }, + .stack_depth = 8, + }, + { + "LDX_MEM_H: operand register aliasing", + .u.insns_int = { + BPF_ST_MEM(BPF_H, R10, -8, 12345), + BPF_MOV64_REG(R0, R10), + BPF_LDX_MEM(BPF_H, R0, R0, -8), + BPF_EXIT_INSN(), + }, + INTERNAL, + { }, + { { 0, 12345 } }, + .stack_depth = 8, + }, + { + "LDX_MEM_W: operand register aliasing", + .u.insns_int = { + BPF_ST_MEM(BPF_W, R10, -8, 123456789), + BPF_MOV64_REG(R0, R10), + BPF_LDX_MEM(BPF_W, R0, R0, -8), + BPF_EXIT_INSN(), + }, + INTERNAL, + { }, + { { 0, 123456789 } }, + .stack_depth = 8, + }, + { + "LDX_MEM_DW: operand register aliasing", + .u.insns_int = { + BPF_LD_IMM64(R1, 0x123456789abcdefULL), + BPF_STX_MEM(BPF_DW, R10, R1, -8), + BPF_MOV64_REG(R0, R10), + BPF_LDX_MEM(BPF_DW, R0, R0, -8), + BPF_ALU64_REG(BPF_SUB, R0, R1), + BPF_MOV64_REG(R1, R0), + BPF_ALU64_IMM(BPF_RSH, R1, 32), + BPF_ALU64_REG(BPF_OR, R0, R1), + BPF_EXIT_INSN(), + }, + INTERNAL, + { }, + { { 0, 0 } }, + .stack_depth = 8, + }, /* * Register (non-)clobbering tests for the case where a JIT implements * complex ALU or ATOMIC operations via function calls. If so, the -- cgit v1.2.3 From 42f355ef59a2f98fa4affb4265d3ba3e2d86baf1 Mon Sep 17 00:00:00 2001 From: Richard Guy Briggs Date: Wed, 19 May 2021 16:00:20 -0400 Subject: audit: replace magic audit syscall class numbers with macros Replace audit syscall class magic numbers with macros. This required putting the macros into new header file include/linux/audit_arch.h since the syscall macros were included for both 64 bit and 32 bit in any compat code, causing redefinition warnings. Link: https://lore.kernel.org/r/2300b1083a32aade7ae7efb95826e8f3f260b1df.1621363275.git.rgb@redhat.com Signed-off-by: Richard Guy Briggs Acked-by: Christian Brauner [PM: renamed header to audit_arch.h after consulting with Richard] Signed-off-by: Paul Moore --- MAINTAINERS | 1 + arch/alpha/kernel/audit.c | 8 ++++---- arch/ia64/kernel/audit.c | 8 ++++---- arch/parisc/kernel/audit.c | 8 ++++---- arch/parisc/kernel/compat_audit.c | 9 +++++---- arch/powerpc/kernel/audit.c | 10 +++++----- arch/powerpc/kernel/compat_audit.c | 11 ++++++----- arch/s390/kernel/audit.c | 10 +++++----- arch/s390/kernel/compat_audit.c | 11 ++++++----- arch/sparc/kernel/audit.c | 10 +++++----- arch/sparc/kernel/compat_audit.c | 11 ++++++----- arch/x86/ia32/audit.c | 11 ++++++----- arch/x86/kernel/audit_64.c | 8 ++++---- include/linux/audit.h | 1 + include/linux/audit_arch.h | 23 +++++++++++++++++++++++ kernel/auditsc.c | 12 ++++++------ lib/audit.c | 10 +++++----- lib/compat_audit.c | 11 ++++++----- 18 files changed, 102 insertions(+), 71 deletions(-) create mode 100644 include/linux/audit_arch.h (limited to 'lib') diff --git a/MAINTAINERS b/MAINTAINERS index eeb4c70b3d5b..94c88bc0502f 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -3113,6 +3113,7 @@ W: https://github.com/linux-audit T: git git://git.kernel.org/pub/scm/linux/kernel/git/pcmoore/audit.git F: include/asm-generic/audit_*.h F: include/linux/audit.h +F: include/linux/audit_arch.h F: include/uapi/linux/audit.h F: kernel/audit* F: lib/*audit.c diff --git a/arch/alpha/kernel/audit.c b/arch/alpha/kernel/audit.c index 96a9d18ff4c4..81cbd804e375 100644 --- a/arch/alpha/kernel/audit.c +++ b/arch/alpha/kernel/audit.c @@ -37,13 +37,13 @@ int audit_classify_syscall(int abi, unsigned syscall) { switch(syscall) { case __NR_open: - return 2; + return AUDITSC_OPEN; case __NR_openat: - return 3; + return AUDITSC_OPENAT; case __NR_execve: - return 5; + return AUDITSC_EXECVE; default: - return 0; + return AUDITSC_NATIVE; } } diff --git a/arch/ia64/kernel/audit.c b/arch/ia64/kernel/audit.c index 5192ca899fe6..dba6a74c9ab3 100644 --- a/arch/ia64/kernel/audit.c +++ b/arch/ia64/kernel/audit.c @@ -38,13 +38,13 @@ int audit_classify_syscall(int abi, unsigned syscall) { switch(syscall) { case __NR_open: - return 2; + return AUDITSC_OPEN; case __NR_openat: - return 3; + return AUDITSC_OPENAT; case __NR_execve: - return 5; + return AUDITSC_EXECVE; default: - return 0; + return AUDITSC_NATIVE; } } diff --git a/arch/parisc/kernel/audit.c b/arch/parisc/kernel/audit.c index 9eb47b2225d2..14244e83db75 100644 --- a/arch/parisc/kernel/audit.c +++ b/arch/parisc/kernel/audit.c @@ -47,13 +47,13 @@ int audit_classify_syscall(int abi, unsigned syscall) #endif switch (syscall) { case __NR_open: - return 2; + return AUDITSC_OPEN; case __NR_openat: - return 3; + return AUDITSC_OPENAT; case __NR_execve: - return 5; + return AUDITSC_EXECVE; default: - return 0; + return AUDITSC_NATIVE; } } diff --git a/arch/parisc/kernel/compat_audit.c b/arch/parisc/kernel/compat_audit.c index 20c39c9d86a9..1991b99f92ba 100644 --- a/arch/parisc/kernel/compat_audit.c +++ b/arch/parisc/kernel/compat_audit.c @@ -1,4 +1,5 @@ // SPDX-License-Identifier: GPL-2.0 +#include #include unsigned int parisc32_dir_class[] = { @@ -30,12 +31,12 @@ int parisc32_classify_syscall(unsigned syscall) { switch (syscall) { case __NR_open: - return 2; + return AUDITSC_OPEN; case __NR_openat: - return 3; + return AUDITSC_OPENAT; case __NR_execve: - return 5; + return AUDITSC_EXECVE; default: - return 1; + return AUDITSC_COMPAT; } } diff --git a/arch/powerpc/kernel/audit.c b/arch/powerpc/kernel/audit.c index a2dddd7f3d09..6eb18ef77dff 100644 --- a/arch/powerpc/kernel/audit.c +++ b/arch/powerpc/kernel/audit.c @@ -47,15 +47,15 @@ int audit_classify_syscall(int abi, unsigned syscall) #endif switch(syscall) { case __NR_open: - return 2; + return AUDITSC_OPEN; case __NR_openat: - return 3; + return AUDITSC_OPENAT; case __NR_socketcall: - return 4; + return AUDITSC_SOCKETCALL; case __NR_execve: - return 5; + return AUDITSC_EXECVE; default: - return 0; + return AUDITSC_NATIVE; } } diff --git a/arch/powerpc/kernel/compat_audit.c b/arch/powerpc/kernel/compat_audit.c index 55c6ccda0a85..216a54f85a12 100644 --- a/arch/powerpc/kernel/compat_audit.c +++ b/arch/powerpc/kernel/compat_audit.c @@ -1,5 +1,6 @@ // SPDX-License-Identifier: GPL-2.0 #undef __powerpc64__ +#include #include unsigned ppc32_dir_class[] = { @@ -31,14 +32,14 @@ int ppc32_classify_syscall(unsigned syscall) { switch(syscall) { case __NR_open: - return 2; + return AUDITSC_OPEN; case __NR_openat: - return 3; + return AUDITSC_OPENAT; case __NR_socketcall: - return 4; + return AUDITSC_SOCKETCALL; case __NR_execve: - return 5; + return AUDITSC_EXECVE; default: - return 1; + return AUDITSC_COMPAT; } } diff --git a/arch/s390/kernel/audit.c b/arch/s390/kernel/audit.c index d395c6c9944c..7e331e1831d4 100644 --- a/arch/s390/kernel/audit.c +++ b/arch/s390/kernel/audit.c @@ -47,15 +47,15 @@ int audit_classify_syscall(int abi, unsigned syscall) #endif switch(syscall) { case __NR_open: - return 2; + return AUDITSC_OPEN; case __NR_openat: - return 3; + return AUDITSC_OPENAT; case __NR_socketcall: - return 4; + return AUDITSC_SOCKETCALL; case __NR_execve: - return 5; + return AUDITSC_EXECVE; default: - return 0; + return AUDITSC_NATIVE; } } diff --git a/arch/s390/kernel/compat_audit.c b/arch/s390/kernel/compat_audit.c index 444fb1f66944..acacc96c57cb 100644 --- a/arch/s390/kernel/compat_audit.c +++ b/arch/s390/kernel/compat_audit.c @@ -1,5 +1,6 @@ // SPDX-License-Identifier: GPL-2.0 #undef __s390x__ +#include #include #include "audit.h" @@ -32,14 +33,14 @@ int s390_classify_syscall(unsigned syscall) { switch(syscall) { case __NR_open: - return 2; + return AUDITSC_OPEN; case __NR_openat: - return 3; + return AUDITSC_OPENAT; case __NR_socketcall: - return 4; + return AUDITSC_SOCKETCALL; case __NR_execve: - return 5; + return AUDITSC_EXECVE; default: - return 1; + return AUDITSC_COMPAT; } } diff --git a/arch/sparc/kernel/audit.c b/arch/sparc/kernel/audit.c index a6e91bf34d48..50fab35bdaba 100644 --- a/arch/sparc/kernel/audit.c +++ b/arch/sparc/kernel/audit.c @@ -48,15 +48,15 @@ int audit_classify_syscall(int abi, unsigned int syscall) #endif switch(syscall) { case __NR_open: - return 2; + return AUDITSC_OPEN; case __NR_openat: - return 3; + return AUDITSC_OPENAT; case __NR_socketcall: - return 4; + return AUDITSC_SOCKETCALL; case __NR_execve: - return 5; + return AUDITSC_EXECVE; default: - return 0; + return AUDITSC_NATIVE; } } diff --git a/arch/sparc/kernel/compat_audit.c b/arch/sparc/kernel/compat_audit.c index 10eeb4f15b20..12515bd655c3 100644 --- a/arch/sparc/kernel/compat_audit.c +++ b/arch/sparc/kernel/compat_audit.c @@ -1,5 +1,6 @@ // SPDX-License-Identifier: GPL-2.0 #define __32bit_syscall_numbers__ +#include #include #include "kernel.h" @@ -32,14 +33,14 @@ int sparc32_classify_syscall(unsigned int syscall) { switch(syscall) { case __NR_open: - return 2; + return AUDITSC_OPEN; case __NR_openat: - return 3; + return AUDITSC_OPENAT; case __NR_socketcall: - return 4; + return AUDITSC_SOCKETCALL; case __NR_execve: - return 5; + return AUDITSC_EXECVE; default: - return 1; + return AUDITSC_COMPAT; } } diff --git a/arch/x86/ia32/audit.c b/arch/x86/ia32/audit.c index 6efe6cb3768a..a5fc3b1385e0 100644 --- a/arch/x86/ia32/audit.c +++ b/arch/x86/ia32/audit.c @@ -1,4 +1,5 @@ // SPDX-License-Identifier: GPL-2.0 +#include #include #include @@ -31,15 +32,15 @@ int ia32_classify_syscall(unsigned syscall) { switch (syscall) { case __NR_open: - return 2; + return AUDITSC_OPEN; case __NR_openat: - return 3; + return AUDITSC_OPENAT; case __NR_socketcall: - return 4; + return AUDITSC_SOCKETCALL; case __NR_execve: case __NR_execveat: - return 5; + return AUDITSC_EXECVE; default: - return 1; + return AUDITSC_COMPAT; } } diff --git a/arch/x86/kernel/audit_64.c b/arch/x86/kernel/audit_64.c index 83d9cad4e68b..2a6cc9c9c881 100644 --- a/arch/x86/kernel/audit_64.c +++ b/arch/x86/kernel/audit_64.c @@ -47,14 +47,14 @@ int audit_classify_syscall(int abi, unsigned syscall) #endif switch(syscall) { case __NR_open: - return 2; + return AUDITSC_OPEN; case __NR_openat: - return 3; + return AUDITSC_OPENAT; case __NR_execve: case __NR_execveat: - return 5; + return AUDITSC_EXECVE; default: - return 0; + return AUDITSC_NATIVE; } } diff --git a/include/linux/audit.h b/include/linux/audit.h index 82b7c1116a85..5fbeeeb6b726 100644 --- a/include/linux/audit.h +++ b/include/linux/audit.h @@ -11,6 +11,7 @@ #include #include +#include #include #include diff --git a/include/linux/audit_arch.h b/include/linux/audit_arch.h new file mode 100644 index 000000000000..d4a506faabb0 --- /dev/null +++ b/include/linux/audit_arch.h @@ -0,0 +1,23 @@ +/* SPDX-License-Identifier: GPL-2.0-or-later */ +/* audit_arch.h -- Arch layer specific support for audit + * + * Copyright 2021 Red Hat Inc., Durham, North Carolina. + * All Rights Reserved. + * + * Author: Richard Guy Briggs + */ +#ifndef _LINUX_AUDIT_ARCH_H_ +#define _LINUX_AUDIT_ARCH_H_ + +enum auditsc_class_t { + AUDITSC_NATIVE = 0, + AUDITSC_COMPAT, + AUDITSC_OPEN, + AUDITSC_OPENAT, + AUDITSC_SOCKETCALL, + AUDITSC_EXECVE, + + AUDITSC_NVALS /* count */ +}; + +#endif diff --git a/kernel/auditsc.c b/kernel/auditsc.c index 1af4e7d0c16e..3f9108101598 100644 --- a/kernel/auditsc.c +++ b/kernel/auditsc.c @@ -153,7 +153,7 @@ static int audit_match_perm(struct audit_context *ctx, int mask) n = ctx->major; switch (audit_classify_syscall(ctx->arch, n)) { - case 0: /* native */ + case AUDITSC_NATIVE: if ((mask & AUDIT_PERM_WRITE) && audit_match_class(AUDIT_CLASS_WRITE, n)) return 1; @@ -164,7 +164,7 @@ static int audit_match_perm(struct audit_context *ctx, int mask) audit_match_class(AUDIT_CLASS_CHATTR, n)) return 1; return 0; - case 1: /* 32bit on biarch */ + case AUDITSC_COMPAT: /* 32bit on biarch */ if ((mask & AUDIT_PERM_WRITE) && audit_match_class(AUDIT_CLASS_WRITE_32, n)) return 1; @@ -175,13 +175,13 @@ static int audit_match_perm(struct audit_context *ctx, int mask) audit_match_class(AUDIT_CLASS_CHATTR_32, n)) return 1; return 0; - case 2: /* open */ + case AUDITSC_OPEN: return mask & ACC_MODE(ctx->argv[1]); - case 3: /* openat */ + case AUDITSC_OPENAT: return mask & ACC_MODE(ctx->argv[2]); - case 4: /* socketcall */ + case AUDITSC_SOCKETCALL: return ((mask & AUDIT_PERM_WRITE) && ctx->argv[0] == SYS_BIND); - case 5: /* execve */ + case AUDITSC_EXECVE: return mask & AUDIT_PERM_EXEC; default: return 0; diff --git a/lib/audit.c b/lib/audit.c index 5004bff928a7..3ec1a94d8d64 100644 --- a/lib/audit.c +++ b/lib/audit.c @@ -45,23 +45,23 @@ int audit_classify_syscall(int abi, unsigned syscall) switch(syscall) { #ifdef __NR_open case __NR_open: - return 2; + return AUDITSC_OPEN; #endif #ifdef __NR_openat case __NR_openat: - return 3; + return AUDITSC_OPENAT; #endif #ifdef __NR_socketcall case __NR_socketcall: - return 4; + return AUDITSC_SOCKETCALL; #endif #ifdef __NR_execveat case __NR_execveat: #endif case __NR_execve: - return 5; + return AUDITSC_EXECVE; default: - return 0; + return AUDITSC_NATIVE; } } diff --git a/lib/compat_audit.c b/lib/compat_audit.c index 77eabad69b4a..d6567d9e8b99 100644 --- a/lib/compat_audit.c +++ b/lib/compat_audit.c @@ -1,6 +1,7 @@ // SPDX-License-Identifier: GPL-2.0 #include #include +#include #include unsigned compat_dir_class[] = { @@ -33,19 +34,19 @@ int audit_classify_compat_syscall(int abi, unsigned syscall) switch (syscall) { #ifdef __NR_open case __NR_open: - return 2; + return AUDITSC_OPEN; #endif #ifdef __NR_openat case __NR_openat: - return 3; + return AUDITSC_OPENAT; #endif #ifdef __NR_socketcall case __NR_socketcall: - return 4; + return AUDITSC_SOCKETCALL; #endif case __NR_execve: - return 5; + return AUDITSC_EXECVE; default: - return 1; + return AUDITSC_COMPAT; } } -- cgit v1.2.3 From 1c30e3af8a79260cdba833a719209b01e6b92300 Mon Sep 17 00:00:00 2001 From: Richard Guy Briggs Date: Wed, 19 May 2021 16:00:21 -0400 Subject: audit: add support for the openat2 syscall The openat2(2) syscall was added in kernel v5.6 with commit fddb5d430ad9 ("open: introduce openat2(2) syscall"). Add the openat2(2) syscall to the audit syscall classifier. Link: https://github.com/linux-audit/audit-kernel/issues/67 Link: https://lore.kernel.org/r/f5f1a4d8699613f8c02ce762807228c841c2e26f.1621363275.git.rgb@redhat.com Signed-off-by: Richard Guy Briggs Acked-by: Christian Brauner [PM: merge fuzz due to previous header rename, commit line wraps] Signed-off-by: Paul Moore --- arch/alpha/kernel/audit.c | 2 ++ arch/ia64/kernel/audit.c | 2 ++ arch/parisc/kernel/audit.c | 2 ++ arch/parisc/kernel/compat_audit.c | 2 ++ arch/powerpc/kernel/audit.c | 2 ++ arch/powerpc/kernel/compat_audit.c | 2 ++ arch/s390/kernel/audit.c | 2 ++ arch/s390/kernel/compat_audit.c | 2 ++ arch/sparc/kernel/audit.c | 2 ++ arch/sparc/kernel/compat_audit.c | 2 ++ arch/x86/ia32/audit.c | 2 ++ arch/x86/kernel/audit_64.c | 2 ++ include/linux/audit_arch.h | 1 + kernel/auditsc.c | 3 +++ lib/audit.c | 4 ++++ lib/compat_audit.c | 4 ++++ 16 files changed, 36 insertions(+) (limited to 'lib') diff --git a/arch/alpha/kernel/audit.c b/arch/alpha/kernel/audit.c index 81cbd804e375..3ab04709784a 100644 --- a/arch/alpha/kernel/audit.c +++ b/arch/alpha/kernel/audit.c @@ -42,6 +42,8 @@ int audit_classify_syscall(int abi, unsigned syscall) return AUDITSC_OPENAT; case __NR_execve: return AUDITSC_EXECVE; + case __NR_openat2: + return AUDITSC_OPENAT2; default: return AUDITSC_NATIVE; } diff --git a/arch/ia64/kernel/audit.c b/arch/ia64/kernel/audit.c index dba6a74c9ab3..ec61f20ca61f 100644 --- a/arch/ia64/kernel/audit.c +++ b/arch/ia64/kernel/audit.c @@ -43,6 +43,8 @@ int audit_classify_syscall(int abi, unsigned syscall) return AUDITSC_OPENAT; case __NR_execve: return AUDITSC_EXECVE; + case __NR_openat2: + return AUDITSC_OPENAT2; default: return AUDITSC_NATIVE; } diff --git a/arch/parisc/kernel/audit.c b/arch/parisc/kernel/audit.c index 14244e83db75..f420b5552140 100644 --- a/arch/parisc/kernel/audit.c +++ b/arch/parisc/kernel/audit.c @@ -52,6 +52,8 @@ int audit_classify_syscall(int abi, unsigned syscall) return AUDITSC_OPENAT; case __NR_execve: return AUDITSC_EXECVE; + case __NR_openat2: + return AUDITSC_OPENAT2; default: return AUDITSC_NATIVE; } diff --git a/arch/parisc/kernel/compat_audit.c b/arch/parisc/kernel/compat_audit.c index 1991b99f92ba..539b16891bdf 100644 --- a/arch/parisc/kernel/compat_audit.c +++ b/arch/parisc/kernel/compat_audit.c @@ -36,6 +36,8 @@ int parisc32_classify_syscall(unsigned syscall) return AUDITSC_OPENAT; case __NR_execve: return AUDITSC_EXECVE; + case __NR_openat2: + return AUDITSC_OPENAT2; default: return AUDITSC_COMPAT; } diff --git a/arch/powerpc/kernel/audit.c b/arch/powerpc/kernel/audit.c index 6eb18ef77dff..1bcfca5fdf67 100644 --- a/arch/powerpc/kernel/audit.c +++ b/arch/powerpc/kernel/audit.c @@ -54,6 +54,8 @@ int audit_classify_syscall(int abi, unsigned syscall) return AUDITSC_SOCKETCALL; case __NR_execve: return AUDITSC_EXECVE; + case __NR_openat2: + return AUDITSC_OPENAT2; default: return AUDITSC_NATIVE; } diff --git a/arch/powerpc/kernel/compat_audit.c b/arch/powerpc/kernel/compat_audit.c index 216a54f85a12..d92ffe4e5dc1 100644 --- a/arch/powerpc/kernel/compat_audit.c +++ b/arch/powerpc/kernel/compat_audit.c @@ -39,6 +39,8 @@ int ppc32_classify_syscall(unsigned syscall) return AUDITSC_SOCKETCALL; case __NR_execve: return AUDITSC_EXECVE; + case __NR_openat2: + return AUDITSC_OPENAT2; default: return AUDITSC_COMPAT; } diff --git a/arch/s390/kernel/audit.c b/arch/s390/kernel/audit.c index 7e331e1831d4..02051a596b87 100644 --- a/arch/s390/kernel/audit.c +++ b/arch/s390/kernel/audit.c @@ -54,6 +54,8 @@ int audit_classify_syscall(int abi, unsigned syscall) return AUDITSC_SOCKETCALL; case __NR_execve: return AUDITSC_EXECVE; + case __NR_openat2: + return AUDITSC_OPENAT2; default: return AUDITSC_NATIVE; } diff --git a/arch/s390/kernel/compat_audit.c b/arch/s390/kernel/compat_audit.c index acacc96c57cb..a7c46e8310f0 100644 --- a/arch/s390/kernel/compat_audit.c +++ b/arch/s390/kernel/compat_audit.c @@ -40,6 +40,8 @@ int s390_classify_syscall(unsigned syscall) return AUDITSC_SOCKETCALL; case __NR_execve: return AUDITSC_EXECVE; + case __NR_openat2: + return AUDITSC_OPENAT2; default: return AUDITSC_COMPAT; } diff --git a/arch/sparc/kernel/audit.c b/arch/sparc/kernel/audit.c index 50fab35bdaba..b092274eca79 100644 --- a/arch/sparc/kernel/audit.c +++ b/arch/sparc/kernel/audit.c @@ -55,6 +55,8 @@ int audit_classify_syscall(int abi, unsigned int syscall) return AUDITSC_SOCKETCALL; case __NR_execve: return AUDITSC_EXECVE; + case __NR_openat2: + return AUDITSC_OPENAT2; default: return AUDITSC_NATIVE; } diff --git a/arch/sparc/kernel/compat_audit.c b/arch/sparc/kernel/compat_audit.c index 12515bd655c3..f1ea0005a729 100644 --- a/arch/sparc/kernel/compat_audit.c +++ b/arch/sparc/kernel/compat_audit.c @@ -40,6 +40,8 @@ int sparc32_classify_syscall(unsigned int syscall) return AUDITSC_SOCKETCALL; case __NR_execve: return AUDITSC_EXECVE; + case __NR_openat2: + return AUDITSC_OPENAT2; default: return AUDITSC_COMPAT; } diff --git a/arch/x86/ia32/audit.c b/arch/x86/ia32/audit.c index a5fc3b1385e0..59e19549e759 100644 --- a/arch/x86/ia32/audit.c +++ b/arch/x86/ia32/audit.c @@ -40,6 +40,8 @@ int ia32_classify_syscall(unsigned syscall) case __NR_execve: case __NR_execveat: return AUDITSC_EXECVE; + case __NR_openat2: + return AUDITSC_OPENAT2; default: return AUDITSC_COMPAT; } diff --git a/arch/x86/kernel/audit_64.c b/arch/x86/kernel/audit_64.c index 2a6cc9c9c881..44c3601cfdc4 100644 --- a/arch/x86/kernel/audit_64.c +++ b/arch/x86/kernel/audit_64.c @@ -53,6 +53,8 @@ int audit_classify_syscall(int abi, unsigned syscall) case __NR_execve: case __NR_execveat: return AUDITSC_EXECVE; + case __NR_openat2: + return AUDITSC_OPENAT2; default: return AUDITSC_NATIVE; } diff --git a/include/linux/audit_arch.h b/include/linux/audit_arch.h index d4a506faabb0..8fdb1afe251a 100644 --- a/include/linux/audit_arch.h +++ b/include/linux/audit_arch.h @@ -16,6 +16,7 @@ enum auditsc_class_t { AUDITSC_OPENAT, AUDITSC_SOCKETCALL, AUDITSC_EXECVE, + AUDITSC_OPENAT2, AUDITSC_NVALS /* count */ }; diff --git a/kernel/auditsc.c b/kernel/auditsc.c index 3f9108101598..8c4335a35274 100644 --- a/kernel/auditsc.c +++ b/kernel/auditsc.c @@ -63,6 +63,7 @@ #include #include #include +#include #include "audit.h" @@ -183,6 +184,8 @@ static int audit_match_perm(struct audit_context *ctx, int mask) return ((mask & AUDIT_PERM_WRITE) && ctx->argv[0] == SYS_BIND); case AUDITSC_EXECVE: return mask & AUDIT_PERM_EXEC; + case AUDITSC_OPENAT2: + return mask & ACC_MODE((u32)((struct open_how *)ctx->argv[2])->flags); default: return 0; } diff --git a/lib/audit.c b/lib/audit.c index 3ec1a94d8d64..738bda22dd39 100644 --- a/lib/audit.c +++ b/lib/audit.c @@ -60,6 +60,10 @@ int audit_classify_syscall(int abi, unsigned syscall) #endif case __NR_execve: return AUDITSC_EXECVE; +#ifdef __NR_openat2 + case __NR_openat2: + return AUDITSC_OPENAT2; +#endif default: return AUDITSC_NATIVE; } diff --git a/lib/compat_audit.c b/lib/compat_audit.c index d6567d9e8b99..3d6b8996f027 100644 --- a/lib/compat_audit.c +++ b/lib/compat_audit.c @@ -46,6 +46,10 @@ int audit_classify_compat_syscall(int abi, unsigned syscall) #endif case __NR_execve: return AUDITSC_EXECVE; +#ifdef __NR_openat2 + case __NR_openat2: + return AUDITSC_OPENAT2; +#endif default: return AUDITSC_COMPAT; } -- cgit v1.2.3 From 2d27e585147395316289c63efc932984675c65c2 Mon Sep 17 00:00:00 2001 From: Vincenzo Frascino Date: Wed, 6 Oct 2021 16:47:51 +0100 Subject: kasan: Extend KASAN mode kernel parameter Architectures supported by KASAN_HW_TAGS can provide an asymmetric mode of execution. On an MTE enabled arm64 hw for example this can be identified with the asymmetric tagging mode of execution. In particular, when such a mode is present, the CPU triggers a fault on a tag mismatch during a load operation and asynchronously updates a register when a tag mismatch is detected during a store operation. Extend the KASAN HW execution mode kernel command line parameter to support asymmetric mode. Cc: Dmitry Vyukov Cc: Andrey Ryabinin Cc: Alexander Potapenko Cc: Andrey Konovalov Signed-off-by: Vincenzo Frascino Reviewed-by: Catalin Marinas Reviewed-by: Andrey Konovalov Link: https://lore.kernel.org/r/20211006154751.4463-6-vincenzo.frascino@arm.com Signed-off-by: Will Deacon --- Documentation/dev-tools/kasan.rst | 7 +++++-- lib/test_kasan.c | 2 +- mm/kasan/hw_tags.c | 29 +++++++++++++++++++---------- mm/kasan/kasan.h | 30 ++++++++++++++++++++++++++---- mm/kasan/report.c | 2 +- 5 files changed, 52 insertions(+), 18 deletions(-) (limited to 'lib') diff --git a/Documentation/dev-tools/kasan.rst b/Documentation/dev-tools/kasan.rst index 21dc03bc10a4..8089c559d339 100644 --- a/Documentation/dev-tools/kasan.rst +++ b/Documentation/dev-tools/kasan.rst @@ -194,14 +194,17 @@ additional boot parameters that allow disabling KASAN or controlling features: - ``kasan=off`` or ``=on`` controls whether KASAN is enabled (default: ``on``). -- ``kasan.mode=sync`` or ``=async`` controls whether KASAN is configured in - synchronous or asynchronous mode of execution (default: ``sync``). +- ``kasan.mode=sync``, ``=async`` or ``=asymm`` controls whether KASAN + is configured in synchronous, asynchronous or asymmetric mode of + execution (default: ``sync``). Synchronous mode: a bad access is detected immediately when a tag check fault occurs. Asynchronous mode: a bad access detection is delayed. When a tag check fault occurs, the information is stored in hardware (in the TFSR_EL1 register for arm64). The kernel periodically checks the hardware and only reports tag faults during these checks. + Asymmetric mode: a bad access is detected synchronously on reads and + asynchronously on writes. - ``kasan.stacktrace=off`` or ``=on`` disables or enables alloc and free stack traces collection (default: ``on``). diff --git a/lib/test_kasan.c b/lib/test_kasan.c index 8835e0784578..ebed755ebf34 100644 --- a/lib/test_kasan.c +++ b/lib/test_kasan.c @@ -88,7 +88,7 @@ static void kasan_test_exit(struct kunit *test) */ #define KUNIT_EXPECT_KASAN_FAIL(test, expression) do { \ if (IS_ENABLED(CONFIG_KASAN_HW_TAGS) && \ - !kasan_async_mode_enabled()) \ + kasan_sync_fault_possible()) \ migrate_disable(); \ KUNIT_EXPECT_FALSE(test, READ_ONCE(fail_data.report_found)); \ barrier(); \ diff --git a/mm/kasan/hw_tags.c b/mm/kasan/hw_tags.c index 05d1e9460e2e..dc892119e88f 100644 --- a/mm/kasan/hw_tags.c +++ b/mm/kasan/hw_tags.c @@ -29,6 +29,7 @@ enum kasan_arg_mode { KASAN_ARG_MODE_DEFAULT, KASAN_ARG_MODE_SYNC, KASAN_ARG_MODE_ASYNC, + KASAN_ARG_MODE_ASYMM, }; enum kasan_arg_stacktrace { @@ -45,9 +46,9 @@ static enum kasan_arg_stacktrace kasan_arg_stacktrace __ro_after_init; DEFINE_STATIC_KEY_FALSE(kasan_flag_enabled); EXPORT_SYMBOL(kasan_flag_enabled); -/* Whether the asynchronous mode is enabled. */ -bool kasan_flag_async __ro_after_init; -EXPORT_SYMBOL_GPL(kasan_flag_async); +/* Whether the selected mode is synchronous/asynchronous/asymmetric.*/ +enum kasan_mode kasan_mode __ro_after_init; +EXPORT_SYMBOL_GPL(kasan_mode); /* Whether to collect alloc/free stack traces. */ DEFINE_STATIC_KEY_FALSE(kasan_flag_stacktrace); @@ -69,7 +70,7 @@ static int __init early_kasan_flag(char *arg) } early_param("kasan", early_kasan_flag); -/* kasan.mode=sync/async */ +/* kasan.mode=sync/async/asymm */ static int __init early_kasan_mode(char *arg) { if (!arg) @@ -79,6 +80,8 @@ static int __init early_kasan_mode(char *arg) kasan_arg_mode = KASAN_ARG_MODE_SYNC; else if (!strcmp(arg, "async")) kasan_arg_mode = KASAN_ARG_MODE_ASYNC; + else if (!strcmp(arg, "asymm")) + kasan_arg_mode = KASAN_ARG_MODE_ASYMM; else return -EINVAL; @@ -116,11 +119,13 @@ void kasan_init_hw_tags_cpu(void) return; /* - * Enable async mode only when explicitly requested through - * the command line. + * Enable async or asymm modes only when explicitly requested + * through the command line. */ if (kasan_arg_mode == KASAN_ARG_MODE_ASYNC) hw_enable_tagging_async(); + else if (kasan_arg_mode == KASAN_ARG_MODE_ASYMM) + hw_enable_tagging_asymm(); else hw_enable_tagging_sync(); } @@ -143,15 +148,19 @@ void __init kasan_init_hw_tags(void) case KASAN_ARG_MODE_DEFAULT: /* * Default to sync mode. - * Do nothing, kasan_flag_async keeps its default value. */ - break; + fallthrough; case KASAN_ARG_MODE_SYNC: - /* Do nothing, kasan_flag_async keeps its default value. */ + /* Sync mode enabled. */ + kasan_mode = KASAN_MODE_SYNC; break; case KASAN_ARG_MODE_ASYNC: /* Async mode enabled. */ - kasan_flag_async = true; + kasan_mode = KASAN_MODE_ASYNC; + break; + case KASAN_ARG_MODE_ASYMM: + /* Asymm mode enabled. */ + kasan_mode = KASAN_MODE_ASYMM; break; } diff --git a/mm/kasan/kasan.h b/mm/kasan/kasan.h index 3639e7c8bb98..b495e17445ad 100644 --- a/mm/kasan/kasan.h +++ b/mm/kasan/kasan.h @@ -13,16 +13,28 @@ #include "../slab.h" DECLARE_STATIC_KEY_FALSE(kasan_flag_stacktrace); -extern bool kasan_flag_async __ro_after_init; + +enum kasan_mode { + KASAN_MODE_SYNC, + KASAN_MODE_ASYNC, + KASAN_MODE_ASYMM, +}; + +extern enum kasan_mode kasan_mode __ro_after_init; static inline bool kasan_stack_collection_enabled(void) { return static_branch_unlikely(&kasan_flag_stacktrace); } -static inline bool kasan_async_mode_enabled(void) +static inline bool kasan_async_fault_possible(void) +{ + return kasan_mode == KASAN_MODE_ASYNC || kasan_mode == KASAN_MODE_ASYMM; +} + +static inline bool kasan_sync_fault_possible(void) { - return kasan_flag_async; + return kasan_mode == KASAN_MODE_SYNC || kasan_mode == KASAN_MODE_ASYMM; } #else @@ -31,11 +43,16 @@ static inline bool kasan_stack_collection_enabled(void) return true; } -static inline bool kasan_async_mode_enabled(void) +static inline bool kasan_async_fault_possible(void) { return false; } +static inline bool kasan_sync_fault_possible(void) +{ + return true; +} + #endif #if defined(CONFIG_KASAN_GENERIC) || defined(CONFIG_KASAN_SW_TAGS) @@ -287,6 +304,9 @@ static inline const void *arch_kasan_set_tag(const void *addr, u8 tag) #ifndef arch_enable_tagging_async #define arch_enable_tagging_async() #endif +#ifndef arch_enable_tagging_asymm +#define arch_enable_tagging_asymm() +#endif #ifndef arch_force_async_tag_fault #define arch_force_async_tag_fault() #endif @@ -302,6 +322,7 @@ static inline const void *arch_kasan_set_tag(const void *addr, u8 tag) #define hw_enable_tagging_sync() arch_enable_tagging_sync() #define hw_enable_tagging_async() arch_enable_tagging_async() +#define hw_enable_tagging_asymm() arch_enable_tagging_asymm() #define hw_force_async_tag_fault() arch_force_async_tag_fault() #define hw_get_random_tag() arch_get_random_tag() #define hw_get_mem_tag(addr) arch_get_mem_tag(addr) @@ -312,6 +333,7 @@ static inline const void *arch_kasan_set_tag(const void *addr, u8 tag) #define hw_enable_tagging_sync() #define hw_enable_tagging_async() +#define hw_enable_tagging_asymm() #endif /* CONFIG_KASAN_HW_TAGS */ diff --git a/mm/kasan/report.c b/mm/kasan/report.c index 884a950c7026..9da071ad930c 100644 --- a/mm/kasan/report.c +++ b/mm/kasan/report.c @@ -112,7 +112,7 @@ static void start_report(unsigned long *flags) static void end_report(unsigned long *flags, unsigned long addr) { - if (!kasan_async_mode_enabled()) + if (!kasan_async_fault_possible()) trace_error_report_end(ERROR_DETECTOR_KASAN, addr); pr_err("==================================================================\n"); add_taint(TAINT_BAD_PAGE, LOCKDEP_NOW_UNRELIABLE); -- cgit v1.2.3 From 0eb4ef88c53f7169c44b1bd2ace5389981409a60 Mon Sep 17 00:00:00 2001 From: Johan Almbladh Date: Thu, 7 Oct 2021 16:30:06 +0200 Subject: bpf, tests: Add more LD_IMM64 tests This patch adds new tests for the two-instruction LD_IMM64. The new tests verify the operation with immediate values of different byte patterns. Mainly intended to cover JITs that want to be clever when loading 64-bit constants. Signed-off-by: Johan Almbladh Signed-off-by: Daniel Borkmann Link: https://lore.kernel.org/bpf/20211007143006.634308-1-johan.almbladh@anyfinetworks.com --- lib/test_bpf.c | 120 +++++++++++++++++++++++++++++++++++++++++++++++++++++++-- 1 file changed, 117 insertions(+), 3 deletions(-) (limited to 'lib') diff --git a/lib/test_bpf.c b/lib/test_bpf.c index b9fc330fc83b..e5b10fdefab5 100644 --- a/lib/test_bpf.c +++ b/lib/test_bpf.c @@ -2134,7 +2134,7 @@ static int bpf_fill_atomic32_cmpxchg_reg_pairs(struct bpf_test *self) * of the immediate value. This is often the case if the native instruction * immediate field width is narrower than 32 bits. */ -static int bpf_fill_ld_imm64(struct bpf_test *self) +static int bpf_fill_ld_imm64_magn(struct bpf_test *self) { int block = 64; /* Increase for more tests per MSB position */ int len = 3 + 8 * 63 * block * 2; @@ -2180,6 +2180,88 @@ static int bpf_fill_ld_imm64(struct bpf_test *self) return 0; } +/* + * Test the two-instruction 64-bit immediate load operation for different + * combinations of bytes. Each byte in the 64-bit word is constructed as + * (base & mask) | (rand() & ~mask), where rand() is a deterministic LCG. + * All patterns (base1, mask1) and (base2, mask2) bytes are tested. + */ +static int __bpf_fill_ld_imm64_bytes(struct bpf_test *self, + u8 base1, u8 mask1, + u8 base2, u8 mask2) +{ + struct bpf_insn *insn; + int len = 3 + 8 * BIT(8); + int pattern, index; + u32 rand = 1; + int i = 0; + + insn = kmalloc_array(len, sizeof(*insn), GFP_KERNEL); + if (!insn) + return -ENOMEM; + + insn[i++] = BPF_ALU64_IMM(BPF_MOV, R0, 0); + + for (pattern = 0; pattern < BIT(8); pattern++) { + u64 imm = 0; + + for (index = 0; index < 8; index++) { + int byte; + + if (pattern & BIT(index)) + byte = (base1 & mask1) | (rand & ~mask1); + else + byte = (base2 & mask2) | (rand & ~mask2); + imm = (imm << 8) | byte; + } + + /* Update our LCG */ + rand = rand * 1664525 + 1013904223; + + /* Perform operation */ + i += __bpf_ld_imm64(&insn[i], R1, imm); + + /* Load reference */ + insn[i++] = BPF_ALU32_IMM(BPF_MOV, R2, imm); + insn[i++] = BPF_ALU32_IMM(BPF_MOV, R3, (u32)(imm >> 32)); + insn[i++] = BPF_ALU64_IMM(BPF_LSH, R3, 32); + insn[i++] = BPF_ALU64_REG(BPF_OR, R2, R3); + + /* Check result */ + insn[i++] = BPF_JMP_REG(BPF_JEQ, R1, R2, 1); + insn[i++] = BPF_EXIT_INSN(); + } + + insn[i++] = BPF_ALU64_IMM(BPF_MOV, R0, 1); + insn[i++] = BPF_EXIT_INSN(); + + self->u.ptr.insns = insn; + self->u.ptr.len = len; + BUG_ON(i != len); + + return 0; +} + +static int bpf_fill_ld_imm64_checker(struct bpf_test *self) +{ + return __bpf_fill_ld_imm64_bytes(self, 0, 0xff, 0xff, 0xff); +} + +static int bpf_fill_ld_imm64_pos_neg(struct bpf_test *self) +{ + return __bpf_fill_ld_imm64_bytes(self, 1, 0x81, 0x80, 0x80); +} + +static int bpf_fill_ld_imm64_pos_zero(struct bpf_test *self) +{ + return __bpf_fill_ld_imm64_bytes(self, 1, 0x81, 0, 0xff); +} + +static int bpf_fill_ld_imm64_neg_zero(struct bpf_test *self) +{ + return __bpf_fill_ld_imm64_bytes(self, 0x80, 0x80, 0, 0xff); +} + /* * Exhaustive tests of JMP operations for all combinations of power-of-two * magnitudes of the operands, both for positive and negative values. The @@ -12401,14 +12483,46 @@ static struct bpf_test tests[] = { .fill_helper = bpf_fill_alu32_mod_reg, .nr_testruns = NR_PATTERN_RUNS, }, - /* LD_IMM64 immediate magnitudes */ + /* LD_IMM64 immediate magnitudes and byte patterns */ { "LD_IMM64: all immediate value magnitudes", { }, INTERNAL | FLAG_NO_DATA, { }, { { 0, 1 } }, - .fill_helper = bpf_fill_ld_imm64, + .fill_helper = bpf_fill_ld_imm64_magn, + }, + { + "LD_IMM64: checker byte patterns", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_ld_imm64_checker, + }, + { + "LD_IMM64: random positive and zero byte patterns", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_ld_imm64_pos_zero, + }, + { + "LD_IMM64: random negative and zero byte patterns", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_ld_imm64_neg_zero, + }, + { + "LD_IMM64: random positive and negative byte patterns", + { }, + INTERNAL | FLAG_NO_DATA, + { }, + { { 0, 1 } }, + .fill_helper = bpf_fill_ld_imm64_pos_neg, }, /* 64-bit ATOMIC register combinations */ { -- cgit v1.2.3 From bdac5c2b243f68ec15f8203c3348ae79fee8e8d8 Mon Sep 17 00:00:00 2001 From: Masami Hiramatsu Date: Thu, 16 Sep 2021 15:23:20 +0900 Subject: bootconfig: Allocate xbc_data inside xbc_init() Allocate 'xbc_data' in the xbc_init() so that it does not need to care about the ownership of the copied data. Link: https://lkml.kernel.org/r/163177339986.682366.898762699429769117.stgit@devnote2 Suggested-by: Steven Rostedt Signed-off-by: Masami Hiramatsu Signed-off-by: Steven Rostedt (VMware) --- include/linux/bootconfig.h | 2 +- init/main.c | 13 ++----------- lib/bootconfig.c | 33 +++++++++++++++++++++------------ tools/bootconfig/main.c | 6 +++--- 4 files changed, 27 insertions(+), 27 deletions(-) (limited to 'lib') diff --git a/include/linux/bootconfig.h b/include/linux/bootconfig.h index 537e1b991f11..62e09b788172 100644 --- a/include/linux/bootconfig.h +++ b/include/linux/bootconfig.h @@ -271,7 +271,7 @@ static inline int __init xbc_node_compose_key(struct xbc_node *node, } /* XBC node initializer */ -int __init xbc_init(char *buf, const char **emsg, int *epos); +int __init xbc_init(const char *buf, size_t size, const char **emsg, int *epos); /* XBC cleanup data structures */ diff --git a/init/main.c b/init/main.c index 81a79a77db46..d894989d86bc 100644 --- a/init/main.c +++ b/init/main.c @@ -409,7 +409,7 @@ static void __init setup_boot_config(void) const char *msg; int pos; u32 size, csum; - char *data, *copy, *err; + char *data, *err; int ret; /* Cut out the bootconfig data even if we have no bootconfig option */ @@ -442,16 +442,7 @@ static void __init setup_boot_config(void) return; } - copy = memblock_alloc(size + 1, SMP_CACHE_BYTES); - if (!copy) { - pr_err("Failed to allocate memory for bootconfig\n"); - return; - } - - memcpy(copy, data, size); - copy[size] = '\0'; - - ret = xbc_init(copy, &msg, &pos); + ret = xbc_init(data, size, &msg, &pos); if (ret < 0) { if (pos < 0) pr_err("Failed to init bootconfig: %s.\n", msg); diff --git a/lib/bootconfig.c b/lib/bootconfig.c index 5ae248b29373..66b02fddfea8 100644 --- a/lib/bootconfig.c +++ b/lib/bootconfig.c @@ -789,6 +789,7 @@ static int __init xbc_verify_tree(void) */ void __init xbc_destroy_all(void) { + memblock_free_ptr(xbc_data, xbc_data_size); xbc_data = NULL; xbc_data_size = 0; xbc_node_num = 0; @@ -799,19 +800,20 @@ void __init xbc_destroy_all(void) /** * xbc_init() - Parse given XBC file and build XBC internal tree - * @buf: boot config text + * @data: The boot config text original data + * @size: The size of @data * @emsg: A pointer of const char * to store the error message * @epos: A pointer of int to store the error position * - * This parses the boot config text in @buf. @buf must be a - * null terminated string and smaller than XBC_DATA_MAX. + * This parses the boot config text in @data. @size must be smaller + * than XBC_DATA_MAX. * Return the number of stored nodes (>0) if succeeded, or -errno * if there is any error. * In error cases, @emsg will be updated with an error message and * @epos will be updated with the error position which is the byte offset * of @buf. If the error is not a parser error, @epos will be -1. */ -int __init xbc_init(char *buf, const char **emsg, int *epos) +int __init xbc_init(const char *data, size_t size, const char **emsg, int *epos) { char *p, *q; int ret, c; @@ -824,28 +826,35 @@ int __init xbc_init(char *buf, const char **emsg, int *epos) *emsg = "Bootconfig is already initialized"; return -EBUSY; } - - ret = strlen(buf); - if (ret > XBC_DATA_MAX - 1 || ret == 0) { + if (size > XBC_DATA_MAX || size == 0) { if (emsg) - *emsg = ret ? "Config data is too big" : + *emsg = size ? "Config data is too big" : "Config data is empty"; return -ERANGE; } + xbc_data = memblock_alloc(size + 1, SMP_CACHE_BYTES); + if (!xbc_data) { + if (emsg) + *emsg = "Failed to allocate bootconfig data"; + return -ENOMEM; + } + memcpy(xbc_data, data, size); + xbc_data[size] = '\0'; + xbc_data_size = size + 1; + xbc_nodes = memblock_alloc(sizeof(struct xbc_node) * XBC_NODE_MAX, SMP_CACHE_BYTES); if (!xbc_nodes) { if (emsg) *emsg = "Failed to allocate bootconfig nodes"; + xbc_destroy_all(); return -ENOMEM; } memset(xbc_nodes, 0, sizeof(struct xbc_node) * XBC_NODE_MAX); - xbc_data = buf; - xbc_data_size = ret + 1; - last_parent = NULL; - p = buf; + last_parent = NULL; + p = xbc_data; do { q = strpbrk(p, "{}=+;:\n#"); if (!q) { diff --git a/tools/bootconfig/main.c b/tools/bootconfig/main.c index fd67496a947f..7269c9e35335 100644 --- a/tools/bootconfig/main.c +++ b/tools/bootconfig/main.c @@ -229,7 +229,7 @@ static int load_xbc_from_initrd(int fd, char **buf) return -EINVAL; } - ret = xbc_init(*buf, &msg, NULL); + ret = xbc_init(*buf, size, &msg, NULL); /* Wrong data */ if (ret < 0) { pr_err("parse error: %s.\n", msg); @@ -269,7 +269,7 @@ static int init_xbc_with_error(char *buf, int len) if (!copy) return -ENOMEM; - ret = xbc_init(buf, &msg, &pos); + ret = xbc_init(buf, len, &msg, &pos); if (ret < 0) show_xbc_error(copy, msg, pos); free(copy); @@ -382,7 +382,7 @@ static int apply_xbc(const char *path, const char *xbc_path) memcpy(data, buf, size); /* Check the data format */ - ret = xbc_init(buf, &msg, &pos); + ret = xbc_init(buf, size, &msg, &pos); if (ret < 0) { show_xbc_error(data, msg, pos); free(data); -- cgit v1.2.3 From e306220cb7b7c2948f191414ab06851e143b54c1 Mon Sep 17 00:00:00 2001 From: Masami Hiramatsu Date: Thu, 16 Sep 2021 15:23:29 +0900 Subject: bootconfig: Add xbc_get_info() for the node information Add xbc_get_info() API which allows user to get the number of used xbc_nodes and the size of bootconfig data. This is also useful for checking the bootconfig is initialized or not. Link: https://lkml.kernel.org/r/163177340877.682366.4360676589783197627.stgit@devnote2 Signed-off-by: Masami Hiramatsu Signed-off-by: Steven Rostedt (VMware) --- include/linux/bootconfig.h | 2 ++ init/main.c | 1 + lib/bootconfig.c | 21 +++++++++++++++++++++ tools/bootconfig/main.c | 1 + 4 files changed, 25 insertions(+) (limited to 'lib') diff --git a/include/linux/bootconfig.h b/include/linux/bootconfig.h index 62e09b788172..f955bb7eabbb 100644 --- a/include/linux/bootconfig.h +++ b/include/linux/bootconfig.h @@ -273,6 +273,8 @@ static inline int __init xbc_node_compose_key(struct xbc_node *node, /* XBC node initializer */ int __init xbc_init(const char *buf, size_t size, const char **emsg, int *epos); +/* XBC node and size information */ +int __init xbc_get_info(int *node_size, size_t *data_size); /* XBC cleanup data structures */ void __init xbc_destroy_all(void); diff --git a/init/main.c b/init/main.c index d894989d86bc..afaed805de19 100644 --- a/init/main.c +++ b/init/main.c @@ -450,6 +450,7 @@ static void __init setup_boot_config(void) pr_err("Failed to parse bootconfig: %s at %d.\n", msg, pos); } else { + xbc_get_info(&ret, NULL); pr_info("Load bootconfig: %d bytes %d nodes\n", size, ret); /* keys starting with "kernel." are passed via cmdline */ extra_command_line = xbc_make_cmdline("kernel"); diff --git a/lib/bootconfig.c b/lib/bootconfig.c index 66b02fddfea8..b088fe5c0001 100644 --- a/lib/bootconfig.c +++ b/lib/bootconfig.c @@ -34,6 +34,27 @@ static int xbc_err_pos __initdata; static int open_brace[XBC_DEPTH_MAX] __initdata; static int brace_index __initdata; +/** + * xbc_get_info() - Get the information of loaded boot config + * node_size: A pointer to store the number of nodes. + * data_size: A pointer to store the size of bootconfig data. + * + * Get the number of used nodes in @node_size if it is not NULL, + * and the size of bootconfig data in @data_size if it is not NULL. + * Return 0 if the boot config is initialized, or return -ENODEV. + */ +int __init xbc_get_info(int *node_size, size_t *data_size) +{ + if (!xbc_data) + return -ENODEV; + + if (node_size) + *node_size = xbc_node_num; + if (data_size) + *data_size = xbc_data_size; + return 0; +} + static int __init xbc_parse_error(const char *msg, const char *p) { xbc_err_msg = msg; diff --git a/tools/bootconfig/main.c b/tools/bootconfig/main.c index 7269c9e35335..4f2a8d884745 100644 --- a/tools/bootconfig/main.c +++ b/tools/bootconfig/main.c @@ -391,6 +391,7 @@ static int apply_xbc(const char *path, const char *xbc_path) return ret; } printf("Apply %s to %s\n", xbc_path, path); + xbc_get_info(&ret, NULL); printf("\tNumber of nodes: %d\n", ret); printf("\tSize: %u bytes\n", (unsigned int)size); printf("\tChecksum: %d\n", (unsigned int)csum); -- cgit v1.2.3 From 115d4d08aeb942133d025a425dd611092893d774 Mon Sep 17 00:00:00 2001 From: Masami Hiramatsu Date: Fri, 17 Sep 2021 19:02:39 +0900 Subject: bootconfig: Rename xbc_destroy_all() to xbc_exit() Avoid using this noisy name and use more calm one. This is just a name change. No functional change. Link: https://lkml.kernel.org/r/163187295918.2366983.5231840238429996027.stgit@devnote2 Signed-off-by: Masami Hiramatsu Signed-off-by: Steven Rostedt (VMware) --- include/linux/bootconfig.h | 2 +- init/main.c | 2 +- lib/bootconfig.c | 8 ++++---- tools/bootconfig/main.c | 2 +- 4 files changed, 7 insertions(+), 7 deletions(-) (limited to 'lib') diff --git a/include/linux/bootconfig.h b/include/linux/bootconfig.h index f955bb7eabbb..7eb7a7f8ade7 100644 --- a/include/linux/bootconfig.h +++ b/include/linux/bootconfig.h @@ -277,7 +277,7 @@ int __init xbc_init(const char *buf, size_t size, const char **emsg, int *epos); int __init xbc_get_info(int *node_size, size_t *data_size); /* XBC cleanup data structures */ -void __init xbc_destroy_all(void); +void __init xbc_exit(void); /* Debug dump functions */ void __init xbc_debug_dump(void); diff --git a/init/main.c b/init/main.c index afaed805de19..f1428234e1e4 100644 --- a/init/main.c +++ b/init/main.c @@ -462,7 +462,7 @@ static void __init setup_boot_config(void) static void __init exit_boot_config(void) { - xbc_destroy_all(); + xbc_exit(); } #else /* !CONFIG_BOOT_CONFIG */ diff --git a/lib/bootconfig.c b/lib/bootconfig.c index b088fe5c0001..a3ce5a0c3ca4 100644 --- a/lib/bootconfig.c +++ b/lib/bootconfig.c @@ -802,13 +802,13 @@ static int __init xbc_verify_tree(void) } /** - * xbc_destroy_all() - Clean up all parsed bootconfig + * xbc_exit() - Clean up all parsed bootconfig * * This clears all data structures of parsed bootconfig on memory. * If you need to reuse xbc_init() with new boot config, you can * use this. */ -void __init xbc_destroy_all(void) +void __init xbc_exit(void) { memblock_free_ptr(xbc_data, xbc_data_size); xbc_data = NULL; @@ -869,7 +869,7 @@ int __init xbc_init(const char *data, size_t size, const char **emsg, int *epos) if (!xbc_nodes) { if (emsg) *emsg = "Failed to allocate bootconfig nodes"; - xbc_destroy_all(); + xbc_exit(); return -ENOMEM; } memset(xbc_nodes, 0, sizeof(struct xbc_node) * XBC_NODE_MAX); @@ -925,7 +925,7 @@ int __init xbc_init(const char *data, size_t size, const char **emsg, int *epos) *epos = xbc_err_pos; if (emsg) *emsg = xbc_err_msg; - xbc_destroy_all(); + xbc_exit(); } else ret = xbc_node_num; diff --git a/tools/bootconfig/main.c b/tools/bootconfig/main.c index 4f2a8d884745..4252c23bd35d 100644 --- a/tools/bootconfig/main.c +++ b/tools/bootconfig/main.c @@ -397,7 +397,7 @@ static int apply_xbc(const char *path, const char *xbc_path) printf("\tChecksum: %d\n", (unsigned int)csum); /* TODO: Check the options by schema */ - xbc_destroy_all(); + xbc_exit(); free(buf); /* Remove old boot config if exists */ -- cgit v1.2.3 From f3668cde8562997b47a9edbc915da32279d4a743 Mon Sep 17 00:00:00 2001 From: Masami Hiramatsu Date: Fri, 17 Sep 2021 19:02:46 +0900 Subject: bootconfig: Split parse-tree part from xbc_init Split bootconfig data parser to build tree code from xbc_init(). This is an internal cosmetic change. Link: https://lkml.kernel.org/r/163187296647.2366983.15590065167920474865.stgit@devnote2 Signed-off-by: Masami Hiramatsu Signed-off-by: Steven Rostedt (VMware) --- lib/bootconfig.c | 99 ++++++++++++++++++++++++++++++-------------------------- 1 file changed, 54 insertions(+), 45 deletions(-) (limited to 'lib') diff --git a/lib/bootconfig.c b/lib/bootconfig.c index a3ce5a0c3ca4..b7e5a32b30d3 100644 --- a/lib/bootconfig.c +++ b/lib/bootconfig.c @@ -801,6 +801,58 @@ static int __init xbc_verify_tree(void) return 0; } +/* Need to setup xbc_data and xbc_nodes before call this. */ +static int __init xbc_parse_tree(void) +{ + char *p, *q; + int ret, c; + + last_parent = NULL; + p = xbc_data; + do { + q = strpbrk(p, "{}=+;:\n#"); + if (!q) { + p = skip_spaces(p); + if (*p != '\0') + ret = xbc_parse_error("No delimiter", p); + break; + } + + c = *q; + *q++ = '\0'; + switch (c) { + case ':': + case '+': + if (*q++ != '=') { + ret = xbc_parse_error(c == '+' ? + "Wrong '+' operator" : + "Wrong ':' operator", + q - 2); + break; + } + fallthrough; + case '=': + ret = xbc_parse_kv(&p, q, c); + break; + case '{': + ret = xbc_open_brace(&p, q); + break; + case '#': + q = skip_comment(q); + fallthrough; + case ';': + case '\n': + ret = xbc_parse_key(&p, q); + break; + case '}': + ret = xbc_close_brace(&p, q); + break; + } + } while (!ret); + + return ret; +} + /** * xbc_exit() - Clean up all parsed bootconfig * @@ -836,8 +888,7 @@ void __init xbc_exit(void) */ int __init xbc_init(const char *data, size_t size, const char **emsg, int *epos) { - char *p, *q; - int ret, c; + int ret; if (epos) *epos = -1; @@ -874,49 +925,7 @@ int __init xbc_init(const char *data, size_t size, const char **emsg, int *epos) } memset(xbc_nodes, 0, sizeof(struct xbc_node) * XBC_NODE_MAX); - last_parent = NULL; - p = xbc_data; - do { - q = strpbrk(p, "{}=+;:\n#"); - if (!q) { - p = skip_spaces(p); - if (*p != '\0') - ret = xbc_parse_error("No delimiter", p); - break; - } - - c = *q; - *q++ = '\0'; - switch (c) { - case ':': - case '+': - if (*q++ != '=') { - ret = xbc_parse_error(c == '+' ? - "Wrong '+' operator" : - "Wrong ':' operator", - q - 2); - break; - } - fallthrough; - case '=': - ret = xbc_parse_kv(&p, q, c); - break; - case '{': - ret = xbc_open_brace(&p, q); - break; - case '#': - q = skip_comment(q); - fallthrough; - case ';': - case '\n': - ret = xbc_parse_key(&p, q); - break; - case '}': - ret = xbc_close_brace(&p, q); - break; - } - } while (!ret); - + ret = xbc_parse_tree(); if (!ret) ret = xbc_verify_tree(); -- cgit v1.2.3 From 9b81c9bfff4651abb28bfa6d83c8b879e467963b Mon Sep 17 00:00:00 2001 From: Masami Hiramatsu Date: Fri, 17 Sep 2021 19:02:53 +0900 Subject: bootconfig: Remove unused debug function Remove unused xbc_debug_dump() from bootconfig for clean up the code. Link: https://lkml.kernel.org/r/163187297371.2366983.12943349701785875450.stgit@devnote2 Signed-off-by: Masami Hiramatsu Signed-off-by: Steven Rostedt (VMware) --- include/linux/bootconfig.h | 3 --- lib/bootconfig.c | 21 --------------------- 2 files changed, 24 deletions(-) (limited to 'lib') diff --git a/include/linux/bootconfig.h b/include/linux/bootconfig.h index 7eb7a7f8ade7..85cdfd381877 100644 --- a/include/linux/bootconfig.h +++ b/include/linux/bootconfig.h @@ -279,7 +279,4 @@ int __init xbc_get_info(int *node_size, size_t *data_size); /* XBC cleanup data structures */ void __init xbc_exit(void); -/* Debug dump functions */ -void __init xbc_debug_dump(void); - #endif diff --git a/lib/bootconfig.c b/lib/bootconfig.c index b7e5a32b30d3..953789171858 100644 --- a/lib/bootconfig.c +++ b/lib/bootconfig.c @@ -4,15 +4,12 @@ * Masami Hiramatsu */ -#define pr_fmt(fmt) "bootconfig: " fmt - #include #include #include #include #include #include -#include #include /* @@ -940,21 +937,3 @@ int __init xbc_init(const char *data, size_t size, const char **emsg, int *epos) return ret; } - -/** - * xbc_debug_dump() - Dump current XBC node list - * - * Dump the current XBC node list on printk buffer for debug. - */ -void __init xbc_debug_dump(void) -{ - int i; - - for (i = 0; i < xbc_node_num; i++) { - pr_debug("[%d] %s (%s) .next=%d, .child=%d .parent=%d\n", i, - xbc_node_get_data(xbc_nodes + i), - xbc_node_is_value(xbc_nodes + i) ? "value" : "key", - xbc_nodes[i].next, xbc_nodes[i].child, - xbc_nodes[i].parent); - } -} -- cgit v1.2.3 From 4f292c4886bfdfc2a7191ef4ff3f7aac69d1cc3f Mon Sep 17 00:00:00 2001 From: Masami Hiramatsu Date: Fri, 17 Sep 2021 19:03:08 +0900 Subject: bootconfig: Replace u16 and u32 with uint16_t and uint32_t Replace u16 and u32 with uint16_t and uint32_t so that the tools/bootconfig only needs . Link: https://lkml.kernel.org/r/163187298835.2366983.9838262576854319669.stgit@devnote2 Signed-off-by: Masami Hiramatsu Signed-off-by: Steven Rostedt (VMware) --- include/linux/bootconfig.h | 12 ++++++------ lib/bootconfig.c | 16 ++++++++-------- tools/bootconfig/include/linux/kernel.h | 4 +--- tools/bootconfig/main.c | 20 ++++++++++---------- 4 files changed, 25 insertions(+), 27 deletions(-) (limited to 'lib') diff --git a/include/linux/bootconfig.h b/include/linux/bootconfig.h index 85cdfd381877..a6f8dc51f168 100644 --- a/include/linux/bootconfig.h +++ b/include/linux/bootconfig.h @@ -25,10 +25,10 @@ * The checksum will be used with the BOOTCONFIG_MAGIC and the size for * embedding the bootconfig in the initrd image. */ -static inline __init u32 xbc_calc_checksum(void *data, u32 size) +static inline __init uint32_t xbc_calc_checksum(void *data, uint32_t size) { unsigned char *p = data; - u32 ret = 0; + uint32_t ret = 0; while (size--) ret += *p++; @@ -38,10 +38,10 @@ static inline __init u32 xbc_calc_checksum(void *data, u32 size) /* XBC tree node */ struct xbc_node { - u16 next; - u16 child; - u16 parent; - u16 data; + uint16_t next; + uint16_t child; + uint16_t parent; + uint16_t data; } __attribute__ ((__packed__)); #define XBC_KEY 0 diff --git a/lib/bootconfig.c b/lib/bootconfig.c index 953789171858..a2f5f582181d 100644 --- a/lib/bootconfig.c +++ b/lib/bootconfig.c @@ -244,7 +244,7 @@ int __init xbc_node_compose_key_after(struct xbc_node *root, struct xbc_node *node, char *buf, size_t size) { - u16 keys[XBC_DEPTH_MAX]; + uint16_t keys[XBC_DEPTH_MAX]; int depth = 0, ret = 0, total = 0; if (!node || node == root) @@ -359,21 +359,21 @@ const char * __init xbc_node_find_next_key_value(struct xbc_node *root, /* XBC parse and tree build */ -static int __init xbc_init_node(struct xbc_node *node, char *data, u32 flag) +static int __init xbc_init_node(struct xbc_node *node, char *data, uint32_t flag) { unsigned long offset = data - xbc_data; if (WARN_ON(offset >= XBC_DATA_MAX)) return -EINVAL; - node->data = (u16)offset | flag; + node->data = (uint16_t)offset | flag; node->child = 0; node->next = 0; return 0; } -static struct xbc_node * __init xbc_add_node(char *data, u32 flag) +static struct xbc_node * __init xbc_add_node(char *data, uint32_t flag) { struct xbc_node *node; @@ -403,7 +403,7 @@ static inline __init struct xbc_node *xbc_last_child(struct xbc_node *node) return node; } -static struct xbc_node * __init __xbc_add_sibling(char *data, u32 flag, bool head) +static struct xbc_node * __init __xbc_add_sibling(char *data, uint32_t flag, bool head) { struct xbc_node *sib, *node = xbc_add_node(data, flag); @@ -430,17 +430,17 @@ static struct xbc_node * __init __xbc_add_sibling(char *data, u32 flag, bool hea return node; } -static inline struct xbc_node * __init xbc_add_sibling(char *data, u32 flag) +static inline struct xbc_node * __init xbc_add_sibling(char *data, uint32_t flag) { return __xbc_add_sibling(data, flag, false); } -static inline struct xbc_node * __init xbc_add_head_sibling(char *data, u32 flag) +static inline struct xbc_node * __init xbc_add_head_sibling(char *data, uint32_t flag) { return __xbc_add_sibling(data, flag, true); } -static inline __init struct xbc_node *xbc_add_child(char *data, u32 flag) +static inline __init struct xbc_node *xbc_add_child(char *data, uint32_t flag) { struct xbc_node *node = xbc_add_sibling(data, flag); diff --git a/tools/bootconfig/include/linux/kernel.h b/tools/bootconfig/include/linux/kernel.h index c4854b8e7023..39f306c18dd0 100644 --- a/tools/bootconfig/include/linux/kernel.h +++ b/tools/bootconfig/include/linux/kernel.h @@ -3,11 +3,9 @@ #define _SKC_LINUX_KERNEL_H #include +#include #include -typedef unsigned short u16; -typedef unsigned int u32; - #define unlikely(cond) (cond) #define __init diff --git a/tools/bootconfig/main.c b/tools/bootconfig/main.c index adc6c6e73fa9..fb7c9fb953d7 100644 --- a/tools/bootconfig/main.c +++ b/tools/bootconfig/main.c @@ -178,7 +178,7 @@ static int load_xbc_from_initrd(int fd, char **buf) { struct stat stat; int ret; - u32 size = 0, csum = 0, rcsum; + uint32_t size = 0, csum = 0, rcsum; char magic[BOOTCONFIG_MAGIC_LEN]; const char *msg; @@ -202,11 +202,11 @@ static int load_xbc_from_initrd(int fd, char **buf) if (lseek(fd, -(8 + BOOTCONFIG_MAGIC_LEN), SEEK_END) < 0) return pr_errno("Failed to lseek for size", -errno); - if (read(fd, &size, sizeof(u32)) < 0) + if (read(fd, &size, sizeof(uint32_t)) < 0) return pr_errno("Failed to read size", -errno); size = le32toh(size); - if (read(fd, &csum, sizeof(u32)) < 0) + if (read(fd, &csum, sizeof(uint32_t)) < 0) return pr_errno("Failed to read checksum", -errno); csum = le32toh(csum); @@ -364,7 +364,7 @@ static int apply_xbc(const char *path, const char *xbc_path) size_t total_size; struct stat stat; const char *msg; - u32 size, csum; + uint32_t size, csum; int pos, pad; int ret, fd; @@ -378,7 +378,7 @@ static int apply_xbc(const char *path, const char *xbc_path) /* Backup the bootconfig data */ data = calloc(size + BOOTCONFIG_ALIGN + - sizeof(u32) + sizeof(u32) + BOOTCONFIG_MAGIC_LEN, 1); + sizeof(uint32_t) + sizeof(uint32_t) + BOOTCONFIG_MAGIC_LEN, 1); if (!data) return -ENOMEM; memcpy(data, buf, size); @@ -426,17 +426,17 @@ static int apply_xbc(const char *path, const char *xbc_path) } /* To align up the total size to BOOTCONFIG_ALIGN, get padding size */ - total_size = stat.st_size + size + sizeof(u32) * 2 + BOOTCONFIG_MAGIC_LEN; + total_size = stat.st_size + size + sizeof(uint32_t) * 2 + BOOTCONFIG_MAGIC_LEN; pad = ((total_size + BOOTCONFIG_ALIGN - 1) & (~BOOTCONFIG_ALIGN_MASK)) - total_size; size += pad; /* Add a footer */ p = data + size; - *(u32 *)p = htole32(size); - p += sizeof(u32); + *(uint32_t *)p = htole32(size); + p += sizeof(uint32_t); - *(u32 *)p = htole32(csum); - p += sizeof(u32); + *(uint32_t *)p = htole32(csum); + p += sizeof(uint32_t); memcpy(p, BOOTCONFIG_MAGIC, BOOTCONFIG_MAGIC_LEN); p += BOOTCONFIG_MAGIC_LEN; -- cgit v1.2.3 From 4ee1b4cac236650979a5d9b745bb0a83efde3a46 Mon Sep 17 00:00:00 2001 From: Masami Hiramatsu Date: Fri, 17 Sep 2021 19:03:16 +0900 Subject: bootconfig: Cleanup dummy headers in tools/bootconfig Cleanup dummy headers in tools/bootconfig/include except for tools/bootconfig/include/linux/bootconfig.h. For this change, I use __KERNEL__ macro to split kernel header #include and introduce xbc_alloc_mem() and xbc_free_mem(). Link: https://lkml.kernel.org/r/163187299574.2366983.18371329724128746091.stgit@devnote2 Signed-off-by: Masami Hiramatsu Signed-off-by: Steven Rostedt (VMware) --- include/linux/bootconfig.h | 10 +++++++ lib/bootconfig.c | 43 +++++++++++++++++++++++---- tools/bootconfig/Makefile | 2 +- tools/bootconfig/include/linux/bootconfig.h | 45 ++++++++++++++++++++++++++++- tools/bootconfig/include/linux/bug.h | 12 -------- tools/bootconfig/include/linux/ctype.h | 7 ----- tools/bootconfig/include/linux/errno.h | 7 ----- tools/bootconfig/include/linux/kernel.h | 14 --------- tools/bootconfig/include/linux/memblock.h | 11 ------- tools/bootconfig/include/linux/string.h | 32 -------------------- tools/bootconfig/main.c | 1 - 11 files changed, 93 insertions(+), 91 deletions(-) delete mode 100644 tools/bootconfig/include/linux/bug.h delete mode 100644 tools/bootconfig/include/linux/ctype.h delete mode 100644 tools/bootconfig/include/linux/errno.h delete mode 100644 tools/bootconfig/include/linux/kernel.h delete mode 100644 tools/bootconfig/include/linux/memblock.h delete mode 100644 tools/bootconfig/include/linux/string.h (limited to 'lib') diff --git a/include/linux/bootconfig.h b/include/linux/bootconfig.h index a6f8dc51f168..a4665c7ab07c 100644 --- a/include/linux/bootconfig.h +++ b/include/linux/bootconfig.h @@ -7,8 +7,18 @@ * Author: Masami Hiramatsu */ +#ifdef __KERNEL__ #include #include +#else /* !__KERNEL__ */ +/* + * NOTE: This is only for tools/bootconfig, because tools/bootconfig will + * run the parser sanity test. + * This does NOT mean linux/bootconfig.h is available in the user space. + * However, if you change this file, please make sure the tools/bootconfig + * has no issue on building and running. + */ +#endif #define BOOTCONFIG_MAGIC "#BOOTCONFIG\n" #define BOOTCONFIG_MAGIC_LEN 12 diff --git a/lib/bootconfig.c b/lib/bootconfig.c index a2f5f582181d..a056ae137750 100644 --- a/lib/bootconfig.c +++ b/lib/bootconfig.c @@ -4,6 +4,7 @@ * Masami Hiramatsu */ +#ifdef __KERNEL__ #include #include #include @@ -11,6 +12,16 @@ #include #include #include +#else /* !__KERNEL__ */ +/* + * NOTE: This is only for tools/bootconfig, because tools/bootconfig will + * run the parser sanity test. + * This does NOT mean lib/bootconfig.c is available in the user space. + * However, if you change this file, please make sure the tools/bootconfig + * has no issue on building and running. + */ +#include +#endif /* * Extra Boot Config (XBC) is given as tree-structured ascii text of @@ -31,6 +42,29 @@ static int xbc_err_pos __initdata; static int open_brace[XBC_DEPTH_MAX] __initdata; static int brace_index __initdata; +#ifdef __KERNEL__ +static inline void *xbc_alloc_mem(size_t size) +{ + return memblock_alloc(size, SMP_CACHE_BYTES); +} + +static inline void xbc_free_mem(void *addr, size_t size) +{ + memblock_free_ptr(addr, size); +} + +#else /* !__KERNEL__ */ + +static inline void *xbc_alloc_mem(size_t size) +{ + return malloc(size); +} + +static inline void xbc_free_mem(void *addr, size_t size) +{ + free(addr); +} +#endif /** * xbc_get_info() - Get the information of loaded boot config * node_size: A pointer to store the number of nodes. @@ -859,11 +893,11 @@ static int __init xbc_parse_tree(void) */ void __init xbc_exit(void) { - memblock_free_ptr(xbc_data, xbc_data_size); + xbc_free_mem(xbc_data, xbc_data_size); xbc_data = NULL; xbc_data_size = 0; xbc_node_num = 0; - memblock_free_ptr(xbc_nodes, sizeof(struct xbc_node) * XBC_NODE_MAX); + xbc_free_mem(xbc_nodes, sizeof(struct xbc_node) * XBC_NODE_MAX); xbc_nodes = NULL; brace_index = 0; } @@ -902,7 +936,7 @@ int __init xbc_init(const char *data, size_t size, const char **emsg, int *epos) return -ERANGE; } - xbc_data = memblock_alloc(size + 1, SMP_CACHE_BYTES); + xbc_data = xbc_alloc_mem(size + 1); if (!xbc_data) { if (emsg) *emsg = "Failed to allocate bootconfig data"; @@ -912,8 +946,7 @@ int __init xbc_init(const char *data, size_t size, const char **emsg, int *epos) xbc_data[size] = '\0'; xbc_data_size = size + 1; - xbc_nodes = memblock_alloc(sizeof(struct xbc_node) * XBC_NODE_MAX, - SMP_CACHE_BYTES); + xbc_nodes = xbc_alloc_mem(sizeof(struct xbc_node) * XBC_NODE_MAX); if (!xbc_nodes) { if (emsg) *emsg = "Failed to allocate bootconfig nodes"; diff --git a/tools/bootconfig/Makefile b/tools/bootconfig/Makefile index f1eec3ccbe18..566c3e0ee561 100644 --- a/tools/bootconfig/Makefile +++ b/tools/bootconfig/Makefile @@ -17,7 +17,7 @@ ALL_PROGRAMS := $(patsubst %,$(OUTPUT)%,$(ALL_TARGETS)) all: $(ALL_PROGRAMS) test -$(OUTPUT)bootconfig: main.c $(LIBSRC) +$(OUTPUT)bootconfig: main.c include/linux/bootconfig.h $(LIBSRC) $(CC) $(filter %.c,$^) $(CFLAGS) -o $@ test: $(ALL_PROGRAMS) test-bootconfig.sh diff --git a/tools/bootconfig/include/linux/bootconfig.h b/tools/bootconfig/include/linux/bootconfig.h index de7f30f99af3..6784296a0692 100644 --- a/tools/bootconfig/include/linux/bootconfig.h +++ b/tools/bootconfig/include/linux/bootconfig.h @@ -2,10 +2,53 @@ #ifndef _BOOTCONFIG_LINUX_BOOTCONFIG_H #define _BOOTCONFIG_LINUX_BOOTCONFIG_H -#include "../../../../include/linux/bootconfig.h" +#include +#include +#include +#include +#include +#include +#include + #ifndef fallthrough # define fallthrough #endif +#define WARN_ON(cond) \ + ((cond) ? printf("Internal warning(%s:%d, %s): %s\n", \ + __FILE__, __LINE__, __func__, #cond) : 0) + +#define unlikely(cond) (cond) + +/* Copied from lib/string.c */ +static inline char *skip_spaces(const char *str) +{ + while (isspace(*str)) + ++str; + return (char *)str; +} + +static inline char *strim(char *s) +{ + size_t size; + char *end; + + size = strlen(s); + if (!size) + return s; + + end = s + size - 1; + while (end >= s && isspace(*end)) + end--; + *(end + 1) = '\0'; + + return skip_spaces(s); +} + +#define __init +#define __initdata + +#include "../../../../include/linux/bootconfig.h" + #endif diff --git a/tools/bootconfig/include/linux/bug.h b/tools/bootconfig/include/linux/bug.h deleted file mode 100644 index 7b65a389c0dd..000000000000 --- a/tools/bootconfig/include/linux/bug.h +++ /dev/null @@ -1,12 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0 */ -#ifndef _SKC_LINUX_BUG_H -#define _SKC_LINUX_BUG_H - -#include -#include - -#define WARN_ON(cond) \ - ((cond) ? printf("Internal warning(%s:%d, %s): %s\n", \ - __FILE__, __LINE__, __func__, #cond) : 0) - -#endif diff --git a/tools/bootconfig/include/linux/ctype.h b/tools/bootconfig/include/linux/ctype.h deleted file mode 100644 index c56ecc136448..000000000000 --- a/tools/bootconfig/include/linux/ctype.h +++ /dev/null @@ -1,7 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0 */ -#ifndef _SKC_LINUX_CTYPE_H -#define _SKC_LINUX_CTYPE_H - -#include - -#endif diff --git a/tools/bootconfig/include/linux/errno.h b/tools/bootconfig/include/linux/errno.h deleted file mode 100644 index 5d9f91ec2fda..000000000000 --- a/tools/bootconfig/include/linux/errno.h +++ /dev/null @@ -1,7 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0 */ -#ifndef _SKC_LINUX_ERRNO_H -#define _SKC_LINUX_ERRNO_H - -#include - -#endif diff --git a/tools/bootconfig/include/linux/kernel.h b/tools/bootconfig/include/linux/kernel.h deleted file mode 100644 index 39f306c18dd0..000000000000 --- a/tools/bootconfig/include/linux/kernel.h +++ /dev/null @@ -1,14 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0 */ -#ifndef _SKC_LINUX_KERNEL_H -#define _SKC_LINUX_KERNEL_H - -#include -#include -#include - -#define unlikely(cond) (cond) - -#define __init -#define __initdata - -#endif diff --git a/tools/bootconfig/include/linux/memblock.h b/tools/bootconfig/include/linux/memblock.h deleted file mode 100644 index f2e506f7d57f..000000000000 --- a/tools/bootconfig/include/linux/memblock.h +++ /dev/null @@ -1,11 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0 */ -#ifndef _XBC_LINUX_MEMBLOCK_H -#define _XBC_LINUX_MEMBLOCK_H - -#include - -#define SMP_CACHE_BYTES 0 -#define memblock_alloc(size, align) malloc(size) -#define memblock_free_ptr(paddr, size) free(paddr) - -#endif diff --git a/tools/bootconfig/include/linux/string.h b/tools/bootconfig/include/linux/string.h deleted file mode 100644 index 8267af75153a..000000000000 --- a/tools/bootconfig/include/linux/string.h +++ /dev/null @@ -1,32 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0 */ -#ifndef _SKC_LINUX_STRING_H -#define _SKC_LINUX_STRING_H - -#include - -/* Copied from lib/string.c */ -static inline char *skip_spaces(const char *str) -{ - while (isspace(*str)) - ++str; - return (char *)str; -} - -static inline char *strim(char *s) -{ - size_t size; - char *end; - - size = strlen(s); - if (!size) - return s; - - end = s + size - 1; - while (end >= s && isspace(*end)) - end--; - *(end + 1) = '\0'; - - return skip_spaces(s); -} - -#endif diff --git a/tools/bootconfig/main.c b/tools/bootconfig/main.c index fb7c9fb953d7..156b62a163c5 100644 --- a/tools/bootconfig/main.c +++ b/tools/bootconfig/main.c @@ -12,7 +12,6 @@ #include #include -#include #include #define pr_err(fmt, ...) fprintf(stderr, fmt, ##__VA_ARGS__) -- cgit v1.2.3 From 814a66741b9ffb5e1ba119e368b178edb0b7322d Mon Sep 17 00:00:00 2001 From: Andreas Gruenbacher Date: Wed, 21 Jul 2021 19:03:47 +0200 Subject: iov_iter: Fix iov_iter_get_pages{,_alloc} page fault return value Both iov_iter_get_pages and iov_iter_get_pages_alloc return the number of bytes of the iovec they could get the pages for. When they cannot get any pages, they're supposed to return 0, but when the start of the iovec isn't page aligned, the calculation goes wrong and they return a negative value. Fix both functions. In addition, change iov_iter_get_pages_alloc to return NULL in that case to prevent resource leaks. Signed-off-by: Andreas Gruenbacher Reviewed-by: Christoph Hellwig --- lib/iov_iter.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) (limited to 'lib') diff --git a/lib/iov_iter.c b/lib/iov_iter.c index 755c10c5138c..60b5e6edfbaa 100644 --- a/lib/iov_iter.c +++ b/lib/iov_iter.c @@ -1488,7 +1488,7 @@ ssize_t iov_iter_get_pages(struct iov_iter *i, res = get_user_pages_fast(addr, n, iov_iter_rw(i) != WRITE ? FOLL_WRITE : 0, pages); - if (unlikely(res < 0)) + if (unlikely(res <= 0)) return res; return (res == n ? len : res * PAGE_SIZE) - *start; } @@ -1612,8 +1612,9 @@ ssize_t iov_iter_get_pages_alloc(struct iov_iter *i, return -ENOMEM; res = get_user_pages_fast(addr, n, iov_iter_rw(i) != WRITE ? FOLL_WRITE : 0, p); - if (unlikely(res < 0)) { + if (unlikely(res <= 0)) { kvfree(p); + *pages = NULL; return res; } *pages = p; -- cgit v1.2.3 From 216a0fc40897d87b4a282bba10b01e96ea6a1c45 Mon Sep 17 00:00:00 2001 From: Jim Cromie Date: Tue, 12 Oct 2021 12:33:06 -0600 Subject: dyndbg: show module in vpr-info in dd-exec-queries dynamic_debug_exec_queries() accepts a separate module arg (so it can support $module.dyndbg boot arg), display that in the vpr-info for a more useful user-debug context. Signed-off-by: Jim Cromie Link: https://lore.kernel.org/r/20211012183310.1016678-2-jim.cromie@gmail.com Signed-off-by: Greg Kroah-Hartman --- lib/dynamic_debug.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'lib') diff --git a/lib/dynamic_debug.c b/lib/dynamic_debug.c index cb5abb42c16a..dfe1e6a857bc 100644 --- a/lib/dynamic_debug.c +++ b/lib/dynamic_debug.c @@ -529,7 +529,7 @@ static int ddebug_exec_queries(char *query, const char *modname) if (!query || !*query || *query == '#') continue; - vpr_info("query %d: \"%s\"\n", i, query); + v2pr_info("query %d: \"%s\" mod:%s\n", i, query, modname ?: "*"); rc = ddebug_exec_query(query, modname); if (rc < 0) { -- cgit v1.2.3 From 2a12e0003580505b8e7d82f9a8fef95f4a1031a8 Mon Sep 17 00:00:00 2001 From: Len Baker Date: Sun, 19 Sep 2021 13:09:13 +0200 Subject: assoc_array: Avoid open coded arithmetic in allocator arguments As noted in the "Deprecated Interfaces, Language Features, Attributes, and Conventions" documentation [1], size calculations (especially multiplication) should not be performed in memory allocator (or similar) function arguments due to the risk of them overflowing. This could lead to values wrapping around and a smaller allocation being made than the caller was expecting. Using those allocations could lead to linear overflows of heap memory and other misbehaviors. So, use the struct_size() helper to do the arithmetic instead of the argument "size + count * size" in the kmalloc() and kzalloc() functions. Also, take the opportunity to refactor the memcpy() calls to use the struct_size() and flex_array_size() helpers. [1] https://www.kernel.org/doc/html/latest/process/deprecated.html#open-coded-arithmetic-in-allocator-arguments Signed-off-by: Len Baker Reviewed-by: Gustavo A. R. Silva Signed-off-by: Gustavo A. R. Silva --- lib/assoc_array.c | 22 ++++++++++------------ 1 file changed, 10 insertions(+), 12 deletions(-) (limited to 'lib') diff --git a/lib/assoc_array.c b/lib/assoc_array.c index 04c98799c3ba..079c72e26493 100644 --- a/lib/assoc_array.c +++ b/lib/assoc_array.c @@ -741,8 +741,7 @@ all_leaves_cluster_together: keylen = round_up(diff, ASSOC_ARRAY_KEY_CHUNK_SIZE); keylen >>= ASSOC_ARRAY_KEY_CHUNK_SHIFT; - new_s0 = kzalloc(sizeof(struct assoc_array_shortcut) + - keylen * sizeof(unsigned long), GFP_KERNEL); + new_s0 = kzalloc(struct_size(new_s0, index_key, keylen), GFP_KERNEL); if (!new_s0) return false; edit->new_meta[2] = assoc_array_shortcut_to_ptr(new_s0); @@ -849,8 +848,8 @@ static bool assoc_array_insert_mid_shortcut(struct assoc_array_edit *edit, keylen = round_up(diff, ASSOC_ARRAY_KEY_CHUNK_SIZE); keylen >>= ASSOC_ARRAY_KEY_CHUNK_SHIFT; - new_s0 = kzalloc(sizeof(struct assoc_array_shortcut) + - keylen * sizeof(unsigned long), GFP_KERNEL); + new_s0 = kzalloc(struct_size(new_s0, index_key, keylen), + GFP_KERNEL); if (!new_s0) return false; edit->new_meta[1] = assoc_array_shortcut_to_ptr(new_s0); @@ -864,7 +863,7 @@ static bool assoc_array_insert_mid_shortcut(struct assoc_array_edit *edit, new_n0->parent_slot = 0; memcpy(new_s0->index_key, shortcut->index_key, - keylen * sizeof(unsigned long)); + flex_array_size(new_s0, index_key, keylen)); blank = ULONG_MAX << (diff & ASSOC_ARRAY_KEY_CHUNK_MASK); pr_devel("blank off [%zu] %d: %lx\n", keylen - 1, diff, blank); @@ -899,8 +898,8 @@ static bool assoc_array_insert_mid_shortcut(struct assoc_array_edit *edit, keylen = round_up(shortcut->skip_to_level, ASSOC_ARRAY_KEY_CHUNK_SIZE); keylen >>= ASSOC_ARRAY_KEY_CHUNK_SHIFT; - new_s1 = kzalloc(sizeof(struct assoc_array_shortcut) + - keylen * sizeof(unsigned long), GFP_KERNEL); + new_s1 = kzalloc(struct_size(new_s1, index_key, keylen), + GFP_KERNEL); if (!new_s1) return false; edit->new_meta[2] = assoc_array_shortcut_to_ptr(new_s1); @@ -913,7 +912,7 @@ static bool assoc_array_insert_mid_shortcut(struct assoc_array_edit *edit, new_n0->slots[sc_slot] = assoc_array_shortcut_to_ptr(new_s1); memcpy(new_s1->index_key, shortcut->index_key, - keylen * sizeof(unsigned long)); + flex_array_size(new_s1, index_key, keylen)); edit->set[1].ptr = &side->back_pointer; edit->set[1].to = assoc_array_shortcut_to_ptr(new_s1); @@ -1490,13 +1489,12 @@ descend: shortcut = assoc_array_ptr_to_shortcut(cursor); keylen = round_up(shortcut->skip_to_level, ASSOC_ARRAY_KEY_CHUNK_SIZE); keylen >>= ASSOC_ARRAY_KEY_CHUNK_SHIFT; - new_s = kmalloc(sizeof(struct assoc_array_shortcut) + - keylen * sizeof(unsigned long), GFP_KERNEL); + new_s = kmalloc(struct_size(new_s, index_key, keylen), + GFP_KERNEL); if (!new_s) goto enomem; pr_devel("dup shortcut %p -> %p\n", shortcut, new_s); - memcpy(new_s, shortcut, (sizeof(struct assoc_array_shortcut) + - keylen * sizeof(unsigned long))); + memcpy(new_s, shortcut, struct_size(new_s, index_key, keylen)); new_s->back_pointer = new_parent; new_s->parent_slot = shortcut->parent_slot; *new_ptr_pp = new_parent = assoc_array_shortcut_to_ptr(new_s); -- cgit v1.2.3 From 5ca173974888368fecfb17ae6fe455df5fd2a9d2 Mon Sep 17 00:00:00 2001 From: Andrew Halaney Date: Wed, 13 Oct 2021 11:40:20 -0400 Subject: dyndbg: make dyndbg a known cli param Right now dyndbg shows up as an unknown parameter if used on boot: Unknown command line parameters: dyndbg=+p That's because it is unknown, it doesn't sit in the __param section, so the processing done to warn users supplying an unknown parameter doesn't think it is legitimate. Install a dummy handler to register it. dynamic debug needs to search the whole command line for modules listed that are currently builtin, so there's no real work to be done in this callback. Fixes: 86d1919a4fb0 ("init: print out unknown kernel parameters") Tested-by: Jim Cromie Signed-off-by: Andrew Halaney Signed-off-by: Jason Baron Link: https://lore.kernel.org/r/1634139622-20667-2-git-send-email-jbaron@akamai.com Signed-off-by: Greg Kroah-Hartman --- lib/dynamic_debug.c | 12 ++++++++++++ 1 file changed, 12 insertions(+) (limited to 'lib') diff --git a/lib/dynamic_debug.c b/lib/dynamic_debug.c index dfe1e6a857bc..cd222a1c1238 100644 --- a/lib/dynamic_debug.c +++ b/lib/dynamic_debug.c @@ -761,6 +761,18 @@ static __init int ddebug_setup_query(char *str) __setup("ddebug_query=", ddebug_setup_query); +/* + * Install a noop handler to make dyndbg look like a normal kernel cli param. + * This avoids warnings about dyndbg being an unknown cli param when supplied + * by a user. + */ +static __init int dyndbg_setup(char *str) +{ + return 1; +} + +__setup("dyndbg=", dyndbg_setup); + /* * File_ops->write method for /dynamic_debug/control. Gathers the * command text from userspace, parses and executes it. -- cgit v1.2.3 From 9c40e1aa84123750773a57c9cf39112459a952dd Mon Sep 17 00:00:00 2001 From: Andrew Halaney Date: Wed, 13 Oct 2021 11:40:21 -0400 Subject: dyndbg: Remove support for ddebug_query param This param has been deprecated for a very long time now, let's rip it out. Signed-off-by: Andrew Halaney Signed-off-by: Jason Baron Link: https://lore.kernel.org/r/1634139622-20667-3-git-send-email-jbaron@akamai.com Signed-off-by: Greg Kroah-Hartman --- Documentation/admin-guide/dynamic-debug-howto.rst | 6 ++---- Documentation/admin-guide/kernel-parameters.txt | 5 ----- lib/dynamic_debug.c | 25 ----------------------- 3 files changed, 2 insertions(+), 34 deletions(-) (limited to 'lib') diff --git a/Documentation/admin-guide/dynamic-debug-howto.rst b/Documentation/admin-guide/dynamic-debug-howto.rst index b119b8277b3e..d0911e7cc271 100644 --- a/Documentation/admin-guide/dynamic-debug-howto.rst +++ b/Documentation/admin-guide/dynamic-debug-howto.rst @@ -249,8 +249,7 @@ Debug messages during Boot Process To activate debug messages for core code and built-in modules during the boot process, even before userspace and debugfs exists, use -``dyndbg="QUERY"``, ``module.dyndbg="QUERY"``, or ``ddebug_query="QUERY"`` -(``ddebug_query`` is obsoleted by ``dyndbg``, and deprecated). QUERY follows +``dyndbg="QUERY"`` or ``module.dyndbg="QUERY"``. QUERY follows the syntax described above, but must not exceed 1023 characters. Your bootloader may impose lower limits. @@ -270,8 +269,7 @@ this boot parameter for debugging purposes. If ``foo`` module is not built-in, ``foo.dyndbg`` will still be processed at boot time, without effect, but will be reprocessed when module is -loaded later. ``ddebug_query=`` and bare ``dyndbg=`` are only processed at -boot. +loaded later. Bare ``dyndbg=`` is only processed at boot. Debug Messages at Module Initialization Time diff --git a/Documentation/admin-guide/kernel-parameters.txt b/Documentation/admin-guide/kernel-parameters.txt index 91ba391f9b32..9c0c57d1ca78 100644 --- a/Documentation/admin-guide/kernel-parameters.txt +++ b/Documentation/admin-guide/kernel-parameters.txt @@ -841,11 +841,6 @@ Format: , See also Documentation/input/devices/joystick-parport.rst - ddebug_query= [KNL,DYNAMIC_DEBUG] Enable debug messages at early boot - time. See - Documentation/admin-guide/dynamic-debug-howto.rst for - details. Deprecated, see dyndbg. - debug [KNL] Enable kernel debugging (events log level). debug_boot_weak_hash diff --git a/lib/dynamic_debug.c b/lib/dynamic_debug.c index cd222a1c1238..a44ae0b4b022 100644 --- a/lib/dynamic_debug.c +++ b/lib/dynamic_debug.c @@ -746,21 +746,6 @@ EXPORT_SYMBOL(__dynamic_ibdev_dbg); #endif -#define DDEBUG_STRING_SIZE 1024 -static __initdata char ddebug_setup_string[DDEBUG_STRING_SIZE]; - -static __init int ddebug_setup_query(char *str) -{ - if (strlen(str) >= DDEBUG_STRING_SIZE) { - pr_warn("ddebug boot param string too large\n"); - return 0; - } - strlcpy(ddebug_setup_string, str, DDEBUG_STRING_SIZE); - return 1; -} - -__setup("ddebug_query=", ddebug_setup_query); - /* * Install a noop handler to make dyndbg look like a normal kernel cli param. * This avoids warnings about dyndbg being an unknown cli param when supplied @@ -1133,16 +1118,6 @@ static int __init dynamic_debug_init(void) entries, modct, (int)((modct * sizeof(struct ddebug_table)) >> 10), (int)((entries * sizeof(struct _ddebug)) >> 10)); - /* apply ddebug_query boot param, dont unload tables on err */ - if (ddebug_setup_string[0] != '\0') { - pr_warn("ddebug_query param name is deprecated, change it to dyndbg\n"); - ret = ddebug_exec_queries(ddebug_setup_string, NULL); - if (ret < 0) - pr_warn("Invalid ddebug boot param %s\n", - ddebug_setup_string); - else - pr_info("%d changes by ddebug_query\n", ret); - } /* now that ddebug tables are loaded, process all boot args * again to find and activate queries given in dyndbg params. * While this has already been done for known boot params, it -- cgit v1.2.3 From 7a5e202dfb8ab7572eba703e43f08dabdc8b7808 Mon Sep 17 00:00:00 2001 From: Jim Cromie Date: Wed, 13 Oct 2021 16:07:25 -0600 Subject: dyndbg: vpr-info on remove-module complete, not starting On qemu --smp 3 runs, remove-module can get called 3 times. So don't print on entry; instead print "removed" after entry is found and removed, so just once. Signed-off-by: Jim Cromie Link: https://lore.kernel.org/r/20211013220726.1280565-3-jim.cromie@gmail.com Signed-off-by: Greg Kroah-Hartman --- lib/dynamic_debug.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'lib') diff --git a/lib/dynamic_debug.c b/lib/dynamic_debug.c index a44ae0b4b022..d706abba8dad 100644 --- a/lib/dynamic_debug.c +++ b/lib/dynamic_debug.c @@ -1025,8 +1025,6 @@ int ddebug_remove_module(const char *mod_name) struct ddebug_table *dt, *nextdt; int ret = -ENOENT; - v2pr_info("removing module \"%s\"\n", mod_name); - mutex_lock(&ddebug_lock); list_for_each_entry_safe(dt, nextdt, &ddebug_tables, link) { if (dt->mod_name == mod_name) { @@ -1036,6 +1034,8 @@ int ddebug_remove_module(const char *mod_name) } } mutex_unlock(&ddebug_lock); + if (!ret) + v2pr_info("removed module \"%s\"\n", mod_name); return ret; } -- cgit v1.2.3 From 7edde0c807855a5bf149c6be3edfa11ecd919571 Mon Sep 17 00:00:00 2001 From: Jim Cromie Date: Wed, 13 Oct 2021 16:07:26 -0600 Subject: dyndbg: no vpr-info on empty queries when `echo $cmd > control` contains multiple queries, extra query separators (;\n) can parse as empty statements. This is normal, and a vpr-info on an empty command is just noise. Signed-off-by: Jim Cromie Link: https://lore.kernel.org/r/20211013220726.1280565-4-jim.cromie@gmail.com Signed-off-by: Greg Kroah-Hartman --- lib/dynamic_debug.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) (limited to 'lib') diff --git a/lib/dynamic_debug.c b/lib/dynamic_debug.c index d706abba8dad..ad80f1346c3f 100644 --- a/lib/dynamic_debug.c +++ b/lib/dynamic_debug.c @@ -540,8 +540,10 @@ static int ddebug_exec_queries(char *query, const char *modname) } i++; } - vpr_info("processed %d queries, with %d matches, %d errs\n", - i, nfound, errs); + + if (i) + vpr_info("processed %d queries, with %d matches, %d errs\n", + i, nfound, errs); if (exitcode) return exitcode; -- cgit v1.2.3 From 1f8818e352f721c49ebea39025f6c98f25756eff Mon Sep 17 00:00:00 2001 From: Jim Cromie Date: Thu, 14 Oct 2021 16:36:14 -0600 Subject: dyndbg: fix spurious vNpr_info change The cited commit inadvertently altered the verbose level of a vpr_info, restore it to original. Fixes: 216a0fc40897 ("dyndbg: show module in vpr-info in dd-exec-queries") Signed-off-By: Jim Cromie Link: https://lore.kernel.org/r/20211014223614.1952171-1-jim.cromie@gmail.com Signed-off-by: Greg Kroah-Hartman --- lib/dynamic_debug.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'lib') diff --git a/lib/dynamic_debug.c b/lib/dynamic_debug.c index ad80f1346c3f..87b1b0121234 100644 --- a/lib/dynamic_debug.c +++ b/lib/dynamic_debug.c @@ -529,7 +529,7 @@ static int ddebug_exec_queries(char *query, const char *modname) if (!query || !*query || *query == '#') continue; - v2pr_info("query %d: \"%s\" mod:%s\n", i, query, modname ?: "*"); + vpr_info("query %d: \"%s\" mod:%s\n", i, query, modname ?: "*"); rc = ddebug_exec_query(query, modname); if (rc < 0) { -- cgit v1.2.3 From be5f1797523004d0d9aaee81a523d86e3b890007 Mon Sep 17 00:00:00 2001 From: "Matthew Wilcox (Oracle)" Date: Sat, 20 Mar 2021 16:34:54 -0400 Subject: flex_proportions: Allow N events instead of 1 When batching events (such as writing back N pages in a single I/O), it is better to do one flex_proportion operation instead of N. There is only one caller of __fprop_inc_percpu_max(), and it's the one we're going to change in the next patch, so rename it instead of adding a compatibility wrapper. Signed-off-by: Matthew Wilcox (Oracle) Reviewed-by: Christoph Hellwig Reviewed-by: Jan Kara --- include/linux/flex_proportions.h | 9 +++++---- lib/flex_proportions.c | 28 +++++++++++++++++++--------- mm/page-writeback.c | 4 ++-- 3 files changed, 26 insertions(+), 15 deletions(-) (limited to 'lib') diff --git a/include/linux/flex_proportions.h b/include/linux/flex_proportions.h index c12df59d3f5f..3e378b1fb0bc 100644 --- a/include/linux/flex_proportions.h +++ b/include/linux/flex_proportions.h @@ -83,9 +83,10 @@ struct fprop_local_percpu { int fprop_local_init_percpu(struct fprop_local_percpu *pl, gfp_t gfp); void fprop_local_destroy_percpu(struct fprop_local_percpu *pl); -void __fprop_inc_percpu(struct fprop_global *p, struct fprop_local_percpu *pl); -void __fprop_inc_percpu_max(struct fprop_global *p, struct fprop_local_percpu *pl, - int max_frac); +void __fprop_add_percpu(struct fprop_global *p, struct fprop_local_percpu *pl, + long nr); +void __fprop_add_percpu_max(struct fprop_global *p, + struct fprop_local_percpu *pl, int max_frac, long nr); void fprop_fraction_percpu(struct fprop_global *p, struct fprop_local_percpu *pl, unsigned long *numerator, unsigned long *denominator); @@ -96,7 +97,7 @@ void fprop_inc_percpu(struct fprop_global *p, struct fprop_local_percpu *pl) unsigned long flags; local_irq_save(flags); - __fprop_inc_percpu(p, pl); + __fprop_add_percpu(p, pl, 1); local_irq_restore(flags); } diff --git a/lib/flex_proportions.c b/lib/flex_proportions.c index 451543937524..53e7eb1dd76c 100644 --- a/lib/flex_proportions.c +++ b/lib/flex_proportions.c @@ -217,11 +217,12 @@ static void fprop_reflect_period_percpu(struct fprop_global *p, } /* Event of type pl happened */ -void __fprop_inc_percpu(struct fprop_global *p, struct fprop_local_percpu *pl) +void __fprop_add_percpu(struct fprop_global *p, struct fprop_local_percpu *pl, + long nr) { fprop_reflect_period_percpu(p, pl); - percpu_counter_add_batch(&pl->events, 1, PROP_BATCH); - percpu_counter_add(&p->events, 1); + percpu_counter_add_batch(&pl->events, nr, PROP_BATCH); + percpu_counter_add(&p->events, nr); } void fprop_fraction_percpu(struct fprop_global *p, @@ -253,20 +254,29 @@ void fprop_fraction_percpu(struct fprop_global *p, } /* - * Like __fprop_inc_percpu() except that event is counted only if the given + * Like __fprop_add_percpu() except that event is counted only if the given * type has fraction smaller than @max_frac/FPROP_FRAC_BASE */ -void __fprop_inc_percpu_max(struct fprop_global *p, - struct fprop_local_percpu *pl, int max_frac) +void __fprop_add_percpu_max(struct fprop_global *p, + struct fprop_local_percpu *pl, int max_frac, long nr) { if (unlikely(max_frac < FPROP_FRAC_BASE)) { unsigned long numerator, denominator; + s64 tmp; fprop_fraction_percpu(p, pl, &numerator, &denominator); - if (numerator > - (((u64)denominator) * max_frac) >> FPROP_FRAC_SHIFT) + /* Adding 'nr' to fraction exceeds max_frac/FPROP_FRAC_BASE? */ + tmp = (u64)denominator * max_frac - + ((u64)numerator << FPROP_FRAC_SHIFT); + if (tmp < 0) { + /* Maximum fraction already exceeded? */ return; + } else if (tmp < nr * (FPROP_FRAC_BASE - max_frac)) { + /* Add just enough for the fraction to saturate */ + nr = div_u64(tmp + FPROP_FRAC_BASE - max_frac - 1, + FPROP_FRAC_BASE - max_frac); + } } - __fprop_inc_percpu(p, pl); + __fprop_add_percpu(p, pl, nr); } diff --git a/mm/page-writeback.c b/mm/page-writeback.c index 1d8f2ee2e065..cb7387d0e77d 100644 --- a/mm/page-writeback.c +++ b/mm/page-writeback.c @@ -566,8 +566,8 @@ static void wb_domain_writeout_inc(struct wb_domain *dom, struct fprop_local_percpu *completions, unsigned int max_prop_frac) { - __fprop_inc_percpu_max(&dom->completions, completions, - max_prop_frac); + __fprop_add_percpu_max(&dom->completions, completions, + max_prop_frac, 1); /* First event after period switching was turned off? */ if (unlikely(!dom->period_time)) { /* -- cgit v1.2.3 From 348332e000697b4ca82ef96719e02876434b8346 Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Mon, 20 Sep 2021 14:33:12 +0200 Subject: mm: don't include in blk-cgroup.h pulls in blkdev.h and thus pretty much all the block headers. Break this dependency chain by turning wbc_blkcg_css into a macro and dropping the blk-cgroup.h include. Signed-off-by: Christoph Hellwig Reviewed-by: Johannes Thumshirn Link: https://lore.kernel.org/r/20210920123328.1399408-2-hch@lst.de Signed-off-by: Jens Axboe --- drivers/gpu/drm/i915/i915_utils.h | 1 + fs/btrfs/inode.c | 1 + fs/quota/quota.c | 1 + include/linux/writeback.h | 14 +++++--------- lib/random32.c | 1 + 5 files changed, 9 insertions(+), 9 deletions(-) (limited to 'lib') diff --git a/drivers/gpu/drm/i915/i915_utils.h b/drivers/gpu/drm/i915/i915_utils.h index 5259edacde38..066a9118c374 100644 --- a/drivers/gpu/drm/i915/i915_utils.h +++ b/drivers/gpu/drm/i915/i915_utils.h @@ -30,6 +30,7 @@ #include #include #include +#include struct drm_i915_private; struct timer_list; diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 487533c35ddb..4a9077c52444 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -6,6 +6,7 @@ #include #include #include +#include #include #include #include diff --git a/fs/quota/quota.c b/fs/quota/quota.c index 2bcc9a6f1bfc..052f143e2e0e 100644 --- a/fs/quota/quota.c +++ b/fs/quota/quota.c @@ -10,6 +10,7 @@ #include #include #include +#include #include #include #include diff --git a/include/linux/writeback.h b/include/linux/writeback.h index d1f65adf6a26..8eb165760752 100644 --- a/include/linux/writeback.h +++ b/include/linux/writeback.h @@ -11,7 +11,6 @@ #include #include #include -#include struct bio; @@ -109,15 +108,12 @@ static inline int wbc_to_write_flags(struct writeback_control *wbc) return flags; } -static inline struct cgroup_subsys_state * -wbc_blkcg_css(struct writeback_control *wbc) -{ #ifdef CONFIG_CGROUP_WRITEBACK - if (wbc->wb) - return wbc->wb->blkcg_css; -#endif - return blkcg_root_css; -} +#define wbc_blkcg_css(wbc) \ + ((wbc)->wb ? (wbc)->wb->blkcg_css : blkcg_root_css) +#else +#define wbc_blkcg_css(wbc) (blkcg_root_css) +#endif /* CONFIG_CGROUP_WRITEBACK */ /* * A wb_domain represents a domain that wb's (bdi_writeback's) belong to diff --git a/lib/random32.c b/lib/random32.c index 4d0e05e471d7..a57a0e18819d 100644 --- a/lib/random32.c +++ b/lib/random32.c @@ -39,6 +39,7 @@ #include #include #include +#include #include #include -- cgit v1.2.3 From 9672b0d43782047b1825a96bafee1b6aefa35bc2 Mon Sep 17 00:00:00 2001 From: Jens Axboe Date: Sat, 9 Oct 2021 13:02:23 -0600 Subject: sbitmap: add __sbitmap_queue_get_batch() The block layer tag allocation batching still calls into sbitmap to get each tag, but we can improve on that. Add __sbitmap_queue_get_batch(), which returns a mask of tags all at once, along with an offset for those tags. An example return would be 0xff, where bits 0..7 are set, with tag_offset == 128. The valid tags in this case would be 128..135. A batch is specific to an individual sbitmap_map, hence it cannot be larger than that. The requested number of tags is automatically reduced to the max that can be satisfied with a single map. On failure, 0 is returned. Caller should fall back to single tag allocation at that point/ Signed-off-by: Jens Axboe --- include/linux/sbitmap.h | 13 +++++++++++++ lib/sbitmap.c | 51 +++++++++++++++++++++++++++++++++++++++++++++++++ 2 files changed, 64 insertions(+) (limited to 'lib') diff --git a/include/linux/sbitmap.h b/include/linux/sbitmap.h index 2713e689ad66..e30b56023ead 100644 --- a/include/linux/sbitmap.h +++ b/include/linux/sbitmap.h @@ -426,6 +426,19 @@ void sbitmap_queue_resize(struct sbitmap_queue *sbq, unsigned int depth); */ int __sbitmap_queue_get(struct sbitmap_queue *sbq); +/** + * __sbitmap_queue_get_batch() - Try to allocate a batch of free bits + * @sbq: Bitmap queue to allocate from. + * @nr_tags: number of tags requested + * @offset: offset to add to returned bits + * + * Return: Mask of allocated tags, 0 if none are found. Each tag allocated is + * a bit in the mask returned, and the caller must add @offset to the value to + * get the absolute tag value. + */ +unsigned long __sbitmap_queue_get_batch(struct sbitmap_queue *sbq, int nr_tags, + unsigned int *offset); + /** * __sbitmap_queue_get_shallow() - Try to allocate a free bit from a &struct * sbitmap_queue, limiting the depth used from each word, with preemption diff --git a/lib/sbitmap.c b/lib/sbitmap.c index b25db9be938a..f398e0ae548e 100644 --- a/lib/sbitmap.c +++ b/lib/sbitmap.c @@ -489,6 +489,57 @@ int __sbitmap_queue_get(struct sbitmap_queue *sbq) } EXPORT_SYMBOL_GPL(__sbitmap_queue_get); +unsigned long __sbitmap_queue_get_batch(struct sbitmap_queue *sbq, int nr_tags, + unsigned int *offset) +{ + struct sbitmap *sb = &sbq->sb; + unsigned int hint, depth; + unsigned long index, nr; + int i; + + if (unlikely(sb->round_robin)) + return 0; + + depth = READ_ONCE(sb->depth); + hint = update_alloc_hint_before_get(sb, depth); + + index = SB_NR_TO_INDEX(sb, hint); + + for (i = 0; i < sb->map_nr; i++) { + struct sbitmap_word *map = &sb->map[index]; + unsigned long get_mask; + + sbitmap_deferred_clear(map); + if (map->word == (1UL << (map->depth - 1)) - 1) + continue; + + nr = find_first_zero_bit(&map->word, map->depth); + if (nr + nr_tags <= map->depth) { + atomic_long_t *ptr = (atomic_long_t *) &map->word; + int map_tags = min_t(int, nr_tags, map->depth); + unsigned long val, ret; + + get_mask = ((1UL << map_tags) - 1) << nr; + do { + val = READ_ONCE(map->word); + ret = atomic_long_cmpxchg(ptr, val, get_mask | val); + } while (ret != val); + get_mask = (get_mask & ~ret) >> nr; + if (get_mask) { + *offset = nr + (index << sb->shift); + update_alloc_hint_after_get(sb, depth, hint, + *offset + map_tags - 1); + return get_mask; + } + } + /* Jump to next index. */ + if (++index >= sb->map_nr) + index = 0; + } + + return 0; +} + int __sbitmap_queue_get_shallow(struct sbitmap_queue *sbq, unsigned int shallow_depth) { -- cgit v1.2.3 From bb523b406c849eef8f265a07cd7f320f1f177743 Mon Sep 17 00:00:00 2001 From: Andreas Gruenbacher Date: Mon, 2 Aug 2021 13:44:20 +0200 Subject: gup: Turn fault_in_pages_{readable,writeable} into fault_in_{readable,writeable} Turn fault_in_pages_{readable,writeable} into versions that return the number of bytes not faulted in, similar to copy_to_user, instead of returning a non-zero value when any of the requested pages couldn't be faulted in. This supports the existing users that require all pages to be faulted in as well as new users that are happy if any pages can be faulted in. Rename the functions to fault_in_{readable,writeable} to make sure this change doesn't silently break things. Neither of these functions is entirely trivial and it doesn't seem useful to inline them, so move them to mm/gup.c. Signed-off-by: Andreas Gruenbacher --- arch/powerpc/kernel/kvm.c | 3 +- arch/powerpc/kernel/signal_32.c | 4 +-- arch/powerpc/kernel/signal_64.c | 2 +- arch/x86/kernel/fpu/signal.c | 7 ++-- drivers/gpu/drm/armada/armada_gem.c | 7 ++-- fs/btrfs/ioctl.c | 5 ++- include/linux/pagemap.h | 57 ++--------------------------- lib/iov_iter.c | 10 +++--- mm/filemap.c | 2 +- mm/gup.c | 72 +++++++++++++++++++++++++++++++++++++ 10 files changed, 93 insertions(+), 76 deletions(-) (limited to 'lib') diff --git a/arch/powerpc/kernel/kvm.c b/arch/powerpc/kernel/kvm.c index d89cf802d9aa..6568823cf306 100644 --- a/arch/powerpc/kernel/kvm.c +++ b/arch/powerpc/kernel/kvm.c @@ -669,7 +669,8 @@ static void __init kvm_use_magic_page(void) on_each_cpu(kvm_map_magic_page, &features, 1); /* Quick self-test to see if the mapping works */ - if (fault_in_pages_readable((const char *)KVM_MAGIC_PAGE, sizeof(u32))) { + if (fault_in_readable((const char __user *)KVM_MAGIC_PAGE, + sizeof(u32))) { kvm_patching_worked = false; return; } diff --git a/arch/powerpc/kernel/signal_32.c b/arch/powerpc/kernel/signal_32.c index 0608581967f0..38c3eae40c14 100644 --- a/arch/powerpc/kernel/signal_32.c +++ b/arch/powerpc/kernel/signal_32.c @@ -1048,7 +1048,7 @@ SYSCALL_DEFINE3(swapcontext, struct ucontext __user *, old_ctx, if (new_ctx == NULL) return 0; if (!access_ok(new_ctx, ctx_size) || - fault_in_pages_readable((u8 __user *)new_ctx, ctx_size)) + fault_in_readable((char __user *)new_ctx, ctx_size)) return -EFAULT; /* @@ -1237,7 +1237,7 @@ SYSCALL_DEFINE3(debug_setcontext, struct ucontext __user *, ctx, #endif if (!access_ok(ctx, sizeof(*ctx)) || - fault_in_pages_readable((u8 __user *)ctx, sizeof(*ctx))) + fault_in_readable((char __user *)ctx, sizeof(*ctx))) return -EFAULT; /* diff --git a/arch/powerpc/kernel/signal_64.c b/arch/powerpc/kernel/signal_64.c index 1831bba0582e..9f471b4a11e3 100644 --- a/arch/powerpc/kernel/signal_64.c +++ b/arch/powerpc/kernel/signal_64.c @@ -688,7 +688,7 @@ SYSCALL_DEFINE3(swapcontext, struct ucontext __user *, old_ctx, if (new_ctx == NULL) return 0; if (!access_ok(new_ctx, ctx_size) || - fault_in_pages_readable((u8 __user *)new_ctx, ctx_size)) + fault_in_readable((char __user *)new_ctx, ctx_size)) return -EFAULT; /* diff --git a/arch/x86/kernel/fpu/signal.c b/arch/x86/kernel/fpu/signal.c index fa17a27390ab..164c96434704 100644 --- a/arch/x86/kernel/fpu/signal.c +++ b/arch/x86/kernel/fpu/signal.c @@ -205,7 +205,7 @@ retry: fpregs_unlock(); if (ret) { - if (!fault_in_pages_writeable(buf_fx, fpu_user_xstate_size)) + if (!fault_in_writeable(buf_fx, fpu_user_xstate_size)) goto retry; return -EFAULT; } @@ -278,10 +278,9 @@ retry: if (ret != -EFAULT) return -EINVAL; - ret = fault_in_pages_readable(buf, size); - if (!ret) + if (!fault_in_readable(buf, size)) goto retry; - return ret; + return -EFAULT; } /* diff --git a/drivers/gpu/drm/armada/armada_gem.c b/drivers/gpu/drm/armada/armada_gem.c index 21909642ee4c..8fbb25913327 100644 --- a/drivers/gpu/drm/armada/armada_gem.c +++ b/drivers/gpu/drm/armada/armada_gem.c @@ -336,7 +336,7 @@ int armada_gem_pwrite_ioctl(struct drm_device *dev, void *data, struct drm_armada_gem_pwrite *args = data; struct armada_gem_object *dobj; char __user *ptr; - int ret; + int ret = 0; DRM_DEBUG_DRIVER("handle %u off %u size %u ptr 0x%llx\n", args->handle, args->offset, args->size, args->ptr); @@ -349,9 +349,8 @@ int armada_gem_pwrite_ioctl(struct drm_device *dev, void *data, if (!access_ok(ptr, args->size)) return -EFAULT; - ret = fault_in_pages_readable(ptr, args->size); - if (ret) - return ret; + if (fault_in_readable(ptr, args->size)) + return -EFAULT; dobj = armada_gem_object_lookup(file, args->handle); if (dobj == NULL) diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c index cc61813213d8..c0739f0af634 100644 --- a/fs/btrfs/ioctl.c +++ b/fs/btrfs/ioctl.c @@ -2261,9 +2261,8 @@ static noinline int search_ioctl(struct inode *inode, key.offset = sk->min_offset; while (1) { - ret = fault_in_pages_writeable(ubuf + sk_offset, - *buf_size - sk_offset); - if (ret) + ret = -EFAULT; + if (fault_in_writeable(ubuf + sk_offset, *buf_size - sk_offset)) break; ret = btrfs_search_forward(root, &key, path, sk->min_transid); diff --git a/include/linux/pagemap.h b/include/linux/pagemap.h index 62db6b0176b9..9fe94f7a4f7e 100644 --- a/include/linux/pagemap.h +++ b/include/linux/pagemap.h @@ -733,61 +733,10 @@ int wait_on_page_private_2_killable(struct page *page); extern void add_page_wait_queue(struct page *page, wait_queue_entry_t *waiter); /* - * Fault everything in given userspace address range in. + * Fault in userspace address range. */ -static inline int fault_in_pages_writeable(char __user *uaddr, size_t size) -{ - char __user *end = uaddr + size - 1; - - if (unlikely(size == 0)) - return 0; - - if (unlikely(uaddr > end)) - return -EFAULT; - /* - * Writing zeroes into userspace here is OK, because we know that if - * the zero gets there, we'll be overwriting it. - */ - do { - if (unlikely(__put_user(0, uaddr) != 0)) - return -EFAULT; - uaddr += PAGE_SIZE; - } while (uaddr <= end); - - /* Check whether the range spilled into the next page. */ - if (((unsigned long)uaddr & PAGE_MASK) == - ((unsigned long)end & PAGE_MASK)) - return __put_user(0, end); - - return 0; -} - -static inline int fault_in_pages_readable(const char __user *uaddr, size_t size) -{ - volatile char c; - const char __user *end = uaddr + size - 1; - - if (unlikely(size == 0)) - return 0; - - if (unlikely(uaddr > end)) - return -EFAULT; - - do { - if (unlikely(__get_user(c, uaddr) != 0)) - return -EFAULT; - uaddr += PAGE_SIZE; - } while (uaddr <= end); - - /* Check whether the range spilled into the next page. */ - if (((unsigned long)uaddr & PAGE_MASK) == - ((unsigned long)end & PAGE_MASK)) { - return __get_user(c, end); - } - - (void)c; - return 0; -} +size_t fault_in_writeable(char __user *uaddr, size_t size); +size_t fault_in_readable(const char __user *uaddr, size_t size); int add_to_page_cache_locked(struct page *page, struct address_space *mapping, pgoff_t index, gfp_t gfp_mask); diff --git a/lib/iov_iter.c b/lib/iov_iter.c index 60b5e6edfbaa..c88908f0f138 100644 --- a/lib/iov_iter.c +++ b/lib/iov_iter.c @@ -191,7 +191,7 @@ static size_t copy_page_to_iter_iovec(struct page *page, size_t offset, size_t b buf = iov->iov_base + skip; copy = min(bytes, iov->iov_len - skip); - if (IS_ENABLED(CONFIG_HIGHMEM) && !fault_in_pages_writeable(buf, copy)) { + if (IS_ENABLED(CONFIG_HIGHMEM) && !fault_in_writeable(buf, copy)) { kaddr = kmap_atomic(page); from = kaddr + offset; @@ -275,7 +275,7 @@ static size_t copy_page_from_iter_iovec(struct page *page, size_t offset, size_t buf = iov->iov_base + skip; copy = min(bytes, iov->iov_len - skip); - if (IS_ENABLED(CONFIG_HIGHMEM) && !fault_in_pages_readable(buf, copy)) { + if (IS_ENABLED(CONFIG_HIGHMEM) && !fault_in_readable(buf, copy)) { kaddr = kmap_atomic(page); to = kaddr + offset; @@ -446,13 +446,11 @@ int iov_iter_fault_in_readable(const struct iov_iter *i, size_t bytes) bytes = i->count; for (p = i->iov, skip = i->iov_offset; bytes; p++, skip = 0) { size_t len = min(bytes, p->iov_len - skip); - int err; if (unlikely(!len)) continue; - err = fault_in_pages_readable(p->iov_base + skip, len); - if (unlikely(err)) - return err; + if (fault_in_readable(p->iov_base + skip, len)) + return -EFAULT; bytes -= len; } } diff --git a/mm/filemap.c b/mm/filemap.c index dae481293b5d..ff34f4087f87 100644 --- a/mm/filemap.c +++ b/mm/filemap.c @@ -90,7 +90,7 @@ * ->lock_page (filemap_fault, access_process_vm) * * ->i_rwsem (generic_perform_write) - * ->mmap_lock (fault_in_pages_readable->do_page_fault) + * ->mmap_lock (fault_in_readable->do_page_fault) * * bdi->wb.list_lock * sb_lock (fs/fs-writeback.c) diff --git a/mm/gup.c b/mm/gup.c index 886d6148d3d0..a7efb027d6cf 100644 --- a/mm/gup.c +++ b/mm/gup.c @@ -1656,6 +1656,78 @@ finish_or_fault: } #endif /* !CONFIG_MMU */ +/** + * fault_in_writeable - fault in userspace address range for writing + * @uaddr: start of address range + * @size: size of address range + * + * Returns the number of bytes not faulted in (like copy_to_user() and + * copy_from_user()). + */ +size_t fault_in_writeable(char __user *uaddr, size_t size) +{ + char __user *start = uaddr, *end; + + if (unlikely(size == 0)) + return 0; + if (!PAGE_ALIGNED(uaddr)) { + if (unlikely(__put_user(0, uaddr) != 0)) + return size; + uaddr = (char __user *)PAGE_ALIGN((unsigned long)uaddr); + } + end = (char __user *)PAGE_ALIGN((unsigned long)start + size); + if (unlikely(end < start)) + end = NULL; + while (uaddr != end) { + if (unlikely(__put_user(0, uaddr) != 0)) + goto out; + uaddr += PAGE_SIZE; + } + +out: + if (size > uaddr - start) + return size - (uaddr - start); + return 0; +} +EXPORT_SYMBOL(fault_in_writeable); + +/** + * fault_in_readable - fault in userspace address range for reading + * @uaddr: start of user address range + * @size: size of user address range + * + * Returns the number of bytes not faulted in (like copy_to_user() and + * copy_from_user()). + */ +size_t fault_in_readable(const char __user *uaddr, size_t size) +{ + const char __user *start = uaddr, *end; + volatile char c; + + if (unlikely(size == 0)) + return 0; + if (!PAGE_ALIGNED(uaddr)) { + if (unlikely(__get_user(c, uaddr) != 0)) + return size; + uaddr = (const char __user *)PAGE_ALIGN((unsigned long)uaddr); + } + end = (const char __user *)PAGE_ALIGN((unsigned long)start + size); + if (unlikely(end < start)) + end = NULL; + while (uaddr != end) { + if (unlikely(__get_user(c, uaddr) != 0)) + goto out; + uaddr += PAGE_SIZE; + } + +out: + (void)c; + if (size > uaddr - start) + return size - (uaddr - start); + return 0; +} +EXPORT_SYMBOL(fault_in_readable); + /** * get_dump_page() - pin user page in memory while writing it to core dump * @addr: user address -- cgit v1.2.3 From a6294593e8a1290091d0b078d5d33da5e0cd3dfe Mon Sep 17 00:00:00 2001 From: Andreas Gruenbacher Date: Mon, 2 Aug 2021 14:54:16 +0200 Subject: iov_iter: Turn iov_iter_fault_in_readable into fault_in_iov_iter_readable Turn iov_iter_fault_in_readable into a function that returns the number of bytes not faulted in, similar to copy_to_user, instead of returning a non-zero value when any of the requested pages couldn't be faulted in. This supports the existing users that require all pages to be faulted in as well as new users that are happy if any pages can be faulted in. Rename iov_iter_fault_in_readable to fault_in_iov_iter_readable to make sure this change doesn't silently break things. Signed-off-by: Andreas Gruenbacher --- fs/btrfs/file.c | 2 +- fs/f2fs/file.c | 2 +- fs/fuse/file.c | 2 +- fs/iomap/buffered-io.c | 2 +- fs/ntfs/file.c | 2 +- fs/ntfs3/file.c | 2 +- include/linux/uio.h | 2 +- lib/iov_iter.c | 33 +++++++++++++++++++++------------ mm/filemap.c | 2 +- 9 files changed, 29 insertions(+), 20 deletions(-) (limited to 'lib') diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c index 7ff577005d0f..f37211d3bb69 100644 --- a/fs/btrfs/file.c +++ b/fs/btrfs/file.c @@ -1710,7 +1710,7 @@ static noinline ssize_t btrfs_buffered_write(struct kiocb *iocb, * Fault pages before locking them in prepare_pages * to avoid recursive lock */ - if (unlikely(iov_iter_fault_in_readable(i, write_bytes))) { + if (unlikely(fault_in_iov_iter_readable(i, write_bytes))) { ret = -EFAULT; break; } diff --git a/fs/f2fs/file.c b/fs/f2fs/file.c index 9c8ef33bd8d3..eb971e1e7227 100644 --- a/fs/f2fs/file.c +++ b/fs/f2fs/file.c @@ -4276,7 +4276,7 @@ static ssize_t f2fs_file_write_iter(struct kiocb *iocb, struct iov_iter *from) size_t target_size = 0; int err; - if (iov_iter_fault_in_readable(from, iov_iter_count(from))) + if (fault_in_iov_iter_readable(from, iov_iter_count(from))) set_inode_flag(inode, FI_NO_PREALLOC); if ((iocb->ki_flags & IOCB_NOWAIT)) { diff --git a/fs/fuse/file.c b/fs/fuse/file.c index 11404f8c21c7..4b6d8e13322d 100644 --- a/fs/fuse/file.c +++ b/fs/fuse/file.c @@ -1164,7 +1164,7 @@ static ssize_t fuse_fill_write_pages(struct fuse_io_args *ia, again: err = -EFAULT; - if (iov_iter_fault_in_readable(ii, bytes)) + if (fault_in_iov_iter_readable(ii, bytes)) break; err = -ENOMEM; diff --git a/fs/iomap/buffered-io.c b/fs/iomap/buffered-io.c index 9cc5798423d1..1753c26c8e76 100644 --- a/fs/iomap/buffered-io.c +++ b/fs/iomap/buffered-io.c @@ -750,7 +750,7 @@ again: * same page as we're writing to, without it being marked * up-to-date. */ - if (unlikely(iov_iter_fault_in_readable(i, bytes))) { + if (unlikely(fault_in_iov_iter_readable(i, bytes))) { status = -EFAULT; break; } diff --git a/fs/ntfs/file.c b/fs/ntfs/file.c index ab4f3362466d..a43adeacd930 100644 --- a/fs/ntfs/file.c +++ b/fs/ntfs/file.c @@ -1829,7 +1829,7 @@ again: * pages being swapped out between us bringing them into memory * and doing the actual copying. */ - if (unlikely(iov_iter_fault_in_readable(i, bytes))) { + if (unlikely(fault_in_iov_iter_readable(i, bytes))) { status = -EFAULT; break; } diff --git a/fs/ntfs3/file.c b/fs/ntfs3/file.c index 424450e77ad5..a52388387175 100644 --- a/fs/ntfs3/file.c +++ b/fs/ntfs3/file.c @@ -987,7 +987,7 @@ static ssize_t ntfs_compress_write(struct kiocb *iocb, struct iov_iter *from) frame_vbo = pos & ~(frame_size - 1); index = frame_vbo >> PAGE_SHIFT; - if (unlikely(iov_iter_fault_in_readable(from, bytes))) { + if (unlikely(fault_in_iov_iter_readable(from, bytes))) { err = -EFAULT; goto out; } diff --git a/include/linux/uio.h b/include/linux/uio.h index 207101a9c5c3..d18458af6681 100644 --- a/include/linux/uio.h +++ b/include/linux/uio.h @@ -133,7 +133,7 @@ size_t copy_page_from_iter_atomic(struct page *page, unsigned offset, size_t bytes, struct iov_iter *i); void iov_iter_advance(struct iov_iter *i, size_t bytes); void iov_iter_revert(struct iov_iter *i, size_t bytes); -int iov_iter_fault_in_readable(const struct iov_iter *i, size_t bytes); +size_t fault_in_iov_iter_readable(const struct iov_iter *i, size_t bytes); size_t iov_iter_single_seg_count(const struct iov_iter *i); size_t copy_page_to_iter(struct page *page, size_t offset, size_t bytes, struct iov_iter *i); diff --git a/lib/iov_iter.c b/lib/iov_iter.c index c88908f0f138..ce3d4f610626 100644 --- a/lib/iov_iter.c +++ b/lib/iov_iter.c @@ -430,33 +430,42 @@ out: } /* + * fault_in_iov_iter_readable - fault in iov iterator for reading + * @i: iterator + * @size: maximum length + * * Fault in one or more iovecs of the given iov_iter, to a maximum length of - * bytes. For each iovec, fault in each page that constitutes the iovec. + * @size. For each iovec, fault in each page that constitutes the iovec. + * + * Returns the number of bytes not faulted in (like copy_to_user() and + * copy_from_user()). * - * Return 0 on success, or non-zero if the memory could not be accessed (i.e. - * because it is an invalid address). + * Always returns 0 for non-userspace iterators. */ -int iov_iter_fault_in_readable(const struct iov_iter *i, size_t bytes) +size_t fault_in_iov_iter_readable(const struct iov_iter *i, size_t size) { if (iter_is_iovec(i)) { + size_t count = min(size, iov_iter_count(i)); const struct iovec *p; size_t skip; - if (bytes > i->count) - bytes = i->count; - for (p = i->iov, skip = i->iov_offset; bytes; p++, skip = 0) { - size_t len = min(bytes, p->iov_len - skip); + size -= count; + for (p = i->iov, skip = i->iov_offset; count; p++, skip = 0) { + size_t len = min(count, p->iov_len - skip); + size_t ret; if (unlikely(!len)) continue; - if (fault_in_readable(p->iov_base + skip, len)) - return -EFAULT; - bytes -= len; + ret = fault_in_readable(p->iov_base + skip, len); + count -= len - ret; + if (ret) + break; } + return count + size; } return 0; } -EXPORT_SYMBOL(iov_iter_fault_in_readable); +EXPORT_SYMBOL(fault_in_iov_iter_readable); void iov_iter_init(struct iov_iter *i, unsigned int direction, const struct iovec *iov, unsigned long nr_segs, diff --git a/mm/filemap.c b/mm/filemap.c index ff34f4087f87..4dd5edcd39fd 100644 --- a/mm/filemap.c +++ b/mm/filemap.c @@ -3757,7 +3757,7 @@ again: * same page as we're writing to, without it being marked * up-to-date. */ - if (unlikely(iov_iter_fault_in_readable(i, bytes))) { + if (unlikely(fault_in_iov_iter_readable(i, bytes))) { status = -EFAULT; break; } -- cgit v1.2.3 From be58f7103700a68d5c7ca60a2bc0b309907599ab Mon Sep 17 00:00:00 2001 From: Kees Cook Date: Tue, 20 Apr 2021 23:22:52 -0700 Subject: fortify: Add compile-time FORTIFY_SOURCE tests While the run-time testing of FORTIFY_SOURCE is already present in LKDTM, there is no testing of the expected compile-time detections. In preparation for correctly supporting FORTIFY_SOURCE under Clang, adding additional FORTIFY_SOURCE defenses, and making sure FORTIFY_SOURCE doesn't silently regress with GCC, introduce a build-time test suite that checks each expected compile-time failure condition. As this is relatively backwards from standard build rules in the sense that a successful test is actually a compile _failure_, create a wrapper script to check for the correct errors, and wire it up as a dummy dependency to lib/string.o, collecting the results into a log file artifact. Signed-off-by: Kees Cook --- MAINTAINERS | 9 ++++ lib/.gitignore | 2 + lib/Makefile | 33 ++++++++++++++ lib/test_fortify/read_overflow-memchr.c | 5 +++ lib/test_fortify/read_overflow-memchr_inv.c | 5 +++ lib/test_fortify/read_overflow-memcmp.c | 5 +++ lib/test_fortify/read_overflow-memscan.c | 5 +++ lib/test_fortify/read_overflow2-memcmp.c | 5 +++ lib/test_fortify/read_overflow2-memcpy.c | 5 +++ lib/test_fortify/read_overflow2-memmove.c | 5 +++ lib/test_fortify/test_fortify.h | 35 +++++++++++++++ lib/test_fortify/write_overflow-memcpy.c | 5 +++ lib/test_fortify/write_overflow-memmove.c | 5 +++ lib/test_fortify/write_overflow-memset.c | 5 +++ lib/test_fortify/write_overflow-strcpy-lit.c | 5 +++ lib/test_fortify/write_overflow-strcpy.c | 5 +++ lib/test_fortify/write_overflow-strlcpy-src.c | 5 +++ lib/test_fortify/write_overflow-strlcpy.c | 5 +++ lib/test_fortify/write_overflow-strncpy-src.c | 5 +++ lib/test_fortify/write_overflow-strncpy.c | 5 +++ lib/test_fortify/write_overflow-strscpy.c | 5 +++ scripts/test_fortify.sh | 62 +++++++++++++++++++++++++++ 22 files changed, 226 insertions(+) create mode 100644 lib/test_fortify/read_overflow-memchr.c create mode 100644 lib/test_fortify/read_overflow-memchr_inv.c create mode 100644 lib/test_fortify/read_overflow-memcmp.c create mode 100644 lib/test_fortify/read_overflow-memscan.c create mode 100644 lib/test_fortify/read_overflow2-memcmp.c create mode 100644 lib/test_fortify/read_overflow2-memcpy.c create mode 100644 lib/test_fortify/read_overflow2-memmove.c create mode 100644 lib/test_fortify/test_fortify.h create mode 100644 lib/test_fortify/write_overflow-memcpy.c create mode 100644 lib/test_fortify/write_overflow-memmove.c create mode 100644 lib/test_fortify/write_overflow-memset.c create mode 100644 lib/test_fortify/write_overflow-strcpy-lit.c create mode 100644 lib/test_fortify/write_overflow-strcpy.c create mode 100644 lib/test_fortify/write_overflow-strlcpy-src.c create mode 100644 lib/test_fortify/write_overflow-strlcpy.c create mode 100644 lib/test_fortify/write_overflow-strncpy-src.c create mode 100644 lib/test_fortify/write_overflow-strncpy.c create mode 100644 lib/test_fortify/write_overflow-strscpy.c create mode 100644 scripts/test_fortify.sh (limited to 'lib') diff --git a/MAINTAINERS b/MAINTAINERS index ca6d6fde85cf..6f4b97dff7ad 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -7323,6 +7323,15 @@ L: netdev@vger.kernel.org S: Maintained F: drivers/net/ethernet/nvidia/* +FORTIFY_SOURCE +M: Kees Cook +L: linux-hardening@vger.kernel.org +S: Supported +F: include/linux/fortify-string.h +F: lib/test_fortify/* +F: scripts/test_fortify.sh +K: \b__NO_FORTIFY\b + FPGA DFL DRIVERS M: Wu Hao R: Tom Rix diff --git a/lib/.gitignore b/lib/.gitignore index 5e7fa54c4536..e5e217b8307b 100644 --- a/lib/.gitignore +++ b/lib/.gitignore @@ -4,3 +4,5 @@ /gen_crc32table /gen_crc64table /oid_registry_data.c +/test_fortify.log +/test_fortify/*.log diff --git a/lib/Makefile b/lib/Makefile index 5efd1b435a37..bd17c2bf43e1 100644 --- a/lib/Makefile +++ b/lib/Makefile @@ -360,3 +360,36 @@ obj-$(CONFIG_CMDLINE_KUNIT_TEST) += cmdline_kunit.o obj-$(CONFIG_SLUB_KUNIT_TEST) += slub_kunit.o obj-$(CONFIG_GENERIC_LIB_DEVMEM_IS_ALLOWED) += devmem_is_allowed.o + +# FORTIFY_SOURCE compile-time behavior tests +TEST_FORTIFY_SRCS = $(wildcard $(srctree)/$(src)/test_fortify/*-*.c) +TEST_FORTIFY_LOGS = $(patsubst $(srctree)/$(src)/%.c, %.log, $(TEST_FORTIFY_SRCS)) +TEST_FORTIFY_LOG = test_fortify.log + +quiet_cmd_test_fortify = TEST $@ + cmd_test_fortify = $(CONFIG_SHELL) $(srctree)/scripts/test_fortify.sh \ + $< $@ "$(NM)" $(CC) $(c_flags) \ + $(call cc-disable-warning,fortify-source) + +targets += $(TEST_FORTIFY_LOGS) +clean-files += $(TEST_FORTIFY_LOGS) +clean-files += $(addsuffix .o, $(TEST_FORTIFY_LOGS)) +$(obj)/test_fortify/%.log: $(src)/test_fortify/%.c \ + $(src)/test_fortify/test_fortify.h \ + $(srctree)/include/linux/fortify-string.h \ + $(srctree)/scripts/test_fortify.sh \ + FORCE + $(call if_changed,test_fortify) + +quiet_cmd_gen_fortify_log = GEN $@ + cmd_gen_fortify_log = cat /dev/null > $@ || true + +targets += $(TEST_FORTIFY_LOG) +clean-files += $(TEST_FORTIFY_LOG) +$(obj)/$(TEST_FORTIFY_LOG): $(addprefix $(obj)/, $(TEST_FORTIFY_LOGS)) FORCE + $(call if_changed,gen_fortify_log) + +# Fake dependency to trigger the fortify tests. +ifeq ($(CONFIG_FORTIFY_SOURCE),y) +$(obj)/string.o: $(obj)/$(TEST_FORTIFY_LOG) +endif diff --git a/lib/test_fortify/read_overflow-memchr.c b/lib/test_fortify/read_overflow-memchr.c new file mode 100644 index 000000000000..2743084b32af --- /dev/null +++ b/lib/test_fortify/read_overflow-memchr.c @@ -0,0 +1,5 @@ +// SPDX-License-Identifier: GPL-2.0-only +#define TEST \ + memchr(small, 0x7A, sizeof(small) + 1) + +#include "test_fortify.h" diff --git a/lib/test_fortify/read_overflow-memchr_inv.c b/lib/test_fortify/read_overflow-memchr_inv.c new file mode 100644 index 000000000000..b26e1f1bc217 --- /dev/null +++ b/lib/test_fortify/read_overflow-memchr_inv.c @@ -0,0 +1,5 @@ +// SPDX-License-Identifier: GPL-2.0-only +#define TEST \ + memchr_inv(small, 0x7A, sizeof(small) + 1) + +#include "test_fortify.h" diff --git a/lib/test_fortify/read_overflow-memcmp.c b/lib/test_fortify/read_overflow-memcmp.c new file mode 100644 index 000000000000..d5d301ff64ef --- /dev/null +++ b/lib/test_fortify/read_overflow-memcmp.c @@ -0,0 +1,5 @@ +// SPDX-License-Identifier: GPL-2.0-only +#define TEST \ + memcmp(small, large, sizeof(small) + 1) + +#include "test_fortify.h" diff --git a/lib/test_fortify/read_overflow-memscan.c b/lib/test_fortify/read_overflow-memscan.c new file mode 100644 index 000000000000..c1a97f2df0f0 --- /dev/null +++ b/lib/test_fortify/read_overflow-memscan.c @@ -0,0 +1,5 @@ +// SPDX-License-Identifier: GPL-2.0-only +#define TEST \ + memscan(small, 0x7A, sizeof(small) + 1) + +#include "test_fortify.h" diff --git a/lib/test_fortify/read_overflow2-memcmp.c b/lib/test_fortify/read_overflow2-memcmp.c new file mode 100644 index 000000000000..c6091e640f76 --- /dev/null +++ b/lib/test_fortify/read_overflow2-memcmp.c @@ -0,0 +1,5 @@ +// SPDX-License-Identifier: GPL-2.0-only +#define TEST \ + memcmp(large, small, sizeof(small) + 1) + +#include "test_fortify.h" diff --git a/lib/test_fortify/read_overflow2-memcpy.c b/lib/test_fortify/read_overflow2-memcpy.c new file mode 100644 index 000000000000..07b62e56cf16 --- /dev/null +++ b/lib/test_fortify/read_overflow2-memcpy.c @@ -0,0 +1,5 @@ +// SPDX-License-Identifier: GPL-2.0-only +#define TEST \ + memcpy(large, instance.buf, sizeof(large)) + +#include "test_fortify.h" diff --git a/lib/test_fortify/read_overflow2-memmove.c b/lib/test_fortify/read_overflow2-memmove.c new file mode 100644 index 000000000000..34edfab040a3 --- /dev/null +++ b/lib/test_fortify/read_overflow2-memmove.c @@ -0,0 +1,5 @@ +// SPDX-License-Identifier: GPL-2.0-only +#define TEST \ + memmove(large, instance.buf, sizeof(large)) + +#include "test_fortify.h" diff --git a/lib/test_fortify/test_fortify.h b/lib/test_fortify/test_fortify.h new file mode 100644 index 000000000000..d22664fff197 --- /dev/null +++ b/lib/test_fortify/test_fortify.h @@ -0,0 +1,35 @@ +/* SPDX-License-Identifier: GPL-2.0-only */ +#include +#include +#include +#include + +void do_fortify_tests(void); + +#define __BUF_SMALL 16 +#define __BUF_LARGE 32 +struct fortify_object { + int a; + char buf[__BUF_SMALL]; + int c; +}; + +#define LITERAL_SMALL "AAAAAAAAAAAAAAA" +#define LITERAL_LARGE "AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA" +const char small_src[__BUF_SMALL] = LITERAL_SMALL; +const char large_src[__BUF_LARGE] = LITERAL_LARGE; + +char small[__BUF_SMALL]; +char large[__BUF_LARGE]; +struct fortify_object instance; +size_t size; + +void do_fortify_tests(void) +{ + /* Normal initializations. */ + memset(&instance, 0x32, sizeof(instance)); + memset(small, 0xA5, sizeof(small)); + memset(large, 0x5A, sizeof(large)); + + TEST; +} diff --git a/lib/test_fortify/write_overflow-memcpy.c b/lib/test_fortify/write_overflow-memcpy.c new file mode 100644 index 000000000000..3b3984e428fb --- /dev/null +++ b/lib/test_fortify/write_overflow-memcpy.c @@ -0,0 +1,5 @@ +// SPDX-License-Identifier: GPL-2.0-only +#define TEST \ + memcpy(instance.buf, large_src, sizeof(large_src)) + +#include "test_fortify.h" diff --git a/lib/test_fortify/write_overflow-memmove.c b/lib/test_fortify/write_overflow-memmove.c new file mode 100644 index 000000000000..640437c3b3e0 --- /dev/null +++ b/lib/test_fortify/write_overflow-memmove.c @@ -0,0 +1,5 @@ +// SPDX-License-Identifier: GPL-2.0-only +#define TEST \ + memmove(instance.buf, large_src, sizeof(large_src)) + +#include "test_fortify.h" diff --git a/lib/test_fortify/write_overflow-memset.c b/lib/test_fortify/write_overflow-memset.c new file mode 100644 index 000000000000..36e34908cfb3 --- /dev/null +++ b/lib/test_fortify/write_overflow-memset.c @@ -0,0 +1,5 @@ +// SPDX-License-Identifier: GPL-2.0-only +#define TEST \ + memset(instance.buf, 0x5A, sizeof(large_src)) + +#include "test_fortify.h" diff --git a/lib/test_fortify/write_overflow-strcpy-lit.c b/lib/test_fortify/write_overflow-strcpy-lit.c new file mode 100644 index 000000000000..51effb3e50f9 --- /dev/null +++ b/lib/test_fortify/write_overflow-strcpy-lit.c @@ -0,0 +1,5 @@ +// SPDX-License-Identifier: GPL-2.0-only +#define TEST \ + strcpy(small, LITERAL_LARGE) + +#include "test_fortify.h" diff --git a/lib/test_fortify/write_overflow-strcpy.c b/lib/test_fortify/write_overflow-strcpy.c new file mode 100644 index 000000000000..84f1c56a64c8 --- /dev/null +++ b/lib/test_fortify/write_overflow-strcpy.c @@ -0,0 +1,5 @@ +// SPDX-License-Identifier: GPL-2.0-only +#define TEST \ + strcpy(small, large_src) + +#include "test_fortify.h" diff --git a/lib/test_fortify/write_overflow-strlcpy-src.c b/lib/test_fortify/write_overflow-strlcpy-src.c new file mode 100644 index 000000000000..91bf83ebd34a --- /dev/null +++ b/lib/test_fortify/write_overflow-strlcpy-src.c @@ -0,0 +1,5 @@ +// SPDX-License-Identifier: GPL-2.0-only +#define TEST \ + strlcpy(small, large_src, sizeof(small) + 1) + +#include "test_fortify.h" diff --git a/lib/test_fortify/write_overflow-strlcpy.c b/lib/test_fortify/write_overflow-strlcpy.c new file mode 100644 index 000000000000..1883db7c0cd6 --- /dev/null +++ b/lib/test_fortify/write_overflow-strlcpy.c @@ -0,0 +1,5 @@ +// SPDX-License-Identifier: GPL-2.0-only +#define TEST \ + strlcpy(instance.buf, large_src, sizeof(instance.buf) + 1) + +#include "test_fortify.h" diff --git a/lib/test_fortify/write_overflow-strncpy-src.c b/lib/test_fortify/write_overflow-strncpy-src.c new file mode 100644 index 000000000000..8dcfb8c788dd --- /dev/null +++ b/lib/test_fortify/write_overflow-strncpy-src.c @@ -0,0 +1,5 @@ +// SPDX-License-Identifier: GPL-2.0-only +#define TEST \ + strncpy(small, large_src, sizeof(small) + 1) + +#include "test_fortify.h" diff --git a/lib/test_fortify/write_overflow-strncpy.c b/lib/test_fortify/write_overflow-strncpy.c new file mode 100644 index 000000000000..b85f079c815d --- /dev/null +++ b/lib/test_fortify/write_overflow-strncpy.c @@ -0,0 +1,5 @@ +// SPDX-License-Identifier: GPL-2.0-only +#define TEST \ + strncpy(instance.buf, large_src, sizeof(instance.buf) + 1) + +#include "test_fortify.h" diff --git a/lib/test_fortify/write_overflow-strscpy.c b/lib/test_fortify/write_overflow-strscpy.c new file mode 100644 index 000000000000..38feddf377dc --- /dev/null +++ b/lib/test_fortify/write_overflow-strscpy.c @@ -0,0 +1,5 @@ +// SPDX-License-Identifier: GPL-2.0-only +#define TEST \ + strscpy(instance.buf, large_src, sizeof(instance.buf) + 1) + +#include "test_fortify.h" diff --git a/scripts/test_fortify.sh b/scripts/test_fortify.sh new file mode 100644 index 000000000000..a4da365508f0 --- /dev/null +++ b/scripts/test_fortify.sh @@ -0,0 +1,62 @@ +#!/bin/sh +# SPDX-License-Identifier: GPL-2.0-only +set -e + +# Argument 1: Source file to build. +IN="$1" +shift +# Extract just the filename for error messages below. +FILE="${IN##*/}" +# Extract the function name for error messages below. +FUNC="${FILE#*-}" +FUNC="${FUNC%%-*}" +FUNC="${FUNC%%.*}" +# Extract the symbol to test for in build/symbol test below. +WANT="__${FILE%%-*}" + +# Argument 2: Where to write the build log. +OUT="$1" +shift +TMP="${OUT}.tmp" + +# Argument 3: Path to "nm" tool. +NM="$1" +shift + +# Remaining arguments are: $(CC) $(c_flags) + +# Clean up temporary file at exit. +__cleanup() { + rm -f "$TMP" +} +trap __cleanup EXIT + +# Function names in warnings are wrapped in backticks under UTF-8 locales. +# Run the commands with LANG=C so that grep output will not change. +export LANG=C + +status= +# Attempt to build a source that is expected to fail with a specific warning. +if "$@" -Werror -c "$IN" -o "$OUT".o 2> "$TMP" ; then + # If the build succeeds, either the test has failed or the + # warning may only happen at link time (Clang). In that case, + # make sure the expected symbol is unresolved in the symbol list. + # If so, FORTIFY is working for this case. + if ! $NM -A "$OUT".o | grep -m1 "\bU ${WANT}$" >>"$TMP" ; then + status="warning: unsafe ${FUNC}() usage lacked '$WANT' symbol in $IN" + fi +else + # If the build failed, check for the warning in the stderr (gcc). + if ! grep -q -m1 "error: call to .\b${WANT}\b." "$TMP" ; then + status="warning: unsafe ${FUNC}() usage lacked '$WANT' warning in $IN" + fi +fi + +if [ -n "$status" ]; then + # Report on failure results, including compilation warnings. + echo "$status" | tee "$OUT" >&2 +else + # Report on good results, and save any compilation output to log. + echo "ok: unsafe ${FUNC}() usage correctly detected with '$WANT' in $IN" >"$OUT" +fi +cat "$TMP" >>"$OUT" -- cgit v1.2.3 From bb95ebbe89a7854368be061acefb22040fbcc486 Mon Sep 17 00:00:00 2001 From: Kees Cook Date: Fri, 25 Jun 2021 17:45:15 -0700 Subject: lib: Introduce CONFIG_MEMCPY_KUNIT_TEST Before changing anything about memcpy(), memmove(), and memset(), add run-time tests to check basic behaviors for any regressions. Signed-off-by: Kees Cook --- lib/Kconfig.debug | 11 +++ lib/Makefile | 1 + lib/memcpy_kunit.c | 265 +++++++++++++++++++++++++++++++++++++++++++++++++++++ 3 files changed, 277 insertions(+) create mode 100644 lib/memcpy_kunit.c (limited to 'lib') diff --git a/lib/Kconfig.debug b/lib/Kconfig.debug index d566f601780f..98426bd78608 100644 --- a/lib/Kconfig.debug +++ b/lib/Kconfig.debug @@ -2452,6 +2452,17 @@ config RATIONAL_KUNIT_TEST If unsure, say N. +config MEMCPY_KUNIT_TEST + tristate "Test memcpy(), memmove(), and memset() functions at runtime" if !KUNIT_ALL_TESTS + depends on KUNIT + default KUNIT_ALL_TESTS + help + Builds unit tests for memcpy(), memmove(), and memset() functions. + For more information on KUnit and unit tests in general please refer + to the KUnit documentation in Documentation/dev-tools/kunit/. + + If unsure, say N. + config TEST_UDELAY tristate "udelay test driver" help diff --git a/lib/Makefile b/lib/Makefile index bd17c2bf43e1..2cfd33917ad5 100644 --- a/lib/Makefile +++ b/lib/Makefile @@ -358,6 +358,7 @@ obj-$(CONFIG_LINEAR_RANGES_TEST) += test_linear_ranges.o obj-$(CONFIG_BITS_TEST) += test_bits.o obj-$(CONFIG_CMDLINE_KUNIT_TEST) += cmdline_kunit.o obj-$(CONFIG_SLUB_KUNIT_TEST) += slub_kunit.o +obj-$(CONFIG_MEMCPY_KUNIT_TEST) += memcpy_kunit.o obj-$(CONFIG_GENERIC_LIB_DEVMEM_IS_ALLOWED) += devmem_is_allowed.o diff --git a/lib/memcpy_kunit.c b/lib/memcpy_kunit.c new file mode 100644 index 000000000000..8b2109bb62df --- /dev/null +++ b/lib/memcpy_kunit.c @@ -0,0 +1,265 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * Test cases for memcpy(), memmove(), and memset(). + */ +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +struct some_bytes { + union { + u8 data[32]; + struct { + u32 one; + u16 two; + u8 three; + /* 1 byte hole */ + u32 four[4]; + }; + }; +}; + +#define check(instance, v) do { \ + int i; \ + BUILD_BUG_ON(sizeof(instance.data) != 32); \ + for (i = 0; i < sizeof(instance.data); i++) { \ + KUNIT_ASSERT_EQ_MSG(test, instance.data[i], v, \ + "line %d: '%s' not initialized to 0x%02x @ %d (saw 0x%02x)\n", \ + __LINE__, #instance, v, i, instance.data[i]); \ + } \ +} while (0) + +#define compare(name, one, two) do { \ + int i; \ + BUILD_BUG_ON(sizeof(one) != sizeof(two)); \ + for (i = 0; i < sizeof(one); i++) { \ + KUNIT_EXPECT_EQ_MSG(test, one.data[i], two.data[i], \ + "line %d: %s.data[%d] (0x%02x) != %s.data[%d] (0x%02x)\n", \ + __LINE__, #one, i, one.data[i], #two, i, two.data[i]); \ + } \ + kunit_info(test, "ok: " TEST_OP "() " name "\n"); \ +} while (0) + +static void memcpy_test(struct kunit *test) +{ +#define TEST_OP "memcpy" + struct some_bytes control = { + .data = { 0x20, 0x20, 0x20, 0x20, 0x20, 0x20, 0x20, 0x20, + 0x20, 0x20, 0x20, 0x20, 0x20, 0x20, 0x20, 0x20, + 0x20, 0x20, 0x20, 0x20, 0x20, 0x20, 0x20, 0x20, + 0x20, 0x20, 0x20, 0x20, 0x20, 0x20, 0x20, 0x20, + }, + }; + struct some_bytes zero = { }; + struct some_bytes middle = { + .data = { 0x20, 0x20, 0x20, 0x20, 0x20, 0x20, 0x20, 0x20, + 0x20, 0x20, 0x20, 0x20, 0x00, 0x00, 0x00, 0x00, + 0x00, 0x00, 0x00, 0x20, 0x20, 0x20, 0x20, 0x20, + 0x20, 0x20, 0x20, 0x20, 0x20, 0x20, 0x20, 0x20, + }, + }; + struct some_bytes three = { + .data = { 0x00, 0x20, 0x20, 0x20, 0x20, 0x20, 0x20, 0x20, + 0x20, 0x00, 0x00, 0x20, 0x20, 0x20, 0x20, 0x20, + 0x20, 0x20, 0x20, 0x20, 0x20, 0x20, 0x20, 0x20, + 0x20, 0x20, 0x20, 0x20, 0x20, 0x20, 0x20, 0x20, + }, + }; + struct some_bytes dest = { }; + int count; + u8 *ptr; + + /* Verify static initializers. */ + check(control, 0x20); + check(zero, 0); + compare("static initializers", dest, zero); + + /* Verify assignment. */ + dest = control; + compare("direct assignment", dest, control); + + /* Verify complete overwrite. */ + memcpy(dest.data, zero.data, sizeof(dest.data)); + compare("complete overwrite", dest, zero); + + /* Verify middle overwrite. */ + dest = control; + memcpy(dest.data + 12, zero.data, 7); + compare("middle overwrite", dest, middle); + + /* Verify argument side-effects aren't repeated. */ + dest = control; + ptr = dest.data; + count = 1; + memcpy(ptr++, zero.data, count++); + ptr += 8; + memcpy(ptr++, zero.data, count++); + compare("argument side-effects", dest, three); +#undef TEST_OP +} + +static void memmove_test(struct kunit *test) +{ +#define TEST_OP "memmove" + struct some_bytes control = { + .data = { 0x99, 0x99, 0x99, 0x99, 0x99, 0x99, 0x99, 0x99, + 0x99, 0x99, 0x99, 0x99, 0x99, 0x99, 0x99, 0x99, + 0x99, 0x99, 0x99, 0x99, 0x99, 0x99, 0x99, 0x99, + 0x99, 0x99, 0x99, 0x99, 0x99, 0x99, 0x99, 0x99, + }, + }; + struct some_bytes zero = { }; + struct some_bytes middle = { + .data = { 0x99, 0x99, 0x99, 0x99, 0x99, 0x99, 0x99, 0x99, + 0x99, 0x99, 0x99, 0x99, 0x00, 0x00, 0x00, 0x00, + 0x00, 0x00, 0x00, 0x99, 0x99, 0x99, 0x99, 0x99, + 0x99, 0x99, 0x99, 0x99, 0x99, 0x99, 0x99, 0x99, + }, + }; + struct some_bytes five = { + .data = { 0x00, 0x00, 0x99, 0x99, 0x99, 0x99, 0x99, 0x99, + 0x99, 0x99, 0x00, 0x00, 0x00, 0x99, 0x99, 0x99, + 0x99, 0x99, 0x99, 0x99, 0x99, 0x99, 0x99, 0x99, + 0x99, 0x99, 0x99, 0x99, 0x99, 0x99, 0x99, 0x99, + }, + }; + struct some_bytes overlap = { + .data = { 0x00, 0x01, 0x02, 0x03, 0x04, 0x05, 0x06, 0x07, + 0x08, 0x09, 0x0A, 0x0B, 0x0C, 0x0D, 0x0E, 0x0F, + 0x99, 0x99, 0x99, 0x99, 0x99, 0x99, 0x99, 0x99, + 0x99, 0x99, 0x99, 0x99, 0x99, 0x99, 0x99, 0x99, + }, + }; + struct some_bytes overlap_expected = { + .data = { 0x00, 0x01, 0x00, 0x01, 0x02, 0x03, 0x04, 0x07, + 0x08, 0x09, 0x0A, 0x0B, 0x0C, 0x0D, 0x0E, 0x0F, + 0x99, 0x99, 0x99, 0x99, 0x99, 0x99, 0x99, 0x99, + 0x99, 0x99, 0x99, 0x99, 0x99, 0x99, 0x99, 0x99, + }, + }; + struct some_bytes dest = { }; + int count; + u8 *ptr; + + /* Verify static initializers. */ + check(control, 0x99); + check(zero, 0); + compare("static initializers", zero, dest); + + /* Verify assignment. */ + dest = control; + compare("direct assignment", dest, control); + + /* Verify complete overwrite. */ + memmove(dest.data, zero.data, sizeof(dest.data)); + compare("complete overwrite", dest, zero); + + /* Verify middle overwrite. */ + dest = control; + memmove(dest.data + 12, zero.data, 7); + compare("middle overwrite", dest, middle); + + /* Verify argument side-effects aren't repeated. */ + dest = control; + ptr = dest.data; + count = 2; + memmove(ptr++, zero.data, count++); + ptr += 9; + memmove(ptr++, zero.data, count++); + compare("argument side-effects", dest, five); + + /* Verify overlapping overwrite is correct. */ + ptr = &overlap.data[2]; + memmove(ptr, overlap.data, 5); + compare("overlapping write", overlap, overlap_expected); +#undef TEST_OP +} + +static void memset_test(struct kunit *test) +{ +#define TEST_OP "memset" + struct some_bytes control = { + .data = { 0x30, 0x30, 0x30, 0x30, 0x30, 0x30, 0x30, 0x30, + 0x30, 0x30, 0x30, 0x30, 0x30, 0x30, 0x30, 0x30, + 0x30, 0x30, 0x30, 0x30, 0x30, 0x30, 0x30, 0x30, + 0x30, 0x30, 0x30, 0x30, 0x30, 0x30, 0x30, 0x30, + }, + }; + struct some_bytes complete = { + .data = { 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, + 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, + 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, + 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, + }, + }; + struct some_bytes middle = { + .data = { 0x30, 0x30, 0x30, 0x30, 0x31, 0x31, 0x31, 0x31, + 0x31, 0x31, 0x31, 0x31, 0x31, 0x31, 0x31, 0x31, + 0x31, 0x31, 0x31, 0x31, 0x30, 0x30, 0x30, 0x30, + 0x30, 0x30, 0x30, 0x30, 0x30, 0x30, 0x30, 0x30, + }, + }; + struct some_bytes three = { + .data = { 0x60, 0x30, 0x30, 0x30, 0x30, 0x30, 0x30, 0x30, + 0x30, 0x61, 0x61, 0x30, 0x30, 0x30, 0x30, 0x30, + 0x30, 0x30, 0x30, 0x30, 0x30, 0x30, 0x30, 0x30, + 0x30, 0x30, 0x30, 0x30, 0x30, 0x30, 0x30, 0x30, + }, + }; + struct some_bytes dest = { }; + int count, value; + u8 *ptr; + + /* Verify static initializers. */ + check(control, 0x30); + check(dest, 0); + + /* Verify assignment. */ + dest = control; + compare("direct assignment", dest, control); + + /* Verify complete overwrite. */ + memset(dest.data, 0xff, sizeof(dest.data)); + compare("complete overwrite", dest, complete); + + /* Verify middle overwrite. */ + dest = control; + memset(dest.data + 4, 0x31, 16); + compare("middle overwrite", dest, middle); + + /* Verify argument side-effects aren't repeated. */ + dest = control; + ptr = dest.data; + value = 0x60; + count = 1; + memset(ptr++, value++, count++); + ptr += 8; + memset(ptr++, value++, count++); + compare("argument side-effects", dest, three); +#undef TEST_OP +} + +static struct kunit_case memcpy_test_cases[] = { + KUNIT_CASE(memset_test), + KUNIT_CASE(memcpy_test), + KUNIT_CASE(memmove_test), + {} +}; + +static struct kunit_suite memcpy_test_suite = { + .name = "memcpy", + .test_cases = memcpy_test_cases, +}; + +kunit_test_suite(memcpy_test_suite); + +MODULE_LICENSE("GPL"); -- cgit v1.2.3 From 4797632f4f1d8af4e0670adcb97bf9800dc3beca Mon Sep 17 00:00:00 2001 From: Kees Cook Date: Mon, 17 May 2021 20:16:57 -0700 Subject: string.h: Introduce memset_after() for wiping trailing members/padding A common idiom in kernel code is to wipe the contents of a structure after a given member. This is especially useful in places where there is trailing padding. These open-coded cases are usually difficult to read and very sensitive to struct layout changes. Introduce a new helper, memset_after() that takes the target struct instance, the byte to write, and the member name after which the zeroing should start. Cc: Steffen Klassert Cc: Herbert Xu Cc: "David S. Miller" Cc: Jakub Kicinski Cc: Andrew Morton Cc: Francis Laniel Cc: Vincenzo Frascino Cc: Daniel Axtens Cc: netdev@vger.kernel.org Signed-off-by: Kees Cook --- include/linux/string.h | 17 +++++++++++++++++ lib/memcpy_kunit.c | 13 +++++++++++++ 2 files changed, 30 insertions(+) (limited to 'lib') diff --git a/include/linux/string.h b/include/linux/string.h index ac1c769a5a80..da490c2154a9 100644 --- a/include/linux/string.h +++ b/include/linux/string.h @@ -271,6 +271,23 @@ static inline void memcpy_and_pad(void *dest, size_t dest_len, memcpy(dest, src, dest_len); } +/** + * memset_after - Set a value after a struct member to the end of a struct + * + * @obj: Address of target struct instance + * @v: Byte value to repeatedly write + * @member: after which struct member to start writing bytes + * + * This is good for clearing padding following the given member. + */ +#define memset_after(obj, v, member) \ +({ \ + u8 *__ptr = (u8 *)(obj); \ + typeof(v) __val = (v); \ + memset(__ptr + offsetofend(typeof(*(obj)), member), __val, \ + sizeof(*(obj)) - offsetofend(typeof(*(obj)), member)); \ +}) + /** * str_has_prefix - Test if a string has a given prefix * @str: The string to test diff --git a/lib/memcpy_kunit.c b/lib/memcpy_kunit.c index 8b2109bb62df..5c5b4f3221d9 100644 --- a/lib/memcpy_kunit.c +++ b/lib/memcpy_kunit.c @@ -215,6 +215,13 @@ static void memset_test(struct kunit *test) 0x30, 0x30, 0x30, 0x30, 0x30, 0x30, 0x30, 0x30, }, }; + struct some_bytes after = { + .data = { 0x30, 0x30, 0x30, 0x30, 0x30, 0x30, 0x30, 0x72, + 0x72, 0x72, 0x72, 0x72, 0x72, 0x72, 0x72, 0x72, + 0x72, 0x72, 0x72, 0x72, 0x72, 0x72, 0x72, 0x72, + 0x72, 0x72, 0x72, 0x72, 0x72, 0x72, 0x72, 0x72, + }, + }; struct some_bytes dest = { }; int count, value; u8 *ptr; @@ -245,6 +252,12 @@ static void memset_test(struct kunit *test) ptr += 8; memset(ptr++, value++, count++); compare("argument side-effects", dest, three); + + /* Verify memset_after() */ + dest = control; + memset_after(&dest, 0x72, three); + compare("memset_after()", dest, after); + #undef TEST_OP } -- cgit v1.2.3 From 6dbefad40815a61aecbcf9b552e87ef57ab8cc7d Mon Sep 17 00:00:00 2001 From: Kees Cook Date: Mon, 17 May 2021 20:16:57 -0700 Subject: string.h: Introduce memset_startat() for wiping trailing members and padding A common idiom in kernel code is to wipe the contents of a structure starting from a given member. These open-coded cases are usually difficult to read and very sensitive to struct layout changes. Like memset_after(), introduce a new helper, memset_startat() that takes the target struct instance, the byte to write, and the member name where zeroing should start. Note that this doesn't zero padding preceding the target member. For those cases, memset_after() should be used on the preceding member. Cc: Steffen Klassert Cc: Herbert Xu Cc: "David S. Miller" Cc: Jakub Kicinski Cc: Andrew Morton Cc: Francis Laniel Cc: Vincenzo Frascino Cc: Daniel Axtens Cc: netdev@vger.kernel.org Signed-off-by: Kees Cook --- include/linux/string.h | 18 ++++++++++++++++++ lib/memcpy_kunit.c | 11 +++++++++++ 2 files changed, 29 insertions(+) (limited to 'lib') diff --git a/include/linux/string.h b/include/linux/string.h index da490c2154a9..5a36608144a9 100644 --- a/include/linux/string.h +++ b/include/linux/string.h @@ -288,6 +288,24 @@ static inline void memcpy_and_pad(void *dest, size_t dest_len, sizeof(*(obj)) - offsetofend(typeof(*(obj)), member)); \ }) +/** + * memset_startat - Set a value starting at a member to the end of a struct + * + * @obj: Address of target struct instance + * @v: Byte value to repeatedly write + * @member: struct member to start writing at + * + * Note that if there is padding between the prior member and the target + * member, memset_after() should be used to clear the prior padding. + */ +#define memset_startat(obj, v, member) \ +({ \ + u8 *__ptr = (u8 *)(obj); \ + typeof(v) __val = (v); \ + memset(__ptr + offsetof(typeof(*(obj)), member), __val, \ + sizeof(*(obj)) - offsetof(typeof(*(obj)), member)); \ +}) + /** * str_has_prefix - Test if a string has a given prefix * @str: The string to test diff --git a/lib/memcpy_kunit.c b/lib/memcpy_kunit.c index 5c5b4f3221d9..62f8ffcbbaa3 100644 --- a/lib/memcpy_kunit.c +++ b/lib/memcpy_kunit.c @@ -222,6 +222,13 @@ static void memset_test(struct kunit *test) 0x72, 0x72, 0x72, 0x72, 0x72, 0x72, 0x72, 0x72, }, }; + struct some_bytes startat = { + .data = { 0x30, 0x30, 0x30, 0x30, 0x30, 0x30, 0x30, 0x30, + 0x79, 0x79, 0x79, 0x79, 0x79, 0x79, 0x79, 0x79, + 0x79, 0x79, 0x79, 0x79, 0x79, 0x79, 0x79, 0x79, + 0x79, 0x79, 0x79, 0x79, 0x79, 0x79, 0x79, 0x79, + }, + }; struct some_bytes dest = { }; int count, value; u8 *ptr; @@ -258,6 +265,10 @@ static void memset_test(struct kunit *test) memset_after(&dest, 0x72, three); compare("memset_after()", dest, after); + /* Verify memset_startat() */ + dest = control; + memset_startat(&dest, 0x79, four); + compare("memset_startat()", dest, startat); #undef TEST_OP } -- cgit v1.2.3 From 1aec5e4a2962f7e0b3fb3e7308dd726be2472c26 Mon Sep 17 00:00:00 2001 From: Jens Axboe Date: Fri, 8 Oct 2021 05:44:23 -0600 Subject: sbitmap: add helper to clear a batch of tags sbitmap currently only supports clearing tags one-by-one, add a helper that allows the caller to pass in an array of tags to clear. Signed-off-by: Jens Axboe --- include/linux/sbitmap.h | 11 +++++++++++ lib/sbitmap.c | 44 +++++++++++++++++++++++++++++++++++++++++--- 2 files changed, 52 insertions(+), 3 deletions(-) (limited to 'lib') diff --git a/include/linux/sbitmap.h b/include/linux/sbitmap.h index e30b56023ead..4a6ff274335a 100644 --- a/include/linux/sbitmap.h +++ b/include/linux/sbitmap.h @@ -528,6 +528,17 @@ void sbitmap_queue_min_shallow_depth(struct sbitmap_queue *sbq, void sbitmap_queue_clear(struct sbitmap_queue *sbq, unsigned int nr, unsigned int cpu); +/** + * sbitmap_queue_clear_batch() - Free a batch of allocated bits + * &struct sbitmap_queue. + * @sbq: Bitmap to free from. + * @offset: offset for each tag in array + * @tags: array of tags + * @nr_tags: number of tags in array + */ +void sbitmap_queue_clear_batch(struct sbitmap_queue *sbq, int offset, + int *tags, int nr_tags); + static inline int sbq_index_inc(int index) { return (index + 1) & (SBQ_WAIT_QUEUES - 1); diff --git a/lib/sbitmap.c b/lib/sbitmap.c index f398e0ae548e..c6e2f1f2c4d2 100644 --- a/lib/sbitmap.c +++ b/lib/sbitmap.c @@ -628,6 +628,46 @@ void sbitmap_queue_wake_up(struct sbitmap_queue *sbq) } EXPORT_SYMBOL_GPL(sbitmap_queue_wake_up); +static inline void sbitmap_update_cpu_hint(struct sbitmap *sb, int cpu, int tag) +{ + if (likely(!sb->round_robin && tag < sb->depth)) + *per_cpu_ptr(sb->alloc_hint, cpu) = tag; +} + +void sbitmap_queue_clear_batch(struct sbitmap_queue *sbq, int offset, + int *tags, int nr_tags) +{ + struct sbitmap *sb = &sbq->sb; + unsigned long *addr = NULL; + unsigned long mask = 0; + int i; + + smp_mb__before_atomic(); + for (i = 0; i < nr_tags; i++) { + const int tag = tags[i] - offset; + unsigned long *this_addr; + + /* since we're clearing a batch, skip the deferred map */ + this_addr = &sb->map[SB_NR_TO_INDEX(sb, tag)].word; + if (!addr) { + addr = this_addr; + } else if (addr != this_addr) { + atomic_long_andnot(mask, (atomic_long_t *) addr); + mask = 0; + addr = this_addr; + } + mask |= (1UL << SB_NR_TO_BIT(sb, tag)); + } + + if (mask) + atomic_long_andnot(mask, (atomic_long_t *) addr); + + smp_mb__after_atomic(); + sbitmap_queue_wake_up(sbq); + sbitmap_update_cpu_hint(&sbq->sb, raw_smp_processor_id(), + tags[nr_tags - 1] - offset); +} + void sbitmap_queue_clear(struct sbitmap_queue *sbq, unsigned int nr, unsigned int cpu) { @@ -652,9 +692,7 @@ void sbitmap_queue_clear(struct sbitmap_queue *sbq, unsigned int nr, */ smp_mb__after_atomic(); sbitmap_queue_wake_up(sbq); - - if (likely(!sbq->sb.round_robin && nr < sbq->sb.depth)) - *per_cpu_ptr(sbq->sb.alloc_hint, cpu) = nr; + sbitmap_update_cpu_hint(&sbq->sb, cpu, nr); } EXPORT_SYMBOL_GPL(sbitmap_queue_clear); -- cgit v1.2.3 From 83d3c4f22a36d005b55f44628f46cc0d319a75e8 Mon Sep 17 00:00:00 2001 From: Lasse Collin Date: Mon, 11 Oct 2021 05:31:39 +0800 Subject: lib/xz: Avoid overlapping memcpy() with invalid input with in-place decompression With valid files, the safety margin described in lib/decompress_unxz.c ensures that these buffers cannot overlap. But if the uncompressed size of the input is larger than the caller thought, which is possible when the input file is invalid/corrupt, the buffers can overlap. Obviously the result will then be garbage (and usually the decoder will return an error too) but no other harm will happen when such an over-run occurs. This change only affects uncompressed LZMA2 chunks and so this should have no effect on performance. Link: https://lore.kernel.org/r/20211010213145.17462-2-xiang@kernel.org Signed-off-by: Lasse Collin Signed-off-by: Gao Xiang --- lib/decompress_unxz.c | 2 +- lib/xz/xz_dec_lzma2.c | 21 +++++++++++++++++++-- 2 files changed, 20 insertions(+), 3 deletions(-) (limited to 'lib') diff --git a/lib/decompress_unxz.c b/lib/decompress_unxz.c index a2f38e23004a..f7a3dc13316a 100644 --- a/lib/decompress_unxz.c +++ b/lib/decompress_unxz.c @@ -167,7 +167,7 @@ * memeq and memzero are not used much and any remotely sane implementation * is fast enough. memcpy/memmove speed matters in multi-call mode, but * the kernel image is decompressed in single-call mode, in which only - * memcpy speed can matter and only if there is a lot of uncompressible data + * memmove speed can matter and only if there is a lot of uncompressible data * (LZMA2 stores uncompressible chunks in uncompressed form). Thus, the * functions below should just be kept small; it's probably not worth * optimizing for speed. diff --git a/lib/xz/xz_dec_lzma2.c b/lib/xz/xz_dec_lzma2.c index 7a6781e3f47b..d548cf0e59fe 100644 --- a/lib/xz/xz_dec_lzma2.c +++ b/lib/xz/xz_dec_lzma2.c @@ -387,7 +387,14 @@ static void dict_uncompressed(struct dictionary *dict, struct xz_buf *b, *left -= copy_size; - memcpy(dict->buf + dict->pos, b->in + b->in_pos, copy_size); + /* + * If doing in-place decompression in single-call mode and the + * uncompressed size of the file is larger than the caller + * thought (i.e. it is invalid input!), the buffers below may + * overlap and cause undefined behavior with memcpy(). + * With valid inputs memcpy() would be fine here. + */ + memmove(dict->buf + dict->pos, b->in + b->in_pos, copy_size); dict->pos += copy_size; if (dict->full < dict->pos) @@ -397,7 +404,11 @@ static void dict_uncompressed(struct dictionary *dict, struct xz_buf *b, if (dict->pos == dict->end) dict->pos = 0; - memcpy(b->out + b->out_pos, b->in + b->in_pos, + /* + * Like above but for multi-call mode: use memmove() + * to avoid undefined behavior with invalid input. + */ + memmove(b->out + b->out_pos, b->in + b->in_pos, copy_size); } @@ -421,6 +432,12 @@ static uint32_t dict_flush(struct dictionary *dict, struct xz_buf *b) if (dict->pos == dict->end) dict->pos = 0; + /* + * These buffers cannot overlap even if doing in-place + * decompression because in multi-call mode dict->buf + * has been allocated by us in this file; it's not + * provided by the caller like in single-call mode. + */ memcpy(b->out + b->out_pos, dict->buf + dict->start, copy_size); } -- cgit v1.2.3 From 4f8d7abaa413c34da9d751289849dbfb7c977d05 Mon Sep 17 00:00:00 2001 From: Lasse Collin Date: Mon, 11 Oct 2021 05:31:40 +0800 Subject: lib/xz: Validate the value before assigning it to an enum variable This might matter, for example, if the underlying type of enum xz_check was a signed char. In such a case the validation wouldn't have caught an unsupported header. I don't know if this problem can occur in the kernel on any arch but it's still good to fix it because some people might copy the XZ code to their own projects from Linux instead of the upstream XZ Embedded repository. This change may increase the code size by a few bytes. An alternative would have been to use an unsigned int instead of enum xz_check but using an enumeration looks cleaner. Link: https://lore.kernel.org/r/20211010213145.17462-3-xiang@kernel.org Signed-off-by: Lasse Collin Signed-off-by: Gao Xiang --- lib/xz/xz_dec_stream.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) (limited to 'lib') diff --git a/lib/xz/xz_dec_stream.c b/lib/xz/xz_dec_stream.c index fea86deaaa01..683570b93a8c 100644 --- a/lib/xz/xz_dec_stream.c +++ b/lib/xz/xz_dec_stream.c @@ -402,12 +402,12 @@ static enum xz_ret dec_stream_header(struct xz_dec *s) * we will accept other check types too, but then the check won't * be verified and a warning (XZ_UNSUPPORTED_CHECK) will be given. */ + if (s->temp.buf[HEADER_MAGIC_SIZE + 1] > XZ_CHECK_MAX) + return XZ_OPTIONS_ERROR; + s->check_type = s->temp.buf[HEADER_MAGIC_SIZE + 1]; #ifdef XZ_DEC_ANY_CHECK - if (s->check_type > XZ_CHECK_MAX) - return XZ_OPTIONS_ERROR; - if (s->check_type > XZ_CHECK_CRC32) return XZ_UNSUPPORTED_CHECK; #else -- cgit v1.2.3 From a98a25408b0e9b0264abcc3dabfafd9ff2ea1046 Mon Sep 17 00:00:00 2001 From: Lasse Collin Date: Mon, 11 Oct 2021 05:31:41 +0800 Subject: lib/xz: Move s->lzma.len = 0 initialization to lzma_reset() It's a more logical place even if the resetting needs to be done only once per LZMA2 stream (if lzma_reset() called in the middle of an LZMA2 stream, .len will already be 0). Link: https://lore.kernel.org/r/20211010213145.17462-4-xiang@kernel.org Signed-off-by: Lasse Collin Signed-off-by: Gao Xiang --- lib/xz/xz_dec_lzma2.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) (limited to 'lib') diff --git a/lib/xz/xz_dec_lzma2.c b/lib/xz/xz_dec_lzma2.c index d548cf0e59fe..22b789645ce5 100644 --- a/lib/xz/xz_dec_lzma2.c +++ b/lib/xz/xz_dec_lzma2.c @@ -791,6 +791,7 @@ static void lzma_reset(struct xz_dec_lzma2 *s) s->lzma.rep1 = 0; s->lzma.rep2 = 0; s->lzma.rep3 = 0; + s->lzma.len = 0; /* * All probabilities are initialized to the same value. This hack @@ -1174,8 +1175,6 @@ XZ_EXTERN enum xz_ret xz_dec_lzma2_reset(struct xz_dec_lzma2 *s, uint8_t props) } } - s->lzma.len = 0; - s->lzma2.sequence = SEQ_CONTROL; s->lzma2.need_dict_reset = true; -- cgit v1.2.3 From aaa2975f2b07b04ee16b2cad1072cbdea3e1c50a Mon Sep 17 00:00:00 2001 From: Lasse Collin Date: Mon, 11 Oct 2021 05:31:42 +0800 Subject: lib/xz: Add MicroLZMA decoder MicroLZMA is a yet another header format variant where the first byte of a raw LZMA stream (without the end of stream marker) has been replaced with a bitwise-negation of the lc/lp/pb properties byte. MicroLZMA was created to be used in EROFS but can be used by other things too where wasting minimal amount of space for headers is important. This is implemented using most of the LZMA2 code as is so the amount of new code is small. The API has a few extra features compared to the XZ decoder. On the other hand, the API lacks XZ_BUF_ERROR support which is important to take into account when using this API. MicroLZMA doesn't support BCJ filters. In theory they could be added later as there are many unused/reserved values for the first byte of the compressed stream but in practice it is somewhat unlikely to happen due to a few implementation reasons. Link: https://lore.kernel.org/r/20211010213145.17462-5-xiang@kernel.org Signed-off-by: Lasse Collin Signed-off-by: Gao Xiang --- include/linux/xz.h | 106 ++++++++++++++++++++++++++++++++++ lib/xz/Kconfig | 13 +++++ lib/xz/xz_dec_lzma2.c | 156 +++++++++++++++++++++++++++++++++++++++++++++++++- lib/xz/xz_dec_syms.c | 9 ++- lib/xz/xz_private.h | 3 + 5 files changed, 284 insertions(+), 3 deletions(-) (limited to 'lib') diff --git a/include/linux/xz.h b/include/linux/xz.h index 9884c8440188..7285ca5d56e9 100644 --- a/include/linux/xz.h +++ b/include/linux/xz.h @@ -233,6 +233,112 @@ XZ_EXTERN void xz_dec_reset(struct xz_dec *s); */ XZ_EXTERN void xz_dec_end(struct xz_dec *s); +/* + * Decompressor for MicroLZMA, an LZMA variant with a very minimal header. + * See xz_dec_microlzma_alloc() below for details. + * + * These functions aren't used or available in preboot code and thus aren't + * marked with XZ_EXTERN. This avoids warnings about static functions that + * are never defined. + */ +/** + * struct xz_dec_microlzma - Opaque type to hold the MicroLZMA decoder state + */ +struct xz_dec_microlzma; + +/** + * xz_dec_microlzma_alloc() - Allocate memory for the MicroLZMA decoder + * @mode XZ_SINGLE or XZ_PREALLOC + * @dict_size LZMA dictionary size. This must be at least 4 KiB and + * at most 3 GiB. + * + * In contrast to xz_dec_init(), this function only allocates the memory + * and remembers the dictionary size. xz_dec_microlzma_reset() must be used + * before calling xz_dec_microlzma_run(). + * + * The amount of allocated memory is a little less than 30 KiB with XZ_SINGLE. + * With XZ_PREALLOC also a dictionary buffer of dict_size bytes is allocated. + * + * On success, xz_dec_microlzma_alloc() returns a pointer to + * struct xz_dec_microlzma. If memory allocation fails or + * dict_size is invalid, NULL is returned. + * + * The compressed format supported by this decoder is a raw LZMA stream + * whose first byte (always 0x00) has been replaced with bitwise-negation + * of the LZMA properties (lc/lp/pb) byte. For example, if lc/lp/pb is + * 3/0/2, the first byte is 0xA2. This way the first byte can never be 0x00. + * Just like with LZMA2, lc + lp <= 4 must be true. The LZMA end-of-stream + * marker must not be used. The unused values are reserved for future use. + * This MicroLZMA header format was created for use in EROFS but may be used + * by others too. + */ +extern struct xz_dec_microlzma *xz_dec_microlzma_alloc(enum xz_mode mode, + uint32_t dict_size); + +/** + * xz_dec_microlzma_reset() - Reset the MicroLZMA decoder state + * @s Decoder state allocated using xz_dec_microlzma_alloc() + * @comp_size Compressed size of the input stream + * @uncomp_size Uncompressed size of the input stream. A value smaller + * than the real uncompressed size of the input stream can + * be specified if uncomp_size_is_exact is set to false. + * uncomp_size can never be set to a value larger than the + * expected real uncompressed size because it would eventually + * result in XZ_DATA_ERROR. + * @uncomp_size_is_exact This is an int instead of bool to avoid + * requiring stdbool.h. This should normally be set to true. + * When this is set to false, error detection is weaker. + */ +extern void xz_dec_microlzma_reset(struct xz_dec_microlzma *s, + uint32_t comp_size, uint32_t uncomp_size, + int uncomp_size_is_exact); + +/** + * xz_dec_microlzma_run() - Run the MicroLZMA decoder + * @s Decoder state initialized using xz_dec_microlzma_reset() + * @b: Input and output buffers + * + * This works similarly to xz_dec_run() with a few important differences. + * Only the differences are documented here. + * + * The only possible return values are XZ_OK, XZ_STREAM_END, and + * XZ_DATA_ERROR. This function cannot return XZ_BUF_ERROR: if no progress + * is possible due to lack of input data or output space, this function will + * keep returning XZ_OK. Thus, the calling code must be written so that it + * will eventually provide input and output space matching (or exceeding) + * comp_size and uncomp_size arguments given to xz_dec_microlzma_reset(). + * If the caller cannot do this (for example, if the input file is truncated + * or otherwise corrupt), the caller must detect this error by itself to + * avoid an infinite loop. + * + * If the compressed data seems to be corrupt, XZ_DATA_ERROR is returned. + * This can happen also when incorrect dictionary, uncompressed, or + * compressed sizes have been specified. + * + * With XZ_PREALLOC only: As an extra feature, b->out may be NULL to skip over + * uncompressed data. This way the caller doesn't need to provide a temporary + * output buffer for the bytes that will be ignored. + * + * With XZ_SINGLE only: In contrast to xz_dec_run(), the return value XZ_OK + * is also possible and thus XZ_SINGLE is actually a limited multi-call mode. + * After XZ_OK the bytes decoded so far may be read from the output buffer. + * It is possible to continue decoding but the variables b->out and b->out_pos + * MUST NOT be changed by the caller. Increasing the value of b->out_size is + * allowed to make more output space available; one doesn't need to provide + * space for the whole uncompressed data on the first call. The input buffer + * may be changed normally like with XZ_PREALLOC. This way input data can be + * provided from non-contiguous memory. + */ +extern enum xz_ret xz_dec_microlzma_run(struct xz_dec_microlzma *s, + struct xz_buf *b); + +/** + * xz_dec_microlzma_end() - Free the memory allocated for the decoder state + * @s: Decoder state allocated using xz_dec_microlzma_alloc(). + * If s is NULL, this function does nothing. + */ +extern void xz_dec_microlzma_end(struct xz_dec_microlzma *s); + /* * Standalone build (userspace build or in-kernel build for boot time use) * needs a CRC32 implementation. For normal in-kernel use, kernel's own diff --git a/lib/xz/Kconfig b/lib/xz/Kconfig index 5cb50245a878..adce22ac18d6 100644 --- a/lib/xz/Kconfig +++ b/lib/xz/Kconfig @@ -39,6 +39,19 @@ config XZ_DEC_SPARC default y select XZ_DEC_BCJ +config XZ_DEC_MICROLZMA + bool "MicroLZMA decoder" + default n + help + MicroLZMA is a header format variant where the first byte + of a raw LZMA stream (without the end of stream marker) has + been replaced with a bitwise-negation of the lc/lp/pb + properties byte. MicroLZMA was created to be used in EROFS + but can be used by other things too where wasting minimal + amount of space for headers is important. + + Unless you know that you need this, say N. + endif config XZ_DEC_BCJ diff --git a/lib/xz/xz_dec_lzma2.c b/lib/xz/xz_dec_lzma2.c index 22b789645ce5..46b186d7eb45 100644 --- a/lib/xz/xz_dec_lzma2.c +++ b/lib/xz/xz_dec_lzma2.c @@ -248,6 +248,10 @@ struct lzma2_dec { * before the first LZMA chunk. */ bool need_props; + +#ifdef XZ_DEC_MICROLZMA + bool pedantic_microlzma; +#endif }; struct xz_dec_lzma2 { @@ -419,6 +423,12 @@ static void dict_uncompressed(struct dictionary *dict, struct xz_buf *b, } } +#ifdef XZ_DEC_MICROLZMA +# define DICT_FLUSH_SUPPORTS_SKIPPING true +#else +# define DICT_FLUSH_SUPPORTS_SKIPPING false +#endif + /* * Flush pending data from dictionary to b->out. It is assumed that there is * enough space in b->out. This is guaranteed because caller uses dict_limit() @@ -437,9 +447,14 @@ static uint32_t dict_flush(struct dictionary *dict, struct xz_buf *b) * decompression because in multi-call mode dict->buf * has been allocated by us in this file; it's not * provided by the caller like in single-call mode. + * + * With MicroLZMA, b->out can be NULL to skip bytes that + * the caller doesn't need. This cannot be done with XZ + * because it would break BCJ filters. */ - memcpy(b->out + b->out_pos, dict->buf + dict->start, - copy_size); + if (!DICT_FLUSH_SUPPORTS_SKIPPING || b->out != NULL) + memcpy(b->out + b->out_pos, dict->buf + dict->start, + copy_size); } dict->start = dict->pos; @@ -1190,3 +1205,140 @@ XZ_EXTERN void xz_dec_lzma2_end(struct xz_dec_lzma2 *s) kfree(s); } + +#ifdef XZ_DEC_MICROLZMA +/* This is a wrapper struct to have a nice struct name in the public API. */ +struct xz_dec_microlzma { + struct xz_dec_lzma2 s; +}; + +enum xz_ret xz_dec_microlzma_run(struct xz_dec_microlzma *s_ptr, + struct xz_buf *b) +{ + struct xz_dec_lzma2 *s = &s_ptr->s; + + /* + * sequence is SEQ_PROPERTIES before the first input byte, + * SEQ_LZMA_PREPARE until a total of five bytes have been read, + * and SEQ_LZMA_RUN for the rest of the input stream. + */ + if (s->lzma2.sequence != SEQ_LZMA_RUN) { + if (s->lzma2.sequence == SEQ_PROPERTIES) { + /* One byte is needed for the props. */ + if (b->in_pos >= b->in_size) + return XZ_OK; + + /* + * Don't increment b->in_pos here. The same byte is + * also passed to rc_read_init() which will ignore it. + */ + if (!lzma_props(s, ~b->in[b->in_pos])) + return XZ_DATA_ERROR; + + s->lzma2.sequence = SEQ_LZMA_PREPARE; + } + + /* + * xz_dec_microlzma_reset() doesn't validate the compressed + * size so we do it here. We have to limit the maximum size + * to avoid integer overflows in lzma2_lzma(). 3 GiB is a nice + * round number and much more than users of this code should + * ever need. + */ + if (s->lzma2.compressed < RC_INIT_BYTES + || s->lzma2.compressed > (3U << 30)) + return XZ_DATA_ERROR; + + if (!rc_read_init(&s->rc, b)) + return XZ_OK; + + s->lzma2.compressed -= RC_INIT_BYTES; + s->lzma2.sequence = SEQ_LZMA_RUN; + + dict_reset(&s->dict, b); + } + + /* This is to allow increasing b->out_size between calls. */ + if (DEC_IS_SINGLE(s->dict.mode)) + s->dict.end = b->out_size - b->out_pos; + + while (true) { + dict_limit(&s->dict, min_t(size_t, b->out_size - b->out_pos, + s->lzma2.uncompressed)); + + if (!lzma2_lzma(s, b)) + return XZ_DATA_ERROR; + + s->lzma2.uncompressed -= dict_flush(&s->dict, b); + + if (s->lzma2.uncompressed == 0) { + if (s->lzma2.pedantic_microlzma) { + if (s->lzma2.compressed > 0 || s->lzma.len > 0 + || !rc_is_finished(&s->rc)) + return XZ_DATA_ERROR; + } + + return XZ_STREAM_END; + } + + if (b->out_pos == b->out_size) + return XZ_OK; + + if (b->in_pos == b->in_size + && s->temp.size < s->lzma2.compressed) + return XZ_OK; + } +} + +struct xz_dec_microlzma *xz_dec_microlzma_alloc(enum xz_mode mode, + uint32_t dict_size) +{ + struct xz_dec_microlzma *s; + + /* Restrict dict_size to the same range as in the LZMA2 code. */ + if (dict_size < 4096 || dict_size > (3U << 30)) + return NULL; + + s = kmalloc(sizeof(*s), GFP_KERNEL); + if (s == NULL) + return NULL; + + s->s.dict.mode = mode; + s->s.dict.size = dict_size; + + if (DEC_IS_MULTI(mode)) { + s->s.dict.end = dict_size; + + s->s.dict.buf = vmalloc(dict_size); + if (s->s.dict.buf == NULL) { + kfree(s); + return NULL; + } + } + + return s; +} + +void xz_dec_microlzma_reset(struct xz_dec_microlzma *s, uint32_t comp_size, + uint32_t uncomp_size, int uncomp_size_is_exact) +{ + /* + * comp_size is validated in xz_dec_microlzma_run(). + * uncomp_size can safely be anything. + */ + s->s.lzma2.compressed = comp_size; + s->s.lzma2.uncompressed = uncomp_size; + s->s.lzma2.pedantic_microlzma = uncomp_size_is_exact; + + s->s.lzma2.sequence = SEQ_PROPERTIES; + s->s.temp.size = 0; +} + +void xz_dec_microlzma_end(struct xz_dec_microlzma *s) +{ + if (DEC_IS_MULTI(s->s.dict.mode)) + vfree(s->s.dict.buf); + + kfree(s); +} +#endif diff --git a/lib/xz/xz_dec_syms.c b/lib/xz/xz_dec_syms.c index 32eb3c03aede..61098c67a413 100644 --- a/lib/xz/xz_dec_syms.c +++ b/lib/xz/xz_dec_syms.c @@ -15,8 +15,15 @@ EXPORT_SYMBOL(xz_dec_reset); EXPORT_SYMBOL(xz_dec_run); EXPORT_SYMBOL(xz_dec_end); +#ifdef CONFIG_XZ_DEC_MICROLZMA +EXPORT_SYMBOL(xz_dec_microlzma_alloc); +EXPORT_SYMBOL(xz_dec_microlzma_reset); +EXPORT_SYMBOL(xz_dec_microlzma_run); +EXPORT_SYMBOL(xz_dec_microlzma_end); +#endif + MODULE_DESCRIPTION("XZ decompressor"); -MODULE_VERSION("1.0"); +MODULE_VERSION("1.1"); MODULE_AUTHOR("Lasse Collin and Igor Pavlov"); /* diff --git a/lib/xz/xz_private.h b/lib/xz/xz_private.h index 09360ebb510e..bf1e94ec7873 100644 --- a/lib/xz/xz_private.h +++ b/lib/xz/xz_private.h @@ -37,6 +37,9 @@ # ifdef CONFIG_XZ_DEC_SPARC # define XZ_DEC_SPARC # endif +# ifdef CONFIG_XZ_DEC_MICROLZMA +# define XZ_DEC_MICROLZMA +# endif # define memeq(a, b, size) (memcmp(a, b, size) == 0) # define memzero(buf, size) memset(buf, 0, size) # endif -- cgit v1.2.3 From 0a434e0a2c9f4395e4560aac22677ef25ab4afd9 Mon Sep 17 00:00:00 2001 From: Lasse Collin Date: Mon, 11 Oct 2021 05:31:43 +0800 Subject: lib/xz, lib/decompress_unxz.c: Fix spelling in comments uncompressible -> incompressible non-splitted -> non-split Link: https://lore.kernel.org/r/20211010213145.17462-6-xiang@kernel.org Signed-off-by: Lasse Collin Signed-off-by: Gao Xiang --- lib/decompress_unxz.c | 10 +++++----- lib/xz/xz_dec_lzma2.c | 2 +- 2 files changed, 6 insertions(+), 6 deletions(-) (limited to 'lib') diff --git a/lib/decompress_unxz.c b/lib/decompress_unxz.c index f7a3dc13316a..9f4262ee33a5 100644 --- a/lib/decompress_unxz.c +++ b/lib/decompress_unxz.c @@ -20,8 +20,8 @@ * * The worst case for in-place decompression is that the beginning of * the file is compressed extremely well, and the rest of the file is - * uncompressible. Thus, we must look for worst-case expansion when the - * compressor is encoding uncompressible data. + * incompressible. Thus, we must look for worst-case expansion when the + * compressor is encoding incompressible data. * * The structure of the .xz file in case of a compressed kernel is as follows. * Sizes (as bytes) of the fields are in parenthesis. @@ -58,7 +58,7 @@ * uncompressed size of the payload is in practice never less than the * payload size itself. The LZMA2 format would allow uncompressed size * to be less than the payload size, but no sane compressor creates such - * files. LZMA2 supports storing uncompressible data in uncompressed form, + * files. LZMA2 supports storing incompressible data in uncompressed form, * so there's never a need to create payloads whose uncompressed size is * smaller than the compressed size. * @@ -167,8 +167,8 @@ * memeq and memzero are not used much and any remotely sane implementation * is fast enough. memcpy/memmove speed matters in multi-call mode, but * the kernel image is decompressed in single-call mode, in which only - * memmove speed can matter and only if there is a lot of uncompressible data - * (LZMA2 stores uncompressible chunks in uncompressed form). Thus, the + * memmove speed can matter and only if there is a lot of incompressible data + * (LZMA2 stores incompressible chunks in uncompressed form). Thus, the * functions below should just be kept small; it's probably not worth * optimizing for speed. */ diff --git a/lib/xz/xz_dec_lzma2.c b/lib/xz/xz_dec_lzma2.c index 46b186d7eb45..27ce34520e78 100644 --- a/lib/xz/xz_dec_lzma2.c +++ b/lib/xz/xz_dec_lzma2.c @@ -520,7 +520,7 @@ static __always_inline void rc_normalize(struct rc_dec *rc) * functions so that the compiler is supposed to be able to more easily avoid * an extra branch. In this particular version of the LZMA decoder, this * doesn't seem to be a good idea (tested with GCC 3.3.6, 3.4.6, and 4.3.3 - * on x86). Using a non-splitted version results in nicer looking code too. + * on x86). Using a non-split version results in nicer looking code too. * * NOTE: This must return an int. Do not make it return a bool or the speed * of the code generated by GCC 3.x decreases 10-15 %. (GCC 4.3 doesn't care, -- cgit v1.2.3 From b7cbaef303c7b9f26c647bcba72da04dd35396c4 Mon Sep 17 00:00:00 2001 From: Daniel Latypov Date: Tue, 14 Sep 2021 14:03:47 -0700 Subject: kunit: drop assumption in kunit-log-test about current suite This test assumes that the declared kunit_suite object is the exact one which is being executed, which KUnit will not guarantee [1]. Specifically, `suite->log` is not initialized until a suite object is executed. So if KUnit makes a copy of the suite and runs that instead, this test dereferences an invalid pointer and (hopefully) segfaults. N.B. since we no longer assume this, we can no longer verify that `suite->log` is *not* allocated during normal execution. An alternative to this patch that would allow us to test that would require exposing an API for the current test to get its current suite. Exposing that for one internal kunit test seems like overkill, and grants users more footguns (e.g. reusing a test case in multiple suites and changing behavior based on the suite name, dynamically modifying the setup/cleanup funcs, storing/reading stuff out of the suite->log, etc.). [1] In a subsequent patch, KUnit will allow running subsets of test cases within a suite by making a copy of the suite w/ the filtered test list. But there are other reasons KUnit might execute a copy, e.g. if it ever wants to support parallel execution of different suites, recovering from errors and restarting suites Signed-off-by: Daniel Latypov Reviewed-by: Brendan Higgins Signed-off-by: Shuah Khan --- lib/kunit/kunit-test.c | 14 ++++++++------ 1 file changed, 8 insertions(+), 6 deletions(-) (limited to 'lib') diff --git a/lib/kunit/kunit-test.c b/lib/kunit/kunit-test.c index d69efcbed624..555601d17f79 100644 --- a/lib/kunit/kunit-test.c +++ b/lib/kunit/kunit-test.c @@ -415,12 +415,15 @@ static struct kunit_suite kunit_log_test_suite = { static void kunit_log_test(struct kunit *test) { - struct kunit_suite *suite = &kunit_log_test_suite; + struct kunit_suite suite; + + suite.log = kunit_kzalloc(test, KUNIT_LOG_SIZE, GFP_KERNEL); + KUNIT_ASSERT_NOT_ERR_OR_NULL(test, suite.log); kunit_log(KERN_INFO, test, "put this in log."); kunit_log(KERN_INFO, test, "this too."); - kunit_log(KERN_INFO, suite, "add to suite log."); - kunit_log(KERN_INFO, suite, "along with this."); + kunit_log(KERN_INFO, &suite, "add to suite log."); + kunit_log(KERN_INFO, &suite, "along with this."); #ifdef CONFIG_KUNIT_DEBUGFS KUNIT_EXPECT_NOT_ERR_OR_NULL(test, @@ -428,12 +431,11 @@ static void kunit_log_test(struct kunit *test) KUNIT_EXPECT_NOT_ERR_OR_NULL(test, strstr(test->log, "this too.")); KUNIT_EXPECT_NOT_ERR_OR_NULL(test, - strstr(suite->log, "add to suite log.")); + strstr(suite.log, "add to suite log.")); KUNIT_EXPECT_NOT_ERR_OR_NULL(test, - strstr(suite->log, "along with this.")); + strstr(suite.log, "along with this.")); #else KUNIT_EXPECT_PTR_EQ(test, test->log, (char *)NULL); - KUNIT_EXPECT_PTR_EQ(test, suite->log, (char *)NULL); #endif } -- cgit v1.2.3 From a127b154a8f231709754b5d56a501163dd837459 Mon Sep 17 00:00:00 2001 From: Daniel Latypov Date: Tue, 14 Sep 2021 14:03:48 -0700 Subject: kunit: tool: allow filtering test cases via glob Commit 1d71307a6f94 ("kunit: add unit test for filtering suites by names") introduced the ability to filter which suites we run via glob. This change extends it so we can also filter individual test cases inside of suites as well. This is quite useful when, e.g. * trying to run just the tests cases you've just added or are working on * trying to debug issues with test hermeticity Examples: $ ./tools/testing/kunit/kunit.py run --kunitconfig=lib/kunit '*exec*.parse*' ... ============================================================ ======== [PASSED] kunit_executor_test ======== [PASSED] parse_filter_test ============================================================ Testing complete. 1 tests run. 0 failed. 0 crashed. $ ./tools/testing/kunit/kunit.py run --kunitconfig=lib/kunit '*.no_matching_tests' ... [ERROR] no tests run! Signed-off-by: Daniel Latypov Reviewed-by: David Gow Reviewed-by: Brendan Higgins Signed-off-by: Shuah Khan --- Documentation/dev-tools/kunit/running_tips.rst | 11 ++- lib/kunit/executor.c | 107 +++++++++++++++++++++--- lib/kunit/executor_test.c | 109 ++++++++++++++++++++++--- tools/testing/kunit/kunit.py | 5 +- 4 files changed, 203 insertions(+), 29 deletions(-) (limited to 'lib') diff --git a/Documentation/dev-tools/kunit/running_tips.rst b/Documentation/dev-tools/kunit/running_tips.rst index 30d2147eb5b5..7b6d26a25959 100644 --- a/Documentation/dev-tools/kunit/running_tips.rst +++ b/Documentation/dev-tools/kunit/running_tips.rst @@ -25,8 +25,8 @@ It can be handy to create a bash function like: Running a subset of tests ------------------------- -``kunit.py run`` accepts an optional glob argument to filter tests. Currently -this only matches against suite names, but this may change in the future. +``kunit.py run`` accepts an optional glob argument to filter tests. The format +is ``"[.test_glob]"``. Say that we wanted to run the sysctl tests, we could do so via: @@ -35,6 +35,13 @@ Say that we wanted to run the sysctl tests, we could do so via: $ echo -e 'CONFIG_KUNIT=y\nCONFIG_KUNIT_ALL_TESTS=y' > .kunit/.kunitconfig $ ./tools/testing/kunit/kunit.py run 'sysctl*' +We can filter down to just the "write" tests via: + +.. code-block:: bash + + $ echo -e 'CONFIG_KUNIT=y\nCONFIG_KUNIT_ALL_TESTS=y' > .kunit/.kunitconfig + $ ./tools/testing/kunit/kunit.py run 'sysctl*.*write*' + We're paying the cost of building more tests than we need this way, but it's easier than fiddling with ``.kunitconfig`` files or commenting out ``kunit_suite``'s. diff --git a/lib/kunit/executor.c b/lib/kunit/executor.c index acd1de436f59..bab3ab940acc 100644 --- a/lib/kunit/executor.c +++ b/lib/kunit/executor.c @@ -17,21 +17,80 @@ extern struct kunit_suite * const * const __kunit_suites_end[]; static char *filter_glob_param; module_param_named(filter_glob, filter_glob_param, charp, 0); MODULE_PARM_DESC(filter_glob, - "Filter which KUnit test suites run at boot-time, e.g. list*"); + "Filter which KUnit test suites/tests run at boot-time, e.g. list* or list*.*del_test"); + +/* glob_match() needs NULL terminated strings, so we need a copy of filter_glob_param. */ +struct kunit_test_filter { + char *suite_glob; + char *test_glob; +}; + +/* Split "suite_glob.test_glob" into two. Assumes filter_glob is not empty. */ +static void kunit_parse_filter_glob(struct kunit_test_filter *parsed, + const char *filter_glob) +{ + const int len = strlen(filter_glob); + const char *period = strchr(filter_glob, '.'); + + if (!period) { + parsed->suite_glob = kmalloc(len, GFP_KERNEL); + parsed->test_glob = NULL; + strcpy(parsed->suite_glob, filter_glob); + return; + } + + parsed->suite_glob = kzalloc(period - filter_glob + 1, GFP_KERNEL); + parsed->test_glob = kzalloc(len - (period - filter_glob) + 1, GFP_KERNEL); + + strncpy(parsed->suite_glob, filter_glob, period - filter_glob); + strncpy(parsed->test_glob, period + 1, len - (period - filter_glob)); +} + +/* Create a copy of suite with only tests that match test_glob. */ +static struct kunit_suite * +kunit_filter_tests(struct kunit_suite *const suite, const char *test_glob) +{ + int n = 0; + struct kunit_case *filtered, *test_case; + struct kunit_suite *copy; + + kunit_suite_for_each_test_case(suite, test_case) { + if (!test_glob || glob_match(test_glob, test_case->name)) + ++n; + } + + if (n == 0) + return NULL; + + /* Use memcpy to workaround copy->name being const. */ + copy = kmalloc(sizeof(*copy), GFP_KERNEL); + memcpy(copy, suite, sizeof(*copy)); + + filtered = kcalloc(n + 1, sizeof(*filtered), GFP_KERNEL); + + n = 0; + kunit_suite_for_each_test_case(suite, test_case) { + if (!test_glob || glob_match(test_glob, test_case->name)) + filtered[n++] = *test_case; + } + + copy->test_cases = filtered; + return copy; +} static char *kunit_shutdown; core_param(kunit_shutdown, kunit_shutdown, charp, 0644); static struct kunit_suite * const * kunit_filter_subsuite(struct kunit_suite * const * const subsuite, - const char *filter_glob) + struct kunit_test_filter *filter) { int i, n = 0; - struct kunit_suite **filtered; + struct kunit_suite **filtered, *filtered_suite; n = 0; - for (i = 0; subsuite[i] != NULL; ++i) { - if (glob_match(filter_glob, subsuite[i]->name)) + for (i = 0; subsuite[i]; ++i) { + if (glob_match(filter->suite_glob, subsuite[i]->name)) ++n; } @@ -44,8 +103,11 @@ kunit_filter_subsuite(struct kunit_suite * const * const subsuite, n = 0; for (i = 0; subsuite[i] != NULL; ++i) { - if (glob_match(filter_glob, subsuite[i]->name)) - filtered[n++] = subsuite[i]; + if (!glob_match(filter->suite_glob, subsuite[i]->name)) + continue; + filtered_suite = kunit_filter_tests(subsuite[i], filter->test_glob); + if (filtered_suite) + filtered[n++] = filtered_suite; } filtered[n] = NULL; @@ -57,12 +119,32 @@ struct suite_set { struct kunit_suite * const * const *end; }; +static void kunit_free_subsuite(struct kunit_suite * const *subsuite) +{ + unsigned int i; + + for (i = 0; subsuite[i]; i++) + kfree(subsuite[i]); + + kfree(subsuite); +} + +static void kunit_free_suite_set(struct suite_set suite_set) +{ + struct kunit_suite * const * const *suites; + + for (suites = suite_set.start; suites < suite_set.end; suites++) + kunit_free_subsuite(*suites); + kfree(suite_set.start); +} + static struct suite_set kunit_filter_suites(const struct suite_set *suite_set, const char *filter_glob) { int i; struct kunit_suite * const **copy, * const *filtered_subsuite; struct suite_set filtered; + struct kunit_test_filter filter; const size_t max = suite_set->end - suite_set->start; @@ -73,12 +155,17 @@ static struct suite_set kunit_filter_suites(const struct suite_set *suite_set, return filtered; } + kunit_parse_filter_glob(&filter, filter_glob); + for (i = 0; i < max; ++i) { - filtered_subsuite = kunit_filter_subsuite(suite_set->start[i], filter_glob); + filtered_subsuite = kunit_filter_subsuite(suite_set->start[i], &filter); if (filtered_subsuite) *copy++ = filtered_subsuite; } filtered.end = copy; + + kfree(filter.suite_glob); + kfree(filter.test_glob); return filtered; } @@ -126,9 +213,7 @@ int kunit_run_all_tests(void) __kunit_test_suites_init(*suites); if (filter_glob_param) { /* a copy was made of each array */ - for (suites = suite_set.start; suites < suite_set.end; suites++) - kfree(*suites); - kfree(suite_set.start); + kunit_free_suite_set(suite_set); } kunit_handle_shutdown(); diff --git a/lib/kunit/executor_test.c b/lib/kunit/executor_test.c index e14a18af573d..edbd8184dcd7 100644 --- a/lib/kunit/executor_test.c +++ b/lib/kunit/executor_test.c @@ -9,38 +9,103 @@ #include static void kfree_at_end(struct kunit *test, const void *to_free); +static void free_subsuite_at_end(struct kunit *test, + struct kunit_suite *const *to_free); static struct kunit_suite *alloc_fake_suite(struct kunit *test, - const char *suite_name); + const char *suite_name, + struct kunit_case *test_cases); + +static void dummy_test(struct kunit *test) {} + +static struct kunit_case dummy_test_cases[] = { + /* .run_case is not important, just needs to be non-NULL */ + { .name = "test1", .run_case = dummy_test }, + { .name = "test2", .run_case = dummy_test }, + {}, +}; + +static void parse_filter_test(struct kunit *test) +{ + struct kunit_test_filter filter = {NULL, NULL}; + + kunit_parse_filter_glob(&filter, "suite"); + KUNIT_EXPECT_STREQ(test, filter.suite_glob, "suite"); + KUNIT_EXPECT_FALSE(test, filter.test_glob); + kfree(filter.suite_glob); + kfree(filter.test_glob); + + kunit_parse_filter_glob(&filter, "suite.test"); + KUNIT_EXPECT_STREQ(test, filter.suite_glob, "suite"); + KUNIT_EXPECT_STREQ(test, filter.test_glob, "test"); + kfree(filter.suite_glob); + kfree(filter.test_glob); +} static void filter_subsuite_test(struct kunit *test) { struct kunit_suite *subsuite[3] = {NULL, NULL, NULL}; struct kunit_suite * const *filtered; + struct kunit_test_filter filter = { + .suite_glob = "suite2", + .test_glob = NULL, + }; - subsuite[0] = alloc_fake_suite(test, "suite1"); - subsuite[1] = alloc_fake_suite(test, "suite2"); + subsuite[0] = alloc_fake_suite(test, "suite1", dummy_test_cases); + subsuite[1] = alloc_fake_suite(test, "suite2", dummy_test_cases); /* Want: suite1, suite2, NULL -> suite2, NULL */ - filtered = kunit_filter_subsuite(subsuite, "suite2*"); + filtered = kunit_filter_subsuite(subsuite, &filter); KUNIT_ASSERT_NOT_ERR_OR_NULL(test, filtered); - kfree_at_end(test, filtered); + free_subsuite_at_end(test, filtered); + /* Validate we just have suite2 */ KUNIT_ASSERT_NOT_ERR_OR_NULL(test, filtered[0]); KUNIT_EXPECT_STREQ(test, (const char *)filtered[0]->name, "suite2"); + KUNIT_EXPECT_FALSE(test, filtered[1]); +} + +static void filter_subsuite_test_glob_test(struct kunit *test) +{ + struct kunit_suite *subsuite[3] = {NULL, NULL, NULL}; + struct kunit_suite * const *filtered; + struct kunit_test_filter filter = { + .suite_glob = "suite2", + .test_glob = "test2", + }; + + subsuite[0] = alloc_fake_suite(test, "suite1", dummy_test_cases); + subsuite[1] = alloc_fake_suite(test, "suite2", dummy_test_cases); + /* Want: suite1, suite2, NULL -> suite2 (just test1), NULL */ + filtered = kunit_filter_subsuite(subsuite, &filter); + KUNIT_ASSERT_NOT_ERR_OR_NULL(test, filtered); + free_subsuite_at_end(test, filtered); + + /* Validate we just have suite2 */ + KUNIT_ASSERT_NOT_ERR_OR_NULL(test, filtered[0]); + KUNIT_EXPECT_STREQ(test, (const char *)filtered[0]->name, "suite2"); KUNIT_EXPECT_FALSE(test, filtered[1]); + + /* Now validate we just have test2 */ + KUNIT_ASSERT_NOT_ERR_OR_NULL(test, filtered[0]->test_cases); + KUNIT_EXPECT_STREQ(test, (const char *)filtered[0]->test_cases[0].name, "test2"); + KUNIT_EXPECT_FALSE(test, filtered[0]->test_cases[1].name); } static void filter_subsuite_to_empty_test(struct kunit *test) { struct kunit_suite *subsuite[3] = {NULL, NULL, NULL}; struct kunit_suite * const *filtered; + struct kunit_test_filter filter = { + .suite_glob = "not_found", + .test_glob = NULL, + }; - subsuite[0] = alloc_fake_suite(test, "suite1"); - subsuite[1] = alloc_fake_suite(test, "suite2"); + subsuite[0] = alloc_fake_suite(test, "suite1", dummy_test_cases); + subsuite[1] = alloc_fake_suite(test, "suite2", dummy_test_cases); - filtered = kunit_filter_subsuite(subsuite, "not_found"); - kfree_at_end(test, filtered); /* just in case */ + filtered = kunit_filter_subsuite(subsuite, &filter); + free_subsuite_at_end(test, filtered); /* just in case */ KUNIT_EXPECT_FALSE_MSG(test, filtered, "should be NULL to indicate no match"); @@ -52,7 +117,7 @@ static void kfree_subsuites_at_end(struct kunit *test, struct suite_set *suite_s kfree_at_end(test, suite_set->start); for (suites = suite_set->start; suites < suite_set->end; suites++) - kfree_at_end(test, *suites); + free_subsuite_at_end(test, *suites); } static void filter_suites_test(struct kunit *test) @@ -74,8 +139,8 @@ static void filter_suites_test(struct kunit *test) struct suite_set filtered = {.start = NULL, .end = NULL}; /* Emulate two files, each having one suite */ - subsuites[0][0] = alloc_fake_suite(test, "suite0"); - subsuites[1][0] = alloc_fake_suite(test, "suite1"); + subsuites[0][0] = alloc_fake_suite(test, "suite0", dummy_test_cases); + subsuites[1][0] = alloc_fake_suite(test, "suite1", dummy_test_cases); /* Filter out suite1 */ filtered = kunit_filter_suites(&suite_set, "suite0"); @@ -88,7 +153,9 @@ static void filter_suites_test(struct kunit *test) } static struct kunit_case executor_test_cases[] = { + KUNIT_CASE(parse_filter_test), KUNIT_CASE(filter_subsuite_test), + KUNIT_CASE(filter_subsuite_test_glob_test), KUNIT_CASE(filter_subsuite_to_empty_test), KUNIT_CASE(filter_suites_test), {} @@ -120,14 +187,30 @@ static void kfree_at_end(struct kunit *test, const void *to_free) (void *)to_free); } +static void free_subsuite_res_free(struct kunit_resource *res) +{ + kunit_free_subsuite(res->data); +} + +static void free_subsuite_at_end(struct kunit *test, + struct kunit_suite *const *to_free) +{ + if (IS_ERR_OR_NULL(to_free)) + return; + kunit_alloc_resource(test, NULL, free_subsuite_res_free, + GFP_KERNEL, (void *)to_free); +} + static struct kunit_suite *alloc_fake_suite(struct kunit *test, - const char *suite_name) + const char *suite_name, + struct kunit_case *test_cases) { struct kunit_suite *suite; /* We normally never expect to allocate suites, hence the non-const cast. */ suite = kunit_kzalloc(test, sizeof(*suite), GFP_KERNEL); strncpy((char *)suite->name, suite_name, sizeof(suite->name) - 1); + suite->test_cases = test_cases; return suite; } diff --git a/tools/testing/kunit/kunit.py b/tools/testing/kunit/kunit.py index ac35c61f65f5..8fa2ccd4f88f 100755 --- a/tools/testing/kunit/kunit.py +++ b/tools/testing/kunit/kunit.py @@ -263,9 +263,8 @@ def add_exec_opts(parser) -> None: default=300, metavar='timeout') parser.add_argument('filter_glob', - help='maximum number of seconds to allow for all tests ' - 'to run. This does not include time taken to build the ' - 'tests.', + help='Filter which KUnit test suites/tests run at ' + 'boot-time, e.g. list* or list*.*del_test', type=str, nargs='?', default='', -- cgit v1.2.3 From cd94fbc2cafb61cc1394cdba3f92b99ce07b03ae Mon Sep 17 00:00:00 2001 From: Daniel Latypov Date: Fri, 1 Oct 2021 18:36:35 -0700 Subject: kunit: fix too small allocation when using suite-only kunit.filter_glob When a user filters by a suite and not a test, e.g. $ ./tools/testing/kunit/kunit.py run 'suite_name' it hits this code const int len = strlen(filter_glob); ... parsed->suite_glob = kmalloc(len, GFP_KERNEL); which fails to allocate space for the terminating NULL. Somehow, it seems like we can't easily reproduce this under UML, so the existing `parse_filter_test()` didn't catch this. Fix this by allocating `len + 1` and switch to kzalloc() just to be a bit more defensive. We're only going to run this code once per kernel boot, and it should never be very long. Also update the unit tests to be a bit more cautious. This bug showed up as a NULL pointer dereference here: > KUNIT_EXPECT_STREQ(test, (const char *)filtered.start[0][0]->name, "suite0"); `filtered.start[0][0]` was NULL, and `name` is at offset 0 in the struct, so `...->name` was also NULL. Fixes: 3b29021ddd10 ("kunit: tool: allow filtering test cases via glob") Reported-by: kernel test robot Signed-off-by: Daniel Latypov Reviewed-by: David Gow Acked-by: Brendan Higgins Signed-off-by: Shuah Khan --- lib/kunit/executor.c | 2 +- lib/kunit/executor_test.c | 1 + 2 files changed, 2 insertions(+), 1 deletion(-) (limited to 'lib') diff --git a/lib/kunit/executor.c b/lib/kunit/executor.c index bab3ab940acc..1d7fecd33261 100644 --- a/lib/kunit/executor.c +++ b/lib/kunit/executor.c @@ -33,7 +33,7 @@ static void kunit_parse_filter_glob(struct kunit_test_filter *parsed, const char *period = strchr(filter_glob, '.'); if (!period) { - parsed->suite_glob = kmalloc(len, GFP_KERNEL); + parsed->suite_glob = kzalloc(len + 1, GFP_KERNEL); parsed->test_glob = NULL; strcpy(parsed->suite_glob, filter_glob); return; diff --git a/lib/kunit/executor_test.c b/lib/kunit/executor_test.c index edbd8184dcd7..4ed57fd94e42 100644 --- a/lib/kunit/executor_test.c +++ b/lib/kunit/executor_test.c @@ -149,6 +149,7 @@ static void filter_suites_test(struct kunit *test) KUNIT_ASSERT_NOT_ERR_OR_NULL(test, filtered.start); KUNIT_ASSERT_NOT_ERR_OR_NULL(test, filtered.start[0]); + KUNIT_ASSERT_NOT_ERR_OR_NULL(test, filtered.start[0][0]); KUNIT_EXPECT_STREQ(test, (const char *)filtered.start[0][0]->name, "suite0"); } -- cgit v1.2.3 From 9c6b0e1d8993e47a3fe437af71c6a23d4ee73e12 Mon Sep 17 00:00:00 2001 From: Daniel Latypov Date: Thu, 30 Sep 2021 15:20:45 -0700 Subject: kunit: add 'kunit.action' param to allow listing out tests Context: It's difficult to map a given .kunitconfig => set of enabled tests. Letting kunit.py figure that out would be useful. This patch: * is intended to be an implementation detail used only by kunit.py * adds a kunit.action module param with one valid non-null value, "list" * for the "list" action, it simply prints out "." * leaves the kunit.py changes to make use of this for another patch. Note: kunit.filter_glob is respected for this and all future actions. Hack: we print a TAP header (but no test plan) to allow kunit.py to use the same code to pick up KUnit output that it does for normal tests. Since this is intended to be an implementation detail, it seems fine for now. Maybe in the future we output each test as SKIPPED or the like. Go with a more generic "action" param, since it seems like we might eventually have more modes besides just running or listing tests, e.g. * perhaps a benchmark mode that reruns test cases and reports timing * perhaps a deflake mode that reruns test cases that failed * perhaps a mode where we randomize test order to try and catch hermeticity bugs like "test a only passes if run after test b" Tested: $ ./tools/testing/kunit/kunit.py run --kernel_arg=kunit.action=list --raw_output=kunit ... TAP version 14 1..1 example.example_simple_test example.example_skip_test example.example_mark_skipped_test reboot: System halted Signed-off-by: Daniel Latypov Reviewed-by: David Gow Reviewed-by: Brendan Higgins Signed-off-by: Shuah Khan --- lib/kunit/executor.c | 45 ++++++++++++++++++++++++++++++++++++++++----- 1 file changed, 40 insertions(+), 5 deletions(-) (limited to 'lib') diff --git a/lib/kunit/executor.c b/lib/kunit/executor.c index 1d7fecd33261..22640c9ee819 100644 --- a/lib/kunit/executor.c +++ b/lib/kunit/executor.c @@ -15,9 +15,16 @@ extern struct kunit_suite * const * const __kunit_suites_end[]; #if IS_BUILTIN(CONFIG_KUNIT) static char *filter_glob_param; +static char *action_param; + module_param_named(filter_glob, filter_glob_param, charp, 0); MODULE_PARM_DESC(filter_glob, "Filter which KUnit test suites/tests run at boot-time, e.g. list* or list*.*del_test"); +module_param_named(action, action_param, charp, 0); +MODULE_PARM_DESC(action, + "Changes KUnit executor behavior, valid values are:\n" + ": run the tests like normal\n" + "'list' to list test names instead of running them.\n"); /* glob_match() needs NULL terminated strings, so we need a copy of filter_glob_param. */ struct kunit_test_filter { @@ -196,9 +203,35 @@ static void kunit_print_tap_header(struct suite_set *suite_set) pr_info("1..%d\n", num_of_suites); } -int kunit_run_all_tests(void) +static void kunit_exec_run_tests(struct suite_set *suite_set) { struct kunit_suite * const * const *suites; + + kunit_print_tap_header(suite_set); + + for (suites = suite_set->start; suites < suite_set->end; suites++) + __kunit_test_suites_init(*suites); +} + +static void kunit_exec_list_tests(struct suite_set *suite_set) +{ + unsigned int i; + struct kunit_suite * const * const *suites; + struct kunit_case *test_case; + + /* Hack: print a tap header so kunit.py can find the start of KUnit output. */ + pr_info("TAP version 14\n"); + + for (suites = suite_set->start; suites < suite_set->end; suites++) + for (i = 0; (*suites)[i] != NULL; i++) { + kunit_suite_for_each_test_case((*suites)[i], test_case) { + pr_info("%s.%s\n", (*suites)[i]->name, test_case->name); + } + } +} + +int kunit_run_all_tests(void) +{ struct suite_set suite_set = { .start = __kunit_suites_start, .end = __kunit_suites_end, @@ -207,10 +240,12 @@ int kunit_run_all_tests(void) if (filter_glob_param) suite_set = kunit_filter_suites(&suite_set, filter_glob_param); - kunit_print_tap_header(&suite_set); - - for (suites = suite_set.start; suites < suite_set.end; suites++) - __kunit_test_suites_init(*suites); + if (!action_param) + kunit_exec_run_tests(&suite_set); + else if (strcmp(action_param, "list") == 0) + kunit_exec_list_tests(&suite_set); + else + pr_err("kunit executor: unknown action '%s'\n", action_param); if (filter_glob_param) { /* a copy was made of each array */ kunit_free_suite_set(suite_set); -- cgit v1.2.3 From 17ac23eb43f0cbefc8bfce44ad51a9f065895f9f Mon Sep 17 00:00:00 2001 From: David Gow Date: Tue, 5 Oct 2021 21:41:11 -0700 Subject: kunit: Reset suite count after running tests There are some KUnit tests (KFENCE, Thunderbolt) which, for various reasons, do not use the kunit_test_suite() macro and end up running before the KUnit executor runs its tests. This means that their results are printed separately, and they aren't included in the suite count used by the executor. This causes the executor output to be invalid TAP, however, as the suite numbers used are no-longer 1-based, and don't match the test plan. kunit_tool, therefore, prints a large number of warnings. While it'd be nice to fix the tests to run in the executor, in the meantime, reset the suite counter to 1 in __kunit_test_suites_exit. Not only does this fix the executor, it means that if there are multiple calls to __kunit_test_suites_init() across different tests, they'll each get their own numbering. kunit_tool likes this better: even if it's lacking the results for those tests which don't use the executor (due to the lack of TAP header), the output for the other tests is valid. Signed-off-by: David Gow Reviewed-by: Daniel Latypov Signed-off-by: Shuah Khan --- lib/kunit/test.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) (limited to 'lib') diff --git a/lib/kunit/test.c b/lib/kunit/test.c index f246b847024e..3bd741e50a2d 100644 --- a/lib/kunit/test.c +++ b/lib/kunit/test.c @@ -190,10 +190,10 @@ enum kunit_status kunit_suite_has_succeeded(struct kunit_suite *suite) } EXPORT_SYMBOL_GPL(kunit_suite_has_succeeded); +static size_t kunit_suite_counter = 1; + static void kunit_print_subtest_end(struct kunit_suite *suite) { - static size_t kunit_suite_counter = 1; - kunit_print_ok_not_ok((void *)suite, false, kunit_suite_has_succeeded(suite), kunit_suite_counter++, @@ -583,6 +583,8 @@ void __kunit_test_suites_exit(struct kunit_suite **suites) for (i = 0; suites[i] != NULL; i++) kunit_exit_suite(suites[i]); + + kunit_suite_counter = 1; } EXPORT_SYMBOL_GPL(__kunit_test_suites_exit); -- cgit v1.2.3 From cdd591fc86e38ad3899196066219fbbd845f3162 Mon Sep 17 00:00:00 2001 From: Andreas Gruenbacher Date: Mon, 5 Jul 2021 17:26:28 +0200 Subject: iov_iter: Introduce fault_in_iov_iter_writeable Introduce a new fault_in_iov_iter_writeable helper for safely faulting in an iterator for writing. Uses get_user_pages() to fault in the pages without actually writing to them, which would be destructive. We'll use fault_in_iov_iter_writeable in gfs2 once we've determined that the iterator passed to .read_iter isn't in memory. Signed-off-by: Andreas Gruenbacher --- include/linux/pagemap.h | 1 + include/linux/uio.h | 1 + lib/iov_iter.c | 39 ++++++++++++++++++++++++++++++ mm/gup.c | 63 +++++++++++++++++++++++++++++++++++++++++++++++++ 4 files changed, 104 insertions(+) (limited to 'lib') diff --git a/include/linux/pagemap.h b/include/linux/pagemap.h index 9fe94f7a4f7e..2f7dd14083d9 100644 --- a/include/linux/pagemap.h +++ b/include/linux/pagemap.h @@ -736,6 +736,7 @@ extern void add_page_wait_queue(struct page *page, wait_queue_entry_t *waiter); * Fault in userspace address range. */ size_t fault_in_writeable(char __user *uaddr, size_t size); +size_t fault_in_safe_writeable(const char __user *uaddr, size_t size); size_t fault_in_readable(const char __user *uaddr, size_t size); int add_to_page_cache_locked(struct page *page, struct address_space *mapping, diff --git a/include/linux/uio.h b/include/linux/uio.h index d18458af6681..25d1c24fd829 100644 --- a/include/linux/uio.h +++ b/include/linux/uio.h @@ -134,6 +134,7 @@ size_t copy_page_from_iter_atomic(struct page *page, unsigned offset, void iov_iter_advance(struct iov_iter *i, size_t bytes); void iov_iter_revert(struct iov_iter *i, size_t bytes); size_t fault_in_iov_iter_readable(const struct iov_iter *i, size_t bytes); +size_t fault_in_iov_iter_writeable(const struct iov_iter *i, size_t bytes); size_t iov_iter_single_seg_count(const struct iov_iter *i); size_t copy_page_to_iter(struct page *page, size_t offset, size_t bytes, struct iov_iter *i); diff --git a/lib/iov_iter.c b/lib/iov_iter.c index ce3d4f610626..ac9a87e727a3 100644 --- a/lib/iov_iter.c +++ b/lib/iov_iter.c @@ -467,6 +467,45 @@ size_t fault_in_iov_iter_readable(const struct iov_iter *i, size_t size) } EXPORT_SYMBOL(fault_in_iov_iter_readable); +/* + * fault_in_iov_iter_writeable - fault in iov iterator for writing + * @i: iterator + * @size: maximum length + * + * Faults in the iterator using get_user_pages(), i.e., without triggering + * hardware page faults. This is primarily useful when we already know that + * some or all of the pages in @i aren't in memory. + * + * Returns the number of bytes not faulted in, like copy_to_user() and + * copy_from_user(). + * + * Always returns 0 for non-user-space iterators. + */ +size_t fault_in_iov_iter_writeable(const struct iov_iter *i, size_t size) +{ + if (iter_is_iovec(i)) { + size_t count = min(size, iov_iter_count(i)); + const struct iovec *p; + size_t skip; + + size -= count; + for (p = i->iov, skip = i->iov_offset; count; p++, skip = 0) { + size_t len = min(count, p->iov_len - skip); + size_t ret; + + if (unlikely(!len)) + continue; + ret = fault_in_safe_writeable(p->iov_base + skip, len); + count -= len - ret; + if (ret) + break; + } + return count + size; + } + return 0; +} +EXPORT_SYMBOL(fault_in_iov_iter_writeable); + void iov_iter_init(struct iov_iter *i, unsigned int direction, const struct iovec *iov, unsigned long nr_segs, size_t count) diff --git a/mm/gup.c b/mm/gup.c index a7efb027d6cf..795f15c410cc 100644 --- a/mm/gup.c +++ b/mm/gup.c @@ -1691,6 +1691,69 @@ out: } EXPORT_SYMBOL(fault_in_writeable); +/* + * fault_in_safe_writeable - fault in an address range for writing + * @uaddr: start of address range + * @size: length of address range + * + * Faults in an address range using get_user_pages, i.e., without triggering + * hardware page faults. This is primarily useful when we already know that + * some or all of the pages in the address range aren't in memory. + * + * Other than fault_in_writeable(), this function is non-destructive. + * + * Note that we don't pin or otherwise hold the pages referenced that we fault + * in. There's no guarantee that they'll stay in memory for any duration of + * time. + * + * Returns the number of bytes not faulted in, like copy_to_user() and + * copy_from_user(). + */ +size_t fault_in_safe_writeable(const char __user *uaddr, size_t size) +{ + unsigned long start = (unsigned long)untagged_addr(uaddr); + unsigned long end, nstart, nend; + struct mm_struct *mm = current->mm; + struct vm_area_struct *vma = NULL; + int locked = 0; + + nstart = start & PAGE_MASK; + end = PAGE_ALIGN(start + size); + if (end < nstart) + end = 0; + for (; nstart != end; nstart = nend) { + unsigned long nr_pages; + long ret; + + if (!locked) { + locked = 1; + mmap_read_lock(mm); + vma = find_vma(mm, nstart); + } else if (nstart >= vma->vm_end) + vma = vma->vm_next; + if (!vma || vma->vm_start >= end) + break; + nend = end ? min(end, vma->vm_end) : vma->vm_end; + if (vma->vm_flags & (VM_IO | VM_PFNMAP)) + continue; + if (nstart < vma->vm_start) + nstart = vma->vm_start; + nr_pages = (nend - nstart) / PAGE_SIZE; + ret = __get_user_pages_locked(mm, nstart, nr_pages, + NULL, NULL, &locked, + FOLL_TOUCH | FOLL_WRITE); + if (ret <= 0) + break; + nend = nstart + ret * PAGE_SIZE; + } + if (locked) + mmap_read_unlock(mm); + if (nstart == end) + return 0; + return size - min_t(size_t, nstart - start, size); +} +EXPORT_SYMBOL(fault_in_safe_writeable); + /** * fault_in_readable - fault in userspace address range for reading * @uaddr: start of user address range -- cgit v1.2.3 From 09ee10ff804ec4bc58eb9849226312ff1c4d7e7b Mon Sep 17 00:00:00 2001 From: Jim Cromie Date: Tue, 19 Oct 2021 15:07:46 -0600 Subject: dyndbg: refine verbosity 1-4 summary-detail adjust current v*pr_info() calls to fit an overview..detail scheme: 1- module level activity: add/remove, etc 2- command ingest, splitting, summary of effects. per >control write 3- command parsing: op, flags, search terms 4- per-site change msg can yield ~3k x 2 logs per echo "+p;-p" > command. Summarize these 4 levels in MODULE_PARM_DESC, and update verbose=3 in Doc. 2- is new, to isolate a problem where a stress-test script (which feeds ~4kb multi-command strings) would produce short writes, truncating last command and causing parsing errors, which confused test results. The script fix was to use syswrite, to deliver full proper commands. 4- gets per-callsite "changed:" pr-infos, which are very noisy during stress tests, and formerly obscured v1-3 messages, and overwhelmed the static-key workload being tested. The verbose parameter has previously seen adjustment: commit 481c0e33f1e7 ("dyndbg: refine debug verbosity; 1 is basic, 2 more chatty") The script driving these adjustments is: !/usr/bin/perl -w =for Doc 1st purpose was to benchmark the effect of wildcard queries on query performance; if wildcards are risk free cheap enough, we can deploy them in the (floating) format search. 1st finding: wildcards take 2x as long to process. 2nd purpose was to benchmark real static-key changes VS simple flag changes. Found ~100x decrease for the hard work. The script maximizes workload per >control by packing it a ~4kb string of "+p; -p;" commands; this uncovered some broken stuff. The 85th query failed, and appears to be truncated, so is gramatically incorrect. Its either an error here, or in the kernel. Its not happening atm, retest. Plot thickens: fail only happens doing +-p, not +-mf, likely load dependent. Error remains consistent. Looks like a short write, longer on writer than kernel-reader. Try syswrite on handle to control this. That fixed short write. =cut use Getopt::Std; getopts('vN:k:', \my %opts) or die <', $ctrl) or die "cant open $ctrl for writing: $!\n"; sub vx { my $arg = shift; my $cmd = "echo $arg > /sys/module/dynamic_debug/parameters/verbose"; system($cmd); warn("vx problem: rc:$? err:$! qry: $cmd\n") if ($?); } sub qryOK { my $qry = shift; print "syntax test: <\n$qry>\n" if $opts{v}; my $bytes = syswrite $CTL, $qry; printf "short read: $bytes / %d\n", length $qry if $bytes < length $qry; if ($?) { warn "rc:$? err:$! qry: $qry\n"; return 0; } return 1; } sub build_queries { my ($cmd, $flags, $ct) = @_; # build experiment and reference queries my $cycle = " $cmd +$flags # on ; $cmd -$flags # off \n"; my $ref = " +$flags ; -$flags \n"; my $len = length $cycle; my $max = int(4096 / $len); # break/fit to buffer size $ct |= $max; print "qry: ct:$max x << \n$cycle >>\n"; return unless qryOK($ref); return unless qryOK($cycle); my $wild = $cycle x $ct; my $empty = $ref x $ct; printf "len: %d, %d\n", length $wild, length $empty; return { trial => $wild, ref => $empty, probe => $cycle, zero => $ref, count => $ct, max => $max }; } my $query_set = build_queries(' file "*" module "*" func "*" ', "mf"); qryOK($query_set->{zero}); qryOK($query_set->{probe}); qryOK($query_set->{ref}); qryOK($query_set->{trial}); use Benchmark; sub dobatch { my ($cmd, $flags, $reps, $ct) = @_; $reps ||= $opts{N}; my $qs = build_queries($cmd, $flags, $ct); timethese($reps, { wildcards => sub { syswrite $CTL, $qs->{trial}; }, no_search => sub { syswrite $CTL, $qs->{ref}; } } ); } sub bench_static_key_toggle { vx 0; dobatch(' file "*" module "*" func "*" ', "mf"); dobatch(' file "*" module "*" func "*" ', "p"); } sub bench_verbose_levels { for my $i (0..4) { vx $i; dobatch(' file "*" module "*" func "*" ', "mf"); } } bench_static_key_toggle(); __END__ Heres how the test-script runs: :: verbose=3 parsing info [ 48.401646] dyndbg: query 95: "file "*" module "*" func "*" -mf # off " mod:* [ 48.402040] dyndbg: split into words: "file" "*" "module" "*" "func" "*" "-mf" [ 48.402456] dyndbg: op='-' [ 48.402615] dyndbg: flags=0x6 [ 48.402779] dyndbg: *flagsp=0x0 *maskp=0xfffffff9 [ 48.403033] dyndbg: parsed: func="*" file="*" module="*" format="" lineno=0-0 [ 48.403674] dyndbg: applied: func="*" file="*" module="*" format="" lineno=0-0 :: verbose=2 >control summary. ~300k site matches/changes per 4kb command [ 48.404063] dyndbg: processed 96 queries, with 296160 matches, 0 errs :: 2 queries against each other, no-search vs all-wildcard-search qry: ct:48 x << file "*" module "*" func "*" +mf # on ; file "*" module "*" func "*" -mf # off >> len: 4080, 576 Benchmark: timing 10 iterations of no_search, wildcards... no_search: 0 wallclock secs ( 0.00 usr + 0.03 sys = 0.03 CPU) @ 333.33/s (n=10) (warning: too few iterations for a reliable count) wildcards: 0 wallclock secs ( 0.00 usr + 0.09 sys = 0.09 CPU) @ 111.11/s (n=10) (warning: too few iterations for a reliable count) :: 2 queries, both doing real work / changing stati-key states. qry: ct:49 x << file "*" module "*" func "*" +p # on ; file "*" module "*" func "*" -p # off >> len: 4067, 490 Benchmark: timing 10 iterations of no_search, wildcards... no_search: 20 wallclock secs ( 0.00 usr + 20.36 sys = 20.36 CPU) @ 0.49/s (n=10) wildcards: 21 wallclock secs ( 0.00 usr + 21.08 sys = 21.08 CPU) @ 0.47/s (n=10) bash-5.1# Thats 150k static-key-toggles / sec ~600x slower than simple flags on qemu --smp 3 run Signed-off-by: Jim Cromie Link: https://lore.kernel.org/r/20211019210746.185307-1-jim.cromie@gmail.com Signed-off-by: Greg Kroah-Hartman --- Documentation/admin-guide/dynamic-debug-howto.rst | 2 +- lib/dynamic_debug.c | 25 +++++++++++++---------- 2 files changed, 15 insertions(+), 12 deletions(-) (limited to 'lib') diff --git a/Documentation/admin-guide/dynamic-debug-howto.rst b/Documentation/admin-guide/dynamic-debug-howto.rst index ae264aab42b6..a89cfa083155 100644 --- a/Documentation/admin-guide/dynamic-debug-howto.rst +++ b/Documentation/admin-guide/dynamic-debug-howto.rst @@ -356,7 +356,7 @@ Examples // boot-args example, with newlines and comments for readability Kernel command line: ... // see whats going on in dyndbg=value processing - dynamic_debug.verbose=1 + dynamic_debug.verbose=3 // enable pr_debugs in the btrfs module (can be builtin or loadable) btrfs.dyndbg="+p" // enable pr_debugs in all files under init/ diff --git a/lib/dynamic_debug.c b/lib/dynamic_debug.c index 87b1b0121234..dd7f56af9aed 100644 --- a/lib/dynamic_debug.c +++ b/lib/dynamic_debug.c @@ -71,6 +71,8 @@ static DEFINE_MUTEX(ddebug_lock); static LIST_HEAD(ddebug_tables); static int verbose; module_param(verbose, int, 0644); +MODULE_PARM_DESC(verbose, " dynamic_debug/control processing " + "( 0 = off (default), 1 = module add/rm, 2 = >control summary, 3 = parsing, 4 = per-site changes)"); /* Return the path relative to source root */ static inline const char *trim_prefix(const char *path) @@ -118,6 +120,8 @@ do { \ #define vpr_info(fmt, ...) vnpr_info(1, fmt, ##__VA_ARGS__) #define v2pr_info(fmt, ...) vnpr_info(2, fmt, ##__VA_ARGS__) +#define v3pr_info(fmt, ...) vnpr_info(3, fmt, ##__VA_ARGS__) +#define v4pr_info(fmt, ...) vnpr_info(4, fmt, ##__VA_ARGS__) static void vpr_info_dq(const struct ddebug_query *query, const char *msg) { @@ -130,7 +134,7 @@ static void vpr_info_dq(const struct ddebug_query *query, const char *msg) fmtlen--; } - vpr_info("%s: func=\"%s\" file=\"%s\" module=\"%s\" format=\"%.*s\" lineno=%u-%u\n", + v3pr_info("%s: func=\"%s\" file=\"%s\" module=\"%s\" format=\"%.*s\" lineno=%u-%u\n", msg, query->function ?: "", query->filename ?: "", @@ -213,7 +217,7 @@ static int ddebug_change(const struct ddebug_query *query, static_branch_enable(&dp->key.dd_key_true); #endif dp->flags = newflags; - v2pr_info("changed %s:%d [%s]%s =%s\n", + v4pr_info("changed %s:%d [%s]%s =%s\n", trim_prefix(dp->filename), dp->lineno, dt->mod_name, dp->function, ddebug_describe_flags(dp->flags, &fbuf)); @@ -273,7 +277,7 @@ static int ddebug_tokenize(char *buf, char *words[], int maxwords) buf = end; } - if (verbose) { + if (verbose >= 3) { int i; pr_info("split into words:"); for (i = 0; i < nwords; i++) @@ -333,7 +337,7 @@ static int parse_linerange(struct ddebug_query *query, const char *first) } else { query->last_lineno = query->first_lineno; } - vpr_info("parsed line %d-%d\n", query->first_lineno, + v3pr_info("parsed line %d-%d\n", query->first_lineno, query->last_lineno); return 0; } @@ -447,7 +451,7 @@ static int ddebug_parse_flags(const char *str, struct flag_settings *modifiers) pr_err("bad flag-op %c, at start of %s\n", *str, str); return -EINVAL; } - vpr_info("op='%c'\n", op); + v3pr_info("op='%c'\n", op); for (; *str ; ++str) { for (i = ARRAY_SIZE(opt_array) - 1; i >= 0; i--) { @@ -461,7 +465,7 @@ static int ddebug_parse_flags(const char *str, struct flag_settings *modifiers) return -EINVAL; } } - vpr_info("flags=0x%x\n", modifiers->flags); + v3pr_info("flags=0x%x\n", modifiers->flags); /* calculate final flags, mask based upon op */ switch (op) { @@ -477,7 +481,7 @@ static int ddebug_parse_flags(const char *str, struct flag_settings *modifiers) modifiers->flags = 0; break; } - vpr_info("*flagsp=0x%x *maskp=0x%x\n", modifiers->flags, modifiers->mask); + v3pr_info("*flagsp=0x%x *maskp=0x%x\n", modifiers->flags, modifiers->mask); return 0; } @@ -540,9 +544,8 @@ static int ddebug_exec_queries(char *query, const char *modname) } i++; } - if (i) - vpr_info("processed %d queries, with %d matches, %d errs\n", + v2pr_info("processed %d queries, with %d matches, %d errs\n", i, nfound, errs); if (exitcode) @@ -780,7 +783,7 @@ static ssize_t ddebug_proc_write(struct file *file, const char __user *ubuf, tmpbuf = memdup_user_nul(ubuf, len); if (IS_ERR(tmpbuf)) return PTR_ERR(tmpbuf); - vpr_info("read %d bytes from userspace\n", (int)len); + v2pr_info("read %zu bytes from userspace\n", len); ret = ddebug_exec_queries(tmpbuf, NULL); kfree(tmpbuf); @@ -968,7 +971,7 @@ int ddebug_add_module(struct _ddebug *tab, unsigned int n, list_add(&dt->link, &ddebug_tables); mutex_unlock(&ddebug_lock); - v2pr_info("%3u debug prints in module %s\n", n, dt->mod_name); + vpr_info("%3u debug prints in module %s\n", n, dt->mod_name); return 0; } -- cgit v1.2.3 From e44e81c5b90f698025eadceb7eef8661eda117d5 Mon Sep 17 00:00:00 2001 From: Sven Schnelle Date: Thu, 21 Oct 2021 09:54:24 +0900 Subject: kprobes: convert tests to kunit This converts the kprobes testcases to use the kunit framework. It adds a dependency on CONFIG_KUNIT, and the output will change to TAP: TAP version 14 1..1 # Subtest: kprobes_test 1..4 random: crng init done ok 1 - test_kprobe ok 2 - test_kprobes ok 3 - test_kretprobe ok 4 - test_kretprobes ok 1 - kprobes_test Note that the kprobes testcases are no longer run immediately after kprobes initialization, but as a late initcall when kunit is initialized. kprobes itself is initialized with an early initcall, so the order is still correct. Signed-off-by: Sven Schnelle Acked-by: Masami Hiramatsu Signed-off-by: Masami Hiramatsu Signed-off-by: Steven Rostedt (VMware) --- kernel/kprobes.c | 3 - kernel/test_kprobes.c | 222 ++++++++++++++------------------------------------ lib/Kconfig.debug | 3 +- 3 files changed, 61 insertions(+), 167 deletions(-) (limited to 'lib') diff --git a/kernel/kprobes.c b/kernel/kprobes.c index b62af9fc3607..4676627cb066 100644 --- a/kernel/kprobes.c +++ b/kernel/kprobes.c @@ -2581,9 +2581,6 @@ static int __init init_kprobes(void) err = register_module_notifier(&kprobe_module_nb); kprobes_initialized = (err == 0); - - if (!err) - init_test_probes(); return err; } early_initcall(init_kprobes); diff --git a/kernel/test_kprobes.c b/kernel/test_kprobes.c index 76c997fdbc9d..e78f18144145 100644 --- a/kernel/test_kprobes.c +++ b/kernel/test_kprobes.c @@ -5,18 +5,17 @@ * Copyright IBM Corp. 2008 */ -#define pr_fmt(fmt) "Kprobe smoke test: " fmt - #include #include #include +#include #define div_factor 3 static u32 rand1, preh_val, posth_val; -static int errors, handler_errors, num_tests; static u32 (*target)(u32 value); static u32 (*target2)(u32 value); +static struct kunit *current_test; static noinline u32 kprobe_target(u32 value) { @@ -25,10 +24,7 @@ static noinline u32 kprobe_target(u32 value) static int kp_pre_handler(struct kprobe *p, struct pt_regs *regs) { - if (preemptible()) { - handler_errors++; - pr_err("pre-handler is preemptible\n"); - } + KUNIT_EXPECT_FALSE(current_test, preemptible()); preh_val = (rand1 / div_factor); return 0; } @@ -36,14 +32,8 @@ static int kp_pre_handler(struct kprobe *p, struct pt_regs *regs) static void kp_post_handler(struct kprobe *p, struct pt_regs *regs, unsigned long flags) { - if (preemptible()) { - handler_errors++; - pr_err("post-handler is preemptible\n"); - } - if (preh_val != (rand1 / div_factor)) { - handler_errors++; - pr_err("incorrect value in post_handler\n"); - } + KUNIT_EXPECT_FALSE(current_test, preemptible()); + KUNIT_EXPECT_EQ(current_test, preh_val, (rand1 / div_factor)); posth_val = preh_val + div_factor; } @@ -53,30 +43,14 @@ static struct kprobe kp = { .post_handler = kp_post_handler }; -static int test_kprobe(void) +static void test_kprobe(struct kunit *test) { - int ret; - - ret = register_kprobe(&kp); - if (ret < 0) { - pr_err("register_kprobe returned %d\n", ret); - return ret; - } - - ret = target(rand1); + current_test = test; + KUNIT_EXPECT_EQ(test, 0, register_kprobe(&kp)); + target(rand1); unregister_kprobe(&kp); - - if (preh_val == 0) { - pr_err("kprobe pre_handler not called\n"); - handler_errors++; - } - - if (posth_val == 0) { - pr_err("kprobe post_handler not called\n"); - handler_errors++; - } - - return 0; + KUNIT_EXPECT_NE(test, 0, preh_val); + KUNIT_EXPECT_NE(test, 0, posth_val); } static noinline u32 kprobe_target2(u32 value) @@ -93,10 +67,7 @@ static int kp_pre_handler2(struct kprobe *p, struct pt_regs *regs) static void kp_post_handler2(struct kprobe *p, struct pt_regs *regs, unsigned long flags) { - if (preh_val != (rand1 / div_factor) + 1) { - handler_errors++; - pr_err("incorrect value in post_handler2\n"); - } + KUNIT_EXPECT_EQ(current_test, preh_val, (rand1 / div_factor) + 1); posth_val = preh_val + div_factor; } @@ -106,51 +77,31 @@ static struct kprobe kp2 = { .post_handler = kp_post_handler2 }; -static int test_kprobes(void) +static void test_kprobes(struct kunit *test) { - int ret; struct kprobe *kps[2] = {&kp, &kp2}; + current_test = test; + /* addr and flags should be cleard for reusing kprobe. */ kp.addr = NULL; kp.flags = 0; - ret = register_kprobes(kps, 2); - if (ret < 0) { - pr_err("register_kprobes returned %d\n", ret); - return ret; - } + KUNIT_EXPECT_EQ(test, 0, register_kprobes(kps, 2)); preh_val = 0; posth_val = 0; - ret = target(rand1); + target(rand1); - if (preh_val == 0) { - pr_err("kprobe pre_handler not called\n"); - handler_errors++; - } - - if (posth_val == 0) { - pr_err("kprobe post_handler not called\n"); - handler_errors++; - } + KUNIT_EXPECT_NE(test, 0, preh_val); + KUNIT_EXPECT_NE(test, 0, posth_val); preh_val = 0; posth_val = 0; - ret = target2(rand1); - - if (preh_val == 0) { - pr_err("kprobe pre_handler2 not called\n"); - handler_errors++; - } - - if (posth_val == 0) { - pr_err("kprobe post_handler2 not called\n"); - handler_errors++; - } + target2(rand1); + KUNIT_EXPECT_NE(test, 0, preh_val); + KUNIT_EXPECT_NE(test, 0, posth_val); unregister_kprobes(kps, 2); - return 0; - } #ifdef CONFIG_KRETPROBES @@ -158,10 +109,7 @@ static u32 krph_val; static int entry_handler(struct kretprobe_instance *ri, struct pt_regs *regs) { - if (preemptible()) { - handler_errors++; - pr_err("kretprobe entry handler is preemptible\n"); - } + KUNIT_EXPECT_FALSE(current_test, preemptible()); krph_val = (rand1 / div_factor); return 0; } @@ -170,19 +118,9 @@ static int return_handler(struct kretprobe_instance *ri, struct pt_regs *regs) { unsigned long ret = regs_return_value(regs); - if (preemptible()) { - handler_errors++; - pr_err("kretprobe return handler is preemptible\n"); - } - if (ret != (rand1 / div_factor)) { - handler_errors++; - pr_err("incorrect value in kretprobe handler\n"); - } - if (krph_val == 0) { - handler_errors++; - pr_err("call to kretprobe entry handler failed\n"); - } - + KUNIT_EXPECT_FALSE(current_test, preemptible()); + KUNIT_EXPECT_EQ(current_test, ret, rand1 / div_factor); + KUNIT_EXPECT_NE(current_test, krph_val, 0); krph_val = rand1; return 0; } @@ -193,39 +131,21 @@ static struct kretprobe rp = { .kp.symbol_name = "kprobe_target" }; -static int test_kretprobe(void) +static void test_kretprobe(struct kunit *test) { - int ret; - - ret = register_kretprobe(&rp); - if (ret < 0) { - pr_err("register_kretprobe returned %d\n", ret); - return ret; - } - - ret = target(rand1); + current_test = test; + KUNIT_EXPECT_EQ(test, 0, register_kretprobe(&rp)); + target(rand1); unregister_kretprobe(&rp); - if (krph_val != rand1) { - pr_err("kretprobe handler not called\n"); - handler_errors++; - } - - return 0; + KUNIT_EXPECT_EQ(test, krph_val, rand1); } static int return_handler2(struct kretprobe_instance *ri, struct pt_regs *regs) { unsigned long ret = regs_return_value(regs); - if (ret != (rand1 / div_factor) + 1) { - handler_errors++; - pr_err("incorrect value in kretprobe handler2\n"); - } - if (krph_val == 0) { - handler_errors++; - pr_err("call to kretprobe entry handler failed\n"); - } - + KUNIT_EXPECT_EQ(current_test, ret, (rand1 / div_factor) + 1); + KUNIT_EXPECT_NE(current_test, krph_val, 0); krph_val = rand1; return 0; } @@ -236,78 +156,54 @@ static struct kretprobe rp2 = { .kp.symbol_name = "kprobe_target2" }; -static int test_kretprobes(void) +static void test_kretprobes(struct kunit *test) { - int ret; struct kretprobe *rps[2] = {&rp, &rp2}; + current_test = test; /* addr and flags should be cleard for reusing kprobe. */ rp.kp.addr = NULL; rp.kp.flags = 0; - ret = register_kretprobes(rps, 2); - if (ret < 0) { - pr_err("register_kretprobe returned %d\n", ret); - return ret; - } + KUNIT_EXPECT_EQ(test, 0, register_kretprobes(rps, 2)); krph_val = 0; - ret = target(rand1); - if (krph_val != rand1) { - pr_err("kretprobe handler not called\n"); - handler_errors++; - } + target(rand1); + KUNIT_EXPECT_EQ(test, krph_val, rand1); krph_val = 0; - ret = target2(rand1); - if (krph_val != rand1) { - pr_err("kretprobe handler2 not called\n"); - handler_errors++; - } + target2(rand1); + KUNIT_EXPECT_EQ(test, krph_val, rand1); unregister_kretprobes(rps, 2); - return 0; } #endif /* CONFIG_KRETPROBES */ -int init_test_probes(void) +static int kprobes_test_init(struct kunit *test) { - int ret; - target = kprobe_target; target2 = kprobe_target2; do { rand1 = prandom_u32(); } while (rand1 <= div_factor); + return 0; +} - pr_info("started\n"); - num_tests++; - ret = test_kprobe(); - if (ret < 0) - errors++; - - num_tests++; - ret = test_kprobes(); - if (ret < 0) - errors++; - +static struct kunit_case kprobes_testcases[] = { + KUNIT_CASE(test_kprobe), + KUNIT_CASE(test_kprobes), #ifdef CONFIG_KRETPROBES - num_tests++; - ret = test_kretprobe(); - if (ret < 0) - errors++; - - num_tests++; - ret = test_kretprobes(); - if (ret < 0) - errors++; -#endif /* CONFIG_KRETPROBES */ + KUNIT_CASE(test_kretprobe), + KUNIT_CASE(test_kretprobes), +#endif + {} +}; - if (errors) - pr_err("BUG: %d out of %d tests failed\n", errors, num_tests); - else if (handler_errors) - pr_err("BUG: %d error(s) running handlers\n", handler_errors); - else - pr_info("passed successfully\n"); +static struct kunit_suite kprobes_test_suite = { + .name = "kprobes_test", + .init = kprobes_test_init, + .test_cases = kprobes_testcases, +}; - return 0; -} +kunit_test_suites(&kprobes_test_suite); + +MODULE_LICENSE("GPL"); diff --git a/lib/Kconfig.debug b/lib/Kconfig.debug index 2a9b6dcdac4f..6ceb11a43e4c 100644 --- a/lib/Kconfig.debug +++ b/lib/Kconfig.debug @@ -2080,9 +2080,10 @@ config TEST_DIV64 If unsure, say N. config KPROBES_SANITY_TEST - bool "Kprobes sanity tests" + tristate "Kprobes sanity tests" depends on DEBUG_KERNEL depends on KPROBES + depends on KUNIT help This option provides for testing basic kprobes functionality on boot. Samples of kprobe and kretprobe are inserted and -- cgit v1.2.3 From 3337ab08d08b1a375f88471d9c8b1cac968cb054 Mon Sep 17 00:00:00 2001 From: Andreas Gruenbacher Date: Mon, 12 Jul 2021 12:06:14 +0200 Subject: iov_iter: Introduce nofault flag to disable page faults Introduce a new nofault flag to indicate to iov_iter_get_pages not to fault in user pages. This is implemented by passing the FOLL_NOFAULT flag to get_user_pages, which causes get_user_pages to fail when it would otherwise fault in a page. We'll use the ->nofault flag to prevent iomap_dio_rw from faulting in pages when page faults are not allowed. Signed-off-by: Andreas Gruenbacher --- include/linux/uio.h | 1 + lib/iov_iter.c | 20 +++++++++++++++----- 2 files changed, 16 insertions(+), 5 deletions(-) (limited to 'lib') diff --git a/include/linux/uio.h b/include/linux/uio.h index 25d1c24fd829..6350354f97e9 100644 --- a/include/linux/uio.h +++ b/include/linux/uio.h @@ -35,6 +35,7 @@ struct iov_iter_state { struct iov_iter { u8 iter_type; + bool nofault; bool data_source; size_t iov_offset; size_t count; diff --git a/lib/iov_iter.c b/lib/iov_iter.c index ac9a87e727a3..66a740e6e153 100644 --- a/lib/iov_iter.c +++ b/lib/iov_iter.c @@ -513,6 +513,7 @@ void iov_iter_init(struct iov_iter *i, unsigned int direction, WARN_ON(direction & ~(READ | WRITE)); *i = (struct iov_iter) { .iter_type = ITER_IOVEC, + .nofault = false, .data_source = direction, .iov = iov, .nr_segs = nr_segs, @@ -1527,13 +1528,17 @@ ssize_t iov_iter_get_pages(struct iov_iter *i, return 0; if (likely(iter_is_iovec(i))) { + unsigned int gup_flags = 0; unsigned long addr; + if (iov_iter_rw(i) != WRITE) + gup_flags |= FOLL_WRITE; + if (i->nofault) + gup_flags |= FOLL_NOFAULT; + addr = first_iovec_segment(i, &len, start, maxsize, maxpages); n = DIV_ROUND_UP(len, PAGE_SIZE); - res = get_user_pages_fast(addr, n, - iov_iter_rw(i) != WRITE ? FOLL_WRITE : 0, - pages); + res = get_user_pages_fast(addr, n, gup_flags, pages); if (unlikely(res <= 0)) return res; return (res == n ? len : res * PAGE_SIZE) - *start; @@ -1649,15 +1654,20 @@ ssize_t iov_iter_get_pages_alloc(struct iov_iter *i, return 0; if (likely(iter_is_iovec(i))) { + unsigned int gup_flags = 0; unsigned long addr; + if (iov_iter_rw(i) != WRITE) + gup_flags |= FOLL_WRITE; + if (i->nofault) + gup_flags |= FOLL_NOFAULT; + addr = first_iovec_segment(i, &len, start, maxsize, ~0U); n = DIV_ROUND_UP(len, PAGE_SIZE); p = get_pages_array(n); if (!p) return -ENOMEM; - res = get_user_pages_fast(addr, n, - iov_iter_rw(i) != WRITE ? FOLL_WRITE : 0, p); + res = get_user_pages_fast(addr, n, gup_flags, p); if (unlikely(res <= 0)) { kvfree(p); *pages = NULL; -- cgit v1.2.3 From 9f8b93a7df4d8e1e8715fb2a45a893cffad9da0b Mon Sep 17 00:00:00 2001 From: Jens Axboe Date: Mon, 25 Oct 2021 10:45:01 -0600 Subject: sbitmap: silence data race warning KCSAN complaints about the sbitmap hint update: ================================================================== BUG: KCSAN: data-race in sbitmap_queue_clear / sbitmap_queue_clear write to 0xffffe8ffffd145b8 of 4 bytes by interrupt on cpu 1: sbitmap_queue_clear+0xca/0xf0 lib/sbitmap.c:606 blk_mq_put_tag+0x82/0x90 __blk_mq_free_request+0x114/0x180 block/blk-mq.c:507 blk_mq_free_request+0x2c8/0x340 block/blk-mq.c:541 __blk_mq_end_request+0x214/0x230 block/blk-mq.c:565 blk_mq_end_request+0x37/0x50 block/blk-mq.c:574 lo_complete_rq+0xca/0x170 drivers/block/loop.c:541 blk_complete_reqs block/blk-mq.c:584 [inline] blk_done_softirq+0x69/0x90 block/blk-mq.c:589 __do_softirq+0x12c/0x26e kernel/softirq.c:558 run_ksoftirqd+0x13/0x20 kernel/softirq.c:920 smpboot_thread_fn+0x22f/0x330 kernel/smpboot.c:164 kthread+0x262/0x280 kernel/kthread.c:319 ret_from_fork+0x1f/0x30 write to 0xffffe8ffffd145b8 of 4 bytes by interrupt on cpu 0: sbitmap_queue_clear+0xca/0xf0 lib/sbitmap.c:606 blk_mq_put_tag+0x82/0x90 __blk_mq_free_request+0x114/0x180 block/blk-mq.c:507 blk_mq_free_request+0x2c8/0x340 block/blk-mq.c:541 __blk_mq_end_request+0x214/0x230 block/blk-mq.c:565 blk_mq_end_request+0x37/0x50 block/blk-mq.c:574 lo_complete_rq+0xca/0x170 drivers/block/loop.c:541 blk_complete_reqs block/blk-mq.c:584 [inline] blk_done_softirq+0x69/0x90 block/blk-mq.c:589 __do_softirq+0x12c/0x26e kernel/softirq.c:558 run_ksoftirqd+0x13/0x20 kernel/softirq.c:920 smpboot_thread_fn+0x22f/0x330 kernel/smpboot.c:164 kthread+0x262/0x280 kernel/kthread.c:319 ret_from_fork+0x1f/0x30 value changed: 0x00000035 -> 0x00000044 Reported by Kernel Concurrency Sanitizer on: CPU: 0 PID: 10 Comm: ksoftirqd/0 Not tainted 5.15.0-rc6-syzkaller #0 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011 ================================================================== which is a data race, but not an important one. This is just updating the percpu alloc hint, and the reader of that hint doesn't ever require it to be valid. Just annotate it with data_race() to silence this one. Reported-by: syzbot+4f8bfd804b4a1f95b8f6@syzkaller.appspotmail.com Acked-by: Marco Elver Signed-off-by: Jens Axboe --- lib/sbitmap.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'lib') diff --git a/lib/sbitmap.c b/lib/sbitmap.c index c6e2f1f2c4d2..2709ab825499 100644 --- a/lib/sbitmap.c +++ b/lib/sbitmap.c @@ -631,7 +631,7 @@ EXPORT_SYMBOL_GPL(sbitmap_queue_wake_up); static inline void sbitmap_update_cpu_hint(struct sbitmap *sb, int cpu, int tag) { if (likely(!sb->round_robin && tag < sb->depth)) - *per_cpu_ptr(sb->alloc_hint, cpu) = tag; + data_race(*per_cpu_ptr(sb->alloc_hint, cpu) = tag); } void sbitmap_queue_clear_batch(struct sbitmap_queue *sbq, int offset, -- cgit v1.2.3 From 4d1c92a4f5ad8454259cfc711c210da6d4cfe8cc Mon Sep 17 00:00:00 2001 From: Masami Hiramatsu Date: Mon, 25 Oct 2021 17:32:37 +0900 Subject: lib/bootconfig: Make xbc_alloc_mem() and xbc_free_mem() as __init function Since the xbc_alloc_mem() and xbc_free_mem() are used from the __init functions and memblock_alloc() is __init function, make them __init functions too. Link: https://lkml.kernel.org/r/163515075747.547467.5746167540626712819.stgit@devnote2 Signed-off-by: Masami Hiramatsu Fixes: 4ee1b4cac236 ("bootconfig: Cleanup dummy headers in tools/bootconfig") Reported-by: Stephen Rothwell Signed-off-by: Steven Rostedt (VMware) --- lib/bootconfig.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'lib') diff --git a/lib/bootconfig.c b/lib/bootconfig.c index a056ae137750..3276675b25e1 100644 --- a/lib/bootconfig.c +++ b/lib/bootconfig.c @@ -43,12 +43,12 @@ static int open_brace[XBC_DEPTH_MAX] __initdata; static int brace_index __initdata; #ifdef __KERNEL__ -static inline void *xbc_alloc_mem(size_t size) +static inline void * __init xbc_alloc_mem(size_t size) { return memblock_alloc(size, SMP_CACHE_BYTES); } -static inline void xbc_free_mem(void *addr, size_t size) +static inline void __init xbc_free_mem(void *addr, size_t size) { memblock_free_ptr(addr, size); } -- cgit v1.2.3 From 010db091b6879786b5d935555b9e19c41e504f71 Mon Sep 17 00:00:00 2001 From: Masami Hiramatsu Date: Tue, 26 Oct 2021 21:21:07 +0900 Subject: lib/bootconfig: Fix the xbc_get_info kerneldoc Fix the kernel doc of xbc_get_info() to add '@' to the parameters. Link: https://lkml.kernel.org/r/163525086738.676803.15352231787913236933.stgit@devnote2 Fixes: e306220cb7b7 ("bootconfig: Add xbc_get_info() for the node information") Reported-by: Stephen Rothwell Signed-off-by: Masami Hiramatsu Signed-off-by: Steven Rostedt (VMware) --- lib/bootconfig.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'lib') diff --git a/lib/bootconfig.c b/lib/bootconfig.c index 3276675b25e1..a10ab25f6fcc 100644 --- a/lib/bootconfig.c +++ b/lib/bootconfig.c @@ -67,8 +67,8 @@ static inline void xbc_free_mem(void *addr, size_t size) #endif /** * xbc_get_info() - Get the information of loaded boot config - * node_size: A pointer to store the number of nodes. - * data_size: A pointer to store the size of bootconfig data. + * @node_size: A pointer to store the number of nodes. + * @data_size: A pointer to store the size of bootconfig data. * * Get the number of used nodes in @node_size if it is not NULL, * and the size of bootconfig data in @data_size if it is not NULL. -- cgit v1.2.3 From b9e94a7bb6fad880ba1ec0d58897480c62f587cf Mon Sep 17 00:00:00 2001 From: Tiezhu Yang Date: Tue, 26 Oct 2021 09:51:30 +0800 Subject: test_kprobes: Move it from kernel/ to lib/ Since config KPROBES_SANITY_TEST is in lib/Kconfig.debug, it is better to let test_kprobes.c in lib/, just like other similar tests found in lib/. Link: https://lkml.kernel.org/r/1635213091-24387-4-git-send-email-yangtiezhu@loongson.cn Signed-off-by: Tiezhu Yang Acked-by: Masami Hiramatsu Signed-off-by: Steven Rostedt (VMware) --- kernel/Makefile | 1 - kernel/test_kprobes.c | 371 -------------------------------------------------- lib/Makefile | 1 + lib/test_kprobes.c | 371 ++++++++++++++++++++++++++++++++++++++++++++++++++ 4 files changed, 372 insertions(+), 372 deletions(-) delete mode 100644 kernel/test_kprobes.c create mode 100644 lib/test_kprobes.c (limited to 'lib') diff --git a/kernel/Makefile b/kernel/Makefile index 4df609be42d0..9e4d33dce8a5 100644 --- a/kernel/Makefile +++ b/kernel/Makefile @@ -85,7 +85,6 @@ obj-$(CONFIG_PID_NS) += pid_namespace.o obj-$(CONFIG_IKCONFIG) += configs.o obj-$(CONFIG_IKHEADERS) += kheaders.o obj-$(CONFIG_SMP) += stop_machine.o -obj-$(CONFIG_KPROBES_SANITY_TEST) += test_kprobes.o obj-$(CONFIG_AUDIT) += audit.o auditfilter.o obj-$(CONFIG_AUDITSYSCALL) += auditsc.o audit_watch.o audit_fsnotify.o audit_tree.o obj-$(CONFIG_GCOV_KERNEL) += gcov/ diff --git a/kernel/test_kprobes.c b/kernel/test_kprobes.c deleted file mode 100644 index a5edc2ebc947..000000000000 --- a/kernel/test_kprobes.c +++ /dev/null @@ -1,371 +0,0 @@ -// SPDX-License-Identifier: GPL-2.0-or-later -/* - * test_kprobes.c - simple sanity test for *probes - * - * Copyright IBM Corp. 2008 - */ - -#include -#include -#include -#include - -#define div_factor 3 - -static u32 rand1, preh_val, posth_val; -static u32 (*target)(u32 value); -static u32 (*target2)(u32 value); -static struct kunit *current_test; - -static unsigned long (*internal_target)(void); -static unsigned long (*stacktrace_target)(void); -static unsigned long (*stacktrace_driver)(void); -static unsigned long target_return_address[2]; - -static noinline u32 kprobe_target(u32 value) -{ - return (value / div_factor); -} - -static int kp_pre_handler(struct kprobe *p, struct pt_regs *regs) -{ - KUNIT_EXPECT_FALSE(current_test, preemptible()); - preh_val = (rand1 / div_factor); - return 0; -} - -static void kp_post_handler(struct kprobe *p, struct pt_regs *regs, - unsigned long flags) -{ - KUNIT_EXPECT_FALSE(current_test, preemptible()); - KUNIT_EXPECT_EQ(current_test, preh_val, (rand1 / div_factor)); - posth_val = preh_val + div_factor; -} - -static struct kprobe kp = { - .symbol_name = "kprobe_target", - .pre_handler = kp_pre_handler, - .post_handler = kp_post_handler -}; - -static void test_kprobe(struct kunit *test) -{ - current_test = test; - KUNIT_EXPECT_EQ(test, 0, register_kprobe(&kp)); - target(rand1); - unregister_kprobe(&kp); - KUNIT_EXPECT_NE(test, 0, preh_val); - KUNIT_EXPECT_NE(test, 0, posth_val); -} - -static noinline u32 kprobe_target2(u32 value) -{ - return (value / div_factor) + 1; -} - -static noinline unsigned long kprobe_stacktrace_internal_target(void) -{ - if (!target_return_address[0]) - target_return_address[0] = (unsigned long)__builtin_return_address(0); - return target_return_address[0]; -} - -static noinline unsigned long kprobe_stacktrace_target(void) -{ - if (!target_return_address[1]) - target_return_address[1] = (unsigned long)__builtin_return_address(0); - - if (internal_target) - internal_target(); - - return target_return_address[1]; -} - -static noinline unsigned long kprobe_stacktrace_driver(void) -{ - if (stacktrace_target) - stacktrace_target(); - - /* This is for preventing inlining the function */ - return (unsigned long)__builtin_return_address(0); -} - -static int kp_pre_handler2(struct kprobe *p, struct pt_regs *regs) -{ - preh_val = (rand1 / div_factor) + 1; - return 0; -} - -static void kp_post_handler2(struct kprobe *p, struct pt_regs *regs, - unsigned long flags) -{ - KUNIT_EXPECT_EQ(current_test, preh_val, (rand1 / div_factor) + 1); - posth_val = preh_val + div_factor; -} - -static struct kprobe kp2 = { - .symbol_name = "kprobe_target2", - .pre_handler = kp_pre_handler2, - .post_handler = kp_post_handler2 -}; - -static void test_kprobes(struct kunit *test) -{ - struct kprobe *kps[2] = {&kp, &kp2}; - - current_test = test; - - /* addr and flags should be cleard for reusing kprobe. */ - kp.addr = NULL; - kp.flags = 0; - - KUNIT_EXPECT_EQ(test, 0, register_kprobes(kps, 2)); - preh_val = 0; - posth_val = 0; - target(rand1); - - KUNIT_EXPECT_NE(test, 0, preh_val); - KUNIT_EXPECT_NE(test, 0, posth_val); - - preh_val = 0; - posth_val = 0; - target2(rand1); - - KUNIT_EXPECT_NE(test, 0, preh_val); - KUNIT_EXPECT_NE(test, 0, posth_val); - unregister_kprobes(kps, 2); -} - -#ifdef CONFIG_KRETPROBES -static u32 krph_val; - -static int entry_handler(struct kretprobe_instance *ri, struct pt_regs *regs) -{ - KUNIT_EXPECT_FALSE(current_test, preemptible()); - krph_val = (rand1 / div_factor); - return 0; -} - -static int return_handler(struct kretprobe_instance *ri, struct pt_regs *regs) -{ - unsigned long ret = regs_return_value(regs); - - KUNIT_EXPECT_FALSE(current_test, preemptible()); - KUNIT_EXPECT_EQ(current_test, ret, rand1 / div_factor); - KUNIT_EXPECT_NE(current_test, krph_val, 0); - krph_val = rand1; - return 0; -} - -static struct kretprobe rp = { - .handler = return_handler, - .entry_handler = entry_handler, - .kp.symbol_name = "kprobe_target" -}; - -static void test_kretprobe(struct kunit *test) -{ - current_test = test; - KUNIT_EXPECT_EQ(test, 0, register_kretprobe(&rp)); - target(rand1); - unregister_kretprobe(&rp); - KUNIT_EXPECT_EQ(test, krph_val, rand1); -} - -static int return_handler2(struct kretprobe_instance *ri, struct pt_regs *regs) -{ - unsigned long ret = regs_return_value(regs); - - KUNIT_EXPECT_EQ(current_test, ret, (rand1 / div_factor) + 1); - KUNIT_EXPECT_NE(current_test, krph_val, 0); - krph_val = rand1; - return 0; -} - -static struct kretprobe rp2 = { - .handler = return_handler2, - .entry_handler = entry_handler, - .kp.symbol_name = "kprobe_target2" -}; - -static void test_kretprobes(struct kunit *test) -{ - struct kretprobe *rps[2] = {&rp, &rp2}; - - current_test = test; - /* addr and flags should be cleard for reusing kprobe. */ - rp.kp.addr = NULL; - rp.kp.flags = 0; - KUNIT_EXPECT_EQ(test, 0, register_kretprobes(rps, 2)); - - krph_val = 0; - target(rand1); - KUNIT_EXPECT_EQ(test, krph_val, rand1); - - krph_val = 0; - target2(rand1); - KUNIT_EXPECT_EQ(test, krph_val, rand1); - unregister_kretprobes(rps, 2); -} - -#ifdef CONFIG_ARCH_CORRECT_STACKTRACE_ON_KRETPROBE -#define STACK_BUF_SIZE 16 -static unsigned long stack_buf[STACK_BUF_SIZE]; - -static int stacktrace_return_handler(struct kretprobe_instance *ri, struct pt_regs *regs) -{ - unsigned long retval = regs_return_value(regs); - int i, ret; - - KUNIT_EXPECT_FALSE(current_test, preemptible()); - KUNIT_EXPECT_EQ(current_test, retval, target_return_address[1]); - - /* - * Test stacktrace inside the kretprobe handler, this will involves - * kretprobe trampoline, but must include correct return address - * of the target function. - */ - ret = stack_trace_save(stack_buf, STACK_BUF_SIZE, 0); - KUNIT_EXPECT_NE(current_test, ret, 0); - - for (i = 0; i < ret; i++) { - if (stack_buf[i] == target_return_address[1]) - break; - } - KUNIT_EXPECT_NE(current_test, i, ret); - -#if !IS_MODULE(CONFIG_KPROBES_SANITY_TEST) - /* - * Test stacktrace from pt_regs at the return address. Thus the stack - * trace must start from the target return address. - */ - ret = stack_trace_save_regs(regs, stack_buf, STACK_BUF_SIZE, 0); - KUNIT_EXPECT_NE(current_test, ret, 0); - KUNIT_EXPECT_EQ(current_test, stack_buf[0], target_return_address[1]); -#endif - - return 0; -} - -static struct kretprobe rp3 = { - .handler = stacktrace_return_handler, - .kp.symbol_name = "kprobe_stacktrace_target" -}; - -static void test_stacktrace_on_kretprobe(struct kunit *test) -{ - unsigned long myretaddr = (unsigned long)__builtin_return_address(0); - - current_test = test; - rp3.kp.addr = NULL; - rp3.kp.flags = 0; - - /* - * Run the stacktrace_driver() to record correct return address in - * stacktrace_target() and ensure stacktrace_driver() call is not - * inlined by checking the return address of stacktrace_driver() - * and the return address of this function is different. - */ - KUNIT_ASSERT_NE(test, myretaddr, stacktrace_driver()); - - KUNIT_ASSERT_EQ(test, 0, register_kretprobe(&rp3)); - KUNIT_ASSERT_NE(test, myretaddr, stacktrace_driver()); - unregister_kretprobe(&rp3); -} - -static int stacktrace_internal_return_handler(struct kretprobe_instance *ri, struct pt_regs *regs) -{ - unsigned long retval = regs_return_value(regs); - int i, ret; - - KUNIT_EXPECT_FALSE(current_test, preemptible()); - KUNIT_EXPECT_EQ(current_test, retval, target_return_address[0]); - - /* - * Test stacktrace inside the kretprobe handler for nested case. - * The unwinder will find the kretprobe_trampoline address on the - * return address, and kretprobe must solve that. - */ - ret = stack_trace_save(stack_buf, STACK_BUF_SIZE, 0); - KUNIT_EXPECT_NE(current_test, ret, 0); - - for (i = 0; i < ret - 1; i++) { - if (stack_buf[i] == target_return_address[0]) { - KUNIT_EXPECT_EQ(current_test, stack_buf[i + 1], target_return_address[1]); - break; - } - } - KUNIT_EXPECT_NE(current_test, i, ret); - -#if !IS_MODULE(CONFIG_KPROBES_SANITY_TEST) - /* Ditto for the regs version. */ - ret = stack_trace_save_regs(regs, stack_buf, STACK_BUF_SIZE, 0); - KUNIT_EXPECT_NE(current_test, ret, 0); - KUNIT_EXPECT_EQ(current_test, stack_buf[0], target_return_address[0]); - KUNIT_EXPECT_EQ(current_test, stack_buf[1], target_return_address[1]); -#endif - - return 0; -} - -static struct kretprobe rp4 = { - .handler = stacktrace_internal_return_handler, - .kp.symbol_name = "kprobe_stacktrace_internal_target" -}; - -static void test_stacktrace_on_nested_kretprobe(struct kunit *test) -{ - unsigned long myretaddr = (unsigned long)__builtin_return_address(0); - struct kretprobe *rps[2] = {&rp3, &rp4}; - - current_test = test; - rp3.kp.addr = NULL; - rp3.kp.flags = 0; - - //KUNIT_ASSERT_NE(test, myretaddr, stacktrace_driver()); - - KUNIT_ASSERT_EQ(test, 0, register_kretprobes(rps, 2)); - KUNIT_ASSERT_NE(test, myretaddr, stacktrace_driver()); - unregister_kretprobes(rps, 2); -} -#endif /* CONFIG_ARCH_CORRECT_STACKTRACE_ON_KRETPROBE */ - -#endif /* CONFIG_KRETPROBES */ - -static int kprobes_test_init(struct kunit *test) -{ - target = kprobe_target; - target2 = kprobe_target2; - stacktrace_target = kprobe_stacktrace_target; - internal_target = kprobe_stacktrace_internal_target; - stacktrace_driver = kprobe_stacktrace_driver; - - do { - rand1 = prandom_u32(); - } while (rand1 <= div_factor); - return 0; -} - -static struct kunit_case kprobes_testcases[] = { - KUNIT_CASE(test_kprobe), - KUNIT_CASE(test_kprobes), -#ifdef CONFIG_KRETPROBES - KUNIT_CASE(test_kretprobe), - KUNIT_CASE(test_kretprobes), -#ifdef CONFIG_ARCH_CORRECT_STACKTRACE_ON_KRETPROBE - KUNIT_CASE(test_stacktrace_on_kretprobe), - KUNIT_CASE(test_stacktrace_on_nested_kretprobe), -#endif -#endif - {} -}; - -static struct kunit_suite kprobes_test_suite = { - .name = "kprobes_test", - .init = kprobes_test_init, - .test_cases = kprobes_testcases, -}; - -kunit_test_suites(&kprobes_test_suite); - -MODULE_LICENSE("GPL"); diff --git a/lib/Makefile b/lib/Makefile index 5efd1b435a37..864ff515814d 100644 --- a/lib/Makefile +++ b/lib/Makefile @@ -100,6 +100,7 @@ obj-$(CONFIG_TEST_MEMINIT) += test_meminit.o obj-$(CONFIG_TEST_LOCKUP) += test_lockup.o obj-$(CONFIG_TEST_HMM) += test_hmm.o obj-$(CONFIG_TEST_FREE_PAGES) += test_free_pages.o +obj-$(CONFIG_KPROBES_SANITY_TEST) += test_kprobes.o # # CFLAGS for compiling floating point code inside the kernel. x86/Makefile turns diff --git a/lib/test_kprobes.c b/lib/test_kprobes.c new file mode 100644 index 000000000000..a5edc2ebc947 --- /dev/null +++ b/lib/test_kprobes.c @@ -0,0 +1,371 @@ +// SPDX-License-Identifier: GPL-2.0-or-later +/* + * test_kprobes.c - simple sanity test for *probes + * + * Copyright IBM Corp. 2008 + */ + +#include +#include +#include +#include + +#define div_factor 3 + +static u32 rand1, preh_val, posth_val; +static u32 (*target)(u32 value); +static u32 (*target2)(u32 value); +static struct kunit *current_test; + +static unsigned long (*internal_target)(void); +static unsigned long (*stacktrace_target)(void); +static unsigned long (*stacktrace_driver)(void); +static unsigned long target_return_address[2]; + +static noinline u32 kprobe_target(u32 value) +{ + return (value / div_factor); +} + +static int kp_pre_handler(struct kprobe *p, struct pt_regs *regs) +{ + KUNIT_EXPECT_FALSE(current_test, preemptible()); + preh_val = (rand1 / div_factor); + return 0; +} + +static void kp_post_handler(struct kprobe *p, struct pt_regs *regs, + unsigned long flags) +{ + KUNIT_EXPECT_FALSE(current_test, preemptible()); + KUNIT_EXPECT_EQ(current_test, preh_val, (rand1 / div_factor)); + posth_val = preh_val + div_factor; +} + +static struct kprobe kp = { + .symbol_name = "kprobe_target", + .pre_handler = kp_pre_handler, + .post_handler = kp_post_handler +}; + +static void test_kprobe(struct kunit *test) +{ + current_test = test; + KUNIT_EXPECT_EQ(test, 0, register_kprobe(&kp)); + target(rand1); + unregister_kprobe(&kp); + KUNIT_EXPECT_NE(test, 0, preh_val); + KUNIT_EXPECT_NE(test, 0, posth_val); +} + +static noinline u32 kprobe_target2(u32 value) +{ + return (value / div_factor) + 1; +} + +static noinline unsigned long kprobe_stacktrace_internal_target(void) +{ + if (!target_return_address[0]) + target_return_address[0] = (unsigned long)__builtin_return_address(0); + return target_return_address[0]; +} + +static noinline unsigned long kprobe_stacktrace_target(void) +{ + if (!target_return_address[1]) + target_return_address[1] = (unsigned long)__builtin_return_address(0); + + if (internal_target) + internal_target(); + + return target_return_address[1]; +} + +static noinline unsigned long kprobe_stacktrace_driver(void) +{ + if (stacktrace_target) + stacktrace_target(); + + /* This is for preventing inlining the function */ + return (unsigned long)__builtin_return_address(0); +} + +static int kp_pre_handler2(struct kprobe *p, struct pt_regs *regs) +{ + preh_val = (rand1 / div_factor) + 1; + return 0; +} + +static void kp_post_handler2(struct kprobe *p, struct pt_regs *regs, + unsigned long flags) +{ + KUNIT_EXPECT_EQ(current_test, preh_val, (rand1 / div_factor) + 1); + posth_val = preh_val + div_factor; +} + +static struct kprobe kp2 = { + .symbol_name = "kprobe_target2", + .pre_handler = kp_pre_handler2, + .post_handler = kp_post_handler2 +}; + +static void test_kprobes(struct kunit *test) +{ + struct kprobe *kps[2] = {&kp, &kp2}; + + current_test = test; + + /* addr and flags should be cleard for reusing kprobe. */ + kp.addr = NULL; + kp.flags = 0; + + KUNIT_EXPECT_EQ(test, 0, register_kprobes(kps, 2)); + preh_val = 0; + posth_val = 0; + target(rand1); + + KUNIT_EXPECT_NE(test, 0, preh_val); + KUNIT_EXPECT_NE(test, 0, posth_val); + + preh_val = 0; + posth_val = 0; + target2(rand1); + + KUNIT_EXPECT_NE(test, 0, preh_val); + KUNIT_EXPECT_NE(test, 0, posth_val); + unregister_kprobes(kps, 2); +} + +#ifdef CONFIG_KRETPROBES +static u32 krph_val; + +static int entry_handler(struct kretprobe_instance *ri, struct pt_regs *regs) +{ + KUNIT_EXPECT_FALSE(current_test, preemptible()); + krph_val = (rand1 / div_factor); + return 0; +} + +static int return_handler(struct kretprobe_instance *ri, struct pt_regs *regs) +{ + unsigned long ret = regs_return_value(regs); + + KUNIT_EXPECT_FALSE(current_test, preemptible()); + KUNIT_EXPECT_EQ(current_test, ret, rand1 / div_factor); + KUNIT_EXPECT_NE(current_test, krph_val, 0); + krph_val = rand1; + return 0; +} + +static struct kretprobe rp = { + .handler = return_handler, + .entry_handler = entry_handler, + .kp.symbol_name = "kprobe_target" +}; + +static void test_kretprobe(struct kunit *test) +{ + current_test = test; + KUNIT_EXPECT_EQ(test, 0, register_kretprobe(&rp)); + target(rand1); + unregister_kretprobe(&rp); + KUNIT_EXPECT_EQ(test, krph_val, rand1); +} + +static int return_handler2(struct kretprobe_instance *ri, struct pt_regs *regs) +{ + unsigned long ret = regs_return_value(regs); + + KUNIT_EXPECT_EQ(current_test, ret, (rand1 / div_factor) + 1); + KUNIT_EXPECT_NE(current_test, krph_val, 0); + krph_val = rand1; + return 0; +} + +static struct kretprobe rp2 = { + .handler = return_handler2, + .entry_handler = entry_handler, + .kp.symbol_name = "kprobe_target2" +}; + +static void test_kretprobes(struct kunit *test) +{ + struct kretprobe *rps[2] = {&rp, &rp2}; + + current_test = test; + /* addr and flags should be cleard for reusing kprobe. */ + rp.kp.addr = NULL; + rp.kp.flags = 0; + KUNIT_EXPECT_EQ(test, 0, register_kretprobes(rps, 2)); + + krph_val = 0; + target(rand1); + KUNIT_EXPECT_EQ(test, krph_val, rand1); + + krph_val = 0; + target2(rand1); + KUNIT_EXPECT_EQ(test, krph_val, rand1); + unregister_kretprobes(rps, 2); +} + +#ifdef CONFIG_ARCH_CORRECT_STACKTRACE_ON_KRETPROBE +#define STACK_BUF_SIZE 16 +static unsigned long stack_buf[STACK_BUF_SIZE]; + +static int stacktrace_return_handler(struct kretprobe_instance *ri, struct pt_regs *regs) +{ + unsigned long retval = regs_return_value(regs); + int i, ret; + + KUNIT_EXPECT_FALSE(current_test, preemptible()); + KUNIT_EXPECT_EQ(current_test, retval, target_return_address[1]); + + /* + * Test stacktrace inside the kretprobe handler, this will involves + * kretprobe trampoline, but must include correct return address + * of the target function. + */ + ret = stack_trace_save(stack_buf, STACK_BUF_SIZE, 0); + KUNIT_EXPECT_NE(current_test, ret, 0); + + for (i = 0; i < ret; i++) { + if (stack_buf[i] == target_return_address[1]) + break; + } + KUNIT_EXPECT_NE(current_test, i, ret); + +#if !IS_MODULE(CONFIG_KPROBES_SANITY_TEST) + /* + * Test stacktrace from pt_regs at the return address. Thus the stack + * trace must start from the target return address. + */ + ret = stack_trace_save_regs(regs, stack_buf, STACK_BUF_SIZE, 0); + KUNIT_EXPECT_NE(current_test, ret, 0); + KUNIT_EXPECT_EQ(current_test, stack_buf[0], target_return_address[1]); +#endif + + return 0; +} + +static struct kretprobe rp3 = { + .handler = stacktrace_return_handler, + .kp.symbol_name = "kprobe_stacktrace_target" +}; + +static void test_stacktrace_on_kretprobe(struct kunit *test) +{ + unsigned long myretaddr = (unsigned long)__builtin_return_address(0); + + current_test = test; + rp3.kp.addr = NULL; + rp3.kp.flags = 0; + + /* + * Run the stacktrace_driver() to record correct return address in + * stacktrace_target() and ensure stacktrace_driver() call is not + * inlined by checking the return address of stacktrace_driver() + * and the return address of this function is different. + */ + KUNIT_ASSERT_NE(test, myretaddr, stacktrace_driver()); + + KUNIT_ASSERT_EQ(test, 0, register_kretprobe(&rp3)); + KUNIT_ASSERT_NE(test, myretaddr, stacktrace_driver()); + unregister_kretprobe(&rp3); +} + +static int stacktrace_internal_return_handler(struct kretprobe_instance *ri, struct pt_regs *regs) +{ + unsigned long retval = regs_return_value(regs); + int i, ret; + + KUNIT_EXPECT_FALSE(current_test, preemptible()); + KUNIT_EXPECT_EQ(current_test, retval, target_return_address[0]); + + /* + * Test stacktrace inside the kretprobe handler for nested case. + * The unwinder will find the kretprobe_trampoline address on the + * return address, and kretprobe must solve that. + */ + ret = stack_trace_save(stack_buf, STACK_BUF_SIZE, 0); + KUNIT_EXPECT_NE(current_test, ret, 0); + + for (i = 0; i < ret - 1; i++) { + if (stack_buf[i] == target_return_address[0]) { + KUNIT_EXPECT_EQ(current_test, stack_buf[i + 1], target_return_address[1]); + break; + } + } + KUNIT_EXPECT_NE(current_test, i, ret); + +#if !IS_MODULE(CONFIG_KPROBES_SANITY_TEST) + /* Ditto for the regs version. */ + ret = stack_trace_save_regs(regs, stack_buf, STACK_BUF_SIZE, 0); + KUNIT_EXPECT_NE(current_test, ret, 0); + KUNIT_EXPECT_EQ(current_test, stack_buf[0], target_return_address[0]); + KUNIT_EXPECT_EQ(current_test, stack_buf[1], target_return_address[1]); +#endif + + return 0; +} + +static struct kretprobe rp4 = { + .handler = stacktrace_internal_return_handler, + .kp.symbol_name = "kprobe_stacktrace_internal_target" +}; + +static void test_stacktrace_on_nested_kretprobe(struct kunit *test) +{ + unsigned long myretaddr = (unsigned long)__builtin_return_address(0); + struct kretprobe *rps[2] = {&rp3, &rp4}; + + current_test = test; + rp3.kp.addr = NULL; + rp3.kp.flags = 0; + + //KUNIT_ASSERT_NE(test, myretaddr, stacktrace_driver()); + + KUNIT_ASSERT_EQ(test, 0, register_kretprobes(rps, 2)); + KUNIT_ASSERT_NE(test, myretaddr, stacktrace_driver()); + unregister_kretprobes(rps, 2); +} +#endif /* CONFIG_ARCH_CORRECT_STACKTRACE_ON_KRETPROBE */ + +#endif /* CONFIG_KRETPROBES */ + +static int kprobes_test_init(struct kunit *test) +{ + target = kprobe_target; + target2 = kprobe_target2; + stacktrace_target = kprobe_stacktrace_target; + internal_target = kprobe_stacktrace_internal_target; + stacktrace_driver = kprobe_stacktrace_driver; + + do { + rand1 = prandom_u32(); + } while (rand1 <= div_factor); + return 0; +} + +static struct kunit_case kprobes_testcases[] = { + KUNIT_CASE(test_kprobe), + KUNIT_CASE(test_kprobes), +#ifdef CONFIG_KRETPROBES + KUNIT_CASE(test_kretprobe), + KUNIT_CASE(test_kretprobes), +#ifdef CONFIG_ARCH_CORRECT_STACKTRACE_ON_KRETPROBE + KUNIT_CASE(test_stacktrace_on_kretprobe), + KUNIT_CASE(test_stacktrace_on_nested_kretprobe), +#endif +#endif + {} +}; + +static struct kunit_suite kprobes_test_suite = { + .name = "kprobes_test", + .init = kprobes_test_init, + .test_cases = kprobes_testcases, +}; + +kunit_test_suites(&kprobes_test_suite); + +MODULE_LICENSE("GPL"); -- cgit v1.2.3 From 7529cc7fbd9c02eda6851f3260416cbe198a321d Mon Sep 17 00:00:00 2001 From: Tariq Toukan Date: Wed, 30 Dec 2020 11:41:52 +0200 Subject: lib: bitmap: Introduce node-aware alloc API Expose new node-aware API for bitmap allocation: bitmap_alloc_node() / bitmap_zalloc_node(). Signed-off-by: Tariq Toukan Reviewed-by: Moshe Shemesh Signed-off-by: Saeed Mahameed --- include/linux/bitmap.h | 2 ++ lib/bitmap.c | 13 +++++++++++++ 2 files changed, 15 insertions(+) (limited to 'lib') diff --git a/include/linux/bitmap.h b/include/linux/bitmap.h index 37f36dad18bd..a241dcf50f39 100644 --- a/include/linux/bitmap.h +++ b/include/linux/bitmap.h @@ -123,6 +123,8 @@ struct device; */ unsigned long *bitmap_alloc(unsigned int nbits, gfp_t flags); unsigned long *bitmap_zalloc(unsigned int nbits, gfp_t flags); +unsigned long *bitmap_alloc_node(unsigned int nbits, gfp_t flags, int node); +unsigned long *bitmap_zalloc_node(unsigned int nbits, gfp_t flags, int node); void bitmap_free(const unsigned long *bitmap); /* Managed variants of the above. */ diff --git a/lib/bitmap.c b/lib/bitmap.c index 663dd81967d4..926408883456 100644 --- a/lib/bitmap.c +++ b/lib/bitmap.c @@ -1398,6 +1398,19 @@ unsigned long *bitmap_zalloc(unsigned int nbits, gfp_t flags) } EXPORT_SYMBOL(bitmap_zalloc); +unsigned long *bitmap_alloc_node(unsigned int nbits, gfp_t flags, int node) +{ + return kmalloc_array_node(BITS_TO_LONGS(nbits), sizeof(unsigned long), + flags, node); +} +EXPORT_SYMBOL(bitmap_alloc_node); + +unsigned long *bitmap_zalloc_node(unsigned int nbits, gfp_t flags, int node) +{ + return bitmap_alloc_node(nbits, flags | __GFP_ZERO, node); +} +EXPORT_SYMBOL(bitmap_zalloc_node); + void bitmap_free(const unsigned long *bitmap) { kfree(bitmap); -- cgit v1.2.3 From c666d447e091be3a742588b49290e7733115769f Mon Sep 17 00:00:00 2001 From: "Matthew Wilcox (Oracle)" Date: Tue, 19 Oct 2021 15:26:17 +0100 Subject: test_printf: Make pft array const Instead of assigning ptf[i].value, leave the values in the on-stack array and then we can make the array const. Signed-off-by: Matthew Wilcox (Oracle) Reviewed-by: Yafang Shao Reviewed-by: Petr Mladek Reviewed-by: Anshuman Khandual Signed-off-by: Petr Mladek Link: https://lore.kernel.org/r/20211019142621.2810043-2-willy@infradead.org --- lib/test_printf.c | 21 ++++++++------------- 1 file changed, 8 insertions(+), 13 deletions(-) (limited to 'lib') diff --git a/lib/test_printf.c b/lib/test_printf.c index 55082432f37e..a52c1c3a55ba 100644 --- a/lib/test_printf.c +++ b/lib/test_printf.c @@ -586,22 +586,21 @@ struct page_flags_test { int width; int shift; int mask; - unsigned long value; const char *fmt; const char *name; }; -static struct page_flags_test pft[] = { +static const struct page_flags_test pft[] = { {SECTIONS_WIDTH, SECTIONS_PGSHIFT, SECTIONS_MASK, - 0, "%d", "section"}, + "%d", "section"}, {NODES_WIDTH, NODES_PGSHIFT, NODES_MASK, - 0, "%d", "node"}, + "%d", "node"}, {ZONES_WIDTH, ZONES_PGSHIFT, ZONES_MASK, - 0, "%d", "zone"}, + "%d", "zone"}, {LAST_CPUPID_WIDTH, LAST_CPUPID_PGSHIFT, LAST_CPUPID_MASK, - 0, "%#x", "lastcpupid"}, + "%#x", "lastcpupid"}, {KASAN_TAG_WIDTH, KASAN_TAG_PGSHIFT, KASAN_TAG_MASK, - 0, "%#x", "kasantag"}, + "%#x", "kasantag"}, }; static void __init @@ -627,10 +626,6 @@ page_flags_test(int section, int node, int zone, int last_cpupid, #endif } - /* Set the test value */ - for (i = 0; i < ARRAY_SIZE(pft); i++) - pft[i].value = values[i]; - for (i = 0; i < ARRAY_SIZE(pft); i++) { if (!pft[i].width) continue; @@ -640,11 +635,11 @@ page_flags_test(int section, int node, int zone, int last_cpupid, size = strlen(cmp_buf); } - page_flags |= (pft[i].value & pft[i].mask) << pft[i].shift; + page_flags |= (values[i] & pft[i].mask) << pft[i].shift; snprintf(cmp_buf + size, BUF_SIZE - size, "%s=", pft[i].name); size = strlen(cmp_buf); snprintf(cmp_buf + size, BUF_SIZE - size, pft[i].fmt, - pft[i].value & pft[i].mask); + values[i] & pft[i].mask); size = strlen(cmp_buf); append = true; } -- cgit v1.2.3 From a25a0854a2264a0c592ba1ea01a165101f8c1a6c Mon Sep 17 00:00:00 2001 From: "Matthew Wilcox (Oracle)" Date: Tue, 19 Oct 2021 15:26:18 +0100 Subject: test_printf: Remove separate page_flags variable Keep flags intact so that we also test what happens when unknown flags are passed to %pGp. Signed-off-by: Matthew Wilcox (Oracle) Reviewed-by: Yafang Shao Reviewed-by: Petr Mladek Signed-off-by: Petr Mladek Link: https://lore.kernel.org/r/20211019142621.2810043-3-willy@infradead.org --- lib/test_printf.c | 12 +++++------- 1 file changed, 5 insertions(+), 7 deletions(-) (limited to 'lib') diff --git a/lib/test_printf.c b/lib/test_printf.c index a52c1c3a55ba..4531063afd45 100644 --- a/lib/test_printf.c +++ b/lib/test_printf.c @@ -605,17 +605,15 @@ static const struct page_flags_test pft[] = { static void __init page_flags_test(int section, int node, int zone, int last_cpupid, - int kasan_tag, int flags, const char *name, char *cmp_buf) + int kasan_tag, unsigned long flags, const char *name, + char *cmp_buf) { unsigned long values[] = {section, node, zone, last_cpupid, kasan_tag}; - unsigned long page_flags = 0; unsigned long size = 0; bool append = false; int i; - flags &= PAGEFLAGS_MASK; - if (flags) { - page_flags |= flags; + if (flags & PAGEFLAGS_MASK) { snprintf(cmp_buf + size, BUF_SIZE - size, "%s", name); size = strlen(cmp_buf); #if SECTIONS_WIDTH || NODES_WIDTH || ZONES_WIDTH || \ @@ -635,7 +633,7 @@ page_flags_test(int section, int node, int zone, int last_cpupid, size = strlen(cmp_buf); } - page_flags |= (values[i] & pft[i].mask) << pft[i].shift; + flags |= (values[i] & pft[i].mask) << pft[i].shift; snprintf(cmp_buf + size, BUF_SIZE - size, "%s=", pft[i].name); size = strlen(cmp_buf); snprintf(cmp_buf + size, BUF_SIZE - size, pft[i].fmt, @@ -644,7 +642,7 @@ page_flags_test(int section, int node, int zone, int last_cpupid, append = true; } - test(cmp_buf, "%pGp", &page_flags); + test(cmp_buf, "%pGp", &flags); } static void __init -- cgit v1.2.3 From 5b358b0de963f822226bfee916fb53c80bae4000 Mon Sep 17 00:00:00 2001 From: "Matthew Wilcox (Oracle)" Date: Tue, 19 Oct 2021 15:26:19 +0100 Subject: test_printf: Remove custom appending of '|' Instead of having an ifdef to decide whether to print a |, use the 'append' functionality of the main loop to print it. Signed-off-by: Matthew Wilcox (Oracle) Reviewed-by: Yafang Shao Reviewed-by: Petr Mladek Reviewed-by: Anshuman Khandual Signed-off-by: Petr Mladek Link: https://lore.kernel.org/r/20211019142621.2810043-4-willy@infradead.org --- lib/test_printf.c | 7 +------ 1 file changed, 1 insertion(+), 6 deletions(-) (limited to 'lib') diff --git a/lib/test_printf.c b/lib/test_printf.c index 4531063afd45..ec584196cb99 100644 --- a/lib/test_printf.c +++ b/lib/test_printf.c @@ -616,12 +616,7 @@ page_flags_test(int section, int node, int zone, int last_cpupid, if (flags & PAGEFLAGS_MASK) { snprintf(cmp_buf + size, BUF_SIZE - size, "%s", name); size = strlen(cmp_buf); -#if SECTIONS_WIDTH || NODES_WIDTH || ZONES_WIDTH || \ - LAST_CPUPID_WIDTH || KASAN_TAG_WIDTH - /* Other information also included in page flags */ - snprintf(cmp_buf + size, BUF_SIZE - size, "|"); - size = strlen(cmp_buf); -#endif + append = true; } for (i = 0; i < ARRAY_SIZE(pft); i++) { -- cgit v1.2.3 From 507f98603607d43cb76ed39c370c4dc1ed6a94f9 Mon Sep 17 00:00:00 2001 From: "Matthew Wilcox (Oracle)" Date: Tue, 19 Oct 2021 15:26:20 +0100 Subject: test_printf: Append strings more efficiently Use scnprintf instead of snprintf + strlen. Signed-off-by: Matthew Wilcox (Oracle) Reviewed-by: Petr Mladek Reviewed-by: Yafang Shao Signed-off-by: Petr Mladek Link: https://lore.kernel.org/r/20211019142621.2810043-5-willy@infradead.org --- lib/test_printf.c | 18 +++++++----------- 1 file changed, 7 insertions(+), 11 deletions(-) (limited to 'lib') diff --git a/lib/test_printf.c b/lib/test_printf.c index ec584196cb99..d09993fca463 100644 --- a/lib/test_printf.c +++ b/lib/test_printf.c @@ -614,8 +614,7 @@ page_flags_test(int section, int node, int zone, int last_cpupid, int i; if (flags & PAGEFLAGS_MASK) { - snprintf(cmp_buf + size, BUF_SIZE - size, "%s", name); - size = strlen(cmp_buf); + size += scnprintf(cmp_buf + size, BUF_SIZE - size, "%s", name); append = true; } @@ -623,17 +622,14 @@ page_flags_test(int section, int node, int zone, int last_cpupid, if (!pft[i].width) continue; - if (append) { - snprintf(cmp_buf + size, BUF_SIZE - size, "|"); - size = strlen(cmp_buf); - } + if (append) + size += scnprintf(cmp_buf + size, BUF_SIZE - size, "|"); flags |= (values[i] & pft[i].mask) << pft[i].shift; - snprintf(cmp_buf + size, BUF_SIZE - size, "%s=", pft[i].name); - size = strlen(cmp_buf); - snprintf(cmp_buf + size, BUF_SIZE - size, pft[i].fmt, - values[i] & pft[i].mask); - size = strlen(cmp_buf); + size += scnprintf(cmp_buf + size, BUF_SIZE - size, "%s=", + pft[i].name); + size += scnprintf(cmp_buf + size, BUF_SIZE - size, pft[i].fmt, + values[i] & pft[i].mask); append = true; } -- cgit v1.2.3 From 23efd0804c0a869dfb1e78470f80a27251317b7e Mon Sep 17 00:00:00 2001 From: "Matthew Wilcox (Oracle)" Date: Tue, 19 Oct 2021 15:26:21 +0100 Subject: vsprintf: Make %pGp print the hex value All existing users of %pGp want the hex value as well as the decoded flag names. This looks awkward (passing the same parameter to printf twice), so move that functionality into the core. If we want, we can make that optional with flag arguments to %pGp in the future. Signed-off-by: Matthew Wilcox (Oracle) Reviewed-by: Yafang Shao Reviewed-by: Petr Mladek Signed-off-by: Petr Mladek Link: https://lore.kernel.org/r/20211019142621.2810043-6-willy@infradead.org --- lib/test_printf.c | 9 +++++++-- lib/vsprintf.c | 8 ++++++++ mm/debug.c | 2 +- mm/memory-failure.c | 8 ++++---- mm/page_owner.c | 4 ++-- mm/slub.c | 4 ++-- 6 files changed, 24 insertions(+), 11 deletions(-) (limited to 'lib') diff --git a/lib/test_printf.c b/lib/test_printf.c index d09993fca463..07309c45f327 100644 --- a/lib/test_printf.c +++ b/lib/test_printf.c @@ -609,10 +609,14 @@ page_flags_test(int section, int node, int zone, int last_cpupid, char *cmp_buf) { unsigned long values[] = {section, node, zone, last_cpupid, kasan_tag}; - unsigned long size = 0; + unsigned long size; bool append = false; int i; + for (i = 0; i < ARRAY_SIZE(values); i++) + flags |= (values[i] & pft[i].mask) << pft[i].shift; + + size = scnprintf(cmp_buf, BUF_SIZE, "%#lx(", flags); if (flags & PAGEFLAGS_MASK) { size += scnprintf(cmp_buf + size, BUF_SIZE - size, "%s", name); append = true; @@ -625,7 +629,6 @@ page_flags_test(int section, int node, int zone, int last_cpupid, if (append) size += scnprintf(cmp_buf + size, BUF_SIZE - size, "|"); - flags |= (values[i] & pft[i].mask) << pft[i].shift; size += scnprintf(cmp_buf + size, BUF_SIZE - size, "%s=", pft[i].name); size += scnprintf(cmp_buf + size, BUF_SIZE - size, pft[i].fmt, @@ -633,6 +636,8 @@ page_flags_test(int section, int node, int zone, int last_cpupid, append = true; } + snprintf(cmp_buf + size, BUF_SIZE - size, ")"); + test(cmp_buf, "%pGp", &flags); } diff --git a/lib/vsprintf.c b/lib/vsprintf.c index d7ad44f2c8f5..214098248610 100644 --- a/lib/vsprintf.c +++ b/lib/vsprintf.c @@ -2023,6 +2023,11 @@ char *format_page_flags(char *buf, char *end, unsigned long flags) bool append = false; int i; + buf = number(buf, end, flags, default_flag_spec); + if (buf < end) + *buf = '('; + buf++; + /* Page flags from the main area. */ if (main_flags) { buf = format_flags(buf, end, main_flags, pageflag_names); @@ -2051,6 +2056,9 @@ char *format_page_flags(char *buf, char *end, unsigned long flags) append = true; } + if (buf < end) + *buf = ')'; + buf++; return buf; } diff --git a/mm/debug.c b/mm/debug.c index e73fe0a8ec3d..ca9611784e4b 100644 --- a/mm/debug.c +++ b/mm/debug.c @@ -160,7 +160,7 @@ static void __dump_page(struct page *page) out_mapping: BUILD_BUG_ON(ARRAY_SIZE(pageflag_names) != __NR_PAGEFLAGS + 1); - pr_warn("%sflags: %#lx(%pGp)%s\n", type, head->flags, &head->flags, + pr_warn("%sflags: %pGp%s\n", type, &head->flags, page_cma ? " CMA" : ""); print_hex_dump(KERN_WARNING, "raw: ", DUMP_PREFIX_NONE, 32, sizeof(unsigned long), page, diff --git a/mm/memory-failure.c b/mm/memory-failure.c index 54879c339024..58ab5161a8ce 100644 --- a/mm/memory-failure.c +++ b/mm/memory-failure.c @@ -2109,14 +2109,14 @@ static int __soft_offline_page(struct page *page) if (!list_empty(&pagelist)) putback_movable_pages(&pagelist); - pr_info("soft offline: %#lx: %s migration failed %d, type %lx (%pGp)\n", - pfn, msg_page[huge], ret, page->flags, &page->flags); + pr_info("soft offline: %#lx: %s migration failed %d, type %pGp\n", + pfn, msg_page[huge], ret, &page->flags); if (ret > 0) ret = -EBUSY; } } else { - pr_info("soft offline: %#lx: %s isolation failed, page count %d, type %lx (%pGp)\n", - pfn, msg_page[huge], page_count(page), page->flags, &page->flags); + pr_info("soft offline: %#lx: %s isolation failed, page count %d, type %pGp\n", + pfn, msg_page[huge], page_count(page), &page->flags); ret = -EBUSY; } return ret; diff --git a/mm/page_owner.c b/mm/page_owner.c index 62402d22539b..4afc713ca525 100644 --- a/mm/page_owner.c +++ b/mm/page_owner.c @@ -351,12 +351,12 @@ print_page_owner(char __user *buf, size_t count, unsigned long pfn, pageblock_mt = get_pageblock_migratetype(page); page_mt = gfp_migratetype(page_owner->gfp_mask); ret += snprintf(kbuf + ret, count - ret, - "PFN %lu type %s Block %lu type %s Flags %#lx(%pGp)\n", + "PFN %lu type %s Block %lu type %s Flags %pGp\n", pfn, migratetype_names[page_mt], pfn >> pageblock_order, migratetype_names[pageblock_mt], - page->flags, &page->flags); + &page->flags); if (ret >= count) goto err; diff --git a/mm/slub.c b/mm/slub.c index 3d2025f7163b..f7ac28646580 100644 --- a/mm/slub.c +++ b/mm/slub.c @@ -763,9 +763,9 @@ void print_tracking(struct kmem_cache *s, void *object) static void print_page_info(struct page *page) { - pr_err("Slab 0x%p objects=%u used=%u fp=0x%p flags=%#lx(%pGp)\n", + pr_err("Slab 0x%p objects=%u used=%u fp=0x%p flags=%pGp\n", page, page->objects, page->inuse, page->freelist, - page->flags, &page->flags); + &page->flags); } -- cgit v1.2.3 From 24a1dffbecafeb00d8830985eb7a318e37aabc4e Mon Sep 17 00:00:00 2001 From: Andy Shevchenko Date: Tue, 26 Oct 2021 17:03:56 +0300 Subject: lib/vsprintf.c: Amend static asserts for format specifier flags There are couple of improvements to static asserts against the format specifier flags: - new static assert for SIGN - fix static assert for SMALL SMALL is not equal to ASCII code of white space, it equals to the bit difference between capital and small letters (however the value is the same, semantically expression means different things). Signed-off-by: Andy Shevchenko Signed-off-by: Petr Mladek Link: https://lore.kernel.org/r/20211026140356.45610-1-andriy.shevchenko@linux.intel.com --- lib/vsprintf.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'lib') diff --git a/lib/vsprintf.c b/lib/vsprintf.c index 26c83943748a..1173930ed9d3 100644 --- a/lib/vsprintf.c +++ b/lib/vsprintf.c @@ -408,8 +408,9 @@ int num_to_str(char *buf, int size, unsigned long long num, unsigned int width) #define SMALL 32 /* use lowercase in hex (must be 32 == 0x20) */ #define SPECIAL 64 /* prefix hex with "0x", octal with "0" */ +static_assert(SIGN == 1); static_assert(ZEROPAD == ('0' - ' ')); -static_assert(SMALL == ' '); +static_assert(SMALL == ('a' ^ 'A')); enum format_type { FORMAT_TYPE_NONE, /* Just a string part */ -- cgit v1.2.3 From 39d9c1c103d3061ac94219ac12c04753860b337e Mon Sep 17 00:00:00 2001 From: "Steven Rostedt (VMware)" Date: Wed, 27 Oct 2021 10:57:53 -0400 Subject: bootconfig: Initialize ret in xbc_parse_tree() The do while loop continues while ret is zero, but ret is never initialized. The check for ret in the loop at the while should always be initialized, but if an empty string were to be passed in, q would be NULL and p would be '\0', and it would break out of the loop without ever setting ret. Set ret to zero, and then xbc_verify_tree() would be called and catch that it is an empty tree and report the proper error. Link: https://lkml.kernel.org/r/20211027105753.6ab9da5f@gandalf.local.home Fixes: bdac5c2b243f ("bootconfig: Allocate xbc_data inside xbc_init()") Reported-by: kernel test robot Reported-by: Andrew Morton Acked-by: Masami Hiramatsu Signed-off-by: Steven Rostedt (VMware) --- lib/bootconfig.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'lib') diff --git a/lib/bootconfig.c b/lib/bootconfig.c index a10ab25f6fcc..70e0d52ffd24 100644 --- a/lib/bootconfig.c +++ b/lib/bootconfig.c @@ -836,7 +836,7 @@ static int __init xbc_verify_tree(void) static int __init xbc_parse_tree(void) { char *p, *q; - int ret, c; + int ret = 0, c; last_parent = NULL; p = xbc_data; -- cgit v1.2.3 From b066abba3ef16a4a085d237e95da0de3f0b87713 Mon Sep 17 00:00:00 2001 From: Tiezhu Yang Date: Thu, 28 Oct 2021 09:25:21 +0800 Subject: bpf, tests: Add module parameter test_suite to test_bpf module After commit 9298e63eafea ("bpf/tests: Add exhaustive tests of ALU operand magnitudes"), when modprobe test_bpf.ko with JIT on mips64, there exists segment fault due to the following reason: [...] ALU64_MOV_X: all register value magnitudes jited:1 Break instruction in kernel code[#1] [...] It seems that the related JIT implementations of some test cases in test_bpf() have problems. At this moment, I do not care about the segment fault while I just want to verify the test cases of tail calls. Based on the above background and motivation, add the following module parameter test_suite to the test_bpf.ko: test_suite=: only the specified test suite will be run, the string can be "test_bpf", "test_tail_calls" or "test_skb_segment". If test_suite is not specified, but test_id, test_name or test_range is specified, set 'test_bpf' as the default test suite. This is useful to only test the corresponding test suite when specifying the valid test_suite string. Any invalid test suite will result in -EINVAL being returned and no tests being run. If the test_suite is not specified or specified as empty string, it does not change the current logic, all of the test cases will be run. Here are some test results: # dmesg -c # modprobe test_bpf # dmesg | grep Summary test_bpf: Summary: 1009 PASSED, 0 FAILED, [0/997 JIT'ed] test_bpf: test_tail_calls: Summary: 8 PASSED, 0 FAILED, [0/8 JIT'ed] test_bpf: test_skb_segment: Summary: 2 PASSED, 0 FAILED # rmmod test_bpf # dmesg -c # modprobe test_bpf test_suite=test_bpf # dmesg | tail -1 test_bpf: Summary: 1009 PASSED, 0 FAILED, [0/997 JIT'ed] # rmmod test_bpf # dmesg -c # modprobe test_bpf test_suite=test_tail_calls # dmesg test_bpf: #0 Tail call leaf jited:0 21 PASS [...] test_bpf: #7 Tail call error path, index out of range jited:0 32 PASS test_bpf: test_tail_calls: Summary: 8 PASSED, 0 FAILED, [0/8 JIT'ed] # rmmod test_bpf # dmesg -c # modprobe test_bpf test_suite=test_skb_segment # dmesg test_bpf: #0 gso_with_rx_frags PASS test_bpf: #1 gso_linear_no_head_frag PASS test_bpf: test_skb_segment: Summary: 2 PASSED, 0 FAILED # rmmod test_bpf # dmesg -c # modprobe test_bpf test_id=1 # dmesg test_bpf: test_bpf: set 'test_bpf' as the default test_suite. test_bpf: #1 TXA jited:0 54 51 50 PASS test_bpf: Summary: 1 PASSED, 0 FAILED, [0/1 JIT'ed] # rmmod test_bpf # dmesg -c # modprobe test_bpf test_suite=test_bpf test_name=TXA # dmesg test_bpf: #1 TXA jited:0 54 50 51 PASS test_bpf: Summary: 1 PASSED, 0 FAILED, [0/1 JIT'ed] # rmmod test_bpf # dmesg -c # modprobe test_bpf test_suite=test_tail_calls test_range=6,7 # dmesg test_bpf: #6 Tail call error path, NULL target jited:0 41 PASS test_bpf: #7 Tail call error path, index out of range jited:0 32 PASS test_bpf: test_tail_calls: Summary: 2 PASSED, 0 FAILED, [0/2 JIT'ed] # rmmod test_bpf # dmesg -c # modprobe test_bpf test_suite=test_skb_segment test_id=1 # dmesg test_bpf: #1 gso_linear_no_head_frag PASS test_bpf: test_skb_segment: Summary: 1 PASSED, 0 FAILED By the way, the above segment fault has been fixed in the latest bpf-next tree which contains the mips64 JIT rework. Signed-off-by: Tiezhu Yang Signed-off-by: Daniel Borkmann Tested-by: Johan Almbladh Acked-by: Johan Almbladh Link: https://lore.kernel.org/bpf/1635384321-28128-1-git-send-email-yangtiezhu@loongson.cn --- lib/test_bpf.c | 212 ++++++++++++++++++++++++++++++++++++--------------------- 1 file changed, 135 insertions(+), 77 deletions(-) (limited to 'lib') diff --git a/lib/test_bpf.c b/lib/test_bpf.c index e5b10fdefab5..adae39567264 100644 --- a/lib/test_bpf.c +++ b/lib/test_bpf.c @@ -14316,72 +14316,9 @@ module_param_string(test_name, test_name, sizeof(test_name), 0); static int test_id = -1; module_param(test_id, int, 0); -static int test_range[2] = { 0, ARRAY_SIZE(tests) - 1 }; +static int test_range[2] = { 0, INT_MAX }; module_param_array(test_range, int, NULL, 0); -static __init int find_test_index(const char *test_name) -{ - int i; - - for (i = 0; i < ARRAY_SIZE(tests); i++) { - if (!strcmp(tests[i].descr, test_name)) - return i; - } - return -1; -} - -static __init int prepare_bpf_tests(void) -{ - if (test_id >= 0) { - /* - * if a test_id was specified, use test_range to - * cover only that test. - */ - if (test_id >= ARRAY_SIZE(tests)) { - pr_err("test_bpf: invalid test_id specified.\n"); - return -EINVAL; - } - - test_range[0] = test_id; - test_range[1] = test_id; - } else if (*test_name) { - /* - * if a test_name was specified, find it and setup - * test_range to cover only that test. - */ - int idx = find_test_index(test_name); - - if (idx < 0) { - pr_err("test_bpf: no test named '%s' found.\n", - test_name); - return -EINVAL; - } - test_range[0] = idx; - test_range[1] = idx; - } else { - /* - * check that the supplied test_range is valid. - */ - if (test_range[0] >= ARRAY_SIZE(tests) || - test_range[1] >= ARRAY_SIZE(tests) || - test_range[0] < 0 || test_range[1] < 0) { - pr_err("test_bpf: test_range is out of bound.\n"); - return -EINVAL; - } - - if (test_range[1] < test_range[0]) { - pr_err("test_bpf: test_range is ending before it starts.\n"); - return -EINVAL; - } - } - - return 0; -} - -static __init void destroy_bpf_tests(void) -{ -} - static bool exclude_test(int test_id) { return test_id < test_range[0] || test_id > test_range[1]; @@ -14553,6 +14490,10 @@ static __init int test_skb_segment(void) for (i = 0; i < ARRAY_SIZE(skb_segment_tests); i++) { const struct skb_segment_test *test = &skb_segment_tests[i]; + cond_resched(); + if (exclude_test(i)) + continue; + pr_info("#%d %s ", i, test->descr); if (test_skb_segment_single(test)) { @@ -14934,6 +14875,8 @@ static __init int test_tail_calls(struct bpf_array *progs) int ret; cond_resched(); + if (exclude_test(i)) + continue; pr_info("#%d %s ", i, test->descr); if (!fp) { @@ -14966,29 +14909,144 @@ static __init int test_tail_calls(struct bpf_array *progs) return err_cnt ? -EINVAL : 0; } +static char test_suite[32]; +module_param_string(test_suite, test_suite, sizeof(test_suite), 0); + +static __init int find_test_index(const char *test_name) +{ + int i; + + if (!strcmp(test_suite, "test_bpf")) { + for (i = 0; i < ARRAY_SIZE(tests); i++) { + if (!strcmp(tests[i].descr, test_name)) + return i; + } + } + + if (!strcmp(test_suite, "test_tail_calls")) { + for (i = 0; i < ARRAY_SIZE(tail_call_tests); i++) { + if (!strcmp(tail_call_tests[i].descr, test_name)) + return i; + } + } + + if (!strcmp(test_suite, "test_skb_segment")) { + for (i = 0; i < ARRAY_SIZE(skb_segment_tests); i++) { + if (!strcmp(skb_segment_tests[i].descr, test_name)) + return i; + } + } + + return -1; +} + +static __init int prepare_test_range(void) +{ + int valid_range; + + if (!strcmp(test_suite, "test_bpf")) + valid_range = ARRAY_SIZE(tests); + else if (!strcmp(test_suite, "test_tail_calls")) + valid_range = ARRAY_SIZE(tail_call_tests); + else if (!strcmp(test_suite, "test_skb_segment")) + valid_range = ARRAY_SIZE(skb_segment_tests); + else + return 0; + + if (test_id >= 0) { + /* + * if a test_id was specified, use test_range to + * cover only that test. + */ + if (test_id >= valid_range) { + pr_err("test_bpf: invalid test_id specified for '%s' suite.\n", + test_suite); + return -EINVAL; + } + + test_range[0] = test_id; + test_range[1] = test_id; + } else if (*test_name) { + /* + * if a test_name was specified, find it and setup + * test_range to cover only that test. + */ + int idx = find_test_index(test_name); + + if (idx < 0) { + pr_err("test_bpf: no test named '%s' found for '%s' suite.\n", + test_name, test_suite); + return -EINVAL; + } + test_range[0] = idx; + test_range[1] = idx; + } else if (test_range[0] != 0 || test_range[1] != INT_MAX) { + /* + * check that the supplied test_range is valid. + */ + if (test_range[0] < 0 || test_range[1] >= valid_range) { + pr_err("test_bpf: test_range is out of bound for '%s' suite.\n", + test_suite); + return -EINVAL; + } + + if (test_range[1] < test_range[0]) { + pr_err("test_bpf: test_range is ending before it starts.\n"); + return -EINVAL; + } + } + + return 0; +} + static int __init test_bpf_init(void) { struct bpf_array *progs = NULL; int ret; - ret = prepare_bpf_tests(); + if (strlen(test_suite) && + strcmp(test_suite, "test_bpf") && + strcmp(test_suite, "test_tail_calls") && + strcmp(test_suite, "test_skb_segment")) { + pr_err("test_bpf: invalid test_suite '%s' specified.\n", test_suite); + return -EINVAL; + } + + /* + * if test_suite is not specified, but test_id, test_name or test_range + * is specified, set 'test_bpf' as the default test suite. + */ + if (!strlen(test_suite) && + (test_id != -1 || strlen(test_name) || + (test_range[0] != 0 || test_range[1] != INT_MAX))) { + pr_info("test_bpf: set 'test_bpf' as the default test_suite.\n"); + strscpy(test_suite, "test_bpf", sizeof(test_suite)); + } + + ret = prepare_test_range(); if (ret < 0) return ret; - ret = test_bpf(); - destroy_bpf_tests(); - if (ret) - return ret; + if (!strlen(test_suite) || !strcmp(test_suite, "test_bpf")) { + ret = test_bpf(); + if (ret) + return ret; + } - ret = prepare_tail_call_tests(&progs); - if (ret) - return ret; - ret = test_tail_calls(progs); - destroy_tail_call_tests(progs); - if (ret) - return ret; + if (!strlen(test_suite) || !strcmp(test_suite, "test_tail_calls")) { + ret = prepare_tail_call_tests(&progs); + if (ret) + return ret; + ret = test_tail_calls(progs); + destroy_tail_call_tests(progs); + if (ret) + return ret; + } - return test_skb_segment(); + if (!strlen(test_suite) || !strcmp(test_suite, "test_skb_segment")) + return test_skb_segment(); + + return 0; } static void __exit test_bpf_exit(void) -- cgit v1.2.3 From 5c4e0a21fae877a7ef89be6dcc6263ec672372b8 Mon Sep 17 00:00:00 2001 From: Guenter Roeck Date: Tue, 2 Nov 2021 07:24:20 -0700 Subject: string: uninline memcpy_and_pad When building m68k:allmodconfig, recent versions of gcc generate the following error if the length of UTS_RELEASE is less than 8 bytes. In function 'memcpy_and_pad', inlined from 'nvmet_execute_disc_identify' at drivers/nvme/target/discovery.c:268:2: arch/m68k/include/asm/string.h:72:25: error: '__builtin_memcpy' reading 8 bytes from a region of size 7 Discussions around the problem suggest that this only happens if an architecture does not provide strlen(), if -ffreestanding is provided as compiler option, and if CONFIG_FORTIFY_SOURCE=n. All of this is the case for m68k. The exact reasons are unknown, but seem to be related to the ability of the compiler to evaluate the return value of strlen() and the resulting execution flow in memcpy_and_pad(). It would be possible to work around the problem by using sizeof(UTS_RELEASE) instead of strlen(UTS_RELEASE), but that would only postpone the problem until the function is called in a similar way. Uninline memcpy_and_pad() instead to solve the problem for good. Suggested-by: Linus Torvalds Reviewed-by: Geert Uytterhoeven Acked-by: Andy Shevchenko Signed-off-by: Guenter Roeck Signed-off-by: Linus Torvalds --- include/linux/string.h | 19 ++----------------- lib/string_helpers.c | 20 ++++++++++++++++++++ 2 files changed, 22 insertions(+), 17 deletions(-) (limited to 'lib') diff --git a/include/linux/string.h b/include/linux/string.h index 5a36608144a9..b6572aeca2f5 100644 --- a/include/linux/string.h +++ b/include/linux/string.h @@ -253,23 +253,8 @@ static inline const char *kbasename(const char *path) #include #endif -/** - * memcpy_and_pad - Copy one buffer to another with padding - * @dest: Where to copy to - * @dest_len: The destination buffer size - * @src: Where to copy from - * @count: The number of bytes to copy - * @pad: Character to use for padding if space is left in destination. - */ -static inline void memcpy_and_pad(void *dest, size_t dest_len, - const void *src, size_t count, int pad) -{ - if (dest_len > count) { - memcpy(dest, src, count); - memset(dest + count, pad, dest_len - count); - } else - memcpy(dest, src, dest_len); -} +void memcpy_and_pad(void *dest, size_t dest_len, const void *src, size_t count, + int pad); /** * memset_after - Set a value after a struct member to the end of a struct diff --git a/lib/string_helpers.c b/lib/string_helpers.c index faa9d8e4e2c5..d5d008f5b1d9 100644 --- a/lib/string_helpers.c +++ b/lib/string_helpers.c @@ -883,6 +883,26 @@ char *strreplace(char *s, char old, char new) } EXPORT_SYMBOL(strreplace); +/** + * memcpy_and_pad - Copy one buffer to another with padding + * @dest: Where to copy to + * @dest_len: The destination buffer size + * @src: Where to copy from + * @count: The number of bytes to copy + * @pad: Character to use for padding if space is left in destination. + */ +void memcpy_and_pad(void *dest, size_t dest_len, const void *src, size_t count, + int pad) +{ + if (dest_len > count) { + memcpy(dest, src, count); + memset(dest + count, pad, dest_len - count); + } else { + memcpy(dest, src, dest_len); + } +} +EXPORT_SYMBOL(memcpy_and_pad); + #ifdef CONFIG_FORTIFY_SOURCE void fortify_panic(const char *name) { -- cgit v1.2.3 From 7f2b8818ea1361e3482d1e3a3c9a824789177d3a Mon Sep 17 00:00:00 2001 From: Marco Elver Date: Fri, 5 Nov 2021 13:35:36 -0700 Subject: lib/stackdepot: remove unused function argument alloc_flags in depot_alloc_stack() is no longer used; remove it. Link: https://lkml.kernel.org/r/20210913112609.2651084-3-elver@google.com Signed-off-by: Marco Elver Tested-by: Shuah Khan Acked-by: Sebastian Andrzej Siewior Reviewed-by: Andrey Konovalov Cc: Alexander Potapenko Cc: Andrey Ryabinin Cc: Dmitry Vyukov Cc: "Gustavo A. R. Silva" Cc: Lai Jiangshan Cc: Taras Madan Cc: Tejun Heo Cc: Thomas Gleixner Cc: Vijayanand Jitta Cc: Vinayak Menon Cc: Walter Wu Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- lib/stackdepot.c | 9 ++++----- 1 file changed, 4 insertions(+), 5 deletions(-) (limited to 'lib') diff --git a/lib/stackdepot.c b/lib/stackdepot.c index 0a2e417f83cb..c80a9f734253 100644 --- a/lib/stackdepot.c +++ b/lib/stackdepot.c @@ -102,8 +102,8 @@ static bool init_stack_slab(void **prealloc) } /* Allocation of a new stack in raw storage */ -static struct stack_record *depot_alloc_stack(unsigned long *entries, int size, - u32 hash, void **prealloc, gfp_t alloc_flags) +static struct stack_record * +depot_alloc_stack(unsigned long *entries, int size, u32 hash, void **prealloc) { struct stack_record *stack; size_t required_size = struct_size(stack, entries, size); @@ -309,9 +309,8 @@ depot_stack_handle_t stack_depot_save(unsigned long *entries, found = find_stack(*bucket, entries, nr_entries, hash); if (!found) { - struct stack_record *new = - depot_alloc_stack(entries, nr_entries, - hash, &prealloc, alloc_flags); + struct stack_record *new = depot_alloc_stack(entries, nr_entries, hash, &prealloc); + if (new) { new->next = *bucket; /* -- cgit v1.2.3 From 11ac25c62cd2f3bb8da9e1df2e71afdebe76f093 Mon Sep 17 00:00:00 2001 From: Marco Elver Date: Fri, 5 Nov 2021 13:35:39 -0700 Subject: lib/stackdepot: introduce __stack_depot_save() Add __stack_depot_save(), which provides more fine-grained control over stackdepot's memory allocation behaviour, in case stackdepot runs out of "stack slabs". Normally stackdepot uses alloc_pages() in case it runs out of space; passing can_alloc==false to __stack_depot_save() prohibits this, at the cost of more likely failure to record a stack trace. Link: https://lkml.kernel.org/r/20210913112609.2651084-4-elver@google.com Signed-off-by: Marco Elver Tested-by: Shuah Khan Acked-by: Sebastian Andrzej Siewior Reviewed-by: Andrey Konovalov Cc: Alexander Potapenko Cc: Andrey Ryabinin Cc: Dmitry Vyukov Cc: "Gustavo A. R. Silva" Cc: Lai Jiangshan Cc: Taras Madan Cc: Tejun Heo Cc: Thomas Gleixner Cc: Vijayanand Jitta Cc: Vinayak Menon Cc: Walter Wu Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/linux/stackdepot.h | 4 ++++ lib/stackdepot.c | 43 +++++++++++++++++++++++++++++++++++++------ 2 files changed, 41 insertions(+), 6 deletions(-) (limited to 'lib') diff --git a/include/linux/stackdepot.h b/include/linux/stackdepot.h index 97b36dc53301..b2f7e7c6ba54 100644 --- a/include/linux/stackdepot.h +++ b/include/linux/stackdepot.h @@ -15,6 +15,10 @@ typedef u32 depot_stack_handle_t; +depot_stack_handle_t __stack_depot_save(unsigned long *entries, + unsigned int nr_entries, + gfp_t gfp_flags, bool can_alloc); + depot_stack_handle_t stack_depot_save(unsigned long *entries, unsigned int nr_entries, gfp_t gfp_flags); diff --git a/lib/stackdepot.c b/lib/stackdepot.c index c80a9f734253..bda58597e375 100644 --- a/lib/stackdepot.c +++ b/lib/stackdepot.c @@ -248,17 +248,28 @@ unsigned int stack_depot_fetch(depot_stack_handle_t handle, EXPORT_SYMBOL_GPL(stack_depot_fetch); /** - * stack_depot_save - Save a stack trace from an array + * __stack_depot_save - Save a stack trace from an array * * @entries: Pointer to storage array * @nr_entries: Size of the storage array * @alloc_flags: Allocation gfp flags + * @can_alloc: Allocate stack slabs (increased chance of failure if false) + * + * Saves a stack trace from @entries array of size @nr_entries. If @can_alloc is + * %true, is allowed to replenish the stack slab pool in case no space is left + * (allocates using GFP flags of @alloc_flags). If @can_alloc is %false, avoids + * any allocations and will fail if no space is left to store the stack trace. + * + * Context: Any context, but setting @can_alloc to %false is required if + * alloc_pages() cannot be used from the current context. Currently + * this is the case from contexts where neither %GFP_ATOMIC nor + * %GFP_NOWAIT can be used (NMI, raw_spin_lock). * - * Return: The handle of the stack struct stored in depot + * Return: The handle of the stack struct stored in depot, 0 on failure. */ -depot_stack_handle_t stack_depot_save(unsigned long *entries, - unsigned int nr_entries, - gfp_t alloc_flags) +depot_stack_handle_t __stack_depot_save(unsigned long *entries, + unsigned int nr_entries, + gfp_t alloc_flags, bool can_alloc) { struct stack_record *found = NULL, **bucket; depot_stack_handle_t retval = 0; @@ -291,7 +302,7 @@ depot_stack_handle_t stack_depot_save(unsigned long *entries, * The smp_load_acquire() here pairs with smp_store_release() to * |next_slab_inited| in depot_alloc_stack() and init_stack_slab(). */ - if (unlikely(!smp_load_acquire(&next_slab_inited))) { + if (unlikely(can_alloc && !smp_load_acquire(&next_slab_inited))) { /* * Zero out zone modifiers, as we don't have specific zone * requirements. Keep the flags related to allocation in atomic @@ -339,6 +350,26 @@ exit: fast_exit: return retval; } +EXPORT_SYMBOL_GPL(__stack_depot_save); + +/** + * stack_depot_save - Save a stack trace from an array + * + * @entries: Pointer to storage array + * @nr_entries: Size of the storage array + * @alloc_flags: Allocation gfp flags + * + * Context: Contexts where allocations via alloc_pages() are allowed. + * See __stack_depot_save() for more details. + * + * Return: The handle of the stack struct stored in depot, 0 on failure. + */ +depot_stack_handle_t stack_depot_save(unsigned long *entries, + unsigned int nr_entries, + gfp_t alloc_flags) +{ + return __stack_depot_save(entries, nr_entries, alloc_flags, true); +} EXPORT_SYMBOL_GPL(stack_depot_save); static inline int in_irqentry_text(unsigned long ptr) -- cgit v1.2.3 From 758cabae312d3aded781aacc6d0c946b299c52df Mon Sep 17 00:00:00 2001 From: Peter Collingbourne Date: Fri, 5 Nov 2021 13:35:56 -0700 Subject: kasan: test: add memcpy test that avoids out-of-bounds write With HW tag-based KASAN, error checks are performed implicitly by the load and store instructions in the memcpy implementation. A failed check results in tag checks being disabled and execution will keep going. As a result, under HW tag-based KASAN, prior to commit 1b0668be62cf ("kasan: test: disable kmalloc_memmove_invalid_size for HW_TAGS"), this memcpy would end up corrupting memory until it hits an inaccessible page and causes a kernel panic. This is a pre-existing issue that was revealed by commit 285133040e6c ("arm64: Import latest memcpy()/memmove() implementation") which changed the memcpy implementation from using signed comparisons (incorrectly, resulting in the memcpy being terminated early for negative sizes) to using unsigned comparisons. It is unclear how this could be handled by memcpy itself in a reasonable way. One possibility would be to add an exception handler that would force memcpy to return if a tag check fault is detected -- this would make the behavior roughly similar to generic and SW tag-based KASAN. However, this wouldn't solve the problem for asynchronous mode and also makes memcpy behavior inconsistent with manually copying data. This test was added as a part of a series that taught KASAN to detect negative sizes in memory operations, see commit 8cceeff48f23 ("kasan: detect negative size in memory operation function"). Therefore we should keep testing for negative sizes with generic and SW tag-based KASAN. But there is some value in testing small memcpy overflows, so let's add another test with memcpy that does not destabilize the kernel by performing out-of-bounds writes, and run it in all modes. Link: https://linux-review.googlesource.com/id/I048d1e6a9aff766c4a53f989fb0c83de68923882 Link: https://lkml.kernel.org/r/20210910211356.3603758-1-pcc@google.com Signed-off-by: Peter Collingbourne Reviewed-by: Andrey Konovalov Acked-by: Marco Elver Cc: Robin Murphy Cc: Will Deacon Cc: Catalin Marinas Cc: Mark Rutland Cc: Evgenii Stepanov Cc: Alexander Potapenko Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- lib/test_kasan.c | 18 +++++++++++++++++- 1 file changed, 17 insertions(+), 1 deletion(-) (limited to 'lib') diff --git a/lib/test_kasan.c b/lib/test_kasan.c index 8835e0784578..aa8e42250219 100644 --- a/lib/test_kasan.c +++ b/lib/test_kasan.c @@ -493,7 +493,7 @@ static void kmalloc_oob_in_memset(struct kunit *test) kfree(ptr); } -static void kmalloc_memmove_invalid_size(struct kunit *test) +static void kmalloc_memmove_negative_size(struct kunit *test) { char *ptr; size_t size = 64; @@ -515,6 +515,21 @@ static void kmalloc_memmove_invalid_size(struct kunit *test) kfree(ptr); } +static void kmalloc_memmove_invalid_size(struct kunit *test) +{ + char *ptr; + size_t size = 64; + volatile size_t invalid_size = size; + + ptr = kmalloc(size, GFP_KERNEL); + KUNIT_ASSERT_NOT_ERR_OR_NULL(test, ptr); + + memset((char *)ptr, 0, 64); + KUNIT_EXPECT_KASAN_FAIL(test, + memmove((char *)ptr, (char *)ptr + 4, invalid_size)); + kfree(ptr); +} + static void kmalloc_uaf(struct kunit *test) { char *ptr; @@ -1129,6 +1144,7 @@ static struct kunit_case kasan_kunit_test_cases[] = { KUNIT_CASE(kmalloc_oob_memset_4), KUNIT_CASE(kmalloc_oob_memset_8), KUNIT_CASE(kmalloc_oob_memset_16), + KUNIT_CASE(kmalloc_memmove_negative_size), KUNIT_CASE(kmalloc_memmove_invalid_size), KUNIT_CASE(kmalloc_uaf), KUNIT_CASE(kmalloc_uaf_memset), -- cgit v1.2.3 From d73dad4eb5ad8c31ac9cf358eb5a55825bafe706 Mon Sep 17 00:00:00 2001 From: Kees Cook Date: Fri, 5 Nov 2021 13:36:12 -0700 Subject: kasan: test: bypass __alloc_size checks Intentional overflows, as performed by the KASAN tests, are detected at compile time[1] (instead of only at run-time) with the addition of __alloc_size. Fix this by forcing the compiler into not being able to trust the size used following the kmalloc()s. [1] https://lore.kernel.org/lkml/20211005184717.65c6d8eb39350395e387b71f@linux-foundation.org Link: https://lkml.kernel.org/r/20211006181544.1670992-1-keescook@chromium.org Signed-off-by: Kees Cook Cc: Andrey Ryabinin Cc: Alexander Potapenko Cc: Andrey Konovalov Cc: Dmitry Vyukov Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- lib/test_kasan.c | 8 +++++++- lib/test_kasan_module.c | 2 ++ 2 files changed, 9 insertions(+), 1 deletion(-) (limited to 'lib') diff --git a/lib/test_kasan.c b/lib/test_kasan.c index aa8e42250219..5475fe396ff7 100644 --- a/lib/test_kasan.c +++ b/lib/test_kasan.c @@ -440,6 +440,7 @@ static void kmalloc_oob_memset_2(struct kunit *test) ptr = kmalloc(size, GFP_KERNEL); KUNIT_ASSERT_NOT_ERR_OR_NULL(test, ptr); + OPTIMIZER_HIDE_VAR(size); KUNIT_EXPECT_KASAN_FAIL(test, memset(ptr + size - 1, 0, 2)); kfree(ptr); } @@ -452,6 +453,7 @@ static void kmalloc_oob_memset_4(struct kunit *test) ptr = kmalloc(size, GFP_KERNEL); KUNIT_ASSERT_NOT_ERR_OR_NULL(test, ptr); + OPTIMIZER_HIDE_VAR(size); KUNIT_EXPECT_KASAN_FAIL(test, memset(ptr + size - 3, 0, 4)); kfree(ptr); } @@ -464,6 +466,7 @@ static void kmalloc_oob_memset_8(struct kunit *test) ptr = kmalloc(size, GFP_KERNEL); KUNIT_ASSERT_NOT_ERR_OR_NULL(test, ptr); + OPTIMIZER_HIDE_VAR(size); KUNIT_EXPECT_KASAN_FAIL(test, memset(ptr + size - 7, 0, 8)); kfree(ptr); } @@ -476,6 +479,7 @@ static void kmalloc_oob_memset_16(struct kunit *test) ptr = kmalloc(size, GFP_KERNEL); KUNIT_ASSERT_NOT_ERR_OR_NULL(test, ptr); + OPTIMIZER_HIDE_VAR(size); KUNIT_EXPECT_KASAN_FAIL(test, memset(ptr + size - 15, 0, 16)); kfree(ptr); } @@ -488,6 +492,7 @@ static void kmalloc_oob_in_memset(struct kunit *test) ptr = kmalloc(size, GFP_KERNEL); KUNIT_ASSERT_NOT_ERR_OR_NULL(test, ptr); + OPTIMIZER_HIDE_VAR(size); KUNIT_EXPECT_KASAN_FAIL(test, memset(ptr, 0, size + KASAN_GRANULE_SIZE)); kfree(ptr); @@ -497,7 +502,7 @@ static void kmalloc_memmove_negative_size(struct kunit *test) { char *ptr; size_t size = 64; - volatile size_t invalid_size = -2; + size_t invalid_size = -2; /* * Hardware tag-based mode doesn't check memmove for negative size. @@ -510,6 +515,7 @@ static void kmalloc_memmove_negative_size(struct kunit *test) KUNIT_ASSERT_NOT_ERR_OR_NULL(test, ptr); memset((char *)ptr, 0, 64); + OPTIMIZER_HIDE_VAR(invalid_size); KUNIT_EXPECT_KASAN_FAIL(test, memmove((char *)ptr, (char *)ptr + 4, invalid_size)); kfree(ptr); diff --git a/lib/test_kasan_module.c b/lib/test_kasan_module.c index 7ebf433edef3..b112cbc835e9 100644 --- a/lib/test_kasan_module.c +++ b/lib/test_kasan_module.c @@ -35,6 +35,8 @@ static noinline void __init copy_user_test(void) return; } + OPTIMIZER_HIDE_VAR(size); + pr_info("out-of-bounds in copy_from_user()\n"); unused = copy_from_user(kmem, usermem, size + 1); -- cgit v1.2.3 From 34b46efd6ec6a6f2d6a57be4216b820c879a0030 Mon Sep 17 00:00:00 2001 From: Changcheng Deng Date: Fri, 5 Nov 2021 13:39:56 -0700 Subject: lib/test_vmalloc.c: use swap() to make code cleaner Use swap() in order to make code cleaner. Issue found by coccinelle. Link: https://lkml.kernel.org/r/20211028111443.15744-1-deng.changcheng@zte.com.cn Signed-off-by: Changcheng Deng Reported-by: Zeal Robot Reviewed-by: Uladzislau Rezki (Sony) Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- lib/test_vmalloc.c | 6 ++---- 1 file changed, 2 insertions(+), 4 deletions(-) (limited to 'lib') diff --git a/lib/test_vmalloc.c b/lib/test_vmalloc.c index e14993bc84d2..cf41fd6df42a 100644 --- a/lib/test_vmalloc.c +++ b/lib/test_vmalloc.c @@ -393,7 +393,7 @@ static struct test_driver { static void shuffle_array(int *arr, int n) { unsigned int rnd; - int i, j, x; + int i, j; for (i = n - 1; i > 0; i--) { get_random_bytes(&rnd, sizeof(rnd)); @@ -402,9 +402,7 @@ static void shuffle_array(int *arr, int n) j = rnd % i; /* Swap indexes. */ - x = arr[i]; - arr[i] = arr[j]; - arr[j] = x; + swap(arr[i], arr[j]); } } -- cgit v1.2.3 From fa27717110ae51b9b9013ced0b5143888257bb79 Mon Sep 17 00:00:00 2001 From: Mike Rapoport Date: Fri, 5 Nov 2021 13:43:13 -0700 Subject: memblock: drop memblock_free_early_nid() and memblock_free_early() memblock_free_early_nid() is unused and memblock_free_early() is an alias for memblock_free(). Replace calls to memblock_free_early() with calls to memblock_free() and remove memblock_free_early() and memblock_free_early_nid(). Link: https://lkml.kernel.org/r/20210930185031.18648-4-rppt@kernel.org Signed-off-by: Mike Rapoport Cc: Christophe Leroy Cc: Juergen Gross Cc: Shahab Vahedi Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/mips/mm/init.c | 2 +- arch/powerpc/platforms/pseries/svm.c | 3 +-- arch/s390/kernel/smp.c | 2 +- drivers/base/arch_numa.c | 2 +- drivers/s390/char/sclp_early.c | 2 +- include/linux/memblock.h | 12 ------------ kernel/dma/swiotlb.c | 2 +- lib/cpumask.c | 2 +- mm/percpu.c | 8 ++++---- mm/sparse.c | 2 +- 10 files changed, 12 insertions(+), 25 deletions(-) (limited to 'lib') diff --git a/arch/mips/mm/init.c b/arch/mips/mm/init.c index 19347dc6bbf8..21a5a7ac0037 100644 --- a/arch/mips/mm/init.c +++ b/arch/mips/mm/init.c @@ -529,7 +529,7 @@ static void * __init pcpu_fc_alloc(unsigned int cpu, size_t size, static void __init pcpu_fc_free(void *ptr, size_t size) { - memblock_free_early(__pa(ptr), size); + memblock_free(__pa(ptr), size); } void __init setup_per_cpu_areas(void) diff --git a/arch/powerpc/platforms/pseries/svm.c b/arch/powerpc/platforms/pseries/svm.c index 87f001b4c4e4..f12229ce7301 100644 --- a/arch/powerpc/platforms/pseries/svm.c +++ b/arch/powerpc/platforms/pseries/svm.c @@ -56,8 +56,7 @@ void __init svm_swiotlb_init(void) return; - memblock_free_early(__pa(vstart), - PAGE_ALIGN(io_tlb_nslabs << IO_TLB_SHIFT)); + memblock_free(__pa(vstart), PAGE_ALIGN(io_tlb_nslabs << IO_TLB_SHIFT)); panic("SVM: Cannot allocate SWIOTLB buffer"); } diff --git a/arch/s390/kernel/smp.c b/arch/s390/kernel/smp.c index 1a04e5bdf655..066efd6d9345 100644 --- a/arch/s390/kernel/smp.c +++ b/arch/s390/kernel/smp.c @@ -880,7 +880,7 @@ void __init smp_detect_cpus(void) /* Add CPUs present at boot */ __smp_rescan_cpus(info, true); - memblock_free_early((unsigned long)info, sizeof(*info)); + memblock_free((unsigned long)info, sizeof(*info)); } /* diff --git a/drivers/base/arch_numa.c b/drivers/base/arch_numa.c index a6491673dc1f..ade8934764f6 100644 --- a/drivers/base/arch_numa.c +++ b/drivers/base/arch_numa.c @@ -166,7 +166,7 @@ static void * __init pcpu_fc_alloc(unsigned int cpu, size_t size, static void __init pcpu_fc_free(void *ptr, size_t size) { - memblock_free_early(__pa(ptr), size); + memblock_free(__pa(ptr), size); } #ifdef CONFIG_NEED_PER_CPU_PAGE_FIRST_CHUNK diff --git a/drivers/s390/char/sclp_early.c b/drivers/s390/char/sclp_early.c index f3d5c7f4c13d..f01d942e1c1d 100644 --- a/drivers/s390/char/sclp_early.c +++ b/drivers/s390/char/sclp_early.c @@ -139,7 +139,7 @@ int __init sclp_early_get_core_info(struct sclp_core_info *info) } sclp_fill_core_info(info, sccb); out: - memblock_free_early((unsigned long)sccb, length); + memblock_free((unsigned long)sccb, length); return rc; } diff --git a/include/linux/memblock.h b/include/linux/memblock.h index 34de69b3b8ba..fc8183be340c 100644 --- a/include/linux/memblock.h +++ b/include/linux/memblock.h @@ -441,18 +441,6 @@ static inline void *memblock_alloc_node(phys_addr_t size, MEMBLOCK_ALLOC_ACCESSIBLE, nid); } -static inline void memblock_free_early(phys_addr_t base, - phys_addr_t size) -{ - memblock_free(base, size); -} - -static inline void memblock_free_early_nid(phys_addr_t base, - phys_addr_t size, int nid) -{ - memblock_free(base, size); -} - static inline void memblock_free_late(phys_addr_t base, phys_addr_t size) { __memblock_free_late(base, size); diff --git a/kernel/dma/swiotlb.c b/kernel/dma/swiotlb.c index 87c40517e822..430d2f78d540 100644 --- a/kernel/dma/swiotlb.c +++ b/kernel/dma/swiotlb.c @@ -247,7 +247,7 @@ swiotlb_init(int verbose) return; fail_free_mem: - memblock_free_early(__pa(tlb), bytes); + memblock_free(__pa(tlb), bytes); fail: pr_warn("Cannot allocate buffer"); } diff --git a/lib/cpumask.c b/lib/cpumask.c index c3c76b833384..045779446a18 100644 --- a/lib/cpumask.c +++ b/lib/cpumask.c @@ -188,7 +188,7 @@ EXPORT_SYMBOL(free_cpumask_var); */ void __init free_bootmem_cpumask_var(cpumask_var_t mask) { - memblock_free_early(__pa(mask), cpumask_size()); + memblock_free(__pa(mask), cpumask_size()); } #endif diff --git a/mm/percpu.c b/mm/percpu.c index e0a986818903..f58318cb04c0 100644 --- a/mm/percpu.c +++ b/mm/percpu.c @@ -2472,7 +2472,7 @@ struct pcpu_alloc_info * __init pcpu_alloc_alloc_info(int nr_groups, */ void __init pcpu_free_alloc_info(struct pcpu_alloc_info *ai) { - memblock_free_early(__pa(ai), ai->__ai_size); + memblock_free(__pa(ai), ai->__ai_size); } /** @@ -3134,7 +3134,7 @@ out_free_areas: out_free: pcpu_free_alloc_info(ai); if (areas) - memblock_free_early(__pa(areas), areas_size); + memblock_free(__pa(areas), areas_size); return rc; } #endif /* BUILD_EMBED_FIRST_CHUNK */ @@ -3256,7 +3256,7 @@ enomem: free_fn(page_address(pages[j]), PAGE_SIZE); rc = -ENOMEM; out_free_ar: - memblock_free_early(__pa(pages), pages_size); + memblock_free(__pa(pages), pages_size); pcpu_free_alloc_info(ai); return rc; } @@ -3286,7 +3286,7 @@ static void * __init pcpu_dfl_fc_alloc(unsigned int cpu, size_t size, static void __init pcpu_dfl_fc_free(void *ptr, size_t size) { - memblock_free_early(__pa(ptr), size); + memblock_free(__pa(ptr), size); } void __init setup_per_cpu_areas(void) diff --git a/mm/sparse.c b/mm/sparse.c index 120bc8ea5293..55fea0c2f927 100644 --- a/mm/sparse.c +++ b/mm/sparse.c @@ -451,7 +451,7 @@ static void *sparsemap_buf_end __meminitdata; static inline void __meminit sparse_buffer_free(unsigned long size) { WARN_ON(!sparsemap_buf || size == 0); - memblock_free_early(__pa(sparsemap_buf), size); + memblock_free(__pa(sparsemap_buf), size); } static void __init sparse_buffer_init(unsigned long size, int nid) -- cgit v1.2.3 From 3ecc68349bbab6bff1d12cbc7951ca6019b2faf6 Mon Sep 17 00:00:00 2001 From: Mike Rapoport Date: Fri, 5 Nov 2021 13:43:19 -0700 Subject: memblock: rename memblock_free to memblock_phys_free Since memblock_free() operates on a physical range, make its name reflect it and rename it to memblock_phys_free(), so it will be a logical counterpart to memblock_phys_alloc(). The callers are updated with the below semantic patch: @@ expression addr; expression size; @@ - memblock_free(addr, size); + memblock_phys_free(addr, size); Link: https://lkml.kernel.org/r/20210930185031.18648-6-rppt@kernel.org Signed-off-by: Mike Rapoport Cc: Christophe Leroy Cc: Juergen Gross Cc: Shahab Vahedi Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/alpha/kernel/core_irongate.c | 3 ++- arch/arc/mm/init.c | 2 +- arch/arm/mach-hisi/platmcpm.c | 2 +- arch/arm/mm/init.c | 2 +- arch/arm64/mm/mmu.c | 4 ++-- arch/mips/mm/init.c | 2 +- arch/mips/sgi-ip30/ip30-setup.c | 6 +++--- arch/powerpc/kernel/dt_cpu_ftrs.c | 4 ++-- arch/powerpc/kernel/paca.c | 8 ++++---- arch/powerpc/kernel/setup-common.c | 2 +- arch/powerpc/kernel/setup_64.c | 2 +- arch/powerpc/platforms/powernv/pci-ioda.c | 2 +- arch/powerpc/platforms/pseries/svm.c | 3 ++- arch/riscv/kernel/setup.c | 5 +++-- arch/s390/kernel/setup.c | 8 ++++---- arch/s390/kernel/smp.c | 4 ++-- arch/s390/kernel/uv.c | 2 +- arch/s390/mm/kasan_init.c | 2 +- arch/sh/boards/mach-ap325rxa/setup.c | 2 +- arch/sh/boards/mach-ecovec24/setup.c | 4 ++-- arch/sh/boards/mach-kfr2r09/setup.c | 2 +- arch/sh/boards/mach-migor/setup.c | 2 +- arch/sh/boards/mach-se/7724/setup.c | 4 ++-- arch/sparc/kernel/smp_64.c | 2 +- arch/um/kernel/mem.c | 2 +- arch/x86/kernel/setup.c | 4 ++-- arch/x86/mm/init.c | 2 +- arch/x86/xen/mmu_pv.c | 6 +++--- arch/x86/xen/setup.c | 6 +++--- drivers/base/arch_numa.c | 2 +- drivers/firmware/efi/memmap.c | 2 +- drivers/of/kexec.c | 3 +-- drivers/of/of_reserved_mem.c | 5 +++-- drivers/s390/char/sclp_early.c | 2 +- drivers/usb/early/xhci-dbc.c | 10 +++++----- drivers/xen/swiotlb-xen.c | 2 +- include/linux/memblock.h | 2 +- init/initramfs.c | 2 +- kernel/dma/swiotlb.c | 2 +- lib/cpumask.c | 2 +- mm/cma.c | 2 +- mm/memblock.c | 8 ++++---- mm/memory_hotplug.c | 2 +- mm/percpu.c | 8 ++++---- mm/sparse.c | 2 +- 45 files changed, 79 insertions(+), 76 deletions(-) (limited to 'lib') diff --git a/arch/alpha/kernel/core_irongate.c b/arch/alpha/kernel/core_irongate.c index 72af1e72d833..ee26dcc49418 100644 --- a/arch/alpha/kernel/core_irongate.c +++ b/arch/alpha/kernel/core_irongate.c @@ -233,7 +233,8 @@ albacore_init_arch(void) unsigned long size; size = initrd_end - initrd_start; - memblock_free(__pa(initrd_start), PAGE_ALIGN(size)); + memblock_phys_free(__pa(initrd_start), + PAGE_ALIGN(size)); if (!move_initrd(pci_mem)) printk("irongate_init_arch: initrd too big " "(%ldK)\ndisabling initrd\n", diff --git a/arch/arc/mm/init.c b/arch/arc/mm/init.c index 699ecf119641..59408f6a02d4 100644 --- a/arch/arc/mm/init.c +++ b/arch/arc/mm/init.c @@ -173,7 +173,7 @@ static void __init highmem_init(void) #ifdef CONFIG_HIGHMEM unsigned long tmp; - memblock_free(high_mem_start, high_mem_sz); + memblock_phys_free(high_mem_start, high_mem_sz); for (tmp = min_high_pfn; tmp < max_high_pfn; tmp++) free_highmem_page(pfn_to_page(tmp)); #endif diff --git a/arch/arm/mach-hisi/platmcpm.c b/arch/arm/mach-hisi/platmcpm.c index 96a484095194..258586e31333 100644 --- a/arch/arm/mach-hisi/platmcpm.c +++ b/arch/arm/mach-hisi/platmcpm.c @@ -339,7 +339,7 @@ err_fabric: err_sysctrl: iounmap(relocation); err_reloc: - memblock_free(hip04_boot_method[0], hip04_boot_method[1]); + memblock_phys_free(hip04_boot_method[0], hip04_boot_method[1]); err: return ret; } diff --git a/arch/arm/mm/init.c b/arch/arm/mm/init.c index 6162a070a410..6d0cb0f7bc54 100644 --- a/arch/arm/mm/init.c +++ b/arch/arm/mm/init.c @@ -158,7 +158,7 @@ phys_addr_t __init arm_memblock_steal(phys_addr_t size, phys_addr_t align) panic("Failed to steal %pa bytes at %pS\n", &size, (void *)_RET_IP_); - memblock_free(phys, size); + memblock_phys_free(phys, size); memblock_remove(phys, size); return phys; diff --git a/arch/arm64/mm/mmu.c b/arch/arm64/mm/mmu.c index cfd9deb347c3..f68c2d953617 100644 --- a/arch/arm64/mm/mmu.c +++ b/arch/arm64/mm/mmu.c @@ -738,8 +738,8 @@ void __init paging_init(void) cpu_replace_ttbr1(lm_alias(swapper_pg_dir)); init_mm.pgd = swapper_pg_dir; - memblock_free(__pa_symbol(init_pg_dir), - __pa_symbol(init_pg_end) - __pa_symbol(init_pg_dir)); + memblock_phys_free(__pa_symbol(init_pg_dir), + __pa_symbol(init_pg_end) - __pa_symbol(init_pg_dir)); memblock_allow_resize(); } diff --git a/arch/mips/mm/init.c b/arch/mips/mm/init.c index 21a5a7ac0037..3be1c29084fa 100644 --- a/arch/mips/mm/init.c +++ b/arch/mips/mm/init.c @@ -529,7 +529,7 @@ static void * __init pcpu_fc_alloc(unsigned int cpu, size_t size, static void __init pcpu_fc_free(void *ptr, size_t size) { - memblock_free(__pa(ptr), size); + memblock_phys_free(__pa(ptr), size); } void __init setup_per_cpu_areas(void) diff --git a/arch/mips/sgi-ip30/ip30-setup.c b/arch/mips/sgi-ip30/ip30-setup.c index 44b1607e964d..75a34684e704 100644 --- a/arch/mips/sgi-ip30/ip30-setup.c +++ b/arch/mips/sgi-ip30/ip30-setup.c @@ -69,10 +69,10 @@ static void __init ip30_mem_init(void) total_mem += size; if (addr >= IP30_REAL_MEMORY_START) - memblock_free(addr, size); + memblock_phys_free(addr, size); else if ((addr + size) > IP30_REAL_MEMORY_START) - memblock_free(IP30_REAL_MEMORY_START, - size - IP30_MAX_PROM_MEMORY); + memblock_phys_free(IP30_REAL_MEMORY_START, + size - IP30_MAX_PROM_MEMORY); } pr_info("Detected %luMB of physical memory.\n", MEM_SHIFT(total_mem)); } diff --git a/arch/powerpc/kernel/dt_cpu_ftrs.c b/arch/powerpc/kernel/dt_cpu_ftrs.c index 358aee7c2d79..42839d6bd486 100644 --- a/arch/powerpc/kernel/dt_cpu_ftrs.c +++ b/arch/powerpc/kernel/dt_cpu_ftrs.c @@ -1095,8 +1095,8 @@ static int __init dt_cpu_ftrs_scan_callback(unsigned long node, const char cpufeatures_setup_finished(); - memblock_free(__pa(dt_cpu_features), - sizeof(struct dt_cpu_feature)*nr_dt_cpu_features); + memblock_phys_free(__pa(dt_cpu_features), + sizeof(struct dt_cpu_feature) * nr_dt_cpu_features); return 0; } diff --git a/arch/powerpc/kernel/paca.c b/arch/powerpc/kernel/paca.c index 9bd30cac852b..4208b4044d12 100644 --- a/arch/powerpc/kernel/paca.c +++ b/arch/powerpc/kernel/paca.c @@ -322,8 +322,8 @@ void __init free_unused_pacas(void) new_ptrs_size = sizeof(struct paca_struct *) * nr_cpu_ids; if (new_ptrs_size < paca_ptrs_size) - memblock_free(__pa(paca_ptrs) + new_ptrs_size, - paca_ptrs_size - new_ptrs_size); + memblock_phys_free(__pa(paca_ptrs) + new_ptrs_size, + paca_ptrs_size - new_ptrs_size); paca_nr_cpu_ids = nr_cpu_ids; paca_ptrs_size = new_ptrs_size; @@ -331,8 +331,8 @@ void __init free_unused_pacas(void) #ifdef CONFIG_PPC_BOOK3S_64 if (early_radix_enabled()) { /* Ugly fixup, see new_slb_shadow() */ - memblock_free(__pa(paca_ptrs[boot_cpuid]->slb_shadow_ptr), - sizeof(struct slb_shadow)); + memblock_phys_free(__pa(paca_ptrs[boot_cpuid]->slb_shadow_ptr), + sizeof(struct slb_shadow)); paca_ptrs[boot_cpuid]->slb_shadow_ptr = NULL; } #endif diff --git a/arch/powerpc/kernel/setup-common.c b/arch/powerpc/kernel/setup-common.c index b1e43b69a559..5af8993a8e6d 100644 --- a/arch/powerpc/kernel/setup-common.c +++ b/arch/powerpc/kernel/setup-common.c @@ -825,7 +825,7 @@ static void __init smp_setup_pacas(void) set_hard_smp_processor_id(cpu, cpu_to_phys_id[cpu]); } - memblock_free(__pa(cpu_to_phys_id), nr_cpu_ids * sizeof(u32)); + memblock_phys_free(__pa(cpu_to_phys_id), nr_cpu_ids * sizeof(u32)); cpu_to_phys_id = NULL; } #endif diff --git a/arch/powerpc/kernel/setup_64.c b/arch/powerpc/kernel/setup_64.c index eaa79a0996d1..75bc294ac40d 100644 --- a/arch/powerpc/kernel/setup_64.c +++ b/arch/powerpc/kernel/setup_64.c @@ -812,7 +812,7 @@ static void * __init pcpu_alloc_bootmem(unsigned int cpu, size_t size, static void __init pcpu_free_bootmem(void *ptr, size_t size) { - memblock_free(__pa(ptr), size); + memblock_phys_free(__pa(ptr), size); } static int pcpu_cpu_distance(unsigned int from, unsigned int to) diff --git a/arch/powerpc/platforms/powernv/pci-ioda.c b/arch/powerpc/platforms/powernv/pci-ioda.c index 3dd35c327d1c..b5a9d343b720 100644 --- a/arch/powerpc/platforms/powernv/pci-ioda.c +++ b/arch/powerpc/platforms/powernv/pci-ioda.c @@ -2981,7 +2981,7 @@ static void __init pnv_pci_init_ioda_phb(struct device_node *np, if (!phb->hose) { pr_err(" Can't allocate PCI controller for %pOF\n", np); - memblock_free(__pa(phb), sizeof(struct pnv_phb)); + memblock_phys_free(__pa(phb), sizeof(struct pnv_phb)); return; } diff --git a/arch/powerpc/platforms/pseries/svm.c b/arch/powerpc/platforms/pseries/svm.c index f12229ce7301..b7c017bb40f7 100644 --- a/arch/powerpc/platforms/pseries/svm.c +++ b/arch/powerpc/platforms/pseries/svm.c @@ -56,7 +56,8 @@ void __init svm_swiotlb_init(void) return; - memblock_free(__pa(vstart), PAGE_ALIGN(io_tlb_nslabs << IO_TLB_SHIFT)); + memblock_phys_free(__pa(vstart), + PAGE_ALIGN(io_tlb_nslabs << IO_TLB_SHIFT)); panic("SVM: Cannot allocate SWIOTLB buffer"); } diff --git a/arch/riscv/kernel/setup.c b/arch/riscv/kernel/setup.c index b9620e5f00ba..6ea7c53b82cd 100644 --- a/arch/riscv/kernel/setup.c +++ b/arch/riscv/kernel/setup.c @@ -230,13 +230,14 @@ static void __init init_resources(void) /* Clean-up any unused pre-allocated resources */ if (res_idx >= 0) - memblock_free(__pa(mem_res), (res_idx + 1) * sizeof(*mem_res)); + memblock_phys_free(__pa(mem_res), + (res_idx + 1) * sizeof(*mem_res)); return; error: /* Better an empty resource tree than an inconsistent one */ release_child_resources(&iomem_resource); - memblock_free(__pa(mem_res), mem_res_sz); + memblock_phys_free(__pa(mem_res), mem_res_sz); } diff --git a/arch/s390/kernel/setup.c b/arch/s390/kernel/setup.c index 67e5fff96ee0..7fc836e9e194 100644 --- a/arch/s390/kernel/setup.c +++ b/arch/s390/kernel/setup.c @@ -693,7 +693,7 @@ static void __init reserve_crashkernel(void) } if (register_memory_notifier(&kdump_mem_nb)) { - memblock_free(crash_base, crash_size); + memblock_phys_free(crash_base, crash_size); return; } @@ -748,7 +748,7 @@ static void __init free_mem_detect_info(void) get_mem_detect_reserved(&start, &size); if (size) - memblock_free(start, size); + memblock_phys_free(start, size); } static const char * __init get_mem_info_source(void) @@ -793,7 +793,7 @@ static void __init check_initrd(void) if (initrd_data.start && initrd_data.size && !memblock_is_region_memory(initrd_data.start, initrd_data.size)) { pr_err("The initial RAM disk does not fit into the memory\n"); - memblock_free(initrd_data.start, initrd_data.size); + memblock_phys_free(initrd_data.start, initrd_data.size); initrd_start = initrd_end = 0; } #endif @@ -890,7 +890,7 @@ static void __init setup_randomness(void) if (stsi(vmms, 3, 2, 2) == 0 && vmms->count) add_device_randomness(&vmms->vm, sizeof(vmms->vm[0]) * vmms->count); - memblock_free((unsigned long) vmms, PAGE_SIZE); + memblock_phys_free((unsigned long)vmms, PAGE_SIZE); } /* diff --git a/arch/s390/kernel/smp.c b/arch/s390/kernel/smp.c index 066efd6d9345..78a8ea6fd582 100644 --- a/arch/s390/kernel/smp.c +++ b/arch/s390/kernel/smp.c @@ -723,7 +723,7 @@ void __init smp_save_dump_cpus(void) /* Get the CPU registers */ smp_save_cpu_regs(sa, addr, is_boot_cpu, page); } - memblock_free(page, PAGE_SIZE); + memblock_phys_free(page, PAGE_SIZE); diag_amode31_ops.diag308_reset(); pcpu_set_smt(0); } @@ -880,7 +880,7 @@ void __init smp_detect_cpus(void) /* Add CPUs present at boot */ __smp_rescan_cpus(info, true); - memblock_free((unsigned long)info, sizeof(*info)); + memblock_phys_free((unsigned long)info, sizeof(*info)); } /* diff --git a/arch/s390/kernel/uv.c b/arch/s390/kernel/uv.c index 5a656c7b7a67..d57457b16fe5 100644 --- a/arch/s390/kernel/uv.c +++ b/arch/s390/kernel/uv.c @@ -64,7 +64,7 @@ void __init setup_uv(void) } if (uv_init(uv_stor_base, uv_info.uv_base_stor_len)) { - memblock_free(uv_stor_base, uv_info.uv_base_stor_len); + memblock_phys_free(uv_stor_base, uv_info.uv_base_stor_len); goto fail; } diff --git a/arch/s390/mm/kasan_init.c b/arch/s390/mm/kasan_init.c index 3e4735168019..483b9dbe0970 100644 --- a/arch/s390/mm/kasan_init.c +++ b/arch/s390/mm/kasan_init.c @@ -399,5 +399,5 @@ void __init kasan_copy_shadow_mapping(void) void __init kasan_free_early_identity(void) { - memblock_free(pgalloc_pos, pgalloc_freeable - pgalloc_pos); + memblock_phys_free(pgalloc_pos, pgalloc_freeable - pgalloc_pos); } diff --git a/arch/sh/boards/mach-ap325rxa/setup.c b/arch/sh/boards/mach-ap325rxa/setup.c index bac8a058ebd7..c77b5f00a66a 100644 --- a/arch/sh/boards/mach-ap325rxa/setup.c +++ b/arch/sh/boards/mach-ap325rxa/setup.c @@ -560,7 +560,7 @@ static void __init ap325rxa_mv_mem_reserve(void) if (!phys) panic("Failed to allocate CEU memory\n"); - memblock_free(phys, size); + memblock_phys_free(phys, size); memblock_remove(phys, size); ceu_dma_membase = phys; diff --git a/arch/sh/boards/mach-ecovec24/setup.c b/arch/sh/boards/mach-ecovec24/setup.c index bab91a99124e..2b22ce792147 100644 --- a/arch/sh/boards/mach-ecovec24/setup.c +++ b/arch/sh/boards/mach-ecovec24/setup.c @@ -1502,7 +1502,7 @@ static void __init ecovec_mv_mem_reserve(void) if (!phys) panic("Failed to allocate CEU0 memory\n"); - memblock_free(phys, size); + memblock_phys_free(phys, size); memblock_remove(phys, size); ceu0_dma_membase = phys; @@ -1510,7 +1510,7 @@ static void __init ecovec_mv_mem_reserve(void) if (!phys) panic("Failed to allocate CEU1 memory\n"); - memblock_free(phys, size); + memblock_phys_free(phys, size); memblock_remove(phys, size); ceu1_dma_membase = phys; } diff --git a/arch/sh/boards/mach-kfr2r09/setup.c b/arch/sh/boards/mach-kfr2r09/setup.c index eeb5ce341efd..20f4db778ed6 100644 --- a/arch/sh/boards/mach-kfr2r09/setup.c +++ b/arch/sh/boards/mach-kfr2r09/setup.c @@ -633,7 +633,7 @@ static void __init kfr2r09_mv_mem_reserve(void) if (!phys) panic("Failed to allocate CEU memory\n"); - memblock_free(phys, size); + memblock_phys_free(phys, size); memblock_remove(phys, size); ceu_dma_membase = phys; diff --git a/arch/sh/boards/mach-migor/setup.c b/arch/sh/boards/mach-migor/setup.c index 6703a2122c0d..f60061283c48 100644 --- a/arch/sh/boards/mach-migor/setup.c +++ b/arch/sh/boards/mach-migor/setup.c @@ -633,7 +633,7 @@ static void __init migor_mv_mem_reserve(void) if (!phys) panic("Failed to allocate CEU memory\n"); - memblock_free(phys, size); + memblock_phys_free(phys, size); memblock_remove(phys, size); ceu_dma_membase = phys; diff --git a/arch/sh/boards/mach-se/7724/setup.c b/arch/sh/boards/mach-se/7724/setup.c index 8d6541ba0186..8bbf5a6aa423 100644 --- a/arch/sh/boards/mach-se/7724/setup.c +++ b/arch/sh/boards/mach-se/7724/setup.c @@ -966,7 +966,7 @@ static void __init ms7724se_mv_mem_reserve(void) if (!phys) panic("Failed to allocate CEU0 memory\n"); - memblock_free(phys, size); + memblock_phys_free(phys, size); memblock_remove(phys, size); ceu0_dma_membase = phys; @@ -974,7 +974,7 @@ static void __init ms7724se_mv_mem_reserve(void) if (!phys) panic("Failed to allocate CEU1 memory\n"); - memblock_free(phys, size); + memblock_phys_free(phys, size); memblock_remove(phys, size); ceu1_dma_membase = phys; } diff --git a/arch/sparc/kernel/smp_64.c b/arch/sparc/kernel/smp_64.c index 0224d8f19ed6..2507549538df 100644 --- a/arch/sparc/kernel/smp_64.c +++ b/arch/sparc/kernel/smp_64.c @@ -1567,7 +1567,7 @@ static void * __init pcpu_alloc_bootmem(unsigned int cpu, size_t size, static void __init pcpu_free_bootmem(void *ptr, size_t size) { - memblock_free(__pa(ptr), size); + memblock_phys_free(__pa(ptr), size); } static int __init pcpu_cpu_distance(unsigned int from, unsigned int to) diff --git a/arch/um/kernel/mem.c b/arch/um/kernel/mem.c index 8e636ce02949..d1710ebb44f4 100644 --- a/arch/um/kernel/mem.c +++ b/arch/um/kernel/mem.c @@ -47,7 +47,7 @@ void __init mem_init(void) */ brk_end = (unsigned long) UML_ROUND_UP(sbrk(0)); map_memory(brk_end, __pa(brk_end), uml_reserved - brk_end, 1, 1, 0); - memblock_free(__pa(brk_end), uml_reserved - brk_end); + memblock_phys_free(__pa(brk_end), uml_reserved - brk_end); uml_reserved = brk_end; /* this will put all low memory onto the freelists */ diff --git a/arch/x86/kernel/setup.c b/arch/x86/kernel/setup.c index 40ed44ead063..49b596db5631 100644 --- a/arch/x86/kernel/setup.c +++ b/arch/x86/kernel/setup.c @@ -322,7 +322,7 @@ static void __init reserve_initrd(void) relocate_initrd(); - memblock_free(ramdisk_image, ramdisk_end - ramdisk_image); + memblock_phys_free(ramdisk_image, ramdisk_end - ramdisk_image); } #else @@ -521,7 +521,7 @@ static void __init reserve_crashkernel(void) } if (crash_base >= (1ULL << 32) && reserve_crashkernel_low()) { - memblock_free(crash_base, crash_size); + memblock_phys_free(crash_base, crash_size); return; } diff --git a/arch/x86/mm/init.c b/arch/x86/mm/init.c index 23a14d82e783..1895986842b9 100644 --- a/arch/x86/mm/init.c +++ b/arch/x86/mm/init.c @@ -618,7 +618,7 @@ static void __init memory_map_top_down(unsigned long map_start, */ addr = memblock_phys_alloc_range(PMD_SIZE, PMD_SIZE, map_start, map_end); - memblock_free(addr, PMD_SIZE); + memblock_phys_free(addr, PMD_SIZE); real_end = addr + PMD_SIZE; /* step_size need to be small so pgt_buf from BRK could cover it */ diff --git a/arch/x86/xen/mmu_pv.c b/arch/x86/xen/mmu_pv.c index 3359c23573c5..676d8d292f8a 100644 --- a/arch/x86/xen/mmu_pv.c +++ b/arch/x86/xen/mmu_pv.c @@ -1025,7 +1025,7 @@ static void __init xen_free_ro_pages(unsigned long paddr, unsigned long size) for (; vaddr < vaddr_end; vaddr += PAGE_SIZE) make_lowmem_page_readwrite(vaddr); - memblock_free(paddr, size); + memblock_phys_free(paddr, size); } static void __init xen_cleanmfnmap_free_pgtbl(void *pgtbl, bool unpin) @@ -1151,7 +1151,7 @@ static void __init xen_pagetable_p2m_free(void) xen_cleanhighmap(addr, addr + size); size = PAGE_ALIGN(xen_start_info->nr_pages * sizeof(unsigned long)); - memblock_free(__pa(addr), size); + memblock_phys_free(__pa(addr), size); } else { xen_cleanmfnmap(addr); } @@ -1955,7 +1955,7 @@ void __init xen_relocate_p2m(void) pfn_end = p2m_pfn_end; } - memblock_free(PFN_PHYS(pfn), PAGE_SIZE * (pfn_end - pfn)); + memblock_phys_free(PFN_PHYS(pfn), PAGE_SIZE * (pfn_end - pfn)); while (pfn < pfn_end) { if (pfn == p2m_pfn) { pfn = p2m_pfn_end; diff --git a/arch/x86/xen/setup.c b/arch/x86/xen/setup.c index 8bfc10330107..f387fc7e5250 100644 --- a/arch/x86/xen/setup.c +++ b/arch/x86/xen/setup.c @@ -153,7 +153,7 @@ static void __init xen_del_extra_mem(unsigned long start_pfn, break; } } - memblock_free(PFN_PHYS(start_pfn), PFN_PHYS(n_pfns)); + memblock_phys_free(PFN_PHYS(start_pfn), PFN_PHYS(n_pfns)); } /* @@ -719,7 +719,7 @@ static void __init xen_reserve_xen_mfnlist(void) return; xen_relocate_p2m(); - memblock_free(start, size); + memblock_phys_free(start, size); } /** @@ -885,7 +885,7 @@ char * __init xen_memory_setup(void) xen_phys_memcpy(new_area, start, size); pr_info("initrd moved from [mem %#010llx-%#010llx] to [mem %#010llx-%#010llx]\n", start, start + size, new_area, new_area + size); - memblock_free(start, size); + memblock_phys_free(start, size); boot_params.hdr.ramdisk_image = new_area; boot_params.ext_ramdisk_image = new_area >> 32; } diff --git a/drivers/base/arch_numa.c b/drivers/base/arch_numa.c index ade8934764f6..712edef03929 100644 --- a/drivers/base/arch_numa.c +++ b/drivers/base/arch_numa.c @@ -166,7 +166,7 @@ static void * __init pcpu_fc_alloc(unsigned int cpu, size_t size, static void __init pcpu_fc_free(void *ptr, size_t size) { - memblock_free(__pa(ptr), size); + memblock_phys_free(__pa(ptr), size); } #ifdef CONFIG_NEED_PER_CPU_PAGE_FIRST_CHUNK diff --git a/drivers/firmware/efi/memmap.c b/drivers/firmware/efi/memmap.c index 2ff1883dc788..4df55a55da84 100644 --- a/drivers/firmware/efi/memmap.c +++ b/drivers/firmware/efi/memmap.c @@ -35,7 +35,7 @@ void __init __efi_memmap_free(u64 phys, unsigned long size, unsigned long flags) if (slab_is_available()) memblock_free_late(phys, size); else - memblock_free(phys, size); + memblock_phys_free(phys, size); } else if (flags & EFI_MEMMAP_SLAB) { struct page *p = pfn_to_page(PHYS_PFN(phys)); unsigned int order = get_order(size); diff --git a/drivers/of/kexec.c b/drivers/of/kexec.c index 053e241f593c..b9bd1cff1793 100644 --- a/drivers/of/kexec.c +++ b/drivers/of/kexec.c @@ -171,8 +171,7 @@ int ima_free_kexec_buffer(void) if (ret) return ret; - return memblock_free(addr, size); - + return memblock_phys_free(addr, size); } /** diff --git a/drivers/of/of_reserved_mem.c b/drivers/of/of_reserved_mem.c index 9da8835ba5a5..9c0fb962c22b 100644 --- a/drivers/of/of_reserved_mem.c +++ b/drivers/of/of_reserved_mem.c @@ -46,7 +46,7 @@ static int __init early_init_dt_alloc_reserved_memory_arch(phys_addr_t size, if (nomap) { err = memblock_mark_nomap(base, size); if (err) - memblock_free(base, size); + memblock_phys_free(base, size); kmemleak_ignore_phys(base); } @@ -284,7 +284,8 @@ void __init fdt_init_reserved_mem(void) if (nomap) memblock_clear_nomap(rmem->base, rmem->size); else - memblock_free(rmem->base, rmem->size); + memblock_phys_free(rmem->base, + rmem->size); } } } diff --git a/drivers/s390/char/sclp_early.c b/drivers/s390/char/sclp_early.c index f01d942e1c1d..c0052655fc4f 100644 --- a/drivers/s390/char/sclp_early.c +++ b/drivers/s390/char/sclp_early.c @@ -139,7 +139,7 @@ int __init sclp_early_get_core_info(struct sclp_core_info *info) } sclp_fill_core_info(info, sccb); out: - memblock_free((unsigned long)sccb, length); + memblock_phys_free((unsigned long)sccb, length); return rc; } diff --git a/drivers/usb/early/xhci-dbc.c b/drivers/usb/early/xhci-dbc.c index be4ecbabdd58..933d77ad0a64 100644 --- a/drivers/usb/early/xhci-dbc.c +++ b/drivers/usb/early/xhci-dbc.c @@ -185,7 +185,7 @@ static void __init xdbc_free_ring(struct xdbc_ring *ring) if (!seg) return; - memblock_free(seg->dma, PAGE_SIZE); + memblock_phys_free(seg->dma, PAGE_SIZE); ring->segment = NULL; } @@ -665,10 +665,10 @@ int __init early_xdbc_setup_hardware(void) xdbc_free_ring(&xdbc.in_ring); if (xdbc.table_dma) - memblock_free(xdbc.table_dma, PAGE_SIZE); + memblock_phys_free(xdbc.table_dma, PAGE_SIZE); if (xdbc.out_dma) - memblock_free(xdbc.out_dma, PAGE_SIZE); + memblock_phys_free(xdbc.out_dma, PAGE_SIZE); xdbc.table_base = NULL; xdbc.out_buf = NULL; @@ -987,8 +987,8 @@ free_and_quit: xdbc_free_ring(&xdbc.evt_ring); xdbc_free_ring(&xdbc.out_ring); xdbc_free_ring(&xdbc.in_ring); - memblock_free(xdbc.table_dma, PAGE_SIZE); - memblock_free(xdbc.out_dma, PAGE_SIZE); + memblock_phys_free(xdbc.table_dma, PAGE_SIZE); + memblock_phys_free(xdbc.out_dma, PAGE_SIZE); writel(0, &xdbc.xdbc_reg->control); early_iounmap(xdbc.xhci_base, xdbc.xhci_length); diff --git a/drivers/xen/swiotlb-xen.c b/drivers/xen/swiotlb-xen.c index e56a5faac395..4b671cc0a7ea 100644 --- a/drivers/xen/swiotlb-xen.c +++ b/drivers/xen/swiotlb-xen.c @@ -241,7 +241,7 @@ retry: */ rc = xen_swiotlb_fixup(start, nslabs); if (rc) { - memblock_free(__pa(start), PAGE_ALIGN(bytes)); + memblock_phys_free(__pa(start), PAGE_ALIGN(bytes)); if (nslabs > 1024 && repeat--) { /* Min is 2MB */ nslabs = max(1024UL, ALIGN(nslabs >> 1, IO_TLB_SEGSIZE)); diff --git a/include/linux/memblock.h b/include/linux/memblock.h index e25f964fdd60..d32d41709513 100644 --- a/include/linux/memblock.h +++ b/include/linux/memblock.h @@ -103,7 +103,7 @@ void memblock_allow_resize(void); int memblock_add_node(phys_addr_t base, phys_addr_t size, int nid); int memblock_add(phys_addr_t base, phys_addr_t size); int memblock_remove(phys_addr_t base, phys_addr_t size); -int memblock_free(phys_addr_t base, phys_addr_t size); +int memblock_phys_free(phys_addr_t base, phys_addr_t size); int memblock_reserve(phys_addr_t base, phys_addr_t size); #ifdef CONFIG_HAVE_MEMBLOCK_PHYS_MAP int memblock_physmem_add(phys_addr_t base, phys_addr_t size); diff --git a/init/initramfs.c b/init/initramfs.c index a842c0544745..1a971f070dd4 100644 --- a/init/initramfs.c +++ b/init/initramfs.c @@ -607,7 +607,7 @@ void __weak __init free_initrd_mem(unsigned long start, unsigned long end) unsigned long aligned_start = ALIGN_DOWN(start, PAGE_SIZE); unsigned long aligned_end = ALIGN(end, PAGE_SIZE); - memblock_free(__pa(aligned_start), aligned_end - aligned_start); + memblock_phys_free(__pa(aligned_start), aligned_end - aligned_start); #endif free_reserved_area((void *)start, (void *)end, POISON_FREE_INITMEM, diff --git a/kernel/dma/swiotlb.c b/kernel/dma/swiotlb.c index 430d2f78d540..b9fa173e5e56 100644 --- a/kernel/dma/swiotlb.c +++ b/kernel/dma/swiotlb.c @@ -247,7 +247,7 @@ swiotlb_init(int verbose) return; fail_free_mem: - memblock_free(__pa(tlb), bytes); + memblock_phys_free(__pa(tlb), bytes); fail: pr_warn("Cannot allocate buffer"); } diff --git a/lib/cpumask.c b/lib/cpumask.c index 045779446a18..a90786b77c1c 100644 --- a/lib/cpumask.c +++ b/lib/cpumask.c @@ -188,7 +188,7 @@ EXPORT_SYMBOL(free_cpumask_var); */ void __init free_bootmem_cpumask_var(cpumask_var_t mask) { - memblock_free(__pa(mask), cpumask_size()); + memblock_phys_free(__pa(mask), cpumask_size()); } #endif diff --git a/mm/cma.c b/mm/cma.c index 11152c3fb23c..bc9ca8f3c487 100644 --- a/mm/cma.c +++ b/mm/cma.c @@ -378,7 +378,7 @@ int __init cma_declare_contiguous_nid(phys_addr_t base, return 0; free_mem: - memblock_free(base, size); + memblock_phys_free(base, size); err: pr_err("Failed to reserve %ld MiB\n", (unsigned long)size / SZ_1M); return ret; diff --git a/mm/memblock.c b/mm/memblock.c index 849060013d3c..52e34abc4abe 100644 --- a/mm/memblock.c +++ b/mm/memblock.c @@ -806,18 +806,18 @@ int __init_memblock memblock_remove(phys_addr_t base, phys_addr_t size) void __init_memblock memblock_free_ptr(void *ptr, size_t size) { if (ptr) - memblock_free(__pa(ptr), size); + memblock_phys_free(__pa(ptr), size); } /** - * memblock_free - free boot memory block + * memblock_phys_free - free boot memory block * @base: phys starting address of the boot memory block * @size: size of the boot memory block in bytes * * Free boot memory block previously allocated by memblock_alloc_xx() API. * The freeing memory will not be released to the buddy allocator. */ -int __init_memblock memblock_free(phys_addr_t base, phys_addr_t size) +int __init_memblock memblock_phys_free(phys_addr_t base, phys_addr_t size) { phys_addr_t end = base + size - 1; @@ -1937,7 +1937,7 @@ static void __init free_memmap(unsigned long start_pfn, unsigned long end_pfn) * memmap array. */ if (pg < pgend) - memblock_free(pg, pgend - pg); + memblock_phys_free(pg, pgend - pg); } /* diff --git a/mm/memory_hotplug.c b/mm/memory_hotplug.c index 9fd0be32a281..feffaa9423fe 100644 --- a/mm/memory_hotplug.c +++ b/mm/memory_hotplug.c @@ -2204,7 +2204,7 @@ static int __ref try_remove_memory(u64 start, u64 size) arch_remove_memory(start, size, altmap); if (IS_ENABLED(CONFIG_ARCH_KEEP_MEMBLOCK)) { - memblock_free(start, size); + memblock_phys_free(start, size); memblock_remove(start, size); } diff --git a/mm/percpu.c b/mm/percpu.c index f58318cb04c0..d65ddf6f2a35 100644 --- a/mm/percpu.c +++ b/mm/percpu.c @@ -2472,7 +2472,7 @@ struct pcpu_alloc_info * __init pcpu_alloc_alloc_info(int nr_groups, */ void __init pcpu_free_alloc_info(struct pcpu_alloc_info *ai) { - memblock_free(__pa(ai), ai->__ai_size); + memblock_phys_free(__pa(ai), ai->__ai_size); } /** @@ -3134,7 +3134,7 @@ out_free_areas: out_free: pcpu_free_alloc_info(ai); if (areas) - memblock_free(__pa(areas), areas_size); + memblock_phys_free(__pa(areas), areas_size); return rc; } #endif /* BUILD_EMBED_FIRST_CHUNK */ @@ -3256,7 +3256,7 @@ enomem: free_fn(page_address(pages[j]), PAGE_SIZE); rc = -ENOMEM; out_free_ar: - memblock_free(__pa(pages), pages_size); + memblock_phys_free(__pa(pages), pages_size); pcpu_free_alloc_info(ai); return rc; } @@ -3286,7 +3286,7 @@ static void * __init pcpu_dfl_fc_alloc(unsigned int cpu, size_t size, static void __init pcpu_dfl_fc_free(void *ptr, size_t size) { - memblock_free(__pa(ptr), size); + memblock_phys_free(__pa(ptr), size); } void __init setup_per_cpu_areas(void) diff --git a/mm/sparse.c b/mm/sparse.c index 55fea0c2f927..fc3ab8d3b6bc 100644 --- a/mm/sparse.c +++ b/mm/sparse.c @@ -451,7 +451,7 @@ static void *sparsemap_buf_end __meminitdata; static inline void __meminit sparse_buffer_free(unsigned long size) { WARN_ON(!sparsemap_buf || size == 0); - memblock_free(__pa(sparsemap_buf), size); + memblock_phys_free(__pa(sparsemap_buf), size); } static void __init sparse_buffer_init(unsigned long size, int nid) -- cgit v1.2.3 From 4421cca0a3e4833b3bf0f20de98eb580ab8c7290 Mon Sep 17 00:00:00 2001 From: Mike Rapoport Date: Fri, 5 Nov 2021 13:43:22 -0700 Subject: memblock: use memblock_free for freeing virtual pointers Rename memblock_free_ptr() to memblock_free() and use memblock_free() when freeing a virtual pointer so that memblock_free() will be a counterpart of memblock_alloc() The callers are updated with the below semantic patch and manual addition of (void *) casting to pointers that are represented by unsigned long variables. @@ identifier vaddr; expression size; @@ ( - memblock_phys_free(__pa(vaddr), size); + memblock_free(vaddr, size); | - memblock_free_ptr(vaddr, size); + memblock_free(vaddr, size); ) [sfr@canb.auug.org.au: fixup] Link: https://lkml.kernel.org/r/20211018192940.3d1d532f@canb.auug.org.au Link: https://lkml.kernel.org/r/20210930185031.18648-7-rppt@kernel.org Signed-off-by: Mike Rapoport Signed-off-by: Stephen Rothwell Cc: Christophe Leroy Cc: Juergen Gross Cc: Shahab Vahedi Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/alpha/kernel/core_irongate.c | 3 +-- arch/mips/mm/init.c | 2 +- arch/powerpc/kernel/dt_cpu_ftrs.c | 4 ++-- arch/powerpc/kernel/setup-common.c | 2 +- arch/powerpc/kernel/setup_64.c | 2 +- arch/powerpc/platforms/powernv/pci-ioda.c | 2 +- arch/powerpc/platforms/pseries/svm.c | 3 +-- arch/riscv/kernel/setup.c | 5 ++--- arch/sparc/kernel/smp_64.c | 2 +- arch/um/kernel/mem.c | 2 +- arch/x86/kernel/setup_percpu.c | 2 +- arch/x86/mm/kasan_init_64.c | 4 ++-- arch/x86/mm/numa.c | 2 +- arch/x86/mm/numa_emulation.c | 2 +- arch/x86/xen/mmu_pv.c | 2 +- arch/x86/xen/p2m.c | 2 +- drivers/base/arch_numa.c | 4 ++-- drivers/macintosh/smu.c | 2 +- drivers/xen/swiotlb-xen.c | 2 +- include/linux/memblock.h | 2 +- init/initramfs.c | 2 +- init/main.c | 4 ++-- kernel/dma/swiotlb.c | 2 +- kernel/printk/printk.c | 4 ++-- lib/bootconfig.c | 2 +- lib/cpumask.c | 2 +- mm/memblock.c | 6 +++--- mm/percpu.c | 8 ++++---- mm/sparse.c | 2 +- 29 files changed, 40 insertions(+), 43 deletions(-) (limited to 'lib') diff --git a/arch/alpha/kernel/core_irongate.c b/arch/alpha/kernel/core_irongate.c index ee26dcc49418..6b8ed12936b6 100644 --- a/arch/alpha/kernel/core_irongate.c +++ b/arch/alpha/kernel/core_irongate.c @@ -233,8 +233,7 @@ albacore_init_arch(void) unsigned long size; size = initrd_end - initrd_start; - memblock_phys_free(__pa(initrd_start), - PAGE_ALIGN(size)); + memblock_free((void *)initrd_start, PAGE_ALIGN(size)); if (!move_initrd(pci_mem)) printk("irongate_init_arch: initrd too big " "(%ldK)\ndisabling initrd\n", diff --git a/arch/mips/mm/init.c b/arch/mips/mm/init.c index 3be1c29084fa..325e1552cbea 100644 --- a/arch/mips/mm/init.c +++ b/arch/mips/mm/init.c @@ -529,7 +529,7 @@ static void * __init pcpu_fc_alloc(unsigned int cpu, size_t size, static void __init pcpu_fc_free(void *ptr, size_t size) { - memblock_phys_free(__pa(ptr), size); + memblock_free(ptr, size); } void __init setup_per_cpu_areas(void) diff --git a/arch/powerpc/kernel/dt_cpu_ftrs.c b/arch/powerpc/kernel/dt_cpu_ftrs.c index 42839d6bd486..ba527fb52993 100644 --- a/arch/powerpc/kernel/dt_cpu_ftrs.c +++ b/arch/powerpc/kernel/dt_cpu_ftrs.c @@ -1095,8 +1095,8 @@ static int __init dt_cpu_ftrs_scan_callback(unsigned long node, const char cpufeatures_setup_finished(); - memblock_phys_free(__pa(dt_cpu_features), - sizeof(struct dt_cpu_feature) * nr_dt_cpu_features); + memblock_free(dt_cpu_features, + sizeof(struct dt_cpu_feature) * nr_dt_cpu_features); return 0; } diff --git a/arch/powerpc/kernel/setup-common.c b/arch/powerpc/kernel/setup-common.c index 5af8993a8e6d..6b1338db8779 100644 --- a/arch/powerpc/kernel/setup-common.c +++ b/arch/powerpc/kernel/setup-common.c @@ -825,7 +825,7 @@ static void __init smp_setup_pacas(void) set_hard_smp_processor_id(cpu, cpu_to_phys_id[cpu]); } - memblock_phys_free(__pa(cpu_to_phys_id), nr_cpu_ids * sizeof(u32)); + memblock_free(cpu_to_phys_id, nr_cpu_ids * sizeof(u32)); cpu_to_phys_id = NULL; } #endif diff --git a/arch/powerpc/kernel/setup_64.c b/arch/powerpc/kernel/setup_64.c index 75bc294ac40d..1777e992b20b 100644 --- a/arch/powerpc/kernel/setup_64.c +++ b/arch/powerpc/kernel/setup_64.c @@ -812,7 +812,7 @@ static void * __init pcpu_alloc_bootmem(unsigned int cpu, size_t size, static void __init pcpu_free_bootmem(void *ptr, size_t size) { - memblock_phys_free(__pa(ptr), size); + memblock_free(ptr, size); } static int pcpu_cpu_distance(unsigned int from, unsigned int to) diff --git a/arch/powerpc/platforms/powernv/pci-ioda.c b/arch/powerpc/platforms/powernv/pci-ioda.c index b5a9d343b720..004cd6a96c8a 100644 --- a/arch/powerpc/platforms/powernv/pci-ioda.c +++ b/arch/powerpc/platforms/powernv/pci-ioda.c @@ -2981,7 +2981,7 @@ static void __init pnv_pci_init_ioda_phb(struct device_node *np, if (!phb->hose) { pr_err(" Can't allocate PCI controller for %pOF\n", np); - memblock_phys_free(__pa(phb), sizeof(struct pnv_phb)); + memblock_free(phb, sizeof(struct pnv_phb)); return; } diff --git a/arch/powerpc/platforms/pseries/svm.c b/arch/powerpc/platforms/pseries/svm.c index b7c017bb40f7..6332365d2891 100644 --- a/arch/powerpc/platforms/pseries/svm.c +++ b/arch/powerpc/platforms/pseries/svm.c @@ -56,8 +56,7 @@ void __init svm_swiotlb_init(void) return; - memblock_phys_free(__pa(vstart), - PAGE_ALIGN(io_tlb_nslabs << IO_TLB_SHIFT)); + memblock_free(vstart, PAGE_ALIGN(io_tlb_nslabs << IO_TLB_SHIFT)); panic("SVM: Cannot allocate SWIOTLB buffer"); } diff --git a/arch/riscv/kernel/setup.c b/arch/riscv/kernel/setup.c index 6ea7c53b82cd..b42bfdc67482 100644 --- a/arch/riscv/kernel/setup.c +++ b/arch/riscv/kernel/setup.c @@ -230,14 +230,13 @@ static void __init init_resources(void) /* Clean-up any unused pre-allocated resources */ if (res_idx >= 0) - memblock_phys_free(__pa(mem_res), - (res_idx + 1) * sizeof(*mem_res)); + memblock_free(mem_res, (res_idx + 1) * sizeof(*mem_res)); return; error: /* Better an empty resource tree than an inconsistent one */ release_child_resources(&iomem_resource); - memblock_phys_free(__pa(mem_res), mem_res_sz); + memblock_free(mem_res, mem_res_sz); } diff --git a/arch/sparc/kernel/smp_64.c b/arch/sparc/kernel/smp_64.c index 2507549538df..b98a7bbe6728 100644 --- a/arch/sparc/kernel/smp_64.c +++ b/arch/sparc/kernel/smp_64.c @@ -1567,7 +1567,7 @@ static void * __init pcpu_alloc_bootmem(unsigned int cpu, size_t size, static void __init pcpu_free_bootmem(void *ptr, size_t size) { - memblock_phys_free(__pa(ptr), size); + memblock_free(ptr, size); } static int __init pcpu_cpu_distance(unsigned int from, unsigned int to) diff --git a/arch/um/kernel/mem.c b/arch/um/kernel/mem.c index d1710ebb44f4..0039771eb01c 100644 --- a/arch/um/kernel/mem.c +++ b/arch/um/kernel/mem.c @@ -47,7 +47,7 @@ void __init mem_init(void) */ brk_end = (unsigned long) UML_ROUND_UP(sbrk(0)); map_memory(brk_end, __pa(brk_end), uml_reserved - brk_end, 1, 1, 0); - memblock_phys_free(__pa(brk_end), uml_reserved - brk_end); + memblock_free((void *)brk_end, uml_reserved - brk_end); uml_reserved = brk_end; /* this will put all low memory onto the freelists */ diff --git a/arch/x86/kernel/setup_percpu.c b/arch/x86/kernel/setup_percpu.c index 5afd98559193..7b65275544b2 100644 --- a/arch/x86/kernel/setup_percpu.c +++ b/arch/x86/kernel/setup_percpu.c @@ -135,7 +135,7 @@ static void * __init pcpu_fc_alloc(unsigned int cpu, size_t size, size_t align) static void __init pcpu_fc_free(void *ptr, size_t size) { - memblock_free_ptr(ptr, size); + memblock_free(ptr, size); } static int __init pcpu_cpu_distance(unsigned int from, unsigned int to) diff --git a/arch/x86/mm/kasan_init_64.c b/arch/x86/mm/kasan_init_64.c index ef885370719a..e7b9b464a82f 100644 --- a/arch/x86/mm/kasan_init_64.c +++ b/arch/x86/mm/kasan_init_64.c @@ -49,7 +49,7 @@ static void __init kasan_populate_pmd(pmd_t *pmd, unsigned long addr, p = early_alloc(PMD_SIZE, nid, false); if (p && pmd_set_huge(pmd, __pa(p), PAGE_KERNEL)) return; - memblock_free_ptr(p, PMD_SIZE); + memblock_free(p, PMD_SIZE); } p = early_alloc(PAGE_SIZE, nid, true); @@ -85,7 +85,7 @@ static void __init kasan_populate_pud(pud_t *pud, unsigned long addr, p = early_alloc(PUD_SIZE, nid, false); if (p && pud_set_huge(pud, __pa(p), PAGE_KERNEL)) return; - memblock_free_ptr(p, PUD_SIZE); + memblock_free(p, PUD_SIZE); } p = early_alloc(PAGE_SIZE, nid, true); diff --git a/arch/x86/mm/numa.c b/arch/x86/mm/numa.c index 1e9b93b088db..c6b1213086d6 100644 --- a/arch/x86/mm/numa.c +++ b/arch/x86/mm/numa.c @@ -355,7 +355,7 @@ void __init numa_reset_distance(void) /* numa_distance could be 1LU marking allocation failure, test cnt */ if (numa_distance_cnt) - memblock_free_ptr(numa_distance, size); + memblock_free(numa_distance, size); numa_distance_cnt = 0; numa_distance = NULL; /* enable table creation */ } diff --git a/arch/x86/mm/numa_emulation.c b/arch/x86/mm/numa_emulation.c index e801e30089c4..1a02b791d273 100644 --- a/arch/x86/mm/numa_emulation.c +++ b/arch/x86/mm/numa_emulation.c @@ -517,7 +517,7 @@ void __init numa_emulation(struct numa_meminfo *numa_meminfo, int numa_dist_cnt) } /* free the copied physical distance table */ - memblock_free_ptr(phys_dist, phys_size); + memblock_free(phys_dist, phys_size); return; no_emu: diff --git a/arch/x86/xen/mmu_pv.c b/arch/x86/xen/mmu_pv.c index 676d8d292f8a..173de1e29bda 100644 --- a/arch/x86/xen/mmu_pv.c +++ b/arch/x86/xen/mmu_pv.c @@ -1151,7 +1151,7 @@ static void __init xen_pagetable_p2m_free(void) xen_cleanhighmap(addr, addr + size); size = PAGE_ALIGN(xen_start_info->nr_pages * sizeof(unsigned long)); - memblock_phys_free(__pa(addr), size); + memblock_free((void *)addr, size); } else { xen_cleanmfnmap(addr); } diff --git a/arch/x86/xen/p2m.c b/arch/x86/xen/p2m.c index 141bb9dbd2fb..58db86f7b384 100644 --- a/arch/x86/xen/p2m.c +++ b/arch/x86/xen/p2m.c @@ -197,7 +197,7 @@ static void * __ref alloc_p2m_page(void) static void __ref free_p2m_page(void *p) { if (unlikely(!slab_is_available())) { - memblock_free_ptr(p, PAGE_SIZE); + memblock_free(p, PAGE_SIZE); return; } diff --git a/drivers/base/arch_numa.c b/drivers/base/arch_numa.c index 712edef03929..bc1876915457 100644 --- a/drivers/base/arch_numa.c +++ b/drivers/base/arch_numa.c @@ -166,7 +166,7 @@ static void * __init pcpu_fc_alloc(unsigned int cpu, size_t size, static void __init pcpu_fc_free(void *ptr, size_t size) { - memblock_phys_free(__pa(ptr), size); + memblock_free(ptr, size); } #ifdef CONFIG_NEED_PER_CPU_PAGE_FIRST_CHUNK @@ -326,7 +326,7 @@ void __init numa_free_distance(void) size = numa_distance_cnt * numa_distance_cnt * sizeof(numa_distance[0]); - memblock_free_ptr(numa_distance, size); + memblock_free(numa_distance, size); numa_distance_cnt = 0; numa_distance = NULL; } diff --git a/drivers/macintosh/smu.c b/drivers/macintosh/smu.c index fe63d5ee201b..f62152111236 100644 --- a/drivers/macintosh/smu.c +++ b/drivers/macintosh/smu.c @@ -570,7 +570,7 @@ fail_msg_node: fail_db_node: of_node_put(smu->db_node); fail_bootmem: - memblock_free_ptr(smu, sizeof(struct smu_device)); + memblock_free(smu, sizeof(struct smu_device)); smu = NULL; fail_np: of_node_put(np); diff --git a/drivers/xen/swiotlb-xen.c b/drivers/xen/swiotlb-xen.c index 4b671cc0a7ea..f083194e2634 100644 --- a/drivers/xen/swiotlb-xen.c +++ b/drivers/xen/swiotlb-xen.c @@ -241,7 +241,7 @@ retry: */ rc = xen_swiotlb_fixup(start, nslabs); if (rc) { - memblock_phys_free(__pa(start), PAGE_ALIGN(bytes)); + memblock_free(start, PAGE_ALIGN(bytes)); if (nslabs > 1024 && repeat--) { /* Min is 2MB */ nslabs = max(1024UL, ALIGN(nslabs >> 1, IO_TLB_SEGSIZE)); diff --git a/include/linux/memblock.h b/include/linux/memblock.h index d32d41709513..484650681bee 100644 --- a/include/linux/memblock.h +++ b/include/linux/memblock.h @@ -118,7 +118,7 @@ int memblock_mark_nomap(phys_addr_t base, phys_addr_t size); int memblock_clear_nomap(phys_addr_t base, phys_addr_t size); void memblock_free_all(void); -void memblock_free_ptr(void *ptr, size_t size); +void memblock_free(void *ptr, size_t size); void reset_node_managed_pages(pg_data_t *pgdat); void reset_all_zones_managed_pages(void); diff --git a/init/initramfs.c b/init/initramfs.c index 1a971f070dd4..2f3d96dc3db6 100644 --- a/init/initramfs.c +++ b/init/initramfs.c @@ -607,7 +607,7 @@ void __weak __init free_initrd_mem(unsigned long start, unsigned long end) unsigned long aligned_start = ALIGN_DOWN(start, PAGE_SIZE); unsigned long aligned_end = ALIGN(end, PAGE_SIZE); - memblock_phys_free(__pa(aligned_start), aligned_end - aligned_start); + memblock_free((void *)aligned_start, aligned_end - aligned_start); #endif free_reserved_area((void *)start, (void *)end, POISON_FREE_INITMEM, diff --git a/init/main.c b/init/main.c index 767ee2672176..f0001af8ebb9 100644 --- a/init/main.c +++ b/init/main.c @@ -382,7 +382,7 @@ static char * __init xbc_make_cmdline(const char *key) ret = xbc_snprint_cmdline(new_cmdline, len + 1, root); if (ret < 0 || ret > len) { pr_err("Failed to print extra kernel cmdline.\n"); - memblock_free_ptr(new_cmdline, len + 1); + memblock_free(new_cmdline, len + 1); return NULL; } @@ -925,7 +925,7 @@ static void __init print_unknown_bootoptions(void) end += sprintf(end, " %s", *p); pr_notice("Unknown command line parameters:%s\n", unknown_options); - memblock_free_ptr(unknown_options, len); + memblock_free(unknown_options, len); } asmlinkage __visible void __init __no_sanitize_address start_kernel(void) diff --git a/kernel/dma/swiotlb.c b/kernel/dma/swiotlb.c index b9fa173e5e56..02656d7ccbfd 100644 --- a/kernel/dma/swiotlb.c +++ b/kernel/dma/swiotlb.c @@ -247,7 +247,7 @@ swiotlb_init(int verbose) return; fail_free_mem: - memblock_phys_free(__pa(tlb), bytes); + memblock_free(tlb, bytes); fail: pr_warn("Cannot allocate buffer"); } diff --git a/kernel/printk/printk.c b/kernel/printk/printk.c index a8d0a58deebc..2cae1bfa6be7 100644 --- a/kernel/printk/printk.c +++ b/kernel/printk/printk.c @@ -1166,9 +1166,9 @@ void __init setup_log_buf(int early) return; err_free_descs: - memblock_free_ptr(new_descs, new_descs_size); + memblock_free(new_descs, new_descs_size); err_free_log_buf: - memblock_free_ptr(new_log_buf, new_log_buf_len); + memblock_free(new_log_buf, new_log_buf_len); } static bool __read_mostly ignore_loglevel; diff --git a/lib/bootconfig.c b/lib/bootconfig.c index 5ae248b29373..547558d80e64 100644 --- a/lib/bootconfig.c +++ b/lib/bootconfig.c @@ -792,7 +792,7 @@ void __init xbc_destroy_all(void) xbc_data = NULL; xbc_data_size = 0; xbc_node_num = 0; - memblock_free_ptr(xbc_nodes, sizeof(struct xbc_node) * XBC_NODE_MAX); + memblock_free(xbc_nodes, sizeof(struct xbc_node) * XBC_NODE_MAX); xbc_nodes = NULL; brace_index = 0; } diff --git a/lib/cpumask.c b/lib/cpumask.c index a90786b77c1c..a971a82d2f43 100644 --- a/lib/cpumask.c +++ b/lib/cpumask.c @@ -188,7 +188,7 @@ EXPORT_SYMBOL(free_cpumask_var); */ void __init free_bootmem_cpumask_var(cpumask_var_t mask) { - memblock_phys_free(__pa(mask), cpumask_size()); + memblock_free(mask, cpumask_size()); } #endif diff --git a/mm/memblock.c b/mm/memblock.c index 52e34abc4abe..fb0c7f48e627 100644 --- a/mm/memblock.c +++ b/mm/memblock.c @@ -472,7 +472,7 @@ static int __init_memblock memblock_double_array(struct memblock_type *type, kfree(old_array); else if (old_array != memblock_memory_init_regions && old_array != memblock_reserved_init_regions) - memblock_free_ptr(old_array, old_alloc_size); + memblock_free(old_array, old_alloc_size); /* * Reserve the new array if that comes from the memblock. Otherwise, we @@ -796,14 +796,14 @@ int __init_memblock memblock_remove(phys_addr_t base, phys_addr_t size) } /** - * memblock_free_ptr - free boot memory allocation + * memblock_free - free boot memory allocation * @ptr: starting address of the boot memory allocation * @size: size of the boot memory block in bytes * * Free boot memory block previously allocated by memblock_alloc_xx() API. * The freeing memory will not be released to the buddy allocator. */ -void __init_memblock memblock_free_ptr(void *ptr, size_t size) +void __init_memblock memblock_free(void *ptr, size_t size) { if (ptr) memblock_phys_free(__pa(ptr), size); diff --git a/mm/percpu.c b/mm/percpu.c index d65ddf6f2a35..f5b2c2ea5a54 100644 --- a/mm/percpu.c +++ b/mm/percpu.c @@ -2472,7 +2472,7 @@ struct pcpu_alloc_info * __init pcpu_alloc_alloc_info(int nr_groups, */ void __init pcpu_free_alloc_info(struct pcpu_alloc_info *ai) { - memblock_phys_free(__pa(ai), ai->__ai_size); + memblock_free(ai, ai->__ai_size); } /** @@ -3134,7 +3134,7 @@ out_free_areas: out_free: pcpu_free_alloc_info(ai); if (areas) - memblock_phys_free(__pa(areas), areas_size); + memblock_free(areas, areas_size); return rc; } #endif /* BUILD_EMBED_FIRST_CHUNK */ @@ -3256,7 +3256,7 @@ enomem: free_fn(page_address(pages[j]), PAGE_SIZE); rc = -ENOMEM; out_free_ar: - memblock_phys_free(__pa(pages), pages_size); + memblock_free(pages, pages_size); pcpu_free_alloc_info(ai); return rc; } @@ -3286,7 +3286,7 @@ static void * __init pcpu_dfl_fc_alloc(unsigned int cpu, size_t size, static void __init pcpu_dfl_fc_free(void *ptr, size_t size) { - memblock_phys_free(__pa(ptr), size); + memblock_free(ptr, size); } void __init setup_per_cpu_areas(void) diff --git a/mm/sparse.c b/mm/sparse.c index fc3ab8d3b6bc..e5c84b0cf0c9 100644 --- a/mm/sparse.c +++ b/mm/sparse.c @@ -451,7 +451,7 @@ static void *sparsemap_buf_end __meminitdata; static inline void __meminit sparse_buffer_free(unsigned long size) { WARN_ON(!sparsemap_buf || size == 0); - memblock_phys_free(__pa(sparsemap_buf), size); + memblock_free(sparsemap_buf, size); } static void __init sparse_buffer_init(unsigned long size, int nid) -- cgit v1.2.3 From 50f9481ed9fb8a2d2a06a155634c7f9eeff9fa61 Mon Sep 17 00:00:00 2001 From: David Hildenbrand Date: Fri, 5 Nov 2021 13:44:24 -0700 Subject: mm/memory_hotplug: remove CONFIG_MEMORY_HOTPLUG_SPARSE CONFIG_MEMORY_HOTPLUG depends on CONFIG_SPARSEMEM, so there is no need for CONFIG_MEMORY_HOTPLUG_SPARSE anymore; adjust all instances to use CONFIG_MEMORY_HOTPLUG and remove CONFIG_MEMORY_HOTPLUG_SPARSE. Link: https://lkml.kernel.org/r/20210929143600.49379-3-david@redhat.com Signed-off-by: David Hildenbrand Acked-by: Shuah Khan [kselftest] Acked-by: Greg Kroah-Hartman Acked-by: Oscar Salvador Cc: Alex Shi Cc: Andy Lutomirski Cc: Benjamin Herrenschmidt Cc: Borislav Petkov Cc: Dave Hansen Cc: "H. Peter Anvin" Cc: Ingo Molnar Cc: Jason Wang Cc: Jonathan Corbet Cc: Michael Ellerman Cc: "Michael S. Tsirkin" Cc: Michal Hocko Cc: Mike Rapoport Cc: Paul Mackerras Cc: Peter Zijlstra Cc: "Rafael J. Wysocki" Cc: Thomas Gleixner Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/powerpc/include/asm/machdep.h | 2 +- arch/powerpc/kernel/setup_64.c | 2 +- arch/powerpc/platforms/powernv/setup.c | 4 ++-- arch/powerpc/platforms/pseries/setup.c | 2 +- drivers/base/Makefile | 2 +- drivers/base/node.c | 9 ++++----- drivers/virtio/Kconfig | 2 +- include/linux/memory.h | 24 ++++++++++-------------- include/linux/node.h | 4 ++-- lib/Kconfig.debug | 2 +- mm/Kconfig | 4 ---- mm/memory_hotplug.c | 2 -- tools/testing/selftests/memory-hotplug/config | 1 - 13 files changed, 24 insertions(+), 36 deletions(-) (limited to 'lib') diff --git a/arch/powerpc/include/asm/machdep.h b/arch/powerpc/include/asm/machdep.h index 764f2732a821..d8a2ca007082 100644 --- a/arch/powerpc/include/asm/machdep.h +++ b/arch/powerpc/include/asm/machdep.h @@ -32,7 +32,7 @@ struct machdep_calls { void (*iommu_save)(void); void (*iommu_restore)(void); #endif -#ifdef CONFIG_MEMORY_HOTPLUG_SPARSE +#ifdef CONFIG_MEMORY_HOTPLUG unsigned long (*memory_block_size)(void); #endif #endif /* CONFIG_PPC64 */ diff --git a/arch/powerpc/kernel/setup_64.c b/arch/powerpc/kernel/setup_64.c index 1777e992b20b..6052f5d5ded3 100644 --- a/arch/powerpc/kernel/setup_64.c +++ b/arch/powerpc/kernel/setup_64.c @@ -912,7 +912,7 @@ void __init setup_per_cpu_areas(void) } #endif -#ifdef CONFIG_MEMORY_HOTPLUG_SPARSE +#ifdef CONFIG_MEMORY_HOTPLUG unsigned long memory_block_size_bytes(void) { if (ppc_md.memory_block_size) diff --git a/arch/powerpc/platforms/powernv/setup.c b/arch/powerpc/platforms/powernv/setup.c index a8db3f153063..ad56a54ac9c5 100644 --- a/arch/powerpc/platforms/powernv/setup.c +++ b/arch/powerpc/platforms/powernv/setup.c @@ -440,7 +440,7 @@ static void pnv_kexec_cpu_down(int crash_shutdown, int secondary) } #endif /* CONFIG_KEXEC_CORE */ -#ifdef CONFIG_MEMORY_HOTPLUG_SPARSE +#ifdef CONFIG_MEMORY_HOTPLUG static unsigned long pnv_memory_block_size(void) { /* @@ -553,7 +553,7 @@ define_machine(powernv) { #ifdef CONFIG_KEXEC_CORE .kexec_cpu_down = pnv_kexec_cpu_down, #endif -#ifdef CONFIG_MEMORY_HOTPLUG_SPARSE +#ifdef CONFIG_MEMORY_HOTPLUG .memory_block_size = pnv_memory_block_size, #endif }; diff --git a/arch/powerpc/platforms/pseries/setup.c b/arch/powerpc/platforms/pseries/setup.c index f79126f16258..d29f6f1f7f37 100644 --- a/arch/powerpc/platforms/pseries/setup.c +++ b/arch/powerpc/platforms/pseries/setup.c @@ -1089,7 +1089,7 @@ define_machine(pseries) { .machine_kexec = pSeries_machine_kexec, .kexec_cpu_down = pseries_kexec_cpu_down, #endif -#ifdef CONFIG_MEMORY_HOTPLUG_SPARSE +#ifdef CONFIG_MEMORY_HOTPLUG .memory_block_size = pseries_memory_block_size, #endif }; diff --git a/drivers/base/Makefile b/drivers/base/Makefile index ef8e44a7d288..02f7f1358e86 100644 --- a/drivers/base/Makefile +++ b/drivers/base/Makefile @@ -13,7 +13,7 @@ obj-y += power/ obj-$(CONFIG_ISA_BUS_API) += isa.o obj-y += firmware_loader/ obj-$(CONFIG_NUMA) += node.o -obj-$(CONFIG_MEMORY_HOTPLUG_SPARSE) += memory.o +obj-$(CONFIG_MEMORY_HOTPLUG) += memory.o ifeq ($(CONFIG_SYSFS),y) obj-$(CONFIG_MODULES) += module.o endif diff --git a/drivers/base/node.c b/drivers/base/node.c index c56d34f8158f..b5a4ba18f9f9 100644 --- a/drivers/base/node.c +++ b/drivers/base/node.c @@ -629,7 +629,7 @@ static void node_device_release(struct device *dev) { struct node *node = to_node(dev); -#if defined(CONFIG_MEMORY_HOTPLUG_SPARSE) && defined(CONFIG_HUGETLBFS) +#if defined(CONFIG_MEMORY_HOTPLUG) && defined(CONFIG_HUGETLBFS) /* * We schedule the work only when a memory section is * onlined/offlined on this node. When we come here, @@ -782,7 +782,7 @@ int unregister_cpu_under_node(unsigned int cpu, unsigned int nid) return 0; } -#ifdef CONFIG_MEMORY_HOTPLUG_SPARSE +#ifdef CONFIG_MEMORY_HOTPLUG static int __ref get_nid_for_pfn(unsigned long pfn) { #ifdef CONFIG_DEFERRED_STRUCT_PAGE_INIT @@ -958,10 +958,9 @@ static int node_memory_callback(struct notifier_block *self, return NOTIFY_OK; } #endif /* CONFIG_HUGETLBFS */ -#endif /* CONFIG_MEMORY_HOTPLUG_SPARSE */ +#endif /* CONFIG_MEMORY_HOTPLUG */ -#if !defined(CONFIG_MEMORY_HOTPLUG_SPARSE) || \ - !defined(CONFIG_HUGETLBFS) +#if !defined(CONFIG_MEMORY_HOTPLUG) || !defined(CONFIG_HUGETLBFS) static inline int node_memory_callback(struct notifier_block *self, unsigned long action, void *arg) { diff --git a/drivers/virtio/Kconfig b/drivers/virtio/Kconfig index ce1b3f6ec325..3654def9915c 100644 --- a/drivers/virtio/Kconfig +++ b/drivers/virtio/Kconfig @@ -98,7 +98,7 @@ config VIRTIO_MEM default m depends on X86_64 depends on VIRTIO - depends on MEMORY_HOTPLUG_SPARSE + depends on MEMORY_HOTPLUG depends on MEMORY_HOTREMOVE depends on CONTIG_ALLOC help diff --git a/include/linux/memory.h b/include/linux/memory.h index 053a530c7bdd..0328ec039c38 100644 --- a/include/linux/memory.h +++ b/include/linux/memory.h @@ -110,7 +110,7 @@ struct mem_section; #define SLAB_CALLBACK_PRI 1 #define IPC_CALLBACK_PRI 10 -#ifndef CONFIG_MEMORY_HOTPLUG_SPARSE +#ifndef CONFIG_MEMORY_HOTPLUG static inline void memory_dev_init(void) { return; @@ -126,7 +126,14 @@ static inline int memory_notify(unsigned long val, void *v) { return 0; } -#else +static inline int hotplug_memory_notifier(notifier_fn_t fn, int pri) +{ + return 0; +} +/* These aren't inline functions due to a GCC bug. */ +#define register_hotmemory_notifier(nb) ({ (void)(nb); 0; }) +#define unregister_hotmemory_notifier(nb) ({ (void)(nb); }) +#else /* CONFIG_MEMORY_HOTPLUG */ extern int register_memory_notifier(struct notifier_block *nb); extern void unregister_memory_notifier(struct notifier_block *nb); int create_memory_block_devices(unsigned long start, unsigned long size, @@ -148,9 +155,6 @@ struct memory_group *memory_group_find_by_id(int mgid); typedef int (*walk_memory_groups_func_t)(struct memory_group *, void *); int walk_dynamic_memory_groups(int nid, walk_memory_groups_func_t func, struct memory_group *excluded, void *arg); -#endif /* CONFIG_MEMORY_HOTPLUG_SPARSE */ - -#ifdef CONFIG_MEMORY_HOTPLUG #define hotplug_memory_notifier(fn, pri) ({ \ static __meminitdata struct notifier_block fn##_mem_nb =\ { .notifier_call = fn, .priority = pri };\ @@ -158,15 +162,7 @@ int walk_dynamic_memory_groups(int nid, walk_memory_groups_func_t func, }) #define register_hotmemory_notifier(nb) register_memory_notifier(nb) #define unregister_hotmemory_notifier(nb) unregister_memory_notifier(nb) -#else -static inline int hotplug_memory_notifier(notifier_fn_t fn, int pri) -{ - return 0; -} -/* These aren't inline functions due to a GCC bug. */ -#define register_hotmemory_notifier(nb) ({ (void)(nb); 0; }) -#define unregister_hotmemory_notifier(nb) ({ (void)(nb); }) -#endif +#endif /* CONFIG_MEMORY_HOTPLUG */ /* * Kernel text modification mutex, used for code patching. Users of this lock diff --git a/include/linux/node.h b/include/linux/node.h index 8e5a29897936..bb21fd631b16 100644 --- a/include/linux/node.h +++ b/include/linux/node.h @@ -85,7 +85,7 @@ struct node { struct device dev; struct list_head access_list; -#if defined(CONFIG_MEMORY_HOTPLUG_SPARSE) && defined(CONFIG_HUGETLBFS) +#if defined(CONFIG_MEMORY_HOTPLUG) && defined(CONFIG_HUGETLBFS) struct work_struct node_work; #endif #ifdef CONFIG_HMEM_REPORTING @@ -98,7 +98,7 @@ struct memory_block; extern struct node *node_devices[]; typedef void (*node_registration_func_t)(struct node *); -#if defined(CONFIG_MEMORY_HOTPLUG_SPARSE) && defined(CONFIG_NUMA) +#if defined(CONFIG_MEMORY_HOTPLUG) && defined(CONFIG_NUMA) void link_mem_sections(int nid, unsigned long start_pfn, unsigned long end_pfn, enum meminit_context context); diff --git a/lib/Kconfig.debug b/lib/Kconfig.debug index 2a9b6dcdac4f..669fee1d26b8 100644 --- a/lib/Kconfig.debug +++ b/lib/Kconfig.debug @@ -877,7 +877,7 @@ config DEBUG_MEMORY_INIT config MEMORY_NOTIFIER_ERROR_INJECT tristate "Memory hotplug notifier error injection module" - depends on MEMORY_HOTPLUG_SPARSE && NOTIFIER_ERROR_INJECTION + depends on MEMORY_HOTPLUG && NOTIFIER_ERROR_INJECTION help This option provides the ability to inject artificial errors to memory hotplug notifier chain callbacks. It is controlled through diff --git a/mm/Kconfig b/mm/Kconfig index b2bf73c90a38..0148a9c4fa2a 100644 --- a/mm/Kconfig +++ b/mm/Kconfig @@ -128,10 +128,6 @@ config MEMORY_HOTPLUG depends on 64BIT || BROKEN select NUMA_KEEP_MEMINFO if NUMA -config MEMORY_HOTPLUG_SPARSE - def_bool y - depends on SPARSEMEM && MEMORY_HOTPLUG - config MEMORY_HOTPLUG_DEFAULT_ONLINE bool "Online the newly added memory blocks by default" depends on MEMORY_HOTPLUG diff --git a/mm/memory_hotplug.c b/mm/memory_hotplug.c index afaae370b8cd..fc07ce7b5842 100644 --- a/mm/memory_hotplug.c +++ b/mm/memory_hotplug.c @@ -220,7 +220,6 @@ static void release_memory_resource(struct resource *res) kfree(res); } -#ifdef CONFIG_MEMORY_HOTPLUG_SPARSE static int check_pfn_span(unsigned long pfn, unsigned long nr_pages, const char *reason) { @@ -1163,7 +1162,6 @@ failed_addition: mem_hotplug_done(); return ret; } -#endif /* CONFIG_MEMORY_HOTPLUG_SPARSE */ static void reset_node_present_pages(pg_data_t *pgdat) { diff --git a/tools/testing/selftests/memory-hotplug/config b/tools/testing/selftests/memory-hotplug/config index a7e8cd5bb265..1eef042a31e1 100644 --- a/tools/testing/selftests/memory-hotplug/config +++ b/tools/testing/selftests/memory-hotplug/config @@ -1,5 +1,4 @@ CONFIG_MEMORY_HOTPLUG=y -CONFIG_MEMORY_HOTPLUG_SPARSE=y CONFIG_NOTIFIER_ERROR_INJECTION=y CONFIG_MEMORY_NOTIFIER_ERROR_INJECT=m CONFIG_MEMORY_HOTREMOVE=y -- cgit v1.2.3 From f39f21b3ddc7fc0f87eb6dc75ddc81b5bbfb7672 Mon Sep 17 00:00:00 2001 From: Marco Elver Date: Fri, 5 Nov 2021 13:45:25 -0700 Subject: stacktrace: move filter_irq_stacks() to kernel/stacktrace.c filter_irq_stacks() has little to do with the stackdepot implementation, except that it is usually used by users (such as KASAN) of stackdepot to reduce the stack trace. However, filter_irq_stacks() itself is not useful without a stack trace as obtained by stack_trace_save() and friends. Therefore, move filter_irq_stacks() to kernel/stacktrace.c, so that new users of filter_irq_stacks() do not have to start depending on STACKDEPOT only for filter_irq_stacks(). Link: https://lkml.kernel.org/r/20210923104803.2620285-1-elver@google.com Signed-off-by: Marco Elver Acked-by: Dmitry Vyukov Cc: Alexander Potapenko Cc: Jann Horn Cc: Aleksandr Nogikh Cc: Taras Madan Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/linux/stackdepot.h | 2 -- include/linux/stacktrace.h | 1 + kernel/stacktrace.c | 30 ++++++++++++++++++++++++++++++ lib/stackdepot.c | 24 ------------------------ 4 files changed, 31 insertions(+), 26 deletions(-) (limited to 'lib') diff --git a/include/linux/stackdepot.h b/include/linux/stackdepot.h index b2f7e7c6ba54..d29860966bc9 100644 --- a/include/linux/stackdepot.h +++ b/include/linux/stackdepot.h @@ -25,8 +25,6 @@ depot_stack_handle_t stack_depot_save(unsigned long *entries, unsigned int stack_depot_fetch(depot_stack_handle_t handle, unsigned long **entries); -unsigned int filter_irq_stacks(unsigned long *entries, unsigned int nr_entries); - #ifdef CONFIG_STACKDEPOT int stack_depot_init(void); #else diff --git a/include/linux/stacktrace.h b/include/linux/stacktrace.h index 9edecb494e9e..bef158815e83 100644 --- a/include/linux/stacktrace.h +++ b/include/linux/stacktrace.h @@ -21,6 +21,7 @@ unsigned int stack_trace_save_tsk(struct task_struct *task, unsigned int stack_trace_save_regs(struct pt_regs *regs, unsigned long *store, unsigned int size, unsigned int skipnr); unsigned int stack_trace_save_user(unsigned long *store, unsigned int size); +unsigned int filter_irq_stacks(unsigned long *entries, unsigned int nr_entries); /* Internal interfaces. Do not use in generic code */ #ifdef CONFIG_ARCH_STACKWALK diff --git a/kernel/stacktrace.c b/kernel/stacktrace.c index 9f8117c7cfdd..9c625257023d 100644 --- a/kernel/stacktrace.c +++ b/kernel/stacktrace.c @@ -13,6 +13,7 @@ #include #include #include +#include /** * stack_trace_print - Print the entries in the stack trace @@ -373,3 +374,32 @@ unsigned int stack_trace_save_user(unsigned long *store, unsigned int size) #endif /* CONFIG_USER_STACKTRACE_SUPPORT */ #endif /* !CONFIG_ARCH_STACKWALK */ + +static inline bool in_irqentry_text(unsigned long ptr) +{ + return (ptr >= (unsigned long)&__irqentry_text_start && + ptr < (unsigned long)&__irqentry_text_end) || + (ptr >= (unsigned long)&__softirqentry_text_start && + ptr < (unsigned long)&__softirqentry_text_end); +} + +/** + * filter_irq_stacks - Find first IRQ stack entry in trace + * @entries: Pointer to stack trace array + * @nr_entries: Number of entries in the storage array + * + * Return: Number of trace entries until IRQ stack starts. + */ +unsigned int filter_irq_stacks(unsigned long *entries, unsigned int nr_entries) +{ + unsigned int i; + + for (i = 0; i < nr_entries; i++) { + if (in_irqentry_text(entries[i])) { + /* Include the irqentry function into the stack. */ + return i + 1; + } + } + return nr_entries; +} +EXPORT_SYMBOL_GPL(filter_irq_stacks); diff --git a/lib/stackdepot.c b/lib/stackdepot.c index bda58597e375..09485dc5bd12 100644 --- a/lib/stackdepot.c +++ b/lib/stackdepot.c @@ -20,7 +20,6 @@ */ #include -#include #include #include #include @@ -371,26 +370,3 @@ depot_stack_handle_t stack_depot_save(unsigned long *entries, return __stack_depot_save(entries, nr_entries, alloc_flags, true); } EXPORT_SYMBOL_GPL(stack_depot_save); - -static inline int in_irqentry_text(unsigned long ptr) -{ - return (ptr >= (unsigned long)&__irqentry_text_start && - ptr < (unsigned long)&__irqentry_text_end) || - (ptr >= (unsigned long)&__softirqentry_text_start && - ptr < (unsigned long)&__softirqentry_text_end); -} - -unsigned int filter_irq_stacks(unsigned long *entries, - unsigned int nr_entries) -{ - unsigned int i; - - for (i = 0; i < nr_entries; i++) { - if (in_irqentry_text(entries[i])) { - /* Include the irqentry function into the stack. */ - return i + 1; - } - } - return nr_entries; -} -EXPORT_SYMBOL_GPL(filter_irq_stacks); -- cgit v1.2.3 From 4f612ed3f748962cbef1316ff3d323e2b9055b6e Mon Sep 17 00:00:00 2001 From: Marco Elver Date: Fri, 5 Nov 2021 13:45:49 -0700 Subject: kfence: default to dynamic branch instead of static keys mode We have observed that on very large machines with newer CPUs, the static key/branch switching delay is on the order of milliseconds. This is due to the required broadcast IPIs, which simply does not scale well to hundreds of CPUs (cores). If done too frequently, this can adversely affect tail latencies of various workloads. One workaround is to increase the sample interval to several seconds, while decreasing sampled allocation coverage, but the problem still exists and could still increase tail latencies. As already noted in the Kconfig help text, there are trade-offs: at lower sample intervals the dynamic branch results in better performance; however, at very large sample intervals, the static keys mode can result in better performance -- careful benchmarking is recommended. Our initial benchmarking showed that with large enough sample intervals and workloads stressing the allocator, the static keys mode was slightly better. Evaluating and observing the possible system-wide side-effects of the static-key-switching induced broadcast IPIs, however, was a blind spot (in particular on large machines with 100s of cores). Therefore, a major downside of the static keys mode is, unfortunately, that it is hard to predict performance on new system architectures and topologies, but also making conclusions about performance of new workloads based on a limited set of benchmarks. Most distributions will simply select the defaults, while targeting a large variety of different workloads and system architectures. As such, the better default is CONFIG_KFENCE_STATIC_KEYS=n, and re-enabling it is only recommended after careful evaluation. For reference, on x86-64 the condition in kfence_alloc() generates exactly 2 instructions in the kmem_cache_alloc() fast-path: | ... | cmpl $0x0,0x1a8021c(%rip) # ffffffff82d560d0 | je ffffffff812d6003 | ... which, given kfence_allocation_gate is infrequently modified, should be well predicted by most CPUs. Link: https://lkml.kernel.org/r/20211019102524.2807208-2-elver@google.com Signed-off-by: Marco Elver Cc: Alexander Potapenko Cc: Dmitry Vyukov Cc: Jann Horn Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- Documentation/dev-tools/kfence.rst | 12 ++++++++---- lib/Kconfig.kfence | 26 +++++++++++++++----------- 2 files changed, 23 insertions(+), 15 deletions(-) (limited to 'lib') diff --git a/Documentation/dev-tools/kfence.rst b/Documentation/dev-tools/kfence.rst index d45f952986ae..ac6b89d1a8c3 100644 --- a/Documentation/dev-tools/kfence.rst +++ b/Documentation/dev-tools/kfence.rst @@ -231,10 +231,14 @@ Guarded allocations are set up based on the sample interval. After expiration of the sample interval, the next allocation through the main allocator (SLAB or SLUB) returns a guarded allocation from the KFENCE object pool (allocation sizes up to PAGE_SIZE are supported). At this point, the timer is reset, and -the next allocation is set up after the expiration of the interval. To "gate" a -KFENCE allocation through the main allocator's fast-path without overhead, -KFENCE relies on static branches via the static keys infrastructure. The static -branch is toggled to redirect the allocation to KFENCE. +the next allocation is set up after the expiration of the interval. + +When using ``CONFIG_KFENCE_STATIC_KEYS=y``, KFENCE allocations are "gated" +through the main allocator's fast-path by relying on static branches via the +static keys infrastructure. The static branch is toggled to redirect the +allocation to KFENCE. Depending on sample interval, target workloads, and +system architecture, this may perform better than the simple dynamic branch. +Careful benchmarking is recommended. KFENCE objects each reside on a dedicated page, at either the left or right page boundaries selected at random. The pages to the left and right of the diff --git a/lib/Kconfig.kfence b/lib/Kconfig.kfence index e641add33947..912f252a41fc 100644 --- a/lib/Kconfig.kfence +++ b/lib/Kconfig.kfence @@ -25,17 +25,6 @@ menuconfig KFENCE if KFENCE -config KFENCE_STATIC_KEYS - bool "Use static keys to set up allocations" - default y - depends on JUMP_LABEL # To ensure performance, require jump labels - help - Use static keys (static branches) to set up KFENCE allocations. Using - static keys is normally recommended, because it avoids a dynamic - branch in the allocator's fast path. However, with very low sample - intervals, or on systems that do not support jump labels, a dynamic - branch may still be an acceptable performance trade-off. - config KFENCE_SAMPLE_INTERVAL int "Default sample interval in milliseconds" default 100 @@ -56,6 +45,21 @@ config KFENCE_NUM_OBJECTS pages are required; with one containing the object and two adjacent ones used as guard pages. +config KFENCE_STATIC_KEYS + bool "Use static keys to set up allocations" if EXPERT + depends on JUMP_LABEL + help + Use static keys (static branches) to set up KFENCE allocations. This + option is only recommended when using very large sample intervals, or + performance has carefully been evaluated with this option. + + Using static keys comes with trade-offs that need to be carefully + evaluated given target workloads and system architectures. Notably, + enabling and disabling static keys invoke IPI broadcasts, the latency + and impact of which is much harder to predict than a dynamic branch. + + Say N if you are unsure. + config KFENCE_STRESS_TEST_FAULTS int "Stress testing of fault handling and error reporting" if EXPERT default 0 -- cgit v1.2.3 From cf30f6a5f0c60ec98a637b836bef6915f602c6ab Mon Sep 17 00:00:00 2001 From: Nick Terrell Date: Fri, 11 Sep 2020 16:49:00 -0700 Subject: lib: zstd: Add kernel-specific API This patch: - Moves `include/linux/zstd.h` -> `include/linux/zstd_lib.h` - Updates modified zstd headers to yearless copyright - Adds a new API in `include/linux/zstd.h` that is functionally equivalent to the in-use subset of the current API. Functions are renamed to avoid symbol collisions with zstd, to make it clear it is not the upstream zstd API, and to follow the kernel style guide. - Updates all callers to use the new API. There are no functional changes in this patch. Since there are no functional change, I felt it was okay to update all the callers in a single patch. Once the API is approved, the callers are mechanically changed. This patch is preparing for the 3rd patch in this series, which updates zstd to version 1.4.10. Since the upstream zstd API is no longer exposed to callers, the update can happen transparently. Signed-off-by: Nick Terrell Tested By: Paul Jones Tested-by: Oleksandr Natalenko Tested-by: Sedat Dilek # LLVM/Clang v13.0.0 on x86-64 Tested-by: Jean-Denis Girard --- crypto/zstd.c | 28 +- fs/btrfs/zstd.c | 68 +-- fs/f2fs/compress.c | 56 +- fs/f2fs/super.c | 2 +- fs/pstore/platform.c | 2 +- fs/squashfs/zstd_wrapper.c | 16 +- include/linux/zstd.h | 1243 ++++++++++---------------------------------- include/linux/zstd_lib.h | 1157 +++++++++++++++++++++++++++++++++++++++++ lib/decompress_unzstd.c | 42 +- lib/zstd/compress.c | 123 +++-- lib/zstd/decompress.c | 112 ++-- 11 files changed, 1691 insertions(+), 1158 deletions(-) create mode 100644 include/linux/zstd_lib.h (limited to 'lib') diff --git a/crypto/zstd.c b/crypto/zstd.c index 1a3309f066f7..154a969c83a8 100644 --- a/crypto/zstd.c +++ b/crypto/zstd.c @@ -18,22 +18,22 @@ #define ZSTD_DEF_LEVEL 3 struct zstd_ctx { - ZSTD_CCtx *cctx; - ZSTD_DCtx *dctx; + zstd_cctx *cctx; + zstd_dctx *dctx; void *cwksp; void *dwksp; }; -static ZSTD_parameters zstd_params(void) +static zstd_parameters zstd_params(void) { - return ZSTD_getParams(ZSTD_DEF_LEVEL, 0, 0); + return zstd_get_params(ZSTD_DEF_LEVEL, 0); } static int zstd_comp_init(struct zstd_ctx *ctx) { int ret = 0; - const ZSTD_parameters params = zstd_params(); - const size_t wksp_size = ZSTD_CCtxWorkspaceBound(params.cParams); + const zstd_parameters params = zstd_params(); + const size_t wksp_size = zstd_cctx_workspace_bound(¶ms.cParams); ctx->cwksp = vzalloc(wksp_size); if (!ctx->cwksp) { @@ -41,7 +41,7 @@ static int zstd_comp_init(struct zstd_ctx *ctx) goto out; } - ctx->cctx = ZSTD_initCCtx(ctx->cwksp, wksp_size); + ctx->cctx = zstd_init_cctx(ctx->cwksp, wksp_size); if (!ctx->cctx) { ret = -EINVAL; goto out_free; @@ -56,7 +56,7 @@ out_free: static int zstd_decomp_init(struct zstd_ctx *ctx) { int ret = 0; - const size_t wksp_size = ZSTD_DCtxWorkspaceBound(); + const size_t wksp_size = zstd_dctx_workspace_bound(); ctx->dwksp = vzalloc(wksp_size); if (!ctx->dwksp) { @@ -64,7 +64,7 @@ static int zstd_decomp_init(struct zstd_ctx *ctx) goto out; } - ctx->dctx = ZSTD_initDCtx(ctx->dwksp, wksp_size); + ctx->dctx = zstd_init_dctx(ctx->dwksp, wksp_size); if (!ctx->dctx) { ret = -EINVAL; goto out_free; @@ -152,10 +152,10 @@ static int __zstd_compress(const u8 *src, unsigned int slen, { size_t out_len; struct zstd_ctx *zctx = ctx; - const ZSTD_parameters params = zstd_params(); + const zstd_parameters params = zstd_params(); - out_len = ZSTD_compressCCtx(zctx->cctx, dst, *dlen, src, slen, params); - if (ZSTD_isError(out_len)) + out_len = zstd_compress_cctx(zctx->cctx, dst, *dlen, src, slen, ¶ms); + if (zstd_is_error(out_len)) return -EINVAL; *dlen = out_len; return 0; @@ -182,8 +182,8 @@ static int __zstd_decompress(const u8 *src, unsigned int slen, size_t out_len; struct zstd_ctx *zctx = ctx; - out_len = ZSTD_decompressDCtx(zctx->dctx, dst, *dlen, src, slen); - if (ZSTD_isError(out_len)) + out_len = zstd_decompress_dctx(zctx->dctx, dst, *dlen, src, slen); + if (zstd_is_error(out_len)) return -EINVAL; *dlen = out_len; return 0; diff --git a/fs/btrfs/zstd.c b/fs/btrfs/zstd.c index f06b68040352..fc42dd0badd7 100644 --- a/fs/btrfs/zstd.c +++ b/fs/btrfs/zstd.c @@ -28,10 +28,10 @@ /* 307s to avoid pathologically clashing with transaction commit */ #define ZSTD_BTRFS_RECLAIM_JIFFIES (307 * HZ) -static ZSTD_parameters zstd_get_btrfs_parameters(unsigned int level, +static zstd_parameters zstd_get_btrfs_parameters(unsigned int level, size_t src_len) { - ZSTD_parameters params = ZSTD_getParams(level, src_len, 0); + zstd_parameters params = zstd_get_params(level, src_len); if (params.cParams.windowLog > ZSTD_BTRFS_MAX_WINDOWLOG) params.cParams.windowLog = ZSTD_BTRFS_MAX_WINDOWLOG; @@ -48,8 +48,8 @@ struct workspace { unsigned long last_used; /* jiffies */ struct list_head list; struct list_head lru_list; - ZSTD_inBuffer in_buf; - ZSTD_outBuffer out_buf; + zstd_in_buffer in_buf; + zstd_out_buffer out_buf; }; /* @@ -155,12 +155,12 @@ static void zstd_calc_ws_mem_sizes(void) unsigned int level; for (level = 1; level <= ZSTD_BTRFS_MAX_LEVEL; level++) { - ZSTD_parameters params = + zstd_parameters params = zstd_get_btrfs_parameters(level, ZSTD_BTRFS_MAX_INPUT); size_t level_size = max_t(size_t, - ZSTD_CStreamWorkspaceBound(params.cParams), - ZSTD_DStreamWorkspaceBound(ZSTD_BTRFS_MAX_INPUT)); + zstd_cstream_workspace_bound(¶ms.cParams), + zstd_dstream_workspace_bound(ZSTD_BTRFS_MAX_INPUT)); max_size = max_t(size_t, max_size, level_size); zstd_ws_mem_sizes[level - 1] = max_size; @@ -371,7 +371,7 @@ int zstd_compress_pages(struct list_head *ws, struct address_space *mapping, unsigned long *total_in, unsigned long *total_out) { struct workspace *workspace = list_entry(ws, struct workspace, list); - ZSTD_CStream *stream; + zstd_cstream *stream; int ret = 0; int nr_pages = 0; struct page *in_page = NULL; /* The current page to read */ @@ -381,7 +381,7 @@ int zstd_compress_pages(struct list_head *ws, struct address_space *mapping, unsigned long len = *total_out; const unsigned long nr_dest_pages = *out_pages; unsigned long max_out = nr_dest_pages * PAGE_SIZE; - ZSTD_parameters params = zstd_get_btrfs_parameters(workspace->req_level, + zstd_parameters params = zstd_get_btrfs_parameters(workspace->req_level, len); *out_pages = 0; @@ -389,10 +389,10 @@ int zstd_compress_pages(struct list_head *ws, struct address_space *mapping, *total_in = 0; /* Initialize the stream */ - stream = ZSTD_initCStream(params, len, workspace->mem, + stream = zstd_init_cstream(¶ms, len, workspace->mem, workspace->size); if (!stream) { - pr_warn("BTRFS: ZSTD_initCStream failed\n"); + pr_warn("BTRFS: zstd_init_cstream failed\n"); ret = -EIO; goto out; } @@ -418,11 +418,11 @@ int zstd_compress_pages(struct list_head *ws, struct address_space *mapping, while (1) { size_t ret2; - ret2 = ZSTD_compressStream(stream, &workspace->out_buf, + ret2 = zstd_compress_stream(stream, &workspace->out_buf, &workspace->in_buf); - if (ZSTD_isError(ret2)) { - pr_debug("BTRFS: ZSTD_compressStream returned %d\n", - ZSTD_getErrorCode(ret2)); + if (zstd_is_error(ret2)) { + pr_debug("BTRFS: zstd_compress_stream returned %d\n", + zstd_get_error_code(ret2)); ret = -EIO; goto out; } @@ -487,10 +487,10 @@ int zstd_compress_pages(struct list_head *ws, struct address_space *mapping, while (1) { size_t ret2; - ret2 = ZSTD_endStream(stream, &workspace->out_buf); - if (ZSTD_isError(ret2)) { - pr_debug("BTRFS: ZSTD_endStream returned %d\n", - ZSTD_getErrorCode(ret2)); + ret2 = zstd_end_stream(stream, &workspace->out_buf); + if (zstd_is_error(ret2)) { + pr_debug("BTRFS: zstd_end_stream returned %d\n", + zstd_get_error_code(ret2)); ret = -EIO; goto out; } @@ -548,17 +548,17 @@ int zstd_decompress_bio(struct list_head *ws, struct compressed_bio *cb) struct workspace *workspace = list_entry(ws, struct workspace, list); struct page **pages_in = cb->compressed_pages; size_t srclen = cb->compressed_len; - ZSTD_DStream *stream; + zstd_dstream *stream; int ret = 0; unsigned long page_in_index = 0; unsigned long total_pages_in = DIV_ROUND_UP(srclen, PAGE_SIZE); unsigned long buf_start; unsigned long total_out = 0; - stream = ZSTD_initDStream( + stream = zstd_init_dstream( ZSTD_BTRFS_MAX_INPUT, workspace->mem, workspace->size); if (!stream) { - pr_debug("BTRFS: ZSTD_initDStream failed\n"); + pr_debug("BTRFS: zstd_init_dstream failed\n"); ret = -EIO; goto done; } @@ -574,11 +574,11 @@ int zstd_decompress_bio(struct list_head *ws, struct compressed_bio *cb) while (1) { size_t ret2; - ret2 = ZSTD_decompressStream(stream, &workspace->out_buf, + ret2 = zstd_decompress_stream(stream, &workspace->out_buf, &workspace->in_buf); - if (ZSTD_isError(ret2)) { - pr_debug("BTRFS: ZSTD_decompressStream returned %d\n", - ZSTD_getErrorCode(ret2)); + if (zstd_is_error(ret2)) { + pr_debug("BTRFS: zstd_decompress_stream returned %d\n", + zstd_get_error_code(ret2)); ret = -EIO; goto done; } @@ -624,16 +624,16 @@ int zstd_decompress(struct list_head *ws, unsigned char *data_in, size_t destlen) { struct workspace *workspace = list_entry(ws, struct workspace, list); - ZSTD_DStream *stream; + zstd_dstream *stream; int ret = 0; size_t ret2; unsigned long total_out = 0; unsigned long pg_offset = 0; - stream = ZSTD_initDStream( + stream = zstd_init_dstream( ZSTD_BTRFS_MAX_INPUT, workspace->mem, workspace->size); if (!stream) { - pr_warn("BTRFS: ZSTD_initDStream failed\n"); + pr_warn("BTRFS: zstd_init_dstream failed\n"); ret = -EIO; goto finish; } @@ -657,15 +657,15 @@ int zstd_decompress(struct list_head *ws, unsigned char *data_in, /* Check if the frame is over and we still need more input */ if (ret2 == 0) { - pr_debug("BTRFS: ZSTD_decompressStream ended early\n"); + pr_debug("BTRFS: zstd_decompress_stream ended early\n"); ret = -EIO; goto finish; } - ret2 = ZSTD_decompressStream(stream, &workspace->out_buf, + ret2 = zstd_decompress_stream(stream, &workspace->out_buf, &workspace->in_buf); - if (ZSTD_isError(ret2)) { - pr_debug("BTRFS: ZSTD_decompressStream returned %d\n", - ZSTD_getErrorCode(ret2)); + if (zstd_is_error(ret2)) { + pr_debug("BTRFS: zstd_decompress_stream returned %d\n", + zstd_get_error_code(ret2)); ret = -EIO; goto finish; } diff --git a/fs/f2fs/compress.c b/fs/f2fs/compress.c index 20a083dc9042..b8d70deddfb7 100644 --- a/fs/f2fs/compress.c +++ b/fs/f2fs/compress.c @@ -336,8 +336,8 @@ static const struct f2fs_compress_ops f2fs_lz4_ops = { static int zstd_init_compress_ctx(struct compress_ctx *cc) { - ZSTD_parameters params; - ZSTD_CStream *stream; + zstd_parameters params; + zstd_cstream *stream; void *workspace; unsigned int workspace_size; unsigned char level = F2FS_I(cc->inode)->i_compress_flag >> @@ -346,17 +346,17 @@ static int zstd_init_compress_ctx(struct compress_ctx *cc) if (!level) level = F2FS_ZSTD_DEFAULT_CLEVEL; - params = ZSTD_getParams(level, cc->rlen, 0); - workspace_size = ZSTD_CStreamWorkspaceBound(params.cParams); + params = zstd_get_params(F2FS_ZSTD_DEFAULT_CLEVEL, cc->rlen); + workspace_size = zstd_cstream_workspace_bound(¶ms.cParams); workspace = f2fs_kvmalloc(F2FS_I_SB(cc->inode), workspace_size, GFP_NOFS); if (!workspace) return -ENOMEM; - stream = ZSTD_initCStream(params, 0, workspace, workspace_size); + stream = zstd_init_cstream(¶ms, 0, workspace, workspace_size); if (!stream) { - printk_ratelimited("%sF2FS-fs (%s): %s ZSTD_initCStream failed\n", + printk_ratelimited("%sF2FS-fs (%s): %s zstd_init_cstream failed\n", KERN_ERR, F2FS_I_SB(cc->inode)->sb->s_id, __func__); kvfree(workspace); @@ -379,9 +379,9 @@ static void zstd_destroy_compress_ctx(struct compress_ctx *cc) static int zstd_compress_pages(struct compress_ctx *cc) { - ZSTD_CStream *stream = cc->private2; - ZSTD_inBuffer inbuf; - ZSTD_outBuffer outbuf; + zstd_cstream *stream = cc->private2; + zstd_in_buffer inbuf; + zstd_out_buffer outbuf; int src_size = cc->rlen; int dst_size = src_size - PAGE_SIZE - COMPRESS_HEADER_SIZE; int ret; @@ -394,19 +394,19 @@ static int zstd_compress_pages(struct compress_ctx *cc) outbuf.dst = cc->cbuf->cdata; outbuf.size = dst_size; - ret = ZSTD_compressStream(stream, &outbuf, &inbuf); - if (ZSTD_isError(ret)) { - printk_ratelimited("%sF2FS-fs (%s): %s ZSTD_compressStream failed, ret: %d\n", + ret = zstd_compress_stream(stream, &outbuf, &inbuf); + if (zstd_is_error(ret)) { + printk_ratelimited("%sF2FS-fs (%s): %s zstd_compress_stream failed, ret: %d\n", KERN_ERR, F2FS_I_SB(cc->inode)->sb->s_id, - __func__, ZSTD_getErrorCode(ret)); + __func__, zstd_get_error_code(ret)); return -EIO; } - ret = ZSTD_endStream(stream, &outbuf); - if (ZSTD_isError(ret)) { - printk_ratelimited("%sF2FS-fs (%s): %s ZSTD_endStream returned %d\n", + ret = zstd_end_stream(stream, &outbuf); + if (zstd_is_error(ret)) { + printk_ratelimited("%sF2FS-fs (%s): %s zstd_end_stream returned %d\n", KERN_ERR, F2FS_I_SB(cc->inode)->sb->s_id, - __func__, ZSTD_getErrorCode(ret)); + __func__, zstd_get_error_code(ret)); return -EIO; } @@ -423,22 +423,22 @@ static int zstd_compress_pages(struct compress_ctx *cc) static int zstd_init_decompress_ctx(struct decompress_io_ctx *dic) { - ZSTD_DStream *stream; + zstd_dstream *stream; void *workspace; unsigned int workspace_size; unsigned int max_window_size = MAX_COMPRESS_WINDOW_SIZE(dic->log_cluster_size); - workspace_size = ZSTD_DStreamWorkspaceBound(max_window_size); + workspace_size = zstd_dstream_workspace_bound(max_window_size); workspace = f2fs_kvmalloc(F2FS_I_SB(dic->inode), workspace_size, GFP_NOFS); if (!workspace) return -ENOMEM; - stream = ZSTD_initDStream(max_window_size, workspace, workspace_size); + stream = zstd_init_dstream(max_window_size, workspace, workspace_size); if (!stream) { - printk_ratelimited("%sF2FS-fs (%s): %s ZSTD_initDStream failed\n", + printk_ratelimited("%sF2FS-fs (%s): %s zstd_init_dstream failed\n", KERN_ERR, F2FS_I_SB(dic->inode)->sb->s_id, __func__); kvfree(workspace); @@ -460,9 +460,9 @@ static void zstd_destroy_decompress_ctx(struct decompress_io_ctx *dic) static int zstd_decompress_pages(struct decompress_io_ctx *dic) { - ZSTD_DStream *stream = dic->private2; - ZSTD_inBuffer inbuf; - ZSTD_outBuffer outbuf; + zstd_dstream *stream = dic->private2; + zstd_in_buffer inbuf; + zstd_out_buffer outbuf; int ret; inbuf.pos = 0; @@ -473,11 +473,11 @@ static int zstd_decompress_pages(struct decompress_io_ctx *dic) outbuf.dst = dic->rbuf; outbuf.size = dic->rlen; - ret = ZSTD_decompressStream(stream, &outbuf, &inbuf); - if (ZSTD_isError(ret)) { - printk_ratelimited("%sF2FS-fs (%s): %s ZSTD_compressStream failed, ret: %d\n", + ret = zstd_decompress_stream(stream, &outbuf, &inbuf); + if (zstd_is_error(ret)) { + printk_ratelimited("%sF2FS-fs (%s): %s zstd_decompress_stream failed, ret: %d\n", KERN_ERR, F2FS_I_SB(dic->inode)->sb->s_id, - __func__, ZSTD_getErrorCode(ret)); + __func__, zstd_get_error_code(ret)); return -EIO; } diff --git a/fs/f2fs/super.c b/fs/f2fs/super.c index cf049a042482..78b64bf5e0f7 100644 --- a/fs/f2fs/super.c +++ b/fs/f2fs/super.c @@ -592,7 +592,7 @@ static int f2fs_set_zstd_level(struct f2fs_sb_info *sbi, const char *str) if (kstrtouint(str + 1, 10, &level)) return -EINVAL; - if (!level || level > ZSTD_maxCLevel()) { + if (!level || level > zstd_max_clevel()) { f2fs_info(sbi, "invalid zstd compress level: %d", level); return -EINVAL; } diff --git a/fs/pstore/platform.c b/fs/pstore/platform.c index b9614db48b1d..f243cb5e6a4f 100644 --- a/fs/pstore/platform.c +++ b/fs/pstore/platform.c @@ -218,7 +218,7 @@ static int zbufsize_842(size_t size) #if IS_ENABLED(CONFIG_PSTORE_ZSTD_COMPRESS) static int zbufsize_zstd(size_t size) { - return ZSTD_compressBound(size); + return zstd_compress_bound(size); } #endif diff --git a/fs/squashfs/zstd_wrapper.c b/fs/squashfs/zstd_wrapper.c index 0015cf8b5582..c40445dbf38c 100644 --- a/fs/squashfs/zstd_wrapper.c +++ b/fs/squashfs/zstd_wrapper.c @@ -34,7 +34,7 @@ static void *zstd_init(struct squashfs_sb_info *msblk, void *buff) goto failed; wksp->window_size = max_t(size_t, msblk->block_size, SQUASHFS_METADATA_SIZE); - wksp->mem_size = ZSTD_DStreamWorkspaceBound(wksp->window_size); + wksp->mem_size = zstd_dstream_workspace_bound(wksp->window_size); wksp->mem = vmalloc(wksp->mem_size); if (wksp->mem == NULL) goto failed; @@ -63,15 +63,15 @@ static int zstd_uncompress(struct squashfs_sb_info *msblk, void *strm, struct squashfs_page_actor *output) { struct workspace *wksp = strm; - ZSTD_DStream *stream; + zstd_dstream *stream; size_t total_out = 0; int error = 0; - ZSTD_inBuffer in_buf = { NULL, 0, 0 }; - ZSTD_outBuffer out_buf = { NULL, 0, 0 }; + zstd_in_buffer in_buf = { NULL, 0, 0 }; + zstd_out_buffer out_buf = { NULL, 0, 0 }; struct bvec_iter_all iter_all = {}; struct bio_vec *bvec = bvec_init_iter_all(&iter_all); - stream = ZSTD_initDStream(wksp->window_size, wksp->mem, wksp->mem_size); + stream = zstd_init_dstream(wksp->window_size, wksp->mem, wksp->mem_size); if (!stream) { ERROR("Failed to initialize zstd decompressor\n"); @@ -116,14 +116,14 @@ static int zstd_uncompress(struct squashfs_sb_info *msblk, void *strm, } total_out -= out_buf.pos; - zstd_err = ZSTD_decompressStream(stream, &out_buf, &in_buf); + zstd_err = zstd_decompress_stream(stream, &out_buf, &in_buf); total_out += out_buf.pos; /* add the additional data produced */ if (zstd_err == 0) break; - if (ZSTD_isError(zstd_err)) { + if (zstd_is_error(zstd_err)) { ERROR("zstd decompression error: %d\n", - (int)ZSTD_getErrorCode(zstd_err)); + (int)zstd_get_error_code(zstd_err)); error = -EIO; break; } diff --git a/include/linux/zstd.h b/include/linux/zstd.h index e87f78c9b19c..9fbc7729b0a0 100644 --- a/include/linux/zstd.h +++ b/include/linux/zstd.h @@ -1,138 +1,96 @@ +/* SPDX-License-Identifier: GPL-2.0-only */ /* - * Copyright (c) 2016-present, Yann Collet, Facebook, Inc. + * Copyright (c) Yann Collet, Facebook, Inc. * All rights reserved. * - * This source code is licensed under the BSD-style license found in the - * LICENSE file in the root directory of https://github.com/facebook/zstd. - * An additional grant of patent rights can be found in the PATENTS file in the - * same directory. - * - * This program is free software; you can redistribute it and/or modify it under - * the terms of the GNU General Public License version 2 as published by the - * Free Software Foundation. This program is dual-licensed; you may select - * either version 2 of the GNU General Public License ("GPL") or BSD license - * ("BSD"). + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of https://github.com/facebook/zstd) and + * the GPLv2 (found in the COPYING file in the root directory of + * https://github.com/facebook/zstd). You may select, at your option, one of the + * above-listed licenses. */ -#ifndef ZSTD_H -#define ZSTD_H +#ifndef LINUX_ZSTD_H +#define LINUX_ZSTD_H -/* ====== Dependency ======*/ -#include /* size_t */ +/** + * This is a kernel-style API that wraps the upstream zstd API, which cannot be + * used directly because the symbols aren't exported. It exposes the minimal + * functionality which is currently required by users of zstd in the kernel. + * Expose extra functions from lib/zstd/zstd.h as needed. + */ +/* ====== Dependency ====== */ +#include +#include -/*-***************************************************************************** - * Introduction +/* ====== Helper Functions ====== */ +/** + * zstd_compress_bound() - maximum compressed size in worst case scenario + * @src_size: The size of the data to compress. * - * zstd, short for Zstandard, is a fast lossless compression algorithm, - * targeting real-time compression scenarios at zlib-level and better - * compression ratios. The zstd compression library provides in-memory - * compression and decompression functions. The library supports compression - * levels from 1 up to ZSTD_maxCLevel() which is 22. Levels >= 20, labeled - * ultra, should be used with caution, as they require more memory. - * Compression can be done in: - * - a single step, reusing a context (described as Explicit memory management) - * - unbounded multiple steps (described as Streaming compression) - * The compression ratio achievable on small data can be highly improved using - * compression with a dictionary in: - * - a single step (described as Simple dictionary API) - * - a single step, reusing a dictionary (described as Fast dictionary API) - ******************************************************************************/ - -/*====== Helper functions ======*/ + * Return: The maximum compressed size in the worst case scenario. + */ +size_t zstd_compress_bound(size_t src_size); /** - * enum ZSTD_ErrorCode - zstd error codes + * zstd_is_error() - tells if a size_t function result is an error code + * @code: The function result to check for error. * - * Functions that return size_t can be checked for errors using ZSTD_isError() - * and the ZSTD_ErrorCode can be extracted using ZSTD_getErrorCode(). + * Return: Non-zero iff the code is an error. + */ +unsigned int zstd_is_error(size_t code); + +/** + * enum zstd_error_code - zstd error codes */ -typedef enum { - ZSTD_error_no_error, - ZSTD_error_GENERIC, - ZSTD_error_prefix_unknown, - ZSTD_error_version_unsupported, - ZSTD_error_parameter_unknown, - ZSTD_error_frameParameter_unsupported, - ZSTD_error_frameParameter_unsupportedBy32bits, - ZSTD_error_frameParameter_windowTooLarge, - ZSTD_error_compressionParameter_unsupported, - ZSTD_error_init_missing, - ZSTD_error_memory_allocation, - ZSTD_error_stage_wrong, - ZSTD_error_dstSize_tooSmall, - ZSTD_error_srcSize_wrong, - ZSTD_error_corruption_detected, - ZSTD_error_checksum_wrong, - ZSTD_error_tableLog_tooLarge, - ZSTD_error_maxSymbolValue_tooLarge, - ZSTD_error_maxSymbolValue_tooSmall, - ZSTD_error_dictionary_corrupted, - ZSTD_error_dictionary_wrong, - ZSTD_error_dictionaryCreation_failed, - ZSTD_error_maxCode -} ZSTD_ErrorCode; +typedef ZSTD_ErrorCode zstd_error_code; /** - * ZSTD_maxCLevel() - maximum compression level available + * zstd_get_error_code() - translates an error function result to an error code + * @code: The function result for which zstd_is_error(code) is true. * - * Return: Maximum compression level available. + * Return: A unique error code for this error. */ -int ZSTD_maxCLevel(void); +zstd_error_code zstd_get_error_code(size_t code); + /** - * ZSTD_compressBound() - maximum compressed size in worst case scenario - * @srcSize: The size of the data to compress. + * zstd_get_error_name() - translates an error function result to a string + * @code: The function result for which zstd_is_error(code) is true. * - * Return: The maximum compressed size in the worst case scenario. + * Return: An error string corresponding to the error code. */ -size_t ZSTD_compressBound(size_t srcSize); +const char *zstd_get_error_name(size_t code); + /** - * ZSTD_isError() - tells if a size_t function result is an error code - * @code: The function result to check for error. + * zstd_min_clevel() - minimum allowed compression level * - * Return: Non-zero iff the code is an error. + * Return: The minimum allowed compression level. */ -static __attribute__((unused)) unsigned int ZSTD_isError(size_t code) -{ - return code > (size_t)-ZSTD_error_maxCode; -} +int zstd_min_clevel(void); + /** - * ZSTD_getErrorCode() - translates an error function result to a ZSTD_ErrorCode - * @functionResult: The result of a function for which ZSTD_isError() is true. + * zstd_max_clevel() - maximum allowed compression level * - * Return: The ZSTD_ErrorCode corresponding to the functionResult or 0 - * if the functionResult isn't an error. + * Return: The maximum allowed compression level. */ -static __attribute__((unused)) ZSTD_ErrorCode ZSTD_getErrorCode( - size_t functionResult) -{ - if (!ZSTD_isError(functionResult)) - return (ZSTD_ErrorCode)0; - return (ZSTD_ErrorCode)(0 - functionResult); -} +int zstd_max_clevel(void); + +/* ====== Parameter Selection ====== */ /** - * enum ZSTD_strategy - zstd compression search strategy + * enum zstd_strategy - zstd compression search strategy * - * From faster to stronger. + * From faster to stronger. See zstd_lib.h. */ -typedef enum { - ZSTD_fast, - ZSTD_dfast, - ZSTD_greedy, - ZSTD_lazy, - ZSTD_lazy2, - ZSTD_btlazy2, - ZSTD_btopt, - ZSTD_btopt2 -} ZSTD_strategy; +typedef ZSTD_strategy zstd_strategy; /** - * struct ZSTD_compressionParameters - zstd compression parameters + * struct zstd_compression_parameters - zstd compression parameters * @windowLog: Log of the largest match distance. Larger means more * compression, and more memory needed during decompression. - * @chainLog: Fully searched segment. Larger means more compression, slower, - * and more memory (useless for fast). + * @chainLog: Fully searched segment. Larger means more compression, + * slower, and more memory (useless for fast). * @hashLog: Dispatch table. Larger means more compression, * slower, and more memory. * @searchLog: Number of searches. Larger means more compression and slower. @@ -141,1017 +99,342 @@ typedef enum { * @targetLength: Acceptable match size for optimal parser (only). Larger means * more compression, and slower. * @strategy: The zstd compression strategy. + * + * See zstd_lib.h. */ -typedef struct { - unsigned int windowLog; - unsigned int chainLog; - unsigned int hashLog; - unsigned int searchLog; - unsigned int searchLength; - unsigned int targetLength; - ZSTD_strategy strategy; -} ZSTD_compressionParameters; +typedef ZSTD_compressionParameters zstd_compression_parameters; /** - * struct ZSTD_frameParameters - zstd frame parameters - * @contentSizeFlag: Controls whether content size will be present in the frame - * header (when known). - * @checksumFlag: Controls whether a 32-bit checksum is generated at the end - * of the frame for error detection. - * @noDictIDFlag: Controls whether dictID will be saved into the frame header - * when using dictionary compression. + * struct zstd_frame_parameters - zstd frame parameters + * @contentSizeFlag: Controls whether content size will be present in the + * frame header (when known). + * @checksumFlag: Controls whether a 32-bit checksum is generated at the + * end of the frame for error detection. + * @noDictIDFlag: Controls whether dictID will be saved into the frame + * header when using dictionary compression. * - * The default value is all fields set to 0. + * The default value is all fields set to 0. See zstd_lib.h. */ -typedef struct { - unsigned int contentSizeFlag; - unsigned int checksumFlag; - unsigned int noDictIDFlag; -} ZSTD_frameParameters; +typedef ZSTD_frameParameters zstd_frame_parameters; /** - * struct ZSTD_parameters - zstd parameters + * struct zstd_parameters - zstd parameters * @cParams: The compression parameters. * @fParams: The frame parameters. */ -typedef struct { - ZSTD_compressionParameters cParams; - ZSTD_frameParameters fParams; -} ZSTD_parameters; +typedef ZSTD_parameters zstd_parameters; /** - * ZSTD_getCParams() - returns ZSTD_compressionParameters for selected level - * @compressionLevel: The compression level from 1 to ZSTD_maxCLevel(). - * @estimatedSrcSize: The estimated source size to compress or 0 if unknown. - * @dictSize: The dictionary size or 0 if a dictionary isn't being used. + * zstd_get_params() - returns zstd_parameters for selected level + * @level: The compression level + * @estimated_src_size: The estimated source size to compress or 0 + * if unknown. * - * Return: The selected ZSTD_compressionParameters. + * Return: The selected zstd_parameters. */ -ZSTD_compressionParameters ZSTD_getCParams(int compressionLevel, - unsigned long long estimatedSrcSize, size_t dictSize); +zstd_parameters zstd_get_params(int level, + unsigned long long estimated_src_size); -/** - * ZSTD_getParams() - returns ZSTD_parameters for selected level - * @compressionLevel: The compression level from 1 to ZSTD_maxCLevel(). - * @estimatedSrcSize: The estimated source size to compress or 0 if unknown. - * @dictSize: The dictionary size or 0 if a dictionary isn't being used. - * - * The same as ZSTD_getCParams() except also selects the default frame - * parameters (all zero). - * - * Return: The selected ZSTD_parameters. - */ -ZSTD_parameters ZSTD_getParams(int compressionLevel, - unsigned long long estimatedSrcSize, size_t dictSize); +/* ====== Single-pass Compression ====== */ -/*-************************************* - * Explicit memory management - **************************************/ +typedef ZSTD_CCtx zstd_cctx; /** - * ZSTD_CCtxWorkspaceBound() - amount of memory needed to initialize a ZSTD_CCtx - * @cParams: The compression parameters to be used for compression. + * zstd_cctx_workspace_bound() - max memory needed to initialize a zstd_cctx + * @parameters: The compression parameters to be used. * * If multiple compression parameters might be used, the caller must call - * ZSTD_CCtxWorkspaceBound() for each set of parameters and use the maximum + * zstd_cctx_workspace_bound() for each set of parameters and use the maximum * size. * - * Return: A lower bound on the size of the workspace that is passed to - * ZSTD_initCCtx(). + * Return: A lower bound on the size of the workspace that is passed to + * zstd_init_cctx(). */ -size_t ZSTD_CCtxWorkspaceBound(ZSTD_compressionParameters cParams); +size_t zstd_cctx_workspace_bound(const zstd_compression_parameters *parameters); /** - * struct ZSTD_CCtx - the zstd compression context - * - * When compressing many times it is recommended to allocate a context just once - * and reuse it for each successive compression operation. - */ -typedef struct ZSTD_CCtx_s ZSTD_CCtx; -/** - * ZSTD_initCCtx() - initialize a zstd compression context - * @workspace: The workspace to emplace the context into. It must outlive - * the returned context. - * @workspaceSize: The size of workspace. Use ZSTD_CCtxWorkspaceBound() to - * determine how large the workspace must be. - * - * Return: A compression context emplaced into workspace. - */ -ZSTD_CCtx *ZSTD_initCCtx(void *workspace, size_t workspaceSize); - -/** - * ZSTD_compressCCtx() - compress src into dst - * @ctx: The context. Must have been initialized with a workspace at - * least as large as ZSTD_CCtxWorkspaceBound(params.cParams). - * @dst: The buffer to compress src into. - * @dstCapacity: The size of the destination buffer. May be any size, but - * ZSTD_compressBound(srcSize) is guaranteed to be large enough. - * @src: The data to compress. - * @srcSize: The size of the data to compress. - * @params: The parameters to use for compression. See ZSTD_getParams(). - * - * Return: The compressed size or an error, which can be checked using - * ZSTD_isError(). - */ -size_t ZSTD_compressCCtx(ZSTD_CCtx *ctx, void *dst, size_t dstCapacity, - const void *src, size_t srcSize, ZSTD_parameters params); - -/** - * ZSTD_DCtxWorkspaceBound() - amount of memory needed to initialize a ZSTD_DCtx - * - * Return: A lower bound on the size of the workspace that is passed to - * ZSTD_initDCtx(). - */ -size_t ZSTD_DCtxWorkspaceBound(void); - -/** - * struct ZSTD_DCtx - the zstd decompression context - * - * When decompressing many times it is recommended to allocate a context just - * once and reuse it for each successive decompression operation. - */ -typedef struct ZSTD_DCtx_s ZSTD_DCtx; -/** - * ZSTD_initDCtx() - initialize a zstd decompression context - * @workspace: The workspace to emplace the context into. It must outlive - * the returned context. - * @workspaceSize: The size of workspace. Use ZSTD_DCtxWorkspaceBound() to - * determine how large the workspace must be. - * - * Return: A decompression context emplaced into workspace. - */ -ZSTD_DCtx *ZSTD_initDCtx(void *workspace, size_t workspaceSize); - -/** - * ZSTD_decompressDCtx() - decompress zstd compressed src into dst - * @ctx: The decompression context. - * @dst: The buffer to decompress src into. - * @dstCapacity: The size of the destination buffer. Must be at least as large - * as the decompressed size. If the caller cannot upper bound the - * decompressed size, then it's better to use the streaming API. - * @src: The zstd compressed data to decompress. Multiple concatenated - * frames and skippable frames are allowed. - * @srcSize: The exact size of the data to decompress. - * - * Return: The decompressed size or an error, which can be checked using - * ZSTD_isError(). - */ -size_t ZSTD_decompressDCtx(ZSTD_DCtx *ctx, void *dst, size_t dstCapacity, - const void *src, size_t srcSize); - -/*-************************ - * Simple dictionary API - **************************/ - -/** - * ZSTD_compress_usingDict() - compress src into dst using a dictionary - * @ctx: The context. Must have been initialized with a workspace at - * least as large as ZSTD_CCtxWorkspaceBound(params.cParams). - * @dst: The buffer to compress src into. - * @dstCapacity: The size of the destination buffer. May be any size, but - * ZSTD_compressBound(srcSize) is guaranteed to be large enough. - * @src: The data to compress. - * @srcSize: The size of the data to compress. - * @dict: The dictionary to use for compression. - * @dictSize: The size of the dictionary. - * @params: The parameters to use for compression. See ZSTD_getParams(). - * - * Compression using a predefined dictionary. The same dictionary must be used - * during decompression. - * - * Return: The compressed size or an error, which can be checked using - * ZSTD_isError(). - */ -size_t ZSTD_compress_usingDict(ZSTD_CCtx *ctx, void *dst, size_t dstCapacity, - const void *src, size_t srcSize, const void *dict, size_t dictSize, - ZSTD_parameters params); - -/** - * ZSTD_decompress_usingDict() - decompress src into dst using a dictionary - * @ctx: The decompression context. - * @dst: The buffer to decompress src into. - * @dstCapacity: The size of the destination buffer. Must be at least as large - * as the decompressed size. If the caller cannot upper bound the - * decompressed size, then it's better to use the streaming API. - * @src: The zstd compressed data to decompress. Multiple concatenated - * frames and skippable frames are allowed. - * @srcSize: The exact size of the data to decompress. - * @dict: The dictionary to use for decompression. The same dictionary - * must've been used to compress the data. - * @dictSize: The size of the dictionary. - * - * Return: The decompressed size or an error, which can be checked using - * ZSTD_isError(). - */ -size_t ZSTD_decompress_usingDict(ZSTD_DCtx *ctx, void *dst, size_t dstCapacity, - const void *src, size_t srcSize, const void *dict, size_t dictSize); - -/*-************************** - * Fast dictionary API - ***************************/ - -/** - * ZSTD_CDictWorkspaceBound() - memory needed to initialize a ZSTD_CDict - * @cParams: The compression parameters to be used for compression. + * zstd_init_cctx() - initialize a zstd compression context + * @workspace: The workspace to emplace the context into. It must outlive + * the returned context. + * @workspace_size: The size of workspace. Use zstd_cctx_workspace_bound() to + * determine how large the workspace must be. * - * Return: A lower bound on the size of the workspace that is passed to - * ZSTD_initCDict(). - */ -size_t ZSTD_CDictWorkspaceBound(ZSTD_compressionParameters cParams); - -/** - * struct ZSTD_CDict - a digested dictionary to be used for compression + * Return: A zstd compression context or NULL on error. */ -typedef struct ZSTD_CDict_s ZSTD_CDict; +zstd_cctx *zstd_init_cctx(void *workspace, size_t workspace_size); /** - * ZSTD_initCDict() - initialize a digested dictionary for compression - * @dictBuffer: The dictionary to digest. The buffer is referenced by the - * ZSTD_CDict so it must outlive the returned ZSTD_CDict. - * @dictSize: The size of the dictionary. - * @params: The parameters to use for compression. See ZSTD_getParams(). - * @workspace: The workspace. It must outlive the returned ZSTD_CDict. - * @workspaceSize: The workspace size. Must be at least - * ZSTD_CDictWorkspaceBound(params.cParams). + * zstd_compress_cctx() - compress src into dst with the initialized parameters + * @cctx: The context. Must have been initialized with zstd_init_cctx(). + * @dst: The buffer to compress src into. + * @dst_capacity: The size of the destination buffer. May be any size, but + * ZSTD_compressBound(srcSize) is guaranteed to be large enough. + * @src: The data to compress. + * @src_size: The size of the data to compress. + * @parameters: The compression parameters to be used. * - * When compressing multiple messages / blocks with the same dictionary it is - * recommended to load it just once. The ZSTD_CDict merely references the - * dictBuffer, so it must outlive the returned ZSTD_CDict. - * - * Return: The digested dictionary emplaced into workspace. + * Return: The compressed size or an error, which can be checked using + * zstd_is_error(). */ -ZSTD_CDict *ZSTD_initCDict(const void *dictBuffer, size_t dictSize, - ZSTD_parameters params, void *workspace, size_t workspaceSize); +size_t zstd_compress_cctx(zstd_cctx *cctx, void *dst, size_t dst_capacity, + const void *src, size_t src_size, const zstd_parameters *parameters); -/** - * ZSTD_compress_usingCDict() - compress src into dst using a ZSTD_CDict - * @ctx: The context. Must have been initialized with a workspace at - * least as large as ZSTD_CCtxWorkspaceBound(cParams) where - * cParams are the compression parameters used to initialize the - * cdict. - * @dst: The buffer to compress src into. - * @dstCapacity: The size of the destination buffer. May be any size, but - * ZSTD_compressBound(srcSize) is guaranteed to be large enough. - * @src: The data to compress. - * @srcSize: The size of the data to compress. - * @cdict: The digested dictionary to use for compression. - * @params: The parameters to use for compression. See ZSTD_getParams(). - * - * Compression using a digested dictionary. The same dictionary must be used - * during decompression. - * - * Return: The compressed size or an error, which can be checked using - * ZSTD_isError(). - */ -size_t ZSTD_compress_usingCDict(ZSTD_CCtx *cctx, void *dst, size_t dstCapacity, - const void *src, size_t srcSize, const ZSTD_CDict *cdict); +/* ====== Single-pass Decompression ====== */ +typedef ZSTD_DCtx zstd_dctx; /** - * ZSTD_DDictWorkspaceBound() - memory needed to initialize a ZSTD_DDict + * zstd_dctx_workspace_bound() - max memory needed to initialize a zstd_dctx * - * Return: A lower bound on the size of the workspace that is passed to - * ZSTD_initDDict(). - */ -size_t ZSTD_DDictWorkspaceBound(void); - -/** - * struct ZSTD_DDict - a digested dictionary to be used for decompression + * Return: A lower bound on the size of the workspace that is passed to + * zstd_init_dctx(). */ -typedef struct ZSTD_DDict_s ZSTD_DDict; +size_t zstd_dctx_workspace_bound(void); /** - * ZSTD_initDDict() - initialize a digested dictionary for decompression - * @dictBuffer: The dictionary to digest. The buffer is referenced by the - * ZSTD_DDict so it must outlive the returned ZSTD_DDict. - * @dictSize: The size of the dictionary. - * @workspace: The workspace. It must outlive the returned ZSTD_DDict. - * @workspaceSize: The workspace size. Must be at least - * ZSTD_DDictWorkspaceBound(). - * - * When decompressing multiple messages / blocks with the same dictionary it is - * recommended to load it just once. The ZSTD_DDict merely references the - * dictBuffer, so it must outlive the returned ZSTD_DDict. + * zstd_init_dctx() - initialize a zstd decompression context + * @workspace: The workspace to emplace the context into. It must outlive + * the returned context. + * @workspace_size: The size of workspace. Use zstd_dctx_workspace_bound() to + * determine how large the workspace must be. * - * Return: The digested dictionary emplaced into workspace. + * Return: A zstd decompression context or NULL on error. */ -ZSTD_DDict *ZSTD_initDDict(const void *dictBuffer, size_t dictSize, - void *workspace, size_t workspaceSize); +zstd_dctx *zstd_init_dctx(void *workspace, size_t workspace_size); /** - * ZSTD_decompress_usingDDict() - decompress src into dst using a ZSTD_DDict - * @ctx: The decompression context. - * @dst: The buffer to decompress src into. - * @dstCapacity: The size of the destination buffer. Must be at least as large - * as the decompressed size. If the caller cannot upper bound the - * decompressed size, then it's better to use the streaming API. - * @src: The zstd compressed data to decompress. Multiple concatenated - * frames and skippable frames are allowed. - * @srcSize: The exact size of the data to decompress. - * @ddict: The digested dictionary to use for decompression. The same - * dictionary must've been used to compress the data. + * zstd_decompress_dctx() - decompress zstd compressed src into dst + * @dctx: The decompression context. + * @dst: The buffer to decompress src into. + * @dst_capacity: The size of the destination buffer. Must be at least as large + * as the decompressed size. If the caller cannot upper bound the + * decompressed size, then it's better to use the streaming API. + * @src: The zstd compressed data to decompress. Multiple concatenated + * frames and skippable frames are allowed. + * @src_size: The exact size of the data to decompress. * - * Return: The decompressed size or an error, which can be checked using - * ZSTD_isError(). + * Return: The decompressed size or an error, which can be checked using + * zstd_is_error(). */ -size_t ZSTD_decompress_usingDDict(ZSTD_DCtx *dctx, void *dst, - size_t dstCapacity, const void *src, size_t srcSize, - const ZSTD_DDict *ddict); +size_t zstd_decompress_dctx(zstd_dctx *dctx, void *dst, size_t dst_capacity, + const void *src, size_t src_size); - -/*-************************** - * Streaming - ***************************/ +/* ====== Streaming Buffers ====== */ /** - * struct ZSTD_inBuffer - input buffer for streaming + * struct zstd_in_buffer - input buffer for streaming * @src: Start of the input buffer. * @size: Size of the input buffer. * @pos: Position where reading stopped. Will be updated. * Necessarily 0 <= pos <= size. + * + * See zstd_lib.h. */ -typedef struct ZSTD_inBuffer_s { - const void *src; - size_t size; - size_t pos; -} ZSTD_inBuffer; +typedef ZSTD_inBuffer zstd_in_buffer; /** - * struct ZSTD_outBuffer - output buffer for streaming + * struct zstd_out_buffer - output buffer for streaming * @dst: Start of the output buffer. * @size: Size of the output buffer. * @pos: Position where writing stopped. Will be updated. * Necessarily 0 <= pos <= size. + * + * See zstd_lib.h. */ -typedef struct ZSTD_outBuffer_s { - void *dst; - size_t size; - size_t pos; -} ZSTD_outBuffer; +typedef ZSTD_outBuffer zstd_out_buffer; +/* ====== Streaming Compression ====== */ - -/*-***************************************************************************** - * Streaming compression - HowTo - * - * A ZSTD_CStream object is required to track streaming operation. - * Use ZSTD_initCStream() to initialize a ZSTD_CStream object. - * ZSTD_CStream objects can be reused multiple times on consecutive compression - * operations. It is recommended to re-use ZSTD_CStream in situations where many - * streaming operations will be achieved consecutively. Use one separate - * ZSTD_CStream per thread for parallel execution. - * - * Use ZSTD_compressStream() repetitively to consume input stream. - * The function will automatically update both `pos` fields. - * Note that it may not consume the entire input, in which case `pos < size`, - * and it's up to the caller to present again remaining data. - * It returns a hint for the preferred number of bytes to use as an input for - * the next function call. - * - * At any moment, it's possible to flush whatever data remains within internal - * buffer, using ZSTD_flushStream(). `output->pos` will be updated. There might - * still be some content left within the internal buffer if `output->size` is - * too small. It returns the number of bytes left in the internal buffer and - * must be called until it returns 0. - * - * ZSTD_endStream() instructs to finish a frame. It will perform a flush and - * write frame epilogue. The epilogue is required for decoders to consider a - * frame completed. Similar to ZSTD_flushStream(), it may not be able to flush - * the full content if `output->size` is too small. In which case, call again - * ZSTD_endStream() to complete the flush. It returns the number of bytes left - * in the internal buffer and must be called until it returns 0. - ******************************************************************************/ +typedef ZSTD_CStream zstd_cstream; /** - * ZSTD_CStreamWorkspaceBound() - memory needed to initialize a ZSTD_CStream - * @cParams: The compression parameters to be used for compression. + * zstd_cstream_workspace_bound() - memory needed to initialize a zstd_cstream + * @cparams: The compression parameters to be used for compression. * * Return: A lower bound on the size of the workspace that is passed to - * ZSTD_initCStream() and ZSTD_initCStream_usingCDict(). - */ -size_t ZSTD_CStreamWorkspaceBound(ZSTD_compressionParameters cParams); - -/** - * struct ZSTD_CStream - the zstd streaming compression context + * zstd_init_cstream(). */ -typedef struct ZSTD_CStream_s ZSTD_CStream; +size_t zstd_cstream_workspace_bound(const zstd_compression_parameters *cparams); -/*===== ZSTD_CStream management functions =====*/ /** - * ZSTD_initCStream() - initialize a zstd streaming compression context - * @params: The zstd compression parameters. - * @pledgedSrcSize: If params.fParams.contentSizeFlag == 1 then the caller must - * pass the source size (zero means empty source). Otherwise, - * the caller may optionally pass the source size, or zero if - * unknown. - * @workspace: The workspace to emplace the context into. It must outlive - * the returned context. - * @workspaceSize: The size of workspace. - * Use ZSTD_CStreamWorkspaceBound(params.cParams) to determine - * how large the workspace must be. + * zstd_init_cstream() - initialize a zstd streaming compression context + * @parameters The zstd parameters to use for compression. + * @pledged_src_size: If params.fParams.contentSizeFlag == 1 then the caller + * must pass the source size (zero means empty source). + * Otherwise, the caller may optionally pass the source + * size, or zero if unknown. + * @workspace: The workspace to emplace the context into. It must outlive + * the returned context. + * @workspace_size: The size of workspace. + * Use zstd_cstream_workspace_bound(params->cparams) to + * determine how large the workspace must be. * - * Return: The zstd streaming compression context. + * Return: The zstd streaming compression context or NULL on error. */ -ZSTD_CStream *ZSTD_initCStream(ZSTD_parameters params, - unsigned long long pledgedSrcSize, void *workspace, - size_t workspaceSize); +zstd_cstream *zstd_init_cstream(const zstd_parameters *parameters, + unsigned long long pledged_src_size, void *workspace, size_t workspace_size); /** - * ZSTD_initCStream_usingCDict() - initialize a streaming compression context - * @cdict: The digested dictionary to use for compression. - * @pledgedSrcSize: Optionally the source size, or zero if unknown. - * @workspace: The workspace to emplace the context into. It must outlive - * the returned context. - * @workspaceSize: The size of workspace. Call ZSTD_CStreamWorkspaceBound() - * with the cParams used to initialize the cdict to determine - * how large the workspace must be. - * - * Return: The zstd streaming compression context. - */ -ZSTD_CStream *ZSTD_initCStream_usingCDict(const ZSTD_CDict *cdict, - unsigned long long pledgedSrcSize, void *workspace, - size_t workspaceSize); - -/*===== Streaming compression functions =====*/ -/** - * ZSTD_resetCStream() - reset the context using parameters from creation - * @zcs: The zstd streaming compression context to reset. - * @pledgedSrcSize: Optionally the source size, or zero if unknown. + * zstd_reset_cstream() - reset the context using parameters from creation + * @cstream: The zstd streaming compression context to reset. + * @pledged_src_size: Optionally the source size, or zero if unknown. * * Resets the context using the parameters from creation. Skips dictionary - * loading, since it can be reused. If `pledgedSrcSize` is non-zero the frame + * loading, since it can be reused. If `pledged_src_size` is non-zero the frame * content size is always written into the frame header. * - * Return: Zero or an error, which can be checked using ZSTD_isError(). + * Return: Zero or an error, which can be checked using + * zstd_is_error(). */ -size_t ZSTD_resetCStream(ZSTD_CStream *zcs, unsigned long long pledgedSrcSize); +size_t zstd_reset_cstream(zstd_cstream *cstream, + unsigned long long pledged_src_size); + /** - * ZSTD_compressStream() - streaming compress some of input into output - * @zcs: The zstd streaming compression context. - * @output: Destination buffer. `output->pos` is updated to indicate how much - * compressed data was written. - * @input: Source buffer. `input->pos` is updated to indicate how much data was - * read. Note that it may not consume the entire input, in which case - * `input->pos < input->size`, and it's up to the caller to present - * remaining data again. + * zstd_compress_stream() - streaming compress some of input into output + * @cstream: The zstd streaming compression context. + * @output: Destination buffer. `output->pos` is updated to indicate how much + * compressed data was written. + * @input: Source buffer. `input->pos` is updated to indicate how much data + * was read. Note that it may not consume the entire input, in which + * case `input->pos < input->size`, and it's up to the caller to + * present remaining data again. * * The `input` and `output` buffers may be any size. Guaranteed to make some * forward progress if `input` and `output` are not empty. * - * Return: A hint for the number of bytes to use as the input for the next - * function call or an error, which can be checked using - * ZSTD_isError(). + * Return: A hint for the number of bytes to use as the input for the next + * function call or an error, which can be checked using + * zstd_is_error(). */ -size_t ZSTD_compressStream(ZSTD_CStream *zcs, ZSTD_outBuffer *output, - ZSTD_inBuffer *input); +size_t zstd_compress_stream(zstd_cstream *cstream, zstd_out_buffer *output, + zstd_in_buffer *input); + /** - * ZSTD_flushStream() - flush internal buffers into output - * @zcs: The zstd streaming compression context. - * @output: Destination buffer. `output->pos` is updated to indicate how much - * compressed data was written. + * zstd_flush_stream() - flush internal buffers into output + * @cstream: The zstd streaming compression context. + * @output: Destination buffer. `output->pos` is updated to indicate how much + * compressed data was written. * - * ZSTD_flushStream() must be called until it returns 0, meaning all the data - * has been flushed. Since ZSTD_flushStream() causes a block to be ended, + * zstd_flush_stream() must be called until it returns 0, meaning all the data + * has been flushed. Since zstd_flush_stream() causes a block to be ended, * calling it too often will degrade the compression ratio. * - * Return: The number of bytes still present within internal buffers or an - * error, which can be checked using ZSTD_isError(). - */ -size_t ZSTD_flushStream(ZSTD_CStream *zcs, ZSTD_outBuffer *output); -/** - * ZSTD_endStream() - flush internal buffers into output and end the frame - * @zcs: The zstd streaming compression context. - * @output: Destination buffer. `output->pos` is updated to indicate how much - * compressed data was written. - * - * ZSTD_endStream() must be called until it returns 0, meaning all the data has - * been flushed and the frame epilogue has been written. - * - * Return: The number of bytes still present within internal buffers or an - * error, which can be checked using ZSTD_isError(). + * Return: The number of bytes still present within internal buffers or an + * error, which can be checked using zstd_is_error(). */ -size_t ZSTD_endStream(ZSTD_CStream *zcs, ZSTD_outBuffer *output); +size_t zstd_flush_stream(zstd_cstream *cstream, zstd_out_buffer *output); /** - * ZSTD_CStreamInSize() - recommended size for the input buffer - * - * Return: The recommended size for the input buffer. - */ -size_t ZSTD_CStreamInSize(void); -/** - * ZSTD_CStreamOutSize() - recommended size for the output buffer + * zstd_end_stream() - flush internal buffers into output and end the frame + * @cstream: The zstd streaming compression context. + * @output: Destination buffer. `output->pos` is updated to indicate how much + * compressed data was written. * - * When the output buffer is at least this large, it is guaranteed to be large - * enough to flush at least one complete compressed block. + * zstd_end_stream() must be called until it returns 0, meaning all the data has + * been flushed and the frame epilogue has been written. * - * Return: The recommended size for the output buffer. + * Return: The number of bytes still present within internal buffers or an + * error, which can be checked using zstd_is_error(). */ -size_t ZSTD_CStreamOutSize(void); +size_t zstd_end_stream(zstd_cstream *cstream, zstd_out_buffer *output); +/* ====== Streaming Decompression ====== */ - -/*-***************************************************************************** - * Streaming decompression - HowTo - * - * A ZSTD_DStream object is required to track streaming operations. - * Use ZSTD_initDStream() to initialize a ZSTD_DStream object. - * ZSTD_DStream objects can be re-used multiple times. - * - * Use ZSTD_decompressStream() repetitively to consume your input. - * The function will update both `pos` fields. - * If `input->pos < input->size`, some input has not been consumed. - * It's up to the caller to present again remaining data. - * If `output->pos < output->size`, decoder has flushed everything it could. - * Returns 0 iff a frame is completely decoded and fully flushed. - * Otherwise it returns a suggested next input size that will never load more - * than the current frame. - ******************************************************************************/ +typedef ZSTD_DStream zstd_dstream; /** - * ZSTD_DStreamWorkspaceBound() - memory needed to initialize a ZSTD_DStream - * @maxWindowSize: The maximum window size allowed for compressed frames. + * zstd_dstream_workspace_bound() - memory needed to initialize a zstd_dstream + * @max_window_size: The maximum window size allowed for compressed frames. * - * Return: A lower bound on the size of the workspace that is passed to - * ZSTD_initDStream() and ZSTD_initDStream_usingDDict(). + * Return: A lower bound on the size of the workspace that is passed + * to zstd_init_dstream(). */ -size_t ZSTD_DStreamWorkspaceBound(size_t maxWindowSize); +size_t zstd_dstream_workspace_bound(size_t max_window_size); /** - * struct ZSTD_DStream - the zstd streaming decompression context - */ -typedef struct ZSTD_DStream_s ZSTD_DStream; -/*===== ZSTD_DStream management functions =====*/ -/** - * ZSTD_initDStream() - initialize a zstd streaming decompression context - * @maxWindowSize: The maximum window size allowed for compressed frames. - * @workspace: The workspace to emplace the context into. It must outlive - * the returned context. - * @workspaceSize: The size of workspace. - * Use ZSTD_DStreamWorkspaceBound(maxWindowSize) to determine - * how large the workspace must be. - * - * Return: The zstd streaming decompression context. - */ -ZSTD_DStream *ZSTD_initDStream(size_t maxWindowSize, void *workspace, - size_t workspaceSize); -/** - * ZSTD_initDStream_usingDDict() - initialize streaming decompression context - * @maxWindowSize: The maximum window size allowed for compressed frames. - * @ddict: The digested dictionary to use for decompression. - * @workspace: The workspace to emplace the context into. It must outlive - * the returned context. - * @workspaceSize: The size of workspace. - * Use ZSTD_DStreamWorkspaceBound(maxWindowSize) to determine - * how large the workspace must be. + * zstd_init_dstream() - initialize a zstd streaming decompression context + * @max_window_size: The maximum window size allowed for compressed frames. + * @workspace: The workspace to emplace the context into. It must outlive + * the returned context. + * @workspaceSize: The size of workspace. + * Use zstd_dstream_workspace_bound(max_window_size) to + * determine how large the workspace must be. * - * Return: The zstd streaming decompression context. + * Return: The zstd streaming decompression context. */ -ZSTD_DStream *ZSTD_initDStream_usingDDict(size_t maxWindowSize, - const ZSTD_DDict *ddict, void *workspace, size_t workspaceSize); +zstd_dstream *zstd_init_dstream(size_t max_window_size, void *workspace, + size_t workspace_size); -/*===== Streaming decompression functions =====*/ /** - * ZSTD_resetDStream() - reset the context using parameters from creation - * @zds: The zstd streaming decompression context to reset. + * zstd_reset_dstream() - reset the context using parameters from creation + * @dstream: The zstd streaming decompression context to reset. * * Resets the context using the parameters from creation. Skips dictionary * loading, since it can be reused. * - * Return: Zero or an error, which can be checked using ZSTD_isError(). + * Return: Zero or an error, which can be checked using zstd_is_error(). */ -size_t ZSTD_resetDStream(ZSTD_DStream *zds); +size_t zstd_reset_dstream(zstd_dstream *dstream); + /** - * ZSTD_decompressStream() - streaming decompress some of input into output - * @zds: The zstd streaming decompression context. - * @output: Destination buffer. `output.pos` is updated to indicate how much - * decompressed data was written. - * @input: Source buffer. `input.pos` is updated to indicate how much data was - * read. Note that it may not consume the entire input, in which case - * `input.pos < input.size`, and it's up to the caller to present - * remaining data again. + * zstd_decompress_stream() - streaming decompress some of input into output + * @dstream: The zstd streaming decompression context. + * @output: Destination buffer. `output.pos` is updated to indicate how much + * decompressed data was written. + * @input: Source buffer. `input.pos` is updated to indicate how much data was + * read. Note that it may not consume the entire input, in which case + * `input.pos < input.size`, and it's up to the caller to present + * remaining data again. * * The `input` and `output` buffers may be any size. Guaranteed to make some * forward progress if `input` and `output` are not empty. - * ZSTD_decompressStream() will not consume the last byte of the frame until + * zstd_decompress_stream() will not consume the last byte of the frame until * the entire frame is flushed. * - * Return: Returns 0 iff a frame is completely decoded and fully flushed. - * Otherwise returns a hint for the number of bytes to use as the input - * for the next function call or an error, which can be checked using - * ZSTD_isError(). The size hint will never load more than the frame. - */ -size_t ZSTD_decompressStream(ZSTD_DStream *zds, ZSTD_outBuffer *output, - ZSTD_inBuffer *input); - -/** - * ZSTD_DStreamInSize() - recommended size for the input buffer - * - * Return: The recommended size for the input buffer. - */ -size_t ZSTD_DStreamInSize(void); -/** - * ZSTD_DStreamOutSize() - recommended size for the output buffer - * - * When the output buffer is at least this large, it is guaranteed to be large - * enough to flush at least one complete decompressed block. - * - * Return: The recommended size for the output buffer. - */ -size_t ZSTD_DStreamOutSize(void); - - -/* --- Constants ---*/ -#define ZSTD_MAGICNUMBER 0xFD2FB528 /* >= v0.8.0 */ -#define ZSTD_MAGIC_SKIPPABLE_START 0x184D2A50U - -#define ZSTD_CONTENTSIZE_UNKNOWN (0ULL - 1) -#define ZSTD_CONTENTSIZE_ERROR (0ULL - 2) - -#define ZSTD_WINDOWLOG_MAX_32 27 -#define ZSTD_WINDOWLOG_MAX_64 27 -#define ZSTD_WINDOWLOG_MAX \ - ((unsigned int)(sizeof(size_t) == 4 \ - ? ZSTD_WINDOWLOG_MAX_32 \ - : ZSTD_WINDOWLOG_MAX_64)) -#define ZSTD_WINDOWLOG_MIN 10 -#define ZSTD_HASHLOG_MAX ZSTD_WINDOWLOG_MAX -#define ZSTD_HASHLOG_MIN 6 -#define ZSTD_CHAINLOG_MAX (ZSTD_WINDOWLOG_MAX+1) -#define ZSTD_CHAINLOG_MIN ZSTD_HASHLOG_MIN -#define ZSTD_HASHLOG3_MAX 17 -#define ZSTD_SEARCHLOG_MAX (ZSTD_WINDOWLOG_MAX-1) -#define ZSTD_SEARCHLOG_MIN 1 -/* only for ZSTD_fast, other strategies are limited to 6 */ -#define ZSTD_SEARCHLENGTH_MAX 7 -/* only for ZSTD_btopt, other strategies are limited to 4 */ -#define ZSTD_SEARCHLENGTH_MIN 3 -#define ZSTD_TARGETLENGTH_MIN 4 -#define ZSTD_TARGETLENGTH_MAX 999 - -/* for static allocation */ -#define ZSTD_FRAMEHEADERSIZE_MAX 18 -#define ZSTD_FRAMEHEADERSIZE_MIN 6 -#define ZSTD_frameHeaderSize_prefix 5 -#define ZSTD_frameHeaderSize_min ZSTD_FRAMEHEADERSIZE_MIN -#define ZSTD_frameHeaderSize_max ZSTD_FRAMEHEADERSIZE_MAX -/* magic number + skippable frame length */ -#define ZSTD_skippableHeaderSize 8 - - -/*-************************************* - * Compressed size functions - **************************************/ - -/** - * ZSTD_findFrameCompressedSize() - returns the size of a compressed frame - * @src: Source buffer. It should point to the start of a zstd encoded frame - * or a skippable frame. - * @srcSize: The size of the source buffer. It must be at least as large as the - * size of the frame. - * - * Return: The compressed size of the frame pointed to by `src` or an error, - * which can be check with ZSTD_isError(). - * Suitable to pass to ZSTD_decompress() or similar functions. - */ -size_t ZSTD_findFrameCompressedSize(const void *src, size_t srcSize); - -/*-************************************* - * Decompressed size functions - **************************************/ -/** - * ZSTD_getFrameContentSize() - returns the content size in a zstd frame header - * @src: It should point to the start of a zstd encoded frame. - * @srcSize: The size of the source buffer. It must be at least as large as the - * frame header. `ZSTD_frameHeaderSize_max` is always large enough. - * - * Return: The frame content size stored in the frame header if known. - * `ZSTD_CONTENTSIZE_UNKNOWN` if the content size isn't stored in the - * frame header. `ZSTD_CONTENTSIZE_ERROR` on invalid input. - */ -unsigned long long ZSTD_getFrameContentSize(const void *src, size_t srcSize); - -/** - * ZSTD_findDecompressedSize() - returns decompressed size of a series of frames - * @src: It should point to the start of a series of zstd encoded and/or - * skippable frames. - * @srcSize: The exact size of the series of frames. - * - * If any zstd encoded frame in the series doesn't have the frame content size - * set, `ZSTD_CONTENTSIZE_UNKNOWN` is returned. But frame content size is always - * set when using ZSTD_compress(). The decompressed size can be very large. - * If the source is untrusted, the decompressed size could be wrong or - * intentionally modified. Always ensure the result fits within the - * application's authorized limits. ZSTD_findDecompressedSize() handles multiple - * frames, and so it must traverse the input to read each frame header. This is - * efficient as most of the data is skipped, however it does mean that all frame - * data must be present and valid. - * - * Return: Decompressed size of all the data contained in the frames if known. - * `ZSTD_CONTENTSIZE_UNKNOWN` if the decompressed size is unknown. - * `ZSTD_CONTENTSIZE_ERROR` if an error occurred. - */ -unsigned long long ZSTD_findDecompressedSize(const void *src, size_t srcSize); - -/*-************************************* - * Advanced compression functions - **************************************/ -/** - * ZSTD_checkCParams() - ensure parameter values remain within authorized range - * @cParams: The zstd compression parameters. - * - * Return: Zero or an error, which can be checked using ZSTD_isError(). - */ -size_t ZSTD_checkCParams(ZSTD_compressionParameters cParams); - -/** - * ZSTD_adjustCParams() - optimize parameters for a given srcSize and dictSize - * @srcSize: Optionally the estimated source size, or zero if unknown. - * @dictSize: Optionally the estimated dictionary size, or zero if unknown. - * - * Return: The optimized parameters. - */ -ZSTD_compressionParameters ZSTD_adjustCParams( - ZSTD_compressionParameters cParams, unsigned long long srcSize, - size_t dictSize); - -/*--- Advanced decompression functions ---*/ - -/** - * ZSTD_isFrame() - returns true iff the buffer starts with a valid frame - * @buffer: The source buffer to check. - * @size: The size of the source buffer, must be at least 4 bytes. - * - * Return: True iff the buffer starts with a zstd or skippable frame identifier. - */ -unsigned int ZSTD_isFrame(const void *buffer, size_t size); - -/** - * ZSTD_getDictID_fromDict() - returns the dictionary id stored in a dictionary - * @dict: The dictionary buffer. - * @dictSize: The size of the dictionary buffer. - * - * Return: The dictionary id stored within the dictionary or 0 if the - * dictionary is not a zstd dictionary. If it returns 0 the - * dictionary can still be loaded as a content-only dictionary. + * Return: Returns 0 iff a frame is completely decoded and fully flushed. + * Otherwise returns a hint for the number of bytes to use as the + * input for the next function call or an error, which can be checked + * using zstd_is_error(). The size hint will never load more than the + * frame. */ -unsigned int ZSTD_getDictID_fromDict(const void *dict, size_t dictSize); +size_t zstd_decompress_stream(zstd_dstream *dstream, zstd_out_buffer *output, + zstd_in_buffer *input); -/** - * ZSTD_getDictID_fromDDict() - returns the dictionary id stored in a ZSTD_DDict - * @ddict: The ddict to find the id of. - * - * Return: The dictionary id stored within `ddict` or 0 if the dictionary is not - * a zstd dictionary. If it returns 0 `ddict` will be loaded as a - * content-only dictionary. - */ -unsigned int ZSTD_getDictID_fromDDict(const ZSTD_DDict *ddict); +/* ====== Frame Inspection Functions ====== */ /** - * ZSTD_getDictID_fromFrame() - returns the dictionary id stored in a zstd frame - * @src: Source buffer. It must be a zstd encoded frame. - * @srcSize: The size of the source buffer. It must be at least as large as the - * frame header. `ZSTD_frameHeaderSize_max` is always large enough. + * zstd_find_frame_compressed_size() - returns the size of a compressed frame + * @src: Source buffer. It should point to the start of a zstd encoded + * frame or a skippable frame. + * @src_size: The size of the source buffer. It must be at least as large as the + * size of the frame. * - * Return: The dictionary id required to decompress the frame stored within - * `src` or 0 if the dictionary id could not be decoded. It can return - * 0 if the frame does not require a dictionary, the dictionary id - * wasn't stored in the frame, `src` is not a zstd frame, or `srcSize` - * is too small. + * Return: The compressed size of the frame pointed to by `src` or an error, + * which can be check with zstd_is_error(). + * Suitable to pass to ZSTD_decompress() or similar functions. */ -unsigned int ZSTD_getDictID_fromFrame(const void *src, size_t srcSize); +size_t zstd_find_frame_compressed_size(const void *src, size_t src_size); /** - * struct ZSTD_frameParams - zstd frame parameters stored in the frame header + * struct zstd_frame_params - zstd frame parameters stored in the frame header * @frameContentSize: The frame content size, or 0 if not present. * @windowSize: The window size, or 0 if the frame is a skippable frame. * @dictID: The dictionary id, or 0 if not present. * @checksumFlag: Whether a checksum was used. */ -typedef struct { - unsigned long long frameContentSize; - unsigned int windowSize; - unsigned int dictID; - unsigned int checksumFlag; -} ZSTD_frameParams; +typedef ZSTD_frameParams zstd_frame_header; /** - * ZSTD_getFrameParams() - extracts parameters from a zstd or skippable frame - * @fparamsPtr: On success the frame parameters are written here. - * @src: The source buffer. It must point to a zstd or skippable frame. - * @srcSize: The size of the source buffer. `ZSTD_frameHeaderSize_max` is - * always large enough to succeed. + * zstd_get_frame_header() - extracts parameters from a zstd or skippable frame + * @params: On success the frame parameters are written here. + * @src: The source buffer. It must point to a zstd or skippable frame. + * @src_size: The size of the source buffer. * - * Return: 0 on success. If more data is required it returns how many bytes - * must be provided to make forward progress. Otherwise it returns - * an error, which can be checked using ZSTD_isError(). + * Return: 0 on success. If more data is required it returns how many bytes + * must be provided to make forward progress. Otherwise it returns + * an error, which can be checked using zstd_is_error(). */ -size_t ZSTD_getFrameParams(ZSTD_frameParams *fparamsPtr, const void *src, - size_t srcSize); - -/*-***************************************************************************** - * Buffer-less and synchronous inner streaming functions - * - * This is an advanced API, giving full control over buffer management, for - * users which need direct control over memory. - * But it's also a complex one, with many restrictions (documented below). - * Prefer using normal streaming API for an easier experience - ******************************************************************************/ - -/*-***************************************************************************** - * Buffer-less streaming compression (synchronous mode) - * - * A ZSTD_CCtx object is required to track streaming operations. - * Use ZSTD_initCCtx() to initialize a context. - * ZSTD_CCtx object can be re-used multiple times within successive compression - * operations. - * - * Start by initializing a context. - * Use ZSTD_compressBegin(), or ZSTD_compressBegin_usingDict() for dictionary - * compression, - * or ZSTD_compressBegin_advanced(), for finer parameter control. - * It's also possible to duplicate a reference context which has already been - * initialized, using ZSTD_copyCCtx() - * - * Then, consume your input using ZSTD_compressContinue(). - * There are some important considerations to keep in mind when using this - * advanced function : - * - ZSTD_compressContinue() has no internal buffer. It uses externally provided - * buffer only. - * - Interface is synchronous : input is consumed entirely and produce 1+ - * (or more) compressed blocks. - * - Caller must ensure there is enough space in `dst` to store compressed data - * under worst case scenario. Worst case evaluation is provided by - * ZSTD_compressBound(). - * ZSTD_compressContinue() doesn't guarantee recover after a failed - * compression. - * - ZSTD_compressContinue() presumes prior input ***is still accessible and - * unmodified*** (up to maximum distance size, see WindowLog). - * It remembers all previous contiguous blocks, plus one separated memory - * segment (which can itself consists of multiple contiguous blocks) - * - ZSTD_compressContinue() detects that prior input has been overwritten when - * `src` buffer overlaps. In which case, it will "discard" the relevant memory - * section from its history. - * - * Finish a frame with ZSTD_compressEnd(), which will write the last block(s) - * and optional checksum. It's possible to use srcSize==0, in which case, it - * will write a final empty block to end the frame. Without last block mark, - * frames will be considered unfinished (corrupted) by decoders. - * - * `ZSTD_CCtx` object can be re-used (ZSTD_compressBegin()) to compress some new - * frame. - ******************************************************************************/ - -/*===== Buffer-less streaming compression functions =====*/ -size_t ZSTD_compressBegin(ZSTD_CCtx *cctx, int compressionLevel); -size_t ZSTD_compressBegin_usingDict(ZSTD_CCtx *cctx, const void *dict, - size_t dictSize, int compressionLevel); -size_t ZSTD_compressBegin_advanced(ZSTD_CCtx *cctx, const void *dict, - size_t dictSize, ZSTD_parameters params, - unsigned long long pledgedSrcSize); -size_t ZSTD_copyCCtx(ZSTD_CCtx *cctx, const ZSTD_CCtx *preparedCCtx, - unsigned long long pledgedSrcSize); -size_t ZSTD_compressBegin_usingCDict(ZSTD_CCtx *cctx, const ZSTD_CDict *cdict, - unsigned long long pledgedSrcSize); -size_t ZSTD_compressContinue(ZSTD_CCtx *cctx, void *dst, size_t dstCapacity, - const void *src, size_t srcSize); -size_t ZSTD_compressEnd(ZSTD_CCtx *cctx, void *dst, size_t dstCapacity, - const void *src, size_t srcSize); - - - -/*-***************************************************************************** - * Buffer-less streaming decompression (synchronous mode) - * - * A ZSTD_DCtx object is required to track streaming operations. - * Use ZSTD_initDCtx() to initialize a context. - * A ZSTD_DCtx object can be re-used multiple times. - * - * First typical operation is to retrieve frame parameters, using - * ZSTD_getFrameParams(). It fills a ZSTD_frameParams structure which provide - * important information to correctly decode the frame, such as the minimum - * rolling buffer size to allocate to decompress data (`windowSize`), and the - * dictionary ID used. - * Note: content size is optional, it may not be present. 0 means unknown. - * Note that these values could be wrong, either because of data malformation, - * or because an attacker is spoofing deliberate false information. As a - * consequence, check that values remain within valid application range, - * especially `windowSize`, before allocation. Each application can set its own - * limit, depending on local restrictions. For extended interoperability, it is - * recommended to support at least 8 MB. - * Frame parameters are extracted from the beginning of the compressed frame. - * Data fragment must be large enough to ensure successful decoding, typically - * `ZSTD_frameHeaderSize_max` bytes. - * Result: 0: successful decoding, the `ZSTD_frameParams` structure is filled. - * >0: `srcSize` is too small, provide at least this many bytes. - * errorCode, which can be tested using ZSTD_isError(). - * - * Start decompression, with ZSTD_decompressBegin() or - * ZSTD_decompressBegin_usingDict(). Alternatively, you can copy a prepared - * context, using ZSTD_copyDCtx(). - * - * Then use ZSTD_nextSrcSizeToDecompress() and ZSTD_decompressContinue() - * alternatively. - * ZSTD_nextSrcSizeToDecompress() tells how many bytes to provide as 'srcSize' - * to ZSTD_decompressContinue(). - * ZSTD_decompressContinue() requires this _exact_ amount of bytes, or it will - * fail. - * - * The result of ZSTD_decompressContinue() is the number of bytes regenerated - * within 'dst' (necessarily <= dstCapacity). It can be zero, which is not an - * error; it just means ZSTD_decompressContinue() has decoded some metadata - * item. It can also be an error code, which can be tested with ZSTD_isError(). - * - * ZSTD_decompressContinue() needs previous data blocks during decompression, up - * to `windowSize`. They should preferably be located contiguously, prior to - * current block. Alternatively, a round buffer of sufficient size is also - * possible. Sufficient size is determined by frame parameters. - * ZSTD_decompressContinue() is very sensitive to contiguity, if 2 blocks don't - * follow each other, make sure that either the compressor breaks contiguity at - * the same place, or that previous contiguous segment is large enough to - * properly handle maximum back-reference. - * - * A frame is fully decoded when ZSTD_nextSrcSizeToDecompress() returns zero. - * Context can then be reset to start a new decompression. - * - * Note: it's possible to know if next input to present is a header or a block, - * using ZSTD_nextInputType(). This information is not required to properly - * decode a frame. - * - * == Special case: skippable frames == - * - * Skippable frames allow integration of user-defined data into a flow of - * concatenated frames. Skippable frames will be ignored (skipped) by a - * decompressor. The format of skippable frames is as follows: - * a) Skippable frame ID - 4 Bytes, Little endian format, any value from - * 0x184D2A50 to 0x184D2A5F - * b) Frame Size - 4 Bytes, Little endian format, unsigned 32-bits - * c) Frame Content - any content (User Data) of length equal to Frame Size - * For skippable frames ZSTD_decompressContinue() always returns 0. - * For skippable frames ZSTD_getFrameParams() returns fparamsPtr->windowLog==0 - * what means that a frame is skippable. - * Note: If fparamsPtr->frameContentSize==0, it is ambiguous: the frame might - * actually be a zstd encoded frame with no content. For purposes of - * decompression, it is valid in both cases to skip the frame using - * ZSTD_findFrameCompressedSize() to find its size in bytes. - * It also returns frame size as fparamsPtr->frameContentSize. - ******************************************************************************/ - -/*===== Buffer-less streaming decompression functions =====*/ -size_t ZSTD_decompressBegin(ZSTD_DCtx *dctx); -size_t ZSTD_decompressBegin_usingDict(ZSTD_DCtx *dctx, const void *dict, - size_t dictSize); -void ZSTD_copyDCtx(ZSTD_DCtx *dctx, const ZSTD_DCtx *preparedDCtx); -size_t ZSTD_nextSrcSizeToDecompress(ZSTD_DCtx *dctx); -size_t ZSTD_decompressContinue(ZSTD_DCtx *dctx, void *dst, size_t dstCapacity, - const void *src, size_t srcSize); -typedef enum { - ZSTDnit_frameHeader, - ZSTDnit_blockHeader, - ZSTDnit_block, - ZSTDnit_lastBlock, - ZSTDnit_checksum, - ZSTDnit_skippableFrame -} ZSTD_nextInputType_e; -ZSTD_nextInputType_e ZSTD_nextInputType(ZSTD_DCtx *dctx); - -/*-***************************************************************************** - * Block functions - * - * Block functions produce and decode raw zstd blocks, without frame metadata. - * Frame metadata cost is typically ~18 bytes, which can be non-negligible for - * very small blocks (< 100 bytes). User will have to take in charge required - * information to regenerate data, such as compressed and content sizes. - * - * A few rules to respect: - * - Compressing and decompressing require a context structure - * + Use ZSTD_initCCtx() and ZSTD_initDCtx() - * - It is necessary to init context before starting - * + compression : ZSTD_compressBegin() - * + decompression : ZSTD_decompressBegin() - * + variants _usingDict() are also allowed - * + copyCCtx() and copyDCtx() work too - * - Block size is limited, it must be <= ZSTD_getBlockSizeMax() - * + If you need to compress more, cut data into multiple blocks - * + Consider using the regular ZSTD_compress() instead, as frame metadata - * costs become negligible when source size is large. - * - When a block is considered not compressible enough, ZSTD_compressBlock() - * result will be zero. In which case, nothing is produced into `dst`. - * + User must test for such outcome and deal directly with uncompressed data - * + ZSTD_decompressBlock() doesn't accept uncompressed data as input!!! - * + In case of multiple successive blocks, decoder must be informed of - * uncompressed block existence to follow proper history. Use - * ZSTD_insertBlock() in such a case. - ******************************************************************************/ - -/* Define for static allocation */ -#define ZSTD_BLOCKSIZE_ABSOLUTEMAX (128 * 1024) -/*===== Raw zstd block functions =====*/ -size_t ZSTD_getBlockSizeMax(ZSTD_CCtx *cctx); -size_t ZSTD_compressBlock(ZSTD_CCtx *cctx, void *dst, size_t dstCapacity, - const void *src, size_t srcSize); -size_t ZSTD_decompressBlock(ZSTD_DCtx *dctx, void *dst, size_t dstCapacity, - const void *src, size_t srcSize); -size_t ZSTD_insertBlock(ZSTD_DCtx *dctx, const void *blockStart, - size_t blockSize); +size_t zstd_get_frame_header(zstd_frame_header *params, const void *src, + size_t src_size); -#endif /* ZSTD_H */ +#endif /* LINUX_ZSTD_H */ diff --git a/include/linux/zstd_lib.h b/include/linux/zstd_lib.h new file mode 100644 index 000000000000..13151c34f725 --- /dev/null +++ b/include/linux/zstd_lib.h @@ -0,0 +1,1157 @@ +/* + * Copyright (c) Yann Collet, Facebook, Inc. + * All rights reserved. + * + * This source code is licensed under the BSD-style license found in the + * LICENSE file in the root directory of https://github.com/facebook/zstd. + * An additional grant of patent rights can be found in the PATENTS file in the + * same directory. + * + * This program is free software; you can redistribute it and/or modify it under + * the terms of the GNU General Public License version 2 as published by the + * Free Software Foundation. This program is dual-licensed; you may select + * either version 2 of the GNU General Public License ("GPL") or BSD license + * ("BSD"). + */ + +#ifndef ZSTD_H +#define ZSTD_H + +/* ====== Dependency ======*/ +#include /* size_t */ + + +/*-***************************************************************************** + * Introduction + * + * zstd, short for Zstandard, is a fast lossless compression algorithm, + * targeting real-time compression scenarios at zlib-level and better + * compression ratios. The zstd compression library provides in-memory + * compression and decompression functions. The library supports compression + * levels from 1 up to ZSTD_maxCLevel() which is 22. Levels >= 20, labeled + * ultra, should be used with caution, as they require more memory. + * Compression can be done in: + * - a single step, reusing a context (described as Explicit memory management) + * - unbounded multiple steps (described as Streaming compression) + * The compression ratio achievable on small data can be highly improved using + * compression with a dictionary in: + * - a single step (described as Simple dictionary API) + * - a single step, reusing a dictionary (described as Fast dictionary API) + ******************************************************************************/ + +/*====== Helper functions ======*/ + +/** + * enum ZSTD_ErrorCode - zstd error codes + * + * Functions that return size_t can be checked for errors using ZSTD_isError() + * and the ZSTD_ErrorCode can be extracted using ZSTD_getErrorCode(). + */ +typedef enum { + ZSTD_error_no_error, + ZSTD_error_GENERIC, + ZSTD_error_prefix_unknown, + ZSTD_error_version_unsupported, + ZSTD_error_parameter_unknown, + ZSTD_error_frameParameter_unsupported, + ZSTD_error_frameParameter_unsupportedBy32bits, + ZSTD_error_frameParameter_windowTooLarge, + ZSTD_error_compressionParameter_unsupported, + ZSTD_error_init_missing, + ZSTD_error_memory_allocation, + ZSTD_error_stage_wrong, + ZSTD_error_dstSize_tooSmall, + ZSTD_error_srcSize_wrong, + ZSTD_error_corruption_detected, + ZSTD_error_checksum_wrong, + ZSTD_error_tableLog_tooLarge, + ZSTD_error_maxSymbolValue_tooLarge, + ZSTD_error_maxSymbolValue_tooSmall, + ZSTD_error_dictionary_corrupted, + ZSTD_error_dictionary_wrong, + ZSTD_error_dictionaryCreation_failed, + ZSTD_error_maxCode +} ZSTD_ErrorCode; + +/** + * ZSTD_maxCLevel() - maximum compression level available + * + * Return: Maximum compression level available. + */ +int ZSTD_maxCLevel(void); +/** + * ZSTD_compressBound() - maximum compressed size in worst case scenario + * @srcSize: The size of the data to compress. + * + * Return: The maximum compressed size in the worst case scenario. + */ +size_t ZSTD_compressBound(size_t srcSize); +/** + * ZSTD_isError() - tells if a size_t function result is an error code + * @code: The function result to check for error. + * + * Return: Non-zero iff the code is an error. + */ +static __attribute__((unused)) unsigned int ZSTD_isError(size_t code) +{ + return code > (size_t)-ZSTD_error_maxCode; +} +/** + * ZSTD_getErrorCode() - translates an error function result to a ZSTD_ErrorCode + * @functionResult: The result of a function for which ZSTD_isError() is true. + * + * Return: The ZSTD_ErrorCode corresponding to the functionResult or 0 + * if the functionResult isn't an error. + */ +static __attribute__((unused)) ZSTD_ErrorCode ZSTD_getErrorCode( + size_t functionResult) +{ + if (!ZSTD_isError(functionResult)) + return (ZSTD_ErrorCode)0; + return (ZSTD_ErrorCode)(0 - functionResult); +} + +/** + * enum ZSTD_strategy - zstd compression search strategy + * + * From faster to stronger. + */ +typedef enum { + ZSTD_fast, + ZSTD_dfast, + ZSTD_greedy, + ZSTD_lazy, + ZSTD_lazy2, + ZSTD_btlazy2, + ZSTD_btopt, + ZSTD_btopt2 +} ZSTD_strategy; + +/** + * struct ZSTD_compressionParameters - zstd compression parameters + * @windowLog: Log of the largest match distance. Larger means more + * compression, and more memory needed during decompression. + * @chainLog: Fully searched segment. Larger means more compression, slower, + * and more memory (useless for fast). + * @hashLog: Dispatch table. Larger means more compression, + * slower, and more memory. + * @searchLog: Number of searches. Larger means more compression and slower. + * @searchLength: Match length searched. Larger means faster decompression, + * sometimes less compression. + * @targetLength: Acceptable match size for optimal parser (only). Larger means + * more compression, and slower. + * @strategy: The zstd compression strategy. + */ +typedef struct { + unsigned int windowLog; + unsigned int chainLog; + unsigned int hashLog; + unsigned int searchLog; + unsigned int searchLength; + unsigned int targetLength; + ZSTD_strategy strategy; +} ZSTD_compressionParameters; + +/** + * struct ZSTD_frameParameters - zstd frame parameters + * @contentSizeFlag: Controls whether content size will be present in the frame + * header (when known). + * @checksumFlag: Controls whether a 32-bit checksum is generated at the end + * of the frame for error detection. + * @noDictIDFlag: Controls whether dictID will be saved into the frame header + * when using dictionary compression. + * + * The default value is all fields set to 0. + */ +typedef struct { + unsigned int contentSizeFlag; + unsigned int checksumFlag; + unsigned int noDictIDFlag; +} ZSTD_frameParameters; + +/** + * struct ZSTD_parameters - zstd parameters + * @cParams: The compression parameters. + * @fParams: The frame parameters. + */ +typedef struct { + ZSTD_compressionParameters cParams; + ZSTD_frameParameters fParams; +} ZSTD_parameters; + +/** + * ZSTD_getCParams() - returns ZSTD_compressionParameters for selected level + * @compressionLevel: The compression level from 1 to ZSTD_maxCLevel(). + * @estimatedSrcSize: The estimated source size to compress or 0 if unknown. + * @dictSize: The dictionary size or 0 if a dictionary isn't being used. + * + * Return: The selected ZSTD_compressionParameters. + */ +ZSTD_compressionParameters ZSTD_getCParams(int compressionLevel, + unsigned long long estimatedSrcSize, size_t dictSize); + +/** + * ZSTD_getParams() - returns ZSTD_parameters for selected level + * @compressionLevel: The compression level from 1 to ZSTD_maxCLevel(). + * @estimatedSrcSize: The estimated source size to compress or 0 if unknown. + * @dictSize: The dictionary size or 0 if a dictionary isn't being used. + * + * The same as ZSTD_getCParams() except also selects the default frame + * parameters (all zero). + * + * Return: The selected ZSTD_parameters. + */ +ZSTD_parameters ZSTD_getParams(int compressionLevel, + unsigned long long estimatedSrcSize, size_t dictSize); + +/*-************************************* + * Explicit memory management + **************************************/ + +/** + * ZSTD_CCtxWorkspaceBound() - amount of memory needed to initialize a ZSTD_CCtx + * @cParams: The compression parameters to be used for compression. + * + * If multiple compression parameters might be used, the caller must call + * ZSTD_CCtxWorkspaceBound() for each set of parameters and use the maximum + * size. + * + * Return: A lower bound on the size of the workspace that is passed to + * ZSTD_initCCtx(). + */ +size_t ZSTD_CCtxWorkspaceBound(ZSTD_compressionParameters cParams); + +/** + * struct ZSTD_CCtx - the zstd compression context + * + * When compressing many times it is recommended to allocate a context just once + * and reuse it for each successive compression operation. + */ +typedef struct ZSTD_CCtx_s ZSTD_CCtx; +/** + * ZSTD_initCCtx() - initialize a zstd compression context + * @workspace: The workspace to emplace the context into. It must outlive + * the returned context. + * @workspaceSize: The size of workspace. Use ZSTD_CCtxWorkspaceBound() to + * determine how large the workspace must be. + * + * Return: A compression context emplaced into workspace. + */ +ZSTD_CCtx *ZSTD_initCCtx(void *workspace, size_t workspaceSize); + +/** + * ZSTD_compressCCtx() - compress src into dst + * @ctx: The context. Must have been initialized with a workspace at + * least as large as ZSTD_CCtxWorkspaceBound(params.cParams). + * @dst: The buffer to compress src into. + * @dstCapacity: The size of the destination buffer. May be any size, but + * ZSTD_compressBound(srcSize) is guaranteed to be large enough. + * @src: The data to compress. + * @srcSize: The size of the data to compress. + * @params: The parameters to use for compression. See ZSTD_getParams(). + * + * Return: The compressed size or an error, which can be checked using + * ZSTD_isError(). + */ +size_t ZSTD_compressCCtx(ZSTD_CCtx *ctx, void *dst, size_t dstCapacity, + const void *src, size_t srcSize, ZSTD_parameters params); + +/** + * ZSTD_DCtxWorkspaceBound() - amount of memory needed to initialize a ZSTD_DCtx + * + * Return: A lower bound on the size of the workspace that is passed to + * ZSTD_initDCtx(). + */ +size_t ZSTD_DCtxWorkspaceBound(void); + +/** + * struct ZSTD_DCtx - the zstd decompression context + * + * When decompressing many times it is recommended to allocate a context just + * once and reuse it for each successive decompression operation. + */ +typedef struct ZSTD_DCtx_s ZSTD_DCtx; +/** + * ZSTD_initDCtx() - initialize a zstd decompression context + * @workspace: The workspace to emplace the context into. It must outlive + * the returned context. + * @workspaceSize: The size of workspace. Use ZSTD_DCtxWorkspaceBound() to + * determine how large the workspace must be. + * + * Return: A decompression context emplaced into workspace. + */ +ZSTD_DCtx *ZSTD_initDCtx(void *workspace, size_t workspaceSize); + +/** + * ZSTD_decompressDCtx() - decompress zstd compressed src into dst + * @ctx: The decompression context. + * @dst: The buffer to decompress src into. + * @dstCapacity: The size of the destination buffer. Must be at least as large + * as the decompressed size. If the caller cannot upper bound the + * decompressed size, then it's better to use the streaming API. + * @src: The zstd compressed data to decompress. Multiple concatenated + * frames and skippable frames are allowed. + * @srcSize: The exact size of the data to decompress. + * + * Return: The decompressed size or an error, which can be checked using + * ZSTD_isError(). + */ +size_t ZSTD_decompressDCtx(ZSTD_DCtx *ctx, void *dst, size_t dstCapacity, + const void *src, size_t srcSize); + +/*-************************ + * Simple dictionary API + **************************/ + +/** + * ZSTD_compress_usingDict() - compress src into dst using a dictionary + * @ctx: The context. Must have been initialized with a workspace at + * least as large as ZSTD_CCtxWorkspaceBound(params.cParams). + * @dst: The buffer to compress src into. + * @dstCapacity: The size of the destination buffer. May be any size, but + * ZSTD_compressBound(srcSize) is guaranteed to be large enough. + * @src: The data to compress. + * @srcSize: The size of the data to compress. + * @dict: The dictionary to use for compression. + * @dictSize: The size of the dictionary. + * @params: The parameters to use for compression. See ZSTD_getParams(). + * + * Compression using a predefined dictionary. The same dictionary must be used + * during decompression. + * + * Return: The compressed size or an error, which can be checked using + * ZSTD_isError(). + */ +size_t ZSTD_compress_usingDict(ZSTD_CCtx *ctx, void *dst, size_t dstCapacity, + const void *src, size_t srcSize, const void *dict, size_t dictSize, + ZSTD_parameters params); + +/** + * ZSTD_decompress_usingDict() - decompress src into dst using a dictionary + * @ctx: The decompression context. + * @dst: The buffer to decompress src into. + * @dstCapacity: The size of the destination buffer. Must be at least as large + * as the decompressed size. If the caller cannot upper bound the + * decompressed size, then it's better to use the streaming API. + * @src: The zstd compressed data to decompress. Multiple concatenated + * frames and skippable frames are allowed. + * @srcSize: The exact size of the data to decompress. + * @dict: The dictionary to use for decompression. The same dictionary + * must've been used to compress the data. + * @dictSize: The size of the dictionary. + * + * Return: The decompressed size or an error, which can be checked using + * ZSTD_isError(). + */ +size_t ZSTD_decompress_usingDict(ZSTD_DCtx *ctx, void *dst, size_t dstCapacity, + const void *src, size_t srcSize, const void *dict, size_t dictSize); + +/*-************************** + * Fast dictionary API + ***************************/ + +/** + * ZSTD_CDictWorkspaceBound() - memory needed to initialize a ZSTD_CDict + * @cParams: The compression parameters to be used for compression. + * + * Return: A lower bound on the size of the workspace that is passed to + * ZSTD_initCDict(). + */ +size_t ZSTD_CDictWorkspaceBound(ZSTD_compressionParameters cParams); + +/** + * struct ZSTD_CDict - a digested dictionary to be used for compression + */ +typedef struct ZSTD_CDict_s ZSTD_CDict; + +/** + * ZSTD_initCDict() - initialize a digested dictionary for compression + * @dictBuffer: The dictionary to digest. The buffer is referenced by the + * ZSTD_CDict so it must outlive the returned ZSTD_CDict. + * @dictSize: The size of the dictionary. + * @params: The parameters to use for compression. See ZSTD_getParams(). + * @workspace: The workspace. It must outlive the returned ZSTD_CDict. + * @workspaceSize: The workspace size. Must be at least + * ZSTD_CDictWorkspaceBound(params.cParams). + * + * When compressing multiple messages / blocks with the same dictionary it is + * recommended to load it just once. The ZSTD_CDict merely references the + * dictBuffer, so it must outlive the returned ZSTD_CDict. + * + * Return: The digested dictionary emplaced into workspace. + */ +ZSTD_CDict *ZSTD_initCDict(const void *dictBuffer, size_t dictSize, + ZSTD_parameters params, void *workspace, size_t workspaceSize); + +/** + * ZSTD_compress_usingCDict() - compress src into dst using a ZSTD_CDict + * @ctx: The context. Must have been initialized with a workspace at + * least as large as ZSTD_CCtxWorkspaceBound(cParams) where + * cParams are the compression parameters used to initialize the + * cdict. + * @dst: The buffer to compress src into. + * @dstCapacity: The size of the destination buffer. May be any size, but + * ZSTD_compressBound(srcSize) is guaranteed to be large enough. + * @src: The data to compress. + * @srcSize: The size of the data to compress. + * @cdict: The digested dictionary to use for compression. + * @params: The parameters to use for compression. See ZSTD_getParams(). + * + * Compression using a digested dictionary. The same dictionary must be used + * during decompression. + * + * Return: The compressed size or an error, which can be checked using + * ZSTD_isError(). + */ +size_t ZSTD_compress_usingCDict(ZSTD_CCtx *cctx, void *dst, size_t dstCapacity, + const void *src, size_t srcSize, const ZSTD_CDict *cdict); + + +/** + * ZSTD_DDictWorkspaceBound() - memory needed to initialize a ZSTD_DDict + * + * Return: A lower bound on the size of the workspace that is passed to + * ZSTD_initDDict(). + */ +size_t ZSTD_DDictWorkspaceBound(void); + +/** + * struct ZSTD_DDict - a digested dictionary to be used for decompression + */ +typedef struct ZSTD_DDict_s ZSTD_DDict; + +/** + * ZSTD_initDDict() - initialize a digested dictionary for decompression + * @dictBuffer: The dictionary to digest. The buffer is referenced by the + * ZSTD_DDict so it must outlive the returned ZSTD_DDict. + * @dictSize: The size of the dictionary. + * @workspace: The workspace. It must outlive the returned ZSTD_DDict. + * @workspaceSize: The workspace size. Must be at least + * ZSTD_DDictWorkspaceBound(). + * + * When decompressing multiple messages / blocks with the same dictionary it is + * recommended to load it just once. The ZSTD_DDict merely references the + * dictBuffer, so it must outlive the returned ZSTD_DDict. + * + * Return: The digested dictionary emplaced into workspace. + */ +ZSTD_DDict *ZSTD_initDDict(const void *dictBuffer, size_t dictSize, + void *workspace, size_t workspaceSize); + +/** + * ZSTD_decompress_usingDDict() - decompress src into dst using a ZSTD_DDict + * @ctx: The decompression context. + * @dst: The buffer to decompress src into. + * @dstCapacity: The size of the destination buffer. Must be at least as large + * as the decompressed size. If the caller cannot upper bound the + * decompressed size, then it's better to use the streaming API. + * @src: The zstd compressed data to decompress. Multiple concatenated + * frames and skippable frames are allowed. + * @srcSize: The exact size of the data to decompress. + * @ddict: The digested dictionary to use for decompression. The same + * dictionary must've been used to compress the data. + * + * Return: The decompressed size or an error, which can be checked using + * ZSTD_isError(). + */ +size_t ZSTD_decompress_usingDDict(ZSTD_DCtx *dctx, void *dst, + size_t dstCapacity, const void *src, size_t srcSize, + const ZSTD_DDict *ddict); + + +/*-************************** + * Streaming + ***************************/ + +/** + * struct ZSTD_inBuffer - input buffer for streaming + * @src: Start of the input buffer. + * @size: Size of the input buffer. + * @pos: Position where reading stopped. Will be updated. + * Necessarily 0 <= pos <= size. + */ +typedef struct ZSTD_inBuffer_s { + const void *src; + size_t size; + size_t pos; +} ZSTD_inBuffer; + +/** + * struct ZSTD_outBuffer - output buffer for streaming + * @dst: Start of the output buffer. + * @size: Size of the output buffer. + * @pos: Position where writing stopped. Will be updated. + * Necessarily 0 <= pos <= size. + */ +typedef struct ZSTD_outBuffer_s { + void *dst; + size_t size; + size_t pos; +} ZSTD_outBuffer; + + + +/*-***************************************************************************** + * Streaming compression - HowTo + * + * A ZSTD_CStream object is required to track streaming operation. + * Use ZSTD_initCStream() to initialize a ZSTD_CStream object. + * ZSTD_CStream objects can be reused multiple times on consecutive compression + * operations. It is recommended to re-use ZSTD_CStream in situations where many + * streaming operations will be achieved consecutively. Use one separate + * ZSTD_CStream per thread for parallel execution. + * + * Use ZSTD_compressStream() repetitively to consume input stream. + * The function will automatically update both `pos` fields. + * Note that it may not consume the entire input, in which case `pos < size`, + * and it's up to the caller to present again remaining data. + * It returns a hint for the preferred number of bytes to use as an input for + * the next function call. + * + * At any moment, it's possible to flush whatever data remains within internal + * buffer, using ZSTD_flushStream(). `output->pos` will be updated. There might + * still be some content left within the internal buffer if `output->size` is + * too small. It returns the number of bytes left in the internal buffer and + * must be called until it returns 0. + * + * ZSTD_endStream() instructs to finish a frame. It will perform a flush and + * write frame epilogue. The epilogue is required for decoders to consider a + * frame completed. Similar to ZSTD_flushStream(), it may not be able to flush + * the full content if `output->size` is too small. In which case, call again + * ZSTD_endStream() to complete the flush. It returns the number of bytes left + * in the internal buffer and must be called until it returns 0. + ******************************************************************************/ + +/** + * ZSTD_CStreamWorkspaceBound() - memory needed to initialize a ZSTD_CStream + * @cParams: The compression parameters to be used for compression. + * + * Return: A lower bound on the size of the workspace that is passed to + * ZSTD_initCStream() and ZSTD_initCStream_usingCDict(). + */ +size_t ZSTD_CStreamWorkspaceBound(ZSTD_compressionParameters cParams); + +/** + * struct ZSTD_CStream - the zstd streaming compression context + */ +typedef struct ZSTD_CStream_s ZSTD_CStream; + +/*===== ZSTD_CStream management functions =====*/ +/** + * ZSTD_initCStream() - initialize a zstd streaming compression context + * @params: The zstd compression parameters. + * @pledgedSrcSize: If params.fParams.contentSizeFlag == 1 then the caller must + * pass the source size (zero means empty source). Otherwise, + * the caller may optionally pass the source size, or zero if + * unknown. + * @workspace: The workspace to emplace the context into. It must outlive + * the returned context. + * @workspaceSize: The size of workspace. + * Use ZSTD_CStreamWorkspaceBound(params.cParams) to determine + * how large the workspace must be. + * + * Return: The zstd streaming compression context. + */ +ZSTD_CStream *ZSTD_initCStream(ZSTD_parameters params, + unsigned long long pledgedSrcSize, void *workspace, + size_t workspaceSize); + +/** + * ZSTD_initCStream_usingCDict() - initialize a streaming compression context + * @cdict: The digested dictionary to use for compression. + * @pledgedSrcSize: Optionally the source size, or zero if unknown. + * @workspace: The workspace to emplace the context into. It must outlive + * the returned context. + * @workspaceSize: The size of workspace. Call ZSTD_CStreamWorkspaceBound() + * with the cParams used to initialize the cdict to determine + * how large the workspace must be. + * + * Return: The zstd streaming compression context. + */ +ZSTD_CStream *ZSTD_initCStream_usingCDict(const ZSTD_CDict *cdict, + unsigned long long pledgedSrcSize, void *workspace, + size_t workspaceSize); + +/*===== Streaming compression functions =====*/ +/** + * ZSTD_resetCStream() - reset the context using parameters from creation + * @zcs: The zstd streaming compression context to reset. + * @pledgedSrcSize: Optionally the source size, or zero if unknown. + * + * Resets the context using the parameters from creation. Skips dictionary + * loading, since it can be reused. If `pledgedSrcSize` is non-zero the frame + * content size is always written into the frame header. + * + * Return: Zero or an error, which can be checked using ZSTD_isError(). + */ +size_t ZSTD_resetCStream(ZSTD_CStream *zcs, unsigned long long pledgedSrcSize); +/** + * ZSTD_compressStream() - streaming compress some of input into output + * @zcs: The zstd streaming compression context. + * @output: Destination buffer. `output->pos` is updated to indicate how much + * compressed data was written. + * @input: Source buffer. `input->pos` is updated to indicate how much data was + * read. Note that it may not consume the entire input, in which case + * `input->pos < input->size`, and it's up to the caller to present + * remaining data again. + * + * The `input` and `output` buffers may be any size. Guaranteed to make some + * forward progress if `input` and `output` are not empty. + * + * Return: A hint for the number of bytes to use as the input for the next + * function call or an error, which can be checked using + * ZSTD_isError(). + */ +size_t ZSTD_compressStream(ZSTD_CStream *zcs, ZSTD_outBuffer *output, + ZSTD_inBuffer *input); +/** + * ZSTD_flushStream() - flush internal buffers into output + * @zcs: The zstd streaming compression context. + * @output: Destination buffer. `output->pos` is updated to indicate how much + * compressed data was written. + * + * ZSTD_flushStream() must be called until it returns 0, meaning all the data + * has been flushed. Since ZSTD_flushStream() causes a block to be ended, + * calling it too often will degrade the compression ratio. + * + * Return: The number of bytes still present within internal buffers or an + * error, which can be checked using ZSTD_isError(). + */ +size_t ZSTD_flushStream(ZSTD_CStream *zcs, ZSTD_outBuffer *output); +/** + * ZSTD_endStream() - flush internal buffers into output and end the frame + * @zcs: The zstd streaming compression context. + * @output: Destination buffer. `output->pos` is updated to indicate how much + * compressed data was written. + * + * ZSTD_endStream() must be called until it returns 0, meaning all the data has + * been flushed and the frame epilogue has been written. + * + * Return: The number of bytes still present within internal buffers or an + * error, which can be checked using ZSTD_isError(). + */ +size_t ZSTD_endStream(ZSTD_CStream *zcs, ZSTD_outBuffer *output); + +/** + * ZSTD_CStreamInSize() - recommended size for the input buffer + * + * Return: The recommended size for the input buffer. + */ +size_t ZSTD_CStreamInSize(void); +/** + * ZSTD_CStreamOutSize() - recommended size for the output buffer + * + * When the output buffer is at least this large, it is guaranteed to be large + * enough to flush at least one complete compressed block. + * + * Return: The recommended size for the output buffer. + */ +size_t ZSTD_CStreamOutSize(void); + + + +/*-***************************************************************************** + * Streaming decompression - HowTo + * + * A ZSTD_DStream object is required to track streaming operations. + * Use ZSTD_initDStream() to initialize a ZSTD_DStream object. + * ZSTD_DStream objects can be re-used multiple times. + * + * Use ZSTD_decompressStream() repetitively to consume your input. + * The function will update both `pos` fields. + * If `input->pos < input->size`, some input has not been consumed. + * It's up to the caller to present again remaining data. + * If `output->pos < output->size`, decoder has flushed everything it could. + * Returns 0 iff a frame is completely decoded and fully flushed. + * Otherwise it returns a suggested next input size that will never load more + * than the current frame. + ******************************************************************************/ + +/** + * ZSTD_DStreamWorkspaceBound() - memory needed to initialize a ZSTD_DStream + * @maxWindowSize: The maximum window size allowed for compressed frames. + * + * Return: A lower bound on the size of the workspace that is passed to + * ZSTD_initDStream() and ZSTD_initDStream_usingDDict(). + */ +size_t ZSTD_DStreamWorkspaceBound(size_t maxWindowSize); + +/** + * struct ZSTD_DStream - the zstd streaming decompression context + */ +typedef struct ZSTD_DStream_s ZSTD_DStream; +/*===== ZSTD_DStream management functions =====*/ +/** + * ZSTD_initDStream() - initialize a zstd streaming decompression context + * @maxWindowSize: The maximum window size allowed for compressed frames. + * @workspace: The workspace to emplace the context into. It must outlive + * the returned context. + * @workspaceSize: The size of workspace. + * Use ZSTD_DStreamWorkspaceBound(maxWindowSize) to determine + * how large the workspace must be. + * + * Return: The zstd streaming decompression context. + */ +ZSTD_DStream *ZSTD_initDStream(size_t maxWindowSize, void *workspace, + size_t workspaceSize); +/** + * ZSTD_initDStream_usingDDict() - initialize streaming decompression context + * @maxWindowSize: The maximum window size allowed for compressed frames. + * @ddict: The digested dictionary to use for decompression. + * @workspace: The workspace to emplace the context into. It must outlive + * the returned context. + * @workspaceSize: The size of workspace. + * Use ZSTD_DStreamWorkspaceBound(maxWindowSize) to determine + * how large the workspace must be. + * + * Return: The zstd streaming decompression context. + */ +ZSTD_DStream *ZSTD_initDStream_usingDDict(size_t maxWindowSize, + const ZSTD_DDict *ddict, void *workspace, size_t workspaceSize); + +/*===== Streaming decompression functions =====*/ +/** + * ZSTD_resetDStream() - reset the context using parameters from creation + * @zds: The zstd streaming decompression context to reset. + * + * Resets the context using the parameters from creation. Skips dictionary + * loading, since it can be reused. + * + * Return: Zero or an error, which can be checked using ZSTD_isError(). + */ +size_t ZSTD_resetDStream(ZSTD_DStream *zds); +/** + * ZSTD_decompressStream() - streaming decompress some of input into output + * @zds: The zstd streaming decompression context. + * @output: Destination buffer. `output.pos` is updated to indicate how much + * decompressed data was written. + * @input: Source buffer. `input.pos` is updated to indicate how much data was + * read. Note that it may not consume the entire input, in which case + * `input.pos < input.size`, and it's up to the caller to present + * remaining data again. + * + * The `input` and `output` buffers may be any size. Guaranteed to make some + * forward progress if `input` and `output` are not empty. + * ZSTD_decompressStream() will not consume the last byte of the frame until + * the entire frame is flushed. + * + * Return: Returns 0 iff a frame is completely decoded and fully flushed. + * Otherwise returns a hint for the number of bytes to use as the input + * for the next function call or an error, which can be checked using + * ZSTD_isError(). The size hint will never load more than the frame. + */ +size_t ZSTD_decompressStream(ZSTD_DStream *zds, ZSTD_outBuffer *output, + ZSTD_inBuffer *input); + +/** + * ZSTD_DStreamInSize() - recommended size for the input buffer + * + * Return: The recommended size for the input buffer. + */ +size_t ZSTD_DStreamInSize(void); +/** + * ZSTD_DStreamOutSize() - recommended size for the output buffer + * + * When the output buffer is at least this large, it is guaranteed to be large + * enough to flush at least one complete decompressed block. + * + * Return: The recommended size for the output buffer. + */ +size_t ZSTD_DStreamOutSize(void); + + +/* --- Constants ---*/ +#define ZSTD_MAGICNUMBER 0xFD2FB528 /* >= v0.8.0 */ +#define ZSTD_MAGIC_SKIPPABLE_START 0x184D2A50U + +#define ZSTD_CONTENTSIZE_UNKNOWN (0ULL - 1) +#define ZSTD_CONTENTSIZE_ERROR (0ULL - 2) + +#define ZSTD_WINDOWLOG_MAX_32 27 +#define ZSTD_WINDOWLOG_MAX_64 27 +#define ZSTD_WINDOWLOG_MAX \ + ((unsigned int)(sizeof(size_t) == 4 \ + ? ZSTD_WINDOWLOG_MAX_32 \ + : ZSTD_WINDOWLOG_MAX_64)) +#define ZSTD_WINDOWLOG_MIN 10 +#define ZSTD_HASHLOG_MAX ZSTD_WINDOWLOG_MAX +#define ZSTD_HASHLOG_MIN 6 +#define ZSTD_CHAINLOG_MAX (ZSTD_WINDOWLOG_MAX+1) +#define ZSTD_CHAINLOG_MIN ZSTD_HASHLOG_MIN +#define ZSTD_HASHLOG3_MAX 17 +#define ZSTD_SEARCHLOG_MAX (ZSTD_WINDOWLOG_MAX-1) +#define ZSTD_SEARCHLOG_MIN 1 +/* only for ZSTD_fast, other strategies are limited to 6 */ +#define ZSTD_SEARCHLENGTH_MAX 7 +/* only for ZSTD_btopt, other strategies are limited to 4 */ +#define ZSTD_SEARCHLENGTH_MIN 3 +#define ZSTD_TARGETLENGTH_MIN 4 +#define ZSTD_TARGETLENGTH_MAX 999 + +/* for static allocation */ +#define ZSTD_FRAMEHEADERSIZE_MAX 18 +#define ZSTD_FRAMEHEADERSIZE_MIN 6 +#define ZSTD_frameHeaderSize_prefix 5 +#define ZSTD_frameHeaderSize_min ZSTD_FRAMEHEADERSIZE_MIN +#define ZSTD_frameHeaderSize_max ZSTD_FRAMEHEADERSIZE_MAX +/* magic number + skippable frame length */ +#define ZSTD_skippableHeaderSize 8 + + +/*-************************************* + * Compressed size functions + **************************************/ + +/** + * ZSTD_findFrameCompressedSize() - returns the size of a compressed frame + * @src: Source buffer. It should point to the start of a zstd encoded frame + * or a skippable frame. + * @srcSize: The size of the source buffer. It must be at least as large as the + * size of the frame. + * + * Return: The compressed size of the frame pointed to by `src` or an error, + * which can be check with ZSTD_isError(). + * Suitable to pass to ZSTD_decompress() or similar functions. + */ +size_t ZSTD_findFrameCompressedSize(const void *src, size_t srcSize); + +/*-************************************* + * Decompressed size functions + **************************************/ +/** + * ZSTD_getFrameContentSize() - returns the content size in a zstd frame header + * @src: It should point to the start of a zstd encoded frame. + * @srcSize: The size of the source buffer. It must be at least as large as the + * frame header. `ZSTD_frameHeaderSize_max` is always large enough. + * + * Return: The frame content size stored in the frame header if known. + * `ZSTD_CONTENTSIZE_UNKNOWN` if the content size isn't stored in the + * frame header. `ZSTD_CONTENTSIZE_ERROR` on invalid input. + */ +unsigned long long ZSTD_getFrameContentSize(const void *src, size_t srcSize); + +/** + * ZSTD_findDecompressedSize() - returns decompressed size of a series of frames + * @src: It should point to the start of a series of zstd encoded and/or + * skippable frames. + * @srcSize: The exact size of the series of frames. + * + * If any zstd encoded frame in the series doesn't have the frame content size + * set, `ZSTD_CONTENTSIZE_UNKNOWN` is returned. But frame content size is always + * set when using ZSTD_compress(). The decompressed size can be very large. + * If the source is untrusted, the decompressed size could be wrong or + * intentionally modified. Always ensure the result fits within the + * application's authorized limits. ZSTD_findDecompressedSize() handles multiple + * frames, and so it must traverse the input to read each frame header. This is + * efficient as most of the data is skipped, however it does mean that all frame + * data must be present and valid. + * + * Return: Decompressed size of all the data contained in the frames if known. + * `ZSTD_CONTENTSIZE_UNKNOWN` if the decompressed size is unknown. + * `ZSTD_CONTENTSIZE_ERROR` if an error occurred. + */ +unsigned long long ZSTD_findDecompressedSize(const void *src, size_t srcSize); + +/*-************************************* + * Advanced compression functions + **************************************/ +/** + * ZSTD_checkCParams() - ensure parameter values remain within authorized range + * @cParams: The zstd compression parameters. + * + * Return: Zero or an error, which can be checked using ZSTD_isError(). + */ +size_t ZSTD_checkCParams(ZSTD_compressionParameters cParams); + +/** + * ZSTD_adjustCParams() - optimize parameters for a given srcSize and dictSize + * @srcSize: Optionally the estimated source size, or zero if unknown. + * @dictSize: Optionally the estimated dictionary size, or zero if unknown. + * + * Return: The optimized parameters. + */ +ZSTD_compressionParameters ZSTD_adjustCParams( + ZSTD_compressionParameters cParams, unsigned long long srcSize, + size_t dictSize); + +/*--- Advanced decompression functions ---*/ + +/** + * ZSTD_isFrame() - returns true iff the buffer starts with a valid frame + * @buffer: The source buffer to check. + * @size: The size of the source buffer, must be at least 4 bytes. + * + * Return: True iff the buffer starts with a zstd or skippable frame identifier. + */ +unsigned int ZSTD_isFrame(const void *buffer, size_t size); + +/** + * ZSTD_getDictID_fromDict() - returns the dictionary id stored in a dictionary + * @dict: The dictionary buffer. + * @dictSize: The size of the dictionary buffer. + * + * Return: The dictionary id stored within the dictionary or 0 if the + * dictionary is not a zstd dictionary. If it returns 0 the + * dictionary can still be loaded as a content-only dictionary. + */ +unsigned int ZSTD_getDictID_fromDict(const void *dict, size_t dictSize); + +/** + * ZSTD_getDictID_fromDDict() - returns the dictionary id stored in a ZSTD_DDict + * @ddict: The ddict to find the id of. + * + * Return: The dictionary id stored within `ddict` or 0 if the dictionary is not + * a zstd dictionary. If it returns 0 `ddict` will be loaded as a + * content-only dictionary. + */ +unsigned int ZSTD_getDictID_fromDDict(const ZSTD_DDict *ddict); + +/** + * ZSTD_getDictID_fromFrame() - returns the dictionary id stored in a zstd frame + * @src: Source buffer. It must be a zstd encoded frame. + * @srcSize: The size of the source buffer. It must be at least as large as the + * frame header. `ZSTD_frameHeaderSize_max` is always large enough. + * + * Return: The dictionary id required to decompress the frame stored within + * `src` or 0 if the dictionary id could not be decoded. It can return + * 0 if the frame does not require a dictionary, the dictionary id + * wasn't stored in the frame, `src` is not a zstd frame, or `srcSize` + * is too small. + */ +unsigned int ZSTD_getDictID_fromFrame(const void *src, size_t srcSize); + +/** + * struct ZSTD_frameParams - zstd frame parameters stored in the frame header + * @frameContentSize: The frame content size, or 0 if not present. + * @windowSize: The window size, or 0 if the frame is a skippable frame. + * @dictID: The dictionary id, or 0 if not present. + * @checksumFlag: Whether a checksum was used. + */ +typedef struct { + unsigned long long frameContentSize; + unsigned int windowSize; + unsigned int dictID; + unsigned int checksumFlag; +} ZSTD_frameParams; + +/** + * ZSTD_getFrameParams() - extracts parameters from a zstd or skippable frame + * @fparamsPtr: On success the frame parameters are written here. + * @src: The source buffer. It must point to a zstd or skippable frame. + * @srcSize: The size of the source buffer. `ZSTD_frameHeaderSize_max` is + * always large enough to succeed. + * + * Return: 0 on success. If more data is required it returns how many bytes + * must be provided to make forward progress. Otherwise it returns + * an error, which can be checked using ZSTD_isError(). + */ +size_t ZSTD_getFrameParams(ZSTD_frameParams *fparamsPtr, const void *src, + size_t srcSize); + +/*-***************************************************************************** + * Buffer-less and synchronous inner streaming functions + * + * This is an advanced API, giving full control over buffer management, for + * users which need direct control over memory. + * But it's also a complex one, with many restrictions (documented below). + * Prefer using normal streaming API for an easier experience + ******************************************************************************/ + +/*-***************************************************************************** + * Buffer-less streaming compression (synchronous mode) + * + * A ZSTD_CCtx object is required to track streaming operations. + * Use ZSTD_initCCtx() to initialize a context. + * ZSTD_CCtx object can be re-used multiple times within successive compression + * operations. + * + * Start by initializing a context. + * Use ZSTD_compressBegin(), or ZSTD_compressBegin_usingDict() for dictionary + * compression, + * or ZSTD_compressBegin_advanced(), for finer parameter control. + * It's also possible to duplicate a reference context which has already been + * initialized, using ZSTD_copyCCtx() + * + * Then, consume your input using ZSTD_compressContinue(). + * There are some important considerations to keep in mind when using this + * advanced function : + * - ZSTD_compressContinue() has no internal buffer. It uses externally provided + * buffer only. + * - Interface is synchronous : input is consumed entirely and produce 1+ + * (or more) compressed blocks. + * - Caller must ensure there is enough space in `dst` to store compressed data + * under worst case scenario. Worst case evaluation is provided by + * ZSTD_compressBound(). + * ZSTD_compressContinue() doesn't guarantee recover after a failed + * compression. + * - ZSTD_compressContinue() presumes prior input ***is still accessible and + * unmodified*** (up to maximum distance size, see WindowLog). + * It remembers all previous contiguous blocks, plus one separated memory + * segment (which can itself consists of multiple contiguous blocks) + * - ZSTD_compressContinue() detects that prior input has been overwritten when + * `src` buffer overlaps. In which case, it will "discard" the relevant memory + * section from its history. + * + * Finish a frame with ZSTD_compressEnd(), which will write the last block(s) + * and optional checksum. It's possible to use srcSize==0, in which case, it + * will write a final empty block to end the frame. Without last block mark, + * frames will be considered unfinished (corrupted) by decoders. + * + * `ZSTD_CCtx` object can be re-used (ZSTD_compressBegin()) to compress some new + * frame. + ******************************************************************************/ + +/*===== Buffer-less streaming compression functions =====*/ +size_t ZSTD_compressBegin(ZSTD_CCtx *cctx, int compressionLevel); +size_t ZSTD_compressBegin_usingDict(ZSTD_CCtx *cctx, const void *dict, + size_t dictSize, int compressionLevel); +size_t ZSTD_compressBegin_advanced(ZSTD_CCtx *cctx, const void *dict, + size_t dictSize, ZSTD_parameters params, + unsigned long long pledgedSrcSize); +size_t ZSTD_copyCCtx(ZSTD_CCtx *cctx, const ZSTD_CCtx *preparedCCtx, + unsigned long long pledgedSrcSize); +size_t ZSTD_compressBegin_usingCDict(ZSTD_CCtx *cctx, const ZSTD_CDict *cdict, + unsigned long long pledgedSrcSize); +size_t ZSTD_compressContinue(ZSTD_CCtx *cctx, void *dst, size_t dstCapacity, + const void *src, size_t srcSize); +size_t ZSTD_compressEnd(ZSTD_CCtx *cctx, void *dst, size_t dstCapacity, + const void *src, size_t srcSize); + + + +/*-***************************************************************************** + * Buffer-less streaming decompression (synchronous mode) + * + * A ZSTD_DCtx object is required to track streaming operations. + * Use ZSTD_initDCtx() to initialize a context. + * A ZSTD_DCtx object can be re-used multiple times. + * + * First typical operation is to retrieve frame parameters, using + * ZSTD_getFrameParams(). It fills a ZSTD_frameParams structure which provide + * important information to correctly decode the frame, such as the minimum + * rolling buffer size to allocate to decompress data (`windowSize`), and the + * dictionary ID used. + * Note: content size is optional, it may not be present. 0 means unknown. + * Note that these values could be wrong, either because of data malformation, + * or because an attacker is spoofing deliberate false information. As a + * consequence, check that values remain within valid application range, + * especially `windowSize`, before allocation. Each application can set its own + * limit, depending on local restrictions. For extended interoperability, it is + * recommended to support at least 8 MB. + * Frame parameters are extracted from the beginning of the compressed frame. + * Data fragment must be large enough to ensure successful decoding, typically + * `ZSTD_frameHeaderSize_max` bytes. + * Result: 0: successful decoding, the `ZSTD_frameParams` structure is filled. + * >0: `srcSize` is too small, provide at least this many bytes. + * errorCode, which can be tested using ZSTD_isError(). + * + * Start decompression, with ZSTD_decompressBegin() or + * ZSTD_decompressBegin_usingDict(). Alternatively, you can copy a prepared + * context, using ZSTD_copyDCtx(). + * + * Then use ZSTD_nextSrcSizeToDecompress() and ZSTD_decompressContinue() + * alternatively. + * ZSTD_nextSrcSizeToDecompress() tells how many bytes to provide as 'srcSize' + * to ZSTD_decompressContinue(). + * ZSTD_decompressContinue() requires this _exact_ amount of bytes, or it will + * fail. + * + * The result of ZSTD_decompressContinue() is the number of bytes regenerated + * within 'dst' (necessarily <= dstCapacity). It can be zero, which is not an + * error; it just means ZSTD_decompressContinue() has decoded some metadata + * item. It can also be an error code, which can be tested with ZSTD_isError(). + * + * ZSTD_decompressContinue() needs previous data blocks during decompression, up + * to `windowSize`. They should preferably be located contiguously, prior to + * current block. Alternatively, a round buffer of sufficient size is also + * possible. Sufficient size is determined by frame parameters. + * ZSTD_decompressContinue() is very sensitive to contiguity, if 2 blocks don't + * follow each other, make sure that either the compressor breaks contiguity at + * the same place, or that previous contiguous segment is large enough to + * properly handle maximum back-reference. + * + * A frame is fully decoded when ZSTD_nextSrcSizeToDecompress() returns zero. + * Context can then be reset to start a new decompression. + * + * Note: it's possible to know if next input to present is a header or a block, + * using ZSTD_nextInputType(). This information is not required to properly + * decode a frame. + * + * == Special case: skippable frames == + * + * Skippable frames allow integration of user-defined data into a flow of + * concatenated frames. Skippable frames will be ignored (skipped) by a + * decompressor. The format of skippable frames is as follows: + * a) Skippable frame ID - 4 Bytes, Little endian format, any value from + * 0x184D2A50 to 0x184D2A5F + * b) Frame Size - 4 Bytes, Little endian format, unsigned 32-bits + * c) Frame Content - any content (User Data) of length equal to Frame Size + * For skippable frames ZSTD_decompressContinue() always returns 0. + * For skippable frames ZSTD_getFrameParams() returns fparamsPtr->windowLog==0 + * what means that a frame is skippable. + * Note: If fparamsPtr->frameContentSize==0, it is ambiguous: the frame might + * actually be a zstd encoded frame with no content. For purposes of + * decompression, it is valid in both cases to skip the frame using + * ZSTD_findFrameCompressedSize() to find its size in bytes. + * It also returns frame size as fparamsPtr->frameContentSize. + ******************************************************************************/ + +/*===== Buffer-less streaming decompression functions =====*/ +size_t ZSTD_decompressBegin(ZSTD_DCtx *dctx); +size_t ZSTD_decompressBegin_usingDict(ZSTD_DCtx *dctx, const void *dict, + size_t dictSize); +void ZSTD_copyDCtx(ZSTD_DCtx *dctx, const ZSTD_DCtx *preparedDCtx); +size_t ZSTD_nextSrcSizeToDecompress(ZSTD_DCtx *dctx); +size_t ZSTD_decompressContinue(ZSTD_DCtx *dctx, void *dst, size_t dstCapacity, + const void *src, size_t srcSize); +typedef enum { + ZSTDnit_frameHeader, + ZSTDnit_blockHeader, + ZSTDnit_block, + ZSTDnit_lastBlock, + ZSTDnit_checksum, + ZSTDnit_skippableFrame +} ZSTD_nextInputType_e; +ZSTD_nextInputType_e ZSTD_nextInputType(ZSTD_DCtx *dctx); + +/*-***************************************************************************** + * Block functions + * + * Block functions produce and decode raw zstd blocks, without frame metadata. + * Frame metadata cost is typically ~18 bytes, which can be non-negligible for + * very small blocks (< 100 bytes). User will have to take in charge required + * information to regenerate data, such as compressed and content sizes. + * + * A few rules to respect: + * - Compressing and decompressing require a context structure + * + Use ZSTD_initCCtx() and ZSTD_initDCtx() + * - It is necessary to init context before starting + * + compression : ZSTD_compressBegin() + * + decompression : ZSTD_decompressBegin() + * + variants _usingDict() are also allowed + * + copyCCtx() and copyDCtx() work too + * - Block size is limited, it must be <= ZSTD_getBlockSizeMax() + * + If you need to compress more, cut data into multiple blocks + * + Consider using the regular ZSTD_compress() instead, as frame metadata + * costs become negligible when source size is large. + * - When a block is considered not compressible enough, ZSTD_compressBlock() + * result will be zero. In which case, nothing is produced into `dst`. + * + User must test for such outcome and deal directly with uncompressed data + * + ZSTD_decompressBlock() doesn't accept uncompressed data as input!!! + * + In case of multiple successive blocks, decoder must be informed of + * uncompressed block existence to follow proper history. Use + * ZSTD_insertBlock() in such a case. + ******************************************************************************/ + +/* Define for static allocation */ +#define ZSTD_BLOCKSIZE_ABSOLUTEMAX (128 * 1024) +/*===== Raw zstd block functions =====*/ +size_t ZSTD_getBlockSizeMax(ZSTD_CCtx *cctx); +size_t ZSTD_compressBlock(ZSTD_CCtx *cctx, void *dst, size_t dstCapacity, + const void *src, size_t srcSize); +size_t ZSTD_decompressBlock(ZSTD_DCtx *dctx, void *dst, size_t dstCapacity, + const void *src, size_t srcSize); +size_t ZSTD_insertBlock(ZSTD_DCtx *dctx, const void *blockStart, + size_t blockSize); + +#endif /* ZSTD_H */ diff --git a/lib/decompress_unzstd.c b/lib/decompress_unzstd.c index 6b629ab31c1e..c076d6f05064 100644 --- a/lib/decompress_unzstd.c +++ b/lib/decompress_unzstd.c @@ -91,11 +91,15 @@ static int INIT handle_zstd_error(size_t ret, void (*error)(char *x)) { - const int err = ZSTD_getErrorCode(ret); + const zstd_error_code err = zstd_get_error_code(ret); - if (!ZSTD_isError(ret)) + if (!zstd_is_error(ret)) return 0; + /* + * zstd_get_error_name() cannot be used because error takes a char * + * not a const char * + */ switch (err) { case ZSTD_error_memory_allocation: error("ZSTD decompressor ran out of memory"); @@ -124,28 +128,28 @@ static int INIT decompress_single(const u8 *in_buf, long in_len, u8 *out_buf, long out_len, long *in_pos, void (*error)(char *x)) { - const size_t wksp_size = ZSTD_DCtxWorkspaceBound(); + const size_t wksp_size = zstd_dctx_workspace_bound(); void *wksp = large_malloc(wksp_size); - ZSTD_DCtx *dctx = ZSTD_initDCtx(wksp, wksp_size); + zstd_dctx *dctx = zstd_init_dctx(wksp, wksp_size); int err; size_t ret; if (dctx == NULL) { - error("Out of memory while allocating ZSTD_DCtx"); + error("Out of memory while allocating zstd_dctx"); err = -1; goto out; } /* * Find out how large the frame actually is, there may be junk at - * the end of the frame that ZSTD_decompressDCtx() can't handle. + * the end of the frame that zstd_decompress_dctx() can't handle. */ - ret = ZSTD_findFrameCompressedSize(in_buf, in_len); + ret = zstd_find_frame_compressed_size(in_buf, in_len); err = handle_zstd_error(ret, error); if (err) goto out; in_len = (long)ret; - ret = ZSTD_decompressDCtx(dctx, out_buf, out_len, in_buf, in_len); + ret = zstd_decompress_dctx(dctx, out_buf, out_len, in_buf, in_len); err = handle_zstd_error(ret, error); if (err) goto out; @@ -167,14 +171,14 @@ static int INIT __unzstd(unsigned char *in_buf, long in_len, long *in_pos, void (*error)(char *x)) { - ZSTD_inBuffer in; - ZSTD_outBuffer out; - ZSTD_frameParams params; + zstd_in_buffer in; + zstd_out_buffer out; + zstd_frame_header header; void *in_allocated = NULL; void *out_allocated = NULL; void *wksp = NULL; size_t wksp_size; - ZSTD_DStream *dstream; + zstd_dstream *dstream; int err; size_t ret; @@ -238,13 +242,13 @@ static int INIT __unzstd(unsigned char *in_buf, long in_len, out.size = out_len; /* - * We need to know the window size to allocate the ZSTD_DStream. + * We need to know the window size to allocate the zstd_dstream. * Since we are streaming, we need to allocate a buffer for the sliding * window. The window size varies from 1 KB to ZSTD_WINDOWSIZE_MAX * (8 MB), so it is important to use the actual value so as not to * waste memory when it is smaller. */ - ret = ZSTD_getFrameParams(¶ms, in.src, in.size); + ret = zstd_get_frame_header(&header, in.src, in.size); err = handle_zstd_error(ret, error); if (err) goto out; @@ -253,19 +257,19 @@ static int INIT __unzstd(unsigned char *in_buf, long in_len, err = -1; goto out; } - if (params.windowSize > ZSTD_WINDOWSIZE_MAX) { + if (header.windowSize > ZSTD_WINDOWSIZE_MAX) { error("ZSTD-compressed data has too large a window size"); err = -1; goto out; } /* - * Allocate the ZSTD_DStream now that we know how much memory is + * Allocate the zstd_dstream now that we know how much memory is * required. */ - wksp_size = ZSTD_DStreamWorkspaceBound(params.windowSize); + wksp_size = zstd_dstream_workspace_bound(header.windowSize); wksp = large_malloc(wksp_size); - dstream = ZSTD_initDStream(params.windowSize, wksp, wksp_size); + dstream = zstd_init_dstream(header.windowSize, wksp, wksp_size); if (dstream == NULL) { error("Out of memory while allocating ZSTD_DStream"); err = -1; @@ -298,7 +302,7 @@ static int INIT __unzstd(unsigned char *in_buf, long in_len, in.size = in_len; } /* Returns zero when the frame is complete. */ - ret = ZSTD_decompressStream(dstream, &out, &in); + ret = zstd_decompress_stream(dstream, &out, &in); err = handle_zstd_error(ret, error); if (err) goto out; diff --git a/lib/zstd/compress.c b/lib/zstd/compress.c index b080264ed3ad..57aaa64306a0 100644 --- a/lib/zstd/compress.c +++ b/lib/zstd/compress.c @@ -3443,43 +3443,92 @@ ZSTD_parameters ZSTD_getParams(int compressionLevel, unsigned long long srcSize, return params; } -EXPORT_SYMBOL(ZSTD_maxCLevel); -EXPORT_SYMBOL(ZSTD_compressBound); - -EXPORT_SYMBOL(ZSTD_CCtxWorkspaceBound); -EXPORT_SYMBOL(ZSTD_initCCtx); -EXPORT_SYMBOL(ZSTD_compressCCtx); -EXPORT_SYMBOL(ZSTD_compress_usingDict); - -EXPORT_SYMBOL(ZSTD_CDictWorkspaceBound); -EXPORT_SYMBOL(ZSTD_initCDict); -EXPORT_SYMBOL(ZSTD_compress_usingCDict); - -EXPORT_SYMBOL(ZSTD_CStreamWorkspaceBound); -EXPORT_SYMBOL(ZSTD_initCStream); -EXPORT_SYMBOL(ZSTD_initCStream_usingCDict); -EXPORT_SYMBOL(ZSTD_resetCStream); -EXPORT_SYMBOL(ZSTD_compressStream); -EXPORT_SYMBOL(ZSTD_flushStream); -EXPORT_SYMBOL(ZSTD_endStream); -EXPORT_SYMBOL(ZSTD_CStreamInSize); -EXPORT_SYMBOL(ZSTD_CStreamOutSize); - -EXPORT_SYMBOL(ZSTD_getCParams); -EXPORT_SYMBOL(ZSTD_getParams); -EXPORT_SYMBOL(ZSTD_checkCParams); -EXPORT_SYMBOL(ZSTD_adjustCParams); - -EXPORT_SYMBOL(ZSTD_compressBegin); -EXPORT_SYMBOL(ZSTD_compressBegin_usingDict); -EXPORT_SYMBOL(ZSTD_compressBegin_advanced); -EXPORT_SYMBOL(ZSTD_copyCCtx); -EXPORT_SYMBOL(ZSTD_compressBegin_usingCDict); -EXPORT_SYMBOL(ZSTD_compressContinue); -EXPORT_SYMBOL(ZSTD_compressEnd); - -EXPORT_SYMBOL(ZSTD_getBlockSizeMax); -EXPORT_SYMBOL(ZSTD_compressBlock); +size_t zstd_compress_bound(size_t src_size) +{ + return ZSTD_compressBound(src_size); +} +EXPORT_SYMBOL(zstd_compress_bound); + +int zstd_min_clevel(void) +{ + /* + * zstd-1.3.1 doesn't implement ZSTD_minCLevel(). + * Return 0 (default level). + */ + return 0; +} +EXPORT_SYMBOL(zstd_min_clevel); + +int zstd_max_clevel(void) +{ + return ZSTD_maxCLevel(); +} +EXPORT_SYMBOL(zstd_max_clevel); + +zstd_parameters zstd_get_params(int level, + unsigned long long estimated_src_size) +{ + return ZSTD_getParams(level, estimated_src_size, 0); +} +EXPORT_SYMBOL(zstd_get_params); + +size_t zstd_cctx_workspace_bound(const zstd_compression_parameters *cparams) +{ + return ZSTD_CCtxWorkspaceBound(*cparams); +} +EXPORT_SYMBOL(zstd_cctx_workspace_bound); + +zstd_cctx *zstd_init_cctx(void *workspace, size_t workspace_size) +{ + return ZSTD_initCCtx(workspace, workspace_size); +} +EXPORT_SYMBOL(zstd_init_cctx); + +size_t zstd_compress_cctx(zstd_cctx *cctx, void *dst, size_t dst_capacity, + const void *src, size_t src_size, const zstd_parameters *parameters) +{ + return ZSTD_compressCCtx(cctx, dst, dst_capacity, src, src_size, *parameters); +} +EXPORT_SYMBOL(zstd_compress_cctx); + +size_t zstd_cstream_workspace_bound(const zstd_compression_parameters *cparams) +{ + return ZSTD_CStreamWorkspaceBound(*cparams); +} +EXPORT_SYMBOL(zstd_cstream_workspace_bound); + +zstd_cstream *zstd_init_cstream(const zstd_parameters *parameters, + unsigned long long pledged_src_size, void *workspace, size_t workspace_size) +{ + return ZSTD_initCStream(*parameters, pledged_src_size, workspace, workspace_size); +} +EXPORT_SYMBOL(zstd_init_cstream); + +size_t zstd_reset_cstream(zstd_cstream *cstream, + unsigned long long pledged_src_size) +{ + return ZSTD_resetCStream(cstream, pledged_src_size); +} +EXPORT_SYMBOL(zstd_reset_cstream); + +size_t zstd_compress_stream(zstd_cstream *cstream, zstd_out_buffer *output, + zstd_in_buffer *input) +{ + return ZSTD_compressStream(cstream, output, input); +} +EXPORT_SYMBOL(zstd_compress_stream); + +size_t zstd_flush_stream(zstd_cstream *cstream, zstd_out_buffer *output) +{ + return ZSTD_flushStream(cstream, output); +} +EXPORT_SYMBOL(zstd_flush_stream); + +size_t zstd_end_stream(zstd_cstream *cstream, zstd_out_buffer *output) +{ + return ZSTD_endStream(cstream, output); +} +EXPORT_SYMBOL(zstd_end_stream); MODULE_LICENSE("Dual BSD/GPL"); MODULE_DESCRIPTION("Zstd Compressor"); diff --git a/lib/zstd/decompress.c b/lib/zstd/decompress.c index 66cd487a326a..02e92c2cbf4f 100644 --- a/lib/zstd/decompress.c +++ b/lib/zstd/decompress.c @@ -2490,42 +2490,82 @@ size_t ZSTD_decompressStream(ZSTD_DStream *zds, ZSTD_outBuffer *output, ZSTD_inB } } -EXPORT_SYMBOL(ZSTD_DCtxWorkspaceBound); -EXPORT_SYMBOL(ZSTD_initDCtx); -EXPORT_SYMBOL(ZSTD_decompressDCtx); -EXPORT_SYMBOL(ZSTD_decompress_usingDict); - -EXPORT_SYMBOL(ZSTD_DDictWorkspaceBound); -EXPORT_SYMBOL(ZSTD_initDDict); -EXPORT_SYMBOL(ZSTD_decompress_usingDDict); - -EXPORT_SYMBOL(ZSTD_DStreamWorkspaceBound); -EXPORT_SYMBOL(ZSTD_initDStream); -EXPORT_SYMBOL(ZSTD_initDStream_usingDDict); -EXPORT_SYMBOL(ZSTD_resetDStream); -EXPORT_SYMBOL(ZSTD_decompressStream); -EXPORT_SYMBOL(ZSTD_DStreamInSize); -EXPORT_SYMBOL(ZSTD_DStreamOutSize); - -EXPORT_SYMBOL(ZSTD_findFrameCompressedSize); -EXPORT_SYMBOL(ZSTD_getFrameContentSize); -EXPORT_SYMBOL(ZSTD_findDecompressedSize); - -EXPORT_SYMBOL(ZSTD_isFrame); -EXPORT_SYMBOL(ZSTD_getDictID_fromDict); -EXPORT_SYMBOL(ZSTD_getDictID_fromDDict); -EXPORT_SYMBOL(ZSTD_getDictID_fromFrame); - -EXPORT_SYMBOL(ZSTD_getFrameParams); -EXPORT_SYMBOL(ZSTD_decompressBegin); -EXPORT_SYMBOL(ZSTD_decompressBegin_usingDict); -EXPORT_SYMBOL(ZSTD_copyDCtx); -EXPORT_SYMBOL(ZSTD_nextSrcSizeToDecompress); -EXPORT_SYMBOL(ZSTD_decompressContinue); -EXPORT_SYMBOL(ZSTD_nextInputType); - -EXPORT_SYMBOL(ZSTD_decompressBlock); -EXPORT_SYMBOL(ZSTD_insertBlock); +unsigned int zstd_is_error(size_t code) +{ + return ZSTD_isError(code); +} +EXPORT_SYMBOL(zstd_is_error); + +zstd_error_code zstd_get_error_code(size_t code) +{ + return ZSTD_getErrorCode(code); +} +EXPORT_SYMBOL(zstd_get_error_code); + +const char *zstd_get_error_name(size_t code) +{ + /* Real implementation in zstd-1.4.6. */ + return "GENERIC"; +} +EXPORT_SYMBOL(zstd_get_error_name); + +size_t zstd_dctx_workspace_bound(void) +{ + return ZSTD_DCtxWorkspaceBound(); +} +EXPORT_SYMBOL(zstd_dctx_workspace_bound); + +zstd_dctx *zstd_init_dctx(void *workspace, size_t workspace_size) +{ + return ZSTD_initDCtx(workspace, workspace_size); +} +EXPORT_SYMBOL(zstd_init_dctx); + +size_t zstd_decompress_dctx(zstd_dctx *dctx, void *dst, size_t dst_capacity, + const void *src, size_t src_size) +{ + return ZSTD_decompressDCtx(dctx, dst, dst_capacity, src, src_size); +} +EXPORT_SYMBOL(zstd_decompress_dctx); + +size_t zstd_dstream_workspace_bound(size_t max_window_size) +{ + return ZSTD_DStreamWorkspaceBound(max_window_size); +} +EXPORT_SYMBOL(zstd_dstream_workspace_bound); + +zstd_dstream *zstd_init_dstream(size_t max_window_size, void *workspace, + size_t workspace_size) +{ + return ZSTD_initDStream(max_window_size, workspace, workspace_size); +} +EXPORT_SYMBOL(zstd_init_dstream); + +size_t zstd_reset_dstream(zstd_dstream *dstream) +{ + return ZSTD_resetDStream(dstream); +} +EXPORT_SYMBOL(zstd_reset_dstream); + +size_t zstd_decompress_stream(zstd_dstream *dstream, zstd_out_buffer *output, + zstd_in_buffer *input) +{ + return ZSTD_decompressStream(dstream, output, input); +} +EXPORT_SYMBOL(zstd_decompress_stream); + +size_t zstd_find_frame_compressed_size(const void *src, size_t src_size) +{ + return ZSTD_findFrameCompressedSize(src, src_size); +} +EXPORT_SYMBOL(zstd_find_frame_compressed_size); + +size_t zstd_get_frame_header(zstd_frame_header *header, const void *src, + size_t src_size) +{ + return ZSTD_getFrameParams(header, src, src_size); +} +EXPORT_SYMBOL(zstd_get_frame_header); MODULE_LICENSE("Dual BSD/GPL"); MODULE_DESCRIPTION("Zstd Decompressor"); -- cgit v1.2.3 From 2479b523898633768e28796238534af31fbd6846 Mon Sep 17 00:00:00 2001 From: Nick Terrell Date: Mon, 14 Sep 2020 12:54:12 -0700 Subject: lib: zstd: Add decompress_sources.h for decompress_unzstd Adds decompress_sources.h which includes every .c file necessary for zstd decompression. This is used in decompress_unzstd.c so the internal structure of the library isn't exposed. This allows us to upgrade the zstd library version without modifying any callers. Instead we just need to update decompress_sources.h. Signed-off-by: Nick Terrell Tested By: Paul Jones Tested-by: Oleksandr Natalenko Tested-by: Sedat Dilek # LLVM/Clang v13.0.0 on x86-64 Tested-by: Jean-Denis Girard --- lib/decompress_unzstd.c | 6 +----- lib/zstd/decompress_sources.h | 23 +++++++++++++++++++++++ 2 files changed, 24 insertions(+), 5 deletions(-) create mode 100644 lib/zstd/decompress_sources.h (limited to 'lib') diff --git a/lib/decompress_unzstd.c b/lib/decompress_unzstd.c index c076d6f05064..a512b99ae16a 100644 --- a/lib/decompress_unzstd.c +++ b/lib/decompress_unzstd.c @@ -68,11 +68,7 @@ #ifdef STATIC # define UNZSTD_PREBOOT # include "xxhash.c" -# include "zstd/entropy_common.c" -# include "zstd/fse_decompress.c" -# include "zstd/huf_decompress.c" -# include "zstd/zstd_common.c" -# include "zstd/decompress.c" +# include "zstd/decompress_sources.h" #endif #include diff --git a/lib/zstd/decompress_sources.h b/lib/zstd/decompress_sources.h new file mode 100644 index 000000000000..9ba367b44149 --- /dev/null +++ b/lib/zstd/decompress_sources.h @@ -0,0 +1,23 @@ +/* SPDX-License-Identifier: GPL-2.0+ OR BSD-3-Clause */ +/* + * Copyright (c) Facebook, Inc. + * All rights reserved. + * + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). + * You may select, at your option, one of the above-listed licenses. + */ + +/* + * This file includes every .c file needed for decompression. + * It is used by lib/decompress_unzstd.c to include the decompression + * source into the translation-unit, so it can be used for kernel + * decompression. + */ + +#include "entropy_common.c" +#include "fse_decompress.c" +#include "huf_decompress.c" +#include "zstd_common.c" +#include "decompress.c" -- cgit v1.2.3 From e0c1b49f5b674cca7b10549c53b3791d0bbc90a8 Mon Sep 17 00:00:00 2001 From: Nick Terrell Date: Fri, 11 Sep 2020 16:37:08 -0700 Subject: lib: zstd: Upgrade to latest upstream zstd version 1.4.10 Upgrade to the latest upstream zstd version 1.4.10. This patch is 100% generated from upstream zstd commit 20821a46f412 [0]. This patch is very large because it is transitioning from the custom kernel zstd to using upstream directly. The new zstd follows upstreams file structure which is different. Future update patches will be much smaller because they will only contain the changes from one upstream zstd release. As an aid for review I've created a commit [1] that shows the diff between upstream zstd as-is (which doesn't compile), and the zstd code imported in this patch. The verion of zstd in this patch is generated from upstream with changes applied by automation to replace upstreams libc dependencies, remove unnecessary portability macros, replace `/**` comments with `/*` comments, and use the kernel's xxhash instead of bundling it. The benefits of this patch are as follows: 1. Using upstream directly with automated script to generate kernel code. This allows us to update the kernel every upstream release, so the kernel gets the latest bug fixes and performance improvements, and doesn't get 3 years out of date again. The automation and the translated code are tested every upstream commit to ensure it continues to work. 2. Upgrades from a custom zstd based on 1.3.1 to 1.4.10, getting 3 years of performance improvements and bug fixes. On x86_64 I've measured 15% faster BtrFS and SquashFS decompression+read speeds, 35% faster kernel decompression, and 30% faster ZRAM decompression+read speeds. 3. Zstd-1.4.10 supports negative compression levels, which allow zstd to match or subsume lzo's performance. 4. Maintains the same kernel-specific wrapper API, so no callers have to be modified with zstd version updates. One concern that was brought up was stack usage. Upstream zstd had already removed most of its heavy stack usage functions, but I just removed the last functions that allocate arrays on the stack. I've measured the high water mark for both compression and decompression before and after this patch. Decompression is approximately neutral, using about 1.2KB of stack space. Compression levels up to 3 regressed from 1.4KB -> 1.6KB, and higher compression levels regressed from 1.5KB -> 2KB. We've added unit tests upstream to prevent further regression. I believe that this is a reasonable increase, and if it does end up causing problems, this commit can be cleanly reverted, because it only touches zstd. I chose the bulk update instead of replaying upstream commits because there have been ~3500 upstream commits since the 1.3.1 release, zstd wasn't ready to be used in the kernel as-is before a month ago, and not all upstream zstd commits build. The bulk update preserves bisectablity because bugs can be bisected to the zstd version update. At that point the update can be reverted, and we can work with upstream to find and fix the bug. Note that upstream zstd release 1.4.10 doesn't exist yet. I have cut a staging branch at 20821a46f412 [0] and will apply any changes requested to the staging branch. Once we're ready to merge this update I will cut a zstd release at the commit we merge, so we have a known zstd release in the kernel. The implementation of the kernel API is contained in zstd_compress_module.c and zstd_decompress_module.c. [0] https://github.com/facebook/zstd/commit/20821a46f4122f9abd7c7b245d28162dde8129c9 [1] https://github.com/terrelln/linux/commit/e0fa481d0e3df26918da0a13749740a1f6777574 Signed-off-by: Nick Terrell Tested By: Paul Jones Tested-by: Oleksandr Natalenko Tested-by: Sedat Dilek # LLVM/Clang v13.0.0 on x86-64 Tested-by: Jean-Denis Girard --- include/linux/zstd.h | 13 +- include/linux/zstd_errors.h | 77 + include/linux/zstd_lib.h | 3367 +++++++++++----- lib/zstd/Makefile | 46 +- lib/zstd/bitstream.h | 380 -- lib/zstd/common/bitstream.h | 437 ++ lib/zstd/common/compiler.h | 170 + lib/zstd/common/cpu.h | 194 + lib/zstd/common/debug.c | 24 + lib/zstd/common/debug.h | 101 + lib/zstd/common/entropy_common.c | 357 ++ lib/zstd/common/error_private.c | 56 + lib/zstd/common/error_private.h | 66 + lib/zstd/common/fse.h | 710 ++++ lib/zstd/common/fse_decompress.c | 390 ++ lib/zstd/common/huf.h | 356 ++ lib/zstd/common/mem.h | 259 ++ lib/zstd/common/zstd_common.c | 83 + lib/zstd/common/zstd_deps.h | 125 + lib/zstd/common/zstd_internal.h | 450 +++ lib/zstd/compress.c | 3534 ---------------- lib/zstd/compress/fse_compress.c | 625 +++ lib/zstd/compress/hist.c | 165 + lib/zstd/compress/hist.h | 75 + lib/zstd/compress/huf_compress.c | 905 +++++ lib/zstd/compress/zstd_compress.c | 5109 ++++++++++++++++++++++++ lib/zstd/compress/zstd_compress_internal.h | 1188 ++++++ lib/zstd/compress/zstd_compress_literals.c | 158 + lib/zstd/compress/zstd_compress_literals.h | 29 + lib/zstd/compress/zstd_compress_sequences.c | 439 ++ lib/zstd/compress/zstd_compress_sequences.h | 54 + lib/zstd/compress/zstd_compress_superblock.c | 850 ++++ lib/zstd/compress/zstd_compress_superblock.h | 32 + lib/zstd/compress/zstd_cwksp.h | 482 +++ lib/zstd/compress/zstd_double_fast.c | 519 +++ lib/zstd/compress/zstd_double_fast.h | 32 + lib/zstd/compress/zstd_fast.c | 496 +++ lib/zstd/compress/zstd_fast.h | 31 + lib/zstd/compress/zstd_lazy.c | 1414 +++++++ lib/zstd/compress/zstd_lazy.h | 81 + lib/zstd/compress/zstd_ldm.c | 686 ++++ lib/zstd/compress/zstd_ldm.h | 110 + lib/zstd/compress/zstd_ldm_geartab.h | 103 + lib/zstd/compress/zstd_opt.c | 1346 +++++++ lib/zstd/compress/zstd_opt.h | 50 + lib/zstd/decompress.c | 2571 ------------ lib/zstd/decompress/huf_decompress.c | 1206 ++++++ lib/zstd/decompress/zstd_ddict.c | 241 ++ lib/zstd/decompress/zstd_ddict.h | 44 + lib/zstd/decompress/zstd_decompress.c | 2085 ++++++++++ lib/zstd/decompress/zstd_decompress_block.c | 1540 +++++++ lib/zstd/decompress/zstd_decompress_block.h | 62 + lib/zstd/decompress/zstd_decompress_internal.h | 202 + lib/zstd/decompress_sources.h | 15 +- lib/zstd/entropy_common.c | 243 -- lib/zstd/error_private.h | 53 - lib/zstd/fse.h | 575 --- lib/zstd/fse_compress.c | 795 ---- lib/zstd/fse_decompress.c | 325 -- lib/zstd/huf.h | 212 - lib/zstd/huf_compress.c | 773 ---- lib/zstd/huf_decompress.c | 960 ----- lib/zstd/mem.h | 151 - lib/zstd/zstd_common.c | 75 - lib/zstd/zstd_compress_module.c | 160 + lib/zstd/zstd_decompress_module.c | 105 + lib/zstd/zstd_internal.h | 273 -- lib/zstd/zstd_opt.h | 1014 ----- 68 files changed, 26861 insertions(+), 12993 deletions(-) create mode 100644 include/linux/zstd_errors.h delete mode 100644 lib/zstd/bitstream.h create mode 100644 lib/zstd/common/bitstream.h create mode 100644 lib/zstd/common/compiler.h create mode 100644 lib/zstd/common/cpu.h create mode 100644 lib/zstd/common/debug.c create mode 100644 lib/zstd/common/debug.h create mode 100644 lib/zstd/common/entropy_common.c create mode 100644 lib/zstd/common/error_private.c create mode 100644 lib/zstd/common/error_private.h create mode 100644 lib/zstd/common/fse.h create mode 100644 lib/zstd/common/fse_decompress.c create mode 100644 lib/zstd/common/huf.h create mode 100644 lib/zstd/common/mem.h create mode 100644 lib/zstd/common/zstd_common.c create mode 100644 lib/zstd/common/zstd_deps.h create mode 100644 lib/zstd/common/zstd_internal.h delete mode 100644 lib/zstd/compress.c create mode 100644 lib/zstd/compress/fse_compress.c create mode 100644 lib/zstd/compress/hist.c create mode 100644 lib/zstd/compress/hist.h create mode 100644 lib/zstd/compress/huf_compress.c create mode 100644 lib/zstd/compress/zstd_compress.c create mode 100644 lib/zstd/compress/zstd_compress_internal.h create mode 100644 lib/zstd/compress/zstd_compress_literals.c create mode 100644 lib/zstd/compress/zstd_compress_literals.h create mode 100644 lib/zstd/compress/zstd_compress_sequences.c create mode 100644 lib/zstd/compress/zstd_compress_sequences.h create mode 100644 lib/zstd/compress/zstd_compress_superblock.c create mode 100644 lib/zstd/compress/zstd_compress_superblock.h create mode 100644 lib/zstd/compress/zstd_cwksp.h create mode 100644 lib/zstd/compress/zstd_double_fast.c create mode 100644 lib/zstd/compress/zstd_double_fast.h create mode 100644 lib/zstd/compress/zstd_fast.c create mode 100644 lib/zstd/compress/zstd_fast.h create mode 100644 lib/zstd/compress/zstd_lazy.c create mode 100644 lib/zstd/compress/zstd_lazy.h create mode 100644 lib/zstd/compress/zstd_ldm.c create mode 100644 lib/zstd/compress/zstd_ldm.h create mode 100644 lib/zstd/compress/zstd_ldm_geartab.h create mode 100644 lib/zstd/compress/zstd_opt.c create mode 100644 lib/zstd/compress/zstd_opt.h delete mode 100644 lib/zstd/decompress.c create mode 100644 lib/zstd/decompress/huf_decompress.c create mode 100644 lib/zstd/decompress/zstd_ddict.c create mode 100644 lib/zstd/decompress/zstd_ddict.h create mode 100644 lib/zstd/decompress/zstd_decompress.c create mode 100644 lib/zstd/decompress/zstd_decompress_block.c create mode 100644 lib/zstd/decompress/zstd_decompress_block.h create mode 100644 lib/zstd/decompress/zstd_decompress_internal.h delete mode 100644 lib/zstd/entropy_common.c delete mode 100644 lib/zstd/error_private.h delete mode 100644 lib/zstd/fse.h delete mode 100644 lib/zstd/fse_compress.c delete mode 100644 lib/zstd/fse_decompress.c delete mode 100644 lib/zstd/huf.h delete mode 100644 lib/zstd/huf_compress.c delete mode 100644 lib/zstd/huf_decompress.c delete mode 100644 lib/zstd/mem.h delete mode 100644 lib/zstd/zstd_common.c create mode 100644 lib/zstd/zstd_compress_module.c create mode 100644 lib/zstd/zstd_decompress_module.c delete mode 100644 lib/zstd/zstd_internal.h delete mode 100644 lib/zstd/zstd_opt.h (limited to 'lib') diff --git a/include/linux/zstd.h b/include/linux/zstd.h index 9fbc7729b0a0..113408eef6ec 100644 --- a/include/linux/zstd.h +++ b/include/linux/zstd.h @@ -1,4 +1,4 @@ -/* SPDX-License-Identifier: GPL-2.0-only */ +/* SPDX-License-Identifier: GPL-2.0+ OR BSD-3-Clause */ /* * Copyright (c) Yann Collet, Facebook, Inc. * All rights reserved. @@ -22,6 +22,7 @@ /* ====== Dependency ====== */ #include +#include #include /* ====== Helper Functions ====== */ @@ -417,12 +418,18 @@ size_t zstd_find_frame_compressed_size(const void *src, size_t src_size); /** * struct zstd_frame_params - zstd frame parameters stored in the frame header - * @frameContentSize: The frame content size, or 0 if not present. + * @frameContentSize: The frame content size, or ZSTD_CONTENTSIZE_UNKNOWN if not + * present. * @windowSize: The window size, or 0 if the frame is a skippable frame. + * @blockSizeMax: The maximum block size. + * @frameType: The frame type (zstd or skippable) + * @headerSize: The size of the frame header. * @dictID: The dictionary id, or 0 if not present. * @checksumFlag: Whether a checksum was used. + * + * See zstd_lib.h. */ -typedef ZSTD_frameParams zstd_frame_header; +typedef ZSTD_frameHeader zstd_frame_header; /** * zstd_get_frame_header() - extracts parameters from a zstd or skippable frame diff --git a/include/linux/zstd_errors.h b/include/linux/zstd_errors.h new file mode 100644 index 000000000000..58b6dd45a969 --- /dev/null +++ b/include/linux/zstd_errors.h @@ -0,0 +1,77 @@ +/* + * Copyright (c) Yann Collet, Facebook, Inc. + * All rights reserved. + * + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). + * You may select, at your option, one of the above-listed licenses. + */ + +#ifndef ZSTD_ERRORS_H_398273423 +#define ZSTD_ERRORS_H_398273423 + + +/*===== dependency =====*/ +#include /* size_t */ + + +/* ===== ZSTDERRORLIB_API : control library symbols visibility ===== */ +#define ZSTDERRORLIB_VISIBILITY +#define ZSTDERRORLIB_API ZSTDERRORLIB_VISIBILITY + +/*-********************************************* + * Error codes list + *-********************************************* + * Error codes _values_ are pinned down since v1.3.1 only. + * Therefore, don't rely on values if you may link to any version < v1.3.1. + * + * Only values < 100 are considered stable. + * + * note 1 : this API shall be used with static linking only. + * dynamic linking is not yet officially supported. + * note 2 : Prefer relying on the enum than on its value whenever possible + * This is the only supported way to use the error list < v1.3.1 + * note 3 : ZSTD_isError() is always correct, whatever the library version. + **********************************************/ +typedef enum { + ZSTD_error_no_error = 0, + ZSTD_error_GENERIC = 1, + ZSTD_error_prefix_unknown = 10, + ZSTD_error_version_unsupported = 12, + ZSTD_error_frameParameter_unsupported = 14, + ZSTD_error_frameParameter_windowTooLarge = 16, + ZSTD_error_corruption_detected = 20, + ZSTD_error_checksum_wrong = 22, + ZSTD_error_dictionary_corrupted = 30, + ZSTD_error_dictionary_wrong = 32, + ZSTD_error_dictionaryCreation_failed = 34, + ZSTD_error_parameter_unsupported = 40, + ZSTD_error_parameter_outOfBound = 42, + ZSTD_error_tableLog_tooLarge = 44, + ZSTD_error_maxSymbolValue_tooLarge = 46, + ZSTD_error_maxSymbolValue_tooSmall = 48, + ZSTD_error_stage_wrong = 60, + ZSTD_error_init_missing = 62, + ZSTD_error_memory_allocation = 64, + ZSTD_error_workSpace_tooSmall= 66, + ZSTD_error_dstSize_tooSmall = 70, + ZSTD_error_srcSize_wrong = 72, + ZSTD_error_dstBuffer_null = 74, + /* following error codes are __NOT STABLE__, they can be removed or changed in future versions */ + ZSTD_error_frameIndex_tooLarge = 100, + ZSTD_error_seekableIO = 102, + ZSTD_error_dstBuffer_wrong = 104, + ZSTD_error_srcBuffer_wrong = 105, + ZSTD_error_maxCode = 120 /* never EVER use this value directly, it can change in future versions! Use ZSTD_isError() instead */ +} ZSTD_ErrorCode; + +/*! ZSTD_getErrorCode() : + convert a `size_t` function result into a `ZSTD_ErrorCode` enum type, + which can be used to compare with enum list published above */ +ZSTDERRORLIB_API ZSTD_ErrorCode ZSTD_getErrorCode(size_t functionResult); +ZSTDERRORLIB_API const char* ZSTD_getErrorString(ZSTD_ErrorCode code); /*< Same as ZSTD_getErrorName, but using a `ZSTD_ErrorCode` enum argument */ + + + +#endif /* ZSTD_ERRORS_H_398273423 */ diff --git a/include/linux/zstd_lib.h b/include/linux/zstd_lib.h index 13151c34f725..b8c7dbf98390 100644 --- a/include/linux/zstd_lib.h +++ b/include/linux/zstd_lib.h @@ -2,1156 +2,2431 @@ * Copyright (c) Yann Collet, Facebook, Inc. * All rights reserved. * - * This source code is licensed under the BSD-style license found in the - * LICENSE file in the root directory of https://github.com/facebook/zstd. - * An additional grant of patent rights can be found in the PATENTS file in the - * same directory. - * - * This program is free software; you can redistribute it and/or modify it under - * the terms of the GNU General Public License version 2 as published by the - * Free Software Foundation. This program is dual-licensed; you may select - * either version 2 of the GNU General Public License ("GPL") or BSD license - * ("BSD"). + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). + * You may select, at your option, one of the above-listed licenses. */ -#ifndef ZSTD_H -#define ZSTD_H +#ifndef ZSTD_H_235446 +#define ZSTD_H_235446 /* ====== Dependency ======*/ +#include /* INT_MAX */ #include /* size_t */ -/*-***************************************************************************** - * Introduction - * - * zstd, short for Zstandard, is a fast lossless compression algorithm, - * targeting real-time compression scenarios at zlib-level and better - * compression ratios. The zstd compression library provides in-memory - * compression and decompression functions. The library supports compression - * levels from 1 up to ZSTD_maxCLevel() which is 22. Levels >= 20, labeled - * ultra, should be used with caution, as they require more memory. - * Compression can be done in: - * - a single step, reusing a context (described as Explicit memory management) - * - unbounded multiple steps (described as Streaming compression) - * The compression ratio achievable on small data can be highly improved using - * compression with a dictionary in: - * - a single step (described as Simple dictionary API) - * - a single step, reusing a dictionary (described as Fast dictionary API) - ******************************************************************************/ +/* ===== ZSTDLIB_API : control library symbols visibility ===== */ +#define ZSTDLIB_VISIBILITY +#define ZSTDLIB_API ZSTDLIB_VISIBILITY + + +/* ***************************************************************************** + Introduction + + zstd, short for Zstandard, is a fast lossless compression algorithm, targeting + real-time compression scenarios at zlib-level and better compression ratios. + The zstd compression library provides in-memory compression and decompression + functions. + + The library supports regular compression levels from 1 up to ZSTD_maxCLevel(), + which is currently 22. Levels >= 20, labeled `--ultra`, should be used with + caution, as they require more memory. The library also offers negative + compression levels, which extend the range of speed vs. ratio preferences. + The lower the level, the faster the speed (at the cost of compression). + + Compression can be done in: + - a single step (described as Simple API) + - a single step, reusing a context (described as Explicit context) + - unbounded multiple steps (described as Streaming compression) + + The compression ratio achievable on small data can be highly improved using + a dictionary. Dictionary compression can be performed in: + - a single step (described as Simple dictionary API) + - a single step, reusing a dictionary (described as Bulk-processing + dictionary API) + + Advanced experimental functions can be accessed using + `#define ZSTD_STATIC_LINKING_ONLY` before including zstd.h. + + Advanced experimental APIs should never be used with a dynamically-linked + library. They are not "stable"; their definitions or signatures may change in + the future. Only static linking is allowed. +*******************************************************************************/ + +/*------ Version ------*/ +#define ZSTD_VERSION_MAJOR 1 +#define ZSTD_VERSION_MINOR 4 +#define ZSTD_VERSION_RELEASE 10 +#define ZSTD_VERSION_NUMBER (ZSTD_VERSION_MAJOR *100*100 + ZSTD_VERSION_MINOR *100 + ZSTD_VERSION_RELEASE) + +/*! ZSTD_versionNumber() : + * Return runtime library version, the value is (MAJOR*100*100 + MINOR*100 + RELEASE). */ +ZSTDLIB_API unsigned ZSTD_versionNumber(void); + +#define ZSTD_LIB_VERSION ZSTD_VERSION_MAJOR.ZSTD_VERSION_MINOR.ZSTD_VERSION_RELEASE +#define ZSTD_QUOTE(str) #str +#define ZSTD_EXPAND_AND_QUOTE(str) ZSTD_QUOTE(str) +#define ZSTD_VERSION_STRING ZSTD_EXPAND_AND_QUOTE(ZSTD_LIB_VERSION) + +/*! ZSTD_versionString() : + * Return runtime library version, like "1.4.5". Requires v1.3.0+. */ +ZSTDLIB_API const char* ZSTD_versionString(void); + +/* ************************************* + * Default constant + ***************************************/ +#ifndef ZSTD_CLEVEL_DEFAULT +# define ZSTD_CLEVEL_DEFAULT 3 +#endif + +/* ************************************* + * Constants + ***************************************/ + +/* All magic numbers are supposed read/written to/from files/memory using little-endian convention */ +#define ZSTD_MAGICNUMBER 0xFD2FB528 /* valid since v0.8.0 */ +#define ZSTD_MAGIC_DICTIONARY 0xEC30A437 /* valid since v0.7.0 */ +#define ZSTD_MAGIC_SKIPPABLE_START 0x184D2A50 /* all 16 values, from 0x184D2A50 to 0x184D2A5F, signal the beginning of a skippable frame */ +#define ZSTD_MAGIC_SKIPPABLE_MASK 0xFFFFFFF0 + +#define ZSTD_BLOCKSIZELOG_MAX 17 +#define ZSTD_BLOCKSIZE_MAX (1<= `ZSTD_compressBound(srcSize)`. + * @return : compressed size written into `dst` (<= `dstCapacity), + * or an error code if it fails (which can be tested using ZSTD_isError()). */ +ZSTDLIB_API size_t ZSTD_compress( void* dst, size_t dstCapacity, + const void* src, size_t srcSize, + int compressionLevel); + +/*! ZSTD_decompress() : + * `compressedSize` : must be the _exact_ size of some number of compressed and/or skippable frames. + * `dstCapacity` is an upper bound of originalSize to regenerate. + * If user cannot imply a maximum upper bound, it's better to use streaming mode to decompress data. + * @return : the number of bytes decompressed into `dst` (<= `dstCapacity`), + * or an errorCode if it fails (which can be tested using ZSTD_isError()). */ +ZSTDLIB_API size_t ZSTD_decompress( void* dst, size_t dstCapacity, + const void* src, size_t compressedSize); + +/*! ZSTD_getFrameContentSize() : requires v1.3.0+ + * `src` should point to the start of a ZSTD encoded frame. + * `srcSize` must be at least as large as the frame header. + * hint : any size >= `ZSTD_frameHeaderSize_max` is large enough. + * @return : - decompressed size of `src` frame content, if known + * - ZSTD_CONTENTSIZE_UNKNOWN if the size cannot be determined + * - ZSTD_CONTENTSIZE_ERROR if an error occurred (e.g. invalid magic number, srcSize too small) + * note 1 : a 0 return value means the frame is valid but "empty". + * note 2 : decompressed size is an optional field, it may not be present, typically in streaming mode. + * When `return==ZSTD_CONTENTSIZE_UNKNOWN`, data to decompress could be any size. + * In which case, it's necessary to use streaming mode to decompress data. + * Optionally, application can rely on some implicit limit, + * as ZSTD_decompress() only needs an upper bound of decompressed size. + * (For example, data could be necessarily cut into blocks <= 16 KB). + * note 3 : decompressed size is always present when compression is completed using single-pass functions, + * such as ZSTD_compress(), ZSTD_compressCCtx() ZSTD_compress_usingDict() or ZSTD_compress_usingCDict(). + * note 4 : decompressed size can be very large (64-bits value), + * potentially larger than what local system can handle as a single memory segment. + * In which case, it's necessary to use streaming mode to decompress data. + * note 5 : If source is untrusted, decompressed size could be wrong or intentionally modified. + * Always ensure return value fits within application's authorized limits. + * Each application can set its own limits. + * note 6 : This function replaces ZSTD_getDecompressedSize() */ +#define ZSTD_CONTENTSIZE_UNKNOWN (0ULL - 1) +#define ZSTD_CONTENTSIZE_ERROR (0ULL - 2) +ZSTDLIB_API unsigned long long ZSTD_getFrameContentSize(const void *src, size_t srcSize); + +/*! ZSTD_getDecompressedSize() : + * NOTE: This function is now obsolete, in favor of ZSTD_getFrameContentSize(). + * Both functions work the same way, but ZSTD_getDecompressedSize() blends + * "empty", "unknown" and "error" results to the same return value (0), + * while ZSTD_getFrameContentSize() gives them separate return values. + * @return : decompressed size of `src` frame content _if known and not empty_, 0 otherwise. */ +ZSTDLIB_API unsigned long long ZSTD_getDecompressedSize(const void* src, size_t srcSize); + +/*! ZSTD_findFrameCompressedSize() : + * `src` should point to the start of a ZSTD frame or skippable frame. + * `srcSize` must be >= first frame size + * @return : the compressed size of the first frame starting at `src`, + * suitable to pass as `srcSize` to `ZSTD_decompress` or similar, + * or an error code if input is invalid */ +ZSTDLIB_API size_t ZSTD_findFrameCompressedSize(const void* src, size_t srcSize); + /*====== Helper functions ======*/ +#define ZSTD_COMPRESSBOUND(srcSize) ((srcSize) + ((srcSize)>>8) + (((srcSize) < (128<<10)) ? (((128<<10) - (srcSize)) >> 11) /* margin, from 64 to 0 */ : 0)) /* this formula ensures that bound(A) + bound(B) <= bound(A+B) as long as A and B >= 128 KB */ +ZSTDLIB_API size_t ZSTD_compressBound(size_t srcSize); /*!< maximum compressed size in worst case single-pass scenario */ +ZSTDLIB_API unsigned ZSTD_isError(size_t code); /*!< tells if a `size_t` function result is an error code */ +ZSTDLIB_API const char* ZSTD_getErrorName(size_t code); /*!< provides readable string from an error code */ +ZSTDLIB_API int ZSTD_minCLevel(void); /*!< minimum negative compression level allowed */ +ZSTDLIB_API int ZSTD_maxCLevel(void); /*!< maximum compression level available */ + + +/* ************************************* +* Explicit context +***************************************/ +/*= Compression context + * When compressing many times, + * it is recommended to allocate a context just once, + * and re-use it for each successive compression operation. + * This will make workload friendlier for system's memory. + * Note : re-using context is just a speed / resource optimization. + * It doesn't change the compression ratio, which remains identical. + * Note 2 : In multi-threaded environments, + * use one different context per thread for parallel execution. + */ +typedef struct ZSTD_CCtx_s ZSTD_CCtx; +ZSTDLIB_API ZSTD_CCtx* ZSTD_createCCtx(void); +ZSTDLIB_API size_t ZSTD_freeCCtx(ZSTD_CCtx* cctx); /* accept NULL pointer */ + +/*! ZSTD_compressCCtx() : + * Same as ZSTD_compress(), using an explicit ZSTD_CCtx. + * Important : in order to behave similarly to `ZSTD_compress()`, + * this function compresses at requested compression level, + * __ignoring any other parameter__ . + * If any advanced parameter was set using the advanced API, + * they will all be reset. Only `compressionLevel` remains. + */ +ZSTDLIB_API size_t ZSTD_compressCCtx(ZSTD_CCtx* cctx, + void* dst, size_t dstCapacity, + const void* src, size_t srcSize, + int compressionLevel); + +/*= Decompression context + * When decompressing many times, + * it is recommended to allocate a context only once, + * and re-use it for each successive compression operation. + * This will make workload friendlier for system's memory. + * Use one context per thread for parallel execution. */ +typedef struct ZSTD_DCtx_s ZSTD_DCtx; +ZSTDLIB_API ZSTD_DCtx* ZSTD_createDCtx(void); +ZSTDLIB_API size_t ZSTD_freeDCtx(ZSTD_DCtx* dctx); /* accept NULL pointer */ + +/*! ZSTD_decompressDCtx() : + * Same as ZSTD_decompress(), + * requires an allocated ZSTD_DCtx. + * Compatible with sticky parameters. + */ +ZSTDLIB_API size_t ZSTD_decompressDCtx(ZSTD_DCtx* dctx, + void* dst, size_t dstCapacity, + const void* src, size_t srcSize); -/** - * enum ZSTD_ErrorCode - zstd error codes + +/* ************************************* +* Advanced compression API +***************************************/ + +/* API design : + * Parameters are pushed one by one into an existing context, + * using ZSTD_CCtx_set*() functions. + * Pushed parameters are sticky : they are valid for next compressed frame, and any subsequent frame. + * "sticky" parameters are applicable to `ZSTD_compress2()` and `ZSTD_compressStream*()` ! + * __They do not apply to "simple" one-shot variants such as ZSTD_compressCCtx()__ . + * + * It's possible to reset all parameters to "default" using ZSTD_CCtx_reset(). * - * Functions that return size_t can be checked for errors using ZSTD_isError() - * and the ZSTD_ErrorCode can be extracted using ZSTD_getErrorCode(). + * This API supercedes all other "advanced" API entry points in the experimental section. + * In the future, we expect to remove from experimental API entry points which are redundant with this API. + */ + + +/* Compression strategies, listed from fastest to strongest */ +typedef enum { ZSTD_fast=1, + ZSTD_dfast=2, + ZSTD_greedy=3, + ZSTD_lazy=4, + ZSTD_lazy2=5, + ZSTD_btlazy2=6, + ZSTD_btopt=7, + ZSTD_btultra=8, + ZSTD_btultra2=9 + /* note : new strategies _might_ be added in the future. + Only the order (from fast to strong) is guaranteed */ +} ZSTD_strategy; + + +typedef enum { + + /* compression parameters + * Note: When compressing with a ZSTD_CDict these parameters are superseded + * by the parameters used to construct the ZSTD_CDict. + * See ZSTD_CCtx_refCDict() for more info (superseded-by-cdict). */ + ZSTD_c_compressionLevel=100, /* Set compression parameters according to pre-defined cLevel table. + * Note that exact compression parameters are dynamically determined, + * depending on both compression level and srcSize (when known). + * Default level is ZSTD_CLEVEL_DEFAULT==3. + * Special: value 0 means default, which is controlled by ZSTD_CLEVEL_DEFAULT. + * Note 1 : it's possible to pass a negative compression level. + * Note 2 : setting a level does not automatically set all other compression parameters + * to default. Setting this will however eventually dynamically impact the compression + * parameters which have not been manually set. The manually set + * ones will 'stick'. */ + /* Advanced compression parameters : + * It's possible to pin down compression parameters to some specific values. + * In which case, these values are no longer dynamically selected by the compressor */ + ZSTD_c_windowLog=101, /* Maximum allowed back-reference distance, expressed as power of 2. + * This will set a memory budget for streaming decompression, + * with larger values requiring more memory + * and typically compressing more. + * Must be clamped between ZSTD_WINDOWLOG_MIN and ZSTD_WINDOWLOG_MAX. + * Special: value 0 means "use default windowLog". + * Note: Using a windowLog greater than ZSTD_WINDOWLOG_LIMIT_DEFAULT + * requires explicitly allowing such size at streaming decompression stage. */ + ZSTD_c_hashLog=102, /* Size of the initial probe table, as a power of 2. + * Resulting memory usage is (1 << (hashLog+2)). + * Must be clamped between ZSTD_HASHLOG_MIN and ZSTD_HASHLOG_MAX. + * Larger tables improve compression ratio of strategies <= dFast, + * and improve speed of strategies > dFast. + * Special: value 0 means "use default hashLog". */ + ZSTD_c_chainLog=103, /* Size of the multi-probe search table, as a power of 2. + * Resulting memory usage is (1 << (chainLog+2)). + * Must be clamped between ZSTD_CHAINLOG_MIN and ZSTD_CHAINLOG_MAX. + * Larger tables result in better and slower compression. + * This parameter is useless for "fast" strategy. + * It's still useful when using "dfast" strategy, + * in which case it defines a secondary probe table. + * Special: value 0 means "use default chainLog". */ + ZSTD_c_searchLog=104, /* Number of search attempts, as a power of 2. + * More attempts result in better and slower compression. + * This parameter is useless for "fast" and "dFast" strategies. + * Special: value 0 means "use default searchLog". */ + ZSTD_c_minMatch=105, /* Minimum size of searched matches. + * Note that Zstandard can still find matches of smaller size, + * it just tweaks its search algorithm to look for this size and larger. + * Larger values increase compression and decompression speed, but decrease ratio. + * Must be clamped between ZSTD_MINMATCH_MIN and ZSTD_MINMATCH_MAX. + * Note that currently, for all strategies < btopt, effective minimum is 4. + * , for all strategies > fast, effective maximum is 6. + * Special: value 0 means "use default minMatchLength". */ + ZSTD_c_targetLength=106, /* Impact of this field depends on strategy. + * For strategies btopt, btultra & btultra2: + * Length of Match considered "good enough" to stop search. + * Larger values make compression stronger, and slower. + * For strategy fast: + * Distance between match sampling. + * Larger values make compression faster, and weaker. + * Special: value 0 means "use default targetLength". */ + ZSTD_c_strategy=107, /* See ZSTD_strategy enum definition. + * The higher the value of selected strategy, the more complex it is, + * resulting in stronger and slower compression. + * Special: value 0 means "use default strategy". */ + + /* LDM mode parameters */ + ZSTD_c_enableLongDistanceMatching=160, /* Enable long distance matching. + * This parameter is designed to improve compression ratio + * for large inputs, by finding large matches at long distance. + * It increases memory usage and window size. + * Note: enabling this parameter increases default ZSTD_c_windowLog to 128 MB + * except when expressly set to a different value. + * Note: will be enabled by default if ZSTD_c_windowLog >= 128 MB and + * compression strategy >= ZSTD_btopt (== compression level 16+) */ + ZSTD_c_ldmHashLog=161, /* Size of the table for long distance matching, as a power of 2. + * Larger values increase memory usage and compression ratio, + * but decrease compression speed. + * Must be clamped between ZSTD_HASHLOG_MIN and ZSTD_HASHLOG_MAX + * default: windowlog - 7. + * Special: value 0 means "automatically determine hashlog". */ + ZSTD_c_ldmMinMatch=162, /* Minimum match size for long distance matcher. + * Larger/too small values usually decrease compression ratio. + * Must be clamped between ZSTD_LDM_MINMATCH_MIN and ZSTD_LDM_MINMATCH_MAX. + * Special: value 0 means "use default value" (default: 64). */ + ZSTD_c_ldmBucketSizeLog=163, /* Log size of each bucket in the LDM hash table for collision resolution. + * Larger values improve collision resolution but decrease compression speed. + * The maximum value is ZSTD_LDM_BUCKETSIZELOG_MAX. + * Special: value 0 means "use default value" (default: 3). */ + ZSTD_c_ldmHashRateLog=164, /* Frequency of inserting/looking up entries into the LDM hash table. + * Must be clamped between 0 and (ZSTD_WINDOWLOG_MAX - ZSTD_HASHLOG_MIN). + * Default is MAX(0, (windowLog - ldmHashLog)), optimizing hash table usage. + * Larger values improve compression speed. + * Deviating far from default value will likely result in a compression ratio decrease. + * Special: value 0 means "automatically determine hashRateLog". */ + + /* frame parameters */ + ZSTD_c_contentSizeFlag=200, /* Content size will be written into frame header _whenever known_ (default:1) + * Content size must be known at the beginning of compression. + * This is automatically the case when using ZSTD_compress2(), + * For streaming scenarios, content size must be provided with ZSTD_CCtx_setPledgedSrcSize() */ + ZSTD_c_checksumFlag=201, /* A 32-bits checksum of content is written at end of frame (default:0) */ + ZSTD_c_dictIDFlag=202, /* When applicable, dictionary's ID is written into frame header (default:1) */ + + /* multi-threading parameters */ + /* These parameters are only active if multi-threading is enabled (compiled with build macro ZSTD_MULTITHREAD). + * Otherwise, trying to set any other value than default (0) will be a no-op and return an error. + * In a situation where it's unknown if the linked library supports multi-threading or not, + * setting ZSTD_c_nbWorkers to any value >= 1 and consulting the return value provides a quick way to check this property. + */ + ZSTD_c_nbWorkers=400, /* Select how many threads will be spawned to compress in parallel. + * When nbWorkers >= 1, triggers asynchronous mode when invoking ZSTD_compressStream*() : + * ZSTD_compressStream*() consumes input and flush output if possible, but immediately gives back control to caller, + * while compression is performed in parallel, within worker thread(s). + * (note : a strong exception to this rule is when first invocation of ZSTD_compressStream2() sets ZSTD_e_end : + * in which case, ZSTD_compressStream2() delegates to ZSTD_compress2(), which is always a blocking call). + * More workers improve speed, but also increase memory usage. + * Default value is `0`, aka "single-threaded mode" : no worker is spawned, + * compression is performed inside Caller's thread, and all invocations are blocking */ + ZSTD_c_jobSize=401, /* Size of a compression job. This value is enforced only when nbWorkers >= 1. + * Each compression job is completed in parallel, so this value can indirectly impact the nb of active threads. + * 0 means default, which is dynamically determined based on compression parameters. + * Job size must be a minimum of overlap size, or 1 MB, whichever is largest. + * The minimum size is automatically and transparently enforced. */ + ZSTD_c_overlapLog=402, /* Control the overlap size, as a fraction of window size. + * The overlap size is an amount of data reloaded from previous job at the beginning of a new job. + * It helps preserve compression ratio, while each job is compressed in parallel. + * This value is enforced only when nbWorkers >= 1. + * Larger values increase compression ratio, but decrease speed. + * Possible values range from 0 to 9 : + * - 0 means "default" : value will be determined by the library, depending on strategy + * - 1 means "no overlap" + * - 9 means "full overlap", using a full window size. + * Each intermediate rank increases/decreases load size by a factor 2 : + * 9: full window; 8: w/2; 7: w/4; 6: w/8; 5:w/16; 4: w/32; 3:w/64; 2:w/128; 1:no overlap; 0:default + * default value varies between 6 and 9, depending on strategy */ + + /* note : additional experimental parameters are also available + * within the experimental section of the API. + * At the time of this writing, they include : + * ZSTD_c_rsyncable + * ZSTD_c_format + * ZSTD_c_forceMaxWindow + * ZSTD_c_forceAttachDict + * ZSTD_c_literalCompressionMode + * ZSTD_c_targetCBlockSize + * ZSTD_c_srcSizeHint + * ZSTD_c_enableDedicatedDictSearch + * ZSTD_c_stableInBuffer + * ZSTD_c_stableOutBuffer + * ZSTD_c_blockDelimiters + * ZSTD_c_validateSequences + * Because they are not stable, it's necessary to define ZSTD_STATIC_LINKING_ONLY to access them. + * note : never ever use experimentalParam? names directly; + * also, the enums values themselves are unstable and can still change. + */ + ZSTD_c_experimentalParam1=500, + ZSTD_c_experimentalParam2=10, + ZSTD_c_experimentalParam3=1000, + ZSTD_c_experimentalParam4=1001, + ZSTD_c_experimentalParam5=1002, + ZSTD_c_experimentalParam6=1003, + ZSTD_c_experimentalParam7=1004, + ZSTD_c_experimentalParam8=1005, + ZSTD_c_experimentalParam9=1006, + ZSTD_c_experimentalParam10=1007, + ZSTD_c_experimentalParam11=1008, + ZSTD_c_experimentalParam12=1009 +} ZSTD_cParameter; + +typedef struct { + size_t error; + int lowerBound; + int upperBound; +} ZSTD_bounds; + +/*! ZSTD_cParam_getBounds() : + * All parameters must belong to an interval with lower and upper bounds, + * otherwise they will either trigger an error or be automatically clamped. + * @return : a structure, ZSTD_bounds, which contains + * - an error status field, which must be tested using ZSTD_isError() + * - lower and upper bounds, both inclusive + */ +ZSTDLIB_API ZSTD_bounds ZSTD_cParam_getBounds(ZSTD_cParameter cParam); + +/*! ZSTD_CCtx_setParameter() : + * Set one compression parameter, selected by enum ZSTD_cParameter. + * All parameters have valid bounds. Bounds can be queried using ZSTD_cParam_getBounds(). + * Providing a value beyond bound will either clamp it, or trigger an error (depending on parameter). + * Setting a parameter is generally only possible during frame initialization (before starting compression). + * Exception : when using multi-threading mode (nbWorkers >= 1), + * the following parameters can be updated _during_ compression (within same frame): + * => compressionLevel, hashLog, chainLog, searchLog, minMatch, targetLength and strategy. + * new parameters will be active for next job only (after a flush()). + * @return : an error code (which can be tested using ZSTD_isError()). */ +ZSTDLIB_API size_t ZSTD_CCtx_setParameter(ZSTD_CCtx* cctx, ZSTD_cParameter param, int value); + +/*! ZSTD_CCtx_setPledgedSrcSize() : + * Total input data size to be compressed as a single frame. + * Value will be written in frame header, unless if explicitly forbidden using ZSTD_c_contentSizeFlag. + * This value will also be controlled at end of frame, and trigger an error if not respected. + * @result : 0, or an error code (which can be tested with ZSTD_isError()). + * Note 1 : pledgedSrcSize==0 actually means zero, aka an empty frame. + * In order to mean "unknown content size", pass constant ZSTD_CONTENTSIZE_UNKNOWN. + * ZSTD_CONTENTSIZE_UNKNOWN is default value for any new frame. + * Note 2 : pledgedSrcSize is only valid once, for the next frame. + * It's discarded at the end of the frame, and replaced by ZSTD_CONTENTSIZE_UNKNOWN. + * Note 3 : Whenever all input data is provided and consumed in a single round, + * for example with ZSTD_compress2(), + * or invoking immediately ZSTD_compressStream2(,,,ZSTD_e_end), + * this value is automatically overridden by srcSize instead. + */ +ZSTDLIB_API size_t ZSTD_CCtx_setPledgedSrcSize(ZSTD_CCtx* cctx, unsigned long long pledgedSrcSize); + typedef enum { - ZSTD_error_no_error, - ZSTD_error_GENERIC, - ZSTD_error_prefix_unknown, - ZSTD_error_version_unsupported, - ZSTD_error_parameter_unknown, - ZSTD_error_frameParameter_unsupported, - ZSTD_error_frameParameter_unsupportedBy32bits, - ZSTD_error_frameParameter_windowTooLarge, - ZSTD_error_compressionParameter_unsupported, - ZSTD_error_init_missing, - ZSTD_error_memory_allocation, - ZSTD_error_stage_wrong, - ZSTD_error_dstSize_tooSmall, - ZSTD_error_srcSize_wrong, - ZSTD_error_corruption_detected, - ZSTD_error_checksum_wrong, - ZSTD_error_tableLog_tooLarge, - ZSTD_error_maxSymbolValue_tooLarge, - ZSTD_error_maxSymbolValue_tooSmall, - ZSTD_error_dictionary_corrupted, - ZSTD_error_dictionary_wrong, - ZSTD_error_dictionaryCreation_failed, - ZSTD_error_maxCode -} ZSTD_ErrorCode; - -/** - * ZSTD_maxCLevel() - maximum compression level available - * - * Return: Maximum compression level available. - */ -int ZSTD_maxCLevel(void); -/** - * ZSTD_compressBound() - maximum compressed size in worst case scenario - * @srcSize: The size of the data to compress. - * - * Return: The maximum compressed size in the worst case scenario. - */ -size_t ZSTD_compressBound(size_t srcSize); -/** - * ZSTD_isError() - tells if a size_t function result is an error code - * @code: The function result to check for error. - * - * Return: Non-zero iff the code is an error. - */ -static __attribute__((unused)) unsigned int ZSTD_isError(size_t code) -{ - return code > (size_t)-ZSTD_error_maxCode; -} -/** - * ZSTD_getErrorCode() - translates an error function result to a ZSTD_ErrorCode - * @functionResult: The result of a function for which ZSTD_isError() is true. - * - * Return: The ZSTD_ErrorCode corresponding to the functionResult or 0 - * if the functionResult isn't an error. - */ -static __attribute__((unused)) ZSTD_ErrorCode ZSTD_getErrorCode( - size_t functionResult) -{ - if (!ZSTD_isError(functionResult)) - return (ZSTD_ErrorCode)0; - return (ZSTD_ErrorCode)(0 - functionResult); -} - -/** - * enum ZSTD_strategy - zstd compression search strategy - * - * From faster to stronger. + ZSTD_reset_session_only = 1, + ZSTD_reset_parameters = 2, + ZSTD_reset_session_and_parameters = 3 +} ZSTD_ResetDirective; + +/*! ZSTD_CCtx_reset() : + * There are 2 different things that can be reset, independently or jointly : + * - The session : will stop compressing current frame, and make CCtx ready to start a new one. + * Useful after an error, or to interrupt any ongoing compression. + * Any internal data not yet flushed is cancelled. + * Compression parameters and dictionary remain unchanged. + * They will be used to compress next frame. + * Resetting session never fails. + * - The parameters : changes all parameters back to "default". + * This removes any reference to any dictionary too. + * Parameters can only be changed between 2 sessions (i.e. no compression is currently ongoing) + * otherwise the reset fails, and function returns an error value (which can be tested using ZSTD_isError()) + * - Both : similar to resetting the session, followed by resetting parameters. */ +ZSTDLIB_API size_t ZSTD_CCtx_reset(ZSTD_CCtx* cctx, ZSTD_ResetDirective reset); + +/*! ZSTD_compress2() : + * Behave the same as ZSTD_compressCCtx(), but compression parameters are set using the advanced API. + * ZSTD_compress2() always starts a new frame. + * Should cctx hold data from a previously unfinished frame, everything about it is forgotten. + * - Compression parameters are pushed into CCtx before starting compression, using ZSTD_CCtx_set*() + * - The function is always blocking, returns when compression is completed. + * Hint : compression runs faster if `dstCapacity` >= `ZSTD_compressBound(srcSize)`. + * @return : compressed size written into `dst` (<= `dstCapacity), + * or an error code if it fails (which can be tested using ZSTD_isError()). + */ +ZSTDLIB_API size_t ZSTD_compress2( ZSTD_CCtx* cctx, + void* dst, size_t dstCapacity, + const void* src, size_t srcSize); + + +/* ************************************* +* Advanced decompression API +***************************************/ + +/* The advanced API pushes parameters one by one into an existing DCtx context. + * Parameters are sticky, and remain valid for all following frames + * using the same DCtx context. + * It's possible to reset parameters to default values using ZSTD_DCtx_reset(). + * Note : This API is compatible with existing ZSTD_decompressDCtx() and ZSTD_decompressStream(). + * Therefore, no new decompression function is necessary. + */ + typedef enum { - ZSTD_fast, - ZSTD_dfast, - ZSTD_greedy, - ZSTD_lazy, - ZSTD_lazy2, - ZSTD_btlazy2, - ZSTD_btopt, - ZSTD_btopt2 -} ZSTD_strategy; -/** - * struct ZSTD_compressionParameters - zstd compression parameters - * @windowLog: Log of the largest match distance. Larger means more - * compression, and more memory needed during decompression. - * @chainLog: Fully searched segment. Larger means more compression, slower, - * and more memory (useless for fast). - * @hashLog: Dispatch table. Larger means more compression, - * slower, and more memory. - * @searchLog: Number of searches. Larger means more compression and slower. - * @searchLength: Match length searched. Larger means faster decompression, - * sometimes less compression. - * @targetLength: Acceptable match size for optimal parser (only). Larger means - * more compression, and slower. - * @strategy: The zstd compression strategy. + ZSTD_d_windowLogMax=100, /* Select a size limit (in power of 2) beyond which + * the streaming API will refuse to allocate memory buffer + * in order to protect the host from unreasonable memory requirements. + * This parameter is only useful in streaming mode, since no internal buffer is allocated in single-pass mode. + * By default, a decompression context accepts window sizes <= (1 << ZSTD_WINDOWLOG_LIMIT_DEFAULT). + * Special: value 0 means "use default maximum windowLog". */ + + /* note : additional experimental parameters are also available + * within the experimental section of the API. + * At the time of this writing, they include : + * ZSTD_d_format + * ZSTD_d_stableOutBuffer + * ZSTD_d_forceIgnoreChecksum + * ZSTD_d_refMultipleDDicts + * Because they are not stable, it's necessary to define ZSTD_STATIC_LINKING_ONLY to access them. + * note : never ever use experimentalParam? names directly + */ + ZSTD_d_experimentalParam1=1000, + ZSTD_d_experimentalParam2=1001, + ZSTD_d_experimentalParam3=1002, + ZSTD_d_experimentalParam4=1003 + +} ZSTD_dParameter; + +/*! ZSTD_dParam_getBounds() : + * All parameters must belong to an interval with lower and upper bounds, + * otherwise they will either trigger an error or be automatically clamped. + * @return : a structure, ZSTD_bounds, which contains + * - an error status field, which must be tested using ZSTD_isError() + * - both lower and upper bounds, inclusive */ -typedef struct { - unsigned int windowLog; - unsigned int chainLog; - unsigned int hashLog; - unsigned int searchLog; - unsigned int searchLength; - unsigned int targetLength; - ZSTD_strategy strategy; -} ZSTD_compressionParameters; +ZSTDLIB_API ZSTD_bounds ZSTD_dParam_getBounds(ZSTD_dParameter dParam); + +/*! ZSTD_DCtx_setParameter() : + * Set one compression parameter, selected by enum ZSTD_dParameter. + * All parameters have valid bounds. Bounds can be queried using ZSTD_dParam_getBounds(). + * Providing a value beyond bound will either clamp it, or trigger an error (depending on parameter). + * Setting a parameter is only possible during frame initialization (before starting decompression). + * @return : 0, or an error code (which can be tested using ZSTD_isError()). + */ +ZSTDLIB_API size_t ZSTD_DCtx_setParameter(ZSTD_DCtx* dctx, ZSTD_dParameter param, int value); + +/*! ZSTD_DCtx_reset() : + * Return a DCtx to clean state. + * Session and parameters can be reset jointly or separately. + * Parameters can only be reset when no active frame is being decompressed. + * @return : 0, or an error code, which can be tested with ZSTD_isError() + */ +ZSTDLIB_API size_t ZSTD_DCtx_reset(ZSTD_DCtx* dctx, ZSTD_ResetDirective reset); + + +/* ************************** +* Streaming +****************************/ + +typedef struct ZSTD_inBuffer_s { + const void* src; /*< start of input buffer */ + size_t size; /*< size of input buffer */ + size_t pos; /*< position where reading stopped. Will be updated. Necessarily 0 <= pos <= size */ +} ZSTD_inBuffer; + +typedef struct ZSTD_outBuffer_s { + void* dst; /*< start of output buffer */ + size_t size; /*< size of output buffer */ + size_t pos; /*< position where writing stopped. Will be updated. Necessarily 0 <= pos <= size */ +} ZSTD_outBuffer; + -/** - * struct ZSTD_frameParameters - zstd frame parameters - * @contentSizeFlag: Controls whether content size will be present in the frame - * header (when known). - * @checksumFlag: Controls whether a 32-bit checksum is generated at the end - * of the frame for error detection. - * @noDictIDFlag: Controls whether dictID will be saved into the frame header - * when using dictionary compression. + +/*-*********************************************************************** +* Streaming compression - HowTo +* +* A ZSTD_CStream object is required to track streaming operation. +* Use ZSTD_createCStream() and ZSTD_freeCStream() to create/release resources. +* ZSTD_CStream objects can be reused multiple times on consecutive compression operations. +* It is recommended to re-use ZSTD_CStream since it will play nicer with system's memory, by re-using already allocated memory. +* +* For parallel execution, use one separate ZSTD_CStream per thread. +* +* note : since v1.3.0, ZSTD_CStream and ZSTD_CCtx are the same thing. +* +* Parameters are sticky : when starting a new compression on the same context, +* it will re-use the same sticky parameters as previous compression session. +* When in doubt, it's recommended to fully initialize the context before usage. +* Use ZSTD_CCtx_reset() to reset the context and ZSTD_CCtx_setParameter(), +* ZSTD_CCtx_setPledgedSrcSize(), or ZSTD_CCtx_loadDictionary() and friends to +* set more specific parameters, the pledged source size, or load a dictionary. +* +* Use ZSTD_compressStream2() with ZSTD_e_continue as many times as necessary to +* consume input stream. The function will automatically update both `pos` +* fields within `input` and `output`. +* Note that the function may not consume the entire input, for example, because +* the output buffer is already full, in which case `input.pos < input.size`. +* The caller must check if input has been entirely consumed. +* If not, the caller must make some room to receive more compressed data, +* and then present again remaining input data. +* note: ZSTD_e_continue is guaranteed to make some forward progress when called, +* but doesn't guarantee maximal forward progress. This is especially relevant +* when compressing with multiple threads. The call won't block if it can +* consume some input, but if it can't it will wait for some, but not all, +* output to be flushed. +* @return : provides a minimum amount of data remaining to be flushed from internal buffers +* or an error code, which can be tested using ZSTD_isError(). +* +* At any moment, it's possible to flush whatever data might remain stuck within internal buffer, +* using ZSTD_compressStream2() with ZSTD_e_flush. `output->pos` will be updated. +* Note that, if `output->size` is too small, a single invocation with ZSTD_e_flush might not be enough (return code > 0). +* In which case, make some room to receive more compressed data, and call again ZSTD_compressStream2() with ZSTD_e_flush. +* You must continue calling ZSTD_compressStream2() with ZSTD_e_flush until it returns 0, at which point you can change the +* operation. +* note: ZSTD_e_flush will flush as much output as possible, meaning when compressing with multiple threads, it will +* block until the flush is complete or the output buffer is full. +* @return : 0 if internal buffers are entirely flushed, +* >0 if some data still present within internal buffer (the value is minimal estimation of remaining size), +* or an error code, which can be tested using ZSTD_isError(). +* +* Calling ZSTD_compressStream2() with ZSTD_e_end instructs to finish a frame. +* It will perform a flush and write frame epilogue. +* The epilogue is required for decoders to consider a frame completed. +* flush operation is the same, and follows same rules as calling ZSTD_compressStream2() with ZSTD_e_flush. +* You must continue calling ZSTD_compressStream2() with ZSTD_e_end until it returns 0, at which point you are free to +* start a new frame. +* note: ZSTD_e_end will flush as much output as possible, meaning when compressing with multiple threads, it will +* block until the flush is complete or the output buffer is full. +* @return : 0 if frame fully completed and fully flushed, +* >0 if some data still present within internal buffer (the value is minimal estimation of remaining size), +* or an error code, which can be tested using ZSTD_isError(). +* +* *******************************************************************/ + +typedef ZSTD_CCtx ZSTD_CStream; /*< CCtx and CStream are now effectively same object (>= v1.3.0) */ + /* Continue to distinguish them for compatibility with older versions <= v1.2.0 */ +/*===== ZSTD_CStream management functions =====*/ +ZSTDLIB_API ZSTD_CStream* ZSTD_createCStream(void); +ZSTDLIB_API size_t ZSTD_freeCStream(ZSTD_CStream* zcs); /* accept NULL pointer */ + +/*===== Streaming compression functions =====*/ +typedef enum { + ZSTD_e_continue=0, /* collect more data, encoder decides when to output compressed result, for optimal compression ratio */ + ZSTD_e_flush=1, /* flush any data provided so far, + * it creates (at least) one new block, that can be decoded immediately on reception; + * frame will continue: any future data can still reference previously compressed data, improving compression. + * note : multithreaded compression will block to flush as much output as possible. */ + ZSTD_e_end=2 /* flush any remaining data _and_ close current frame. + * note that frame is only closed after compressed data is fully flushed (return value == 0). + * After that point, any additional data starts a new frame. + * note : each frame is independent (does not reference any content from previous frame). + : note : multithreaded compression will block to flush as much output as possible. */ +} ZSTD_EndDirective; + +/*! ZSTD_compressStream2() : + * Behaves about the same as ZSTD_compressStream, with additional control on end directive. + * - Compression parameters are pushed into CCtx before starting compression, using ZSTD_CCtx_set*() + * - Compression parameters cannot be changed once compression is started (save a list of exceptions in multi-threading mode) + * - output->pos must be <= dstCapacity, input->pos must be <= srcSize + * - output->pos and input->pos will be updated. They are guaranteed to remain below their respective limit. + * - endOp must be a valid directive + * - When nbWorkers==0 (default), function is blocking : it completes its job before returning to caller. + * - When nbWorkers>=1, function is non-blocking : it copies a portion of input, distributes jobs to internal worker threads, flush to output whatever is available, + * and then immediately returns, just indicating that there is some data remaining to be flushed. + * The function nonetheless guarantees forward progress : it will return only after it reads or write at least 1+ byte. + * - Exception : if the first call requests a ZSTD_e_end directive and provides enough dstCapacity, the function delegates to ZSTD_compress2() which is always blocking. + * - @return provides a minimum amount of data remaining to be flushed from internal buffers + * or an error code, which can be tested using ZSTD_isError(). + * if @return != 0, flush is not fully completed, there is still some data left within internal buffers. + * This is useful for ZSTD_e_flush, since in this case more flushes are necessary to empty all buffers. + * For ZSTD_e_end, @return == 0 when internal buffers are fully flushed and frame is completed. + * - after a ZSTD_e_end directive, if internal buffer is not fully flushed (@return != 0), + * only ZSTD_e_end or ZSTD_e_flush operations are allowed. + * Before starting a new compression job, or changing compression parameters, + * it is required to fully flush internal buffers. + */ +ZSTDLIB_API size_t ZSTD_compressStream2( ZSTD_CCtx* cctx, + ZSTD_outBuffer* output, + ZSTD_inBuffer* input, + ZSTD_EndDirective endOp); + + +/* These buffer sizes are softly recommended. + * They are not required : ZSTD_compressStream*() happily accepts any buffer size, for both input and output. + * Respecting the recommended size just makes it a bit easier for ZSTD_compressStream*(), + * reducing the amount of memory shuffling and buffering, resulting in minor performance savings. + * + * However, note that these recommendations are from the perspective of a C caller program. + * If the streaming interface is invoked from some other language, + * especially managed ones such as Java or Go, through a foreign function interface such as jni or cgo, + * a major performance rule is to reduce crossing such interface to an absolute minimum. + * It's not rare that performance ends being spent more into the interface, rather than compression itself. + * In which cases, prefer using large buffers, as large as practical, + * for both input and output, to reduce the nb of roundtrips. + */ +ZSTDLIB_API size_t ZSTD_CStreamInSize(void); /*< recommended size for input buffer */ +ZSTDLIB_API size_t ZSTD_CStreamOutSize(void); /*< recommended size for output buffer. Guarantee to successfully flush at least one complete compressed block. */ + + +/* ***************************************************************************** + * This following is a legacy streaming API. + * It can be replaced by ZSTD_CCtx_reset() and ZSTD_compressStream2(). + * It is redundant, but remains fully supported. + * Advanced parameters and dictionary compression can only be used through the + * new API. + ******************************************************************************/ + +/*! + * Equivalent to: * - * The default value is all fields set to 0. + * ZSTD_CCtx_reset(zcs, ZSTD_reset_session_only); + * ZSTD_CCtx_refCDict(zcs, NULL); // clear the dictionary (if any) + * ZSTD_CCtx_setParameter(zcs, ZSTD_c_compressionLevel, compressionLevel); + */ +ZSTDLIB_API size_t ZSTD_initCStream(ZSTD_CStream* zcs, int compressionLevel); +/*! + * Alternative for ZSTD_compressStream2(zcs, output, input, ZSTD_e_continue). + * NOTE: The return value is different. ZSTD_compressStream() returns a hint for + * the next read size (if non-zero and not an error). ZSTD_compressStream2() + * returns the minimum nb of bytes left to flush (if non-zero and not an error). */ +ZSTDLIB_API size_t ZSTD_compressStream(ZSTD_CStream* zcs, ZSTD_outBuffer* output, ZSTD_inBuffer* input); +/*! Equivalent to ZSTD_compressStream2(zcs, output, &emptyInput, ZSTD_e_flush). */ +ZSTDLIB_API size_t ZSTD_flushStream(ZSTD_CStream* zcs, ZSTD_outBuffer* output); +/*! Equivalent to ZSTD_compressStream2(zcs, output, &emptyInput, ZSTD_e_end). */ +ZSTDLIB_API size_t ZSTD_endStream(ZSTD_CStream* zcs, ZSTD_outBuffer* output); + + +/*-*************************************************************************** +* Streaming decompression - HowTo +* +* A ZSTD_DStream object is required to track streaming operations. +* Use ZSTD_createDStream() and ZSTD_freeDStream() to create/release resources. +* ZSTD_DStream objects can be re-used multiple times. +* +* Use ZSTD_initDStream() to start a new decompression operation. +* @return : recommended first input size +* Alternatively, use advanced API to set specific properties. +* +* Use ZSTD_decompressStream() repetitively to consume your input. +* The function will update both `pos` fields. +* If `input.pos < input.size`, some input has not been consumed. +* It's up to the caller to present again remaining data. +* The function tries to flush all data decoded immediately, respecting output buffer size. +* If `output.pos < output.size`, decoder has flushed everything it could. +* But if `output.pos == output.size`, there might be some data left within internal buffers., +* In which case, call ZSTD_decompressStream() again to flush whatever remains in the buffer. +* Note : with no additional input provided, amount of data flushed is necessarily <= ZSTD_BLOCKSIZE_MAX. +* @return : 0 when a frame is completely decoded and fully flushed, +* or an error code, which can be tested using ZSTD_isError(), +* or any other value > 0, which means there is still some decoding or flushing to do to complete current frame : +* the return value is a suggested next input size (just a hint for better latency) +* that will never request more than the remaining frame size. +* *******************************************************************************/ + +typedef ZSTD_DCtx ZSTD_DStream; /*< DCtx and DStream are now effectively same object (>= v1.3.0) */ + /* For compatibility with versions <= v1.2.0, prefer differentiating them. */ +/*===== ZSTD_DStream management functions =====*/ +ZSTDLIB_API ZSTD_DStream* ZSTD_createDStream(void); +ZSTDLIB_API size_t ZSTD_freeDStream(ZSTD_DStream* zds); /* accept NULL pointer */ + +/*===== Streaming decompression functions =====*/ + +/* This function is redundant with the advanced API and equivalent to: + * + * ZSTD_DCtx_reset(zds, ZSTD_reset_session_only); + * ZSTD_DCtx_refDDict(zds, NULL); + */ +ZSTDLIB_API size_t ZSTD_initDStream(ZSTD_DStream* zds); + +ZSTDLIB_API size_t ZSTD_decompressStream(ZSTD_DStream* zds, ZSTD_outBuffer* output, ZSTD_inBuffer* input); + +ZSTDLIB_API size_t ZSTD_DStreamInSize(void); /*!< recommended size for input buffer */ +ZSTDLIB_API size_t ZSTD_DStreamOutSize(void); /*!< recommended size for output buffer. Guarantee to successfully flush at least one complete block in all circumstances. */ + + +/* ************************ +* Simple dictionary API +***************************/ +/*! ZSTD_compress_usingDict() : + * Compression at an explicit compression level using a Dictionary. + * A dictionary can be any arbitrary data segment (also called a prefix), + * or a buffer with specified information (see dictBuilder/zdict.h). + * Note : This function loads the dictionary, resulting in significant startup delay. + * It's intended for a dictionary used only once. + * Note 2 : When `dict == NULL || dictSize < 8` no dictionary is used. */ +ZSTDLIB_API size_t ZSTD_compress_usingDict(ZSTD_CCtx* ctx, + void* dst, size_t dstCapacity, + const void* src, size_t srcSize, + const void* dict,size_t dictSize, + int compressionLevel); + +/*! ZSTD_decompress_usingDict() : + * Decompression using a known Dictionary. + * Dictionary must be identical to the one used during compression. + * Note : This function loads the dictionary, resulting in significant startup delay. + * It's intended for a dictionary used only once. + * Note : When `dict == NULL || dictSize < 8` no dictionary is used. */ +ZSTDLIB_API size_t ZSTD_decompress_usingDict(ZSTD_DCtx* dctx, + void* dst, size_t dstCapacity, + const void* src, size_t srcSize, + const void* dict,size_t dictSize); + + +/* ********************************* + * Bulk processing dictionary API + **********************************/ +typedef struct ZSTD_CDict_s ZSTD_CDict; + +/*! ZSTD_createCDict() : + * When compressing multiple messages or blocks using the same dictionary, + * it's recommended to digest the dictionary only once, since it's a costly operation. + * ZSTD_createCDict() will create a state from digesting a dictionary. + * The resulting state can be used for future compression operations with very limited startup cost. + * ZSTD_CDict can be created once and shared by multiple threads concurrently, since its usage is read-only. + * @dictBuffer can be released after ZSTD_CDict creation, because its content is copied within CDict. + * Note 1 : Consider experimental function `ZSTD_createCDict_byReference()` if you prefer to not duplicate @dictBuffer content. + * Note 2 : A ZSTD_CDict can be created from an empty @dictBuffer, + * in which case the only thing that it transports is the @compressionLevel. + * This can be useful in a pipeline featuring ZSTD_compress_usingCDict() exclusively, + * expecting a ZSTD_CDict parameter with any data, including those without a known dictionary. */ +ZSTDLIB_API ZSTD_CDict* ZSTD_createCDict(const void* dictBuffer, size_t dictSize, + int compressionLevel); + +/*! ZSTD_freeCDict() : + * Function frees memory allocated by ZSTD_createCDict(). + * If a NULL pointer is passed, no operation is performed. */ +ZSTDLIB_API size_t ZSTD_freeCDict(ZSTD_CDict* CDict); + +/*! ZSTD_compress_usingCDict() : + * Compression using a digested Dictionary. + * Recommended when same dictionary is used multiple times. + * Note : compression level is _decided at dictionary creation time_, + * and frame parameters are hardcoded (dictID=yes, contentSize=yes, checksum=no) */ +ZSTDLIB_API size_t ZSTD_compress_usingCDict(ZSTD_CCtx* cctx, + void* dst, size_t dstCapacity, + const void* src, size_t srcSize, + const ZSTD_CDict* cdict); + + +typedef struct ZSTD_DDict_s ZSTD_DDict; + +/*! ZSTD_createDDict() : + * Create a digested dictionary, ready to start decompression operation without startup delay. + * dictBuffer can be released after DDict creation, as its content is copied inside DDict. */ +ZSTDLIB_API ZSTD_DDict* ZSTD_createDDict(const void* dictBuffer, size_t dictSize); + +/*! ZSTD_freeDDict() : + * Function frees memory allocated with ZSTD_createDDict() + * If a NULL pointer is passed, no operation is performed. */ +ZSTDLIB_API size_t ZSTD_freeDDict(ZSTD_DDict* ddict); + +/*! ZSTD_decompress_usingDDict() : + * Decompression using a digested Dictionary. + * Recommended when same dictionary is used multiple times. */ +ZSTDLIB_API size_t ZSTD_decompress_usingDDict(ZSTD_DCtx* dctx, + void* dst, size_t dstCapacity, + const void* src, size_t srcSize, + const ZSTD_DDict* ddict); + + +/* ****************************** + * Dictionary helper functions + *******************************/ + +/*! ZSTD_getDictID_fromDict() : + * Provides the dictID stored within dictionary. + * if @return == 0, the dictionary is not conformant with Zstandard specification. + * It can still be loaded, but as a content-only dictionary. */ +ZSTDLIB_API unsigned ZSTD_getDictID_fromDict(const void* dict, size_t dictSize); + +/*! ZSTD_getDictID_fromDDict() : + * Provides the dictID of the dictionary loaded into `ddict`. + * If @return == 0, the dictionary is not conformant to Zstandard specification, or empty. + * Non-conformant dictionaries can still be loaded, but as content-only dictionaries. */ +ZSTDLIB_API unsigned ZSTD_getDictID_fromDDict(const ZSTD_DDict* ddict); + +/*! ZSTD_getDictID_fromFrame() : + * Provides the dictID required to decompressed the frame stored within `src`. + * If @return == 0, the dictID could not be decoded. + * This could for one of the following reasons : + * - The frame does not require a dictionary to be decoded (most common case). + * - The frame was built with dictID intentionally removed. Whatever dictionary is necessary is a hidden information. + * Note : this use case also happens when using a non-conformant dictionary. + * - `srcSize` is too small, and as a result, the frame header could not be decoded (only possible if `srcSize < ZSTD_FRAMEHEADERSIZE_MAX`). + * - This is not a Zstandard frame. + * When identifying the exact failure cause, it's possible to use ZSTD_getFrameHeader(), which will provide a more precise error code. */ +ZSTDLIB_API unsigned ZSTD_getDictID_fromFrame(const void* src, size_t srcSize); + + +/* ***************************************************************************** + * Advanced dictionary and prefix API + * + * This API allows dictionaries to be used with ZSTD_compress2(), + * ZSTD_compressStream2(), and ZSTD_decompress(). Dictionaries are sticky, and + * only reset with the context is reset with ZSTD_reset_parameters or + * ZSTD_reset_session_and_parameters. Prefixes are single-use. + ******************************************************************************/ + + +/*! ZSTD_CCtx_loadDictionary() : + * Create an internal CDict from `dict` buffer. + * Decompression will have to use same dictionary. + * @result : 0, or an error code (which can be tested with ZSTD_isError()). + * Special: Loading a NULL (or 0-size) dictionary invalidates previous dictionary, + * meaning "return to no-dictionary mode". + * Note 1 : Dictionary is sticky, it will be used for all future compressed frames. + * To return to "no-dictionary" situation, load a NULL dictionary (or reset parameters). + * Note 2 : Loading a dictionary involves building tables. + * It's also a CPU consuming operation, with non-negligible impact on latency. + * Tables are dependent on compression parameters, and for this reason, + * compression parameters can no longer be changed after loading a dictionary. + * Note 3 :`dict` content will be copied internally. + * Use experimental ZSTD_CCtx_loadDictionary_byReference() to reference content instead. + * In such a case, dictionary buffer must outlive its users. + * Note 4 : Use ZSTD_CCtx_loadDictionary_advanced() + * to precisely select how dictionary content must be interpreted. */ +ZSTDLIB_API size_t ZSTD_CCtx_loadDictionary(ZSTD_CCtx* cctx, const void* dict, size_t dictSize); + +/*! ZSTD_CCtx_refCDict() : + * Reference a prepared dictionary, to be used for all next compressed frames. + * Note that compression parameters are enforced from within CDict, + * and supersede any compression parameter previously set within CCtx. + * The parameters ignored are labelled as "superseded-by-cdict" in the ZSTD_cParameter enum docs. + * The ignored parameters will be used again if the CCtx is returned to no-dictionary mode. + * The dictionary will remain valid for future compressed frames using same CCtx. + * @result : 0, or an error code (which can be tested with ZSTD_isError()). + * Special : Referencing a NULL CDict means "return to no-dictionary mode". + * Note 1 : Currently, only one dictionary can be managed. + * Referencing a new dictionary effectively "discards" any previous one. + * Note 2 : CDict is just referenced, its lifetime must outlive its usage within CCtx. */ +ZSTDLIB_API size_t ZSTD_CCtx_refCDict(ZSTD_CCtx* cctx, const ZSTD_CDict* cdict); + +/*! ZSTD_CCtx_refPrefix() : + * Reference a prefix (single-usage dictionary) for next compressed frame. + * A prefix is **only used once**. Tables are discarded at end of frame (ZSTD_e_end). + * Decompression will need same prefix to properly regenerate data. + * Compressing with a prefix is similar in outcome as performing a diff and compressing it, + * but performs much faster, especially during decompression (compression speed is tunable with compression level). + * @result : 0, or an error code (which can be tested with ZSTD_isError()). + * Special: Adding any prefix (including NULL) invalidates any previous prefix or dictionary + * Note 1 : Prefix buffer is referenced. It **must** outlive compression. + * Its content must remain unmodified during compression. + * Note 2 : If the intention is to diff some large src data blob with some prior version of itself, + * ensure that the window size is large enough to contain the entire source. + * See ZSTD_c_windowLog. + * Note 3 : Referencing a prefix involves building tables, which are dependent on compression parameters. + * It's a CPU consuming operation, with non-negligible impact on latency. + * If there is a need to use the same prefix multiple times, consider loadDictionary instead. + * Note 4 : By default, the prefix is interpreted as raw content (ZSTD_dct_rawContent). + * Use experimental ZSTD_CCtx_refPrefix_advanced() to alter dictionary interpretation. */ +ZSTDLIB_API size_t ZSTD_CCtx_refPrefix(ZSTD_CCtx* cctx, + const void* prefix, size_t prefixSize); + +/*! ZSTD_DCtx_loadDictionary() : + * Create an internal DDict from dict buffer, + * to be used to decompress next frames. + * The dictionary remains valid for all future frames, until explicitly invalidated. + * @result : 0, or an error code (which can be tested with ZSTD_isError()). + * Special : Adding a NULL (or 0-size) dictionary invalidates any previous dictionary, + * meaning "return to no-dictionary mode". + * Note 1 : Loading a dictionary involves building tables, + * which has a non-negligible impact on CPU usage and latency. + * It's recommended to "load once, use many times", to amortize the cost + * Note 2 :`dict` content will be copied internally, so `dict` can be released after loading. + * Use ZSTD_DCtx_loadDictionary_byReference() to reference dictionary content instead. + * Note 3 : Use ZSTD_DCtx_loadDictionary_advanced() to take control of + * how dictionary content is loaded and interpreted. + */ +ZSTDLIB_API size_t ZSTD_DCtx_loadDictionary(ZSTD_DCtx* dctx, const void* dict, size_t dictSize); + +/*! ZSTD_DCtx_refDDict() : + * Reference a prepared dictionary, to be used to decompress next frames. + * The dictionary remains active for decompression of future frames using same DCtx. + * + * If called with ZSTD_d_refMultipleDDicts enabled, repeated calls of this function + * will store the DDict references in a table, and the DDict used for decompression + * will be determined at decompression time, as per the dict ID in the frame. + * The memory for the table is allocated on the first call to refDDict, and can be + * freed with ZSTD_freeDCtx(). + * + * @result : 0, or an error code (which can be tested with ZSTD_isError()). + * Note 1 : Currently, only one dictionary can be managed. + * Referencing a new dictionary effectively "discards" any previous one. + * Special: referencing a NULL DDict means "return to no-dictionary mode". + * Note 2 : DDict is just referenced, its lifetime must outlive its usage from DCtx. + */ +ZSTDLIB_API size_t ZSTD_DCtx_refDDict(ZSTD_DCtx* dctx, const ZSTD_DDict* ddict); + +/*! ZSTD_DCtx_refPrefix() : + * Reference a prefix (single-usage dictionary) to decompress next frame. + * This is the reverse operation of ZSTD_CCtx_refPrefix(), + * and must use the same prefix as the one used during compression. + * Prefix is **only used once**. Reference is discarded at end of frame. + * End of frame is reached when ZSTD_decompressStream() returns 0. + * @result : 0, or an error code (which can be tested with ZSTD_isError()). + * Note 1 : Adding any prefix (including NULL) invalidates any previously set prefix or dictionary + * Note 2 : Prefix buffer is referenced. It **must** outlive decompression. + * Prefix buffer must remain unmodified up to the end of frame, + * reached when ZSTD_decompressStream() returns 0. + * Note 3 : By default, the prefix is treated as raw content (ZSTD_dct_rawContent). + * Use ZSTD_CCtx_refPrefix_advanced() to alter dictMode (Experimental section) + * Note 4 : Referencing a raw content prefix has almost no cpu nor memory cost. + * A full dictionary is more costly, as it requires building tables. + */ +ZSTDLIB_API size_t ZSTD_DCtx_refPrefix(ZSTD_DCtx* dctx, + const void* prefix, size_t prefixSize); + +/* === Memory management === */ + +/*! ZSTD_sizeof_*() : + * These functions give the _current_ memory usage of selected object. + * Note that object memory usage can evolve (increase or decrease) over time. */ +ZSTDLIB_API size_t ZSTD_sizeof_CCtx(const ZSTD_CCtx* cctx); +ZSTDLIB_API size_t ZSTD_sizeof_DCtx(const ZSTD_DCtx* dctx); +ZSTDLIB_API size_t ZSTD_sizeof_CStream(const ZSTD_CStream* zcs); +ZSTDLIB_API size_t ZSTD_sizeof_DStream(const ZSTD_DStream* zds); +ZSTDLIB_API size_t ZSTD_sizeof_CDict(const ZSTD_CDict* cdict); +ZSTDLIB_API size_t ZSTD_sizeof_DDict(const ZSTD_DDict* ddict); + +#endif /* ZSTD_H_235446 */ + + +/* ************************************************************************************** + * ADVANCED AND EXPERIMENTAL FUNCTIONS + **************************************************************************************** + * The definitions in the following section are considered experimental. + * They are provided for advanced scenarios. + * They should never be used with a dynamic library, as prototypes may change in the future. + * Use them only in association with static linking. + * ***************************************************************************************/ + +#if !defined(ZSTD_H_ZSTD_STATIC_LINKING_ONLY) +#define ZSTD_H_ZSTD_STATIC_LINKING_ONLY + +/* ************************************************************************************** + * experimental API (static linking only) + **************************************************************************************** + * The following symbols and constants + * are not planned to join "stable API" status in the near future. + * They can still change in future versions. + * Some of them are planned to remain in the static_only section indefinitely. + * Some of them might be removed in the future (especially when redundant with existing stable functions) + * ***************************************************************************************/ + +#define ZSTD_FRAMEHEADERSIZE_PREFIX(format) ((format) == ZSTD_f_zstd1 ? 5 : 1) /* minimum input size required to query frame header size */ +#define ZSTD_FRAMEHEADERSIZE_MIN(format) ((format) == ZSTD_f_zstd1 ? 6 : 2) +#define ZSTD_FRAMEHEADERSIZE_MAX 18 /* can be useful for static allocation */ +#define ZSTD_SKIPPABLEHEADERSIZE 8 + +/* compression parameter bounds */ +#define ZSTD_WINDOWLOG_MAX_32 30 +#define ZSTD_WINDOWLOG_MAX_64 31 +#define ZSTD_WINDOWLOG_MAX ((int)(sizeof(size_t) == 4 ? ZSTD_WINDOWLOG_MAX_32 : ZSTD_WINDOWLOG_MAX_64)) +#define ZSTD_WINDOWLOG_MIN 10 +#define ZSTD_HASHLOG_MAX ((ZSTD_WINDOWLOG_MAX < 30) ? ZSTD_WINDOWLOG_MAX : 30) +#define ZSTD_HASHLOG_MIN 6 +#define ZSTD_CHAINLOG_MAX_32 29 +#define ZSTD_CHAINLOG_MAX_64 30 +#define ZSTD_CHAINLOG_MAX ((int)(sizeof(size_t) == 4 ? ZSTD_CHAINLOG_MAX_32 : ZSTD_CHAINLOG_MAX_64)) +#define ZSTD_CHAINLOG_MIN ZSTD_HASHLOG_MIN +#define ZSTD_SEARCHLOG_MAX (ZSTD_WINDOWLOG_MAX-1) +#define ZSTD_SEARCHLOG_MIN 1 +#define ZSTD_MINMATCH_MAX 7 /* only for ZSTD_fast, other strategies are limited to 6 */ +#define ZSTD_MINMATCH_MIN 3 /* only for ZSTD_btopt+, faster strategies are limited to 4 */ +#define ZSTD_TARGETLENGTH_MAX ZSTD_BLOCKSIZE_MAX +#define ZSTD_TARGETLENGTH_MIN 0 /* note : comparing this constant to an unsigned results in a tautological test */ +#define ZSTD_STRATEGY_MIN ZSTD_fast +#define ZSTD_STRATEGY_MAX ZSTD_btultra2 + + +#define ZSTD_OVERLAPLOG_MIN 0 +#define ZSTD_OVERLAPLOG_MAX 9 + +#define ZSTD_WINDOWLOG_LIMIT_DEFAULT 27 /* by default, the streaming decoder will refuse any frame + * requiring larger than (1< 0: + * If litLength != 0: + * rep == 1 --> offset == repeat_offset_1 + * rep == 2 --> offset == repeat_offset_2 + * rep == 3 --> offset == repeat_offset_3 + * If litLength == 0: + * rep == 1 --> offset == repeat_offset_2 + * rep == 2 --> offset == repeat_offset_3 + * rep == 3 --> offset == repeat_offset_1 - 1 + * + * Note: This field is optional. ZSTD_generateSequences() will calculate the value of + * 'rep', but repeat offsets do not necessarily need to be calculated from an external + * sequence provider's perspective. For example, ZSTD_compressSequences() does not + * use this 'rep' field at all (as of now). + */ +} ZSTD_Sequence; + +typedef struct { + unsigned windowLog; /*< largest match distance : larger == more compression, more memory needed during decompression */ + unsigned chainLog; /*< fully searched segment : larger == more compression, slower, more memory (useless for fast) */ + unsigned hashLog; /*< dispatch table : larger == faster, more memory */ + unsigned searchLog; /*< nb of searches : larger == more compression, slower */ + unsigned minMatch; /*< match length searched : larger == faster decompression, sometimes less compression */ + unsigned targetLength; /*< acceptable match size for optimal parser (only) : larger == more compression, slower */ + ZSTD_strategy strategy; /*< see ZSTD_strategy definition above */ +} ZSTD_compressionParameters; + +typedef struct { + int contentSizeFlag; /*< 1: content size will be in frame header (when known) */ + int checksumFlag; /*< 1: generate a 32-bits checksum using XXH64 algorithm at end of frame, for error detection */ + int noDictIDFlag; /*< 1: no dictID will be saved into frame header (dictID is only useful for dictionary compression) */ } ZSTD_frameParameters; -/** - * struct ZSTD_parameters - zstd parameters - * @cParams: The compression parameters. - * @fParams: The frame parameters. - */ typedef struct { - ZSTD_compressionParameters cParams; - ZSTD_frameParameters fParams; + ZSTD_compressionParameters cParams; + ZSTD_frameParameters fParams; } ZSTD_parameters; -/** - * ZSTD_getCParams() - returns ZSTD_compressionParameters for selected level - * @compressionLevel: The compression level from 1 to ZSTD_maxCLevel(). - * @estimatedSrcSize: The estimated source size to compress or 0 if unknown. - * @dictSize: The dictionary size or 0 if a dictionary isn't being used. - * - * Return: The selected ZSTD_compressionParameters. +typedef enum { + ZSTD_dct_auto = 0, /* dictionary is "full" when starting with ZSTD_MAGIC_DICTIONARY, otherwise it is "rawContent" */ + ZSTD_dct_rawContent = 1, /* ensures dictionary is always loaded as rawContent, even if it starts with ZSTD_MAGIC_DICTIONARY */ + ZSTD_dct_fullDict = 2 /* refuses to load a dictionary if it does not respect Zstandard's specification, starting with ZSTD_MAGIC_DICTIONARY */ +} ZSTD_dictContentType_e; + +typedef enum { + ZSTD_dlm_byCopy = 0, /*< Copy dictionary content internally */ + ZSTD_dlm_byRef = 1 /*< Reference dictionary content -- the dictionary buffer must outlive its users. */ +} ZSTD_dictLoadMethod_e; + +typedef enum { + ZSTD_f_zstd1 = 0, /* zstd frame format, specified in zstd_compression_format.md (default) */ + ZSTD_f_zstd1_magicless = 1 /* Variant of zstd frame format, without initial 4-bytes magic number. + * Useful to save 4 bytes per generated frame. + * Decoder cannot recognise automatically this format, requiring this instruction. */ +} ZSTD_format_e; + +typedef enum { + /* Note: this enum controls ZSTD_d_forceIgnoreChecksum */ + ZSTD_d_validateChecksum = 0, + ZSTD_d_ignoreChecksum = 1 +} ZSTD_forceIgnoreChecksum_e; + +typedef enum { + /* Note: this enum controls ZSTD_d_refMultipleDDicts */ + ZSTD_rmd_refSingleDDict = 0, + ZSTD_rmd_refMultipleDDicts = 1 +} ZSTD_refMultipleDDicts_e; + +typedef enum { + /* Note: this enum and the behavior it controls are effectively internal + * implementation details of the compressor. They are expected to continue + * to evolve and should be considered only in the context of extremely + * advanced performance tuning. + * + * Zstd currently supports the use of a CDict in three ways: + * + * - The contents of the CDict can be copied into the working context. This + * means that the compression can search both the dictionary and input + * while operating on a single set of internal tables. This makes + * the compression faster per-byte of input. However, the initial copy of + * the CDict's tables incurs a fixed cost at the beginning of the + * compression. For small compressions (< 8 KB), that copy can dominate + * the cost of the compression. + * + * - The CDict's tables can be used in-place. In this model, compression is + * slower per input byte, because the compressor has to search two sets of + * tables. However, this model incurs no start-up cost (as long as the + * working context's tables can be reused). For small inputs, this can be + * faster than copying the CDict's tables. + * + * - The CDict's tables are not used at all, and instead we use the working + * context alone to reload the dictionary and use params based on the source + * size. See ZSTD_compress_insertDictionary() and ZSTD_compress_usingDict(). + * This method is effective when the dictionary sizes are very small relative + * to the input size, and the input size is fairly large to begin with. + * + * Zstd has a simple internal heuristic that selects which strategy to use + * at the beginning of a compression. However, if experimentation shows that + * Zstd is making poor choices, it is possible to override that choice with + * this enum. + */ + ZSTD_dictDefaultAttach = 0, /* Use the default heuristic. */ + ZSTD_dictForceAttach = 1, /* Never copy the dictionary. */ + ZSTD_dictForceCopy = 2, /* Always copy the dictionary. */ + ZSTD_dictForceLoad = 3 /* Always reload the dictionary */ +} ZSTD_dictAttachPref_e; + +typedef enum { + ZSTD_lcm_auto = 0, /*< Automatically determine the compression mode based on the compression level. + * Negative compression levels will be uncompressed, and positive compression + * levels will be compressed. */ + ZSTD_lcm_huffman = 1, /*< Always attempt Huffman compression. Uncompressed literals will still be + * emitted if Huffman compression is not profitable. */ + ZSTD_lcm_uncompressed = 2 /*< Always emit uncompressed literals. */ +} ZSTD_literalCompressionMode_e; + + +/* ************************************* +* Frame size functions +***************************************/ + +/*! ZSTD_findDecompressedSize() : + * `src` should point to the start of a series of ZSTD encoded and/or skippable frames + * `srcSize` must be the _exact_ size of this series + * (i.e. there should be a frame boundary at `src + srcSize`) + * @return : - decompressed size of all data in all successive frames + * - if the decompressed size cannot be determined: ZSTD_CONTENTSIZE_UNKNOWN + * - if an error occurred: ZSTD_CONTENTSIZE_ERROR + * + * note 1 : decompressed size is an optional field, that may not be present, especially in streaming mode. + * When `return==ZSTD_CONTENTSIZE_UNKNOWN`, data to decompress could be any size. + * In which case, it's necessary to use streaming mode to decompress data. + * note 2 : decompressed size is always present when compression is done with ZSTD_compress() + * note 3 : decompressed size can be very large (64-bits value), + * potentially larger than what local system can handle as a single memory segment. + * In which case, it's necessary to use streaming mode to decompress data. + * note 4 : If source is untrusted, decompressed size could be wrong or intentionally modified. + * Always ensure result fits within application's authorized limits. + * Each application can set its own limits. + * note 5 : ZSTD_findDecompressedSize handles multiple frames, and so it must traverse the input to + * read each contained frame header. This is fast as most of the data is skipped, + * however it does mean that all frame data must be present and valid. */ +ZSTDLIB_API unsigned long long ZSTD_findDecompressedSize(const void* src, size_t srcSize); + +/*! ZSTD_decompressBound() : + * `src` should point to the start of a series of ZSTD encoded and/or skippable frames + * `srcSize` must be the _exact_ size of this series + * (i.e. there should be a frame boundary at `src + srcSize`) + * @return : - upper-bound for the decompressed size of all data in all successive frames + * - if an error occurred: ZSTD_CONTENTSIZE_ERROR + * + * note 1 : an error can occur if `src` contains an invalid or incorrectly formatted frame. + * note 2 : the upper-bound is exact when the decompressed size field is available in every ZSTD encoded frame of `src`. + * in this case, `ZSTD_findDecompressedSize` and `ZSTD_decompressBound` return the same value. + * note 3 : when the decompressed size field isn't available, the upper-bound for that frame is calculated by: + * upper-bound = # blocks * min(128 KB, Window_Size) */ -ZSTD_compressionParameters ZSTD_getCParams(int compressionLevel, - unsigned long long estimatedSrcSize, size_t dictSize); +ZSTDLIB_API unsigned long long ZSTD_decompressBound(const void* src, size_t srcSize); -/** - * ZSTD_getParams() - returns ZSTD_parameters for selected level - * @compressionLevel: The compression level from 1 to ZSTD_maxCLevel(). - * @estimatedSrcSize: The estimated source size to compress or 0 if unknown. - * @dictSize: The dictionary size or 0 if a dictionary isn't being used. +/*! ZSTD_frameHeaderSize() : + * srcSize must be >= ZSTD_FRAMEHEADERSIZE_PREFIX. + * @return : size of the Frame Header, + * or an error code (if srcSize is too small) */ +ZSTDLIB_API size_t ZSTD_frameHeaderSize(const void* src, size_t srcSize); + +typedef enum { + ZSTD_sf_noBlockDelimiters = 0, /* Representation of ZSTD_Sequence has no block delimiters, sequences only */ + ZSTD_sf_explicitBlockDelimiters = 1 /* Representation of ZSTD_Sequence contains explicit block delimiters */ +} ZSTD_sequenceFormat_e; + +/*! ZSTD_generateSequences() : + * Generate sequences using ZSTD_compress2, given a source buffer. + * + * Each block will end with a dummy sequence + * with offset == 0, matchLength == 0, and litLength == length of last literals. + * litLength may be == 0, and if so, then the sequence of (of: 0 ml: 0 ll: 0) + * simply acts as a block delimiter. * - * The same as ZSTD_getCParams() except also selects the default frame - * parameters (all zero). + * zc can be used to insert custom compression params. + * This function invokes ZSTD_compress2 * - * Return: The selected ZSTD_parameters. + * The output of this function can be fed into ZSTD_compressSequences() with CCtx + * setting of ZSTD_c_blockDelimiters as ZSTD_sf_explicitBlockDelimiters + * @return : number of sequences generated */ -ZSTD_parameters ZSTD_getParams(int compressionLevel, - unsigned long long estimatedSrcSize, size_t dictSize); -/*-************************************* - * Explicit memory management - **************************************/ +ZSTDLIB_API size_t ZSTD_generateSequences(ZSTD_CCtx* zc, ZSTD_Sequence* outSeqs, + size_t outSeqsSize, const void* src, size_t srcSize); -/** - * ZSTD_CCtxWorkspaceBound() - amount of memory needed to initialize a ZSTD_CCtx - * @cParams: The compression parameters to be used for compression. +/*! ZSTD_mergeBlockDelimiters() : + * Given an array of ZSTD_Sequence, remove all sequences that represent block delimiters/last literals + * by merging them into into the literals of the next sequence. * - * If multiple compression parameters might be used, the caller must call - * ZSTD_CCtxWorkspaceBound() for each set of parameters and use the maximum - * size. + * As such, the final generated result has no explicit representation of block boundaries, + * and the final last literals segment is not represented in the sequences. * - * Return: A lower bound on the size of the workspace that is passed to - * ZSTD_initCCtx(). + * The output of this function can be fed into ZSTD_compressSequences() with CCtx + * setting of ZSTD_c_blockDelimiters as ZSTD_sf_noBlockDelimiters + * @return : number of sequences left after merging + */ +ZSTDLIB_API size_t ZSTD_mergeBlockDelimiters(ZSTD_Sequence* sequences, size_t seqsSize); + +/*! ZSTD_compressSequences() : + * Compress an array of ZSTD_Sequence, generated from the original source buffer, into dst. + * If a dictionary is included, then the cctx should reference the dict. (see: ZSTD_CCtx_refCDict(), ZSTD_CCtx_loadDictionary(), etc.) + * The entire source is compressed into a single frame. + * + * The compression behavior changes based on cctx params. In particular: + * If ZSTD_c_blockDelimiters == ZSTD_sf_noBlockDelimiters, the array of ZSTD_Sequence is expected to contain + * no block delimiters (defined in ZSTD_Sequence). Block boundaries are roughly determined based on + * the block size derived from the cctx, and sequences may be split. This is the default setting. + * + * If ZSTD_c_blockDelimiters == ZSTD_sf_explicitBlockDelimiters, the array of ZSTD_Sequence is expected to contain + * block delimiters (defined in ZSTD_Sequence). Behavior is undefined if no block delimiters are provided. + * + * If ZSTD_c_validateSequences == 0, this function will blindly accept the sequences provided. Invalid sequences cause undefined + * behavior. If ZSTD_c_validateSequences == 1, then if sequence is invalid (see doc/zstd_compression_format.md for + * specifics regarding offset/matchlength requirements) then the function will bail out and return an error. + * + * In addition to the two adjustable experimental params, there are other important cctx params. + * - ZSTD_c_minMatch MUST be set as less than or equal to the smallest match generated by the match finder. It has a minimum value of ZSTD_MINMATCH_MIN. + * - ZSTD_c_compressionLevel accordingly adjusts the strength of the entropy coder, as it would in typical compression. + * - ZSTD_c_windowLog affects offset validation: this function will return an error at higher debug levels if a provided offset + * is larger than what the spec allows for a given window log and dictionary (if present). See: doc/zstd_compression_format.md + * + * Note: Repcodes are, as of now, always re-calculated within this function, so ZSTD_Sequence::rep is unused. + * Note 2: Once we integrate ability to ingest repcodes, the explicit block delims mode must respect those repcodes exactly, + * and cannot emit an RLE block that disagrees with the repcode history + * @return : final compressed size or a ZSTD error. */ -size_t ZSTD_CCtxWorkspaceBound(ZSTD_compressionParameters cParams); +ZSTDLIB_API size_t ZSTD_compressSequences(ZSTD_CCtx* const cctx, void* dst, size_t dstSize, + const ZSTD_Sequence* inSeqs, size_t inSeqsSize, + const void* src, size_t srcSize); -/** - * struct ZSTD_CCtx - the zstd compression context + +/*! ZSTD_writeSkippableFrame() : + * Generates a zstd skippable frame containing data given by src, and writes it to dst buffer. * - * When compressing many times it is recommended to allocate a context just once - * and reuse it for each successive compression operation. - */ -typedef struct ZSTD_CCtx_s ZSTD_CCtx; -/** - * ZSTD_initCCtx() - initialize a zstd compression context - * @workspace: The workspace to emplace the context into. It must outlive - * the returned context. - * @workspaceSize: The size of workspace. Use ZSTD_CCtxWorkspaceBound() to - * determine how large the workspace must be. + * Skippable frames begin with a a 4-byte magic number. There are 16 possible choices of magic number, + * ranging from ZSTD_MAGIC_SKIPPABLE_START to ZSTD_MAGIC_SKIPPABLE_START+15. + * As such, the parameter magicVariant controls the exact skippable frame magic number variant used, so + * the magic number used will be ZSTD_MAGIC_SKIPPABLE_START + magicVariant. * - * Return: A compression context emplaced into workspace. - */ -ZSTD_CCtx *ZSTD_initCCtx(void *workspace, size_t workspaceSize); - -/** - * ZSTD_compressCCtx() - compress src into dst - * @ctx: The context. Must have been initialized with a workspace at - * least as large as ZSTD_CCtxWorkspaceBound(params.cParams). - * @dst: The buffer to compress src into. - * @dstCapacity: The size of the destination buffer. May be any size, but - * ZSTD_compressBound(srcSize) is guaranteed to be large enough. - * @src: The data to compress. - * @srcSize: The size of the data to compress. - * @params: The parameters to use for compression. See ZSTD_getParams(). + * Returns an error if destination buffer is not large enough, if the source size is not representable + * with a 4-byte unsigned int, or if the parameter magicVariant is greater than 15 (and therefore invalid). * - * Return: The compressed size or an error, which can be checked using - * ZSTD_isError(). + * @return : number of bytes written or a ZSTD error. */ -size_t ZSTD_compressCCtx(ZSTD_CCtx *ctx, void *dst, size_t dstCapacity, - const void *src, size_t srcSize, ZSTD_parameters params); - -/** - * ZSTD_DCtxWorkspaceBound() - amount of memory needed to initialize a ZSTD_DCtx - * - * Return: A lower bound on the size of the workspace that is passed to - * ZSTD_initDCtx(). +ZSTDLIB_API size_t ZSTD_writeSkippableFrame(void* dst, size_t dstCapacity, + const void* src, size_t srcSize, unsigned magicVariant); + + +/* ************************************* +* Memory management +***************************************/ + +/*! ZSTD_estimate*() : + * These functions make it possible to estimate memory usage + * of a future {D,C}Ctx, before its creation. + * + * ZSTD_estimateCCtxSize() will provide a memory budget large enough + * for any compression level up to selected one. + * Note : Unlike ZSTD_estimateCStreamSize*(), this estimate + * does not include space for a window buffer. + * Therefore, the estimation is only guaranteed for single-shot compressions, not streaming. + * The estimate will assume the input may be arbitrarily large, + * which is the worst case. + * + * When srcSize can be bound by a known and rather "small" value, + * this fact can be used to provide a tighter estimation + * because the CCtx compression context will need less memory. + * This tighter estimation can be provided by more advanced functions + * ZSTD_estimateCCtxSize_usingCParams(), which can be used in tandem with ZSTD_getCParams(), + * and ZSTD_estimateCCtxSize_usingCCtxParams(), which can be used in tandem with ZSTD_CCtxParams_setParameter(). + * Both can be used to estimate memory using custom compression parameters and arbitrary srcSize limits. + * + * Note 2 : only single-threaded compression is supported. + * ZSTD_estimateCCtxSize_usingCCtxParams() will return an error code if ZSTD_c_nbWorkers is >= 1. + */ +ZSTDLIB_API size_t ZSTD_estimateCCtxSize(int compressionLevel); +ZSTDLIB_API size_t ZSTD_estimateCCtxSize_usingCParams(ZSTD_compressionParameters cParams); +ZSTDLIB_API size_t ZSTD_estimateCCtxSize_usingCCtxParams(const ZSTD_CCtx_params* params); +ZSTDLIB_API size_t ZSTD_estimateDCtxSize(void); + +/*! ZSTD_estimateCStreamSize() : + * ZSTD_estimateCStreamSize() will provide a budget large enough for any compression level up to selected one. + * It will also consider src size to be arbitrarily "large", which is worst case. + * If srcSize is known to always be small, ZSTD_estimateCStreamSize_usingCParams() can provide a tighter estimation. + * ZSTD_estimateCStreamSize_usingCParams() can be used in tandem with ZSTD_getCParams() to create cParams from compressionLevel. + * ZSTD_estimateCStreamSize_usingCCtxParams() can be used in tandem with ZSTD_CCtxParams_setParameter(). Only single-threaded compression is supported. This function will return an error code if ZSTD_c_nbWorkers is >= 1. + * Note : CStream size estimation is only correct for single-threaded compression. + * ZSTD_DStream memory budget depends on window Size. + * This information can be passed manually, using ZSTD_estimateDStreamSize, + * or deducted from a valid frame Header, using ZSTD_estimateDStreamSize_fromFrame(); + * Note : if streaming is init with function ZSTD_init?Stream_usingDict(), + * an internal ?Dict will be created, which additional size is not estimated here. + * In this case, get total size by adding ZSTD_estimate?DictSize */ +ZSTDLIB_API size_t ZSTD_estimateCStreamSize(int compressionLevel); +ZSTDLIB_API size_t ZSTD_estimateCStreamSize_usingCParams(ZSTD_compressionParameters cParams); +ZSTDLIB_API size_t ZSTD_estimateCStreamSize_usingCCtxParams(const ZSTD_CCtx_params* params); +ZSTDLIB_API size_t ZSTD_estimateDStreamSize(size_t windowSize); +ZSTDLIB_API size_t ZSTD_estimateDStreamSize_fromFrame(const void* src, size_t srcSize); + +/*! ZSTD_estimate?DictSize() : + * ZSTD_estimateCDictSize() will bet that src size is relatively "small", and content is copied, like ZSTD_createCDict(). + * ZSTD_estimateCDictSize_advanced() makes it possible to control compression parameters precisely, like ZSTD_createCDict_advanced(). + * Note : dictionaries created by reference (`ZSTD_dlm_byRef`) are logically smaller. + */ +ZSTDLIB_API size_t ZSTD_estimateCDictSize(size_t dictSize, int compressionLevel); +ZSTDLIB_API size_t ZSTD_estimateCDictSize_advanced(size_t dictSize, ZSTD_compressionParameters cParams, ZSTD_dictLoadMethod_e dictLoadMethod); +ZSTDLIB_API size_t ZSTD_estimateDDictSize(size_t dictSize, ZSTD_dictLoadMethod_e dictLoadMethod); + +/*! ZSTD_initStatic*() : + * Initialize an object using a pre-allocated fixed-size buffer. + * workspace: The memory area to emplace the object into. + * Provided pointer *must be 8-bytes aligned*. + * Buffer must outlive object. + * workspaceSize: Use ZSTD_estimate*Size() to determine + * how large workspace must be to support target scenario. + * @return : pointer to object (same address as workspace, just different type), + * or NULL if error (size too small, incorrect alignment, etc.) + * Note : zstd will never resize nor malloc() when using a static buffer. + * If the object requires more memory than available, + * zstd will just error out (typically ZSTD_error_memory_allocation). + * Note 2 : there is no corresponding "free" function. + * Since workspace is allocated externally, it must be freed externally too. + * Note 3 : cParams : use ZSTD_getCParams() to convert a compression level + * into its associated cParams. + * Limitation 1 : currently not compatible with internal dictionary creation, triggered by + * ZSTD_CCtx_loadDictionary(), ZSTD_initCStream_usingDict() or ZSTD_initDStream_usingDict(). + * Limitation 2 : static cctx currently not compatible with multi-threading. + * Limitation 3 : static dctx is incompatible with legacy support. */ -size_t ZSTD_DCtxWorkspaceBound(void); +ZSTDLIB_API ZSTD_CCtx* ZSTD_initStaticCCtx(void* workspace, size_t workspaceSize); +ZSTDLIB_API ZSTD_CStream* ZSTD_initStaticCStream(void* workspace, size_t workspaceSize); /*< same as ZSTD_initStaticCCtx() */ + +ZSTDLIB_API ZSTD_DCtx* ZSTD_initStaticDCtx(void* workspace, size_t workspaceSize); +ZSTDLIB_API ZSTD_DStream* ZSTD_initStaticDStream(void* workspace, size_t workspaceSize); /*< same as ZSTD_initStaticDCtx() */ + +ZSTDLIB_API const ZSTD_CDict* ZSTD_initStaticCDict( + void* workspace, size_t workspaceSize, + const void* dict, size_t dictSize, + ZSTD_dictLoadMethod_e dictLoadMethod, + ZSTD_dictContentType_e dictContentType, + ZSTD_compressionParameters cParams); + +ZSTDLIB_API const ZSTD_DDict* ZSTD_initStaticDDict( + void* workspace, size_t workspaceSize, + const void* dict, size_t dictSize, + ZSTD_dictLoadMethod_e dictLoadMethod, + ZSTD_dictContentType_e dictContentType); + + +/*! Custom memory allocation : + * These prototypes make it possible to pass your own allocation/free functions. + * ZSTD_customMem is provided at creation time, using ZSTD_create*_advanced() variants listed below. + * All allocation/free operations will be completed using these custom variants instead of regular ones. + */ +typedef void* (*ZSTD_allocFunction) (void* opaque, size_t size); +typedef void (*ZSTD_freeFunction) (void* opaque, void* address); +typedef struct { ZSTD_allocFunction customAlloc; ZSTD_freeFunction customFree; void* opaque; } ZSTD_customMem; +static +__attribute__((__unused__)) +ZSTD_customMem const ZSTD_defaultCMem = { NULL, NULL, NULL }; /*< this constant defers to stdlib's functions */ + +ZSTDLIB_API ZSTD_CCtx* ZSTD_createCCtx_advanced(ZSTD_customMem customMem); +ZSTDLIB_API ZSTD_CStream* ZSTD_createCStream_advanced(ZSTD_customMem customMem); +ZSTDLIB_API ZSTD_DCtx* ZSTD_createDCtx_advanced(ZSTD_customMem customMem); +ZSTDLIB_API ZSTD_DStream* ZSTD_createDStream_advanced(ZSTD_customMem customMem); + +ZSTDLIB_API ZSTD_CDict* ZSTD_createCDict_advanced(const void* dict, size_t dictSize, + ZSTD_dictLoadMethod_e dictLoadMethod, + ZSTD_dictContentType_e dictContentType, + ZSTD_compressionParameters cParams, + ZSTD_customMem customMem); + +/* ! Thread pool : + * These prototypes make it possible to share a thread pool among multiple compression contexts. + * This can limit resources for applications with multiple threads where each one uses + * a threaded compression mode (via ZSTD_c_nbWorkers parameter). + * ZSTD_createThreadPool creates a new thread pool with a given number of threads. + * Note that the lifetime of such pool must exist while being used. + * ZSTD_CCtx_refThreadPool assigns a thread pool to a context (use NULL argument value + * to use an internal thread pool). + * ZSTD_freeThreadPool frees a thread pool, accepts NULL pointer. + */ +typedef struct POOL_ctx_s ZSTD_threadPool; +ZSTDLIB_API ZSTD_threadPool* ZSTD_createThreadPool(size_t numThreads); +ZSTDLIB_API void ZSTD_freeThreadPool (ZSTD_threadPool* pool); /* accept NULL pointer */ +ZSTDLIB_API size_t ZSTD_CCtx_refThreadPool(ZSTD_CCtx* cctx, ZSTD_threadPool* pool); -/** - * struct ZSTD_DCtx - the zstd decompression context - * - * When decompressing many times it is recommended to allocate a context just - * once and reuse it for each successive decompression operation. + +/* + * This API is temporary and is expected to change or disappear in the future! */ -typedef struct ZSTD_DCtx_s ZSTD_DCtx; -/** - * ZSTD_initDCtx() - initialize a zstd decompression context - * @workspace: The workspace to emplace the context into. It must outlive - * the returned context. - * @workspaceSize: The size of workspace. Use ZSTD_DCtxWorkspaceBound() to - * determine how large the workspace must be. - * - * Return: A decompression context emplaced into workspace. - */ -ZSTD_DCtx *ZSTD_initDCtx(void *workspace, size_t workspaceSize); - -/** - * ZSTD_decompressDCtx() - decompress zstd compressed src into dst - * @ctx: The decompression context. - * @dst: The buffer to decompress src into. - * @dstCapacity: The size of the destination buffer. Must be at least as large - * as the decompressed size. If the caller cannot upper bound the - * decompressed size, then it's better to use the streaming API. - * @src: The zstd compressed data to decompress. Multiple concatenated - * frames and skippable frames are allowed. - * @srcSize: The exact size of the data to decompress. - * - * Return: The decompressed size or an error, which can be checked using - * ZSTD_isError(). - */ -size_t ZSTD_decompressDCtx(ZSTD_DCtx *ctx, void *dst, size_t dstCapacity, - const void *src, size_t srcSize); - -/*-************************ - * Simple dictionary API - **************************/ - -/** - * ZSTD_compress_usingDict() - compress src into dst using a dictionary - * @ctx: The context. Must have been initialized with a workspace at - * least as large as ZSTD_CCtxWorkspaceBound(params.cParams). - * @dst: The buffer to compress src into. - * @dstCapacity: The size of the destination buffer. May be any size, but - * ZSTD_compressBound(srcSize) is guaranteed to be large enough. - * @src: The data to compress. - * @srcSize: The size of the data to compress. - * @dict: The dictionary to use for compression. - * @dictSize: The size of the dictionary. - * @params: The parameters to use for compression. See ZSTD_getParams(). - * - * Compression using a predefined dictionary. The same dictionary must be used - * during decompression. - * - * Return: The compressed size or an error, which can be checked using - * ZSTD_isError(). - */ -size_t ZSTD_compress_usingDict(ZSTD_CCtx *ctx, void *dst, size_t dstCapacity, - const void *src, size_t srcSize, const void *dict, size_t dictSize, - ZSTD_parameters params); - -/** - * ZSTD_decompress_usingDict() - decompress src into dst using a dictionary - * @ctx: The decompression context. - * @dst: The buffer to decompress src into. - * @dstCapacity: The size of the destination buffer. Must be at least as large - * as the decompressed size. If the caller cannot upper bound the - * decompressed size, then it's better to use the streaming API. - * @src: The zstd compressed data to decompress. Multiple concatenated - * frames and skippable frames are allowed. - * @srcSize: The exact size of the data to decompress. - * @dict: The dictionary to use for decompression. The same dictionary - * must've been used to compress the data. - * @dictSize: The size of the dictionary. - * - * Return: The decompressed size or an error, which can be checked using - * ZSTD_isError(). - */ -size_t ZSTD_decompress_usingDict(ZSTD_DCtx *ctx, void *dst, size_t dstCapacity, - const void *src, size_t srcSize, const void *dict, size_t dictSize); - -/*-************************** - * Fast dictionary API - ***************************/ - -/** - * ZSTD_CDictWorkspaceBound() - memory needed to initialize a ZSTD_CDict - * @cParams: The compression parameters to be used for compression. - * - * Return: A lower bound on the size of the workspace that is passed to - * ZSTD_initCDict(). - */ -size_t ZSTD_CDictWorkspaceBound(ZSTD_compressionParameters cParams); - -/** - * struct ZSTD_CDict - a digested dictionary to be used for compression +ZSTDLIB_API ZSTD_CDict* ZSTD_createCDict_advanced2( + const void* dict, size_t dictSize, + ZSTD_dictLoadMethod_e dictLoadMethod, + ZSTD_dictContentType_e dictContentType, + const ZSTD_CCtx_params* cctxParams, + ZSTD_customMem customMem); + +ZSTDLIB_API ZSTD_DDict* ZSTD_createDDict_advanced( + const void* dict, size_t dictSize, + ZSTD_dictLoadMethod_e dictLoadMethod, + ZSTD_dictContentType_e dictContentType, + ZSTD_customMem customMem); + + +/* ************************************* +* Advanced compression functions +***************************************/ + +/*! ZSTD_createCDict_byReference() : + * Create a digested dictionary for compression + * Dictionary content is just referenced, not duplicated. + * As a consequence, `dictBuffer` **must** outlive CDict, + * and its content must remain unmodified throughout the lifetime of CDict. + * note: equivalent to ZSTD_createCDict_advanced(), with dictLoadMethod==ZSTD_dlm_byRef */ +ZSTDLIB_API ZSTD_CDict* ZSTD_createCDict_byReference(const void* dictBuffer, size_t dictSize, int compressionLevel); + +/*! ZSTD_getDictID_fromCDict() : + * Provides the dictID of the dictionary loaded into `cdict`. + * If @return == 0, the dictionary is not conformant to Zstandard specification, or empty. + * Non-conformant dictionaries can still be loaded, but as content-only dictionaries. */ +ZSTDLIB_API unsigned ZSTD_getDictID_fromCDict(const ZSTD_CDict* cdict); + +/*! ZSTD_getCParams() : + * @return ZSTD_compressionParameters structure for a selected compression level and estimated srcSize. + * `estimatedSrcSize` value is optional, select 0 if not known */ +ZSTDLIB_API ZSTD_compressionParameters ZSTD_getCParams(int compressionLevel, unsigned long long estimatedSrcSize, size_t dictSize); + +/*! ZSTD_getParams() : + * same as ZSTD_getCParams(), but @return a full `ZSTD_parameters` object instead of sub-component `ZSTD_compressionParameters`. + * All fields of `ZSTD_frameParameters` are set to default : contentSize=1, checksum=0, noDictID=0 */ +ZSTDLIB_API ZSTD_parameters ZSTD_getParams(int compressionLevel, unsigned long long estimatedSrcSize, size_t dictSize); + +/*! ZSTD_checkCParams() : + * Ensure param values remain within authorized range. + * @return 0 on success, or an error code (can be checked with ZSTD_isError()) */ +ZSTDLIB_API size_t ZSTD_checkCParams(ZSTD_compressionParameters params); + +/*! ZSTD_adjustCParams() : + * optimize params for a given `srcSize` and `dictSize`. + * `srcSize` can be unknown, in which case use ZSTD_CONTENTSIZE_UNKNOWN. + * `dictSize` must be `0` when there is no dictionary. + * cPar can be invalid : all parameters will be clamped within valid range in the @return struct. + * This function never fails (wide contract) */ +ZSTDLIB_API ZSTD_compressionParameters ZSTD_adjustCParams(ZSTD_compressionParameters cPar, unsigned long long srcSize, size_t dictSize); + +/*! ZSTD_compress_advanced() : + * Note : this function is now DEPRECATED. + * It can be replaced by ZSTD_compress2(), in combination with ZSTD_CCtx_setParameter() and other parameter setters. + * This prototype will be marked as deprecated and generate compilation warning on reaching v1.5.x */ +ZSTDLIB_API size_t ZSTD_compress_advanced(ZSTD_CCtx* cctx, + void* dst, size_t dstCapacity, + const void* src, size_t srcSize, + const void* dict,size_t dictSize, + ZSTD_parameters params); + +/*! ZSTD_compress_usingCDict_advanced() : + * Note : this function is now REDUNDANT. + * It can be replaced by ZSTD_compress2(), in combination with ZSTD_CCtx_loadDictionary() and other parameter setters. + * This prototype will be marked as deprecated and generate compilation warning in some future version */ +ZSTDLIB_API size_t ZSTD_compress_usingCDict_advanced(ZSTD_CCtx* cctx, + void* dst, size_t dstCapacity, + const void* src, size_t srcSize, + const ZSTD_CDict* cdict, + ZSTD_frameParameters fParams); + + +/*! ZSTD_CCtx_loadDictionary_byReference() : + * Same as ZSTD_CCtx_loadDictionary(), but dictionary content is referenced, instead of being copied into CCtx. + * It saves some memory, but also requires that `dict` outlives its usage within `cctx` */ +ZSTDLIB_API size_t ZSTD_CCtx_loadDictionary_byReference(ZSTD_CCtx* cctx, const void* dict, size_t dictSize); + +/*! ZSTD_CCtx_loadDictionary_advanced() : + * Same as ZSTD_CCtx_loadDictionary(), but gives finer control over + * how to load the dictionary (by copy ? by reference ?) + * and how to interpret it (automatic ? force raw mode ? full mode only ?) */ +ZSTDLIB_API size_t ZSTD_CCtx_loadDictionary_advanced(ZSTD_CCtx* cctx, const void* dict, size_t dictSize, ZSTD_dictLoadMethod_e dictLoadMethod, ZSTD_dictContentType_e dictContentType); + +/*! ZSTD_CCtx_refPrefix_advanced() : + * Same as ZSTD_CCtx_refPrefix(), but gives finer control over + * how to interpret prefix content (automatic ? force raw mode (default) ? full mode only ?) */ +ZSTDLIB_API size_t ZSTD_CCtx_refPrefix_advanced(ZSTD_CCtx* cctx, const void* prefix, size_t prefixSize, ZSTD_dictContentType_e dictContentType); + +/* === experimental parameters === */ +/* these parameters can be used with ZSTD_setParameter() + * they are not guaranteed to remain supported in the future */ + + /* Enables rsyncable mode, + * which makes compressed files more rsync friendly + * by adding periodic synchronization points to the compressed data. + * The target average block size is ZSTD_c_jobSize / 2. + * It's possible to modify the job size to increase or decrease + * the granularity of the synchronization point. + * Once the jobSize is smaller than the window size, + * it will result in compression ratio degradation. + * NOTE 1: rsyncable mode only works when multithreading is enabled. + * NOTE 2: rsyncable performs poorly in combination with long range mode, + * since it will decrease the effectiveness of synchronization points, + * though mileage may vary. + * NOTE 3: Rsyncable mode limits maximum compression speed to ~400 MB/s. + * If the selected compression level is already running significantly slower, + * the overall speed won't be significantly impacted. + */ + #define ZSTD_c_rsyncable ZSTD_c_experimentalParam1 + +/* Select a compression format. + * The value must be of type ZSTD_format_e. + * See ZSTD_format_e enum definition for details */ +#define ZSTD_c_format ZSTD_c_experimentalParam2 + +/* Force back-reference distances to remain < windowSize, + * even when referencing into Dictionary content (default:0) */ +#define ZSTD_c_forceMaxWindow ZSTD_c_experimentalParam3 + +/* Controls whether the contents of a CDict + * are used in place, or copied into the working context. + * Accepts values from the ZSTD_dictAttachPref_e enum. + * See the comments on that enum for an explanation of the feature. */ +#define ZSTD_c_forceAttachDict ZSTD_c_experimentalParam4 + +/* Controls how the literals are compressed (default is auto). + * The value must be of type ZSTD_literalCompressionMode_e. + * See ZSTD_literalCompressionMode_t enum definition for details. */ -typedef struct ZSTD_CDict_s ZSTD_CDict; +#define ZSTD_c_literalCompressionMode ZSTD_c_experimentalParam5 + +/* Tries to fit compressed block size to be around targetCBlockSize. + * No target when targetCBlockSize == 0. + * There is no guarantee on compressed block size (default:0) */ +#define ZSTD_c_targetCBlockSize ZSTD_c_experimentalParam6 + +/* User's best guess of source size. + * Hint is not valid when srcSizeHint == 0. + * There is no guarantee that hint is close to actual source size, + * but compression ratio may regress significantly if guess considerably underestimates */ +#define ZSTD_c_srcSizeHint ZSTD_c_experimentalParam7 + +/* Controls whether the new and experimental "dedicated dictionary search + * structure" can be used. This feature is still rough around the edges, be + * prepared for surprising behavior! + * + * How to use it: + * + * When using a CDict, whether to use this feature or not is controlled at + * CDict creation, and it must be set in a CCtxParams set passed into that + * construction (via ZSTD_createCDict_advanced2()). A compression will then + * use the feature or not based on how the CDict was constructed; the value of + * this param, set in the CCtx, will have no effect. + * + * However, when a dictionary buffer is passed into a CCtx, such as via + * ZSTD_CCtx_loadDictionary(), this param can be set on the CCtx to control + * whether the CDict that is created internally can use the feature or not. + * + * What it does: + * + * Normally, the internal data structures of the CDict are analogous to what + * would be stored in a CCtx after compressing the contents of a dictionary. + * To an approximation, a compression using a dictionary can then use those + * data structures to simply continue what is effectively a streaming + * compression where the simulated compression of the dictionary left off. + * Which is to say, the search structures in the CDict are normally the same + * format as in the CCtx. + * + * It is possible to do better, since the CDict is not like a CCtx: the search + * structures are written once during CDict creation, and then are only read + * after that, while the search structures in the CCtx are both read and + * written as the compression goes along. This means we can choose a search + * structure for the dictionary that is read-optimized. + * + * This feature enables the use of that different structure. + * + * Note that some of the members of the ZSTD_compressionParameters struct have + * different semantics and constraints in the dedicated search structure. It is + * highly recommended that you simply set a compression level in the CCtxParams + * you pass into the CDict creation call, and avoid messing with the cParams + * directly. + * + * Effects: + * + * This will only have any effect when the selected ZSTD_strategy + * implementation supports this feature. Currently, that's limited to + * ZSTD_greedy, ZSTD_lazy, and ZSTD_lazy2. + * + * Note that this means that the CDict tables can no longer be copied into the + * CCtx, so the dict attachment mode ZSTD_dictForceCopy will no longer be + * useable. The dictionary can only be attached or reloaded. + * + * In general, you should expect compression to be faster--sometimes very much + * so--and CDict creation to be slightly slower. Eventually, we will probably + * make this mode the default. + */ +#define ZSTD_c_enableDedicatedDictSearch ZSTD_c_experimentalParam8 + +/* ZSTD_c_stableInBuffer + * Experimental parameter. + * Default is 0 == disabled. Set to 1 to enable. + * + * Tells the compressor that the ZSTD_inBuffer will ALWAYS be the same + * between calls, except for the modifications that zstd makes to pos (the + * caller must not modify pos). This is checked by the compressor, and + * compression will fail if it ever changes. This means the only flush + * mode that makes sense is ZSTD_e_end, so zstd will error if ZSTD_e_end + * is not used. The data in the ZSTD_inBuffer in the range [src, src + pos) + * MUST not be modified during compression or you will get data corruption. + * + * When this flag is enabled zstd won't allocate an input window buffer, + * because the user guarantees it can reference the ZSTD_inBuffer until + * the frame is complete. But, it will still allocate an output buffer + * large enough to fit a block (see ZSTD_c_stableOutBuffer). This will also + * avoid the memcpy() from the input buffer to the input window buffer. + * + * NOTE: ZSTD_compressStream2() will error if ZSTD_e_end is not used. + * That means this flag cannot be used with ZSTD_compressStream(). + * + * NOTE: So long as the ZSTD_inBuffer always points to valid memory, using + * this flag is ALWAYS memory safe, and will never access out-of-bounds + * memory. However, compression WILL fail if you violate the preconditions. + * + * WARNING: The data in the ZSTD_inBuffer in the range [dst, dst + pos) MUST + * not be modified during compression or you will get data corruption. This + * is because zstd needs to reference data in the ZSTD_inBuffer to find + * matches. Normally zstd maintains its own window buffer for this purpose, + * but passing this flag tells zstd to use the user provided buffer. + */ +#define ZSTD_c_stableInBuffer ZSTD_c_experimentalParam9 -/** - * ZSTD_initCDict() - initialize a digested dictionary for compression - * @dictBuffer: The dictionary to digest. The buffer is referenced by the - * ZSTD_CDict so it must outlive the returned ZSTD_CDict. - * @dictSize: The size of the dictionary. - * @params: The parameters to use for compression. See ZSTD_getParams(). - * @workspace: The workspace. It must outlive the returned ZSTD_CDict. - * @workspaceSize: The workspace size. Must be at least - * ZSTD_CDictWorkspaceBound(params.cParams). +/* ZSTD_c_stableOutBuffer + * Experimental parameter. + * Default is 0 == disabled. Set to 1 to enable. * - * When compressing multiple messages / blocks with the same dictionary it is - * recommended to load it just once. The ZSTD_CDict merely references the - * dictBuffer, so it must outlive the returned ZSTD_CDict. + * Tells he compressor that the ZSTD_outBuffer will not be resized between + * calls. Specifically: (out.size - out.pos) will never grow. This gives the + * compressor the freedom to say: If the compressed data doesn't fit in the + * output buffer then return ZSTD_error_dstSizeTooSmall. This allows us to + * always decompress directly into the output buffer, instead of decompressing + * into an internal buffer and copying to the output buffer. * - * Return: The digested dictionary emplaced into workspace. + * When this flag is enabled zstd won't allocate an output buffer, because + * it can write directly to the ZSTD_outBuffer. It will still allocate the + * input window buffer (see ZSTD_c_stableInBuffer). + * + * Zstd will check that (out.size - out.pos) never grows and return an error + * if it does. While not strictly necessary, this should prevent surprises. */ -ZSTD_CDict *ZSTD_initCDict(const void *dictBuffer, size_t dictSize, - ZSTD_parameters params, void *workspace, size_t workspaceSize); +#define ZSTD_c_stableOutBuffer ZSTD_c_experimentalParam10 -/** - * ZSTD_compress_usingCDict() - compress src into dst using a ZSTD_CDict - * @ctx: The context. Must have been initialized with a workspace at - * least as large as ZSTD_CCtxWorkspaceBound(cParams) where - * cParams are the compression parameters used to initialize the - * cdict. - * @dst: The buffer to compress src into. - * @dstCapacity: The size of the destination buffer. May be any size, but - * ZSTD_compressBound(srcSize) is guaranteed to be large enough. - * @src: The data to compress. - * @srcSize: The size of the data to compress. - * @cdict: The digested dictionary to use for compression. - * @params: The parameters to use for compression. See ZSTD_getParams(). +/* ZSTD_c_blockDelimiters + * Default is 0 == ZSTD_sf_noBlockDelimiters. * - * Compression using a digested dictionary. The same dictionary must be used - * during decompression. + * For use with sequence compression API: ZSTD_compressSequences(). * - * Return: The compressed size or an error, which can be checked using - * ZSTD_isError(). + * Designates whether or not the given array of ZSTD_Sequence contains block delimiters + * and last literals, which are defined as sequences with offset == 0 and matchLength == 0. + * See the definition of ZSTD_Sequence for more specifics. */ -size_t ZSTD_compress_usingCDict(ZSTD_CCtx *cctx, void *dst, size_t dstCapacity, - const void *src, size_t srcSize, const ZSTD_CDict *cdict); - +#define ZSTD_c_blockDelimiters ZSTD_c_experimentalParam11 -/** - * ZSTD_DDictWorkspaceBound() - memory needed to initialize a ZSTD_DDict +/* ZSTD_c_validateSequences + * Default is 0 == disabled. Set to 1 to enable sequence validation. + * + * For use with sequence compression API: ZSTD_compressSequences(). + * Designates whether or not we validate sequences provided to ZSTD_compressSequences() + * during function execution. + * + * Without validation, providing a sequence that does not conform to the zstd spec will cause + * undefined behavior, and may produce a corrupted block. + * + * With validation enabled, a if sequence is invalid (see doc/zstd_compression_format.md for + * specifics regarding offset/matchlength requirements) then the function will bail out and + * return an error. * - * Return: A lower bound on the size of the workspace that is passed to - * ZSTD_initDDict(). */ -size_t ZSTD_DDictWorkspaceBound(void); +#define ZSTD_c_validateSequences ZSTD_c_experimentalParam12 -/** - * struct ZSTD_DDict - a digested dictionary to be used for decompression +/*! ZSTD_CCtx_getParameter() : + * Get the requested compression parameter value, selected by enum ZSTD_cParameter, + * and store it into int* value. + * @return : 0, or an error code (which can be tested with ZSTD_isError()). */ -typedef struct ZSTD_DDict_s ZSTD_DDict; - -/** - * ZSTD_initDDict() - initialize a digested dictionary for decompression - * @dictBuffer: The dictionary to digest. The buffer is referenced by the - * ZSTD_DDict so it must outlive the returned ZSTD_DDict. - * @dictSize: The size of the dictionary. - * @workspace: The workspace. It must outlive the returned ZSTD_DDict. - * @workspaceSize: The workspace size. Must be at least - * ZSTD_DDictWorkspaceBound(). - * - * When decompressing multiple messages / blocks with the same dictionary it is - * recommended to load it just once. The ZSTD_DDict merely references the - * dictBuffer, so it must outlive the returned ZSTD_DDict. - * - * Return: The digested dictionary emplaced into workspace. - */ -ZSTD_DDict *ZSTD_initDDict(const void *dictBuffer, size_t dictSize, - void *workspace, size_t workspaceSize); - -/** - * ZSTD_decompress_usingDDict() - decompress src into dst using a ZSTD_DDict - * @ctx: The decompression context. - * @dst: The buffer to decompress src into. - * @dstCapacity: The size of the destination buffer. Must be at least as large - * as the decompressed size. If the caller cannot upper bound the - * decompressed size, then it's better to use the streaming API. - * @src: The zstd compressed data to decompress. Multiple concatenated - * frames and skippable frames are allowed. - * @srcSize: The exact size of the data to decompress. - * @ddict: The digested dictionary to use for decompression. The same - * dictionary must've been used to compress the data. - * - * Return: The decompressed size or an error, which can be checked using - * ZSTD_isError(). - */ -size_t ZSTD_decompress_usingDDict(ZSTD_DCtx *dctx, void *dst, - size_t dstCapacity, const void *src, size_t srcSize, - const ZSTD_DDict *ddict); - - -/*-************************** - * Streaming - ***************************/ - -/** - * struct ZSTD_inBuffer - input buffer for streaming - * @src: Start of the input buffer. - * @size: Size of the input buffer. - * @pos: Position where reading stopped. Will be updated. - * Necessarily 0 <= pos <= size. +ZSTDLIB_API size_t ZSTD_CCtx_getParameter(const ZSTD_CCtx* cctx, ZSTD_cParameter param, int* value); + + +/*! ZSTD_CCtx_params : + * Quick howto : + * - ZSTD_createCCtxParams() : Create a ZSTD_CCtx_params structure + * - ZSTD_CCtxParams_setParameter() : Push parameters one by one into + * an existing ZSTD_CCtx_params structure. + * This is similar to + * ZSTD_CCtx_setParameter(). + * - ZSTD_CCtx_setParametersUsingCCtxParams() : Apply parameters to + * an existing CCtx. + * These parameters will be applied to + * all subsequent frames. + * - ZSTD_compressStream2() : Do compression using the CCtx. + * - ZSTD_freeCCtxParams() : Free the memory, accept NULL pointer. + * + * This can be used with ZSTD_estimateCCtxSize_advanced_usingCCtxParams() + * for static allocation of CCtx for single-threaded compression. */ -typedef struct ZSTD_inBuffer_s { - const void *src; - size_t size; - size_t pos; -} ZSTD_inBuffer; +ZSTDLIB_API ZSTD_CCtx_params* ZSTD_createCCtxParams(void); +ZSTDLIB_API size_t ZSTD_freeCCtxParams(ZSTD_CCtx_params* params); /* accept NULL pointer */ -/** - * struct ZSTD_outBuffer - output buffer for streaming - * @dst: Start of the output buffer. - * @size: Size of the output buffer. - * @pos: Position where writing stopped. Will be updated. - * Necessarily 0 <= pos <= size. +/*! ZSTD_CCtxParams_reset() : + * Reset params to default values. */ -typedef struct ZSTD_outBuffer_s { - void *dst; - size_t size; - size_t pos; -} ZSTD_outBuffer; - - +ZSTDLIB_API size_t ZSTD_CCtxParams_reset(ZSTD_CCtx_params* params); -/*-***************************************************************************** - * Streaming compression - HowTo - * - * A ZSTD_CStream object is required to track streaming operation. - * Use ZSTD_initCStream() to initialize a ZSTD_CStream object. - * ZSTD_CStream objects can be reused multiple times on consecutive compression - * operations. It is recommended to re-use ZSTD_CStream in situations where many - * streaming operations will be achieved consecutively. Use one separate - * ZSTD_CStream per thread for parallel execution. - * - * Use ZSTD_compressStream() repetitively to consume input stream. - * The function will automatically update both `pos` fields. - * Note that it may not consume the entire input, in which case `pos < size`, - * and it's up to the caller to present again remaining data. - * It returns a hint for the preferred number of bytes to use as an input for - * the next function call. - * - * At any moment, it's possible to flush whatever data remains within internal - * buffer, using ZSTD_flushStream(). `output->pos` will be updated. There might - * still be some content left within the internal buffer if `output->size` is - * too small. It returns the number of bytes left in the internal buffer and - * must be called until it returns 0. - * - * ZSTD_endStream() instructs to finish a frame. It will perform a flush and - * write frame epilogue. The epilogue is required for decoders to consider a - * frame completed. Similar to ZSTD_flushStream(), it may not be able to flush - * the full content if `output->size` is too small. In which case, call again - * ZSTD_endStream() to complete the flush. It returns the number of bytes left - * in the internal buffer and must be called until it returns 0. - ******************************************************************************/ +/*! ZSTD_CCtxParams_init() : + * Initializes the compression parameters of cctxParams according to + * compression level. All other parameters are reset to their default values. + */ +ZSTDLIB_API size_t ZSTD_CCtxParams_init(ZSTD_CCtx_params* cctxParams, int compressionLevel); -/** - * ZSTD_CStreamWorkspaceBound() - memory needed to initialize a ZSTD_CStream - * @cParams: The compression parameters to be used for compression. - * - * Return: A lower bound on the size of the workspace that is passed to - * ZSTD_initCStream() and ZSTD_initCStream_usingCDict(). +/*! ZSTD_CCtxParams_init_advanced() : + * Initializes the compression and frame parameters of cctxParams according to + * params. All other parameters are reset to their default values. */ -size_t ZSTD_CStreamWorkspaceBound(ZSTD_compressionParameters cParams); +ZSTDLIB_API size_t ZSTD_CCtxParams_init_advanced(ZSTD_CCtx_params* cctxParams, ZSTD_parameters params); + +/*! ZSTD_CCtxParams_setParameter() : + * Similar to ZSTD_CCtx_setParameter. + * Set one compression parameter, selected by enum ZSTD_cParameter. + * Parameters must be applied to a ZSTD_CCtx using + * ZSTD_CCtx_setParametersUsingCCtxParams(). + * @result : a code representing success or failure (which can be tested with + * ZSTD_isError()). + */ +ZSTDLIB_API size_t ZSTD_CCtxParams_setParameter(ZSTD_CCtx_params* params, ZSTD_cParameter param, int value); -/** - * struct ZSTD_CStream - the zstd streaming compression context +/*! ZSTD_CCtxParams_getParameter() : + * Similar to ZSTD_CCtx_getParameter. + * Get the requested value of one compression parameter, selected by enum ZSTD_cParameter. + * @result : 0, or an error code (which can be tested with ZSTD_isError()). + */ +ZSTDLIB_API size_t ZSTD_CCtxParams_getParameter(const ZSTD_CCtx_params* params, ZSTD_cParameter param, int* value); + +/*! ZSTD_CCtx_setParametersUsingCCtxParams() : + * Apply a set of ZSTD_CCtx_params to the compression context. + * This can be done even after compression is started, + * if nbWorkers==0, this will have no impact until a new compression is started. + * if nbWorkers>=1, new parameters will be picked up at next job, + * with a few restrictions (windowLog, pledgedSrcSize, nbWorkers, jobSize, and overlapLog are not updated). + */ +ZSTDLIB_API size_t ZSTD_CCtx_setParametersUsingCCtxParams( + ZSTD_CCtx* cctx, const ZSTD_CCtx_params* params); + +/*! ZSTD_compressStream2_simpleArgs() : + * Same as ZSTD_compressStream2(), + * but using only integral types as arguments. + * This variant might be helpful for binders from dynamic languages + * which have troubles handling structures containing memory pointers. */ -typedef struct ZSTD_CStream_s ZSTD_CStream; +ZSTDLIB_API size_t ZSTD_compressStream2_simpleArgs ( + ZSTD_CCtx* cctx, + void* dst, size_t dstCapacity, size_t* dstPos, + const void* src, size_t srcSize, size_t* srcPos, + ZSTD_EndDirective endOp); + + +/* ************************************* +* Advanced decompression functions +***************************************/ + +/*! ZSTD_isFrame() : + * Tells if the content of `buffer` starts with a valid Frame Identifier. + * Note : Frame Identifier is 4 bytes. If `size < 4`, @return will always be 0. + * Note 2 : Legacy Frame Identifiers are considered valid only if Legacy Support is enabled. + * Note 3 : Skippable Frame Identifiers are considered valid. */ +ZSTDLIB_API unsigned ZSTD_isFrame(const void* buffer, size_t size); + +/*! ZSTD_createDDict_byReference() : + * Create a digested dictionary, ready to start decompression operation without startup delay. + * Dictionary content is referenced, and therefore stays in dictBuffer. + * It is important that dictBuffer outlives DDict, + * it must remain read accessible throughout the lifetime of DDict */ +ZSTDLIB_API ZSTD_DDict* ZSTD_createDDict_byReference(const void* dictBuffer, size_t dictSize); + +/*! ZSTD_DCtx_loadDictionary_byReference() : + * Same as ZSTD_DCtx_loadDictionary(), + * but references `dict` content instead of copying it into `dctx`. + * This saves memory if `dict` remains around., + * However, it's imperative that `dict` remains accessible (and unmodified) while being used, so it must outlive decompression. */ +ZSTDLIB_API size_t ZSTD_DCtx_loadDictionary_byReference(ZSTD_DCtx* dctx, const void* dict, size_t dictSize); + +/*! ZSTD_DCtx_loadDictionary_advanced() : + * Same as ZSTD_DCtx_loadDictionary(), + * but gives direct control over + * how to load the dictionary (by copy ? by reference ?) + * and how to interpret it (automatic ? force raw mode ? full mode only ?). */ +ZSTDLIB_API size_t ZSTD_DCtx_loadDictionary_advanced(ZSTD_DCtx* dctx, const void* dict, size_t dictSize, ZSTD_dictLoadMethod_e dictLoadMethod, ZSTD_dictContentType_e dictContentType); + +/*! ZSTD_DCtx_refPrefix_advanced() : + * Same as ZSTD_DCtx_refPrefix(), but gives finer control over + * how to interpret prefix content (automatic ? force raw mode (default) ? full mode only ?) */ +ZSTDLIB_API size_t ZSTD_DCtx_refPrefix_advanced(ZSTD_DCtx* dctx, const void* prefix, size_t prefixSize, ZSTD_dictContentType_e dictContentType); + +/*! ZSTD_DCtx_setMaxWindowSize() : + * Refuses allocating internal buffers for frames requiring a window size larger than provided limit. + * This protects a decoder context from reserving too much memory for itself (potential attack scenario). + * This parameter is only useful in streaming mode, since no internal buffer is allocated in single-pass mode. + * By default, a decompression context accepts all window sizes <= (1 << ZSTD_WINDOWLOG_LIMIT_DEFAULT) + * @return : 0, or an error code (which can be tested using ZSTD_isError()). + */ +ZSTDLIB_API size_t ZSTD_DCtx_setMaxWindowSize(ZSTD_DCtx* dctx, size_t maxWindowSize); -/*===== ZSTD_CStream management functions =====*/ -/** - * ZSTD_initCStream() - initialize a zstd streaming compression context - * @params: The zstd compression parameters. - * @pledgedSrcSize: If params.fParams.contentSizeFlag == 1 then the caller must - * pass the source size (zero means empty source). Otherwise, - * the caller may optionally pass the source size, or zero if - * unknown. - * @workspace: The workspace to emplace the context into. It must outlive - * the returned context. - * @workspaceSize: The size of workspace. - * Use ZSTD_CStreamWorkspaceBound(params.cParams) to determine - * how large the workspace must be. - * - * Return: The zstd streaming compression context. - */ -ZSTD_CStream *ZSTD_initCStream(ZSTD_parameters params, - unsigned long long pledgedSrcSize, void *workspace, - size_t workspaceSize); - -/** - * ZSTD_initCStream_usingCDict() - initialize a streaming compression context - * @cdict: The digested dictionary to use for compression. - * @pledgedSrcSize: Optionally the source size, or zero if unknown. - * @workspace: The workspace to emplace the context into. It must outlive - * the returned context. - * @workspaceSize: The size of workspace. Call ZSTD_CStreamWorkspaceBound() - * with the cParams used to initialize the cdict to determine - * how large the workspace must be. - * - * Return: The zstd streaming compression context. - */ -ZSTD_CStream *ZSTD_initCStream_usingCDict(const ZSTD_CDict *cdict, - unsigned long long pledgedSrcSize, void *workspace, - size_t workspaceSize); +/*! ZSTD_DCtx_getParameter() : + * Get the requested decompression parameter value, selected by enum ZSTD_dParameter, + * and store it into int* value. + * @return : 0, or an error code (which can be tested with ZSTD_isError()). + */ +ZSTDLIB_API size_t ZSTD_DCtx_getParameter(ZSTD_DCtx* dctx, ZSTD_dParameter param, int* value); -/*===== Streaming compression functions =====*/ -/** - * ZSTD_resetCStream() - reset the context using parameters from creation - * @zcs: The zstd streaming compression context to reset. - * @pledgedSrcSize: Optionally the source size, or zero if unknown. - * - * Resets the context using the parameters from creation. Skips dictionary - * loading, since it can be reused. If `pledgedSrcSize` is non-zero the frame - * content size is always written into the frame header. - * - * Return: Zero or an error, which can be checked using ZSTD_isError(). +/* ZSTD_d_format + * experimental parameter, + * allowing selection between ZSTD_format_e input compression formats */ -size_t ZSTD_resetCStream(ZSTD_CStream *zcs, unsigned long long pledgedSrcSize); -/** - * ZSTD_compressStream() - streaming compress some of input into output - * @zcs: The zstd streaming compression context. - * @output: Destination buffer. `output->pos` is updated to indicate how much - * compressed data was written. - * @input: Source buffer. `input->pos` is updated to indicate how much data was - * read. Note that it may not consume the entire input, in which case - * `input->pos < input->size`, and it's up to the caller to present - * remaining data again. - * - * The `input` and `output` buffers may be any size. Guaranteed to make some - * forward progress if `input` and `output` are not empty. - * - * Return: A hint for the number of bytes to use as the input for the next - * function call or an error, which can be checked using - * ZSTD_isError(). +#define ZSTD_d_format ZSTD_d_experimentalParam1 +/* ZSTD_d_stableOutBuffer + * Experimental parameter. + * Default is 0 == disabled. Set to 1 to enable. + * + * Tells the decompressor that the ZSTD_outBuffer will ALWAYS be the same + * between calls, except for the modifications that zstd makes to pos (the + * caller must not modify pos). This is checked by the decompressor, and + * decompression will fail if it ever changes. Therefore the ZSTD_outBuffer + * MUST be large enough to fit the entire decompressed frame. This will be + * checked when the frame content size is known. The data in the ZSTD_outBuffer + * in the range [dst, dst + pos) MUST not be modified during decompression + * or you will get data corruption. + * + * When this flags is enabled zstd won't allocate an output buffer, because + * it can write directly to the ZSTD_outBuffer, but it will still allocate + * an input buffer large enough to fit any compressed block. This will also + * avoid the memcpy() from the internal output buffer to the ZSTD_outBuffer. + * If you need to avoid the input buffer allocation use the buffer-less + * streaming API. + * + * NOTE: So long as the ZSTD_outBuffer always points to valid memory, using + * this flag is ALWAYS memory safe, and will never access out-of-bounds + * memory. However, decompression WILL fail if you violate the preconditions. + * + * WARNING: The data in the ZSTD_outBuffer in the range [dst, dst + pos) MUST + * not be modified during decompression or you will get data corruption. This + * is because zstd needs to reference data in the ZSTD_outBuffer to regenerate + * matches. Normally zstd maintains its own buffer for this purpose, but passing + * this flag tells zstd to use the user provided buffer. */ -size_t ZSTD_compressStream(ZSTD_CStream *zcs, ZSTD_outBuffer *output, - ZSTD_inBuffer *input); -/** - * ZSTD_flushStream() - flush internal buffers into output - * @zcs: The zstd streaming compression context. - * @output: Destination buffer. `output->pos` is updated to indicate how much - * compressed data was written. - * - * ZSTD_flushStream() must be called until it returns 0, meaning all the data - * has been flushed. Since ZSTD_flushStream() causes a block to be ended, - * calling it too often will degrade the compression ratio. +#define ZSTD_d_stableOutBuffer ZSTD_d_experimentalParam2 + +/* ZSTD_d_forceIgnoreChecksum + * Experimental parameter. + * Default is 0 == disabled. Set to 1 to enable * - * Return: The number of bytes still present within internal buffers or an - * error, which can be checked using ZSTD_isError(). + * Tells the decompressor to skip checksum validation during decompression, regardless + * of whether checksumming was specified during compression. This offers some + * slight performance benefits, and may be useful for debugging. + * Param has values of type ZSTD_forceIgnoreChecksum_e */ -size_t ZSTD_flushStream(ZSTD_CStream *zcs, ZSTD_outBuffer *output); -/** - * ZSTD_endStream() - flush internal buffers into output and end the frame - * @zcs: The zstd streaming compression context. - * @output: Destination buffer. `output->pos` is updated to indicate how much - * compressed data was written. +#define ZSTD_d_forceIgnoreChecksum ZSTD_d_experimentalParam3 + +/* ZSTD_d_refMultipleDDicts + * Experimental parameter. + * Default is 0 == disabled. Set to 1 to enable * - * ZSTD_endStream() must be called until it returns 0, meaning all the data has - * been flushed and the frame epilogue has been written. + * If enabled and dctx is allocated on the heap, then additional memory will be allocated + * to store references to multiple ZSTD_DDict. That is, multiple calls of ZSTD_refDDict() + * using a given ZSTD_DCtx, rather than overwriting the previous DDict reference, will instead + * store all references. At decompression time, the appropriate dictID is selected + * from the set of DDicts based on the dictID in the frame. * - * Return: The number of bytes still present within internal buffers or an - * error, which can be checked using ZSTD_isError(). - */ -size_t ZSTD_endStream(ZSTD_CStream *zcs, ZSTD_outBuffer *output); - -/** - * ZSTD_CStreamInSize() - recommended size for the input buffer + * Usage is simply calling ZSTD_refDDict() on multiple dict buffers. * - * Return: The recommended size for the input buffer. - */ -size_t ZSTD_CStreamInSize(void); -/** - * ZSTD_CStreamOutSize() - recommended size for the output buffer + * Param has values of byte ZSTD_refMultipleDDicts_e * - * When the output buffer is at least this large, it is guaranteed to be large - * enough to flush at least one complete compressed block. + * WARNING: Enabling this parameter and calling ZSTD_DCtx_refDDict(), will trigger memory + * allocation for the hash table. ZSTD_freeDCtx() also frees this memory. + * Memory is allocated as per ZSTD_DCtx::customMem. * - * Return: The recommended size for the output buffer. + * Although this function allocates memory for the table, the user is still responsible for + * memory management of the underlying ZSTD_DDict* themselves. */ -size_t ZSTD_CStreamOutSize(void); - +#define ZSTD_d_refMultipleDDicts ZSTD_d_experimentalParam4 -/*-***************************************************************************** - * Streaming decompression - HowTo - * - * A ZSTD_DStream object is required to track streaming operations. - * Use ZSTD_initDStream() to initialize a ZSTD_DStream object. - * ZSTD_DStream objects can be re-used multiple times. - * - * Use ZSTD_decompressStream() repetitively to consume your input. - * The function will update both `pos` fields. - * If `input->pos < input->size`, some input has not been consumed. - * It's up to the caller to present again remaining data. - * If `output->pos < output->size`, decoder has flushed everything it could. - * Returns 0 iff a frame is completely decoded and fully flushed. - * Otherwise it returns a suggested next input size that will never load more - * than the current frame. - ******************************************************************************/ +/*! ZSTD_DCtx_setFormat() : + * Instruct the decoder context about what kind of data to decode next. + * This instruction is mandatory to decode data without a fully-formed header, + * such ZSTD_f_zstd1_magicless for example. + * @return : 0, or an error code (which can be tested using ZSTD_isError()). */ +ZSTDLIB_API size_t ZSTD_DCtx_setFormat(ZSTD_DCtx* dctx, ZSTD_format_e format); -/** - * ZSTD_DStreamWorkspaceBound() - memory needed to initialize a ZSTD_DStream - * @maxWindowSize: The maximum window size allowed for compressed frames. - * - * Return: A lower bound on the size of the workspace that is passed to - * ZSTD_initDStream() and ZSTD_initDStream_usingDDict(). +/*! ZSTD_decompressStream_simpleArgs() : + * Same as ZSTD_decompressStream(), + * but using only integral types as arguments. + * This can be helpful for binders from dynamic languages + * which have troubles handling structures containing memory pointers. */ -size_t ZSTD_DStreamWorkspaceBound(size_t maxWindowSize); +ZSTDLIB_API size_t ZSTD_decompressStream_simpleArgs ( + ZSTD_DCtx* dctx, + void* dst, size_t dstCapacity, size_t* dstPos, + const void* src, size_t srcSize, size_t* srcPos); + + +/* ****************************************************************** +* Advanced streaming functions +* Warning : most of these functions are now redundant with the Advanced API. +* Once Advanced API reaches "stable" status, +* redundant functions will be deprecated, and then at some point removed. +********************************************************************/ + +/*===== Advanced Streaming compression functions =====*/ + +/*! ZSTD_initCStream_srcSize() : + * This function is deprecated, and equivalent to: + * ZSTD_CCtx_reset(zcs, ZSTD_reset_session_only); + * ZSTD_CCtx_refCDict(zcs, NULL); // clear the dictionary (if any) + * ZSTD_CCtx_setParameter(zcs, ZSTD_c_compressionLevel, compressionLevel); + * ZSTD_CCtx_setPledgedSrcSize(zcs, pledgedSrcSize); + * + * pledgedSrcSize must be correct. If it is not known at init time, use + * ZSTD_CONTENTSIZE_UNKNOWN. Note that, for compatibility with older programs, + * "0" also disables frame content size field. It may be enabled in the future. + * Note : this prototype will be marked as deprecated and generate compilation warnings on reaching v1.5.x + */ +ZSTDLIB_API size_t +ZSTD_initCStream_srcSize(ZSTD_CStream* zcs, + int compressionLevel, + unsigned long long pledgedSrcSize); + +/*! ZSTD_initCStream_usingDict() : + * This function is deprecated, and is equivalent to: + * ZSTD_CCtx_reset(zcs, ZSTD_reset_session_only); + * ZSTD_CCtx_setParameter(zcs, ZSTD_c_compressionLevel, compressionLevel); + * ZSTD_CCtx_loadDictionary(zcs, dict, dictSize); + * + * Creates of an internal CDict (incompatible with static CCtx), except if + * dict == NULL or dictSize < 8, in which case no dict is used. + * Note: dict is loaded with ZSTD_dct_auto (treated as a full zstd dictionary if + * it begins with ZSTD_MAGIC_DICTIONARY, else as raw content) and ZSTD_dlm_byCopy. + * Note : this prototype will be marked as deprecated and generate compilation warnings on reaching v1.5.x + */ +ZSTDLIB_API size_t +ZSTD_initCStream_usingDict(ZSTD_CStream* zcs, + const void* dict, size_t dictSize, + int compressionLevel); + +/*! ZSTD_initCStream_advanced() : + * This function is deprecated, and is approximately equivalent to: + * ZSTD_CCtx_reset(zcs, ZSTD_reset_session_only); + * // Pseudocode: Set each zstd parameter and leave the rest as-is. + * for ((param, value) : params) { + * ZSTD_CCtx_setParameter(zcs, param, value); + * } + * ZSTD_CCtx_setPledgedSrcSize(zcs, pledgedSrcSize); + * ZSTD_CCtx_loadDictionary(zcs, dict, dictSize); + * + * dict is loaded with ZSTD_dct_auto and ZSTD_dlm_byCopy. + * pledgedSrcSize must be correct. + * If srcSize is not known at init time, use value ZSTD_CONTENTSIZE_UNKNOWN. + * Note : this prototype will be marked as deprecated and generate compilation warnings on reaching v1.5.x + */ +ZSTDLIB_API size_t +ZSTD_initCStream_advanced(ZSTD_CStream* zcs, + const void* dict, size_t dictSize, + ZSTD_parameters params, + unsigned long long pledgedSrcSize); + +/*! ZSTD_initCStream_usingCDict() : + * This function is deprecated, and equivalent to: + * ZSTD_CCtx_reset(zcs, ZSTD_reset_session_only); + * ZSTD_CCtx_refCDict(zcs, cdict); + * + * note : cdict will just be referenced, and must outlive compression session + * Note : this prototype will be marked as deprecated and generate compilation warnings on reaching v1.5.x + */ +ZSTDLIB_API size_t ZSTD_initCStream_usingCDict(ZSTD_CStream* zcs, const ZSTD_CDict* cdict); + +/*! ZSTD_initCStream_usingCDict_advanced() : + * This function is DEPRECATED, and is approximately equivalent to: + * ZSTD_CCtx_reset(zcs, ZSTD_reset_session_only); + * // Pseudocode: Set each zstd frame parameter and leave the rest as-is. + * for ((fParam, value) : fParams) { + * ZSTD_CCtx_setParameter(zcs, fParam, value); + * } + * ZSTD_CCtx_setPledgedSrcSize(zcs, pledgedSrcSize); + * ZSTD_CCtx_refCDict(zcs, cdict); + * + * same as ZSTD_initCStream_usingCDict(), with control over frame parameters. + * pledgedSrcSize must be correct. If srcSize is not known at init time, use + * value ZSTD_CONTENTSIZE_UNKNOWN. + * Note : this prototype will be marked as deprecated and generate compilation warnings on reaching v1.5.x + */ +ZSTDLIB_API size_t +ZSTD_initCStream_usingCDict_advanced(ZSTD_CStream* zcs, + const ZSTD_CDict* cdict, + ZSTD_frameParameters fParams, + unsigned long long pledgedSrcSize); + +/*! ZSTD_resetCStream() : + * This function is deprecated, and is equivalent to: + * ZSTD_CCtx_reset(zcs, ZSTD_reset_session_only); + * ZSTD_CCtx_setPledgedSrcSize(zcs, pledgedSrcSize); + * + * start a new frame, using same parameters from previous frame. + * This is typically useful to skip dictionary loading stage, since it will re-use it in-place. + * Note that zcs must be init at least once before using ZSTD_resetCStream(). + * If pledgedSrcSize is not known at reset time, use macro ZSTD_CONTENTSIZE_UNKNOWN. + * If pledgedSrcSize > 0, its value must be correct, as it will be written in header, and controlled at the end. + * For the time being, pledgedSrcSize==0 is interpreted as "srcSize unknown" for compatibility with older programs, + * but it will change to mean "empty" in future version, so use macro ZSTD_CONTENTSIZE_UNKNOWN instead. + * @return : 0, or an error code (which can be tested using ZSTD_isError()) + * Note : this prototype will be marked as deprecated and generate compilation warnings on reaching v1.5.x + */ +ZSTDLIB_API size_t ZSTD_resetCStream(ZSTD_CStream* zcs, unsigned long long pledgedSrcSize); -/** - * struct ZSTD_DStream - the zstd streaming decompression context + +typedef struct { + unsigned long long ingested; /* nb input bytes read and buffered */ + unsigned long long consumed; /* nb input bytes actually compressed */ + unsigned long long produced; /* nb of compressed bytes generated and buffered */ + unsigned long long flushed; /* nb of compressed bytes flushed : not provided; can be tracked from caller side */ + unsigned currentJobID; /* MT only : latest started job nb */ + unsigned nbActiveWorkers; /* MT only : nb of workers actively compressing at probe time */ +} ZSTD_frameProgression; + +/* ZSTD_getFrameProgression() : + * tells how much data has been ingested (read from input) + * consumed (input actually compressed) and produced (output) for current frame. + * Note : (ingested - consumed) is amount of input data buffered internally, not yet compressed. + * Aggregates progression inside active worker threads. */ -typedef struct ZSTD_DStream_s ZSTD_DStream; -/*===== ZSTD_DStream management functions =====*/ -/** - * ZSTD_initDStream() - initialize a zstd streaming decompression context - * @maxWindowSize: The maximum window size allowed for compressed frames. - * @workspace: The workspace to emplace the context into. It must outlive - * the returned context. - * @workspaceSize: The size of workspace. - * Use ZSTD_DStreamWorkspaceBound(maxWindowSize) to determine - * how large the workspace must be. - * - * Return: The zstd streaming decompression context. - */ -ZSTD_DStream *ZSTD_initDStream(size_t maxWindowSize, void *workspace, - size_t workspaceSize); -/** - * ZSTD_initDStream_usingDDict() - initialize streaming decompression context - * @maxWindowSize: The maximum window size allowed for compressed frames. - * @ddict: The digested dictionary to use for decompression. - * @workspace: The workspace to emplace the context into. It must outlive - * the returned context. - * @workspaceSize: The size of workspace. - * Use ZSTD_DStreamWorkspaceBound(maxWindowSize) to determine - * how large the workspace must be. - * - * Return: The zstd streaming decompression context. - */ -ZSTD_DStream *ZSTD_initDStream_usingDDict(size_t maxWindowSize, - const ZSTD_DDict *ddict, void *workspace, size_t workspaceSize); +ZSTDLIB_API ZSTD_frameProgression ZSTD_getFrameProgression(const ZSTD_CCtx* cctx); + +/*! ZSTD_toFlushNow() : + * Tell how many bytes are ready to be flushed immediately. + * Useful for multithreading scenarios (nbWorkers >= 1). + * Probe the oldest active job, defined as oldest job not yet entirely flushed, + * and check its output buffer. + * @return : amount of data stored in oldest job and ready to be flushed immediately. + * if @return == 0, it means either : + * + there is no active job (could be checked with ZSTD_frameProgression()), or + * + oldest job is still actively compressing data, + * but everything it has produced has also been flushed so far, + * therefore flush speed is limited by production speed of oldest job + * irrespective of the speed of concurrent (and newer) jobs. + */ +ZSTDLIB_API size_t ZSTD_toFlushNow(ZSTD_CCtx* cctx); -/*===== Streaming decompression functions =====*/ -/** - * ZSTD_resetDStream() - reset the context using parameters from creation - * @zds: The zstd streaming decompression context to reset. + +/*===== Advanced Streaming decompression functions =====*/ + +/*! + * This function is deprecated, and is equivalent to: * - * Resets the context using the parameters from creation. Skips dictionary - * loading, since it can be reused. + * ZSTD_DCtx_reset(zds, ZSTD_reset_session_only); + * ZSTD_DCtx_loadDictionary(zds, dict, dictSize); * - * Return: Zero or an error, which can be checked using ZSTD_isError(). + * note: no dictionary will be used if dict == NULL or dictSize < 8 + * Note : this prototype will be marked as deprecated and generate compilation warnings on reaching v1.5.x */ -size_t ZSTD_resetDStream(ZSTD_DStream *zds); -/** - * ZSTD_decompressStream() - streaming decompress some of input into output - * @zds: The zstd streaming decompression context. - * @output: Destination buffer. `output.pos` is updated to indicate how much - * decompressed data was written. - * @input: Source buffer. `input.pos` is updated to indicate how much data was - * read. Note that it may not consume the entire input, in which case - * `input.pos < input.size`, and it's up to the caller to present - * remaining data again. +ZSTDLIB_API size_t ZSTD_initDStream_usingDict(ZSTD_DStream* zds, const void* dict, size_t dictSize); + +/*! + * This function is deprecated, and is equivalent to: * - * The `input` and `output` buffers may be any size. Guaranteed to make some - * forward progress if `input` and `output` are not empty. - * ZSTD_decompressStream() will not consume the last byte of the frame until - * the entire frame is flushed. + * ZSTD_DCtx_reset(zds, ZSTD_reset_session_only); + * ZSTD_DCtx_refDDict(zds, ddict); * - * Return: Returns 0 iff a frame is completely decoded and fully flushed. - * Otherwise returns a hint for the number of bytes to use as the input - * for the next function call or an error, which can be checked using - * ZSTD_isError(). The size hint will never load more than the frame. + * note : ddict is referenced, it must outlive decompression session + * Note : this prototype will be marked as deprecated and generate compilation warnings on reaching v1.5.x */ -size_t ZSTD_decompressStream(ZSTD_DStream *zds, ZSTD_outBuffer *output, - ZSTD_inBuffer *input); +ZSTDLIB_API size_t ZSTD_initDStream_usingDDict(ZSTD_DStream* zds, const ZSTD_DDict* ddict); -/** - * ZSTD_DStreamInSize() - recommended size for the input buffer - * - * Return: The recommended size for the input buffer. - */ -size_t ZSTD_DStreamInSize(void); -/** - * ZSTD_DStreamOutSize() - recommended size for the output buffer +/*! + * This function is deprecated, and is equivalent to: * - * When the output buffer is at least this large, it is guaranteed to be large - * enough to flush at least one complete decompressed block. + * ZSTD_DCtx_reset(zds, ZSTD_reset_session_only); * - * Return: The recommended size for the output buffer. + * re-use decompression parameters from previous init; saves dictionary loading + * Note : this prototype will be marked as deprecated and generate compilation warnings on reaching v1.5.x */ -size_t ZSTD_DStreamOutSize(void); +ZSTDLIB_API size_t ZSTD_resetDStream(ZSTD_DStream* zds); -/* --- Constants ---*/ -#define ZSTD_MAGICNUMBER 0xFD2FB528 /* >= v0.8.0 */ -#define ZSTD_MAGIC_SKIPPABLE_START 0x184D2A50U +/* ******************************************************************* +* Buffer-less and synchronous inner streaming functions +* +* This is an advanced API, giving full control over buffer management, for users which need direct control over memory. +* But it's also a complex one, with several restrictions, documented below. +* Prefer normal streaming API for an easier experience. +********************************************************************* */ -#define ZSTD_CONTENTSIZE_UNKNOWN (0ULL - 1) -#define ZSTD_CONTENTSIZE_ERROR (0ULL - 2) +/* + Buffer-less streaming compression (synchronous mode) + + A ZSTD_CCtx object is required to track streaming operations. + Use ZSTD_createCCtx() / ZSTD_freeCCtx() to manage resource. + ZSTD_CCtx object can be re-used multiple times within successive compression operations. + + Start by initializing a context. + Use ZSTD_compressBegin(), or ZSTD_compressBegin_usingDict() for dictionary compression, + or ZSTD_compressBegin_advanced(), for finer parameter control. + It's also possible to duplicate a reference context which has already been initialized, using ZSTD_copyCCtx() + + Then, consume your input using ZSTD_compressContinue(). + There are some important considerations to keep in mind when using this advanced function : + - ZSTD_compressContinue() has no internal buffer. It uses externally provided buffers only. + - Interface is synchronous : input is consumed entirely and produces 1+ compressed blocks. + - Caller must ensure there is enough space in `dst` to store compressed data under worst case scenario. + Worst case evaluation is provided by ZSTD_compressBound(). + ZSTD_compressContinue() doesn't guarantee recover after a failed compression. + - ZSTD_compressContinue() presumes prior input ***is still accessible and unmodified*** (up to maximum distance size, see WindowLog). + It remembers all previous contiguous blocks, plus one separated memory segment (which can itself consists of multiple contiguous blocks) + - ZSTD_compressContinue() detects that prior input has been overwritten when `src` buffer overlaps. + In which case, it will "discard" the relevant memory section from its history. + + Finish a frame with ZSTD_compressEnd(), which will write the last block(s) and optional checksum. + It's possible to use srcSize==0, in which case, it will write a final empty block to end the frame. + Without last block mark, frames are considered unfinished (hence corrupted) by compliant decoders. + + `ZSTD_CCtx` object can be re-used (ZSTD_compressBegin()) to compress again. +*/ -#define ZSTD_WINDOWLOG_MAX_32 27 -#define ZSTD_WINDOWLOG_MAX_64 27 -#define ZSTD_WINDOWLOG_MAX \ - ((unsigned int)(sizeof(size_t) == 4 \ - ? ZSTD_WINDOWLOG_MAX_32 \ - : ZSTD_WINDOWLOG_MAX_64)) -#define ZSTD_WINDOWLOG_MIN 10 -#define ZSTD_HASHLOG_MAX ZSTD_WINDOWLOG_MAX -#define ZSTD_HASHLOG_MIN 6 -#define ZSTD_CHAINLOG_MAX (ZSTD_WINDOWLOG_MAX+1) -#define ZSTD_CHAINLOG_MIN ZSTD_HASHLOG_MIN -#define ZSTD_HASHLOG3_MAX 17 -#define ZSTD_SEARCHLOG_MAX (ZSTD_WINDOWLOG_MAX-1) -#define ZSTD_SEARCHLOG_MIN 1 -/* only for ZSTD_fast, other strategies are limited to 6 */ -#define ZSTD_SEARCHLENGTH_MAX 7 -/* only for ZSTD_btopt, other strategies are limited to 4 */ -#define ZSTD_SEARCHLENGTH_MIN 3 -#define ZSTD_TARGETLENGTH_MIN 4 -#define ZSTD_TARGETLENGTH_MAX 999 - -/* for static allocation */ -#define ZSTD_FRAMEHEADERSIZE_MAX 18 -#define ZSTD_FRAMEHEADERSIZE_MIN 6 -#define ZSTD_frameHeaderSize_prefix 5 -#define ZSTD_frameHeaderSize_min ZSTD_FRAMEHEADERSIZE_MIN -#define ZSTD_frameHeaderSize_max ZSTD_FRAMEHEADERSIZE_MAX -/* magic number + skippable frame length */ -#define ZSTD_skippableHeaderSize 8 - - -/*-************************************* - * Compressed size functions - **************************************/ - -/** - * ZSTD_findFrameCompressedSize() - returns the size of a compressed frame - * @src: Source buffer. It should point to the start of a zstd encoded frame - * or a skippable frame. - * @srcSize: The size of the source buffer. It must be at least as large as the - * size of the frame. - * - * Return: The compressed size of the frame pointed to by `src` or an error, - * which can be check with ZSTD_isError(). - * Suitable to pass to ZSTD_decompress() or similar functions. - */ -size_t ZSTD_findFrameCompressedSize(const void *src, size_t srcSize); - -/*-************************************* - * Decompressed size functions - **************************************/ -/** - * ZSTD_getFrameContentSize() - returns the content size in a zstd frame header - * @src: It should point to the start of a zstd encoded frame. - * @srcSize: The size of the source buffer. It must be at least as large as the - * frame header. `ZSTD_frameHeaderSize_max` is always large enough. - * - * Return: The frame content size stored in the frame header if known. - * `ZSTD_CONTENTSIZE_UNKNOWN` if the content size isn't stored in the - * frame header. `ZSTD_CONTENTSIZE_ERROR` on invalid input. - */ -unsigned long long ZSTD_getFrameContentSize(const void *src, size_t srcSize); - -/** - * ZSTD_findDecompressedSize() - returns decompressed size of a series of frames - * @src: It should point to the start of a series of zstd encoded and/or - * skippable frames. - * @srcSize: The exact size of the series of frames. - * - * If any zstd encoded frame in the series doesn't have the frame content size - * set, `ZSTD_CONTENTSIZE_UNKNOWN` is returned. But frame content size is always - * set when using ZSTD_compress(). The decompressed size can be very large. - * If the source is untrusted, the decompressed size could be wrong or - * intentionally modified. Always ensure the result fits within the - * application's authorized limits. ZSTD_findDecompressedSize() handles multiple - * frames, and so it must traverse the input to read each frame header. This is - * efficient as most of the data is skipped, however it does mean that all frame - * data must be present and valid. - * - * Return: Decompressed size of all the data contained in the frames if known. - * `ZSTD_CONTENTSIZE_UNKNOWN` if the decompressed size is unknown. - * `ZSTD_CONTENTSIZE_ERROR` if an error occurred. - */ -unsigned long long ZSTD_findDecompressedSize(const void *src, size_t srcSize); - -/*-************************************* - * Advanced compression functions - **************************************/ -/** - * ZSTD_checkCParams() - ensure parameter values remain within authorized range - * @cParams: The zstd compression parameters. - * - * Return: Zero or an error, which can be checked using ZSTD_isError(). - */ -size_t ZSTD_checkCParams(ZSTD_compressionParameters cParams); - -/** - * ZSTD_adjustCParams() - optimize parameters for a given srcSize and dictSize - * @srcSize: Optionally the estimated source size, or zero if unknown. - * @dictSize: Optionally the estimated dictionary size, or zero if unknown. - * - * Return: The optimized parameters. - */ -ZSTD_compressionParameters ZSTD_adjustCParams( - ZSTD_compressionParameters cParams, unsigned long long srcSize, - size_t dictSize); - -/*--- Advanced decompression functions ---*/ - -/** - * ZSTD_isFrame() - returns true iff the buffer starts with a valid frame - * @buffer: The source buffer to check. - * @size: The size of the source buffer, must be at least 4 bytes. - * - * Return: True iff the buffer starts with a zstd or skippable frame identifier. - */ -unsigned int ZSTD_isFrame(const void *buffer, size_t size); - -/** - * ZSTD_getDictID_fromDict() - returns the dictionary id stored in a dictionary - * @dict: The dictionary buffer. - * @dictSize: The size of the dictionary buffer. - * - * Return: The dictionary id stored within the dictionary or 0 if the - * dictionary is not a zstd dictionary. If it returns 0 the - * dictionary can still be loaded as a content-only dictionary. - */ -unsigned int ZSTD_getDictID_fromDict(const void *dict, size_t dictSize); +/*===== Buffer-less streaming compression functions =====*/ +ZSTDLIB_API size_t ZSTD_compressBegin(ZSTD_CCtx* cctx, int compressionLevel); +ZSTDLIB_API size_t ZSTD_compressBegin_usingDict(ZSTD_CCtx* cctx, const void* dict, size_t dictSize, int compressionLevel); +ZSTDLIB_API size_t ZSTD_compressBegin_advanced(ZSTD_CCtx* cctx, const void* dict, size_t dictSize, ZSTD_parameters params, unsigned long long pledgedSrcSize); /*< pledgedSrcSize : If srcSize is not known at init time, use ZSTD_CONTENTSIZE_UNKNOWN */ +ZSTDLIB_API size_t ZSTD_compressBegin_usingCDict(ZSTD_CCtx* cctx, const ZSTD_CDict* cdict); /*< note: fails if cdict==NULL */ +ZSTDLIB_API size_t ZSTD_compressBegin_usingCDict_advanced(ZSTD_CCtx* const cctx, const ZSTD_CDict* const cdict, ZSTD_frameParameters const fParams, unsigned long long const pledgedSrcSize); /* compression parameters are already set within cdict. pledgedSrcSize must be correct. If srcSize is not known, use macro ZSTD_CONTENTSIZE_UNKNOWN */ +ZSTDLIB_API size_t ZSTD_copyCCtx(ZSTD_CCtx* cctx, const ZSTD_CCtx* preparedCCtx, unsigned long long pledgedSrcSize); /*< note: if pledgedSrcSize is not known, use ZSTD_CONTENTSIZE_UNKNOWN */ -/** - * ZSTD_getDictID_fromDDict() - returns the dictionary id stored in a ZSTD_DDict - * @ddict: The ddict to find the id of. - * - * Return: The dictionary id stored within `ddict` or 0 if the dictionary is not - * a zstd dictionary. If it returns 0 `ddict` will be loaded as a - * content-only dictionary. - */ -unsigned int ZSTD_getDictID_fromDDict(const ZSTD_DDict *ddict); +ZSTDLIB_API size_t ZSTD_compressContinue(ZSTD_CCtx* cctx, void* dst, size_t dstCapacity, const void* src, size_t srcSize); +ZSTDLIB_API size_t ZSTD_compressEnd(ZSTD_CCtx* cctx, void* dst, size_t dstCapacity, const void* src, size_t srcSize); -/** - * ZSTD_getDictID_fromFrame() - returns the dictionary id stored in a zstd frame - * @src: Source buffer. It must be a zstd encoded frame. - * @srcSize: The size of the source buffer. It must be at least as large as the - * frame header. `ZSTD_frameHeaderSize_max` is always large enough. - * - * Return: The dictionary id required to decompress the frame stored within - * `src` or 0 if the dictionary id could not be decoded. It can return - * 0 if the frame does not require a dictionary, the dictionary id - * wasn't stored in the frame, `src` is not a zstd frame, or `srcSize` - * is too small. - */ -unsigned int ZSTD_getDictID_fromFrame(const void *src, size_t srcSize); -/** - * struct ZSTD_frameParams - zstd frame parameters stored in the frame header - * @frameContentSize: The frame content size, or 0 if not present. - * @windowSize: The window size, or 0 if the frame is a skippable frame. - * @dictID: The dictionary id, or 0 if not present. - * @checksumFlag: Whether a checksum was used. - */ +/* + Buffer-less streaming decompression (synchronous mode) + + A ZSTD_DCtx object is required to track streaming operations. + Use ZSTD_createDCtx() / ZSTD_freeDCtx() to manage it. + A ZSTD_DCtx object can be re-used multiple times. + + First typical operation is to retrieve frame parameters, using ZSTD_getFrameHeader(). + Frame header is extracted from the beginning of compressed frame, so providing only the frame's beginning is enough. + Data fragment must be large enough to ensure successful decoding. + `ZSTD_frameHeaderSize_max` bytes is guaranteed to always be large enough. + @result : 0 : successful decoding, the `ZSTD_frameHeader` structure is correctly filled. + >0 : `srcSize` is too small, please provide at least @result bytes on next attempt. + errorCode, which can be tested using ZSTD_isError(). + + It fills a ZSTD_frameHeader structure with important information to correctly decode the frame, + such as the dictionary ID, content size, or maximum back-reference distance (`windowSize`). + Note that these values could be wrong, either because of data corruption, or because a 3rd party deliberately spoofs false information. + As a consequence, check that values remain within valid application range. + For example, do not allocate memory blindly, check that `windowSize` is within expectation. + Each application can set its own limits, depending on local restrictions. + For extended interoperability, it is recommended to support `windowSize` of at least 8 MB. + + ZSTD_decompressContinue() needs previous data blocks during decompression, up to `windowSize` bytes. + ZSTD_decompressContinue() is very sensitive to contiguity, + if 2 blocks don't follow each other, make sure that either the compressor breaks contiguity at the same place, + or that previous contiguous segment is large enough to properly handle maximum back-reference distance. + There are multiple ways to guarantee this condition. + + The most memory efficient way is to use a round buffer of sufficient size. + Sufficient size is determined by invoking ZSTD_decodingBufferSize_min(), + which can @return an error code if required value is too large for current system (in 32-bits mode). + In a round buffer methodology, ZSTD_decompressContinue() decompresses each block next to previous one, + up to the moment there is not enough room left in the buffer to guarantee decoding another full block, + which maximum size is provided in `ZSTD_frameHeader` structure, field `blockSizeMax`. + At which point, decoding can resume from the beginning of the buffer. + Note that already decoded data stored in the buffer should be flushed before being overwritten. + + There are alternatives possible, for example using two or more buffers of size `windowSize` each, though they consume more memory. + + Finally, if you control the compression process, you can also ignore all buffer size rules, + as long as the encoder and decoder progress in "lock-step", + aka use exactly the same buffer sizes, break contiguity at the same place, etc. + + Once buffers are setup, start decompression, with ZSTD_decompressBegin(). + If decompression requires a dictionary, use ZSTD_decompressBegin_usingDict() or ZSTD_decompressBegin_usingDDict(). + + Then use ZSTD_nextSrcSizeToDecompress() and ZSTD_decompressContinue() alternatively. + ZSTD_nextSrcSizeToDecompress() tells how many bytes to provide as 'srcSize' to ZSTD_decompressContinue(). + ZSTD_decompressContinue() requires this _exact_ amount of bytes, or it will fail. + + @result of ZSTD_decompressContinue() is the number of bytes regenerated within 'dst' (necessarily <= dstCapacity). + It can be zero : it just means ZSTD_decompressContinue() has decoded some metadata item. + It can also be an error code, which can be tested with ZSTD_isError(). + + A frame is fully decoded when ZSTD_nextSrcSizeToDecompress() returns zero. + Context can then be reset to start a new decompression. + + Note : it's possible to know if next input to present is a header or a block, using ZSTD_nextInputType(). + This information is not required to properly decode a frame. + + == Special case : skippable frames == + + Skippable frames allow integration of user-defined data into a flow of concatenated frames. + Skippable frames will be ignored (skipped) by decompressor. + The format of skippable frames is as follows : + a) Skippable frame ID - 4 Bytes, Little endian format, any value from 0x184D2A50 to 0x184D2A5F + b) Frame Size - 4 Bytes, Little endian format, unsigned 32-bits + c) Frame Content - any content (User Data) of length equal to Frame Size + For skippable frames ZSTD_getFrameHeader() returns zfhPtr->frameType==ZSTD_skippableFrame. + For skippable frames ZSTD_decompressContinue() always returns 0 : it only skips the content. +*/ + +/*===== Buffer-less streaming decompression functions =====*/ +typedef enum { ZSTD_frame, ZSTD_skippableFrame } ZSTD_frameType_e; typedef struct { - unsigned long long frameContentSize; - unsigned int windowSize; - unsigned int dictID; - unsigned int checksumFlag; -} ZSTD_frameParams; - -/** - * ZSTD_getFrameParams() - extracts parameters from a zstd or skippable frame - * @fparamsPtr: On success the frame parameters are written here. - * @src: The source buffer. It must point to a zstd or skippable frame. - * @srcSize: The size of the source buffer. `ZSTD_frameHeaderSize_max` is - * always large enough to succeed. - * - * Return: 0 on success. If more data is required it returns how many bytes - * must be provided to make forward progress. Otherwise it returns - * an error, which can be checked using ZSTD_isError(). - */ -size_t ZSTD_getFrameParams(ZSTD_frameParams *fparamsPtr, const void *src, - size_t srcSize); - -/*-***************************************************************************** - * Buffer-less and synchronous inner streaming functions - * - * This is an advanced API, giving full control over buffer management, for - * users which need direct control over memory. - * But it's also a complex one, with many restrictions (documented below). - * Prefer using normal streaming API for an easier experience - ******************************************************************************/ + unsigned long long frameContentSize; /* if == ZSTD_CONTENTSIZE_UNKNOWN, it means this field is not available. 0 means "empty" */ + unsigned long long windowSize; /* can be very large, up to <= frameContentSize */ + unsigned blockSizeMax; + ZSTD_frameType_e frameType; /* if == ZSTD_skippableFrame, frameContentSize is the size of skippable content */ + unsigned headerSize; + unsigned dictID; + unsigned checksumFlag; +} ZSTD_frameHeader; + +/*! ZSTD_getFrameHeader() : + * decode Frame Header, or requires larger `srcSize`. + * @return : 0, `zfhPtr` is correctly filled, + * >0, `srcSize` is too small, value is wanted `srcSize` amount, + * or an error code, which can be tested using ZSTD_isError() */ +ZSTDLIB_API size_t ZSTD_getFrameHeader(ZSTD_frameHeader* zfhPtr, const void* src, size_t srcSize); /*< doesn't consume input */ +/*! ZSTD_getFrameHeader_advanced() : + * same as ZSTD_getFrameHeader(), + * with added capability to select a format (like ZSTD_f_zstd1_magicless) */ +ZSTDLIB_API size_t ZSTD_getFrameHeader_advanced(ZSTD_frameHeader* zfhPtr, const void* src, size_t srcSize, ZSTD_format_e format); +ZSTDLIB_API size_t ZSTD_decodingBufferSize_min(unsigned long long windowSize, unsigned long long frameContentSize); /*< when frame content size is not known, pass in frameContentSize == ZSTD_CONTENTSIZE_UNKNOWN */ + +ZSTDLIB_API size_t ZSTD_decompressBegin(ZSTD_DCtx* dctx); +ZSTDLIB_API size_t ZSTD_decompressBegin_usingDict(ZSTD_DCtx* dctx, const void* dict, size_t dictSize); +ZSTDLIB_API size_t ZSTD_decompressBegin_usingDDict(ZSTD_DCtx* dctx, const ZSTD_DDict* ddict); + +ZSTDLIB_API size_t ZSTD_nextSrcSizeToDecompress(ZSTD_DCtx* dctx); +ZSTDLIB_API size_t ZSTD_decompressContinue(ZSTD_DCtx* dctx, void* dst, size_t dstCapacity, const void* src, size_t srcSize); + +/* misc */ +ZSTDLIB_API void ZSTD_copyDCtx(ZSTD_DCtx* dctx, const ZSTD_DCtx* preparedDCtx); +typedef enum { ZSTDnit_frameHeader, ZSTDnit_blockHeader, ZSTDnit_block, ZSTDnit_lastBlock, ZSTDnit_checksum, ZSTDnit_skippableFrame } ZSTD_nextInputType_e; +ZSTDLIB_API ZSTD_nextInputType_e ZSTD_nextInputType(ZSTD_DCtx* dctx); + + + + +/* ============================ */ +/* Block level API */ +/* ============================ */ + +/*! + Block functions produce and decode raw zstd blocks, without frame metadata. + Frame metadata cost is typically ~12 bytes, which can be non-negligible for very small blocks (< 100 bytes). + But users will have to take in charge needed metadata to regenerate data, such as compressed and content sizes. + + A few rules to respect : + - Compressing and decompressing require a context structure + + Use ZSTD_createCCtx() and ZSTD_createDCtx() + - It is necessary to init context before starting + + compression : any ZSTD_compressBegin*() variant, including with dictionary + + decompression : any ZSTD_decompressBegin*() variant, including with dictionary + + copyCCtx() and copyDCtx() can be used too + - Block size is limited, it must be <= ZSTD_getBlockSize() <= ZSTD_BLOCKSIZE_MAX == 128 KB + + If input is larger than a block size, it's necessary to split input data into multiple blocks + + For inputs larger than a single block, consider using regular ZSTD_compress() instead. + Frame metadata is not that costly, and quickly becomes negligible as source size grows larger than a block. + - When a block is considered not compressible enough, ZSTD_compressBlock() result will be 0 (zero) ! + ===> In which case, nothing is produced into `dst` ! + + User __must__ test for such outcome and deal directly with uncompressed data + + A block cannot be declared incompressible if ZSTD_compressBlock() return value was != 0. + Doing so would mess up with statistics history, leading to potential data corruption. + + ZSTD_decompressBlock() _doesn't accept uncompressed data as input_ !! + + In case of multiple successive blocks, should some of them be uncompressed, + decoder must be informed of their existence in order to follow proper history. + Use ZSTD_insertBlock() for such a case. +*/ -/*-***************************************************************************** - * Buffer-less streaming compression (synchronous mode) - * - * A ZSTD_CCtx object is required to track streaming operations. - * Use ZSTD_initCCtx() to initialize a context. - * ZSTD_CCtx object can be re-used multiple times within successive compression - * operations. - * - * Start by initializing a context. - * Use ZSTD_compressBegin(), or ZSTD_compressBegin_usingDict() for dictionary - * compression, - * or ZSTD_compressBegin_advanced(), for finer parameter control. - * It's also possible to duplicate a reference context which has already been - * initialized, using ZSTD_copyCCtx() - * - * Then, consume your input using ZSTD_compressContinue(). - * There are some important considerations to keep in mind when using this - * advanced function : - * - ZSTD_compressContinue() has no internal buffer. It uses externally provided - * buffer only. - * - Interface is synchronous : input is consumed entirely and produce 1+ - * (or more) compressed blocks. - * - Caller must ensure there is enough space in `dst` to store compressed data - * under worst case scenario. Worst case evaluation is provided by - * ZSTD_compressBound(). - * ZSTD_compressContinue() doesn't guarantee recover after a failed - * compression. - * - ZSTD_compressContinue() presumes prior input ***is still accessible and - * unmodified*** (up to maximum distance size, see WindowLog). - * It remembers all previous contiguous blocks, plus one separated memory - * segment (which can itself consists of multiple contiguous blocks) - * - ZSTD_compressContinue() detects that prior input has been overwritten when - * `src` buffer overlaps. In which case, it will "discard" the relevant memory - * section from its history. - * - * Finish a frame with ZSTD_compressEnd(), which will write the last block(s) - * and optional checksum. It's possible to use srcSize==0, in which case, it - * will write a final empty block to end the frame. Without last block mark, - * frames will be considered unfinished (corrupted) by decoders. - * - * `ZSTD_CCtx` object can be re-used (ZSTD_compressBegin()) to compress some new - * frame. - ******************************************************************************/ +/*===== Raw zstd block functions =====*/ +ZSTDLIB_API size_t ZSTD_getBlockSize (const ZSTD_CCtx* cctx); +ZSTDLIB_API size_t ZSTD_compressBlock (ZSTD_CCtx* cctx, void* dst, size_t dstCapacity, const void* src, size_t srcSize); +ZSTDLIB_API size_t ZSTD_decompressBlock(ZSTD_DCtx* dctx, void* dst, size_t dstCapacity, const void* src, size_t srcSize); +ZSTDLIB_API size_t ZSTD_insertBlock (ZSTD_DCtx* dctx, const void* blockStart, size_t blockSize); /*< insert uncompressed block into `dctx` history. Useful for multi-blocks decompression. */ -/*===== Buffer-less streaming compression functions =====*/ -size_t ZSTD_compressBegin(ZSTD_CCtx *cctx, int compressionLevel); -size_t ZSTD_compressBegin_usingDict(ZSTD_CCtx *cctx, const void *dict, - size_t dictSize, int compressionLevel); -size_t ZSTD_compressBegin_advanced(ZSTD_CCtx *cctx, const void *dict, - size_t dictSize, ZSTD_parameters params, - unsigned long long pledgedSrcSize); -size_t ZSTD_copyCCtx(ZSTD_CCtx *cctx, const ZSTD_CCtx *preparedCCtx, - unsigned long long pledgedSrcSize); -size_t ZSTD_compressBegin_usingCDict(ZSTD_CCtx *cctx, const ZSTD_CDict *cdict, - unsigned long long pledgedSrcSize); -size_t ZSTD_compressContinue(ZSTD_CCtx *cctx, void *dst, size_t dstCapacity, - const void *src, size_t srcSize); -size_t ZSTD_compressEnd(ZSTD_CCtx *cctx, void *dst, size_t dstCapacity, - const void *src, size_t srcSize); - - - -/*-***************************************************************************** - * Buffer-less streaming decompression (synchronous mode) - * - * A ZSTD_DCtx object is required to track streaming operations. - * Use ZSTD_initDCtx() to initialize a context. - * A ZSTD_DCtx object can be re-used multiple times. - * - * First typical operation is to retrieve frame parameters, using - * ZSTD_getFrameParams(). It fills a ZSTD_frameParams structure which provide - * important information to correctly decode the frame, such as the minimum - * rolling buffer size to allocate to decompress data (`windowSize`), and the - * dictionary ID used. - * Note: content size is optional, it may not be present. 0 means unknown. - * Note that these values could be wrong, either because of data malformation, - * or because an attacker is spoofing deliberate false information. As a - * consequence, check that values remain within valid application range, - * especially `windowSize`, before allocation. Each application can set its own - * limit, depending on local restrictions. For extended interoperability, it is - * recommended to support at least 8 MB. - * Frame parameters are extracted from the beginning of the compressed frame. - * Data fragment must be large enough to ensure successful decoding, typically - * `ZSTD_frameHeaderSize_max` bytes. - * Result: 0: successful decoding, the `ZSTD_frameParams` structure is filled. - * >0: `srcSize` is too small, provide at least this many bytes. - * errorCode, which can be tested using ZSTD_isError(). - * - * Start decompression, with ZSTD_decompressBegin() or - * ZSTD_decompressBegin_usingDict(). Alternatively, you can copy a prepared - * context, using ZSTD_copyDCtx(). - * - * Then use ZSTD_nextSrcSizeToDecompress() and ZSTD_decompressContinue() - * alternatively. - * ZSTD_nextSrcSizeToDecompress() tells how many bytes to provide as 'srcSize' - * to ZSTD_decompressContinue(). - * ZSTD_decompressContinue() requires this _exact_ amount of bytes, or it will - * fail. - * - * The result of ZSTD_decompressContinue() is the number of bytes regenerated - * within 'dst' (necessarily <= dstCapacity). It can be zero, which is not an - * error; it just means ZSTD_decompressContinue() has decoded some metadata - * item. It can also be an error code, which can be tested with ZSTD_isError(). - * - * ZSTD_decompressContinue() needs previous data blocks during decompression, up - * to `windowSize`. They should preferably be located contiguously, prior to - * current block. Alternatively, a round buffer of sufficient size is also - * possible. Sufficient size is determined by frame parameters. - * ZSTD_decompressContinue() is very sensitive to contiguity, if 2 blocks don't - * follow each other, make sure that either the compressor breaks contiguity at - * the same place, or that previous contiguous segment is large enough to - * properly handle maximum back-reference. - * - * A frame is fully decoded when ZSTD_nextSrcSizeToDecompress() returns zero. - * Context can then be reset to start a new decompression. - * - * Note: it's possible to know if next input to present is a header or a block, - * using ZSTD_nextInputType(). This information is not required to properly - * decode a frame. - * - * == Special case: skippable frames == - * - * Skippable frames allow integration of user-defined data into a flow of - * concatenated frames. Skippable frames will be ignored (skipped) by a - * decompressor. The format of skippable frames is as follows: - * a) Skippable frame ID - 4 Bytes, Little endian format, any value from - * 0x184D2A50 to 0x184D2A5F - * b) Frame Size - 4 Bytes, Little endian format, unsigned 32-bits - * c) Frame Content - any content (User Data) of length equal to Frame Size - * For skippable frames ZSTD_decompressContinue() always returns 0. - * For skippable frames ZSTD_getFrameParams() returns fparamsPtr->windowLog==0 - * what means that a frame is skippable. - * Note: If fparamsPtr->frameContentSize==0, it is ambiguous: the frame might - * actually be a zstd encoded frame with no content. For purposes of - * decompression, it is valid in both cases to skip the frame using - * ZSTD_findFrameCompressedSize() to find its size in bytes. - * It also returns frame size as fparamsPtr->frameContentSize. - ******************************************************************************/ -/*===== Buffer-less streaming decompression functions =====*/ -size_t ZSTD_decompressBegin(ZSTD_DCtx *dctx); -size_t ZSTD_decompressBegin_usingDict(ZSTD_DCtx *dctx, const void *dict, - size_t dictSize); -void ZSTD_copyDCtx(ZSTD_DCtx *dctx, const ZSTD_DCtx *preparedDCtx); -size_t ZSTD_nextSrcSizeToDecompress(ZSTD_DCtx *dctx); -size_t ZSTD_decompressContinue(ZSTD_DCtx *dctx, void *dst, size_t dstCapacity, - const void *src, size_t srcSize); -typedef enum { - ZSTDnit_frameHeader, - ZSTDnit_blockHeader, - ZSTDnit_block, - ZSTDnit_lastBlock, - ZSTDnit_checksum, - ZSTDnit_skippableFrame -} ZSTD_nextInputType_e; -ZSTD_nextInputType_e ZSTD_nextInputType(ZSTD_DCtx *dctx); - -/*-***************************************************************************** - * Block functions - * - * Block functions produce and decode raw zstd blocks, without frame metadata. - * Frame metadata cost is typically ~18 bytes, which can be non-negligible for - * very small blocks (< 100 bytes). User will have to take in charge required - * information to regenerate data, such as compressed and content sizes. - * - * A few rules to respect: - * - Compressing and decompressing require a context structure - * + Use ZSTD_initCCtx() and ZSTD_initDCtx() - * - It is necessary to init context before starting - * + compression : ZSTD_compressBegin() - * + decompression : ZSTD_decompressBegin() - * + variants _usingDict() are also allowed - * + copyCCtx() and copyDCtx() work too - * - Block size is limited, it must be <= ZSTD_getBlockSizeMax() - * + If you need to compress more, cut data into multiple blocks - * + Consider using the regular ZSTD_compress() instead, as frame metadata - * costs become negligible when source size is large. - * - When a block is considered not compressible enough, ZSTD_compressBlock() - * result will be zero. In which case, nothing is produced into `dst`. - * + User must test for such outcome and deal directly with uncompressed data - * + ZSTD_decompressBlock() doesn't accept uncompressed data as input!!! - * + In case of multiple successive blocks, decoder must be informed of - * uncompressed block existence to follow proper history. Use - * ZSTD_insertBlock() in such a case. - ******************************************************************************/ +#endif /* ZSTD_H_ZSTD_STATIC_LINKING_ONLY */ -/* Define for static allocation */ -#define ZSTD_BLOCKSIZE_ABSOLUTEMAX (128 * 1024) -/*===== Raw zstd block functions =====*/ -size_t ZSTD_getBlockSizeMax(ZSTD_CCtx *cctx); -size_t ZSTD_compressBlock(ZSTD_CCtx *cctx, void *dst, size_t dstCapacity, - const void *src, size_t srcSize); -size_t ZSTD_decompressBlock(ZSTD_DCtx *dctx, void *dst, size_t dstCapacity, - const void *src, size_t srcSize); -size_t ZSTD_insertBlock(ZSTD_DCtx *dctx, const void *blockStart, - size_t blockSize); - -#endif /* ZSTD_H */ diff --git a/lib/zstd/Makefile b/lib/zstd/Makefile index f5d778e7e5c7..65218ec5b8f2 100644 --- a/lib/zstd/Makefile +++ b/lib/zstd/Makefile @@ -1,10 +1,46 @@ -# SPDX-License-Identifier: GPL-2.0-only +# SPDX-License-Identifier: GPL-2.0+ OR BSD-3-Clause +# ################################################################ +# Copyright (c) Facebook, Inc. +# All rights reserved. +# +# This source code is licensed under both the BSD-style license (found in the +# LICENSE file in the root directory of this source tree) and the GPLv2 (found +# in the COPYING file in the root directory of this source tree). +# You may select, at your option, one of the above-listed licenses. +# ################################################################ obj-$(CONFIG_ZSTD_COMPRESS) += zstd_compress.o obj-$(CONFIG_ZSTD_DECOMPRESS) += zstd_decompress.o ccflags-y += -O3 -zstd_compress-y := fse_compress.o huf_compress.o compress.o \ - entropy_common.o fse_decompress.o zstd_common.o -zstd_decompress-y := huf_decompress.o decompress.o \ - entropy_common.o fse_decompress.o zstd_common.o +zstd_compress-y := \ + zstd_compress_module.o \ + common/debug.o \ + common/entropy_common.o \ + common/error_private.o \ + common/fse_decompress.o \ + common/zstd_common.o \ + compress/fse_compress.o \ + compress/hist.o \ + compress/huf_compress.o \ + compress/zstd_compress.o \ + compress/zstd_compress_literals.o \ + compress/zstd_compress_sequences.o \ + compress/zstd_compress_superblock.o \ + compress/zstd_double_fast.o \ + compress/zstd_fast.o \ + compress/zstd_lazy.o \ + compress/zstd_ldm.o \ + compress/zstd_opt.o \ + +zstd_decompress-y := \ + zstd_decompress_module.o \ + common/debug.o \ + common/entropy_common.o \ + common/error_private.o \ + common/fse_decompress.o \ + common/zstd_common.o \ + decompress/huf_decompress.o \ + decompress/zstd_ddict.o \ + decompress/zstd_decompress.o \ + decompress/zstd_decompress_block.o \ diff --git a/lib/zstd/bitstream.h b/lib/zstd/bitstream.h deleted file mode 100644 index 5d6343c1a909..000000000000 --- a/lib/zstd/bitstream.h +++ /dev/null @@ -1,380 +0,0 @@ -/* - * bitstream - * Part of FSE library - * header file (to include) - * Copyright (C) 2013-2016, Yann Collet. - * - * BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php) - * - * Redistribution and use in source and binary forms, with or without - * modification, are permitted provided that the following conditions are - * met: - * - * * Redistributions of source code must retain the above copyright - * notice, this list of conditions and the following disclaimer. - * * Redistributions in binary form must reproduce the above - * copyright notice, this list of conditions and the following disclaimer - * in the documentation and/or other materials provided with the - * distribution. - * - * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS - * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT - * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR - * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT - * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, - * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT - * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, - * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY - * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT - * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE - * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. - * - * This program is free software; you can redistribute it and/or modify it under - * the terms of the GNU General Public License version 2 as published by the - * Free Software Foundation. This program is dual-licensed; you may select - * either version 2 of the GNU General Public License ("GPL") or BSD license - * ("BSD"). - * - * You can contact the author at : - * - Source repository : https://github.com/Cyan4973/FiniteStateEntropy - */ -#ifndef BITSTREAM_H_MODULE -#define BITSTREAM_H_MODULE - -/* -* This API consists of small unitary functions, which must be inlined for best performance. -* Since link-time-optimization is not available for all compilers, -* these functions are defined into a .h to be included. -*/ - -/*-**************************************** -* Dependencies -******************************************/ -#include "error_private.h" /* error codes and messages */ -#include "mem.h" /* unaligned access routines */ - -/*========================================= -* Target specific -=========================================*/ -#define STREAM_ACCUMULATOR_MIN_32 25 -#define STREAM_ACCUMULATOR_MIN_64 57 -#define STREAM_ACCUMULATOR_MIN ((U32)(ZSTD_32bits() ? STREAM_ACCUMULATOR_MIN_32 : STREAM_ACCUMULATOR_MIN_64)) - -/*-****************************************** -* bitStream encoding API (write forward) -********************************************/ -/* bitStream can mix input from multiple sources. -* A critical property of these streams is that they encode and decode in **reverse** direction. -* So the first bit sequence you add will be the last to be read, like a LIFO stack. -*/ -typedef struct { - size_t bitContainer; - int bitPos; - char *startPtr; - char *ptr; - char *endPtr; -} BIT_CStream_t; - -ZSTD_STATIC size_t BIT_initCStream(BIT_CStream_t *bitC, void *dstBuffer, size_t dstCapacity); -ZSTD_STATIC void BIT_addBits(BIT_CStream_t *bitC, size_t value, unsigned nbBits); -ZSTD_STATIC void BIT_flushBits(BIT_CStream_t *bitC); -ZSTD_STATIC size_t BIT_closeCStream(BIT_CStream_t *bitC); - -/* Start with initCStream, providing the size of buffer to write into. -* bitStream will never write outside of this buffer. -* `dstCapacity` must be >= sizeof(bitD->bitContainer), otherwise @return will be an error code. -* -* bits are first added to a local register. -* Local register is size_t, hence 64-bits on 64-bits systems, or 32-bits on 32-bits systems. -* Writing data into memory is an explicit operation, performed by the flushBits function. -* Hence keep track how many bits are potentially stored into local register to avoid register overflow. -* After a flushBits, a maximum of 7 bits might still be stored into local register. -* -* Avoid storing elements of more than 24 bits if you want compatibility with 32-bits bitstream readers. -* -* Last operation is to close the bitStream. -* The function returns the final size of CStream in bytes. -* If data couldn't fit into `dstBuffer`, it will return a 0 ( == not storable) -*/ - -/*-******************************************** -* bitStream decoding API (read backward) -**********************************************/ -typedef struct { - size_t bitContainer; - unsigned bitsConsumed; - const char *ptr; - const char *start; -} BIT_DStream_t; - -typedef enum { - BIT_DStream_unfinished = 0, - BIT_DStream_endOfBuffer = 1, - BIT_DStream_completed = 2, - BIT_DStream_overflow = 3 -} BIT_DStream_status; /* result of BIT_reloadDStream() */ -/* 1,2,4,8 would be better for bitmap combinations, but slows down performance a bit ... :( */ - -ZSTD_STATIC size_t BIT_initDStream(BIT_DStream_t *bitD, const void *srcBuffer, size_t srcSize); -ZSTD_STATIC size_t BIT_readBits(BIT_DStream_t *bitD, unsigned nbBits); -ZSTD_STATIC BIT_DStream_status BIT_reloadDStream(BIT_DStream_t *bitD); -ZSTD_STATIC unsigned BIT_endOfDStream(const BIT_DStream_t *bitD); - -/* Start by invoking BIT_initDStream(). -* A chunk of the bitStream is then stored into a local register. -* Local register size is 64-bits on 64-bits systems, 32-bits on 32-bits systems (size_t). -* You can then retrieve bitFields stored into the local register, **in reverse order**. -* Local register is explicitly reloaded from memory by the BIT_reloadDStream() method. -* A reload guarantee a minimum of ((8*sizeof(bitD->bitContainer))-7) bits when its result is BIT_DStream_unfinished. -* Otherwise, it can be less than that, so proceed accordingly. -* Checking if DStream has reached its end can be performed with BIT_endOfDStream(). -*/ - -/*-**************************************** -* unsafe API -******************************************/ -ZSTD_STATIC void BIT_addBitsFast(BIT_CStream_t *bitC, size_t value, unsigned nbBits); -/* faster, but works only if value is "clean", meaning all high bits above nbBits are 0 */ - -ZSTD_STATIC void BIT_flushBitsFast(BIT_CStream_t *bitC); -/* unsafe version; does not check buffer overflow */ - -ZSTD_STATIC size_t BIT_readBitsFast(BIT_DStream_t *bitD, unsigned nbBits); -/* faster, but works only if nbBits >= 1 */ - -/*-************************************************************** -* Internal functions -****************************************************************/ -ZSTD_STATIC unsigned BIT_highbit32(register U32 val) { return 31 - __builtin_clz(val); } - -/*===== Local Constants =====*/ -static const unsigned BIT_mask[] = {0, 1, 3, 7, 0xF, 0x1F, 0x3F, 0x7F, 0xFF, - 0x1FF, 0x3FF, 0x7FF, 0xFFF, 0x1FFF, 0x3FFF, 0x7FFF, 0xFFFF, 0x1FFFF, - 0x3FFFF, 0x7FFFF, 0xFFFFF, 0x1FFFFF, 0x3FFFFF, 0x7FFFFF, 0xFFFFFF, 0x1FFFFFF, 0x3FFFFFF}; /* up to 26 bits */ - -/*-************************************************************** -* bitStream encoding -****************************************************************/ -/*! BIT_initCStream() : - * `dstCapacity` must be > sizeof(void*) - * @return : 0 if success, - otherwise an error code (can be tested using ERR_isError() ) */ -ZSTD_STATIC size_t BIT_initCStream(BIT_CStream_t *bitC, void *startPtr, size_t dstCapacity) -{ - bitC->bitContainer = 0; - bitC->bitPos = 0; - bitC->startPtr = (char *)startPtr; - bitC->ptr = bitC->startPtr; - bitC->endPtr = bitC->startPtr + dstCapacity - sizeof(bitC->ptr); - if (dstCapacity <= sizeof(bitC->ptr)) - return ERROR(dstSize_tooSmall); - return 0; -} - -/*! BIT_addBits() : - can add up to 26 bits into `bitC`. - Does not check for register overflow ! */ -ZSTD_STATIC void BIT_addBits(BIT_CStream_t *bitC, size_t value, unsigned nbBits) -{ - bitC->bitContainer |= (value & BIT_mask[nbBits]) << bitC->bitPos; - bitC->bitPos += nbBits; -} - -/*! BIT_addBitsFast() : - * works only if `value` is _clean_, meaning all high bits above nbBits are 0 */ -ZSTD_STATIC void BIT_addBitsFast(BIT_CStream_t *bitC, size_t value, unsigned nbBits) -{ - bitC->bitContainer |= value << bitC->bitPos; - bitC->bitPos += nbBits; -} - -/*! BIT_flushBitsFast() : - * unsafe version; does not check buffer overflow */ -ZSTD_STATIC void BIT_flushBitsFast(BIT_CStream_t *bitC) -{ - size_t const nbBytes = bitC->bitPos >> 3; - ZSTD_writeLEST(bitC->ptr, bitC->bitContainer); - bitC->ptr += nbBytes; - bitC->bitPos &= 7; - bitC->bitContainer >>= nbBytes * 8; /* if bitPos >= sizeof(bitContainer)*8 --> undefined behavior */ -} - -/*! BIT_flushBits() : - * safe version; check for buffer overflow, and prevents it. - * note : does not signal buffer overflow. This will be revealed later on using BIT_closeCStream() */ -ZSTD_STATIC void BIT_flushBits(BIT_CStream_t *bitC) -{ - size_t const nbBytes = bitC->bitPos >> 3; - ZSTD_writeLEST(bitC->ptr, bitC->bitContainer); - bitC->ptr += nbBytes; - if (bitC->ptr > bitC->endPtr) - bitC->ptr = bitC->endPtr; - bitC->bitPos &= 7; - bitC->bitContainer >>= nbBytes * 8; /* if bitPos >= sizeof(bitContainer)*8 --> undefined behavior */ -} - -/*! BIT_closeCStream() : - * @return : size of CStream, in bytes, - or 0 if it could not fit into dstBuffer */ -ZSTD_STATIC size_t BIT_closeCStream(BIT_CStream_t *bitC) -{ - BIT_addBitsFast(bitC, 1, 1); /* endMark */ - BIT_flushBits(bitC); - - if (bitC->ptr >= bitC->endPtr) - return 0; /* doesn't fit within authorized budget : cancel */ - - return (bitC->ptr - bitC->startPtr) + (bitC->bitPos > 0); -} - -/*-******************************************************** -* bitStream decoding -**********************************************************/ -/*! BIT_initDStream() : -* Initialize a BIT_DStream_t. -* `bitD` : a pointer to an already allocated BIT_DStream_t structure. -* `srcSize` must be the *exact* size of the bitStream, in bytes. -* @return : size of stream (== srcSize) or an errorCode if a problem is detected -*/ -ZSTD_STATIC size_t BIT_initDStream(BIT_DStream_t *bitD, const void *srcBuffer, size_t srcSize) -{ - if (srcSize < 1) { - memset(bitD, 0, sizeof(*bitD)); - return ERROR(srcSize_wrong); - } - - if (srcSize >= sizeof(bitD->bitContainer)) { /* normal case */ - bitD->start = (const char *)srcBuffer; - bitD->ptr = (const char *)srcBuffer + srcSize - sizeof(bitD->bitContainer); - bitD->bitContainer = ZSTD_readLEST(bitD->ptr); - { - BYTE const lastByte = ((const BYTE *)srcBuffer)[srcSize - 1]; - bitD->bitsConsumed = lastByte ? 8 - BIT_highbit32(lastByte) : 0; /* ensures bitsConsumed is always set */ - if (lastByte == 0) - return ERROR(GENERIC); /* endMark not present */ - } - } else { - bitD->start = (const char *)srcBuffer; - bitD->ptr = bitD->start; - bitD->bitContainer = *(const BYTE *)(bitD->start); - switch (srcSize) { - case 7: bitD->bitContainer += (size_t)(((const BYTE *)(srcBuffer))[6]) << (sizeof(bitD->bitContainer) * 8 - 16); - fallthrough; - case 6: bitD->bitContainer += (size_t)(((const BYTE *)(srcBuffer))[5]) << (sizeof(bitD->bitContainer) * 8 - 24); - fallthrough; - case 5: bitD->bitContainer += (size_t)(((const BYTE *)(srcBuffer))[4]) << (sizeof(bitD->bitContainer) * 8 - 32); - fallthrough; - case 4: bitD->bitContainer += (size_t)(((const BYTE *)(srcBuffer))[3]) << 24; - fallthrough; - case 3: bitD->bitContainer += (size_t)(((const BYTE *)(srcBuffer))[2]) << 16; - fallthrough; - case 2: bitD->bitContainer += (size_t)(((const BYTE *)(srcBuffer))[1]) << 8; - fallthrough; - default:; - } - { - BYTE const lastByte = ((const BYTE *)srcBuffer)[srcSize - 1]; - bitD->bitsConsumed = lastByte ? 8 - BIT_highbit32(lastByte) : 0; - if (lastByte == 0) - return ERROR(GENERIC); /* endMark not present */ - } - bitD->bitsConsumed += (U32)(sizeof(bitD->bitContainer) - srcSize) * 8; - } - - return srcSize; -} - -ZSTD_STATIC size_t BIT_getUpperBits(size_t bitContainer, U32 const start) { return bitContainer >> start; } - -ZSTD_STATIC size_t BIT_getMiddleBits(size_t bitContainer, U32 const start, U32 const nbBits) { return (bitContainer >> start) & BIT_mask[nbBits]; } - -ZSTD_STATIC size_t BIT_getLowerBits(size_t bitContainer, U32 const nbBits) { return bitContainer & BIT_mask[nbBits]; } - -/*! BIT_lookBits() : - * Provides next n bits from local register. - * local register is not modified. - * On 32-bits, maxNbBits==24. - * On 64-bits, maxNbBits==56. - * @return : value extracted - */ -ZSTD_STATIC size_t BIT_lookBits(const BIT_DStream_t *bitD, U32 nbBits) -{ - U32 const bitMask = sizeof(bitD->bitContainer) * 8 - 1; - return ((bitD->bitContainer << (bitD->bitsConsumed & bitMask)) >> 1) >> ((bitMask - nbBits) & bitMask); -} - -/*! BIT_lookBitsFast() : -* unsafe version; only works only if nbBits >= 1 */ -ZSTD_STATIC size_t BIT_lookBitsFast(const BIT_DStream_t *bitD, U32 nbBits) -{ - U32 const bitMask = sizeof(bitD->bitContainer) * 8 - 1; - return (bitD->bitContainer << (bitD->bitsConsumed & bitMask)) >> (((bitMask + 1) - nbBits) & bitMask); -} - -ZSTD_STATIC void BIT_skipBits(BIT_DStream_t *bitD, U32 nbBits) { bitD->bitsConsumed += nbBits; } - -/*! BIT_readBits() : - * Read (consume) next n bits from local register and update. - * Pay attention to not read more than nbBits contained into local register. - * @return : extracted value. - */ -ZSTD_STATIC size_t BIT_readBits(BIT_DStream_t *bitD, U32 nbBits) -{ - size_t const value = BIT_lookBits(bitD, nbBits); - BIT_skipBits(bitD, nbBits); - return value; -} - -/*! BIT_readBitsFast() : -* unsafe version; only works only if nbBits >= 1 */ -ZSTD_STATIC size_t BIT_readBitsFast(BIT_DStream_t *bitD, U32 nbBits) -{ - size_t const value = BIT_lookBitsFast(bitD, nbBits); - BIT_skipBits(bitD, nbBits); - return value; -} - -/*! BIT_reloadDStream() : -* Refill `bitD` from buffer previously set in BIT_initDStream() . -* This function is safe, it guarantees it will not read beyond src buffer. -* @return : status of `BIT_DStream_t` internal register. - if status == BIT_DStream_unfinished, internal register is filled with >= (sizeof(bitD->bitContainer)*8 - 7) bits */ -ZSTD_STATIC BIT_DStream_status BIT_reloadDStream(BIT_DStream_t *bitD) -{ - if (bitD->bitsConsumed > (sizeof(bitD->bitContainer) * 8)) /* should not happen => corruption detected */ - return BIT_DStream_overflow; - - if (bitD->ptr >= bitD->start + sizeof(bitD->bitContainer)) { - bitD->ptr -= bitD->bitsConsumed >> 3; - bitD->bitsConsumed &= 7; - bitD->bitContainer = ZSTD_readLEST(bitD->ptr); - return BIT_DStream_unfinished; - } - if (bitD->ptr == bitD->start) { - if (bitD->bitsConsumed < sizeof(bitD->bitContainer) * 8) - return BIT_DStream_endOfBuffer; - return BIT_DStream_completed; - } - { - U32 nbBytes = bitD->bitsConsumed >> 3; - BIT_DStream_status result = BIT_DStream_unfinished; - if (bitD->ptr - nbBytes < bitD->start) { - nbBytes = (U32)(bitD->ptr - bitD->start); /* ptr > start */ - result = BIT_DStream_endOfBuffer; - } - bitD->ptr -= nbBytes; - bitD->bitsConsumed -= nbBytes * 8; - bitD->bitContainer = ZSTD_readLEST(bitD->ptr); /* reminder : srcSize > sizeof(bitD) */ - return result; - } -} - -/*! BIT_endOfDStream() : -* @return Tells if DStream has exactly reached its end (all bits consumed). -*/ -ZSTD_STATIC unsigned BIT_endOfDStream(const BIT_DStream_t *DStream) -{ - return ((DStream->ptr == DStream->start) && (DStream->bitsConsumed == sizeof(DStream->bitContainer) * 8)); -} - -#endif /* BITSTREAM_H_MODULE */ diff --git a/lib/zstd/common/bitstream.h b/lib/zstd/common/bitstream.h new file mode 100644 index 000000000000..28248abe8612 --- /dev/null +++ b/lib/zstd/common/bitstream.h @@ -0,0 +1,437 @@ +/* ****************************************************************** + * bitstream + * Part of FSE library + * Copyright (c) Yann Collet, Facebook, Inc. + * + * You can contact the author at : + * - Source repository : https://github.com/Cyan4973/FiniteStateEntropy + * + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). + * You may select, at your option, one of the above-listed licenses. +****************************************************************** */ +#ifndef BITSTREAM_H_MODULE +#define BITSTREAM_H_MODULE + +/* +* This API consists of small unitary functions, which must be inlined for best performance. +* Since link-time-optimization is not available for all compilers, +* these functions are defined into a .h to be included. +*/ + +/*-**************************************** +* Dependencies +******************************************/ +#include "mem.h" /* unaligned access routines */ +#include "compiler.h" /* UNLIKELY() */ +#include "debug.h" /* assert(), DEBUGLOG(), RAWLOG() */ +#include "error_private.h" /* error codes and messages */ + + +/*========================================= +* Target specific +=========================================*/ + +#define STREAM_ACCUMULATOR_MIN_32 25 +#define STREAM_ACCUMULATOR_MIN_64 57 +#define STREAM_ACCUMULATOR_MIN ((U32)(MEM_32bits() ? STREAM_ACCUMULATOR_MIN_32 : STREAM_ACCUMULATOR_MIN_64)) + + +/*-****************************************** +* bitStream encoding API (write forward) +********************************************/ +/* bitStream can mix input from multiple sources. + * A critical property of these streams is that they encode and decode in **reverse** direction. + * So the first bit sequence you add will be the last to be read, like a LIFO stack. + */ +typedef struct { + size_t bitContainer; + unsigned bitPos; + char* startPtr; + char* ptr; + char* endPtr; +} BIT_CStream_t; + +MEM_STATIC size_t BIT_initCStream(BIT_CStream_t* bitC, void* dstBuffer, size_t dstCapacity); +MEM_STATIC void BIT_addBits(BIT_CStream_t* bitC, size_t value, unsigned nbBits); +MEM_STATIC void BIT_flushBits(BIT_CStream_t* bitC); +MEM_STATIC size_t BIT_closeCStream(BIT_CStream_t* bitC); + +/* Start with initCStream, providing the size of buffer to write into. +* bitStream will never write outside of this buffer. +* `dstCapacity` must be >= sizeof(bitD->bitContainer), otherwise @return will be an error code. +* +* bits are first added to a local register. +* Local register is size_t, hence 64-bits on 64-bits systems, or 32-bits on 32-bits systems. +* Writing data into memory is an explicit operation, performed by the flushBits function. +* Hence keep track how many bits are potentially stored into local register to avoid register overflow. +* After a flushBits, a maximum of 7 bits might still be stored into local register. +* +* Avoid storing elements of more than 24 bits if you want compatibility with 32-bits bitstream readers. +* +* Last operation is to close the bitStream. +* The function returns the final size of CStream in bytes. +* If data couldn't fit into `dstBuffer`, it will return a 0 ( == not storable) +*/ + + +/*-******************************************** +* bitStream decoding API (read backward) +**********************************************/ +typedef struct { + size_t bitContainer; + unsigned bitsConsumed; + const char* ptr; + const char* start; + const char* limitPtr; +} BIT_DStream_t; + +typedef enum { BIT_DStream_unfinished = 0, + BIT_DStream_endOfBuffer = 1, + BIT_DStream_completed = 2, + BIT_DStream_overflow = 3 } BIT_DStream_status; /* result of BIT_reloadDStream() */ + /* 1,2,4,8 would be better for bitmap combinations, but slows down performance a bit ... :( */ + +MEM_STATIC size_t BIT_initDStream(BIT_DStream_t* bitD, const void* srcBuffer, size_t srcSize); +MEM_STATIC size_t BIT_readBits(BIT_DStream_t* bitD, unsigned nbBits); +MEM_STATIC BIT_DStream_status BIT_reloadDStream(BIT_DStream_t* bitD); +MEM_STATIC unsigned BIT_endOfDStream(const BIT_DStream_t* bitD); + + +/* Start by invoking BIT_initDStream(). +* A chunk of the bitStream is then stored into a local register. +* Local register size is 64-bits on 64-bits systems, 32-bits on 32-bits systems (size_t). +* You can then retrieve bitFields stored into the local register, **in reverse order**. +* Local register is explicitly reloaded from memory by the BIT_reloadDStream() method. +* A reload guarantee a minimum of ((8*sizeof(bitD->bitContainer))-7) bits when its result is BIT_DStream_unfinished. +* Otherwise, it can be less than that, so proceed accordingly. +* Checking if DStream has reached its end can be performed with BIT_endOfDStream(). +*/ + + +/*-**************************************** +* unsafe API +******************************************/ +MEM_STATIC void BIT_addBitsFast(BIT_CStream_t* bitC, size_t value, unsigned nbBits); +/* faster, but works only if value is "clean", meaning all high bits above nbBits are 0 */ + +MEM_STATIC void BIT_flushBitsFast(BIT_CStream_t* bitC); +/* unsafe version; does not check buffer overflow */ + +MEM_STATIC size_t BIT_readBitsFast(BIT_DStream_t* bitD, unsigned nbBits); +/* faster, but works only if nbBits >= 1 */ + + + +/*-************************************************************** +* Internal functions +****************************************************************/ +MEM_STATIC unsigned BIT_highbit32 (U32 val) +{ + assert(val != 0); + { +# if (__GNUC__ >= 3) /* Use GCC Intrinsic */ + return __builtin_clz (val) ^ 31; +# else /* Software version */ + static const unsigned DeBruijnClz[32] = { 0, 9, 1, 10, 13, 21, 2, 29, + 11, 14, 16, 18, 22, 25, 3, 30, + 8, 12, 20, 28, 15, 17, 24, 7, + 19, 27, 23, 6, 26, 5, 4, 31 }; + U32 v = val; + v |= v >> 1; + v |= v >> 2; + v |= v >> 4; + v |= v >> 8; + v |= v >> 16; + return DeBruijnClz[ (U32) (v * 0x07C4ACDDU) >> 27]; +# endif + } +} + +/*===== Local Constants =====*/ +static const unsigned BIT_mask[] = { + 0, 1, 3, 7, 0xF, 0x1F, + 0x3F, 0x7F, 0xFF, 0x1FF, 0x3FF, 0x7FF, + 0xFFF, 0x1FFF, 0x3FFF, 0x7FFF, 0xFFFF, 0x1FFFF, + 0x3FFFF, 0x7FFFF, 0xFFFFF, 0x1FFFFF, 0x3FFFFF, 0x7FFFFF, + 0xFFFFFF, 0x1FFFFFF, 0x3FFFFFF, 0x7FFFFFF, 0xFFFFFFF, 0x1FFFFFFF, + 0x3FFFFFFF, 0x7FFFFFFF}; /* up to 31 bits */ +#define BIT_MASK_SIZE (sizeof(BIT_mask) / sizeof(BIT_mask[0])) + +/*-************************************************************** +* bitStream encoding +****************************************************************/ +/*! BIT_initCStream() : + * `dstCapacity` must be > sizeof(size_t) + * @return : 0 if success, + * otherwise an error code (can be tested using ERR_isError()) */ +MEM_STATIC size_t BIT_initCStream(BIT_CStream_t* bitC, + void* startPtr, size_t dstCapacity) +{ + bitC->bitContainer = 0; + bitC->bitPos = 0; + bitC->startPtr = (char*)startPtr; + bitC->ptr = bitC->startPtr; + bitC->endPtr = bitC->startPtr + dstCapacity - sizeof(bitC->bitContainer); + if (dstCapacity <= sizeof(bitC->bitContainer)) return ERROR(dstSize_tooSmall); + return 0; +} + +/*! BIT_addBits() : + * can add up to 31 bits into `bitC`. + * Note : does not check for register overflow ! */ +MEM_STATIC void BIT_addBits(BIT_CStream_t* bitC, + size_t value, unsigned nbBits) +{ + DEBUG_STATIC_ASSERT(BIT_MASK_SIZE == 32); + assert(nbBits < BIT_MASK_SIZE); + assert(nbBits + bitC->bitPos < sizeof(bitC->bitContainer) * 8); + bitC->bitContainer |= (value & BIT_mask[nbBits]) << bitC->bitPos; + bitC->bitPos += nbBits; +} + +/*! BIT_addBitsFast() : + * works only if `value` is _clean_, + * meaning all high bits above nbBits are 0 */ +MEM_STATIC void BIT_addBitsFast(BIT_CStream_t* bitC, + size_t value, unsigned nbBits) +{ + assert((value>>nbBits) == 0); + assert(nbBits + bitC->bitPos < sizeof(bitC->bitContainer) * 8); + bitC->bitContainer |= value << bitC->bitPos; + bitC->bitPos += nbBits; +} + +/*! BIT_flushBitsFast() : + * assumption : bitContainer has not overflowed + * unsafe version; does not check buffer overflow */ +MEM_STATIC void BIT_flushBitsFast(BIT_CStream_t* bitC) +{ + size_t const nbBytes = bitC->bitPos >> 3; + assert(bitC->bitPos < sizeof(bitC->bitContainer) * 8); + assert(bitC->ptr <= bitC->endPtr); + MEM_writeLEST(bitC->ptr, bitC->bitContainer); + bitC->ptr += nbBytes; + bitC->bitPos &= 7; + bitC->bitContainer >>= nbBytes*8; +} + +/*! BIT_flushBits() : + * assumption : bitContainer has not overflowed + * safe version; check for buffer overflow, and prevents it. + * note : does not signal buffer overflow. + * overflow will be revealed later on using BIT_closeCStream() */ +MEM_STATIC void BIT_flushBits(BIT_CStream_t* bitC) +{ + size_t const nbBytes = bitC->bitPos >> 3; + assert(bitC->bitPos < sizeof(bitC->bitContainer) * 8); + assert(bitC->ptr <= bitC->endPtr); + MEM_writeLEST(bitC->ptr, bitC->bitContainer); + bitC->ptr += nbBytes; + if (bitC->ptr > bitC->endPtr) bitC->ptr = bitC->endPtr; + bitC->bitPos &= 7; + bitC->bitContainer >>= nbBytes*8; +} + +/*! BIT_closeCStream() : + * @return : size of CStream, in bytes, + * or 0 if it could not fit into dstBuffer */ +MEM_STATIC size_t BIT_closeCStream(BIT_CStream_t* bitC) +{ + BIT_addBitsFast(bitC, 1, 1); /* endMark */ + BIT_flushBits(bitC); + if (bitC->ptr >= bitC->endPtr) return 0; /* overflow detected */ + return (bitC->ptr - bitC->startPtr) + (bitC->bitPos > 0); +} + + +/*-******************************************************** +* bitStream decoding +**********************************************************/ +/*! BIT_initDStream() : + * Initialize a BIT_DStream_t. + * `bitD` : a pointer to an already allocated BIT_DStream_t structure. + * `srcSize` must be the *exact* size of the bitStream, in bytes. + * @return : size of stream (== srcSize), or an errorCode if a problem is detected + */ +MEM_STATIC size_t BIT_initDStream(BIT_DStream_t* bitD, const void* srcBuffer, size_t srcSize) +{ + if (srcSize < 1) { ZSTD_memset(bitD, 0, sizeof(*bitD)); return ERROR(srcSize_wrong); } + + bitD->start = (const char*)srcBuffer; + bitD->limitPtr = bitD->start + sizeof(bitD->bitContainer); + + if (srcSize >= sizeof(bitD->bitContainer)) { /* normal case */ + bitD->ptr = (const char*)srcBuffer + srcSize - sizeof(bitD->bitContainer); + bitD->bitContainer = MEM_readLEST(bitD->ptr); + { BYTE const lastByte = ((const BYTE*)srcBuffer)[srcSize-1]; + bitD->bitsConsumed = lastByte ? 8 - BIT_highbit32(lastByte) : 0; /* ensures bitsConsumed is always set */ + if (lastByte == 0) return ERROR(GENERIC); /* endMark not present */ } + } else { + bitD->ptr = bitD->start; + bitD->bitContainer = *(const BYTE*)(bitD->start); + switch(srcSize) + { + case 7: bitD->bitContainer += (size_t)(((const BYTE*)(srcBuffer))[6]) << (sizeof(bitD->bitContainer)*8 - 16); + ZSTD_FALLTHROUGH; + + case 6: bitD->bitContainer += (size_t)(((const BYTE*)(srcBuffer))[5]) << (sizeof(bitD->bitContainer)*8 - 24); + ZSTD_FALLTHROUGH; + + case 5: bitD->bitContainer += (size_t)(((const BYTE*)(srcBuffer))[4]) << (sizeof(bitD->bitContainer)*8 - 32); + ZSTD_FALLTHROUGH; + + case 4: bitD->bitContainer += (size_t)(((const BYTE*)(srcBuffer))[3]) << 24; + ZSTD_FALLTHROUGH; + + case 3: bitD->bitContainer += (size_t)(((const BYTE*)(srcBuffer))[2]) << 16; + ZSTD_FALLTHROUGH; + + case 2: bitD->bitContainer += (size_t)(((const BYTE*)(srcBuffer))[1]) << 8; + ZSTD_FALLTHROUGH; + + default: break; + } + { BYTE const lastByte = ((const BYTE*)srcBuffer)[srcSize-1]; + bitD->bitsConsumed = lastByte ? 8 - BIT_highbit32(lastByte) : 0; + if (lastByte == 0) return ERROR(corruption_detected); /* endMark not present */ + } + bitD->bitsConsumed += (U32)(sizeof(bitD->bitContainer) - srcSize)*8; + } + + return srcSize; +} + +MEM_STATIC FORCE_INLINE_ATTR size_t BIT_getUpperBits(size_t bitContainer, U32 const start) +{ + return bitContainer >> start; +} + +MEM_STATIC FORCE_INLINE_ATTR size_t BIT_getMiddleBits(size_t bitContainer, U32 const start, U32 const nbBits) +{ + U32 const regMask = sizeof(bitContainer)*8 - 1; + /* if start > regMask, bitstream is corrupted, and result is undefined */ + assert(nbBits < BIT_MASK_SIZE); + return (bitContainer >> (start & regMask)) & BIT_mask[nbBits]; +} + +MEM_STATIC FORCE_INLINE_ATTR size_t BIT_getLowerBits(size_t bitContainer, U32 const nbBits) +{ + assert(nbBits < BIT_MASK_SIZE); + return bitContainer & BIT_mask[nbBits]; +} + +/*! BIT_lookBits() : + * Provides next n bits from local register. + * local register is not modified. + * On 32-bits, maxNbBits==24. + * On 64-bits, maxNbBits==56. + * @return : value extracted */ +MEM_STATIC FORCE_INLINE_ATTR size_t BIT_lookBits(const BIT_DStream_t* bitD, U32 nbBits) +{ + /* arbitrate between double-shift and shift+mask */ +#if 1 + /* if bitD->bitsConsumed + nbBits > sizeof(bitD->bitContainer)*8, + * bitstream is likely corrupted, and result is undefined */ + return BIT_getMiddleBits(bitD->bitContainer, (sizeof(bitD->bitContainer)*8) - bitD->bitsConsumed - nbBits, nbBits); +#else + /* this code path is slower on my os-x laptop */ + U32 const regMask = sizeof(bitD->bitContainer)*8 - 1; + return ((bitD->bitContainer << (bitD->bitsConsumed & regMask)) >> 1) >> ((regMask-nbBits) & regMask); +#endif +} + +/*! BIT_lookBitsFast() : + * unsafe version; only works if nbBits >= 1 */ +MEM_STATIC size_t BIT_lookBitsFast(const BIT_DStream_t* bitD, U32 nbBits) +{ + U32 const regMask = sizeof(bitD->bitContainer)*8 - 1; + assert(nbBits >= 1); + return (bitD->bitContainer << (bitD->bitsConsumed & regMask)) >> (((regMask+1)-nbBits) & regMask); +} + +MEM_STATIC FORCE_INLINE_ATTR void BIT_skipBits(BIT_DStream_t* bitD, U32 nbBits) +{ + bitD->bitsConsumed += nbBits; +} + +/*! BIT_readBits() : + * Read (consume) next n bits from local register and update. + * Pay attention to not read more than nbBits contained into local register. + * @return : extracted value. */ +MEM_STATIC FORCE_INLINE_ATTR size_t BIT_readBits(BIT_DStream_t* bitD, unsigned nbBits) +{ + size_t const value = BIT_lookBits(bitD, nbBits); + BIT_skipBits(bitD, nbBits); + return value; +} + +/*! BIT_readBitsFast() : + * unsafe version; only works only if nbBits >= 1 */ +MEM_STATIC size_t BIT_readBitsFast(BIT_DStream_t* bitD, unsigned nbBits) +{ + size_t const value = BIT_lookBitsFast(bitD, nbBits); + assert(nbBits >= 1); + BIT_skipBits(bitD, nbBits); + return value; +} + +/*! BIT_reloadDStreamFast() : + * Similar to BIT_reloadDStream(), but with two differences: + * 1. bitsConsumed <= sizeof(bitD->bitContainer)*8 must hold! + * 2. Returns BIT_DStream_overflow when bitD->ptr < bitD->limitPtr, at this + * point you must use BIT_reloadDStream() to reload. + */ +MEM_STATIC BIT_DStream_status BIT_reloadDStreamFast(BIT_DStream_t* bitD) +{ + if (UNLIKELY(bitD->ptr < bitD->limitPtr)) + return BIT_DStream_overflow; + assert(bitD->bitsConsumed <= sizeof(bitD->bitContainer)*8); + bitD->ptr -= bitD->bitsConsumed >> 3; + bitD->bitsConsumed &= 7; + bitD->bitContainer = MEM_readLEST(bitD->ptr); + return BIT_DStream_unfinished; +} + +/*! BIT_reloadDStream() : + * Refill `bitD` from buffer previously set in BIT_initDStream() . + * This function is safe, it guarantees it will not read beyond src buffer. + * @return : status of `BIT_DStream_t` internal register. + * when status == BIT_DStream_unfinished, internal register is filled with at least 25 or 57 bits */ +MEM_STATIC BIT_DStream_status BIT_reloadDStream(BIT_DStream_t* bitD) +{ + if (bitD->bitsConsumed > (sizeof(bitD->bitContainer)*8)) /* overflow detected, like end of stream */ + return BIT_DStream_overflow; + + if (bitD->ptr >= bitD->limitPtr) { + return BIT_reloadDStreamFast(bitD); + } + if (bitD->ptr == bitD->start) { + if (bitD->bitsConsumed < sizeof(bitD->bitContainer)*8) return BIT_DStream_endOfBuffer; + return BIT_DStream_completed; + } + /* start < ptr < limitPtr */ + { U32 nbBytes = bitD->bitsConsumed >> 3; + BIT_DStream_status result = BIT_DStream_unfinished; + if (bitD->ptr - nbBytes < bitD->start) { + nbBytes = (U32)(bitD->ptr - bitD->start); /* ptr > start */ + result = BIT_DStream_endOfBuffer; + } + bitD->ptr -= nbBytes; + bitD->bitsConsumed -= nbBytes*8; + bitD->bitContainer = MEM_readLEST(bitD->ptr); /* reminder : srcSize > sizeof(bitD->bitContainer), otherwise bitD->ptr == bitD->start */ + return result; + } +} + +/*! BIT_endOfDStream() : + * @return : 1 if DStream has _exactly_ reached its end (all bits consumed). + */ +MEM_STATIC unsigned BIT_endOfDStream(const BIT_DStream_t* DStream) +{ + return ((DStream->ptr == DStream->start) && (DStream->bitsConsumed == sizeof(DStream->bitContainer)*8)); +} + + +#endif /* BITSTREAM_H_MODULE */ diff --git a/lib/zstd/common/compiler.h b/lib/zstd/common/compiler.h new file mode 100644 index 000000000000..a1a051e4bce6 --- /dev/null +++ b/lib/zstd/common/compiler.h @@ -0,0 +1,170 @@ +/* + * Copyright (c) Yann Collet, Facebook, Inc. + * All rights reserved. + * + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). + * You may select, at your option, one of the above-listed licenses. + */ + +#ifndef ZSTD_COMPILER_H +#define ZSTD_COMPILER_H + +/*-******************************************************* +* Compiler specifics +*********************************************************/ +/* force inlining */ + +#if (defined(__GNUC__) && !defined(__STRICT_ANSI__)) || defined(__cplusplus) || defined(__STDC_VERSION__) && __STDC_VERSION__ >= 199901L /* C99 */ +# define INLINE_KEYWORD inline +#else +# define INLINE_KEYWORD +#endif + +#define FORCE_INLINE_ATTR __attribute__((always_inline)) + + +/* + On MSVC qsort requires that functions passed into it use the __cdecl calling conversion(CC). + This explictly marks such functions as __cdecl so that the code will still compile + if a CC other than __cdecl has been made the default. +*/ +#define WIN_CDECL + +/* + * FORCE_INLINE_TEMPLATE is used to define C "templates", which take constant + * parameters. They must be inlined for the compiler to eliminate the constant + * branches. + */ +#define FORCE_INLINE_TEMPLATE static INLINE_KEYWORD FORCE_INLINE_ATTR +/* + * HINT_INLINE is used to help the compiler generate better code. It is *not* + * used for "templates", so it can be tweaked based on the compilers + * performance. + * + * gcc-4.8 and gcc-4.9 have been shown to benefit from leaving off the + * always_inline attribute. + * + * clang up to 5.0.0 (trunk) benefit tremendously from the always_inline + * attribute. + */ +#if !defined(__clang__) && defined(__GNUC__) && __GNUC__ >= 4 && __GNUC_MINOR__ >= 8 && __GNUC__ < 5 +# define HINT_INLINE static INLINE_KEYWORD +#else +# define HINT_INLINE static INLINE_KEYWORD FORCE_INLINE_ATTR +#endif + +/* UNUSED_ATTR tells the compiler it is okay if the function is unused. */ +#define UNUSED_ATTR __attribute__((unused)) + +/* force no inlining */ +#define FORCE_NOINLINE static __attribute__((__noinline__)) + + +/* target attribute */ +#ifndef __has_attribute + #define __has_attribute(x) 0 /* Compatibility with non-clang compilers. */ +#endif +#define TARGET_ATTRIBUTE(target) __attribute__((__target__(target))) + +/* Enable runtime BMI2 dispatch based on the CPU. + * Enabled for clang & gcc >=4.8 on x86 when BMI2 isn't enabled by default. + */ +#ifndef DYNAMIC_BMI2 + #if ((defined(__clang__) && __has_attribute(__target__)) \ + || (defined(__GNUC__) \ + && (__GNUC__ >= 5 || (__GNUC__ == 4 && __GNUC_MINOR__ >= 8)))) \ + && (defined(__x86_64__) || defined(_M_X86)) \ + && !defined(__BMI2__) + # define DYNAMIC_BMI2 1 + #else + # define DYNAMIC_BMI2 0 + #endif +#endif + +/* prefetch + * can be disabled, by declaring NO_PREFETCH build macro */ +#if ( (__GNUC__ >= 4) || ( (__GNUC__ == 3) && (__GNUC_MINOR__ >= 1) ) ) +# define PREFETCH_L1(ptr) __builtin_prefetch((ptr), 0 /* rw==read */, 3 /* locality */) +# define PREFETCH_L2(ptr) __builtin_prefetch((ptr), 0 /* rw==read */, 2 /* locality */) +#elif defined(__aarch64__) +# define PREFETCH_L1(ptr) __asm__ __volatile__("prfm pldl1keep, %0" ::"Q"(*(ptr))) +# define PREFETCH_L2(ptr) __asm__ __volatile__("prfm pldl2keep, %0" ::"Q"(*(ptr))) +#else +# define PREFETCH_L1(ptr) (void)(ptr) /* disabled */ +# define PREFETCH_L2(ptr) (void)(ptr) /* disabled */ +#endif /* NO_PREFETCH */ + +#define CACHELINE_SIZE 64 + +#define PREFETCH_AREA(p, s) { \ + const char* const _ptr = (const char*)(p); \ + size_t const _size = (size_t)(s); \ + size_t _pos; \ + for (_pos=0; _pos<_size; _pos+=CACHELINE_SIZE) { \ + PREFETCH_L2(_ptr + _pos); \ + } \ +} + +/* vectorization + * older GCC (pre gcc-4.3 picked as the cutoff) uses a different syntax */ +#if !defined(__INTEL_COMPILER) && !defined(__clang__) && defined(__GNUC__) +# if (__GNUC__ == 4 && __GNUC_MINOR__ > 3) || (__GNUC__ >= 5) +# define DONT_VECTORIZE __attribute__((optimize("no-tree-vectorize"))) +# else +# define DONT_VECTORIZE _Pragma("GCC optimize(\"no-tree-vectorize\")") +# endif +#else +# define DONT_VECTORIZE +#endif + +/* Tell the compiler that a branch is likely or unlikely. + * Only use these macros if it causes the compiler to generate better code. + * If you can remove a LIKELY/UNLIKELY annotation without speed changes in gcc + * and clang, please do. + */ +#define LIKELY(x) (__builtin_expect((x), 1)) +#define UNLIKELY(x) (__builtin_expect((x), 0)) + +/* disable warnings */ + +/*Like DYNAMIC_BMI2 but for compile time determination of BMI2 support*/ + + +/* compat. with non-clang compilers */ +#ifndef __has_builtin +# define __has_builtin(x) 0 +#endif + +/* compat. with non-clang compilers */ +#ifndef __has_feature +# define __has_feature(x) 0 +#endif + +/* C-language Attributes are added in C23. */ +#if defined(__STDC_VERSION__) && (__STDC_VERSION__ > 201710L) && defined(__has_c_attribute) +# define ZSTD_HAS_C_ATTRIBUTE(x) __has_c_attribute(x) +#else +# define ZSTD_HAS_C_ATTRIBUTE(x) 0 +#endif + +/* Only use C++ attributes in C++. Some compilers report support for C++ + * attributes when compiling with C. + */ +#define ZSTD_HAS_CPP_ATTRIBUTE(x) 0 + +/* Define ZSTD_FALLTHROUGH macro for annotating switch case with the 'fallthrough' attribute. + * - C23: https://en.cppreference.com/w/c/language/attributes/fallthrough + * - CPP17: https://en.cppreference.com/w/cpp/language/attributes/fallthrough + * - Else: __attribute__((__fallthrough__)) + */ +#define ZSTD_FALLTHROUGH fallthrough + +/* detects whether we are being compiled under msan */ + + +/* detects whether we are being compiled under asan */ + + +#endif /* ZSTD_COMPILER_H */ diff --git a/lib/zstd/common/cpu.h b/lib/zstd/common/cpu.h new file mode 100644 index 000000000000..0db7b42407ee --- /dev/null +++ b/lib/zstd/common/cpu.h @@ -0,0 +1,194 @@ +/* + * Copyright (c) Facebook, Inc. + * All rights reserved. + * + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). + * You may select, at your option, one of the above-listed licenses. + */ + +#ifndef ZSTD_COMMON_CPU_H +#define ZSTD_COMMON_CPU_H + +/* + * Implementation taken from folly/CpuId.h + * https://github.com/facebook/folly/blob/master/folly/CpuId.h + */ + +#include "mem.h" + + +typedef struct { + U32 f1c; + U32 f1d; + U32 f7b; + U32 f7c; +} ZSTD_cpuid_t; + +MEM_STATIC ZSTD_cpuid_t ZSTD_cpuid(void) { + U32 f1c = 0; + U32 f1d = 0; + U32 f7b = 0; + U32 f7c = 0; +#if defined(__i386__) && defined(__PIC__) && !defined(__clang__) && defined(__GNUC__) + /* The following block like the normal cpuid branch below, but gcc + * reserves ebx for use of its pic register so we must specially + * handle the save and restore to avoid clobbering the register + */ + U32 n; + __asm__( + "pushl %%ebx\n\t" + "cpuid\n\t" + "popl %%ebx\n\t" + : "=a"(n) + : "a"(0) + : "ecx", "edx"); + if (n >= 1) { + U32 f1a; + __asm__( + "pushl %%ebx\n\t" + "cpuid\n\t" + "popl %%ebx\n\t" + : "=a"(f1a), "=c"(f1c), "=d"(f1d) + : "a"(1)); + } + if (n >= 7) { + __asm__( + "pushl %%ebx\n\t" + "cpuid\n\t" + "movl %%ebx, %%eax\n\t" + "popl %%ebx" + : "=a"(f7b), "=c"(f7c) + : "a"(7), "c"(0) + : "edx"); + } +#elif defined(__x86_64__) || defined(_M_X64) || defined(__i386__) + U32 n; + __asm__("cpuid" : "=a"(n) : "a"(0) : "ebx", "ecx", "edx"); + if (n >= 1) { + U32 f1a; + __asm__("cpuid" : "=a"(f1a), "=c"(f1c), "=d"(f1d) : "a"(1) : "ebx"); + } + if (n >= 7) { + U32 f7a; + __asm__("cpuid" + : "=a"(f7a), "=b"(f7b), "=c"(f7c) + : "a"(7), "c"(0) + : "edx"); + } +#endif + { + ZSTD_cpuid_t cpuid; + cpuid.f1c = f1c; + cpuid.f1d = f1d; + cpuid.f7b = f7b; + cpuid.f7c = f7c; + return cpuid; + } +} + +#define X(name, r, bit) \ + MEM_STATIC int ZSTD_cpuid_##name(ZSTD_cpuid_t const cpuid) { \ + return ((cpuid.r) & (1U << bit)) != 0; \ + } + +/* cpuid(1): Processor Info and Feature Bits. */ +#define C(name, bit) X(name, f1c, bit) + C(sse3, 0) + C(pclmuldq, 1) + C(dtes64, 2) + C(monitor, 3) + C(dscpl, 4) + C(vmx, 5) + C(smx, 6) + C(eist, 7) + C(tm2, 8) + C(ssse3, 9) + C(cnxtid, 10) + C(fma, 12) + C(cx16, 13) + C(xtpr, 14) + C(pdcm, 15) + C(pcid, 17) + C(dca, 18) + C(sse41, 19) + C(sse42, 20) + C(x2apic, 21) + C(movbe, 22) + C(popcnt, 23) + C(tscdeadline, 24) + C(aes, 25) + C(xsave, 26) + C(osxsave, 27) + C(avx, 28) + C(f16c, 29) + C(rdrand, 30) +#undef C +#define D(name, bit) X(name, f1d, bit) + D(fpu, 0) + D(vme, 1) + D(de, 2) + D(pse, 3) + D(tsc, 4) + D(msr, 5) + D(pae, 6) + D(mce, 7) + D(cx8, 8) + D(apic, 9) + D(sep, 11) + D(mtrr, 12) + D(pge, 13) + D(mca, 14) + D(cmov, 15) + D(pat, 16) + D(pse36, 17) + D(psn, 18) + D(clfsh, 19) + D(ds, 21) + D(acpi, 22) + D(mmx, 23) + D(fxsr, 24) + D(sse, 25) + D(sse2, 26) + D(ss, 27) + D(htt, 28) + D(tm, 29) + D(pbe, 31) +#undef D + +/* cpuid(7): Extended Features. */ +#define B(name, bit) X(name, f7b, bit) + B(bmi1, 3) + B(hle, 4) + B(avx2, 5) + B(smep, 7) + B(bmi2, 8) + B(erms, 9) + B(invpcid, 10) + B(rtm, 11) + B(mpx, 14) + B(avx512f, 16) + B(avx512dq, 17) + B(rdseed, 18) + B(adx, 19) + B(smap, 20) + B(avx512ifma, 21) + B(pcommit, 22) + B(clflushopt, 23) + B(clwb, 24) + B(avx512pf, 26) + B(avx512er, 27) + B(avx512cd, 28) + B(sha, 29) + B(avx512bw, 30) + B(avx512vl, 31) +#undef B +#define C(name, bit) X(name, f7c, bit) + C(prefetchwt1, 0) + C(avx512vbmi, 1) +#undef C + +#undef X + +#endif /* ZSTD_COMMON_CPU_H */ diff --git a/lib/zstd/common/debug.c b/lib/zstd/common/debug.c new file mode 100644 index 000000000000..bb863c9ea616 --- /dev/null +++ b/lib/zstd/common/debug.c @@ -0,0 +1,24 @@ +/* ****************************************************************** + * debug + * Part of FSE library + * Copyright (c) Yann Collet, Facebook, Inc. + * + * You can contact the author at : + * - Source repository : https://github.com/Cyan4973/FiniteStateEntropy + * + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). + * You may select, at your option, one of the above-listed licenses. +****************************************************************** */ + + +/* + * This module only hosts one global variable + * which can be used to dynamically influence the verbosity of traces, + * such as DEBUGLOG and RAWLOG + */ + +#include "debug.h" + +int g_debuglevel = DEBUGLEVEL; diff --git a/lib/zstd/common/debug.h b/lib/zstd/common/debug.h new file mode 100644 index 000000000000..6dd88d1fbd02 --- /dev/null +++ b/lib/zstd/common/debug.h @@ -0,0 +1,101 @@ +/* ****************************************************************** + * debug + * Part of FSE library + * Copyright (c) Yann Collet, Facebook, Inc. + * + * You can contact the author at : + * - Source repository : https://github.com/Cyan4973/FiniteStateEntropy + * + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). + * You may select, at your option, one of the above-listed licenses. +****************************************************************** */ + + +/* + * The purpose of this header is to enable debug functions. + * They regroup assert(), DEBUGLOG() and RAWLOG() for run-time, + * and DEBUG_STATIC_ASSERT() for compile-time. + * + * By default, DEBUGLEVEL==0, which means run-time debug is disabled. + * + * Level 1 enables assert() only. + * Starting level 2, traces can be generated and pushed to stderr. + * The higher the level, the more verbose the traces. + * + * It's possible to dynamically adjust level using variable g_debug_level, + * which is only declared if DEBUGLEVEL>=2, + * and is a global variable, not multi-thread protected (use with care) + */ + +#ifndef DEBUG_H_12987983217 +#define DEBUG_H_12987983217 + + + +/* static assert is triggered at compile time, leaving no runtime artefact. + * static assert only works with compile-time constants. + * Also, this variant can only be used inside a function. */ +#define DEBUG_STATIC_ASSERT(c) (void)sizeof(char[(c) ? 1 : -1]) + + +/* DEBUGLEVEL is expected to be defined externally, + * typically through compiler command line. + * Value must be a number. */ +#ifndef DEBUGLEVEL +# define DEBUGLEVEL 0 +#endif + + +/* recommended values for DEBUGLEVEL : + * 0 : release mode, no debug, all run-time checks disabled + * 1 : enables assert() only, no display + * 2 : reserved, for currently active debug path + * 3 : events once per object lifetime (CCtx, CDict, etc.) + * 4 : events once per frame + * 5 : events once per block + * 6 : events once per sequence (verbose) + * 7+: events at every position (*very* verbose) + * + * It's generally inconvenient to output traces > 5. + * In which case, it's possible to selectively trigger high verbosity levels + * by modifying g_debug_level. + */ + +#if (DEBUGLEVEL>=1) +# define ZSTD_DEPS_NEED_ASSERT +# include "zstd_deps.h" +#else +# ifndef assert /* assert may be already defined, due to prior #include */ +# define assert(condition) ((void)0) /* disable assert (default) */ +# endif +#endif + +#if (DEBUGLEVEL>=2) +# define ZSTD_DEPS_NEED_IO +# include "zstd_deps.h" +extern int g_debuglevel; /* the variable is only declared, + it actually lives in debug.c, + and is shared by the whole process. + It's not thread-safe. + It's useful when enabling very verbose levels + on selective conditions (such as position in src) */ + +# define RAWLOG(l, ...) { \ + if (l<=g_debuglevel) { \ + ZSTD_DEBUG_PRINT(__VA_ARGS__); \ + } } +# define DEBUGLOG(l, ...) { \ + if (l<=g_debuglevel) { \ + ZSTD_DEBUG_PRINT(__FILE__ ": " __VA_ARGS__); \ + ZSTD_DEBUG_PRINT(" \n"); \ + } } +#else +# define RAWLOG(l, ...) {} /* disabled */ +# define DEBUGLOG(l, ...) {} /* disabled */ +#endif + + + +#endif /* DEBUG_H_12987983217 */ diff --git a/lib/zstd/common/entropy_common.c b/lib/zstd/common/entropy_common.c new file mode 100644 index 000000000000..53b47a2b52ff --- /dev/null +++ b/lib/zstd/common/entropy_common.c @@ -0,0 +1,357 @@ +/* ****************************************************************** + * Common functions of New Generation Entropy library + * Copyright (c) Yann Collet, Facebook, Inc. + * + * You can contact the author at : + * - FSE+HUF source repository : https://github.com/Cyan4973/FiniteStateEntropy + * - Public forum : https://groups.google.com/forum/#!forum/lz4c + * + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). + * You may select, at your option, one of the above-listed licenses. +****************************************************************** */ + +/* ************************************* +* Dependencies +***************************************/ +#include "mem.h" +#include "error_private.h" /* ERR_*, ERROR */ +#define FSE_STATIC_LINKING_ONLY /* FSE_MIN_TABLELOG */ +#include "fse.h" +#define HUF_STATIC_LINKING_ONLY /* HUF_TABLELOG_ABSOLUTEMAX */ +#include "huf.h" + + +/*=== Version ===*/ +unsigned FSE_versionNumber(void) { return FSE_VERSION_NUMBER; } + + +/*=== Error Management ===*/ +unsigned FSE_isError(size_t code) { return ERR_isError(code); } +const char* FSE_getErrorName(size_t code) { return ERR_getErrorName(code); } + +unsigned HUF_isError(size_t code) { return ERR_isError(code); } +const char* HUF_getErrorName(size_t code) { return ERR_getErrorName(code); } + + +/*-************************************************************** +* FSE NCount encoding-decoding +****************************************************************/ +static U32 FSE_ctz(U32 val) +{ + assert(val != 0); + { +# if (__GNUC__ >= 3) /* GCC Intrinsic */ + return __builtin_ctz(val); +# else /* Software version */ + U32 count = 0; + while ((val & 1) == 0) { + val >>= 1; + ++count; + } + return count; +# endif + } +} + +FORCE_INLINE_TEMPLATE +size_t FSE_readNCount_body(short* normalizedCounter, unsigned* maxSVPtr, unsigned* tableLogPtr, + const void* headerBuffer, size_t hbSize) +{ + const BYTE* const istart = (const BYTE*) headerBuffer; + const BYTE* const iend = istart + hbSize; + const BYTE* ip = istart; + int nbBits; + int remaining; + int threshold; + U32 bitStream; + int bitCount; + unsigned charnum = 0; + unsigned const maxSV1 = *maxSVPtr + 1; + int previous0 = 0; + + if (hbSize < 8) { + /* This function only works when hbSize >= 8 */ + char buffer[8] = {0}; + ZSTD_memcpy(buffer, headerBuffer, hbSize); + { size_t const countSize = FSE_readNCount(normalizedCounter, maxSVPtr, tableLogPtr, + buffer, sizeof(buffer)); + if (FSE_isError(countSize)) return countSize; + if (countSize > hbSize) return ERROR(corruption_detected); + return countSize; + } } + assert(hbSize >= 8); + + /* init */ + ZSTD_memset(normalizedCounter, 0, (*maxSVPtr+1) * sizeof(normalizedCounter[0])); /* all symbols not present in NCount have a frequency of 0 */ + bitStream = MEM_readLE32(ip); + nbBits = (bitStream & 0xF) + FSE_MIN_TABLELOG; /* extract tableLog */ + if (nbBits > FSE_TABLELOG_ABSOLUTE_MAX) return ERROR(tableLog_tooLarge); + bitStream >>= 4; + bitCount = 4; + *tableLogPtr = nbBits; + remaining = (1<> 1; + while (repeats >= 12) { + charnum += 3 * 12; + if (LIKELY(ip <= iend-7)) { + ip += 3; + } else { + bitCount -= (int)(8 * (iend - 7 - ip)); + bitCount &= 31; + ip = iend - 4; + } + bitStream = MEM_readLE32(ip) >> bitCount; + repeats = FSE_ctz(~bitStream | 0x80000000) >> 1; + } + charnum += 3 * repeats; + bitStream >>= 2 * repeats; + bitCount += 2 * repeats; + + /* Add the final repeat which isn't 0b11. */ + assert((bitStream & 3) < 3); + charnum += bitStream & 3; + bitCount += 2; + + /* This is an error, but break and return an error + * at the end, because returning out of a loop makes + * it harder for the compiler to optimize. + */ + if (charnum >= maxSV1) break; + + /* We don't need to set the normalized count to 0 + * because we already memset the whole buffer to 0. + */ + + if (LIKELY(ip <= iend-7) || (ip + (bitCount>>3) <= iend-4)) { + assert((bitCount >> 3) <= 3); /* For first condition to work */ + ip += bitCount>>3; + bitCount &= 7; + } else { + bitCount -= (int)(8 * (iend - 4 - ip)); + bitCount &= 31; + ip = iend - 4; + } + bitStream = MEM_readLE32(ip) >> bitCount; + } + { + int const max = (2*threshold-1) - remaining; + int count; + + if ((bitStream & (threshold-1)) < (U32)max) { + count = bitStream & (threshold-1); + bitCount += nbBits-1; + } else { + count = bitStream & (2*threshold-1); + if (count >= threshold) count -= max; + bitCount += nbBits; + } + + count--; /* extra accuracy */ + /* When it matters (small blocks), this is a + * predictable branch, because we don't use -1. + */ + if (count >= 0) { + remaining -= count; + } else { + assert(count == -1); + remaining += count; + } + normalizedCounter[charnum++] = (short)count; + previous0 = !count; + + assert(threshold > 1); + if (remaining < threshold) { + /* This branch can be folded into the + * threshold update condition because we + * know that threshold > 1. + */ + if (remaining <= 1) break; + nbBits = BIT_highbit32(remaining) + 1; + threshold = 1 << (nbBits - 1); + } + if (charnum >= maxSV1) break; + + if (LIKELY(ip <= iend-7) || (ip + (bitCount>>3) <= iend-4)) { + ip += bitCount>>3; + bitCount &= 7; + } else { + bitCount -= (int)(8 * (iend - 4 - ip)); + bitCount &= 31; + ip = iend - 4; + } + bitStream = MEM_readLE32(ip) >> bitCount; + } } + if (remaining != 1) return ERROR(corruption_detected); + /* Only possible when there are too many zeros. */ + if (charnum > maxSV1) return ERROR(maxSymbolValue_tooSmall); + if (bitCount > 32) return ERROR(corruption_detected); + *maxSVPtr = charnum-1; + + ip += (bitCount+7)>>3; + return ip-istart; +} + +/* Avoids the FORCE_INLINE of the _body() function. */ +static size_t FSE_readNCount_body_default( + short* normalizedCounter, unsigned* maxSVPtr, unsigned* tableLogPtr, + const void* headerBuffer, size_t hbSize) +{ + return FSE_readNCount_body(normalizedCounter, maxSVPtr, tableLogPtr, headerBuffer, hbSize); +} + +#if DYNAMIC_BMI2 +TARGET_ATTRIBUTE("bmi2") static size_t FSE_readNCount_body_bmi2( + short* normalizedCounter, unsigned* maxSVPtr, unsigned* tableLogPtr, + const void* headerBuffer, size_t hbSize) +{ + return FSE_readNCount_body(normalizedCounter, maxSVPtr, tableLogPtr, headerBuffer, hbSize); +} +#endif + +size_t FSE_readNCount_bmi2( + short* normalizedCounter, unsigned* maxSVPtr, unsigned* tableLogPtr, + const void* headerBuffer, size_t hbSize, int bmi2) +{ +#if DYNAMIC_BMI2 + if (bmi2) { + return FSE_readNCount_body_bmi2(normalizedCounter, maxSVPtr, tableLogPtr, headerBuffer, hbSize); + } +#endif + (void)bmi2; + return FSE_readNCount_body_default(normalizedCounter, maxSVPtr, tableLogPtr, headerBuffer, hbSize); +} + +size_t FSE_readNCount( + short* normalizedCounter, unsigned* maxSVPtr, unsigned* tableLogPtr, + const void* headerBuffer, size_t hbSize) +{ + return FSE_readNCount_bmi2(normalizedCounter, maxSVPtr, tableLogPtr, headerBuffer, hbSize, /* bmi2 */ 0); +} + + +/*! HUF_readStats() : + Read compact Huffman tree, saved by HUF_writeCTable(). + `huffWeight` is destination buffer. + `rankStats` is assumed to be a table of at least HUF_TABLELOG_MAX U32. + @return : size read from `src` , or an error Code . + Note : Needed by HUF_readCTable() and HUF_readDTableX?() . +*/ +size_t HUF_readStats(BYTE* huffWeight, size_t hwSize, U32* rankStats, + U32* nbSymbolsPtr, U32* tableLogPtr, + const void* src, size_t srcSize) +{ + U32 wksp[HUF_READ_STATS_WORKSPACE_SIZE_U32]; + return HUF_readStats_wksp(huffWeight, hwSize, rankStats, nbSymbolsPtr, tableLogPtr, src, srcSize, wksp, sizeof(wksp), /* bmi2 */ 0); +} + +FORCE_INLINE_TEMPLATE size_t +HUF_readStats_body(BYTE* huffWeight, size_t hwSize, U32* rankStats, + U32* nbSymbolsPtr, U32* tableLogPtr, + const void* src, size_t srcSize, + void* workSpace, size_t wkspSize, + int bmi2) +{ + U32 weightTotal; + const BYTE* ip = (const BYTE*) src; + size_t iSize; + size_t oSize; + + if (!srcSize) return ERROR(srcSize_wrong); + iSize = ip[0]; + /* ZSTD_memset(huffWeight, 0, hwSize); *//* is not necessary, even though some analyzer complain ... */ + + if (iSize >= 128) { /* special header */ + oSize = iSize - 127; + iSize = ((oSize+1)/2); + if (iSize+1 > srcSize) return ERROR(srcSize_wrong); + if (oSize >= hwSize) return ERROR(corruption_detected); + ip += 1; + { U32 n; + for (n=0; n> 4; + huffWeight[n+1] = ip[n/2] & 15; + } } } + else { /* header compressed with FSE (normal case) */ + if (iSize+1 > srcSize) return ERROR(srcSize_wrong); + /* max (hwSize-1) values decoded, as last one is implied */ + oSize = FSE_decompress_wksp_bmi2(huffWeight, hwSize-1, ip+1, iSize, 6, workSpace, wkspSize, bmi2); + if (FSE_isError(oSize)) return oSize; + } + + /* collect weight stats */ + ZSTD_memset(rankStats, 0, (HUF_TABLELOG_MAX + 1) * sizeof(U32)); + weightTotal = 0; + { U32 n; for (n=0; n= HUF_TABLELOG_MAX) return ERROR(corruption_detected); + rankStats[huffWeight[n]]++; + weightTotal += (1 << huffWeight[n]) >> 1; + } } + if (weightTotal == 0) return ERROR(corruption_detected); + + /* get last non-null symbol weight (implied, total must be 2^n) */ + { U32 const tableLog = BIT_highbit32(weightTotal) + 1; + if (tableLog > HUF_TABLELOG_MAX) return ERROR(corruption_detected); + *tableLogPtr = tableLog; + /* determine last weight */ + { U32 const total = 1 << tableLog; + U32 const rest = total - weightTotal; + U32 const verif = 1 << BIT_highbit32(rest); + U32 const lastWeight = BIT_highbit32(rest) + 1; + if (verif != rest) return ERROR(corruption_detected); /* last value must be a clean power of 2 */ + huffWeight[oSize] = (BYTE)lastWeight; + rankStats[lastWeight]++; + } } + + /* check tree construction validity */ + if ((rankStats[1] < 2) || (rankStats[1] & 1)) return ERROR(corruption_detected); /* by construction : at least 2 elts of rank 1, must be even */ + + /* results */ + *nbSymbolsPtr = (U32)(oSize+1); + return iSize+1; +} + +/* Avoids the FORCE_INLINE of the _body() function. */ +static size_t HUF_readStats_body_default(BYTE* huffWeight, size_t hwSize, U32* rankStats, + U32* nbSymbolsPtr, U32* tableLogPtr, + const void* src, size_t srcSize, + void* workSpace, size_t wkspSize) +{ + return HUF_readStats_body(huffWeight, hwSize, rankStats, nbSymbolsPtr, tableLogPtr, src, srcSize, workSpace, wkspSize, 0); +} + +#if DYNAMIC_BMI2 +static TARGET_ATTRIBUTE("bmi2") size_t HUF_readStats_body_bmi2(BYTE* huffWeight, size_t hwSize, U32* rankStats, + U32* nbSymbolsPtr, U32* tableLogPtr, + const void* src, size_t srcSize, + void* workSpace, size_t wkspSize) +{ + return HUF_readStats_body(huffWeight, hwSize, rankStats, nbSymbolsPtr, tableLogPtr, src, srcSize, workSpace, wkspSize, 1); +} +#endif + +size_t HUF_readStats_wksp(BYTE* huffWeight, size_t hwSize, U32* rankStats, + U32* nbSymbolsPtr, U32* tableLogPtr, + const void* src, size_t srcSize, + void* workSpace, size_t wkspSize, + int bmi2) +{ +#if DYNAMIC_BMI2 + if (bmi2) { + return HUF_readStats_body_bmi2(huffWeight, hwSize, rankStats, nbSymbolsPtr, tableLogPtr, src, srcSize, workSpace, wkspSize); + } +#endif + (void)bmi2; + return HUF_readStats_body_default(huffWeight, hwSize, rankStats, nbSymbolsPtr, tableLogPtr, src, srcSize, workSpace, wkspSize); +} diff --git a/lib/zstd/common/error_private.c b/lib/zstd/common/error_private.c new file mode 100644 index 000000000000..6d1135f8c373 --- /dev/null +++ b/lib/zstd/common/error_private.c @@ -0,0 +1,56 @@ +/* + * Copyright (c) Yann Collet, Facebook, Inc. + * All rights reserved. + * + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). + * You may select, at your option, one of the above-listed licenses. + */ + +/* The purpose of this file is to have a single list of error strings embedded in binary */ + +#include "error_private.h" + +const char* ERR_getErrorString(ERR_enum code) +{ +#ifdef ZSTD_STRIP_ERROR_STRINGS + (void)code; + return "Error strings stripped"; +#else + static const char* const notErrorCode = "Unspecified error code"; + switch( code ) + { + case PREFIX(no_error): return "No error detected"; + case PREFIX(GENERIC): return "Error (generic)"; + case PREFIX(prefix_unknown): return "Unknown frame descriptor"; + case PREFIX(version_unsupported): return "Version not supported"; + case PREFIX(frameParameter_unsupported): return "Unsupported frame parameter"; + case PREFIX(frameParameter_windowTooLarge): return "Frame requires too much memory for decoding"; + case PREFIX(corruption_detected): return "Corrupted block detected"; + case PREFIX(checksum_wrong): return "Restored data doesn't match checksum"; + case PREFIX(parameter_unsupported): return "Unsupported parameter"; + case PREFIX(parameter_outOfBound): return "Parameter is out of bound"; + case PREFIX(init_missing): return "Context should be init first"; + case PREFIX(memory_allocation): return "Allocation error : not enough memory"; + case PREFIX(workSpace_tooSmall): return "workSpace buffer is not large enough"; + case PREFIX(stage_wrong): return "Operation not authorized at current processing stage"; + case PREFIX(tableLog_tooLarge): return "tableLog requires too much memory : unsupported"; + case PREFIX(maxSymbolValue_tooLarge): return "Unsupported max Symbol Value : too large"; + case PREFIX(maxSymbolValue_tooSmall): return "Specified maxSymbolValue is too small"; + case PREFIX(dictionary_corrupted): return "Dictionary is corrupted"; + case PREFIX(dictionary_wrong): return "Dictionary mismatch"; + case PREFIX(dictionaryCreation_failed): return "Cannot create Dictionary from provided samples"; + case PREFIX(dstSize_tooSmall): return "Destination buffer is too small"; + case PREFIX(srcSize_wrong): return "Src size is incorrect"; + case PREFIX(dstBuffer_null): return "Operation on NULL destination buffer"; + /* following error codes are not stable and may be removed or changed in a future version */ + case PREFIX(frameIndex_tooLarge): return "Frame index is too large"; + case PREFIX(seekableIO): return "An I/O error occurred when reading/seeking"; + case PREFIX(dstBuffer_wrong): return "Destination buffer is wrong"; + case PREFIX(srcBuffer_wrong): return "Source buffer is wrong"; + case PREFIX(maxCode): + default: return notErrorCode; + } +#endif +} diff --git a/lib/zstd/common/error_private.h b/lib/zstd/common/error_private.h new file mode 100644 index 000000000000..d14e686adf95 --- /dev/null +++ b/lib/zstd/common/error_private.h @@ -0,0 +1,66 @@ +/* + * Copyright (c) Yann Collet, Facebook, Inc. + * All rights reserved. + * + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). + * You may select, at your option, one of the above-listed licenses. + */ + +/* Note : this module is expected to remain private, do not expose it */ + +#ifndef ERROR_H_MODULE +#define ERROR_H_MODULE + + + +/* **************************************** +* Dependencies +******************************************/ +#include "zstd_deps.h" /* size_t */ +#include /* enum list */ + + +/* **************************************** +* Compiler-specific +******************************************/ +#define ERR_STATIC static __attribute__((unused)) + + +/*-**************************************** +* Customization (error_public.h) +******************************************/ +typedef ZSTD_ErrorCode ERR_enum; +#define PREFIX(name) ZSTD_error_##name + + +/*-**************************************** +* Error codes handling +******************************************/ +#undef ERROR /* already defined on Visual Studio */ +#define ERROR(name) ZSTD_ERROR(name) +#define ZSTD_ERROR(name) ((size_t)-PREFIX(name)) + +ERR_STATIC unsigned ERR_isError(size_t code) { return (code > ERROR(maxCode)); } + +ERR_STATIC ERR_enum ERR_getErrorCode(size_t code) { if (!ERR_isError(code)) return (ERR_enum)0; return (ERR_enum) (0-code); } + +/* check and forward error code */ +#define CHECK_V_F(e, f) size_t const e = f; if (ERR_isError(e)) return e +#define CHECK_F(f) { CHECK_V_F(_var_err__, f); } + + +/*-**************************************** +* Error Strings +******************************************/ + +const char* ERR_getErrorString(ERR_enum code); /* error_private.c */ + +ERR_STATIC const char* ERR_getErrorName(size_t code) +{ + return ERR_getErrorString(ERR_getErrorCode(code)); +} + + +#endif /* ERROR_H_MODULE */ diff --git a/lib/zstd/common/fse.h b/lib/zstd/common/fse.h new file mode 100644 index 000000000000..0bb174c2c367 --- /dev/null +++ b/lib/zstd/common/fse.h @@ -0,0 +1,710 @@ +/* ****************************************************************** + * FSE : Finite State Entropy codec + * Public Prototypes declaration + * Copyright (c) Yann Collet, Facebook, Inc. + * + * You can contact the author at : + * - Source repository : https://github.com/Cyan4973/FiniteStateEntropy + * + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). + * You may select, at your option, one of the above-listed licenses. +****************************************************************** */ + + +#ifndef FSE_H +#define FSE_H + + +/*-***************************************** +* Dependencies +******************************************/ +#include "zstd_deps.h" /* size_t, ptrdiff_t */ + + +/*-***************************************** +* FSE_PUBLIC_API : control library symbols visibility +******************************************/ +#if defined(FSE_DLL_EXPORT) && (FSE_DLL_EXPORT==1) && defined(__GNUC__) && (__GNUC__ >= 4) +# define FSE_PUBLIC_API __attribute__ ((visibility ("default"))) +#elif defined(FSE_DLL_EXPORT) && (FSE_DLL_EXPORT==1) /* Visual expected */ +# define FSE_PUBLIC_API __declspec(dllexport) +#elif defined(FSE_DLL_IMPORT) && (FSE_DLL_IMPORT==1) +# define FSE_PUBLIC_API __declspec(dllimport) /* It isn't required but allows to generate better code, saving a function pointer load from the IAT and an indirect jump.*/ +#else +# define FSE_PUBLIC_API +#endif + +/*------ Version ------*/ +#define FSE_VERSION_MAJOR 0 +#define FSE_VERSION_MINOR 9 +#define FSE_VERSION_RELEASE 0 + +#define FSE_LIB_VERSION FSE_VERSION_MAJOR.FSE_VERSION_MINOR.FSE_VERSION_RELEASE +#define FSE_QUOTE(str) #str +#define FSE_EXPAND_AND_QUOTE(str) FSE_QUOTE(str) +#define FSE_VERSION_STRING FSE_EXPAND_AND_QUOTE(FSE_LIB_VERSION) + +#define FSE_VERSION_NUMBER (FSE_VERSION_MAJOR *100*100 + FSE_VERSION_MINOR *100 + FSE_VERSION_RELEASE) +FSE_PUBLIC_API unsigned FSE_versionNumber(void); /*< library version number; to be used when checking dll version */ + + +/*-**************************************** +* FSE simple functions +******************************************/ +/*! FSE_compress() : + Compress content of buffer 'src', of size 'srcSize', into destination buffer 'dst'. + 'dst' buffer must be already allocated. Compression runs faster is dstCapacity >= FSE_compressBound(srcSize). + @return : size of compressed data (<= dstCapacity). + Special values : if return == 0, srcData is not compressible => Nothing is stored within dst !!! + if return == 1, srcData is a single byte symbol * srcSize times. Use RLE compression instead. + if FSE_isError(return), compression failed (more details using FSE_getErrorName()) +*/ +FSE_PUBLIC_API size_t FSE_compress(void* dst, size_t dstCapacity, + const void* src, size_t srcSize); + +/*! FSE_decompress(): + Decompress FSE data from buffer 'cSrc', of size 'cSrcSize', + into already allocated destination buffer 'dst', of size 'dstCapacity'. + @return : size of regenerated data (<= maxDstSize), + or an error code, which can be tested using FSE_isError() . + + ** Important ** : FSE_decompress() does not decompress non-compressible nor RLE data !!! + Why ? : making this distinction requires a header. + Header management is intentionally delegated to the user layer, which can better manage special cases. +*/ +FSE_PUBLIC_API size_t FSE_decompress(void* dst, size_t dstCapacity, + const void* cSrc, size_t cSrcSize); + + +/*-***************************************** +* Tool functions +******************************************/ +FSE_PUBLIC_API size_t FSE_compressBound(size_t size); /* maximum compressed size */ + +/* Error Management */ +FSE_PUBLIC_API unsigned FSE_isError(size_t code); /* tells if a return value is an error code */ +FSE_PUBLIC_API const char* FSE_getErrorName(size_t code); /* provides error code string (useful for debugging) */ + + +/*-***************************************** +* FSE advanced functions +******************************************/ +/*! FSE_compress2() : + Same as FSE_compress(), but allows the selection of 'maxSymbolValue' and 'tableLog' + Both parameters can be defined as '0' to mean : use default value + @return : size of compressed data + Special values : if return == 0, srcData is not compressible => Nothing is stored within cSrc !!! + if return == 1, srcData is a single byte symbol * srcSize times. Use RLE compression. + if FSE_isError(return), it's an error code. +*/ +FSE_PUBLIC_API size_t FSE_compress2 (void* dst, size_t dstSize, const void* src, size_t srcSize, unsigned maxSymbolValue, unsigned tableLog); + + +/*-***************************************** +* FSE detailed API +******************************************/ +/*! +FSE_compress() does the following: +1. count symbol occurrence from source[] into table count[] (see hist.h) +2. normalize counters so that sum(count[]) == Power_of_2 (2^tableLog) +3. save normalized counters to memory buffer using writeNCount() +4. build encoding table 'CTable' from normalized counters +5. encode the data stream using encoding table 'CTable' + +FSE_decompress() does the following: +1. read normalized counters with readNCount() +2. build decoding table 'DTable' from normalized counters +3. decode the data stream using decoding table 'DTable' + +The following API allows targeting specific sub-functions for advanced tasks. +For example, it's possible to compress several blocks using the same 'CTable', +or to save and provide normalized distribution using external method. +*/ + +/* *** COMPRESSION *** */ + +/*! FSE_optimalTableLog(): + dynamically downsize 'tableLog' when conditions are met. + It saves CPU time, by using smaller tables, while preserving or even improving compression ratio. + @return : recommended tableLog (necessarily <= 'maxTableLog') */ +FSE_PUBLIC_API unsigned FSE_optimalTableLog(unsigned maxTableLog, size_t srcSize, unsigned maxSymbolValue); + +/*! FSE_normalizeCount(): + normalize counts so that sum(count[]) == Power_of_2 (2^tableLog) + 'normalizedCounter' is a table of short, of minimum size (maxSymbolValue+1). + useLowProbCount is a boolean parameter which trades off compressed size for + faster header decoding. When it is set to 1, the compressed data will be slightly + smaller. And when it is set to 0, FSE_readNCount() and FSE_buildDTable() will be + faster. If you are compressing a small amount of data (< 2 KB) then useLowProbCount=0 + is a good default, since header deserialization makes a big speed difference. + Otherwise, useLowProbCount=1 is a good default, since the speed difference is small. + @return : tableLog, + or an errorCode, which can be tested using FSE_isError() */ +FSE_PUBLIC_API size_t FSE_normalizeCount(short* normalizedCounter, unsigned tableLog, + const unsigned* count, size_t srcSize, unsigned maxSymbolValue, unsigned useLowProbCount); + +/*! FSE_NCountWriteBound(): + Provides the maximum possible size of an FSE normalized table, given 'maxSymbolValue' and 'tableLog'. + Typically useful for allocation purpose. */ +FSE_PUBLIC_API size_t FSE_NCountWriteBound(unsigned maxSymbolValue, unsigned tableLog); + +/*! FSE_writeNCount(): + Compactly save 'normalizedCounter' into 'buffer'. + @return : size of the compressed table, + or an errorCode, which can be tested using FSE_isError(). */ +FSE_PUBLIC_API size_t FSE_writeNCount (void* buffer, size_t bufferSize, + const short* normalizedCounter, + unsigned maxSymbolValue, unsigned tableLog); + +/*! Constructor and Destructor of FSE_CTable. + Note that FSE_CTable size depends on 'tableLog' and 'maxSymbolValue' */ +typedef unsigned FSE_CTable; /* don't allocate that. It's only meant to be more restrictive than void* */ +FSE_PUBLIC_API FSE_CTable* FSE_createCTable (unsigned maxSymbolValue, unsigned tableLog); +FSE_PUBLIC_API void FSE_freeCTable (FSE_CTable* ct); + +/*! FSE_buildCTable(): + Builds `ct`, which must be already allocated, using FSE_createCTable(). + @return : 0, or an errorCode, which can be tested using FSE_isError() */ +FSE_PUBLIC_API size_t FSE_buildCTable(FSE_CTable* ct, const short* normalizedCounter, unsigned maxSymbolValue, unsigned tableLog); + +/*! FSE_compress_usingCTable(): + Compress `src` using `ct` into `dst` which must be already allocated. + @return : size of compressed data (<= `dstCapacity`), + or 0 if compressed data could not fit into `dst`, + or an errorCode, which can be tested using FSE_isError() */ +FSE_PUBLIC_API size_t FSE_compress_usingCTable (void* dst, size_t dstCapacity, const void* src, size_t srcSize, const FSE_CTable* ct); + +/*! +Tutorial : +---------- +The first step is to count all symbols. FSE_count() does this job very fast. +Result will be saved into 'count', a table of unsigned int, which must be already allocated, and have 'maxSymbolValuePtr[0]+1' cells. +'src' is a table of bytes of size 'srcSize'. All values within 'src' MUST be <= maxSymbolValuePtr[0] +maxSymbolValuePtr[0] will be updated, with its real value (necessarily <= original value) +FSE_count() will return the number of occurrence of the most frequent symbol. +This can be used to know if there is a single symbol within 'src', and to quickly evaluate its compressibility. +If there is an error, the function will return an ErrorCode (which can be tested using FSE_isError()). + +The next step is to normalize the frequencies. +FSE_normalizeCount() will ensure that sum of frequencies is == 2 ^'tableLog'. +It also guarantees a minimum of 1 to any Symbol with frequency >= 1. +You can use 'tableLog'==0 to mean "use default tableLog value". +If you are unsure of which tableLog value to use, you can ask FSE_optimalTableLog(), +which will provide the optimal valid tableLog given sourceSize, maxSymbolValue, and a user-defined maximum (0 means "default"). + +The result of FSE_normalizeCount() will be saved into a table, +called 'normalizedCounter', which is a table of signed short. +'normalizedCounter' must be already allocated, and have at least 'maxSymbolValue+1' cells. +The return value is tableLog if everything proceeded as expected. +It is 0 if there is a single symbol within distribution. +If there is an error (ex: invalid tableLog value), the function will return an ErrorCode (which can be tested using FSE_isError()). + +'normalizedCounter' can be saved in a compact manner to a memory area using FSE_writeNCount(). +'buffer' must be already allocated. +For guaranteed success, buffer size must be at least FSE_headerBound(). +The result of the function is the number of bytes written into 'buffer'. +If there is an error, the function will return an ErrorCode (which can be tested using FSE_isError(); ex : buffer size too small). + +'normalizedCounter' can then be used to create the compression table 'CTable'. +The space required by 'CTable' must be already allocated, using FSE_createCTable(). +You can then use FSE_buildCTable() to fill 'CTable'. +If there is an error, both functions will return an ErrorCode (which can be tested using FSE_isError()). + +'CTable' can then be used to compress 'src', with FSE_compress_usingCTable(). +Similar to FSE_count(), the convention is that 'src' is assumed to be a table of char of size 'srcSize' +The function returns the size of compressed data (without header), necessarily <= `dstCapacity`. +If it returns '0', compressed data could not fit into 'dst'. +If there is an error, the function will return an ErrorCode (which can be tested using FSE_isError()). +*/ + + +/* *** DECOMPRESSION *** */ + +/*! FSE_readNCount(): + Read compactly saved 'normalizedCounter' from 'rBuffer'. + @return : size read from 'rBuffer', + or an errorCode, which can be tested using FSE_isError(). + maxSymbolValuePtr[0] and tableLogPtr[0] will also be updated with their respective values */ +FSE_PUBLIC_API size_t FSE_readNCount (short* normalizedCounter, + unsigned* maxSymbolValuePtr, unsigned* tableLogPtr, + const void* rBuffer, size_t rBuffSize); + +/*! FSE_readNCount_bmi2(): + * Same as FSE_readNCount() but pass bmi2=1 when your CPU supports BMI2 and 0 otherwise. + */ +FSE_PUBLIC_API size_t FSE_readNCount_bmi2(short* normalizedCounter, + unsigned* maxSymbolValuePtr, unsigned* tableLogPtr, + const void* rBuffer, size_t rBuffSize, int bmi2); + +/*! Constructor and Destructor of FSE_DTable. + Note that its size depends on 'tableLog' */ +typedef unsigned FSE_DTable; /* don't allocate that. It's just a way to be more restrictive than void* */ +FSE_PUBLIC_API FSE_DTable* FSE_createDTable(unsigned tableLog); +FSE_PUBLIC_API void FSE_freeDTable(FSE_DTable* dt); + +/*! FSE_buildDTable(): + Builds 'dt', which must be already allocated, using FSE_createDTable(). + return : 0, or an errorCode, which can be tested using FSE_isError() */ +FSE_PUBLIC_API size_t FSE_buildDTable (FSE_DTable* dt, const short* normalizedCounter, unsigned maxSymbolValue, unsigned tableLog); + +/*! FSE_decompress_usingDTable(): + Decompress compressed source `cSrc` of size `cSrcSize` using `dt` + into `dst` which must be already allocated. + @return : size of regenerated data (necessarily <= `dstCapacity`), + or an errorCode, which can be tested using FSE_isError() */ +FSE_PUBLIC_API size_t FSE_decompress_usingDTable(void* dst, size_t dstCapacity, const void* cSrc, size_t cSrcSize, const FSE_DTable* dt); + +/*! +Tutorial : +---------- +(Note : these functions only decompress FSE-compressed blocks. + If block is uncompressed, use memcpy() instead + If block is a single repeated byte, use memset() instead ) + +The first step is to obtain the normalized frequencies of symbols. +This can be performed by FSE_readNCount() if it was saved using FSE_writeNCount(). +'normalizedCounter' must be already allocated, and have at least 'maxSymbolValuePtr[0]+1' cells of signed short. +In practice, that means it's necessary to know 'maxSymbolValue' beforehand, +or size the table to handle worst case situations (typically 256). +FSE_readNCount() will provide 'tableLog' and 'maxSymbolValue'. +The result of FSE_readNCount() is the number of bytes read from 'rBuffer'. +Note that 'rBufferSize' must be at least 4 bytes, even if useful information is less than that. +If there is an error, the function will return an error code, which can be tested using FSE_isError(). + +The next step is to build the decompression tables 'FSE_DTable' from 'normalizedCounter'. +This is performed by the function FSE_buildDTable(). +The space required by 'FSE_DTable' must be already allocated using FSE_createDTable(). +If there is an error, the function will return an error code, which can be tested using FSE_isError(). + +`FSE_DTable` can then be used to decompress `cSrc`, with FSE_decompress_usingDTable(). +`cSrcSize` must be strictly correct, otherwise decompression will fail. +FSE_decompress_usingDTable() result will tell how many bytes were regenerated (<=`dstCapacity`). +If there is an error, the function will return an error code, which can be tested using FSE_isError(). (ex: dst buffer too small) +*/ + +#endif /* FSE_H */ + +#if !defined(FSE_H_FSE_STATIC_LINKING_ONLY) +#define FSE_H_FSE_STATIC_LINKING_ONLY + +/* *** Dependency *** */ +#include "bitstream.h" + + +/* ***************************************** +* Static allocation +*******************************************/ +/* FSE buffer bounds */ +#define FSE_NCOUNTBOUND 512 +#define FSE_BLOCKBOUND(size) ((size) + ((size)>>7) + 4 /* fse states */ + sizeof(size_t) /* bitContainer */) +#define FSE_COMPRESSBOUND(size) (FSE_NCOUNTBOUND + FSE_BLOCKBOUND(size)) /* Macro version, useful for static allocation */ + +/* It is possible to statically allocate FSE CTable/DTable as a table of FSE_CTable/FSE_DTable using below macros */ +#define FSE_CTABLE_SIZE_U32(maxTableLog, maxSymbolValue) (1 + (1<<((maxTableLog)-1)) + (((maxSymbolValue)+1)*2)) +#define FSE_DTABLE_SIZE_U32(maxTableLog) (1 + (1<<(maxTableLog))) + +/* or use the size to malloc() space directly. Pay attention to alignment restrictions though */ +#define FSE_CTABLE_SIZE(maxTableLog, maxSymbolValue) (FSE_CTABLE_SIZE_U32(maxTableLog, maxSymbolValue) * sizeof(FSE_CTable)) +#define FSE_DTABLE_SIZE(maxTableLog) (FSE_DTABLE_SIZE_U32(maxTableLog) * sizeof(FSE_DTable)) + + +/* ***************************************** + * FSE advanced API + ***************************************** */ + +unsigned FSE_optimalTableLog_internal(unsigned maxTableLog, size_t srcSize, unsigned maxSymbolValue, unsigned minus); +/*< same as FSE_optimalTableLog(), which used `minus==2` */ + +/* FSE_compress_wksp() : + * Same as FSE_compress2(), but using an externally allocated scratch buffer (`workSpace`). + * FSE_COMPRESS_WKSP_SIZE_U32() provides the minimum size required for `workSpace` as a table of FSE_CTable. + */ +#define FSE_COMPRESS_WKSP_SIZE_U32(maxTableLog, maxSymbolValue) ( FSE_CTABLE_SIZE_U32(maxTableLog, maxSymbolValue) + ((maxTableLog > 12) ? (1 << (maxTableLog - 2)) : 1024) ) +size_t FSE_compress_wksp (void* dst, size_t dstSize, const void* src, size_t srcSize, unsigned maxSymbolValue, unsigned tableLog, void* workSpace, size_t wkspSize); + +size_t FSE_buildCTable_raw (FSE_CTable* ct, unsigned nbBits); +/*< build a fake FSE_CTable, designed for a flat distribution, where each symbol uses nbBits */ + +size_t FSE_buildCTable_rle (FSE_CTable* ct, unsigned char symbolValue); +/*< build a fake FSE_CTable, designed to compress always the same symbolValue */ + +/* FSE_buildCTable_wksp() : + * Same as FSE_buildCTable(), but using an externally allocated scratch buffer (`workSpace`). + * `wkspSize` must be >= `FSE_BUILD_CTABLE_WORKSPACE_SIZE_U32(maxSymbolValue, tableLog)` of `unsigned`. + */ +#define FSE_BUILD_CTABLE_WORKSPACE_SIZE_U32(maxSymbolValue, tableLog) (maxSymbolValue + 2 + (1ull << (tableLog - 2))) +#define FSE_BUILD_CTABLE_WORKSPACE_SIZE(maxSymbolValue, tableLog) (sizeof(unsigned) * FSE_BUILD_CTABLE_WORKSPACE_SIZE_U32(maxSymbolValue, tableLog)) +size_t FSE_buildCTable_wksp(FSE_CTable* ct, const short* normalizedCounter, unsigned maxSymbolValue, unsigned tableLog, void* workSpace, size_t wkspSize); + +#define FSE_BUILD_DTABLE_WKSP_SIZE(maxTableLog, maxSymbolValue) (sizeof(short) * (maxSymbolValue + 1) + (1ULL << maxTableLog) + 8) +#define FSE_BUILD_DTABLE_WKSP_SIZE_U32(maxTableLog, maxSymbolValue) ((FSE_BUILD_DTABLE_WKSP_SIZE(maxTableLog, maxSymbolValue) + sizeof(unsigned) - 1) / sizeof(unsigned)) +FSE_PUBLIC_API size_t FSE_buildDTable_wksp(FSE_DTable* dt, const short* normalizedCounter, unsigned maxSymbolValue, unsigned tableLog, void* workSpace, size_t wkspSize); +/*< Same as FSE_buildDTable(), using an externally allocated `workspace` produced with `FSE_BUILD_DTABLE_WKSP_SIZE_U32(maxSymbolValue)` */ + +size_t FSE_buildDTable_raw (FSE_DTable* dt, unsigned nbBits); +/*< build a fake FSE_DTable, designed to read a flat distribution where each symbol uses nbBits */ + +size_t FSE_buildDTable_rle (FSE_DTable* dt, unsigned char symbolValue); +/*< build a fake FSE_DTable, designed to always generate the same symbolValue */ + +#define FSE_DECOMPRESS_WKSP_SIZE_U32(maxTableLog, maxSymbolValue) (FSE_DTABLE_SIZE_U32(maxTableLog) + FSE_BUILD_DTABLE_WKSP_SIZE_U32(maxTableLog, maxSymbolValue) + (FSE_MAX_SYMBOL_VALUE + 1) / 2 + 1) +#define FSE_DECOMPRESS_WKSP_SIZE(maxTableLog, maxSymbolValue) (FSE_DECOMPRESS_WKSP_SIZE_U32(maxTableLog, maxSymbolValue) * sizeof(unsigned)) +size_t FSE_decompress_wksp(void* dst, size_t dstCapacity, const void* cSrc, size_t cSrcSize, unsigned maxLog, void* workSpace, size_t wkspSize); +/*< same as FSE_decompress(), using an externally allocated `workSpace` produced with `FSE_DECOMPRESS_WKSP_SIZE_U32(maxLog, maxSymbolValue)` */ + +size_t FSE_decompress_wksp_bmi2(void* dst, size_t dstCapacity, const void* cSrc, size_t cSrcSize, unsigned maxLog, void* workSpace, size_t wkspSize, int bmi2); +/*< Same as FSE_decompress_wksp() but with dynamic BMI2 support. Pass 1 if your CPU supports BMI2 or 0 if it doesn't. */ + +typedef enum { + FSE_repeat_none, /*< Cannot use the previous table */ + FSE_repeat_check, /*< Can use the previous table but it must be checked */ + FSE_repeat_valid /*< Can use the previous table and it is assumed to be valid */ + } FSE_repeat; + +/* ***************************************** +* FSE symbol compression API +*******************************************/ +/*! + This API consists of small unitary functions, which highly benefit from being inlined. + Hence their body are included in next section. +*/ +typedef struct { + ptrdiff_t value; + const void* stateTable; + const void* symbolTT; + unsigned stateLog; +} FSE_CState_t; + +static void FSE_initCState(FSE_CState_t* CStatePtr, const FSE_CTable* ct); + +static void FSE_encodeSymbol(BIT_CStream_t* bitC, FSE_CState_t* CStatePtr, unsigned symbol); + +static void FSE_flushCState(BIT_CStream_t* bitC, const FSE_CState_t* CStatePtr); + +/*< +These functions are inner components of FSE_compress_usingCTable(). +They allow the creation of custom streams, mixing multiple tables and bit sources. + +A key property to keep in mind is that encoding and decoding are done **in reverse direction**. +So the first symbol you will encode is the last you will decode, like a LIFO stack. + +You will need a few variables to track your CStream. They are : + +FSE_CTable ct; // Provided by FSE_buildCTable() +BIT_CStream_t bitStream; // bitStream tracking structure +FSE_CState_t state; // State tracking structure (can have several) + + +The first thing to do is to init bitStream and state. + size_t errorCode = BIT_initCStream(&bitStream, dstBuffer, maxDstSize); + FSE_initCState(&state, ct); + +Note that BIT_initCStream() can produce an error code, so its result should be tested, using FSE_isError(); +You can then encode your input data, byte after byte. +FSE_encodeSymbol() outputs a maximum of 'tableLog' bits at a time. +Remember decoding will be done in reverse direction. + FSE_encodeByte(&bitStream, &state, symbol); + +At any time, you can also add any bit sequence. +Note : maximum allowed nbBits is 25, for compatibility with 32-bits decoders + BIT_addBits(&bitStream, bitField, nbBits); + +The above methods don't commit data to memory, they just store it into local register, for speed. +Local register size is 64-bits on 64-bits systems, 32-bits on 32-bits systems (size_t). +Writing data to memory is a manual operation, performed by the flushBits function. + BIT_flushBits(&bitStream); + +Your last FSE encoding operation shall be to flush your last state value(s). + FSE_flushState(&bitStream, &state); + +Finally, you must close the bitStream. +The function returns the size of CStream in bytes. +If data couldn't fit into dstBuffer, it will return a 0 ( == not compressible) +If there is an error, it returns an errorCode (which can be tested using FSE_isError()). + size_t size = BIT_closeCStream(&bitStream); +*/ + + +/* ***************************************** +* FSE symbol decompression API +*******************************************/ +typedef struct { + size_t state; + const void* table; /* precise table may vary, depending on U16 */ +} FSE_DState_t; + + +static void FSE_initDState(FSE_DState_t* DStatePtr, BIT_DStream_t* bitD, const FSE_DTable* dt); + +static unsigned char FSE_decodeSymbol(FSE_DState_t* DStatePtr, BIT_DStream_t* bitD); + +static unsigned FSE_endOfDState(const FSE_DState_t* DStatePtr); + +/*< +Let's now decompose FSE_decompress_usingDTable() into its unitary components. +You will decode FSE-encoded symbols from the bitStream, +and also any other bitFields you put in, **in reverse order**. + +You will need a few variables to track your bitStream. They are : + +BIT_DStream_t DStream; // Stream context +FSE_DState_t DState; // State context. Multiple ones are possible +FSE_DTable* DTablePtr; // Decoding table, provided by FSE_buildDTable() + +The first thing to do is to init the bitStream. + errorCode = BIT_initDStream(&DStream, srcBuffer, srcSize); + +You should then retrieve your initial state(s) +(in reverse flushing order if you have several ones) : + errorCode = FSE_initDState(&DState, &DStream, DTablePtr); + +You can then decode your data, symbol after symbol. +For information the maximum number of bits read by FSE_decodeSymbol() is 'tableLog'. +Keep in mind that symbols are decoded in reverse order, like a LIFO stack (last in, first out). + unsigned char symbol = FSE_decodeSymbol(&DState, &DStream); + +You can retrieve any bitfield you eventually stored into the bitStream (in reverse order) +Note : maximum allowed nbBits is 25, for 32-bits compatibility + size_t bitField = BIT_readBits(&DStream, nbBits); + +All above operations only read from local register (which size depends on size_t). +Refueling the register from memory is manually performed by the reload method. + endSignal = FSE_reloadDStream(&DStream); + +BIT_reloadDStream() result tells if there is still some more data to read from DStream. +BIT_DStream_unfinished : there is still some data left into the DStream. +BIT_DStream_endOfBuffer : Dstream reached end of buffer. Its container may no longer be completely filled. +BIT_DStream_completed : Dstream reached its exact end, corresponding in general to decompression completed. +BIT_DStream_tooFar : Dstream went too far. Decompression result is corrupted. + +When reaching end of buffer (BIT_DStream_endOfBuffer), progress slowly, notably if you decode multiple symbols per loop, +to properly detect the exact end of stream. +After each decoded symbol, check if DStream is fully consumed using this simple test : + BIT_reloadDStream(&DStream) >= BIT_DStream_completed + +When it's done, verify decompression is fully completed, by checking both DStream and the relevant states. +Checking if DStream has reached its end is performed by : + BIT_endOfDStream(&DStream); +Check also the states. There might be some symbols left there, if some high probability ones (>50%) are possible. + FSE_endOfDState(&DState); +*/ + + +/* ***************************************** +* FSE unsafe API +*******************************************/ +static unsigned char FSE_decodeSymbolFast(FSE_DState_t* DStatePtr, BIT_DStream_t* bitD); +/* faster, but works only if nbBits is always >= 1 (otherwise, result will be corrupted) */ + + +/* ***************************************** +* Implementation of inlined functions +*******************************************/ +typedef struct { + int deltaFindState; + U32 deltaNbBits; +} FSE_symbolCompressionTransform; /* total 8 bytes */ + +MEM_STATIC void FSE_initCState(FSE_CState_t* statePtr, const FSE_CTable* ct) +{ + const void* ptr = ct; + const U16* u16ptr = (const U16*) ptr; + const U32 tableLog = MEM_read16(ptr); + statePtr->value = (ptrdiff_t)1<stateTable = u16ptr+2; + statePtr->symbolTT = ct + 1 + (tableLog ? (1<<(tableLog-1)) : 1); + statePtr->stateLog = tableLog; +} + + +/*! FSE_initCState2() : +* Same as FSE_initCState(), but the first symbol to include (which will be the last to be read) +* uses the smallest state value possible, saving the cost of this symbol */ +MEM_STATIC void FSE_initCState2(FSE_CState_t* statePtr, const FSE_CTable* ct, U32 symbol) +{ + FSE_initCState(statePtr, ct); + { const FSE_symbolCompressionTransform symbolTT = ((const FSE_symbolCompressionTransform*)(statePtr->symbolTT))[symbol]; + const U16* stateTable = (const U16*)(statePtr->stateTable); + U32 nbBitsOut = (U32)((symbolTT.deltaNbBits + (1<<15)) >> 16); + statePtr->value = (nbBitsOut << 16) - symbolTT.deltaNbBits; + statePtr->value = stateTable[(statePtr->value >> nbBitsOut) + symbolTT.deltaFindState]; + } +} + +MEM_STATIC void FSE_encodeSymbol(BIT_CStream_t* bitC, FSE_CState_t* statePtr, unsigned symbol) +{ + FSE_symbolCompressionTransform const symbolTT = ((const FSE_symbolCompressionTransform*)(statePtr->symbolTT))[symbol]; + const U16* const stateTable = (const U16*)(statePtr->stateTable); + U32 const nbBitsOut = (U32)((statePtr->value + symbolTT.deltaNbBits) >> 16); + BIT_addBits(bitC, statePtr->value, nbBitsOut); + statePtr->value = stateTable[ (statePtr->value >> nbBitsOut) + symbolTT.deltaFindState]; +} + +MEM_STATIC void FSE_flushCState(BIT_CStream_t* bitC, const FSE_CState_t* statePtr) +{ + BIT_addBits(bitC, statePtr->value, statePtr->stateLog); + BIT_flushBits(bitC); +} + + +/* FSE_getMaxNbBits() : + * Approximate maximum cost of a symbol, in bits. + * Fractional get rounded up (i.e : a symbol with a normalized frequency of 3 gives the same result as a frequency of 2) + * note 1 : assume symbolValue is valid (<= maxSymbolValue) + * note 2 : if freq[symbolValue]==0, @return a fake cost of tableLog+1 bits */ +MEM_STATIC U32 FSE_getMaxNbBits(const void* symbolTTPtr, U32 symbolValue) +{ + const FSE_symbolCompressionTransform* symbolTT = (const FSE_symbolCompressionTransform*) symbolTTPtr; + return (symbolTT[symbolValue].deltaNbBits + ((1<<16)-1)) >> 16; +} + +/* FSE_bitCost() : + * Approximate symbol cost, as fractional value, using fixed-point format (accuracyLog fractional bits) + * note 1 : assume symbolValue is valid (<= maxSymbolValue) + * note 2 : if freq[symbolValue]==0, @return a fake cost of tableLog+1 bits */ +MEM_STATIC U32 FSE_bitCost(const void* symbolTTPtr, U32 tableLog, U32 symbolValue, U32 accuracyLog) +{ + const FSE_symbolCompressionTransform* symbolTT = (const FSE_symbolCompressionTransform*) symbolTTPtr; + U32 const minNbBits = symbolTT[symbolValue].deltaNbBits >> 16; + U32 const threshold = (minNbBits+1) << 16; + assert(tableLog < 16); + assert(accuracyLog < 31-tableLog); /* ensure enough room for renormalization double shift */ + { U32 const tableSize = 1 << tableLog; + U32 const deltaFromThreshold = threshold - (symbolTT[symbolValue].deltaNbBits + tableSize); + U32 const normalizedDeltaFromThreshold = (deltaFromThreshold << accuracyLog) >> tableLog; /* linear interpolation (very approximate) */ + U32 const bitMultiplier = 1 << accuracyLog; + assert(symbolTT[symbolValue].deltaNbBits + tableSize <= threshold); + assert(normalizedDeltaFromThreshold <= bitMultiplier); + return (minNbBits+1)*bitMultiplier - normalizedDeltaFromThreshold; + } +} + + +/* ====== Decompression ====== */ + +typedef struct { + U16 tableLog; + U16 fastMode; +} FSE_DTableHeader; /* sizeof U32 */ + +typedef struct +{ + unsigned short newState; + unsigned char symbol; + unsigned char nbBits; +} FSE_decode_t; /* size == U32 */ + +MEM_STATIC void FSE_initDState(FSE_DState_t* DStatePtr, BIT_DStream_t* bitD, const FSE_DTable* dt) +{ + const void* ptr = dt; + const FSE_DTableHeader* const DTableH = (const FSE_DTableHeader*)ptr; + DStatePtr->state = BIT_readBits(bitD, DTableH->tableLog); + BIT_reloadDStream(bitD); + DStatePtr->table = dt + 1; +} + +MEM_STATIC BYTE FSE_peekSymbol(const FSE_DState_t* DStatePtr) +{ + FSE_decode_t const DInfo = ((const FSE_decode_t*)(DStatePtr->table))[DStatePtr->state]; + return DInfo.symbol; +} + +MEM_STATIC void FSE_updateState(FSE_DState_t* DStatePtr, BIT_DStream_t* bitD) +{ + FSE_decode_t const DInfo = ((const FSE_decode_t*)(DStatePtr->table))[DStatePtr->state]; + U32 const nbBits = DInfo.nbBits; + size_t const lowBits = BIT_readBits(bitD, nbBits); + DStatePtr->state = DInfo.newState + lowBits; +} + +MEM_STATIC BYTE FSE_decodeSymbol(FSE_DState_t* DStatePtr, BIT_DStream_t* bitD) +{ + FSE_decode_t const DInfo = ((const FSE_decode_t*)(DStatePtr->table))[DStatePtr->state]; + U32 const nbBits = DInfo.nbBits; + BYTE const symbol = DInfo.symbol; + size_t const lowBits = BIT_readBits(bitD, nbBits); + + DStatePtr->state = DInfo.newState + lowBits; + return symbol; +} + +/*! FSE_decodeSymbolFast() : + unsafe, only works if no symbol has a probability > 50% */ +MEM_STATIC BYTE FSE_decodeSymbolFast(FSE_DState_t* DStatePtr, BIT_DStream_t* bitD) +{ + FSE_decode_t const DInfo = ((const FSE_decode_t*)(DStatePtr->table))[DStatePtr->state]; + U32 const nbBits = DInfo.nbBits; + BYTE const symbol = DInfo.symbol; + size_t const lowBits = BIT_readBitsFast(bitD, nbBits); + + DStatePtr->state = DInfo.newState + lowBits; + return symbol; +} + +MEM_STATIC unsigned FSE_endOfDState(const FSE_DState_t* DStatePtr) +{ + return DStatePtr->state == 0; +} + + + +#ifndef FSE_COMMONDEFS_ONLY + +/* ************************************************************** +* Tuning parameters +****************************************************************/ +/*!MEMORY_USAGE : +* Memory usage formula : N->2^N Bytes (examples : 10 -> 1KB; 12 -> 4KB ; 16 -> 64KB; 20 -> 1MB; etc.) +* Increasing memory usage improves compression ratio +* Reduced memory usage can improve speed, due to cache effect +* Recommended max value is 14, for 16KB, which nicely fits into Intel x86 L1 cache */ +#ifndef FSE_MAX_MEMORY_USAGE +# define FSE_MAX_MEMORY_USAGE 14 +#endif +#ifndef FSE_DEFAULT_MEMORY_USAGE +# define FSE_DEFAULT_MEMORY_USAGE 13 +#endif +#if (FSE_DEFAULT_MEMORY_USAGE > FSE_MAX_MEMORY_USAGE) +# error "FSE_DEFAULT_MEMORY_USAGE must be <= FSE_MAX_MEMORY_USAGE" +#endif + +/*!FSE_MAX_SYMBOL_VALUE : +* Maximum symbol value authorized. +* Required for proper stack allocation */ +#ifndef FSE_MAX_SYMBOL_VALUE +# define FSE_MAX_SYMBOL_VALUE 255 +#endif + +/* ************************************************************** +* template functions type & suffix +****************************************************************/ +#define FSE_FUNCTION_TYPE BYTE +#define FSE_FUNCTION_EXTENSION +#define FSE_DECODE_TYPE FSE_decode_t + + +#endif /* !FSE_COMMONDEFS_ONLY */ + + +/* *************************************************************** +* Constants +*****************************************************************/ +#define FSE_MAX_TABLELOG (FSE_MAX_MEMORY_USAGE-2) +#define FSE_MAX_TABLESIZE (1U< FSE_TABLELOG_ABSOLUTE_MAX +# error "FSE_MAX_TABLELOG > FSE_TABLELOG_ABSOLUTE_MAX is not supported" +#endif + +#define FSE_TABLESTEP(tableSize) (((tableSize)>>1) + ((tableSize)>>3) + 3) + + +#endif /* FSE_STATIC_LINKING_ONLY */ + + diff --git a/lib/zstd/common/fse_decompress.c b/lib/zstd/common/fse_decompress.c new file mode 100644 index 000000000000..2c8bbe3e4c14 --- /dev/null +++ b/lib/zstd/common/fse_decompress.c @@ -0,0 +1,390 @@ +/* ****************************************************************** + * FSE : Finite State Entropy decoder + * Copyright (c) Yann Collet, Facebook, Inc. + * + * You can contact the author at : + * - FSE source repository : https://github.com/Cyan4973/FiniteStateEntropy + * - Public forum : https://groups.google.com/forum/#!forum/lz4c + * + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). + * You may select, at your option, one of the above-listed licenses. +****************************************************************** */ + + +/* ************************************************************** +* Includes +****************************************************************/ +#include "debug.h" /* assert */ +#include "bitstream.h" +#include "compiler.h" +#define FSE_STATIC_LINKING_ONLY +#include "fse.h" +#include "error_private.h" +#define ZSTD_DEPS_NEED_MALLOC +#include "zstd_deps.h" + + +/* ************************************************************** +* Error Management +****************************************************************/ +#define FSE_isError ERR_isError +#define FSE_STATIC_ASSERT(c) DEBUG_STATIC_ASSERT(c) /* use only *after* variable declarations */ + + +/* ************************************************************** +* Templates +****************************************************************/ +/* + designed to be included + for type-specific functions (template emulation in C) + Objective is to write these functions only once, for improved maintenance +*/ + +/* safety checks */ +#ifndef FSE_FUNCTION_EXTENSION +# error "FSE_FUNCTION_EXTENSION must be defined" +#endif +#ifndef FSE_FUNCTION_TYPE +# error "FSE_FUNCTION_TYPE must be defined" +#endif + +/* Function names */ +#define FSE_CAT(X,Y) X##Y +#define FSE_FUNCTION_NAME(X,Y) FSE_CAT(X,Y) +#define FSE_TYPE_NAME(X,Y) FSE_CAT(X,Y) + + +/* Function templates */ +FSE_DTable* FSE_createDTable (unsigned tableLog) +{ + if (tableLog > FSE_TABLELOG_ABSOLUTE_MAX) tableLog = FSE_TABLELOG_ABSOLUTE_MAX; + return (FSE_DTable*)ZSTD_malloc( FSE_DTABLE_SIZE_U32(tableLog) * sizeof (U32) ); +} + +void FSE_freeDTable (FSE_DTable* dt) +{ + ZSTD_free(dt); +} + +static size_t FSE_buildDTable_internal(FSE_DTable* dt, const short* normalizedCounter, unsigned maxSymbolValue, unsigned tableLog, void* workSpace, size_t wkspSize) +{ + void* const tdPtr = dt+1; /* because *dt is unsigned, 32-bits aligned on 32-bits */ + FSE_DECODE_TYPE* const tableDecode = (FSE_DECODE_TYPE*) (tdPtr); + U16* symbolNext = (U16*)workSpace; + BYTE* spread = (BYTE*)(symbolNext + maxSymbolValue + 1); + + U32 const maxSV1 = maxSymbolValue + 1; + U32 const tableSize = 1 << tableLog; + U32 highThreshold = tableSize-1; + + /* Sanity Checks */ + if (FSE_BUILD_DTABLE_WKSP_SIZE(tableLog, maxSymbolValue) > wkspSize) return ERROR(maxSymbolValue_tooLarge); + if (maxSymbolValue > FSE_MAX_SYMBOL_VALUE) return ERROR(maxSymbolValue_tooLarge); + if (tableLog > FSE_MAX_TABLELOG) return ERROR(tableLog_tooLarge); + + /* Init, lay down lowprob symbols */ + { FSE_DTableHeader DTableH; + DTableH.tableLog = (U16)tableLog; + DTableH.fastMode = 1; + { S16 const largeLimit= (S16)(1 << (tableLog-1)); + U32 s; + for (s=0; s= largeLimit) DTableH.fastMode=0; + symbolNext[s] = normalizedCounter[s]; + } } } + ZSTD_memcpy(dt, &DTableH, sizeof(DTableH)); + } + + /* Spread symbols */ + if (highThreshold == tableSize - 1) { + size_t const tableMask = tableSize-1; + size_t const step = FSE_TABLESTEP(tableSize); + /* First lay down the symbols in order. + * We use a uint64_t to lay down 8 bytes at a time. This reduces branch + * misses since small blocks generally have small table logs, so nearly + * all symbols have counts <= 8. We ensure we have 8 bytes at the end of + * our buffer to handle the over-write. + */ + { + U64 const add = 0x0101010101010101ull; + size_t pos = 0; + U64 sv = 0; + U32 s; + for (s=0; s highThreshold) position = (position + step) & tableMask; /* lowprob area */ + } } + if (position!=0) return ERROR(GENERIC); /* position must reach all cells once, otherwise normalizedCounter is incorrect */ + } + + /* Build Decoding table */ + { U32 u; + for (u=0; utableLog = 0; + DTableH->fastMode = 0; + + cell->newState = 0; + cell->symbol = symbolValue; + cell->nbBits = 0; + + return 0; +} + + +size_t FSE_buildDTable_raw (FSE_DTable* dt, unsigned nbBits) +{ + void* ptr = dt; + FSE_DTableHeader* const DTableH = (FSE_DTableHeader*)ptr; + void* dPtr = dt + 1; + FSE_decode_t* const dinfo = (FSE_decode_t*)dPtr; + const unsigned tableSize = 1 << nbBits; + const unsigned tableMask = tableSize - 1; + const unsigned maxSV1 = tableMask+1; + unsigned s; + + /* Sanity checks */ + if (nbBits < 1) return ERROR(GENERIC); /* min size */ + + /* Build Decoding Table */ + DTableH->tableLog = (U16)nbBits; + DTableH->fastMode = 1; + for (s=0; s sizeof(bitD.bitContainer)*8) /* This test must be static */ + BIT_reloadDStream(&bitD); + + op[1] = FSE_GETSYMBOL(&state2); + + if (FSE_MAX_TABLELOG*4+7 > sizeof(bitD.bitContainer)*8) /* This test must be static */ + { if (BIT_reloadDStream(&bitD) > BIT_DStream_unfinished) { op+=2; break; } } + + op[2] = FSE_GETSYMBOL(&state1); + + if (FSE_MAX_TABLELOG*2+7 > sizeof(bitD.bitContainer)*8) /* This test must be static */ + BIT_reloadDStream(&bitD); + + op[3] = FSE_GETSYMBOL(&state2); + } + + /* tail */ + /* note : BIT_reloadDStream(&bitD) >= FSE_DStream_partiallyFilled; Ends at exactly BIT_DStream_completed */ + while (1) { + if (op>(omax-2)) return ERROR(dstSize_tooSmall); + *op++ = FSE_GETSYMBOL(&state1); + if (BIT_reloadDStream(&bitD)==BIT_DStream_overflow) { + *op++ = FSE_GETSYMBOL(&state2); + break; + } + + if (op>(omax-2)) return ERROR(dstSize_tooSmall); + *op++ = FSE_GETSYMBOL(&state2); + if (BIT_reloadDStream(&bitD)==BIT_DStream_overflow) { + *op++ = FSE_GETSYMBOL(&state1); + break; + } } + + return op-ostart; +} + + +size_t FSE_decompress_usingDTable(void* dst, size_t originalSize, + const void* cSrc, size_t cSrcSize, + const FSE_DTable* dt) +{ + const void* ptr = dt; + const FSE_DTableHeader* DTableH = (const FSE_DTableHeader*)ptr; + const U32 fastMode = DTableH->fastMode; + + /* select fast mode (static) */ + if (fastMode) return FSE_decompress_usingDTable_generic(dst, originalSize, cSrc, cSrcSize, dt, 1); + return FSE_decompress_usingDTable_generic(dst, originalSize, cSrc, cSrcSize, dt, 0); +} + + +size_t FSE_decompress_wksp(void* dst, size_t dstCapacity, const void* cSrc, size_t cSrcSize, unsigned maxLog, void* workSpace, size_t wkspSize) +{ + return FSE_decompress_wksp_bmi2(dst, dstCapacity, cSrc, cSrcSize, maxLog, workSpace, wkspSize, /* bmi2 */ 0); +} + +typedef struct { + short ncount[FSE_MAX_SYMBOL_VALUE + 1]; + FSE_DTable dtable[1]; /* Dynamically sized */ +} FSE_DecompressWksp; + + +FORCE_INLINE_TEMPLATE size_t FSE_decompress_wksp_body( + void* dst, size_t dstCapacity, + const void* cSrc, size_t cSrcSize, + unsigned maxLog, void* workSpace, size_t wkspSize, + int bmi2) +{ + const BYTE* const istart = (const BYTE*)cSrc; + const BYTE* ip = istart; + unsigned tableLog; + unsigned maxSymbolValue = FSE_MAX_SYMBOL_VALUE; + FSE_DecompressWksp* const wksp = (FSE_DecompressWksp*)workSpace; + + DEBUG_STATIC_ASSERT((FSE_MAX_SYMBOL_VALUE + 1) % 2 == 0); + if (wkspSize < sizeof(*wksp)) return ERROR(GENERIC); + + /* normal FSE decoding mode */ + { + size_t const NCountLength = FSE_readNCount_bmi2(wksp->ncount, &maxSymbolValue, &tableLog, istart, cSrcSize, bmi2); + if (FSE_isError(NCountLength)) return NCountLength; + if (tableLog > maxLog) return ERROR(tableLog_tooLarge); + assert(NCountLength <= cSrcSize); + ip += NCountLength; + cSrcSize -= NCountLength; + } + + if (FSE_DECOMPRESS_WKSP_SIZE(tableLog, maxSymbolValue) > wkspSize) return ERROR(tableLog_tooLarge); + workSpace = wksp->dtable + FSE_DTABLE_SIZE_U32(tableLog); + wkspSize -= sizeof(*wksp) + FSE_DTABLE_SIZE(tableLog); + + CHECK_F( FSE_buildDTable_internal(wksp->dtable, wksp->ncount, maxSymbolValue, tableLog, workSpace, wkspSize) ); + + { + const void* ptr = wksp->dtable; + const FSE_DTableHeader* DTableH = (const FSE_DTableHeader*)ptr; + const U32 fastMode = DTableH->fastMode; + + /* select fast mode (static) */ + if (fastMode) return FSE_decompress_usingDTable_generic(dst, dstCapacity, ip, cSrcSize, wksp->dtable, 1); + return FSE_decompress_usingDTable_generic(dst, dstCapacity, ip, cSrcSize, wksp->dtable, 0); + } +} + +/* Avoids the FORCE_INLINE of the _body() function. */ +static size_t FSE_decompress_wksp_body_default(void* dst, size_t dstCapacity, const void* cSrc, size_t cSrcSize, unsigned maxLog, void* workSpace, size_t wkspSize) +{ + return FSE_decompress_wksp_body(dst, dstCapacity, cSrc, cSrcSize, maxLog, workSpace, wkspSize, 0); +} + +#if DYNAMIC_BMI2 +TARGET_ATTRIBUTE("bmi2") static size_t FSE_decompress_wksp_body_bmi2(void* dst, size_t dstCapacity, const void* cSrc, size_t cSrcSize, unsigned maxLog, void* workSpace, size_t wkspSize) +{ + return FSE_decompress_wksp_body(dst, dstCapacity, cSrc, cSrcSize, maxLog, workSpace, wkspSize, 1); +} +#endif + +size_t FSE_decompress_wksp_bmi2(void* dst, size_t dstCapacity, const void* cSrc, size_t cSrcSize, unsigned maxLog, void* workSpace, size_t wkspSize, int bmi2) +{ +#if DYNAMIC_BMI2 + if (bmi2) { + return FSE_decompress_wksp_body_bmi2(dst, dstCapacity, cSrc, cSrcSize, maxLog, workSpace, wkspSize); + } +#endif + (void)bmi2; + return FSE_decompress_wksp_body_default(dst, dstCapacity, cSrc, cSrcSize, maxLog, workSpace, wkspSize); +} + + +typedef FSE_DTable DTable_max_t[FSE_DTABLE_SIZE_U32(FSE_MAX_TABLELOG)]; + + + +#endif /* FSE_COMMONDEFS_ONLY */ diff --git a/lib/zstd/common/huf.h b/lib/zstd/common/huf.h new file mode 100644 index 000000000000..88c5586646aa --- /dev/null +++ b/lib/zstd/common/huf.h @@ -0,0 +1,356 @@ +/* ****************************************************************** + * huff0 huffman codec, + * part of Finite State Entropy library + * Copyright (c) Yann Collet, Facebook, Inc. + * + * You can contact the author at : + * - Source repository : https://github.com/Cyan4973/FiniteStateEntropy + * + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). + * You may select, at your option, one of the above-listed licenses. +****************************************************************** */ + + +#ifndef HUF_H_298734234 +#define HUF_H_298734234 + +/* *** Dependencies *** */ +#include "zstd_deps.h" /* size_t */ + + +/* *** library symbols visibility *** */ +/* Note : when linking with -fvisibility=hidden on gcc, or by default on Visual, + * HUF symbols remain "private" (internal symbols for library only). + * Set macro FSE_DLL_EXPORT to 1 if you want HUF symbols visible on DLL interface */ +#if defined(FSE_DLL_EXPORT) && (FSE_DLL_EXPORT==1) && defined(__GNUC__) && (__GNUC__ >= 4) +# define HUF_PUBLIC_API __attribute__ ((visibility ("default"))) +#elif defined(FSE_DLL_EXPORT) && (FSE_DLL_EXPORT==1) /* Visual expected */ +# define HUF_PUBLIC_API __declspec(dllexport) +#elif defined(FSE_DLL_IMPORT) && (FSE_DLL_IMPORT==1) +# define HUF_PUBLIC_API __declspec(dllimport) /* not required, just to generate faster code (saves a function pointer load from IAT and an indirect jump) */ +#else +# define HUF_PUBLIC_API +#endif + + +/* ========================== */ +/* *** simple functions *** */ +/* ========================== */ + +/* HUF_compress() : + * Compress content from buffer 'src', of size 'srcSize', into buffer 'dst'. + * 'dst' buffer must be already allocated. + * Compression runs faster if `dstCapacity` >= HUF_compressBound(srcSize). + * `srcSize` must be <= `HUF_BLOCKSIZE_MAX` == 128 KB. + * @return : size of compressed data (<= `dstCapacity`). + * Special values : if return == 0, srcData is not compressible => Nothing is stored within dst !!! + * if HUF_isError(return), compression failed (more details using HUF_getErrorName()) + */ +HUF_PUBLIC_API size_t HUF_compress(void* dst, size_t dstCapacity, + const void* src, size_t srcSize); + +/* HUF_decompress() : + * Decompress HUF data from buffer 'cSrc', of size 'cSrcSize', + * into already allocated buffer 'dst', of minimum size 'dstSize'. + * `originalSize` : **must** be the ***exact*** size of original (uncompressed) data. + * Note : in contrast with FSE, HUF_decompress can regenerate + * RLE (cSrcSize==1) and uncompressed (cSrcSize==dstSize) data, + * because it knows size to regenerate (originalSize). + * @return : size of regenerated data (== originalSize), + * or an error code, which can be tested using HUF_isError() + */ +HUF_PUBLIC_API size_t HUF_decompress(void* dst, size_t originalSize, + const void* cSrc, size_t cSrcSize); + + +/* *** Tool functions *** */ +#define HUF_BLOCKSIZE_MAX (128 * 1024) /*< maximum input size for a single block compressed with HUF_compress */ +HUF_PUBLIC_API size_t HUF_compressBound(size_t size); /*< maximum compressed size (worst case) */ + +/* Error Management */ +HUF_PUBLIC_API unsigned HUF_isError(size_t code); /*< tells if a return value is an error code */ +HUF_PUBLIC_API const char* HUF_getErrorName(size_t code); /*< provides error code string (useful for debugging) */ + + +/* *** Advanced function *** */ + +/* HUF_compress2() : + * Same as HUF_compress(), but offers control over `maxSymbolValue` and `tableLog`. + * `maxSymbolValue` must be <= HUF_SYMBOLVALUE_MAX . + * `tableLog` must be `<= HUF_TABLELOG_MAX` . */ +HUF_PUBLIC_API size_t HUF_compress2 (void* dst, size_t dstCapacity, + const void* src, size_t srcSize, + unsigned maxSymbolValue, unsigned tableLog); + +/* HUF_compress4X_wksp() : + * Same as HUF_compress2(), but uses externally allocated `workSpace`. + * `workspace` must have minimum alignment of 4, and be at least as large as HUF_WORKSPACE_SIZE */ +#define HUF_WORKSPACE_SIZE ((6 << 10) + 256) +#define HUF_WORKSPACE_SIZE_U32 (HUF_WORKSPACE_SIZE / sizeof(U32)) +HUF_PUBLIC_API size_t HUF_compress4X_wksp (void* dst, size_t dstCapacity, + const void* src, size_t srcSize, + unsigned maxSymbolValue, unsigned tableLog, + void* workSpace, size_t wkspSize); + +#endif /* HUF_H_298734234 */ + +/* ****************************************************************** + * WARNING !! + * The following section contains advanced and experimental definitions + * which shall never be used in the context of a dynamic library, + * because they are not guaranteed to remain stable in the future. + * Only consider them in association with static linking. + * *****************************************************************/ +#if !defined(HUF_H_HUF_STATIC_LINKING_ONLY) +#define HUF_H_HUF_STATIC_LINKING_ONLY + +/* *** Dependencies *** */ +#include "mem.h" /* U32 */ +#define FSE_STATIC_LINKING_ONLY +#include "fse.h" + + +/* *** Constants *** */ +#define HUF_TABLELOG_MAX 12 /* max runtime value of tableLog (due to static allocation); can be modified up to HUF_ABSOLUTEMAX_TABLELOG */ +#define HUF_TABLELOG_DEFAULT 11 /* default tableLog value when none specified */ +#define HUF_SYMBOLVALUE_MAX 255 + +#define HUF_TABLELOG_ABSOLUTEMAX 15 /* absolute limit of HUF_MAX_TABLELOG. Beyond that value, code does not work */ +#if (HUF_TABLELOG_MAX > HUF_TABLELOG_ABSOLUTEMAX) +# error "HUF_TABLELOG_MAX is too large !" +#endif + + +/* **************************************** +* Static allocation +******************************************/ +/* HUF buffer bounds */ +#define HUF_CTABLEBOUND 129 +#define HUF_BLOCKBOUND(size) (size + (size>>8) + 8) /* only true when incompressible is pre-filtered with fast heuristic */ +#define HUF_COMPRESSBOUND(size) (HUF_CTABLEBOUND + HUF_BLOCKBOUND(size)) /* Macro version, useful for static allocation */ + +/* static allocation of HUF's Compression Table */ +/* this is a private definition, just exposed for allocation and strict aliasing purpose. never EVER access its members directly */ +struct HUF_CElt_s { + U16 val; + BYTE nbBits; +}; /* typedef'd to HUF_CElt */ +typedef struct HUF_CElt_s HUF_CElt; /* consider it an incomplete type */ +#define HUF_CTABLE_SIZE_U32(maxSymbolValue) ((maxSymbolValue)+1) /* Use tables of U32, for proper alignment */ +#define HUF_CTABLE_SIZE(maxSymbolValue) (HUF_CTABLE_SIZE_U32(maxSymbolValue) * sizeof(U32)) +#define HUF_CREATE_STATIC_CTABLE(name, maxSymbolValue) \ + HUF_CElt name[HUF_CTABLE_SIZE_U32(maxSymbolValue)] /* no final ; */ + +/* static allocation of HUF's DTable */ +typedef U32 HUF_DTable; +#define HUF_DTABLE_SIZE(maxTableLog) (1 + (1<<(maxTableLog))) +#define HUF_CREATE_STATIC_DTABLEX1(DTable, maxTableLog) \ + HUF_DTable DTable[HUF_DTABLE_SIZE((maxTableLog)-1)] = { ((U32)((maxTableLog)-1) * 0x01000001) } +#define HUF_CREATE_STATIC_DTABLEX2(DTable, maxTableLog) \ + HUF_DTable DTable[HUF_DTABLE_SIZE(maxTableLog)] = { ((U32)(maxTableLog) * 0x01000001) } + + +/* **************************************** +* Advanced decompression functions +******************************************/ +size_t HUF_decompress4X1 (void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize); /*< single-symbol decoder */ +#ifndef HUF_FORCE_DECOMPRESS_X1 +size_t HUF_decompress4X2 (void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize); /*< double-symbols decoder */ +#endif + +size_t HUF_decompress4X_DCtx (HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize); /*< decodes RLE and uncompressed */ +size_t HUF_decompress4X_hufOnly(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize); /*< considers RLE and uncompressed as errors */ +size_t HUF_decompress4X_hufOnly_wksp(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize, void* workSpace, size_t wkspSize); /*< considers RLE and uncompressed as errors */ +size_t HUF_decompress4X1_DCtx(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize); /*< single-symbol decoder */ +size_t HUF_decompress4X1_DCtx_wksp(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize, void* workSpace, size_t wkspSize); /*< single-symbol decoder */ +#ifndef HUF_FORCE_DECOMPRESS_X1 +size_t HUF_decompress4X2_DCtx(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize); /*< double-symbols decoder */ +size_t HUF_decompress4X2_DCtx_wksp(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize, void* workSpace, size_t wkspSize); /*< double-symbols decoder */ +#endif + + +/* **************************************** + * HUF detailed API + * ****************************************/ + +/*! HUF_compress() does the following: + * 1. count symbol occurrence from source[] into table count[] using FSE_count() (exposed within "fse.h") + * 2. (optional) refine tableLog using HUF_optimalTableLog() + * 3. build Huffman table from count using HUF_buildCTable() + * 4. save Huffman table to memory buffer using HUF_writeCTable() + * 5. encode the data stream using HUF_compress4X_usingCTable() + * + * The following API allows targeting specific sub-functions for advanced tasks. + * For example, it's possible to compress several blocks using the same 'CTable', + * or to save and regenerate 'CTable' using external methods. + */ +unsigned HUF_optimalTableLog(unsigned maxTableLog, size_t srcSize, unsigned maxSymbolValue); +size_t HUF_buildCTable (HUF_CElt* CTable, const unsigned* count, unsigned maxSymbolValue, unsigned maxNbBits); /* @return : maxNbBits; CTable and count can overlap. In which case, CTable will overwrite count content */ +size_t HUF_writeCTable (void* dst, size_t maxDstSize, const HUF_CElt* CTable, unsigned maxSymbolValue, unsigned huffLog); +size_t HUF_writeCTable_wksp(void* dst, size_t maxDstSize, const HUF_CElt* CTable, unsigned maxSymbolValue, unsigned huffLog, void* workspace, size_t workspaceSize); +size_t HUF_compress4X_usingCTable(void* dst, size_t dstSize, const void* src, size_t srcSize, const HUF_CElt* CTable); +size_t HUF_estimateCompressedSize(const HUF_CElt* CTable, const unsigned* count, unsigned maxSymbolValue); +int HUF_validateCTable(const HUF_CElt* CTable, const unsigned* count, unsigned maxSymbolValue); + +typedef enum { + HUF_repeat_none, /*< Cannot use the previous table */ + HUF_repeat_check, /*< Can use the previous table but it must be checked. Note : The previous table must have been constructed by HUF_compress{1, 4}X_repeat */ + HUF_repeat_valid /*< Can use the previous table and it is assumed to be valid */ + } HUF_repeat; +/* HUF_compress4X_repeat() : + * Same as HUF_compress4X_wksp(), but considers using hufTable if *repeat != HUF_repeat_none. + * If it uses hufTable it does not modify hufTable or repeat. + * If it doesn't, it sets *repeat = HUF_repeat_none, and it sets hufTable to the table used. + * If preferRepeat then the old table will always be used if valid. */ +size_t HUF_compress4X_repeat(void* dst, size_t dstSize, + const void* src, size_t srcSize, + unsigned maxSymbolValue, unsigned tableLog, + void* workSpace, size_t wkspSize, /*< `workSpace` must be aligned on 4-bytes boundaries, `wkspSize` must be >= HUF_WORKSPACE_SIZE */ + HUF_CElt* hufTable, HUF_repeat* repeat, int preferRepeat, int bmi2); + +/* HUF_buildCTable_wksp() : + * Same as HUF_buildCTable(), but using externally allocated scratch buffer. + * `workSpace` must be aligned on 4-bytes boundaries, and its size must be >= HUF_CTABLE_WORKSPACE_SIZE. + */ +#define HUF_CTABLE_WORKSPACE_SIZE_U32 (2*HUF_SYMBOLVALUE_MAX +1 +1) +#define HUF_CTABLE_WORKSPACE_SIZE (HUF_CTABLE_WORKSPACE_SIZE_U32 * sizeof(unsigned)) +size_t HUF_buildCTable_wksp (HUF_CElt* tree, + const unsigned* count, U32 maxSymbolValue, U32 maxNbBits, + void* workSpace, size_t wkspSize); + +/*! HUF_readStats() : + * Read compact Huffman tree, saved by HUF_writeCTable(). + * `huffWeight` is destination buffer. + * @return : size read from `src` , or an error Code . + * Note : Needed by HUF_readCTable() and HUF_readDTableXn() . */ +size_t HUF_readStats(BYTE* huffWeight, size_t hwSize, + U32* rankStats, U32* nbSymbolsPtr, U32* tableLogPtr, + const void* src, size_t srcSize); + +/*! HUF_readStats_wksp() : + * Same as HUF_readStats() but takes an external workspace which must be + * 4-byte aligned and its size must be >= HUF_READ_STATS_WORKSPACE_SIZE. + * If the CPU has BMI2 support, pass bmi2=1, otherwise pass bmi2=0. + */ +#define HUF_READ_STATS_WORKSPACE_SIZE_U32 FSE_DECOMPRESS_WKSP_SIZE_U32(6, HUF_TABLELOG_MAX-1) +#define HUF_READ_STATS_WORKSPACE_SIZE (HUF_READ_STATS_WORKSPACE_SIZE_U32 * sizeof(unsigned)) +size_t HUF_readStats_wksp(BYTE* huffWeight, size_t hwSize, + U32* rankStats, U32* nbSymbolsPtr, U32* tableLogPtr, + const void* src, size_t srcSize, + void* workspace, size_t wkspSize, + int bmi2); + +/* HUF_readCTable() : + * Loading a CTable saved with HUF_writeCTable() */ +size_t HUF_readCTable (HUF_CElt* CTable, unsigned* maxSymbolValuePtr, const void* src, size_t srcSize, unsigned *hasZeroWeights); + +/* HUF_getNbBits() : + * Read nbBits from CTable symbolTable, for symbol `symbolValue` presumed <= HUF_SYMBOLVALUE_MAX + * Note 1 : is not inlined, as HUF_CElt definition is private + * Note 2 : const void* used, so that it can provide a statically allocated table as argument (which uses type U32) */ +U32 HUF_getNbBits(const void* symbolTable, U32 symbolValue); + +/* + * HUF_decompress() does the following: + * 1. select the decompression algorithm (X1, X2) based on pre-computed heuristics + * 2. build Huffman table from save, using HUF_readDTableX?() + * 3. decode 1 or 4 segments in parallel using HUF_decompress?X?_usingDTable() + */ + +/* HUF_selectDecoder() : + * Tells which decoder is likely to decode faster, + * based on a set of pre-computed metrics. + * @return : 0==HUF_decompress4X1, 1==HUF_decompress4X2 . + * Assumption : 0 < dstSize <= 128 KB */ +U32 HUF_selectDecoder (size_t dstSize, size_t cSrcSize); + +/* + * The minimum workspace size for the `workSpace` used in + * HUF_readDTableX1_wksp() and HUF_readDTableX2_wksp(). + * + * The space used depends on HUF_TABLELOG_MAX, ranging from ~1500 bytes when + * HUF_TABLE_LOG_MAX=12 to ~1850 bytes when HUF_TABLE_LOG_MAX=15. + * Buffer overflow errors may potentially occur if code modifications result in + * a required workspace size greater than that specified in the following + * macro. + */ +#define HUF_DECOMPRESS_WORKSPACE_SIZE ((2 << 10) + (1 << 9)) +#define HUF_DECOMPRESS_WORKSPACE_SIZE_U32 (HUF_DECOMPRESS_WORKSPACE_SIZE / sizeof(U32)) + +#ifndef HUF_FORCE_DECOMPRESS_X2 +size_t HUF_readDTableX1 (HUF_DTable* DTable, const void* src, size_t srcSize); +size_t HUF_readDTableX1_wksp (HUF_DTable* DTable, const void* src, size_t srcSize, void* workSpace, size_t wkspSize); +#endif +#ifndef HUF_FORCE_DECOMPRESS_X1 +size_t HUF_readDTableX2 (HUF_DTable* DTable, const void* src, size_t srcSize); +size_t HUF_readDTableX2_wksp (HUF_DTable* DTable, const void* src, size_t srcSize, void* workSpace, size_t wkspSize); +#endif + +size_t HUF_decompress4X_usingDTable(void* dst, size_t maxDstSize, const void* cSrc, size_t cSrcSize, const HUF_DTable* DTable); +#ifndef HUF_FORCE_DECOMPRESS_X2 +size_t HUF_decompress4X1_usingDTable(void* dst, size_t maxDstSize, const void* cSrc, size_t cSrcSize, const HUF_DTable* DTable); +#endif +#ifndef HUF_FORCE_DECOMPRESS_X1 +size_t HUF_decompress4X2_usingDTable(void* dst, size_t maxDstSize, const void* cSrc, size_t cSrcSize, const HUF_DTable* DTable); +#endif + + +/* ====================== */ +/* single stream variants */ +/* ====================== */ + +size_t HUF_compress1X (void* dst, size_t dstSize, const void* src, size_t srcSize, unsigned maxSymbolValue, unsigned tableLog); +size_t HUF_compress1X_wksp (void* dst, size_t dstSize, const void* src, size_t srcSize, unsigned maxSymbolValue, unsigned tableLog, void* workSpace, size_t wkspSize); /*< `workSpace` must be a table of at least HUF_WORKSPACE_SIZE_U32 unsigned */ +size_t HUF_compress1X_usingCTable(void* dst, size_t dstSize, const void* src, size_t srcSize, const HUF_CElt* CTable); +/* HUF_compress1X_repeat() : + * Same as HUF_compress1X_wksp(), but considers using hufTable if *repeat != HUF_repeat_none. + * If it uses hufTable it does not modify hufTable or repeat. + * If it doesn't, it sets *repeat = HUF_repeat_none, and it sets hufTable to the table used. + * If preferRepeat then the old table will always be used if valid. */ +size_t HUF_compress1X_repeat(void* dst, size_t dstSize, + const void* src, size_t srcSize, + unsigned maxSymbolValue, unsigned tableLog, + void* workSpace, size_t wkspSize, /*< `workSpace` must be aligned on 4-bytes boundaries, `wkspSize` must be >= HUF_WORKSPACE_SIZE */ + HUF_CElt* hufTable, HUF_repeat* repeat, int preferRepeat, int bmi2); + +size_t HUF_decompress1X1 (void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize); /* single-symbol decoder */ +#ifndef HUF_FORCE_DECOMPRESS_X1 +size_t HUF_decompress1X2 (void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize); /* double-symbol decoder */ +#endif + +size_t HUF_decompress1X_DCtx (HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize); +size_t HUF_decompress1X_DCtx_wksp (HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize, void* workSpace, size_t wkspSize); +#ifndef HUF_FORCE_DECOMPRESS_X2 +size_t HUF_decompress1X1_DCtx(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize); /*< single-symbol decoder */ +size_t HUF_decompress1X1_DCtx_wksp(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize, void* workSpace, size_t wkspSize); /*< single-symbol decoder */ +#endif +#ifndef HUF_FORCE_DECOMPRESS_X1 +size_t HUF_decompress1X2_DCtx(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize); /*< double-symbols decoder */ +size_t HUF_decompress1X2_DCtx_wksp(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize, void* workSpace, size_t wkspSize); /*< double-symbols decoder */ +#endif + +size_t HUF_decompress1X_usingDTable(void* dst, size_t maxDstSize, const void* cSrc, size_t cSrcSize, const HUF_DTable* DTable); /*< automatic selection of sing or double symbol decoder, based on DTable */ +#ifndef HUF_FORCE_DECOMPRESS_X2 +size_t HUF_decompress1X1_usingDTable(void* dst, size_t maxDstSize, const void* cSrc, size_t cSrcSize, const HUF_DTable* DTable); +#endif +#ifndef HUF_FORCE_DECOMPRESS_X1 +size_t HUF_decompress1X2_usingDTable(void* dst, size_t maxDstSize, const void* cSrc, size_t cSrcSize, const HUF_DTable* DTable); +#endif + +/* BMI2 variants. + * If the CPU has BMI2 support, pass bmi2=1, otherwise pass bmi2=0. + */ +size_t HUF_decompress1X_usingDTable_bmi2(void* dst, size_t maxDstSize, const void* cSrc, size_t cSrcSize, const HUF_DTable* DTable, int bmi2); +#ifndef HUF_FORCE_DECOMPRESS_X2 +size_t HUF_decompress1X1_DCtx_wksp_bmi2(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize, void* workSpace, size_t wkspSize, int bmi2); +#endif +size_t HUF_decompress4X_usingDTable_bmi2(void* dst, size_t maxDstSize, const void* cSrc, size_t cSrcSize, const HUF_DTable* DTable, int bmi2); +size_t HUF_decompress4X_hufOnly_wksp_bmi2(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize, void* workSpace, size_t wkspSize, int bmi2); +#ifndef HUF_FORCE_DECOMPRESS_X2 +size_t HUF_readDTableX1_wksp_bmi2(HUF_DTable* DTable, const void* src, size_t srcSize, void* workSpace, size_t wkspSize, int bmi2); +#endif + +#endif /* HUF_STATIC_LINKING_ONLY */ + diff --git a/lib/zstd/common/mem.h b/lib/zstd/common/mem.h new file mode 100644 index 000000000000..dcdd586a9fd9 --- /dev/null +++ b/lib/zstd/common/mem.h @@ -0,0 +1,259 @@ +/* SPDX-License-Identifier: GPL-2.0+ OR BSD-3-Clause */ +/* + * Copyright (c) Yann Collet, Facebook, Inc. + * All rights reserved. + * + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). + * You may select, at your option, one of the above-listed licenses. + */ + +#ifndef MEM_H_MODULE +#define MEM_H_MODULE + +/*-**************************************** +* Dependencies +******************************************/ +#include /* get_unaligned, put_unaligned* */ +#include /* inline */ +#include /* swab32, swab64 */ +#include /* size_t, ptrdiff_t */ +#include "debug.h" /* DEBUG_STATIC_ASSERT */ + +/*-**************************************** +* Compiler specifics +******************************************/ +#define MEM_STATIC static inline + +/*-************************************************************** +* Basic Types +*****************************************************************/ +typedef uint8_t BYTE; +typedef uint16_t U16; +typedef int16_t S16; +typedef uint32_t U32; +typedef int32_t S32; +typedef uint64_t U64; +typedef int64_t S64; + +/*-************************************************************** +* Memory I/O API +*****************************************************************/ +/*=== Static platform detection ===*/ +MEM_STATIC unsigned MEM_32bits(void); +MEM_STATIC unsigned MEM_64bits(void); +MEM_STATIC unsigned MEM_isLittleEndian(void); + +/*=== Native unaligned read/write ===*/ +MEM_STATIC U16 MEM_read16(const void* memPtr); +MEM_STATIC U32 MEM_read32(const void* memPtr); +MEM_STATIC U64 MEM_read64(const void* memPtr); +MEM_STATIC size_t MEM_readST(const void* memPtr); + +MEM_STATIC void MEM_write16(void* memPtr, U16 value); +MEM_STATIC void MEM_write32(void* memPtr, U32 value); +MEM_STATIC void MEM_write64(void* memPtr, U64 value); + +/*=== Little endian unaligned read/write ===*/ +MEM_STATIC U16 MEM_readLE16(const void* memPtr); +MEM_STATIC U32 MEM_readLE24(const void* memPtr); +MEM_STATIC U32 MEM_readLE32(const void* memPtr); +MEM_STATIC U64 MEM_readLE64(const void* memPtr); +MEM_STATIC size_t MEM_readLEST(const void* memPtr); + +MEM_STATIC void MEM_writeLE16(void* memPtr, U16 val); +MEM_STATIC void MEM_writeLE24(void* memPtr, U32 val); +MEM_STATIC void MEM_writeLE32(void* memPtr, U32 val32); +MEM_STATIC void MEM_writeLE64(void* memPtr, U64 val64); +MEM_STATIC void MEM_writeLEST(void* memPtr, size_t val); + +/*=== Big endian unaligned read/write ===*/ +MEM_STATIC U32 MEM_readBE32(const void* memPtr); +MEM_STATIC U64 MEM_readBE64(const void* memPtr); +MEM_STATIC size_t MEM_readBEST(const void* memPtr); + +MEM_STATIC void MEM_writeBE32(void* memPtr, U32 val32); +MEM_STATIC void MEM_writeBE64(void* memPtr, U64 val64); +MEM_STATIC void MEM_writeBEST(void* memPtr, size_t val); + +/*=== Byteswap ===*/ +MEM_STATIC U32 MEM_swap32(U32 in); +MEM_STATIC U64 MEM_swap64(U64 in); +MEM_STATIC size_t MEM_swapST(size_t in); + +/*-************************************************************** +* Memory I/O Implementation +*****************************************************************/ +MEM_STATIC unsigned MEM_32bits(void) +{ + return sizeof(size_t) == 4; +} + +MEM_STATIC unsigned MEM_64bits(void) +{ + return sizeof(size_t) == 8; +} + +#if defined(__LITTLE_ENDIAN) +#define MEM_LITTLE_ENDIAN 1 +#else +#define MEM_LITTLE_ENDIAN 0 +#endif + +MEM_STATIC unsigned MEM_isLittleEndian(void) +{ + return MEM_LITTLE_ENDIAN; +} + +MEM_STATIC U16 MEM_read16(const void *memPtr) +{ + return get_unaligned((const U16 *)memPtr); +} + +MEM_STATIC U32 MEM_read32(const void *memPtr) +{ + return get_unaligned((const U32 *)memPtr); +} + +MEM_STATIC U64 MEM_read64(const void *memPtr) +{ + return get_unaligned((const U64 *)memPtr); +} + +MEM_STATIC size_t MEM_readST(const void *memPtr) +{ + return get_unaligned((const size_t *)memPtr); +} + +MEM_STATIC void MEM_write16(void *memPtr, U16 value) +{ + put_unaligned(value, (U16 *)memPtr); +} + +MEM_STATIC void MEM_write32(void *memPtr, U32 value) +{ + put_unaligned(value, (U32 *)memPtr); +} + +MEM_STATIC void MEM_write64(void *memPtr, U64 value) +{ + put_unaligned(value, (U64 *)memPtr); +} + +/*=== Little endian r/w ===*/ + +MEM_STATIC U16 MEM_readLE16(const void *memPtr) +{ + return get_unaligned_le16(memPtr); +} + +MEM_STATIC void MEM_writeLE16(void *memPtr, U16 val) +{ + put_unaligned_le16(val, memPtr); +} + +MEM_STATIC U32 MEM_readLE24(const void *memPtr) +{ + return MEM_readLE16(memPtr) + (((const BYTE *)memPtr)[2] << 16); +} + +MEM_STATIC void MEM_writeLE24(void *memPtr, U32 val) +{ + MEM_writeLE16(memPtr, (U16)val); + ((BYTE *)memPtr)[2] = (BYTE)(val >> 16); +} + +MEM_STATIC U32 MEM_readLE32(const void *memPtr) +{ + return get_unaligned_le32(memPtr); +} + +MEM_STATIC void MEM_writeLE32(void *memPtr, U32 val32) +{ + put_unaligned_le32(val32, memPtr); +} + +MEM_STATIC U64 MEM_readLE64(const void *memPtr) +{ + return get_unaligned_le64(memPtr); +} + +MEM_STATIC void MEM_writeLE64(void *memPtr, U64 val64) +{ + put_unaligned_le64(val64, memPtr); +} + +MEM_STATIC size_t MEM_readLEST(const void *memPtr) +{ + if (MEM_32bits()) + return (size_t)MEM_readLE32(memPtr); + else + return (size_t)MEM_readLE64(memPtr); +} + +MEM_STATIC void MEM_writeLEST(void *memPtr, size_t val) +{ + if (MEM_32bits()) + MEM_writeLE32(memPtr, (U32)val); + else + MEM_writeLE64(memPtr, (U64)val); +} + +/*=== Big endian r/w ===*/ + +MEM_STATIC U32 MEM_readBE32(const void *memPtr) +{ + return get_unaligned_be32(memPtr); +} + +MEM_STATIC void MEM_writeBE32(void *memPtr, U32 val32) +{ + put_unaligned_be32(val32, memPtr); +} + +MEM_STATIC U64 MEM_readBE64(const void *memPtr) +{ + return get_unaligned_be64(memPtr); +} + +MEM_STATIC void MEM_writeBE64(void *memPtr, U64 val64) +{ + put_unaligned_be64(val64, memPtr); +} + +MEM_STATIC size_t MEM_readBEST(const void *memPtr) +{ + if (MEM_32bits()) + return (size_t)MEM_readBE32(memPtr); + else + return (size_t)MEM_readBE64(memPtr); +} + +MEM_STATIC void MEM_writeBEST(void *memPtr, size_t val) +{ + if (MEM_32bits()) + MEM_writeBE32(memPtr, (U32)val); + else + MEM_writeBE64(memPtr, (U64)val); +} + +MEM_STATIC U32 MEM_swap32(U32 in) +{ + return swab32(in); +} + +MEM_STATIC U64 MEM_swap64(U64 in) +{ + return swab64(in); +} + +MEM_STATIC size_t MEM_swapST(size_t in) +{ + if (MEM_32bits()) + return (size_t)MEM_swap32((U32)in); + else + return (size_t)MEM_swap64((U64)in); +} + +#endif /* MEM_H_MODULE */ diff --git a/lib/zstd/common/zstd_common.c b/lib/zstd/common/zstd_common.c new file mode 100644 index 000000000000..3d7e35b309b5 --- /dev/null +++ b/lib/zstd/common/zstd_common.c @@ -0,0 +1,83 @@ +/* + * Copyright (c) Yann Collet, Facebook, Inc. + * All rights reserved. + * + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). + * You may select, at your option, one of the above-listed licenses. + */ + + + +/*-************************************* +* Dependencies +***************************************/ +#define ZSTD_DEPS_NEED_MALLOC +#include "zstd_deps.h" /* ZSTD_malloc, ZSTD_calloc, ZSTD_free, ZSTD_memset */ +#include "error_private.h" +#include "zstd_internal.h" + + +/*-**************************************** +* Version +******************************************/ +unsigned ZSTD_versionNumber(void) { return ZSTD_VERSION_NUMBER; } + +const char* ZSTD_versionString(void) { return ZSTD_VERSION_STRING; } + + +/*-**************************************** +* ZSTD Error Management +******************************************/ +#undef ZSTD_isError /* defined within zstd_internal.h */ +/*! ZSTD_isError() : + * tells if a return value is an error code + * symbol is required for external callers */ +unsigned ZSTD_isError(size_t code) { return ERR_isError(code); } + +/*! ZSTD_getErrorName() : + * provides error code string from function result (useful for debugging) */ +const char* ZSTD_getErrorName(size_t code) { return ERR_getErrorName(code); } + +/*! ZSTD_getError() : + * convert a `size_t` function result into a proper ZSTD_errorCode enum */ +ZSTD_ErrorCode ZSTD_getErrorCode(size_t code) { return ERR_getErrorCode(code); } + +/*! ZSTD_getErrorString() : + * provides error code string from enum */ +const char* ZSTD_getErrorString(ZSTD_ErrorCode code) { return ERR_getErrorString(code); } + + + +/*=************************************************************** +* Custom allocator +****************************************************************/ +void* ZSTD_customMalloc(size_t size, ZSTD_customMem customMem) +{ + if (customMem.customAlloc) + return customMem.customAlloc(customMem.opaque, size); + return ZSTD_malloc(size); +} + +void* ZSTD_customCalloc(size_t size, ZSTD_customMem customMem) +{ + if (customMem.customAlloc) { + /* calloc implemented as malloc+memset; + * not as efficient as calloc, but next best guess for custom malloc */ + void* const ptr = customMem.customAlloc(customMem.opaque, size); + ZSTD_memset(ptr, 0, size); + return ptr; + } + return ZSTD_calloc(1, size); +} + +void ZSTD_customFree(void* ptr, ZSTD_customMem customMem) +{ + if (ptr!=NULL) { + if (customMem.customFree) + customMem.customFree(customMem.opaque, ptr); + else + ZSTD_free(ptr); + } +} diff --git a/lib/zstd/common/zstd_deps.h b/lib/zstd/common/zstd_deps.h new file mode 100644 index 000000000000..7a5bf44839c9 --- /dev/null +++ b/lib/zstd/common/zstd_deps.h @@ -0,0 +1,125 @@ +/* SPDX-License-Identifier: GPL-2.0+ OR BSD-3-Clause */ +/* + * Copyright (c) Facebook, Inc. + * All rights reserved. + * + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). + * You may select, at your option, one of the above-listed licenses. + */ + +/* + * This file provides common libc dependencies that zstd requires. + * The purpose is to allow replacing this file with a custom implementation + * to compile zstd without libc support. + */ + +/* Need: + * NULL + * INT_MAX + * UINT_MAX + * ZSTD_memcpy() + * ZSTD_memset() + * ZSTD_memmove() + */ +#ifndef ZSTD_DEPS_COMMON +#define ZSTD_DEPS_COMMON + +#include +#include + +#define ZSTD_memcpy(d,s,n) __builtin_memcpy((d),(s),(n)) +#define ZSTD_memmove(d,s,n) __builtin_memmove((d),(s),(n)) +#define ZSTD_memset(d,s,n) __builtin_memset((d),(s),(n)) + +#endif /* ZSTD_DEPS_COMMON */ + +/* + * Define malloc as always failing. That means the user must + * either use ZSTD_customMem or statically allocate memory. + * Need: + * ZSTD_malloc() + * ZSTD_free() + * ZSTD_calloc() + */ +#ifdef ZSTD_DEPS_NEED_MALLOC +#ifndef ZSTD_DEPS_MALLOC +#define ZSTD_DEPS_MALLOC + +#define ZSTD_malloc(s) ({ (void)(s); NULL; }) +#define ZSTD_free(p) ((void)(p)) +#define ZSTD_calloc(n,s) ({ (void)(n); (void)(s); NULL; }) + +#endif /* ZSTD_DEPS_MALLOC */ +#endif /* ZSTD_DEPS_NEED_MALLOC */ + +/* + * Provides 64-bit math support. + * Need: + * U64 ZSTD_div64(U64 dividend, U32 divisor) + */ +#ifdef ZSTD_DEPS_NEED_MATH64 +#ifndef ZSTD_DEPS_MATH64 +#define ZSTD_DEPS_MATH64 + +#include + +static uint64_t ZSTD_div64(uint64_t dividend, uint32_t divisor) { + return div_u64(dividend, divisor); +} + +#endif /* ZSTD_DEPS_MATH64 */ +#endif /* ZSTD_DEPS_NEED_MATH64 */ + +/* + * This is only requested when DEBUGLEVEL >= 1, meaning + * it is disabled in production. + * Need: + * assert() + */ +#ifdef ZSTD_DEPS_NEED_ASSERT +#ifndef ZSTD_DEPS_ASSERT +#define ZSTD_DEPS_ASSERT + +#include + +#define assert(x) WARN_ON((x)) + +#endif /* ZSTD_DEPS_ASSERT */ +#endif /* ZSTD_DEPS_NEED_ASSERT */ + +/* + * This is only requested when DEBUGLEVEL >= 2, meaning + * it is disabled in production. + * Need: + * ZSTD_DEBUG_PRINT() + */ +#ifdef ZSTD_DEPS_NEED_IO +#ifndef ZSTD_DEPS_IO +#define ZSTD_DEPS_IO + +#include + +#define ZSTD_DEBUG_PRINT(...) pr_debug(__VA_ARGS__) + +#endif /* ZSTD_DEPS_IO */ +#endif /* ZSTD_DEPS_NEED_IO */ + +/* + * Only requested when MSAN is enabled. + * Need: + * intptr_t + */ +#ifdef ZSTD_DEPS_NEED_STDINT +#ifndef ZSTD_DEPS_STDINT +#define ZSTD_DEPS_STDINT + +/* + * The Linux Kernel doesn't provide intptr_t, only uintptr_t, which + * is an unsigned long. + */ +typedef long intptr_t; + +#endif /* ZSTD_DEPS_STDINT */ +#endif /* ZSTD_DEPS_NEED_STDINT */ diff --git a/lib/zstd/common/zstd_internal.h b/lib/zstd/common/zstd_internal.h new file mode 100644 index 000000000000..fc6f3a9b40c0 --- /dev/null +++ b/lib/zstd/common/zstd_internal.h @@ -0,0 +1,450 @@ +/* + * Copyright (c) Yann Collet, Facebook, Inc. + * All rights reserved. + * + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). + * You may select, at your option, one of the above-listed licenses. + */ + +#ifndef ZSTD_CCOMMON_H_MODULE +#define ZSTD_CCOMMON_H_MODULE + +/* this module contains definitions which must be identical + * across compression, decompression and dictBuilder. + * It also contains a few functions useful to at least 2 of them + * and which benefit from being inlined */ + +/*-************************************* +* Dependencies +***************************************/ +#include "compiler.h" +#include "mem.h" +#include "debug.h" /* assert, DEBUGLOG, RAWLOG, g_debuglevel */ +#include "error_private.h" +#define ZSTD_STATIC_LINKING_ONLY +#include +#define FSE_STATIC_LINKING_ONLY +#include "fse.h" +#define HUF_STATIC_LINKING_ONLY +#include "huf.h" +#include /* XXH_reset, update, digest */ +#define ZSTD_TRACE 0 + + +/* ---- static assert (debug) --- */ +#define ZSTD_STATIC_ASSERT(c) DEBUG_STATIC_ASSERT(c) +#define ZSTD_isError ERR_isError /* for inlining */ +#define FSE_isError ERR_isError +#define HUF_isError ERR_isError + + +/*-************************************* +* shared macros +***************************************/ +#undef MIN +#undef MAX +#define MIN(a,b) ((a)<(b) ? (a) : (b)) +#define MAX(a,b) ((a)>(b) ? (a) : (b)) + +/* + * Ignore: this is an internal helper. + * + * This is a helper function to help force C99-correctness during compilation. + * Under strict compilation modes, variadic macro arguments can't be empty. + * However, variadic function arguments can be. Using a function therefore lets + * us statically check that at least one (string) argument was passed, + * independent of the compilation flags. + */ +static INLINE_KEYWORD UNUSED_ATTR +void _force_has_format_string(const char *format, ...) { + (void)format; +} + +/* + * Ignore: this is an internal helper. + * + * We want to force this function invocation to be syntactically correct, but + * we don't want to force runtime evaluation of its arguments. + */ +#define _FORCE_HAS_FORMAT_STRING(...) \ + if (0) { \ + _force_has_format_string(__VA_ARGS__); \ + } + +/* + * Return the specified error if the condition evaluates to true. + * + * In debug modes, prints additional information. + * In order to do that (particularly, printing the conditional that failed), + * this can't just wrap RETURN_ERROR(). + */ +#define RETURN_ERROR_IF(cond, err, ...) \ + if (cond) { \ + RAWLOG(3, "%s:%d: ERROR!: check %s failed, returning %s", \ + __FILE__, __LINE__, ZSTD_QUOTE(cond), ZSTD_QUOTE(ERROR(err))); \ + _FORCE_HAS_FORMAT_STRING(__VA_ARGS__); \ + RAWLOG(3, ": " __VA_ARGS__); \ + RAWLOG(3, "\n"); \ + return ERROR(err); \ + } + +/* + * Unconditionally return the specified error. + * + * In debug modes, prints additional information. + */ +#define RETURN_ERROR(err, ...) \ + do { \ + RAWLOG(3, "%s:%d: ERROR!: unconditional check failed, returning %s", \ + __FILE__, __LINE__, ZSTD_QUOTE(ERROR(err))); \ + _FORCE_HAS_FORMAT_STRING(__VA_ARGS__); \ + RAWLOG(3, ": " __VA_ARGS__); \ + RAWLOG(3, "\n"); \ + return ERROR(err); \ + } while(0); + +/* + * If the provided expression evaluates to an error code, returns that error code. + * + * In debug modes, prints additional information. + */ +#define FORWARD_IF_ERROR(err, ...) \ + do { \ + size_t const err_code = (err); \ + if (ERR_isError(err_code)) { \ + RAWLOG(3, "%s:%d: ERROR!: forwarding error in %s: %s", \ + __FILE__, __LINE__, ZSTD_QUOTE(err), ERR_getErrorName(err_code)); \ + _FORCE_HAS_FORMAT_STRING(__VA_ARGS__); \ + RAWLOG(3, ": " __VA_ARGS__); \ + RAWLOG(3, "\n"); \ + return err_code; \ + } \ + } while(0); + + +/*-************************************* +* Common constants +***************************************/ +#define ZSTD_OPT_NUM (1<<12) + +#define ZSTD_REP_NUM 3 /* number of repcodes */ +#define ZSTD_REP_MOVE (ZSTD_REP_NUM-1) +static UNUSED_ATTR const U32 repStartValue[ZSTD_REP_NUM] = { 1, 4, 8 }; + +#define KB *(1 <<10) +#define MB *(1 <<20) +#define GB *(1U<<30) + +#define BIT7 128 +#define BIT6 64 +#define BIT5 32 +#define BIT4 16 +#define BIT1 2 +#define BIT0 1 + +#define ZSTD_WINDOWLOG_ABSOLUTEMIN 10 +static UNUSED_ATTR const size_t ZSTD_fcs_fieldSize[4] = { 0, 2, 4, 8 }; +static UNUSED_ATTR const size_t ZSTD_did_fieldSize[4] = { 0, 1, 2, 4 }; + +#define ZSTD_FRAMEIDSIZE 4 /* magic number size */ + +#define ZSTD_BLOCKHEADERSIZE 3 /* C standard doesn't allow `static const` variable to be init using another `static const` variable */ +static UNUSED_ATTR const size_t ZSTD_blockHeaderSize = ZSTD_BLOCKHEADERSIZE; +typedef enum { bt_raw, bt_rle, bt_compressed, bt_reserved } blockType_e; + +#define ZSTD_FRAMECHECKSUMSIZE 4 + +#define MIN_SEQUENCES_SIZE 1 /* nbSeq==0 */ +#define MIN_CBLOCK_SIZE (1 /*litCSize*/ + 1 /* RLE or RAW */ + MIN_SEQUENCES_SIZE /* nbSeq==0 */) /* for a non-null block */ + +#define HufLog 12 +typedef enum { set_basic, set_rle, set_compressed, set_repeat } symbolEncodingType_e; + +#define LONGNBSEQ 0x7F00 + +#define MINMATCH 3 + +#define Litbits 8 +#define MaxLit ((1<= 8 || (ovtype == ZSTD_no_overlap && diff <= -WILDCOPY_VECLEN)); + + if (ovtype == ZSTD_overlap_src_before_dst && diff < WILDCOPY_VECLEN) { + /* Handle short offset copies. */ + do { + COPY8(op, ip) + } while (op < oend); + } else { + assert(diff >= WILDCOPY_VECLEN || diff <= -WILDCOPY_VECLEN); + /* Separate out the first COPY16() call because the copy length is + * almost certain to be short, so the branches have different + * probabilities. Since it is almost certain to be short, only do + * one COPY16() in the first call. Then, do two calls per loop since + * at that point it is more likely to have a high trip count. + */ +#ifdef __aarch64__ + do { + COPY16(op, ip); + } + while (op < oend); +#else + ZSTD_copy16(op, ip); + if (16 >= length) return; + op += 16; + ip += 16; + do { + COPY16(op, ip); + COPY16(op, ip); + } + while (op < oend); +#endif + } +} + +MEM_STATIC size_t ZSTD_limitCopy(void* dst, size_t dstCapacity, const void* src, size_t srcSize) +{ + size_t const length = MIN(dstCapacity, srcSize); + if (length > 0) { + ZSTD_memcpy(dst, src, length); + } + return length; +} + +/* define "workspace is too large" as this number of times larger than needed */ +#define ZSTD_WORKSPACETOOLARGE_FACTOR 3 + +/* when workspace is continuously too large + * during at least this number of times, + * context's memory usage is considered wasteful, + * because it's sized to handle a worst case scenario which rarely happens. + * In which case, resize it down to free some memory */ +#define ZSTD_WORKSPACETOOLARGE_MAXDURATION 128 + +/* Controls whether the input/output buffer is buffered or stable. */ +typedef enum { + ZSTD_bm_buffered = 0, /* Buffer the input/output */ + ZSTD_bm_stable = 1 /* ZSTD_inBuffer/ZSTD_outBuffer is stable */ +} ZSTD_bufferMode_e; + + +/*-******************************************* +* Private declarations +*********************************************/ +typedef struct seqDef_s { + U32 offset; /* Offset code of the sequence */ + U16 litLength; + U16 matchLength; +} seqDef; + +typedef struct { + seqDef* sequencesStart; + seqDef* sequences; /* ptr to end of sequences */ + BYTE* litStart; + BYTE* lit; /* ptr to end of literals */ + BYTE* llCode; + BYTE* mlCode; + BYTE* ofCode; + size_t maxNbSeq; + size_t maxNbLit; + + /* longLengthPos and longLengthID to allow us to represent either a single litLength or matchLength + * in the seqStore that has a value larger than U16 (if it exists). To do so, we increment + * the existing value of the litLength or matchLength by 0x10000. + */ + U32 longLengthID; /* 0 == no longLength; 1 == Represent the long literal; 2 == Represent the long match; */ + U32 longLengthPos; /* Index of the sequence to apply long length modification to */ +} seqStore_t; + +typedef struct { + U32 litLength; + U32 matchLength; +} ZSTD_sequenceLength; + +/* + * Returns the ZSTD_sequenceLength for the given sequences. It handles the decoding of long sequences + * indicated by longLengthPos and longLengthID, and adds MINMATCH back to matchLength. + */ +MEM_STATIC ZSTD_sequenceLength ZSTD_getSequenceLength(seqStore_t const* seqStore, seqDef const* seq) +{ + ZSTD_sequenceLength seqLen; + seqLen.litLength = seq->litLength; + seqLen.matchLength = seq->matchLength + MINMATCH; + if (seqStore->longLengthPos == (U32)(seq - seqStore->sequencesStart)) { + if (seqStore->longLengthID == 1) { + seqLen.litLength += 0xFFFF; + } + if (seqStore->longLengthID == 2) { + seqLen.matchLength += 0xFFFF; + } + } + return seqLen; +} + +/* + * Contains the compressed frame size and an upper-bound for the decompressed frame size. + * Note: before using `compressedSize`, check for errors using ZSTD_isError(). + * similarly, before using `decompressedBound`, check for errors using: + * `decompressedBound != ZSTD_CONTENTSIZE_ERROR` + */ +typedef struct { + size_t compressedSize; + unsigned long long decompressedBound; +} ZSTD_frameSizeInfo; /* decompress & legacy */ + +const seqStore_t* ZSTD_getSeqStore(const ZSTD_CCtx* ctx); /* compress & dictBuilder */ +void ZSTD_seqToCodes(const seqStore_t* seqStorePtr); /* compress, dictBuilder, decodeCorpus (shouldn't get its definition from here) */ + +/* custom memory allocation functions */ +void* ZSTD_customMalloc(size_t size, ZSTD_customMem customMem); +void* ZSTD_customCalloc(size_t size, ZSTD_customMem customMem); +void ZSTD_customFree(void* ptr, ZSTD_customMem customMem); + + +MEM_STATIC U32 ZSTD_highbit32(U32 val) /* compress, dictBuilder, decodeCorpus */ +{ + assert(val != 0); + { +# if (__GNUC__ >= 3) /* GCC Intrinsic */ + return __builtin_clz (val) ^ 31; +# else /* Software version */ + static const U32 DeBruijnClz[32] = { 0, 9, 1, 10, 13, 21, 2, 29, 11, 14, 16, 18, 22, 25, 3, 30, 8, 12, 20, 28, 15, 17, 24, 7, 19, 27, 23, 6, 26, 5, 4, 31 }; + U32 v = val; + v |= v >> 1; + v |= v >> 2; + v |= v >> 4; + v |= v >> 8; + v |= v >> 16; + return DeBruijnClz[(v * 0x07C4ACDDU) >> 27]; +# endif + } +} + + +/* ZSTD_invalidateRepCodes() : + * ensures next compression will not use repcodes from previous block. + * Note : only works with regular variant; + * do not use with extDict variant ! */ +void ZSTD_invalidateRepCodes(ZSTD_CCtx* cctx); /* zstdmt, adaptive_compression (shouldn't get this definition from here) */ + + +typedef struct { + blockType_e blockType; + U32 lastBlock; + U32 origSize; +} blockProperties_t; /* declared here for decompress and fullbench */ + +/*! ZSTD_getcBlockSize() : + * Provides the size of compressed block from block header `src` */ +/* Used by: decompress, fullbench (does not get its definition from here) */ +size_t ZSTD_getcBlockSize(const void* src, size_t srcSize, + blockProperties_t* bpPtr); + +/*! ZSTD_decodeSeqHeaders() : + * decode sequence header from src */ +/* Used by: decompress, fullbench (does not get its definition from here) */ +size_t ZSTD_decodeSeqHeaders(ZSTD_DCtx* dctx, int* nbSeqPtr, + const void* src, size_t srcSize); + + + +#endif /* ZSTD_CCOMMON_H_MODULE */ diff --git a/lib/zstd/compress.c b/lib/zstd/compress.c deleted file mode 100644 index 57aaa64306a0..000000000000 --- a/lib/zstd/compress.c +++ /dev/null @@ -1,3534 +0,0 @@ -/** - * Copyright (c) 2016-present, Yann Collet, Facebook, Inc. - * All rights reserved. - * - * This source code is licensed under the BSD-style license found in the - * LICENSE file in the root directory of https://github.com/facebook/zstd. - * An additional grant of patent rights can be found in the PATENTS file in the - * same directory. - * - * This program is free software; you can redistribute it and/or modify it under - * the terms of the GNU General Public License version 2 as published by the - * Free Software Foundation. This program is dual-licensed; you may select - * either version 2 of the GNU General Public License ("GPL") or BSD license - * ("BSD"). - */ - -/*-************************************* -* Dependencies -***************************************/ -#include "fse.h" -#include "huf.h" -#include "mem.h" -#include "zstd_internal.h" /* includes zstd.h */ -#include -#include -#include /* memset */ - -/*-************************************* -* Constants -***************************************/ -static const U32 g_searchStrength = 8; /* control skip over incompressible data */ -#define HASH_READ_SIZE 8 -typedef enum { ZSTDcs_created = 0, ZSTDcs_init, ZSTDcs_ongoing, ZSTDcs_ending } ZSTD_compressionStage_e; - -/*-************************************* -* Helper functions -***************************************/ -size_t ZSTD_compressBound(size_t srcSize) { return FSE_compressBound(srcSize) + 12; } - -/*-************************************* -* Sequence storage -***************************************/ -static void ZSTD_resetSeqStore(seqStore_t *ssPtr) -{ - ssPtr->lit = ssPtr->litStart; - ssPtr->sequences = ssPtr->sequencesStart; - ssPtr->longLengthID = 0; -} - -/*-************************************* -* Context memory management -***************************************/ -struct ZSTD_CCtx_s { - const BYTE *nextSrc; /* next block here to continue on curr prefix */ - const BYTE *base; /* All regular indexes relative to this position */ - const BYTE *dictBase; /* extDict indexes relative to this position */ - U32 dictLimit; /* below that point, need extDict */ - U32 lowLimit; /* below that point, no more data */ - U32 nextToUpdate; /* index from which to continue dictionary update */ - U32 nextToUpdate3; /* index from which to continue dictionary update */ - U32 hashLog3; /* dispatch table : larger == faster, more memory */ - U32 loadedDictEnd; /* index of end of dictionary */ - U32 forceWindow; /* force back-references to respect limit of 1< 3) ? 0 : MIN(ZSTD_HASHLOG3_MAX, cParams.windowLog); - size_t const h3Size = ((size_t)1) << hashLog3; - size_t const tableSpace = (chainSize + hSize + h3Size) * sizeof(U32); - size_t const optSpace = - ((MaxML + 1) + (MaxLL + 1) + (MaxOff + 1) + (1 << Litbits)) * sizeof(U32) + (ZSTD_OPT_NUM + 1) * (sizeof(ZSTD_match_t) + sizeof(ZSTD_optimal_t)); - size_t const workspaceSize = tableSpace + (256 * sizeof(U32)) /* huffTable */ + tokenSpace + - (((cParams.strategy == ZSTD_btopt) || (cParams.strategy == ZSTD_btopt2)) ? optSpace : 0); - - return ZSTD_ALIGN(sizeof(ZSTD_stack)) + ZSTD_ALIGN(sizeof(ZSTD_CCtx)) + ZSTD_ALIGN(workspaceSize); -} - -static ZSTD_CCtx *ZSTD_createCCtx_advanced(ZSTD_customMem customMem) -{ - ZSTD_CCtx *cctx; - if (!customMem.customAlloc || !customMem.customFree) - return NULL; - cctx = (ZSTD_CCtx *)ZSTD_malloc(sizeof(ZSTD_CCtx), customMem); - if (!cctx) - return NULL; - memset(cctx, 0, sizeof(ZSTD_CCtx)); - cctx->customMem = customMem; - return cctx; -} - -ZSTD_CCtx *ZSTD_initCCtx(void *workspace, size_t workspaceSize) -{ - ZSTD_customMem const stackMem = ZSTD_initStack(workspace, workspaceSize); - ZSTD_CCtx *cctx = ZSTD_createCCtx_advanced(stackMem); - if (cctx) { - cctx->workSpace = ZSTD_stackAllocAll(cctx->customMem.opaque, &cctx->workSpaceSize); - } - return cctx; -} - -size_t ZSTD_freeCCtx(ZSTD_CCtx *cctx) -{ - if (cctx == NULL) - return 0; /* support free on NULL */ - ZSTD_free(cctx->workSpace, cctx->customMem); - ZSTD_free(cctx, cctx->customMem); - return 0; /* reserved as a potential error code in the future */ -} - -const seqStore_t *ZSTD_getSeqStore(const ZSTD_CCtx *ctx) /* hidden interface */ { return &(ctx->seqStore); } - -static ZSTD_parameters ZSTD_getParamsFromCCtx(const ZSTD_CCtx *cctx) { return cctx->params; } - -/** ZSTD_checkParams() : - ensure param values remain within authorized range. - @return : 0, or an error code if one value is beyond authorized range */ -size_t ZSTD_checkCParams(ZSTD_compressionParameters cParams) -{ -#define CLAMPCHECK(val, min, max) \ - { \ - if ((val < min) | (val > max)) \ - return ERROR(compressionParameter_unsupported); \ - } - CLAMPCHECK(cParams.windowLog, ZSTD_WINDOWLOG_MIN, ZSTD_WINDOWLOG_MAX); - CLAMPCHECK(cParams.chainLog, ZSTD_CHAINLOG_MIN, ZSTD_CHAINLOG_MAX); - CLAMPCHECK(cParams.hashLog, ZSTD_HASHLOG_MIN, ZSTD_HASHLOG_MAX); - CLAMPCHECK(cParams.searchLog, ZSTD_SEARCHLOG_MIN, ZSTD_SEARCHLOG_MAX); - CLAMPCHECK(cParams.searchLength, ZSTD_SEARCHLENGTH_MIN, ZSTD_SEARCHLENGTH_MAX); - CLAMPCHECK(cParams.targetLength, ZSTD_TARGETLENGTH_MIN, ZSTD_TARGETLENGTH_MAX); - if ((U32)(cParams.strategy) > (U32)ZSTD_btopt2) - return ERROR(compressionParameter_unsupported); - return 0; -} - -/** ZSTD_cycleLog() : - * condition for correct operation : hashLog > 1 */ -static U32 ZSTD_cycleLog(U32 hashLog, ZSTD_strategy strat) -{ - U32 const btScale = ((U32)strat >= (U32)ZSTD_btlazy2); - return hashLog - btScale; -} - -/** ZSTD_adjustCParams() : - optimize `cPar` for a given input (`srcSize` and `dictSize`). - mostly downsizing to reduce memory consumption and initialization. - Both `srcSize` and `dictSize` are optional (use 0 if unknown), - but if both are 0, no optimization can be done. - Note : cPar is considered validated at this stage. Use ZSTD_checkParams() to ensure that. */ -ZSTD_compressionParameters ZSTD_adjustCParams(ZSTD_compressionParameters cPar, unsigned long long srcSize, size_t dictSize) -{ - if (srcSize + dictSize == 0) - return cPar; /* no size information available : no adjustment */ - - /* resize params, to use less memory when necessary */ - { - U32 const minSrcSize = (srcSize == 0) ? 500 : 0; - U64 const rSize = srcSize + dictSize + minSrcSize; - if (rSize < ((U64)1 << ZSTD_WINDOWLOG_MAX)) { - U32 const srcLog = MAX(ZSTD_HASHLOG_MIN, ZSTD_highbit32((U32)(rSize)-1) + 1); - if (cPar.windowLog > srcLog) - cPar.windowLog = srcLog; - } - } - if (cPar.hashLog > cPar.windowLog) - cPar.hashLog = cPar.windowLog; - { - U32 const cycleLog = ZSTD_cycleLog(cPar.chainLog, cPar.strategy); - if (cycleLog > cPar.windowLog) - cPar.chainLog -= (cycleLog - cPar.windowLog); - } - - if (cPar.windowLog < ZSTD_WINDOWLOG_ABSOLUTEMIN) - cPar.windowLog = ZSTD_WINDOWLOG_ABSOLUTEMIN; /* required for frame header */ - - return cPar; -} - -static U32 ZSTD_equivalentParams(ZSTD_parameters param1, ZSTD_parameters param2) -{ - return (param1.cParams.hashLog == param2.cParams.hashLog) & (param1.cParams.chainLog == param2.cParams.chainLog) & - (param1.cParams.strategy == param2.cParams.strategy) & ((param1.cParams.searchLength == 3) == (param2.cParams.searchLength == 3)); -} - -/*! ZSTD_continueCCtx() : - reuse CCtx without reset (note : requires no dictionary) */ -static size_t ZSTD_continueCCtx(ZSTD_CCtx *cctx, ZSTD_parameters params, U64 frameContentSize) -{ - U32 const end = (U32)(cctx->nextSrc - cctx->base); - cctx->params = params; - cctx->frameContentSize = frameContentSize; - cctx->lowLimit = end; - cctx->dictLimit = end; - cctx->nextToUpdate = end + 1; - cctx->stage = ZSTDcs_init; - cctx->dictID = 0; - cctx->loadedDictEnd = 0; - { - int i; - for (i = 0; i < ZSTD_REP_NUM; i++) - cctx->rep[i] = repStartValue[i]; - } - cctx->seqStore.litLengthSum = 0; /* force reset of btopt stats */ - xxh64_reset(&cctx->xxhState, 0); - return 0; -} - -typedef enum { ZSTDcrp_continue, ZSTDcrp_noMemset, ZSTDcrp_fullReset } ZSTD_compResetPolicy_e; - -/*! ZSTD_resetCCtx_advanced() : - note : `params` must be validated */ -static size_t ZSTD_resetCCtx_advanced(ZSTD_CCtx *zc, ZSTD_parameters params, U64 frameContentSize, ZSTD_compResetPolicy_e const crp) -{ - if (crp == ZSTDcrp_continue) - if (ZSTD_equivalentParams(params, zc->params)) { - zc->flagStaticTables = 0; - zc->flagStaticHufTable = HUF_repeat_none; - return ZSTD_continueCCtx(zc, params, frameContentSize); - } - - { - size_t const blockSize = MIN(ZSTD_BLOCKSIZE_ABSOLUTEMAX, (size_t)1 << params.cParams.windowLog); - U32 const divider = (params.cParams.searchLength == 3) ? 3 : 4; - size_t const maxNbSeq = blockSize / divider; - size_t const tokenSpace = blockSize + 11 * maxNbSeq; - size_t const chainSize = (params.cParams.strategy == ZSTD_fast) ? 0 : (1 << params.cParams.chainLog); - size_t const hSize = ((size_t)1) << params.cParams.hashLog; - U32 const hashLog3 = (params.cParams.searchLength > 3) ? 0 : MIN(ZSTD_HASHLOG3_MAX, params.cParams.windowLog); - size_t const h3Size = ((size_t)1) << hashLog3; - size_t const tableSpace = (chainSize + hSize + h3Size) * sizeof(U32); - void *ptr; - - /* Check if workSpace is large enough, alloc a new one if needed */ - { - size_t const optSpace = ((MaxML + 1) + (MaxLL + 1) + (MaxOff + 1) + (1 << Litbits)) * sizeof(U32) + - (ZSTD_OPT_NUM + 1) * (sizeof(ZSTD_match_t) + sizeof(ZSTD_optimal_t)); - size_t const neededSpace = tableSpace + (256 * sizeof(U32)) /* huffTable */ + tokenSpace + - (((params.cParams.strategy == ZSTD_btopt) || (params.cParams.strategy == ZSTD_btopt2)) ? optSpace : 0); - if (zc->workSpaceSize < neededSpace) { - ZSTD_free(zc->workSpace, zc->customMem); - zc->workSpace = ZSTD_malloc(neededSpace, zc->customMem); - if (zc->workSpace == NULL) - return ERROR(memory_allocation); - zc->workSpaceSize = neededSpace; - } - } - - if (crp != ZSTDcrp_noMemset) - memset(zc->workSpace, 0, tableSpace); /* reset tables only */ - xxh64_reset(&zc->xxhState, 0); - zc->hashLog3 = hashLog3; - zc->hashTable = (U32 *)(zc->workSpace); - zc->chainTable = zc->hashTable + hSize; - zc->hashTable3 = zc->chainTable + chainSize; - ptr = zc->hashTable3 + h3Size; - zc->hufTable = (HUF_CElt *)ptr; - zc->flagStaticTables = 0; - zc->flagStaticHufTable = HUF_repeat_none; - ptr = ((U32 *)ptr) + 256; /* note : HUF_CElt* is incomplete type, size is simulated using U32 */ - - zc->nextToUpdate = 1; - zc->nextSrc = NULL; - zc->base = NULL; - zc->dictBase = NULL; - zc->dictLimit = 0; - zc->lowLimit = 0; - zc->params = params; - zc->blockSize = blockSize; - zc->frameContentSize = frameContentSize; - { - int i; - for (i = 0; i < ZSTD_REP_NUM; i++) - zc->rep[i] = repStartValue[i]; - } - - if ((params.cParams.strategy == ZSTD_btopt) || (params.cParams.strategy == ZSTD_btopt2)) { - zc->seqStore.litFreq = (U32 *)ptr; - zc->seqStore.litLengthFreq = zc->seqStore.litFreq + (1 << Litbits); - zc->seqStore.matchLengthFreq = zc->seqStore.litLengthFreq + (MaxLL + 1); - zc->seqStore.offCodeFreq = zc->seqStore.matchLengthFreq + (MaxML + 1); - ptr = zc->seqStore.offCodeFreq + (MaxOff + 1); - zc->seqStore.matchTable = (ZSTD_match_t *)ptr; - ptr = zc->seqStore.matchTable + ZSTD_OPT_NUM + 1; - zc->seqStore.priceTable = (ZSTD_optimal_t *)ptr; - ptr = zc->seqStore.priceTable + ZSTD_OPT_NUM + 1; - zc->seqStore.litLengthSum = 0; - } - zc->seqStore.sequencesStart = (seqDef *)ptr; - ptr = zc->seqStore.sequencesStart + maxNbSeq; - zc->seqStore.llCode = (BYTE *)ptr; - zc->seqStore.mlCode = zc->seqStore.llCode + maxNbSeq; - zc->seqStore.ofCode = zc->seqStore.mlCode + maxNbSeq; - zc->seqStore.litStart = zc->seqStore.ofCode + maxNbSeq; - - zc->stage = ZSTDcs_init; - zc->dictID = 0; - zc->loadedDictEnd = 0; - - return 0; - } -} - -/* ZSTD_invalidateRepCodes() : - * ensures next compression will not use repcodes from previous block. - * Note : only works with regular variant; - * do not use with extDict variant ! */ -void ZSTD_invalidateRepCodes(ZSTD_CCtx *cctx) -{ - int i; - for (i = 0; i < ZSTD_REP_NUM; i++) - cctx->rep[i] = 0; -} - -/*! ZSTD_copyCCtx() : -* Duplicate an existing context `srcCCtx` into another one `dstCCtx`. -* Only works during stage ZSTDcs_init (i.e. after creation, but before first call to ZSTD_compressContinue()). -* @return : 0, or an error code */ -size_t ZSTD_copyCCtx(ZSTD_CCtx *dstCCtx, const ZSTD_CCtx *srcCCtx, unsigned long long pledgedSrcSize) -{ - if (srcCCtx->stage != ZSTDcs_init) - return ERROR(stage_wrong); - - memcpy(&dstCCtx->customMem, &srcCCtx->customMem, sizeof(ZSTD_customMem)); - { - ZSTD_parameters params = srcCCtx->params; - params.fParams.contentSizeFlag = (pledgedSrcSize > 0); - ZSTD_resetCCtx_advanced(dstCCtx, params, pledgedSrcSize, ZSTDcrp_noMemset); - } - - /* copy tables */ - { - size_t const chainSize = (srcCCtx->params.cParams.strategy == ZSTD_fast) ? 0 : (1 << srcCCtx->params.cParams.chainLog); - size_t const hSize = ((size_t)1) << srcCCtx->params.cParams.hashLog; - size_t const h3Size = (size_t)1 << srcCCtx->hashLog3; - size_t const tableSpace = (chainSize + hSize + h3Size) * sizeof(U32); - memcpy(dstCCtx->workSpace, srcCCtx->workSpace, tableSpace); - } - - /* copy dictionary offsets */ - dstCCtx->nextToUpdate = srcCCtx->nextToUpdate; - dstCCtx->nextToUpdate3 = srcCCtx->nextToUpdate3; - dstCCtx->nextSrc = srcCCtx->nextSrc; - dstCCtx->base = srcCCtx->base; - dstCCtx->dictBase = srcCCtx->dictBase; - dstCCtx->dictLimit = srcCCtx->dictLimit; - dstCCtx->lowLimit = srcCCtx->lowLimit; - dstCCtx->loadedDictEnd = srcCCtx->loadedDictEnd; - dstCCtx->dictID = srcCCtx->dictID; - - /* copy entropy tables */ - dstCCtx->flagStaticTables = srcCCtx->flagStaticTables; - dstCCtx->flagStaticHufTable = srcCCtx->flagStaticHufTable; - if (srcCCtx->flagStaticTables) { - memcpy(dstCCtx->litlengthCTable, srcCCtx->litlengthCTable, sizeof(dstCCtx->litlengthCTable)); - memcpy(dstCCtx->matchlengthCTable, srcCCtx->matchlengthCTable, sizeof(dstCCtx->matchlengthCTable)); - memcpy(dstCCtx->offcodeCTable, srcCCtx->offcodeCTable, sizeof(dstCCtx->offcodeCTable)); - } - if (srcCCtx->flagStaticHufTable) { - memcpy(dstCCtx->hufTable, srcCCtx->hufTable, 256 * 4); - } - - return 0; -} - -/*! ZSTD_reduceTable() : -* reduce table indexes by `reducerValue` */ -static void ZSTD_reduceTable(U32 *const table, U32 const size, U32 const reducerValue) -{ - U32 u; - for (u = 0; u < size; u++) { - if (table[u] < reducerValue) - table[u] = 0; - else - table[u] -= reducerValue; - } -} - -/*! ZSTD_reduceIndex() : -* rescale all indexes to avoid future overflow (indexes are U32) */ -static void ZSTD_reduceIndex(ZSTD_CCtx *zc, const U32 reducerValue) -{ - { - U32 const hSize = 1 << zc->params.cParams.hashLog; - ZSTD_reduceTable(zc->hashTable, hSize, reducerValue); - } - - { - U32 const chainSize = (zc->params.cParams.strategy == ZSTD_fast) ? 0 : (1 << zc->params.cParams.chainLog); - ZSTD_reduceTable(zc->chainTable, chainSize, reducerValue); - } - - { - U32 const h3Size = (zc->hashLog3) ? 1 << zc->hashLog3 : 0; - ZSTD_reduceTable(zc->hashTable3, h3Size, reducerValue); - } -} - -/*-******************************************************* -* Block entropic compression -*********************************************************/ - -/* See doc/zstd_compression_format.md for detailed format description */ - -size_t ZSTD_noCompressBlock(void *dst, size_t dstCapacity, const void *src, size_t srcSize) -{ - if (srcSize + ZSTD_blockHeaderSize > dstCapacity) - return ERROR(dstSize_tooSmall); - memcpy((BYTE *)dst + ZSTD_blockHeaderSize, src, srcSize); - ZSTD_writeLE24(dst, (U32)(srcSize << 2) + (U32)bt_raw); - return ZSTD_blockHeaderSize + srcSize; -} - -static size_t ZSTD_noCompressLiterals(void *dst, size_t dstCapacity, const void *src, size_t srcSize) -{ - BYTE *const ostart = (BYTE * const)dst; - U32 const flSize = 1 + (srcSize > 31) + (srcSize > 4095); - - if (srcSize + flSize > dstCapacity) - return ERROR(dstSize_tooSmall); - - switch (flSize) { - case 1: /* 2 - 1 - 5 */ ostart[0] = (BYTE)((U32)set_basic + (srcSize << 3)); break; - case 2: /* 2 - 2 - 12 */ ZSTD_writeLE16(ostart, (U16)((U32)set_basic + (1 << 2) + (srcSize << 4))); break; - default: /*note : should not be necessary : flSize is within {1,2,3} */ - case 3: /* 2 - 2 - 20 */ ZSTD_writeLE32(ostart, (U32)((U32)set_basic + (3 << 2) + (srcSize << 4))); break; - } - - memcpy(ostart + flSize, src, srcSize); - return srcSize + flSize; -} - -static size_t ZSTD_compressRleLiteralsBlock(void *dst, size_t dstCapacity, const void *src, size_t srcSize) -{ - BYTE *const ostart = (BYTE * const)dst; - U32 const flSize = 1 + (srcSize > 31) + (srcSize > 4095); - - (void)dstCapacity; /* dstCapacity already guaranteed to be >=4, hence large enough */ - - switch (flSize) { - case 1: /* 2 - 1 - 5 */ ostart[0] = (BYTE)((U32)set_rle + (srcSize << 3)); break; - case 2: /* 2 - 2 - 12 */ ZSTD_writeLE16(ostart, (U16)((U32)set_rle + (1 << 2) + (srcSize << 4))); break; - default: /*note : should not be necessary : flSize is necessarily within {1,2,3} */ - case 3: /* 2 - 2 - 20 */ ZSTD_writeLE32(ostart, (U32)((U32)set_rle + (3 << 2) + (srcSize << 4))); break; - } - - ostart[flSize] = *(const BYTE *)src; - return flSize + 1; -} - -static size_t ZSTD_minGain(size_t srcSize) { return (srcSize >> 6) + 2; } - -static size_t ZSTD_compressLiterals(ZSTD_CCtx *zc, void *dst, size_t dstCapacity, const void *src, size_t srcSize) -{ - size_t const minGain = ZSTD_minGain(srcSize); - size_t const lhSize = 3 + (srcSize >= 1 KB) + (srcSize >= 16 KB); - BYTE *const ostart = (BYTE *)dst; - U32 singleStream = srcSize < 256; - symbolEncodingType_e hType = set_compressed; - size_t cLitSize; - -/* small ? don't even attempt compression (speed opt) */ -#define LITERAL_NOENTROPY 63 - { - size_t const minLitSize = zc->flagStaticHufTable == HUF_repeat_valid ? 6 : LITERAL_NOENTROPY; - if (srcSize <= minLitSize) - return ZSTD_noCompressLiterals(dst, dstCapacity, src, srcSize); - } - - if (dstCapacity < lhSize + 1) - return ERROR(dstSize_tooSmall); /* not enough space for compression */ - { - HUF_repeat repeat = zc->flagStaticHufTable; - int const preferRepeat = zc->params.cParams.strategy < ZSTD_lazy ? srcSize <= 1024 : 0; - if (repeat == HUF_repeat_valid && lhSize == 3) - singleStream = 1; - cLitSize = singleStream ? HUF_compress1X_repeat(ostart + lhSize, dstCapacity - lhSize, src, srcSize, 255, 11, zc->tmpCounters, - sizeof(zc->tmpCounters), zc->hufTable, &repeat, preferRepeat) - : HUF_compress4X_repeat(ostart + lhSize, dstCapacity - lhSize, src, srcSize, 255, 11, zc->tmpCounters, - sizeof(zc->tmpCounters), zc->hufTable, &repeat, preferRepeat); - if (repeat != HUF_repeat_none) { - hType = set_repeat; - } /* reused the existing table */ - else { - zc->flagStaticHufTable = HUF_repeat_check; - } /* now have a table to reuse */ - } - - if ((cLitSize == 0) | (cLitSize >= srcSize - minGain)) { - zc->flagStaticHufTable = HUF_repeat_none; - return ZSTD_noCompressLiterals(dst, dstCapacity, src, srcSize); - } - if (cLitSize == 1) { - zc->flagStaticHufTable = HUF_repeat_none; - return ZSTD_compressRleLiteralsBlock(dst, dstCapacity, src, srcSize); - } - - /* Build header */ - switch (lhSize) { - case 3: /* 2 - 2 - 10 - 10 */ - { - U32 const lhc = hType + ((!singleStream) << 2) + ((U32)srcSize << 4) + ((U32)cLitSize << 14); - ZSTD_writeLE24(ostart, lhc); - break; - } - case 4: /* 2 - 2 - 14 - 14 */ - { - U32 const lhc = hType + (2 << 2) + ((U32)srcSize << 4) + ((U32)cLitSize << 18); - ZSTD_writeLE32(ostart, lhc); - break; - } - default: /* should not be necessary, lhSize is only {3,4,5} */ - case 5: /* 2 - 2 - 18 - 18 */ - { - U32 const lhc = hType + (3 << 2) + ((U32)srcSize << 4) + ((U32)cLitSize << 22); - ZSTD_writeLE32(ostart, lhc); - ostart[4] = (BYTE)(cLitSize >> 10); - break; - } - } - return lhSize + cLitSize; -} - -static const BYTE LL_Code[64] = {0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 16, 17, 17, 18, 18, - 19, 19, 20, 20, 20, 20, 21, 21, 21, 21, 22, 22, 22, 22, 22, 22, 22, 22, 23, 23, 23, 23, - 23, 23, 23, 23, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24}; - -static const BYTE ML_Code[128] = {0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, - 26, 27, 28, 29, 30, 31, 32, 32, 33, 33, 34, 34, 35, 35, 36, 36, 36, 36, 37, 37, 37, 37, 38, 38, 38, 38, - 38, 38, 38, 38, 39, 39, 39, 39, 39, 39, 39, 39, 40, 40, 40, 40, 40, 40, 40, 40, 40, 40, 40, 40, 40, 40, - 40, 40, 41, 41, 41, 41, 41, 41, 41, 41, 41, 41, 41, 41, 41, 41, 41, 41, 42, 42, 42, 42, 42, 42, 42, 42, - 42, 42, 42, 42, 42, 42, 42, 42, 42, 42, 42, 42, 42, 42, 42, 42, 42, 42, 42, 42, 42, 42, 42, 42}; - -void ZSTD_seqToCodes(const seqStore_t *seqStorePtr) -{ - BYTE const LL_deltaCode = 19; - BYTE const ML_deltaCode = 36; - const seqDef *const sequences = seqStorePtr->sequencesStart; - BYTE *const llCodeTable = seqStorePtr->llCode; - BYTE *const ofCodeTable = seqStorePtr->ofCode; - BYTE *const mlCodeTable = seqStorePtr->mlCode; - U32 const nbSeq = (U32)(seqStorePtr->sequences - seqStorePtr->sequencesStart); - U32 u; - for (u = 0; u < nbSeq; u++) { - U32 const llv = sequences[u].litLength; - U32 const mlv = sequences[u].matchLength; - llCodeTable[u] = (llv > 63) ? (BYTE)ZSTD_highbit32(llv) + LL_deltaCode : LL_Code[llv]; - ofCodeTable[u] = (BYTE)ZSTD_highbit32(sequences[u].offset); - mlCodeTable[u] = (mlv > 127) ? (BYTE)ZSTD_highbit32(mlv) + ML_deltaCode : ML_Code[mlv]; - } - if (seqStorePtr->longLengthID == 1) - llCodeTable[seqStorePtr->longLengthPos] = MaxLL; - if (seqStorePtr->longLengthID == 2) - mlCodeTable[seqStorePtr->longLengthPos] = MaxML; -} - -ZSTD_STATIC size_t ZSTD_compressSequences_internal(ZSTD_CCtx *zc, void *dst, size_t dstCapacity) -{ - const int longOffsets = zc->params.cParams.windowLog > STREAM_ACCUMULATOR_MIN; - const seqStore_t *seqStorePtr = &(zc->seqStore); - FSE_CTable *CTable_LitLength = zc->litlengthCTable; - FSE_CTable *CTable_OffsetBits = zc->offcodeCTable; - FSE_CTable *CTable_MatchLength = zc->matchlengthCTable; - U32 LLtype, Offtype, MLtype; /* compressed, raw or rle */ - const seqDef *const sequences = seqStorePtr->sequencesStart; - const BYTE *const ofCodeTable = seqStorePtr->ofCode; - const BYTE *const llCodeTable = seqStorePtr->llCode; - const BYTE *const mlCodeTable = seqStorePtr->mlCode; - BYTE *const ostart = (BYTE *)dst; - BYTE *const oend = ostart + dstCapacity; - BYTE *op = ostart; - size_t const nbSeq = seqStorePtr->sequences - seqStorePtr->sequencesStart; - BYTE *seqHead; - - U32 *count; - S16 *norm; - U32 *workspace; - size_t workspaceSize = sizeof(zc->tmpCounters); - { - size_t spaceUsed32 = 0; - count = (U32 *)zc->tmpCounters + spaceUsed32; - spaceUsed32 += MaxSeq + 1; - norm = (S16 *)((U32 *)zc->tmpCounters + spaceUsed32); - spaceUsed32 += ALIGN(sizeof(S16) * (MaxSeq + 1), sizeof(U32)) >> 2; - - workspace = (U32 *)zc->tmpCounters + spaceUsed32; - workspaceSize -= (spaceUsed32 << 2); - } - - /* Compress literals */ - { - const BYTE *const literals = seqStorePtr->litStart; - size_t const litSize = seqStorePtr->lit - literals; - size_t const cSize = ZSTD_compressLiterals(zc, op, dstCapacity, literals, litSize); - if (ZSTD_isError(cSize)) - return cSize; - op += cSize; - } - - /* Sequences Header */ - if ((oend - op) < 3 /*max nbSeq Size*/ + 1 /*seqHead */) - return ERROR(dstSize_tooSmall); - if (nbSeq < 0x7F) - *op++ = (BYTE)nbSeq; - else if (nbSeq < LONGNBSEQ) - op[0] = (BYTE)((nbSeq >> 8) + 0x80), op[1] = (BYTE)nbSeq, op += 2; - else - op[0] = 0xFF, ZSTD_writeLE16(op + 1, (U16)(nbSeq - LONGNBSEQ)), op += 3; - if (nbSeq == 0) - return op - ostart; - - /* seqHead : flags for FSE encoding type */ - seqHead = op++; - -#define MIN_SEQ_FOR_DYNAMIC_FSE 64 -#define MAX_SEQ_FOR_STATIC_FSE 1000 - - /* convert length/distances into codes */ - ZSTD_seqToCodes(seqStorePtr); - - /* CTable for Literal Lengths */ - { - U32 max = MaxLL; - size_t const mostFrequent = FSE_countFast_wksp(count, &max, llCodeTable, nbSeq, workspace); - if ((mostFrequent == nbSeq) && (nbSeq > 2)) { - *op++ = llCodeTable[0]; - FSE_buildCTable_rle(CTable_LitLength, (BYTE)max); - LLtype = set_rle; - } else if ((zc->flagStaticTables) && (nbSeq < MAX_SEQ_FOR_STATIC_FSE)) { - LLtype = set_repeat; - } else if ((nbSeq < MIN_SEQ_FOR_DYNAMIC_FSE) || (mostFrequent < (nbSeq >> (LL_defaultNormLog - 1)))) { - FSE_buildCTable_wksp(CTable_LitLength, LL_defaultNorm, MaxLL, LL_defaultNormLog, workspace, workspaceSize); - LLtype = set_basic; - } else { - size_t nbSeq_1 = nbSeq; - const U32 tableLog = FSE_optimalTableLog(LLFSELog, nbSeq, max); - if (count[llCodeTable[nbSeq - 1]] > 1) { - count[llCodeTable[nbSeq - 1]]--; - nbSeq_1--; - } - FSE_normalizeCount(norm, tableLog, count, nbSeq_1, max); - { - size_t const NCountSize = FSE_writeNCount(op, oend - op, norm, max, tableLog); /* overflow protected */ - if (FSE_isError(NCountSize)) - return NCountSize; - op += NCountSize; - } - FSE_buildCTable_wksp(CTable_LitLength, norm, max, tableLog, workspace, workspaceSize); - LLtype = set_compressed; - } - } - - /* CTable for Offsets */ - { - U32 max = MaxOff; - size_t const mostFrequent = FSE_countFast_wksp(count, &max, ofCodeTable, nbSeq, workspace); - if ((mostFrequent == nbSeq) && (nbSeq > 2)) { - *op++ = ofCodeTable[0]; - FSE_buildCTable_rle(CTable_OffsetBits, (BYTE)max); - Offtype = set_rle; - } else if ((zc->flagStaticTables) && (nbSeq < MAX_SEQ_FOR_STATIC_FSE)) { - Offtype = set_repeat; - } else if ((nbSeq < MIN_SEQ_FOR_DYNAMIC_FSE) || (mostFrequent < (nbSeq >> (OF_defaultNormLog - 1)))) { - FSE_buildCTable_wksp(CTable_OffsetBits, OF_defaultNorm, MaxOff, OF_defaultNormLog, workspace, workspaceSize); - Offtype = set_basic; - } else { - size_t nbSeq_1 = nbSeq; - const U32 tableLog = FSE_optimalTableLog(OffFSELog, nbSeq, max); - if (count[ofCodeTable[nbSeq - 1]] > 1) { - count[ofCodeTable[nbSeq - 1]]--; - nbSeq_1--; - } - FSE_normalizeCount(norm, tableLog, count, nbSeq_1, max); - { - size_t const NCountSize = FSE_writeNCount(op, oend - op, norm, max, tableLog); /* overflow protected */ - if (FSE_isError(NCountSize)) - return NCountSize; - op += NCountSize; - } - FSE_buildCTable_wksp(CTable_OffsetBits, norm, max, tableLog, workspace, workspaceSize); - Offtype = set_compressed; - } - } - - /* CTable for MatchLengths */ - { - U32 max = MaxML; - size_t const mostFrequent = FSE_countFast_wksp(count, &max, mlCodeTable, nbSeq, workspace); - if ((mostFrequent == nbSeq) && (nbSeq > 2)) { - *op++ = *mlCodeTable; - FSE_buildCTable_rle(CTable_MatchLength, (BYTE)max); - MLtype = set_rle; - } else if ((zc->flagStaticTables) && (nbSeq < MAX_SEQ_FOR_STATIC_FSE)) { - MLtype = set_repeat; - } else if ((nbSeq < MIN_SEQ_FOR_DYNAMIC_FSE) || (mostFrequent < (nbSeq >> (ML_defaultNormLog - 1)))) { - FSE_buildCTable_wksp(CTable_MatchLength, ML_defaultNorm, MaxML, ML_defaultNormLog, workspace, workspaceSize); - MLtype = set_basic; - } else { - size_t nbSeq_1 = nbSeq; - const U32 tableLog = FSE_optimalTableLog(MLFSELog, nbSeq, max); - if (count[mlCodeTable[nbSeq - 1]] > 1) { - count[mlCodeTable[nbSeq - 1]]--; - nbSeq_1--; - } - FSE_normalizeCount(norm, tableLog, count, nbSeq_1, max); - { - size_t const NCountSize = FSE_writeNCount(op, oend - op, norm, max, tableLog); /* overflow protected */ - if (FSE_isError(NCountSize)) - return NCountSize; - op += NCountSize; - } - FSE_buildCTable_wksp(CTable_MatchLength, norm, max, tableLog, workspace, workspaceSize); - MLtype = set_compressed; - } - } - - *seqHead = (BYTE)((LLtype << 6) + (Offtype << 4) + (MLtype << 2)); - zc->flagStaticTables = 0; - - /* Encoding Sequences */ - { - BIT_CStream_t blockStream; - FSE_CState_t stateMatchLength; - FSE_CState_t stateOffsetBits; - FSE_CState_t stateLitLength; - - CHECK_E(BIT_initCStream(&blockStream, op, oend - op), dstSize_tooSmall); /* not enough space remaining */ - - /* first symbols */ - FSE_initCState2(&stateMatchLength, CTable_MatchLength, mlCodeTable[nbSeq - 1]); - FSE_initCState2(&stateOffsetBits, CTable_OffsetBits, ofCodeTable[nbSeq - 1]); - FSE_initCState2(&stateLitLength, CTable_LitLength, llCodeTable[nbSeq - 1]); - BIT_addBits(&blockStream, sequences[nbSeq - 1].litLength, LL_bits[llCodeTable[nbSeq - 1]]); - if (ZSTD_32bits()) - BIT_flushBits(&blockStream); - BIT_addBits(&blockStream, sequences[nbSeq - 1].matchLength, ML_bits[mlCodeTable[nbSeq - 1]]); - if (ZSTD_32bits()) - BIT_flushBits(&blockStream); - if (longOffsets) { - U32 const ofBits = ofCodeTable[nbSeq - 1]; - int const extraBits = ofBits - MIN(ofBits, STREAM_ACCUMULATOR_MIN - 1); - if (extraBits) { - BIT_addBits(&blockStream, sequences[nbSeq - 1].offset, extraBits); - BIT_flushBits(&blockStream); - } - BIT_addBits(&blockStream, sequences[nbSeq - 1].offset >> extraBits, ofBits - extraBits); - } else { - BIT_addBits(&blockStream, sequences[nbSeq - 1].offset, ofCodeTable[nbSeq - 1]); - } - BIT_flushBits(&blockStream); - - { - size_t n; - for (n = nbSeq - 2; n < nbSeq; n--) { /* intentional underflow */ - BYTE const llCode = llCodeTable[n]; - BYTE const ofCode = ofCodeTable[n]; - BYTE const mlCode = mlCodeTable[n]; - U32 const llBits = LL_bits[llCode]; - U32 const ofBits = ofCode; /* 32b*/ /* 64b*/ - U32 const mlBits = ML_bits[mlCode]; - /* (7)*/ /* (7)*/ - FSE_encodeSymbol(&blockStream, &stateOffsetBits, ofCode); /* 15 */ /* 15 */ - FSE_encodeSymbol(&blockStream, &stateMatchLength, mlCode); /* 24 */ /* 24 */ - if (ZSTD_32bits()) - BIT_flushBits(&blockStream); /* (7)*/ - FSE_encodeSymbol(&blockStream, &stateLitLength, llCode); /* 16 */ /* 33 */ - if (ZSTD_32bits() || (ofBits + mlBits + llBits >= 64 - 7 - (LLFSELog + MLFSELog + OffFSELog))) - BIT_flushBits(&blockStream); /* (7)*/ - BIT_addBits(&blockStream, sequences[n].litLength, llBits); - if (ZSTD_32bits() && ((llBits + mlBits) > 24)) - BIT_flushBits(&blockStream); - BIT_addBits(&blockStream, sequences[n].matchLength, mlBits); - if (ZSTD_32bits()) - BIT_flushBits(&blockStream); /* (7)*/ - if (longOffsets) { - int const extraBits = ofBits - MIN(ofBits, STREAM_ACCUMULATOR_MIN - 1); - if (extraBits) { - BIT_addBits(&blockStream, sequences[n].offset, extraBits); - BIT_flushBits(&blockStream); /* (7)*/ - } - BIT_addBits(&blockStream, sequences[n].offset >> extraBits, ofBits - extraBits); /* 31 */ - } else { - BIT_addBits(&blockStream, sequences[n].offset, ofBits); /* 31 */ - } - BIT_flushBits(&blockStream); /* (7)*/ - } - } - - FSE_flushCState(&blockStream, &stateMatchLength); - FSE_flushCState(&blockStream, &stateOffsetBits); - FSE_flushCState(&blockStream, &stateLitLength); - - { - size_t const streamSize = BIT_closeCStream(&blockStream); - if (streamSize == 0) - return ERROR(dstSize_tooSmall); /* not enough space */ - op += streamSize; - } - } - return op - ostart; -} - -ZSTD_STATIC size_t ZSTD_compressSequences(ZSTD_CCtx *zc, void *dst, size_t dstCapacity, size_t srcSize) -{ - size_t const cSize = ZSTD_compressSequences_internal(zc, dst, dstCapacity); - size_t const minGain = ZSTD_minGain(srcSize); - size_t const maxCSize = srcSize - minGain; - /* If the srcSize <= dstCapacity, then there is enough space to write a - * raw uncompressed block. Since we ran out of space, the block must not - * be compressible, so fall back to a raw uncompressed block. - */ - int const uncompressibleError = cSize == ERROR(dstSize_tooSmall) && srcSize <= dstCapacity; - int i; - - if (ZSTD_isError(cSize) && !uncompressibleError) - return cSize; - if (cSize >= maxCSize || uncompressibleError) { - zc->flagStaticHufTable = HUF_repeat_none; - return 0; - } - /* confirm repcodes */ - for (i = 0; i < ZSTD_REP_NUM; i++) - zc->rep[i] = zc->repToConfirm[i]; - return cSize; -} - -/*! ZSTD_storeSeq() : - Store a sequence (literal length, literals, offset code and match length code) into seqStore_t. - `offsetCode` : distance to match, or 0 == repCode. - `matchCode` : matchLength - MINMATCH -*/ -ZSTD_STATIC void ZSTD_storeSeq(seqStore_t *seqStorePtr, size_t litLength, const void *literals, U32 offsetCode, size_t matchCode) -{ - /* copy Literals */ - ZSTD_wildcopy(seqStorePtr->lit, literals, litLength); - seqStorePtr->lit += litLength; - - /* literal Length */ - if (litLength > 0xFFFF) { - seqStorePtr->longLengthID = 1; - seqStorePtr->longLengthPos = (U32)(seqStorePtr->sequences - seqStorePtr->sequencesStart); - } - seqStorePtr->sequences[0].litLength = (U16)litLength; - - /* match offset */ - seqStorePtr->sequences[0].offset = offsetCode + 1; - - /* match Length */ - if (matchCode > 0xFFFF) { - seqStorePtr->longLengthID = 2; - seqStorePtr->longLengthPos = (U32)(seqStorePtr->sequences - seqStorePtr->sequencesStart); - } - seqStorePtr->sequences[0].matchLength = (U16)matchCode; - - seqStorePtr->sequences++; -} - -/*-************************************* -* Match length counter -***************************************/ -static unsigned ZSTD_NbCommonBytes(register size_t val) -{ - if (ZSTD_isLittleEndian()) { - if (ZSTD_64bits()) { - return (__builtin_ctzll((U64)val) >> 3); - } else { /* 32 bits */ - return (__builtin_ctz((U32)val) >> 3); - } - } else { /* Big Endian CPU */ - if (ZSTD_64bits()) { - return (__builtin_clzll(val) >> 3); - } else { /* 32 bits */ - return (__builtin_clz((U32)val) >> 3); - } - } -} - -static size_t ZSTD_count(const BYTE *pIn, const BYTE *pMatch, const BYTE *const pInLimit) -{ - const BYTE *const pStart = pIn; - const BYTE *const pInLoopLimit = pInLimit - (sizeof(size_t) - 1); - - while (pIn < pInLoopLimit) { - size_t const diff = ZSTD_readST(pMatch) ^ ZSTD_readST(pIn); - if (!diff) { - pIn += sizeof(size_t); - pMatch += sizeof(size_t); - continue; - } - pIn += ZSTD_NbCommonBytes(diff); - return (size_t)(pIn - pStart); - } - if (ZSTD_64bits()) - if ((pIn < (pInLimit - 3)) && (ZSTD_read32(pMatch) == ZSTD_read32(pIn))) { - pIn += 4; - pMatch += 4; - } - if ((pIn < (pInLimit - 1)) && (ZSTD_read16(pMatch) == ZSTD_read16(pIn))) { - pIn += 2; - pMatch += 2; - } - if ((pIn < pInLimit) && (*pMatch == *pIn)) - pIn++; - return (size_t)(pIn - pStart); -} - -/** ZSTD_count_2segments() : -* can count match length with `ip` & `match` in 2 different segments. -* convention : on reaching mEnd, match count continue starting from iStart -*/ -static size_t ZSTD_count_2segments(const BYTE *ip, const BYTE *match, const BYTE *iEnd, const BYTE *mEnd, const BYTE *iStart) -{ - const BYTE *const vEnd = MIN(ip + (mEnd - match), iEnd); - size_t const matchLength = ZSTD_count(ip, match, vEnd); - if (match + matchLength != mEnd) - return matchLength; - return matchLength + ZSTD_count(ip + matchLength, iStart, iEnd); -} - -/*-************************************* -* Hashes -***************************************/ -static const U32 prime3bytes = 506832829U; -static U32 ZSTD_hash3(U32 u, U32 h) { return ((u << (32 - 24)) * prime3bytes) >> (32 - h); } -ZSTD_STATIC size_t ZSTD_hash3Ptr(const void *ptr, U32 h) { return ZSTD_hash3(ZSTD_readLE32(ptr), h); } /* only in zstd_opt.h */ - -static const U32 prime4bytes = 2654435761U; -static U32 ZSTD_hash4(U32 u, U32 h) { return (u * prime4bytes) >> (32 - h); } -static size_t ZSTD_hash4Ptr(const void *ptr, U32 h) { return ZSTD_hash4(ZSTD_read32(ptr), h); } - -static const U64 prime5bytes = 889523592379ULL; -static size_t ZSTD_hash5(U64 u, U32 h) { return (size_t)(((u << (64 - 40)) * prime5bytes) >> (64 - h)); } -static size_t ZSTD_hash5Ptr(const void *p, U32 h) { return ZSTD_hash5(ZSTD_readLE64(p), h); } - -static const U64 prime6bytes = 227718039650203ULL; -static size_t ZSTD_hash6(U64 u, U32 h) { return (size_t)(((u << (64 - 48)) * prime6bytes) >> (64 - h)); } -static size_t ZSTD_hash6Ptr(const void *p, U32 h) { return ZSTD_hash6(ZSTD_readLE64(p), h); } - -static const U64 prime7bytes = 58295818150454627ULL; -static size_t ZSTD_hash7(U64 u, U32 h) { return (size_t)(((u << (64 - 56)) * prime7bytes) >> (64 - h)); } -static size_t ZSTD_hash7Ptr(const void *p, U32 h) { return ZSTD_hash7(ZSTD_readLE64(p), h); } - -static const U64 prime8bytes = 0xCF1BBCDCB7A56463ULL; -static size_t ZSTD_hash8(U64 u, U32 h) { return (size_t)(((u)*prime8bytes) >> (64 - h)); } -static size_t ZSTD_hash8Ptr(const void *p, U32 h) { return ZSTD_hash8(ZSTD_readLE64(p), h); } - -static size_t ZSTD_hashPtr(const void *p, U32 hBits, U32 mls) -{ - switch (mls) { - // case 3: return ZSTD_hash3Ptr(p, hBits); - default: - case 4: return ZSTD_hash4Ptr(p, hBits); - case 5: return ZSTD_hash5Ptr(p, hBits); - case 6: return ZSTD_hash6Ptr(p, hBits); - case 7: return ZSTD_hash7Ptr(p, hBits); - case 8: return ZSTD_hash8Ptr(p, hBits); - } -} - -/*-************************************* -* Fast Scan -***************************************/ -static void ZSTD_fillHashTable(ZSTD_CCtx *zc, const void *end, const U32 mls) -{ - U32 *const hashTable = zc->hashTable; - U32 const hBits = zc->params.cParams.hashLog; - const BYTE *const base = zc->base; - const BYTE *ip = base + zc->nextToUpdate; - const BYTE *const iend = ((const BYTE *)end) - HASH_READ_SIZE; - const size_t fastHashFillStep = 3; - - while (ip <= iend) { - hashTable[ZSTD_hashPtr(ip, hBits, mls)] = (U32)(ip - base); - ip += fastHashFillStep; - } -} - -FORCE_INLINE -void ZSTD_compressBlock_fast_generic(ZSTD_CCtx *cctx, const void *src, size_t srcSize, const U32 mls) -{ - U32 *const hashTable = cctx->hashTable; - U32 const hBits = cctx->params.cParams.hashLog; - seqStore_t *seqStorePtr = &(cctx->seqStore); - const BYTE *const base = cctx->base; - const BYTE *const istart = (const BYTE *)src; - const BYTE *ip = istart; - const BYTE *anchor = istart; - const U32 lowestIndex = cctx->dictLimit; - const BYTE *const lowest = base + lowestIndex; - const BYTE *const iend = istart + srcSize; - const BYTE *const ilimit = iend - HASH_READ_SIZE; - U32 offset_1 = cctx->rep[0], offset_2 = cctx->rep[1]; - U32 offsetSaved = 0; - - /* init */ - ip += (ip == lowest); - { - U32 const maxRep = (U32)(ip - lowest); - if (offset_2 > maxRep) - offsetSaved = offset_2, offset_2 = 0; - if (offset_1 > maxRep) - offsetSaved = offset_1, offset_1 = 0; - } - - /* Main Search Loop */ - while (ip < ilimit) { /* < instead of <=, because repcode check at (ip+1) */ - size_t mLength; - size_t const h = ZSTD_hashPtr(ip, hBits, mls); - U32 const curr = (U32)(ip - base); - U32 const matchIndex = hashTable[h]; - const BYTE *match = base + matchIndex; - hashTable[h] = curr; /* update hash table */ - - if ((offset_1 > 0) & (ZSTD_read32(ip + 1 - offset_1) == ZSTD_read32(ip + 1))) { - mLength = ZSTD_count(ip + 1 + 4, ip + 1 + 4 - offset_1, iend) + 4; - ip++; - ZSTD_storeSeq(seqStorePtr, ip - anchor, anchor, 0, mLength - MINMATCH); - } else { - U32 offset; - if ((matchIndex <= lowestIndex) || (ZSTD_read32(match) != ZSTD_read32(ip))) { - ip += ((ip - anchor) >> g_searchStrength) + 1; - continue; - } - mLength = ZSTD_count(ip + 4, match + 4, iend) + 4; - offset = (U32)(ip - match); - while (((ip > anchor) & (match > lowest)) && (ip[-1] == match[-1])) { - ip--; - match--; - mLength++; - } /* catch up */ - offset_2 = offset_1; - offset_1 = offset; - - ZSTD_storeSeq(seqStorePtr, ip - anchor, anchor, offset + ZSTD_REP_MOVE, mLength - MINMATCH); - } - - /* match found */ - ip += mLength; - anchor = ip; - - if (ip <= ilimit) { - /* Fill Table */ - hashTable[ZSTD_hashPtr(base + curr + 2, hBits, mls)] = curr + 2; /* here because curr+2 could be > iend-8 */ - hashTable[ZSTD_hashPtr(ip - 2, hBits, mls)] = (U32)(ip - 2 - base); - /* check immediate repcode */ - while ((ip <= ilimit) && ((offset_2 > 0) & (ZSTD_read32(ip) == ZSTD_read32(ip - offset_2)))) { - /* store sequence */ - size_t const rLength = ZSTD_count(ip + 4, ip + 4 - offset_2, iend) + 4; - { - U32 const tmpOff = offset_2; - offset_2 = offset_1; - offset_1 = tmpOff; - } /* swap offset_2 <=> offset_1 */ - hashTable[ZSTD_hashPtr(ip, hBits, mls)] = (U32)(ip - base); - ZSTD_storeSeq(seqStorePtr, 0, anchor, 0, rLength - MINMATCH); - ip += rLength; - anchor = ip; - continue; /* faster when present ... (?) */ - } - } - } - - /* save reps for next block */ - cctx->repToConfirm[0] = offset_1 ? offset_1 : offsetSaved; - cctx->repToConfirm[1] = offset_2 ? offset_2 : offsetSaved; - - /* Last Literals */ - { - size_t const lastLLSize = iend - anchor; - memcpy(seqStorePtr->lit, anchor, lastLLSize); - seqStorePtr->lit += lastLLSize; - } -} - -static void ZSTD_compressBlock_fast(ZSTD_CCtx *ctx, const void *src, size_t srcSize) -{ - const U32 mls = ctx->params.cParams.searchLength; - switch (mls) { - default: /* includes case 3 */ - case 4: ZSTD_compressBlock_fast_generic(ctx, src, srcSize, 4); return; - case 5: ZSTD_compressBlock_fast_generic(ctx, src, srcSize, 5); return; - case 6: ZSTD_compressBlock_fast_generic(ctx, src, srcSize, 6); return; - case 7: ZSTD_compressBlock_fast_generic(ctx, src, srcSize, 7); return; - } -} - -static void ZSTD_compressBlock_fast_extDict_generic(ZSTD_CCtx *ctx, const void *src, size_t srcSize, const U32 mls) -{ - U32 *hashTable = ctx->hashTable; - const U32 hBits = ctx->params.cParams.hashLog; - seqStore_t *seqStorePtr = &(ctx->seqStore); - const BYTE *const base = ctx->base; - const BYTE *const dictBase = ctx->dictBase; - const BYTE *const istart = (const BYTE *)src; - const BYTE *ip = istart; - const BYTE *anchor = istart; - const U32 lowestIndex = ctx->lowLimit; - const BYTE *const dictStart = dictBase + lowestIndex; - const U32 dictLimit = ctx->dictLimit; - const BYTE *const lowPrefixPtr = base + dictLimit; - const BYTE *const dictEnd = dictBase + dictLimit; - const BYTE *const iend = istart + srcSize; - const BYTE *const ilimit = iend - 8; - U32 offset_1 = ctx->rep[0], offset_2 = ctx->rep[1]; - - /* Search Loop */ - while (ip < ilimit) { /* < instead of <=, because (ip+1) */ - const size_t h = ZSTD_hashPtr(ip, hBits, mls); - const U32 matchIndex = hashTable[h]; - const BYTE *matchBase = matchIndex < dictLimit ? dictBase : base; - const BYTE *match = matchBase + matchIndex; - const U32 curr = (U32)(ip - base); - const U32 repIndex = curr + 1 - offset_1; /* offset_1 expected <= curr +1 */ - const BYTE *repBase = repIndex < dictLimit ? dictBase : base; - const BYTE *repMatch = repBase + repIndex; - size_t mLength; - hashTable[h] = curr; /* update hash table */ - - if ((((U32)((dictLimit - 1) - repIndex) >= 3) /* intentional underflow */ & (repIndex > lowestIndex)) && - (ZSTD_read32(repMatch) == ZSTD_read32(ip + 1))) { - const BYTE *repMatchEnd = repIndex < dictLimit ? dictEnd : iend; - mLength = ZSTD_count_2segments(ip + 1 + EQUAL_READ32, repMatch + EQUAL_READ32, iend, repMatchEnd, lowPrefixPtr) + EQUAL_READ32; - ip++; - ZSTD_storeSeq(seqStorePtr, ip - anchor, anchor, 0, mLength - MINMATCH); - } else { - if ((matchIndex < lowestIndex) || (ZSTD_read32(match) != ZSTD_read32(ip))) { - ip += ((ip - anchor) >> g_searchStrength) + 1; - continue; - } - { - const BYTE *matchEnd = matchIndex < dictLimit ? dictEnd : iend; - const BYTE *lowMatchPtr = matchIndex < dictLimit ? dictStart : lowPrefixPtr; - U32 offset; - mLength = ZSTD_count_2segments(ip + EQUAL_READ32, match + EQUAL_READ32, iend, matchEnd, lowPrefixPtr) + EQUAL_READ32; - while (((ip > anchor) & (match > lowMatchPtr)) && (ip[-1] == match[-1])) { - ip--; - match--; - mLength++; - } /* catch up */ - offset = curr - matchIndex; - offset_2 = offset_1; - offset_1 = offset; - ZSTD_storeSeq(seqStorePtr, ip - anchor, anchor, offset + ZSTD_REP_MOVE, mLength - MINMATCH); - } - } - - /* found a match : store it */ - ip += mLength; - anchor = ip; - - if (ip <= ilimit) { - /* Fill Table */ - hashTable[ZSTD_hashPtr(base + curr + 2, hBits, mls)] = curr + 2; - hashTable[ZSTD_hashPtr(ip - 2, hBits, mls)] = (U32)(ip - 2 - base); - /* check immediate repcode */ - while (ip <= ilimit) { - U32 const curr2 = (U32)(ip - base); - U32 const repIndex2 = curr2 - offset_2; - const BYTE *repMatch2 = repIndex2 < dictLimit ? dictBase + repIndex2 : base + repIndex2; - if ((((U32)((dictLimit - 1) - repIndex2) >= 3) & (repIndex2 > lowestIndex)) /* intentional overflow */ - && (ZSTD_read32(repMatch2) == ZSTD_read32(ip))) { - const BYTE *const repEnd2 = repIndex2 < dictLimit ? dictEnd : iend; - size_t repLength2 = - ZSTD_count_2segments(ip + EQUAL_READ32, repMatch2 + EQUAL_READ32, iend, repEnd2, lowPrefixPtr) + EQUAL_READ32; - U32 tmpOffset = offset_2; - offset_2 = offset_1; - offset_1 = tmpOffset; /* swap offset_2 <=> offset_1 */ - ZSTD_storeSeq(seqStorePtr, 0, anchor, 0, repLength2 - MINMATCH); - hashTable[ZSTD_hashPtr(ip, hBits, mls)] = curr2; - ip += repLength2; - anchor = ip; - continue; - } - break; - } - } - } - - /* save reps for next block */ - ctx->repToConfirm[0] = offset_1; - ctx->repToConfirm[1] = offset_2; - - /* Last Literals */ - { - size_t const lastLLSize = iend - anchor; - memcpy(seqStorePtr->lit, anchor, lastLLSize); - seqStorePtr->lit += lastLLSize; - } -} - -static void ZSTD_compressBlock_fast_extDict(ZSTD_CCtx *ctx, const void *src, size_t srcSize) -{ - U32 const mls = ctx->params.cParams.searchLength; - switch (mls) { - default: /* includes case 3 */ - case 4: ZSTD_compressBlock_fast_extDict_generic(ctx, src, srcSize, 4); return; - case 5: ZSTD_compressBlock_fast_extDict_generic(ctx, src, srcSize, 5); return; - case 6: ZSTD_compressBlock_fast_extDict_generic(ctx, src, srcSize, 6); return; - case 7: ZSTD_compressBlock_fast_extDict_generic(ctx, src, srcSize, 7); return; - } -} - -/*-************************************* -* Double Fast -***************************************/ -static void ZSTD_fillDoubleHashTable(ZSTD_CCtx *cctx, const void *end, const U32 mls) -{ - U32 *const hashLarge = cctx->hashTable; - U32 const hBitsL = cctx->params.cParams.hashLog; - U32 *const hashSmall = cctx->chainTable; - U32 const hBitsS = cctx->params.cParams.chainLog; - const BYTE *const base = cctx->base; - const BYTE *ip = base + cctx->nextToUpdate; - const BYTE *const iend = ((const BYTE *)end) - HASH_READ_SIZE; - const size_t fastHashFillStep = 3; - - while (ip <= iend) { - hashSmall[ZSTD_hashPtr(ip, hBitsS, mls)] = (U32)(ip - base); - hashLarge[ZSTD_hashPtr(ip, hBitsL, 8)] = (U32)(ip - base); - ip += fastHashFillStep; - } -} - -FORCE_INLINE -void ZSTD_compressBlock_doubleFast_generic(ZSTD_CCtx *cctx, const void *src, size_t srcSize, const U32 mls) -{ - U32 *const hashLong = cctx->hashTable; - const U32 hBitsL = cctx->params.cParams.hashLog; - U32 *const hashSmall = cctx->chainTable; - const U32 hBitsS = cctx->params.cParams.chainLog; - seqStore_t *seqStorePtr = &(cctx->seqStore); - const BYTE *const base = cctx->base; - const BYTE *const istart = (const BYTE *)src; - const BYTE *ip = istart; - const BYTE *anchor = istart; - const U32 lowestIndex = cctx->dictLimit; - const BYTE *const lowest = base + lowestIndex; - const BYTE *const iend = istart + srcSize; - const BYTE *const ilimit = iend - HASH_READ_SIZE; - U32 offset_1 = cctx->rep[0], offset_2 = cctx->rep[1]; - U32 offsetSaved = 0; - - /* init */ - ip += (ip == lowest); - { - U32 const maxRep = (U32)(ip - lowest); - if (offset_2 > maxRep) - offsetSaved = offset_2, offset_2 = 0; - if (offset_1 > maxRep) - offsetSaved = offset_1, offset_1 = 0; - } - - /* Main Search Loop */ - while (ip < ilimit) { /* < instead of <=, because repcode check at (ip+1) */ - size_t mLength; - size_t const h2 = ZSTD_hashPtr(ip, hBitsL, 8); - size_t const h = ZSTD_hashPtr(ip, hBitsS, mls); - U32 const curr = (U32)(ip - base); - U32 const matchIndexL = hashLong[h2]; - U32 const matchIndexS = hashSmall[h]; - const BYTE *matchLong = base + matchIndexL; - const BYTE *match = base + matchIndexS; - hashLong[h2] = hashSmall[h] = curr; /* update hash tables */ - - if ((offset_1 > 0) & (ZSTD_read32(ip + 1 - offset_1) == ZSTD_read32(ip + 1))) { /* note : by construction, offset_1 <= curr */ - mLength = ZSTD_count(ip + 1 + 4, ip + 1 + 4 - offset_1, iend) + 4; - ip++; - ZSTD_storeSeq(seqStorePtr, ip - anchor, anchor, 0, mLength - MINMATCH); - } else { - U32 offset; - if ((matchIndexL > lowestIndex) && (ZSTD_read64(matchLong) == ZSTD_read64(ip))) { - mLength = ZSTD_count(ip + 8, matchLong + 8, iend) + 8; - offset = (U32)(ip - matchLong); - while (((ip > anchor) & (matchLong > lowest)) && (ip[-1] == matchLong[-1])) { - ip--; - matchLong--; - mLength++; - } /* catch up */ - } else if ((matchIndexS > lowestIndex) && (ZSTD_read32(match) == ZSTD_read32(ip))) { - size_t const h3 = ZSTD_hashPtr(ip + 1, hBitsL, 8); - U32 const matchIndex3 = hashLong[h3]; - const BYTE *match3 = base + matchIndex3; - hashLong[h3] = curr + 1; - if ((matchIndex3 > lowestIndex) && (ZSTD_read64(match3) == ZSTD_read64(ip + 1))) { - mLength = ZSTD_count(ip + 9, match3 + 8, iend) + 8; - ip++; - offset = (U32)(ip - match3); - while (((ip > anchor) & (match3 > lowest)) && (ip[-1] == match3[-1])) { - ip--; - match3--; - mLength++; - } /* catch up */ - } else { - mLength = ZSTD_count(ip + 4, match + 4, iend) + 4; - offset = (U32)(ip - match); - while (((ip > anchor) & (match > lowest)) && (ip[-1] == match[-1])) { - ip--; - match--; - mLength++; - } /* catch up */ - } - } else { - ip += ((ip - anchor) >> g_searchStrength) + 1; - continue; - } - - offset_2 = offset_1; - offset_1 = offset; - - ZSTD_storeSeq(seqStorePtr, ip - anchor, anchor, offset + ZSTD_REP_MOVE, mLength - MINMATCH); - } - - /* match found */ - ip += mLength; - anchor = ip; - - if (ip <= ilimit) { - /* Fill Table */ - hashLong[ZSTD_hashPtr(base + curr + 2, hBitsL, 8)] = hashSmall[ZSTD_hashPtr(base + curr + 2, hBitsS, mls)] = - curr + 2; /* here because curr+2 could be > iend-8 */ - hashLong[ZSTD_hashPtr(ip - 2, hBitsL, 8)] = hashSmall[ZSTD_hashPtr(ip - 2, hBitsS, mls)] = (U32)(ip - 2 - base); - - /* check immediate repcode */ - while ((ip <= ilimit) && ((offset_2 > 0) & (ZSTD_read32(ip) == ZSTD_read32(ip - offset_2)))) { - /* store sequence */ - size_t const rLength = ZSTD_count(ip + 4, ip + 4 - offset_2, iend) + 4; - { - U32 const tmpOff = offset_2; - offset_2 = offset_1; - offset_1 = tmpOff; - } /* swap offset_2 <=> offset_1 */ - hashSmall[ZSTD_hashPtr(ip, hBitsS, mls)] = (U32)(ip - base); - hashLong[ZSTD_hashPtr(ip, hBitsL, 8)] = (U32)(ip - base); - ZSTD_storeSeq(seqStorePtr, 0, anchor, 0, rLength - MINMATCH); - ip += rLength; - anchor = ip; - continue; /* faster when present ... (?) */ - } - } - } - - /* save reps for next block */ - cctx->repToConfirm[0] = offset_1 ? offset_1 : offsetSaved; - cctx->repToConfirm[1] = offset_2 ? offset_2 : offsetSaved; - - /* Last Literals */ - { - size_t const lastLLSize = iend - anchor; - memcpy(seqStorePtr->lit, anchor, lastLLSize); - seqStorePtr->lit += lastLLSize; - } -} - -static void ZSTD_compressBlock_doubleFast(ZSTD_CCtx *ctx, const void *src, size_t srcSize) -{ - const U32 mls = ctx->params.cParams.searchLength; - switch (mls) { - default: /* includes case 3 */ - case 4: ZSTD_compressBlock_doubleFast_generic(ctx, src, srcSize, 4); return; - case 5: ZSTD_compressBlock_doubleFast_generic(ctx, src, srcSize, 5); return; - case 6: ZSTD_compressBlock_doubleFast_generic(ctx, src, srcSize, 6); return; - case 7: ZSTD_compressBlock_doubleFast_generic(ctx, src, srcSize, 7); return; - } -} - -static void ZSTD_compressBlock_doubleFast_extDict_generic(ZSTD_CCtx *ctx, const void *src, size_t srcSize, const U32 mls) -{ - U32 *const hashLong = ctx->hashTable; - U32 const hBitsL = ctx->params.cParams.hashLog; - U32 *const hashSmall = ctx->chainTable; - U32 const hBitsS = ctx->params.cParams.chainLog; - seqStore_t *seqStorePtr = &(ctx->seqStore); - const BYTE *const base = ctx->base; - const BYTE *const dictBase = ctx->dictBase; - const BYTE *const istart = (const BYTE *)src; - const BYTE *ip = istart; - const BYTE *anchor = istart; - const U32 lowestIndex = ctx->lowLimit; - const BYTE *const dictStart = dictBase + lowestIndex; - const U32 dictLimit = ctx->dictLimit; - const BYTE *const lowPrefixPtr = base + dictLimit; - const BYTE *const dictEnd = dictBase + dictLimit; - const BYTE *const iend = istart + srcSize; - const BYTE *const ilimit = iend - 8; - U32 offset_1 = ctx->rep[0], offset_2 = ctx->rep[1]; - - /* Search Loop */ - while (ip < ilimit) { /* < instead of <=, because (ip+1) */ - const size_t hSmall = ZSTD_hashPtr(ip, hBitsS, mls); - const U32 matchIndex = hashSmall[hSmall]; - const BYTE *matchBase = matchIndex < dictLimit ? dictBase : base; - const BYTE *match = matchBase + matchIndex; - - const size_t hLong = ZSTD_hashPtr(ip, hBitsL, 8); - const U32 matchLongIndex = hashLong[hLong]; - const BYTE *matchLongBase = matchLongIndex < dictLimit ? dictBase : base; - const BYTE *matchLong = matchLongBase + matchLongIndex; - - const U32 curr = (U32)(ip - base); - const U32 repIndex = curr + 1 - offset_1; /* offset_1 expected <= curr +1 */ - const BYTE *repBase = repIndex < dictLimit ? dictBase : base; - const BYTE *repMatch = repBase + repIndex; - size_t mLength; - hashSmall[hSmall] = hashLong[hLong] = curr; /* update hash table */ - - if ((((U32)((dictLimit - 1) - repIndex) >= 3) /* intentional underflow */ & (repIndex > lowestIndex)) && - (ZSTD_read32(repMatch) == ZSTD_read32(ip + 1))) { - const BYTE *repMatchEnd = repIndex < dictLimit ? dictEnd : iend; - mLength = ZSTD_count_2segments(ip + 1 + 4, repMatch + 4, iend, repMatchEnd, lowPrefixPtr) + 4; - ip++; - ZSTD_storeSeq(seqStorePtr, ip - anchor, anchor, 0, mLength - MINMATCH); - } else { - if ((matchLongIndex > lowestIndex) && (ZSTD_read64(matchLong) == ZSTD_read64(ip))) { - const BYTE *matchEnd = matchLongIndex < dictLimit ? dictEnd : iend; - const BYTE *lowMatchPtr = matchLongIndex < dictLimit ? dictStart : lowPrefixPtr; - U32 offset; - mLength = ZSTD_count_2segments(ip + 8, matchLong + 8, iend, matchEnd, lowPrefixPtr) + 8; - offset = curr - matchLongIndex; - while (((ip > anchor) & (matchLong > lowMatchPtr)) && (ip[-1] == matchLong[-1])) { - ip--; - matchLong--; - mLength++; - } /* catch up */ - offset_2 = offset_1; - offset_1 = offset; - ZSTD_storeSeq(seqStorePtr, ip - anchor, anchor, offset + ZSTD_REP_MOVE, mLength - MINMATCH); - - } else if ((matchIndex > lowestIndex) && (ZSTD_read32(match) == ZSTD_read32(ip))) { - size_t const h3 = ZSTD_hashPtr(ip + 1, hBitsL, 8); - U32 const matchIndex3 = hashLong[h3]; - const BYTE *const match3Base = matchIndex3 < dictLimit ? dictBase : base; - const BYTE *match3 = match3Base + matchIndex3; - U32 offset; - hashLong[h3] = curr + 1; - if ((matchIndex3 > lowestIndex) && (ZSTD_read64(match3) == ZSTD_read64(ip + 1))) { - const BYTE *matchEnd = matchIndex3 < dictLimit ? dictEnd : iend; - const BYTE *lowMatchPtr = matchIndex3 < dictLimit ? dictStart : lowPrefixPtr; - mLength = ZSTD_count_2segments(ip + 9, match3 + 8, iend, matchEnd, lowPrefixPtr) + 8; - ip++; - offset = curr + 1 - matchIndex3; - while (((ip > anchor) & (match3 > lowMatchPtr)) && (ip[-1] == match3[-1])) { - ip--; - match3--; - mLength++; - } /* catch up */ - } else { - const BYTE *matchEnd = matchIndex < dictLimit ? dictEnd : iend; - const BYTE *lowMatchPtr = matchIndex < dictLimit ? dictStart : lowPrefixPtr; - mLength = ZSTD_count_2segments(ip + 4, match + 4, iend, matchEnd, lowPrefixPtr) + 4; - offset = curr - matchIndex; - while (((ip > anchor) & (match > lowMatchPtr)) && (ip[-1] == match[-1])) { - ip--; - match--; - mLength++; - } /* catch up */ - } - offset_2 = offset_1; - offset_1 = offset; - ZSTD_storeSeq(seqStorePtr, ip - anchor, anchor, offset + ZSTD_REP_MOVE, mLength - MINMATCH); - - } else { - ip += ((ip - anchor) >> g_searchStrength) + 1; - continue; - } - } - - /* found a match : store it */ - ip += mLength; - anchor = ip; - - if (ip <= ilimit) { - /* Fill Table */ - hashSmall[ZSTD_hashPtr(base + curr + 2, hBitsS, mls)] = curr + 2; - hashLong[ZSTD_hashPtr(base + curr + 2, hBitsL, 8)] = curr + 2; - hashSmall[ZSTD_hashPtr(ip - 2, hBitsS, mls)] = (U32)(ip - 2 - base); - hashLong[ZSTD_hashPtr(ip - 2, hBitsL, 8)] = (U32)(ip - 2 - base); - /* check immediate repcode */ - while (ip <= ilimit) { - U32 const curr2 = (U32)(ip - base); - U32 const repIndex2 = curr2 - offset_2; - const BYTE *repMatch2 = repIndex2 < dictLimit ? dictBase + repIndex2 : base + repIndex2; - if ((((U32)((dictLimit - 1) - repIndex2) >= 3) & (repIndex2 > lowestIndex)) /* intentional overflow */ - && (ZSTD_read32(repMatch2) == ZSTD_read32(ip))) { - const BYTE *const repEnd2 = repIndex2 < dictLimit ? dictEnd : iend; - size_t const repLength2 = - ZSTD_count_2segments(ip + EQUAL_READ32, repMatch2 + EQUAL_READ32, iend, repEnd2, lowPrefixPtr) + EQUAL_READ32; - U32 tmpOffset = offset_2; - offset_2 = offset_1; - offset_1 = tmpOffset; /* swap offset_2 <=> offset_1 */ - ZSTD_storeSeq(seqStorePtr, 0, anchor, 0, repLength2 - MINMATCH); - hashSmall[ZSTD_hashPtr(ip, hBitsS, mls)] = curr2; - hashLong[ZSTD_hashPtr(ip, hBitsL, 8)] = curr2; - ip += repLength2; - anchor = ip; - continue; - } - break; - } - } - } - - /* save reps for next block */ - ctx->repToConfirm[0] = offset_1; - ctx->repToConfirm[1] = offset_2; - - /* Last Literals */ - { - size_t const lastLLSize = iend - anchor; - memcpy(seqStorePtr->lit, anchor, lastLLSize); - seqStorePtr->lit += lastLLSize; - } -} - -static void ZSTD_compressBlock_doubleFast_extDict(ZSTD_CCtx *ctx, const void *src, size_t srcSize) -{ - U32 const mls = ctx->params.cParams.searchLength; - switch (mls) { - default: /* includes case 3 */ - case 4: ZSTD_compressBlock_doubleFast_extDict_generic(ctx, src, srcSize, 4); return; - case 5: ZSTD_compressBlock_doubleFast_extDict_generic(ctx, src, srcSize, 5); return; - case 6: ZSTD_compressBlock_doubleFast_extDict_generic(ctx, src, srcSize, 6); return; - case 7: ZSTD_compressBlock_doubleFast_extDict_generic(ctx, src, srcSize, 7); return; - } -} - -/*-************************************* -* Binary Tree search -***************************************/ -/** ZSTD_insertBt1() : add one or multiple positions to tree. -* ip : assumed <= iend-8 . -* @return : nb of positions added */ -static U32 ZSTD_insertBt1(ZSTD_CCtx *zc, const BYTE *const ip, const U32 mls, const BYTE *const iend, U32 nbCompares, U32 extDict) -{ - U32 *const hashTable = zc->hashTable; - U32 const hashLog = zc->params.cParams.hashLog; - size_t const h = ZSTD_hashPtr(ip, hashLog, mls); - U32 *const bt = zc->chainTable; - U32 const btLog = zc->params.cParams.chainLog - 1; - U32 const btMask = (1 << btLog) - 1; - U32 matchIndex = hashTable[h]; - size_t commonLengthSmaller = 0, commonLengthLarger = 0; - const BYTE *const base = zc->base; - const BYTE *const dictBase = zc->dictBase; - const U32 dictLimit = zc->dictLimit; - const BYTE *const dictEnd = dictBase + dictLimit; - const BYTE *const prefixStart = base + dictLimit; - const BYTE *match; - const U32 curr = (U32)(ip - base); - const U32 btLow = btMask >= curr ? 0 : curr - btMask; - U32 *smallerPtr = bt + 2 * (curr & btMask); - U32 *largerPtr = smallerPtr + 1; - U32 dummy32; /* to be nullified at the end */ - U32 const windowLow = zc->lowLimit; - U32 matchEndIdx = curr + 8; - size_t bestLength = 8; - - hashTable[h] = curr; /* Update Hash Table */ - - while (nbCompares-- && (matchIndex > windowLow)) { - U32 *const nextPtr = bt + 2 * (matchIndex & btMask); - size_t matchLength = MIN(commonLengthSmaller, commonLengthLarger); /* guaranteed minimum nb of common bytes */ - - if ((!extDict) || (matchIndex + matchLength >= dictLimit)) { - match = base + matchIndex; - if (match[matchLength] == ip[matchLength]) - matchLength += ZSTD_count(ip + matchLength + 1, match + matchLength + 1, iend) + 1; - } else { - match = dictBase + matchIndex; - matchLength += ZSTD_count_2segments(ip + matchLength, match + matchLength, iend, dictEnd, prefixStart); - if (matchIndex + matchLength >= dictLimit) - match = base + matchIndex; /* to prepare for next usage of match[matchLength] */ - } - - if (matchLength > bestLength) { - bestLength = matchLength; - if (matchLength > matchEndIdx - matchIndex) - matchEndIdx = matchIndex + (U32)matchLength; - } - - if (ip + matchLength == iend) /* equal : no way to know if inf or sup */ - break; /* drop , to guarantee consistency ; miss a bit of compression, but other solutions can corrupt the tree */ - - if (match[matchLength] < ip[matchLength]) { /* necessarily within correct buffer */ - /* match is smaller than curr */ - *smallerPtr = matchIndex; /* update smaller idx */ - commonLengthSmaller = matchLength; /* all smaller will now have at least this guaranteed common length */ - if (matchIndex <= btLow) { - smallerPtr = &dummy32; - break; - } /* beyond tree size, stop the search */ - smallerPtr = nextPtr + 1; /* new "smaller" => larger of match */ - matchIndex = nextPtr[1]; /* new matchIndex larger than previous (closer to curr) */ - } else { - /* match is larger than curr */ - *largerPtr = matchIndex; - commonLengthLarger = matchLength; - if (matchIndex <= btLow) { - largerPtr = &dummy32; - break; - } /* beyond tree size, stop the search */ - largerPtr = nextPtr; - matchIndex = nextPtr[0]; - } - } - - *smallerPtr = *largerPtr = 0; - if (bestLength > 384) - return MIN(192, (U32)(bestLength - 384)); /* speed optimization */ - if (matchEndIdx > curr + 8) - return matchEndIdx - curr - 8; - return 1; -} - -static size_t ZSTD_insertBtAndFindBestMatch(ZSTD_CCtx *zc, const BYTE *const ip, const BYTE *const iend, size_t *offsetPtr, U32 nbCompares, const U32 mls, - U32 extDict) -{ - U32 *const hashTable = zc->hashTable; - U32 const hashLog = zc->params.cParams.hashLog; - size_t const h = ZSTD_hashPtr(ip, hashLog, mls); - U32 *const bt = zc->chainTable; - U32 const btLog = zc->params.cParams.chainLog - 1; - U32 const btMask = (1 << btLog) - 1; - U32 matchIndex = hashTable[h]; - size_t commonLengthSmaller = 0, commonLengthLarger = 0; - const BYTE *const base = zc->base; - const BYTE *const dictBase = zc->dictBase; - const U32 dictLimit = zc->dictLimit; - const BYTE *const dictEnd = dictBase + dictLimit; - const BYTE *const prefixStart = base + dictLimit; - const U32 curr = (U32)(ip - base); - const U32 btLow = btMask >= curr ? 0 : curr - btMask; - const U32 windowLow = zc->lowLimit; - U32 *smallerPtr = bt + 2 * (curr & btMask); - U32 *largerPtr = bt + 2 * (curr & btMask) + 1; - U32 matchEndIdx = curr + 8; - U32 dummy32; /* to be nullified at the end */ - size_t bestLength = 0; - - hashTable[h] = curr; /* Update Hash Table */ - - while (nbCompares-- && (matchIndex > windowLow)) { - U32 *const nextPtr = bt + 2 * (matchIndex & btMask); - size_t matchLength = MIN(commonLengthSmaller, commonLengthLarger); /* guaranteed minimum nb of common bytes */ - const BYTE *match; - - if ((!extDict) || (matchIndex + matchLength >= dictLimit)) { - match = base + matchIndex; - if (match[matchLength] == ip[matchLength]) - matchLength += ZSTD_count(ip + matchLength + 1, match + matchLength + 1, iend) + 1; - } else { - match = dictBase + matchIndex; - matchLength += ZSTD_count_2segments(ip + matchLength, match + matchLength, iend, dictEnd, prefixStart); - if (matchIndex + matchLength >= dictLimit) - match = base + matchIndex; /* to prepare for next usage of match[matchLength] */ - } - - if (matchLength > bestLength) { - if (matchLength > matchEndIdx - matchIndex) - matchEndIdx = matchIndex + (U32)matchLength; - if ((4 * (int)(matchLength - bestLength)) > (int)(ZSTD_highbit32(curr - matchIndex + 1) - ZSTD_highbit32((U32)offsetPtr[0] + 1))) - bestLength = matchLength, *offsetPtr = ZSTD_REP_MOVE + curr - matchIndex; - if (ip + matchLength == iend) /* equal : no way to know if inf or sup */ - break; /* drop, to guarantee consistency (miss a little bit of compression) */ - } - - if (match[matchLength] < ip[matchLength]) { - /* match is smaller than curr */ - *smallerPtr = matchIndex; /* update smaller idx */ - commonLengthSmaller = matchLength; /* all smaller will now have at least this guaranteed common length */ - if (matchIndex <= btLow) { - smallerPtr = &dummy32; - break; - } /* beyond tree size, stop the search */ - smallerPtr = nextPtr + 1; /* new "smaller" => larger of match */ - matchIndex = nextPtr[1]; /* new matchIndex larger than previous (closer to curr) */ - } else { - /* match is larger than curr */ - *largerPtr = matchIndex; - commonLengthLarger = matchLength; - if (matchIndex <= btLow) { - largerPtr = &dummy32; - break; - } /* beyond tree size, stop the search */ - largerPtr = nextPtr; - matchIndex = nextPtr[0]; - } - } - - *smallerPtr = *largerPtr = 0; - - zc->nextToUpdate = (matchEndIdx > curr + 8) ? matchEndIdx - 8 : curr + 1; - return bestLength; -} - -static void ZSTD_updateTree(ZSTD_CCtx *zc, const BYTE *const ip, const BYTE *const iend, const U32 nbCompares, const U32 mls) -{ - const BYTE *const base = zc->base; - const U32 target = (U32)(ip - base); - U32 idx = zc->nextToUpdate; - - while (idx < target) - idx += ZSTD_insertBt1(zc, base + idx, mls, iend, nbCompares, 0); -} - -/** ZSTD_BtFindBestMatch() : Tree updater, providing best match */ -static size_t ZSTD_BtFindBestMatch(ZSTD_CCtx *zc, const BYTE *const ip, const BYTE *const iLimit, size_t *offsetPtr, const U32 maxNbAttempts, const U32 mls) -{ - if (ip < zc->base + zc->nextToUpdate) - return 0; /* skipped area */ - ZSTD_updateTree(zc, ip, iLimit, maxNbAttempts, mls); - return ZSTD_insertBtAndFindBestMatch(zc, ip, iLimit, offsetPtr, maxNbAttempts, mls, 0); -} - -static size_t ZSTD_BtFindBestMatch_selectMLS(ZSTD_CCtx *zc, /* Index table will be updated */ - const BYTE *ip, const BYTE *const iLimit, size_t *offsetPtr, const U32 maxNbAttempts, const U32 matchLengthSearch) -{ - switch (matchLengthSearch) { - default: /* includes case 3 */ - case 4: return ZSTD_BtFindBestMatch(zc, ip, iLimit, offsetPtr, maxNbAttempts, 4); - case 5: return ZSTD_BtFindBestMatch(zc, ip, iLimit, offsetPtr, maxNbAttempts, 5); - case 7: - case 6: return ZSTD_BtFindBestMatch(zc, ip, iLimit, offsetPtr, maxNbAttempts, 6); - } -} - -static void ZSTD_updateTree_extDict(ZSTD_CCtx *zc, const BYTE *const ip, const BYTE *const iend, const U32 nbCompares, const U32 mls) -{ - const BYTE *const base = zc->base; - const U32 target = (U32)(ip - base); - U32 idx = zc->nextToUpdate; - - while (idx < target) - idx += ZSTD_insertBt1(zc, base + idx, mls, iend, nbCompares, 1); -} - -/** Tree updater, providing best match */ -static size_t ZSTD_BtFindBestMatch_extDict(ZSTD_CCtx *zc, const BYTE *const ip, const BYTE *const iLimit, size_t *offsetPtr, const U32 maxNbAttempts, - const U32 mls) -{ - if (ip < zc->base + zc->nextToUpdate) - return 0; /* skipped area */ - ZSTD_updateTree_extDict(zc, ip, iLimit, maxNbAttempts, mls); - return ZSTD_insertBtAndFindBestMatch(zc, ip, iLimit, offsetPtr, maxNbAttempts, mls, 1); -} - -static size_t ZSTD_BtFindBestMatch_selectMLS_extDict(ZSTD_CCtx *zc, /* Index table will be updated */ - const BYTE *ip, const BYTE *const iLimit, size_t *offsetPtr, const U32 maxNbAttempts, - const U32 matchLengthSearch) -{ - switch (matchLengthSearch) { - default: /* includes case 3 */ - case 4: return ZSTD_BtFindBestMatch_extDict(zc, ip, iLimit, offsetPtr, maxNbAttempts, 4); - case 5: return ZSTD_BtFindBestMatch_extDict(zc, ip, iLimit, offsetPtr, maxNbAttempts, 5); - case 7: - case 6: return ZSTD_BtFindBestMatch_extDict(zc, ip, iLimit, offsetPtr, maxNbAttempts, 6); - } -} - -/* ********************************* -* Hash Chain -***********************************/ -#define NEXT_IN_CHAIN(d, mask) chainTable[(d)&mask] - -/* Update chains up to ip (excluded) - Assumption : always within prefix (i.e. not within extDict) */ -FORCE_INLINE -U32 ZSTD_insertAndFindFirstIndex(ZSTD_CCtx *zc, const BYTE *ip, U32 mls) -{ - U32 *const hashTable = zc->hashTable; - const U32 hashLog = zc->params.cParams.hashLog; - U32 *const chainTable = zc->chainTable; - const U32 chainMask = (1 << zc->params.cParams.chainLog) - 1; - const BYTE *const base = zc->base; - const U32 target = (U32)(ip - base); - U32 idx = zc->nextToUpdate; - - while (idx < target) { /* catch up */ - size_t const h = ZSTD_hashPtr(base + idx, hashLog, mls); - NEXT_IN_CHAIN(idx, chainMask) = hashTable[h]; - hashTable[h] = idx; - idx++; - } - - zc->nextToUpdate = target; - return hashTable[ZSTD_hashPtr(ip, hashLog, mls)]; -} - -/* inlining is important to hardwire a hot branch (template emulation) */ -FORCE_INLINE -size_t ZSTD_HcFindBestMatch_generic(ZSTD_CCtx *zc, /* Index table will be updated */ - const BYTE *const ip, const BYTE *const iLimit, size_t *offsetPtr, const U32 maxNbAttempts, const U32 mls, - const U32 extDict) -{ - U32 *const chainTable = zc->chainTable; - const U32 chainSize = (1 << zc->params.cParams.chainLog); - const U32 chainMask = chainSize - 1; - const BYTE *const base = zc->base; - const BYTE *const dictBase = zc->dictBase; - const U32 dictLimit = zc->dictLimit; - const BYTE *const prefixStart = base + dictLimit; - const BYTE *const dictEnd = dictBase + dictLimit; - const U32 lowLimit = zc->lowLimit; - const U32 curr = (U32)(ip - base); - const U32 minChain = curr > chainSize ? curr - chainSize : 0; - int nbAttempts = maxNbAttempts; - size_t ml = EQUAL_READ32 - 1; - - /* HC4 match finder */ - U32 matchIndex = ZSTD_insertAndFindFirstIndex(zc, ip, mls); - - for (; (matchIndex > lowLimit) & (nbAttempts > 0); nbAttempts--) { - const BYTE *match; - size_t currMl = 0; - if ((!extDict) || matchIndex >= dictLimit) { - match = base + matchIndex; - if (match[ml] == ip[ml]) /* potentially better */ - currMl = ZSTD_count(ip, match, iLimit); - } else { - match = dictBase + matchIndex; - if (ZSTD_read32(match) == ZSTD_read32(ip)) /* assumption : matchIndex <= dictLimit-4 (by table construction) */ - currMl = ZSTD_count_2segments(ip + EQUAL_READ32, match + EQUAL_READ32, iLimit, dictEnd, prefixStart) + EQUAL_READ32; - } - - /* save best solution */ - if (currMl > ml) { - ml = currMl; - *offsetPtr = curr - matchIndex + ZSTD_REP_MOVE; - if (ip + currMl == iLimit) - break; /* best possible, and avoid read overflow*/ - } - - if (matchIndex <= minChain) - break; - matchIndex = NEXT_IN_CHAIN(matchIndex, chainMask); - } - - return ml; -} - -FORCE_INLINE size_t ZSTD_HcFindBestMatch_selectMLS(ZSTD_CCtx *zc, const BYTE *ip, const BYTE *const iLimit, size_t *offsetPtr, const U32 maxNbAttempts, - const U32 matchLengthSearch) -{ - switch (matchLengthSearch) { - default: /* includes case 3 */ - case 4: return ZSTD_HcFindBestMatch_generic(zc, ip, iLimit, offsetPtr, maxNbAttempts, 4, 0); - case 5: return ZSTD_HcFindBestMatch_generic(zc, ip, iLimit, offsetPtr, maxNbAttempts, 5, 0); - case 7: - case 6: return ZSTD_HcFindBestMatch_generic(zc, ip, iLimit, offsetPtr, maxNbAttempts, 6, 0); - } -} - -FORCE_INLINE size_t ZSTD_HcFindBestMatch_extDict_selectMLS(ZSTD_CCtx *zc, const BYTE *ip, const BYTE *const iLimit, size_t *offsetPtr, const U32 maxNbAttempts, - const U32 matchLengthSearch) -{ - switch (matchLengthSearch) { - default: /* includes case 3 */ - case 4: return ZSTD_HcFindBestMatch_generic(zc, ip, iLimit, offsetPtr, maxNbAttempts, 4, 1); - case 5: return ZSTD_HcFindBestMatch_generic(zc, ip, iLimit, offsetPtr, maxNbAttempts, 5, 1); - case 7: - case 6: return ZSTD_HcFindBestMatch_generic(zc, ip, iLimit, offsetPtr, maxNbAttempts, 6, 1); - } -} - -/* ******************************* -* Common parser - lazy strategy -*********************************/ -FORCE_INLINE -void ZSTD_compressBlock_lazy_generic(ZSTD_CCtx *ctx, const void *src, size_t srcSize, const U32 searchMethod, const U32 depth) -{ - seqStore_t *seqStorePtr = &(ctx->seqStore); - const BYTE *const istart = (const BYTE *)src; - const BYTE *ip = istart; - const BYTE *anchor = istart; - const BYTE *const iend = istart + srcSize; - const BYTE *const ilimit = iend - 8; - const BYTE *const base = ctx->base + ctx->dictLimit; - - U32 const maxSearches = 1 << ctx->params.cParams.searchLog; - U32 const mls = ctx->params.cParams.searchLength; - - typedef size_t (*searchMax_f)(ZSTD_CCtx * zc, const BYTE *ip, const BYTE *iLimit, size_t *offsetPtr, U32 maxNbAttempts, U32 matchLengthSearch); - searchMax_f const searchMax = searchMethod ? ZSTD_BtFindBestMatch_selectMLS : ZSTD_HcFindBestMatch_selectMLS; - U32 offset_1 = ctx->rep[0], offset_2 = ctx->rep[1], savedOffset = 0; - - /* init */ - ip += (ip == base); - ctx->nextToUpdate3 = ctx->nextToUpdate; - { - U32 const maxRep = (U32)(ip - base); - if (offset_2 > maxRep) - savedOffset = offset_2, offset_2 = 0; - if (offset_1 > maxRep) - savedOffset = offset_1, offset_1 = 0; - } - - /* Match Loop */ - while (ip < ilimit) { - size_t matchLength = 0; - size_t offset = 0; - const BYTE *start = ip + 1; - - /* check repCode */ - if ((offset_1 > 0) & (ZSTD_read32(ip + 1) == ZSTD_read32(ip + 1 - offset_1))) { - /* repcode : we take it */ - matchLength = ZSTD_count(ip + 1 + EQUAL_READ32, ip + 1 + EQUAL_READ32 - offset_1, iend) + EQUAL_READ32; - if (depth == 0) - goto _storeSequence; - } - - /* first search (depth 0) */ - { - size_t offsetFound = 99999999; - size_t const ml2 = searchMax(ctx, ip, iend, &offsetFound, maxSearches, mls); - if (ml2 > matchLength) - matchLength = ml2, start = ip, offset = offsetFound; - } - - if (matchLength < EQUAL_READ32) { - ip += ((ip - anchor) >> g_searchStrength) + 1; /* jump faster over incompressible sections */ - continue; - } - - /* let's try to find a better solution */ - if (depth >= 1) - while (ip < ilimit) { - ip++; - if ((offset) && ((offset_1 > 0) & (ZSTD_read32(ip) == ZSTD_read32(ip - offset_1)))) { - size_t const mlRep = ZSTD_count(ip + EQUAL_READ32, ip + EQUAL_READ32 - offset_1, iend) + EQUAL_READ32; - int const gain2 = (int)(mlRep * 3); - int const gain1 = (int)(matchLength * 3 - ZSTD_highbit32((U32)offset + 1) + 1); - if ((mlRep >= EQUAL_READ32) && (gain2 > gain1)) - matchLength = mlRep, offset = 0, start = ip; - } - { - size_t offset2 = 99999999; - size_t const ml2 = searchMax(ctx, ip, iend, &offset2, maxSearches, mls); - int const gain2 = (int)(ml2 * 4 - ZSTD_highbit32((U32)offset2 + 1)); /* raw approx */ - int const gain1 = (int)(matchLength * 4 - ZSTD_highbit32((U32)offset + 1) + 4); - if ((ml2 >= EQUAL_READ32) && (gain2 > gain1)) { - matchLength = ml2, offset = offset2, start = ip; - continue; /* search a better one */ - } - } - - /* let's find an even better one */ - if ((depth == 2) && (ip < ilimit)) { - ip++; - if ((offset) && ((offset_1 > 0) & (ZSTD_read32(ip) == ZSTD_read32(ip - offset_1)))) { - size_t const ml2 = ZSTD_count(ip + EQUAL_READ32, ip + EQUAL_READ32 - offset_1, iend) + EQUAL_READ32; - int const gain2 = (int)(ml2 * 4); - int const gain1 = (int)(matchLength * 4 - ZSTD_highbit32((U32)offset + 1) + 1); - if ((ml2 >= EQUAL_READ32) && (gain2 > gain1)) - matchLength = ml2, offset = 0, start = ip; - } - { - size_t offset2 = 99999999; - size_t const ml2 = searchMax(ctx, ip, iend, &offset2, maxSearches, mls); - int const gain2 = (int)(ml2 * 4 - ZSTD_highbit32((U32)offset2 + 1)); /* raw approx */ - int const gain1 = (int)(matchLength * 4 - ZSTD_highbit32((U32)offset + 1) + 7); - if ((ml2 >= EQUAL_READ32) && (gain2 > gain1)) { - matchLength = ml2, offset = offset2, start = ip; - continue; - } - } - } - break; /* nothing found : store previous solution */ - } - - /* NOTE: - * start[-offset+ZSTD_REP_MOVE-1] is undefined behavior. - * (-offset+ZSTD_REP_MOVE-1) is unsigned, and is added to start, which - * overflows the pointer, which is undefined behavior. - */ - /* catch up */ - if (offset) { - while ((start > anchor) && (start > base + offset - ZSTD_REP_MOVE) && - (start[-1] == (start-offset+ZSTD_REP_MOVE)[-1])) /* only search for offset within prefix */ - { - start--; - matchLength++; - } - offset_2 = offset_1; - offset_1 = (U32)(offset - ZSTD_REP_MOVE); - } - - /* store sequence */ -_storeSequence: - { - size_t const litLength = start - anchor; - ZSTD_storeSeq(seqStorePtr, litLength, anchor, (U32)offset, matchLength - MINMATCH); - anchor = ip = start + matchLength; - } - - /* check immediate repcode */ - while ((ip <= ilimit) && ((offset_2 > 0) & (ZSTD_read32(ip) == ZSTD_read32(ip - offset_2)))) { - /* store sequence */ - matchLength = ZSTD_count(ip + EQUAL_READ32, ip + EQUAL_READ32 - offset_2, iend) + EQUAL_READ32; - offset = offset_2; - offset_2 = offset_1; - offset_1 = (U32)offset; /* swap repcodes */ - ZSTD_storeSeq(seqStorePtr, 0, anchor, 0, matchLength - MINMATCH); - ip += matchLength; - anchor = ip; - continue; /* faster when present ... (?) */ - } - } - - /* Save reps for next block */ - ctx->repToConfirm[0] = offset_1 ? offset_1 : savedOffset; - ctx->repToConfirm[1] = offset_2 ? offset_2 : savedOffset; - - /* Last Literals */ - { - size_t const lastLLSize = iend - anchor; - memcpy(seqStorePtr->lit, anchor, lastLLSize); - seqStorePtr->lit += lastLLSize; - } -} - -static void ZSTD_compressBlock_btlazy2(ZSTD_CCtx *ctx, const void *src, size_t srcSize) { ZSTD_compressBlock_lazy_generic(ctx, src, srcSize, 1, 2); } - -static void ZSTD_compressBlock_lazy2(ZSTD_CCtx *ctx, const void *src, size_t srcSize) { ZSTD_compressBlock_lazy_generic(ctx, src, srcSize, 0, 2); } - -static void ZSTD_compressBlock_lazy(ZSTD_CCtx *ctx, const void *src, size_t srcSize) { ZSTD_compressBlock_lazy_generic(ctx, src, srcSize, 0, 1); } - -static void ZSTD_compressBlock_greedy(ZSTD_CCtx *ctx, const void *src, size_t srcSize) { ZSTD_compressBlock_lazy_generic(ctx, src, srcSize, 0, 0); } - -FORCE_INLINE -void ZSTD_compressBlock_lazy_extDict_generic(ZSTD_CCtx *ctx, const void *src, size_t srcSize, const U32 searchMethod, const U32 depth) -{ - seqStore_t *seqStorePtr = &(ctx->seqStore); - const BYTE *const istart = (const BYTE *)src; - const BYTE *ip = istart; - const BYTE *anchor = istart; - const BYTE *const iend = istart + srcSize; - const BYTE *const ilimit = iend - 8; - const BYTE *const base = ctx->base; - const U32 dictLimit = ctx->dictLimit; - const U32 lowestIndex = ctx->lowLimit; - const BYTE *const prefixStart = base + dictLimit; - const BYTE *const dictBase = ctx->dictBase; - const BYTE *const dictEnd = dictBase + dictLimit; - const BYTE *const dictStart = dictBase + ctx->lowLimit; - - const U32 maxSearches = 1 << ctx->params.cParams.searchLog; - const U32 mls = ctx->params.cParams.searchLength; - - typedef size_t (*searchMax_f)(ZSTD_CCtx * zc, const BYTE *ip, const BYTE *iLimit, size_t *offsetPtr, U32 maxNbAttempts, U32 matchLengthSearch); - searchMax_f searchMax = searchMethod ? ZSTD_BtFindBestMatch_selectMLS_extDict : ZSTD_HcFindBestMatch_extDict_selectMLS; - - U32 offset_1 = ctx->rep[0], offset_2 = ctx->rep[1]; - - /* init */ - ctx->nextToUpdate3 = ctx->nextToUpdate; - ip += (ip == prefixStart); - - /* Match Loop */ - while (ip < ilimit) { - size_t matchLength = 0; - size_t offset = 0; - const BYTE *start = ip + 1; - U32 curr = (U32)(ip - base); - - /* check repCode */ - { - const U32 repIndex = (U32)(curr + 1 - offset_1); - const BYTE *const repBase = repIndex < dictLimit ? dictBase : base; - const BYTE *const repMatch = repBase + repIndex; - if (((U32)((dictLimit - 1) - repIndex) >= 3) & (repIndex > lowestIndex)) /* intentional overflow */ - if (ZSTD_read32(ip + 1) == ZSTD_read32(repMatch)) { - /* repcode detected we should take it */ - const BYTE *const repEnd = repIndex < dictLimit ? dictEnd : iend; - matchLength = - ZSTD_count_2segments(ip + 1 + EQUAL_READ32, repMatch + EQUAL_READ32, iend, repEnd, prefixStart) + EQUAL_READ32; - if (depth == 0) - goto _storeSequence; - } - } - - /* first search (depth 0) */ - { - size_t offsetFound = 99999999; - size_t const ml2 = searchMax(ctx, ip, iend, &offsetFound, maxSearches, mls); - if (ml2 > matchLength) - matchLength = ml2, start = ip, offset = offsetFound; - } - - if (matchLength < EQUAL_READ32) { - ip += ((ip - anchor) >> g_searchStrength) + 1; /* jump faster over incompressible sections */ - continue; - } - - /* let's try to find a better solution */ - if (depth >= 1) - while (ip < ilimit) { - ip++; - curr++; - /* check repCode */ - if (offset) { - const U32 repIndex = (U32)(curr - offset_1); - const BYTE *const repBase = repIndex < dictLimit ? dictBase : base; - const BYTE *const repMatch = repBase + repIndex; - if (((U32)((dictLimit - 1) - repIndex) >= 3) & (repIndex > lowestIndex)) /* intentional overflow */ - if (ZSTD_read32(ip) == ZSTD_read32(repMatch)) { - /* repcode detected */ - const BYTE *const repEnd = repIndex < dictLimit ? dictEnd : iend; - size_t const repLength = - ZSTD_count_2segments(ip + EQUAL_READ32, repMatch + EQUAL_READ32, iend, repEnd, prefixStart) + - EQUAL_READ32; - int const gain2 = (int)(repLength * 3); - int const gain1 = (int)(matchLength * 3 - ZSTD_highbit32((U32)offset + 1) + 1); - if ((repLength >= EQUAL_READ32) && (gain2 > gain1)) - matchLength = repLength, offset = 0, start = ip; - } - } - - /* search match, depth 1 */ - { - size_t offset2 = 99999999; - size_t const ml2 = searchMax(ctx, ip, iend, &offset2, maxSearches, mls); - int const gain2 = (int)(ml2 * 4 - ZSTD_highbit32((U32)offset2 + 1)); /* raw approx */ - int const gain1 = (int)(matchLength * 4 - ZSTD_highbit32((U32)offset + 1) + 4); - if ((ml2 >= EQUAL_READ32) && (gain2 > gain1)) { - matchLength = ml2, offset = offset2, start = ip; - continue; /* search a better one */ - } - } - - /* let's find an even better one */ - if ((depth == 2) && (ip < ilimit)) { - ip++; - curr++; - /* check repCode */ - if (offset) { - const U32 repIndex = (U32)(curr - offset_1); - const BYTE *const repBase = repIndex < dictLimit ? dictBase : base; - const BYTE *const repMatch = repBase + repIndex; - if (((U32)((dictLimit - 1) - repIndex) >= 3) & (repIndex > lowestIndex)) /* intentional overflow */ - if (ZSTD_read32(ip) == ZSTD_read32(repMatch)) { - /* repcode detected */ - const BYTE *const repEnd = repIndex < dictLimit ? dictEnd : iend; - size_t repLength = ZSTD_count_2segments(ip + EQUAL_READ32, repMatch + EQUAL_READ32, iend, - repEnd, prefixStart) + - EQUAL_READ32; - int gain2 = (int)(repLength * 4); - int gain1 = (int)(matchLength * 4 - ZSTD_highbit32((U32)offset + 1) + 1); - if ((repLength >= EQUAL_READ32) && (gain2 > gain1)) - matchLength = repLength, offset = 0, start = ip; - } - } - - /* search match, depth 2 */ - { - size_t offset2 = 99999999; - size_t const ml2 = searchMax(ctx, ip, iend, &offset2, maxSearches, mls); - int const gain2 = (int)(ml2 * 4 - ZSTD_highbit32((U32)offset2 + 1)); /* raw approx */ - int const gain1 = (int)(matchLength * 4 - ZSTD_highbit32((U32)offset + 1) + 7); - if ((ml2 >= EQUAL_READ32) && (gain2 > gain1)) { - matchLength = ml2, offset = offset2, start = ip; - continue; - } - } - } - break; /* nothing found : store previous solution */ - } - - /* catch up */ - if (offset) { - U32 const matchIndex = (U32)((start - base) - (offset - ZSTD_REP_MOVE)); - const BYTE *match = (matchIndex < dictLimit) ? dictBase + matchIndex : base + matchIndex; - const BYTE *const mStart = (matchIndex < dictLimit) ? dictStart : prefixStart; - while ((start > anchor) && (match > mStart) && (start[-1] == match[-1])) { - start--; - match--; - matchLength++; - } /* catch up */ - offset_2 = offset_1; - offset_1 = (U32)(offset - ZSTD_REP_MOVE); - } - - /* store sequence */ - _storeSequence : { - size_t const litLength = start - anchor; - ZSTD_storeSeq(seqStorePtr, litLength, anchor, (U32)offset, matchLength - MINMATCH); - anchor = ip = start + matchLength; - } - - /* check immediate repcode */ - while (ip <= ilimit) { - const U32 repIndex = (U32)((ip - base) - offset_2); - const BYTE *const repBase = repIndex < dictLimit ? dictBase : base; - const BYTE *const repMatch = repBase + repIndex; - if (((U32)((dictLimit - 1) - repIndex) >= 3) & (repIndex > lowestIndex)) /* intentional overflow */ - if (ZSTD_read32(ip) == ZSTD_read32(repMatch)) { - /* repcode detected we should take it */ - const BYTE *const repEnd = repIndex < dictLimit ? dictEnd : iend; - matchLength = - ZSTD_count_2segments(ip + EQUAL_READ32, repMatch + EQUAL_READ32, iend, repEnd, prefixStart) + EQUAL_READ32; - offset = offset_2; - offset_2 = offset_1; - offset_1 = (U32)offset; /* swap offset history */ - ZSTD_storeSeq(seqStorePtr, 0, anchor, 0, matchLength - MINMATCH); - ip += matchLength; - anchor = ip; - continue; /* faster when present ... (?) */ - } - break; - } - } - - /* Save reps for next block */ - ctx->repToConfirm[0] = offset_1; - ctx->repToConfirm[1] = offset_2; - - /* Last Literals */ - { - size_t const lastLLSize = iend - anchor; - memcpy(seqStorePtr->lit, anchor, lastLLSize); - seqStorePtr->lit += lastLLSize; - } -} - -void ZSTD_compressBlock_greedy_extDict(ZSTD_CCtx *ctx, const void *src, size_t srcSize) { ZSTD_compressBlock_lazy_extDict_generic(ctx, src, srcSize, 0, 0); } - -static void ZSTD_compressBlock_lazy_extDict(ZSTD_CCtx *ctx, const void *src, size_t srcSize) -{ - ZSTD_compressBlock_lazy_extDict_generic(ctx, src, srcSize, 0, 1); -} - -static void ZSTD_compressBlock_lazy2_extDict(ZSTD_CCtx *ctx, const void *src, size_t srcSize) -{ - ZSTD_compressBlock_lazy_extDict_generic(ctx, src, srcSize, 0, 2); -} - -static void ZSTD_compressBlock_btlazy2_extDict(ZSTD_CCtx *ctx, const void *src, size_t srcSize) -{ - ZSTD_compressBlock_lazy_extDict_generic(ctx, src, srcSize, 1, 2); -} - -/* The optimal parser */ -#include "zstd_opt.h" - -static void ZSTD_compressBlock_btopt(ZSTD_CCtx *ctx, const void *src, size_t srcSize) -{ -#ifdef ZSTD_OPT_H_91842398743 - ZSTD_compressBlock_opt_generic(ctx, src, srcSize, 0); -#else - (void)ctx; - (void)src; - (void)srcSize; - return; -#endif -} - -static void ZSTD_compressBlock_btopt2(ZSTD_CCtx *ctx, const void *src, size_t srcSize) -{ -#ifdef ZSTD_OPT_H_91842398743 - ZSTD_compressBlock_opt_generic(ctx, src, srcSize, 1); -#else - (void)ctx; - (void)src; - (void)srcSize; - return; -#endif -} - -static void ZSTD_compressBlock_btopt_extDict(ZSTD_CCtx *ctx, const void *src, size_t srcSize) -{ -#ifdef ZSTD_OPT_H_91842398743 - ZSTD_compressBlock_opt_extDict_generic(ctx, src, srcSize, 0); -#else - (void)ctx; - (void)src; - (void)srcSize; - return; -#endif -} - -static void ZSTD_compressBlock_btopt2_extDict(ZSTD_CCtx *ctx, const void *src, size_t srcSize) -{ -#ifdef ZSTD_OPT_H_91842398743 - ZSTD_compressBlock_opt_extDict_generic(ctx, src, srcSize, 1); -#else - (void)ctx; - (void)src; - (void)srcSize; - return; -#endif -} - -typedef void (*ZSTD_blockCompressor)(ZSTD_CCtx *ctx, const void *src, size_t srcSize); - -static ZSTD_blockCompressor ZSTD_selectBlockCompressor(ZSTD_strategy strat, int extDict) -{ - static const ZSTD_blockCompressor blockCompressor[2][8] = { - {ZSTD_compressBlock_fast, ZSTD_compressBlock_doubleFast, ZSTD_compressBlock_greedy, ZSTD_compressBlock_lazy, ZSTD_compressBlock_lazy2, - ZSTD_compressBlock_btlazy2, ZSTD_compressBlock_btopt, ZSTD_compressBlock_btopt2}, - {ZSTD_compressBlock_fast_extDict, ZSTD_compressBlock_doubleFast_extDict, ZSTD_compressBlock_greedy_extDict, ZSTD_compressBlock_lazy_extDict, - ZSTD_compressBlock_lazy2_extDict, ZSTD_compressBlock_btlazy2_extDict, ZSTD_compressBlock_btopt_extDict, ZSTD_compressBlock_btopt2_extDict}}; - - return blockCompressor[extDict][(U32)strat]; -} - -static size_t ZSTD_compressBlock_internal(ZSTD_CCtx *zc, void *dst, size_t dstCapacity, const void *src, size_t srcSize) -{ - ZSTD_blockCompressor const blockCompressor = ZSTD_selectBlockCompressor(zc->params.cParams.strategy, zc->lowLimit < zc->dictLimit); - const BYTE *const base = zc->base; - const BYTE *const istart = (const BYTE *)src; - const U32 curr = (U32)(istart - base); - if (srcSize < MIN_CBLOCK_SIZE + ZSTD_blockHeaderSize + 1) - return 0; /* don't even attempt compression below a certain srcSize */ - ZSTD_resetSeqStore(&(zc->seqStore)); - if (curr > zc->nextToUpdate + 384) - zc->nextToUpdate = curr - MIN(192, (U32)(curr - zc->nextToUpdate - 384)); /* update tree not updated after finding very long rep matches */ - blockCompressor(zc, src, srcSize); - return ZSTD_compressSequences(zc, dst, dstCapacity, srcSize); -} - -/*! ZSTD_compress_generic() : -* Compress a chunk of data into one or multiple blocks. -* All blocks will be terminated, all input will be consumed. -* Function will issue an error if there is not enough `dstCapacity` to hold the compressed content. -* Frame is supposed already started (header already produced) -* @return : compressed size, or an error code -*/ -static size_t ZSTD_compress_generic(ZSTD_CCtx *cctx, void *dst, size_t dstCapacity, const void *src, size_t srcSize, U32 lastFrameChunk) -{ - size_t blockSize = cctx->blockSize; - size_t remaining = srcSize; - const BYTE *ip = (const BYTE *)src; - BYTE *const ostart = (BYTE *)dst; - BYTE *op = ostart; - U32 const maxDist = 1 << cctx->params.cParams.windowLog; - - if (cctx->params.fParams.checksumFlag && srcSize) - xxh64_update(&cctx->xxhState, src, srcSize); - - while (remaining) { - U32 const lastBlock = lastFrameChunk & (blockSize >= remaining); - size_t cSize; - - if (dstCapacity < ZSTD_blockHeaderSize + MIN_CBLOCK_SIZE) - return ERROR(dstSize_tooSmall); /* not enough space to store compressed block */ - if (remaining < blockSize) - blockSize = remaining; - - /* preemptive overflow correction */ - if (cctx->lowLimit > (3U << 29)) { - U32 const cycleMask = (1 << ZSTD_cycleLog(cctx->params.cParams.hashLog, cctx->params.cParams.strategy)) - 1; - U32 const curr = (U32)(ip - cctx->base); - U32 const newCurr = (curr & cycleMask) + (1 << cctx->params.cParams.windowLog); - U32 const correction = curr - newCurr; - ZSTD_STATIC_ASSERT(ZSTD_WINDOWLOG_MAX_64 <= 30); - ZSTD_reduceIndex(cctx, correction); - cctx->base += correction; - cctx->dictBase += correction; - cctx->lowLimit -= correction; - cctx->dictLimit -= correction; - if (cctx->nextToUpdate < correction) - cctx->nextToUpdate = 0; - else - cctx->nextToUpdate -= correction; - } - - if ((U32)(ip + blockSize - cctx->base) > cctx->loadedDictEnd + maxDist) { - /* enforce maxDist */ - U32 const newLowLimit = (U32)(ip + blockSize - cctx->base) - maxDist; - if (cctx->lowLimit < newLowLimit) - cctx->lowLimit = newLowLimit; - if (cctx->dictLimit < cctx->lowLimit) - cctx->dictLimit = cctx->lowLimit; - } - - cSize = ZSTD_compressBlock_internal(cctx, op + ZSTD_blockHeaderSize, dstCapacity - ZSTD_blockHeaderSize, ip, blockSize); - if (ZSTD_isError(cSize)) - return cSize; - - if (cSize == 0) { /* block is not compressible */ - U32 const cBlockHeader24 = lastBlock + (((U32)bt_raw) << 1) + (U32)(blockSize << 3); - if (blockSize + ZSTD_blockHeaderSize > dstCapacity) - return ERROR(dstSize_tooSmall); - ZSTD_writeLE32(op, cBlockHeader24); /* no pb, 4th byte will be overwritten */ - memcpy(op + ZSTD_blockHeaderSize, ip, blockSize); - cSize = ZSTD_blockHeaderSize + blockSize; - } else { - U32 const cBlockHeader24 = lastBlock + (((U32)bt_compressed) << 1) + (U32)(cSize << 3); - ZSTD_writeLE24(op, cBlockHeader24); - cSize += ZSTD_blockHeaderSize; - } - - remaining -= blockSize; - dstCapacity -= cSize; - ip += blockSize; - op += cSize; - } - - if (lastFrameChunk && (op > ostart)) - cctx->stage = ZSTDcs_ending; - return op - ostart; -} - -static size_t ZSTD_writeFrameHeader(void *dst, size_t dstCapacity, ZSTD_parameters params, U64 pledgedSrcSize, U32 dictID) -{ - BYTE *const op = (BYTE *)dst; - U32 const dictIDSizeCode = (dictID > 0) + (dictID >= 256) + (dictID >= 65536); /* 0-3 */ - U32 const checksumFlag = params.fParams.checksumFlag > 0; - U32 const windowSize = 1U << params.cParams.windowLog; - U32 const singleSegment = params.fParams.contentSizeFlag && (windowSize >= pledgedSrcSize); - BYTE const windowLogByte = (BYTE)((params.cParams.windowLog - ZSTD_WINDOWLOG_ABSOLUTEMIN) << 3); - U32 const fcsCode = - params.fParams.contentSizeFlag ? (pledgedSrcSize >= 256) + (pledgedSrcSize >= 65536 + 256) + (pledgedSrcSize >= 0xFFFFFFFFU) : 0; /* 0-3 */ - BYTE const frameHeaderDecriptionByte = (BYTE)(dictIDSizeCode + (checksumFlag << 2) + (singleSegment << 5) + (fcsCode << 6)); - size_t pos; - - if (dstCapacity < ZSTD_frameHeaderSize_max) - return ERROR(dstSize_tooSmall); - - ZSTD_writeLE32(dst, ZSTD_MAGICNUMBER); - op[4] = frameHeaderDecriptionByte; - pos = 5; - if (!singleSegment) - op[pos++] = windowLogByte; - switch (dictIDSizeCode) { - default: /* impossible */ - case 0: break; - case 1: - op[pos] = (BYTE)(dictID); - pos++; - break; - case 2: - ZSTD_writeLE16(op + pos, (U16)dictID); - pos += 2; - break; - case 3: - ZSTD_writeLE32(op + pos, dictID); - pos += 4; - break; - } - switch (fcsCode) { - default: /* impossible */ - case 0: - if (singleSegment) - op[pos++] = (BYTE)(pledgedSrcSize); - break; - case 1: - ZSTD_writeLE16(op + pos, (U16)(pledgedSrcSize - 256)); - pos += 2; - break; - case 2: - ZSTD_writeLE32(op + pos, (U32)(pledgedSrcSize)); - pos += 4; - break; - case 3: - ZSTD_writeLE64(op + pos, (U64)(pledgedSrcSize)); - pos += 8; - break; - } - return pos; -} - -static size_t ZSTD_compressContinue_internal(ZSTD_CCtx *cctx, void *dst, size_t dstCapacity, const void *src, size_t srcSize, U32 frame, U32 lastFrameChunk) -{ - const BYTE *const ip = (const BYTE *)src; - size_t fhSize = 0; - - if (cctx->stage == ZSTDcs_created) - return ERROR(stage_wrong); /* missing init (ZSTD_compressBegin) */ - - if (frame && (cctx->stage == ZSTDcs_init)) { - fhSize = ZSTD_writeFrameHeader(dst, dstCapacity, cctx->params, cctx->frameContentSize, cctx->dictID); - if (ZSTD_isError(fhSize)) - return fhSize; - dstCapacity -= fhSize; - dst = (char *)dst + fhSize; - cctx->stage = ZSTDcs_ongoing; - } - - /* Check if blocks follow each other */ - if (src != cctx->nextSrc) { - /* not contiguous */ - ptrdiff_t const delta = cctx->nextSrc - ip; - cctx->lowLimit = cctx->dictLimit; - cctx->dictLimit = (U32)(cctx->nextSrc - cctx->base); - cctx->dictBase = cctx->base; - cctx->base -= delta; - cctx->nextToUpdate = cctx->dictLimit; - if (cctx->dictLimit - cctx->lowLimit < HASH_READ_SIZE) - cctx->lowLimit = cctx->dictLimit; /* too small extDict */ - } - - /* if input and dictionary overlap : reduce dictionary (area presumed modified by input) */ - if ((ip + srcSize > cctx->dictBase + cctx->lowLimit) & (ip < cctx->dictBase + cctx->dictLimit)) { - ptrdiff_t const highInputIdx = (ip + srcSize) - cctx->dictBase; - U32 const lowLimitMax = (highInputIdx > (ptrdiff_t)cctx->dictLimit) ? cctx->dictLimit : (U32)highInputIdx; - cctx->lowLimit = lowLimitMax; - } - - cctx->nextSrc = ip + srcSize; - - if (srcSize) { - size_t const cSize = frame ? ZSTD_compress_generic(cctx, dst, dstCapacity, src, srcSize, lastFrameChunk) - : ZSTD_compressBlock_internal(cctx, dst, dstCapacity, src, srcSize); - if (ZSTD_isError(cSize)) - return cSize; - return cSize + fhSize; - } else - return fhSize; -} - -size_t ZSTD_compressContinue(ZSTD_CCtx *cctx, void *dst, size_t dstCapacity, const void *src, size_t srcSize) -{ - return ZSTD_compressContinue_internal(cctx, dst, dstCapacity, src, srcSize, 1, 0); -} - -size_t ZSTD_getBlockSizeMax(ZSTD_CCtx *cctx) { return MIN(ZSTD_BLOCKSIZE_ABSOLUTEMAX, 1 << cctx->params.cParams.windowLog); } - -size_t ZSTD_compressBlock(ZSTD_CCtx *cctx, void *dst, size_t dstCapacity, const void *src, size_t srcSize) -{ - size_t const blockSizeMax = ZSTD_getBlockSizeMax(cctx); - if (srcSize > blockSizeMax) - return ERROR(srcSize_wrong); - return ZSTD_compressContinue_internal(cctx, dst, dstCapacity, src, srcSize, 0, 0); -} - -/*! ZSTD_loadDictionaryContent() : - * @return : 0, or an error code - */ -static size_t ZSTD_loadDictionaryContent(ZSTD_CCtx *zc, const void *src, size_t srcSize) -{ - const BYTE *const ip = (const BYTE *)src; - const BYTE *const iend = ip + srcSize; - - /* input becomes curr prefix */ - zc->lowLimit = zc->dictLimit; - zc->dictLimit = (U32)(zc->nextSrc - zc->base); - zc->dictBase = zc->base; - zc->base += ip - zc->nextSrc; - zc->nextToUpdate = zc->dictLimit; - zc->loadedDictEnd = zc->forceWindow ? 0 : (U32)(iend - zc->base); - - zc->nextSrc = iend; - if (srcSize <= HASH_READ_SIZE) - return 0; - - switch (zc->params.cParams.strategy) { - case ZSTD_fast: ZSTD_fillHashTable(zc, iend, zc->params.cParams.searchLength); break; - - case ZSTD_dfast: ZSTD_fillDoubleHashTable(zc, iend, zc->params.cParams.searchLength); break; - - case ZSTD_greedy: - case ZSTD_lazy: - case ZSTD_lazy2: - if (srcSize >= HASH_READ_SIZE) - ZSTD_insertAndFindFirstIndex(zc, iend - HASH_READ_SIZE, zc->params.cParams.searchLength); - break; - - case ZSTD_btlazy2: - case ZSTD_btopt: - case ZSTD_btopt2: - if (srcSize >= HASH_READ_SIZE) - ZSTD_updateTree(zc, iend - HASH_READ_SIZE, iend, 1 << zc->params.cParams.searchLog, zc->params.cParams.searchLength); - break; - - default: - return ERROR(GENERIC); /* strategy doesn't exist; impossible */ - } - - zc->nextToUpdate = (U32)(iend - zc->base); - return 0; -} - -/* Dictionaries that assign zero probability to symbols that show up causes problems - when FSE encoding. Refuse dictionaries that assign zero probability to symbols - that we may encounter during compression. - NOTE: This behavior is not standard and could be improved in the future. */ -static size_t ZSTD_checkDictNCount(short *normalizedCounter, unsigned dictMaxSymbolValue, unsigned maxSymbolValue) -{ - U32 s; - if (dictMaxSymbolValue < maxSymbolValue) - return ERROR(dictionary_corrupted); - for (s = 0; s <= maxSymbolValue; ++s) { - if (normalizedCounter[s] == 0) - return ERROR(dictionary_corrupted); - } - return 0; -} - -/* Dictionary format : - * See : - * https://github.com/facebook/zstd/blob/master/doc/zstd_compression_format.md#dictionary-format - */ -/*! ZSTD_loadZstdDictionary() : - * @return : 0, or an error code - * assumptions : magic number supposed already checked - * dictSize supposed > 8 - */ -static size_t ZSTD_loadZstdDictionary(ZSTD_CCtx *cctx, const void *dict, size_t dictSize) -{ - const BYTE *dictPtr = (const BYTE *)dict; - const BYTE *const dictEnd = dictPtr + dictSize; - short offcodeNCount[MaxOff + 1]; - unsigned offcodeMaxValue = MaxOff; - - dictPtr += 4; /* skip magic number */ - cctx->dictID = cctx->params.fParams.noDictIDFlag ? 0 : ZSTD_readLE32(dictPtr); - dictPtr += 4; - - { - size_t const hufHeaderSize = HUF_readCTable_wksp(cctx->hufTable, 255, dictPtr, dictEnd - dictPtr, cctx->tmpCounters, sizeof(cctx->tmpCounters)); - if (HUF_isError(hufHeaderSize)) - return ERROR(dictionary_corrupted); - dictPtr += hufHeaderSize; - } - - { - unsigned offcodeLog; - size_t const offcodeHeaderSize = FSE_readNCount(offcodeNCount, &offcodeMaxValue, &offcodeLog, dictPtr, dictEnd - dictPtr); - if (FSE_isError(offcodeHeaderSize)) - return ERROR(dictionary_corrupted); - if (offcodeLog > OffFSELog) - return ERROR(dictionary_corrupted); - /* Defer checking offcodeMaxValue because we need to know the size of the dictionary content */ - CHECK_E(FSE_buildCTable_wksp(cctx->offcodeCTable, offcodeNCount, offcodeMaxValue, offcodeLog, cctx->tmpCounters, sizeof(cctx->tmpCounters)), - dictionary_corrupted); - dictPtr += offcodeHeaderSize; - } - - { - short matchlengthNCount[MaxML + 1]; - unsigned matchlengthMaxValue = MaxML, matchlengthLog; - size_t const matchlengthHeaderSize = FSE_readNCount(matchlengthNCount, &matchlengthMaxValue, &matchlengthLog, dictPtr, dictEnd - dictPtr); - if (FSE_isError(matchlengthHeaderSize)) - return ERROR(dictionary_corrupted); - if (matchlengthLog > MLFSELog) - return ERROR(dictionary_corrupted); - /* Every match length code must have non-zero probability */ - CHECK_F(ZSTD_checkDictNCount(matchlengthNCount, matchlengthMaxValue, MaxML)); - CHECK_E( - FSE_buildCTable_wksp(cctx->matchlengthCTable, matchlengthNCount, matchlengthMaxValue, matchlengthLog, cctx->tmpCounters, sizeof(cctx->tmpCounters)), - dictionary_corrupted); - dictPtr += matchlengthHeaderSize; - } - - { - short litlengthNCount[MaxLL + 1]; - unsigned litlengthMaxValue = MaxLL, litlengthLog; - size_t const litlengthHeaderSize = FSE_readNCount(litlengthNCount, &litlengthMaxValue, &litlengthLog, dictPtr, dictEnd - dictPtr); - if (FSE_isError(litlengthHeaderSize)) - return ERROR(dictionary_corrupted); - if (litlengthLog > LLFSELog) - return ERROR(dictionary_corrupted); - /* Every literal length code must have non-zero probability */ - CHECK_F(ZSTD_checkDictNCount(litlengthNCount, litlengthMaxValue, MaxLL)); - CHECK_E(FSE_buildCTable_wksp(cctx->litlengthCTable, litlengthNCount, litlengthMaxValue, litlengthLog, cctx->tmpCounters, sizeof(cctx->tmpCounters)), - dictionary_corrupted); - dictPtr += litlengthHeaderSize; - } - - if (dictPtr + 12 > dictEnd) - return ERROR(dictionary_corrupted); - cctx->rep[0] = ZSTD_readLE32(dictPtr + 0); - cctx->rep[1] = ZSTD_readLE32(dictPtr + 4); - cctx->rep[2] = ZSTD_readLE32(dictPtr + 8); - dictPtr += 12; - - { - size_t const dictContentSize = (size_t)(dictEnd - dictPtr); - U32 offcodeMax = MaxOff; - if (dictContentSize <= ((U32)-1) - 128 KB) { - U32 const maxOffset = (U32)dictContentSize + 128 KB; /* The maximum offset that must be supported */ - offcodeMax = ZSTD_highbit32(maxOffset); /* Calculate minimum offset code required to represent maxOffset */ - } - /* All offset values <= dictContentSize + 128 KB must be representable */ - CHECK_F(ZSTD_checkDictNCount(offcodeNCount, offcodeMaxValue, MIN(offcodeMax, MaxOff))); - /* All repCodes must be <= dictContentSize and != 0*/ - { - U32 u; - for (u = 0; u < 3; u++) { - if (cctx->rep[u] == 0) - return ERROR(dictionary_corrupted); - if (cctx->rep[u] > dictContentSize) - return ERROR(dictionary_corrupted); - } - } - - cctx->flagStaticTables = 1; - cctx->flagStaticHufTable = HUF_repeat_valid; - return ZSTD_loadDictionaryContent(cctx, dictPtr, dictContentSize); - } -} - -/** ZSTD_compress_insertDictionary() : -* @return : 0, or an error code */ -static size_t ZSTD_compress_insertDictionary(ZSTD_CCtx *cctx, const void *dict, size_t dictSize) -{ - if ((dict == NULL) || (dictSize <= 8)) - return 0; - - /* dict as pure content */ - if ((ZSTD_readLE32(dict) != ZSTD_DICT_MAGIC) || (cctx->forceRawDict)) - return ZSTD_loadDictionaryContent(cctx, dict, dictSize); - - /* dict as zstd dictionary */ - return ZSTD_loadZstdDictionary(cctx, dict, dictSize); -} - -/*! ZSTD_compressBegin_internal() : -* @return : 0, or an error code */ -static size_t ZSTD_compressBegin_internal(ZSTD_CCtx *cctx, const void *dict, size_t dictSize, ZSTD_parameters params, U64 pledgedSrcSize) -{ - ZSTD_compResetPolicy_e const crp = dictSize ? ZSTDcrp_fullReset : ZSTDcrp_continue; - CHECK_F(ZSTD_resetCCtx_advanced(cctx, params, pledgedSrcSize, crp)); - return ZSTD_compress_insertDictionary(cctx, dict, dictSize); -} - -/*! ZSTD_compressBegin_advanced() : -* @return : 0, or an error code */ -size_t ZSTD_compressBegin_advanced(ZSTD_CCtx *cctx, const void *dict, size_t dictSize, ZSTD_parameters params, unsigned long long pledgedSrcSize) -{ - /* compression parameters verification and optimization */ - CHECK_F(ZSTD_checkCParams(params.cParams)); - return ZSTD_compressBegin_internal(cctx, dict, dictSize, params, pledgedSrcSize); -} - -size_t ZSTD_compressBegin_usingDict(ZSTD_CCtx *cctx, const void *dict, size_t dictSize, int compressionLevel) -{ - ZSTD_parameters const params = ZSTD_getParams(compressionLevel, 0, dictSize); - return ZSTD_compressBegin_internal(cctx, dict, dictSize, params, 0); -} - -size_t ZSTD_compressBegin(ZSTD_CCtx *cctx, int compressionLevel) { return ZSTD_compressBegin_usingDict(cctx, NULL, 0, compressionLevel); } - -/*! ZSTD_writeEpilogue() : -* Ends a frame. -* @return : nb of bytes written into dst (or an error code) */ -static size_t ZSTD_writeEpilogue(ZSTD_CCtx *cctx, void *dst, size_t dstCapacity) -{ - BYTE *const ostart = (BYTE *)dst; - BYTE *op = ostart; - size_t fhSize = 0; - - if (cctx->stage == ZSTDcs_created) - return ERROR(stage_wrong); /* init missing */ - - /* special case : empty frame */ - if (cctx->stage == ZSTDcs_init) { - fhSize = ZSTD_writeFrameHeader(dst, dstCapacity, cctx->params, 0, 0); - if (ZSTD_isError(fhSize)) - return fhSize; - dstCapacity -= fhSize; - op += fhSize; - cctx->stage = ZSTDcs_ongoing; - } - - if (cctx->stage != ZSTDcs_ending) { - /* write one last empty block, make it the "last" block */ - U32 const cBlockHeader24 = 1 /* last block */ + (((U32)bt_raw) << 1) + 0; - if (dstCapacity < 4) - return ERROR(dstSize_tooSmall); - ZSTD_writeLE32(op, cBlockHeader24); - op += ZSTD_blockHeaderSize; - dstCapacity -= ZSTD_blockHeaderSize; - } - - if (cctx->params.fParams.checksumFlag) { - U32 const checksum = (U32)xxh64_digest(&cctx->xxhState); - if (dstCapacity < 4) - return ERROR(dstSize_tooSmall); - ZSTD_writeLE32(op, checksum); - op += 4; - } - - cctx->stage = ZSTDcs_created; /* return to "created but no init" status */ - return op - ostart; -} - -size_t ZSTD_compressEnd(ZSTD_CCtx *cctx, void *dst, size_t dstCapacity, const void *src, size_t srcSize) -{ - size_t endResult; - size_t const cSize = ZSTD_compressContinue_internal(cctx, dst, dstCapacity, src, srcSize, 1, 1); - if (ZSTD_isError(cSize)) - return cSize; - endResult = ZSTD_writeEpilogue(cctx, (char *)dst + cSize, dstCapacity - cSize); - if (ZSTD_isError(endResult)) - return endResult; - return cSize + endResult; -} - -static size_t ZSTD_compress_internal(ZSTD_CCtx *cctx, void *dst, size_t dstCapacity, const void *src, size_t srcSize, const void *dict, size_t dictSize, - ZSTD_parameters params) -{ - CHECK_F(ZSTD_compressBegin_internal(cctx, dict, dictSize, params, srcSize)); - return ZSTD_compressEnd(cctx, dst, dstCapacity, src, srcSize); -} - -size_t ZSTD_compress_usingDict(ZSTD_CCtx *ctx, void *dst, size_t dstCapacity, const void *src, size_t srcSize, const void *dict, size_t dictSize, - ZSTD_parameters params) -{ - return ZSTD_compress_internal(ctx, dst, dstCapacity, src, srcSize, dict, dictSize, params); -} - -size_t ZSTD_compressCCtx(ZSTD_CCtx *ctx, void *dst, size_t dstCapacity, const void *src, size_t srcSize, ZSTD_parameters params) -{ - return ZSTD_compress_internal(ctx, dst, dstCapacity, src, srcSize, NULL, 0, params); -} - -/* ===== Dictionary API ===== */ - -struct ZSTD_CDict_s { - void *dictBuffer; - const void *dictContent; - size_t dictContentSize; - ZSTD_CCtx *refContext; -}; /* typedef'd tp ZSTD_CDict within "zstd.h" */ - -size_t ZSTD_CDictWorkspaceBound(ZSTD_compressionParameters cParams) { return ZSTD_CCtxWorkspaceBound(cParams) + ZSTD_ALIGN(sizeof(ZSTD_CDict)); } - -static ZSTD_CDict *ZSTD_createCDict_advanced(const void *dictBuffer, size_t dictSize, unsigned byReference, ZSTD_parameters params, ZSTD_customMem customMem) -{ - if (!customMem.customAlloc || !customMem.customFree) - return NULL; - - { - ZSTD_CDict *const cdict = (ZSTD_CDict *)ZSTD_malloc(sizeof(ZSTD_CDict), customMem); - ZSTD_CCtx *const cctx = ZSTD_createCCtx_advanced(customMem); - - if (!cdict || !cctx) { - ZSTD_free(cdict, customMem); - ZSTD_freeCCtx(cctx); - return NULL; - } - - if ((byReference) || (!dictBuffer) || (!dictSize)) { - cdict->dictBuffer = NULL; - cdict->dictContent = dictBuffer; - } else { - void *const internalBuffer = ZSTD_malloc(dictSize, customMem); - if (!internalBuffer) { - ZSTD_free(cctx, customMem); - ZSTD_free(cdict, customMem); - return NULL; - } - memcpy(internalBuffer, dictBuffer, dictSize); - cdict->dictBuffer = internalBuffer; - cdict->dictContent = internalBuffer; - } - - { - size_t const errorCode = ZSTD_compressBegin_advanced(cctx, cdict->dictContent, dictSize, params, 0); - if (ZSTD_isError(errorCode)) { - ZSTD_free(cdict->dictBuffer, customMem); - ZSTD_free(cdict, customMem); - ZSTD_freeCCtx(cctx); - return NULL; - } - } - - cdict->refContext = cctx; - cdict->dictContentSize = dictSize; - return cdict; - } -} - -ZSTD_CDict *ZSTD_initCDict(const void *dict, size_t dictSize, ZSTD_parameters params, void *workspace, size_t workspaceSize) -{ - ZSTD_customMem const stackMem = ZSTD_initStack(workspace, workspaceSize); - return ZSTD_createCDict_advanced(dict, dictSize, 1, params, stackMem); -} - -size_t ZSTD_freeCDict(ZSTD_CDict *cdict) -{ - if (cdict == NULL) - return 0; /* support free on NULL */ - { - ZSTD_customMem const cMem = cdict->refContext->customMem; - ZSTD_freeCCtx(cdict->refContext); - ZSTD_free(cdict->dictBuffer, cMem); - ZSTD_free(cdict, cMem); - return 0; - } -} - -static ZSTD_parameters ZSTD_getParamsFromCDict(const ZSTD_CDict *cdict) { return ZSTD_getParamsFromCCtx(cdict->refContext); } - -size_t ZSTD_compressBegin_usingCDict(ZSTD_CCtx *cctx, const ZSTD_CDict *cdict, unsigned long long pledgedSrcSize) -{ - if (cdict->dictContentSize) - CHECK_F(ZSTD_copyCCtx(cctx, cdict->refContext, pledgedSrcSize)) - else { - ZSTD_parameters params = cdict->refContext->params; - params.fParams.contentSizeFlag = (pledgedSrcSize > 0); - CHECK_F(ZSTD_compressBegin_advanced(cctx, NULL, 0, params, pledgedSrcSize)); - } - return 0; -} - -/*! ZSTD_compress_usingCDict() : -* Compression using a digested Dictionary. -* Faster startup than ZSTD_compress_usingDict(), recommended when same dictionary is used multiple times. -* Note that compression level is decided during dictionary creation */ -size_t ZSTD_compress_usingCDict(ZSTD_CCtx *cctx, void *dst, size_t dstCapacity, const void *src, size_t srcSize, const ZSTD_CDict *cdict) -{ - CHECK_F(ZSTD_compressBegin_usingCDict(cctx, cdict, srcSize)); - - if (cdict->refContext->params.fParams.contentSizeFlag == 1) { - cctx->params.fParams.contentSizeFlag = 1; - cctx->frameContentSize = srcSize; - } else { - cctx->params.fParams.contentSizeFlag = 0; - } - - return ZSTD_compressEnd(cctx, dst, dstCapacity, src, srcSize); -} - -/* ****************************************************************** -* Streaming -********************************************************************/ - -typedef enum { zcss_init, zcss_load, zcss_flush, zcss_final } ZSTD_cStreamStage; - -struct ZSTD_CStream_s { - ZSTD_CCtx *cctx; - ZSTD_CDict *cdictLocal; - const ZSTD_CDict *cdict; - char *inBuff; - size_t inBuffSize; - size_t inToCompress; - size_t inBuffPos; - size_t inBuffTarget; - size_t blockSize; - char *outBuff; - size_t outBuffSize; - size_t outBuffContentSize; - size_t outBuffFlushedSize; - ZSTD_cStreamStage stage; - U32 checksum; - U32 frameEnded; - U64 pledgedSrcSize; - U64 inputProcessed; - ZSTD_parameters params; - ZSTD_customMem customMem; -}; /* typedef'd to ZSTD_CStream within "zstd.h" */ - -size_t ZSTD_CStreamWorkspaceBound(ZSTD_compressionParameters cParams) -{ - size_t const inBuffSize = (size_t)1 << cParams.windowLog; - size_t const blockSize = MIN(ZSTD_BLOCKSIZE_ABSOLUTEMAX, inBuffSize); - size_t const outBuffSize = ZSTD_compressBound(blockSize) + 1; - - return ZSTD_CCtxWorkspaceBound(cParams) + ZSTD_ALIGN(sizeof(ZSTD_CStream)) + ZSTD_ALIGN(inBuffSize) + ZSTD_ALIGN(outBuffSize); -} - -ZSTD_CStream *ZSTD_createCStream_advanced(ZSTD_customMem customMem) -{ - ZSTD_CStream *zcs; - - if (!customMem.customAlloc || !customMem.customFree) - return NULL; - - zcs = (ZSTD_CStream *)ZSTD_malloc(sizeof(ZSTD_CStream), customMem); - if (zcs == NULL) - return NULL; - memset(zcs, 0, sizeof(ZSTD_CStream)); - memcpy(&zcs->customMem, &customMem, sizeof(ZSTD_customMem)); - zcs->cctx = ZSTD_createCCtx_advanced(customMem); - if (zcs->cctx == NULL) { - ZSTD_freeCStream(zcs); - return NULL; - } - return zcs; -} - -size_t ZSTD_freeCStream(ZSTD_CStream *zcs) -{ - if (zcs == NULL) - return 0; /* support free on NULL */ - { - ZSTD_customMem const cMem = zcs->customMem; - ZSTD_freeCCtx(zcs->cctx); - zcs->cctx = NULL; - ZSTD_freeCDict(zcs->cdictLocal); - zcs->cdictLocal = NULL; - ZSTD_free(zcs->inBuff, cMem); - zcs->inBuff = NULL; - ZSTD_free(zcs->outBuff, cMem); - zcs->outBuff = NULL; - ZSTD_free(zcs, cMem); - return 0; - } -} - -/*====== Initialization ======*/ - -size_t ZSTD_CStreamInSize(void) { return ZSTD_BLOCKSIZE_ABSOLUTEMAX; } -size_t ZSTD_CStreamOutSize(void) { return ZSTD_compressBound(ZSTD_BLOCKSIZE_ABSOLUTEMAX) + ZSTD_blockHeaderSize + 4 /* 32-bits hash */; } - -static size_t ZSTD_resetCStream_internal(ZSTD_CStream *zcs, unsigned long long pledgedSrcSize) -{ - if (zcs->inBuffSize == 0) - return ERROR(stage_wrong); /* zcs has not been init at least once => can't reset */ - - if (zcs->cdict) - CHECK_F(ZSTD_compressBegin_usingCDict(zcs->cctx, zcs->cdict, pledgedSrcSize)) - else - CHECK_F(ZSTD_compressBegin_advanced(zcs->cctx, NULL, 0, zcs->params, pledgedSrcSize)); - - zcs->inToCompress = 0; - zcs->inBuffPos = 0; - zcs->inBuffTarget = zcs->blockSize; - zcs->outBuffContentSize = zcs->outBuffFlushedSize = 0; - zcs->stage = zcss_load; - zcs->frameEnded = 0; - zcs->pledgedSrcSize = pledgedSrcSize; - zcs->inputProcessed = 0; - return 0; /* ready to go */ -} - -size_t ZSTD_resetCStream(ZSTD_CStream *zcs, unsigned long long pledgedSrcSize) -{ - - zcs->params.fParams.contentSizeFlag = (pledgedSrcSize > 0); - - return ZSTD_resetCStream_internal(zcs, pledgedSrcSize); -} - -static size_t ZSTD_initCStream_advanced(ZSTD_CStream *zcs, const void *dict, size_t dictSize, ZSTD_parameters params, unsigned long long pledgedSrcSize) -{ - /* allocate buffers */ - { - size_t const neededInBuffSize = (size_t)1 << params.cParams.windowLog; - if (zcs->inBuffSize < neededInBuffSize) { - zcs->inBuffSize = neededInBuffSize; - ZSTD_free(zcs->inBuff, zcs->customMem); - zcs->inBuff = (char *)ZSTD_malloc(neededInBuffSize, zcs->customMem); - if (zcs->inBuff == NULL) - return ERROR(memory_allocation); - } - zcs->blockSize = MIN(ZSTD_BLOCKSIZE_ABSOLUTEMAX, neededInBuffSize); - } - if (zcs->outBuffSize < ZSTD_compressBound(zcs->blockSize) + 1) { - zcs->outBuffSize = ZSTD_compressBound(zcs->blockSize) + 1; - ZSTD_free(zcs->outBuff, zcs->customMem); - zcs->outBuff = (char *)ZSTD_malloc(zcs->outBuffSize, zcs->customMem); - if (zcs->outBuff == NULL) - return ERROR(memory_allocation); - } - - if (dict && dictSize >= 8) { - ZSTD_freeCDict(zcs->cdictLocal); - zcs->cdictLocal = ZSTD_createCDict_advanced(dict, dictSize, 0, params, zcs->customMem); - if (zcs->cdictLocal == NULL) - return ERROR(memory_allocation); - zcs->cdict = zcs->cdictLocal; - } else - zcs->cdict = NULL; - - zcs->checksum = params.fParams.checksumFlag > 0; - zcs->params = params; - - return ZSTD_resetCStream_internal(zcs, pledgedSrcSize); -} - -ZSTD_CStream *ZSTD_initCStream(ZSTD_parameters params, unsigned long long pledgedSrcSize, void *workspace, size_t workspaceSize) -{ - ZSTD_customMem const stackMem = ZSTD_initStack(workspace, workspaceSize); - ZSTD_CStream *const zcs = ZSTD_createCStream_advanced(stackMem); - if (zcs) { - size_t const code = ZSTD_initCStream_advanced(zcs, NULL, 0, params, pledgedSrcSize); - if (ZSTD_isError(code)) { - return NULL; - } - } - return zcs; -} - -ZSTD_CStream *ZSTD_initCStream_usingCDict(const ZSTD_CDict *cdict, unsigned long long pledgedSrcSize, void *workspace, size_t workspaceSize) -{ - ZSTD_parameters const params = ZSTD_getParamsFromCDict(cdict); - ZSTD_CStream *const zcs = ZSTD_initCStream(params, pledgedSrcSize, workspace, workspaceSize); - if (zcs) { - zcs->cdict = cdict; - if (ZSTD_isError(ZSTD_resetCStream_internal(zcs, pledgedSrcSize))) { - return NULL; - } - } - return zcs; -} - -/*====== Compression ======*/ - -typedef enum { zsf_gather, zsf_flush, zsf_end } ZSTD_flush_e; - -ZSTD_STATIC size_t ZSTD_limitCopy(void *dst, size_t dstCapacity, const void *src, size_t srcSize) -{ - size_t const length = MIN(dstCapacity, srcSize); - memcpy(dst, src, length); - return length; -} - -static size_t ZSTD_compressStream_generic(ZSTD_CStream *zcs, void *dst, size_t *dstCapacityPtr, const void *src, size_t *srcSizePtr, ZSTD_flush_e const flush) -{ - U32 someMoreWork = 1; - const char *const istart = (const char *)src; - const char *const iend = istart + *srcSizePtr; - const char *ip = istart; - char *const ostart = (char *)dst; - char *const oend = ostart + *dstCapacityPtr; - char *op = ostart; - - while (someMoreWork) { - switch (zcs->stage) { - case zcss_init: - return ERROR(init_missing); /* call ZBUFF_compressInit() first ! */ - - case zcss_load: - /* complete inBuffer */ - { - size_t const toLoad = zcs->inBuffTarget - zcs->inBuffPos; - size_t const loaded = ZSTD_limitCopy(zcs->inBuff + zcs->inBuffPos, toLoad, ip, iend - ip); - zcs->inBuffPos += loaded; - ip += loaded; - if ((zcs->inBuffPos == zcs->inToCompress) || (!flush && (toLoad != loaded))) { - someMoreWork = 0; - break; /* not enough input to get a full block : stop there, wait for more */ - } - } - /* compress curr block (note : this stage cannot be stopped in the middle) */ - { - void *cDst; - size_t cSize; - size_t const iSize = zcs->inBuffPos - zcs->inToCompress; - size_t oSize = oend - op; - if (oSize >= ZSTD_compressBound(iSize)) - cDst = op; /* compress directly into output buffer (avoid flush stage) */ - else - cDst = zcs->outBuff, oSize = zcs->outBuffSize; - cSize = (flush == zsf_end) ? ZSTD_compressEnd(zcs->cctx, cDst, oSize, zcs->inBuff + zcs->inToCompress, iSize) - : ZSTD_compressContinue(zcs->cctx, cDst, oSize, zcs->inBuff + zcs->inToCompress, iSize); - if (ZSTD_isError(cSize)) - return cSize; - if (flush == zsf_end) - zcs->frameEnded = 1; - /* prepare next block */ - zcs->inBuffTarget = zcs->inBuffPos + zcs->blockSize; - if (zcs->inBuffTarget > zcs->inBuffSize) - zcs->inBuffPos = 0, zcs->inBuffTarget = zcs->blockSize; /* note : inBuffSize >= blockSize */ - zcs->inToCompress = zcs->inBuffPos; - if (cDst == op) { - op += cSize; - break; - } /* no need to flush */ - zcs->outBuffContentSize = cSize; - zcs->outBuffFlushedSize = 0; - zcs->stage = zcss_flush; /* pass-through to flush stage */ - } - fallthrough; - - case zcss_flush: { - size_t const toFlush = zcs->outBuffContentSize - zcs->outBuffFlushedSize; - size_t const flushed = ZSTD_limitCopy(op, oend - op, zcs->outBuff + zcs->outBuffFlushedSize, toFlush); - op += flushed; - zcs->outBuffFlushedSize += flushed; - if (toFlush != flushed) { - someMoreWork = 0; - break; - } /* dst too small to store flushed data : stop there */ - zcs->outBuffContentSize = zcs->outBuffFlushedSize = 0; - zcs->stage = zcss_load; - break; - } - - case zcss_final: - someMoreWork = 0; /* do nothing */ - break; - - default: - return ERROR(GENERIC); /* impossible */ - } - } - - *srcSizePtr = ip - istart; - *dstCapacityPtr = op - ostart; - zcs->inputProcessed += *srcSizePtr; - if (zcs->frameEnded) - return 0; - { - size_t hintInSize = zcs->inBuffTarget - zcs->inBuffPos; - if (hintInSize == 0) - hintInSize = zcs->blockSize; - return hintInSize; - } -} - -size_t ZSTD_compressStream(ZSTD_CStream *zcs, ZSTD_outBuffer *output, ZSTD_inBuffer *input) -{ - size_t sizeRead = input->size - input->pos; - size_t sizeWritten = output->size - output->pos; - size_t const result = - ZSTD_compressStream_generic(zcs, (char *)(output->dst) + output->pos, &sizeWritten, (const char *)(input->src) + input->pos, &sizeRead, zsf_gather); - input->pos += sizeRead; - output->pos += sizeWritten; - return result; -} - -/*====== Finalize ======*/ - -/*! ZSTD_flushStream() : -* @return : amount of data remaining to flush */ -size_t ZSTD_flushStream(ZSTD_CStream *zcs, ZSTD_outBuffer *output) -{ - size_t srcSize = 0; - size_t sizeWritten = output->size - output->pos; - size_t const result = ZSTD_compressStream_generic(zcs, (char *)(output->dst) + output->pos, &sizeWritten, &srcSize, - &srcSize, /* use a valid src address instead of NULL */ - zsf_flush); - output->pos += sizeWritten; - if (ZSTD_isError(result)) - return result; - return zcs->outBuffContentSize - zcs->outBuffFlushedSize; /* remaining to flush */ -} - -size_t ZSTD_endStream(ZSTD_CStream *zcs, ZSTD_outBuffer *output) -{ - BYTE *const ostart = (BYTE *)(output->dst) + output->pos; - BYTE *const oend = (BYTE *)(output->dst) + output->size; - BYTE *op = ostart; - - if ((zcs->pledgedSrcSize) && (zcs->inputProcessed != zcs->pledgedSrcSize)) - return ERROR(srcSize_wrong); /* pledgedSrcSize not respected */ - - if (zcs->stage != zcss_final) { - /* flush whatever remains */ - size_t srcSize = 0; - size_t sizeWritten = output->size - output->pos; - size_t const notEnded = - ZSTD_compressStream_generic(zcs, ostart, &sizeWritten, &srcSize, &srcSize, zsf_end); /* use a valid src address instead of NULL */ - size_t const remainingToFlush = zcs->outBuffContentSize - zcs->outBuffFlushedSize; - op += sizeWritten; - if (remainingToFlush) { - output->pos += sizeWritten; - return remainingToFlush + ZSTD_BLOCKHEADERSIZE /* final empty block */ + (zcs->checksum * 4); - } - /* create epilogue */ - zcs->stage = zcss_final; - zcs->outBuffContentSize = !notEnded ? 0 : ZSTD_compressEnd(zcs->cctx, zcs->outBuff, zcs->outBuffSize, NULL, - 0); /* write epilogue, including final empty block, into outBuff */ - } - - /* flush epilogue */ - { - size_t const toFlush = zcs->outBuffContentSize - zcs->outBuffFlushedSize; - size_t const flushed = ZSTD_limitCopy(op, oend - op, zcs->outBuff + zcs->outBuffFlushedSize, toFlush); - op += flushed; - zcs->outBuffFlushedSize += flushed; - output->pos += op - ostart; - if (toFlush == flushed) - zcs->stage = zcss_init; /* end reached */ - return toFlush - flushed; - } -} - -/*-===== Pre-defined compression levels =====-*/ - -#define ZSTD_DEFAULT_CLEVEL 1 -#define ZSTD_MAX_CLEVEL 22 -int ZSTD_maxCLevel(void) { return ZSTD_MAX_CLEVEL; } - -static const ZSTD_compressionParameters ZSTD_defaultCParameters[4][ZSTD_MAX_CLEVEL + 1] = { - { - /* "default" */ - /* W, C, H, S, L, TL, strat */ - {18, 12, 12, 1, 7, 16, ZSTD_fast}, /* level 0 - never used */ - {19, 13, 14, 1, 7, 16, ZSTD_fast}, /* level 1 */ - {19, 15, 16, 1, 6, 16, ZSTD_fast}, /* level 2 */ - {20, 16, 17, 1, 5, 16, ZSTD_dfast}, /* level 3.*/ - {20, 18, 18, 1, 5, 16, ZSTD_dfast}, /* level 4.*/ - {20, 15, 18, 3, 5, 16, ZSTD_greedy}, /* level 5 */ - {21, 16, 19, 2, 5, 16, ZSTD_lazy}, /* level 6 */ - {21, 17, 20, 3, 5, 16, ZSTD_lazy}, /* level 7 */ - {21, 18, 20, 3, 5, 16, ZSTD_lazy2}, /* level 8 */ - {21, 20, 20, 3, 5, 16, ZSTD_lazy2}, /* level 9 */ - {21, 19, 21, 4, 5, 16, ZSTD_lazy2}, /* level 10 */ - {22, 20, 22, 4, 5, 16, ZSTD_lazy2}, /* level 11 */ - {22, 20, 22, 5, 5, 16, ZSTD_lazy2}, /* level 12 */ - {22, 21, 22, 5, 5, 16, ZSTD_lazy2}, /* level 13 */ - {22, 21, 22, 6, 5, 16, ZSTD_lazy2}, /* level 14 */ - {22, 21, 21, 5, 5, 16, ZSTD_btlazy2}, /* level 15 */ - {23, 22, 22, 5, 5, 16, ZSTD_btlazy2}, /* level 16 */ - {23, 21, 22, 4, 5, 24, ZSTD_btopt}, /* level 17 */ - {23, 23, 22, 6, 5, 32, ZSTD_btopt}, /* level 18 */ - {23, 23, 22, 6, 3, 48, ZSTD_btopt}, /* level 19 */ - {25, 25, 23, 7, 3, 64, ZSTD_btopt2}, /* level 20 */ - {26, 26, 23, 7, 3, 256, ZSTD_btopt2}, /* level 21 */ - {27, 27, 25, 9, 3, 512, ZSTD_btopt2}, /* level 22 */ - }, - { - /* for srcSize <= 256 KB */ - /* W, C, H, S, L, T, strat */ - {0, 0, 0, 0, 0, 0, ZSTD_fast}, /* level 0 - not used */ - {18, 13, 14, 1, 6, 8, ZSTD_fast}, /* level 1 */ - {18, 14, 13, 1, 5, 8, ZSTD_dfast}, /* level 2 */ - {18, 16, 15, 1, 5, 8, ZSTD_dfast}, /* level 3 */ - {18, 15, 17, 1, 5, 8, ZSTD_greedy}, /* level 4.*/ - {18, 16, 17, 4, 5, 8, ZSTD_greedy}, /* level 5.*/ - {18, 16, 17, 3, 5, 8, ZSTD_lazy}, /* level 6.*/ - {18, 17, 17, 4, 4, 8, ZSTD_lazy}, /* level 7 */ - {18, 17, 17, 4, 4, 8, ZSTD_lazy2}, /* level 8 */ - {18, 17, 17, 5, 4, 8, ZSTD_lazy2}, /* level 9 */ - {18, 17, 17, 6, 4, 8, ZSTD_lazy2}, /* level 10 */ - {18, 18, 17, 6, 4, 8, ZSTD_lazy2}, /* level 11.*/ - {18, 18, 17, 7, 4, 8, ZSTD_lazy2}, /* level 12.*/ - {18, 19, 17, 6, 4, 8, ZSTD_btlazy2}, /* level 13 */ - {18, 18, 18, 4, 4, 16, ZSTD_btopt}, /* level 14.*/ - {18, 18, 18, 4, 3, 16, ZSTD_btopt}, /* level 15.*/ - {18, 19, 18, 6, 3, 32, ZSTD_btopt}, /* level 16.*/ - {18, 19, 18, 8, 3, 64, ZSTD_btopt}, /* level 17.*/ - {18, 19, 18, 9, 3, 128, ZSTD_btopt}, /* level 18.*/ - {18, 19, 18, 10, 3, 256, ZSTD_btopt}, /* level 19.*/ - {18, 19, 18, 11, 3, 512, ZSTD_btopt2}, /* level 20.*/ - {18, 19, 18, 12, 3, 512, ZSTD_btopt2}, /* level 21.*/ - {18, 19, 18, 13, 3, 512, ZSTD_btopt2}, /* level 22.*/ - }, - { - /* for srcSize <= 128 KB */ - /* W, C, H, S, L, T, strat */ - {17, 12, 12, 1, 7, 8, ZSTD_fast}, /* level 0 - not used */ - {17, 12, 13, 1, 6, 8, ZSTD_fast}, /* level 1 */ - {17, 13, 16, 1, 5, 8, ZSTD_fast}, /* level 2 */ - {17, 16, 16, 2, 5, 8, ZSTD_dfast}, /* level 3 */ - {17, 13, 15, 3, 4, 8, ZSTD_greedy}, /* level 4 */ - {17, 15, 17, 4, 4, 8, ZSTD_greedy}, /* level 5 */ - {17, 16, 17, 3, 4, 8, ZSTD_lazy}, /* level 6 */ - {17, 15, 17, 4, 4, 8, ZSTD_lazy2}, /* level 7 */ - {17, 17, 17, 4, 4, 8, ZSTD_lazy2}, /* level 8 */ - {17, 17, 17, 5, 4, 8, ZSTD_lazy2}, /* level 9 */ - {17, 17, 17, 6, 4, 8, ZSTD_lazy2}, /* level 10 */ - {17, 17, 17, 7, 4, 8, ZSTD_lazy2}, /* level 11 */ - {17, 17, 17, 8, 4, 8, ZSTD_lazy2}, /* level 12 */ - {17, 18, 17, 6, 4, 8, ZSTD_btlazy2}, /* level 13.*/ - {17, 17, 17, 7, 3, 8, ZSTD_btopt}, /* level 14.*/ - {17, 17, 17, 7, 3, 16, ZSTD_btopt}, /* level 15.*/ - {17, 18, 17, 7, 3, 32, ZSTD_btopt}, /* level 16.*/ - {17, 18, 17, 7, 3, 64, ZSTD_btopt}, /* level 17.*/ - {17, 18, 17, 7, 3, 256, ZSTD_btopt}, /* level 18.*/ - {17, 18, 17, 8, 3, 256, ZSTD_btopt}, /* level 19.*/ - {17, 18, 17, 9, 3, 256, ZSTD_btopt2}, /* level 20.*/ - {17, 18, 17, 10, 3, 256, ZSTD_btopt2}, /* level 21.*/ - {17, 18, 17, 11, 3, 512, ZSTD_btopt2}, /* level 22.*/ - }, - { - /* for srcSize <= 16 KB */ - /* W, C, H, S, L, T, strat */ - {14, 12, 12, 1, 7, 6, ZSTD_fast}, /* level 0 - not used */ - {14, 14, 14, 1, 6, 6, ZSTD_fast}, /* level 1 */ - {14, 14, 14, 1, 4, 6, ZSTD_fast}, /* level 2 */ - {14, 14, 14, 1, 4, 6, ZSTD_dfast}, /* level 3.*/ - {14, 14, 14, 4, 4, 6, ZSTD_greedy}, /* level 4.*/ - {14, 14, 14, 3, 4, 6, ZSTD_lazy}, /* level 5.*/ - {14, 14, 14, 4, 4, 6, ZSTD_lazy2}, /* level 6 */ - {14, 14, 14, 5, 4, 6, ZSTD_lazy2}, /* level 7 */ - {14, 14, 14, 6, 4, 6, ZSTD_lazy2}, /* level 8.*/ - {14, 15, 14, 6, 4, 6, ZSTD_btlazy2}, /* level 9.*/ - {14, 15, 14, 3, 3, 6, ZSTD_btopt}, /* level 10.*/ - {14, 15, 14, 6, 3, 8, ZSTD_btopt}, /* level 11.*/ - {14, 15, 14, 6, 3, 16, ZSTD_btopt}, /* level 12.*/ - {14, 15, 14, 6, 3, 24, ZSTD_btopt}, /* level 13.*/ - {14, 15, 15, 6, 3, 48, ZSTD_btopt}, /* level 14.*/ - {14, 15, 15, 6, 3, 64, ZSTD_btopt}, /* level 15.*/ - {14, 15, 15, 6, 3, 96, ZSTD_btopt}, /* level 16.*/ - {14, 15, 15, 6, 3, 128, ZSTD_btopt}, /* level 17.*/ - {14, 15, 15, 6, 3, 256, ZSTD_btopt}, /* level 18.*/ - {14, 15, 15, 7, 3, 256, ZSTD_btopt}, /* level 19.*/ - {14, 15, 15, 8, 3, 256, ZSTD_btopt2}, /* level 20.*/ - {14, 15, 15, 9, 3, 256, ZSTD_btopt2}, /* level 21.*/ - {14, 15, 15, 10, 3, 256, ZSTD_btopt2}, /* level 22.*/ - }, -}; - -/*! ZSTD_getCParams() : -* @return ZSTD_compressionParameters structure for a selected compression level, `srcSize` and `dictSize`. -* Size values are optional, provide 0 if not known or unused */ -ZSTD_compressionParameters ZSTD_getCParams(int compressionLevel, unsigned long long srcSize, size_t dictSize) -{ - ZSTD_compressionParameters cp; - size_t const addedSize = srcSize ? 0 : 500; - U64 const rSize = srcSize + dictSize ? srcSize + dictSize + addedSize : (U64)-1; - U32 const tableID = (rSize <= 256 KB) + (rSize <= 128 KB) + (rSize <= 16 KB); /* intentional underflow for srcSizeHint == 0 */ - if (compressionLevel <= 0) - compressionLevel = ZSTD_DEFAULT_CLEVEL; /* 0 == default; no negative compressionLevel yet */ - if (compressionLevel > ZSTD_MAX_CLEVEL) - compressionLevel = ZSTD_MAX_CLEVEL; - cp = ZSTD_defaultCParameters[tableID][compressionLevel]; - if (ZSTD_32bits()) { /* auto-correction, for 32-bits mode */ - if (cp.windowLog > ZSTD_WINDOWLOG_MAX) - cp.windowLog = ZSTD_WINDOWLOG_MAX; - if (cp.chainLog > ZSTD_CHAINLOG_MAX) - cp.chainLog = ZSTD_CHAINLOG_MAX; - if (cp.hashLog > ZSTD_HASHLOG_MAX) - cp.hashLog = ZSTD_HASHLOG_MAX; - } - cp = ZSTD_adjustCParams(cp, srcSize, dictSize); - return cp; -} - -/*! ZSTD_getParams() : -* same as ZSTD_getCParams(), but @return a `ZSTD_parameters` object (instead of `ZSTD_compressionParameters`). -* All fields of `ZSTD_frameParameters` are set to default (0) */ -ZSTD_parameters ZSTD_getParams(int compressionLevel, unsigned long long srcSize, size_t dictSize) -{ - ZSTD_parameters params; - ZSTD_compressionParameters const cParams = ZSTD_getCParams(compressionLevel, srcSize, dictSize); - memset(¶ms, 0, sizeof(params)); - params.cParams = cParams; - return params; -} - -size_t zstd_compress_bound(size_t src_size) -{ - return ZSTD_compressBound(src_size); -} -EXPORT_SYMBOL(zstd_compress_bound); - -int zstd_min_clevel(void) -{ - /* - * zstd-1.3.1 doesn't implement ZSTD_minCLevel(). - * Return 0 (default level). - */ - return 0; -} -EXPORT_SYMBOL(zstd_min_clevel); - -int zstd_max_clevel(void) -{ - return ZSTD_maxCLevel(); -} -EXPORT_SYMBOL(zstd_max_clevel); - -zstd_parameters zstd_get_params(int level, - unsigned long long estimated_src_size) -{ - return ZSTD_getParams(level, estimated_src_size, 0); -} -EXPORT_SYMBOL(zstd_get_params); - -size_t zstd_cctx_workspace_bound(const zstd_compression_parameters *cparams) -{ - return ZSTD_CCtxWorkspaceBound(*cparams); -} -EXPORT_SYMBOL(zstd_cctx_workspace_bound); - -zstd_cctx *zstd_init_cctx(void *workspace, size_t workspace_size) -{ - return ZSTD_initCCtx(workspace, workspace_size); -} -EXPORT_SYMBOL(zstd_init_cctx); - -size_t zstd_compress_cctx(zstd_cctx *cctx, void *dst, size_t dst_capacity, - const void *src, size_t src_size, const zstd_parameters *parameters) -{ - return ZSTD_compressCCtx(cctx, dst, dst_capacity, src, src_size, *parameters); -} -EXPORT_SYMBOL(zstd_compress_cctx); - -size_t zstd_cstream_workspace_bound(const zstd_compression_parameters *cparams) -{ - return ZSTD_CStreamWorkspaceBound(*cparams); -} -EXPORT_SYMBOL(zstd_cstream_workspace_bound); - -zstd_cstream *zstd_init_cstream(const zstd_parameters *parameters, - unsigned long long pledged_src_size, void *workspace, size_t workspace_size) -{ - return ZSTD_initCStream(*parameters, pledged_src_size, workspace, workspace_size); -} -EXPORT_SYMBOL(zstd_init_cstream); - -size_t zstd_reset_cstream(zstd_cstream *cstream, - unsigned long long pledged_src_size) -{ - return ZSTD_resetCStream(cstream, pledged_src_size); -} -EXPORT_SYMBOL(zstd_reset_cstream); - -size_t zstd_compress_stream(zstd_cstream *cstream, zstd_out_buffer *output, - zstd_in_buffer *input) -{ - return ZSTD_compressStream(cstream, output, input); -} -EXPORT_SYMBOL(zstd_compress_stream); - -size_t zstd_flush_stream(zstd_cstream *cstream, zstd_out_buffer *output) -{ - return ZSTD_flushStream(cstream, output); -} -EXPORT_SYMBOL(zstd_flush_stream); - -size_t zstd_end_stream(zstd_cstream *cstream, zstd_out_buffer *output) -{ - return ZSTD_endStream(cstream, output); -} -EXPORT_SYMBOL(zstd_end_stream); - -MODULE_LICENSE("Dual BSD/GPL"); -MODULE_DESCRIPTION("Zstd Compressor"); diff --git a/lib/zstd/compress/fse_compress.c b/lib/zstd/compress/fse_compress.c new file mode 100644 index 000000000000..436985b620e5 --- /dev/null +++ b/lib/zstd/compress/fse_compress.c @@ -0,0 +1,625 @@ +/* ****************************************************************** + * FSE : Finite State Entropy encoder + * Copyright (c) Yann Collet, Facebook, Inc. + * + * You can contact the author at : + * - FSE source repository : https://github.com/Cyan4973/FiniteStateEntropy + * - Public forum : https://groups.google.com/forum/#!forum/lz4c + * + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). + * You may select, at your option, one of the above-listed licenses. +****************************************************************** */ + +/* ************************************************************** +* Includes +****************************************************************/ +#include "../common/compiler.h" +#include "../common/mem.h" /* U32, U16, etc. */ +#include "../common/debug.h" /* assert, DEBUGLOG */ +#include "hist.h" /* HIST_count_wksp */ +#include "../common/bitstream.h" +#define FSE_STATIC_LINKING_ONLY +#include "../common/fse.h" +#include "../common/error_private.h" +#define ZSTD_DEPS_NEED_MALLOC +#define ZSTD_DEPS_NEED_MATH64 +#include "../common/zstd_deps.h" /* ZSTD_malloc, ZSTD_free, ZSTD_memcpy, ZSTD_memset */ + + +/* ************************************************************** +* Error Management +****************************************************************/ +#define FSE_isError ERR_isError + + +/* ************************************************************** +* Templates +****************************************************************/ +/* + designed to be included + for type-specific functions (template emulation in C) + Objective is to write these functions only once, for improved maintenance +*/ + +/* safety checks */ +#ifndef FSE_FUNCTION_EXTENSION +# error "FSE_FUNCTION_EXTENSION must be defined" +#endif +#ifndef FSE_FUNCTION_TYPE +# error "FSE_FUNCTION_TYPE must be defined" +#endif + +/* Function names */ +#define FSE_CAT(X,Y) X##Y +#define FSE_FUNCTION_NAME(X,Y) FSE_CAT(X,Y) +#define FSE_TYPE_NAME(X,Y) FSE_CAT(X,Y) + + +/* Function templates */ + +/* FSE_buildCTable_wksp() : + * Same as FSE_buildCTable(), but using an externally allocated scratch buffer (`workSpace`). + * wkspSize should be sized to handle worst case situation, which is `1<>1 : 1) ; + FSE_symbolCompressionTransform* const symbolTT = (FSE_symbolCompressionTransform*) (FSCT); + U32 const step = FSE_TABLESTEP(tableSize); + + U32* cumul = (U32*)workSpace; + FSE_FUNCTION_TYPE* tableSymbol = (FSE_FUNCTION_TYPE*)(cumul + (maxSymbolValue + 2)); + + U32 highThreshold = tableSize-1; + + if ((size_t)workSpace & 3) return ERROR(GENERIC); /* Must be 4 byte aligned */ + if (FSE_BUILD_CTABLE_WORKSPACE_SIZE(maxSymbolValue, tableLog) > wkspSize) return ERROR(tableLog_tooLarge); + /* CTable header */ + tableU16[-2] = (U16) tableLog; + tableU16[-1] = (U16) maxSymbolValue; + assert(tableLog < 16); /* required for threshold strategy to work */ + + /* For explanations on how to distribute symbol values over the table : + * http://fastcompression.blogspot.fr/2014/02/fse-distributing-symbol-values.html */ + + #ifdef __clang_analyzer__ + ZSTD_memset(tableSymbol, 0, sizeof(*tableSymbol) * tableSize); /* useless initialization, just to keep scan-build happy */ + #endif + + /* symbol start positions */ + { U32 u; + cumul[0] = 0; + for (u=1; u <= maxSymbolValue+1; u++) { + if (normalizedCounter[u-1]==-1) { /* Low proba symbol */ + cumul[u] = cumul[u-1] + 1; + tableSymbol[highThreshold--] = (FSE_FUNCTION_TYPE)(u-1); + } else { + cumul[u] = cumul[u-1] + normalizedCounter[u-1]; + } } + cumul[maxSymbolValue+1] = tableSize+1; + } + + /* Spread symbols */ + { U32 position = 0; + U32 symbol; + for (symbol=0; symbol<=maxSymbolValue; symbol++) { + int nbOccurrences; + int const freq = normalizedCounter[symbol]; + for (nbOccurrences=0; nbOccurrences highThreshold) + position = (position + step) & tableMask; /* Low proba area */ + } } + + assert(position==0); /* Must have initialized all positions */ + } + + /* Build table */ + { U32 u; for (u=0; u> 3) + 3; + return maxSymbolValue ? maxHeaderSize : FSE_NCOUNTBOUND; /* maxSymbolValue==0 ? use default */ +} + +static size_t +FSE_writeNCount_generic (void* header, size_t headerBufferSize, + const short* normalizedCounter, unsigned maxSymbolValue, unsigned tableLog, + unsigned writeIsSafe) +{ + BYTE* const ostart = (BYTE*) header; + BYTE* out = ostart; + BYTE* const oend = ostart + headerBufferSize; + int nbBits; + const int tableSize = 1 << tableLog; + int remaining; + int threshold; + U32 bitStream = 0; + int bitCount = 0; + unsigned symbol = 0; + unsigned const alphabetSize = maxSymbolValue + 1; + int previousIs0 = 0; + + /* Table Size */ + bitStream += (tableLog-FSE_MIN_TABLELOG) << bitCount; + bitCount += 4; + + /* Init */ + remaining = tableSize+1; /* +1 for extra accuracy */ + threshold = tableSize; + nbBits = tableLog+1; + + while ((symbol < alphabetSize) && (remaining>1)) { /* stops at 1 */ + if (previousIs0) { + unsigned start = symbol; + while ((symbol < alphabetSize) && !normalizedCounter[symbol]) symbol++; + if (symbol == alphabetSize) break; /* incorrect distribution */ + while (symbol >= start+24) { + start+=24; + bitStream += 0xFFFFU << bitCount; + if ((!writeIsSafe) && (out > oend-2)) + return ERROR(dstSize_tooSmall); /* Buffer overflow */ + out[0] = (BYTE) bitStream; + out[1] = (BYTE)(bitStream>>8); + out+=2; + bitStream>>=16; + } + while (symbol >= start+3) { + start+=3; + bitStream += 3 << bitCount; + bitCount += 2; + } + bitStream += (symbol-start) << bitCount; + bitCount += 2; + if (bitCount>16) { + if ((!writeIsSafe) && (out > oend - 2)) + return ERROR(dstSize_tooSmall); /* Buffer overflow */ + out[0] = (BYTE)bitStream; + out[1] = (BYTE)(bitStream>>8); + out += 2; + bitStream >>= 16; + bitCount -= 16; + } } + { int count = normalizedCounter[symbol++]; + int const max = (2*threshold-1) - remaining; + remaining -= count < 0 ? -count : count; + count++; /* +1 for extra accuracy */ + if (count>=threshold) + count += max; /* [0..max[ [max..threshold[ (...) [threshold+max 2*threshold[ */ + bitStream += count << bitCount; + bitCount += nbBits; + bitCount -= (count>=1; } + } + if (bitCount>16) { + if ((!writeIsSafe) && (out > oend - 2)) + return ERROR(dstSize_tooSmall); /* Buffer overflow */ + out[0] = (BYTE)bitStream; + out[1] = (BYTE)(bitStream>>8); + out += 2; + bitStream >>= 16; + bitCount -= 16; + } } + + if (remaining != 1) + return ERROR(GENERIC); /* incorrect normalized distribution */ + assert(symbol <= alphabetSize); + + /* flush remaining bitStream */ + if ((!writeIsSafe) && (out > oend - 2)) + return ERROR(dstSize_tooSmall); /* Buffer overflow */ + out[0] = (BYTE)bitStream; + out[1] = (BYTE)(bitStream>>8); + out+= (bitCount+7) /8; + + return (out-ostart); +} + + +size_t FSE_writeNCount (void* buffer, size_t bufferSize, + const short* normalizedCounter, unsigned maxSymbolValue, unsigned tableLog) +{ + if (tableLog > FSE_MAX_TABLELOG) return ERROR(tableLog_tooLarge); /* Unsupported */ + if (tableLog < FSE_MIN_TABLELOG) return ERROR(GENERIC); /* Unsupported */ + + if (bufferSize < FSE_NCountWriteBound(maxSymbolValue, tableLog)) + return FSE_writeNCount_generic(buffer, bufferSize, normalizedCounter, maxSymbolValue, tableLog, 0); + + return FSE_writeNCount_generic(buffer, bufferSize, normalizedCounter, maxSymbolValue, tableLog, 1 /* write in buffer is safe */); +} + + +/*-************************************************************** +* FSE Compression Code +****************************************************************/ + +FSE_CTable* FSE_createCTable (unsigned maxSymbolValue, unsigned tableLog) +{ + size_t size; + if (tableLog > FSE_TABLELOG_ABSOLUTE_MAX) tableLog = FSE_TABLELOG_ABSOLUTE_MAX; + size = FSE_CTABLE_SIZE_U32 (tableLog, maxSymbolValue) * sizeof(U32); + return (FSE_CTable*)ZSTD_malloc(size); +} + +void FSE_freeCTable (FSE_CTable* ct) { ZSTD_free(ct); } + +/* provides the minimum logSize to safely represent a distribution */ +static unsigned FSE_minTableLog(size_t srcSize, unsigned maxSymbolValue) +{ + U32 minBitsSrc = BIT_highbit32((U32)(srcSize)) + 1; + U32 minBitsSymbols = BIT_highbit32(maxSymbolValue) + 2; + U32 minBits = minBitsSrc < minBitsSymbols ? minBitsSrc : minBitsSymbols; + assert(srcSize > 1); /* Not supported, RLE should be used instead */ + return minBits; +} + +unsigned FSE_optimalTableLog_internal(unsigned maxTableLog, size_t srcSize, unsigned maxSymbolValue, unsigned minus) +{ + U32 maxBitsSrc = BIT_highbit32((U32)(srcSize - 1)) - minus; + U32 tableLog = maxTableLog; + U32 minBits = FSE_minTableLog(srcSize, maxSymbolValue); + assert(srcSize > 1); /* Not supported, RLE should be used instead */ + if (tableLog==0) tableLog = FSE_DEFAULT_TABLELOG; + if (maxBitsSrc < tableLog) tableLog = maxBitsSrc; /* Accuracy can be reduced */ + if (minBits > tableLog) tableLog = minBits; /* Need a minimum to safely represent all symbol values */ + if (tableLog < FSE_MIN_TABLELOG) tableLog = FSE_MIN_TABLELOG; + if (tableLog > FSE_MAX_TABLELOG) tableLog = FSE_MAX_TABLELOG; + return tableLog; +} + +unsigned FSE_optimalTableLog(unsigned maxTableLog, size_t srcSize, unsigned maxSymbolValue) +{ + return FSE_optimalTableLog_internal(maxTableLog, srcSize, maxSymbolValue, 2); +} + +/* Secondary normalization method. + To be used when primary method fails. */ + +static size_t FSE_normalizeM2(short* norm, U32 tableLog, const unsigned* count, size_t total, U32 maxSymbolValue, short lowProbCount) +{ + short const NOT_YET_ASSIGNED = -2; + U32 s; + U32 distributed = 0; + U32 ToDistribute; + + /* Init */ + U32 const lowThreshold = (U32)(total >> tableLog); + U32 lowOne = (U32)((total * 3) >> (tableLog + 1)); + + for (s=0; s<=maxSymbolValue; s++) { + if (count[s] == 0) { + norm[s]=0; + continue; + } + if (count[s] <= lowThreshold) { + norm[s] = lowProbCount; + distributed++; + total -= count[s]; + continue; + } + if (count[s] <= lowOne) { + norm[s] = 1; + distributed++; + total -= count[s]; + continue; + } + + norm[s]=NOT_YET_ASSIGNED; + } + ToDistribute = (1 << tableLog) - distributed; + + if (ToDistribute == 0) + return 0; + + if ((total / ToDistribute) > lowOne) { + /* risk of rounding to zero */ + lowOne = (U32)((total * 3) / (ToDistribute * 2)); + for (s=0; s<=maxSymbolValue; s++) { + if ((norm[s] == NOT_YET_ASSIGNED) && (count[s] <= lowOne)) { + norm[s] = 1; + distributed++; + total -= count[s]; + continue; + } } + ToDistribute = (1 << tableLog) - distributed; + } + + if (distributed == maxSymbolValue+1) { + /* all values are pretty poor; + probably incompressible data (should have already been detected); + find max, then give all remaining points to max */ + U32 maxV = 0, maxC = 0; + for (s=0; s<=maxSymbolValue; s++) + if (count[s] > maxC) { maxV=s; maxC=count[s]; } + norm[maxV] += (short)ToDistribute; + return 0; + } + + if (total == 0) { + /* all of the symbols were low enough for the lowOne or lowThreshold */ + for (s=0; ToDistribute > 0; s = (s+1)%(maxSymbolValue+1)) + if (norm[s] > 0) { ToDistribute--; norm[s]++; } + return 0; + } + + { U64 const vStepLog = 62 - tableLog; + U64 const mid = (1ULL << (vStepLog-1)) - 1; + U64 const rStep = ZSTD_div64((((U64)1<> vStepLog); + U32 const sEnd = (U32)(end >> vStepLog); + U32 const weight = sEnd - sStart; + if (weight < 1) + return ERROR(GENERIC); + norm[s] = (short)weight; + tmpTotal = end; + } } } + + return 0; +} + +size_t FSE_normalizeCount (short* normalizedCounter, unsigned tableLog, + const unsigned* count, size_t total, + unsigned maxSymbolValue, unsigned useLowProbCount) +{ + /* Sanity checks */ + if (tableLog==0) tableLog = FSE_DEFAULT_TABLELOG; + if (tableLog < FSE_MIN_TABLELOG) return ERROR(GENERIC); /* Unsupported size */ + if (tableLog > FSE_MAX_TABLELOG) return ERROR(tableLog_tooLarge); /* Unsupported size */ + if (tableLog < FSE_minTableLog(total, maxSymbolValue)) return ERROR(GENERIC); /* Too small tableLog, compression potentially impossible */ + + { static U32 const rtbTable[] = { 0, 473195, 504333, 520860, 550000, 700000, 750000, 830000 }; + short const lowProbCount = useLowProbCount ? -1 : 1; + U64 const scale = 62 - tableLog; + U64 const step = ZSTD_div64((U64)1<<62, (U32)total); /* <== here, one division ! */ + U64 const vStep = 1ULL<<(scale-20); + int stillToDistribute = 1<> tableLog); + + for (s=0; s<=maxSymbolValue; s++) { + if (count[s] == total) return 0; /* rle special case */ + if (count[s] == 0) { normalizedCounter[s]=0; continue; } + if (count[s] <= lowThreshold) { + normalizedCounter[s] = lowProbCount; + stillToDistribute--; + } else { + short proba = (short)((count[s]*step) >> scale); + if (proba<8) { + U64 restToBeat = vStep * rtbTable[proba]; + proba += (count[s]*step) - ((U64)proba< restToBeat; + } + if (proba > largestP) { largestP=proba; largest=s; } + normalizedCounter[s] = proba; + stillToDistribute -= proba; + } } + if (-stillToDistribute >= (normalizedCounter[largest] >> 1)) { + /* corner case, need another normalization method */ + size_t const errorCode = FSE_normalizeM2(normalizedCounter, tableLog, count, total, maxSymbolValue, lowProbCount); + if (FSE_isError(errorCode)) return errorCode; + } + else normalizedCounter[largest] += (short)stillToDistribute; + } + +#if 0 + { /* Print Table (debug) */ + U32 s; + U32 nTotal = 0; + for (s=0; s<=maxSymbolValue; s++) + RAWLOG(2, "%3i: %4i \n", s, normalizedCounter[s]); + for (s=0; s<=maxSymbolValue; s++) + nTotal += abs(normalizedCounter[s]); + if (nTotal != (1U<>1); /* assumption : tableLog >= 1 */ + FSE_symbolCompressionTransform* const symbolTT = (FSE_symbolCompressionTransform*) (FSCT); + unsigned s; + + /* Sanity checks */ + if (nbBits < 1) return ERROR(GENERIC); /* min size */ + + /* header */ + tableU16[-2] = (U16) nbBits; + tableU16[-1] = (U16) maxSymbolValue; + + /* Build table */ + for (s=0; s FSE_MAX_TABLELOG*4+7 ) && (srcSize & 2)) { /* test bit 2 */ + FSE_encodeSymbol(&bitC, &CState2, *--ip); + FSE_encodeSymbol(&bitC, &CState1, *--ip); + FSE_FLUSHBITS(&bitC); + } + + /* 2 or 4 encoding per loop */ + while ( ip>istart ) { + + FSE_encodeSymbol(&bitC, &CState2, *--ip); + + if (sizeof(bitC.bitContainer)*8 < FSE_MAX_TABLELOG*2+7 ) /* this test must be static */ + FSE_FLUSHBITS(&bitC); + + FSE_encodeSymbol(&bitC, &CState1, *--ip); + + if (sizeof(bitC.bitContainer)*8 > FSE_MAX_TABLELOG*4+7 ) { /* this test must be static */ + FSE_encodeSymbol(&bitC, &CState2, *--ip); + FSE_encodeSymbol(&bitC, &CState1, *--ip); + } + + FSE_FLUSHBITS(&bitC); + } + + FSE_flushCState(&bitC, &CState2); + FSE_flushCState(&bitC, &CState1); + return BIT_closeCStream(&bitC); +} + +size_t FSE_compress_usingCTable (void* dst, size_t dstSize, + const void* src, size_t srcSize, + const FSE_CTable* ct) +{ + unsigned const fast = (dstSize >= FSE_BLOCKBOUND(srcSize)); + + if (fast) + return FSE_compress_usingCTable_generic(dst, dstSize, src, srcSize, ct, 1); + else + return FSE_compress_usingCTable_generic(dst, dstSize, src, srcSize, ct, 0); +} + + +size_t FSE_compressBound(size_t size) { return FSE_COMPRESSBOUND(size); } + + +#endif /* FSE_COMMONDEFS_ONLY */ diff --git a/lib/zstd/compress/hist.c b/lib/zstd/compress/hist.c new file mode 100644 index 000000000000..3ddc6dfb6894 --- /dev/null +++ b/lib/zstd/compress/hist.c @@ -0,0 +1,165 @@ +/* ****************************************************************** + * hist : Histogram functions + * part of Finite State Entropy project + * Copyright (c) Yann Collet, Facebook, Inc. + * + * You can contact the author at : + * - FSE source repository : https://github.com/Cyan4973/FiniteStateEntropy + * - Public forum : https://groups.google.com/forum/#!forum/lz4c + * + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). + * You may select, at your option, one of the above-listed licenses. +****************************************************************** */ + +/* --- dependencies --- */ +#include "../common/mem.h" /* U32, BYTE, etc. */ +#include "../common/debug.h" /* assert, DEBUGLOG */ +#include "../common/error_private.h" /* ERROR */ +#include "hist.h" + + +/* --- Error management --- */ +unsigned HIST_isError(size_t code) { return ERR_isError(code); } + +/*-************************************************************** + * Histogram functions + ****************************************************************/ +unsigned HIST_count_simple(unsigned* count, unsigned* maxSymbolValuePtr, + const void* src, size_t srcSize) +{ + const BYTE* ip = (const BYTE*)src; + const BYTE* const end = ip + srcSize; + unsigned maxSymbolValue = *maxSymbolValuePtr; + unsigned largestCount=0; + + ZSTD_memset(count, 0, (maxSymbolValue+1) * sizeof(*count)); + if (srcSize==0) { *maxSymbolValuePtr = 0; return 0; } + + while (ip largestCount) largestCount = count[s]; + } + + return largestCount; +} + +typedef enum { trustInput, checkMaxSymbolValue } HIST_checkInput_e; + +/* HIST_count_parallel_wksp() : + * store histogram into 4 intermediate tables, recombined at the end. + * this design makes better use of OoO cpus, + * and is noticeably faster when some values are heavily repeated. + * But it needs some additional workspace for intermediate tables. + * `workSpace` must be a U32 table of size >= HIST_WKSP_SIZE_U32. + * @return : largest histogram frequency, + * or an error code (notably when histogram's alphabet is larger than *maxSymbolValuePtr) */ +static size_t HIST_count_parallel_wksp( + unsigned* count, unsigned* maxSymbolValuePtr, + const void* source, size_t sourceSize, + HIST_checkInput_e check, + U32* const workSpace) +{ + const BYTE* ip = (const BYTE*)source; + const BYTE* const iend = ip+sourceSize; + size_t const countSize = (*maxSymbolValuePtr + 1) * sizeof(*count); + unsigned max=0; + U32* const Counting1 = workSpace; + U32* const Counting2 = Counting1 + 256; + U32* const Counting3 = Counting2 + 256; + U32* const Counting4 = Counting3 + 256; + + /* safety checks */ + assert(*maxSymbolValuePtr <= 255); + if (!sourceSize) { + ZSTD_memset(count, 0, countSize); + *maxSymbolValuePtr = 0; + return 0; + } + ZSTD_memset(workSpace, 0, 4*256*sizeof(unsigned)); + + /* by stripes of 16 bytes */ + { U32 cached = MEM_read32(ip); ip += 4; + while (ip < iend-15) { + U32 c = cached; cached = MEM_read32(ip); ip += 4; + Counting1[(BYTE) c ]++; + Counting2[(BYTE)(c>>8) ]++; + Counting3[(BYTE)(c>>16)]++; + Counting4[ c>>24 ]++; + c = cached; cached = MEM_read32(ip); ip += 4; + Counting1[(BYTE) c ]++; + Counting2[(BYTE)(c>>8) ]++; + Counting3[(BYTE)(c>>16)]++; + Counting4[ c>>24 ]++; + c = cached; cached = MEM_read32(ip); ip += 4; + Counting1[(BYTE) c ]++; + Counting2[(BYTE)(c>>8) ]++; + Counting3[(BYTE)(c>>16)]++; + Counting4[ c>>24 ]++; + c = cached; cached = MEM_read32(ip); ip += 4; + Counting1[(BYTE) c ]++; + Counting2[(BYTE)(c>>8) ]++; + Counting3[(BYTE)(c>>16)]++; + Counting4[ c>>24 ]++; + } + ip-=4; + } + + /* finish last symbols */ + while (ip max) max = Counting1[s]; + } } + + { unsigned maxSymbolValue = 255; + while (!Counting1[maxSymbolValue]) maxSymbolValue--; + if (check && maxSymbolValue > *maxSymbolValuePtr) return ERROR(maxSymbolValue_tooSmall); + *maxSymbolValuePtr = maxSymbolValue; + ZSTD_memmove(count, Counting1, countSize); /* in case count & Counting1 are overlapping */ + } + return (size_t)max; +} + +/* HIST_countFast_wksp() : + * Same as HIST_countFast(), but using an externally provided scratch buffer. + * `workSpace` is a writable buffer which must be 4-bytes aligned, + * `workSpaceSize` must be >= HIST_WKSP_SIZE + */ +size_t HIST_countFast_wksp(unsigned* count, unsigned* maxSymbolValuePtr, + const void* source, size_t sourceSize, + void* workSpace, size_t workSpaceSize) +{ + if (sourceSize < 1500) /* heuristic threshold */ + return HIST_count_simple(count, maxSymbolValuePtr, source, sourceSize); + if ((size_t)workSpace & 3) return ERROR(GENERIC); /* must be aligned on 4-bytes boundaries */ + if (workSpaceSize < HIST_WKSP_SIZE) return ERROR(workSpace_tooSmall); + return HIST_count_parallel_wksp(count, maxSymbolValuePtr, source, sourceSize, trustInput, (U32*)workSpace); +} + +/* HIST_count_wksp() : + * Same as HIST_count(), but using an externally provided scratch buffer. + * `workSpace` size must be table of >= HIST_WKSP_SIZE_U32 unsigned */ +size_t HIST_count_wksp(unsigned* count, unsigned* maxSymbolValuePtr, + const void* source, size_t sourceSize, + void* workSpace, size_t workSpaceSize) +{ + if ((size_t)workSpace & 3) return ERROR(GENERIC); /* must be aligned on 4-bytes boundaries */ + if (workSpaceSize < HIST_WKSP_SIZE) return ERROR(workSpace_tooSmall); + if (*maxSymbolValuePtr < 255) + return HIST_count_parallel_wksp(count, maxSymbolValuePtr, source, sourceSize, checkMaxSymbolValue, (U32*)workSpace); + *maxSymbolValuePtr = 255; + return HIST_countFast_wksp(count, maxSymbolValuePtr, source, sourceSize, workSpace, workSpaceSize); +} + diff --git a/lib/zstd/compress/hist.h b/lib/zstd/compress/hist.h new file mode 100644 index 000000000000..fc1830abc9c6 --- /dev/null +++ b/lib/zstd/compress/hist.h @@ -0,0 +1,75 @@ +/* ****************************************************************** + * hist : Histogram functions + * part of Finite State Entropy project + * Copyright (c) Yann Collet, Facebook, Inc. + * + * You can contact the author at : + * - FSE source repository : https://github.com/Cyan4973/FiniteStateEntropy + * - Public forum : https://groups.google.com/forum/#!forum/lz4c + * + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). + * You may select, at your option, one of the above-listed licenses. +****************************************************************** */ + +/* --- dependencies --- */ +#include "../common/zstd_deps.h" /* size_t */ + + +/* --- simple histogram functions --- */ + +/*! HIST_count(): + * Provides the precise count of each byte within a table 'count'. + * 'count' is a table of unsigned int, of minimum size (*maxSymbolValuePtr+1). + * Updates *maxSymbolValuePtr with actual largest symbol value detected. + * @return : count of the most frequent symbol (which isn't identified). + * or an error code, which can be tested using HIST_isError(). + * note : if return == srcSize, there is only one symbol. + */ +size_t HIST_count(unsigned* count, unsigned* maxSymbolValuePtr, + const void* src, size_t srcSize); + +unsigned HIST_isError(size_t code); /*< tells if a return value is an error code */ + + +/* --- advanced histogram functions --- */ + +#define HIST_WKSP_SIZE_U32 1024 +#define HIST_WKSP_SIZE (HIST_WKSP_SIZE_U32 * sizeof(unsigned)) +/* HIST_count_wksp() : + * Same as HIST_count(), but using an externally provided scratch buffer. + * Benefit is this function will use very little stack space. + * `workSpace` is a writable buffer which must be 4-bytes aligned, + * `workSpaceSize` must be >= HIST_WKSP_SIZE + */ +size_t HIST_count_wksp(unsigned* count, unsigned* maxSymbolValuePtr, + const void* src, size_t srcSize, + void* workSpace, size_t workSpaceSize); + +/* HIST_countFast() : + * same as HIST_count(), but blindly trusts that all byte values within src are <= *maxSymbolValuePtr. + * This function is unsafe, and will segfault if any value within `src` is `> *maxSymbolValuePtr` + */ +size_t HIST_countFast(unsigned* count, unsigned* maxSymbolValuePtr, + const void* src, size_t srcSize); + +/* HIST_countFast_wksp() : + * Same as HIST_countFast(), but using an externally provided scratch buffer. + * `workSpace` is a writable buffer which must be 4-bytes aligned, + * `workSpaceSize` must be >= HIST_WKSP_SIZE + */ +size_t HIST_countFast_wksp(unsigned* count, unsigned* maxSymbolValuePtr, + const void* src, size_t srcSize, + void* workSpace, size_t workSpaceSize); + +/*! HIST_count_simple() : + * Same as HIST_countFast(), this function is unsafe, + * and will segfault if any value within `src` is `> *maxSymbolValuePtr`. + * It is also a bit slower for large inputs. + * However, it does not need any additional memory (not even on stack). + * @return : count of the most frequent symbol. + * Note this function doesn't produce any error (i.e. it must succeed). + */ +unsigned HIST_count_simple(unsigned* count, unsigned* maxSymbolValuePtr, + const void* src, size_t srcSize); diff --git a/lib/zstd/compress/huf_compress.c b/lib/zstd/compress/huf_compress.c new file mode 100644 index 000000000000..f76a526bfa54 --- /dev/null +++ b/lib/zstd/compress/huf_compress.c @@ -0,0 +1,905 @@ +/* ****************************************************************** + * Huffman encoder, part of New Generation Entropy library + * Copyright (c) Yann Collet, Facebook, Inc. + * + * You can contact the author at : + * - FSE+HUF source repository : https://github.com/Cyan4973/FiniteStateEntropy + * - Public forum : https://groups.google.com/forum/#!forum/lz4c + * + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). + * You may select, at your option, one of the above-listed licenses. +****************************************************************** */ + +/* ************************************************************** +* Compiler specifics +****************************************************************/ + + +/* ************************************************************** +* Includes +****************************************************************/ +#include "../common/zstd_deps.h" /* ZSTD_memcpy, ZSTD_memset */ +#include "../common/compiler.h" +#include "../common/bitstream.h" +#include "hist.h" +#define FSE_STATIC_LINKING_ONLY /* FSE_optimalTableLog_internal */ +#include "../common/fse.h" /* header compression */ +#define HUF_STATIC_LINKING_ONLY +#include "../common/huf.h" +#include "../common/error_private.h" + + +/* ************************************************************** +* Error Management +****************************************************************/ +#define HUF_isError ERR_isError +#define HUF_STATIC_ASSERT(c) DEBUG_STATIC_ASSERT(c) /* use only *after* variable declarations */ + + +/* ************************************************************** +* Utils +****************************************************************/ +unsigned HUF_optimalTableLog(unsigned maxTableLog, size_t srcSize, unsigned maxSymbolValue) +{ + return FSE_optimalTableLog_internal(maxTableLog, srcSize, maxSymbolValue, 1); +} + + +/* ******************************************************* +* HUF : Huffman block compression +*********************************************************/ +/* HUF_compressWeights() : + * Same as FSE_compress(), but dedicated to huff0's weights compression. + * The use case needs much less stack memory. + * Note : all elements within weightTable are supposed to be <= HUF_TABLELOG_MAX. + */ +#define MAX_FSE_TABLELOG_FOR_HUFF_HEADER 6 + +typedef struct { + FSE_CTable CTable[FSE_CTABLE_SIZE_U32(MAX_FSE_TABLELOG_FOR_HUFF_HEADER, HUF_TABLELOG_MAX)]; + U32 scratchBuffer[FSE_BUILD_CTABLE_WORKSPACE_SIZE_U32(HUF_TABLELOG_MAX, MAX_FSE_TABLELOG_FOR_HUFF_HEADER)]; + unsigned count[HUF_TABLELOG_MAX+1]; + S16 norm[HUF_TABLELOG_MAX+1]; +} HUF_CompressWeightsWksp; + +static size_t HUF_compressWeights(void* dst, size_t dstSize, const void* weightTable, size_t wtSize, void* workspace, size_t workspaceSize) +{ + BYTE* const ostart = (BYTE*) dst; + BYTE* op = ostart; + BYTE* const oend = ostart + dstSize; + + unsigned maxSymbolValue = HUF_TABLELOG_MAX; + U32 tableLog = MAX_FSE_TABLELOG_FOR_HUFF_HEADER; + HUF_CompressWeightsWksp* wksp = (HUF_CompressWeightsWksp*)workspace; + + if (workspaceSize < sizeof(HUF_CompressWeightsWksp)) return ERROR(GENERIC); + + /* init conditions */ + if (wtSize <= 1) return 0; /* Not compressible */ + + /* Scan input and build symbol stats */ + { unsigned const maxCount = HIST_count_simple(wksp->count, &maxSymbolValue, weightTable, wtSize); /* never fails */ + if (maxCount == wtSize) return 1; /* only a single symbol in src : rle */ + if (maxCount == 1) return 0; /* each symbol present maximum once => not compressible */ + } + + tableLog = FSE_optimalTableLog(tableLog, wtSize, maxSymbolValue); + CHECK_F( FSE_normalizeCount(wksp->norm, tableLog, wksp->count, wtSize, maxSymbolValue, /* useLowProbCount */ 0) ); + + /* Write table description header */ + { CHECK_V_F(hSize, FSE_writeNCount(op, (size_t)(oend-op), wksp->norm, maxSymbolValue, tableLog) ); + op += hSize; + } + + /* Compress */ + CHECK_F( FSE_buildCTable_wksp(wksp->CTable, wksp->norm, maxSymbolValue, tableLog, wksp->scratchBuffer, sizeof(wksp->scratchBuffer)) ); + { CHECK_V_F(cSize, FSE_compress_usingCTable(op, (size_t)(oend - op), weightTable, wtSize, wksp->CTable) ); + if (cSize == 0) return 0; /* not enough space for compressed data */ + op += cSize; + } + + return (size_t)(op-ostart); +} + + +typedef struct { + HUF_CompressWeightsWksp wksp; + BYTE bitsToWeight[HUF_TABLELOG_MAX + 1]; /* precomputed conversion table */ + BYTE huffWeight[HUF_SYMBOLVALUE_MAX]; +} HUF_WriteCTableWksp; + +size_t HUF_writeCTable_wksp(void* dst, size_t maxDstSize, + const HUF_CElt* CTable, unsigned maxSymbolValue, unsigned huffLog, + void* workspace, size_t workspaceSize) +{ + BYTE* op = (BYTE*)dst; + U32 n; + HUF_WriteCTableWksp* wksp = (HUF_WriteCTableWksp*)workspace; + + /* check conditions */ + if (workspaceSize < sizeof(HUF_WriteCTableWksp)) return ERROR(GENERIC); + if (maxSymbolValue > HUF_SYMBOLVALUE_MAX) return ERROR(maxSymbolValue_tooLarge); + + /* convert to weight */ + wksp->bitsToWeight[0] = 0; + for (n=1; nbitsToWeight[n] = (BYTE)(huffLog + 1 - n); + for (n=0; nhuffWeight[n] = wksp->bitsToWeight[CTable[n].nbBits]; + + /* attempt weights compression by FSE */ + { CHECK_V_F(hSize, HUF_compressWeights(op+1, maxDstSize-1, wksp->huffWeight, maxSymbolValue, &wksp->wksp, sizeof(wksp->wksp)) ); + if ((hSize>1) & (hSize < maxSymbolValue/2)) { /* FSE compressed */ + op[0] = (BYTE)hSize; + return hSize+1; + } } + + /* write raw values as 4-bits (max : 15) */ + if (maxSymbolValue > (256-128)) return ERROR(GENERIC); /* should not happen : likely means source cannot be compressed */ + if (((maxSymbolValue+1)/2) + 1 > maxDstSize) return ERROR(dstSize_tooSmall); /* not enough space within dst buffer */ + op[0] = (BYTE)(128 /*special case*/ + (maxSymbolValue-1)); + wksp->huffWeight[maxSymbolValue] = 0; /* to be sure it doesn't cause msan issue in final combination */ + for (n=0; nhuffWeight[n] << 4) + wksp->huffWeight[n+1]); + return ((maxSymbolValue+1)/2) + 1; +} + +/*! HUF_writeCTable() : + `CTable` : Huffman tree to save, using huf representation. + @return : size of saved CTable */ +size_t HUF_writeCTable (void* dst, size_t maxDstSize, + const HUF_CElt* CTable, unsigned maxSymbolValue, unsigned huffLog) +{ + HUF_WriteCTableWksp wksp; + return HUF_writeCTable_wksp(dst, maxDstSize, CTable, maxSymbolValue, huffLog, &wksp, sizeof(wksp)); +} + + +size_t HUF_readCTable (HUF_CElt* CTable, unsigned* maxSymbolValuePtr, const void* src, size_t srcSize, unsigned* hasZeroWeights) +{ + BYTE huffWeight[HUF_SYMBOLVALUE_MAX + 1]; /* init not required, even though some static analyzer may complain */ + U32 rankVal[HUF_TABLELOG_ABSOLUTEMAX + 1]; /* large enough for values from 0 to 16 */ + U32 tableLog = 0; + U32 nbSymbols = 0; + + /* get symbol weights */ + CHECK_V_F(readSize, HUF_readStats(huffWeight, HUF_SYMBOLVALUE_MAX+1, rankVal, &nbSymbols, &tableLog, src, srcSize)); + *hasZeroWeights = (rankVal[0] > 0); + + /* check result */ + if (tableLog > HUF_TABLELOG_MAX) return ERROR(tableLog_tooLarge); + if (nbSymbols > *maxSymbolValuePtr+1) return ERROR(maxSymbolValue_tooSmall); + + /* Prepare base value per rank */ + { U32 n, nextRankStart = 0; + for (n=1; n<=tableLog; n++) { + U32 curr = nextRankStart; + nextRankStart += (rankVal[n] << (n-1)); + rankVal[n] = curr; + } } + + /* fill nbBits */ + { U32 n; for (n=0; nn=tableLog+1 */ + U16 valPerRank[HUF_TABLELOG_MAX+2] = {0}; + { U32 n; for (n=0; n0; n--) { /* start at n=tablelog <-> w=1 */ + valPerRank[n] = min; /* get starting value within each rank */ + min += nbPerRank[n]; + min >>= 1; + } } + /* assign value within rank, symbol order */ + { U32 n; for (n=0; n maxNbBits to be maxNbBits. Then it adjusts + * the tree to so that it is a valid canonical Huffman tree. + * + * @pre The sum of the ranks of each symbol == 2^largestBits, + * where largestBits == huffNode[lastNonNull].nbBits. + * @post The sum of the ranks of each symbol == 2^largestBits, + * where largestBits is the return value <= maxNbBits. + * + * @param huffNode The Huffman tree modified in place to enforce maxNbBits. + * @param lastNonNull The symbol with the lowest count in the Huffman tree. + * @param maxNbBits The maximum allowed number of bits, which the Huffman tree + * may not respect. After this function the Huffman tree will + * respect maxNbBits. + * @return The maximum number of bits of the Huffman tree after adjustment, + * necessarily no more than maxNbBits. + */ +static U32 HUF_setMaxHeight(nodeElt* huffNode, U32 lastNonNull, U32 maxNbBits) +{ + const U32 largestBits = huffNode[lastNonNull].nbBits; + /* early exit : no elt > maxNbBits, so the tree is already valid. */ + if (largestBits <= maxNbBits) return largestBits; + + /* there are several too large elements (at least >= 2) */ + { int totalCost = 0; + const U32 baseCost = 1 << (largestBits - maxNbBits); + int n = (int)lastNonNull; + + /* Adjust any ranks > maxNbBits to maxNbBits. + * Compute totalCost, which is how far the sum of the ranks is + * we are over 2^largestBits after adjust the offending ranks. + */ + while (huffNode[n].nbBits > maxNbBits) { + totalCost += baseCost - (1 << (largestBits - huffNode[n].nbBits)); + huffNode[n].nbBits = (BYTE)maxNbBits; + n--; + } + /* n stops at huffNode[n].nbBits <= maxNbBits */ + assert(huffNode[n].nbBits <= maxNbBits); + /* n end at index of smallest symbol using < maxNbBits */ + while (huffNode[n].nbBits == maxNbBits) --n; + + /* renorm totalCost from 2^largestBits to 2^maxNbBits + * note : totalCost is necessarily a multiple of baseCost */ + assert((totalCost & (baseCost - 1)) == 0); + totalCost >>= (largestBits - maxNbBits); + assert(totalCost > 0); + + /* repay normalized cost */ + { U32 const noSymbol = 0xF0F0F0F0; + U32 rankLast[HUF_TABLELOG_MAX+2]; + + /* Get pos of last (smallest = lowest cum. count) symbol per rank */ + ZSTD_memset(rankLast, 0xF0, sizeof(rankLast)); + { U32 currentNbBits = maxNbBits; + int pos; + for (pos=n ; pos >= 0; pos--) { + if (huffNode[pos].nbBits >= currentNbBits) continue; + currentNbBits = huffNode[pos].nbBits; /* < maxNbBits */ + rankLast[maxNbBits-currentNbBits] = (U32)pos; + } } + + while (totalCost > 0) { + /* Try to reduce the next power of 2 above totalCost because we + * gain back half the rank. + */ + U32 nBitsToDecrease = BIT_highbit32((U32)totalCost) + 1; + for ( ; nBitsToDecrease > 1; nBitsToDecrease--) { + U32 const highPos = rankLast[nBitsToDecrease]; + U32 const lowPos = rankLast[nBitsToDecrease-1]; + if (highPos == noSymbol) continue; + /* Decrease highPos if no symbols of lowPos or if it is + * not cheaper to remove 2 lowPos than highPos. + */ + if (lowPos == noSymbol) break; + { U32 const highTotal = huffNode[highPos].count; + U32 const lowTotal = 2 * huffNode[lowPos].count; + if (highTotal <= lowTotal) break; + } } + /* only triggered when no more rank 1 symbol left => find closest one (note : there is necessarily at least one !) */ + assert(rankLast[nBitsToDecrease] != noSymbol || nBitsToDecrease == 1); + /* HUF_MAX_TABLELOG test just to please gcc 5+; but it should not be necessary */ + while ((nBitsToDecrease<=HUF_TABLELOG_MAX) && (rankLast[nBitsToDecrease] == noSymbol)) + nBitsToDecrease++; + assert(rankLast[nBitsToDecrease] != noSymbol); + /* Increase the number of bits to gain back half the rank cost. */ + totalCost -= 1 << (nBitsToDecrease-1); + huffNode[rankLast[nBitsToDecrease]].nbBits++; + + /* Fix up the new rank. + * If the new rank was empty, this symbol is now its smallest. + * Otherwise, this symbol will be the largest in the new rank so no adjustment. + */ + if (rankLast[nBitsToDecrease-1] == noSymbol) + rankLast[nBitsToDecrease-1] = rankLast[nBitsToDecrease]; + /* Fix up the old rank. + * If the symbol was at position 0, meaning it was the highest weight symbol in the tree, + * it must be the only symbol in its rank, so the old rank now has no symbols. + * Otherwise, since the Huffman nodes are sorted by count, the previous position is now + * the smallest node in the rank. If the previous position belongs to a different rank, + * then the rank is now empty. + */ + if (rankLast[nBitsToDecrease] == 0) /* special case, reached largest symbol */ + rankLast[nBitsToDecrease] = noSymbol; + else { + rankLast[nBitsToDecrease]--; + if (huffNode[rankLast[nBitsToDecrease]].nbBits != maxNbBits-nBitsToDecrease) + rankLast[nBitsToDecrease] = noSymbol; /* this rank is now empty */ + } + } /* while (totalCost > 0) */ + + /* If we've removed too much weight, then we have to add it back. + * To avoid overshooting again, we only adjust the smallest rank. + * We take the largest nodes from the lowest rank 0 and move them + * to rank 1. There's guaranteed to be enough rank 0 symbols because + * TODO. + */ + while (totalCost < 0) { /* Sometimes, cost correction overshoot */ + /* special case : no rank 1 symbol (using maxNbBits-1); + * let's create one from largest rank 0 (using maxNbBits). + */ + if (rankLast[1] == noSymbol) { + while (huffNode[n].nbBits == maxNbBits) n--; + huffNode[n+1].nbBits--; + assert(n >= 0); + rankLast[1] = (U32)(n+1); + totalCost++; + continue; + } + huffNode[ rankLast[1] + 1 ].nbBits--; + rankLast[1]++; + totalCost ++; + } + } /* repay normalized cost */ + } /* there are several too large elements (at least >= 2) */ + + return maxNbBits; +} + +typedef struct { + U32 base; + U32 curr; +} rankPos; + +typedef nodeElt huffNodeTable[HUF_CTABLE_WORKSPACE_SIZE_U32]; + +#define RANK_POSITION_TABLE_SIZE 32 + +typedef struct { + huffNodeTable huffNodeTbl; + rankPos rankPosition[RANK_POSITION_TABLE_SIZE]; +} HUF_buildCTable_wksp_tables; + +/* + * HUF_sort(): + * Sorts the symbols [0, maxSymbolValue] by count[symbol] in decreasing order. + * + * @param[out] huffNode Sorted symbols by decreasing count. Only members `.count` and `.byte` are filled. + * Must have (maxSymbolValue + 1) entries. + * @param[in] count Histogram of the symbols. + * @param[in] maxSymbolValue Maximum symbol value. + * @param rankPosition This is a scratch workspace. Must have RANK_POSITION_TABLE_SIZE entries. + */ +static void HUF_sort(nodeElt* huffNode, const unsigned* count, U32 maxSymbolValue, rankPos* rankPosition) +{ + int n; + int const maxSymbolValue1 = (int)maxSymbolValue + 1; + + /* Compute base and set curr to base. + * For symbol s let lowerRank = BIT_highbit32(count[n]+1) and rank = lowerRank + 1. + * Then 2^lowerRank <= count[n]+1 <= 2^rank. + * We attribute each symbol to lowerRank's base value, because we want to know where + * each rank begins in the output, so for rank R we want to count ranks R+1 and above. + */ + ZSTD_memset(rankPosition, 0, sizeof(*rankPosition) * RANK_POSITION_TABLE_SIZE); + for (n = 0; n < maxSymbolValue1; ++n) { + U32 lowerRank = BIT_highbit32(count[n] + 1); + rankPosition[lowerRank].base++; + } + assert(rankPosition[RANK_POSITION_TABLE_SIZE - 1].base == 0); + for (n = RANK_POSITION_TABLE_SIZE - 1; n > 0; --n) { + rankPosition[n-1].base += rankPosition[n].base; + rankPosition[n-1].curr = rankPosition[n-1].base; + } + /* Sort */ + for (n = 0; n < maxSymbolValue1; ++n) { + U32 const c = count[n]; + U32 const r = BIT_highbit32(c+1) + 1; + U32 pos = rankPosition[r].curr++; + /* Insert into the correct position in the rank. + * We have at most 256 symbols, so this insertion should be fine. + */ + while ((pos > rankPosition[r].base) && (c > huffNode[pos-1].count)) { + huffNode[pos] = huffNode[pos-1]; + pos--; + } + huffNode[pos].count = c; + huffNode[pos].byte = (BYTE)n; + } +} + + +/* HUF_buildCTable_wksp() : + * Same as HUF_buildCTable(), but using externally allocated scratch buffer. + * `workSpace` must be aligned on 4-bytes boundaries, and be at least as large as sizeof(HUF_buildCTable_wksp_tables). + */ +#define STARTNODE (HUF_SYMBOLVALUE_MAX+1) + +/* HUF_buildTree(): + * Takes the huffNode array sorted by HUF_sort() and builds an unlimited-depth Huffman tree. + * + * @param huffNode The array sorted by HUF_sort(). Builds the Huffman tree in this array. + * @param maxSymbolValue The maximum symbol value. + * @return The smallest node in the Huffman tree (by count). + */ +static int HUF_buildTree(nodeElt* huffNode, U32 maxSymbolValue) +{ + nodeElt* const huffNode0 = huffNode - 1; + int nonNullRank; + int lowS, lowN; + int nodeNb = STARTNODE; + int n, nodeRoot; + /* init for parents */ + nonNullRank = (int)maxSymbolValue; + while(huffNode[nonNullRank].count == 0) nonNullRank--; + lowS = nonNullRank; nodeRoot = nodeNb + lowS - 1; lowN = nodeNb; + huffNode[nodeNb].count = huffNode[lowS].count + huffNode[lowS-1].count; + huffNode[lowS].parent = huffNode[lowS-1].parent = (U16)nodeNb; + nodeNb++; lowS-=2; + for (n=nodeNb; n<=nodeRoot; n++) huffNode[n].count = (U32)(1U<<30); + huffNode0[0].count = (U32)(1U<<31); /* fake entry, strong barrier */ + + /* create parents */ + while (nodeNb <= nodeRoot) { + int const n1 = (huffNode[lowS].count < huffNode[lowN].count) ? lowS-- : lowN++; + int const n2 = (huffNode[lowS].count < huffNode[lowN].count) ? lowS-- : lowN++; + huffNode[nodeNb].count = huffNode[n1].count + huffNode[n2].count; + huffNode[n1].parent = huffNode[n2].parent = (U16)nodeNb; + nodeNb++; + } + + /* distribute weights (unlimited tree height) */ + huffNode[nodeRoot].nbBits = 0; + for (n=nodeRoot-1; n>=STARTNODE; n--) + huffNode[n].nbBits = huffNode[ huffNode[n].parent ].nbBits + 1; + for (n=0; n<=nonNullRank; n++) + huffNode[n].nbBits = huffNode[ huffNode[n].parent ].nbBits + 1; + + return nonNullRank; +} + +/* + * HUF_buildCTableFromTree(): + * Build the CTable given the Huffman tree in huffNode. + * + * @param[out] CTable The output Huffman CTable. + * @param huffNode The Huffman tree. + * @param nonNullRank The last and smallest node in the Huffman tree. + * @param maxSymbolValue The maximum symbol value. + * @param maxNbBits The exact maximum number of bits used in the Huffman tree. + */ +static void HUF_buildCTableFromTree(HUF_CElt* CTable, nodeElt const* huffNode, int nonNullRank, U32 maxSymbolValue, U32 maxNbBits) +{ + /* fill result into ctable (val, nbBits) */ + int n; + U16 nbPerRank[HUF_TABLELOG_MAX+1] = {0}; + U16 valPerRank[HUF_TABLELOG_MAX+1] = {0}; + int const alphabetSize = (int)(maxSymbolValue + 1); + for (n=0; n<=nonNullRank; n++) + nbPerRank[huffNode[n].nbBits]++; + /* determine starting value per rank */ + { U16 min = 0; + for (n=(int)maxNbBits; n>0; n--) { + valPerRank[n] = min; /* get starting value within each rank */ + min += nbPerRank[n]; + min >>= 1; + } } + for (n=0; nhuffNodeTbl; + nodeElt* const huffNode = huffNode0+1; + int nonNullRank; + + /* safety checks */ + if (((size_t)workSpace & 3) != 0) return ERROR(GENERIC); /* must be aligned on 4-bytes boundaries */ + if (wkspSize < sizeof(HUF_buildCTable_wksp_tables)) + return ERROR(workSpace_tooSmall); + if (maxNbBits == 0) maxNbBits = HUF_TABLELOG_DEFAULT; + if (maxSymbolValue > HUF_SYMBOLVALUE_MAX) + return ERROR(maxSymbolValue_tooLarge); + ZSTD_memset(huffNode0, 0, sizeof(huffNodeTable)); + + /* sort, decreasing order */ + HUF_sort(huffNode, count, maxSymbolValue, wksp_tables->rankPosition); + + /* build tree */ + nonNullRank = HUF_buildTree(huffNode, maxSymbolValue); + + /* enforce maxTableLog */ + maxNbBits = HUF_setMaxHeight(huffNode, (U32)nonNullRank, maxNbBits); + if (maxNbBits > HUF_TABLELOG_MAX) return ERROR(GENERIC); /* check fit into table */ + + HUF_buildCTableFromTree(tree, huffNode, nonNullRank, maxSymbolValue, maxNbBits); + + return maxNbBits; +} + +size_t HUF_estimateCompressedSize(const HUF_CElt* CTable, const unsigned* count, unsigned maxSymbolValue) +{ + size_t nbBits = 0; + int s; + for (s = 0; s <= (int)maxSymbolValue; ++s) { + nbBits += CTable[s].nbBits * count[s]; + } + return nbBits >> 3; +} + +int HUF_validateCTable(const HUF_CElt* CTable, const unsigned* count, unsigned maxSymbolValue) { + int bad = 0; + int s; + for (s = 0; s <= (int)maxSymbolValue; ++s) { + bad |= (count[s] != 0) & (CTable[s].nbBits == 0); + } + return !bad; +} + +size_t HUF_compressBound(size_t size) { return HUF_COMPRESSBOUND(size); } + +FORCE_INLINE_TEMPLATE void +HUF_encodeSymbol(BIT_CStream_t* bitCPtr, U32 symbol, const HUF_CElt* CTable) +{ + BIT_addBitsFast(bitCPtr, CTable[symbol].val, CTable[symbol].nbBits); +} + +#define HUF_FLUSHBITS(s) BIT_flushBits(s) + +#define HUF_FLUSHBITS_1(stream) \ + if (sizeof((stream)->bitContainer)*8 < HUF_TABLELOG_MAX*2+7) HUF_FLUSHBITS(stream) + +#define HUF_FLUSHBITS_2(stream) \ + if (sizeof((stream)->bitContainer)*8 < HUF_TABLELOG_MAX*4+7) HUF_FLUSHBITS(stream) + +FORCE_INLINE_TEMPLATE size_t +HUF_compress1X_usingCTable_internal_body(void* dst, size_t dstSize, + const void* src, size_t srcSize, + const HUF_CElt* CTable) +{ + const BYTE* ip = (const BYTE*) src; + BYTE* const ostart = (BYTE*)dst; + BYTE* const oend = ostart + dstSize; + BYTE* op = ostart; + size_t n; + BIT_CStream_t bitC; + + /* init */ + if (dstSize < 8) return 0; /* not enough space to compress */ + { size_t const initErr = BIT_initCStream(&bitC, op, (size_t)(oend-op)); + if (HUF_isError(initErr)) return 0; } + + n = srcSize & ~3; /* join to mod 4 */ + switch (srcSize & 3) + { + case 3: + HUF_encodeSymbol(&bitC, ip[n+ 2], CTable); + HUF_FLUSHBITS_2(&bitC); + ZSTD_FALLTHROUGH; + case 2: + HUF_encodeSymbol(&bitC, ip[n+ 1], CTable); + HUF_FLUSHBITS_1(&bitC); + ZSTD_FALLTHROUGH; + case 1: + HUF_encodeSymbol(&bitC, ip[n+ 0], CTable); + HUF_FLUSHBITS(&bitC); + ZSTD_FALLTHROUGH; + case 0: ZSTD_FALLTHROUGH; + default: break; + } + + for (; n>0; n-=4) { /* note : n&3==0 at this stage */ + HUF_encodeSymbol(&bitC, ip[n- 1], CTable); + HUF_FLUSHBITS_1(&bitC); + HUF_encodeSymbol(&bitC, ip[n- 2], CTable); + HUF_FLUSHBITS_2(&bitC); + HUF_encodeSymbol(&bitC, ip[n- 3], CTable); + HUF_FLUSHBITS_1(&bitC); + HUF_encodeSymbol(&bitC, ip[n- 4], CTable); + HUF_FLUSHBITS(&bitC); + } + + return BIT_closeCStream(&bitC); +} + +#if DYNAMIC_BMI2 + +static TARGET_ATTRIBUTE("bmi2") size_t +HUF_compress1X_usingCTable_internal_bmi2(void* dst, size_t dstSize, + const void* src, size_t srcSize, + const HUF_CElt* CTable) +{ + return HUF_compress1X_usingCTable_internal_body(dst, dstSize, src, srcSize, CTable); +} + +static size_t +HUF_compress1X_usingCTable_internal_default(void* dst, size_t dstSize, + const void* src, size_t srcSize, + const HUF_CElt* CTable) +{ + return HUF_compress1X_usingCTable_internal_body(dst, dstSize, src, srcSize, CTable); +} + +static size_t +HUF_compress1X_usingCTable_internal(void* dst, size_t dstSize, + const void* src, size_t srcSize, + const HUF_CElt* CTable, const int bmi2) +{ + if (bmi2) { + return HUF_compress1X_usingCTable_internal_bmi2(dst, dstSize, src, srcSize, CTable); + } + return HUF_compress1X_usingCTable_internal_default(dst, dstSize, src, srcSize, CTable); +} + +#else + +static size_t +HUF_compress1X_usingCTable_internal(void* dst, size_t dstSize, + const void* src, size_t srcSize, + const HUF_CElt* CTable, const int bmi2) +{ + (void)bmi2; + return HUF_compress1X_usingCTable_internal_body(dst, dstSize, src, srcSize, CTable); +} + +#endif + +size_t HUF_compress1X_usingCTable(void* dst, size_t dstSize, const void* src, size_t srcSize, const HUF_CElt* CTable) +{ + return HUF_compress1X_usingCTable_internal(dst, dstSize, src, srcSize, CTable, /* bmi2 */ 0); +} + + +static size_t +HUF_compress4X_usingCTable_internal(void* dst, size_t dstSize, + const void* src, size_t srcSize, + const HUF_CElt* CTable, int bmi2) +{ + size_t const segmentSize = (srcSize+3)/4; /* first 3 segments */ + const BYTE* ip = (const BYTE*) src; + const BYTE* const iend = ip + srcSize; + BYTE* const ostart = (BYTE*) dst; + BYTE* const oend = ostart + dstSize; + BYTE* op = ostart; + + if (dstSize < 6 + 1 + 1 + 1 + 8) return 0; /* minimum space to compress successfully */ + if (srcSize < 12) return 0; /* no saving possible : too small input */ + op += 6; /* jumpTable */ + + assert(op <= oend); + { CHECK_V_F(cSize, HUF_compress1X_usingCTable_internal(op, (size_t)(oend-op), ip, segmentSize, CTable, bmi2) ); + if (cSize==0) return 0; + assert(cSize <= 65535); + MEM_writeLE16(ostart, (U16)cSize); + op += cSize; + } + + ip += segmentSize; + assert(op <= oend); + { CHECK_V_F(cSize, HUF_compress1X_usingCTable_internal(op, (size_t)(oend-op), ip, segmentSize, CTable, bmi2) ); + if (cSize==0) return 0; + assert(cSize <= 65535); + MEM_writeLE16(ostart+2, (U16)cSize); + op += cSize; + } + + ip += segmentSize; + assert(op <= oend); + { CHECK_V_F(cSize, HUF_compress1X_usingCTable_internal(op, (size_t)(oend-op), ip, segmentSize, CTable, bmi2) ); + if (cSize==0) return 0; + assert(cSize <= 65535); + MEM_writeLE16(ostart+4, (U16)cSize); + op += cSize; + } + + ip += segmentSize; + assert(op <= oend); + assert(ip <= iend); + { CHECK_V_F(cSize, HUF_compress1X_usingCTable_internal(op, (size_t)(oend-op), ip, (size_t)(iend-ip), CTable, bmi2) ); + if (cSize==0) return 0; + op += cSize; + } + + return (size_t)(op-ostart); +} + +size_t HUF_compress4X_usingCTable(void* dst, size_t dstSize, const void* src, size_t srcSize, const HUF_CElt* CTable) +{ + return HUF_compress4X_usingCTable_internal(dst, dstSize, src, srcSize, CTable, /* bmi2 */ 0); +} + +typedef enum { HUF_singleStream, HUF_fourStreams } HUF_nbStreams_e; + +static size_t HUF_compressCTable_internal( + BYTE* const ostart, BYTE* op, BYTE* const oend, + const void* src, size_t srcSize, + HUF_nbStreams_e nbStreams, const HUF_CElt* CTable, const int bmi2) +{ + size_t const cSize = (nbStreams==HUF_singleStream) ? + HUF_compress1X_usingCTable_internal(op, (size_t)(oend - op), src, srcSize, CTable, bmi2) : + HUF_compress4X_usingCTable_internal(op, (size_t)(oend - op), src, srcSize, CTable, bmi2); + if (HUF_isError(cSize)) { return cSize; } + if (cSize==0) { return 0; } /* uncompressible */ + op += cSize; + /* check compressibility */ + assert(op >= ostart); + if ((size_t)(op-ostart) >= srcSize-1) { return 0; } + return (size_t)(op-ostart); +} + +typedef struct { + unsigned count[HUF_SYMBOLVALUE_MAX + 1]; + HUF_CElt CTable[HUF_SYMBOLVALUE_MAX + 1]; + union { + HUF_buildCTable_wksp_tables buildCTable_wksp; + HUF_WriteCTableWksp writeCTable_wksp; + } wksps; +} HUF_compress_tables_t; + +/* HUF_compress_internal() : + * `workSpace_align4` must be aligned on 4-bytes boundaries, + * and occupies the same space as a table of HUF_WORKSPACE_SIZE_U32 unsigned */ +static size_t +HUF_compress_internal (void* dst, size_t dstSize, + const void* src, size_t srcSize, + unsigned maxSymbolValue, unsigned huffLog, + HUF_nbStreams_e nbStreams, + void* workSpace_align4, size_t wkspSize, + HUF_CElt* oldHufTable, HUF_repeat* repeat, int preferRepeat, + const int bmi2) +{ + HUF_compress_tables_t* const table = (HUF_compress_tables_t*)workSpace_align4; + BYTE* const ostart = (BYTE*)dst; + BYTE* const oend = ostart + dstSize; + BYTE* op = ostart; + + HUF_STATIC_ASSERT(sizeof(*table) <= HUF_WORKSPACE_SIZE); + assert(((size_t)workSpace_align4 & 3) == 0); /* must be aligned on 4-bytes boundaries */ + + /* checks & inits */ + if (wkspSize < HUF_WORKSPACE_SIZE) return ERROR(workSpace_tooSmall); + if (!srcSize) return 0; /* Uncompressed */ + if (!dstSize) return 0; /* cannot fit anything within dst budget */ + if (srcSize > HUF_BLOCKSIZE_MAX) return ERROR(srcSize_wrong); /* current block size limit */ + if (huffLog > HUF_TABLELOG_MAX) return ERROR(tableLog_tooLarge); + if (maxSymbolValue > HUF_SYMBOLVALUE_MAX) return ERROR(maxSymbolValue_tooLarge); + if (!maxSymbolValue) maxSymbolValue = HUF_SYMBOLVALUE_MAX; + if (!huffLog) huffLog = HUF_TABLELOG_DEFAULT; + + /* Heuristic : If old table is valid, use it for small inputs */ + if (preferRepeat && repeat && *repeat == HUF_repeat_valid) { + return HUF_compressCTable_internal(ostart, op, oend, + src, srcSize, + nbStreams, oldHufTable, bmi2); + } + + /* Scan input and build symbol stats */ + { CHECK_V_F(largest, HIST_count_wksp (table->count, &maxSymbolValue, (const BYTE*)src, srcSize, workSpace_align4, wkspSize) ); + if (largest == srcSize) { *ostart = ((const BYTE*)src)[0]; return 1; } /* single symbol, rle */ + if (largest <= (srcSize >> 7)+4) return 0; /* heuristic : probably not compressible enough */ + } + + /* Check validity of previous table */ + if ( repeat + && *repeat == HUF_repeat_check + && !HUF_validateCTable(oldHufTable, table->count, maxSymbolValue)) { + *repeat = HUF_repeat_none; + } + /* Heuristic : use existing table for small inputs */ + if (preferRepeat && repeat && *repeat != HUF_repeat_none) { + return HUF_compressCTable_internal(ostart, op, oend, + src, srcSize, + nbStreams, oldHufTable, bmi2); + } + + /* Build Huffman Tree */ + huffLog = HUF_optimalTableLog(huffLog, srcSize, maxSymbolValue); + { size_t const maxBits = HUF_buildCTable_wksp(table->CTable, table->count, + maxSymbolValue, huffLog, + &table->wksps.buildCTable_wksp, sizeof(table->wksps.buildCTable_wksp)); + CHECK_F(maxBits); + huffLog = (U32)maxBits; + /* Zero unused symbols in CTable, so we can check it for validity */ + ZSTD_memset(table->CTable + (maxSymbolValue + 1), 0, + sizeof(table->CTable) - ((maxSymbolValue + 1) * sizeof(HUF_CElt))); + } + + /* Write table description header */ + { CHECK_V_F(hSize, HUF_writeCTable_wksp(op, dstSize, table->CTable, maxSymbolValue, huffLog, + &table->wksps.writeCTable_wksp, sizeof(table->wksps.writeCTable_wksp)) ); + /* Check if using previous huffman table is beneficial */ + if (repeat && *repeat != HUF_repeat_none) { + size_t const oldSize = HUF_estimateCompressedSize(oldHufTable, table->count, maxSymbolValue); + size_t const newSize = HUF_estimateCompressedSize(table->CTable, table->count, maxSymbolValue); + if (oldSize <= hSize + newSize || hSize + 12 >= srcSize) { + return HUF_compressCTable_internal(ostart, op, oend, + src, srcSize, + nbStreams, oldHufTable, bmi2); + } } + + /* Use the new huffman table */ + if (hSize + 12ul >= srcSize) { return 0; } + op += hSize; + if (repeat) { *repeat = HUF_repeat_none; } + if (oldHufTable) + ZSTD_memcpy(oldHufTable, table->CTable, sizeof(table->CTable)); /* Save new table */ + } + return HUF_compressCTable_internal(ostart, op, oend, + src, srcSize, + nbStreams, table->CTable, bmi2); +} + + +size_t HUF_compress1X_wksp (void* dst, size_t dstSize, + const void* src, size_t srcSize, + unsigned maxSymbolValue, unsigned huffLog, + void* workSpace, size_t wkspSize) +{ + return HUF_compress_internal(dst, dstSize, src, srcSize, + maxSymbolValue, huffLog, HUF_singleStream, + workSpace, wkspSize, + NULL, NULL, 0, 0 /*bmi2*/); +} + +size_t HUF_compress1X_repeat (void* dst, size_t dstSize, + const void* src, size_t srcSize, + unsigned maxSymbolValue, unsigned huffLog, + void* workSpace, size_t wkspSize, + HUF_CElt* hufTable, HUF_repeat* repeat, int preferRepeat, int bmi2) +{ + return HUF_compress_internal(dst, dstSize, src, srcSize, + maxSymbolValue, huffLog, HUF_singleStream, + workSpace, wkspSize, hufTable, + repeat, preferRepeat, bmi2); +} + +/* HUF_compress4X_repeat(): + * compress input using 4 streams. + * provide workspace to generate compression tables */ +size_t HUF_compress4X_wksp (void* dst, size_t dstSize, + const void* src, size_t srcSize, + unsigned maxSymbolValue, unsigned huffLog, + void* workSpace, size_t wkspSize) +{ + return HUF_compress_internal(dst, dstSize, src, srcSize, + maxSymbolValue, huffLog, HUF_fourStreams, + workSpace, wkspSize, + NULL, NULL, 0, 0 /*bmi2*/); +} + +/* HUF_compress4X_repeat(): + * compress input using 4 streams. + * re-use an existing huffman compression table */ +size_t HUF_compress4X_repeat (void* dst, size_t dstSize, + const void* src, size_t srcSize, + unsigned maxSymbolValue, unsigned huffLog, + void* workSpace, size_t wkspSize, + HUF_CElt* hufTable, HUF_repeat* repeat, int preferRepeat, int bmi2) +{ + return HUF_compress_internal(dst, dstSize, src, srcSize, + maxSymbolValue, huffLog, HUF_fourStreams, + workSpace, wkspSize, + hufTable, repeat, preferRepeat, bmi2); +} + diff --git a/lib/zstd/compress/zstd_compress.c b/lib/zstd/compress/zstd_compress.c new file mode 100644 index 000000000000..a4e916008b3a --- /dev/null +++ b/lib/zstd/compress/zstd_compress.c @@ -0,0 +1,5109 @@ +/* + * Copyright (c) Yann Collet, Facebook, Inc. + * All rights reserved. + * + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). + * You may select, at your option, one of the above-listed licenses. + */ + +/*-************************************* +* Dependencies +***************************************/ +#include "../common/zstd_deps.h" /* INT_MAX, ZSTD_memset, ZSTD_memcpy */ +#include "../common/cpu.h" +#include "../common/mem.h" +#include "hist.h" /* HIST_countFast_wksp */ +#define FSE_STATIC_LINKING_ONLY /* FSE_encodeSymbol */ +#include "../common/fse.h" +#define HUF_STATIC_LINKING_ONLY +#include "../common/huf.h" +#include "zstd_compress_internal.h" +#include "zstd_compress_sequences.h" +#include "zstd_compress_literals.h" +#include "zstd_fast.h" +#include "zstd_double_fast.h" +#include "zstd_lazy.h" +#include "zstd_opt.h" +#include "zstd_ldm.h" +#include "zstd_compress_superblock.h" + +/* *************************************************************** +* Tuning parameters +*****************************************************************/ +/*! + * COMPRESS_HEAPMODE : + * Select how default decompression function ZSTD_compress() allocates its context, + * on stack (0, default), or into heap (1). + * Note that functions with explicit context such as ZSTD_compressCCtx() are unaffected. + */ + + +/*-************************************* +* Helper functions +***************************************/ +/* ZSTD_compressBound() + * Note that the result from this function is only compatible with the "normal" + * full-block strategy. + * When there are a lot of small blocks due to frequent flush in streaming mode + * the overhead of headers can make the compressed data to be larger than the + * return value of ZSTD_compressBound(). + */ +size_t ZSTD_compressBound(size_t srcSize) { + return ZSTD_COMPRESSBOUND(srcSize); +} + + +/*-************************************* +* Context memory management +***************************************/ +struct ZSTD_CDict_s { + const void* dictContent; + size_t dictContentSize; + ZSTD_dictContentType_e dictContentType; /* The dictContentType the CDict was created with */ + U32* entropyWorkspace; /* entropy workspace of HUF_WORKSPACE_SIZE bytes */ + ZSTD_cwksp workspace; + ZSTD_matchState_t matchState; + ZSTD_compressedBlockState_t cBlockState; + ZSTD_customMem customMem; + U32 dictID; + int compressionLevel; /* 0 indicates that advanced API was used to select CDict params */ +}; /* typedef'd to ZSTD_CDict within "zstd.h" */ + +ZSTD_CCtx* ZSTD_createCCtx(void) +{ + return ZSTD_createCCtx_advanced(ZSTD_defaultCMem); +} + +static void ZSTD_initCCtx(ZSTD_CCtx* cctx, ZSTD_customMem memManager) +{ + assert(cctx != NULL); + ZSTD_memset(cctx, 0, sizeof(*cctx)); + cctx->customMem = memManager; + cctx->bmi2 = ZSTD_cpuid_bmi2(ZSTD_cpuid()); + { size_t const err = ZSTD_CCtx_reset(cctx, ZSTD_reset_parameters); + assert(!ZSTD_isError(err)); + (void)err; + } +} + +ZSTD_CCtx* ZSTD_createCCtx_advanced(ZSTD_customMem customMem) +{ + ZSTD_STATIC_ASSERT(zcss_init==0); + ZSTD_STATIC_ASSERT(ZSTD_CONTENTSIZE_UNKNOWN==(0ULL - 1)); + if ((!customMem.customAlloc) ^ (!customMem.customFree)) return NULL; + { ZSTD_CCtx* const cctx = (ZSTD_CCtx*)ZSTD_customMalloc(sizeof(ZSTD_CCtx), customMem); + if (!cctx) return NULL; + ZSTD_initCCtx(cctx, customMem); + return cctx; + } +} + +ZSTD_CCtx* ZSTD_initStaticCCtx(void* workspace, size_t workspaceSize) +{ + ZSTD_cwksp ws; + ZSTD_CCtx* cctx; + if (workspaceSize <= sizeof(ZSTD_CCtx)) return NULL; /* minimum size */ + if ((size_t)workspace & 7) return NULL; /* must be 8-aligned */ + ZSTD_cwksp_init(&ws, workspace, workspaceSize, ZSTD_cwksp_static_alloc); + + cctx = (ZSTD_CCtx*)ZSTD_cwksp_reserve_object(&ws, sizeof(ZSTD_CCtx)); + if (cctx == NULL) return NULL; + + ZSTD_memset(cctx, 0, sizeof(ZSTD_CCtx)); + ZSTD_cwksp_move(&cctx->workspace, &ws); + cctx->staticSize = workspaceSize; + + /* statically sized space. entropyWorkspace never moves (but prev/next block swap places) */ + if (!ZSTD_cwksp_check_available(&cctx->workspace, ENTROPY_WORKSPACE_SIZE + 2 * sizeof(ZSTD_compressedBlockState_t))) return NULL; + cctx->blockState.prevCBlock = (ZSTD_compressedBlockState_t*)ZSTD_cwksp_reserve_object(&cctx->workspace, sizeof(ZSTD_compressedBlockState_t)); + cctx->blockState.nextCBlock = (ZSTD_compressedBlockState_t*)ZSTD_cwksp_reserve_object(&cctx->workspace, sizeof(ZSTD_compressedBlockState_t)); + cctx->entropyWorkspace = (U32*)ZSTD_cwksp_reserve_object(&cctx->workspace, ENTROPY_WORKSPACE_SIZE); + cctx->bmi2 = ZSTD_cpuid_bmi2(ZSTD_cpuid()); + return cctx; +} + +/* + * Clears and frees all of the dictionaries in the CCtx. + */ +static void ZSTD_clearAllDicts(ZSTD_CCtx* cctx) +{ + ZSTD_customFree(cctx->localDict.dictBuffer, cctx->customMem); + ZSTD_freeCDict(cctx->localDict.cdict); + ZSTD_memset(&cctx->localDict, 0, sizeof(cctx->localDict)); + ZSTD_memset(&cctx->prefixDict, 0, sizeof(cctx->prefixDict)); + cctx->cdict = NULL; +} + +static size_t ZSTD_sizeof_localDict(ZSTD_localDict dict) +{ + size_t const bufferSize = dict.dictBuffer != NULL ? dict.dictSize : 0; + size_t const cdictSize = ZSTD_sizeof_CDict(dict.cdict); + return bufferSize + cdictSize; +} + +static void ZSTD_freeCCtxContent(ZSTD_CCtx* cctx) +{ + assert(cctx != NULL); + assert(cctx->staticSize == 0); + ZSTD_clearAllDicts(cctx); + ZSTD_cwksp_free(&cctx->workspace, cctx->customMem); +} + +size_t ZSTD_freeCCtx(ZSTD_CCtx* cctx) +{ + if (cctx==NULL) return 0; /* support free on NULL */ + RETURN_ERROR_IF(cctx->staticSize, memory_allocation, + "not compatible with static CCtx"); + { + int cctxInWorkspace = ZSTD_cwksp_owns_buffer(&cctx->workspace, cctx); + ZSTD_freeCCtxContent(cctx); + if (!cctxInWorkspace) { + ZSTD_customFree(cctx, cctx->customMem); + } + } + return 0; +} + + +static size_t ZSTD_sizeof_mtctx(const ZSTD_CCtx* cctx) +{ + (void)cctx; + return 0; +} + + +size_t ZSTD_sizeof_CCtx(const ZSTD_CCtx* cctx) +{ + if (cctx==NULL) return 0; /* support sizeof on NULL */ + /* cctx may be in the workspace */ + return (cctx->workspace.workspace == cctx ? 0 : sizeof(*cctx)) + + ZSTD_cwksp_sizeof(&cctx->workspace) + + ZSTD_sizeof_localDict(cctx->localDict) + + ZSTD_sizeof_mtctx(cctx); +} + +size_t ZSTD_sizeof_CStream(const ZSTD_CStream* zcs) +{ + return ZSTD_sizeof_CCtx(zcs); /* same object */ +} + +/* private API call, for dictBuilder only */ +const seqStore_t* ZSTD_getSeqStore(const ZSTD_CCtx* ctx) { return &(ctx->seqStore); } + +/* Returns 1 if compression parameters are such that we should + * enable long distance matching (wlog >= 27, strategy >= btopt). + * Returns 0 otherwise. + */ +static U32 ZSTD_CParams_shouldEnableLdm(const ZSTD_compressionParameters* const cParams) { + return cParams->strategy >= ZSTD_btopt && cParams->windowLog >= 27; +} + +static ZSTD_CCtx_params ZSTD_makeCCtxParamsFromCParams( + ZSTD_compressionParameters cParams) +{ + ZSTD_CCtx_params cctxParams; + /* should not matter, as all cParams are presumed properly defined */ + ZSTD_CCtxParams_init(&cctxParams, ZSTD_CLEVEL_DEFAULT); + cctxParams.cParams = cParams; + + if (ZSTD_CParams_shouldEnableLdm(&cParams)) { + DEBUGLOG(4, "ZSTD_makeCCtxParamsFromCParams(): Including LDM into cctx params"); + cctxParams.ldmParams.enableLdm = 1; + /* LDM is enabled by default for optimal parser and window size >= 128MB */ + ZSTD_ldm_adjustParameters(&cctxParams.ldmParams, &cParams); + assert(cctxParams.ldmParams.hashLog >= cctxParams.ldmParams.bucketSizeLog); + assert(cctxParams.ldmParams.hashRateLog < 32); + } + + assert(!ZSTD_checkCParams(cParams)); + return cctxParams; +} + +static ZSTD_CCtx_params* ZSTD_createCCtxParams_advanced( + ZSTD_customMem customMem) +{ + ZSTD_CCtx_params* params; + if ((!customMem.customAlloc) ^ (!customMem.customFree)) return NULL; + params = (ZSTD_CCtx_params*)ZSTD_customCalloc( + sizeof(ZSTD_CCtx_params), customMem); + if (!params) { return NULL; } + ZSTD_CCtxParams_init(params, ZSTD_CLEVEL_DEFAULT); + params->customMem = customMem; + return params; +} + +ZSTD_CCtx_params* ZSTD_createCCtxParams(void) +{ + return ZSTD_createCCtxParams_advanced(ZSTD_defaultCMem); +} + +size_t ZSTD_freeCCtxParams(ZSTD_CCtx_params* params) +{ + if (params == NULL) { return 0; } + ZSTD_customFree(params, params->customMem); + return 0; +} + +size_t ZSTD_CCtxParams_reset(ZSTD_CCtx_params* params) +{ + return ZSTD_CCtxParams_init(params, ZSTD_CLEVEL_DEFAULT); +} + +size_t ZSTD_CCtxParams_init(ZSTD_CCtx_params* cctxParams, int compressionLevel) { + RETURN_ERROR_IF(!cctxParams, GENERIC, "NULL pointer!"); + ZSTD_memset(cctxParams, 0, sizeof(*cctxParams)); + cctxParams->compressionLevel = compressionLevel; + cctxParams->fParams.contentSizeFlag = 1; + return 0; +} + +#define ZSTD_NO_CLEVEL 0 + +/* + * Initializes the cctxParams from params and compressionLevel. + * @param compressionLevel If params are derived from a compression level then that compression level, otherwise ZSTD_NO_CLEVEL. + */ +static void ZSTD_CCtxParams_init_internal(ZSTD_CCtx_params* cctxParams, ZSTD_parameters const* params, int compressionLevel) +{ + assert(!ZSTD_checkCParams(params->cParams)); + ZSTD_memset(cctxParams, 0, sizeof(*cctxParams)); + cctxParams->cParams = params->cParams; + cctxParams->fParams = params->fParams; + /* Should not matter, as all cParams are presumed properly defined. + * But, set it for tracing anyway. + */ + cctxParams->compressionLevel = compressionLevel; +} + +size_t ZSTD_CCtxParams_init_advanced(ZSTD_CCtx_params* cctxParams, ZSTD_parameters params) +{ + RETURN_ERROR_IF(!cctxParams, GENERIC, "NULL pointer!"); + FORWARD_IF_ERROR( ZSTD_checkCParams(params.cParams) , ""); + ZSTD_CCtxParams_init_internal(cctxParams, ¶ms, ZSTD_NO_CLEVEL); + return 0; +} + +/* + * Sets cctxParams' cParams and fParams from params, but otherwise leaves them alone. + * @param param Validated zstd parameters. + */ +static void ZSTD_CCtxParams_setZstdParams( + ZSTD_CCtx_params* cctxParams, const ZSTD_parameters* params) +{ + assert(!ZSTD_checkCParams(params->cParams)); + cctxParams->cParams = params->cParams; + cctxParams->fParams = params->fParams; + /* Should not matter, as all cParams are presumed properly defined. + * But, set it for tracing anyway. + */ + cctxParams->compressionLevel = ZSTD_NO_CLEVEL; +} + +ZSTD_bounds ZSTD_cParam_getBounds(ZSTD_cParameter param) +{ + ZSTD_bounds bounds = { 0, 0, 0 }; + + switch(param) + { + case ZSTD_c_compressionLevel: + bounds.lowerBound = ZSTD_minCLevel(); + bounds.upperBound = ZSTD_maxCLevel(); + return bounds; + + case ZSTD_c_windowLog: + bounds.lowerBound = ZSTD_WINDOWLOG_MIN; + bounds.upperBound = ZSTD_WINDOWLOG_MAX; + return bounds; + + case ZSTD_c_hashLog: + bounds.lowerBound = ZSTD_HASHLOG_MIN; + bounds.upperBound = ZSTD_HASHLOG_MAX; + return bounds; + + case ZSTD_c_chainLog: + bounds.lowerBound = ZSTD_CHAINLOG_MIN; + bounds.upperBound = ZSTD_CHAINLOG_MAX; + return bounds; + + case ZSTD_c_searchLog: + bounds.lowerBound = ZSTD_SEARCHLOG_MIN; + bounds.upperBound = ZSTD_SEARCHLOG_MAX; + return bounds; + + case ZSTD_c_minMatch: + bounds.lowerBound = ZSTD_MINMATCH_MIN; + bounds.upperBound = ZSTD_MINMATCH_MAX; + return bounds; + + case ZSTD_c_targetLength: + bounds.lowerBound = ZSTD_TARGETLENGTH_MIN; + bounds.upperBound = ZSTD_TARGETLENGTH_MAX; + return bounds; + + case ZSTD_c_strategy: + bounds.lowerBound = ZSTD_STRATEGY_MIN; + bounds.upperBound = ZSTD_STRATEGY_MAX; + return bounds; + + case ZSTD_c_contentSizeFlag: + bounds.lowerBound = 0; + bounds.upperBound = 1; + return bounds; + + case ZSTD_c_checksumFlag: + bounds.lowerBound = 0; + bounds.upperBound = 1; + return bounds; + + case ZSTD_c_dictIDFlag: + bounds.lowerBound = 0; + bounds.upperBound = 1; + return bounds; + + case ZSTD_c_nbWorkers: + bounds.lowerBound = 0; + bounds.upperBound = 0; + return bounds; + + case ZSTD_c_jobSize: + bounds.lowerBound = 0; + bounds.upperBound = 0; + return bounds; + + case ZSTD_c_overlapLog: + bounds.lowerBound = 0; + bounds.upperBound = 0; + return bounds; + + case ZSTD_c_enableDedicatedDictSearch: + bounds.lowerBound = 0; + bounds.upperBound = 1; + return bounds; + + case ZSTD_c_enableLongDistanceMatching: + bounds.lowerBound = 0; + bounds.upperBound = 1; + return bounds; + + case ZSTD_c_ldmHashLog: + bounds.lowerBound = ZSTD_LDM_HASHLOG_MIN; + bounds.upperBound = ZSTD_LDM_HASHLOG_MAX; + return bounds; + + case ZSTD_c_ldmMinMatch: + bounds.lowerBound = ZSTD_LDM_MINMATCH_MIN; + bounds.upperBound = ZSTD_LDM_MINMATCH_MAX; + return bounds; + + case ZSTD_c_ldmBucketSizeLog: + bounds.lowerBound = ZSTD_LDM_BUCKETSIZELOG_MIN; + bounds.upperBound = ZSTD_LDM_BUCKETSIZELOG_MAX; + return bounds; + + case ZSTD_c_ldmHashRateLog: + bounds.lowerBound = ZSTD_LDM_HASHRATELOG_MIN; + bounds.upperBound = ZSTD_LDM_HASHRATELOG_MAX; + return bounds; + + /* experimental parameters */ + case ZSTD_c_rsyncable: + bounds.lowerBound = 0; + bounds.upperBound = 1; + return bounds; + + case ZSTD_c_forceMaxWindow : + bounds.lowerBound = 0; + bounds.upperBound = 1; + return bounds; + + case ZSTD_c_format: + ZSTD_STATIC_ASSERT(ZSTD_f_zstd1 < ZSTD_f_zstd1_magicless); + bounds.lowerBound = ZSTD_f_zstd1; + bounds.upperBound = ZSTD_f_zstd1_magicless; /* note : how to ensure at compile time that this is the highest value enum ? */ + return bounds; + + case ZSTD_c_forceAttachDict: + ZSTD_STATIC_ASSERT(ZSTD_dictDefaultAttach < ZSTD_dictForceLoad); + bounds.lowerBound = ZSTD_dictDefaultAttach; + bounds.upperBound = ZSTD_dictForceLoad; /* note : how to ensure at compile time that this is the highest value enum ? */ + return bounds; + + case ZSTD_c_literalCompressionMode: + ZSTD_STATIC_ASSERT(ZSTD_lcm_auto < ZSTD_lcm_huffman && ZSTD_lcm_huffman < ZSTD_lcm_uncompressed); + bounds.lowerBound = ZSTD_lcm_auto; + bounds.upperBound = ZSTD_lcm_uncompressed; + return bounds; + + case ZSTD_c_targetCBlockSize: + bounds.lowerBound = ZSTD_TARGETCBLOCKSIZE_MIN; + bounds.upperBound = ZSTD_TARGETCBLOCKSIZE_MAX; + return bounds; + + case ZSTD_c_srcSizeHint: + bounds.lowerBound = ZSTD_SRCSIZEHINT_MIN; + bounds.upperBound = ZSTD_SRCSIZEHINT_MAX; + return bounds; + + case ZSTD_c_stableInBuffer: + case ZSTD_c_stableOutBuffer: + bounds.lowerBound = (int)ZSTD_bm_buffered; + bounds.upperBound = (int)ZSTD_bm_stable; + return bounds; + + case ZSTD_c_blockDelimiters: + bounds.lowerBound = (int)ZSTD_sf_noBlockDelimiters; + bounds.upperBound = (int)ZSTD_sf_explicitBlockDelimiters; + return bounds; + + case ZSTD_c_validateSequences: + bounds.lowerBound = 0; + bounds.upperBound = 1; + return bounds; + + default: + bounds.error = ERROR(parameter_unsupported); + return bounds; + } +} + +/* ZSTD_cParam_clampBounds: + * Clamps the value into the bounded range. + */ +static size_t ZSTD_cParam_clampBounds(ZSTD_cParameter cParam, int* value) +{ + ZSTD_bounds const bounds = ZSTD_cParam_getBounds(cParam); + if (ZSTD_isError(bounds.error)) return bounds.error; + if (*value < bounds.lowerBound) *value = bounds.lowerBound; + if (*value > bounds.upperBound) *value = bounds.upperBound; + return 0; +} + +#define BOUNDCHECK(cParam, val) { \ + RETURN_ERROR_IF(!ZSTD_cParam_withinBounds(cParam,val), \ + parameter_outOfBound, "Param out of bounds"); \ +} + + +static int ZSTD_isUpdateAuthorized(ZSTD_cParameter param) +{ + switch(param) + { + case ZSTD_c_compressionLevel: + case ZSTD_c_hashLog: + case ZSTD_c_chainLog: + case ZSTD_c_searchLog: + case ZSTD_c_minMatch: + case ZSTD_c_targetLength: + case ZSTD_c_strategy: + return 1; + + case ZSTD_c_format: + case ZSTD_c_windowLog: + case ZSTD_c_contentSizeFlag: + case ZSTD_c_checksumFlag: + case ZSTD_c_dictIDFlag: + case ZSTD_c_forceMaxWindow : + case ZSTD_c_nbWorkers: + case ZSTD_c_jobSize: + case ZSTD_c_overlapLog: + case ZSTD_c_rsyncable: + case ZSTD_c_enableDedicatedDictSearch: + case ZSTD_c_enableLongDistanceMatching: + case ZSTD_c_ldmHashLog: + case ZSTD_c_ldmMinMatch: + case ZSTD_c_ldmBucketSizeLog: + case ZSTD_c_ldmHashRateLog: + case ZSTD_c_forceAttachDict: + case ZSTD_c_literalCompressionMode: + case ZSTD_c_targetCBlockSize: + case ZSTD_c_srcSizeHint: + case ZSTD_c_stableInBuffer: + case ZSTD_c_stableOutBuffer: + case ZSTD_c_blockDelimiters: + case ZSTD_c_validateSequences: + default: + return 0; + } +} + +size_t ZSTD_CCtx_setParameter(ZSTD_CCtx* cctx, ZSTD_cParameter param, int value) +{ + DEBUGLOG(4, "ZSTD_CCtx_setParameter (%i, %i)", (int)param, value); + if (cctx->streamStage != zcss_init) { + if (ZSTD_isUpdateAuthorized(param)) { + cctx->cParamsChanged = 1; + } else { + RETURN_ERROR(stage_wrong, "can only set params in ctx init stage"); + } } + + switch(param) + { + case ZSTD_c_nbWorkers: + RETURN_ERROR_IF((value!=0) && cctx->staticSize, parameter_unsupported, + "MT not compatible with static alloc"); + break; + + case ZSTD_c_compressionLevel: + case ZSTD_c_windowLog: + case ZSTD_c_hashLog: + case ZSTD_c_chainLog: + case ZSTD_c_searchLog: + case ZSTD_c_minMatch: + case ZSTD_c_targetLength: + case ZSTD_c_strategy: + case ZSTD_c_ldmHashRateLog: + case ZSTD_c_format: + case ZSTD_c_contentSizeFlag: + case ZSTD_c_checksumFlag: + case ZSTD_c_dictIDFlag: + case ZSTD_c_forceMaxWindow: + case ZSTD_c_forceAttachDict: + case ZSTD_c_literalCompressionMode: + case ZSTD_c_jobSize: + case ZSTD_c_overlapLog: + case ZSTD_c_rsyncable: + case ZSTD_c_enableDedicatedDictSearch: + case ZSTD_c_enableLongDistanceMatching: + case ZSTD_c_ldmHashLog: + case ZSTD_c_ldmMinMatch: + case ZSTD_c_ldmBucketSizeLog: + case ZSTD_c_targetCBlockSize: + case ZSTD_c_srcSizeHint: + case ZSTD_c_stableInBuffer: + case ZSTD_c_stableOutBuffer: + case ZSTD_c_blockDelimiters: + case ZSTD_c_validateSequences: + break; + + default: RETURN_ERROR(parameter_unsupported, "unknown parameter"); + } + return ZSTD_CCtxParams_setParameter(&cctx->requestedParams, param, value); +} + +size_t ZSTD_CCtxParams_setParameter(ZSTD_CCtx_params* CCtxParams, + ZSTD_cParameter param, int value) +{ + DEBUGLOG(4, "ZSTD_CCtxParams_setParameter (%i, %i)", (int)param, value); + switch(param) + { + case ZSTD_c_format : + BOUNDCHECK(ZSTD_c_format, value); + CCtxParams->format = (ZSTD_format_e)value; + return (size_t)CCtxParams->format; + + case ZSTD_c_compressionLevel : { + FORWARD_IF_ERROR(ZSTD_cParam_clampBounds(param, &value), ""); + if (value == 0) + CCtxParams->compressionLevel = ZSTD_CLEVEL_DEFAULT; /* 0 == default */ + else + CCtxParams->compressionLevel = value; + if (CCtxParams->compressionLevel >= 0) return (size_t)CCtxParams->compressionLevel; + return 0; /* return type (size_t) cannot represent negative values */ + } + + case ZSTD_c_windowLog : + if (value!=0) /* 0 => use default */ + BOUNDCHECK(ZSTD_c_windowLog, value); + CCtxParams->cParams.windowLog = (U32)value; + return CCtxParams->cParams.windowLog; + + case ZSTD_c_hashLog : + if (value!=0) /* 0 => use default */ + BOUNDCHECK(ZSTD_c_hashLog, value); + CCtxParams->cParams.hashLog = (U32)value; + return CCtxParams->cParams.hashLog; + + case ZSTD_c_chainLog : + if (value!=0) /* 0 => use default */ + BOUNDCHECK(ZSTD_c_chainLog, value); + CCtxParams->cParams.chainLog = (U32)value; + return CCtxParams->cParams.chainLog; + + case ZSTD_c_searchLog : + if (value!=0) /* 0 => use default */ + BOUNDCHECK(ZSTD_c_searchLog, value); + CCtxParams->cParams.searchLog = (U32)value; + return (size_t)value; + + case ZSTD_c_minMatch : + if (value!=0) /* 0 => use default */ + BOUNDCHECK(ZSTD_c_minMatch, value); + CCtxParams->cParams.minMatch = value; + return CCtxParams->cParams.minMatch; + + case ZSTD_c_targetLength : + BOUNDCHECK(ZSTD_c_targetLength, value); + CCtxParams->cParams.targetLength = value; + return CCtxParams->cParams.targetLength; + + case ZSTD_c_strategy : + if (value!=0) /* 0 => use default */ + BOUNDCHECK(ZSTD_c_strategy, value); + CCtxParams->cParams.strategy = (ZSTD_strategy)value; + return (size_t)CCtxParams->cParams.strategy; + + case ZSTD_c_contentSizeFlag : + /* Content size written in frame header _when known_ (default:1) */ + DEBUGLOG(4, "set content size flag = %u", (value!=0)); + CCtxParams->fParams.contentSizeFlag = value != 0; + return CCtxParams->fParams.contentSizeFlag; + + case ZSTD_c_checksumFlag : + /* A 32-bits content checksum will be calculated and written at end of frame (default:0) */ + CCtxParams->fParams.checksumFlag = value != 0; + return CCtxParams->fParams.checksumFlag; + + case ZSTD_c_dictIDFlag : /* When applicable, dictionary's dictID is provided in frame header (default:1) */ + DEBUGLOG(4, "set dictIDFlag = %u", (value!=0)); + CCtxParams->fParams.noDictIDFlag = !value; + return !CCtxParams->fParams.noDictIDFlag; + + case ZSTD_c_forceMaxWindow : + CCtxParams->forceWindow = (value != 0); + return CCtxParams->forceWindow; + + case ZSTD_c_forceAttachDict : { + const ZSTD_dictAttachPref_e pref = (ZSTD_dictAttachPref_e)value; + BOUNDCHECK(ZSTD_c_forceAttachDict, pref); + CCtxParams->attachDictPref = pref; + return CCtxParams->attachDictPref; + } + + case ZSTD_c_literalCompressionMode : { + const ZSTD_literalCompressionMode_e lcm = (ZSTD_literalCompressionMode_e)value; + BOUNDCHECK(ZSTD_c_literalCompressionMode, lcm); + CCtxParams->literalCompressionMode = lcm; + return CCtxParams->literalCompressionMode; + } + + case ZSTD_c_nbWorkers : + RETURN_ERROR_IF(value!=0, parameter_unsupported, "not compiled with multithreading"); + return 0; + + case ZSTD_c_jobSize : + RETURN_ERROR_IF(value!=0, parameter_unsupported, "not compiled with multithreading"); + return 0; + + case ZSTD_c_overlapLog : + RETURN_ERROR_IF(value!=0, parameter_unsupported, "not compiled with multithreading"); + return 0; + + case ZSTD_c_rsyncable : + RETURN_ERROR_IF(value!=0, parameter_unsupported, "not compiled with multithreading"); + return 0; + + case ZSTD_c_enableDedicatedDictSearch : + CCtxParams->enableDedicatedDictSearch = (value!=0); + return CCtxParams->enableDedicatedDictSearch; + + case ZSTD_c_enableLongDistanceMatching : + CCtxParams->ldmParams.enableLdm = (value!=0); + return CCtxParams->ldmParams.enableLdm; + + case ZSTD_c_ldmHashLog : + if (value!=0) /* 0 ==> auto */ + BOUNDCHECK(ZSTD_c_ldmHashLog, value); + CCtxParams->ldmParams.hashLog = value; + return CCtxParams->ldmParams.hashLog; + + case ZSTD_c_ldmMinMatch : + if (value!=0) /* 0 ==> default */ + BOUNDCHECK(ZSTD_c_ldmMinMatch, value); + CCtxParams->ldmParams.minMatchLength = value; + return CCtxParams->ldmParams.minMatchLength; + + case ZSTD_c_ldmBucketSizeLog : + if (value!=0) /* 0 ==> default */ + BOUNDCHECK(ZSTD_c_ldmBucketSizeLog, value); + CCtxParams->ldmParams.bucketSizeLog = value; + return CCtxParams->ldmParams.bucketSizeLog; + + case ZSTD_c_ldmHashRateLog : + if (value!=0) /* 0 ==> default */ + BOUNDCHECK(ZSTD_c_ldmHashRateLog, value); + CCtxParams->ldmParams.hashRateLog = value; + return CCtxParams->ldmParams.hashRateLog; + + case ZSTD_c_targetCBlockSize : + if (value!=0) /* 0 ==> default */ + BOUNDCHECK(ZSTD_c_targetCBlockSize, value); + CCtxParams->targetCBlockSize = value; + return CCtxParams->targetCBlockSize; + + case ZSTD_c_srcSizeHint : + if (value!=0) /* 0 ==> default */ + BOUNDCHECK(ZSTD_c_srcSizeHint, value); + CCtxParams->srcSizeHint = value; + return CCtxParams->srcSizeHint; + + case ZSTD_c_stableInBuffer: + BOUNDCHECK(ZSTD_c_stableInBuffer, value); + CCtxParams->inBufferMode = (ZSTD_bufferMode_e)value; + return CCtxParams->inBufferMode; + + case ZSTD_c_stableOutBuffer: + BOUNDCHECK(ZSTD_c_stableOutBuffer, value); + CCtxParams->outBufferMode = (ZSTD_bufferMode_e)value; + return CCtxParams->outBufferMode; + + case ZSTD_c_blockDelimiters: + BOUNDCHECK(ZSTD_c_blockDelimiters, value); + CCtxParams->blockDelimiters = (ZSTD_sequenceFormat_e)value; + return CCtxParams->blockDelimiters; + + case ZSTD_c_validateSequences: + BOUNDCHECK(ZSTD_c_validateSequences, value); + CCtxParams->validateSequences = value; + return CCtxParams->validateSequences; + + default: RETURN_ERROR(parameter_unsupported, "unknown parameter"); + } +} + +size_t ZSTD_CCtx_getParameter(ZSTD_CCtx const* cctx, ZSTD_cParameter param, int* value) +{ + return ZSTD_CCtxParams_getParameter(&cctx->requestedParams, param, value); +} + +size_t ZSTD_CCtxParams_getParameter( + ZSTD_CCtx_params const* CCtxParams, ZSTD_cParameter param, int* value) +{ + switch(param) + { + case ZSTD_c_format : + *value = CCtxParams->format; + break; + case ZSTD_c_compressionLevel : + *value = CCtxParams->compressionLevel; + break; + case ZSTD_c_windowLog : + *value = (int)CCtxParams->cParams.windowLog; + break; + case ZSTD_c_hashLog : + *value = (int)CCtxParams->cParams.hashLog; + break; + case ZSTD_c_chainLog : + *value = (int)CCtxParams->cParams.chainLog; + break; + case ZSTD_c_searchLog : + *value = CCtxParams->cParams.searchLog; + break; + case ZSTD_c_minMatch : + *value = CCtxParams->cParams.minMatch; + break; + case ZSTD_c_targetLength : + *value = CCtxParams->cParams.targetLength; + break; + case ZSTD_c_strategy : + *value = (unsigned)CCtxParams->cParams.strategy; + break; + case ZSTD_c_contentSizeFlag : + *value = CCtxParams->fParams.contentSizeFlag; + break; + case ZSTD_c_checksumFlag : + *value = CCtxParams->fParams.checksumFlag; + break; + case ZSTD_c_dictIDFlag : + *value = !CCtxParams->fParams.noDictIDFlag; + break; + case ZSTD_c_forceMaxWindow : + *value = CCtxParams->forceWindow; + break; + case ZSTD_c_forceAttachDict : + *value = CCtxParams->attachDictPref; + break; + case ZSTD_c_literalCompressionMode : + *value = CCtxParams->literalCompressionMode; + break; + case ZSTD_c_nbWorkers : + assert(CCtxParams->nbWorkers == 0); + *value = CCtxParams->nbWorkers; + break; + case ZSTD_c_jobSize : + RETURN_ERROR(parameter_unsupported, "not compiled with multithreading"); + case ZSTD_c_overlapLog : + RETURN_ERROR(parameter_unsupported, "not compiled with multithreading"); + case ZSTD_c_rsyncable : + RETURN_ERROR(parameter_unsupported, "not compiled with multithreading"); + case ZSTD_c_enableDedicatedDictSearch : + *value = CCtxParams->enableDedicatedDictSearch; + break; + case ZSTD_c_enableLongDistanceMatching : + *value = CCtxParams->ldmParams.enableLdm; + break; + case ZSTD_c_ldmHashLog : + *value = CCtxParams->ldmParams.hashLog; + break; + case ZSTD_c_ldmMinMatch : + *value = CCtxParams->ldmParams.minMatchLength; + break; + case ZSTD_c_ldmBucketSizeLog : + *value = CCtxParams->ldmParams.bucketSizeLog; + break; + case ZSTD_c_ldmHashRateLog : + *value = CCtxParams->ldmParams.hashRateLog; + break; + case ZSTD_c_targetCBlockSize : + *value = (int)CCtxParams->targetCBlockSize; + break; + case ZSTD_c_srcSizeHint : + *value = (int)CCtxParams->srcSizeHint; + break; + case ZSTD_c_stableInBuffer : + *value = (int)CCtxParams->inBufferMode; + break; + case ZSTD_c_stableOutBuffer : + *value = (int)CCtxParams->outBufferMode; + break; + case ZSTD_c_blockDelimiters : + *value = (int)CCtxParams->blockDelimiters; + break; + case ZSTD_c_validateSequences : + *value = (int)CCtxParams->validateSequences; + break; + default: RETURN_ERROR(parameter_unsupported, "unknown parameter"); + } + return 0; +} + +/* ZSTD_CCtx_setParametersUsingCCtxParams() : + * just applies `params` into `cctx` + * no action is performed, parameters are merely stored. + * If ZSTDMT is enabled, parameters are pushed to cctx->mtctx. + * This is possible even if a compression is ongoing. + * In which case, new parameters will be applied on the fly, starting with next compression job. + */ +size_t ZSTD_CCtx_setParametersUsingCCtxParams( + ZSTD_CCtx* cctx, const ZSTD_CCtx_params* params) +{ + DEBUGLOG(4, "ZSTD_CCtx_setParametersUsingCCtxParams"); + RETURN_ERROR_IF(cctx->streamStage != zcss_init, stage_wrong, + "The context is in the wrong stage!"); + RETURN_ERROR_IF(cctx->cdict, stage_wrong, + "Can't override parameters with cdict attached (some must " + "be inherited from the cdict)."); + + cctx->requestedParams = *params; + return 0; +} + +ZSTDLIB_API size_t ZSTD_CCtx_setPledgedSrcSize(ZSTD_CCtx* cctx, unsigned long long pledgedSrcSize) +{ + DEBUGLOG(4, "ZSTD_CCtx_setPledgedSrcSize to %u bytes", (U32)pledgedSrcSize); + RETURN_ERROR_IF(cctx->streamStage != zcss_init, stage_wrong, + "Can't set pledgedSrcSize when not in init stage."); + cctx->pledgedSrcSizePlusOne = pledgedSrcSize+1; + return 0; +} + +static ZSTD_compressionParameters ZSTD_dedicatedDictSearch_getCParams( + int const compressionLevel, + size_t const dictSize); +static int ZSTD_dedicatedDictSearch_isSupported( + const ZSTD_compressionParameters* cParams); +static void ZSTD_dedicatedDictSearch_revertCParams( + ZSTD_compressionParameters* cParams); + +/* + * Initializes the local dict using the requested parameters. + * NOTE: This does not use the pledged src size, because it may be used for more + * than one compression. + */ +static size_t ZSTD_initLocalDict(ZSTD_CCtx* cctx) +{ + ZSTD_localDict* const dl = &cctx->localDict; + if (dl->dict == NULL) { + /* No local dictionary. */ + assert(dl->dictBuffer == NULL); + assert(dl->cdict == NULL); + assert(dl->dictSize == 0); + return 0; + } + if (dl->cdict != NULL) { + assert(cctx->cdict == dl->cdict); + /* Local dictionary already initialized. */ + return 0; + } + assert(dl->dictSize > 0); + assert(cctx->cdict == NULL); + assert(cctx->prefixDict.dict == NULL); + + dl->cdict = ZSTD_createCDict_advanced2( + dl->dict, + dl->dictSize, + ZSTD_dlm_byRef, + dl->dictContentType, + &cctx->requestedParams, + cctx->customMem); + RETURN_ERROR_IF(!dl->cdict, memory_allocation, "ZSTD_createCDict_advanced failed"); + cctx->cdict = dl->cdict; + return 0; +} + +size_t ZSTD_CCtx_loadDictionary_advanced( + ZSTD_CCtx* cctx, const void* dict, size_t dictSize, + ZSTD_dictLoadMethod_e dictLoadMethod, ZSTD_dictContentType_e dictContentType) +{ + RETURN_ERROR_IF(cctx->streamStage != zcss_init, stage_wrong, + "Can't load a dictionary when ctx is not in init stage."); + DEBUGLOG(4, "ZSTD_CCtx_loadDictionary_advanced (size: %u)", (U32)dictSize); + ZSTD_clearAllDicts(cctx); /* in case one already exists */ + if (dict == NULL || dictSize == 0) /* no dictionary mode */ + return 0; + if (dictLoadMethod == ZSTD_dlm_byRef) { + cctx->localDict.dict = dict; + } else { + void* dictBuffer; + RETURN_ERROR_IF(cctx->staticSize, memory_allocation, + "no malloc for static CCtx"); + dictBuffer = ZSTD_customMalloc(dictSize, cctx->customMem); + RETURN_ERROR_IF(!dictBuffer, memory_allocation, "NULL pointer!"); + ZSTD_memcpy(dictBuffer, dict, dictSize); + cctx->localDict.dictBuffer = dictBuffer; + cctx->localDict.dict = dictBuffer; + } + cctx->localDict.dictSize = dictSize; + cctx->localDict.dictContentType = dictContentType; + return 0; +} + +ZSTDLIB_API size_t ZSTD_CCtx_loadDictionary_byReference( + ZSTD_CCtx* cctx, const void* dict, size_t dictSize) +{ + return ZSTD_CCtx_loadDictionary_advanced( + cctx, dict, dictSize, ZSTD_dlm_byRef, ZSTD_dct_auto); +} + +ZSTDLIB_API size_t ZSTD_CCtx_loadDictionary(ZSTD_CCtx* cctx, const void* dict, size_t dictSize) +{ + return ZSTD_CCtx_loadDictionary_advanced( + cctx, dict, dictSize, ZSTD_dlm_byCopy, ZSTD_dct_auto); +} + + +size_t ZSTD_CCtx_refCDict(ZSTD_CCtx* cctx, const ZSTD_CDict* cdict) +{ + RETURN_ERROR_IF(cctx->streamStage != zcss_init, stage_wrong, + "Can't ref a dict when ctx not in init stage."); + /* Free the existing local cdict (if any) to save memory. */ + ZSTD_clearAllDicts(cctx); + cctx->cdict = cdict; + return 0; +} + +size_t ZSTD_CCtx_refThreadPool(ZSTD_CCtx* cctx, ZSTD_threadPool* pool) +{ + RETURN_ERROR_IF(cctx->streamStage != zcss_init, stage_wrong, + "Can't ref a pool when ctx not in init stage."); + cctx->pool = pool; + return 0; +} + +size_t ZSTD_CCtx_refPrefix(ZSTD_CCtx* cctx, const void* prefix, size_t prefixSize) +{ + return ZSTD_CCtx_refPrefix_advanced(cctx, prefix, prefixSize, ZSTD_dct_rawContent); +} + +size_t ZSTD_CCtx_refPrefix_advanced( + ZSTD_CCtx* cctx, const void* prefix, size_t prefixSize, ZSTD_dictContentType_e dictContentType) +{ + RETURN_ERROR_IF(cctx->streamStage != zcss_init, stage_wrong, + "Can't ref a prefix when ctx not in init stage."); + ZSTD_clearAllDicts(cctx); + if (prefix != NULL && prefixSize > 0) { + cctx->prefixDict.dict = prefix; + cctx->prefixDict.dictSize = prefixSize; + cctx->prefixDict.dictContentType = dictContentType; + } + return 0; +} + +/*! ZSTD_CCtx_reset() : + * Also dumps dictionary */ +size_t ZSTD_CCtx_reset(ZSTD_CCtx* cctx, ZSTD_ResetDirective reset) +{ + if ( (reset == ZSTD_reset_session_only) + || (reset == ZSTD_reset_session_and_parameters) ) { + cctx->streamStage = zcss_init; + cctx->pledgedSrcSizePlusOne = 0; + } + if ( (reset == ZSTD_reset_parameters) + || (reset == ZSTD_reset_session_and_parameters) ) { + RETURN_ERROR_IF(cctx->streamStage != zcss_init, stage_wrong, + "Can't reset parameters only when not in init stage."); + ZSTD_clearAllDicts(cctx); + return ZSTD_CCtxParams_reset(&cctx->requestedParams); + } + return 0; +} + + +/* ZSTD_checkCParams() : + control CParam values remain within authorized range. + @return : 0, or an error code if one value is beyond authorized range */ +size_t ZSTD_checkCParams(ZSTD_compressionParameters cParams) +{ + BOUNDCHECK(ZSTD_c_windowLog, (int)cParams.windowLog); + BOUNDCHECK(ZSTD_c_chainLog, (int)cParams.chainLog); + BOUNDCHECK(ZSTD_c_hashLog, (int)cParams.hashLog); + BOUNDCHECK(ZSTD_c_searchLog, (int)cParams.searchLog); + BOUNDCHECK(ZSTD_c_minMatch, (int)cParams.minMatch); + BOUNDCHECK(ZSTD_c_targetLength,(int)cParams.targetLength); + BOUNDCHECK(ZSTD_c_strategy, cParams.strategy); + return 0; +} + +/* ZSTD_clampCParams() : + * make CParam values within valid range. + * @return : valid CParams */ +static ZSTD_compressionParameters +ZSTD_clampCParams(ZSTD_compressionParameters cParams) +{ +# define CLAMP_TYPE(cParam, val, type) { \ + ZSTD_bounds const bounds = ZSTD_cParam_getBounds(cParam); \ + if ((int)valbounds.upperBound) val=(type)bounds.upperBound; \ + } +# define CLAMP(cParam, val) CLAMP_TYPE(cParam, val, unsigned) + CLAMP(ZSTD_c_windowLog, cParams.windowLog); + CLAMP(ZSTD_c_chainLog, cParams.chainLog); + CLAMP(ZSTD_c_hashLog, cParams.hashLog); + CLAMP(ZSTD_c_searchLog, cParams.searchLog); + CLAMP(ZSTD_c_minMatch, cParams.minMatch); + CLAMP(ZSTD_c_targetLength,cParams.targetLength); + CLAMP_TYPE(ZSTD_c_strategy,cParams.strategy, ZSTD_strategy); + return cParams; +} + +/* ZSTD_cycleLog() : + * condition for correct operation : hashLog > 1 */ +U32 ZSTD_cycleLog(U32 hashLog, ZSTD_strategy strat) +{ + U32 const btScale = ((U32)strat >= (U32)ZSTD_btlazy2); + return hashLog - btScale; +} + +/* ZSTD_dictAndWindowLog() : + * Returns an adjusted window log that is large enough to fit the source and the dictionary. + * The zstd format says that the entire dictionary is valid if one byte of the dictionary + * is within the window. So the hashLog and chainLog should be large enough to reference both + * the dictionary and the window. So we must use this adjusted dictAndWindowLog when downsizing + * the hashLog and windowLog. + * NOTE: srcSize must not be ZSTD_CONTENTSIZE_UNKNOWN. + */ +static U32 ZSTD_dictAndWindowLog(U32 windowLog, U64 srcSize, U64 dictSize) +{ + const U64 maxWindowSize = 1ULL << ZSTD_WINDOWLOG_MAX; + /* No dictionary ==> No change */ + if (dictSize == 0) { + return windowLog; + } + assert(windowLog <= ZSTD_WINDOWLOG_MAX); + assert(srcSize != ZSTD_CONTENTSIZE_UNKNOWN); /* Handled in ZSTD_adjustCParams_internal() */ + { + U64 const windowSize = 1ULL << windowLog; + U64 const dictAndWindowSize = dictSize + windowSize; + /* If the window size is already large enough to fit both the source and the dictionary + * then just use the window size. Otherwise adjust so that it fits the dictionary and + * the window. + */ + if (windowSize >= dictSize + srcSize) { + return windowLog; /* Window size large enough already */ + } else if (dictAndWindowSize >= maxWindowSize) { + return ZSTD_WINDOWLOG_MAX; /* Larger than max window log */ + } else { + return ZSTD_highbit32((U32)dictAndWindowSize - 1) + 1; + } + } +} + +/* ZSTD_adjustCParams_internal() : + * optimize `cPar` for a specified input (`srcSize` and `dictSize`). + * mostly downsize to reduce memory consumption and initialization latency. + * `srcSize` can be ZSTD_CONTENTSIZE_UNKNOWN when not known. + * `mode` is the mode for parameter adjustment. See docs for `ZSTD_cParamMode_e`. + * note : `srcSize==0` means 0! + * condition : cPar is presumed validated (can be checked using ZSTD_checkCParams()). */ +static ZSTD_compressionParameters +ZSTD_adjustCParams_internal(ZSTD_compressionParameters cPar, + unsigned long long srcSize, + size_t dictSize, + ZSTD_cParamMode_e mode) +{ + const U64 minSrcSize = 513; /* (1<<9) + 1 */ + const U64 maxWindowResize = 1ULL << (ZSTD_WINDOWLOG_MAX-1); + assert(ZSTD_checkCParams(cPar)==0); + + switch (mode) { + case ZSTD_cpm_unknown: + case ZSTD_cpm_noAttachDict: + /* If we don't know the source size, don't make any + * assumptions about it. We will already have selected + * smaller parameters if a dictionary is in use. + */ + break; + case ZSTD_cpm_createCDict: + /* Assume a small source size when creating a dictionary + * with an unkown source size. + */ + if (dictSize && srcSize == ZSTD_CONTENTSIZE_UNKNOWN) + srcSize = minSrcSize; + break; + case ZSTD_cpm_attachDict: + /* Dictionary has its own dedicated parameters which have + * already been selected. We are selecting parameters + * for only the source. + */ + dictSize = 0; + break; + default: + assert(0); + break; + } + + /* resize windowLog if input is small enough, to use less memory */ + if ( (srcSize < maxWindowResize) + && (dictSize < maxWindowResize) ) { + U32 const tSize = (U32)(srcSize + dictSize); + static U32 const hashSizeMin = 1 << ZSTD_HASHLOG_MIN; + U32 const srcLog = (tSize < hashSizeMin) ? ZSTD_HASHLOG_MIN : + ZSTD_highbit32(tSize-1) + 1; + if (cPar.windowLog > srcLog) cPar.windowLog = srcLog; + } + if (srcSize != ZSTD_CONTENTSIZE_UNKNOWN) { + U32 const dictAndWindowLog = ZSTD_dictAndWindowLog(cPar.windowLog, (U64)srcSize, (U64)dictSize); + U32 const cycleLog = ZSTD_cycleLog(cPar.chainLog, cPar.strategy); + if (cPar.hashLog > dictAndWindowLog+1) cPar.hashLog = dictAndWindowLog+1; + if (cycleLog > dictAndWindowLog) + cPar.chainLog -= (cycleLog - dictAndWindowLog); + } + + if (cPar.windowLog < ZSTD_WINDOWLOG_ABSOLUTEMIN) + cPar.windowLog = ZSTD_WINDOWLOG_ABSOLUTEMIN; /* minimum wlog required for valid frame header */ + + return cPar; +} + +ZSTD_compressionParameters +ZSTD_adjustCParams(ZSTD_compressionParameters cPar, + unsigned long long srcSize, + size_t dictSize) +{ + cPar = ZSTD_clampCParams(cPar); /* resulting cPar is necessarily valid (all parameters within range) */ + if (srcSize == 0) srcSize = ZSTD_CONTENTSIZE_UNKNOWN; + return ZSTD_adjustCParams_internal(cPar, srcSize, dictSize, ZSTD_cpm_unknown); +} + +static ZSTD_compressionParameters ZSTD_getCParams_internal(int compressionLevel, unsigned long long srcSizeHint, size_t dictSize, ZSTD_cParamMode_e mode); +static ZSTD_parameters ZSTD_getParams_internal(int compressionLevel, unsigned long long srcSizeHint, size_t dictSize, ZSTD_cParamMode_e mode); + +static void ZSTD_overrideCParams( + ZSTD_compressionParameters* cParams, + const ZSTD_compressionParameters* overrides) +{ + if (overrides->windowLog) cParams->windowLog = overrides->windowLog; + if (overrides->hashLog) cParams->hashLog = overrides->hashLog; + if (overrides->chainLog) cParams->chainLog = overrides->chainLog; + if (overrides->searchLog) cParams->searchLog = overrides->searchLog; + if (overrides->minMatch) cParams->minMatch = overrides->minMatch; + if (overrides->targetLength) cParams->targetLength = overrides->targetLength; + if (overrides->strategy) cParams->strategy = overrides->strategy; +} + +ZSTD_compressionParameters ZSTD_getCParamsFromCCtxParams( + const ZSTD_CCtx_params* CCtxParams, U64 srcSizeHint, size_t dictSize, ZSTD_cParamMode_e mode) +{ + ZSTD_compressionParameters cParams; + if (srcSizeHint == ZSTD_CONTENTSIZE_UNKNOWN && CCtxParams->srcSizeHint > 0) { + srcSizeHint = CCtxParams->srcSizeHint; + } + cParams = ZSTD_getCParams_internal(CCtxParams->compressionLevel, srcSizeHint, dictSize, mode); + if (CCtxParams->ldmParams.enableLdm) cParams.windowLog = ZSTD_LDM_DEFAULT_WINDOW_LOG; + ZSTD_overrideCParams(&cParams, &CCtxParams->cParams); + assert(!ZSTD_checkCParams(cParams)); + /* srcSizeHint == 0 means 0 */ + return ZSTD_adjustCParams_internal(cParams, srcSizeHint, dictSize, mode); +} + +static size_t +ZSTD_sizeof_matchState(const ZSTD_compressionParameters* const cParams, + const U32 forCCtx) +{ + size_t const chainSize = (cParams->strategy == ZSTD_fast) ? 0 : ((size_t)1 << cParams->chainLog); + size_t const hSize = ((size_t)1) << cParams->hashLog; + U32 const hashLog3 = (forCCtx && cParams->minMatch==3) ? MIN(ZSTD_HASHLOG3_MAX, cParams->windowLog) : 0; + size_t const h3Size = hashLog3 ? ((size_t)1) << hashLog3 : 0; + /* We don't use ZSTD_cwksp_alloc_size() here because the tables aren't + * surrounded by redzones in ASAN. */ + size_t const tableSpace = chainSize * sizeof(U32) + + hSize * sizeof(U32) + + h3Size * sizeof(U32); + size_t const optPotentialSpace = + ZSTD_cwksp_alloc_size((MaxML+1) * sizeof(U32)) + + ZSTD_cwksp_alloc_size((MaxLL+1) * sizeof(U32)) + + ZSTD_cwksp_alloc_size((MaxOff+1) * sizeof(U32)) + + ZSTD_cwksp_alloc_size((1<strategy >= ZSTD_btopt)) + ? optPotentialSpace + : 0; + DEBUGLOG(4, "chainSize: %u - hSize: %u - h3Size: %u", + (U32)chainSize, (U32)hSize, (U32)h3Size); + return tableSpace + optSpace; +} + +static size_t ZSTD_estimateCCtxSize_usingCCtxParams_internal( + const ZSTD_compressionParameters* cParams, + const ldmParams_t* ldmParams, + const int isStatic, + const size_t buffInSize, + const size_t buffOutSize, + const U64 pledgedSrcSize) +{ + size_t const windowSize = MAX(1, (size_t)MIN(((U64)1 << cParams->windowLog), pledgedSrcSize)); + size_t const blockSize = MIN(ZSTD_BLOCKSIZE_MAX, windowSize); + U32 const divider = (cParams->minMatch==3) ? 3 : 4; + size_t const maxNbSeq = blockSize / divider; + size_t const tokenSpace = ZSTD_cwksp_alloc_size(WILDCOPY_OVERLENGTH + blockSize) + + ZSTD_cwksp_alloc_size(maxNbSeq * sizeof(seqDef)) + + 3 * ZSTD_cwksp_alloc_size(maxNbSeq * sizeof(BYTE)); + size_t const entropySpace = ZSTD_cwksp_alloc_size(ENTROPY_WORKSPACE_SIZE); + size_t const blockStateSpace = 2 * ZSTD_cwksp_alloc_size(sizeof(ZSTD_compressedBlockState_t)); + size_t const matchStateSize = ZSTD_sizeof_matchState(cParams, /* forCCtx */ 1); + + size_t const ldmSpace = ZSTD_ldm_getTableSize(*ldmParams); + size_t const maxNbLdmSeq = ZSTD_ldm_getMaxNbSeq(*ldmParams, blockSize); + size_t const ldmSeqSpace = ldmParams->enableLdm ? + ZSTD_cwksp_alloc_size(maxNbLdmSeq * sizeof(rawSeq)) : 0; + + + size_t const bufferSpace = ZSTD_cwksp_alloc_size(buffInSize) + + ZSTD_cwksp_alloc_size(buffOutSize); + + size_t const cctxSpace = isStatic ? ZSTD_cwksp_alloc_size(sizeof(ZSTD_CCtx)) : 0; + + size_t const neededSpace = + cctxSpace + + entropySpace + + blockStateSpace + + ldmSpace + + ldmSeqSpace + + matchStateSize + + tokenSpace + + bufferSpace; + + DEBUGLOG(5, "estimate workspace : %u", (U32)neededSpace); + return neededSpace; +} + +size_t ZSTD_estimateCCtxSize_usingCCtxParams(const ZSTD_CCtx_params* params) +{ + ZSTD_compressionParameters const cParams = + ZSTD_getCParamsFromCCtxParams(params, ZSTD_CONTENTSIZE_UNKNOWN, 0, ZSTD_cpm_noAttachDict); + + RETURN_ERROR_IF(params->nbWorkers > 0, GENERIC, "Estimate CCtx size is supported for single-threaded compression only."); + /* estimateCCtxSize is for one-shot compression. So no buffers should + * be needed. However, we still allocate two 0-sized buffers, which can + * take space under ASAN. */ + return ZSTD_estimateCCtxSize_usingCCtxParams_internal( + &cParams, ¶ms->ldmParams, 1, 0, 0, ZSTD_CONTENTSIZE_UNKNOWN); +} + +size_t ZSTD_estimateCCtxSize_usingCParams(ZSTD_compressionParameters cParams) +{ + ZSTD_CCtx_params const params = ZSTD_makeCCtxParamsFromCParams(cParams); + return ZSTD_estimateCCtxSize_usingCCtxParams(¶ms); +} + +static size_t ZSTD_estimateCCtxSize_internal(int compressionLevel) +{ + int tier = 0; + size_t largestSize = 0; + static const unsigned long long srcSizeTiers[4] = {16 KB, 128 KB, 256 KB, ZSTD_CONTENTSIZE_UNKNOWN}; + for (; tier < 4; ++tier) { + /* Choose the set of cParams for a given level across all srcSizes that give the largest cctxSize */ + ZSTD_compressionParameters const cParams = ZSTD_getCParams_internal(compressionLevel, srcSizeTiers[tier], 0, ZSTD_cpm_noAttachDict); + largestSize = MAX(ZSTD_estimateCCtxSize_usingCParams(cParams), largestSize); + } + return largestSize; +} + +size_t ZSTD_estimateCCtxSize(int compressionLevel) +{ + int level; + size_t memBudget = 0; + for (level=MIN(compressionLevel, 1); level<=compressionLevel; level++) { + /* Ensure monotonically increasing memory usage as compression level increases */ + size_t const newMB = ZSTD_estimateCCtxSize_internal(level); + if (newMB > memBudget) memBudget = newMB; + } + return memBudget; +} + +size_t ZSTD_estimateCStreamSize_usingCCtxParams(const ZSTD_CCtx_params* params) +{ + RETURN_ERROR_IF(params->nbWorkers > 0, GENERIC, "Estimate CCtx size is supported for single-threaded compression only."); + { ZSTD_compressionParameters const cParams = + ZSTD_getCParamsFromCCtxParams(params, ZSTD_CONTENTSIZE_UNKNOWN, 0, ZSTD_cpm_noAttachDict); + size_t const blockSize = MIN(ZSTD_BLOCKSIZE_MAX, (size_t)1 << cParams.windowLog); + size_t const inBuffSize = (params->inBufferMode == ZSTD_bm_buffered) + ? ((size_t)1 << cParams.windowLog) + blockSize + : 0; + size_t const outBuffSize = (params->outBufferMode == ZSTD_bm_buffered) + ? ZSTD_compressBound(blockSize) + 1 + : 0; + + return ZSTD_estimateCCtxSize_usingCCtxParams_internal( + &cParams, ¶ms->ldmParams, 1, inBuffSize, outBuffSize, + ZSTD_CONTENTSIZE_UNKNOWN); + } +} + +size_t ZSTD_estimateCStreamSize_usingCParams(ZSTD_compressionParameters cParams) +{ + ZSTD_CCtx_params const params = ZSTD_makeCCtxParamsFromCParams(cParams); + return ZSTD_estimateCStreamSize_usingCCtxParams(¶ms); +} + +static size_t ZSTD_estimateCStreamSize_internal(int compressionLevel) +{ + ZSTD_compressionParameters const cParams = ZSTD_getCParams_internal(compressionLevel, ZSTD_CONTENTSIZE_UNKNOWN, 0, ZSTD_cpm_noAttachDict); + return ZSTD_estimateCStreamSize_usingCParams(cParams); +} + +size_t ZSTD_estimateCStreamSize(int compressionLevel) +{ + int level; + size_t memBudget = 0; + for (level=MIN(compressionLevel, 1); level<=compressionLevel; level++) { + size_t const newMB = ZSTD_estimateCStreamSize_internal(level); + if (newMB > memBudget) memBudget = newMB; + } + return memBudget; +} + +/* ZSTD_getFrameProgression(): + * tells how much data has been consumed (input) and produced (output) for current frame. + * able to count progression inside worker threads (non-blocking mode). + */ +ZSTD_frameProgression ZSTD_getFrameProgression(const ZSTD_CCtx* cctx) +{ + { ZSTD_frameProgression fp; + size_t const buffered = (cctx->inBuff == NULL) ? 0 : + cctx->inBuffPos - cctx->inToCompress; + if (buffered) assert(cctx->inBuffPos >= cctx->inToCompress); + assert(buffered <= ZSTD_BLOCKSIZE_MAX); + fp.ingested = cctx->consumedSrcSize + buffered; + fp.consumed = cctx->consumedSrcSize; + fp.produced = cctx->producedCSize; + fp.flushed = cctx->producedCSize; /* simplified; some data might still be left within streaming output buffer */ + fp.currentJobID = 0; + fp.nbActiveWorkers = 0; + return fp; +} } + +/*! ZSTD_toFlushNow() + * Only useful for multithreading scenarios currently (nbWorkers >= 1). + */ +size_t ZSTD_toFlushNow(ZSTD_CCtx* cctx) +{ + (void)cctx; + return 0; /* over-simplification; could also check if context is currently running in streaming mode, and in which case, report how many bytes are left to be flushed within output buffer */ +} + +static void ZSTD_assertEqualCParams(ZSTD_compressionParameters cParams1, + ZSTD_compressionParameters cParams2) +{ + (void)cParams1; + (void)cParams2; + assert(cParams1.windowLog == cParams2.windowLog); + assert(cParams1.chainLog == cParams2.chainLog); + assert(cParams1.hashLog == cParams2.hashLog); + assert(cParams1.searchLog == cParams2.searchLog); + assert(cParams1.minMatch == cParams2.minMatch); + assert(cParams1.targetLength == cParams2.targetLength); + assert(cParams1.strategy == cParams2.strategy); +} + +void ZSTD_reset_compressedBlockState(ZSTD_compressedBlockState_t* bs) +{ + int i; + for (i = 0; i < ZSTD_REP_NUM; ++i) + bs->rep[i] = repStartValue[i]; + bs->entropy.huf.repeatMode = HUF_repeat_none; + bs->entropy.fse.offcode_repeatMode = FSE_repeat_none; + bs->entropy.fse.matchlength_repeatMode = FSE_repeat_none; + bs->entropy.fse.litlength_repeatMode = FSE_repeat_none; +} + +/*! ZSTD_invalidateMatchState() + * Invalidate all the matches in the match finder tables. + * Requires nextSrc and base to be set (can be NULL). + */ +static void ZSTD_invalidateMatchState(ZSTD_matchState_t* ms) +{ + ZSTD_window_clear(&ms->window); + + ms->nextToUpdate = ms->window.dictLimit; + ms->loadedDictEnd = 0; + ms->opt.litLengthSum = 0; /* force reset of btopt stats */ + ms->dictMatchState = NULL; +} + +/* + * Controls, for this matchState reset, whether the tables need to be cleared / + * prepared for the coming compression (ZSTDcrp_makeClean), or whether the + * tables can be left unclean (ZSTDcrp_leaveDirty), because we know that a + * subsequent operation will overwrite the table space anyways (e.g., copying + * the matchState contents in from a CDict). + */ +typedef enum { + ZSTDcrp_makeClean, + ZSTDcrp_leaveDirty +} ZSTD_compResetPolicy_e; + +/* + * Controls, for this matchState reset, whether indexing can continue where it + * left off (ZSTDirp_continue), or whether it needs to be restarted from zero + * (ZSTDirp_reset). + */ +typedef enum { + ZSTDirp_continue, + ZSTDirp_reset +} ZSTD_indexResetPolicy_e; + +typedef enum { + ZSTD_resetTarget_CDict, + ZSTD_resetTarget_CCtx +} ZSTD_resetTarget_e; + +static size_t +ZSTD_reset_matchState(ZSTD_matchState_t* ms, + ZSTD_cwksp* ws, + const ZSTD_compressionParameters* cParams, + const ZSTD_compResetPolicy_e crp, + const ZSTD_indexResetPolicy_e forceResetIndex, + const ZSTD_resetTarget_e forWho) +{ + size_t const chainSize = (cParams->strategy == ZSTD_fast) ? 0 : ((size_t)1 << cParams->chainLog); + size_t const hSize = ((size_t)1) << cParams->hashLog; + U32 const hashLog3 = ((forWho == ZSTD_resetTarget_CCtx) && cParams->minMatch==3) ? MIN(ZSTD_HASHLOG3_MAX, cParams->windowLog) : 0; + size_t const h3Size = hashLog3 ? ((size_t)1) << hashLog3 : 0; + + DEBUGLOG(4, "reset indices : %u", forceResetIndex == ZSTDirp_reset); + if (forceResetIndex == ZSTDirp_reset) { + ZSTD_window_init(&ms->window); + ZSTD_cwksp_mark_tables_dirty(ws); + } + + ms->hashLog3 = hashLog3; + + ZSTD_invalidateMatchState(ms); + + assert(!ZSTD_cwksp_reserve_failed(ws)); /* check that allocation hasn't already failed */ + + ZSTD_cwksp_clear_tables(ws); + + DEBUGLOG(5, "reserving table space"); + /* table Space */ + ms->hashTable = (U32*)ZSTD_cwksp_reserve_table(ws, hSize * sizeof(U32)); + ms->chainTable = (U32*)ZSTD_cwksp_reserve_table(ws, chainSize * sizeof(U32)); + ms->hashTable3 = (U32*)ZSTD_cwksp_reserve_table(ws, h3Size * sizeof(U32)); + RETURN_ERROR_IF(ZSTD_cwksp_reserve_failed(ws), memory_allocation, + "failed a workspace allocation in ZSTD_reset_matchState"); + + DEBUGLOG(4, "reset table : %u", crp!=ZSTDcrp_leaveDirty); + if (crp!=ZSTDcrp_leaveDirty) { + /* reset tables only */ + ZSTD_cwksp_clean_tables(ws); + } + + /* opt parser space */ + if ((forWho == ZSTD_resetTarget_CCtx) && (cParams->strategy >= ZSTD_btopt)) { + DEBUGLOG(4, "reserving optimal parser space"); + ms->opt.litFreq = (unsigned*)ZSTD_cwksp_reserve_aligned(ws, (1<opt.litLengthFreq = (unsigned*)ZSTD_cwksp_reserve_aligned(ws, (MaxLL+1) * sizeof(unsigned)); + ms->opt.matchLengthFreq = (unsigned*)ZSTD_cwksp_reserve_aligned(ws, (MaxML+1) * sizeof(unsigned)); + ms->opt.offCodeFreq = (unsigned*)ZSTD_cwksp_reserve_aligned(ws, (MaxOff+1) * sizeof(unsigned)); + ms->opt.matchTable = (ZSTD_match_t*)ZSTD_cwksp_reserve_aligned(ws, (ZSTD_OPT_NUM+1) * sizeof(ZSTD_match_t)); + ms->opt.priceTable = (ZSTD_optimal_t*)ZSTD_cwksp_reserve_aligned(ws, (ZSTD_OPT_NUM+1) * sizeof(ZSTD_optimal_t)); + } + + ms->cParams = *cParams; + + RETURN_ERROR_IF(ZSTD_cwksp_reserve_failed(ws), memory_allocation, + "failed a workspace allocation in ZSTD_reset_matchState"); + + return 0; +} + +/* ZSTD_indexTooCloseToMax() : + * minor optimization : prefer memset() rather than reduceIndex() + * which is measurably slow in some circumstances (reported for Visual Studio). + * Works when re-using a context for a lot of smallish inputs : + * if all inputs are smaller than ZSTD_INDEXOVERFLOW_MARGIN, + * memset() will be triggered before reduceIndex(). + */ +#define ZSTD_INDEXOVERFLOW_MARGIN (16 MB) +static int ZSTD_indexTooCloseToMax(ZSTD_window_t w) +{ + return (size_t)(w.nextSrc - w.base) > (ZSTD_CURRENT_MAX - ZSTD_INDEXOVERFLOW_MARGIN); +} + +/*! ZSTD_resetCCtx_internal() : + note : `params` are assumed fully validated at this stage */ +static size_t ZSTD_resetCCtx_internal(ZSTD_CCtx* zc, + ZSTD_CCtx_params params, + U64 const pledgedSrcSize, + ZSTD_compResetPolicy_e const crp, + ZSTD_buffered_policy_e const zbuff) +{ + ZSTD_cwksp* const ws = &zc->workspace; + DEBUGLOG(4, "ZSTD_resetCCtx_internal: pledgedSrcSize=%u, wlog=%u", + (U32)pledgedSrcSize, params.cParams.windowLog); + assert(!ZSTD_isError(ZSTD_checkCParams(params.cParams))); + + zc->isFirstBlock = 1; + + if (params.ldmParams.enableLdm) { + /* Adjust long distance matching parameters */ + ZSTD_ldm_adjustParameters(¶ms.ldmParams, ¶ms.cParams); + assert(params.ldmParams.hashLog >= params.ldmParams.bucketSizeLog); + assert(params.ldmParams.hashRateLog < 32); + } + + { size_t const windowSize = MAX(1, (size_t)MIN(((U64)1 << params.cParams.windowLog), pledgedSrcSize)); + size_t const blockSize = MIN(ZSTD_BLOCKSIZE_MAX, windowSize); + U32 const divider = (params.cParams.minMatch==3) ? 3 : 4; + size_t const maxNbSeq = blockSize / divider; + size_t const buffOutSize = (zbuff == ZSTDb_buffered && params.outBufferMode == ZSTD_bm_buffered) + ? ZSTD_compressBound(blockSize) + 1 + : 0; + size_t const buffInSize = (zbuff == ZSTDb_buffered && params.inBufferMode == ZSTD_bm_buffered) + ? windowSize + blockSize + : 0; + size_t const maxNbLdmSeq = ZSTD_ldm_getMaxNbSeq(params.ldmParams, blockSize); + + int const indexTooClose = ZSTD_indexTooCloseToMax(zc->blockState.matchState.window); + ZSTD_indexResetPolicy_e needsIndexReset = + (!indexTooClose && zc->initialized) ? ZSTDirp_continue : ZSTDirp_reset; + + size_t const neededSpace = + ZSTD_estimateCCtxSize_usingCCtxParams_internal( + ¶ms.cParams, ¶ms.ldmParams, zc->staticSize != 0, + buffInSize, buffOutSize, pledgedSrcSize); + FORWARD_IF_ERROR(neededSpace, "cctx size estimate failed!"); + + if (!zc->staticSize) ZSTD_cwksp_bump_oversized_duration(ws, 0); + + /* Check if workspace is large enough, alloc a new one if needed */ + { + int const workspaceTooSmall = ZSTD_cwksp_sizeof(ws) < neededSpace; + int const workspaceWasteful = ZSTD_cwksp_check_wasteful(ws, neededSpace); + + DEBUGLOG(4, "Need %zu B workspace", neededSpace); + DEBUGLOG(4, "windowSize: %zu - blockSize: %zu", windowSize, blockSize); + + if (workspaceTooSmall || workspaceWasteful) { + DEBUGLOG(4, "Resize workspaceSize from %zuKB to %zuKB", + ZSTD_cwksp_sizeof(ws) >> 10, + neededSpace >> 10); + + RETURN_ERROR_IF(zc->staticSize, memory_allocation, "static cctx : no resize"); + + needsIndexReset = ZSTDirp_reset; + + ZSTD_cwksp_free(ws, zc->customMem); + FORWARD_IF_ERROR(ZSTD_cwksp_create(ws, neededSpace, zc->customMem), ""); + + DEBUGLOG(5, "reserving object space"); + /* Statically sized space. + * entropyWorkspace never moves, + * though prev/next block swap places */ + assert(ZSTD_cwksp_check_available(ws, 2 * sizeof(ZSTD_compressedBlockState_t))); + zc->blockState.prevCBlock = (ZSTD_compressedBlockState_t*) ZSTD_cwksp_reserve_object(ws, sizeof(ZSTD_compressedBlockState_t)); + RETURN_ERROR_IF(zc->blockState.prevCBlock == NULL, memory_allocation, "couldn't allocate prevCBlock"); + zc->blockState.nextCBlock = (ZSTD_compressedBlockState_t*) ZSTD_cwksp_reserve_object(ws, sizeof(ZSTD_compressedBlockState_t)); + RETURN_ERROR_IF(zc->blockState.nextCBlock == NULL, memory_allocation, "couldn't allocate nextCBlock"); + zc->entropyWorkspace = (U32*) ZSTD_cwksp_reserve_object(ws, ENTROPY_WORKSPACE_SIZE); + RETURN_ERROR_IF(zc->blockState.nextCBlock == NULL, memory_allocation, "couldn't allocate entropyWorkspace"); + } } + + ZSTD_cwksp_clear(ws); + + /* init params */ + zc->appliedParams = params; + zc->blockState.matchState.cParams = params.cParams; + zc->pledgedSrcSizePlusOne = pledgedSrcSize+1; + zc->consumedSrcSize = 0; + zc->producedCSize = 0; + if (pledgedSrcSize == ZSTD_CONTENTSIZE_UNKNOWN) + zc->appliedParams.fParams.contentSizeFlag = 0; + DEBUGLOG(4, "pledged content size : %u ; flag : %u", + (unsigned)pledgedSrcSize, zc->appliedParams.fParams.contentSizeFlag); + zc->blockSize = blockSize; + + xxh64_reset(&zc->xxhState, 0); + zc->stage = ZSTDcs_init; + zc->dictID = 0; + zc->dictContentSize = 0; + + ZSTD_reset_compressedBlockState(zc->blockState.prevCBlock); + + /* ZSTD_wildcopy() is used to copy into the literals buffer, + * so we have to oversize the buffer by WILDCOPY_OVERLENGTH bytes. + */ + zc->seqStore.litStart = ZSTD_cwksp_reserve_buffer(ws, blockSize + WILDCOPY_OVERLENGTH); + zc->seqStore.maxNbLit = blockSize; + + /* buffers */ + zc->bufferedPolicy = zbuff; + zc->inBuffSize = buffInSize; + zc->inBuff = (char*)ZSTD_cwksp_reserve_buffer(ws, buffInSize); + zc->outBuffSize = buffOutSize; + zc->outBuff = (char*)ZSTD_cwksp_reserve_buffer(ws, buffOutSize); + + /* ldm bucketOffsets table */ + if (params.ldmParams.enableLdm) { + /* TODO: avoid memset? */ + size_t const numBuckets = + ((size_t)1) << (params.ldmParams.hashLog - + params.ldmParams.bucketSizeLog); + zc->ldmState.bucketOffsets = ZSTD_cwksp_reserve_buffer(ws, numBuckets); + ZSTD_memset(zc->ldmState.bucketOffsets, 0, numBuckets); + } + + /* sequences storage */ + ZSTD_referenceExternalSequences(zc, NULL, 0); + zc->seqStore.maxNbSeq = maxNbSeq; + zc->seqStore.llCode = ZSTD_cwksp_reserve_buffer(ws, maxNbSeq * sizeof(BYTE)); + zc->seqStore.mlCode = ZSTD_cwksp_reserve_buffer(ws, maxNbSeq * sizeof(BYTE)); + zc->seqStore.ofCode = ZSTD_cwksp_reserve_buffer(ws, maxNbSeq * sizeof(BYTE)); + zc->seqStore.sequencesStart = (seqDef*)ZSTD_cwksp_reserve_aligned(ws, maxNbSeq * sizeof(seqDef)); + + FORWARD_IF_ERROR(ZSTD_reset_matchState( + &zc->blockState.matchState, + ws, + ¶ms.cParams, + crp, + needsIndexReset, + ZSTD_resetTarget_CCtx), ""); + + /* ldm hash table */ + if (params.ldmParams.enableLdm) { + /* TODO: avoid memset? */ + size_t const ldmHSize = ((size_t)1) << params.ldmParams.hashLog; + zc->ldmState.hashTable = (ldmEntry_t*)ZSTD_cwksp_reserve_aligned(ws, ldmHSize * sizeof(ldmEntry_t)); + ZSTD_memset(zc->ldmState.hashTable, 0, ldmHSize * sizeof(ldmEntry_t)); + zc->ldmSequences = (rawSeq*)ZSTD_cwksp_reserve_aligned(ws, maxNbLdmSeq * sizeof(rawSeq)); + zc->maxNbLdmSequences = maxNbLdmSeq; + + ZSTD_window_init(&zc->ldmState.window); + ZSTD_window_clear(&zc->ldmState.window); + zc->ldmState.loadedDictEnd = 0; + } + + /* Due to alignment, when reusing a workspace, we can actually consume + * up to 3 extra bytes for alignment. See the comments in zstd_cwksp.h + */ + assert(ZSTD_cwksp_used(ws) >= neededSpace && + ZSTD_cwksp_used(ws) <= neededSpace + 3); + + DEBUGLOG(3, "wksp: finished allocating, %zd bytes remain available", ZSTD_cwksp_available_space(ws)); + zc->initialized = 1; + + return 0; + } +} + +/* ZSTD_invalidateRepCodes() : + * ensures next compression will not use repcodes from previous block. + * Note : only works with regular variant; + * do not use with extDict variant ! */ +void ZSTD_invalidateRepCodes(ZSTD_CCtx* cctx) { + int i; + for (i=0; iblockState.prevCBlock->rep[i] = 0; + assert(!ZSTD_window_hasExtDict(cctx->blockState.matchState.window)); +} + +/* These are the approximate sizes for each strategy past which copying the + * dictionary tables into the working context is faster than using them + * in-place. + */ +static const size_t attachDictSizeCutoffs[ZSTD_STRATEGY_MAX+1] = { + 8 KB, /* unused */ + 8 KB, /* ZSTD_fast */ + 16 KB, /* ZSTD_dfast */ + 32 KB, /* ZSTD_greedy */ + 32 KB, /* ZSTD_lazy */ + 32 KB, /* ZSTD_lazy2 */ + 32 KB, /* ZSTD_btlazy2 */ + 32 KB, /* ZSTD_btopt */ + 8 KB, /* ZSTD_btultra */ + 8 KB /* ZSTD_btultra2 */ +}; + +static int ZSTD_shouldAttachDict(const ZSTD_CDict* cdict, + const ZSTD_CCtx_params* params, + U64 pledgedSrcSize) +{ + size_t cutoff = attachDictSizeCutoffs[cdict->matchState.cParams.strategy]; + int const dedicatedDictSearch = cdict->matchState.dedicatedDictSearch; + return dedicatedDictSearch + || ( ( pledgedSrcSize <= cutoff + || pledgedSrcSize == ZSTD_CONTENTSIZE_UNKNOWN + || params->attachDictPref == ZSTD_dictForceAttach ) + && params->attachDictPref != ZSTD_dictForceCopy + && !params->forceWindow ); /* dictMatchState isn't correctly + * handled in _enforceMaxDist */ +} + +static size_t +ZSTD_resetCCtx_byAttachingCDict(ZSTD_CCtx* cctx, + const ZSTD_CDict* cdict, + ZSTD_CCtx_params params, + U64 pledgedSrcSize, + ZSTD_buffered_policy_e zbuff) +{ + { + ZSTD_compressionParameters adjusted_cdict_cParams = cdict->matchState.cParams; + unsigned const windowLog = params.cParams.windowLog; + assert(windowLog != 0); + /* Resize working context table params for input only, since the dict + * has its own tables. */ + /* pledgedSrcSize == 0 means 0! */ + + if (cdict->matchState.dedicatedDictSearch) { + ZSTD_dedicatedDictSearch_revertCParams(&adjusted_cdict_cParams); + } + + params.cParams = ZSTD_adjustCParams_internal(adjusted_cdict_cParams, pledgedSrcSize, + cdict->dictContentSize, ZSTD_cpm_attachDict); + params.cParams.windowLog = windowLog; + FORWARD_IF_ERROR(ZSTD_resetCCtx_internal(cctx, params, pledgedSrcSize, + ZSTDcrp_makeClean, zbuff), ""); + assert(cctx->appliedParams.cParams.strategy == adjusted_cdict_cParams.strategy); + } + + { const U32 cdictEnd = (U32)( cdict->matchState.window.nextSrc + - cdict->matchState.window.base); + const U32 cdictLen = cdictEnd - cdict->matchState.window.dictLimit; + if (cdictLen == 0) { + /* don't even attach dictionaries with no contents */ + DEBUGLOG(4, "skipping attaching empty dictionary"); + } else { + DEBUGLOG(4, "attaching dictionary into context"); + cctx->blockState.matchState.dictMatchState = &cdict->matchState; + + /* prep working match state so dict matches never have negative indices + * when they are translated to the working context's index space. */ + if (cctx->blockState.matchState.window.dictLimit < cdictEnd) { + cctx->blockState.matchState.window.nextSrc = + cctx->blockState.matchState.window.base + cdictEnd; + ZSTD_window_clear(&cctx->blockState.matchState.window); + } + /* loadedDictEnd is expressed within the referential of the active context */ + cctx->blockState.matchState.loadedDictEnd = cctx->blockState.matchState.window.dictLimit; + } } + + cctx->dictID = cdict->dictID; + cctx->dictContentSize = cdict->dictContentSize; + + /* copy block state */ + ZSTD_memcpy(cctx->blockState.prevCBlock, &cdict->cBlockState, sizeof(cdict->cBlockState)); + + return 0; +} + +static size_t ZSTD_resetCCtx_byCopyingCDict(ZSTD_CCtx* cctx, + const ZSTD_CDict* cdict, + ZSTD_CCtx_params params, + U64 pledgedSrcSize, + ZSTD_buffered_policy_e zbuff) +{ + const ZSTD_compressionParameters *cdict_cParams = &cdict->matchState.cParams; + + assert(!cdict->matchState.dedicatedDictSearch); + + DEBUGLOG(4, "copying dictionary into context"); + + { unsigned const windowLog = params.cParams.windowLog; + assert(windowLog != 0); + /* Copy only compression parameters related to tables. */ + params.cParams = *cdict_cParams; + params.cParams.windowLog = windowLog; + FORWARD_IF_ERROR(ZSTD_resetCCtx_internal(cctx, params, pledgedSrcSize, + ZSTDcrp_leaveDirty, zbuff), ""); + assert(cctx->appliedParams.cParams.strategy == cdict_cParams->strategy); + assert(cctx->appliedParams.cParams.hashLog == cdict_cParams->hashLog); + assert(cctx->appliedParams.cParams.chainLog == cdict_cParams->chainLog); + } + + ZSTD_cwksp_mark_tables_dirty(&cctx->workspace); + + /* copy tables */ + { size_t const chainSize = (cdict_cParams->strategy == ZSTD_fast) ? 0 : ((size_t)1 << cdict_cParams->chainLog); + size_t const hSize = (size_t)1 << cdict_cParams->hashLog; + + ZSTD_memcpy(cctx->blockState.matchState.hashTable, + cdict->matchState.hashTable, + hSize * sizeof(U32)); + ZSTD_memcpy(cctx->blockState.matchState.chainTable, + cdict->matchState.chainTable, + chainSize * sizeof(U32)); + } + + /* Zero the hashTable3, since the cdict never fills it */ + { int const h3log = cctx->blockState.matchState.hashLog3; + size_t const h3Size = h3log ? ((size_t)1 << h3log) : 0; + assert(cdict->matchState.hashLog3 == 0); + ZSTD_memset(cctx->blockState.matchState.hashTable3, 0, h3Size * sizeof(U32)); + } + + ZSTD_cwksp_mark_tables_clean(&cctx->workspace); + + /* copy dictionary offsets */ + { ZSTD_matchState_t const* srcMatchState = &cdict->matchState; + ZSTD_matchState_t* dstMatchState = &cctx->blockState.matchState; + dstMatchState->window = srcMatchState->window; + dstMatchState->nextToUpdate = srcMatchState->nextToUpdate; + dstMatchState->loadedDictEnd= srcMatchState->loadedDictEnd; + } + + cctx->dictID = cdict->dictID; + cctx->dictContentSize = cdict->dictContentSize; + + /* copy block state */ + ZSTD_memcpy(cctx->blockState.prevCBlock, &cdict->cBlockState, sizeof(cdict->cBlockState)); + + return 0; +} + +/* We have a choice between copying the dictionary context into the working + * context, or referencing the dictionary context from the working context + * in-place. We decide here which strategy to use. */ +static size_t ZSTD_resetCCtx_usingCDict(ZSTD_CCtx* cctx, + const ZSTD_CDict* cdict, + const ZSTD_CCtx_params* params, + U64 pledgedSrcSize, + ZSTD_buffered_policy_e zbuff) +{ + + DEBUGLOG(4, "ZSTD_resetCCtx_usingCDict (pledgedSrcSize=%u)", + (unsigned)pledgedSrcSize); + + if (ZSTD_shouldAttachDict(cdict, params, pledgedSrcSize)) { + return ZSTD_resetCCtx_byAttachingCDict( + cctx, cdict, *params, pledgedSrcSize, zbuff); + } else { + return ZSTD_resetCCtx_byCopyingCDict( + cctx, cdict, *params, pledgedSrcSize, zbuff); + } +} + +/*! ZSTD_copyCCtx_internal() : + * Duplicate an existing context `srcCCtx` into another one `dstCCtx`. + * Only works during stage ZSTDcs_init (i.e. after creation, but before first call to ZSTD_compressContinue()). + * The "context", in this case, refers to the hash and chain tables, + * entropy tables, and dictionary references. + * `windowLog` value is enforced if != 0, otherwise value is copied from srcCCtx. + * @return : 0, or an error code */ +static size_t ZSTD_copyCCtx_internal(ZSTD_CCtx* dstCCtx, + const ZSTD_CCtx* srcCCtx, + ZSTD_frameParameters fParams, + U64 pledgedSrcSize, + ZSTD_buffered_policy_e zbuff) +{ + DEBUGLOG(5, "ZSTD_copyCCtx_internal"); + RETURN_ERROR_IF(srcCCtx->stage!=ZSTDcs_init, stage_wrong, + "Can't copy a ctx that's not in init stage."); + + ZSTD_memcpy(&dstCCtx->customMem, &srcCCtx->customMem, sizeof(ZSTD_customMem)); + { ZSTD_CCtx_params params = dstCCtx->requestedParams; + /* Copy only compression parameters related to tables. */ + params.cParams = srcCCtx->appliedParams.cParams; + params.fParams = fParams; + ZSTD_resetCCtx_internal(dstCCtx, params, pledgedSrcSize, + ZSTDcrp_leaveDirty, zbuff); + assert(dstCCtx->appliedParams.cParams.windowLog == srcCCtx->appliedParams.cParams.windowLog); + assert(dstCCtx->appliedParams.cParams.strategy == srcCCtx->appliedParams.cParams.strategy); + assert(dstCCtx->appliedParams.cParams.hashLog == srcCCtx->appliedParams.cParams.hashLog); + assert(dstCCtx->appliedParams.cParams.chainLog == srcCCtx->appliedParams.cParams.chainLog); + assert(dstCCtx->blockState.matchState.hashLog3 == srcCCtx->blockState.matchState.hashLog3); + } + + ZSTD_cwksp_mark_tables_dirty(&dstCCtx->workspace); + + /* copy tables */ + { size_t const chainSize = (srcCCtx->appliedParams.cParams.strategy == ZSTD_fast) ? 0 : ((size_t)1 << srcCCtx->appliedParams.cParams.chainLog); + size_t const hSize = (size_t)1 << srcCCtx->appliedParams.cParams.hashLog; + int const h3log = srcCCtx->blockState.matchState.hashLog3; + size_t const h3Size = h3log ? ((size_t)1 << h3log) : 0; + + ZSTD_memcpy(dstCCtx->blockState.matchState.hashTable, + srcCCtx->blockState.matchState.hashTable, + hSize * sizeof(U32)); + ZSTD_memcpy(dstCCtx->blockState.matchState.chainTable, + srcCCtx->blockState.matchState.chainTable, + chainSize * sizeof(U32)); + ZSTD_memcpy(dstCCtx->blockState.matchState.hashTable3, + srcCCtx->blockState.matchState.hashTable3, + h3Size * sizeof(U32)); + } + + ZSTD_cwksp_mark_tables_clean(&dstCCtx->workspace); + + /* copy dictionary offsets */ + { + const ZSTD_matchState_t* srcMatchState = &srcCCtx->blockState.matchState; + ZSTD_matchState_t* dstMatchState = &dstCCtx->blockState.matchState; + dstMatchState->window = srcMatchState->window; + dstMatchState->nextToUpdate = srcMatchState->nextToUpdate; + dstMatchState->loadedDictEnd= srcMatchState->loadedDictEnd; + } + dstCCtx->dictID = srcCCtx->dictID; + dstCCtx->dictContentSize = srcCCtx->dictContentSize; + + /* copy block state */ + ZSTD_memcpy(dstCCtx->blockState.prevCBlock, srcCCtx->blockState.prevCBlock, sizeof(*srcCCtx->blockState.prevCBlock)); + + return 0; +} + +/*! ZSTD_copyCCtx() : + * Duplicate an existing context `srcCCtx` into another one `dstCCtx`. + * Only works during stage ZSTDcs_init (i.e. after creation, but before first call to ZSTD_compressContinue()). + * pledgedSrcSize==0 means "unknown". +* @return : 0, or an error code */ +size_t ZSTD_copyCCtx(ZSTD_CCtx* dstCCtx, const ZSTD_CCtx* srcCCtx, unsigned long long pledgedSrcSize) +{ + ZSTD_frameParameters fParams = { 1 /*content*/, 0 /*checksum*/, 0 /*noDictID*/ }; + ZSTD_buffered_policy_e const zbuff = srcCCtx->bufferedPolicy; + ZSTD_STATIC_ASSERT((U32)ZSTDb_buffered==1); + if (pledgedSrcSize==0) pledgedSrcSize = ZSTD_CONTENTSIZE_UNKNOWN; + fParams.contentSizeFlag = (pledgedSrcSize != ZSTD_CONTENTSIZE_UNKNOWN); + + return ZSTD_copyCCtx_internal(dstCCtx, srcCCtx, + fParams, pledgedSrcSize, + zbuff); +} + + +#define ZSTD_ROWSIZE 16 +/*! ZSTD_reduceTable() : + * reduce table indexes by `reducerValue`, or squash to zero. + * PreserveMark preserves "unsorted mark" for btlazy2 strategy. + * It must be set to a clear 0/1 value, to remove branch during inlining. + * Presume table size is a multiple of ZSTD_ROWSIZE + * to help auto-vectorization */ +FORCE_INLINE_TEMPLATE void +ZSTD_reduceTable_internal (U32* const table, U32 const size, U32 const reducerValue, int const preserveMark) +{ + int const nbRows = (int)size / ZSTD_ROWSIZE; + int cellNb = 0; + int rowNb; + assert((size & (ZSTD_ROWSIZE-1)) == 0); /* multiple of ZSTD_ROWSIZE */ + assert(size < (1U<<31)); /* can be casted to int */ + + + for (rowNb=0 ; rowNb < nbRows ; rowNb++) { + int column; + for (column=0; columncParams.hashLog; + ZSTD_reduceTable(ms->hashTable, hSize, reducerValue); + } + + if (params->cParams.strategy != ZSTD_fast) { + U32 const chainSize = (U32)1 << params->cParams.chainLog; + if (params->cParams.strategy == ZSTD_btlazy2) + ZSTD_reduceTable_btlazy2(ms->chainTable, chainSize, reducerValue); + else + ZSTD_reduceTable(ms->chainTable, chainSize, reducerValue); + } + + if (ms->hashLog3) { + U32 const h3Size = (U32)1 << ms->hashLog3; + ZSTD_reduceTable(ms->hashTable3, h3Size, reducerValue); + } +} + + +/*-******************************************************* +* Block entropic compression +*********************************************************/ + +/* See doc/zstd_compression_format.md for detailed format description */ + +void ZSTD_seqToCodes(const seqStore_t* seqStorePtr) +{ + const seqDef* const sequences = seqStorePtr->sequencesStart; + BYTE* const llCodeTable = seqStorePtr->llCode; + BYTE* const ofCodeTable = seqStorePtr->ofCode; + BYTE* const mlCodeTable = seqStorePtr->mlCode; + U32 const nbSeq = (U32)(seqStorePtr->sequences - seqStorePtr->sequencesStart); + U32 u; + assert(nbSeq <= seqStorePtr->maxNbSeq); + for (u=0; ulongLengthID==1) + llCodeTable[seqStorePtr->longLengthPos] = MaxLL; + if (seqStorePtr->longLengthID==2) + mlCodeTable[seqStorePtr->longLengthPos] = MaxML; +} + +/* ZSTD_useTargetCBlockSize(): + * Returns if target compressed block size param is being used. + * If used, compression will do best effort to make a compressed block size to be around targetCBlockSize. + * Returns 1 if true, 0 otherwise. */ +static int ZSTD_useTargetCBlockSize(const ZSTD_CCtx_params* cctxParams) +{ + DEBUGLOG(5, "ZSTD_useTargetCBlockSize (targetCBlockSize=%zu)", cctxParams->targetCBlockSize); + return (cctxParams->targetCBlockSize != 0); +} + +/* ZSTD_entropyCompressSequences_internal(): + * actually compresses both literals and sequences */ +MEM_STATIC size_t +ZSTD_entropyCompressSequences_internal(seqStore_t* seqStorePtr, + const ZSTD_entropyCTables_t* prevEntropy, + ZSTD_entropyCTables_t* nextEntropy, + const ZSTD_CCtx_params* cctxParams, + void* dst, size_t dstCapacity, + void* entropyWorkspace, size_t entropyWkspSize, + const int bmi2) +{ + const int longOffsets = cctxParams->cParams.windowLog > STREAM_ACCUMULATOR_MIN; + ZSTD_strategy const strategy = cctxParams->cParams.strategy; + unsigned* count = (unsigned*)entropyWorkspace; + FSE_CTable* CTable_LitLength = nextEntropy->fse.litlengthCTable; + FSE_CTable* CTable_OffsetBits = nextEntropy->fse.offcodeCTable; + FSE_CTable* CTable_MatchLength = nextEntropy->fse.matchlengthCTable; + U32 LLtype, Offtype, MLtype; /* compressed, raw or rle */ + const seqDef* const sequences = seqStorePtr->sequencesStart; + const BYTE* const ofCodeTable = seqStorePtr->ofCode; + const BYTE* const llCodeTable = seqStorePtr->llCode; + const BYTE* const mlCodeTable = seqStorePtr->mlCode; + BYTE* const ostart = (BYTE*)dst; + BYTE* const oend = ostart + dstCapacity; + BYTE* op = ostart; + size_t const nbSeq = (size_t)(seqStorePtr->sequences - seqStorePtr->sequencesStart); + BYTE* seqHead; + BYTE* lastNCount = NULL; + + entropyWorkspace = count + (MaxSeq + 1); + entropyWkspSize -= (MaxSeq + 1) * sizeof(*count); + + DEBUGLOG(4, "ZSTD_entropyCompressSequences_internal (nbSeq=%zu)", nbSeq); + ZSTD_STATIC_ASSERT(HUF_WORKSPACE_SIZE >= (1<= HUF_WORKSPACE_SIZE); + + /* Compress literals */ + { const BYTE* const literals = seqStorePtr->litStart; + size_t const litSize = (size_t)(seqStorePtr->lit - literals); + size_t const cSize = ZSTD_compressLiterals( + &prevEntropy->huf, &nextEntropy->huf, + cctxParams->cParams.strategy, + ZSTD_disableLiteralsCompression(cctxParams), + op, dstCapacity, + literals, litSize, + entropyWorkspace, entropyWkspSize, + bmi2); + FORWARD_IF_ERROR(cSize, "ZSTD_compressLiterals failed"); + assert(cSize <= dstCapacity); + op += cSize; + } + + /* Sequences Header */ + RETURN_ERROR_IF((oend-op) < 3 /*max nbSeq Size*/ + 1 /*seqHead*/, + dstSize_tooSmall, "Can't fit seq hdr in output buf!"); + if (nbSeq < 128) { + *op++ = (BYTE)nbSeq; + } else if (nbSeq < LONGNBSEQ) { + op[0] = (BYTE)((nbSeq>>8) + 0x80); + op[1] = (BYTE)nbSeq; + op+=2; + } else { + op[0]=0xFF; + MEM_writeLE16(op+1, (U16)(nbSeq - LONGNBSEQ)); + op+=3; + } + assert(op <= oend); + if (nbSeq==0) { + /* Copy the old tables over as if we repeated them */ + ZSTD_memcpy(&nextEntropy->fse, &prevEntropy->fse, sizeof(prevEntropy->fse)); + return (size_t)(op - ostart); + } + + /* seqHead : flags for FSE encoding type */ + seqHead = op++; + assert(op <= oend); + + /* convert length/distances into codes */ + ZSTD_seqToCodes(seqStorePtr); + /* build CTable for Literal Lengths */ + { unsigned max = MaxLL; + size_t const mostFrequent = HIST_countFast_wksp(count, &max, llCodeTable, nbSeq, entropyWorkspace, entropyWkspSize); /* can't fail */ + DEBUGLOG(5, "Building LL table"); + nextEntropy->fse.litlength_repeatMode = prevEntropy->fse.litlength_repeatMode; + LLtype = ZSTD_selectEncodingType(&nextEntropy->fse.litlength_repeatMode, + count, max, mostFrequent, nbSeq, + LLFSELog, prevEntropy->fse.litlengthCTable, + LL_defaultNorm, LL_defaultNormLog, + ZSTD_defaultAllowed, strategy); + assert(set_basic < set_compressed && set_rle < set_compressed); + assert(!(LLtype < set_compressed && nextEntropy->fse.litlength_repeatMode != FSE_repeat_none)); /* We don't copy tables */ + { size_t const countSize = ZSTD_buildCTable( + op, (size_t)(oend - op), + CTable_LitLength, LLFSELog, (symbolEncodingType_e)LLtype, + count, max, llCodeTable, nbSeq, + LL_defaultNorm, LL_defaultNormLog, MaxLL, + prevEntropy->fse.litlengthCTable, + sizeof(prevEntropy->fse.litlengthCTable), + entropyWorkspace, entropyWkspSize); + FORWARD_IF_ERROR(countSize, "ZSTD_buildCTable for LitLens failed"); + if (LLtype == set_compressed) + lastNCount = op; + op += countSize; + assert(op <= oend); + } } + /* build CTable for Offsets */ + { unsigned max = MaxOff; + size_t const mostFrequent = HIST_countFast_wksp( + count, &max, ofCodeTable, nbSeq, entropyWorkspace, entropyWkspSize); /* can't fail */ + /* We can only use the basic table if max <= DefaultMaxOff, otherwise the offsets are too large */ + ZSTD_defaultPolicy_e const defaultPolicy = (max <= DefaultMaxOff) ? ZSTD_defaultAllowed : ZSTD_defaultDisallowed; + DEBUGLOG(5, "Building OF table"); + nextEntropy->fse.offcode_repeatMode = prevEntropy->fse.offcode_repeatMode; + Offtype = ZSTD_selectEncodingType(&nextEntropy->fse.offcode_repeatMode, + count, max, mostFrequent, nbSeq, + OffFSELog, prevEntropy->fse.offcodeCTable, + OF_defaultNorm, OF_defaultNormLog, + defaultPolicy, strategy); + assert(!(Offtype < set_compressed && nextEntropy->fse.offcode_repeatMode != FSE_repeat_none)); /* We don't copy tables */ + { size_t const countSize = ZSTD_buildCTable( + op, (size_t)(oend - op), + CTable_OffsetBits, OffFSELog, (symbolEncodingType_e)Offtype, + count, max, ofCodeTable, nbSeq, + OF_defaultNorm, OF_defaultNormLog, DefaultMaxOff, + prevEntropy->fse.offcodeCTable, + sizeof(prevEntropy->fse.offcodeCTable), + entropyWorkspace, entropyWkspSize); + FORWARD_IF_ERROR(countSize, "ZSTD_buildCTable for Offsets failed"); + if (Offtype == set_compressed) + lastNCount = op; + op += countSize; + assert(op <= oend); + } } + /* build CTable for MatchLengths */ + { unsigned max = MaxML; + size_t const mostFrequent = HIST_countFast_wksp( + count, &max, mlCodeTable, nbSeq, entropyWorkspace, entropyWkspSize); /* can't fail */ + DEBUGLOG(5, "Building ML table (remaining space : %i)", (int)(oend-op)); + nextEntropy->fse.matchlength_repeatMode = prevEntropy->fse.matchlength_repeatMode; + MLtype = ZSTD_selectEncodingType(&nextEntropy->fse.matchlength_repeatMode, + count, max, mostFrequent, nbSeq, + MLFSELog, prevEntropy->fse.matchlengthCTable, + ML_defaultNorm, ML_defaultNormLog, + ZSTD_defaultAllowed, strategy); + assert(!(MLtype < set_compressed && nextEntropy->fse.matchlength_repeatMode != FSE_repeat_none)); /* We don't copy tables */ + { size_t const countSize = ZSTD_buildCTable( + op, (size_t)(oend - op), + CTable_MatchLength, MLFSELog, (symbolEncodingType_e)MLtype, + count, max, mlCodeTable, nbSeq, + ML_defaultNorm, ML_defaultNormLog, MaxML, + prevEntropy->fse.matchlengthCTable, + sizeof(prevEntropy->fse.matchlengthCTable), + entropyWorkspace, entropyWkspSize); + FORWARD_IF_ERROR(countSize, "ZSTD_buildCTable for MatchLengths failed"); + if (MLtype == set_compressed) + lastNCount = op; + op += countSize; + assert(op <= oend); + } } + + *seqHead = (BYTE)((LLtype<<6) + (Offtype<<4) + (MLtype<<2)); + + { size_t const bitstreamSize = ZSTD_encodeSequences( + op, (size_t)(oend - op), + CTable_MatchLength, mlCodeTable, + CTable_OffsetBits, ofCodeTable, + CTable_LitLength, llCodeTable, + sequences, nbSeq, + longOffsets, bmi2); + FORWARD_IF_ERROR(bitstreamSize, "ZSTD_encodeSequences failed"); + op += bitstreamSize; + assert(op <= oend); + /* zstd versions <= 1.3.4 mistakenly report corruption when + * FSE_readNCount() receives a buffer < 4 bytes. + * Fixed by https://github.com/facebook/zstd/pull/1146. + * This can happen when the last set_compressed table present is 2 + * bytes and the bitstream is only one byte. + * In this exceedingly rare case, we will simply emit an uncompressed + * block, since it isn't worth optimizing. + */ + if (lastNCount && (op - lastNCount) < 4) { + /* NCountSize >= 2 && bitstreamSize > 0 ==> lastCountSize == 3 */ + assert(op - lastNCount == 3); + DEBUGLOG(5, "Avoiding bug in zstd decoder in versions <= 1.3.4 by " + "emitting an uncompressed block."); + return 0; + } + } + + DEBUGLOG(5, "compressed block size : %u", (unsigned)(op - ostart)); + return (size_t)(op - ostart); +} + +MEM_STATIC size_t +ZSTD_entropyCompressSequences(seqStore_t* seqStorePtr, + const ZSTD_entropyCTables_t* prevEntropy, + ZSTD_entropyCTables_t* nextEntropy, + const ZSTD_CCtx_params* cctxParams, + void* dst, size_t dstCapacity, + size_t srcSize, + void* entropyWorkspace, size_t entropyWkspSize, + int bmi2) +{ + size_t const cSize = ZSTD_entropyCompressSequences_internal( + seqStorePtr, prevEntropy, nextEntropy, cctxParams, + dst, dstCapacity, + entropyWorkspace, entropyWkspSize, bmi2); + if (cSize == 0) return 0; + /* When srcSize <= dstCapacity, there is enough space to write a raw uncompressed block. + * Since we ran out of space, block must be not compressible, so fall back to raw uncompressed block. + */ + if ((cSize == ERROR(dstSize_tooSmall)) & (srcSize <= dstCapacity)) + return 0; /* block not compressed */ + FORWARD_IF_ERROR(cSize, "ZSTD_entropyCompressSequences_internal failed"); + + /* Check compressibility */ + { size_t const maxCSize = srcSize - ZSTD_minGain(srcSize, cctxParams->cParams.strategy); + if (cSize >= maxCSize) return 0; /* block not compressed */ + } + DEBUGLOG(4, "ZSTD_entropyCompressSequences() cSize: %zu\n", cSize); + return cSize; +} + +/* ZSTD_selectBlockCompressor() : + * Not static, but internal use only (used by long distance matcher) + * assumption : strat is a valid strategy */ +ZSTD_blockCompressor ZSTD_selectBlockCompressor(ZSTD_strategy strat, ZSTD_dictMode_e dictMode) +{ + static const ZSTD_blockCompressor blockCompressor[4][ZSTD_STRATEGY_MAX+1] = { + { ZSTD_compressBlock_fast /* default for 0 */, + ZSTD_compressBlock_fast, + ZSTD_compressBlock_doubleFast, + ZSTD_compressBlock_greedy, + ZSTD_compressBlock_lazy, + ZSTD_compressBlock_lazy2, + ZSTD_compressBlock_btlazy2, + ZSTD_compressBlock_btopt, + ZSTD_compressBlock_btultra, + ZSTD_compressBlock_btultra2 }, + { ZSTD_compressBlock_fast_extDict /* default for 0 */, + ZSTD_compressBlock_fast_extDict, + ZSTD_compressBlock_doubleFast_extDict, + ZSTD_compressBlock_greedy_extDict, + ZSTD_compressBlock_lazy_extDict, + ZSTD_compressBlock_lazy2_extDict, + ZSTD_compressBlock_btlazy2_extDict, + ZSTD_compressBlock_btopt_extDict, + ZSTD_compressBlock_btultra_extDict, + ZSTD_compressBlock_btultra_extDict }, + { ZSTD_compressBlock_fast_dictMatchState /* default for 0 */, + ZSTD_compressBlock_fast_dictMatchState, + ZSTD_compressBlock_doubleFast_dictMatchState, + ZSTD_compressBlock_greedy_dictMatchState, + ZSTD_compressBlock_lazy_dictMatchState, + ZSTD_compressBlock_lazy2_dictMatchState, + ZSTD_compressBlock_btlazy2_dictMatchState, + ZSTD_compressBlock_btopt_dictMatchState, + ZSTD_compressBlock_btultra_dictMatchState, + ZSTD_compressBlock_btultra_dictMatchState }, + { NULL /* default for 0 */, + NULL, + NULL, + ZSTD_compressBlock_greedy_dedicatedDictSearch, + ZSTD_compressBlock_lazy_dedicatedDictSearch, + ZSTD_compressBlock_lazy2_dedicatedDictSearch, + NULL, + NULL, + NULL, + NULL } + }; + ZSTD_blockCompressor selectedCompressor; + ZSTD_STATIC_ASSERT((unsigned)ZSTD_fast == 1); + + assert(ZSTD_cParam_withinBounds(ZSTD_c_strategy, strat)); + selectedCompressor = blockCompressor[(int)dictMode][(int)strat]; + assert(selectedCompressor != NULL); + return selectedCompressor; +} + +static void ZSTD_storeLastLiterals(seqStore_t* seqStorePtr, + const BYTE* anchor, size_t lastLLSize) +{ + ZSTD_memcpy(seqStorePtr->lit, anchor, lastLLSize); + seqStorePtr->lit += lastLLSize; +} + +void ZSTD_resetSeqStore(seqStore_t* ssPtr) +{ + ssPtr->lit = ssPtr->litStart; + ssPtr->sequences = ssPtr->sequencesStart; + ssPtr->longLengthID = 0; +} + +typedef enum { ZSTDbss_compress, ZSTDbss_noCompress } ZSTD_buildSeqStore_e; + +static size_t ZSTD_buildSeqStore(ZSTD_CCtx* zc, const void* src, size_t srcSize) +{ + ZSTD_matchState_t* const ms = &zc->blockState.matchState; + DEBUGLOG(5, "ZSTD_buildSeqStore (srcSize=%zu)", srcSize); + assert(srcSize <= ZSTD_BLOCKSIZE_MAX); + /* Assert that we have correctly flushed the ctx params into the ms's copy */ + ZSTD_assertEqualCParams(zc->appliedParams.cParams, ms->cParams); + if (srcSize < MIN_CBLOCK_SIZE+ZSTD_blockHeaderSize+1) { + if (zc->appliedParams.cParams.strategy >= ZSTD_btopt) { + ZSTD_ldm_skipRawSeqStoreBytes(&zc->externSeqStore, srcSize); + } else { + ZSTD_ldm_skipSequences(&zc->externSeqStore, srcSize, zc->appliedParams.cParams.minMatch); + } + return ZSTDbss_noCompress; /* don't even attempt compression below a certain srcSize */ + } + ZSTD_resetSeqStore(&(zc->seqStore)); + /* required for optimal parser to read stats from dictionary */ + ms->opt.symbolCosts = &zc->blockState.prevCBlock->entropy; + /* tell the optimal parser how we expect to compress literals */ + ms->opt.literalCompressionMode = zc->appliedParams.literalCompressionMode; + /* a gap between an attached dict and the current window is not safe, + * they must remain adjacent, + * and when that stops being the case, the dict must be unset */ + assert(ms->dictMatchState == NULL || ms->loadedDictEnd == ms->window.dictLimit); + + /* limited update after a very long match */ + { const BYTE* const base = ms->window.base; + const BYTE* const istart = (const BYTE*)src; + const U32 curr = (U32)(istart-base); + if (sizeof(ptrdiff_t)==8) assert(istart - base < (ptrdiff_t)(U32)(-1)); /* ensure no overflow */ + if (curr > ms->nextToUpdate + 384) + ms->nextToUpdate = curr - MIN(192, (U32)(curr - ms->nextToUpdate - 384)); + } + + /* select and store sequences */ + { ZSTD_dictMode_e const dictMode = ZSTD_matchState_dictMode(ms); + size_t lastLLSize; + { int i; + for (i = 0; i < ZSTD_REP_NUM; ++i) + zc->blockState.nextCBlock->rep[i] = zc->blockState.prevCBlock->rep[i]; + } + if (zc->externSeqStore.pos < zc->externSeqStore.size) { + assert(!zc->appliedParams.ldmParams.enableLdm); + /* Updates ldmSeqStore.pos */ + lastLLSize = + ZSTD_ldm_blockCompress(&zc->externSeqStore, + ms, &zc->seqStore, + zc->blockState.nextCBlock->rep, + src, srcSize); + assert(zc->externSeqStore.pos <= zc->externSeqStore.size); + } else if (zc->appliedParams.ldmParams.enableLdm) { + rawSeqStore_t ldmSeqStore = kNullRawSeqStore; + + ldmSeqStore.seq = zc->ldmSequences; + ldmSeqStore.capacity = zc->maxNbLdmSequences; + /* Updates ldmSeqStore.size */ + FORWARD_IF_ERROR(ZSTD_ldm_generateSequences(&zc->ldmState, &ldmSeqStore, + &zc->appliedParams.ldmParams, + src, srcSize), ""); + /* Updates ldmSeqStore.pos */ + lastLLSize = + ZSTD_ldm_blockCompress(&ldmSeqStore, + ms, &zc->seqStore, + zc->blockState.nextCBlock->rep, + src, srcSize); + assert(ldmSeqStore.pos == ldmSeqStore.size); + } else { /* not long range mode */ + ZSTD_blockCompressor const blockCompressor = ZSTD_selectBlockCompressor(zc->appliedParams.cParams.strategy, dictMode); + ms->ldmSeqStore = NULL; + lastLLSize = blockCompressor(ms, &zc->seqStore, zc->blockState.nextCBlock->rep, src, srcSize); + } + { const BYTE* const lastLiterals = (const BYTE*)src + srcSize - lastLLSize; + ZSTD_storeLastLiterals(&zc->seqStore, lastLiterals, lastLLSize); + } } + return ZSTDbss_compress; +} + +static void ZSTD_copyBlockSequences(ZSTD_CCtx* zc) +{ + const seqStore_t* seqStore = ZSTD_getSeqStore(zc); + const seqDef* seqStoreSeqs = seqStore->sequencesStart; + size_t seqStoreSeqSize = seqStore->sequences - seqStoreSeqs; + size_t seqStoreLiteralsSize = (size_t)(seqStore->lit - seqStore->litStart); + size_t literalsRead = 0; + size_t lastLLSize; + + ZSTD_Sequence* outSeqs = &zc->seqCollector.seqStart[zc->seqCollector.seqIndex]; + size_t i; + repcodes_t updatedRepcodes; + + assert(zc->seqCollector.seqIndex + 1 < zc->seqCollector.maxSequences); + /* Ensure we have enough space for last literals "sequence" */ + assert(zc->seqCollector.maxSequences >= seqStoreSeqSize + 1); + ZSTD_memcpy(updatedRepcodes.rep, zc->blockState.prevCBlock->rep, sizeof(repcodes_t)); + for (i = 0; i < seqStoreSeqSize; ++i) { + U32 rawOffset = seqStoreSeqs[i].offset - ZSTD_REP_NUM; + outSeqs[i].litLength = seqStoreSeqs[i].litLength; + outSeqs[i].matchLength = seqStoreSeqs[i].matchLength + MINMATCH; + outSeqs[i].rep = 0; + + if (i == seqStore->longLengthPos) { + if (seqStore->longLengthID == 1) { + outSeqs[i].litLength += 0x10000; + } else if (seqStore->longLengthID == 2) { + outSeqs[i].matchLength += 0x10000; + } + } + + if (seqStoreSeqs[i].offset <= ZSTD_REP_NUM) { + /* Derive the correct offset corresponding to a repcode */ + outSeqs[i].rep = seqStoreSeqs[i].offset; + if (outSeqs[i].litLength != 0) { + rawOffset = updatedRepcodes.rep[outSeqs[i].rep - 1]; + } else { + if (outSeqs[i].rep == 3) { + rawOffset = updatedRepcodes.rep[0] - 1; + } else { + rawOffset = updatedRepcodes.rep[outSeqs[i].rep]; + } + } + } + outSeqs[i].offset = rawOffset; + /* seqStoreSeqs[i].offset == offCode+1, and ZSTD_updateRep() expects offCode + so we provide seqStoreSeqs[i].offset - 1 */ + updatedRepcodes = ZSTD_updateRep(updatedRepcodes.rep, + seqStoreSeqs[i].offset - 1, + seqStoreSeqs[i].litLength == 0); + literalsRead += outSeqs[i].litLength; + } + /* Insert last literals (if any exist) in the block as a sequence with ml == off == 0. + * If there are no last literals, then we'll emit (of: 0, ml: 0, ll: 0), which is a marker + * for the block boundary, according to the API. + */ + assert(seqStoreLiteralsSize >= literalsRead); + lastLLSize = seqStoreLiteralsSize - literalsRead; + outSeqs[i].litLength = (U32)lastLLSize; + outSeqs[i].matchLength = outSeqs[i].offset = outSeqs[i].rep = 0; + seqStoreSeqSize++; + zc->seqCollector.seqIndex += seqStoreSeqSize; +} + +size_t ZSTD_generateSequences(ZSTD_CCtx* zc, ZSTD_Sequence* outSeqs, + size_t outSeqsSize, const void* src, size_t srcSize) +{ + const size_t dstCapacity = ZSTD_compressBound(srcSize); + void* dst = ZSTD_customMalloc(dstCapacity, ZSTD_defaultCMem); + SeqCollector seqCollector; + + RETURN_ERROR_IF(dst == NULL, memory_allocation, "NULL pointer!"); + + seqCollector.collectSequences = 1; + seqCollector.seqStart = outSeqs; + seqCollector.seqIndex = 0; + seqCollector.maxSequences = outSeqsSize; + zc->seqCollector = seqCollector; + + ZSTD_compress2(zc, dst, dstCapacity, src, srcSize); + ZSTD_customFree(dst, ZSTD_defaultCMem); + return zc->seqCollector.seqIndex; +} + +size_t ZSTD_mergeBlockDelimiters(ZSTD_Sequence* sequences, size_t seqsSize) { + size_t in = 0; + size_t out = 0; + for (; in < seqsSize; ++in) { + if (sequences[in].offset == 0 && sequences[in].matchLength == 0) { + if (in != seqsSize - 1) { + sequences[in+1].litLength += sequences[in].litLength; + } + } else { + sequences[out] = sequences[in]; + ++out; + } + } + return out; +} + +/* Unrolled loop to read four size_ts of input at a time. Returns 1 if is RLE, 0 if not. */ +static int ZSTD_isRLE(const BYTE* src, size_t length) { + const BYTE* ip = src; + const BYTE value = ip[0]; + const size_t valueST = (size_t)((U64)value * 0x0101010101010101ULL); + const size_t unrollSize = sizeof(size_t) * 4; + const size_t unrollMask = unrollSize - 1; + const size_t prefixLength = length & unrollMask; + size_t i; + size_t u; + if (length == 1) return 1; + /* Check if prefix is RLE first before using unrolled loop */ + if (prefixLength && ZSTD_count(ip+1, ip, ip+prefixLength) != prefixLength-1) { + return 0; + } + for (i = prefixLength; i != length; i += unrollSize) { + for (u = 0; u < unrollSize; u += sizeof(size_t)) { + if (MEM_readST(ip + i + u) != valueST) { + return 0; + } + } + } + return 1; +} + +/* Returns true if the given block may be RLE. + * This is just a heuristic based on the compressibility. + * It may return both false positives and false negatives. + */ +static int ZSTD_maybeRLE(seqStore_t const* seqStore) +{ + size_t const nbSeqs = (size_t)(seqStore->sequences - seqStore->sequencesStart); + size_t const nbLits = (size_t)(seqStore->lit - seqStore->litStart); + + return nbSeqs < 4 && nbLits < 10; +} + +static void ZSTD_confirmRepcodesAndEntropyTables(ZSTD_CCtx* zc) +{ + ZSTD_compressedBlockState_t* const tmp = zc->blockState.prevCBlock; + zc->blockState.prevCBlock = zc->blockState.nextCBlock; + zc->blockState.nextCBlock = tmp; +} + +static size_t ZSTD_compressBlock_internal(ZSTD_CCtx* zc, + void* dst, size_t dstCapacity, + const void* src, size_t srcSize, U32 frame) +{ + /* This the upper bound for the length of an rle block. + * This isn't the actual upper bound. Finding the real threshold + * needs further investigation. + */ + const U32 rleMaxLength = 25; + size_t cSize; + const BYTE* ip = (const BYTE*)src; + BYTE* op = (BYTE*)dst; + DEBUGLOG(5, "ZSTD_compressBlock_internal (dstCapacity=%u, dictLimit=%u, nextToUpdate=%u)", + (unsigned)dstCapacity, (unsigned)zc->blockState.matchState.window.dictLimit, + (unsigned)zc->blockState.matchState.nextToUpdate); + + { const size_t bss = ZSTD_buildSeqStore(zc, src, srcSize); + FORWARD_IF_ERROR(bss, "ZSTD_buildSeqStore failed"); + if (bss == ZSTDbss_noCompress) { cSize = 0; goto out; } + } + + if (zc->seqCollector.collectSequences) { + ZSTD_copyBlockSequences(zc); + ZSTD_confirmRepcodesAndEntropyTables(zc); + return 0; + } + + /* encode sequences and literals */ + cSize = ZSTD_entropyCompressSequences(&zc->seqStore, + &zc->blockState.prevCBlock->entropy, &zc->blockState.nextCBlock->entropy, + &zc->appliedParams, + dst, dstCapacity, + srcSize, + zc->entropyWorkspace, ENTROPY_WORKSPACE_SIZE /* statically allocated in resetCCtx */, + zc->bmi2); + + if (zc->seqCollector.collectSequences) { + ZSTD_copyBlockSequences(zc); + return 0; + } + + + if (frame && + /* We don't want to emit our first block as a RLE even if it qualifies because + * doing so will cause the decoder (cli only) to throw a "should consume all input error." + * This is only an issue for zstd <= v1.4.3 + */ + !zc->isFirstBlock && + cSize < rleMaxLength && + ZSTD_isRLE(ip, srcSize)) + { + cSize = 1; + op[0] = ip[0]; + } + +out: + if (!ZSTD_isError(cSize) && cSize > 1) { + ZSTD_confirmRepcodesAndEntropyTables(zc); + } + /* We check that dictionaries have offset codes available for the first + * block. After the first block, the offcode table might not have large + * enough codes to represent the offsets in the data. + */ + if (zc->blockState.prevCBlock->entropy.fse.offcode_repeatMode == FSE_repeat_valid) + zc->blockState.prevCBlock->entropy.fse.offcode_repeatMode = FSE_repeat_check; + + return cSize; +} + +static size_t ZSTD_compressBlock_targetCBlockSize_body(ZSTD_CCtx* zc, + void* dst, size_t dstCapacity, + const void* src, size_t srcSize, + const size_t bss, U32 lastBlock) +{ + DEBUGLOG(6, "Attempting ZSTD_compressSuperBlock()"); + if (bss == ZSTDbss_compress) { + if (/* We don't want to emit our first block as a RLE even if it qualifies because + * doing so will cause the decoder (cli only) to throw a "should consume all input error." + * This is only an issue for zstd <= v1.4.3 + */ + !zc->isFirstBlock && + ZSTD_maybeRLE(&zc->seqStore) && + ZSTD_isRLE((BYTE const*)src, srcSize)) + { + return ZSTD_rleCompressBlock(dst, dstCapacity, *(BYTE const*)src, srcSize, lastBlock); + } + /* Attempt superblock compression. + * + * Note that compressed size of ZSTD_compressSuperBlock() is not bound by the + * standard ZSTD_compressBound(). This is a problem, because even if we have + * space now, taking an extra byte now could cause us to run out of space later + * and violate ZSTD_compressBound(). + * + * Define blockBound(blockSize) = blockSize + ZSTD_blockHeaderSize. + * + * In order to respect ZSTD_compressBound() we must attempt to emit a raw + * uncompressed block in these cases: + * * cSize == 0: Return code for an uncompressed block. + * * cSize == dstSize_tooSmall: We may have expanded beyond blockBound(srcSize). + * ZSTD_noCompressBlock() will return dstSize_tooSmall if we are really out of + * output space. + * * cSize >= blockBound(srcSize): We have expanded the block too much so + * emit an uncompressed block. + */ + { + size_t const cSize = ZSTD_compressSuperBlock(zc, dst, dstCapacity, src, srcSize, lastBlock); + if (cSize != ERROR(dstSize_tooSmall)) { + size_t const maxCSize = srcSize - ZSTD_minGain(srcSize, zc->appliedParams.cParams.strategy); + FORWARD_IF_ERROR(cSize, "ZSTD_compressSuperBlock failed"); + if (cSize != 0 && cSize < maxCSize + ZSTD_blockHeaderSize) { + ZSTD_confirmRepcodesAndEntropyTables(zc); + return cSize; + } + } + } + } + + DEBUGLOG(6, "Resorting to ZSTD_noCompressBlock()"); + /* Superblock compression failed, attempt to emit a single no compress block. + * The decoder will be able to stream this block since it is uncompressed. + */ + return ZSTD_noCompressBlock(dst, dstCapacity, src, srcSize, lastBlock); +} + +static size_t ZSTD_compressBlock_targetCBlockSize(ZSTD_CCtx* zc, + void* dst, size_t dstCapacity, + const void* src, size_t srcSize, + U32 lastBlock) +{ + size_t cSize = 0; + const size_t bss = ZSTD_buildSeqStore(zc, src, srcSize); + DEBUGLOG(5, "ZSTD_compressBlock_targetCBlockSize (dstCapacity=%u, dictLimit=%u, nextToUpdate=%u, srcSize=%zu)", + (unsigned)dstCapacity, (unsigned)zc->blockState.matchState.window.dictLimit, (unsigned)zc->blockState.matchState.nextToUpdate, srcSize); + FORWARD_IF_ERROR(bss, "ZSTD_buildSeqStore failed"); + + cSize = ZSTD_compressBlock_targetCBlockSize_body(zc, dst, dstCapacity, src, srcSize, bss, lastBlock); + FORWARD_IF_ERROR(cSize, "ZSTD_compressBlock_targetCBlockSize_body failed"); + + if (zc->blockState.prevCBlock->entropy.fse.offcode_repeatMode == FSE_repeat_valid) + zc->blockState.prevCBlock->entropy.fse.offcode_repeatMode = FSE_repeat_check; + + return cSize; +} + +static void ZSTD_overflowCorrectIfNeeded(ZSTD_matchState_t* ms, + ZSTD_cwksp* ws, + ZSTD_CCtx_params const* params, + void const* ip, + void const* iend) +{ + if (ZSTD_window_needOverflowCorrection(ms->window, iend)) { + U32 const maxDist = (U32)1 << params->cParams.windowLog; + U32 const cycleLog = ZSTD_cycleLog(params->cParams.chainLog, params->cParams.strategy); + U32 const correction = ZSTD_window_correctOverflow(&ms->window, cycleLog, maxDist, ip); + ZSTD_STATIC_ASSERT(ZSTD_CHAINLOG_MAX <= 30); + ZSTD_STATIC_ASSERT(ZSTD_WINDOWLOG_MAX_32 <= 30); + ZSTD_STATIC_ASSERT(ZSTD_WINDOWLOG_MAX <= 31); + ZSTD_cwksp_mark_tables_dirty(ws); + ZSTD_reduceIndex(ms, params, correction); + ZSTD_cwksp_mark_tables_clean(ws); + if (ms->nextToUpdate < correction) ms->nextToUpdate = 0; + else ms->nextToUpdate -= correction; + /* invalidate dictionaries on overflow correction */ + ms->loadedDictEnd = 0; + ms->dictMatchState = NULL; + } +} + +/*! ZSTD_compress_frameChunk() : +* Compress a chunk of data into one or multiple blocks. +* All blocks will be terminated, all input will be consumed. +* Function will issue an error if there is not enough `dstCapacity` to hold the compressed content. +* Frame is supposed already started (header already produced) +* @return : compressed size, or an error code +*/ +static size_t ZSTD_compress_frameChunk (ZSTD_CCtx* cctx, + void* dst, size_t dstCapacity, + const void* src, size_t srcSize, + U32 lastFrameChunk) +{ + size_t blockSize = cctx->blockSize; + size_t remaining = srcSize; + const BYTE* ip = (const BYTE*)src; + BYTE* const ostart = (BYTE*)dst; + BYTE* op = ostart; + U32 const maxDist = (U32)1 << cctx->appliedParams.cParams.windowLog; + + assert(cctx->appliedParams.cParams.windowLog <= ZSTD_WINDOWLOG_MAX); + + DEBUGLOG(4, "ZSTD_compress_frameChunk (blockSize=%u)", (unsigned)blockSize); + if (cctx->appliedParams.fParams.checksumFlag && srcSize) + xxh64_update(&cctx->xxhState, src, srcSize); + + while (remaining) { + ZSTD_matchState_t* const ms = &cctx->blockState.matchState; + U32 const lastBlock = lastFrameChunk & (blockSize >= remaining); + + RETURN_ERROR_IF(dstCapacity < ZSTD_blockHeaderSize + MIN_CBLOCK_SIZE, + dstSize_tooSmall, + "not enough space to store compressed block"); + if (remaining < blockSize) blockSize = remaining; + + ZSTD_overflowCorrectIfNeeded( + ms, &cctx->workspace, &cctx->appliedParams, ip, ip + blockSize); + ZSTD_checkDictValidity(&ms->window, ip + blockSize, maxDist, &ms->loadedDictEnd, &ms->dictMatchState); + + /* Ensure hash/chain table insertion resumes no sooner than lowlimit */ + if (ms->nextToUpdate < ms->window.lowLimit) ms->nextToUpdate = ms->window.lowLimit; + + { size_t cSize; + if (ZSTD_useTargetCBlockSize(&cctx->appliedParams)) { + cSize = ZSTD_compressBlock_targetCBlockSize(cctx, op, dstCapacity, ip, blockSize, lastBlock); + FORWARD_IF_ERROR(cSize, "ZSTD_compressBlock_targetCBlockSize failed"); + assert(cSize > 0); + assert(cSize <= blockSize + ZSTD_blockHeaderSize); + } else { + cSize = ZSTD_compressBlock_internal(cctx, + op+ZSTD_blockHeaderSize, dstCapacity-ZSTD_blockHeaderSize, + ip, blockSize, 1 /* frame */); + FORWARD_IF_ERROR(cSize, "ZSTD_compressBlock_internal failed"); + + if (cSize == 0) { /* block is not compressible */ + cSize = ZSTD_noCompressBlock(op, dstCapacity, ip, blockSize, lastBlock); + FORWARD_IF_ERROR(cSize, "ZSTD_noCompressBlock failed"); + } else { + U32 const cBlockHeader = cSize == 1 ? + lastBlock + (((U32)bt_rle)<<1) + (U32)(blockSize << 3) : + lastBlock + (((U32)bt_compressed)<<1) + (U32)(cSize << 3); + MEM_writeLE24(op, cBlockHeader); + cSize += ZSTD_blockHeaderSize; + } + } + + + ip += blockSize; + assert(remaining >= blockSize); + remaining -= blockSize; + op += cSize; + assert(dstCapacity >= cSize); + dstCapacity -= cSize; + cctx->isFirstBlock = 0; + DEBUGLOG(5, "ZSTD_compress_frameChunk: adding a block of size %u", + (unsigned)cSize); + } } + + if (lastFrameChunk && (op>ostart)) cctx->stage = ZSTDcs_ending; + return (size_t)(op-ostart); +} + + +static size_t ZSTD_writeFrameHeader(void* dst, size_t dstCapacity, + const ZSTD_CCtx_params* params, U64 pledgedSrcSize, U32 dictID) +{ BYTE* const op = (BYTE*)dst; + U32 const dictIDSizeCodeLength = (dictID>0) + (dictID>=256) + (dictID>=65536); /* 0-3 */ + U32 const dictIDSizeCode = params->fParams.noDictIDFlag ? 0 : dictIDSizeCodeLength; /* 0-3 */ + U32 const checksumFlag = params->fParams.checksumFlag>0; + U32 const windowSize = (U32)1 << params->cParams.windowLog; + U32 const singleSegment = params->fParams.contentSizeFlag && (windowSize >= pledgedSrcSize); + BYTE const windowLogByte = (BYTE)((params->cParams.windowLog - ZSTD_WINDOWLOG_ABSOLUTEMIN) << 3); + U32 const fcsCode = params->fParams.contentSizeFlag ? + (pledgedSrcSize>=256) + (pledgedSrcSize>=65536+256) + (pledgedSrcSize>=0xFFFFFFFFU) : 0; /* 0-3 */ + BYTE const frameHeaderDescriptionByte = (BYTE)(dictIDSizeCode + (checksumFlag<<2) + (singleSegment<<5) + (fcsCode<<6) ); + size_t pos=0; + + assert(!(params->fParams.contentSizeFlag && pledgedSrcSize == ZSTD_CONTENTSIZE_UNKNOWN)); + RETURN_ERROR_IF(dstCapacity < ZSTD_FRAMEHEADERSIZE_MAX, dstSize_tooSmall, + "dst buf is too small to fit worst-case frame header size."); + DEBUGLOG(4, "ZSTD_writeFrameHeader : dictIDFlag : %u ; dictID : %u ; dictIDSizeCode : %u", + !params->fParams.noDictIDFlag, (unsigned)dictID, (unsigned)dictIDSizeCode); + if (params->format == ZSTD_f_zstd1) { + MEM_writeLE32(dst, ZSTD_MAGICNUMBER); + pos = 4; + } + op[pos++] = frameHeaderDescriptionByte; + if (!singleSegment) op[pos++] = windowLogByte; + switch(dictIDSizeCode) + { + default: + assert(0); /* impossible */ + ZSTD_FALLTHROUGH; + case 0 : break; + case 1 : op[pos] = (BYTE)(dictID); pos++; break; + case 2 : MEM_writeLE16(op+pos, (U16)dictID); pos+=2; break; + case 3 : MEM_writeLE32(op+pos, dictID); pos+=4; break; + } + switch(fcsCode) + { + default: + assert(0); /* impossible */ + ZSTD_FALLTHROUGH; + case 0 : if (singleSegment) op[pos++] = (BYTE)(pledgedSrcSize); break; + case 1 : MEM_writeLE16(op+pos, (U16)(pledgedSrcSize-256)); pos+=2; break; + case 2 : MEM_writeLE32(op+pos, (U32)(pledgedSrcSize)); pos+=4; break; + case 3 : MEM_writeLE64(op+pos, (U64)(pledgedSrcSize)); pos+=8; break; + } + return pos; +} + +/* ZSTD_writeSkippableFrame_advanced() : + * Writes out a skippable frame with the specified magic number variant (16 are supported), + * from ZSTD_MAGIC_SKIPPABLE_START to ZSTD_MAGIC_SKIPPABLE_START+15, and the desired source data. + * + * Returns the total number of bytes written, or a ZSTD error code. + */ +size_t ZSTD_writeSkippableFrame(void* dst, size_t dstCapacity, + const void* src, size_t srcSize, unsigned magicVariant) { + BYTE* op = (BYTE*)dst; + RETURN_ERROR_IF(dstCapacity < srcSize + ZSTD_SKIPPABLEHEADERSIZE /* Skippable frame overhead */, + dstSize_tooSmall, "Not enough room for skippable frame"); + RETURN_ERROR_IF(srcSize > (unsigned)0xFFFFFFFF, srcSize_wrong, "Src size too large for skippable frame"); + RETURN_ERROR_IF(magicVariant > 15, parameter_outOfBound, "Skippable frame magic number variant not supported"); + + MEM_writeLE32(op, (U32)(ZSTD_MAGIC_SKIPPABLE_START + magicVariant)); + MEM_writeLE32(op+4, (U32)srcSize); + ZSTD_memcpy(op+8, src, srcSize); + return srcSize + ZSTD_SKIPPABLEHEADERSIZE; +} + +/* ZSTD_writeLastEmptyBlock() : + * output an empty Block with end-of-frame mark to complete a frame + * @return : size of data written into `dst` (== ZSTD_blockHeaderSize (defined in zstd_internal.h)) + * or an error code if `dstCapacity` is too small (stage != ZSTDcs_init, stage_wrong, + "wrong cctx stage"); + RETURN_ERROR_IF(cctx->appliedParams.ldmParams.enableLdm, + parameter_unsupported, + "incompatible with ldm"); + cctx->externSeqStore.seq = seq; + cctx->externSeqStore.size = nbSeq; + cctx->externSeqStore.capacity = nbSeq; + cctx->externSeqStore.pos = 0; + cctx->externSeqStore.posInSequence = 0; + return 0; +} + + +static size_t ZSTD_compressContinue_internal (ZSTD_CCtx* cctx, + void* dst, size_t dstCapacity, + const void* src, size_t srcSize, + U32 frame, U32 lastFrameChunk) +{ + ZSTD_matchState_t* const ms = &cctx->blockState.matchState; + size_t fhSize = 0; + + DEBUGLOG(5, "ZSTD_compressContinue_internal, stage: %u, srcSize: %u", + cctx->stage, (unsigned)srcSize); + RETURN_ERROR_IF(cctx->stage==ZSTDcs_created, stage_wrong, + "missing init (ZSTD_compressBegin)"); + + if (frame && (cctx->stage==ZSTDcs_init)) { + fhSize = ZSTD_writeFrameHeader(dst, dstCapacity, &cctx->appliedParams, + cctx->pledgedSrcSizePlusOne-1, cctx->dictID); + FORWARD_IF_ERROR(fhSize, "ZSTD_writeFrameHeader failed"); + assert(fhSize <= dstCapacity); + dstCapacity -= fhSize; + dst = (char*)dst + fhSize; + cctx->stage = ZSTDcs_ongoing; + } + + if (!srcSize) return fhSize; /* do not generate an empty block if no input */ + + if (!ZSTD_window_update(&ms->window, src, srcSize)) { + ms->nextToUpdate = ms->window.dictLimit; + } + if (cctx->appliedParams.ldmParams.enableLdm) { + ZSTD_window_update(&cctx->ldmState.window, src, srcSize); + } + + if (!frame) { + /* overflow check and correction for block mode */ + ZSTD_overflowCorrectIfNeeded( + ms, &cctx->workspace, &cctx->appliedParams, + src, (BYTE const*)src + srcSize); + } + + DEBUGLOG(5, "ZSTD_compressContinue_internal (blockSize=%u)", (unsigned)cctx->blockSize); + { size_t const cSize = frame ? + ZSTD_compress_frameChunk (cctx, dst, dstCapacity, src, srcSize, lastFrameChunk) : + ZSTD_compressBlock_internal (cctx, dst, dstCapacity, src, srcSize, 0 /* frame */); + FORWARD_IF_ERROR(cSize, "%s", frame ? "ZSTD_compress_frameChunk failed" : "ZSTD_compressBlock_internal failed"); + cctx->consumedSrcSize += srcSize; + cctx->producedCSize += (cSize + fhSize); + assert(!(cctx->appliedParams.fParams.contentSizeFlag && cctx->pledgedSrcSizePlusOne == 0)); + if (cctx->pledgedSrcSizePlusOne != 0) { /* control src size */ + ZSTD_STATIC_ASSERT(ZSTD_CONTENTSIZE_UNKNOWN == (unsigned long long)-1); + RETURN_ERROR_IF( + cctx->consumedSrcSize+1 > cctx->pledgedSrcSizePlusOne, + srcSize_wrong, + "error : pledgedSrcSize = %u, while realSrcSize >= %u", + (unsigned)cctx->pledgedSrcSizePlusOne-1, + (unsigned)cctx->consumedSrcSize); + } + return cSize + fhSize; + } +} + +size_t ZSTD_compressContinue (ZSTD_CCtx* cctx, + void* dst, size_t dstCapacity, + const void* src, size_t srcSize) +{ + DEBUGLOG(5, "ZSTD_compressContinue (srcSize=%u)", (unsigned)srcSize); + return ZSTD_compressContinue_internal(cctx, dst, dstCapacity, src, srcSize, 1 /* frame mode */, 0 /* last chunk */); +} + + +size_t ZSTD_getBlockSize(const ZSTD_CCtx* cctx) +{ + ZSTD_compressionParameters const cParams = cctx->appliedParams.cParams; + assert(!ZSTD_checkCParams(cParams)); + return MIN (ZSTD_BLOCKSIZE_MAX, (U32)1 << cParams.windowLog); +} + +size_t ZSTD_compressBlock(ZSTD_CCtx* cctx, void* dst, size_t dstCapacity, const void* src, size_t srcSize) +{ + DEBUGLOG(5, "ZSTD_compressBlock: srcSize = %u", (unsigned)srcSize); + { size_t const blockSizeMax = ZSTD_getBlockSize(cctx); + RETURN_ERROR_IF(srcSize > blockSizeMax, srcSize_wrong, "input is larger than a block"); } + + return ZSTD_compressContinue_internal(cctx, dst, dstCapacity, src, srcSize, 0 /* frame mode */, 0 /* last chunk */); +} + +/*! ZSTD_loadDictionaryContent() : + * @return : 0, or an error code + */ +static size_t ZSTD_loadDictionaryContent(ZSTD_matchState_t* ms, + ldmState_t* ls, + ZSTD_cwksp* ws, + ZSTD_CCtx_params const* params, + const void* src, size_t srcSize, + ZSTD_dictTableLoadMethod_e dtlm) +{ + const BYTE* ip = (const BYTE*) src; + const BYTE* const iend = ip + srcSize; + + ZSTD_window_update(&ms->window, src, srcSize); + ms->loadedDictEnd = params->forceWindow ? 0 : (U32)(iend - ms->window.base); + + if (params->ldmParams.enableLdm && ls != NULL) { + ZSTD_window_update(&ls->window, src, srcSize); + ls->loadedDictEnd = params->forceWindow ? 0 : (U32)(iend - ls->window.base); + } + + /* Assert that we the ms params match the params we're being given */ + ZSTD_assertEqualCParams(params->cParams, ms->cParams); + + if (srcSize <= HASH_READ_SIZE) return 0; + + while (iend - ip > HASH_READ_SIZE) { + size_t const remaining = (size_t)(iend - ip); + size_t const chunk = MIN(remaining, ZSTD_CHUNKSIZE_MAX); + const BYTE* const ichunk = ip + chunk; + + ZSTD_overflowCorrectIfNeeded(ms, ws, params, ip, ichunk); + + if (params->ldmParams.enableLdm && ls != NULL) + ZSTD_ldm_fillHashTable(ls, (const BYTE*)src, (const BYTE*)src + srcSize, ¶ms->ldmParams); + + switch(params->cParams.strategy) + { + case ZSTD_fast: + ZSTD_fillHashTable(ms, ichunk, dtlm); + break; + case ZSTD_dfast: + ZSTD_fillDoubleHashTable(ms, ichunk, dtlm); + break; + + case ZSTD_greedy: + case ZSTD_lazy: + case ZSTD_lazy2: + if (chunk >= HASH_READ_SIZE && ms->dedicatedDictSearch) { + assert(chunk == remaining); /* must load everything in one go */ + ZSTD_dedicatedDictSearch_lazy_loadDictionary(ms, ichunk-HASH_READ_SIZE); + } else if (chunk >= HASH_READ_SIZE) { + ZSTD_insertAndFindFirstIndex(ms, ichunk-HASH_READ_SIZE); + } + break; + + case ZSTD_btlazy2: /* we want the dictionary table fully sorted */ + case ZSTD_btopt: + case ZSTD_btultra: + case ZSTD_btultra2: + if (chunk >= HASH_READ_SIZE) + ZSTD_updateTree(ms, ichunk-HASH_READ_SIZE, ichunk); + break; + + default: + assert(0); /* not possible : not a valid strategy id */ + } + + ip = ichunk; + } + + ms->nextToUpdate = (U32)(iend - ms->window.base); + return 0; +} + + +/* Dictionaries that assign zero probability to symbols that show up causes problems + * when FSE encoding. Mark dictionaries with zero probability symbols as FSE_repeat_check + * and only dictionaries with 100% valid symbols can be assumed valid. + */ +static FSE_repeat ZSTD_dictNCountRepeat(short* normalizedCounter, unsigned dictMaxSymbolValue, unsigned maxSymbolValue) +{ + U32 s; + if (dictMaxSymbolValue < maxSymbolValue) { + return FSE_repeat_check; + } + for (s = 0; s <= maxSymbolValue; ++s) { + if (normalizedCounter[s] == 0) { + return FSE_repeat_check; + } + } + return FSE_repeat_valid; +} + +size_t ZSTD_loadCEntropy(ZSTD_compressedBlockState_t* bs, void* workspace, + const void* const dict, size_t dictSize) +{ + short offcodeNCount[MaxOff+1]; + unsigned offcodeMaxValue = MaxOff; + const BYTE* dictPtr = (const BYTE*)dict; /* skip magic num and dict ID */ + const BYTE* const dictEnd = dictPtr + dictSize; + dictPtr += 8; + bs->entropy.huf.repeatMode = HUF_repeat_check; + + { unsigned maxSymbolValue = 255; + unsigned hasZeroWeights = 1; + size_t const hufHeaderSize = HUF_readCTable((HUF_CElt*)bs->entropy.huf.CTable, &maxSymbolValue, dictPtr, + dictEnd-dictPtr, &hasZeroWeights); + + /* We only set the loaded table as valid if it contains all non-zero + * weights. Otherwise, we set it to check */ + if (!hasZeroWeights) + bs->entropy.huf.repeatMode = HUF_repeat_valid; + + RETURN_ERROR_IF(HUF_isError(hufHeaderSize), dictionary_corrupted, ""); + RETURN_ERROR_IF(maxSymbolValue < 255, dictionary_corrupted, ""); + dictPtr += hufHeaderSize; + } + + { unsigned offcodeLog; + size_t const offcodeHeaderSize = FSE_readNCount(offcodeNCount, &offcodeMaxValue, &offcodeLog, dictPtr, dictEnd-dictPtr); + RETURN_ERROR_IF(FSE_isError(offcodeHeaderSize), dictionary_corrupted, ""); + RETURN_ERROR_IF(offcodeLog > OffFSELog, dictionary_corrupted, ""); + /* fill all offset symbols to avoid garbage at end of table */ + RETURN_ERROR_IF(FSE_isError(FSE_buildCTable_wksp( + bs->entropy.fse.offcodeCTable, + offcodeNCount, MaxOff, offcodeLog, + workspace, HUF_WORKSPACE_SIZE)), + dictionary_corrupted, ""); + /* Defer checking offcodeMaxValue because we need to know the size of the dictionary content */ + dictPtr += offcodeHeaderSize; + } + + { short matchlengthNCount[MaxML+1]; + unsigned matchlengthMaxValue = MaxML, matchlengthLog; + size_t const matchlengthHeaderSize = FSE_readNCount(matchlengthNCount, &matchlengthMaxValue, &matchlengthLog, dictPtr, dictEnd-dictPtr); + RETURN_ERROR_IF(FSE_isError(matchlengthHeaderSize), dictionary_corrupted, ""); + RETURN_ERROR_IF(matchlengthLog > MLFSELog, dictionary_corrupted, ""); + RETURN_ERROR_IF(FSE_isError(FSE_buildCTable_wksp( + bs->entropy.fse.matchlengthCTable, + matchlengthNCount, matchlengthMaxValue, matchlengthLog, + workspace, HUF_WORKSPACE_SIZE)), + dictionary_corrupted, ""); + bs->entropy.fse.matchlength_repeatMode = ZSTD_dictNCountRepeat(matchlengthNCount, matchlengthMaxValue, MaxML); + dictPtr += matchlengthHeaderSize; + } + + { short litlengthNCount[MaxLL+1]; + unsigned litlengthMaxValue = MaxLL, litlengthLog; + size_t const litlengthHeaderSize = FSE_readNCount(litlengthNCount, &litlengthMaxValue, &litlengthLog, dictPtr, dictEnd-dictPtr); + RETURN_ERROR_IF(FSE_isError(litlengthHeaderSize), dictionary_corrupted, ""); + RETURN_ERROR_IF(litlengthLog > LLFSELog, dictionary_corrupted, ""); + RETURN_ERROR_IF(FSE_isError(FSE_buildCTable_wksp( + bs->entropy.fse.litlengthCTable, + litlengthNCount, litlengthMaxValue, litlengthLog, + workspace, HUF_WORKSPACE_SIZE)), + dictionary_corrupted, ""); + bs->entropy.fse.litlength_repeatMode = ZSTD_dictNCountRepeat(litlengthNCount, litlengthMaxValue, MaxLL); + dictPtr += litlengthHeaderSize; + } + + RETURN_ERROR_IF(dictPtr+12 > dictEnd, dictionary_corrupted, ""); + bs->rep[0] = MEM_readLE32(dictPtr+0); + bs->rep[1] = MEM_readLE32(dictPtr+4); + bs->rep[2] = MEM_readLE32(dictPtr+8); + dictPtr += 12; + + { size_t const dictContentSize = (size_t)(dictEnd - dictPtr); + U32 offcodeMax = MaxOff; + if (dictContentSize <= ((U32)-1) - 128 KB) { + U32 const maxOffset = (U32)dictContentSize + 128 KB; /* The maximum offset that must be supported */ + offcodeMax = ZSTD_highbit32(maxOffset); /* Calculate minimum offset code required to represent maxOffset */ + } + /* All offset values <= dictContentSize + 128 KB must be representable for a valid table */ + bs->entropy.fse.offcode_repeatMode = ZSTD_dictNCountRepeat(offcodeNCount, offcodeMaxValue, MIN(offcodeMax, MaxOff)); + + /* All repCodes must be <= dictContentSize and != 0 */ + { U32 u; + for (u=0; u<3; u++) { + RETURN_ERROR_IF(bs->rep[u] == 0, dictionary_corrupted, ""); + RETURN_ERROR_IF(bs->rep[u] > dictContentSize, dictionary_corrupted, ""); + } } } + + return dictPtr - (const BYTE*)dict; +} + +/* Dictionary format : + * See : + * https://github.com/facebook/zstd/blob/release/doc/zstd_compression_format.md#dictionary-format + */ +/*! ZSTD_loadZstdDictionary() : + * @return : dictID, or an error code + * assumptions : magic number supposed already checked + * dictSize supposed >= 8 + */ +static size_t ZSTD_loadZstdDictionary(ZSTD_compressedBlockState_t* bs, + ZSTD_matchState_t* ms, + ZSTD_cwksp* ws, + ZSTD_CCtx_params const* params, + const void* dict, size_t dictSize, + ZSTD_dictTableLoadMethod_e dtlm, + void* workspace) +{ + const BYTE* dictPtr = (const BYTE*)dict; + const BYTE* const dictEnd = dictPtr + dictSize; + size_t dictID; + size_t eSize; + + ZSTD_STATIC_ASSERT(HUF_WORKSPACE_SIZE >= (1<= 8); + assert(MEM_readLE32(dictPtr) == ZSTD_MAGIC_DICTIONARY); + + dictID = params->fParams.noDictIDFlag ? 0 : MEM_readLE32(dictPtr + 4 /* skip magic number */ ); + eSize = ZSTD_loadCEntropy(bs, workspace, dict, dictSize); + FORWARD_IF_ERROR(eSize, "ZSTD_loadCEntropy failed"); + dictPtr += eSize; + + { + size_t const dictContentSize = (size_t)(dictEnd - dictPtr); + FORWARD_IF_ERROR(ZSTD_loadDictionaryContent( + ms, NULL, ws, params, dictPtr, dictContentSize, dtlm), ""); + } + return dictID; +} + +/* ZSTD_compress_insertDictionary() : +* @return : dictID, or an error code */ +static size_t +ZSTD_compress_insertDictionary(ZSTD_compressedBlockState_t* bs, + ZSTD_matchState_t* ms, + ldmState_t* ls, + ZSTD_cwksp* ws, + const ZSTD_CCtx_params* params, + const void* dict, size_t dictSize, + ZSTD_dictContentType_e dictContentType, + ZSTD_dictTableLoadMethod_e dtlm, + void* workspace) +{ + DEBUGLOG(4, "ZSTD_compress_insertDictionary (dictSize=%u)", (U32)dictSize); + if ((dict==NULL) || (dictSize<8)) { + RETURN_ERROR_IF(dictContentType == ZSTD_dct_fullDict, dictionary_wrong, ""); + return 0; + } + + ZSTD_reset_compressedBlockState(bs); + + /* dict restricted modes */ + if (dictContentType == ZSTD_dct_rawContent) + return ZSTD_loadDictionaryContent(ms, ls, ws, params, dict, dictSize, dtlm); + + if (MEM_readLE32(dict) != ZSTD_MAGIC_DICTIONARY) { + if (dictContentType == ZSTD_dct_auto) { + DEBUGLOG(4, "raw content dictionary detected"); + return ZSTD_loadDictionaryContent( + ms, ls, ws, params, dict, dictSize, dtlm); + } + RETURN_ERROR_IF(dictContentType == ZSTD_dct_fullDict, dictionary_wrong, ""); + assert(0); /* impossible */ + } + + /* dict as full zstd dictionary */ + return ZSTD_loadZstdDictionary( + bs, ms, ws, params, dict, dictSize, dtlm, workspace); +} + +#define ZSTD_USE_CDICT_PARAMS_SRCSIZE_CUTOFF (128 KB) +#define ZSTD_USE_CDICT_PARAMS_DICTSIZE_MULTIPLIER (6ULL) + +/*! ZSTD_compressBegin_internal() : + * @return : 0, or an error code */ +static size_t ZSTD_compressBegin_internal(ZSTD_CCtx* cctx, + const void* dict, size_t dictSize, + ZSTD_dictContentType_e dictContentType, + ZSTD_dictTableLoadMethod_e dtlm, + const ZSTD_CDict* cdict, + const ZSTD_CCtx_params* params, U64 pledgedSrcSize, + ZSTD_buffered_policy_e zbuff) +{ + DEBUGLOG(4, "ZSTD_compressBegin_internal: wlog=%u", params->cParams.windowLog); + /* params are supposed to be fully validated at this point */ + assert(!ZSTD_isError(ZSTD_checkCParams(params->cParams))); + assert(!((dict) && (cdict))); /* either dict or cdict, not both */ + if ( (cdict) + && (cdict->dictContentSize > 0) + && ( pledgedSrcSize < ZSTD_USE_CDICT_PARAMS_SRCSIZE_CUTOFF + || pledgedSrcSize < cdict->dictContentSize * ZSTD_USE_CDICT_PARAMS_DICTSIZE_MULTIPLIER + || pledgedSrcSize == ZSTD_CONTENTSIZE_UNKNOWN + || cdict->compressionLevel == 0) + && (params->attachDictPref != ZSTD_dictForceLoad) ) { + return ZSTD_resetCCtx_usingCDict(cctx, cdict, params, pledgedSrcSize, zbuff); + } + + FORWARD_IF_ERROR( ZSTD_resetCCtx_internal(cctx, *params, pledgedSrcSize, + ZSTDcrp_makeClean, zbuff) , ""); + { size_t const dictID = cdict ? + ZSTD_compress_insertDictionary( + cctx->blockState.prevCBlock, &cctx->blockState.matchState, + &cctx->ldmState, &cctx->workspace, &cctx->appliedParams, cdict->dictContent, + cdict->dictContentSize, cdict->dictContentType, dtlm, + cctx->entropyWorkspace) + : ZSTD_compress_insertDictionary( + cctx->blockState.prevCBlock, &cctx->blockState.matchState, + &cctx->ldmState, &cctx->workspace, &cctx->appliedParams, dict, dictSize, + dictContentType, dtlm, cctx->entropyWorkspace); + FORWARD_IF_ERROR(dictID, "ZSTD_compress_insertDictionary failed"); + assert(dictID <= UINT_MAX); + cctx->dictID = (U32)dictID; + cctx->dictContentSize = cdict ? cdict->dictContentSize : dictSize; + } + return 0; +} + +size_t ZSTD_compressBegin_advanced_internal(ZSTD_CCtx* cctx, + const void* dict, size_t dictSize, + ZSTD_dictContentType_e dictContentType, + ZSTD_dictTableLoadMethod_e dtlm, + const ZSTD_CDict* cdict, + const ZSTD_CCtx_params* params, + unsigned long long pledgedSrcSize) +{ + DEBUGLOG(4, "ZSTD_compressBegin_advanced_internal: wlog=%u", params->cParams.windowLog); + /* compression parameters verification and optimization */ + FORWARD_IF_ERROR( ZSTD_checkCParams(params->cParams) , ""); + return ZSTD_compressBegin_internal(cctx, + dict, dictSize, dictContentType, dtlm, + cdict, + params, pledgedSrcSize, + ZSTDb_not_buffered); +} + +/*! ZSTD_compressBegin_advanced() : +* @return : 0, or an error code */ +size_t ZSTD_compressBegin_advanced(ZSTD_CCtx* cctx, + const void* dict, size_t dictSize, + ZSTD_parameters params, unsigned long long pledgedSrcSize) +{ + ZSTD_CCtx_params cctxParams; + ZSTD_CCtxParams_init_internal(&cctxParams, ¶ms, ZSTD_NO_CLEVEL); + return ZSTD_compressBegin_advanced_internal(cctx, + dict, dictSize, ZSTD_dct_auto, ZSTD_dtlm_fast, + NULL /*cdict*/, + &cctxParams, pledgedSrcSize); +} + +size_t ZSTD_compressBegin_usingDict(ZSTD_CCtx* cctx, const void* dict, size_t dictSize, int compressionLevel) +{ + ZSTD_CCtx_params cctxParams; + { + ZSTD_parameters const params = ZSTD_getParams_internal(compressionLevel, ZSTD_CONTENTSIZE_UNKNOWN, dictSize, ZSTD_cpm_noAttachDict); + ZSTD_CCtxParams_init_internal(&cctxParams, ¶ms, (compressionLevel == 0) ? ZSTD_CLEVEL_DEFAULT : compressionLevel); + } + DEBUGLOG(4, "ZSTD_compressBegin_usingDict (dictSize=%u)", (unsigned)dictSize); + return ZSTD_compressBegin_internal(cctx, dict, dictSize, ZSTD_dct_auto, ZSTD_dtlm_fast, NULL, + &cctxParams, ZSTD_CONTENTSIZE_UNKNOWN, ZSTDb_not_buffered); +} + +size_t ZSTD_compressBegin(ZSTD_CCtx* cctx, int compressionLevel) +{ + return ZSTD_compressBegin_usingDict(cctx, NULL, 0, compressionLevel); +} + + +/*! ZSTD_writeEpilogue() : +* Ends a frame. +* @return : nb of bytes written into dst (or an error code) */ +static size_t ZSTD_writeEpilogue(ZSTD_CCtx* cctx, void* dst, size_t dstCapacity) +{ + BYTE* const ostart = (BYTE*)dst; + BYTE* op = ostart; + size_t fhSize = 0; + + DEBUGLOG(4, "ZSTD_writeEpilogue"); + RETURN_ERROR_IF(cctx->stage == ZSTDcs_created, stage_wrong, "init missing"); + + /* special case : empty frame */ + if (cctx->stage == ZSTDcs_init) { + fhSize = ZSTD_writeFrameHeader(dst, dstCapacity, &cctx->appliedParams, 0, 0); + FORWARD_IF_ERROR(fhSize, "ZSTD_writeFrameHeader failed"); + dstCapacity -= fhSize; + op += fhSize; + cctx->stage = ZSTDcs_ongoing; + } + + if (cctx->stage != ZSTDcs_ending) { + /* write one last empty block, make it the "last" block */ + U32 const cBlockHeader24 = 1 /* last block */ + (((U32)bt_raw)<<1) + 0; + RETURN_ERROR_IF(dstCapacity<4, dstSize_tooSmall, "no room for epilogue"); + MEM_writeLE32(op, cBlockHeader24); + op += ZSTD_blockHeaderSize; + dstCapacity -= ZSTD_blockHeaderSize; + } + + if (cctx->appliedParams.fParams.checksumFlag) { + U32 const checksum = (U32) xxh64_digest(&cctx->xxhState); + RETURN_ERROR_IF(dstCapacity<4, dstSize_tooSmall, "no room for checksum"); + DEBUGLOG(4, "ZSTD_writeEpilogue: write checksum : %08X", (unsigned)checksum); + MEM_writeLE32(op, checksum); + op += 4; + } + + cctx->stage = ZSTDcs_created; /* return to "created but no init" status */ + return op-ostart; +} + +void ZSTD_CCtx_trace(ZSTD_CCtx* cctx, size_t extraCSize) +{ + (void)cctx; + (void)extraCSize; +} + +size_t ZSTD_compressEnd (ZSTD_CCtx* cctx, + void* dst, size_t dstCapacity, + const void* src, size_t srcSize) +{ + size_t endResult; + size_t const cSize = ZSTD_compressContinue_internal(cctx, + dst, dstCapacity, src, srcSize, + 1 /* frame mode */, 1 /* last chunk */); + FORWARD_IF_ERROR(cSize, "ZSTD_compressContinue_internal failed"); + endResult = ZSTD_writeEpilogue(cctx, (char*)dst + cSize, dstCapacity-cSize); + FORWARD_IF_ERROR(endResult, "ZSTD_writeEpilogue failed"); + assert(!(cctx->appliedParams.fParams.contentSizeFlag && cctx->pledgedSrcSizePlusOne == 0)); + if (cctx->pledgedSrcSizePlusOne != 0) { /* control src size */ + ZSTD_STATIC_ASSERT(ZSTD_CONTENTSIZE_UNKNOWN == (unsigned long long)-1); + DEBUGLOG(4, "end of frame : controlling src size"); + RETURN_ERROR_IF( + cctx->pledgedSrcSizePlusOne != cctx->consumedSrcSize+1, + srcSize_wrong, + "error : pledgedSrcSize = %u, while realSrcSize = %u", + (unsigned)cctx->pledgedSrcSizePlusOne-1, + (unsigned)cctx->consumedSrcSize); + } + ZSTD_CCtx_trace(cctx, endResult); + return cSize + endResult; +} + +size_t ZSTD_compress_advanced (ZSTD_CCtx* cctx, + void* dst, size_t dstCapacity, + const void* src, size_t srcSize, + const void* dict,size_t dictSize, + ZSTD_parameters params) +{ + ZSTD_CCtx_params cctxParams; + DEBUGLOG(4, "ZSTD_compress_advanced"); + FORWARD_IF_ERROR(ZSTD_checkCParams(params.cParams), ""); + ZSTD_CCtxParams_init_internal(&cctxParams, ¶ms, ZSTD_NO_CLEVEL); + return ZSTD_compress_advanced_internal(cctx, + dst, dstCapacity, + src, srcSize, + dict, dictSize, + &cctxParams); +} + +/* Internal */ +size_t ZSTD_compress_advanced_internal( + ZSTD_CCtx* cctx, + void* dst, size_t dstCapacity, + const void* src, size_t srcSize, + const void* dict,size_t dictSize, + const ZSTD_CCtx_params* params) +{ + DEBUGLOG(4, "ZSTD_compress_advanced_internal (srcSize:%u)", (unsigned)srcSize); + FORWARD_IF_ERROR( ZSTD_compressBegin_internal(cctx, + dict, dictSize, ZSTD_dct_auto, ZSTD_dtlm_fast, NULL, + params, srcSize, ZSTDb_not_buffered) , ""); + return ZSTD_compressEnd(cctx, dst, dstCapacity, src, srcSize); +} + +size_t ZSTD_compress_usingDict(ZSTD_CCtx* cctx, + void* dst, size_t dstCapacity, + const void* src, size_t srcSize, + const void* dict, size_t dictSize, + int compressionLevel) +{ + ZSTD_CCtx_params cctxParams; + { + ZSTD_parameters const params = ZSTD_getParams_internal(compressionLevel, srcSize, dict ? dictSize : 0, ZSTD_cpm_noAttachDict); + assert(params.fParams.contentSizeFlag == 1); + ZSTD_CCtxParams_init_internal(&cctxParams, ¶ms, (compressionLevel == 0) ? ZSTD_CLEVEL_DEFAULT: compressionLevel); + } + DEBUGLOG(4, "ZSTD_compress_usingDict (srcSize=%u)", (unsigned)srcSize); + return ZSTD_compress_advanced_internal(cctx, dst, dstCapacity, src, srcSize, dict, dictSize, &cctxParams); +} + +size_t ZSTD_compressCCtx(ZSTD_CCtx* cctx, + void* dst, size_t dstCapacity, + const void* src, size_t srcSize, + int compressionLevel) +{ + DEBUGLOG(4, "ZSTD_compressCCtx (srcSize=%u)", (unsigned)srcSize); + assert(cctx != NULL); + return ZSTD_compress_usingDict(cctx, dst, dstCapacity, src, srcSize, NULL, 0, compressionLevel); +} + +size_t ZSTD_compress(void* dst, size_t dstCapacity, + const void* src, size_t srcSize, + int compressionLevel) +{ + size_t result; + ZSTD_CCtx* cctx = ZSTD_createCCtx(); + RETURN_ERROR_IF(!cctx, memory_allocation, "ZSTD_createCCtx failed"); + result = ZSTD_compressCCtx(cctx, dst, dstCapacity, src, srcSize, compressionLevel); + ZSTD_freeCCtx(cctx); + return result; +} + + +/* ===== Dictionary API ===== */ + +/*! ZSTD_estimateCDictSize_advanced() : + * Estimate amount of memory that will be needed to create a dictionary with following arguments */ +size_t ZSTD_estimateCDictSize_advanced( + size_t dictSize, ZSTD_compressionParameters cParams, + ZSTD_dictLoadMethod_e dictLoadMethod) +{ + DEBUGLOG(5, "sizeof(ZSTD_CDict) : %u", (unsigned)sizeof(ZSTD_CDict)); + return ZSTD_cwksp_alloc_size(sizeof(ZSTD_CDict)) + + ZSTD_cwksp_alloc_size(HUF_WORKSPACE_SIZE) + + ZSTD_sizeof_matchState(&cParams, /* forCCtx */ 0) + + (dictLoadMethod == ZSTD_dlm_byRef ? 0 + : ZSTD_cwksp_alloc_size(ZSTD_cwksp_align(dictSize, sizeof(void *)))); +} + +size_t ZSTD_estimateCDictSize(size_t dictSize, int compressionLevel) +{ + ZSTD_compressionParameters const cParams = ZSTD_getCParams_internal(compressionLevel, ZSTD_CONTENTSIZE_UNKNOWN, dictSize, ZSTD_cpm_createCDict); + return ZSTD_estimateCDictSize_advanced(dictSize, cParams, ZSTD_dlm_byCopy); +} + +size_t ZSTD_sizeof_CDict(const ZSTD_CDict* cdict) +{ + if (cdict==NULL) return 0; /* support sizeof on NULL */ + DEBUGLOG(5, "sizeof(*cdict) : %u", (unsigned)sizeof(*cdict)); + /* cdict may be in the workspace */ + return (cdict->workspace.workspace == cdict ? 0 : sizeof(*cdict)) + + ZSTD_cwksp_sizeof(&cdict->workspace); +} + +static size_t ZSTD_initCDict_internal( + ZSTD_CDict* cdict, + const void* dictBuffer, size_t dictSize, + ZSTD_dictLoadMethod_e dictLoadMethod, + ZSTD_dictContentType_e dictContentType, + ZSTD_CCtx_params params) +{ + DEBUGLOG(3, "ZSTD_initCDict_internal (dictContentType:%u)", (unsigned)dictContentType); + assert(!ZSTD_checkCParams(params.cParams)); + cdict->matchState.cParams = params.cParams; + cdict->matchState.dedicatedDictSearch = params.enableDedicatedDictSearch; + if (cdict->matchState.dedicatedDictSearch && dictSize > ZSTD_CHUNKSIZE_MAX) { + cdict->matchState.dedicatedDictSearch = 0; + } + if ((dictLoadMethod == ZSTD_dlm_byRef) || (!dictBuffer) || (!dictSize)) { + cdict->dictContent = dictBuffer; + } else { + void *internalBuffer = ZSTD_cwksp_reserve_object(&cdict->workspace, ZSTD_cwksp_align(dictSize, sizeof(void*))); + RETURN_ERROR_IF(!internalBuffer, memory_allocation, "NULL pointer!"); + cdict->dictContent = internalBuffer; + ZSTD_memcpy(internalBuffer, dictBuffer, dictSize); + } + cdict->dictContentSize = dictSize; + cdict->dictContentType = dictContentType; + + cdict->entropyWorkspace = (U32*)ZSTD_cwksp_reserve_object(&cdict->workspace, HUF_WORKSPACE_SIZE); + + + /* Reset the state to no dictionary */ + ZSTD_reset_compressedBlockState(&cdict->cBlockState); + FORWARD_IF_ERROR(ZSTD_reset_matchState( + &cdict->matchState, + &cdict->workspace, + ¶ms.cParams, + ZSTDcrp_makeClean, + ZSTDirp_reset, + ZSTD_resetTarget_CDict), ""); + /* (Maybe) load the dictionary + * Skips loading the dictionary if it is < 8 bytes. + */ + { params.compressionLevel = ZSTD_CLEVEL_DEFAULT; + params.fParams.contentSizeFlag = 1; + { size_t const dictID = ZSTD_compress_insertDictionary( + &cdict->cBlockState, &cdict->matchState, NULL, &cdict->workspace, + ¶ms, cdict->dictContent, cdict->dictContentSize, + dictContentType, ZSTD_dtlm_full, cdict->entropyWorkspace); + FORWARD_IF_ERROR(dictID, "ZSTD_compress_insertDictionary failed"); + assert(dictID <= (size_t)(U32)-1); + cdict->dictID = (U32)dictID; + } + } + + return 0; +} + +static ZSTD_CDict* ZSTD_createCDict_advanced_internal(size_t dictSize, + ZSTD_dictLoadMethod_e dictLoadMethod, + ZSTD_compressionParameters cParams, ZSTD_customMem customMem) +{ + if ((!customMem.customAlloc) ^ (!customMem.customFree)) return NULL; + + { size_t const workspaceSize = + ZSTD_cwksp_alloc_size(sizeof(ZSTD_CDict)) + + ZSTD_cwksp_alloc_size(HUF_WORKSPACE_SIZE) + + ZSTD_sizeof_matchState(&cParams, /* forCCtx */ 0) + + (dictLoadMethod == ZSTD_dlm_byRef ? 0 + : ZSTD_cwksp_alloc_size(ZSTD_cwksp_align(dictSize, sizeof(void*)))); + void* const workspace = ZSTD_customMalloc(workspaceSize, customMem); + ZSTD_cwksp ws; + ZSTD_CDict* cdict; + + if (!workspace) { + ZSTD_customFree(workspace, customMem); + return NULL; + } + + ZSTD_cwksp_init(&ws, workspace, workspaceSize, ZSTD_cwksp_dynamic_alloc); + + cdict = (ZSTD_CDict*)ZSTD_cwksp_reserve_object(&ws, sizeof(ZSTD_CDict)); + assert(cdict != NULL); + ZSTD_cwksp_move(&cdict->workspace, &ws); + cdict->customMem = customMem; + cdict->compressionLevel = ZSTD_NO_CLEVEL; /* signals advanced API usage */ + + return cdict; + } +} + +ZSTD_CDict* ZSTD_createCDict_advanced(const void* dictBuffer, size_t dictSize, + ZSTD_dictLoadMethod_e dictLoadMethod, + ZSTD_dictContentType_e dictContentType, + ZSTD_compressionParameters cParams, + ZSTD_customMem customMem) +{ + ZSTD_CCtx_params cctxParams; + ZSTD_memset(&cctxParams, 0, sizeof(cctxParams)); + ZSTD_CCtxParams_init(&cctxParams, 0); + cctxParams.cParams = cParams; + cctxParams.customMem = customMem; + return ZSTD_createCDict_advanced2( + dictBuffer, dictSize, + dictLoadMethod, dictContentType, + &cctxParams, customMem); +} + +ZSTDLIB_API ZSTD_CDict* ZSTD_createCDict_advanced2( + const void* dict, size_t dictSize, + ZSTD_dictLoadMethod_e dictLoadMethod, + ZSTD_dictContentType_e dictContentType, + const ZSTD_CCtx_params* originalCctxParams, + ZSTD_customMem customMem) +{ + ZSTD_CCtx_params cctxParams = *originalCctxParams; + ZSTD_compressionParameters cParams; + ZSTD_CDict* cdict; + + DEBUGLOG(3, "ZSTD_createCDict_advanced2, mode %u", (unsigned)dictContentType); + if (!customMem.customAlloc ^ !customMem.customFree) return NULL; + + if (cctxParams.enableDedicatedDictSearch) { + cParams = ZSTD_dedicatedDictSearch_getCParams( + cctxParams.compressionLevel, dictSize); + ZSTD_overrideCParams(&cParams, &cctxParams.cParams); + } else { + cParams = ZSTD_getCParamsFromCCtxParams( + &cctxParams, ZSTD_CONTENTSIZE_UNKNOWN, dictSize, ZSTD_cpm_createCDict); + } + + if (!ZSTD_dedicatedDictSearch_isSupported(&cParams)) { + /* Fall back to non-DDSS params */ + cctxParams.enableDedicatedDictSearch = 0; + cParams = ZSTD_getCParamsFromCCtxParams( + &cctxParams, ZSTD_CONTENTSIZE_UNKNOWN, dictSize, ZSTD_cpm_createCDict); + } + + cctxParams.cParams = cParams; + + cdict = ZSTD_createCDict_advanced_internal(dictSize, + dictLoadMethod, cctxParams.cParams, + customMem); + + if (ZSTD_isError( ZSTD_initCDict_internal(cdict, + dict, dictSize, + dictLoadMethod, dictContentType, + cctxParams) )) { + ZSTD_freeCDict(cdict); + return NULL; + } + + return cdict; +} + +ZSTD_CDict* ZSTD_createCDict(const void* dict, size_t dictSize, int compressionLevel) +{ + ZSTD_compressionParameters cParams = ZSTD_getCParams_internal(compressionLevel, ZSTD_CONTENTSIZE_UNKNOWN, dictSize, ZSTD_cpm_createCDict); + ZSTD_CDict* const cdict = ZSTD_createCDict_advanced(dict, dictSize, + ZSTD_dlm_byCopy, ZSTD_dct_auto, + cParams, ZSTD_defaultCMem); + if (cdict) + cdict->compressionLevel = (compressionLevel == 0) ? ZSTD_CLEVEL_DEFAULT : compressionLevel; + return cdict; +} + +ZSTD_CDict* ZSTD_createCDict_byReference(const void* dict, size_t dictSize, int compressionLevel) +{ + ZSTD_compressionParameters cParams = ZSTD_getCParams_internal(compressionLevel, ZSTD_CONTENTSIZE_UNKNOWN, dictSize, ZSTD_cpm_createCDict); + ZSTD_CDict* const cdict = ZSTD_createCDict_advanced(dict, dictSize, + ZSTD_dlm_byRef, ZSTD_dct_auto, + cParams, ZSTD_defaultCMem); + if (cdict) + cdict->compressionLevel = (compressionLevel == 0) ? ZSTD_CLEVEL_DEFAULT : compressionLevel; + return cdict; +} + +size_t ZSTD_freeCDict(ZSTD_CDict* cdict) +{ + if (cdict==NULL) return 0; /* support free on NULL */ + { ZSTD_customMem const cMem = cdict->customMem; + int cdictInWorkspace = ZSTD_cwksp_owns_buffer(&cdict->workspace, cdict); + ZSTD_cwksp_free(&cdict->workspace, cMem); + if (!cdictInWorkspace) { + ZSTD_customFree(cdict, cMem); + } + return 0; + } +} + +/*! ZSTD_initStaticCDict_advanced() : + * Generate a digested dictionary in provided memory area. + * workspace: The memory area to emplace the dictionary into. + * Provided pointer must 8-bytes aligned. + * It must outlive dictionary usage. + * workspaceSize: Use ZSTD_estimateCDictSize() + * to determine how large workspace must be. + * cParams : use ZSTD_getCParams() to transform a compression level + * into its relevants cParams. + * @return : pointer to ZSTD_CDict*, or NULL if error (size too small) + * Note : there is no corresponding "free" function. + * Since workspace was allocated externally, it must be freed externally. + */ +const ZSTD_CDict* ZSTD_initStaticCDict( + void* workspace, size_t workspaceSize, + const void* dict, size_t dictSize, + ZSTD_dictLoadMethod_e dictLoadMethod, + ZSTD_dictContentType_e dictContentType, + ZSTD_compressionParameters cParams) +{ + size_t const matchStateSize = ZSTD_sizeof_matchState(&cParams, /* forCCtx */ 0); + size_t const neededSize = ZSTD_cwksp_alloc_size(sizeof(ZSTD_CDict)) + + (dictLoadMethod == ZSTD_dlm_byRef ? 0 + : ZSTD_cwksp_alloc_size(ZSTD_cwksp_align(dictSize, sizeof(void*)))) + + ZSTD_cwksp_alloc_size(HUF_WORKSPACE_SIZE) + + matchStateSize; + ZSTD_CDict* cdict; + ZSTD_CCtx_params params; + + if ((size_t)workspace & 7) return NULL; /* 8-aligned */ + + { + ZSTD_cwksp ws; + ZSTD_cwksp_init(&ws, workspace, workspaceSize, ZSTD_cwksp_static_alloc); + cdict = (ZSTD_CDict*)ZSTD_cwksp_reserve_object(&ws, sizeof(ZSTD_CDict)); + if (cdict == NULL) return NULL; + ZSTD_cwksp_move(&cdict->workspace, &ws); + } + + DEBUGLOG(4, "(workspaceSize < neededSize) : (%u < %u) => %u", + (unsigned)workspaceSize, (unsigned)neededSize, (unsigned)(workspaceSize < neededSize)); + if (workspaceSize < neededSize) return NULL; + + ZSTD_CCtxParams_init(¶ms, 0); + params.cParams = cParams; + + if (ZSTD_isError( ZSTD_initCDict_internal(cdict, + dict, dictSize, + dictLoadMethod, dictContentType, + params) )) + return NULL; + + return cdict; +} + +ZSTD_compressionParameters ZSTD_getCParamsFromCDict(const ZSTD_CDict* cdict) +{ + assert(cdict != NULL); + return cdict->matchState.cParams; +} + +/*! ZSTD_getDictID_fromCDict() : + * Provides the dictID of the dictionary loaded into `cdict`. + * If @return == 0, the dictionary is not conformant to Zstandard specification, or empty. + * Non-conformant dictionaries can still be loaded, but as content-only dictionaries. */ +unsigned ZSTD_getDictID_fromCDict(const ZSTD_CDict* cdict) +{ + if (cdict==NULL) return 0; + return cdict->dictID; +} + + +/* ZSTD_compressBegin_usingCDict_advanced() : + * cdict must be != NULL */ +size_t ZSTD_compressBegin_usingCDict_advanced( + ZSTD_CCtx* const cctx, const ZSTD_CDict* const cdict, + ZSTD_frameParameters const fParams, unsigned long long const pledgedSrcSize) +{ + ZSTD_CCtx_params cctxParams; + DEBUGLOG(4, "ZSTD_compressBegin_usingCDict_advanced"); + RETURN_ERROR_IF(cdict==NULL, dictionary_wrong, "NULL pointer!"); + /* Initialize the cctxParams from the cdict */ + { + ZSTD_parameters params; + params.fParams = fParams; + params.cParams = ( pledgedSrcSize < ZSTD_USE_CDICT_PARAMS_SRCSIZE_CUTOFF + || pledgedSrcSize < cdict->dictContentSize * ZSTD_USE_CDICT_PARAMS_DICTSIZE_MULTIPLIER + || pledgedSrcSize == ZSTD_CONTENTSIZE_UNKNOWN + || cdict->compressionLevel == 0 ) ? + ZSTD_getCParamsFromCDict(cdict) + : ZSTD_getCParams(cdict->compressionLevel, + pledgedSrcSize, + cdict->dictContentSize); + ZSTD_CCtxParams_init_internal(&cctxParams, ¶ms, cdict->compressionLevel); + } + /* Increase window log to fit the entire dictionary and source if the + * source size is known. Limit the increase to 19, which is the + * window log for compression level 1 with the largest source size. + */ + if (pledgedSrcSize != ZSTD_CONTENTSIZE_UNKNOWN) { + U32 const limitedSrcSize = (U32)MIN(pledgedSrcSize, 1U << 19); + U32 const limitedSrcLog = limitedSrcSize > 1 ? ZSTD_highbit32(limitedSrcSize - 1) + 1 : 1; + cctxParams.cParams.windowLog = MAX(cctxParams.cParams.windowLog, limitedSrcLog); + } + return ZSTD_compressBegin_internal(cctx, + NULL, 0, ZSTD_dct_auto, ZSTD_dtlm_fast, + cdict, + &cctxParams, pledgedSrcSize, + ZSTDb_not_buffered); +} + +/* ZSTD_compressBegin_usingCDict() : + * pledgedSrcSize=0 means "unknown" + * if pledgedSrcSize>0, it will enable contentSizeFlag */ +size_t ZSTD_compressBegin_usingCDict(ZSTD_CCtx* cctx, const ZSTD_CDict* cdict) +{ + ZSTD_frameParameters const fParams = { 0 /*content*/, 0 /*checksum*/, 0 /*noDictID*/ }; + DEBUGLOG(4, "ZSTD_compressBegin_usingCDict : dictIDFlag == %u", !fParams.noDictIDFlag); + return ZSTD_compressBegin_usingCDict_advanced(cctx, cdict, fParams, ZSTD_CONTENTSIZE_UNKNOWN); +} + +size_t ZSTD_compress_usingCDict_advanced(ZSTD_CCtx* cctx, + void* dst, size_t dstCapacity, + const void* src, size_t srcSize, + const ZSTD_CDict* cdict, ZSTD_frameParameters fParams) +{ + FORWARD_IF_ERROR(ZSTD_compressBegin_usingCDict_advanced(cctx, cdict, fParams, srcSize), ""); /* will check if cdict != NULL */ + return ZSTD_compressEnd(cctx, dst, dstCapacity, src, srcSize); +} + +/*! ZSTD_compress_usingCDict() : + * Compression using a digested Dictionary. + * Faster startup than ZSTD_compress_usingDict(), recommended when same dictionary is used multiple times. + * Note that compression parameters are decided at CDict creation time + * while frame parameters are hardcoded */ +size_t ZSTD_compress_usingCDict(ZSTD_CCtx* cctx, + void* dst, size_t dstCapacity, + const void* src, size_t srcSize, + const ZSTD_CDict* cdict) +{ + ZSTD_frameParameters const fParams = { 1 /*content*/, 0 /*checksum*/, 0 /*noDictID*/ }; + return ZSTD_compress_usingCDict_advanced(cctx, dst, dstCapacity, src, srcSize, cdict, fParams); +} + + + +/* ****************************************************************** +* Streaming +********************************************************************/ + +ZSTD_CStream* ZSTD_createCStream(void) +{ + DEBUGLOG(3, "ZSTD_createCStream"); + return ZSTD_createCStream_advanced(ZSTD_defaultCMem); +} + +ZSTD_CStream* ZSTD_initStaticCStream(void *workspace, size_t workspaceSize) +{ + return ZSTD_initStaticCCtx(workspace, workspaceSize); +} + +ZSTD_CStream* ZSTD_createCStream_advanced(ZSTD_customMem customMem) +{ /* CStream and CCtx are now same object */ + return ZSTD_createCCtx_advanced(customMem); +} + +size_t ZSTD_freeCStream(ZSTD_CStream* zcs) +{ + return ZSTD_freeCCtx(zcs); /* same object */ +} + + + +/*====== Initialization ======*/ + +size_t ZSTD_CStreamInSize(void) { return ZSTD_BLOCKSIZE_MAX; } + +size_t ZSTD_CStreamOutSize(void) +{ + return ZSTD_compressBound(ZSTD_BLOCKSIZE_MAX) + ZSTD_blockHeaderSize + 4 /* 32-bits hash */ ; +} + +static ZSTD_cParamMode_e ZSTD_getCParamMode(ZSTD_CDict const* cdict, ZSTD_CCtx_params const* params, U64 pledgedSrcSize) +{ + if (cdict != NULL && ZSTD_shouldAttachDict(cdict, params, pledgedSrcSize)) + return ZSTD_cpm_attachDict; + else + return ZSTD_cpm_noAttachDict; +} + +/* ZSTD_resetCStream(): + * pledgedSrcSize == 0 means "unknown" */ +size_t ZSTD_resetCStream(ZSTD_CStream* zcs, unsigned long long pss) +{ + /* temporary : 0 interpreted as "unknown" during transition period. + * Users willing to specify "unknown" **must** use ZSTD_CONTENTSIZE_UNKNOWN. + * 0 will be interpreted as "empty" in the future. + */ + U64 const pledgedSrcSize = (pss==0) ? ZSTD_CONTENTSIZE_UNKNOWN : pss; + DEBUGLOG(4, "ZSTD_resetCStream: pledgedSrcSize = %u", (unsigned)pledgedSrcSize); + FORWARD_IF_ERROR( ZSTD_CCtx_reset(zcs, ZSTD_reset_session_only) , ""); + FORWARD_IF_ERROR( ZSTD_CCtx_setPledgedSrcSize(zcs, pledgedSrcSize) , ""); + return 0; +} + +/*! ZSTD_initCStream_internal() : + * Note : for lib/compress only. Used by zstdmt_compress.c. + * Assumption 1 : params are valid + * Assumption 2 : either dict, or cdict, is defined, not both */ +size_t ZSTD_initCStream_internal(ZSTD_CStream* zcs, + const void* dict, size_t dictSize, const ZSTD_CDict* cdict, + const ZSTD_CCtx_params* params, + unsigned long long pledgedSrcSize) +{ + DEBUGLOG(4, "ZSTD_initCStream_internal"); + FORWARD_IF_ERROR( ZSTD_CCtx_reset(zcs, ZSTD_reset_session_only) , ""); + FORWARD_IF_ERROR( ZSTD_CCtx_setPledgedSrcSize(zcs, pledgedSrcSize) , ""); + assert(!ZSTD_isError(ZSTD_checkCParams(params->cParams))); + zcs->requestedParams = *params; + assert(!((dict) && (cdict))); /* either dict or cdict, not both */ + if (dict) { + FORWARD_IF_ERROR( ZSTD_CCtx_loadDictionary(zcs, dict, dictSize) , ""); + } else { + /* Dictionary is cleared if !cdict */ + FORWARD_IF_ERROR( ZSTD_CCtx_refCDict(zcs, cdict) , ""); + } + return 0; +} + +/* ZSTD_initCStream_usingCDict_advanced() : + * same as ZSTD_initCStream_usingCDict(), with control over frame parameters */ +size_t ZSTD_initCStream_usingCDict_advanced(ZSTD_CStream* zcs, + const ZSTD_CDict* cdict, + ZSTD_frameParameters fParams, + unsigned long long pledgedSrcSize) +{ + DEBUGLOG(4, "ZSTD_initCStream_usingCDict_advanced"); + FORWARD_IF_ERROR( ZSTD_CCtx_reset(zcs, ZSTD_reset_session_only) , ""); + FORWARD_IF_ERROR( ZSTD_CCtx_setPledgedSrcSize(zcs, pledgedSrcSize) , ""); + zcs->requestedParams.fParams = fParams; + FORWARD_IF_ERROR( ZSTD_CCtx_refCDict(zcs, cdict) , ""); + return 0; +} + +/* note : cdict must outlive compression session */ +size_t ZSTD_initCStream_usingCDict(ZSTD_CStream* zcs, const ZSTD_CDict* cdict) +{ + DEBUGLOG(4, "ZSTD_initCStream_usingCDict"); + FORWARD_IF_ERROR( ZSTD_CCtx_reset(zcs, ZSTD_reset_session_only) , ""); + FORWARD_IF_ERROR( ZSTD_CCtx_refCDict(zcs, cdict) , ""); + return 0; +} + + +/* ZSTD_initCStream_advanced() : + * pledgedSrcSize must be exact. + * if srcSize is not known at init time, use value ZSTD_CONTENTSIZE_UNKNOWN. + * dict is loaded with default parameters ZSTD_dct_auto and ZSTD_dlm_byCopy. */ +size_t ZSTD_initCStream_advanced(ZSTD_CStream* zcs, + const void* dict, size_t dictSize, + ZSTD_parameters params, unsigned long long pss) +{ + /* for compatibility with older programs relying on this behavior. + * Users should now specify ZSTD_CONTENTSIZE_UNKNOWN. + * This line will be removed in the future. + */ + U64 const pledgedSrcSize = (pss==0 && params.fParams.contentSizeFlag==0) ? ZSTD_CONTENTSIZE_UNKNOWN : pss; + DEBUGLOG(4, "ZSTD_initCStream_advanced"); + FORWARD_IF_ERROR( ZSTD_CCtx_reset(zcs, ZSTD_reset_session_only) , ""); + FORWARD_IF_ERROR( ZSTD_CCtx_setPledgedSrcSize(zcs, pledgedSrcSize) , ""); + FORWARD_IF_ERROR( ZSTD_checkCParams(params.cParams) , ""); + ZSTD_CCtxParams_setZstdParams(&zcs->requestedParams, ¶ms); + FORWARD_IF_ERROR( ZSTD_CCtx_loadDictionary(zcs, dict, dictSize) , ""); + return 0; +} + +size_t ZSTD_initCStream_usingDict(ZSTD_CStream* zcs, const void* dict, size_t dictSize, int compressionLevel) +{ + DEBUGLOG(4, "ZSTD_initCStream_usingDict"); + FORWARD_IF_ERROR( ZSTD_CCtx_reset(zcs, ZSTD_reset_session_only) , ""); + FORWARD_IF_ERROR( ZSTD_CCtx_setParameter(zcs, ZSTD_c_compressionLevel, compressionLevel) , ""); + FORWARD_IF_ERROR( ZSTD_CCtx_loadDictionary(zcs, dict, dictSize) , ""); + return 0; +} + +size_t ZSTD_initCStream_srcSize(ZSTD_CStream* zcs, int compressionLevel, unsigned long long pss) +{ + /* temporary : 0 interpreted as "unknown" during transition period. + * Users willing to specify "unknown" **must** use ZSTD_CONTENTSIZE_UNKNOWN. + * 0 will be interpreted as "empty" in the future. + */ + U64 const pledgedSrcSize = (pss==0) ? ZSTD_CONTENTSIZE_UNKNOWN : pss; + DEBUGLOG(4, "ZSTD_initCStream_srcSize"); + FORWARD_IF_ERROR( ZSTD_CCtx_reset(zcs, ZSTD_reset_session_only) , ""); + FORWARD_IF_ERROR( ZSTD_CCtx_refCDict(zcs, NULL) , ""); + FORWARD_IF_ERROR( ZSTD_CCtx_setParameter(zcs, ZSTD_c_compressionLevel, compressionLevel) , ""); + FORWARD_IF_ERROR( ZSTD_CCtx_setPledgedSrcSize(zcs, pledgedSrcSize) , ""); + return 0; +} + +size_t ZSTD_initCStream(ZSTD_CStream* zcs, int compressionLevel) +{ + DEBUGLOG(4, "ZSTD_initCStream"); + FORWARD_IF_ERROR( ZSTD_CCtx_reset(zcs, ZSTD_reset_session_only) , ""); + FORWARD_IF_ERROR( ZSTD_CCtx_refCDict(zcs, NULL) , ""); + FORWARD_IF_ERROR( ZSTD_CCtx_setParameter(zcs, ZSTD_c_compressionLevel, compressionLevel) , ""); + return 0; +} + +/*====== Compression ======*/ + +static size_t ZSTD_nextInputSizeHint(const ZSTD_CCtx* cctx) +{ + size_t hintInSize = cctx->inBuffTarget - cctx->inBuffPos; + if (hintInSize==0) hintInSize = cctx->blockSize; + return hintInSize; +} + +/* ZSTD_compressStream_generic(): + * internal function for all *compressStream*() variants + * non-static, because can be called from zstdmt_compress.c + * @return : hint size for next input */ +static size_t ZSTD_compressStream_generic(ZSTD_CStream* zcs, + ZSTD_outBuffer* output, + ZSTD_inBuffer* input, + ZSTD_EndDirective const flushMode) +{ + const char* const istart = (const char*)input->src; + const char* const iend = input->size != 0 ? istart + input->size : istart; + const char* ip = input->pos != 0 ? istart + input->pos : istart; + char* const ostart = (char*)output->dst; + char* const oend = output->size != 0 ? ostart + output->size : ostart; + char* op = output->pos != 0 ? ostart + output->pos : ostart; + U32 someMoreWork = 1; + + /* check expectations */ + DEBUGLOG(5, "ZSTD_compressStream_generic, flush=%u", (unsigned)flushMode); + if (zcs->appliedParams.inBufferMode == ZSTD_bm_buffered) { + assert(zcs->inBuff != NULL); + assert(zcs->inBuffSize > 0); + } + if (zcs->appliedParams.outBufferMode == ZSTD_bm_buffered) { + assert(zcs->outBuff != NULL); + assert(zcs->outBuffSize > 0); + } + assert(output->pos <= output->size); + assert(input->pos <= input->size); + assert((U32)flushMode <= (U32)ZSTD_e_end); + + while (someMoreWork) { + switch(zcs->streamStage) + { + case zcss_init: + RETURN_ERROR(init_missing, "call ZSTD_initCStream() first!"); + + case zcss_load: + if ( (flushMode == ZSTD_e_end) + && ( (size_t)(oend-op) >= ZSTD_compressBound(iend-ip) /* Enough output space */ + || zcs->appliedParams.outBufferMode == ZSTD_bm_stable) /* OR we are allowed to return dstSizeTooSmall */ + && (zcs->inBuffPos == 0) ) { + /* shortcut to compression pass directly into output buffer */ + size_t const cSize = ZSTD_compressEnd(zcs, + op, oend-op, ip, iend-ip); + DEBUGLOG(4, "ZSTD_compressEnd : cSize=%u", (unsigned)cSize); + FORWARD_IF_ERROR(cSize, "ZSTD_compressEnd failed"); + ip = iend; + op += cSize; + zcs->frameEnded = 1; + ZSTD_CCtx_reset(zcs, ZSTD_reset_session_only); + someMoreWork = 0; break; + } + /* complete loading into inBuffer in buffered mode */ + if (zcs->appliedParams.inBufferMode == ZSTD_bm_buffered) { + size_t const toLoad = zcs->inBuffTarget - zcs->inBuffPos; + size_t const loaded = ZSTD_limitCopy( + zcs->inBuff + zcs->inBuffPos, toLoad, + ip, iend-ip); + zcs->inBuffPos += loaded; + if (loaded != 0) + ip += loaded; + if ( (flushMode == ZSTD_e_continue) + && (zcs->inBuffPos < zcs->inBuffTarget) ) { + /* not enough input to fill full block : stop here */ + someMoreWork = 0; break; + } + if ( (flushMode == ZSTD_e_flush) + && (zcs->inBuffPos == zcs->inToCompress) ) { + /* empty */ + someMoreWork = 0; break; + } + } + /* compress current block (note : this stage cannot be stopped in the middle) */ + DEBUGLOG(5, "stream compression stage (flushMode==%u)", flushMode); + { int const inputBuffered = (zcs->appliedParams.inBufferMode == ZSTD_bm_buffered); + void* cDst; + size_t cSize; + size_t oSize = oend-op; + size_t const iSize = inputBuffered + ? zcs->inBuffPos - zcs->inToCompress + : MIN((size_t)(iend - ip), zcs->blockSize); + if (oSize >= ZSTD_compressBound(iSize) || zcs->appliedParams.outBufferMode == ZSTD_bm_stable) + cDst = op; /* compress into output buffer, to skip flush stage */ + else + cDst = zcs->outBuff, oSize = zcs->outBuffSize; + if (inputBuffered) { + unsigned const lastBlock = (flushMode == ZSTD_e_end) && (ip==iend); + cSize = lastBlock ? + ZSTD_compressEnd(zcs, cDst, oSize, + zcs->inBuff + zcs->inToCompress, iSize) : + ZSTD_compressContinue(zcs, cDst, oSize, + zcs->inBuff + zcs->inToCompress, iSize); + FORWARD_IF_ERROR(cSize, "%s", lastBlock ? "ZSTD_compressEnd failed" : "ZSTD_compressContinue failed"); + zcs->frameEnded = lastBlock; + /* prepare next block */ + zcs->inBuffTarget = zcs->inBuffPos + zcs->blockSize; + if (zcs->inBuffTarget > zcs->inBuffSize) + zcs->inBuffPos = 0, zcs->inBuffTarget = zcs->blockSize; + DEBUGLOG(5, "inBuffTarget:%u / inBuffSize:%u", + (unsigned)zcs->inBuffTarget, (unsigned)zcs->inBuffSize); + if (!lastBlock) + assert(zcs->inBuffTarget <= zcs->inBuffSize); + zcs->inToCompress = zcs->inBuffPos; + } else { + unsigned const lastBlock = (ip + iSize == iend); + assert(flushMode == ZSTD_e_end /* Already validated */); + cSize = lastBlock ? + ZSTD_compressEnd(zcs, cDst, oSize, ip, iSize) : + ZSTD_compressContinue(zcs, cDst, oSize, ip, iSize); + /* Consume the input prior to error checking to mirror buffered mode. */ + if (iSize > 0) + ip += iSize; + FORWARD_IF_ERROR(cSize, "%s", lastBlock ? "ZSTD_compressEnd failed" : "ZSTD_compressContinue failed"); + zcs->frameEnded = lastBlock; + if (lastBlock) + assert(ip == iend); + } + if (cDst == op) { /* no need to flush */ + op += cSize; + if (zcs->frameEnded) { + DEBUGLOG(5, "Frame completed directly in outBuffer"); + someMoreWork = 0; + ZSTD_CCtx_reset(zcs, ZSTD_reset_session_only); + } + break; + } + zcs->outBuffContentSize = cSize; + zcs->outBuffFlushedSize = 0; + zcs->streamStage = zcss_flush; /* pass-through to flush stage */ + } + ZSTD_FALLTHROUGH; + case zcss_flush: + DEBUGLOG(5, "flush stage"); + assert(zcs->appliedParams.outBufferMode == ZSTD_bm_buffered); + { size_t const toFlush = zcs->outBuffContentSize - zcs->outBuffFlushedSize; + size_t const flushed = ZSTD_limitCopy(op, (size_t)(oend-op), + zcs->outBuff + zcs->outBuffFlushedSize, toFlush); + DEBUGLOG(5, "toFlush: %u into %u ==> flushed: %u", + (unsigned)toFlush, (unsigned)(oend-op), (unsigned)flushed); + if (flushed) + op += flushed; + zcs->outBuffFlushedSize += flushed; + if (toFlush!=flushed) { + /* flush not fully completed, presumably because dst is too small */ + assert(op==oend); + someMoreWork = 0; + break; + } + zcs->outBuffContentSize = zcs->outBuffFlushedSize = 0; + if (zcs->frameEnded) { + DEBUGLOG(5, "Frame completed on flush"); + someMoreWork = 0; + ZSTD_CCtx_reset(zcs, ZSTD_reset_session_only); + break; + } + zcs->streamStage = zcss_load; + break; + } + + default: /* impossible */ + assert(0); + } + } + + input->pos = ip - istart; + output->pos = op - ostart; + if (zcs->frameEnded) return 0; + return ZSTD_nextInputSizeHint(zcs); +} + +static size_t ZSTD_nextInputSizeHint_MTorST(const ZSTD_CCtx* cctx) +{ + return ZSTD_nextInputSizeHint(cctx); + +} + +size_t ZSTD_compressStream(ZSTD_CStream* zcs, ZSTD_outBuffer* output, ZSTD_inBuffer* input) +{ + FORWARD_IF_ERROR( ZSTD_compressStream2(zcs, output, input, ZSTD_e_continue) , ""); + return ZSTD_nextInputSizeHint_MTorST(zcs); +} + +/* After a compression call set the expected input/output buffer. + * This is validated at the start of the next compression call. + */ +static void ZSTD_setBufferExpectations(ZSTD_CCtx* cctx, ZSTD_outBuffer const* output, ZSTD_inBuffer const* input) +{ + if (cctx->appliedParams.inBufferMode == ZSTD_bm_stable) { + cctx->expectedInBuffer = *input; + } + if (cctx->appliedParams.outBufferMode == ZSTD_bm_stable) { + cctx->expectedOutBufferSize = output->size - output->pos; + } +} + +/* Validate that the input/output buffers match the expectations set by + * ZSTD_setBufferExpectations. + */ +static size_t ZSTD_checkBufferStability(ZSTD_CCtx const* cctx, + ZSTD_outBuffer const* output, + ZSTD_inBuffer const* input, + ZSTD_EndDirective endOp) +{ + if (cctx->appliedParams.inBufferMode == ZSTD_bm_stable) { + ZSTD_inBuffer const expect = cctx->expectedInBuffer; + if (expect.src != input->src || expect.pos != input->pos || expect.size != input->size) + RETURN_ERROR(srcBuffer_wrong, "ZSTD_c_stableInBuffer enabled but input differs!"); + if (endOp != ZSTD_e_end) + RETURN_ERROR(srcBuffer_wrong, "ZSTD_c_stableInBuffer can only be used with ZSTD_e_end!"); + } + if (cctx->appliedParams.outBufferMode == ZSTD_bm_stable) { + size_t const outBufferSize = output->size - output->pos; + if (cctx->expectedOutBufferSize != outBufferSize) + RETURN_ERROR(dstBuffer_wrong, "ZSTD_c_stableOutBuffer enabled but output size differs!"); + } + return 0; +} + +static size_t ZSTD_CCtx_init_compressStream2(ZSTD_CCtx* cctx, + ZSTD_EndDirective endOp, + size_t inSize) { + ZSTD_CCtx_params params = cctx->requestedParams; + ZSTD_prefixDict const prefixDict = cctx->prefixDict; + FORWARD_IF_ERROR( ZSTD_initLocalDict(cctx) , ""); /* Init the local dict if present. */ + ZSTD_memset(&cctx->prefixDict, 0, sizeof(cctx->prefixDict)); /* single usage */ + assert(prefixDict.dict==NULL || cctx->cdict==NULL); /* only one can be set */ + if (cctx->cdict) + params.compressionLevel = cctx->cdict->compressionLevel; /* let cdict take priority in terms of compression level */ + DEBUGLOG(4, "ZSTD_compressStream2 : transparent init stage"); + if (endOp == ZSTD_e_end) cctx->pledgedSrcSizePlusOne = inSize + 1; /* auto-fix pledgedSrcSize */ + { + size_t const dictSize = prefixDict.dict + ? prefixDict.dictSize + : (cctx->cdict ? cctx->cdict->dictContentSize : 0); + ZSTD_cParamMode_e const mode = ZSTD_getCParamMode(cctx->cdict, ¶ms, cctx->pledgedSrcSizePlusOne - 1); + params.cParams = ZSTD_getCParamsFromCCtxParams( + ¶ms, cctx->pledgedSrcSizePlusOne-1, + dictSize, mode); + } + + if (ZSTD_CParams_shouldEnableLdm(¶ms.cParams)) { + /* Enable LDM by default for optimal parser and window size >= 128MB */ + DEBUGLOG(4, "LDM enabled by default (window size >= 128MB, strategy >= btopt)"); + params.ldmParams.enableLdm = 1; + } + + { U64 const pledgedSrcSize = cctx->pledgedSrcSizePlusOne - 1; + assert(!ZSTD_isError(ZSTD_checkCParams(params.cParams))); + FORWARD_IF_ERROR( ZSTD_compressBegin_internal(cctx, + prefixDict.dict, prefixDict.dictSize, prefixDict.dictContentType, ZSTD_dtlm_fast, + cctx->cdict, + ¶ms, pledgedSrcSize, + ZSTDb_buffered) , ""); + assert(cctx->appliedParams.nbWorkers == 0); + cctx->inToCompress = 0; + cctx->inBuffPos = 0; + if (cctx->appliedParams.inBufferMode == ZSTD_bm_buffered) { + /* for small input: avoid automatic flush on reaching end of block, since + * it would require to add a 3-bytes null block to end frame + */ + cctx->inBuffTarget = cctx->blockSize + (cctx->blockSize == pledgedSrcSize); + } else { + cctx->inBuffTarget = 0; + } + cctx->outBuffContentSize = cctx->outBuffFlushedSize = 0; + cctx->streamStage = zcss_load; + cctx->frameEnded = 0; + } + return 0; +} + +size_t ZSTD_compressStream2( ZSTD_CCtx* cctx, + ZSTD_outBuffer* output, + ZSTD_inBuffer* input, + ZSTD_EndDirective endOp) +{ + DEBUGLOG(5, "ZSTD_compressStream2, endOp=%u ", (unsigned)endOp); + /* check conditions */ + RETURN_ERROR_IF(output->pos > output->size, dstSize_tooSmall, "invalid output buffer"); + RETURN_ERROR_IF(input->pos > input->size, srcSize_wrong, "invalid input buffer"); + RETURN_ERROR_IF((U32)endOp > (U32)ZSTD_e_end, parameter_outOfBound, "invalid endDirective"); + assert(cctx != NULL); + + /* transparent initialization stage */ + if (cctx->streamStage == zcss_init) { + FORWARD_IF_ERROR(ZSTD_CCtx_init_compressStream2(cctx, endOp, input->size), "CompressStream2 initialization failed"); + ZSTD_setBufferExpectations(cctx, output, input); /* Set initial buffer expectations now that we've initialized */ + } + /* end of transparent initialization stage */ + + FORWARD_IF_ERROR(ZSTD_checkBufferStability(cctx, output, input, endOp), "invalid buffers"); + /* compression stage */ + FORWARD_IF_ERROR( ZSTD_compressStream_generic(cctx, output, input, endOp) , ""); + DEBUGLOG(5, "completed ZSTD_compressStream2"); + ZSTD_setBufferExpectations(cctx, output, input); + return cctx->outBuffContentSize - cctx->outBuffFlushedSize; /* remaining to flush */ +} + +size_t ZSTD_compressStream2_simpleArgs ( + ZSTD_CCtx* cctx, + void* dst, size_t dstCapacity, size_t* dstPos, + const void* src, size_t srcSize, size_t* srcPos, + ZSTD_EndDirective endOp) +{ + ZSTD_outBuffer output = { dst, dstCapacity, *dstPos }; + ZSTD_inBuffer input = { src, srcSize, *srcPos }; + /* ZSTD_compressStream2() will check validity of dstPos and srcPos */ + size_t const cErr = ZSTD_compressStream2(cctx, &output, &input, endOp); + *dstPos = output.pos; + *srcPos = input.pos; + return cErr; +} + +size_t ZSTD_compress2(ZSTD_CCtx* cctx, + void* dst, size_t dstCapacity, + const void* src, size_t srcSize) +{ + ZSTD_bufferMode_e const originalInBufferMode = cctx->requestedParams.inBufferMode; + ZSTD_bufferMode_e const originalOutBufferMode = cctx->requestedParams.outBufferMode; + DEBUGLOG(4, "ZSTD_compress2 (srcSize=%u)", (unsigned)srcSize); + ZSTD_CCtx_reset(cctx, ZSTD_reset_session_only); + /* Enable stable input/output buffers. */ + cctx->requestedParams.inBufferMode = ZSTD_bm_stable; + cctx->requestedParams.outBufferMode = ZSTD_bm_stable; + { size_t oPos = 0; + size_t iPos = 0; + size_t const result = ZSTD_compressStream2_simpleArgs(cctx, + dst, dstCapacity, &oPos, + src, srcSize, &iPos, + ZSTD_e_end); + /* Reset to the original values. */ + cctx->requestedParams.inBufferMode = originalInBufferMode; + cctx->requestedParams.outBufferMode = originalOutBufferMode; + FORWARD_IF_ERROR(result, "ZSTD_compressStream2_simpleArgs failed"); + if (result != 0) { /* compression not completed, due to lack of output space */ + assert(oPos == dstCapacity); + RETURN_ERROR(dstSize_tooSmall, ""); + } + assert(iPos == srcSize); /* all input is expected consumed */ + return oPos; + } +} + +typedef struct { + U32 idx; /* Index in array of ZSTD_Sequence */ + U32 posInSequence; /* Position within sequence at idx */ + size_t posInSrc; /* Number of bytes given by sequences provided so far */ +} ZSTD_sequencePosition; + +/* Returns a ZSTD error code if sequence is not valid */ +static size_t ZSTD_validateSequence(U32 offCode, U32 matchLength, + size_t posInSrc, U32 windowLog, size_t dictSize, U32 minMatch) { + size_t offsetBound; + U32 windowSize = 1 << windowLog; + /* posInSrc represents the amount of data the the decoder would decode up to this point. + * As long as the amount of data decoded is less than or equal to window size, offsets may be + * larger than the total length of output decoded in order to reference the dict, even larger than + * window size. After output surpasses windowSize, we're limited to windowSize offsets again. + */ + offsetBound = posInSrc > windowSize ? (size_t)windowSize : posInSrc + (size_t)dictSize; + RETURN_ERROR_IF(offCode > offsetBound + ZSTD_REP_MOVE, corruption_detected, "Offset too large!"); + RETURN_ERROR_IF(matchLength < minMatch, corruption_detected, "Matchlength too small"); + return 0; +} + +/* Returns an offset code, given a sequence's raw offset, the ongoing repcode array, and whether litLength == 0 */ +static U32 ZSTD_finalizeOffCode(U32 rawOffset, const U32 rep[ZSTD_REP_NUM], U32 ll0) { + U32 offCode = rawOffset + ZSTD_REP_MOVE; + U32 repCode = 0; + + if (!ll0 && rawOffset == rep[0]) { + repCode = 1; + } else if (rawOffset == rep[1]) { + repCode = 2 - ll0; + } else if (rawOffset == rep[2]) { + repCode = 3 - ll0; + } else if (ll0 && rawOffset == rep[0] - 1) { + repCode = 3; + } + if (repCode) { + /* ZSTD_storeSeq expects a number in the range [0, 2] to represent a repcode */ + offCode = repCode - 1; + } + return offCode; +} + +/* Returns 0 on success, and a ZSTD_error otherwise. This function scans through an array of + * ZSTD_Sequence, storing the sequences it finds, until it reaches a block delimiter. + */ +static size_t ZSTD_copySequencesToSeqStoreExplicitBlockDelim(ZSTD_CCtx* cctx, ZSTD_sequencePosition* seqPos, + const ZSTD_Sequence* const inSeqs, size_t inSeqsSize, + const void* src, size_t blockSize) { + U32 idx = seqPos->idx; + BYTE const* ip = (BYTE const*)(src); + const BYTE* const iend = ip + blockSize; + repcodes_t updatedRepcodes; + U32 dictSize; + U32 litLength; + U32 matchLength; + U32 ll0; + U32 offCode; + + if (cctx->cdict) { + dictSize = (U32)cctx->cdict->dictContentSize; + } else if (cctx->prefixDict.dict) { + dictSize = (U32)cctx->prefixDict.dictSize; + } else { + dictSize = 0; + } + ZSTD_memcpy(updatedRepcodes.rep, cctx->blockState.prevCBlock->rep, sizeof(repcodes_t)); + for (; (inSeqs[idx].matchLength != 0 || inSeqs[idx].offset != 0) && idx < inSeqsSize; ++idx) { + litLength = inSeqs[idx].litLength; + matchLength = inSeqs[idx].matchLength; + ll0 = litLength == 0; + offCode = ZSTD_finalizeOffCode(inSeqs[idx].offset, updatedRepcodes.rep, ll0); + updatedRepcodes = ZSTD_updateRep(updatedRepcodes.rep, offCode, ll0); + + DEBUGLOG(6, "Storing sequence: (of: %u, ml: %u, ll: %u)", offCode, matchLength, litLength); + if (cctx->appliedParams.validateSequences) { + seqPos->posInSrc += litLength + matchLength; + FORWARD_IF_ERROR(ZSTD_validateSequence(offCode, matchLength, seqPos->posInSrc, + cctx->appliedParams.cParams.windowLog, dictSize, + cctx->appliedParams.cParams.minMatch), + "Sequence validation failed"); + } + RETURN_ERROR_IF(idx - seqPos->idx > cctx->seqStore.maxNbSeq, memory_allocation, + "Not enough memory allocated. Try adjusting ZSTD_c_minMatch."); + ZSTD_storeSeq(&cctx->seqStore, litLength, ip, iend, offCode, matchLength - MINMATCH); + ip += matchLength + litLength; + } + ZSTD_memcpy(cctx->blockState.nextCBlock->rep, updatedRepcodes.rep, sizeof(repcodes_t)); + + if (inSeqs[idx].litLength) { + DEBUGLOG(6, "Storing last literals of size: %u", inSeqs[idx].litLength); + ZSTD_storeLastLiterals(&cctx->seqStore, ip, inSeqs[idx].litLength); + ip += inSeqs[idx].litLength; + seqPos->posInSrc += inSeqs[idx].litLength; + } + RETURN_ERROR_IF(ip != iend, corruption_detected, "Blocksize doesn't agree with block delimiter!"); + seqPos->idx = idx+1; + return 0; +} + +/* Returns the number of bytes to move the current read position back by. Only non-zero + * if we ended up splitting a sequence. Otherwise, it may return a ZSTD error if something + * went wrong. + * + * This function will attempt to scan through blockSize bytes represented by the sequences + * in inSeqs, storing any (partial) sequences. + * + * Occasionally, we may want to change the actual number of bytes we consumed from inSeqs to + * avoid splitting a match, or to avoid splitting a match such that it would produce a match + * smaller than MINMATCH. In this case, we return the number of bytes that we didn't read from this block. + */ +static size_t ZSTD_copySequencesToSeqStoreNoBlockDelim(ZSTD_CCtx* cctx, ZSTD_sequencePosition* seqPos, + const ZSTD_Sequence* const inSeqs, size_t inSeqsSize, + const void* src, size_t blockSize) { + U32 idx = seqPos->idx; + U32 startPosInSequence = seqPos->posInSequence; + U32 endPosInSequence = seqPos->posInSequence + (U32)blockSize; + size_t dictSize; + BYTE const* ip = (BYTE const*)(src); + BYTE const* iend = ip + blockSize; /* May be adjusted if we decide to process fewer than blockSize bytes */ + repcodes_t updatedRepcodes; + U32 bytesAdjustment = 0; + U32 finalMatchSplit = 0; + U32 litLength; + U32 matchLength; + U32 rawOffset; + U32 offCode; + + if (cctx->cdict) { + dictSize = cctx->cdict->dictContentSize; + } else if (cctx->prefixDict.dict) { + dictSize = cctx->prefixDict.dictSize; + } else { + dictSize = 0; + } + DEBUGLOG(5, "ZSTD_copySequencesToSeqStore: idx: %u PIS: %u blockSize: %zu", idx, startPosInSequence, blockSize); + DEBUGLOG(5, "Start seq: idx: %u (of: %u ml: %u ll: %u)", idx, inSeqs[idx].offset, inSeqs[idx].matchLength, inSeqs[idx].litLength); + ZSTD_memcpy(updatedRepcodes.rep, cctx->blockState.prevCBlock->rep, sizeof(repcodes_t)); + while (endPosInSequence && idx < inSeqsSize && !finalMatchSplit) { + const ZSTD_Sequence currSeq = inSeqs[idx]; + litLength = currSeq.litLength; + matchLength = currSeq.matchLength; + rawOffset = currSeq.offset; + + /* Modify the sequence depending on where endPosInSequence lies */ + if (endPosInSequence >= currSeq.litLength + currSeq.matchLength) { + if (startPosInSequence >= litLength) { + startPosInSequence -= litLength; + litLength = 0; + matchLength -= startPosInSequence; + } else { + litLength -= startPosInSequence; + } + /* Move to the next sequence */ + endPosInSequence -= currSeq.litLength + currSeq.matchLength; + startPosInSequence = 0; + idx++; + } else { + /* This is the final (partial) sequence we're adding from inSeqs, and endPosInSequence + does not reach the end of the match. So, we have to split the sequence */ + DEBUGLOG(6, "Require a split: diff: %u, idx: %u PIS: %u", + currSeq.litLength + currSeq.matchLength - endPosInSequence, idx, endPosInSequence); + if (endPosInSequence > litLength) { + U32 firstHalfMatchLength; + litLength = startPosInSequence >= litLength ? 0 : litLength - startPosInSequence; + firstHalfMatchLength = endPosInSequence - startPosInSequence - litLength; + if (matchLength > blockSize && firstHalfMatchLength >= cctx->appliedParams.cParams.minMatch) { + /* Only ever split the match if it is larger than the block size */ + U32 secondHalfMatchLength = currSeq.matchLength + currSeq.litLength - endPosInSequence; + if (secondHalfMatchLength < cctx->appliedParams.cParams.minMatch) { + /* Move the endPosInSequence backward so that it creates match of minMatch length */ + endPosInSequence -= cctx->appliedParams.cParams.minMatch - secondHalfMatchLength; + bytesAdjustment = cctx->appliedParams.cParams.minMatch - secondHalfMatchLength; + firstHalfMatchLength -= bytesAdjustment; + } + matchLength = firstHalfMatchLength; + /* Flag that we split the last match - after storing the sequence, exit the loop, + but keep the value of endPosInSequence */ + finalMatchSplit = 1; + } else { + /* Move the position in sequence backwards so that we don't split match, and break to store + * the last literals. We use the original currSeq.litLength as a marker for where endPosInSequence + * should go. We prefer to do this whenever it is not necessary to split the match, or if doing so + * would cause the first half of the match to be too small + */ + bytesAdjustment = endPosInSequence - currSeq.litLength; + endPosInSequence = currSeq.litLength; + break; + } + } else { + /* This sequence ends inside the literals, break to store the last literals */ + break; + } + } + /* Check if this offset can be represented with a repcode */ + { U32 ll0 = (litLength == 0); + offCode = ZSTD_finalizeOffCode(rawOffset, updatedRepcodes.rep, ll0); + updatedRepcodes = ZSTD_updateRep(updatedRepcodes.rep, offCode, ll0); + } + + if (cctx->appliedParams.validateSequences) { + seqPos->posInSrc += litLength + matchLength; + FORWARD_IF_ERROR(ZSTD_validateSequence(offCode, matchLength, seqPos->posInSrc, + cctx->appliedParams.cParams.windowLog, dictSize, + cctx->appliedParams.cParams.minMatch), + "Sequence validation failed"); + } + DEBUGLOG(6, "Storing sequence: (of: %u, ml: %u, ll: %u)", offCode, matchLength, litLength); + RETURN_ERROR_IF(idx - seqPos->idx > cctx->seqStore.maxNbSeq, memory_allocation, + "Not enough memory allocated. Try adjusting ZSTD_c_minMatch."); + ZSTD_storeSeq(&cctx->seqStore, litLength, ip, iend, offCode, matchLength - MINMATCH); + ip += matchLength + litLength; + } + DEBUGLOG(5, "Ending seq: idx: %u (of: %u ml: %u ll: %u)", idx, inSeqs[idx].offset, inSeqs[idx].matchLength, inSeqs[idx].litLength); + assert(idx == inSeqsSize || endPosInSequence <= inSeqs[idx].litLength + inSeqs[idx].matchLength); + seqPos->idx = idx; + seqPos->posInSequence = endPosInSequence; + ZSTD_memcpy(cctx->blockState.nextCBlock->rep, updatedRepcodes.rep, sizeof(repcodes_t)); + + iend -= bytesAdjustment; + if (ip != iend) { + /* Store any last literals */ + U32 lastLLSize = (U32)(iend - ip); + assert(ip <= iend); + DEBUGLOG(6, "Storing last literals of size: %u", lastLLSize); + ZSTD_storeLastLiterals(&cctx->seqStore, ip, lastLLSize); + seqPos->posInSrc += lastLLSize; + } + + return bytesAdjustment; +} + +typedef size_t (*ZSTD_sequenceCopier) (ZSTD_CCtx* cctx, ZSTD_sequencePosition* seqPos, + const ZSTD_Sequence* const inSeqs, size_t inSeqsSize, + const void* src, size_t blockSize); +static ZSTD_sequenceCopier ZSTD_selectSequenceCopier(ZSTD_sequenceFormat_e mode) { + ZSTD_sequenceCopier sequenceCopier = NULL; + assert(ZSTD_cParam_withinBounds(ZSTD_c_blockDelimiters, mode)); + if (mode == ZSTD_sf_explicitBlockDelimiters) { + return ZSTD_copySequencesToSeqStoreExplicitBlockDelim; + } else if (mode == ZSTD_sf_noBlockDelimiters) { + return ZSTD_copySequencesToSeqStoreNoBlockDelim; + } + assert(sequenceCopier != NULL); + return sequenceCopier; +} + +/* Compress, block-by-block, all of the sequences given. + * + * Returns the cumulative size of all compressed blocks (including their headers), otherwise a ZSTD error. + */ +static size_t ZSTD_compressSequences_internal(ZSTD_CCtx* cctx, + void* dst, size_t dstCapacity, + const ZSTD_Sequence* inSeqs, size_t inSeqsSize, + const void* src, size_t srcSize) { + size_t cSize = 0; + U32 lastBlock; + size_t blockSize; + size_t compressedSeqsSize; + size_t remaining = srcSize; + ZSTD_sequencePosition seqPos = {0, 0, 0}; + + BYTE const* ip = (BYTE const*)src; + BYTE* op = (BYTE*)dst; + ZSTD_sequenceCopier sequenceCopier = ZSTD_selectSequenceCopier(cctx->appliedParams.blockDelimiters); + + DEBUGLOG(4, "ZSTD_compressSequences_internal srcSize: %zu, inSeqsSize: %zu", srcSize, inSeqsSize); + /* Special case: empty frame */ + if (remaining == 0) { + U32 const cBlockHeader24 = 1 /* last block */ + (((U32)bt_raw)<<1); + RETURN_ERROR_IF(dstCapacity<4, dstSize_tooSmall, "No room for empty frame block header"); + MEM_writeLE32(op, cBlockHeader24); + op += ZSTD_blockHeaderSize; + dstCapacity -= ZSTD_blockHeaderSize; + cSize += ZSTD_blockHeaderSize; + } + + while (remaining) { + size_t cBlockSize; + size_t additionalByteAdjustment; + lastBlock = remaining <= cctx->blockSize; + blockSize = lastBlock ? (U32)remaining : (U32)cctx->blockSize; + ZSTD_resetSeqStore(&cctx->seqStore); + DEBUGLOG(4, "Working on new block. Blocksize: %zu", blockSize); + + additionalByteAdjustment = sequenceCopier(cctx, &seqPos, inSeqs, inSeqsSize, ip, blockSize); + FORWARD_IF_ERROR(additionalByteAdjustment, "Bad sequence copy"); + blockSize -= additionalByteAdjustment; + + /* If blocks are too small, emit as a nocompress block */ + if (blockSize < MIN_CBLOCK_SIZE+ZSTD_blockHeaderSize+1) { + cBlockSize = ZSTD_noCompressBlock(op, dstCapacity, ip, blockSize, lastBlock); + FORWARD_IF_ERROR(cBlockSize, "Nocompress block failed"); + DEBUGLOG(4, "Block too small, writing out nocompress block: cSize: %zu", cBlockSize); + cSize += cBlockSize; + ip += blockSize; + op += cBlockSize; + remaining -= blockSize; + dstCapacity -= cBlockSize; + continue; + } + + compressedSeqsSize = ZSTD_entropyCompressSequences(&cctx->seqStore, + &cctx->blockState.prevCBlock->entropy, &cctx->blockState.nextCBlock->entropy, + &cctx->appliedParams, + op + ZSTD_blockHeaderSize /* Leave space for block header */, dstCapacity - ZSTD_blockHeaderSize, + blockSize, + cctx->entropyWorkspace, ENTROPY_WORKSPACE_SIZE /* statically allocated in resetCCtx */, + cctx->bmi2); + FORWARD_IF_ERROR(compressedSeqsSize, "Compressing sequences of block failed"); + DEBUGLOG(4, "Compressed sequences size: %zu", compressedSeqsSize); + + if (!cctx->isFirstBlock && + ZSTD_maybeRLE(&cctx->seqStore) && + ZSTD_isRLE((BYTE const*)src, srcSize)) { + /* We don't want to emit our first block as a RLE even if it qualifies because + * doing so will cause the decoder (cli only) to throw a "should consume all input error." + * This is only an issue for zstd <= v1.4.3 + */ + compressedSeqsSize = 1; + } + + if (compressedSeqsSize == 0) { + /* ZSTD_noCompressBlock writes the block header as well */ + cBlockSize = ZSTD_noCompressBlock(op, dstCapacity, ip, blockSize, lastBlock); + FORWARD_IF_ERROR(cBlockSize, "Nocompress block failed"); + DEBUGLOG(4, "Writing out nocompress block, size: %zu", cBlockSize); + } else if (compressedSeqsSize == 1) { + cBlockSize = ZSTD_rleCompressBlock(op, dstCapacity, *ip, blockSize, lastBlock); + FORWARD_IF_ERROR(cBlockSize, "RLE compress block failed"); + DEBUGLOG(4, "Writing out RLE block, size: %zu", cBlockSize); + } else { + U32 cBlockHeader; + /* Error checking and repcodes update */ + ZSTD_confirmRepcodesAndEntropyTables(cctx); + if (cctx->blockState.prevCBlock->entropy.fse.offcode_repeatMode == FSE_repeat_valid) + cctx->blockState.prevCBlock->entropy.fse.offcode_repeatMode = FSE_repeat_check; + + /* Write block header into beginning of block*/ + cBlockHeader = lastBlock + (((U32)bt_compressed)<<1) + (U32)(compressedSeqsSize << 3); + MEM_writeLE24(op, cBlockHeader); + cBlockSize = ZSTD_blockHeaderSize + compressedSeqsSize; + DEBUGLOG(4, "Writing out compressed block, size: %zu", cBlockSize); + } + + cSize += cBlockSize; + DEBUGLOG(4, "cSize running total: %zu", cSize); + + if (lastBlock) { + break; + } else { + ip += blockSize; + op += cBlockSize; + remaining -= blockSize; + dstCapacity -= cBlockSize; + cctx->isFirstBlock = 0; + } + } + + return cSize; +} + +size_t ZSTD_compressSequences(ZSTD_CCtx* const cctx, void* dst, size_t dstCapacity, + const ZSTD_Sequence* inSeqs, size_t inSeqsSize, + const void* src, size_t srcSize) { + BYTE* op = (BYTE*)dst; + size_t cSize = 0; + size_t compressedBlocksSize = 0; + size_t frameHeaderSize = 0; + + /* Transparent initialization stage, same as compressStream2() */ + DEBUGLOG(3, "ZSTD_compressSequences()"); + assert(cctx != NULL); + FORWARD_IF_ERROR(ZSTD_CCtx_init_compressStream2(cctx, ZSTD_e_end, srcSize), "CCtx initialization failed"); + /* Begin writing output, starting with frame header */ + frameHeaderSize = ZSTD_writeFrameHeader(op, dstCapacity, &cctx->appliedParams, srcSize, cctx->dictID); + op += frameHeaderSize; + dstCapacity -= frameHeaderSize; + cSize += frameHeaderSize; + if (cctx->appliedParams.fParams.checksumFlag && srcSize) { + xxh64_update(&cctx->xxhState, src, srcSize); + } + /* cSize includes block header size and compressed sequences size */ + compressedBlocksSize = ZSTD_compressSequences_internal(cctx, + op, dstCapacity, + inSeqs, inSeqsSize, + src, srcSize); + FORWARD_IF_ERROR(compressedBlocksSize, "Compressing blocks failed!"); + cSize += compressedBlocksSize; + dstCapacity -= compressedBlocksSize; + + if (cctx->appliedParams.fParams.checksumFlag) { + U32 const checksum = (U32) xxh64_digest(&cctx->xxhState); + RETURN_ERROR_IF(dstCapacity<4, dstSize_tooSmall, "no room for checksum"); + DEBUGLOG(4, "Write checksum : %08X", (unsigned)checksum); + MEM_writeLE32((char*)dst + cSize, checksum); + cSize += 4; + } + + DEBUGLOG(3, "Final compressed size: %zu", cSize); + return cSize; +} + +/*====== Finalize ======*/ + +/*! ZSTD_flushStream() : + * @return : amount of data remaining to flush */ +size_t ZSTD_flushStream(ZSTD_CStream* zcs, ZSTD_outBuffer* output) +{ + ZSTD_inBuffer input = { NULL, 0, 0 }; + return ZSTD_compressStream2(zcs, output, &input, ZSTD_e_flush); +} + + +size_t ZSTD_endStream(ZSTD_CStream* zcs, ZSTD_outBuffer* output) +{ + ZSTD_inBuffer input = { NULL, 0, 0 }; + size_t const remainingToFlush = ZSTD_compressStream2(zcs, output, &input, ZSTD_e_end); + FORWARD_IF_ERROR( remainingToFlush , "ZSTD_compressStream2 failed"); + if (zcs->appliedParams.nbWorkers > 0) return remainingToFlush; /* minimal estimation */ + /* single thread mode : attempt to calculate remaining to flush more precisely */ + { size_t const lastBlockSize = zcs->frameEnded ? 0 : ZSTD_BLOCKHEADERSIZE; + size_t const checksumSize = (size_t)(zcs->frameEnded ? 0 : zcs->appliedParams.fParams.checksumFlag * 4); + size_t const toFlush = remainingToFlush + lastBlockSize + checksumSize; + DEBUGLOG(4, "ZSTD_endStream : remaining to flush : %u", (unsigned)toFlush); + return toFlush; + } +} + + +/*-===== Pre-defined compression levels =====-*/ + +#define ZSTD_MAX_CLEVEL 22 +int ZSTD_maxCLevel(void) { return ZSTD_MAX_CLEVEL; } +int ZSTD_minCLevel(void) { return (int)-ZSTD_TARGETLENGTH_MAX; } + +static const ZSTD_compressionParameters ZSTD_defaultCParameters[4][ZSTD_MAX_CLEVEL+1] = { +{ /* "default" - for any srcSize > 256 KB */ + /* W, C, H, S, L, TL, strat */ + { 19, 12, 13, 1, 6, 1, ZSTD_fast }, /* base for negative levels */ + { 19, 13, 14, 1, 7, 0, ZSTD_fast }, /* level 1 */ + { 20, 15, 16, 1, 6, 0, ZSTD_fast }, /* level 2 */ + { 21, 16, 17, 1, 5, 0, ZSTD_dfast }, /* level 3 */ + { 21, 18, 18, 1, 5, 0, ZSTD_dfast }, /* level 4 */ + { 21, 18, 19, 2, 5, 2, ZSTD_greedy }, /* level 5 */ + { 21, 19, 19, 3, 5, 4, ZSTD_greedy }, /* level 6 */ + { 21, 19, 19, 3, 5, 8, ZSTD_lazy }, /* level 7 */ + { 21, 19, 19, 3, 5, 16, ZSTD_lazy2 }, /* level 8 */ + { 21, 19, 20, 4, 5, 16, ZSTD_lazy2 }, /* level 9 */ + { 22, 20, 21, 4, 5, 16, ZSTD_lazy2 }, /* level 10 */ + { 22, 21, 22, 4, 5, 16, ZSTD_lazy2 }, /* level 11 */ + { 22, 21, 22, 5, 5, 16, ZSTD_lazy2 }, /* level 12 */ + { 22, 21, 22, 5, 5, 32, ZSTD_btlazy2 }, /* level 13 */ + { 22, 22, 23, 5, 5, 32, ZSTD_btlazy2 }, /* level 14 */ + { 22, 23, 23, 6, 5, 32, ZSTD_btlazy2 }, /* level 15 */ + { 22, 22, 22, 5, 5, 48, ZSTD_btopt }, /* level 16 */ + { 23, 23, 22, 5, 4, 64, ZSTD_btopt }, /* level 17 */ + { 23, 23, 22, 6, 3, 64, ZSTD_btultra }, /* level 18 */ + { 23, 24, 22, 7, 3,256, ZSTD_btultra2}, /* level 19 */ + { 25, 25, 23, 7, 3,256, ZSTD_btultra2}, /* level 20 */ + { 26, 26, 24, 7, 3,512, ZSTD_btultra2}, /* level 21 */ + { 27, 27, 25, 9, 3,999, ZSTD_btultra2}, /* level 22 */ +}, +{ /* for srcSize <= 256 KB */ + /* W, C, H, S, L, T, strat */ + { 18, 12, 13, 1, 5, 1, ZSTD_fast }, /* base for negative levels */ + { 18, 13, 14, 1, 6, 0, ZSTD_fast }, /* level 1 */ + { 18, 14, 14, 1, 5, 0, ZSTD_dfast }, /* level 2 */ + { 18, 16, 16, 1, 4, 0, ZSTD_dfast }, /* level 3 */ + { 18, 16, 17, 2, 5, 2, ZSTD_greedy }, /* level 4.*/ + { 18, 18, 18, 3, 5, 2, ZSTD_greedy }, /* level 5.*/ + { 18, 18, 19, 3, 5, 4, ZSTD_lazy }, /* level 6.*/ + { 18, 18, 19, 4, 4, 4, ZSTD_lazy }, /* level 7 */ + { 18, 18, 19, 4, 4, 8, ZSTD_lazy2 }, /* level 8 */ + { 18, 18, 19, 5, 4, 8, ZSTD_lazy2 }, /* level 9 */ + { 18, 18, 19, 6, 4, 8, ZSTD_lazy2 }, /* level 10 */ + { 18, 18, 19, 5, 4, 12, ZSTD_btlazy2 }, /* level 11.*/ + { 18, 19, 19, 7, 4, 12, ZSTD_btlazy2 }, /* level 12.*/ + { 18, 18, 19, 4, 4, 16, ZSTD_btopt }, /* level 13 */ + { 18, 18, 19, 4, 3, 32, ZSTD_btopt }, /* level 14.*/ + { 18, 18, 19, 6, 3,128, ZSTD_btopt }, /* level 15.*/ + { 18, 19, 19, 6, 3,128, ZSTD_btultra }, /* level 16.*/ + { 18, 19, 19, 8, 3,256, ZSTD_btultra }, /* level 17.*/ + { 18, 19, 19, 6, 3,128, ZSTD_btultra2}, /* level 18.*/ + { 18, 19, 19, 8, 3,256, ZSTD_btultra2}, /* level 19.*/ + { 18, 19, 19, 10, 3,512, ZSTD_btultra2}, /* level 20.*/ + { 18, 19, 19, 12, 3,512, ZSTD_btultra2}, /* level 21.*/ + { 18, 19, 19, 13, 3,999, ZSTD_btultra2}, /* level 22.*/ +}, +{ /* for srcSize <= 128 KB */ + /* W, C, H, S, L, T, strat */ + { 17, 12, 12, 1, 5, 1, ZSTD_fast }, /* base for negative levels */ + { 17, 12, 13, 1, 6, 0, ZSTD_fast }, /* level 1 */ + { 17, 13, 15, 1, 5, 0, ZSTD_fast }, /* level 2 */ + { 17, 15, 16, 2, 5, 0, ZSTD_dfast }, /* level 3 */ + { 17, 17, 17, 2, 4, 0, ZSTD_dfast }, /* level 4 */ + { 17, 16, 17, 3, 4, 2, ZSTD_greedy }, /* level 5 */ + { 17, 17, 17, 3, 4, 4, ZSTD_lazy }, /* level 6 */ + { 17, 17, 17, 3, 4, 8, ZSTD_lazy2 }, /* level 7 */ + { 17, 17, 17, 4, 4, 8, ZSTD_lazy2 }, /* level 8 */ + { 17, 17, 17, 5, 4, 8, ZSTD_lazy2 }, /* level 9 */ + { 17, 17, 17, 6, 4, 8, ZSTD_lazy2 }, /* level 10 */ + { 17, 17, 17, 5, 4, 8, ZSTD_btlazy2 }, /* level 11 */ + { 17, 18, 17, 7, 4, 12, ZSTD_btlazy2 }, /* level 12 */ + { 17, 18, 17, 3, 4, 12, ZSTD_btopt }, /* level 13.*/ + { 17, 18, 17, 4, 3, 32, ZSTD_btopt }, /* level 14.*/ + { 17, 18, 17, 6, 3,256, ZSTD_btopt }, /* level 15.*/ + { 17, 18, 17, 6, 3,128, ZSTD_btultra }, /* level 16.*/ + { 17, 18, 17, 8, 3,256, ZSTD_btultra }, /* level 17.*/ + { 17, 18, 17, 10, 3,512, ZSTD_btultra }, /* level 18.*/ + { 17, 18, 17, 5, 3,256, ZSTD_btultra2}, /* level 19.*/ + { 17, 18, 17, 7, 3,512, ZSTD_btultra2}, /* level 20.*/ + { 17, 18, 17, 9, 3,512, ZSTD_btultra2}, /* level 21.*/ + { 17, 18, 17, 11, 3,999, ZSTD_btultra2}, /* level 22.*/ +}, +{ /* for srcSize <= 16 KB */ + /* W, C, H, S, L, T, strat */ + { 14, 12, 13, 1, 5, 1, ZSTD_fast }, /* base for negative levels */ + { 14, 14, 15, 1, 5, 0, ZSTD_fast }, /* level 1 */ + { 14, 14, 15, 1, 4, 0, ZSTD_fast }, /* level 2 */ + { 14, 14, 15, 2, 4, 0, ZSTD_dfast }, /* level 3 */ + { 14, 14, 14, 4, 4, 2, ZSTD_greedy }, /* level 4 */ + { 14, 14, 14, 3, 4, 4, ZSTD_lazy }, /* level 5.*/ + { 14, 14, 14, 4, 4, 8, ZSTD_lazy2 }, /* level 6 */ + { 14, 14, 14, 6, 4, 8, ZSTD_lazy2 }, /* level 7 */ + { 14, 14, 14, 8, 4, 8, ZSTD_lazy2 }, /* level 8.*/ + { 14, 15, 14, 5, 4, 8, ZSTD_btlazy2 }, /* level 9.*/ + { 14, 15, 14, 9, 4, 8, ZSTD_btlazy2 }, /* level 10.*/ + { 14, 15, 14, 3, 4, 12, ZSTD_btopt }, /* level 11.*/ + { 14, 15, 14, 4, 3, 24, ZSTD_btopt }, /* level 12.*/ + { 14, 15, 14, 5, 3, 32, ZSTD_btultra }, /* level 13.*/ + { 14, 15, 15, 6, 3, 64, ZSTD_btultra }, /* level 14.*/ + { 14, 15, 15, 7, 3,256, ZSTD_btultra }, /* level 15.*/ + { 14, 15, 15, 5, 3, 48, ZSTD_btultra2}, /* level 16.*/ + { 14, 15, 15, 6, 3,128, ZSTD_btultra2}, /* level 17.*/ + { 14, 15, 15, 7, 3,256, ZSTD_btultra2}, /* level 18.*/ + { 14, 15, 15, 8, 3,256, ZSTD_btultra2}, /* level 19.*/ + { 14, 15, 15, 8, 3,512, ZSTD_btultra2}, /* level 20.*/ + { 14, 15, 15, 9, 3,512, ZSTD_btultra2}, /* level 21.*/ + { 14, 15, 15, 10, 3,999, ZSTD_btultra2}, /* level 22.*/ +}, +}; + +static ZSTD_compressionParameters ZSTD_dedicatedDictSearch_getCParams(int const compressionLevel, size_t const dictSize) +{ + ZSTD_compressionParameters cParams = ZSTD_getCParams_internal(compressionLevel, 0, dictSize, ZSTD_cpm_createCDict); + switch (cParams.strategy) { + case ZSTD_fast: + case ZSTD_dfast: + break; + case ZSTD_greedy: + case ZSTD_lazy: + case ZSTD_lazy2: + cParams.hashLog += ZSTD_LAZY_DDSS_BUCKET_LOG; + break; + case ZSTD_btlazy2: + case ZSTD_btopt: + case ZSTD_btultra: + case ZSTD_btultra2: + break; + } + return cParams; +} + +static int ZSTD_dedicatedDictSearch_isSupported( + ZSTD_compressionParameters const* cParams) +{ + return (cParams->strategy >= ZSTD_greedy) + && (cParams->strategy <= ZSTD_lazy2) + && (cParams->hashLog >= cParams->chainLog) + && (cParams->chainLog <= 24); +} + +/* + * Reverses the adjustment applied to cparams when enabling dedicated dict + * search. This is used to recover the params set to be used in the working + * context. (Otherwise, those tables would also grow.) + */ +static void ZSTD_dedicatedDictSearch_revertCParams( + ZSTD_compressionParameters* cParams) { + switch (cParams->strategy) { + case ZSTD_fast: + case ZSTD_dfast: + break; + case ZSTD_greedy: + case ZSTD_lazy: + case ZSTD_lazy2: + cParams->hashLog -= ZSTD_LAZY_DDSS_BUCKET_LOG; + break; + case ZSTD_btlazy2: + case ZSTD_btopt: + case ZSTD_btultra: + case ZSTD_btultra2: + break; + } +} + +static U64 ZSTD_getCParamRowSize(U64 srcSizeHint, size_t dictSize, ZSTD_cParamMode_e mode) +{ + switch (mode) { + case ZSTD_cpm_unknown: + case ZSTD_cpm_noAttachDict: + case ZSTD_cpm_createCDict: + break; + case ZSTD_cpm_attachDict: + dictSize = 0; + break; + default: + assert(0); + break; + } + { int const unknown = srcSizeHint == ZSTD_CONTENTSIZE_UNKNOWN; + size_t const addedSize = unknown && dictSize > 0 ? 500 : 0; + return unknown && dictSize == 0 ? ZSTD_CONTENTSIZE_UNKNOWN : srcSizeHint+dictSize+addedSize; + } +} + +/*! ZSTD_getCParams_internal() : + * @return ZSTD_compressionParameters structure for a selected compression level, srcSize and dictSize. + * Note: srcSizeHint 0 means 0, use ZSTD_CONTENTSIZE_UNKNOWN for unknown. + * Use dictSize == 0 for unknown or unused. + * Note: `mode` controls how we treat the `dictSize`. See docs for `ZSTD_cParamMode_e`. */ +static ZSTD_compressionParameters ZSTD_getCParams_internal(int compressionLevel, unsigned long long srcSizeHint, size_t dictSize, ZSTD_cParamMode_e mode) +{ + U64 const rSize = ZSTD_getCParamRowSize(srcSizeHint, dictSize, mode); + U32 const tableID = (rSize <= 256 KB) + (rSize <= 128 KB) + (rSize <= 16 KB); + int row; + DEBUGLOG(5, "ZSTD_getCParams_internal (cLevel=%i)", compressionLevel); + + /* row */ + if (compressionLevel == 0) row = ZSTD_CLEVEL_DEFAULT; /* 0 == default */ + else if (compressionLevel < 0) row = 0; /* entry 0 is baseline for fast mode */ + else if (compressionLevel > ZSTD_MAX_CLEVEL) row = ZSTD_MAX_CLEVEL; + else row = compressionLevel; + + { ZSTD_compressionParameters cp = ZSTD_defaultCParameters[tableID][row]; + /* acceleration factor */ + if (compressionLevel < 0) { + int const clampedCompressionLevel = MAX(ZSTD_minCLevel(), compressionLevel); + cp.targetLength = (unsigned)(-clampedCompressionLevel); + } + /* refine parameters based on srcSize & dictSize */ + return ZSTD_adjustCParams_internal(cp, srcSizeHint, dictSize, mode); + } +} + +/*! ZSTD_getCParams() : + * @return ZSTD_compressionParameters structure for a selected compression level, srcSize and dictSize. + * Size values are optional, provide 0 if not known or unused */ +ZSTD_compressionParameters ZSTD_getCParams(int compressionLevel, unsigned long long srcSizeHint, size_t dictSize) +{ + if (srcSizeHint == 0) srcSizeHint = ZSTD_CONTENTSIZE_UNKNOWN; + return ZSTD_getCParams_internal(compressionLevel, srcSizeHint, dictSize, ZSTD_cpm_unknown); +} + +/*! ZSTD_getParams() : + * same idea as ZSTD_getCParams() + * @return a `ZSTD_parameters` structure (instead of `ZSTD_compressionParameters`). + * Fields of `ZSTD_frameParameters` are set to default values */ +static ZSTD_parameters ZSTD_getParams_internal(int compressionLevel, unsigned long long srcSizeHint, size_t dictSize, ZSTD_cParamMode_e mode) { + ZSTD_parameters params; + ZSTD_compressionParameters const cParams = ZSTD_getCParams_internal(compressionLevel, srcSizeHint, dictSize, mode); + DEBUGLOG(5, "ZSTD_getParams (cLevel=%i)", compressionLevel); + ZSTD_memset(¶ms, 0, sizeof(params)); + params.cParams = cParams; + params.fParams.contentSizeFlag = 1; + return params; +} + +/*! ZSTD_getParams() : + * same idea as ZSTD_getCParams() + * @return a `ZSTD_parameters` structure (instead of `ZSTD_compressionParameters`). + * Fields of `ZSTD_frameParameters` are set to default values */ +ZSTD_parameters ZSTD_getParams(int compressionLevel, unsigned long long srcSizeHint, size_t dictSize) { + if (srcSizeHint == 0) srcSizeHint = ZSTD_CONTENTSIZE_UNKNOWN; + return ZSTD_getParams_internal(compressionLevel, srcSizeHint, dictSize, ZSTD_cpm_unknown); +} diff --git a/lib/zstd/compress/zstd_compress_internal.h b/lib/zstd/compress/zstd_compress_internal.h new file mode 100644 index 000000000000..685d2f996cc2 --- /dev/null +++ b/lib/zstd/compress/zstd_compress_internal.h @@ -0,0 +1,1188 @@ +/* + * Copyright (c) Yann Collet, Facebook, Inc. + * All rights reserved. + * + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). + * You may select, at your option, one of the above-listed licenses. + */ + +/* This header contains definitions + * that shall **only** be used by modules within lib/compress. + */ + +#ifndef ZSTD_COMPRESS_H +#define ZSTD_COMPRESS_H + +/*-************************************* +* Dependencies +***************************************/ +#include "../common/zstd_internal.h" +#include "zstd_cwksp.h" + + +/*-************************************* +* Constants +***************************************/ +#define kSearchStrength 8 +#define HASH_READ_SIZE 8 +#define ZSTD_DUBT_UNSORTED_MARK 1 /* For btlazy2 strategy, index ZSTD_DUBT_UNSORTED_MARK==1 means "unsorted". + It could be confused for a real successor at index "1", if sorted as larger than its predecessor. + It's not a big deal though : candidate will just be sorted again. + Additionally, candidate position 1 will be lost. + But candidate 1 cannot hide a large tree of candidates, so it's a minimal loss. + The benefit is that ZSTD_DUBT_UNSORTED_MARK cannot be mishandled after table re-use with a different strategy. + This constant is required by ZSTD_compressBlock_btlazy2() and ZSTD_reduceTable_internal() */ + + +/*-************************************* +* Context memory management +***************************************/ +typedef enum { ZSTDcs_created=0, ZSTDcs_init, ZSTDcs_ongoing, ZSTDcs_ending } ZSTD_compressionStage_e; +typedef enum { zcss_init=0, zcss_load, zcss_flush } ZSTD_cStreamStage; + +typedef struct ZSTD_prefixDict_s { + const void* dict; + size_t dictSize; + ZSTD_dictContentType_e dictContentType; +} ZSTD_prefixDict; + +typedef struct { + void* dictBuffer; + void const* dict; + size_t dictSize; + ZSTD_dictContentType_e dictContentType; + ZSTD_CDict* cdict; +} ZSTD_localDict; + +typedef struct { + HUF_CElt CTable[HUF_CTABLE_SIZE_U32(255)]; + HUF_repeat repeatMode; +} ZSTD_hufCTables_t; + +typedef struct { + FSE_CTable offcodeCTable[FSE_CTABLE_SIZE_U32(OffFSELog, MaxOff)]; + FSE_CTable matchlengthCTable[FSE_CTABLE_SIZE_U32(MLFSELog, MaxML)]; + FSE_CTable litlengthCTable[FSE_CTABLE_SIZE_U32(LLFSELog, MaxLL)]; + FSE_repeat offcode_repeatMode; + FSE_repeat matchlength_repeatMode; + FSE_repeat litlength_repeatMode; +} ZSTD_fseCTables_t; + +typedef struct { + ZSTD_hufCTables_t huf; + ZSTD_fseCTables_t fse; +} ZSTD_entropyCTables_t; + +typedef struct { + U32 off; /* Offset code (offset + ZSTD_REP_MOVE) for the match */ + U32 len; /* Raw length of match */ +} ZSTD_match_t; + +typedef struct { + U32 offset; /* Offset of sequence */ + U32 litLength; /* Length of literals prior to match */ + U32 matchLength; /* Raw length of match */ +} rawSeq; + +typedef struct { + rawSeq* seq; /* The start of the sequences */ + size_t pos; /* The index in seq where reading stopped. pos <= size. */ + size_t posInSequence; /* The position within the sequence at seq[pos] where reading + stopped. posInSequence <= seq[pos].litLength + seq[pos].matchLength */ + size_t size; /* The number of sequences. <= capacity. */ + size_t capacity; /* The capacity starting from `seq` pointer */ +} rawSeqStore_t; + +UNUSED_ATTR static const rawSeqStore_t kNullRawSeqStore = {NULL, 0, 0, 0, 0}; + +typedef struct { + int price; + U32 off; + U32 mlen; + U32 litlen; + U32 rep[ZSTD_REP_NUM]; +} ZSTD_optimal_t; + +typedef enum { zop_dynamic=0, zop_predef } ZSTD_OptPrice_e; + +typedef struct { + /* All tables are allocated inside cctx->workspace by ZSTD_resetCCtx_internal() */ + unsigned* litFreq; /* table of literals statistics, of size 256 */ + unsigned* litLengthFreq; /* table of litLength statistics, of size (MaxLL+1) */ + unsigned* matchLengthFreq; /* table of matchLength statistics, of size (MaxML+1) */ + unsigned* offCodeFreq; /* table of offCode statistics, of size (MaxOff+1) */ + ZSTD_match_t* matchTable; /* list of found matches, of size ZSTD_OPT_NUM+1 */ + ZSTD_optimal_t* priceTable; /* All positions tracked by optimal parser, of size ZSTD_OPT_NUM+1 */ + + U32 litSum; /* nb of literals */ + U32 litLengthSum; /* nb of litLength codes */ + U32 matchLengthSum; /* nb of matchLength codes */ + U32 offCodeSum; /* nb of offset codes */ + U32 litSumBasePrice; /* to compare to log2(litfreq) */ + U32 litLengthSumBasePrice; /* to compare to log2(llfreq) */ + U32 matchLengthSumBasePrice;/* to compare to log2(mlfreq) */ + U32 offCodeSumBasePrice; /* to compare to log2(offreq) */ + ZSTD_OptPrice_e priceType; /* prices can be determined dynamically, or follow a pre-defined cost structure */ + const ZSTD_entropyCTables_t* symbolCosts; /* pre-calculated dictionary statistics */ + ZSTD_literalCompressionMode_e literalCompressionMode; +} optState_t; + +typedef struct { + ZSTD_entropyCTables_t entropy; + U32 rep[ZSTD_REP_NUM]; +} ZSTD_compressedBlockState_t; + +typedef struct { + BYTE const* nextSrc; /* next block here to continue on current prefix */ + BYTE const* base; /* All regular indexes relative to this position */ + BYTE const* dictBase; /* extDict indexes relative to this position */ + U32 dictLimit; /* below that point, need extDict */ + U32 lowLimit; /* below that point, no more valid data */ +} ZSTD_window_t; + +typedef struct ZSTD_matchState_t ZSTD_matchState_t; +struct ZSTD_matchState_t { + ZSTD_window_t window; /* State for window round buffer management */ + U32 loadedDictEnd; /* index of end of dictionary, within context's referential. + * When loadedDictEnd != 0, a dictionary is in use, and still valid. + * This relies on a mechanism to set loadedDictEnd=0 when dictionary is no longer within distance. + * Such mechanism is provided within ZSTD_window_enforceMaxDist() and ZSTD_checkDictValidity(). + * When dict referential is copied into active context (i.e. not attached), + * loadedDictEnd == dictSize, since referential starts from zero. + */ + U32 nextToUpdate; /* index from which to continue table update */ + U32 hashLog3; /* dispatch table for matches of len==3 : larger == faster, more memory */ + U32* hashTable; + U32* hashTable3; + U32* chainTable; + int dedicatedDictSearch; /* Indicates whether this matchState is using the + * dedicated dictionary search structure. + */ + optState_t opt; /* optimal parser state */ + const ZSTD_matchState_t* dictMatchState; + ZSTD_compressionParameters cParams; + const rawSeqStore_t* ldmSeqStore; +}; + +typedef struct { + ZSTD_compressedBlockState_t* prevCBlock; + ZSTD_compressedBlockState_t* nextCBlock; + ZSTD_matchState_t matchState; +} ZSTD_blockState_t; + +typedef struct { + U32 offset; + U32 checksum; +} ldmEntry_t; + +typedef struct { + BYTE const* split; + U32 hash; + U32 checksum; + ldmEntry_t* bucket; +} ldmMatchCandidate_t; + +#define LDM_BATCH_SIZE 64 + +typedef struct { + ZSTD_window_t window; /* State for the window round buffer management */ + ldmEntry_t* hashTable; + U32 loadedDictEnd; + BYTE* bucketOffsets; /* Next position in bucket to insert entry */ + size_t splitIndices[LDM_BATCH_SIZE]; + ldmMatchCandidate_t matchCandidates[LDM_BATCH_SIZE]; +} ldmState_t; + +typedef struct { + U32 enableLdm; /* 1 if enable long distance matching */ + U32 hashLog; /* Log size of hashTable */ + U32 bucketSizeLog; /* Log bucket size for collision resolution, at most 8 */ + U32 minMatchLength; /* Minimum match length */ + U32 hashRateLog; /* Log number of entries to skip */ + U32 windowLog; /* Window log for the LDM */ +} ldmParams_t; + +typedef struct { + int collectSequences; + ZSTD_Sequence* seqStart; + size_t seqIndex; + size_t maxSequences; +} SeqCollector; + +struct ZSTD_CCtx_params_s { + ZSTD_format_e format; + ZSTD_compressionParameters cParams; + ZSTD_frameParameters fParams; + + int compressionLevel; + int forceWindow; /* force back-references to respect limit of + * 1< 63) ? ZSTD_highbit32(litLength) + LL_deltaCode : LL_Code[litLength]; +} + +/* ZSTD_MLcode() : + * note : mlBase = matchLength - MINMATCH; + * because it's the format it's stored in seqStore->sequences */ +MEM_STATIC U32 ZSTD_MLcode(U32 mlBase) +{ + static const BYTE ML_Code[128] = { 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, + 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, + 32, 32, 33, 33, 34, 34, 35, 35, 36, 36, 36, 36, 37, 37, 37, 37, + 38, 38, 38, 38, 38, 38, 38, 38, 39, 39, 39, 39, 39, 39, 39, 39, + 40, 40, 40, 40, 40, 40, 40, 40, 40, 40, 40, 40, 40, 40, 40, 40, + 41, 41, 41, 41, 41, 41, 41, 41, 41, 41, 41, 41, 41, 41, 41, 41, + 42, 42, 42, 42, 42, 42, 42, 42, 42, 42, 42, 42, 42, 42, 42, 42, + 42, 42, 42, 42, 42, 42, 42, 42, 42, 42, 42, 42, 42, 42, 42, 42 }; + static const U32 ML_deltaCode = 36; + return (mlBase > 127) ? ZSTD_highbit32(mlBase) + ML_deltaCode : ML_Code[mlBase]; +} + +typedef struct repcodes_s { + U32 rep[3]; +} repcodes_t; + +MEM_STATIC repcodes_t ZSTD_updateRep(U32 const rep[3], U32 const offset, U32 const ll0) +{ + repcodes_t newReps; + if (offset >= ZSTD_REP_NUM) { /* full offset */ + newReps.rep[2] = rep[1]; + newReps.rep[1] = rep[0]; + newReps.rep[0] = offset - ZSTD_REP_MOVE; + } else { /* repcode */ + U32 const repCode = offset + ll0; + if (repCode > 0) { /* note : if repCode==0, no change */ + U32 const currentOffset = (repCode==ZSTD_REP_NUM) ? (rep[0] - 1) : rep[repCode]; + newReps.rep[2] = (repCode >= 2) ? rep[1] : rep[2]; + newReps.rep[1] = rep[0]; + newReps.rep[0] = currentOffset; + } else { /* repCode == 0 */ + ZSTD_memcpy(&newReps, rep, sizeof(newReps)); + } + } + return newReps; +} + +/* ZSTD_cParam_withinBounds: + * @return 1 if value is within cParam bounds, + * 0 otherwise */ +MEM_STATIC int ZSTD_cParam_withinBounds(ZSTD_cParameter cParam, int value) +{ + ZSTD_bounds const bounds = ZSTD_cParam_getBounds(cParam); + if (ZSTD_isError(bounds.error)) return 0; + if (value < bounds.lowerBound) return 0; + if (value > bounds.upperBound) return 0; + return 1; +} + +/* ZSTD_noCompressBlock() : + * Writes uncompressed block to dst buffer from given src. + * Returns the size of the block */ +MEM_STATIC size_t ZSTD_noCompressBlock (void* dst, size_t dstCapacity, const void* src, size_t srcSize, U32 lastBlock) +{ + U32 const cBlockHeader24 = lastBlock + (((U32)bt_raw)<<1) + (U32)(srcSize << 3); + RETURN_ERROR_IF(srcSize + ZSTD_blockHeaderSize > dstCapacity, + dstSize_tooSmall, "dst buf too small for uncompressed block"); + MEM_writeLE24(dst, cBlockHeader24); + ZSTD_memcpy((BYTE*)dst + ZSTD_blockHeaderSize, src, srcSize); + return ZSTD_blockHeaderSize + srcSize; +} + +MEM_STATIC size_t ZSTD_rleCompressBlock (void* dst, size_t dstCapacity, BYTE src, size_t srcSize, U32 lastBlock) +{ + BYTE* const op = (BYTE*)dst; + U32 const cBlockHeader = lastBlock + (((U32)bt_rle)<<1) + (U32)(srcSize << 3); + RETURN_ERROR_IF(dstCapacity < 4, dstSize_tooSmall, ""); + MEM_writeLE24(op, cBlockHeader); + op[3] = src; + return 4; +} + + +/* ZSTD_minGain() : + * minimum compression required + * to generate a compress block or a compressed literals section. + * note : use same formula for both situations */ +MEM_STATIC size_t ZSTD_minGain(size_t srcSize, ZSTD_strategy strat) +{ + U32 const minlog = (strat>=ZSTD_btultra) ? (U32)(strat) - 1 : 6; + ZSTD_STATIC_ASSERT(ZSTD_btultra == 8); + assert(ZSTD_cParam_withinBounds(ZSTD_c_strategy, strat)); + return (srcSize >> minlog) + 2; +} + +MEM_STATIC int ZSTD_disableLiteralsCompression(const ZSTD_CCtx_params* cctxParams) +{ + switch (cctxParams->literalCompressionMode) { + case ZSTD_lcm_huffman: + return 0; + case ZSTD_lcm_uncompressed: + return 1; + default: + assert(0 /* impossible: pre-validated */); + ZSTD_FALLTHROUGH; + case ZSTD_lcm_auto: + return (cctxParams->cParams.strategy == ZSTD_fast) && (cctxParams->cParams.targetLength > 0); + } +} + +/*! ZSTD_safecopyLiterals() : + * memcpy() function that won't read beyond more than WILDCOPY_OVERLENGTH bytes past ilimit_w. + * Only called when the sequence ends past ilimit_w, so it only needs to be optimized for single + * large copies. + */ +static void ZSTD_safecopyLiterals(BYTE* op, BYTE const* ip, BYTE const* const iend, BYTE const* ilimit_w) { + assert(iend > ilimit_w); + if (ip <= ilimit_w) { + ZSTD_wildcopy(op, ip, ilimit_w - ip, ZSTD_no_overlap); + op += ilimit_w - ip; + ip = ilimit_w; + } + while (ip < iend) *op++ = *ip++; +} + +/*! ZSTD_storeSeq() : + * Store a sequence (litlen, litPtr, offCode and mlBase) into seqStore_t. + * `offCode` : distance to match + ZSTD_REP_MOVE (values <= ZSTD_REP_MOVE are repCodes). + * `mlBase` : matchLength - MINMATCH + * Allowed to overread literals up to litLimit. +*/ +HINT_INLINE UNUSED_ATTR +void ZSTD_storeSeq(seqStore_t* seqStorePtr, size_t litLength, const BYTE* literals, const BYTE* litLimit, U32 offCode, size_t mlBase) +{ + BYTE const* const litLimit_w = litLimit - WILDCOPY_OVERLENGTH; + BYTE const* const litEnd = literals + litLength; +#if defined(DEBUGLEVEL) && (DEBUGLEVEL >= 6) + static const BYTE* g_start = NULL; + if (g_start==NULL) g_start = (const BYTE*)literals; /* note : index only works for compression within a single segment */ + { U32 const pos = (U32)((const BYTE*)literals - g_start); + DEBUGLOG(6, "Cpos%7u :%3u literals, match%4u bytes at offCode%7u", + pos, (U32)litLength, (U32)mlBase+MINMATCH, (U32)offCode); + } +#endif + assert((size_t)(seqStorePtr->sequences - seqStorePtr->sequencesStart) < seqStorePtr->maxNbSeq); + /* copy Literals */ + assert(seqStorePtr->maxNbLit <= 128 KB); + assert(seqStorePtr->lit + litLength <= seqStorePtr->litStart + seqStorePtr->maxNbLit); + assert(literals + litLength <= litLimit); + if (litEnd <= litLimit_w) { + /* Common case we can use wildcopy. + * First copy 16 bytes, because literals are likely short. + */ + assert(WILDCOPY_OVERLENGTH >= 16); + ZSTD_copy16(seqStorePtr->lit, literals); + if (litLength > 16) { + ZSTD_wildcopy(seqStorePtr->lit+16, literals+16, (ptrdiff_t)litLength-16, ZSTD_no_overlap); + } + } else { + ZSTD_safecopyLiterals(seqStorePtr->lit, literals, litEnd, litLimit_w); + } + seqStorePtr->lit += litLength; + + /* literal Length */ + if (litLength>0xFFFF) { + assert(seqStorePtr->longLengthID == 0); /* there can only be a single long length */ + seqStorePtr->longLengthID = 1; + seqStorePtr->longLengthPos = (U32)(seqStorePtr->sequences - seqStorePtr->sequencesStart); + } + seqStorePtr->sequences[0].litLength = (U16)litLength; + + /* match offset */ + seqStorePtr->sequences[0].offset = offCode + 1; + + /* match Length */ + if (mlBase>0xFFFF) { + assert(seqStorePtr->longLengthID == 0); /* there can only be a single long length */ + seqStorePtr->longLengthID = 2; + seqStorePtr->longLengthPos = (U32)(seqStorePtr->sequences - seqStorePtr->sequencesStart); + } + seqStorePtr->sequences[0].matchLength = (U16)mlBase; + + seqStorePtr->sequences++; +} + + +/*-************************************* +* Match length counter +***************************************/ +static unsigned ZSTD_NbCommonBytes (size_t val) +{ + if (MEM_isLittleEndian()) { + if (MEM_64bits()) { +# if (__GNUC__ >= 4) + return (__builtin_ctzll((U64)val) >> 3); +# else + static const int DeBruijnBytePos[64] = { 0, 0, 0, 0, 0, 1, 1, 2, + 0, 3, 1, 3, 1, 4, 2, 7, + 0, 2, 3, 6, 1, 5, 3, 5, + 1, 3, 4, 4, 2, 5, 6, 7, + 7, 0, 1, 2, 3, 3, 4, 6, + 2, 6, 5, 5, 3, 4, 5, 6, + 7, 1, 2, 4, 6, 4, 4, 5, + 7, 2, 6, 5, 7, 6, 7, 7 }; + return DeBruijnBytePos[((U64)((val & -(long long)val) * 0x0218A392CDABBD3FULL)) >> 58]; +# endif + } else { /* 32 bits */ +# if (__GNUC__ >= 3) + return (__builtin_ctz((U32)val) >> 3); +# else + static const int DeBruijnBytePos[32] = { 0, 0, 3, 0, 3, 1, 3, 0, + 3, 2, 2, 1, 3, 2, 0, 1, + 3, 3, 1, 2, 2, 2, 2, 0, + 3, 1, 2, 0, 1, 0, 1, 1 }; + return DeBruijnBytePos[((U32)((val & -(S32)val) * 0x077CB531U)) >> 27]; +# endif + } + } else { /* Big Endian CPU */ + if (MEM_64bits()) { +# if (__GNUC__ >= 4) + return (__builtin_clzll(val) >> 3); +# else + unsigned r; + const unsigned n32 = sizeof(size_t)*4; /* calculate this way due to compiler complaining in 32-bits mode */ + if (!(val>>n32)) { r=4; } else { r=0; val>>=n32; } + if (!(val>>16)) { r+=2; val>>=8; } else { val>>=24; } + r += (!val); + return r; +# endif + } else { /* 32 bits */ +# if (__GNUC__ >= 3) + return (__builtin_clz((U32)val) >> 3); +# else + unsigned r; + if (!(val>>16)) { r=2; val>>=8; } else { r=0; val>>=24; } + r += (!val); + return r; +# endif + } } +} + + +MEM_STATIC size_t ZSTD_count(const BYTE* pIn, const BYTE* pMatch, const BYTE* const pInLimit) +{ + const BYTE* const pStart = pIn; + const BYTE* const pInLoopLimit = pInLimit - (sizeof(size_t)-1); + + if (pIn < pInLoopLimit) { + { size_t const diff = MEM_readST(pMatch) ^ MEM_readST(pIn); + if (diff) return ZSTD_NbCommonBytes(diff); } + pIn+=sizeof(size_t); pMatch+=sizeof(size_t); + while (pIn < pInLoopLimit) { + size_t const diff = MEM_readST(pMatch) ^ MEM_readST(pIn); + if (!diff) { pIn+=sizeof(size_t); pMatch+=sizeof(size_t); continue; } + pIn += ZSTD_NbCommonBytes(diff); + return (size_t)(pIn - pStart); + } } + if (MEM_64bits() && (pIn<(pInLimit-3)) && (MEM_read32(pMatch) == MEM_read32(pIn))) { pIn+=4; pMatch+=4; } + if ((pIn<(pInLimit-1)) && (MEM_read16(pMatch) == MEM_read16(pIn))) { pIn+=2; pMatch+=2; } + if ((pIn> (32-h) ; } +MEM_STATIC size_t ZSTD_hash3Ptr(const void* ptr, U32 h) { return ZSTD_hash3(MEM_readLE32(ptr), h); } /* only in zstd_opt.h */ + +static const U32 prime4bytes = 2654435761U; +static U32 ZSTD_hash4(U32 u, U32 h) { return (u * prime4bytes) >> (32-h) ; } +static size_t ZSTD_hash4Ptr(const void* ptr, U32 h) { return ZSTD_hash4(MEM_read32(ptr), h); } + +static const U64 prime5bytes = 889523592379ULL; +static size_t ZSTD_hash5(U64 u, U32 h) { return (size_t)(((u << (64-40)) * prime5bytes) >> (64-h)) ; } +static size_t ZSTD_hash5Ptr(const void* p, U32 h) { return ZSTD_hash5(MEM_readLE64(p), h); } + +static const U64 prime6bytes = 227718039650203ULL; +static size_t ZSTD_hash6(U64 u, U32 h) { return (size_t)(((u << (64-48)) * prime6bytes) >> (64-h)) ; } +static size_t ZSTD_hash6Ptr(const void* p, U32 h) { return ZSTD_hash6(MEM_readLE64(p), h); } + +static const U64 prime7bytes = 58295818150454627ULL; +static size_t ZSTD_hash7(U64 u, U32 h) { return (size_t)(((u << (64-56)) * prime7bytes) >> (64-h)) ; } +static size_t ZSTD_hash7Ptr(const void* p, U32 h) { return ZSTD_hash7(MEM_readLE64(p), h); } + +static const U64 prime8bytes = 0xCF1BBCDCB7A56463ULL; +static size_t ZSTD_hash8(U64 u, U32 h) { return (size_t)(((u) * prime8bytes) >> (64-h)) ; } +static size_t ZSTD_hash8Ptr(const void* p, U32 h) { return ZSTD_hash8(MEM_readLE64(p), h); } + +MEM_STATIC FORCE_INLINE_ATTR +size_t ZSTD_hashPtr(const void* p, U32 hBits, U32 mls) +{ + switch(mls) + { + default: + case 4: return ZSTD_hash4Ptr(p, hBits); + case 5: return ZSTD_hash5Ptr(p, hBits); + case 6: return ZSTD_hash6Ptr(p, hBits); + case 7: return ZSTD_hash7Ptr(p, hBits); + case 8: return ZSTD_hash8Ptr(p, hBits); + } +} + +/* ZSTD_ipow() : + * Return base^exponent. + */ +static U64 ZSTD_ipow(U64 base, U64 exponent) +{ + U64 power = 1; + while (exponent) { + if (exponent & 1) power *= base; + exponent >>= 1; + base *= base; + } + return power; +} + +#define ZSTD_ROLL_HASH_CHAR_OFFSET 10 + +/* ZSTD_rollingHash_append() : + * Add the buffer to the hash value. + */ +static U64 ZSTD_rollingHash_append(U64 hash, void const* buf, size_t size) +{ + BYTE const* istart = (BYTE const*)buf; + size_t pos; + for (pos = 0; pos < size; ++pos) { + hash *= prime8bytes; + hash += istart[pos] + ZSTD_ROLL_HASH_CHAR_OFFSET; + } + return hash; +} + +/* ZSTD_rollingHash_compute() : + * Compute the rolling hash value of the buffer. + */ +MEM_STATIC U64 ZSTD_rollingHash_compute(void const* buf, size_t size) +{ + return ZSTD_rollingHash_append(0, buf, size); +} + +/* ZSTD_rollingHash_primePower() : + * Compute the primePower to be passed to ZSTD_rollingHash_rotate() for a hash + * over a window of length bytes. + */ +MEM_STATIC U64 ZSTD_rollingHash_primePower(U32 length) +{ + return ZSTD_ipow(prime8bytes, length - 1); +} + +/* ZSTD_rollingHash_rotate() : + * Rotate the rolling hash by one byte. + */ +MEM_STATIC U64 ZSTD_rollingHash_rotate(U64 hash, BYTE toRemove, BYTE toAdd, U64 primePower) +{ + hash -= (toRemove + ZSTD_ROLL_HASH_CHAR_OFFSET) * primePower; + hash *= prime8bytes; + hash += toAdd + ZSTD_ROLL_HASH_CHAR_OFFSET; + return hash; +} + +/*-************************************* +* Round buffer management +***************************************/ +#if (ZSTD_WINDOWLOG_MAX_64 > 31) +# error "ZSTD_WINDOWLOG_MAX is too large : would overflow ZSTD_CURRENT_MAX" +#endif +/* Max current allowed */ +#define ZSTD_CURRENT_MAX ((3U << 29) + (1U << ZSTD_WINDOWLOG_MAX)) +/* Maximum chunk size before overflow correction needs to be called again */ +#define ZSTD_CHUNKSIZE_MAX \ + ( ((U32)-1) /* Maximum ending current index */ \ + - ZSTD_CURRENT_MAX) /* Maximum beginning lowLimit */ + +/* + * ZSTD_window_clear(): + * Clears the window containing the history by simply setting it to empty. + */ +MEM_STATIC void ZSTD_window_clear(ZSTD_window_t* window) +{ + size_t const endT = (size_t)(window->nextSrc - window->base); + U32 const end = (U32)endT; + + window->lowLimit = end; + window->dictLimit = end; +} + +/* + * ZSTD_window_hasExtDict(): + * Returns non-zero if the window has a non-empty extDict. + */ +MEM_STATIC U32 ZSTD_window_hasExtDict(ZSTD_window_t const window) +{ + return window.lowLimit < window.dictLimit; +} + +/* + * ZSTD_matchState_dictMode(): + * Inspects the provided matchState and figures out what dictMode should be + * passed to the compressor. + */ +MEM_STATIC ZSTD_dictMode_e ZSTD_matchState_dictMode(const ZSTD_matchState_t *ms) +{ + return ZSTD_window_hasExtDict(ms->window) ? + ZSTD_extDict : + ms->dictMatchState != NULL ? + (ms->dictMatchState->dedicatedDictSearch ? ZSTD_dedicatedDictSearch : ZSTD_dictMatchState) : + ZSTD_noDict; +} + +/* + * ZSTD_window_needOverflowCorrection(): + * Returns non-zero if the indices are getting too large and need overflow + * protection. + */ +MEM_STATIC U32 ZSTD_window_needOverflowCorrection(ZSTD_window_t const window, + void const* srcEnd) +{ + U32 const curr = (U32)((BYTE const*)srcEnd - window.base); + return curr > ZSTD_CURRENT_MAX; +} + +/* + * ZSTD_window_correctOverflow(): + * Reduces the indices to protect from index overflow. + * Returns the correction made to the indices, which must be applied to every + * stored index. + * + * The least significant cycleLog bits of the indices must remain the same, + * which may be 0. Every index up to maxDist in the past must be valid. + * NOTE: (maxDist & cycleMask) must be zero. + */ +MEM_STATIC U32 ZSTD_window_correctOverflow(ZSTD_window_t* window, U32 cycleLog, + U32 maxDist, void const* src) +{ + /* preemptive overflow correction: + * 1. correction is large enough: + * lowLimit > (3<<29) ==> current > 3<<29 + 1< (3<<29 + 1< (3<<29) - (1< (3<<29) - (1<<30) (NOTE: chainLog <= 30) + * > 1<<29 + * + * 2. (ip+ZSTD_CHUNKSIZE_MAX - cctx->base) doesn't overflow: + * After correction, current is less than (1<base < 1<<32. + * 3. (cctx->lowLimit + 1< 3<<29 + 1<base); + U32 const currentCycle0 = curr & cycleMask; + /* Exclude zero so that newCurrent - maxDist >= 1. */ + U32 const currentCycle1 = currentCycle0 == 0 ? (1U << cycleLog) : currentCycle0; + U32 const newCurrent = currentCycle1 + maxDist; + U32 const correction = curr - newCurrent; + assert((maxDist & cycleMask) == 0); + assert(curr > newCurrent); + /* Loose bound, should be around 1<<29 (see above) */ + assert(correction > 1<<28); + + window->base += correction; + window->dictBase += correction; + if (window->lowLimit <= correction) window->lowLimit = 1; + else window->lowLimit -= correction; + if (window->dictLimit <= correction) window->dictLimit = 1; + else window->dictLimit -= correction; + + /* Ensure we can still reference the full window. */ + assert(newCurrent >= maxDist); + assert(newCurrent - maxDist >= 1); + /* Ensure that lowLimit and dictLimit didn't underflow. */ + assert(window->lowLimit <= newCurrent); + assert(window->dictLimit <= newCurrent); + + DEBUGLOG(4, "Correction of 0x%x bytes to lowLimit=0x%x", correction, + window->lowLimit); + return correction; +} + +/* + * ZSTD_window_enforceMaxDist(): + * Updates lowLimit so that: + * (srcEnd - base) - lowLimit == maxDist + loadedDictEnd + * + * It ensures index is valid as long as index >= lowLimit. + * This must be called before a block compression call. + * + * loadedDictEnd is only defined if a dictionary is in use for current compression. + * As the name implies, loadedDictEnd represents the index at end of dictionary. + * The value lies within context's referential, it can be directly compared to blockEndIdx. + * + * If loadedDictEndPtr is NULL, no dictionary is in use, and we use loadedDictEnd == 0. + * If loadedDictEndPtr is not NULL, we set it to zero after updating lowLimit. + * This is because dictionaries are allowed to be referenced fully + * as long as the last byte of the dictionary is in the window. + * Once input has progressed beyond window size, dictionary cannot be referenced anymore. + * + * In normal dict mode, the dictionary lies between lowLimit and dictLimit. + * In dictMatchState mode, lowLimit and dictLimit are the same, + * and the dictionary is below them. + * forceWindow and dictMatchState are therefore incompatible. + */ +MEM_STATIC void +ZSTD_window_enforceMaxDist(ZSTD_window_t* window, + const void* blockEnd, + U32 maxDist, + U32* loadedDictEndPtr, + const ZSTD_matchState_t** dictMatchStatePtr) +{ + U32 const blockEndIdx = (U32)((BYTE const*)blockEnd - window->base); + U32 const loadedDictEnd = (loadedDictEndPtr != NULL) ? *loadedDictEndPtr : 0; + DEBUGLOG(5, "ZSTD_window_enforceMaxDist: blockEndIdx=%u, maxDist=%u, loadedDictEnd=%u", + (unsigned)blockEndIdx, (unsigned)maxDist, (unsigned)loadedDictEnd); + + /* - When there is no dictionary : loadedDictEnd == 0. + In which case, the test (blockEndIdx > maxDist) is merely to avoid + overflowing next operation `newLowLimit = blockEndIdx - maxDist`. + - When there is a standard dictionary : + Index referential is copied from the dictionary, + which means it starts from 0. + In which case, loadedDictEnd == dictSize, + and it makes sense to compare `blockEndIdx > maxDist + dictSize` + since `blockEndIdx` also starts from zero. + - When there is an attached dictionary : + loadedDictEnd is expressed within the referential of the context, + so it can be directly compared against blockEndIdx. + */ + if (blockEndIdx > maxDist + loadedDictEnd) { + U32 const newLowLimit = blockEndIdx - maxDist; + if (window->lowLimit < newLowLimit) window->lowLimit = newLowLimit; + if (window->dictLimit < window->lowLimit) { + DEBUGLOG(5, "Update dictLimit to match lowLimit, from %u to %u", + (unsigned)window->dictLimit, (unsigned)window->lowLimit); + window->dictLimit = window->lowLimit; + } + /* On reaching window size, dictionaries are invalidated */ + if (loadedDictEndPtr) *loadedDictEndPtr = 0; + if (dictMatchStatePtr) *dictMatchStatePtr = NULL; + } +} + +/* Similar to ZSTD_window_enforceMaxDist(), + * but only invalidates dictionary + * when input progresses beyond window size. + * assumption : loadedDictEndPtr and dictMatchStatePtr are valid (non NULL) + * loadedDictEnd uses same referential as window->base + * maxDist is the window size */ +MEM_STATIC void +ZSTD_checkDictValidity(const ZSTD_window_t* window, + const void* blockEnd, + U32 maxDist, + U32* loadedDictEndPtr, + const ZSTD_matchState_t** dictMatchStatePtr) +{ + assert(loadedDictEndPtr != NULL); + assert(dictMatchStatePtr != NULL); + { U32 const blockEndIdx = (U32)((BYTE const*)blockEnd - window->base); + U32 const loadedDictEnd = *loadedDictEndPtr; + DEBUGLOG(5, "ZSTD_checkDictValidity: blockEndIdx=%u, maxDist=%u, loadedDictEnd=%u", + (unsigned)blockEndIdx, (unsigned)maxDist, (unsigned)loadedDictEnd); + assert(blockEndIdx >= loadedDictEnd); + + if (blockEndIdx > loadedDictEnd + maxDist) { + /* On reaching window size, dictionaries are invalidated. + * For simplification, if window size is reached anywhere within next block, + * the dictionary is invalidated for the full block. + */ + DEBUGLOG(6, "invalidating dictionary for current block (distance > windowSize)"); + *loadedDictEndPtr = 0; + *dictMatchStatePtr = NULL; + } else { + if (*loadedDictEndPtr != 0) { + DEBUGLOG(6, "dictionary considered valid for current block"); + } } } +} + +MEM_STATIC void ZSTD_window_init(ZSTD_window_t* window) { + ZSTD_memset(window, 0, sizeof(*window)); + window->base = (BYTE const*)""; + window->dictBase = (BYTE const*)""; + window->dictLimit = 1; /* start from 1, so that 1st position is valid */ + window->lowLimit = 1; /* it ensures first and later CCtx usages compress the same */ + window->nextSrc = window->base + 1; /* see issue #1241 */ +} + +/* + * ZSTD_window_update(): + * Updates the window by appending [src, src + srcSize) to the window. + * If it is not contiguous, the current prefix becomes the extDict, and we + * forget about the extDict. Handles overlap of the prefix and extDict. + * Returns non-zero if the segment is contiguous. + */ +MEM_STATIC U32 ZSTD_window_update(ZSTD_window_t* window, + void const* src, size_t srcSize) +{ + BYTE const* const ip = (BYTE const*)src; + U32 contiguous = 1; + DEBUGLOG(5, "ZSTD_window_update"); + if (srcSize == 0) + return contiguous; + assert(window->base != NULL); + assert(window->dictBase != NULL); + /* Check if blocks follow each other */ + if (src != window->nextSrc) { + /* not contiguous */ + size_t const distanceFromBase = (size_t)(window->nextSrc - window->base); + DEBUGLOG(5, "Non contiguous blocks, new segment starts at %u", window->dictLimit); + window->lowLimit = window->dictLimit; + assert(distanceFromBase == (size_t)(U32)distanceFromBase); /* should never overflow */ + window->dictLimit = (U32)distanceFromBase; + window->dictBase = window->base; + window->base = ip - distanceFromBase; + /* ms->nextToUpdate = window->dictLimit; */ + if (window->dictLimit - window->lowLimit < HASH_READ_SIZE) window->lowLimit = window->dictLimit; /* too small extDict */ + contiguous = 0; + } + window->nextSrc = ip + srcSize; + /* if input and dictionary overlap : reduce dictionary (area presumed modified by input) */ + if ( (ip+srcSize > window->dictBase + window->lowLimit) + & (ip < window->dictBase + window->dictLimit)) { + ptrdiff_t const highInputIdx = (ip + srcSize) - window->dictBase; + U32 const lowLimitMax = (highInputIdx > (ptrdiff_t)window->dictLimit) ? window->dictLimit : (U32)highInputIdx; + window->lowLimit = lowLimitMax; + DEBUGLOG(5, "Overlapping extDict and input : new lowLimit = %u", window->lowLimit); + } + return contiguous; +} + +/* + * Returns the lowest allowed match index. It may either be in the ext-dict or the prefix. + */ +MEM_STATIC U32 ZSTD_getLowestMatchIndex(const ZSTD_matchState_t* ms, U32 curr, unsigned windowLog) +{ + U32 const maxDistance = 1U << windowLog; + U32 const lowestValid = ms->window.lowLimit; + U32 const withinWindow = (curr - lowestValid > maxDistance) ? curr - maxDistance : lowestValid; + U32 const isDictionary = (ms->loadedDictEnd != 0); + /* When using a dictionary the entire dictionary is valid if a single byte of the dictionary + * is within the window. We invalidate the dictionary (and set loadedDictEnd to 0) when it isn't + * valid for the entire block. So this check is sufficient to find the lowest valid match index. + */ + U32 const matchLowest = isDictionary ? lowestValid : withinWindow; + return matchLowest; +} + +/* + * Returns the lowest allowed match index in the prefix. + */ +MEM_STATIC U32 ZSTD_getLowestPrefixIndex(const ZSTD_matchState_t* ms, U32 curr, unsigned windowLog) +{ + U32 const maxDistance = 1U << windowLog; + U32 const lowestValid = ms->window.dictLimit; + U32 const withinWindow = (curr - lowestValid > maxDistance) ? curr - maxDistance : lowestValid; + U32 const isDictionary = (ms->loadedDictEnd != 0); + /* When computing the lowest prefix index we need to take the dictionary into account to handle + * the edge case where the dictionary and the source are contiguous in memory. + */ + U32 const matchLowest = isDictionary ? lowestValid : withinWindow; + return matchLowest; +} + + + +/* debug functions */ +#if (DEBUGLEVEL>=2) + +MEM_STATIC double ZSTD_fWeight(U32 rawStat) +{ + U32 const fp_accuracy = 8; + U32 const fp_multiplier = (1 << fp_accuracy); + U32 const newStat = rawStat + 1; + U32 const hb = ZSTD_highbit32(newStat); + U32 const BWeight = hb * fp_multiplier; + U32 const FWeight = (newStat << fp_accuracy) >> hb; + U32 const weight = BWeight + FWeight; + assert(hb + fp_accuracy < 31); + return (double)weight / fp_multiplier; +} + +/* display a table content, + * listing each element, its frequency, and its predicted bit cost */ +MEM_STATIC void ZSTD_debugTable(const U32* table, U32 max) +{ + unsigned u, sum; + for (u=0, sum=0; u<=max; u++) sum += table[u]; + DEBUGLOG(2, "total nb elts: %u", sum); + for (u=0; u<=max; u++) { + DEBUGLOG(2, "%2u: %5u (%.2f)", + u, table[u], ZSTD_fWeight(sum) - ZSTD_fWeight(table[u]) ); + } +} + +#endif + + + +/* =============================================================== + * Shared internal declarations + * These prototypes may be called from sources not in lib/compress + * =============================================================== */ + +/* ZSTD_loadCEntropy() : + * dict : must point at beginning of a valid zstd dictionary. + * return : size of dictionary header (size of magic number + dict ID + entropy tables) + * assumptions : magic number supposed already checked + * and dictSize >= 8 */ +size_t ZSTD_loadCEntropy(ZSTD_compressedBlockState_t* bs, void* workspace, + const void* const dict, size_t dictSize); + +void ZSTD_reset_compressedBlockState(ZSTD_compressedBlockState_t* bs); + +/* ============================================================== + * Private declarations + * These prototypes shall only be called from within lib/compress + * ============================================================== */ + +/* ZSTD_getCParamsFromCCtxParams() : + * cParams are built depending on compressionLevel, src size hints, + * LDM and manually set compression parameters. + * Note: srcSizeHint == 0 means 0! + */ +ZSTD_compressionParameters ZSTD_getCParamsFromCCtxParams( + const ZSTD_CCtx_params* CCtxParams, U64 srcSizeHint, size_t dictSize, ZSTD_cParamMode_e mode); + +/*! ZSTD_initCStream_internal() : + * Private use only. Init streaming operation. + * expects params to be valid. + * must receive dict, or cdict, or none, but not both. + * @return : 0, or an error code */ +size_t ZSTD_initCStream_internal(ZSTD_CStream* zcs, + const void* dict, size_t dictSize, + const ZSTD_CDict* cdict, + const ZSTD_CCtx_params* params, unsigned long long pledgedSrcSize); + +void ZSTD_resetSeqStore(seqStore_t* ssPtr); + +/*! ZSTD_getCParamsFromCDict() : + * as the name implies */ +ZSTD_compressionParameters ZSTD_getCParamsFromCDict(const ZSTD_CDict* cdict); + +/* ZSTD_compressBegin_advanced_internal() : + * Private use only. To be called from zstdmt_compress.c. */ +size_t ZSTD_compressBegin_advanced_internal(ZSTD_CCtx* cctx, + const void* dict, size_t dictSize, + ZSTD_dictContentType_e dictContentType, + ZSTD_dictTableLoadMethod_e dtlm, + const ZSTD_CDict* cdict, + const ZSTD_CCtx_params* params, + unsigned long long pledgedSrcSize); + +/* ZSTD_compress_advanced_internal() : + * Private use only. To be called from zstdmt_compress.c. */ +size_t ZSTD_compress_advanced_internal(ZSTD_CCtx* cctx, + void* dst, size_t dstCapacity, + const void* src, size_t srcSize, + const void* dict,size_t dictSize, + const ZSTD_CCtx_params* params); + + +/* ZSTD_writeLastEmptyBlock() : + * output an empty Block with end-of-frame mark to complete a frame + * @return : size of data written into `dst` (== ZSTD_blockHeaderSize (defined in zstd_internal.h)) + * or an error code if `dstCapacity` is too small ( 1 */ +U32 ZSTD_cycleLog(U32 hashLog, ZSTD_strategy strat); + +/* ZSTD_CCtx_trace() : + * Trace the end of a compression call. + */ +void ZSTD_CCtx_trace(ZSTD_CCtx* cctx, size_t extraCSize); + +#endif /* ZSTD_COMPRESS_H */ diff --git a/lib/zstd/compress/zstd_compress_literals.c b/lib/zstd/compress/zstd_compress_literals.c new file mode 100644 index 000000000000..655bcda4d1f1 --- /dev/null +++ b/lib/zstd/compress/zstd_compress_literals.c @@ -0,0 +1,158 @@ +/* + * Copyright (c) Yann Collet, Facebook, Inc. + * All rights reserved. + * + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). + * You may select, at your option, one of the above-listed licenses. + */ + + /*-************************************* + * Dependencies + ***************************************/ +#include "zstd_compress_literals.h" + +size_t ZSTD_noCompressLiterals (void* dst, size_t dstCapacity, const void* src, size_t srcSize) +{ + BYTE* const ostart = (BYTE*)dst; + U32 const flSize = 1 + (srcSize>31) + (srcSize>4095); + + RETURN_ERROR_IF(srcSize + flSize > dstCapacity, dstSize_tooSmall, ""); + + switch(flSize) + { + case 1: /* 2 - 1 - 5 */ + ostart[0] = (BYTE)((U32)set_basic + (srcSize<<3)); + break; + case 2: /* 2 - 2 - 12 */ + MEM_writeLE16(ostart, (U16)((U32)set_basic + (1<<2) + (srcSize<<4))); + break; + case 3: /* 2 - 2 - 20 */ + MEM_writeLE32(ostart, (U32)((U32)set_basic + (3<<2) + (srcSize<<4))); + break; + default: /* not necessary : flSize is {1,2,3} */ + assert(0); + } + + ZSTD_memcpy(ostart + flSize, src, srcSize); + DEBUGLOG(5, "Raw literals: %u -> %u", (U32)srcSize, (U32)(srcSize + flSize)); + return srcSize + flSize; +} + +size_t ZSTD_compressRleLiteralsBlock (void* dst, size_t dstCapacity, const void* src, size_t srcSize) +{ + BYTE* const ostart = (BYTE*)dst; + U32 const flSize = 1 + (srcSize>31) + (srcSize>4095); + + (void)dstCapacity; /* dstCapacity already guaranteed to be >=4, hence large enough */ + + switch(flSize) + { + case 1: /* 2 - 1 - 5 */ + ostart[0] = (BYTE)((U32)set_rle + (srcSize<<3)); + break; + case 2: /* 2 - 2 - 12 */ + MEM_writeLE16(ostart, (U16)((U32)set_rle + (1<<2) + (srcSize<<4))); + break; + case 3: /* 2 - 2 - 20 */ + MEM_writeLE32(ostart, (U32)((U32)set_rle + (3<<2) + (srcSize<<4))); + break; + default: /* not necessary : flSize is {1,2,3} */ + assert(0); + } + + ostart[flSize] = *(const BYTE*)src; + DEBUGLOG(5, "RLE literals: %u -> %u", (U32)srcSize, (U32)flSize + 1); + return flSize+1; +} + +size_t ZSTD_compressLiterals (ZSTD_hufCTables_t const* prevHuf, + ZSTD_hufCTables_t* nextHuf, + ZSTD_strategy strategy, int disableLiteralCompression, + void* dst, size_t dstCapacity, + const void* src, size_t srcSize, + void* entropyWorkspace, size_t entropyWorkspaceSize, + const int bmi2) +{ + size_t const minGain = ZSTD_minGain(srcSize, strategy); + size_t const lhSize = 3 + (srcSize >= 1 KB) + (srcSize >= 16 KB); + BYTE* const ostart = (BYTE*)dst; + U32 singleStream = srcSize < 256; + symbolEncodingType_e hType = set_compressed; + size_t cLitSize; + + DEBUGLOG(5,"ZSTD_compressLiterals (disableLiteralCompression=%i srcSize=%u)", + disableLiteralCompression, (U32)srcSize); + + /* Prepare nextEntropy assuming reusing the existing table */ + ZSTD_memcpy(nextHuf, prevHuf, sizeof(*prevHuf)); + + if (disableLiteralCompression) + return ZSTD_noCompressLiterals(dst, dstCapacity, src, srcSize); + + /* small ? don't even attempt compression (speed opt) */ +# define COMPRESS_LITERALS_SIZE_MIN 63 + { size_t const minLitSize = (prevHuf->repeatMode == HUF_repeat_valid) ? 6 : COMPRESS_LITERALS_SIZE_MIN; + if (srcSize <= minLitSize) return ZSTD_noCompressLiterals(dst, dstCapacity, src, srcSize); + } + + RETURN_ERROR_IF(dstCapacity < lhSize+1, dstSize_tooSmall, "not enough space for compression"); + { HUF_repeat repeat = prevHuf->repeatMode; + int const preferRepeat = strategy < ZSTD_lazy ? srcSize <= 1024 : 0; + if (repeat == HUF_repeat_valid && lhSize == 3) singleStream = 1; + cLitSize = singleStream ? + HUF_compress1X_repeat( + ostart+lhSize, dstCapacity-lhSize, src, srcSize, + HUF_SYMBOLVALUE_MAX, HUF_TABLELOG_DEFAULT, entropyWorkspace, entropyWorkspaceSize, + (HUF_CElt*)nextHuf->CTable, &repeat, preferRepeat, bmi2) : + HUF_compress4X_repeat( + ostart+lhSize, dstCapacity-lhSize, src, srcSize, + HUF_SYMBOLVALUE_MAX, HUF_TABLELOG_DEFAULT, entropyWorkspace, entropyWorkspaceSize, + (HUF_CElt*)nextHuf->CTable, &repeat, preferRepeat, bmi2); + if (repeat != HUF_repeat_none) { + /* reused the existing table */ + DEBUGLOG(5, "Reusing previous huffman table"); + hType = set_repeat; + } + } + + if ((cLitSize==0) | (cLitSize >= srcSize - minGain) | ERR_isError(cLitSize)) { + ZSTD_memcpy(nextHuf, prevHuf, sizeof(*prevHuf)); + return ZSTD_noCompressLiterals(dst, dstCapacity, src, srcSize); + } + if (cLitSize==1) { + ZSTD_memcpy(nextHuf, prevHuf, sizeof(*prevHuf)); + return ZSTD_compressRleLiteralsBlock(dst, dstCapacity, src, srcSize); + } + + if (hType == set_compressed) { + /* using a newly constructed table */ + nextHuf->repeatMode = HUF_repeat_check; + } + + /* Build header */ + switch(lhSize) + { + case 3: /* 2 - 2 - 10 - 10 */ + { U32 const lhc = hType + ((!singleStream) << 2) + ((U32)srcSize<<4) + ((U32)cLitSize<<14); + MEM_writeLE24(ostart, lhc); + break; + } + case 4: /* 2 - 2 - 14 - 14 */ + { U32 const lhc = hType + (2 << 2) + ((U32)srcSize<<4) + ((U32)cLitSize<<18); + MEM_writeLE32(ostart, lhc); + break; + } + case 5: /* 2 - 2 - 18 - 18 */ + { U32 const lhc = hType + (3 << 2) + ((U32)srcSize<<4) + ((U32)cLitSize<<22); + MEM_writeLE32(ostart, lhc); + ostart[4] = (BYTE)(cLitSize >> 10); + break; + } + default: /* not possible : lhSize is {3,4,5} */ + assert(0); + } + DEBUGLOG(5, "Compressed literals: %u -> %u", (U32)srcSize, (U32)(lhSize+cLitSize)); + return lhSize+cLitSize; +} diff --git a/lib/zstd/compress/zstd_compress_literals.h b/lib/zstd/compress/zstd_compress_literals.h new file mode 100644 index 000000000000..9904c0cd30a0 --- /dev/null +++ b/lib/zstd/compress/zstd_compress_literals.h @@ -0,0 +1,29 @@ +/* + * Copyright (c) Yann Collet, Facebook, Inc. + * All rights reserved. + * + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). + * You may select, at your option, one of the above-listed licenses. + */ + +#ifndef ZSTD_COMPRESS_LITERALS_H +#define ZSTD_COMPRESS_LITERALS_H + +#include "zstd_compress_internal.h" /* ZSTD_hufCTables_t, ZSTD_minGain() */ + + +size_t ZSTD_noCompressLiterals (void* dst, size_t dstCapacity, const void* src, size_t srcSize); + +size_t ZSTD_compressRleLiteralsBlock (void* dst, size_t dstCapacity, const void* src, size_t srcSize); + +size_t ZSTD_compressLiterals (ZSTD_hufCTables_t const* prevHuf, + ZSTD_hufCTables_t* nextHuf, + ZSTD_strategy strategy, int disableLiteralCompression, + void* dst, size_t dstCapacity, + const void* src, size_t srcSize, + void* entropyWorkspace, size_t entropyWorkspaceSize, + const int bmi2); + +#endif /* ZSTD_COMPRESS_LITERALS_H */ diff --git a/lib/zstd/compress/zstd_compress_sequences.c b/lib/zstd/compress/zstd_compress_sequences.c new file mode 100644 index 000000000000..dcfcdc9cc5e8 --- /dev/null +++ b/lib/zstd/compress/zstd_compress_sequences.c @@ -0,0 +1,439 @@ +/* + * Copyright (c) Yann Collet, Facebook, Inc. + * All rights reserved. + * + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). + * You may select, at your option, one of the above-listed licenses. + */ + + /*-************************************* + * Dependencies + ***************************************/ +#include "zstd_compress_sequences.h" + +/* + * -log2(x / 256) lookup table for x in [0, 256). + * If x == 0: Return 0 + * Else: Return floor(-log2(x / 256) * 256) + */ +static unsigned const kInverseProbabilityLog256[256] = { + 0, 2048, 1792, 1642, 1536, 1453, 1386, 1329, 1280, 1236, 1197, 1162, + 1130, 1100, 1073, 1047, 1024, 1001, 980, 960, 941, 923, 906, 889, + 874, 859, 844, 830, 817, 804, 791, 779, 768, 756, 745, 734, + 724, 714, 704, 694, 685, 676, 667, 658, 650, 642, 633, 626, + 618, 610, 603, 595, 588, 581, 574, 567, 561, 554, 548, 542, + 535, 529, 523, 517, 512, 506, 500, 495, 489, 484, 478, 473, + 468, 463, 458, 453, 448, 443, 438, 434, 429, 424, 420, 415, + 411, 407, 402, 398, 394, 390, 386, 382, 377, 373, 370, 366, + 362, 358, 354, 350, 347, 343, 339, 336, 332, 329, 325, 322, + 318, 315, 311, 308, 305, 302, 298, 295, 292, 289, 286, 282, + 279, 276, 273, 270, 267, 264, 261, 258, 256, 253, 250, 247, + 244, 241, 239, 236, 233, 230, 228, 225, 222, 220, 217, 215, + 212, 209, 207, 204, 202, 199, 197, 194, 192, 190, 187, 185, + 182, 180, 178, 175, 173, 171, 168, 166, 164, 162, 159, 157, + 155, 153, 151, 149, 146, 144, 142, 140, 138, 136, 134, 132, + 130, 128, 126, 123, 121, 119, 117, 115, 114, 112, 110, 108, + 106, 104, 102, 100, 98, 96, 94, 93, 91, 89, 87, 85, + 83, 82, 80, 78, 76, 74, 73, 71, 69, 67, 66, 64, + 62, 61, 59, 57, 55, 54, 52, 50, 49, 47, 46, 44, + 42, 41, 39, 37, 36, 34, 33, 31, 30, 28, 26, 25, + 23, 22, 20, 19, 17, 16, 14, 13, 11, 10, 8, 7, + 5, 4, 2, 1, +}; + +static unsigned ZSTD_getFSEMaxSymbolValue(FSE_CTable const* ctable) { + void const* ptr = ctable; + U16 const* u16ptr = (U16 const*)ptr; + U32 const maxSymbolValue = MEM_read16(u16ptr + 1); + return maxSymbolValue; +} + +/* + * Returns true if we should use ncount=-1 else we should + * use ncount=1 for low probability symbols instead. + */ +static unsigned ZSTD_useLowProbCount(size_t const nbSeq) +{ + /* Heuristic: This should cover most blocks <= 16K and + * start to fade out after 16K to about 32K depending on + * comprssibility. + */ + return nbSeq >= 2048; +} + +/* + * Returns the cost in bytes of encoding the normalized count header. + * Returns an error if any of the helper functions return an error. + */ +static size_t ZSTD_NCountCost(unsigned const* count, unsigned const max, + size_t const nbSeq, unsigned const FSELog) +{ + BYTE wksp[FSE_NCOUNTBOUND]; + S16 norm[MaxSeq + 1]; + const U32 tableLog = FSE_optimalTableLog(FSELog, nbSeq, max); + FORWARD_IF_ERROR(FSE_normalizeCount(norm, tableLog, count, nbSeq, max, ZSTD_useLowProbCount(nbSeq)), ""); + return FSE_writeNCount(wksp, sizeof(wksp), norm, max, tableLog); +} + +/* + * Returns the cost in bits of encoding the distribution described by count + * using the entropy bound. + */ +static size_t ZSTD_entropyCost(unsigned const* count, unsigned const max, size_t const total) +{ + unsigned cost = 0; + unsigned s; + for (s = 0; s <= max; ++s) { + unsigned norm = (unsigned)((256 * count[s]) / total); + if (count[s] != 0 && norm == 0) + norm = 1; + assert(count[s] < total); + cost += count[s] * kInverseProbabilityLog256[norm]; + } + return cost >> 8; +} + +/* + * Returns the cost in bits of encoding the distribution in count using ctable. + * Returns an error if ctable cannot represent all the symbols in count. + */ +size_t ZSTD_fseBitCost( + FSE_CTable const* ctable, + unsigned const* count, + unsigned const max) +{ + unsigned const kAccuracyLog = 8; + size_t cost = 0; + unsigned s; + FSE_CState_t cstate; + FSE_initCState(&cstate, ctable); + if (ZSTD_getFSEMaxSymbolValue(ctable) < max) { + DEBUGLOG(5, "Repeat FSE_CTable has maxSymbolValue %u < %u", + ZSTD_getFSEMaxSymbolValue(ctable), max); + return ERROR(GENERIC); + } + for (s = 0; s <= max; ++s) { + unsigned const tableLog = cstate.stateLog; + unsigned const badCost = (tableLog + 1) << kAccuracyLog; + unsigned const bitCost = FSE_bitCost(cstate.symbolTT, tableLog, s, kAccuracyLog); + if (count[s] == 0) + continue; + if (bitCost >= badCost) { + DEBUGLOG(5, "Repeat FSE_CTable has Prob[%u] == 0", s); + return ERROR(GENERIC); + } + cost += (size_t)count[s] * bitCost; + } + return cost >> kAccuracyLog; +} + +/* + * Returns the cost in bits of encoding the distribution in count using the + * table described by norm. The max symbol support by norm is assumed >= max. + * norm must be valid for every symbol with non-zero probability in count. + */ +size_t ZSTD_crossEntropyCost(short const* norm, unsigned accuracyLog, + unsigned const* count, unsigned const max) +{ + unsigned const shift = 8 - accuracyLog; + size_t cost = 0; + unsigned s; + assert(accuracyLog <= 8); + for (s = 0; s <= max; ++s) { + unsigned const normAcc = (norm[s] != -1) ? (unsigned)norm[s] : 1; + unsigned const norm256 = normAcc << shift; + assert(norm256 > 0); + assert(norm256 < 256); + cost += count[s] * kInverseProbabilityLog256[norm256]; + } + return cost >> 8; +} + +symbolEncodingType_e +ZSTD_selectEncodingType( + FSE_repeat* repeatMode, unsigned const* count, unsigned const max, + size_t const mostFrequent, size_t nbSeq, unsigned const FSELog, + FSE_CTable const* prevCTable, + short const* defaultNorm, U32 defaultNormLog, + ZSTD_defaultPolicy_e const isDefaultAllowed, + ZSTD_strategy const strategy) +{ + ZSTD_STATIC_ASSERT(ZSTD_defaultDisallowed == 0 && ZSTD_defaultAllowed != 0); + if (mostFrequent == nbSeq) { + *repeatMode = FSE_repeat_none; + if (isDefaultAllowed && nbSeq <= 2) { + /* Prefer set_basic over set_rle when there are 2 or less symbols, + * since RLE uses 1 byte, but set_basic uses 5-6 bits per symbol. + * If basic encoding isn't possible, always choose RLE. + */ + DEBUGLOG(5, "Selected set_basic"); + return set_basic; + } + DEBUGLOG(5, "Selected set_rle"); + return set_rle; + } + if (strategy < ZSTD_lazy) { + if (isDefaultAllowed) { + size_t const staticFse_nbSeq_max = 1000; + size_t const mult = 10 - strategy; + size_t const baseLog = 3; + size_t const dynamicFse_nbSeq_min = (((size_t)1 << defaultNormLog) * mult) >> baseLog; /* 28-36 for offset, 56-72 for lengths */ + assert(defaultNormLog >= 5 && defaultNormLog <= 6); /* xx_DEFAULTNORMLOG */ + assert(mult <= 9 && mult >= 7); + if ( (*repeatMode == FSE_repeat_valid) + && (nbSeq < staticFse_nbSeq_max) ) { + DEBUGLOG(5, "Selected set_repeat"); + return set_repeat; + } + if ( (nbSeq < dynamicFse_nbSeq_min) + || (mostFrequent < (nbSeq >> (defaultNormLog-1))) ) { + DEBUGLOG(5, "Selected set_basic"); + /* The format allows default tables to be repeated, but it isn't useful. + * When using simple heuristics to select encoding type, we don't want + * to confuse these tables with dictionaries. When running more careful + * analysis, we don't need to waste time checking both repeating tables + * and default tables. + */ + *repeatMode = FSE_repeat_none; + return set_basic; + } + } + } else { + size_t const basicCost = isDefaultAllowed ? ZSTD_crossEntropyCost(defaultNorm, defaultNormLog, count, max) : ERROR(GENERIC); + size_t const repeatCost = *repeatMode != FSE_repeat_none ? ZSTD_fseBitCost(prevCTable, count, max) : ERROR(GENERIC); + size_t const NCountCost = ZSTD_NCountCost(count, max, nbSeq, FSELog); + size_t const compressedCost = (NCountCost << 3) + ZSTD_entropyCost(count, max, nbSeq); + + if (isDefaultAllowed) { + assert(!ZSTD_isError(basicCost)); + assert(!(*repeatMode == FSE_repeat_valid && ZSTD_isError(repeatCost))); + } + assert(!ZSTD_isError(NCountCost)); + assert(compressedCost < ERROR(maxCode)); + DEBUGLOG(5, "Estimated bit costs: basic=%u\trepeat=%u\tcompressed=%u", + (unsigned)basicCost, (unsigned)repeatCost, (unsigned)compressedCost); + if (basicCost <= repeatCost && basicCost <= compressedCost) { + DEBUGLOG(5, "Selected set_basic"); + assert(isDefaultAllowed); + *repeatMode = FSE_repeat_none; + return set_basic; + } + if (repeatCost <= compressedCost) { + DEBUGLOG(5, "Selected set_repeat"); + assert(!ZSTD_isError(repeatCost)); + return set_repeat; + } + assert(compressedCost < basicCost && compressedCost < repeatCost); + } + DEBUGLOG(5, "Selected set_compressed"); + *repeatMode = FSE_repeat_check; + return set_compressed; +} + +typedef struct { + S16 norm[MaxSeq + 1]; + U32 wksp[FSE_BUILD_CTABLE_WORKSPACE_SIZE_U32(MaxSeq, MaxFSELog)]; +} ZSTD_BuildCTableWksp; + +size_t +ZSTD_buildCTable(void* dst, size_t dstCapacity, + FSE_CTable* nextCTable, U32 FSELog, symbolEncodingType_e type, + unsigned* count, U32 max, + const BYTE* codeTable, size_t nbSeq, + const S16* defaultNorm, U32 defaultNormLog, U32 defaultMax, + const FSE_CTable* prevCTable, size_t prevCTableSize, + void* entropyWorkspace, size_t entropyWorkspaceSize) +{ + BYTE* op = (BYTE*)dst; + const BYTE* const oend = op + dstCapacity; + DEBUGLOG(6, "ZSTD_buildCTable (dstCapacity=%u)", (unsigned)dstCapacity); + + switch (type) { + case set_rle: + FORWARD_IF_ERROR(FSE_buildCTable_rle(nextCTable, (BYTE)max), ""); + RETURN_ERROR_IF(dstCapacity==0, dstSize_tooSmall, "not enough space"); + *op = codeTable[0]; + return 1; + case set_repeat: + ZSTD_memcpy(nextCTable, prevCTable, prevCTableSize); + return 0; + case set_basic: + FORWARD_IF_ERROR(FSE_buildCTable_wksp(nextCTable, defaultNorm, defaultMax, defaultNormLog, entropyWorkspace, entropyWorkspaceSize), ""); /* note : could be pre-calculated */ + return 0; + case set_compressed: { + ZSTD_BuildCTableWksp* wksp = (ZSTD_BuildCTableWksp*)entropyWorkspace; + size_t nbSeq_1 = nbSeq; + const U32 tableLog = FSE_optimalTableLog(FSELog, nbSeq, max); + if (count[codeTable[nbSeq-1]] > 1) { + count[codeTable[nbSeq-1]]--; + nbSeq_1--; + } + assert(nbSeq_1 > 1); + assert(entropyWorkspaceSize >= sizeof(ZSTD_BuildCTableWksp)); + (void)entropyWorkspaceSize; + FORWARD_IF_ERROR(FSE_normalizeCount(wksp->norm, tableLog, count, nbSeq_1, max, ZSTD_useLowProbCount(nbSeq_1)), ""); + { size_t const NCountSize = FSE_writeNCount(op, oend - op, wksp->norm, max, tableLog); /* overflow protected */ + FORWARD_IF_ERROR(NCountSize, "FSE_writeNCount failed"); + FORWARD_IF_ERROR(FSE_buildCTable_wksp(nextCTable, wksp->norm, max, tableLog, wksp->wksp, sizeof(wksp->wksp)), ""); + return NCountSize; + } + } + default: assert(0); RETURN_ERROR(GENERIC, "impossible to reach"); + } +} + +FORCE_INLINE_TEMPLATE size_t +ZSTD_encodeSequences_body( + void* dst, size_t dstCapacity, + FSE_CTable const* CTable_MatchLength, BYTE const* mlCodeTable, + FSE_CTable const* CTable_OffsetBits, BYTE const* ofCodeTable, + FSE_CTable const* CTable_LitLength, BYTE const* llCodeTable, + seqDef const* sequences, size_t nbSeq, int longOffsets) +{ + BIT_CStream_t blockStream; + FSE_CState_t stateMatchLength; + FSE_CState_t stateOffsetBits; + FSE_CState_t stateLitLength; + + RETURN_ERROR_IF( + ERR_isError(BIT_initCStream(&blockStream, dst, dstCapacity)), + dstSize_tooSmall, "not enough space remaining"); + DEBUGLOG(6, "available space for bitstream : %i (dstCapacity=%u)", + (int)(blockStream.endPtr - blockStream.startPtr), + (unsigned)dstCapacity); + + /* first symbols */ + FSE_initCState2(&stateMatchLength, CTable_MatchLength, mlCodeTable[nbSeq-1]); + FSE_initCState2(&stateOffsetBits, CTable_OffsetBits, ofCodeTable[nbSeq-1]); + FSE_initCState2(&stateLitLength, CTable_LitLength, llCodeTable[nbSeq-1]); + BIT_addBits(&blockStream, sequences[nbSeq-1].litLength, LL_bits[llCodeTable[nbSeq-1]]); + if (MEM_32bits()) BIT_flushBits(&blockStream); + BIT_addBits(&blockStream, sequences[nbSeq-1].matchLength, ML_bits[mlCodeTable[nbSeq-1]]); + if (MEM_32bits()) BIT_flushBits(&blockStream); + if (longOffsets) { + U32 const ofBits = ofCodeTable[nbSeq-1]; + unsigned const extraBits = ofBits - MIN(ofBits, STREAM_ACCUMULATOR_MIN-1); + if (extraBits) { + BIT_addBits(&blockStream, sequences[nbSeq-1].offset, extraBits); + BIT_flushBits(&blockStream); + } + BIT_addBits(&blockStream, sequences[nbSeq-1].offset >> extraBits, + ofBits - extraBits); + } else { + BIT_addBits(&blockStream, sequences[nbSeq-1].offset, ofCodeTable[nbSeq-1]); + } + BIT_flushBits(&blockStream); + + { size_t n; + for (n=nbSeq-2 ; n= 64-7-(LLFSELog+MLFSELog+OffFSELog))) + BIT_flushBits(&blockStream); /* (7)*/ + BIT_addBits(&blockStream, sequences[n].litLength, llBits); + if (MEM_32bits() && ((llBits+mlBits)>24)) BIT_flushBits(&blockStream); + BIT_addBits(&blockStream, sequences[n].matchLength, mlBits); + if (MEM_32bits() || (ofBits+mlBits+llBits > 56)) BIT_flushBits(&blockStream); + if (longOffsets) { + unsigned const extraBits = ofBits - MIN(ofBits, STREAM_ACCUMULATOR_MIN-1); + if (extraBits) { + BIT_addBits(&blockStream, sequences[n].offset, extraBits); + BIT_flushBits(&blockStream); /* (7)*/ + } + BIT_addBits(&blockStream, sequences[n].offset >> extraBits, + ofBits - extraBits); /* 31 */ + } else { + BIT_addBits(&blockStream, sequences[n].offset, ofBits); /* 31 */ + } + BIT_flushBits(&blockStream); /* (7)*/ + DEBUGLOG(7, "remaining space : %i", (int)(blockStream.endPtr - blockStream.ptr)); + } } + + DEBUGLOG(6, "ZSTD_encodeSequences: flushing ML state with %u bits", stateMatchLength.stateLog); + FSE_flushCState(&blockStream, &stateMatchLength); + DEBUGLOG(6, "ZSTD_encodeSequences: flushing Off state with %u bits", stateOffsetBits.stateLog); + FSE_flushCState(&blockStream, &stateOffsetBits); + DEBUGLOG(6, "ZSTD_encodeSequences: flushing LL state with %u bits", stateLitLength.stateLog); + FSE_flushCState(&blockStream, &stateLitLength); + + { size_t const streamSize = BIT_closeCStream(&blockStream); + RETURN_ERROR_IF(streamSize==0, dstSize_tooSmall, "not enough space"); + return streamSize; + } +} + +static size_t +ZSTD_encodeSequences_default( + void* dst, size_t dstCapacity, + FSE_CTable const* CTable_MatchLength, BYTE const* mlCodeTable, + FSE_CTable const* CTable_OffsetBits, BYTE const* ofCodeTable, + FSE_CTable const* CTable_LitLength, BYTE const* llCodeTable, + seqDef const* sequences, size_t nbSeq, int longOffsets) +{ + return ZSTD_encodeSequences_body(dst, dstCapacity, + CTable_MatchLength, mlCodeTable, + CTable_OffsetBits, ofCodeTable, + CTable_LitLength, llCodeTable, + sequences, nbSeq, longOffsets); +} + + +#if DYNAMIC_BMI2 + +static TARGET_ATTRIBUTE("bmi2") size_t +ZSTD_encodeSequences_bmi2( + void* dst, size_t dstCapacity, + FSE_CTable const* CTable_MatchLength, BYTE const* mlCodeTable, + FSE_CTable const* CTable_OffsetBits, BYTE const* ofCodeTable, + FSE_CTable const* CTable_LitLength, BYTE const* llCodeTable, + seqDef const* sequences, size_t nbSeq, int longOffsets) +{ + return ZSTD_encodeSequences_body(dst, dstCapacity, + CTable_MatchLength, mlCodeTable, + CTable_OffsetBits, ofCodeTable, + CTable_LitLength, llCodeTable, + sequences, nbSeq, longOffsets); +} + +#endif + +size_t ZSTD_encodeSequences( + void* dst, size_t dstCapacity, + FSE_CTable const* CTable_MatchLength, BYTE const* mlCodeTable, + FSE_CTable const* CTable_OffsetBits, BYTE const* ofCodeTable, + FSE_CTable const* CTable_LitLength, BYTE const* llCodeTable, + seqDef const* sequences, size_t nbSeq, int longOffsets, int bmi2) +{ + DEBUGLOG(5, "ZSTD_encodeSequences: dstCapacity = %u", (unsigned)dstCapacity); +#if DYNAMIC_BMI2 + if (bmi2) { + return ZSTD_encodeSequences_bmi2(dst, dstCapacity, + CTable_MatchLength, mlCodeTable, + CTable_OffsetBits, ofCodeTable, + CTable_LitLength, llCodeTable, + sequences, nbSeq, longOffsets); + } +#endif + (void)bmi2; + return ZSTD_encodeSequences_default(dst, dstCapacity, + CTable_MatchLength, mlCodeTable, + CTable_OffsetBits, ofCodeTable, + CTable_LitLength, llCodeTable, + sequences, nbSeq, longOffsets); +} diff --git a/lib/zstd/compress/zstd_compress_sequences.h b/lib/zstd/compress/zstd_compress_sequences.h new file mode 100644 index 000000000000..7991364c2f71 --- /dev/null +++ b/lib/zstd/compress/zstd_compress_sequences.h @@ -0,0 +1,54 @@ +/* + * Copyright (c) Yann Collet, Facebook, Inc. + * All rights reserved. + * + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). + * You may select, at your option, one of the above-listed licenses. + */ + +#ifndef ZSTD_COMPRESS_SEQUENCES_H +#define ZSTD_COMPRESS_SEQUENCES_H + +#include "../common/fse.h" /* FSE_repeat, FSE_CTable */ +#include "../common/zstd_internal.h" /* symbolEncodingType_e, ZSTD_strategy */ + +typedef enum { + ZSTD_defaultDisallowed = 0, + ZSTD_defaultAllowed = 1 +} ZSTD_defaultPolicy_e; + +symbolEncodingType_e +ZSTD_selectEncodingType( + FSE_repeat* repeatMode, unsigned const* count, unsigned const max, + size_t const mostFrequent, size_t nbSeq, unsigned const FSELog, + FSE_CTable const* prevCTable, + short const* defaultNorm, U32 defaultNormLog, + ZSTD_defaultPolicy_e const isDefaultAllowed, + ZSTD_strategy const strategy); + +size_t +ZSTD_buildCTable(void* dst, size_t dstCapacity, + FSE_CTable* nextCTable, U32 FSELog, symbolEncodingType_e type, + unsigned* count, U32 max, + const BYTE* codeTable, size_t nbSeq, + const S16* defaultNorm, U32 defaultNormLog, U32 defaultMax, + const FSE_CTable* prevCTable, size_t prevCTableSize, + void* entropyWorkspace, size_t entropyWorkspaceSize); + +size_t ZSTD_encodeSequences( + void* dst, size_t dstCapacity, + FSE_CTable const* CTable_MatchLength, BYTE const* mlCodeTable, + FSE_CTable const* CTable_OffsetBits, BYTE const* ofCodeTable, + FSE_CTable const* CTable_LitLength, BYTE const* llCodeTable, + seqDef const* sequences, size_t nbSeq, int longOffsets, int bmi2); + +size_t ZSTD_fseBitCost( + FSE_CTable const* ctable, + unsigned const* count, + unsigned const max); + +size_t ZSTD_crossEntropyCost(short const* norm, unsigned accuracyLog, + unsigned const* count, unsigned const max); +#endif /* ZSTD_COMPRESS_SEQUENCES_H */ diff --git a/lib/zstd/compress/zstd_compress_superblock.c b/lib/zstd/compress/zstd_compress_superblock.c new file mode 100644 index 000000000000..ee03e0aedb03 --- /dev/null +++ b/lib/zstd/compress/zstd_compress_superblock.c @@ -0,0 +1,850 @@ +/* + * Copyright (c) Yann Collet, Facebook, Inc. + * All rights reserved. + * + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). + * You may select, at your option, one of the above-listed licenses. + */ + + /*-************************************* + * Dependencies + ***************************************/ +#include "zstd_compress_superblock.h" + +#include "../common/zstd_internal.h" /* ZSTD_getSequenceLength */ +#include "hist.h" /* HIST_countFast_wksp */ +#include "zstd_compress_internal.h" +#include "zstd_compress_sequences.h" +#include "zstd_compress_literals.h" + +/*-************************************* +* Superblock entropy buffer structs +***************************************/ +/* ZSTD_hufCTablesMetadata_t : + * Stores Literals Block Type for a super-block in hType, and + * huffman tree description in hufDesBuffer. + * hufDesSize refers to the size of huffman tree description in bytes. + * This metadata is populated in ZSTD_buildSuperBlockEntropy_literal() */ +typedef struct { + symbolEncodingType_e hType; + BYTE hufDesBuffer[ZSTD_MAX_HUF_HEADER_SIZE]; + size_t hufDesSize; +} ZSTD_hufCTablesMetadata_t; + +/* ZSTD_fseCTablesMetadata_t : + * Stores symbol compression modes for a super-block in {ll, ol, ml}Type, and + * fse tables in fseTablesBuffer. + * fseTablesSize refers to the size of fse tables in bytes. + * This metadata is populated in ZSTD_buildSuperBlockEntropy_sequences() */ +typedef struct { + symbolEncodingType_e llType; + symbolEncodingType_e ofType; + symbolEncodingType_e mlType; + BYTE fseTablesBuffer[ZSTD_MAX_FSE_HEADERS_SIZE]; + size_t fseTablesSize; + size_t lastCountSize; /* This is to account for bug in 1.3.4. More detail in ZSTD_compressSubBlock_sequences() */ +} ZSTD_fseCTablesMetadata_t; + +typedef struct { + ZSTD_hufCTablesMetadata_t hufMetadata; + ZSTD_fseCTablesMetadata_t fseMetadata; +} ZSTD_entropyCTablesMetadata_t; + + +/* ZSTD_buildSuperBlockEntropy_literal() : + * Builds entropy for the super-block literals. + * Stores literals block type (raw, rle, compressed, repeat) and + * huffman description table to hufMetadata. + * @return : size of huffman description table or error code */ +static size_t ZSTD_buildSuperBlockEntropy_literal(void* const src, size_t srcSize, + const ZSTD_hufCTables_t* prevHuf, + ZSTD_hufCTables_t* nextHuf, + ZSTD_hufCTablesMetadata_t* hufMetadata, + const int disableLiteralsCompression, + void* workspace, size_t wkspSize) +{ + BYTE* const wkspStart = (BYTE*)workspace; + BYTE* const wkspEnd = wkspStart + wkspSize; + BYTE* const countWkspStart = wkspStart; + unsigned* const countWksp = (unsigned*)workspace; + const size_t countWkspSize = (HUF_SYMBOLVALUE_MAX + 1) * sizeof(unsigned); + BYTE* const nodeWksp = countWkspStart + countWkspSize; + const size_t nodeWkspSize = wkspEnd-nodeWksp; + unsigned maxSymbolValue = 255; + unsigned huffLog = HUF_TABLELOG_DEFAULT; + HUF_repeat repeat = prevHuf->repeatMode; + + DEBUGLOG(5, "ZSTD_buildSuperBlockEntropy_literal (srcSize=%zu)", srcSize); + + /* Prepare nextEntropy assuming reusing the existing table */ + ZSTD_memcpy(nextHuf, prevHuf, sizeof(*prevHuf)); + + if (disableLiteralsCompression) { + DEBUGLOG(5, "set_basic - disabled"); + hufMetadata->hType = set_basic; + return 0; + } + + /* small ? don't even attempt compression (speed opt) */ +# define COMPRESS_LITERALS_SIZE_MIN 63 + { size_t const minLitSize = (prevHuf->repeatMode == HUF_repeat_valid) ? 6 : COMPRESS_LITERALS_SIZE_MIN; + if (srcSize <= minLitSize) { + DEBUGLOG(5, "set_basic - too small"); + hufMetadata->hType = set_basic; + return 0; + } + } + + /* Scan input and build symbol stats */ + { size_t const largest = HIST_count_wksp (countWksp, &maxSymbolValue, (const BYTE*)src, srcSize, workspace, wkspSize); + FORWARD_IF_ERROR(largest, "HIST_count_wksp failed"); + if (largest == srcSize) { + DEBUGLOG(5, "set_rle"); + hufMetadata->hType = set_rle; + return 0; + } + if (largest <= (srcSize >> 7)+4) { + DEBUGLOG(5, "set_basic - no gain"); + hufMetadata->hType = set_basic; + return 0; + } + } + + /* Validate the previous Huffman table */ + if (repeat == HUF_repeat_check && !HUF_validateCTable((HUF_CElt const*)prevHuf->CTable, countWksp, maxSymbolValue)) { + repeat = HUF_repeat_none; + } + + /* Build Huffman Tree */ + ZSTD_memset(nextHuf->CTable, 0, sizeof(nextHuf->CTable)); + huffLog = HUF_optimalTableLog(huffLog, srcSize, maxSymbolValue); + { size_t const maxBits = HUF_buildCTable_wksp((HUF_CElt*)nextHuf->CTable, countWksp, + maxSymbolValue, huffLog, + nodeWksp, nodeWkspSize); + FORWARD_IF_ERROR(maxBits, "HUF_buildCTable_wksp"); + huffLog = (U32)maxBits; + { /* Build and write the CTable */ + size_t const newCSize = HUF_estimateCompressedSize( + (HUF_CElt*)nextHuf->CTable, countWksp, maxSymbolValue); + size_t const hSize = HUF_writeCTable_wksp( + hufMetadata->hufDesBuffer, sizeof(hufMetadata->hufDesBuffer), + (HUF_CElt*)nextHuf->CTable, maxSymbolValue, huffLog, + nodeWksp, nodeWkspSize); + /* Check against repeating the previous CTable */ + if (repeat != HUF_repeat_none) { + size_t const oldCSize = HUF_estimateCompressedSize( + (HUF_CElt const*)prevHuf->CTable, countWksp, maxSymbolValue); + if (oldCSize < srcSize && (oldCSize <= hSize + newCSize || hSize + 12 >= srcSize)) { + DEBUGLOG(5, "set_repeat - smaller"); + ZSTD_memcpy(nextHuf, prevHuf, sizeof(*prevHuf)); + hufMetadata->hType = set_repeat; + return 0; + } + } + if (newCSize + hSize >= srcSize) { + DEBUGLOG(5, "set_basic - no gains"); + ZSTD_memcpy(nextHuf, prevHuf, sizeof(*prevHuf)); + hufMetadata->hType = set_basic; + return 0; + } + DEBUGLOG(5, "set_compressed (hSize=%u)", (U32)hSize); + hufMetadata->hType = set_compressed; + nextHuf->repeatMode = HUF_repeat_check; + return hSize; + } + } +} + +/* ZSTD_buildSuperBlockEntropy_sequences() : + * Builds entropy for the super-block sequences. + * Stores symbol compression modes and fse table to fseMetadata. + * @return : size of fse tables or error code */ +static size_t ZSTD_buildSuperBlockEntropy_sequences(seqStore_t* seqStorePtr, + const ZSTD_fseCTables_t* prevEntropy, + ZSTD_fseCTables_t* nextEntropy, + const ZSTD_CCtx_params* cctxParams, + ZSTD_fseCTablesMetadata_t* fseMetadata, + void* workspace, size_t wkspSize) +{ + BYTE* const wkspStart = (BYTE*)workspace; + BYTE* const wkspEnd = wkspStart + wkspSize; + BYTE* const countWkspStart = wkspStart; + unsigned* const countWksp = (unsigned*)workspace; + const size_t countWkspSize = (MaxSeq + 1) * sizeof(unsigned); + BYTE* const cTableWksp = countWkspStart + countWkspSize; + const size_t cTableWkspSize = wkspEnd-cTableWksp; + ZSTD_strategy const strategy = cctxParams->cParams.strategy; + FSE_CTable* CTable_LitLength = nextEntropy->litlengthCTable; + FSE_CTable* CTable_OffsetBits = nextEntropy->offcodeCTable; + FSE_CTable* CTable_MatchLength = nextEntropy->matchlengthCTable; + const BYTE* const ofCodeTable = seqStorePtr->ofCode; + const BYTE* const llCodeTable = seqStorePtr->llCode; + const BYTE* const mlCodeTable = seqStorePtr->mlCode; + size_t const nbSeq = seqStorePtr->sequences - seqStorePtr->sequencesStart; + BYTE* const ostart = fseMetadata->fseTablesBuffer; + BYTE* const oend = ostart + sizeof(fseMetadata->fseTablesBuffer); + BYTE* op = ostart; + + assert(cTableWkspSize >= (1 << MaxFSELog) * sizeof(FSE_FUNCTION_TYPE)); + DEBUGLOG(5, "ZSTD_buildSuperBlockEntropy_sequences (nbSeq=%zu)", nbSeq); + ZSTD_memset(workspace, 0, wkspSize); + + fseMetadata->lastCountSize = 0; + /* convert length/distances into codes */ + ZSTD_seqToCodes(seqStorePtr); + /* build CTable for Literal Lengths */ + { U32 LLtype; + unsigned max = MaxLL; + size_t const mostFrequent = HIST_countFast_wksp(countWksp, &max, llCodeTable, nbSeq, workspace, wkspSize); /* can't fail */ + DEBUGLOG(5, "Building LL table"); + nextEntropy->litlength_repeatMode = prevEntropy->litlength_repeatMode; + LLtype = ZSTD_selectEncodingType(&nextEntropy->litlength_repeatMode, + countWksp, max, mostFrequent, nbSeq, + LLFSELog, prevEntropy->litlengthCTable, + LL_defaultNorm, LL_defaultNormLog, + ZSTD_defaultAllowed, strategy); + assert(set_basic < set_compressed && set_rle < set_compressed); + assert(!(LLtype < set_compressed && nextEntropy->litlength_repeatMode != FSE_repeat_none)); /* We don't copy tables */ + { size_t const countSize = ZSTD_buildCTable(op, oend - op, CTable_LitLength, LLFSELog, (symbolEncodingType_e)LLtype, + countWksp, max, llCodeTable, nbSeq, LL_defaultNorm, LL_defaultNormLog, MaxLL, + prevEntropy->litlengthCTable, sizeof(prevEntropy->litlengthCTable), + cTableWksp, cTableWkspSize); + FORWARD_IF_ERROR(countSize, "ZSTD_buildCTable for LitLens failed"); + if (LLtype == set_compressed) + fseMetadata->lastCountSize = countSize; + op += countSize; + fseMetadata->llType = (symbolEncodingType_e) LLtype; + } } + /* build CTable for Offsets */ + { U32 Offtype; + unsigned max = MaxOff; + size_t const mostFrequent = HIST_countFast_wksp(countWksp, &max, ofCodeTable, nbSeq, workspace, wkspSize); /* can't fail */ + /* We can only use the basic table if max <= DefaultMaxOff, otherwise the offsets are too large */ + ZSTD_defaultPolicy_e const defaultPolicy = (max <= DefaultMaxOff) ? ZSTD_defaultAllowed : ZSTD_defaultDisallowed; + DEBUGLOG(5, "Building OF table"); + nextEntropy->offcode_repeatMode = prevEntropy->offcode_repeatMode; + Offtype = ZSTD_selectEncodingType(&nextEntropy->offcode_repeatMode, + countWksp, max, mostFrequent, nbSeq, + OffFSELog, prevEntropy->offcodeCTable, + OF_defaultNorm, OF_defaultNormLog, + defaultPolicy, strategy); + assert(!(Offtype < set_compressed && nextEntropy->offcode_repeatMode != FSE_repeat_none)); /* We don't copy tables */ + { size_t const countSize = ZSTD_buildCTable(op, oend - op, CTable_OffsetBits, OffFSELog, (symbolEncodingType_e)Offtype, + countWksp, max, ofCodeTable, nbSeq, OF_defaultNorm, OF_defaultNormLog, DefaultMaxOff, + prevEntropy->offcodeCTable, sizeof(prevEntropy->offcodeCTable), + cTableWksp, cTableWkspSize); + FORWARD_IF_ERROR(countSize, "ZSTD_buildCTable for Offsets failed"); + if (Offtype == set_compressed) + fseMetadata->lastCountSize = countSize; + op += countSize; + fseMetadata->ofType = (symbolEncodingType_e) Offtype; + } } + /* build CTable for MatchLengths */ + { U32 MLtype; + unsigned max = MaxML; + size_t const mostFrequent = HIST_countFast_wksp(countWksp, &max, mlCodeTable, nbSeq, workspace, wkspSize); /* can't fail */ + DEBUGLOG(5, "Building ML table (remaining space : %i)", (int)(oend-op)); + nextEntropy->matchlength_repeatMode = prevEntropy->matchlength_repeatMode; + MLtype = ZSTD_selectEncodingType(&nextEntropy->matchlength_repeatMode, + countWksp, max, mostFrequent, nbSeq, + MLFSELog, prevEntropy->matchlengthCTable, + ML_defaultNorm, ML_defaultNormLog, + ZSTD_defaultAllowed, strategy); + assert(!(MLtype < set_compressed && nextEntropy->matchlength_repeatMode != FSE_repeat_none)); /* We don't copy tables */ + { size_t const countSize = ZSTD_buildCTable(op, oend - op, CTable_MatchLength, MLFSELog, (symbolEncodingType_e)MLtype, + countWksp, max, mlCodeTable, nbSeq, ML_defaultNorm, ML_defaultNormLog, MaxML, + prevEntropy->matchlengthCTable, sizeof(prevEntropy->matchlengthCTable), + cTableWksp, cTableWkspSize); + FORWARD_IF_ERROR(countSize, "ZSTD_buildCTable for MatchLengths failed"); + if (MLtype == set_compressed) + fseMetadata->lastCountSize = countSize; + op += countSize; + fseMetadata->mlType = (symbolEncodingType_e) MLtype; + } } + assert((size_t) (op-ostart) <= sizeof(fseMetadata->fseTablesBuffer)); + return op-ostart; +} + + +/* ZSTD_buildSuperBlockEntropy() : + * Builds entropy for the super-block. + * @return : 0 on success or error code */ +static size_t +ZSTD_buildSuperBlockEntropy(seqStore_t* seqStorePtr, + const ZSTD_entropyCTables_t* prevEntropy, + ZSTD_entropyCTables_t* nextEntropy, + const ZSTD_CCtx_params* cctxParams, + ZSTD_entropyCTablesMetadata_t* entropyMetadata, + void* workspace, size_t wkspSize) +{ + size_t const litSize = seqStorePtr->lit - seqStorePtr->litStart; + DEBUGLOG(5, "ZSTD_buildSuperBlockEntropy"); + entropyMetadata->hufMetadata.hufDesSize = + ZSTD_buildSuperBlockEntropy_literal(seqStorePtr->litStart, litSize, + &prevEntropy->huf, &nextEntropy->huf, + &entropyMetadata->hufMetadata, + ZSTD_disableLiteralsCompression(cctxParams), + workspace, wkspSize); + FORWARD_IF_ERROR(entropyMetadata->hufMetadata.hufDesSize, "ZSTD_buildSuperBlockEntropy_literal failed"); + entropyMetadata->fseMetadata.fseTablesSize = + ZSTD_buildSuperBlockEntropy_sequences(seqStorePtr, + &prevEntropy->fse, &nextEntropy->fse, + cctxParams, + &entropyMetadata->fseMetadata, + workspace, wkspSize); + FORWARD_IF_ERROR(entropyMetadata->fseMetadata.fseTablesSize, "ZSTD_buildSuperBlockEntropy_sequences failed"); + return 0; +} + +/* ZSTD_compressSubBlock_literal() : + * Compresses literals section for a sub-block. + * When we have to write the Huffman table we will sometimes choose a header + * size larger than necessary. This is because we have to pick the header size + * before we know the table size + compressed size, so we have a bound on the + * table size. If we guessed incorrectly, we fall back to uncompressed literals. + * + * We write the header when writeEntropy=1 and set entropyWritten=1 when we succeeded + * in writing the header, otherwise it is set to 0. + * + * hufMetadata->hType has literals block type info. + * If it is set_basic, all sub-blocks literals section will be Raw_Literals_Block. + * If it is set_rle, all sub-blocks literals section will be RLE_Literals_Block. + * If it is set_compressed, first sub-block's literals section will be Compressed_Literals_Block + * If it is set_compressed, first sub-block's literals section will be Treeless_Literals_Block + * and the following sub-blocks' literals sections will be Treeless_Literals_Block. + * @return : compressed size of literals section of a sub-block + * Or 0 if it unable to compress. + * Or error code */ +static size_t ZSTD_compressSubBlock_literal(const HUF_CElt* hufTable, + const ZSTD_hufCTablesMetadata_t* hufMetadata, + const BYTE* literals, size_t litSize, + void* dst, size_t dstSize, + const int bmi2, int writeEntropy, int* entropyWritten) +{ + size_t const header = writeEntropy ? 200 : 0; + size_t const lhSize = 3 + (litSize >= (1 KB - header)) + (litSize >= (16 KB - header)); + BYTE* const ostart = (BYTE*)dst; + BYTE* const oend = ostart + dstSize; + BYTE* op = ostart + lhSize; + U32 const singleStream = lhSize == 3; + symbolEncodingType_e hType = writeEntropy ? hufMetadata->hType : set_repeat; + size_t cLitSize = 0; + + (void)bmi2; /* TODO bmi2... */ + + DEBUGLOG(5, "ZSTD_compressSubBlock_literal (litSize=%zu, lhSize=%zu, writeEntropy=%d)", litSize, lhSize, writeEntropy); + + *entropyWritten = 0; + if (litSize == 0 || hufMetadata->hType == set_basic) { + DEBUGLOG(5, "ZSTD_compressSubBlock_literal using raw literal"); + return ZSTD_noCompressLiterals(dst, dstSize, literals, litSize); + } else if (hufMetadata->hType == set_rle) { + DEBUGLOG(5, "ZSTD_compressSubBlock_literal using rle literal"); + return ZSTD_compressRleLiteralsBlock(dst, dstSize, literals, litSize); + } + + assert(litSize > 0); + assert(hufMetadata->hType == set_compressed || hufMetadata->hType == set_repeat); + + if (writeEntropy && hufMetadata->hType == set_compressed) { + ZSTD_memcpy(op, hufMetadata->hufDesBuffer, hufMetadata->hufDesSize); + op += hufMetadata->hufDesSize; + cLitSize += hufMetadata->hufDesSize; + DEBUGLOG(5, "ZSTD_compressSubBlock_literal (hSize=%zu)", hufMetadata->hufDesSize); + } + + /* TODO bmi2 */ + { const size_t cSize = singleStream ? HUF_compress1X_usingCTable(op, oend-op, literals, litSize, hufTable) + : HUF_compress4X_usingCTable(op, oend-op, literals, litSize, hufTable); + op += cSize; + cLitSize += cSize; + if (cSize == 0 || ERR_isError(cSize)) { + DEBUGLOG(5, "Failed to write entropy tables %s", ZSTD_getErrorName(cSize)); + return 0; + } + /* If we expand and we aren't writing a header then emit uncompressed */ + if (!writeEntropy && cLitSize >= litSize) { + DEBUGLOG(5, "ZSTD_compressSubBlock_literal using raw literal because uncompressible"); + return ZSTD_noCompressLiterals(dst, dstSize, literals, litSize); + } + /* If we are writing headers then allow expansion that doesn't change our header size. */ + if (lhSize < (size_t)(3 + (cLitSize >= 1 KB) + (cLitSize >= 16 KB))) { + assert(cLitSize > litSize); + DEBUGLOG(5, "Literals expanded beyond allowed header size"); + return ZSTD_noCompressLiterals(dst, dstSize, literals, litSize); + } + DEBUGLOG(5, "ZSTD_compressSubBlock_literal (cSize=%zu)", cSize); + } + + /* Build header */ + switch(lhSize) + { + case 3: /* 2 - 2 - 10 - 10 */ + { U32 const lhc = hType + ((!singleStream) << 2) + ((U32)litSize<<4) + ((U32)cLitSize<<14); + MEM_writeLE24(ostart, lhc); + break; + } + case 4: /* 2 - 2 - 14 - 14 */ + { U32 const lhc = hType + (2 << 2) + ((U32)litSize<<4) + ((U32)cLitSize<<18); + MEM_writeLE32(ostart, lhc); + break; + } + case 5: /* 2 - 2 - 18 - 18 */ + { U32 const lhc = hType + (3 << 2) + ((U32)litSize<<4) + ((U32)cLitSize<<22); + MEM_writeLE32(ostart, lhc); + ostart[4] = (BYTE)(cLitSize >> 10); + break; + } + default: /* not possible : lhSize is {3,4,5} */ + assert(0); + } + *entropyWritten = 1; + DEBUGLOG(5, "Compressed literals: %u -> %u", (U32)litSize, (U32)(op-ostart)); + return op-ostart; +} + +static size_t ZSTD_seqDecompressedSize(seqStore_t const* seqStore, const seqDef* sequences, size_t nbSeq, size_t litSize, int lastSequence) { + const seqDef* const sstart = sequences; + const seqDef* const send = sequences + nbSeq; + const seqDef* sp = sstart; + size_t matchLengthSum = 0; + size_t litLengthSum = 0; + while (send-sp > 0) { + ZSTD_sequenceLength const seqLen = ZSTD_getSequenceLength(seqStore, sp); + litLengthSum += seqLen.litLength; + matchLengthSum += seqLen.matchLength; + sp++; + } + assert(litLengthSum <= litSize); + if (!lastSequence) { + assert(litLengthSum == litSize); + } + return matchLengthSum + litSize; +} + +/* ZSTD_compressSubBlock_sequences() : + * Compresses sequences section for a sub-block. + * fseMetadata->llType, fseMetadata->ofType, and fseMetadata->mlType have + * symbol compression modes for the super-block. + * The first successfully compressed block will have these in its header. + * We set entropyWritten=1 when we succeed in compressing the sequences. + * The following sub-blocks will always have repeat mode. + * @return : compressed size of sequences section of a sub-block + * Or 0 if it is unable to compress + * Or error code. */ +static size_t ZSTD_compressSubBlock_sequences(const ZSTD_fseCTables_t* fseTables, + const ZSTD_fseCTablesMetadata_t* fseMetadata, + const seqDef* sequences, size_t nbSeq, + const BYTE* llCode, const BYTE* mlCode, const BYTE* ofCode, + const ZSTD_CCtx_params* cctxParams, + void* dst, size_t dstCapacity, + const int bmi2, int writeEntropy, int* entropyWritten) +{ + const int longOffsets = cctxParams->cParams.windowLog > STREAM_ACCUMULATOR_MIN; + BYTE* const ostart = (BYTE*)dst; + BYTE* const oend = ostart + dstCapacity; + BYTE* op = ostart; + BYTE* seqHead; + + DEBUGLOG(5, "ZSTD_compressSubBlock_sequences (nbSeq=%zu, writeEntropy=%d, longOffsets=%d)", nbSeq, writeEntropy, longOffsets); + + *entropyWritten = 0; + /* Sequences Header */ + RETURN_ERROR_IF((oend-op) < 3 /*max nbSeq Size*/ + 1 /*seqHead*/, + dstSize_tooSmall, ""); + if (nbSeq < 0x7F) + *op++ = (BYTE)nbSeq; + else if (nbSeq < LONGNBSEQ) + op[0] = (BYTE)((nbSeq>>8) + 0x80), op[1] = (BYTE)nbSeq, op+=2; + else + op[0]=0xFF, MEM_writeLE16(op+1, (U16)(nbSeq - LONGNBSEQ)), op+=3; + if (nbSeq==0) { + return op - ostart; + } + + /* seqHead : flags for FSE encoding type */ + seqHead = op++; + + DEBUGLOG(5, "ZSTD_compressSubBlock_sequences (seqHeadSize=%u)", (unsigned)(op-ostart)); + + if (writeEntropy) { + const U32 LLtype = fseMetadata->llType; + const U32 Offtype = fseMetadata->ofType; + const U32 MLtype = fseMetadata->mlType; + DEBUGLOG(5, "ZSTD_compressSubBlock_sequences (fseTablesSize=%zu)", fseMetadata->fseTablesSize); + *seqHead = (BYTE)((LLtype<<6) + (Offtype<<4) + (MLtype<<2)); + ZSTD_memcpy(op, fseMetadata->fseTablesBuffer, fseMetadata->fseTablesSize); + op += fseMetadata->fseTablesSize; + } else { + const U32 repeat = set_repeat; + *seqHead = (BYTE)((repeat<<6) + (repeat<<4) + (repeat<<2)); + } + + { size_t const bitstreamSize = ZSTD_encodeSequences( + op, oend - op, + fseTables->matchlengthCTable, mlCode, + fseTables->offcodeCTable, ofCode, + fseTables->litlengthCTable, llCode, + sequences, nbSeq, + longOffsets, bmi2); + FORWARD_IF_ERROR(bitstreamSize, "ZSTD_encodeSequences failed"); + op += bitstreamSize; + /* zstd versions <= 1.3.4 mistakenly report corruption when + * FSE_readNCount() receives a buffer < 4 bytes. + * Fixed by https://github.com/facebook/zstd/pull/1146. + * This can happen when the last set_compressed table present is 2 + * bytes and the bitstream is only one byte. + * In this exceedingly rare case, we will simply emit an uncompressed + * block, since it isn't worth optimizing. + */ +#ifndef FUZZING_BUILD_MODE_UNSAFE_FOR_PRODUCTION + if (writeEntropy && fseMetadata->lastCountSize && fseMetadata->lastCountSize + bitstreamSize < 4) { + /* NCountSize >= 2 && bitstreamSize > 0 ==> lastCountSize == 3 */ + assert(fseMetadata->lastCountSize + bitstreamSize == 3); + DEBUGLOG(5, "Avoiding bug in zstd decoder in versions <= 1.3.4 by " + "emitting an uncompressed block."); + return 0; + } +#endif + DEBUGLOG(5, "ZSTD_compressSubBlock_sequences (bitstreamSize=%zu)", bitstreamSize); + } + + /* zstd versions <= 1.4.0 mistakenly report error when + * sequences section body size is less than 3 bytes. + * Fixed by https://github.com/facebook/zstd/pull/1664. + * This can happen when the previous sequences section block is compressed + * with rle mode and the current block's sequences section is compressed + * with repeat mode where sequences section body size can be 1 byte. + */ +#ifndef FUZZING_BUILD_MODE_UNSAFE_FOR_PRODUCTION + if (op-seqHead < 4) { + DEBUGLOG(5, "Avoiding bug in zstd decoder in versions <= 1.4.0 by emitting " + "an uncompressed block when sequences are < 4 bytes"); + return 0; + } +#endif + + *entropyWritten = 1; + return op - ostart; +} + +/* ZSTD_compressSubBlock() : + * Compresses a single sub-block. + * @return : compressed size of the sub-block + * Or 0 if it failed to compress. */ +static size_t ZSTD_compressSubBlock(const ZSTD_entropyCTables_t* entropy, + const ZSTD_entropyCTablesMetadata_t* entropyMetadata, + const seqDef* sequences, size_t nbSeq, + const BYTE* literals, size_t litSize, + const BYTE* llCode, const BYTE* mlCode, const BYTE* ofCode, + const ZSTD_CCtx_params* cctxParams, + void* dst, size_t dstCapacity, + const int bmi2, + int writeLitEntropy, int writeSeqEntropy, + int* litEntropyWritten, int* seqEntropyWritten, + U32 lastBlock) +{ + BYTE* const ostart = (BYTE*)dst; + BYTE* const oend = ostart + dstCapacity; + BYTE* op = ostart + ZSTD_blockHeaderSize; + DEBUGLOG(5, "ZSTD_compressSubBlock (litSize=%zu, nbSeq=%zu, writeLitEntropy=%d, writeSeqEntropy=%d, lastBlock=%d)", + litSize, nbSeq, writeLitEntropy, writeSeqEntropy, lastBlock); + { size_t cLitSize = ZSTD_compressSubBlock_literal((const HUF_CElt*)entropy->huf.CTable, + &entropyMetadata->hufMetadata, literals, litSize, + op, oend-op, bmi2, writeLitEntropy, litEntropyWritten); + FORWARD_IF_ERROR(cLitSize, "ZSTD_compressSubBlock_literal failed"); + if (cLitSize == 0) return 0; + op += cLitSize; + } + { size_t cSeqSize = ZSTD_compressSubBlock_sequences(&entropy->fse, + &entropyMetadata->fseMetadata, + sequences, nbSeq, + llCode, mlCode, ofCode, + cctxParams, + op, oend-op, + bmi2, writeSeqEntropy, seqEntropyWritten); + FORWARD_IF_ERROR(cSeqSize, "ZSTD_compressSubBlock_sequences failed"); + if (cSeqSize == 0) return 0; + op += cSeqSize; + } + /* Write block header */ + { size_t cSize = (op-ostart)-ZSTD_blockHeaderSize; + U32 const cBlockHeader24 = lastBlock + (((U32)bt_compressed)<<1) + (U32)(cSize << 3); + MEM_writeLE24(ostart, cBlockHeader24); + } + return op-ostart; +} + +static size_t ZSTD_estimateSubBlockSize_literal(const BYTE* literals, size_t litSize, + const ZSTD_hufCTables_t* huf, + const ZSTD_hufCTablesMetadata_t* hufMetadata, + void* workspace, size_t wkspSize, + int writeEntropy) +{ + unsigned* const countWksp = (unsigned*)workspace; + unsigned maxSymbolValue = 255; + size_t literalSectionHeaderSize = 3; /* Use hard coded size of 3 bytes */ + + if (hufMetadata->hType == set_basic) return litSize; + else if (hufMetadata->hType == set_rle) return 1; + else if (hufMetadata->hType == set_compressed || hufMetadata->hType == set_repeat) { + size_t const largest = HIST_count_wksp (countWksp, &maxSymbolValue, (const BYTE*)literals, litSize, workspace, wkspSize); + if (ZSTD_isError(largest)) return litSize; + { size_t cLitSizeEstimate = HUF_estimateCompressedSize((const HUF_CElt*)huf->CTable, countWksp, maxSymbolValue); + if (writeEntropy) cLitSizeEstimate += hufMetadata->hufDesSize; + return cLitSizeEstimate + literalSectionHeaderSize; + } } + assert(0); /* impossible */ + return 0; +} + +static size_t ZSTD_estimateSubBlockSize_symbolType(symbolEncodingType_e type, + const BYTE* codeTable, unsigned maxCode, + size_t nbSeq, const FSE_CTable* fseCTable, + const U32* additionalBits, + short const* defaultNorm, U32 defaultNormLog, U32 defaultMax, + void* workspace, size_t wkspSize) +{ + unsigned* const countWksp = (unsigned*)workspace; + const BYTE* ctp = codeTable; + const BYTE* const ctStart = ctp; + const BYTE* const ctEnd = ctStart + nbSeq; + size_t cSymbolTypeSizeEstimateInBits = 0; + unsigned max = maxCode; + + HIST_countFast_wksp(countWksp, &max, codeTable, nbSeq, workspace, wkspSize); /* can't fail */ + if (type == set_basic) { + /* We selected this encoding type, so it must be valid. */ + assert(max <= defaultMax); + cSymbolTypeSizeEstimateInBits = max <= defaultMax + ? ZSTD_crossEntropyCost(defaultNorm, defaultNormLog, countWksp, max) + : ERROR(GENERIC); + } else if (type == set_rle) { + cSymbolTypeSizeEstimateInBits = 0; + } else if (type == set_compressed || type == set_repeat) { + cSymbolTypeSizeEstimateInBits = ZSTD_fseBitCost(fseCTable, countWksp, max); + } + if (ZSTD_isError(cSymbolTypeSizeEstimateInBits)) return nbSeq * 10; + while (ctp < ctEnd) { + if (additionalBits) cSymbolTypeSizeEstimateInBits += additionalBits[*ctp]; + else cSymbolTypeSizeEstimateInBits += *ctp; /* for offset, offset code is also the number of additional bits */ + ctp++; + } + return cSymbolTypeSizeEstimateInBits / 8; +} + +static size_t ZSTD_estimateSubBlockSize_sequences(const BYTE* ofCodeTable, + const BYTE* llCodeTable, + const BYTE* mlCodeTable, + size_t nbSeq, + const ZSTD_fseCTables_t* fseTables, + const ZSTD_fseCTablesMetadata_t* fseMetadata, + void* workspace, size_t wkspSize, + int writeEntropy) +{ + size_t sequencesSectionHeaderSize = 3; /* Use hard coded size of 3 bytes */ + size_t cSeqSizeEstimate = 0; + cSeqSizeEstimate += ZSTD_estimateSubBlockSize_symbolType(fseMetadata->ofType, ofCodeTable, MaxOff, + nbSeq, fseTables->offcodeCTable, NULL, + OF_defaultNorm, OF_defaultNormLog, DefaultMaxOff, + workspace, wkspSize); + cSeqSizeEstimate += ZSTD_estimateSubBlockSize_symbolType(fseMetadata->llType, llCodeTable, MaxLL, + nbSeq, fseTables->litlengthCTable, LL_bits, + LL_defaultNorm, LL_defaultNormLog, MaxLL, + workspace, wkspSize); + cSeqSizeEstimate += ZSTD_estimateSubBlockSize_symbolType(fseMetadata->mlType, mlCodeTable, MaxML, + nbSeq, fseTables->matchlengthCTable, ML_bits, + ML_defaultNorm, ML_defaultNormLog, MaxML, + workspace, wkspSize); + if (writeEntropy) cSeqSizeEstimate += fseMetadata->fseTablesSize; + return cSeqSizeEstimate + sequencesSectionHeaderSize; +} + +static size_t ZSTD_estimateSubBlockSize(const BYTE* literals, size_t litSize, + const BYTE* ofCodeTable, + const BYTE* llCodeTable, + const BYTE* mlCodeTable, + size_t nbSeq, + const ZSTD_entropyCTables_t* entropy, + const ZSTD_entropyCTablesMetadata_t* entropyMetadata, + void* workspace, size_t wkspSize, + int writeLitEntropy, int writeSeqEntropy) { + size_t cSizeEstimate = 0; + cSizeEstimate += ZSTD_estimateSubBlockSize_literal(literals, litSize, + &entropy->huf, &entropyMetadata->hufMetadata, + workspace, wkspSize, writeLitEntropy); + cSizeEstimate += ZSTD_estimateSubBlockSize_sequences(ofCodeTable, llCodeTable, mlCodeTable, + nbSeq, &entropy->fse, &entropyMetadata->fseMetadata, + workspace, wkspSize, writeSeqEntropy); + return cSizeEstimate + ZSTD_blockHeaderSize; +} + +static int ZSTD_needSequenceEntropyTables(ZSTD_fseCTablesMetadata_t const* fseMetadata) +{ + if (fseMetadata->llType == set_compressed || fseMetadata->llType == set_rle) + return 1; + if (fseMetadata->mlType == set_compressed || fseMetadata->mlType == set_rle) + return 1; + if (fseMetadata->ofType == set_compressed || fseMetadata->ofType == set_rle) + return 1; + return 0; +} + +/* ZSTD_compressSubBlock_multi() : + * Breaks super-block into multiple sub-blocks and compresses them. + * Entropy will be written to the first block. + * The following blocks will use repeat mode to compress. + * All sub-blocks are compressed blocks (no raw or rle blocks). + * @return : compressed size of the super block (which is multiple ZSTD blocks) + * Or 0 if it failed to compress. */ +static size_t ZSTD_compressSubBlock_multi(const seqStore_t* seqStorePtr, + const ZSTD_compressedBlockState_t* prevCBlock, + ZSTD_compressedBlockState_t* nextCBlock, + const ZSTD_entropyCTablesMetadata_t* entropyMetadata, + const ZSTD_CCtx_params* cctxParams, + void* dst, size_t dstCapacity, + const void* src, size_t srcSize, + const int bmi2, U32 lastBlock, + void* workspace, size_t wkspSize) +{ + const seqDef* const sstart = seqStorePtr->sequencesStart; + const seqDef* const send = seqStorePtr->sequences; + const seqDef* sp = sstart; + const BYTE* const lstart = seqStorePtr->litStart; + const BYTE* const lend = seqStorePtr->lit; + const BYTE* lp = lstart; + BYTE const* ip = (BYTE const*)src; + BYTE const* const iend = ip + srcSize; + BYTE* const ostart = (BYTE*)dst; + BYTE* const oend = ostart + dstCapacity; + BYTE* op = ostart; + const BYTE* llCodePtr = seqStorePtr->llCode; + const BYTE* mlCodePtr = seqStorePtr->mlCode; + const BYTE* ofCodePtr = seqStorePtr->ofCode; + size_t targetCBlockSize = cctxParams->targetCBlockSize; + size_t litSize, seqCount; + int writeLitEntropy = entropyMetadata->hufMetadata.hType == set_compressed; + int writeSeqEntropy = 1; + int lastSequence = 0; + + DEBUGLOG(5, "ZSTD_compressSubBlock_multi (litSize=%u, nbSeq=%u)", + (unsigned)(lend-lp), (unsigned)(send-sstart)); + + litSize = 0; + seqCount = 0; + do { + size_t cBlockSizeEstimate = 0; + if (sstart == send) { + lastSequence = 1; + } else { + const seqDef* const sequence = sp + seqCount; + lastSequence = sequence == send - 1; + litSize += ZSTD_getSequenceLength(seqStorePtr, sequence).litLength; + seqCount++; + } + if (lastSequence) { + assert(lp <= lend); + assert(litSize <= (size_t)(lend - lp)); + litSize = (size_t)(lend - lp); + } + /* I think there is an optimization opportunity here. + * Calling ZSTD_estimateSubBlockSize for every sequence can be wasteful + * since it recalculates estimate from scratch. + * For example, it would recount literal distribution and symbol codes everytime. + */ + cBlockSizeEstimate = ZSTD_estimateSubBlockSize(lp, litSize, ofCodePtr, llCodePtr, mlCodePtr, seqCount, + &nextCBlock->entropy, entropyMetadata, + workspace, wkspSize, writeLitEntropy, writeSeqEntropy); + if (cBlockSizeEstimate > targetCBlockSize || lastSequence) { + int litEntropyWritten = 0; + int seqEntropyWritten = 0; + const size_t decompressedSize = ZSTD_seqDecompressedSize(seqStorePtr, sp, seqCount, litSize, lastSequence); + const size_t cSize = ZSTD_compressSubBlock(&nextCBlock->entropy, entropyMetadata, + sp, seqCount, + lp, litSize, + llCodePtr, mlCodePtr, ofCodePtr, + cctxParams, + op, oend-op, + bmi2, writeLitEntropy, writeSeqEntropy, + &litEntropyWritten, &seqEntropyWritten, + lastBlock && lastSequence); + FORWARD_IF_ERROR(cSize, "ZSTD_compressSubBlock failed"); + if (cSize > 0 && cSize < decompressedSize) { + DEBUGLOG(5, "Committed the sub-block"); + assert(ip + decompressedSize <= iend); + ip += decompressedSize; + sp += seqCount; + lp += litSize; + op += cSize; + llCodePtr += seqCount; + mlCodePtr += seqCount; + ofCodePtr += seqCount; + litSize = 0; + seqCount = 0; + /* Entropy only needs to be written once */ + if (litEntropyWritten) { + writeLitEntropy = 0; + } + if (seqEntropyWritten) { + writeSeqEntropy = 0; + } + } + } + } while (!lastSequence); + if (writeLitEntropy) { + DEBUGLOG(5, "ZSTD_compressSubBlock_multi has literal entropy tables unwritten"); + ZSTD_memcpy(&nextCBlock->entropy.huf, &prevCBlock->entropy.huf, sizeof(prevCBlock->entropy.huf)); + } + if (writeSeqEntropy && ZSTD_needSequenceEntropyTables(&entropyMetadata->fseMetadata)) { + /* If we haven't written our entropy tables, then we've violated our contract and + * must emit an uncompressed block. + */ + DEBUGLOG(5, "ZSTD_compressSubBlock_multi has sequence entropy tables unwritten"); + return 0; + } + if (ip < iend) { + size_t const cSize = ZSTD_noCompressBlock(op, oend - op, ip, iend - ip, lastBlock); + DEBUGLOG(5, "ZSTD_compressSubBlock_multi last sub-block uncompressed, %zu bytes", (size_t)(iend - ip)); + FORWARD_IF_ERROR(cSize, "ZSTD_noCompressBlock failed"); + assert(cSize != 0); + op += cSize; + /* We have to regenerate the repcodes because we've skipped some sequences */ + if (sp < send) { + seqDef const* seq; + repcodes_t rep; + ZSTD_memcpy(&rep, prevCBlock->rep, sizeof(rep)); + for (seq = sstart; seq < sp; ++seq) { + rep = ZSTD_updateRep(rep.rep, seq->offset - 1, ZSTD_getSequenceLength(seqStorePtr, seq).litLength == 0); + } + ZSTD_memcpy(nextCBlock->rep, &rep, sizeof(rep)); + } + } + DEBUGLOG(5, "ZSTD_compressSubBlock_multi compressed"); + return op-ostart; +} + +size_t ZSTD_compressSuperBlock(ZSTD_CCtx* zc, + void* dst, size_t dstCapacity, + void const* src, size_t srcSize, + unsigned lastBlock) { + ZSTD_entropyCTablesMetadata_t entropyMetadata; + + FORWARD_IF_ERROR(ZSTD_buildSuperBlockEntropy(&zc->seqStore, + &zc->blockState.prevCBlock->entropy, + &zc->blockState.nextCBlock->entropy, + &zc->appliedParams, + &entropyMetadata, + zc->entropyWorkspace, ENTROPY_WORKSPACE_SIZE /* statically allocated in resetCCtx */), ""); + + return ZSTD_compressSubBlock_multi(&zc->seqStore, + zc->blockState.prevCBlock, + zc->blockState.nextCBlock, + &entropyMetadata, + &zc->appliedParams, + dst, dstCapacity, + src, srcSize, + zc->bmi2, lastBlock, + zc->entropyWorkspace, ENTROPY_WORKSPACE_SIZE /* statically allocated in resetCCtx */); +} diff --git a/lib/zstd/compress/zstd_compress_superblock.h b/lib/zstd/compress/zstd_compress_superblock.h new file mode 100644 index 000000000000..224ece79546e --- /dev/null +++ b/lib/zstd/compress/zstd_compress_superblock.h @@ -0,0 +1,32 @@ +/* + * Copyright (c) Yann Collet, Facebook, Inc. + * All rights reserved. + * + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). + * You may select, at your option, one of the above-listed licenses. + */ + +#ifndef ZSTD_COMPRESS_ADVANCED_H +#define ZSTD_COMPRESS_ADVANCED_H + +/*-************************************* +* Dependencies +***************************************/ + +#include /* ZSTD_CCtx */ + +/*-************************************* +* Target Compressed Block Size +***************************************/ + +/* ZSTD_compressSuperBlock() : + * Used to compress a super block when targetCBlockSize is being used. + * The given block will be compressed into multiple sub blocks that are around targetCBlockSize. */ +size_t ZSTD_compressSuperBlock(ZSTD_CCtx* zc, + void* dst, size_t dstCapacity, + void const* src, size_t srcSize, + unsigned lastBlock); + +#endif /* ZSTD_COMPRESS_ADVANCED_H */ diff --git a/lib/zstd/compress/zstd_cwksp.h b/lib/zstd/compress/zstd_cwksp.h new file mode 100644 index 000000000000..98e359adf5d4 --- /dev/null +++ b/lib/zstd/compress/zstd_cwksp.h @@ -0,0 +1,482 @@ +/* + * Copyright (c) Yann Collet, Facebook, Inc. + * All rights reserved. + * + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). + * You may select, at your option, one of the above-listed licenses. + */ + +#ifndef ZSTD_CWKSP_H +#define ZSTD_CWKSP_H + +/*-************************************* +* Dependencies +***************************************/ +#include "../common/zstd_internal.h" + + +/*-************************************* +* Constants +***************************************/ + +/* Since the workspace is effectively its own little malloc implementation / + * arena, when we run under ASAN, we should similarly insert redzones between + * each internal element of the workspace, so ASAN will catch overruns that + * reach outside an object but that stay inside the workspace. + * + * This defines the size of that redzone. + */ +#ifndef ZSTD_CWKSP_ASAN_REDZONE_SIZE +#define ZSTD_CWKSP_ASAN_REDZONE_SIZE 128 +#endif + +/*-************************************* +* Structures +***************************************/ +typedef enum { + ZSTD_cwksp_alloc_objects, + ZSTD_cwksp_alloc_buffers, + ZSTD_cwksp_alloc_aligned +} ZSTD_cwksp_alloc_phase_e; + +/* + * Used to describe whether the workspace is statically allocated (and will not + * necessarily ever be freed), or if it's dynamically allocated and we can + * expect a well-formed caller to free this. + */ +typedef enum { + ZSTD_cwksp_dynamic_alloc, + ZSTD_cwksp_static_alloc +} ZSTD_cwksp_static_alloc_e; + +/* + * Zstd fits all its internal datastructures into a single continuous buffer, + * so that it only needs to perform a single OS allocation (or so that a buffer + * can be provided to it and it can perform no allocations at all). This buffer + * is called the workspace. + * + * Several optimizations complicate that process of allocating memory ranges + * from this workspace for each internal datastructure: + * + * - These different internal datastructures have different setup requirements: + * + * - The static objects need to be cleared once and can then be trivially + * reused for each compression. + * + * - Various buffers don't need to be initialized at all--they are always + * written into before they're read. + * + * - The matchstate tables have a unique requirement that they don't need + * their memory to be totally cleared, but they do need the memory to have + * some bound, i.e., a guarantee that all values in the memory they've been + * allocated is less than some maximum value (which is the starting value + * for the indices that they will then use for compression). When this + * guarantee is provided to them, they can use the memory without any setup + * work. When it can't, they have to clear the area. + * + * - These buffers also have different alignment requirements. + * + * - We would like to reuse the objects in the workspace for multiple + * compressions without having to perform any expensive reallocation or + * reinitialization work. + * + * - We would like to be able to efficiently reuse the workspace across + * multiple compressions **even when the compression parameters change** and + * we need to resize some of the objects (where possible). + * + * To attempt to manage this buffer, given these constraints, the ZSTD_cwksp + * abstraction was created. It works as follows: + * + * Workspace Layout: + * + * [ ... workspace ... ] + * [objects][tables ... ->] free space [<- ... aligned][<- ... buffers] + * + * The various objects that live in the workspace are divided into the + * following categories, and are allocated separately: + * + * - Static objects: this is optionally the enclosing ZSTD_CCtx or ZSTD_CDict, + * so that literally everything fits in a single buffer. Note: if present, + * this must be the first object in the workspace, since ZSTD_customFree{CCtx, + * CDict}() rely on a pointer comparison to see whether one or two frees are + * required. + * + * - Fixed size objects: these are fixed-size, fixed-count objects that are + * nonetheless "dynamically" allocated in the workspace so that we can + * control how they're initialized separately from the broader ZSTD_CCtx. + * Examples: + * - Entropy Workspace + * - 2 x ZSTD_compressedBlockState_t + * - CDict dictionary contents + * + * - Tables: these are any of several different datastructures (hash tables, + * chain tables, binary trees) that all respect a common format: they are + * uint32_t arrays, all of whose values are between 0 and (nextSrc - base). + * Their sizes depend on the cparams. + * + * - Aligned: these buffers are used for various purposes that require 4 byte + * alignment, but don't require any initialization before they're used. + * + * - Buffers: these buffers are used for various purposes that don't require + * any alignment or initialization before they're used. This means they can + * be moved around at no cost for a new compression. + * + * Allocating Memory: + * + * The various types of objects must be allocated in order, so they can be + * correctly packed into the workspace buffer. That order is: + * + * 1. Objects + * 2. Buffers + * 3. Aligned + * 4. Tables + * + * Attempts to reserve objects of different types out of order will fail. + */ +typedef struct { + void* workspace; + void* workspaceEnd; + + void* objectEnd; + void* tableEnd; + void* tableValidEnd; + void* allocStart; + + BYTE allocFailed; + int workspaceOversizedDuration; + ZSTD_cwksp_alloc_phase_e phase; + ZSTD_cwksp_static_alloc_e isStatic; +} ZSTD_cwksp; + +/*-************************************* +* Functions +***************************************/ + +MEM_STATIC size_t ZSTD_cwksp_available_space(ZSTD_cwksp* ws); + +MEM_STATIC void ZSTD_cwksp_assert_internal_consistency(ZSTD_cwksp* ws) { + (void)ws; + assert(ws->workspace <= ws->objectEnd); + assert(ws->objectEnd <= ws->tableEnd); + assert(ws->objectEnd <= ws->tableValidEnd); + assert(ws->tableEnd <= ws->allocStart); + assert(ws->tableValidEnd <= ws->allocStart); + assert(ws->allocStart <= ws->workspaceEnd); +} + +/* + * Align must be a power of 2. + */ +MEM_STATIC size_t ZSTD_cwksp_align(size_t size, size_t const align) { + size_t const mask = align - 1; + assert((align & mask) == 0); + return (size + mask) & ~mask; +} + +/* + * Use this to determine how much space in the workspace we will consume to + * allocate this object. (Normally it should be exactly the size of the object, + * but under special conditions, like ASAN, where we pad each object, it might + * be larger.) + * + * Since tables aren't currently redzoned, you don't need to call through this + * to figure out how much space you need for the matchState tables. Everything + * else is though. + */ +MEM_STATIC size_t ZSTD_cwksp_alloc_size(size_t size) { + if (size == 0) + return 0; + return size; +} + +MEM_STATIC void ZSTD_cwksp_internal_advance_phase( + ZSTD_cwksp* ws, ZSTD_cwksp_alloc_phase_e phase) { + assert(phase >= ws->phase); + if (phase > ws->phase) { + if (ws->phase < ZSTD_cwksp_alloc_buffers && + phase >= ZSTD_cwksp_alloc_buffers) { + ws->tableValidEnd = ws->objectEnd; + } + if (ws->phase < ZSTD_cwksp_alloc_aligned && + phase >= ZSTD_cwksp_alloc_aligned) { + /* If unaligned allocations down from a too-large top have left us + * unaligned, we need to realign our alloc ptr. Technically, this + * can consume space that is unaccounted for in the neededSpace + * calculation. However, I believe this can only happen when the + * workspace is too large, and specifically when it is too large + * by a larger margin than the space that will be consumed. */ + /* TODO: cleaner, compiler warning friendly way to do this??? */ + ws->allocStart = (BYTE*)ws->allocStart - ((size_t)ws->allocStart & (sizeof(U32)-1)); + if (ws->allocStart < ws->tableValidEnd) { + ws->tableValidEnd = ws->allocStart; + } + } + ws->phase = phase; + } +} + +/* + * Returns whether this object/buffer/etc was allocated in this workspace. + */ +MEM_STATIC int ZSTD_cwksp_owns_buffer(const ZSTD_cwksp* ws, const void* ptr) { + return (ptr != NULL) && (ws->workspace <= ptr) && (ptr <= ws->workspaceEnd); +} + +/* + * Internal function. Do not use directly. + */ +MEM_STATIC void* ZSTD_cwksp_reserve_internal( + ZSTD_cwksp* ws, size_t bytes, ZSTD_cwksp_alloc_phase_e phase) { + void* alloc; + void* bottom = ws->tableEnd; + ZSTD_cwksp_internal_advance_phase(ws, phase); + alloc = (BYTE *)ws->allocStart - bytes; + + if (bytes == 0) + return NULL; + + + DEBUGLOG(5, "cwksp: reserving %p %zd bytes, %zd bytes remaining", + alloc, bytes, ZSTD_cwksp_available_space(ws) - bytes); + ZSTD_cwksp_assert_internal_consistency(ws); + assert(alloc >= bottom); + if (alloc < bottom) { + DEBUGLOG(4, "cwksp: alloc failed!"); + ws->allocFailed = 1; + return NULL; + } + if (alloc < ws->tableValidEnd) { + ws->tableValidEnd = alloc; + } + ws->allocStart = alloc; + + + return alloc; +} + +/* + * Reserves and returns unaligned memory. + */ +MEM_STATIC BYTE* ZSTD_cwksp_reserve_buffer(ZSTD_cwksp* ws, size_t bytes) { + return (BYTE*)ZSTD_cwksp_reserve_internal(ws, bytes, ZSTD_cwksp_alloc_buffers); +} + +/* + * Reserves and returns memory sized on and aligned on sizeof(unsigned). + */ +MEM_STATIC void* ZSTD_cwksp_reserve_aligned(ZSTD_cwksp* ws, size_t bytes) { + assert((bytes & (sizeof(U32)-1)) == 0); + return ZSTD_cwksp_reserve_internal(ws, ZSTD_cwksp_align(bytes, sizeof(U32)), ZSTD_cwksp_alloc_aligned); +} + +/* + * Aligned on sizeof(unsigned). These buffers have the special property that + * their values remain constrained, allowing us to re-use them without + * memset()-ing them. + */ +MEM_STATIC void* ZSTD_cwksp_reserve_table(ZSTD_cwksp* ws, size_t bytes) { + const ZSTD_cwksp_alloc_phase_e phase = ZSTD_cwksp_alloc_aligned; + void* alloc = ws->tableEnd; + void* end = (BYTE *)alloc + bytes; + void* top = ws->allocStart; + + DEBUGLOG(5, "cwksp: reserving %p table %zd bytes, %zd bytes remaining", + alloc, bytes, ZSTD_cwksp_available_space(ws) - bytes); + assert((bytes & (sizeof(U32)-1)) == 0); + ZSTD_cwksp_internal_advance_phase(ws, phase); + ZSTD_cwksp_assert_internal_consistency(ws); + assert(end <= top); + if (end > top) { + DEBUGLOG(4, "cwksp: table alloc failed!"); + ws->allocFailed = 1; + return NULL; + } + ws->tableEnd = end; + + + return alloc; +} + +/* + * Aligned on sizeof(void*). + */ +MEM_STATIC void* ZSTD_cwksp_reserve_object(ZSTD_cwksp* ws, size_t bytes) { + size_t roundedBytes = ZSTD_cwksp_align(bytes, sizeof(void*)); + void* alloc = ws->objectEnd; + void* end = (BYTE*)alloc + roundedBytes; + + + DEBUGLOG(5, + "cwksp: reserving %p object %zd bytes (rounded to %zd), %zd bytes remaining", + alloc, bytes, roundedBytes, ZSTD_cwksp_available_space(ws) - roundedBytes); + assert(((size_t)alloc & (sizeof(void*)-1)) == 0); + assert((bytes & (sizeof(void*)-1)) == 0); + ZSTD_cwksp_assert_internal_consistency(ws); + /* we must be in the first phase, no advance is possible */ + if (ws->phase != ZSTD_cwksp_alloc_objects || end > ws->workspaceEnd) { + DEBUGLOG(4, "cwksp: object alloc failed!"); + ws->allocFailed = 1; + return NULL; + } + ws->objectEnd = end; + ws->tableEnd = end; + ws->tableValidEnd = end; + + + return alloc; +} + +MEM_STATIC void ZSTD_cwksp_mark_tables_dirty(ZSTD_cwksp* ws) { + DEBUGLOG(4, "cwksp: ZSTD_cwksp_mark_tables_dirty"); + + + assert(ws->tableValidEnd >= ws->objectEnd); + assert(ws->tableValidEnd <= ws->allocStart); + ws->tableValidEnd = ws->objectEnd; + ZSTD_cwksp_assert_internal_consistency(ws); +} + +MEM_STATIC void ZSTD_cwksp_mark_tables_clean(ZSTD_cwksp* ws) { + DEBUGLOG(4, "cwksp: ZSTD_cwksp_mark_tables_clean"); + assert(ws->tableValidEnd >= ws->objectEnd); + assert(ws->tableValidEnd <= ws->allocStart); + if (ws->tableValidEnd < ws->tableEnd) { + ws->tableValidEnd = ws->tableEnd; + } + ZSTD_cwksp_assert_internal_consistency(ws); +} + +/* + * Zero the part of the allocated tables not already marked clean. + */ +MEM_STATIC void ZSTD_cwksp_clean_tables(ZSTD_cwksp* ws) { + DEBUGLOG(4, "cwksp: ZSTD_cwksp_clean_tables"); + assert(ws->tableValidEnd >= ws->objectEnd); + assert(ws->tableValidEnd <= ws->allocStart); + if (ws->tableValidEnd < ws->tableEnd) { + ZSTD_memset(ws->tableValidEnd, 0, (BYTE*)ws->tableEnd - (BYTE*)ws->tableValidEnd); + } + ZSTD_cwksp_mark_tables_clean(ws); +} + +/* + * Invalidates table allocations. + * All other allocations remain valid. + */ +MEM_STATIC void ZSTD_cwksp_clear_tables(ZSTD_cwksp* ws) { + DEBUGLOG(4, "cwksp: clearing tables!"); + + + ws->tableEnd = ws->objectEnd; + ZSTD_cwksp_assert_internal_consistency(ws); +} + +/* + * Invalidates all buffer, aligned, and table allocations. + * Object allocations remain valid. + */ +MEM_STATIC void ZSTD_cwksp_clear(ZSTD_cwksp* ws) { + DEBUGLOG(4, "cwksp: clearing!"); + + + + ws->tableEnd = ws->objectEnd; + ws->allocStart = ws->workspaceEnd; + ws->allocFailed = 0; + if (ws->phase > ZSTD_cwksp_alloc_buffers) { + ws->phase = ZSTD_cwksp_alloc_buffers; + } + ZSTD_cwksp_assert_internal_consistency(ws); +} + +/* + * The provided workspace takes ownership of the buffer [start, start+size). + * Any existing values in the workspace are ignored (the previously managed + * buffer, if present, must be separately freed). + */ +MEM_STATIC void ZSTD_cwksp_init(ZSTD_cwksp* ws, void* start, size_t size, ZSTD_cwksp_static_alloc_e isStatic) { + DEBUGLOG(4, "cwksp: init'ing workspace with %zd bytes", size); + assert(((size_t)start & (sizeof(void*)-1)) == 0); /* ensure correct alignment */ + ws->workspace = start; + ws->workspaceEnd = (BYTE*)start + size; + ws->objectEnd = ws->workspace; + ws->tableValidEnd = ws->objectEnd; + ws->phase = ZSTD_cwksp_alloc_objects; + ws->isStatic = isStatic; + ZSTD_cwksp_clear(ws); + ws->workspaceOversizedDuration = 0; + ZSTD_cwksp_assert_internal_consistency(ws); +} + +MEM_STATIC size_t ZSTD_cwksp_create(ZSTD_cwksp* ws, size_t size, ZSTD_customMem customMem) { + void* workspace = ZSTD_customMalloc(size, customMem); + DEBUGLOG(4, "cwksp: creating new workspace with %zd bytes", size); + RETURN_ERROR_IF(workspace == NULL, memory_allocation, "NULL pointer!"); + ZSTD_cwksp_init(ws, workspace, size, ZSTD_cwksp_dynamic_alloc); + return 0; +} + +MEM_STATIC void ZSTD_cwksp_free(ZSTD_cwksp* ws, ZSTD_customMem customMem) { + void *ptr = ws->workspace; + DEBUGLOG(4, "cwksp: freeing workspace"); + ZSTD_memset(ws, 0, sizeof(ZSTD_cwksp)); + ZSTD_customFree(ptr, customMem); +} + +/* + * Moves the management of a workspace from one cwksp to another. The src cwksp + * is left in an invalid state (src must be re-init()'ed before it's used again). + */ +MEM_STATIC void ZSTD_cwksp_move(ZSTD_cwksp* dst, ZSTD_cwksp* src) { + *dst = *src; + ZSTD_memset(src, 0, sizeof(ZSTD_cwksp)); +} + +MEM_STATIC size_t ZSTD_cwksp_sizeof(const ZSTD_cwksp* ws) { + return (size_t)((BYTE*)ws->workspaceEnd - (BYTE*)ws->workspace); +} + +MEM_STATIC size_t ZSTD_cwksp_used(const ZSTD_cwksp* ws) { + return (size_t)((BYTE*)ws->tableEnd - (BYTE*)ws->workspace) + + (size_t)((BYTE*)ws->workspaceEnd - (BYTE*)ws->allocStart); +} + +MEM_STATIC int ZSTD_cwksp_reserve_failed(const ZSTD_cwksp* ws) { + return ws->allocFailed; +} + +/*-************************************* +* Functions Checking Free Space +***************************************/ + +MEM_STATIC size_t ZSTD_cwksp_available_space(ZSTD_cwksp* ws) { + return (size_t)((BYTE*)ws->allocStart - (BYTE*)ws->tableEnd); +} + +MEM_STATIC int ZSTD_cwksp_check_available(ZSTD_cwksp* ws, size_t additionalNeededSpace) { + return ZSTD_cwksp_available_space(ws) >= additionalNeededSpace; +} + +MEM_STATIC int ZSTD_cwksp_check_too_large(ZSTD_cwksp* ws, size_t additionalNeededSpace) { + return ZSTD_cwksp_check_available( + ws, additionalNeededSpace * ZSTD_WORKSPACETOOLARGE_FACTOR); +} + +MEM_STATIC int ZSTD_cwksp_check_wasteful(ZSTD_cwksp* ws, size_t additionalNeededSpace) { + return ZSTD_cwksp_check_too_large(ws, additionalNeededSpace) + && ws->workspaceOversizedDuration > ZSTD_WORKSPACETOOLARGE_MAXDURATION; +} + +MEM_STATIC void ZSTD_cwksp_bump_oversized_duration( + ZSTD_cwksp* ws, size_t additionalNeededSpace) { + if (ZSTD_cwksp_check_too_large(ws, additionalNeededSpace)) { + ws->workspaceOversizedDuration++; + } else { + ws->workspaceOversizedDuration = 0; + } +} + + +#endif /* ZSTD_CWKSP_H */ diff --git a/lib/zstd/compress/zstd_double_fast.c b/lib/zstd/compress/zstd_double_fast.c new file mode 100644 index 000000000000..b0424d23ac57 --- /dev/null +++ b/lib/zstd/compress/zstd_double_fast.c @@ -0,0 +1,519 @@ +/* + * Copyright (c) Yann Collet, Facebook, Inc. + * All rights reserved. + * + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). + * You may select, at your option, one of the above-listed licenses. + */ + +#include "zstd_compress_internal.h" +#include "zstd_double_fast.h" + + +void ZSTD_fillDoubleHashTable(ZSTD_matchState_t* ms, + void const* end, ZSTD_dictTableLoadMethod_e dtlm) +{ + const ZSTD_compressionParameters* const cParams = &ms->cParams; + U32* const hashLarge = ms->hashTable; + U32 const hBitsL = cParams->hashLog; + U32 const mls = cParams->minMatch; + U32* const hashSmall = ms->chainTable; + U32 const hBitsS = cParams->chainLog; + const BYTE* const base = ms->window.base; + const BYTE* ip = base + ms->nextToUpdate; + const BYTE* const iend = ((const BYTE*)end) - HASH_READ_SIZE; + const U32 fastHashFillStep = 3; + + /* Always insert every fastHashFillStep position into the hash tables. + * Insert the other positions into the large hash table if their entry + * is empty. + */ + for (; ip + fastHashFillStep - 1 <= iend; ip += fastHashFillStep) { + U32 const curr = (U32)(ip - base); + U32 i; + for (i = 0; i < fastHashFillStep; ++i) { + size_t const smHash = ZSTD_hashPtr(ip + i, hBitsS, mls); + size_t const lgHash = ZSTD_hashPtr(ip + i, hBitsL, 8); + if (i == 0) + hashSmall[smHash] = curr + i; + if (i == 0 || hashLarge[lgHash] == 0) + hashLarge[lgHash] = curr + i; + /* Only load extra positions for ZSTD_dtlm_full */ + if (dtlm == ZSTD_dtlm_fast) + break; + } } +} + + +FORCE_INLINE_TEMPLATE +size_t ZSTD_compressBlock_doubleFast_generic( + ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM], + void const* src, size_t srcSize, + U32 const mls /* template */, ZSTD_dictMode_e const dictMode) +{ + ZSTD_compressionParameters const* cParams = &ms->cParams; + U32* const hashLong = ms->hashTable; + const U32 hBitsL = cParams->hashLog; + U32* const hashSmall = ms->chainTable; + const U32 hBitsS = cParams->chainLog; + const BYTE* const base = ms->window.base; + const BYTE* const istart = (const BYTE*)src; + const BYTE* ip = istart; + const BYTE* anchor = istart; + const U32 endIndex = (U32)((size_t)(istart - base) + srcSize); + /* presumes that, if there is a dictionary, it must be using Attach mode */ + const U32 prefixLowestIndex = ZSTD_getLowestPrefixIndex(ms, endIndex, cParams->windowLog); + const BYTE* const prefixLowest = base + prefixLowestIndex; + const BYTE* const iend = istart + srcSize; + const BYTE* const ilimit = iend - HASH_READ_SIZE; + U32 offset_1=rep[0], offset_2=rep[1]; + U32 offsetSaved = 0; + + const ZSTD_matchState_t* const dms = ms->dictMatchState; + const ZSTD_compressionParameters* const dictCParams = + dictMode == ZSTD_dictMatchState ? + &dms->cParams : NULL; + const U32* const dictHashLong = dictMode == ZSTD_dictMatchState ? + dms->hashTable : NULL; + const U32* const dictHashSmall = dictMode == ZSTD_dictMatchState ? + dms->chainTable : NULL; + const U32 dictStartIndex = dictMode == ZSTD_dictMatchState ? + dms->window.dictLimit : 0; + const BYTE* const dictBase = dictMode == ZSTD_dictMatchState ? + dms->window.base : NULL; + const BYTE* const dictStart = dictMode == ZSTD_dictMatchState ? + dictBase + dictStartIndex : NULL; + const BYTE* const dictEnd = dictMode == ZSTD_dictMatchState ? + dms->window.nextSrc : NULL; + const U32 dictIndexDelta = dictMode == ZSTD_dictMatchState ? + prefixLowestIndex - (U32)(dictEnd - dictBase) : + 0; + const U32 dictHBitsL = dictMode == ZSTD_dictMatchState ? + dictCParams->hashLog : hBitsL; + const U32 dictHBitsS = dictMode == ZSTD_dictMatchState ? + dictCParams->chainLog : hBitsS; + const U32 dictAndPrefixLength = (U32)((ip - prefixLowest) + (dictEnd - dictStart)); + + DEBUGLOG(5, "ZSTD_compressBlock_doubleFast_generic"); + + assert(dictMode == ZSTD_noDict || dictMode == ZSTD_dictMatchState); + + /* if a dictionary is attached, it must be within window range */ + if (dictMode == ZSTD_dictMatchState) { + assert(ms->window.dictLimit + (1U << cParams->windowLog) >= endIndex); + } + + /* init */ + ip += (dictAndPrefixLength == 0); + if (dictMode == ZSTD_noDict) { + U32 const curr = (U32)(ip - base); + U32 const windowLow = ZSTD_getLowestPrefixIndex(ms, curr, cParams->windowLog); + U32 const maxRep = curr - windowLow; + if (offset_2 > maxRep) offsetSaved = offset_2, offset_2 = 0; + if (offset_1 > maxRep) offsetSaved = offset_1, offset_1 = 0; + } + if (dictMode == ZSTD_dictMatchState) { + /* dictMatchState repCode checks don't currently handle repCode == 0 + * disabling. */ + assert(offset_1 <= dictAndPrefixLength); + assert(offset_2 <= dictAndPrefixLength); + } + + /* Main Search Loop */ + while (ip < ilimit) { /* < instead of <=, because repcode check at (ip+1) */ + size_t mLength; + U32 offset; + size_t const h2 = ZSTD_hashPtr(ip, hBitsL, 8); + size_t const h = ZSTD_hashPtr(ip, hBitsS, mls); + size_t const dictHL = ZSTD_hashPtr(ip, dictHBitsL, 8); + size_t const dictHS = ZSTD_hashPtr(ip, dictHBitsS, mls); + U32 const curr = (U32)(ip-base); + U32 const matchIndexL = hashLong[h2]; + U32 matchIndexS = hashSmall[h]; + const BYTE* matchLong = base + matchIndexL; + const BYTE* match = base + matchIndexS; + const U32 repIndex = curr + 1 - offset_1; + const BYTE* repMatch = (dictMode == ZSTD_dictMatchState + && repIndex < prefixLowestIndex) ? + dictBase + (repIndex - dictIndexDelta) : + base + repIndex; + hashLong[h2] = hashSmall[h] = curr; /* update hash tables */ + + /* check dictMatchState repcode */ + if (dictMode == ZSTD_dictMatchState + && ((U32)((prefixLowestIndex-1) - repIndex) >= 3 /* intentional underflow */) + && (MEM_read32(repMatch) == MEM_read32(ip+1)) ) { + const BYTE* repMatchEnd = repIndex < prefixLowestIndex ? dictEnd : iend; + mLength = ZSTD_count_2segments(ip+1+4, repMatch+4, iend, repMatchEnd, prefixLowest) + 4; + ip++; + ZSTD_storeSeq(seqStore, (size_t)(ip-anchor), anchor, iend, 0, mLength-MINMATCH); + goto _match_stored; + } + + /* check noDict repcode */ + if ( dictMode == ZSTD_noDict + && ((offset_1 > 0) & (MEM_read32(ip+1-offset_1) == MEM_read32(ip+1)))) { + mLength = ZSTD_count(ip+1+4, ip+1+4-offset_1, iend) + 4; + ip++; + ZSTD_storeSeq(seqStore, (size_t)(ip-anchor), anchor, iend, 0, mLength-MINMATCH); + goto _match_stored; + } + + if (matchIndexL > prefixLowestIndex) { + /* check prefix long match */ + if (MEM_read64(matchLong) == MEM_read64(ip)) { + mLength = ZSTD_count(ip+8, matchLong+8, iend) + 8; + offset = (U32)(ip-matchLong); + while (((ip>anchor) & (matchLong>prefixLowest)) && (ip[-1] == matchLong[-1])) { ip--; matchLong--; mLength++; } /* catch up */ + goto _match_found; + } + } else if (dictMode == ZSTD_dictMatchState) { + /* check dictMatchState long match */ + U32 const dictMatchIndexL = dictHashLong[dictHL]; + const BYTE* dictMatchL = dictBase + dictMatchIndexL; + assert(dictMatchL < dictEnd); + + if (dictMatchL > dictStart && MEM_read64(dictMatchL) == MEM_read64(ip)) { + mLength = ZSTD_count_2segments(ip+8, dictMatchL+8, iend, dictEnd, prefixLowest) + 8; + offset = (U32)(curr - dictMatchIndexL - dictIndexDelta); + while (((ip>anchor) & (dictMatchL>dictStart)) && (ip[-1] == dictMatchL[-1])) { ip--; dictMatchL--; mLength++; } /* catch up */ + goto _match_found; + } } + + if (matchIndexS > prefixLowestIndex) { + /* check prefix short match */ + if (MEM_read32(match) == MEM_read32(ip)) { + goto _search_next_long; + } + } else if (dictMode == ZSTD_dictMatchState) { + /* check dictMatchState short match */ + U32 const dictMatchIndexS = dictHashSmall[dictHS]; + match = dictBase + dictMatchIndexS; + matchIndexS = dictMatchIndexS + dictIndexDelta; + + if (match > dictStart && MEM_read32(match) == MEM_read32(ip)) { + goto _search_next_long; + } } + + ip += ((ip-anchor) >> kSearchStrength) + 1; +#if defined(__aarch64__) + PREFETCH_L1(ip+256); +#endif + continue; + +_search_next_long: + + { size_t const hl3 = ZSTD_hashPtr(ip+1, hBitsL, 8); + size_t const dictHLNext = ZSTD_hashPtr(ip+1, dictHBitsL, 8); + U32 const matchIndexL3 = hashLong[hl3]; + const BYTE* matchL3 = base + matchIndexL3; + hashLong[hl3] = curr + 1; + + /* check prefix long +1 match */ + if (matchIndexL3 > prefixLowestIndex) { + if (MEM_read64(matchL3) == MEM_read64(ip+1)) { + mLength = ZSTD_count(ip+9, matchL3+8, iend) + 8; + ip++; + offset = (U32)(ip-matchL3); + while (((ip>anchor) & (matchL3>prefixLowest)) && (ip[-1] == matchL3[-1])) { ip--; matchL3--; mLength++; } /* catch up */ + goto _match_found; + } + } else if (dictMode == ZSTD_dictMatchState) { + /* check dict long +1 match */ + U32 const dictMatchIndexL3 = dictHashLong[dictHLNext]; + const BYTE* dictMatchL3 = dictBase + dictMatchIndexL3; + assert(dictMatchL3 < dictEnd); + if (dictMatchL3 > dictStart && MEM_read64(dictMatchL3) == MEM_read64(ip+1)) { + mLength = ZSTD_count_2segments(ip+1+8, dictMatchL3+8, iend, dictEnd, prefixLowest) + 8; + ip++; + offset = (U32)(curr + 1 - dictMatchIndexL3 - dictIndexDelta); + while (((ip>anchor) & (dictMatchL3>dictStart)) && (ip[-1] == dictMatchL3[-1])) { ip--; dictMatchL3--; mLength++; } /* catch up */ + goto _match_found; + } } } + + /* if no long +1 match, explore the short match we found */ + if (dictMode == ZSTD_dictMatchState && matchIndexS < prefixLowestIndex) { + mLength = ZSTD_count_2segments(ip+4, match+4, iend, dictEnd, prefixLowest) + 4; + offset = (U32)(curr - matchIndexS); + while (((ip>anchor) & (match>dictStart)) && (ip[-1] == match[-1])) { ip--; match--; mLength++; } /* catch up */ + } else { + mLength = ZSTD_count(ip+4, match+4, iend) + 4; + offset = (U32)(ip - match); + while (((ip>anchor) & (match>prefixLowest)) && (ip[-1] == match[-1])) { ip--; match--; mLength++; } /* catch up */ + } + +_match_found: + offset_2 = offset_1; + offset_1 = offset; + + ZSTD_storeSeq(seqStore, (size_t)(ip-anchor), anchor, iend, offset + ZSTD_REP_MOVE, mLength-MINMATCH); + +_match_stored: + /* match found */ + ip += mLength; + anchor = ip; + + if (ip <= ilimit) { + /* Complementary insertion */ + /* done after iLimit test, as candidates could be > iend-8 */ + { U32 const indexToInsert = curr+2; + hashLong[ZSTD_hashPtr(base+indexToInsert, hBitsL, 8)] = indexToInsert; + hashLong[ZSTD_hashPtr(ip-2, hBitsL, 8)] = (U32)(ip-2-base); + hashSmall[ZSTD_hashPtr(base+indexToInsert, hBitsS, mls)] = indexToInsert; + hashSmall[ZSTD_hashPtr(ip-1, hBitsS, mls)] = (U32)(ip-1-base); + } + + /* check immediate repcode */ + if (dictMode == ZSTD_dictMatchState) { + while (ip <= ilimit) { + U32 const current2 = (U32)(ip-base); + U32 const repIndex2 = current2 - offset_2; + const BYTE* repMatch2 = dictMode == ZSTD_dictMatchState + && repIndex2 < prefixLowestIndex ? + dictBase + repIndex2 - dictIndexDelta : + base + repIndex2; + if ( ((U32)((prefixLowestIndex-1) - (U32)repIndex2) >= 3 /* intentional overflow */) + && (MEM_read32(repMatch2) == MEM_read32(ip)) ) { + const BYTE* const repEnd2 = repIndex2 < prefixLowestIndex ? dictEnd : iend; + size_t const repLength2 = ZSTD_count_2segments(ip+4, repMatch2+4, iend, repEnd2, prefixLowest) + 4; + U32 tmpOffset = offset_2; offset_2 = offset_1; offset_1 = tmpOffset; /* swap offset_2 <=> offset_1 */ + ZSTD_storeSeq(seqStore, 0, anchor, iend, 0, repLength2-MINMATCH); + hashSmall[ZSTD_hashPtr(ip, hBitsS, mls)] = current2; + hashLong[ZSTD_hashPtr(ip, hBitsL, 8)] = current2; + ip += repLength2; + anchor = ip; + continue; + } + break; + } } + + if (dictMode == ZSTD_noDict) { + while ( (ip <= ilimit) + && ( (offset_2>0) + & (MEM_read32(ip) == MEM_read32(ip - offset_2)) )) { + /* store sequence */ + size_t const rLength = ZSTD_count(ip+4, ip+4-offset_2, iend) + 4; + U32 const tmpOff = offset_2; offset_2 = offset_1; offset_1 = tmpOff; /* swap offset_2 <=> offset_1 */ + hashSmall[ZSTD_hashPtr(ip, hBitsS, mls)] = (U32)(ip-base); + hashLong[ZSTD_hashPtr(ip, hBitsL, 8)] = (U32)(ip-base); + ZSTD_storeSeq(seqStore, 0, anchor, iend, 0, rLength-MINMATCH); + ip += rLength; + anchor = ip; + continue; /* faster when present ... (?) */ + } } } + } /* while (ip < ilimit) */ + + /* save reps for next block */ + rep[0] = offset_1 ? offset_1 : offsetSaved; + rep[1] = offset_2 ? offset_2 : offsetSaved; + + /* Return the last literals size */ + return (size_t)(iend - anchor); +} + + +size_t ZSTD_compressBlock_doubleFast( + ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM], + void const* src, size_t srcSize) +{ + const U32 mls = ms->cParams.minMatch; + switch(mls) + { + default: /* includes case 3 */ + case 4 : + return ZSTD_compressBlock_doubleFast_generic(ms, seqStore, rep, src, srcSize, 4, ZSTD_noDict); + case 5 : + return ZSTD_compressBlock_doubleFast_generic(ms, seqStore, rep, src, srcSize, 5, ZSTD_noDict); + case 6 : + return ZSTD_compressBlock_doubleFast_generic(ms, seqStore, rep, src, srcSize, 6, ZSTD_noDict); + case 7 : + return ZSTD_compressBlock_doubleFast_generic(ms, seqStore, rep, src, srcSize, 7, ZSTD_noDict); + } +} + + +size_t ZSTD_compressBlock_doubleFast_dictMatchState( + ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM], + void const* src, size_t srcSize) +{ + const U32 mls = ms->cParams.minMatch; + switch(mls) + { + default: /* includes case 3 */ + case 4 : + return ZSTD_compressBlock_doubleFast_generic(ms, seqStore, rep, src, srcSize, 4, ZSTD_dictMatchState); + case 5 : + return ZSTD_compressBlock_doubleFast_generic(ms, seqStore, rep, src, srcSize, 5, ZSTD_dictMatchState); + case 6 : + return ZSTD_compressBlock_doubleFast_generic(ms, seqStore, rep, src, srcSize, 6, ZSTD_dictMatchState); + case 7 : + return ZSTD_compressBlock_doubleFast_generic(ms, seqStore, rep, src, srcSize, 7, ZSTD_dictMatchState); + } +} + + +static size_t ZSTD_compressBlock_doubleFast_extDict_generic( + ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM], + void const* src, size_t srcSize, + U32 const mls /* template */) +{ + ZSTD_compressionParameters const* cParams = &ms->cParams; + U32* const hashLong = ms->hashTable; + U32 const hBitsL = cParams->hashLog; + U32* const hashSmall = ms->chainTable; + U32 const hBitsS = cParams->chainLog; + const BYTE* const istart = (const BYTE*)src; + const BYTE* ip = istart; + const BYTE* anchor = istart; + const BYTE* const iend = istart + srcSize; + const BYTE* const ilimit = iend - 8; + const BYTE* const base = ms->window.base; + const U32 endIndex = (U32)((size_t)(istart - base) + srcSize); + const U32 lowLimit = ZSTD_getLowestMatchIndex(ms, endIndex, cParams->windowLog); + const U32 dictStartIndex = lowLimit; + const U32 dictLimit = ms->window.dictLimit; + const U32 prefixStartIndex = (dictLimit > lowLimit) ? dictLimit : lowLimit; + const BYTE* const prefixStart = base + prefixStartIndex; + const BYTE* const dictBase = ms->window.dictBase; + const BYTE* const dictStart = dictBase + dictStartIndex; + const BYTE* const dictEnd = dictBase + prefixStartIndex; + U32 offset_1=rep[0], offset_2=rep[1]; + + DEBUGLOG(5, "ZSTD_compressBlock_doubleFast_extDict_generic (srcSize=%zu)", srcSize); + + /* if extDict is invalidated due to maxDistance, switch to "regular" variant */ + if (prefixStartIndex == dictStartIndex) + return ZSTD_compressBlock_doubleFast_generic(ms, seqStore, rep, src, srcSize, mls, ZSTD_noDict); + + /* Search Loop */ + while (ip < ilimit) { /* < instead of <=, because (ip+1) */ + const size_t hSmall = ZSTD_hashPtr(ip, hBitsS, mls); + const U32 matchIndex = hashSmall[hSmall]; + const BYTE* const matchBase = matchIndex < prefixStartIndex ? dictBase : base; + const BYTE* match = matchBase + matchIndex; + + const size_t hLong = ZSTD_hashPtr(ip, hBitsL, 8); + const U32 matchLongIndex = hashLong[hLong]; + const BYTE* const matchLongBase = matchLongIndex < prefixStartIndex ? dictBase : base; + const BYTE* matchLong = matchLongBase + matchLongIndex; + + const U32 curr = (U32)(ip-base); + const U32 repIndex = curr + 1 - offset_1; /* offset_1 expected <= curr +1 */ + const BYTE* const repBase = repIndex < prefixStartIndex ? dictBase : base; + const BYTE* const repMatch = repBase + repIndex; + size_t mLength; + hashSmall[hSmall] = hashLong[hLong] = curr; /* update hash table */ + + if ((((U32)((prefixStartIndex-1) - repIndex) >= 3) /* intentional underflow : ensure repIndex doesn't overlap dict + prefix */ + & (repIndex > dictStartIndex)) + && (MEM_read32(repMatch) == MEM_read32(ip+1)) ) { + const BYTE* repMatchEnd = repIndex < prefixStartIndex ? dictEnd : iend; + mLength = ZSTD_count_2segments(ip+1+4, repMatch+4, iend, repMatchEnd, prefixStart) + 4; + ip++; + ZSTD_storeSeq(seqStore, (size_t)(ip-anchor), anchor, iend, 0, mLength-MINMATCH); + } else { + if ((matchLongIndex > dictStartIndex) && (MEM_read64(matchLong) == MEM_read64(ip))) { + const BYTE* const matchEnd = matchLongIndex < prefixStartIndex ? dictEnd : iend; + const BYTE* const lowMatchPtr = matchLongIndex < prefixStartIndex ? dictStart : prefixStart; + U32 offset; + mLength = ZSTD_count_2segments(ip+8, matchLong+8, iend, matchEnd, prefixStart) + 8; + offset = curr - matchLongIndex; + while (((ip>anchor) & (matchLong>lowMatchPtr)) && (ip[-1] == matchLong[-1])) { ip--; matchLong--; mLength++; } /* catch up */ + offset_2 = offset_1; + offset_1 = offset; + ZSTD_storeSeq(seqStore, (size_t)(ip-anchor), anchor, iend, offset + ZSTD_REP_MOVE, mLength-MINMATCH); + + } else if ((matchIndex > dictStartIndex) && (MEM_read32(match) == MEM_read32(ip))) { + size_t const h3 = ZSTD_hashPtr(ip+1, hBitsL, 8); + U32 const matchIndex3 = hashLong[h3]; + const BYTE* const match3Base = matchIndex3 < prefixStartIndex ? dictBase : base; + const BYTE* match3 = match3Base + matchIndex3; + U32 offset; + hashLong[h3] = curr + 1; + if ( (matchIndex3 > dictStartIndex) && (MEM_read64(match3) == MEM_read64(ip+1)) ) { + const BYTE* const matchEnd = matchIndex3 < prefixStartIndex ? dictEnd : iend; + const BYTE* const lowMatchPtr = matchIndex3 < prefixStartIndex ? dictStart : prefixStart; + mLength = ZSTD_count_2segments(ip+9, match3+8, iend, matchEnd, prefixStart) + 8; + ip++; + offset = curr+1 - matchIndex3; + while (((ip>anchor) & (match3>lowMatchPtr)) && (ip[-1] == match3[-1])) { ip--; match3--; mLength++; } /* catch up */ + } else { + const BYTE* const matchEnd = matchIndex < prefixStartIndex ? dictEnd : iend; + const BYTE* const lowMatchPtr = matchIndex < prefixStartIndex ? dictStart : prefixStart; + mLength = ZSTD_count_2segments(ip+4, match+4, iend, matchEnd, prefixStart) + 4; + offset = curr - matchIndex; + while (((ip>anchor) & (match>lowMatchPtr)) && (ip[-1] == match[-1])) { ip--; match--; mLength++; } /* catch up */ + } + offset_2 = offset_1; + offset_1 = offset; + ZSTD_storeSeq(seqStore, (size_t)(ip-anchor), anchor, iend, offset + ZSTD_REP_MOVE, mLength-MINMATCH); + + } else { + ip += ((ip-anchor) >> kSearchStrength) + 1; + continue; + } } + + /* move to next sequence start */ + ip += mLength; + anchor = ip; + + if (ip <= ilimit) { + /* Complementary insertion */ + /* done after iLimit test, as candidates could be > iend-8 */ + { U32 const indexToInsert = curr+2; + hashLong[ZSTD_hashPtr(base+indexToInsert, hBitsL, 8)] = indexToInsert; + hashLong[ZSTD_hashPtr(ip-2, hBitsL, 8)] = (U32)(ip-2-base); + hashSmall[ZSTD_hashPtr(base+indexToInsert, hBitsS, mls)] = indexToInsert; + hashSmall[ZSTD_hashPtr(ip-1, hBitsS, mls)] = (U32)(ip-1-base); + } + + /* check immediate repcode */ + while (ip <= ilimit) { + U32 const current2 = (U32)(ip-base); + U32 const repIndex2 = current2 - offset_2; + const BYTE* repMatch2 = repIndex2 < prefixStartIndex ? dictBase + repIndex2 : base + repIndex2; + if ( (((U32)((prefixStartIndex-1) - repIndex2) >= 3) /* intentional overflow : ensure repIndex2 doesn't overlap dict + prefix */ + & (repIndex2 > dictStartIndex)) + && (MEM_read32(repMatch2) == MEM_read32(ip)) ) { + const BYTE* const repEnd2 = repIndex2 < prefixStartIndex ? dictEnd : iend; + size_t const repLength2 = ZSTD_count_2segments(ip+4, repMatch2+4, iend, repEnd2, prefixStart) + 4; + U32 const tmpOffset = offset_2; offset_2 = offset_1; offset_1 = tmpOffset; /* swap offset_2 <=> offset_1 */ + ZSTD_storeSeq(seqStore, 0, anchor, iend, 0, repLength2-MINMATCH); + hashSmall[ZSTD_hashPtr(ip, hBitsS, mls)] = current2; + hashLong[ZSTD_hashPtr(ip, hBitsL, 8)] = current2; + ip += repLength2; + anchor = ip; + continue; + } + break; + } } } + + /* save reps for next block */ + rep[0] = offset_1; + rep[1] = offset_2; + + /* Return the last literals size */ + return (size_t)(iend - anchor); +} + + +size_t ZSTD_compressBlock_doubleFast_extDict( + ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM], + void const* src, size_t srcSize) +{ + U32 const mls = ms->cParams.minMatch; + switch(mls) + { + default: /* includes case 3 */ + case 4 : + return ZSTD_compressBlock_doubleFast_extDict_generic(ms, seqStore, rep, src, srcSize, 4); + case 5 : + return ZSTD_compressBlock_doubleFast_extDict_generic(ms, seqStore, rep, src, srcSize, 5); + case 6 : + return ZSTD_compressBlock_doubleFast_extDict_generic(ms, seqStore, rep, src, srcSize, 6); + case 7 : + return ZSTD_compressBlock_doubleFast_extDict_generic(ms, seqStore, rep, src, srcSize, 7); + } +} diff --git a/lib/zstd/compress/zstd_double_fast.h b/lib/zstd/compress/zstd_double_fast.h new file mode 100644 index 000000000000..6822bde65a1d --- /dev/null +++ b/lib/zstd/compress/zstd_double_fast.h @@ -0,0 +1,32 @@ +/* + * Copyright (c) Yann Collet, Facebook, Inc. + * All rights reserved. + * + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). + * You may select, at your option, one of the above-listed licenses. + */ + +#ifndef ZSTD_DOUBLE_FAST_H +#define ZSTD_DOUBLE_FAST_H + + +#include "../common/mem.h" /* U32 */ +#include "zstd_compress_internal.h" /* ZSTD_CCtx, size_t */ + +void ZSTD_fillDoubleHashTable(ZSTD_matchState_t* ms, + void const* end, ZSTD_dictTableLoadMethod_e dtlm); +size_t ZSTD_compressBlock_doubleFast( + ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM], + void const* src, size_t srcSize); +size_t ZSTD_compressBlock_doubleFast_dictMatchState( + ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM], + void const* src, size_t srcSize); +size_t ZSTD_compressBlock_doubleFast_extDict( + ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM], + void const* src, size_t srcSize); + + + +#endif /* ZSTD_DOUBLE_FAST_H */ diff --git a/lib/zstd/compress/zstd_fast.c b/lib/zstd/compress/zstd_fast.c new file mode 100644 index 000000000000..96b7d48e2868 --- /dev/null +++ b/lib/zstd/compress/zstd_fast.c @@ -0,0 +1,496 @@ +/* + * Copyright (c) Yann Collet, Facebook, Inc. + * All rights reserved. + * + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). + * You may select, at your option, one of the above-listed licenses. + */ + +#include "zstd_compress_internal.h" /* ZSTD_hashPtr, ZSTD_count, ZSTD_storeSeq */ +#include "zstd_fast.h" + + +void ZSTD_fillHashTable(ZSTD_matchState_t* ms, + const void* const end, + ZSTD_dictTableLoadMethod_e dtlm) +{ + const ZSTD_compressionParameters* const cParams = &ms->cParams; + U32* const hashTable = ms->hashTable; + U32 const hBits = cParams->hashLog; + U32 const mls = cParams->minMatch; + const BYTE* const base = ms->window.base; + const BYTE* ip = base + ms->nextToUpdate; + const BYTE* const iend = ((const BYTE*)end) - HASH_READ_SIZE; + const U32 fastHashFillStep = 3; + + /* Always insert every fastHashFillStep position into the hash table. + * Insert the other positions if their hash entry is empty. + */ + for ( ; ip + fastHashFillStep < iend + 2; ip += fastHashFillStep) { + U32 const curr = (U32)(ip - base); + size_t const hash0 = ZSTD_hashPtr(ip, hBits, mls); + hashTable[hash0] = curr; + if (dtlm == ZSTD_dtlm_fast) continue; + /* Only load extra positions for ZSTD_dtlm_full */ + { U32 p; + for (p = 1; p < fastHashFillStep; ++p) { + size_t const hash = ZSTD_hashPtr(ip + p, hBits, mls); + if (hashTable[hash] == 0) { /* not yet filled */ + hashTable[hash] = curr + p; + } } } } +} + + +FORCE_INLINE_TEMPLATE size_t +ZSTD_compressBlock_fast_generic( + ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM], + void const* src, size_t srcSize, + U32 const mls) +{ + const ZSTD_compressionParameters* const cParams = &ms->cParams; + U32* const hashTable = ms->hashTable; + U32 const hlog = cParams->hashLog; + /* support stepSize of 0 */ + size_t const stepSize = cParams->targetLength + !(cParams->targetLength) + 1; + const BYTE* const base = ms->window.base; + const BYTE* const istart = (const BYTE*)src; + /* We check ip0 (ip + 0) and ip1 (ip + 1) each loop */ + const BYTE* ip0 = istart; + const BYTE* ip1; + const BYTE* anchor = istart; + const U32 endIndex = (U32)((size_t)(istart - base) + srcSize); + const U32 prefixStartIndex = ZSTD_getLowestPrefixIndex(ms, endIndex, cParams->windowLog); + const BYTE* const prefixStart = base + prefixStartIndex; + const BYTE* const iend = istart + srcSize; + const BYTE* const ilimit = iend - HASH_READ_SIZE; + U32 offset_1=rep[0], offset_2=rep[1]; + U32 offsetSaved = 0; + + /* init */ + DEBUGLOG(5, "ZSTD_compressBlock_fast_generic"); + ip0 += (ip0 == prefixStart); + ip1 = ip0 + 1; + { U32 const curr = (U32)(ip0 - base); + U32 const windowLow = ZSTD_getLowestPrefixIndex(ms, curr, cParams->windowLog); + U32 const maxRep = curr - windowLow; + if (offset_2 > maxRep) offsetSaved = offset_2, offset_2 = 0; + if (offset_1 > maxRep) offsetSaved = offset_1, offset_1 = 0; + } + + /* Main Search Loop */ +#ifdef __INTEL_COMPILER + /* From intel 'The vector pragma indicates that the loop should be + * vectorized if it is legal to do so'. Can be used together with + * #pragma ivdep (but have opted to exclude that because intel + * warns against using it).*/ + #pragma vector always +#endif + while (ip1 < ilimit) { /* < instead of <=, because check at ip0+2 */ + size_t mLength; + BYTE const* ip2 = ip0 + 2; + size_t const h0 = ZSTD_hashPtr(ip0, hlog, mls); + U32 const val0 = MEM_read32(ip0); + size_t const h1 = ZSTD_hashPtr(ip1, hlog, mls); + U32 const val1 = MEM_read32(ip1); + U32 const current0 = (U32)(ip0-base); + U32 const current1 = (U32)(ip1-base); + U32 const matchIndex0 = hashTable[h0]; + U32 const matchIndex1 = hashTable[h1]; + BYTE const* repMatch = ip2 - offset_1; + const BYTE* match0 = base + matchIndex0; + const BYTE* match1 = base + matchIndex1; + U32 offcode; + +#if defined(__aarch64__) + PREFETCH_L1(ip0+256); +#endif + + hashTable[h0] = current0; /* update hash table */ + hashTable[h1] = current1; /* update hash table */ + + assert(ip0 + 1 == ip1); + + if ((offset_1 > 0) & (MEM_read32(repMatch) == MEM_read32(ip2))) { + mLength = (ip2[-1] == repMatch[-1]) ? 1 : 0; + ip0 = ip2 - mLength; + match0 = repMatch - mLength; + mLength += 4; + offcode = 0; + goto _match; + } + if ((matchIndex0 > prefixStartIndex) && MEM_read32(match0) == val0) { + /* found a regular match */ + goto _offset; + } + if ((matchIndex1 > prefixStartIndex) && MEM_read32(match1) == val1) { + /* found a regular match after one literal */ + ip0 = ip1; + match0 = match1; + goto _offset; + } + { size_t const step = ((size_t)(ip0-anchor) >> (kSearchStrength - 1)) + stepSize; + assert(step >= 2); + ip0 += step; + ip1 += step; + continue; + } +_offset: /* Requires: ip0, match0 */ + /* Compute the offset code */ + offset_2 = offset_1; + offset_1 = (U32)(ip0-match0); + offcode = offset_1 + ZSTD_REP_MOVE; + mLength = 4; + /* Count the backwards match length */ + while (((ip0>anchor) & (match0>prefixStart)) + && (ip0[-1] == match0[-1])) { ip0--; match0--; mLength++; } /* catch up */ + +_match: /* Requires: ip0, match0, offcode */ + /* Count the forward length */ + mLength += ZSTD_count(ip0+mLength, match0+mLength, iend); + ZSTD_storeSeq(seqStore, (size_t)(ip0-anchor), anchor, iend, offcode, mLength-MINMATCH); + /* match found */ + ip0 += mLength; + anchor = ip0; + + if (ip0 <= ilimit) { + /* Fill Table */ + assert(base+current0+2 > istart); /* check base overflow */ + hashTable[ZSTD_hashPtr(base+current0+2, hlog, mls)] = current0+2; /* here because current+2 could be > iend-8 */ + hashTable[ZSTD_hashPtr(ip0-2, hlog, mls)] = (U32)(ip0-2-base); + + if (offset_2 > 0) { /* offset_2==0 means offset_2 is invalidated */ + while ( (ip0 <= ilimit) && (MEM_read32(ip0) == MEM_read32(ip0 - offset_2)) ) { + /* store sequence */ + size_t const rLength = ZSTD_count(ip0+4, ip0+4-offset_2, iend) + 4; + { U32 const tmpOff = offset_2; offset_2 = offset_1; offset_1 = tmpOff; } /* swap offset_2 <=> offset_1 */ + hashTable[ZSTD_hashPtr(ip0, hlog, mls)] = (U32)(ip0-base); + ip0 += rLength; + ZSTD_storeSeq(seqStore, 0 /*litLen*/, anchor, iend, 0 /*offCode*/, rLength-MINMATCH); + anchor = ip0; + continue; /* faster when present (confirmed on gcc-8) ... (?) */ + } } } + ip1 = ip0 + 1; + } + + /* save reps for next block */ + rep[0] = offset_1 ? offset_1 : offsetSaved; + rep[1] = offset_2 ? offset_2 : offsetSaved; + + /* Return the last literals size */ + return (size_t)(iend - anchor); +} + + +size_t ZSTD_compressBlock_fast( + ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM], + void const* src, size_t srcSize) +{ + U32 const mls = ms->cParams.minMatch; + assert(ms->dictMatchState == NULL); + switch(mls) + { + default: /* includes case 3 */ + case 4 : + return ZSTD_compressBlock_fast_generic(ms, seqStore, rep, src, srcSize, 4); + case 5 : + return ZSTD_compressBlock_fast_generic(ms, seqStore, rep, src, srcSize, 5); + case 6 : + return ZSTD_compressBlock_fast_generic(ms, seqStore, rep, src, srcSize, 6); + case 7 : + return ZSTD_compressBlock_fast_generic(ms, seqStore, rep, src, srcSize, 7); + } +} + +FORCE_INLINE_TEMPLATE +size_t ZSTD_compressBlock_fast_dictMatchState_generic( + ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM], + void const* src, size_t srcSize, U32 const mls) +{ + const ZSTD_compressionParameters* const cParams = &ms->cParams; + U32* const hashTable = ms->hashTable; + U32 const hlog = cParams->hashLog; + /* support stepSize of 0 */ + U32 const stepSize = cParams->targetLength + !(cParams->targetLength); + const BYTE* const base = ms->window.base; + const BYTE* const istart = (const BYTE*)src; + const BYTE* ip = istart; + const BYTE* anchor = istart; + const U32 prefixStartIndex = ms->window.dictLimit; + const BYTE* const prefixStart = base + prefixStartIndex; + const BYTE* const iend = istart + srcSize; + const BYTE* const ilimit = iend - HASH_READ_SIZE; + U32 offset_1=rep[0], offset_2=rep[1]; + U32 offsetSaved = 0; + + const ZSTD_matchState_t* const dms = ms->dictMatchState; + const ZSTD_compressionParameters* const dictCParams = &dms->cParams ; + const U32* const dictHashTable = dms->hashTable; + const U32 dictStartIndex = dms->window.dictLimit; + const BYTE* const dictBase = dms->window.base; + const BYTE* const dictStart = dictBase + dictStartIndex; + const BYTE* const dictEnd = dms->window.nextSrc; + const U32 dictIndexDelta = prefixStartIndex - (U32)(dictEnd - dictBase); + const U32 dictAndPrefixLength = (U32)(ip - prefixStart + dictEnd - dictStart); + const U32 dictHLog = dictCParams->hashLog; + + /* if a dictionary is still attached, it necessarily means that + * it is within window size. So we just check it. */ + const U32 maxDistance = 1U << cParams->windowLog; + const U32 endIndex = (U32)((size_t)(ip - base) + srcSize); + assert(endIndex - prefixStartIndex <= maxDistance); + (void)maxDistance; (void)endIndex; /* these variables are not used when assert() is disabled */ + + /* ensure there will be no underflow + * when translating a dict index into a local index */ + assert(prefixStartIndex >= (U32)(dictEnd - dictBase)); + + /* init */ + DEBUGLOG(5, "ZSTD_compressBlock_fast_dictMatchState_generic"); + ip += (dictAndPrefixLength == 0); + /* dictMatchState repCode checks don't currently handle repCode == 0 + * disabling. */ + assert(offset_1 <= dictAndPrefixLength); + assert(offset_2 <= dictAndPrefixLength); + + /* Main Search Loop */ + while (ip < ilimit) { /* < instead of <=, because repcode check at (ip+1) */ + size_t mLength; + size_t const h = ZSTD_hashPtr(ip, hlog, mls); + U32 const curr = (U32)(ip-base); + U32 const matchIndex = hashTable[h]; + const BYTE* match = base + matchIndex; + const U32 repIndex = curr + 1 - offset_1; + const BYTE* repMatch = (repIndex < prefixStartIndex) ? + dictBase + (repIndex - dictIndexDelta) : + base + repIndex; + hashTable[h] = curr; /* update hash table */ + + if ( ((U32)((prefixStartIndex-1) - repIndex) >= 3) /* intentional underflow : ensure repIndex isn't overlapping dict + prefix */ + && (MEM_read32(repMatch) == MEM_read32(ip+1)) ) { + const BYTE* const repMatchEnd = repIndex < prefixStartIndex ? dictEnd : iend; + mLength = ZSTD_count_2segments(ip+1+4, repMatch+4, iend, repMatchEnd, prefixStart) + 4; + ip++; + ZSTD_storeSeq(seqStore, (size_t)(ip-anchor), anchor, iend, 0, mLength-MINMATCH); + } else if ( (matchIndex <= prefixStartIndex) ) { + size_t const dictHash = ZSTD_hashPtr(ip, dictHLog, mls); + U32 const dictMatchIndex = dictHashTable[dictHash]; + const BYTE* dictMatch = dictBase + dictMatchIndex; + if (dictMatchIndex <= dictStartIndex || + MEM_read32(dictMatch) != MEM_read32(ip)) { + assert(stepSize >= 1); + ip += ((ip-anchor) >> kSearchStrength) + stepSize; + continue; + } else { + /* found a dict match */ + U32 const offset = (U32)(curr-dictMatchIndex-dictIndexDelta); + mLength = ZSTD_count_2segments(ip+4, dictMatch+4, iend, dictEnd, prefixStart) + 4; + while (((ip>anchor) & (dictMatch>dictStart)) + && (ip[-1] == dictMatch[-1])) { + ip--; dictMatch--; mLength++; + } /* catch up */ + offset_2 = offset_1; + offset_1 = offset; + ZSTD_storeSeq(seqStore, (size_t)(ip-anchor), anchor, iend, offset + ZSTD_REP_MOVE, mLength-MINMATCH); + } + } else if (MEM_read32(match) != MEM_read32(ip)) { + /* it's not a match, and we're not going to check the dictionary */ + assert(stepSize >= 1); + ip += ((ip-anchor) >> kSearchStrength) + stepSize; + continue; + } else { + /* found a regular match */ + U32 const offset = (U32)(ip-match); + mLength = ZSTD_count(ip+4, match+4, iend) + 4; + while (((ip>anchor) & (match>prefixStart)) + && (ip[-1] == match[-1])) { ip--; match--; mLength++; } /* catch up */ + offset_2 = offset_1; + offset_1 = offset; + ZSTD_storeSeq(seqStore, (size_t)(ip-anchor), anchor, iend, offset + ZSTD_REP_MOVE, mLength-MINMATCH); + } + + /* match found */ + ip += mLength; + anchor = ip; + + if (ip <= ilimit) { + /* Fill Table */ + assert(base+curr+2 > istart); /* check base overflow */ + hashTable[ZSTD_hashPtr(base+curr+2, hlog, mls)] = curr+2; /* here because curr+2 could be > iend-8 */ + hashTable[ZSTD_hashPtr(ip-2, hlog, mls)] = (U32)(ip-2-base); + + /* check immediate repcode */ + while (ip <= ilimit) { + U32 const current2 = (U32)(ip-base); + U32 const repIndex2 = current2 - offset_2; + const BYTE* repMatch2 = repIndex2 < prefixStartIndex ? + dictBase - dictIndexDelta + repIndex2 : + base + repIndex2; + if ( ((U32)((prefixStartIndex-1) - (U32)repIndex2) >= 3 /* intentional overflow */) + && (MEM_read32(repMatch2) == MEM_read32(ip)) ) { + const BYTE* const repEnd2 = repIndex2 < prefixStartIndex ? dictEnd : iend; + size_t const repLength2 = ZSTD_count_2segments(ip+4, repMatch2+4, iend, repEnd2, prefixStart) + 4; + U32 tmpOffset = offset_2; offset_2 = offset_1; offset_1 = tmpOffset; /* swap offset_2 <=> offset_1 */ + ZSTD_storeSeq(seqStore, 0, anchor, iend, 0, repLength2-MINMATCH); + hashTable[ZSTD_hashPtr(ip, hlog, mls)] = current2; + ip += repLength2; + anchor = ip; + continue; + } + break; + } + } + } + + /* save reps for next block */ + rep[0] = offset_1 ? offset_1 : offsetSaved; + rep[1] = offset_2 ? offset_2 : offsetSaved; + + /* Return the last literals size */ + return (size_t)(iend - anchor); +} + +size_t ZSTD_compressBlock_fast_dictMatchState( + ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM], + void const* src, size_t srcSize) +{ + U32 const mls = ms->cParams.minMatch; + assert(ms->dictMatchState != NULL); + switch(mls) + { + default: /* includes case 3 */ + case 4 : + return ZSTD_compressBlock_fast_dictMatchState_generic(ms, seqStore, rep, src, srcSize, 4); + case 5 : + return ZSTD_compressBlock_fast_dictMatchState_generic(ms, seqStore, rep, src, srcSize, 5); + case 6 : + return ZSTD_compressBlock_fast_dictMatchState_generic(ms, seqStore, rep, src, srcSize, 6); + case 7 : + return ZSTD_compressBlock_fast_dictMatchState_generic(ms, seqStore, rep, src, srcSize, 7); + } +} + + +static size_t ZSTD_compressBlock_fast_extDict_generic( + ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM], + void const* src, size_t srcSize, U32 const mls) +{ + const ZSTD_compressionParameters* const cParams = &ms->cParams; + U32* const hashTable = ms->hashTable; + U32 const hlog = cParams->hashLog; + /* support stepSize of 0 */ + U32 const stepSize = cParams->targetLength + !(cParams->targetLength); + const BYTE* const base = ms->window.base; + const BYTE* const dictBase = ms->window.dictBase; + const BYTE* const istart = (const BYTE*)src; + const BYTE* ip = istart; + const BYTE* anchor = istart; + const U32 endIndex = (U32)((size_t)(istart - base) + srcSize); + const U32 lowLimit = ZSTD_getLowestMatchIndex(ms, endIndex, cParams->windowLog); + const U32 dictStartIndex = lowLimit; + const BYTE* const dictStart = dictBase + dictStartIndex; + const U32 dictLimit = ms->window.dictLimit; + const U32 prefixStartIndex = dictLimit < lowLimit ? lowLimit : dictLimit; + const BYTE* const prefixStart = base + prefixStartIndex; + const BYTE* const dictEnd = dictBase + prefixStartIndex; + const BYTE* const iend = istart + srcSize; + const BYTE* const ilimit = iend - 8; + U32 offset_1=rep[0], offset_2=rep[1]; + + DEBUGLOG(5, "ZSTD_compressBlock_fast_extDict_generic (offset_1=%u)", offset_1); + + /* switch to "regular" variant if extDict is invalidated due to maxDistance */ + if (prefixStartIndex == dictStartIndex) + return ZSTD_compressBlock_fast_generic(ms, seqStore, rep, src, srcSize, mls); + + /* Search Loop */ + while (ip < ilimit) { /* < instead of <=, because (ip+1) */ + const size_t h = ZSTD_hashPtr(ip, hlog, mls); + const U32 matchIndex = hashTable[h]; + const BYTE* const matchBase = matchIndex < prefixStartIndex ? dictBase : base; + const BYTE* match = matchBase + matchIndex; + const U32 curr = (U32)(ip-base); + const U32 repIndex = curr + 1 - offset_1; + const BYTE* const repBase = repIndex < prefixStartIndex ? dictBase : base; + const BYTE* const repMatch = repBase + repIndex; + hashTable[h] = curr; /* update hash table */ + DEBUGLOG(7, "offset_1 = %u , curr = %u", offset_1, curr); + assert(offset_1 <= curr +1); /* check repIndex */ + + if ( (((U32)((prefixStartIndex-1) - repIndex) >= 3) /* intentional underflow */ & (repIndex > dictStartIndex)) + && (MEM_read32(repMatch) == MEM_read32(ip+1)) ) { + const BYTE* const repMatchEnd = repIndex < prefixStartIndex ? dictEnd : iend; + size_t const rLength = ZSTD_count_2segments(ip+1 +4, repMatch +4, iend, repMatchEnd, prefixStart) + 4; + ip++; + ZSTD_storeSeq(seqStore, (size_t)(ip-anchor), anchor, iend, 0, rLength-MINMATCH); + ip += rLength; + anchor = ip; + } else { + if ( (matchIndex < dictStartIndex) || + (MEM_read32(match) != MEM_read32(ip)) ) { + assert(stepSize >= 1); + ip += ((ip-anchor) >> kSearchStrength) + stepSize; + continue; + } + { const BYTE* const matchEnd = matchIndex < prefixStartIndex ? dictEnd : iend; + const BYTE* const lowMatchPtr = matchIndex < prefixStartIndex ? dictStart : prefixStart; + U32 const offset = curr - matchIndex; + size_t mLength = ZSTD_count_2segments(ip+4, match+4, iend, matchEnd, prefixStart) + 4; + while (((ip>anchor) & (match>lowMatchPtr)) && (ip[-1] == match[-1])) { ip--; match--; mLength++; } /* catch up */ + offset_2 = offset_1; offset_1 = offset; /* update offset history */ + ZSTD_storeSeq(seqStore, (size_t)(ip-anchor), anchor, iend, offset + ZSTD_REP_MOVE, mLength-MINMATCH); + ip += mLength; + anchor = ip; + } } + + if (ip <= ilimit) { + /* Fill Table */ + hashTable[ZSTD_hashPtr(base+curr+2, hlog, mls)] = curr+2; + hashTable[ZSTD_hashPtr(ip-2, hlog, mls)] = (U32)(ip-2-base); + /* check immediate repcode */ + while (ip <= ilimit) { + U32 const current2 = (U32)(ip-base); + U32 const repIndex2 = current2 - offset_2; + const BYTE* const repMatch2 = repIndex2 < prefixStartIndex ? dictBase + repIndex2 : base + repIndex2; + if ( (((U32)((prefixStartIndex-1) - repIndex2) >= 3) & (repIndex2 > dictStartIndex)) /* intentional overflow */ + && (MEM_read32(repMatch2) == MEM_read32(ip)) ) { + const BYTE* const repEnd2 = repIndex2 < prefixStartIndex ? dictEnd : iend; + size_t const repLength2 = ZSTD_count_2segments(ip+4, repMatch2+4, iend, repEnd2, prefixStart) + 4; + { U32 const tmpOffset = offset_2; offset_2 = offset_1; offset_1 = tmpOffset; } /* swap offset_2 <=> offset_1 */ + ZSTD_storeSeq(seqStore, 0 /*litlen*/, anchor, iend, 0 /*offcode*/, repLength2-MINMATCH); + hashTable[ZSTD_hashPtr(ip, hlog, mls)] = current2; + ip += repLength2; + anchor = ip; + continue; + } + break; + } } } + + /* save reps for next block */ + rep[0] = offset_1; + rep[1] = offset_2; + + /* Return the last literals size */ + return (size_t)(iend - anchor); +} + + +size_t ZSTD_compressBlock_fast_extDict( + ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM], + void const* src, size_t srcSize) +{ + U32 const mls = ms->cParams.minMatch; + switch(mls) + { + default: /* includes case 3 */ + case 4 : + return ZSTD_compressBlock_fast_extDict_generic(ms, seqStore, rep, src, srcSize, 4); + case 5 : + return ZSTD_compressBlock_fast_extDict_generic(ms, seqStore, rep, src, srcSize, 5); + case 6 : + return ZSTD_compressBlock_fast_extDict_generic(ms, seqStore, rep, src, srcSize, 6); + case 7 : + return ZSTD_compressBlock_fast_extDict_generic(ms, seqStore, rep, src, srcSize, 7); + } +} diff --git a/lib/zstd/compress/zstd_fast.h b/lib/zstd/compress/zstd_fast.h new file mode 100644 index 000000000000..fddc2f532d21 --- /dev/null +++ b/lib/zstd/compress/zstd_fast.h @@ -0,0 +1,31 @@ +/* + * Copyright (c) Yann Collet, Facebook, Inc. + * All rights reserved. + * + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). + * You may select, at your option, one of the above-listed licenses. + */ + +#ifndef ZSTD_FAST_H +#define ZSTD_FAST_H + + +#include "../common/mem.h" /* U32 */ +#include "zstd_compress_internal.h" + +void ZSTD_fillHashTable(ZSTD_matchState_t* ms, + void const* end, ZSTD_dictTableLoadMethod_e dtlm); +size_t ZSTD_compressBlock_fast( + ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM], + void const* src, size_t srcSize); +size_t ZSTD_compressBlock_fast_dictMatchState( + ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM], + void const* src, size_t srcSize); +size_t ZSTD_compressBlock_fast_extDict( + ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM], + void const* src, size_t srcSize); + + +#endif /* ZSTD_FAST_H */ diff --git a/lib/zstd/compress/zstd_lazy.c b/lib/zstd/compress/zstd_lazy.c new file mode 100644 index 000000000000..fb54d4e28a2b --- /dev/null +++ b/lib/zstd/compress/zstd_lazy.c @@ -0,0 +1,1414 @@ +/* + * Copyright (c) Yann Collet, Facebook, Inc. + * All rights reserved. + * + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). + * You may select, at your option, one of the above-listed licenses. + */ + +#include "zstd_compress_internal.h" +#include "zstd_lazy.h" + + +/*-************************************* +* Binary Tree search +***************************************/ + +static void +ZSTD_updateDUBT(ZSTD_matchState_t* ms, + const BYTE* ip, const BYTE* iend, + U32 mls) +{ + const ZSTD_compressionParameters* const cParams = &ms->cParams; + U32* const hashTable = ms->hashTable; + U32 const hashLog = cParams->hashLog; + + U32* const bt = ms->chainTable; + U32 const btLog = cParams->chainLog - 1; + U32 const btMask = (1 << btLog) - 1; + + const BYTE* const base = ms->window.base; + U32 const target = (U32)(ip - base); + U32 idx = ms->nextToUpdate; + + if (idx != target) + DEBUGLOG(7, "ZSTD_updateDUBT, from %u to %u (dictLimit:%u)", + idx, target, ms->window.dictLimit); + assert(ip + 8 <= iend); /* condition for ZSTD_hashPtr */ + (void)iend; + + assert(idx >= ms->window.dictLimit); /* condition for valid base+idx */ + for ( ; idx < target ; idx++) { + size_t const h = ZSTD_hashPtr(base + idx, hashLog, mls); /* assumption : ip + 8 <= iend */ + U32 const matchIndex = hashTable[h]; + + U32* const nextCandidatePtr = bt + 2*(idx&btMask); + U32* const sortMarkPtr = nextCandidatePtr + 1; + + DEBUGLOG(8, "ZSTD_updateDUBT: insert %u", idx); + hashTable[h] = idx; /* Update Hash Table */ + *nextCandidatePtr = matchIndex; /* update BT like a chain */ + *sortMarkPtr = ZSTD_DUBT_UNSORTED_MARK; + } + ms->nextToUpdate = target; +} + + +/* ZSTD_insertDUBT1() : + * sort one already inserted but unsorted position + * assumption : curr >= btlow == (curr - btmask) + * doesn't fail */ +static void +ZSTD_insertDUBT1(ZSTD_matchState_t* ms, + U32 curr, const BYTE* inputEnd, + U32 nbCompares, U32 btLow, + const ZSTD_dictMode_e dictMode) +{ + const ZSTD_compressionParameters* const cParams = &ms->cParams; + U32* const bt = ms->chainTable; + U32 const btLog = cParams->chainLog - 1; + U32 const btMask = (1 << btLog) - 1; + size_t commonLengthSmaller=0, commonLengthLarger=0; + const BYTE* const base = ms->window.base; + const BYTE* const dictBase = ms->window.dictBase; + const U32 dictLimit = ms->window.dictLimit; + const BYTE* const ip = (curr>=dictLimit) ? base + curr : dictBase + curr; + const BYTE* const iend = (curr>=dictLimit) ? inputEnd : dictBase + dictLimit; + const BYTE* const dictEnd = dictBase + dictLimit; + const BYTE* const prefixStart = base + dictLimit; + const BYTE* match; + U32* smallerPtr = bt + 2*(curr&btMask); + U32* largerPtr = smallerPtr + 1; + U32 matchIndex = *smallerPtr; /* this candidate is unsorted : next sorted candidate is reached through *smallerPtr, while *largerPtr contains previous unsorted candidate (which is already saved and can be overwritten) */ + U32 dummy32; /* to be nullified at the end */ + U32 const windowValid = ms->window.lowLimit; + U32 const maxDistance = 1U << cParams->windowLog; + U32 const windowLow = (curr - windowValid > maxDistance) ? curr - maxDistance : windowValid; + + + DEBUGLOG(8, "ZSTD_insertDUBT1(%u) (dictLimit=%u, lowLimit=%u)", + curr, dictLimit, windowLow); + assert(curr >= btLow); + assert(ip < iend); /* condition for ZSTD_count */ + + for (; nbCompares && (matchIndex > windowLow); --nbCompares) { + U32* const nextPtr = bt + 2*(matchIndex & btMask); + size_t matchLength = MIN(commonLengthSmaller, commonLengthLarger); /* guaranteed minimum nb of common bytes */ + assert(matchIndex < curr); + /* note : all candidates are now supposed sorted, + * but it's still possible to have nextPtr[1] == ZSTD_DUBT_UNSORTED_MARK + * when a real index has the same value as ZSTD_DUBT_UNSORTED_MARK */ + + if ( (dictMode != ZSTD_extDict) + || (matchIndex+matchLength >= dictLimit) /* both in current segment*/ + || (curr < dictLimit) /* both in extDict */) { + const BYTE* const mBase = ( (dictMode != ZSTD_extDict) + || (matchIndex+matchLength >= dictLimit)) ? + base : dictBase; + assert( (matchIndex+matchLength >= dictLimit) /* might be wrong if extDict is incorrectly set to 0 */ + || (curr < dictLimit) ); + match = mBase + matchIndex; + matchLength += ZSTD_count(ip+matchLength, match+matchLength, iend); + } else { + match = dictBase + matchIndex; + matchLength += ZSTD_count_2segments(ip+matchLength, match+matchLength, iend, dictEnd, prefixStart); + if (matchIndex+matchLength >= dictLimit) + match = base + matchIndex; /* preparation for next read of match[matchLength] */ + } + + DEBUGLOG(8, "ZSTD_insertDUBT1: comparing %u with %u : found %u common bytes ", + curr, matchIndex, (U32)matchLength); + + if (ip+matchLength == iend) { /* equal : no way to know if inf or sup */ + break; /* drop , to guarantee consistency ; miss a bit of compression, but other solutions can corrupt tree */ + } + + if (match[matchLength] < ip[matchLength]) { /* necessarily within buffer */ + /* match is smaller than current */ + *smallerPtr = matchIndex; /* update smaller idx */ + commonLengthSmaller = matchLength; /* all smaller will now have at least this guaranteed common length */ + if (matchIndex <= btLow) { smallerPtr=&dummy32; break; } /* beyond tree size, stop searching */ + DEBUGLOG(8, "ZSTD_insertDUBT1: %u (>btLow=%u) is smaller : next => %u", + matchIndex, btLow, nextPtr[1]); + smallerPtr = nextPtr+1; /* new "candidate" => larger than match, which was smaller than target */ + matchIndex = nextPtr[1]; /* new matchIndex, larger than previous and closer to current */ + } else { + /* match is larger than current */ + *largerPtr = matchIndex; + commonLengthLarger = matchLength; + if (matchIndex <= btLow) { largerPtr=&dummy32; break; } /* beyond tree size, stop searching */ + DEBUGLOG(8, "ZSTD_insertDUBT1: %u (>btLow=%u) is larger => %u", + matchIndex, btLow, nextPtr[0]); + largerPtr = nextPtr; + matchIndex = nextPtr[0]; + } } + + *smallerPtr = *largerPtr = 0; +} + + +static size_t +ZSTD_DUBT_findBetterDictMatch ( + ZSTD_matchState_t* ms, + const BYTE* const ip, const BYTE* const iend, + size_t* offsetPtr, + size_t bestLength, + U32 nbCompares, + U32 const mls, + const ZSTD_dictMode_e dictMode) +{ + const ZSTD_matchState_t * const dms = ms->dictMatchState; + const ZSTD_compressionParameters* const dmsCParams = &dms->cParams; + const U32 * const dictHashTable = dms->hashTable; + U32 const hashLog = dmsCParams->hashLog; + size_t const h = ZSTD_hashPtr(ip, hashLog, mls); + U32 dictMatchIndex = dictHashTable[h]; + + const BYTE* const base = ms->window.base; + const BYTE* const prefixStart = base + ms->window.dictLimit; + U32 const curr = (U32)(ip-base); + const BYTE* const dictBase = dms->window.base; + const BYTE* const dictEnd = dms->window.nextSrc; + U32 const dictHighLimit = (U32)(dms->window.nextSrc - dms->window.base); + U32 const dictLowLimit = dms->window.lowLimit; + U32 const dictIndexDelta = ms->window.lowLimit - dictHighLimit; + + U32* const dictBt = dms->chainTable; + U32 const btLog = dmsCParams->chainLog - 1; + U32 const btMask = (1 << btLog) - 1; + U32 const btLow = (btMask >= dictHighLimit - dictLowLimit) ? dictLowLimit : dictHighLimit - btMask; + + size_t commonLengthSmaller=0, commonLengthLarger=0; + + (void)dictMode; + assert(dictMode == ZSTD_dictMatchState); + + for (; nbCompares && (dictMatchIndex > dictLowLimit); --nbCompares) { + U32* const nextPtr = dictBt + 2*(dictMatchIndex & btMask); + size_t matchLength = MIN(commonLengthSmaller, commonLengthLarger); /* guaranteed minimum nb of common bytes */ + const BYTE* match = dictBase + dictMatchIndex; + matchLength += ZSTD_count_2segments(ip+matchLength, match+matchLength, iend, dictEnd, prefixStart); + if (dictMatchIndex+matchLength >= dictHighLimit) + match = base + dictMatchIndex + dictIndexDelta; /* to prepare for next usage of match[matchLength] */ + + if (matchLength > bestLength) { + U32 matchIndex = dictMatchIndex + dictIndexDelta; + if ( (4*(int)(matchLength-bestLength)) > (int)(ZSTD_highbit32(curr-matchIndex+1) - ZSTD_highbit32((U32)offsetPtr[0]+1)) ) { + DEBUGLOG(9, "ZSTD_DUBT_findBetterDictMatch(%u) : found better match length %u -> %u and offsetCode %u -> %u (dictMatchIndex %u, matchIndex %u)", + curr, (U32)bestLength, (U32)matchLength, (U32)*offsetPtr, ZSTD_REP_MOVE + curr - matchIndex, dictMatchIndex, matchIndex); + bestLength = matchLength, *offsetPtr = ZSTD_REP_MOVE + curr - matchIndex; + } + if (ip+matchLength == iend) { /* reached end of input : ip[matchLength] is not valid, no way to know if it's larger or smaller than match */ + break; /* drop, to guarantee consistency (miss a little bit of compression) */ + } + } + + if (match[matchLength] < ip[matchLength]) { + if (dictMatchIndex <= btLow) { break; } /* beyond tree size, stop the search */ + commonLengthSmaller = matchLength; /* all smaller will now have at least this guaranteed common length */ + dictMatchIndex = nextPtr[1]; /* new matchIndex larger than previous (closer to current) */ + } else { + /* match is larger than current */ + if (dictMatchIndex <= btLow) { break; } /* beyond tree size, stop the search */ + commonLengthLarger = matchLength; + dictMatchIndex = nextPtr[0]; + } + } + + if (bestLength >= MINMATCH) { + U32 const mIndex = curr - ((U32)*offsetPtr - ZSTD_REP_MOVE); (void)mIndex; + DEBUGLOG(8, "ZSTD_DUBT_findBetterDictMatch(%u) : found match of length %u and offsetCode %u (pos %u)", + curr, (U32)bestLength, (U32)*offsetPtr, mIndex); + } + return bestLength; + +} + + +static size_t +ZSTD_DUBT_findBestMatch(ZSTD_matchState_t* ms, + const BYTE* const ip, const BYTE* const iend, + size_t* offsetPtr, + U32 const mls, + const ZSTD_dictMode_e dictMode) +{ + const ZSTD_compressionParameters* const cParams = &ms->cParams; + U32* const hashTable = ms->hashTable; + U32 const hashLog = cParams->hashLog; + size_t const h = ZSTD_hashPtr(ip, hashLog, mls); + U32 matchIndex = hashTable[h]; + + const BYTE* const base = ms->window.base; + U32 const curr = (U32)(ip-base); + U32 const windowLow = ZSTD_getLowestMatchIndex(ms, curr, cParams->windowLog); + + U32* const bt = ms->chainTable; + U32 const btLog = cParams->chainLog - 1; + U32 const btMask = (1 << btLog) - 1; + U32 const btLow = (btMask >= curr) ? 0 : curr - btMask; + U32 const unsortLimit = MAX(btLow, windowLow); + + U32* nextCandidate = bt + 2*(matchIndex&btMask); + U32* unsortedMark = bt + 2*(matchIndex&btMask) + 1; + U32 nbCompares = 1U << cParams->searchLog; + U32 nbCandidates = nbCompares; + U32 previousCandidate = 0; + + DEBUGLOG(7, "ZSTD_DUBT_findBestMatch (%u) ", curr); + assert(ip <= iend-8); /* required for h calculation */ + assert(dictMode != ZSTD_dedicatedDictSearch); + + /* reach end of unsorted candidates list */ + while ( (matchIndex > unsortLimit) + && (*unsortedMark == ZSTD_DUBT_UNSORTED_MARK) + && (nbCandidates > 1) ) { + DEBUGLOG(8, "ZSTD_DUBT_findBestMatch: candidate %u is unsorted", + matchIndex); + *unsortedMark = previousCandidate; /* the unsortedMark becomes a reversed chain, to move up back to original position */ + previousCandidate = matchIndex; + matchIndex = *nextCandidate; + nextCandidate = bt + 2*(matchIndex&btMask); + unsortedMark = bt + 2*(matchIndex&btMask) + 1; + nbCandidates --; + } + + /* nullify last candidate if it's still unsorted + * simplification, detrimental to compression ratio, beneficial for speed */ + if ( (matchIndex > unsortLimit) + && (*unsortedMark==ZSTD_DUBT_UNSORTED_MARK) ) { + DEBUGLOG(7, "ZSTD_DUBT_findBestMatch: nullify last unsorted candidate %u", + matchIndex); + *nextCandidate = *unsortedMark = 0; + } + + /* batch sort stacked candidates */ + matchIndex = previousCandidate; + while (matchIndex) { /* will end on matchIndex == 0 */ + U32* const nextCandidateIdxPtr = bt + 2*(matchIndex&btMask) + 1; + U32 const nextCandidateIdx = *nextCandidateIdxPtr; + ZSTD_insertDUBT1(ms, matchIndex, iend, + nbCandidates, unsortLimit, dictMode); + matchIndex = nextCandidateIdx; + nbCandidates++; + } + + /* find longest match */ + { size_t commonLengthSmaller = 0, commonLengthLarger = 0; + const BYTE* const dictBase = ms->window.dictBase; + const U32 dictLimit = ms->window.dictLimit; + const BYTE* const dictEnd = dictBase + dictLimit; + const BYTE* const prefixStart = base + dictLimit; + U32* smallerPtr = bt + 2*(curr&btMask); + U32* largerPtr = bt + 2*(curr&btMask) + 1; + U32 matchEndIdx = curr + 8 + 1; + U32 dummy32; /* to be nullified at the end */ + size_t bestLength = 0; + + matchIndex = hashTable[h]; + hashTable[h] = curr; /* Update Hash Table */ + + for (; nbCompares && (matchIndex > windowLow); --nbCompares) { + U32* const nextPtr = bt + 2*(matchIndex & btMask); + size_t matchLength = MIN(commonLengthSmaller, commonLengthLarger); /* guaranteed minimum nb of common bytes */ + const BYTE* match; + + if ((dictMode != ZSTD_extDict) || (matchIndex+matchLength >= dictLimit)) { + match = base + matchIndex; + matchLength += ZSTD_count(ip+matchLength, match+matchLength, iend); + } else { + match = dictBase + matchIndex; + matchLength += ZSTD_count_2segments(ip+matchLength, match+matchLength, iend, dictEnd, prefixStart); + if (matchIndex+matchLength >= dictLimit) + match = base + matchIndex; /* to prepare for next usage of match[matchLength] */ + } + + if (matchLength > bestLength) { + if (matchLength > matchEndIdx - matchIndex) + matchEndIdx = matchIndex + (U32)matchLength; + if ( (4*(int)(matchLength-bestLength)) > (int)(ZSTD_highbit32(curr-matchIndex+1) - ZSTD_highbit32((U32)offsetPtr[0]+1)) ) + bestLength = matchLength, *offsetPtr = ZSTD_REP_MOVE + curr - matchIndex; + if (ip+matchLength == iend) { /* equal : no way to know if inf or sup */ + if (dictMode == ZSTD_dictMatchState) { + nbCompares = 0; /* in addition to avoiding checking any + * further in this loop, make sure we + * skip checking in the dictionary. */ + } + break; /* drop, to guarantee consistency (miss a little bit of compression) */ + } + } + + if (match[matchLength] < ip[matchLength]) { + /* match is smaller than current */ + *smallerPtr = matchIndex; /* update smaller idx */ + commonLengthSmaller = matchLength; /* all smaller will now have at least this guaranteed common length */ + if (matchIndex <= btLow) { smallerPtr=&dummy32; break; } /* beyond tree size, stop the search */ + smallerPtr = nextPtr+1; /* new "smaller" => larger of match */ + matchIndex = nextPtr[1]; /* new matchIndex larger than previous (closer to current) */ + } else { + /* match is larger than current */ + *largerPtr = matchIndex; + commonLengthLarger = matchLength; + if (matchIndex <= btLow) { largerPtr=&dummy32; break; } /* beyond tree size, stop the search */ + largerPtr = nextPtr; + matchIndex = nextPtr[0]; + } } + + *smallerPtr = *largerPtr = 0; + + assert(nbCompares <= (1U << ZSTD_SEARCHLOG_MAX)); /* Check we haven't underflowed. */ + if (dictMode == ZSTD_dictMatchState && nbCompares) { + bestLength = ZSTD_DUBT_findBetterDictMatch( + ms, ip, iend, + offsetPtr, bestLength, nbCompares, + mls, dictMode); + } + + assert(matchEndIdx > curr+8); /* ensure nextToUpdate is increased */ + ms->nextToUpdate = matchEndIdx - 8; /* skip repetitive patterns */ + if (bestLength >= MINMATCH) { + U32 const mIndex = curr - ((U32)*offsetPtr - ZSTD_REP_MOVE); (void)mIndex; + DEBUGLOG(8, "ZSTD_DUBT_findBestMatch(%u) : found match of length %u and offsetCode %u (pos %u)", + curr, (U32)bestLength, (U32)*offsetPtr, mIndex); + } + return bestLength; + } +} + + +/* ZSTD_BtFindBestMatch() : Tree updater, providing best match */ +FORCE_INLINE_TEMPLATE size_t +ZSTD_BtFindBestMatch( ZSTD_matchState_t* ms, + const BYTE* const ip, const BYTE* const iLimit, + size_t* offsetPtr, + const U32 mls /* template */, + const ZSTD_dictMode_e dictMode) +{ + DEBUGLOG(7, "ZSTD_BtFindBestMatch"); + if (ip < ms->window.base + ms->nextToUpdate) return 0; /* skipped area */ + ZSTD_updateDUBT(ms, ip, iLimit, mls); + return ZSTD_DUBT_findBestMatch(ms, ip, iLimit, offsetPtr, mls, dictMode); +} + + +static size_t +ZSTD_BtFindBestMatch_selectMLS ( ZSTD_matchState_t* ms, + const BYTE* ip, const BYTE* const iLimit, + size_t* offsetPtr) +{ + switch(ms->cParams.minMatch) + { + default : /* includes case 3 */ + case 4 : return ZSTD_BtFindBestMatch(ms, ip, iLimit, offsetPtr, 4, ZSTD_noDict); + case 5 : return ZSTD_BtFindBestMatch(ms, ip, iLimit, offsetPtr, 5, ZSTD_noDict); + case 7 : + case 6 : return ZSTD_BtFindBestMatch(ms, ip, iLimit, offsetPtr, 6, ZSTD_noDict); + } +} + + +static size_t ZSTD_BtFindBestMatch_dictMatchState_selectMLS ( + ZSTD_matchState_t* ms, + const BYTE* ip, const BYTE* const iLimit, + size_t* offsetPtr) +{ + switch(ms->cParams.minMatch) + { + default : /* includes case 3 */ + case 4 : return ZSTD_BtFindBestMatch(ms, ip, iLimit, offsetPtr, 4, ZSTD_dictMatchState); + case 5 : return ZSTD_BtFindBestMatch(ms, ip, iLimit, offsetPtr, 5, ZSTD_dictMatchState); + case 7 : + case 6 : return ZSTD_BtFindBestMatch(ms, ip, iLimit, offsetPtr, 6, ZSTD_dictMatchState); + } +} + + +static size_t ZSTD_BtFindBestMatch_extDict_selectMLS ( + ZSTD_matchState_t* ms, + const BYTE* ip, const BYTE* const iLimit, + size_t* offsetPtr) +{ + switch(ms->cParams.minMatch) + { + default : /* includes case 3 */ + case 4 : return ZSTD_BtFindBestMatch(ms, ip, iLimit, offsetPtr, 4, ZSTD_extDict); + case 5 : return ZSTD_BtFindBestMatch(ms, ip, iLimit, offsetPtr, 5, ZSTD_extDict); + case 7 : + case 6 : return ZSTD_BtFindBestMatch(ms, ip, iLimit, offsetPtr, 6, ZSTD_extDict); + } +} + + + +/* ********************************* +* Hash Chain +***********************************/ +#define NEXT_IN_CHAIN(d, mask) chainTable[(d) & (mask)] + +/* Update chains up to ip (excluded) + Assumption : always within prefix (i.e. not within extDict) */ +FORCE_INLINE_TEMPLATE U32 ZSTD_insertAndFindFirstIndex_internal( + ZSTD_matchState_t* ms, + const ZSTD_compressionParameters* const cParams, + const BYTE* ip, U32 const mls) +{ + U32* const hashTable = ms->hashTable; + const U32 hashLog = cParams->hashLog; + U32* const chainTable = ms->chainTable; + const U32 chainMask = (1 << cParams->chainLog) - 1; + const BYTE* const base = ms->window.base; + const U32 target = (U32)(ip - base); + U32 idx = ms->nextToUpdate; + + while(idx < target) { /* catch up */ + size_t const h = ZSTD_hashPtr(base+idx, hashLog, mls); + NEXT_IN_CHAIN(idx, chainMask) = hashTable[h]; + hashTable[h] = idx; + idx++; + } + + ms->nextToUpdate = target; + return hashTable[ZSTD_hashPtr(ip, hashLog, mls)]; +} + +U32 ZSTD_insertAndFindFirstIndex(ZSTD_matchState_t* ms, const BYTE* ip) { + const ZSTD_compressionParameters* const cParams = &ms->cParams; + return ZSTD_insertAndFindFirstIndex_internal(ms, cParams, ip, ms->cParams.minMatch); +} + +void ZSTD_dedicatedDictSearch_lazy_loadDictionary(ZSTD_matchState_t* ms, const BYTE* const ip) +{ + const BYTE* const base = ms->window.base; + U32 const target = (U32)(ip - base); + U32* const hashTable = ms->hashTable; + U32* const chainTable = ms->chainTable; + U32 const chainSize = 1 << ms->cParams.chainLog; + U32 idx = ms->nextToUpdate; + U32 const minChain = chainSize < target ? target - chainSize : idx; + U32 const bucketSize = 1 << ZSTD_LAZY_DDSS_BUCKET_LOG; + U32 const cacheSize = bucketSize - 1; + U32 const chainAttempts = (1 << ms->cParams.searchLog) - cacheSize; + U32 const chainLimit = chainAttempts > 255 ? 255 : chainAttempts; + + /* We know the hashtable is oversized by a factor of `bucketSize`. + * We are going to temporarily pretend `bucketSize == 1`, keeping only a + * single entry. We will use the rest of the space to construct a temporary + * chaintable. + */ + U32 const hashLog = ms->cParams.hashLog - ZSTD_LAZY_DDSS_BUCKET_LOG; + U32* const tmpHashTable = hashTable; + U32* const tmpChainTable = hashTable + ((size_t)1 << hashLog); + U32 const tmpChainSize = ((1 << ZSTD_LAZY_DDSS_BUCKET_LOG) - 1) << hashLog; + U32 const tmpMinChain = tmpChainSize < target ? target - tmpChainSize : idx; + + U32 hashIdx; + + assert(ms->cParams.chainLog <= 24); + assert(ms->cParams.hashLog >= ms->cParams.chainLog); + assert(idx != 0); + assert(tmpMinChain <= minChain); + + /* fill conventional hash table and conventional chain table */ + for ( ; idx < target; idx++) { + U32 const h = (U32)ZSTD_hashPtr(base + idx, hashLog, ms->cParams.minMatch); + if (idx >= tmpMinChain) { + tmpChainTable[idx - tmpMinChain] = hashTable[h]; + } + tmpHashTable[h] = idx; + } + + /* sort chains into ddss chain table */ + { + U32 chainPos = 0; + for (hashIdx = 0; hashIdx < (1U << hashLog); hashIdx++) { + U32 count; + U32 countBeyondMinChain = 0; + U32 i = tmpHashTable[hashIdx]; + for (count = 0; i >= tmpMinChain && count < cacheSize; count++) { + /* skip through the chain to the first position that won't be + * in the hash cache bucket */ + if (i < minChain) { + countBeyondMinChain++; + } + i = tmpChainTable[i - tmpMinChain]; + } + if (count == cacheSize) { + for (count = 0; count < chainLimit;) { + if (i < minChain) { + if (!i || countBeyondMinChain++ > cacheSize) { + /* only allow pulling `cacheSize` number of entries + * into the cache or chainTable beyond `minChain`, + * to replace the entries pulled out of the + * chainTable into the cache. This lets us reach + * back further without increasing the total number + * of entries in the chainTable, guaranteeing the + * DDSS chain table will fit into the space + * allocated for the regular one. */ + break; + } + } + chainTable[chainPos++] = i; + count++; + if (i < tmpMinChain) { + break; + } + i = tmpChainTable[i - tmpMinChain]; + } + } else { + count = 0; + } + if (count) { + tmpHashTable[hashIdx] = ((chainPos - count) << 8) + count; + } else { + tmpHashTable[hashIdx] = 0; + } + } + assert(chainPos <= chainSize); /* I believe this is guaranteed... */ + } + + /* move chain pointers into the last entry of each hash bucket */ + for (hashIdx = (1 << hashLog); hashIdx; ) { + U32 const bucketIdx = --hashIdx << ZSTD_LAZY_DDSS_BUCKET_LOG; + U32 const chainPackedPointer = tmpHashTable[hashIdx]; + U32 i; + for (i = 0; i < cacheSize; i++) { + hashTable[bucketIdx + i] = 0; + } + hashTable[bucketIdx + bucketSize - 1] = chainPackedPointer; + } + + /* fill the buckets of the hash table */ + for (idx = ms->nextToUpdate; idx < target; idx++) { + U32 const h = (U32)ZSTD_hashPtr(base + idx, hashLog, ms->cParams.minMatch) + << ZSTD_LAZY_DDSS_BUCKET_LOG; + U32 i; + /* Shift hash cache down 1. */ + for (i = cacheSize - 1; i; i--) + hashTable[h + i] = hashTable[h + i - 1]; + hashTable[h] = idx; + } + + ms->nextToUpdate = target; +} + + +/* inlining is important to hardwire a hot branch (template emulation) */ +FORCE_INLINE_TEMPLATE +size_t ZSTD_HcFindBestMatch_generic ( + ZSTD_matchState_t* ms, + const BYTE* const ip, const BYTE* const iLimit, + size_t* offsetPtr, + const U32 mls, const ZSTD_dictMode_e dictMode) +{ + const ZSTD_compressionParameters* const cParams = &ms->cParams; + U32* const chainTable = ms->chainTable; + const U32 chainSize = (1 << cParams->chainLog); + const U32 chainMask = chainSize-1; + const BYTE* const base = ms->window.base; + const BYTE* const dictBase = ms->window.dictBase; + const U32 dictLimit = ms->window.dictLimit; + const BYTE* const prefixStart = base + dictLimit; + const BYTE* const dictEnd = dictBase + dictLimit; + const U32 curr = (U32)(ip-base); + const U32 maxDistance = 1U << cParams->windowLog; + const U32 lowestValid = ms->window.lowLimit; + const U32 withinMaxDistance = (curr - lowestValid > maxDistance) ? curr - maxDistance : lowestValid; + const U32 isDictionary = (ms->loadedDictEnd != 0); + const U32 lowLimit = isDictionary ? lowestValid : withinMaxDistance; + const U32 minChain = curr > chainSize ? curr - chainSize : 0; + U32 nbAttempts = 1U << cParams->searchLog; + size_t ml=4-1; + + const ZSTD_matchState_t* const dms = ms->dictMatchState; + const U32 ddsHashLog = dictMode == ZSTD_dedicatedDictSearch + ? dms->cParams.hashLog - ZSTD_LAZY_DDSS_BUCKET_LOG : 0; + const size_t ddsIdx = dictMode == ZSTD_dedicatedDictSearch + ? ZSTD_hashPtr(ip, ddsHashLog, mls) << ZSTD_LAZY_DDSS_BUCKET_LOG : 0; + + U32 matchIndex; + + if (dictMode == ZSTD_dedicatedDictSearch) { + const U32* entry = &dms->hashTable[ddsIdx]; + PREFETCH_L1(entry); + } + + /* HC4 match finder */ + matchIndex = ZSTD_insertAndFindFirstIndex_internal(ms, cParams, ip, mls); + + for ( ; (matchIndex>=lowLimit) & (nbAttempts>0) ; nbAttempts--) { + size_t currentMl=0; + if ((dictMode != ZSTD_extDict) || matchIndex >= dictLimit) { + const BYTE* const match = base + matchIndex; + assert(matchIndex >= dictLimit); /* ensures this is true if dictMode != ZSTD_extDict */ + if (match[ml] == ip[ml]) /* potentially better */ + currentMl = ZSTD_count(ip, match, iLimit); + } else { + const BYTE* const match = dictBase + matchIndex; + assert(match+4 <= dictEnd); + if (MEM_read32(match) == MEM_read32(ip)) /* assumption : matchIndex <= dictLimit-4 (by table construction) */ + currentMl = ZSTD_count_2segments(ip+4, match+4, iLimit, dictEnd, prefixStart) + 4; + } + + /* save best solution */ + if (currentMl > ml) { + ml = currentMl; + *offsetPtr = curr - matchIndex + ZSTD_REP_MOVE; + if (ip+currentMl == iLimit) break; /* best possible, avoids read overflow on next attempt */ + } + + if (matchIndex <= minChain) break; + matchIndex = NEXT_IN_CHAIN(matchIndex, chainMask); + } + + assert(nbAttempts <= (1U << ZSTD_SEARCHLOG_MAX)); /* Check we haven't underflowed. */ + if (dictMode == ZSTD_dedicatedDictSearch) { + const U32 ddsLowestIndex = dms->window.dictLimit; + const BYTE* const ddsBase = dms->window.base; + const BYTE* const ddsEnd = dms->window.nextSrc; + const U32 ddsSize = (U32)(ddsEnd - ddsBase); + const U32 ddsIndexDelta = dictLimit - ddsSize; + const U32 bucketSize = (1 << ZSTD_LAZY_DDSS_BUCKET_LOG); + const U32 bucketLimit = nbAttempts < bucketSize - 1 ? nbAttempts : bucketSize - 1; + U32 ddsAttempt; + + for (ddsAttempt = 0; ddsAttempt < bucketSize - 1; ddsAttempt++) { + PREFETCH_L1(ddsBase + dms->hashTable[ddsIdx + ddsAttempt]); + } + + { + U32 const chainPackedPointer = dms->hashTable[ddsIdx + bucketSize - 1]; + U32 const chainIndex = chainPackedPointer >> 8; + + PREFETCH_L1(&dms->chainTable[chainIndex]); + } + + for (ddsAttempt = 0; ddsAttempt < bucketLimit; ddsAttempt++) { + size_t currentMl=0; + const BYTE* match; + matchIndex = dms->hashTable[ddsIdx + ddsAttempt]; + match = ddsBase + matchIndex; + + if (!matchIndex) { + return ml; + } + + /* guaranteed by table construction */ + (void)ddsLowestIndex; + assert(matchIndex >= ddsLowestIndex); + assert(match+4 <= ddsEnd); + if (MEM_read32(match) == MEM_read32(ip)) { + /* assumption : matchIndex <= dictLimit-4 (by table construction) */ + currentMl = ZSTD_count_2segments(ip+4, match+4, iLimit, ddsEnd, prefixStart) + 4; + } + + /* save best solution */ + if (currentMl > ml) { + ml = currentMl; + *offsetPtr = curr - (matchIndex + ddsIndexDelta) + ZSTD_REP_MOVE; + if (ip+currentMl == iLimit) { + /* best possible, avoids read overflow on next attempt */ + return ml; + } + } + } + + { + U32 const chainPackedPointer = dms->hashTable[ddsIdx + bucketSize - 1]; + U32 chainIndex = chainPackedPointer >> 8; + U32 const chainLength = chainPackedPointer & 0xFF; + U32 const chainAttempts = nbAttempts - ddsAttempt; + U32 const chainLimit = chainAttempts > chainLength ? chainLength : chainAttempts; + U32 chainAttempt; + + for (chainAttempt = 0 ; chainAttempt < chainLimit; chainAttempt++) { + PREFETCH_L1(ddsBase + dms->chainTable[chainIndex + chainAttempt]); + } + + for (chainAttempt = 0 ; chainAttempt < chainLimit; chainAttempt++, chainIndex++) { + size_t currentMl=0; + const BYTE* match; + matchIndex = dms->chainTable[chainIndex]; + match = ddsBase + matchIndex; + + /* guaranteed by table construction */ + assert(matchIndex >= ddsLowestIndex); + assert(match+4 <= ddsEnd); + if (MEM_read32(match) == MEM_read32(ip)) { + /* assumption : matchIndex <= dictLimit-4 (by table construction) */ + currentMl = ZSTD_count_2segments(ip+4, match+4, iLimit, ddsEnd, prefixStart) + 4; + } + + /* save best solution */ + if (currentMl > ml) { + ml = currentMl; + *offsetPtr = curr - (matchIndex + ddsIndexDelta) + ZSTD_REP_MOVE; + if (ip+currentMl == iLimit) break; /* best possible, avoids read overflow on next attempt */ + } + } + } + } else if (dictMode == ZSTD_dictMatchState) { + const U32* const dmsChainTable = dms->chainTable; + const U32 dmsChainSize = (1 << dms->cParams.chainLog); + const U32 dmsChainMask = dmsChainSize - 1; + const U32 dmsLowestIndex = dms->window.dictLimit; + const BYTE* const dmsBase = dms->window.base; + const BYTE* const dmsEnd = dms->window.nextSrc; + const U32 dmsSize = (U32)(dmsEnd - dmsBase); + const U32 dmsIndexDelta = dictLimit - dmsSize; + const U32 dmsMinChain = dmsSize > dmsChainSize ? dmsSize - dmsChainSize : 0; + + matchIndex = dms->hashTable[ZSTD_hashPtr(ip, dms->cParams.hashLog, mls)]; + + for ( ; (matchIndex>=dmsLowestIndex) & (nbAttempts>0) ; nbAttempts--) { + size_t currentMl=0; + const BYTE* const match = dmsBase + matchIndex; + assert(match+4 <= dmsEnd); + if (MEM_read32(match) == MEM_read32(ip)) /* assumption : matchIndex <= dictLimit-4 (by table construction) */ + currentMl = ZSTD_count_2segments(ip+4, match+4, iLimit, dmsEnd, prefixStart) + 4; + + /* save best solution */ + if (currentMl > ml) { + ml = currentMl; + *offsetPtr = curr - (matchIndex + dmsIndexDelta) + ZSTD_REP_MOVE; + if (ip+currentMl == iLimit) break; /* best possible, avoids read overflow on next attempt */ + } + + if (matchIndex <= dmsMinChain) break; + + matchIndex = dmsChainTable[matchIndex & dmsChainMask]; + } + } + + return ml; +} + + +FORCE_INLINE_TEMPLATE size_t ZSTD_HcFindBestMatch_selectMLS ( + ZSTD_matchState_t* ms, + const BYTE* ip, const BYTE* const iLimit, + size_t* offsetPtr) +{ + switch(ms->cParams.minMatch) + { + default : /* includes case 3 */ + case 4 : return ZSTD_HcFindBestMatch_generic(ms, ip, iLimit, offsetPtr, 4, ZSTD_noDict); + case 5 : return ZSTD_HcFindBestMatch_generic(ms, ip, iLimit, offsetPtr, 5, ZSTD_noDict); + case 7 : + case 6 : return ZSTD_HcFindBestMatch_generic(ms, ip, iLimit, offsetPtr, 6, ZSTD_noDict); + } +} + + +static size_t ZSTD_HcFindBestMatch_dictMatchState_selectMLS ( + ZSTD_matchState_t* ms, + const BYTE* ip, const BYTE* const iLimit, + size_t* offsetPtr) +{ + switch(ms->cParams.minMatch) + { + default : /* includes case 3 */ + case 4 : return ZSTD_HcFindBestMatch_generic(ms, ip, iLimit, offsetPtr, 4, ZSTD_dictMatchState); + case 5 : return ZSTD_HcFindBestMatch_generic(ms, ip, iLimit, offsetPtr, 5, ZSTD_dictMatchState); + case 7 : + case 6 : return ZSTD_HcFindBestMatch_generic(ms, ip, iLimit, offsetPtr, 6, ZSTD_dictMatchState); + } +} + + +static size_t ZSTD_HcFindBestMatch_dedicatedDictSearch_selectMLS ( + ZSTD_matchState_t* ms, + const BYTE* ip, const BYTE* const iLimit, + size_t* offsetPtr) +{ + switch(ms->cParams.minMatch) + { + default : /* includes case 3 */ + case 4 : return ZSTD_HcFindBestMatch_generic(ms, ip, iLimit, offsetPtr, 4, ZSTD_dedicatedDictSearch); + case 5 : return ZSTD_HcFindBestMatch_generic(ms, ip, iLimit, offsetPtr, 5, ZSTD_dedicatedDictSearch); + case 7 : + case 6 : return ZSTD_HcFindBestMatch_generic(ms, ip, iLimit, offsetPtr, 6, ZSTD_dedicatedDictSearch); + } +} + + +FORCE_INLINE_TEMPLATE size_t ZSTD_HcFindBestMatch_extDict_selectMLS ( + ZSTD_matchState_t* ms, + const BYTE* ip, const BYTE* const iLimit, + size_t* offsetPtr) +{ + switch(ms->cParams.minMatch) + { + default : /* includes case 3 */ + case 4 : return ZSTD_HcFindBestMatch_generic(ms, ip, iLimit, offsetPtr, 4, ZSTD_extDict); + case 5 : return ZSTD_HcFindBestMatch_generic(ms, ip, iLimit, offsetPtr, 5, ZSTD_extDict); + case 7 : + case 6 : return ZSTD_HcFindBestMatch_generic(ms, ip, iLimit, offsetPtr, 6, ZSTD_extDict); + } +} + + +/* ******************************* +* Common parser - lazy strategy +*********************************/ +typedef enum { search_hashChain, search_binaryTree } searchMethod_e; + +FORCE_INLINE_TEMPLATE size_t +ZSTD_compressBlock_lazy_generic( + ZSTD_matchState_t* ms, seqStore_t* seqStore, + U32 rep[ZSTD_REP_NUM], + const void* src, size_t srcSize, + const searchMethod_e searchMethod, const U32 depth, + ZSTD_dictMode_e const dictMode) +{ + const BYTE* const istart = (const BYTE*)src; + const BYTE* ip = istart; + const BYTE* anchor = istart; + const BYTE* const iend = istart + srcSize; + const BYTE* const ilimit = iend - 8; + const BYTE* const base = ms->window.base; + const U32 prefixLowestIndex = ms->window.dictLimit; + const BYTE* const prefixLowest = base + prefixLowestIndex; + + typedef size_t (*searchMax_f)( + ZSTD_matchState_t* ms, + const BYTE* ip, const BYTE* iLimit, size_t* offsetPtr); + + /* + * This table is indexed first by the four ZSTD_dictMode_e values, and then + * by the two searchMethod_e values. NULLs are placed for configurations + * that should never occur (extDict modes go to the other implementation + * below and there is no DDSS for binary tree search yet). + */ + const searchMax_f searchFuncs[4][2] = { + { + ZSTD_HcFindBestMatch_selectMLS, + ZSTD_BtFindBestMatch_selectMLS + }, + { + NULL, + NULL + }, + { + ZSTD_HcFindBestMatch_dictMatchState_selectMLS, + ZSTD_BtFindBestMatch_dictMatchState_selectMLS + }, + { + ZSTD_HcFindBestMatch_dedicatedDictSearch_selectMLS, + NULL + } + }; + + searchMax_f const searchMax = searchFuncs[dictMode][searchMethod == search_binaryTree]; + U32 offset_1 = rep[0], offset_2 = rep[1], savedOffset=0; + + const int isDMS = dictMode == ZSTD_dictMatchState; + const int isDDS = dictMode == ZSTD_dedicatedDictSearch; + const int isDxS = isDMS || isDDS; + const ZSTD_matchState_t* const dms = ms->dictMatchState; + const U32 dictLowestIndex = isDxS ? dms->window.dictLimit : 0; + const BYTE* const dictBase = isDxS ? dms->window.base : NULL; + const BYTE* const dictLowest = isDxS ? dictBase + dictLowestIndex : NULL; + const BYTE* const dictEnd = isDxS ? dms->window.nextSrc : NULL; + const U32 dictIndexDelta = isDxS ? + prefixLowestIndex - (U32)(dictEnd - dictBase) : + 0; + const U32 dictAndPrefixLength = (U32)((ip - prefixLowest) + (dictEnd - dictLowest)); + + assert(searchMax != NULL); + + DEBUGLOG(5, "ZSTD_compressBlock_lazy_generic (dictMode=%u)", (U32)dictMode); + + /* init */ + ip += (dictAndPrefixLength == 0); + if (dictMode == ZSTD_noDict) { + U32 const curr = (U32)(ip - base); + U32 const windowLow = ZSTD_getLowestPrefixIndex(ms, curr, ms->cParams.windowLog); + U32 const maxRep = curr - windowLow; + if (offset_2 > maxRep) savedOffset = offset_2, offset_2 = 0; + if (offset_1 > maxRep) savedOffset = offset_1, offset_1 = 0; + } + if (isDxS) { + /* dictMatchState repCode checks don't currently handle repCode == 0 + * disabling. */ + assert(offset_1 <= dictAndPrefixLength); + assert(offset_2 <= dictAndPrefixLength); + } + + /* Match Loop */ +#if defined(__x86_64__) + /* I've measured random a 5% speed loss on levels 5 & 6 (greedy) when the + * code alignment is perturbed. To fix the instability align the loop on 32-bytes. + */ + __asm__(".p2align 5"); +#endif + while (ip < ilimit) { + size_t matchLength=0; + size_t offset=0; + const BYTE* start=ip+1; + + /* check repCode */ + if (isDxS) { + const U32 repIndex = (U32)(ip - base) + 1 - offset_1; + const BYTE* repMatch = ((dictMode == ZSTD_dictMatchState || dictMode == ZSTD_dedicatedDictSearch) + && repIndex < prefixLowestIndex) ? + dictBase + (repIndex - dictIndexDelta) : + base + repIndex; + if (((U32)((prefixLowestIndex-1) - repIndex) >= 3 /* intentional underflow */) + && (MEM_read32(repMatch) == MEM_read32(ip+1)) ) { + const BYTE* repMatchEnd = repIndex < prefixLowestIndex ? dictEnd : iend; + matchLength = ZSTD_count_2segments(ip+1+4, repMatch+4, iend, repMatchEnd, prefixLowest) + 4; + if (depth==0) goto _storeSequence; + } + } + if ( dictMode == ZSTD_noDict + && ((offset_1 > 0) & (MEM_read32(ip+1-offset_1) == MEM_read32(ip+1)))) { + matchLength = ZSTD_count(ip+1+4, ip+1+4-offset_1, iend) + 4; + if (depth==0) goto _storeSequence; + } + + /* first search (depth 0) */ + { size_t offsetFound = 999999999; + size_t const ml2 = searchMax(ms, ip, iend, &offsetFound); + if (ml2 > matchLength) + matchLength = ml2, start = ip, offset=offsetFound; + } + + if (matchLength < 4) { + ip += ((ip-anchor) >> kSearchStrength) + 1; /* jump faster over incompressible sections */ + continue; + } + + /* let's try to find a better solution */ + if (depth>=1) + while (ip0) & (MEM_read32(ip) == MEM_read32(ip - offset_1)))) { + size_t const mlRep = ZSTD_count(ip+4, ip+4-offset_1, iend) + 4; + int const gain2 = (int)(mlRep * 3); + int const gain1 = (int)(matchLength*3 - ZSTD_highbit32((U32)offset+1) + 1); + if ((mlRep >= 4) && (gain2 > gain1)) + matchLength = mlRep, offset = 0, start = ip; + } + if (isDxS) { + const U32 repIndex = (U32)(ip - base) - offset_1; + const BYTE* repMatch = repIndex < prefixLowestIndex ? + dictBase + (repIndex - dictIndexDelta) : + base + repIndex; + if (((U32)((prefixLowestIndex-1) - repIndex) >= 3 /* intentional underflow */) + && (MEM_read32(repMatch) == MEM_read32(ip)) ) { + const BYTE* repMatchEnd = repIndex < prefixLowestIndex ? dictEnd : iend; + size_t const mlRep = ZSTD_count_2segments(ip+4, repMatch+4, iend, repMatchEnd, prefixLowest) + 4; + int const gain2 = (int)(mlRep * 3); + int const gain1 = (int)(matchLength*3 - ZSTD_highbit32((U32)offset+1) + 1); + if ((mlRep >= 4) && (gain2 > gain1)) + matchLength = mlRep, offset = 0, start = ip; + } + } + { size_t offset2=999999999; + size_t const ml2 = searchMax(ms, ip, iend, &offset2); + int const gain2 = (int)(ml2*4 - ZSTD_highbit32((U32)offset2+1)); /* raw approx */ + int const gain1 = (int)(matchLength*4 - ZSTD_highbit32((U32)offset+1) + 4); + if ((ml2 >= 4) && (gain2 > gain1)) { + matchLength = ml2, offset = offset2, start = ip; + continue; /* search a better one */ + } } + + /* let's find an even better one */ + if ((depth==2) && (ip0) & (MEM_read32(ip) == MEM_read32(ip - offset_1)))) { + size_t const mlRep = ZSTD_count(ip+4, ip+4-offset_1, iend) + 4; + int const gain2 = (int)(mlRep * 4); + int const gain1 = (int)(matchLength*4 - ZSTD_highbit32((U32)offset+1) + 1); + if ((mlRep >= 4) && (gain2 > gain1)) + matchLength = mlRep, offset = 0, start = ip; + } + if (isDxS) { + const U32 repIndex = (U32)(ip - base) - offset_1; + const BYTE* repMatch = repIndex < prefixLowestIndex ? + dictBase + (repIndex - dictIndexDelta) : + base + repIndex; + if (((U32)((prefixLowestIndex-1) - repIndex) >= 3 /* intentional underflow */) + && (MEM_read32(repMatch) == MEM_read32(ip)) ) { + const BYTE* repMatchEnd = repIndex < prefixLowestIndex ? dictEnd : iend; + size_t const mlRep = ZSTD_count_2segments(ip+4, repMatch+4, iend, repMatchEnd, prefixLowest) + 4; + int const gain2 = (int)(mlRep * 4); + int const gain1 = (int)(matchLength*4 - ZSTD_highbit32((U32)offset+1) + 1); + if ((mlRep >= 4) && (gain2 > gain1)) + matchLength = mlRep, offset = 0, start = ip; + } + } + { size_t offset2=999999999; + size_t const ml2 = searchMax(ms, ip, iend, &offset2); + int const gain2 = (int)(ml2*4 - ZSTD_highbit32((U32)offset2+1)); /* raw approx */ + int const gain1 = (int)(matchLength*4 - ZSTD_highbit32((U32)offset+1) + 7); + if ((ml2 >= 4) && (gain2 > gain1)) { + matchLength = ml2, offset = offset2, start = ip; + continue; + } } } + break; /* nothing found : store previous solution */ + } + + /* NOTE: + * start[-offset+ZSTD_REP_MOVE-1] is undefined behavior. + * (-offset+ZSTD_REP_MOVE-1) is unsigned, and is added to start, which + * overflows the pointer, which is undefined behavior. + */ + /* catch up */ + if (offset) { + if (dictMode == ZSTD_noDict) { + while ( ((start > anchor) & (start - (offset-ZSTD_REP_MOVE) > prefixLowest)) + && (start[-1] == (start-(offset-ZSTD_REP_MOVE))[-1]) ) /* only search for offset within prefix */ + { start--; matchLength++; } + } + if (isDxS) { + U32 const matchIndex = (U32)((start-base) - (offset - ZSTD_REP_MOVE)); + const BYTE* match = (matchIndex < prefixLowestIndex) ? dictBase + matchIndex - dictIndexDelta : base + matchIndex; + const BYTE* const mStart = (matchIndex < prefixLowestIndex) ? dictLowest : prefixLowest; + while ((start>anchor) && (match>mStart) && (start[-1] == match[-1])) { start--; match--; matchLength++; } /* catch up */ + } + offset_2 = offset_1; offset_1 = (U32)(offset - ZSTD_REP_MOVE); + } + /* store sequence */ +_storeSequence: + { size_t const litLength = start - anchor; + ZSTD_storeSeq(seqStore, litLength, anchor, iend, (U32)offset, matchLength-MINMATCH); + anchor = ip = start + matchLength; + } + + /* check immediate repcode */ + if (isDxS) { + while (ip <= ilimit) { + U32 const current2 = (U32)(ip-base); + U32 const repIndex = current2 - offset_2; + const BYTE* repMatch = repIndex < prefixLowestIndex ? + dictBase - dictIndexDelta + repIndex : + base + repIndex; + if ( ((U32)((prefixLowestIndex-1) - (U32)repIndex) >= 3 /* intentional overflow */) + && (MEM_read32(repMatch) == MEM_read32(ip)) ) { + const BYTE* const repEnd2 = repIndex < prefixLowestIndex ? dictEnd : iend; + matchLength = ZSTD_count_2segments(ip+4, repMatch+4, iend, repEnd2, prefixLowest) + 4; + offset = offset_2; offset_2 = offset_1; offset_1 = (U32)offset; /* swap offset_2 <=> offset_1 */ + ZSTD_storeSeq(seqStore, 0, anchor, iend, 0, matchLength-MINMATCH); + ip += matchLength; + anchor = ip; + continue; + } + break; + } + } + + if (dictMode == ZSTD_noDict) { + while ( ((ip <= ilimit) & (offset_2>0)) + && (MEM_read32(ip) == MEM_read32(ip - offset_2)) ) { + /* store sequence */ + matchLength = ZSTD_count(ip+4, ip+4-offset_2, iend) + 4; + offset = offset_2; offset_2 = offset_1; offset_1 = (U32)offset; /* swap repcodes */ + ZSTD_storeSeq(seqStore, 0, anchor, iend, 0, matchLength-MINMATCH); + ip += matchLength; + anchor = ip; + continue; /* faster when present ... (?) */ + } } } + + /* Save reps for next block */ + rep[0] = offset_1 ? offset_1 : savedOffset; + rep[1] = offset_2 ? offset_2 : savedOffset; + + /* Return the last literals size */ + return (size_t)(iend - anchor); +} + + +size_t ZSTD_compressBlock_btlazy2( + ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM], + void const* src, size_t srcSize) +{ + return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, search_binaryTree, 2, ZSTD_noDict); +} + +size_t ZSTD_compressBlock_lazy2( + ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM], + void const* src, size_t srcSize) +{ + return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, search_hashChain, 2, ZSTD_noDict); +} + +size_t ZSTD_compressBlock_lazy( + ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM], + void const* src, size_t srcSize) +{ + return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, search_hashChain, 1, ZSTD_noDict); +} + +size_t ZSTD_compressBlock_greedy( + ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM], + void const* src, size_t srcSize) +{ + return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, search_hashChain, 0, ZSTD_noDict); +} + +size_t ZSTD_compressBlock_btlazy2_dictMatchState( + ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM], + void const* src, size_t srcSize) +{ + return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, search_binaryTree, 2, ZSTD_dictMatchState); +} + +size_t ZSTD_compressBlock_lazy2_dictMatchState( + ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM], + void const* src, size_t srcSize) +{ + return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, search_hashChain, 2, ZSTD_dictMatchState); +} + +size_t ZSTD_compressBlock_lazy_dictMatchState( + ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM], + void const* src, size_t srcSize) +{ + return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, search_hashChain, 1, ZSTD_dictMatchState); +} + +size_t ZSTD_compressBlock_greedy_dictMatchState( + ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM], + void const* src, size_t srcSize) +{ + return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, search_hashChain, 0, ZSTD_dictMatchState); +} + + +size_t ZSTD_compressBlock_lazy2_dedicatedDictSearch( + ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM], + void const* src, size_t srcSize) +{ + return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, search_hashChain, 2, ZSTD_dedicatedDictSearch); +} + +size_t ZSTD_compressBlock_lazy_dedicatedDictSearch( + ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM], + void const* src, size_t srcSize) +{ + return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, search_hashChain, 1, ZSTD_dedicatedDictSearch); +} + +size_t ZSTD_compressBlock_greedy_dedicatedDictSearch( + ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM], + void const* src, size_t srcSize) +{ + return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, search_hashChain, 0, ZSTD_dedicatedDictSearch); +} + + +FORCE_INLINE_TEMPLATE +size_t ZSTD_compressBlock_lazy_extDict_generic( + ZSTD_matchState_t* ms, seqStore_t* seqStore, + U32 rep[ZSTD_REP_NUM], + const void* src, size_t srcSize, + const searchMethod_e searchMethod, const U32 depth) +{ + const BYTE* const istart = (const BYTE*)src; + const BYTE* ip = istart; + const BYTE* anchor = istart; + const BYTE* const iend = istart + srcSize; + const BYTE* const ilimit = iend - 8; + const BYTE* const base = ms->window.base; + const U32 dictLimit = ms->window.dictLimit; + const BYTE* const prefixStart = base + dictLimit; + const BYTE* const dictBase = ms->window.dictBase; + const BYTE* const dictEnd = dictBase + dictLimit; + const BYTE* const dictStart = dictBase + ms->window.lowLimit; + const U32 windowLog = ms->cParams.windowLog; + + typedef size_t (*searchMax_f)( + ZSTD_matchState_t* ms, + const BYTE* ip, const BYTE* iLimit, size_t* offsetPtr); + searchMax_f searchMax = searchMethod==search_binaryTree ? ZSTD_BtFindBestMatch_extDict_selectMLS : ZSTD_HcFindBestMatch_extDict_selectMLS; + + U32 offset_1 = rep[0], offset_2 = rep[1]; + + DEBUGLOG(5, "ZSTD_compressBlock_lazy_extDict_generic"); + + /* init */ + ip += (ip == prefixStart); + + /* Match Loop */ +#if defined(__x86_64__) + /* I've measured random a 5% speed loss on levels 5 & 6 (greedy) when the + * code alignment is perturbed. To fix the instability align the loop on 32-bytes. + */ + __asm__(".p2align 5"); +#endif + while (ip < ilimit) { + size_t matchLength=0; + size_t offset=0; + const BYTE* start=ip+1; + U32 curr = (U32)(ip-base); + + /* check repCode */ + { const U32 windowLow = ZSTD_getLowestMatchIndex(ms, curr+1, windowLog); + const U32 repIndex = (U32)(curr+1 - offset_1); + const BYTE* const repBase = repIndex < dictLimit ? dictBase : base; + const BYTE* const repMatch = repBase + repIndex; + if (((U32)((dictLimit-1) - repIndex) >= 3) & (repIndex > windowLow)) /* intentional overflow */ + if (MEM_read32(ip+1) == MEM_read32(repMatch)) { + /* repcode detected we should take it */ + const BYTE* const repEnd = repIndex < dictLimit ? dictEnd : iend; + matchLength = ZSTD_count_2segments(ip+1+4, repMatch+4, iend, repEnd, prefixStart) + 4; + if (depth==0) goto _storeSequence; + } } + + /* first search (depth 0) */ + { size_t offsetFound = 999999999; + size_t const ml2 = searchMax(ms, ip, iend, &offsetFound); + if (ml2 > matchLength) + matchLength = ml2, start = ip, offset=offsetFound; + } + + if (matchLength < 4) { + ip += ((ip-anchor) >> kSearchStrength) + 1; /* jump faster over incompressible sections */ + continue; + } + + /* let's try to find a better solution */ + if (depth>=1) + while (ip= 3) & (repIndex > windowLow)) /* intentional overflow */ + if (MEM_read32(ip) == MEM_read32(repMatch)) { + /* repcode detected */ + const BYTE* const repEnd = repIndex < dictLimit ? dictEnd : iend; + size_t const repLength = ZSTD_count_2segments(ip+4, repMatch+4, iend, repEnd, prefixStart) + 4; + int const gain2 = (int)(repLength * 3); + int const gain1 = (int)(matchLength*3 - ZSTD_highbit32((U32)offset+1) + 1); + if ((repLength >= 4) && (gain2 > gain1)) + matchLength = repLength, offset = 0, start = ip; + } } + + /* search match, depth 1 */ + { size_t offset2=999999999; + size_t const ml2 = searchMax(ms, ip, iend, &offset2); + int const gain2 = (int)(ml2*4 - ZSTD_highbit32((U32)offset2+1)); /* raw approx */ + int const gain1 = (int)(matchLength*4 - ZSTD_highbit32((U32)offset+1) + 4); + if ((ml2 >= 4) && (gain2 > gain1)) { + matchLength = ml2, offset = offset2, start = ip; + continue; /* search a better one */ + } } + + /* let's find an even better one */ + if ((depth==2) && (ip= 3) & (repIndex > windowLow)) /* intentional overflow */ + if (MEM_read32(ip) == MEM_read32(repMatch)) { + /* repcode detected */ + const BYTE* const repEnd = repIndex < dictLimit ? dictEnd : iend; + size_t const repLength = ZSTD_count_2segments(ip+4, repMatch+4, iend, repEnd, prefixStart) + 4; + int const gain2 = (int)(repLength * 4); + int const gain1 = (int)(matchLength*4 - ZSTD_highbit32((U32)offset+1) + 1); + if ((repLength >= 4) && (gain2 > gain1)) + matchLength = repLength, offset = 0, start = ip; + } } + + /* search match, depth 2 */ + { size_t offset2=999999999; + size_t const ml2 = searchMax(ms, ip, iend, &offset2); + int const gain2 = (int)(ml2*4 - ZSTD_highbit32((U32)offset2+1)); /* raw approx */ + int const gain1 = (int)(matchLength*4 - ZSTD_highbit32((U32)offset+1) + 7); + if ((ml2 >= 4) && (gain2 > gain1)) { + matchLength = ml2, offset = offset2, start = ip; + continue; + } } } + break; /* nothing found : store previous solution */ + } + + /* catch up */ + if (offset) { + U32 const matchIndex = (U32)((start-base) - (offset - ZSTD_REP_MOVE)); + const BYTE* match = (matchIndex < dictLimit) ? dictBase + matchIndex : base + matchIndex; + const BYTE* const mStart = (matchIndex < dictLimit) ? dictStart : prefixStart; + while ((start>anchor) && (match>mStart) && (start[-1] == match[-1])) { start--; match--; matchLength++; } /* catch up */ + offset_2 = offset_1; offset_1 = (U32)(offset - ZSTD_REP_MOVE); + } + + /* store sequence */ +_storeSequence: + { size_t const litLength = start - anchor; + ZSTD_storeSeq(seqStore, litLength, anchor, iend, (U32)offset, matchLength-MINMATCH); + anchor = ip = start + matchLength; + } + + /* check immediate repcode */ + while (ip <= ilimit) { + const U32 repCurrent = (U32)(ip-base); + const U32 windowLow = ZSTD_getLowestMatchIndex(ms, repCurrent, windowLog); + const U32 repIndex = repCurrent - offset_2; + const BYTE* const repBase = repIndex < dictLimit ? dictBase : base; + const BYTE* const repMatch = repBase + repIndex; + if (((U32)((dictLimit-1) - repIndex) >= 3) & (repIndex > windowLow)) /* intentional overflow */ + if (MEM_read32(ip) == MEM_read32(repMatch)) { + /* repcode detected we should take it */ + const BYTE* const repEnd = repIndex < dictLimit ? dictEnd : iend; + matchLength = ZSTD_count_2segments(ip+4, repMatch+4, iend, repEnd, prefixStart) + 4; + offset = offset_2; offset_2 = offset_1; offset_1 = (U32)offset; /* swap offset history */ + ZSTD_storeSeq(seqStore, 0, anchor, iend, 0, matchLength-MINMATCH); + ip += matchLength; + anchor = ip; + continue; /* faster when present ... (?) */ + } + break; + } } + + /* Save reps for next block */ + rep[0] = offset_1; + rep[1] = offset_2; + + /* Return the last literals size */ + return (size_t)(iend - anchor); +} + + +size_t ZSTD_compressBlock_greedy_extDict( + ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM], + void const* src, size_t srcSize) +{ + return ZSTD_compressBlock_lazy_extDict_generic(ms, seqStore, rep, src, srcSize, search_hashChain, 0); +} + +size_t ZSTD_compressBlock_lazy_extDict( + ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM], + void const* src, size_t srcSize) + +{ + return ZSTD_compressBlock_lazy_extDict_generic(ms, seqStore, rep, src, srcSize, search_hashChain, 1); +} + +size_t ZSTD_compressBlock_lazy2_extDict( + ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM], + void const* src, size_t srcSize) + +{ + return ZSTD_compressBlock_lazy_extDict_generic(ms, seqStore, rep, src, srcSize, search_hashChain, 2); +} + +size_t ZSTD_compressBlock_btlazy2_extDict( + ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM], + void const* src, size_t srcSize) + +{ + return ZSTD_compressBlock_lazy_extDict_generic(ms, seqStore, rep, src, srcSize, search_binaryTree, 2); +} diff --git a/lib/zstd/compress/zstd_lazy.h b/lib/zstd/compress/zstd_lazy.h new file mode 100644 index 000000000000..2fc5a6182134 --- /dev/null +++ b/lib/zstd/compress/zstd_lazy.h @@ -0,0 +1,81 @@ +/* + * Copyright (c) Yann Collet, Facebook, Inc. + * All rights reserved. + * + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). + * You may select, at your option, one of the above-listed licenses. + */ + +#ifndef ZSTD_LAZY_H +#define ZSTD_LAZY_H + + +#include "zstd_compress_internal.h" + +/* + * Dedicated Dictionary Search Structure bucket log. In the + * ZSTD_dedicatedDictSearch mode, the hashTable has + * 2 ** ZSTD_LAZY_DDSS_BUCKET_LOG entries in each bucket, rather than just + * one. + */ +#define ZSTD_LAZY_DDSS_BUCKET_LOG 2 + +U32 ZSTD_insertAndFindFirstIndex(ZSTD_matchState_t* ms, const BYTE* ip); + +void ZSTD_dedicatedDictSearch_lazy_loadDictionary(ZSTD_matchState_t* ms, const BYTE* const ip); + +void ZSTD_preserveUnsortedMark (U32* const table, U32 const size, U32 const reducerValue); /*! used in ZSTD_reduceIndex(). preemptively increase value of ZSTD_DUBT_UNSORTED_MARK */ + +size_t ZSTD_compressBlock_btlazy2( + ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM], + void const* src, size_t srcSize); +size_t ZSTD_compressBlock_lazy2( + ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM], + void const* src, size_t srcSize); +size_t ZSTD_compressBlock_lazy( + ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM], + void const* src, size_t srcSize); +size_t ZSTD_compressBlock_greedy( + ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM], + void const* src, size_t srcSize); + +size_t ZSTD_compressBlock_btlazy2_dictMatchState( + ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM], + void const* src, size_t srcSize); +size_t ZSTD_compressBlock_lazy2_dictMatchState( + ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM], + void const* src, size_t srcSize); +size_t ZSTD_compressBlock_lazy_dictMatchState( + ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM], + void const* src, size_t srcSize); +size_t ZSTD_compressBlock_greedy_dictMatchState( + ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM], + void const* src, size_t srcSize); + +size_t ZSTD_compressBlock_lazy2_dedicatedDictSearch( + ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM], + void const* src, size_t srcSize); +size_t ZSTD_compressBlock_lazy_dedicatedDictSearch( + ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM], + void const* src, size_t srcSize); +size_t ZSTD_compressBlock_greedy_dedicatedDictSearch( + ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM], + void const* src, size_t srcSize); + +size_t ZSTD_compressBlock_greedy_extDict( + ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM], + void const* src, size_t srcSize); +size_t ZSTD_compressBlock_lazy_extDict( + ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM], + void const* src, size_t srcSize); +size_t ZSTD_compressBlock_lazy2_extDict( + ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM], + void const* src, size_t srcSize); +size_t ZSTD_compressBlock_btlazy2_extDict( + ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM], + void const* src, size_t srcSize); + + +#endif /* ZSTD_LAZY_H */ diff --git a/lib/zstd/compress/zstd_ldm.c b/lib/zstd/compress/zstd_ldm.c new file mode 100644 index 000000000000..8ef7e88a5add --- /dev/null +++ b/lib/zstd/compress/zstd_ldm.c @@ -0,0 +1,686 @@ +/* + * Copyright (c) Yann Collet, Facebook, Inc. + * All rights reserved. + * + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). + * You may select, at your option, one of the above-listed licenses. + */ + +#include "zstd_ldm.h" + +#include "../common/debug.h" +#include +#include "zstd_fast.h" /* ZSTD_fillHashTable() */ +#include "zstd_double_fast.h" /* ZSTD_fillDoubleHashTable() */ +#include "zstd_ldm_geartab.h" + +#define LDM_BUCKET_SIZE_LOG 3 +#define LDM_MIN_MATCH_LENGTH 64 +#define LDM_HASH_RLOG 7 + +typedef struct { + U64 rolling; + U64 stopMask; +} ldmRollingHashState_t; + +/* ZSTD_ldm_gear_init(): + * + * Initializes the rolling hash state such that it will honor the + * settings in params. */ +static void ZSTD_ldm_gear_init(ldmRollingHashState_t* state, ldmParams_t const* params) +{ + unsigned maxBitsInMask = MIN(params->minMatchLength, 64); + unsigned hashRateLog = params->hashRateLog; + + state->rolling = ~(U32)0; + + /* The choice of the splitting criterion is subject to two conditions: + * 1. it has to trigger on average every 2^(hashRateLog) bytes; + * 2. ideally, it has to depend on a window of minMatchLength bytes. + * + * In the gear hash algorithm, bit n depends on the last n bytes; + * so in order to obtain a good quality splitting criterion it is + * preferable to use bits with high weight. + * + * To match condition 1 we use a mask with hashRateLog bits set + * and, because of the previous remark, we make sure these bits + * have the highest possible weight while still respecting + * condition 2. + */ + if (hashRateLog > 0 && hashRateLog <= maxBitsInMask) { + state->stopMask = (((U64)1 << hashRateLog) - 1) << (maxBitsInMask - hashRateLog); + } else { + /* In this degenerate case we simply honor the hash rate. */ + state->stopMask = ((U64)1 << hashRateLog) - 1; + } +} + +/* ZSTD_ldm_gear_feed(): + * + * Registers in the splits array all the split points found in the first + * size bytes following the data pointer. This function terminates when + * either all the data has been processed or LDM_BATCH_SIZE splits are + * present in the splits array. + * + * Precondition: The splits array must not be full. + * Returns: The number of bytes processed. */ +static size_t ZSTD_ldm_gear_feed(ldmRollingHashState_t* state, + BYTE const* data, size_t size, + size_t* splits, unsigned* numSplits) +{ + size_t n; + U64 hash, mask; + + hash = state->rolling; + mask = state->stopMask; + n = 0; + +#define GEAR_ITER_ONCE() do { \ + hash = (hash << 1) + ZSTD_ldm_gearTab[data[n] & 0xff]; \ + n += 1; \ + if (UNLIKELY((hash & mask) == 0)) { \ + splits[*numSplits] = n; \ + *numSplits += 1; \ + if (*numSplits == LDM_BATCH_SIZE) \ + goto done; \ + } \ + } while (0) + + while (n + 3 < size) { + GEAR_ITER_ONCE(); + GEAR_ITER_ONCE(); + GEAR_ITER_ONCE(); + GEAR_ITER_ONCE(); + } + while (n < size) { + GEAR_ITER_ONCE(); + } + +#undef GEAR_ITER_ONCE + +done: + state->rolling = hash; + return n; +} + +void ZSTD_ldm_adjustParameters(ldmParams_t* params, + ZSTD_compressionParameters const* cParams) +{ + params->windowLog = cParams->windowLog; + ZSTD_STATIC_ASSERT(LDM_BUCKET_SIZE_LOG <= ZSTD_LDM_BUCKETSIZELOG_MAX); + DEBUGLOG(4, "ZSTD_ldm_adjustParameters"); + if (!params->bucketSizeLog) params->bucketSizeLog = LDM_BUCKET_SIZE_LOG; + if (!params->minMatchLength) params->minMatchLength = LDM_MIN_MATCH_LENGTH; + if (params->hashLog == 0) { + params->hashLog = MAX(ZSTD_HASHLOG_MIN, params->windowLog - LDM_HASH_RLOG); + assert(params->hashLog <= ZSTD_HASHLOG_MAX); + } + if (params->hashRateLog == 0) { + params->hashRateLog = params->windowLog < params->hashLog + ? 0 + : params->windowLog - params->hashLog; + } + params->bucketSizeLog = MIN(params->bucketSizeLog, params->hashLog); +} + +size_t ZSTD_ldm_getTableSize(ldmParams_t params) +{ + size_t const ldmHSize = ((size_t)1) << params.hashLog; + size_t const ldmBucketSizeLog = MIN(params.bucketSizeLog, params.hashLog); + size_t const ldmBucketSize = ((size_t)1) << (params.hashLog - ldmBucketSizeLog); + size_t const totalSize = ZSTD_cwksp_alloc_size(ldmBucketSize) + + ZSTD_cwksp_alloc_size(ldmHSize * sizeof(ldmEntry_t)); + return params.enableLdm ? totalSize : 0; +} + +size_t ZSTD_ldm_getMaxNbSeq(ldmParams_t params, size_t maxChunkSize) +{ + return params.enableLdm ? (maxChunkSize / params.minMatchLength) : 0; +} + +/* ZSTD_ldm_getBucket() : + * Returns a pointer to the start of the bucket associated with hash. */ +static ldmEntry_t* ZSTD_ldm_getBucket( + ldmState_t* ldmState, size_t hash, ldmParams_t const ldmParams) +{ + return ldmState->hashTable + (hash << ldmParams.bucketSizeLog); +} + +/* ZSTD_ldm_insertEntry() : + * Insert the entry with corresponding hash into the hash table */ +static void ZSTD_ldm_insertEntry(ldmState_t* ldmState, + size_t const hash, const ldmEntry_t entry, + ldmParams_t const ldmParams) +{ + BYTE* const pOffset = ldmState->bucketOffsets + hash; + unsigned const offset = *pOffset; + + *(ZSTD_ldm_getBucket(ldmState, hash, ldmParams) + offset) = entry; + *pOffset = (BYTE)((offset + 1) & ((1u << ldmParams.bucketSizeLog) - 1)); + +} + +/* ZSTD_ldm_countBackwardsMatch() : + * Returns the number of bytes that match backwards before pIn and pMatch. + * + * We count only bytes where pMatch >= pBase and pIn >= pAnchor. */ +static size_t ZSTD_ldm_countBackwardsMatch( + const BYTE* pIn, const BYTE* pAnchor, + const BYTE* pMatch, const BYTE* pMatchBase) +{ + size_t matchLength = 0; + while (pIn > pAnchor && pMatch > pMatchBase && pIn[-1] == pMatch[-1]) { + pIn--; + pMatch--; + matchLength++; + } + return matchLength; +} + +/* ZSTD_ldm_countBackwardsMatch_2segments() : + * Returns the number of bytes that match backwards from pMatch, + * even with the backwards match spanning 2 different segments. + * + * On reaching `pMatchBase`, start counting from mEnd */ +static size_t ZSTD_ldm_countBackwardsMatch_2segments( + const BYTE* pIn, const BYTE* pAnchor, + const BYTE* pMatch, const BYTE* pMatchBase, + const BYTE* pExtDictStart, const BYTE* pExtDictEnd) +{ + size_t matchLength = ZSTD_ldm_countBackwardsMatch(pIn, pAnchor, pMatch, pMatchBase); + if (pMatch - matchLength != pMatchBase || pMatchBase == pExtDictStart) { + /* If backwards match is entirely in the extDict or prefix, immediately return */ + return matchLength; + } + DEBUGLOG(7, "ZSTD_ldm_countBackwardsMatch_2segments: found 2-parts backwards match (length in prefix==%zu)", matchLength); + matchLength += ZSTD_ldm_countBackwardsMatch(pIn - matchLength, pAnchor, pExtDictEnd, pExtDictStart); + DEBUGLOG(7, "final backwards match length = %zu", matchLength); + return matchLength; +} + +/* ZSTD_ldm_fillFastTables() : + * + * Fills the relevant tables for the ZSTD_fast and ZSTD_dfast strategies. + * This is similar to ZSTD_loadDictionaryContent. + * + * The tables for the other strategies are filled within their + * block compressors. */ +static size_t ZSTD_ldm_fillFastTables(ZSTD_matchState_t* ms, + void const* end) +{ + const BYTE* const iend = (const BYTE*)end; + + switch(ms->cParams.strategy) + { + case ZSTD_fast: + ZSTD_fillHashTable(ms, iend, ZSTD_dtlm_fast); + break; + + case ZSTD_dfast: + ZSTD_fillDoubleHashTable(ms, iend, ZSTD_dtlm_fast); + break; + + case ZSTD_greedy: + case ZSTD_lazy: + case ZSTD_lazy2: + case ZSTD_btlazy2: + case ZSTD_btopt: + case ZSTD_btultra: + case ZSTD_btultra2: + break; + default: + assert(0); /* not possible : not a valid strategy id */ + } + + return 0; +} + +void ZSTD_ldm_fillHashTable( + ldmState_t* ldmState, const BYTE* ip, + const BYTE* iend, ldmParams_t const* params) +{ + U32 const minMatchLength = params->minMatchLength; + U32 const hBits = params->hashLog - params->bucketSizeLog; + BYTE const* const base = ldmState->window.base; + BYTE const* const istart = ip; + ldmRollingHashState_t hashState; + size_t* const splits = ldmState->splitIndices; + unsigned numSplits; + + DEBUGLOG(5, "ZSTD_ldm_fillHashTable"); + + ZSTD_ldm_gear_init(&hashState, params); + while (ip < iend) { + size_t hashed; + unsigned n; + + numSplits = 0; + hashed = ZSTD_ldm_gear_feed(&hashState, ip, iend - ip, splits, &numSplits); + + for (n = 0; n < numSplits; n++) { + if (ip + splits[n] >= istart + minMatchLength) { + BYTE const* const split = ip + splits[n] - minMatchLength; + U64 const xxhash = xxh64(split, minMatchLength, 0); + U32 const hash = (U32)(xxhash & (((U32)1 << hBits) - 1)); + ldmEntry_t entry; + + entry.offset = (U32)(split - base); + entry.checksum = (U32)(xxhash >> 32); + ZSTD_ldm_insertEntry(ldmState, hash, entry, *params); + } + } + + ip += hashed; + } +} + + +/* ZSTD_ldm_limitTableUpdate() : + * + * Sets cctx->nextToUpdate to a position corresponding closer to anchor + * if it is far way + * (after a long match, only update tables a limited amount). */ +static void ZSTD_ldm_limitTableUpdate(ZSTD_matchState_t* ms, const BYTE* anchor) +{ + U32 const curr = (U32)(anchor - ms->window.base); + if (curr > ms->nextToUpdate + 1024) { + ms->nextToUpdate = + curr - MIN(512, curr - ms->nextToUpdate - 1024); + } +} + +static size_t ZSTD_ldm_generateSequences_internal( + ldmState_t* ldmState, rawSeqStore_t* rawSeqStore, + ldmParams_t const* params, void const* src, size_t srcSize) +{ + /* LDM parameters */ + int const extDict = ZSTD_window_hasExtDict(ldmState->window); + U32 const minMatchLength = params->minMatchLength; + U32 const entsPerBucket = 1U << params->bucketSizeLog; + U32 const hBits = params->hashLog - params->bucketSizeLog; + /* Prefix and extDict parameters */ + U32 const dictLimit = ldmState->window.dictLimit; + U32 const lowestIndex = extDict ? ldmState->window.lowLimit : dictLimit; + BYTE const* const base = ldmState->window.base; + BYTE const* const dictBase = extDict ? ldmState->window.dictBase : NULL; + BYTE const* const dictStart = extDict ? dictBase + lowestIndex : NULL; + BYTE const* const dictEnd = extDict ? dictBase + dictLimit : NULL; + BYTE const* const lowPrefixPtr = base + dictLimit; + /* Input bounds */ + BYTE const* const istart = (BYTE const*)src; + BYTE const* const iend = istart + srcSize; + BYTE const* const ilimit = iend - HASH_READ_SIZE; + /* Input positions */ + BYTE const* anchor = istart; + BYTE const* ip = istart; + /* Rolling hash state */ + ldmRollingHashState_t hashState; + /* Arrays for staged-processing */ + size_t* const splits = ldmState->splitIndices; + ldmMatchCandidate_t* const candidates = ldmState->matchCandidates; + unsigned numSplits; + + if (srcSize < minMatchLength) + return iend - anchor; + + /* Initialize the rolling hash state with the first minMatchLength bytes */ + ZSTD_ldm_gear_init(&hashState, params); + { + size_t n = 0; + + while (n < minMatchLength) { + numSplits = 0; + n += ZSTD_ldm_gear_feed(&hashState, ip + n, minMatchLength - n, + splits, &numSplits); + } + ip += minMatchLength; + } + + while (ip < ilimit) { + size_t hashed; + unsigned n; + + numSplits = 0; + hashed = ZSTD_ldm_gear_feed(&hashState, ip, ilimit - ip, + splits, &numSplits); + + for (n = 0; n < numSplits; n++) { + BYTE const* const split = ip + splits[n] - minMatchLength; + U64 const xxhash = xxh64(split, minMatchLength, 0); + U32 const hash = (U32)(xxhash & (((U32)1 << hBits) - 1)); + + candidates[n].split = split; + candidates[n].hash = hash; + candidates[n].checksum = (U32)(xxhash >> 32); + candidates[n].bucket = ZSTD_ldm_getBucket(ldmState, hash, *params); + PREFETCH_L1(candidates[n].bucket); + } + + for (n = 0; n < numSplits; n++) { + size_t forwardMatchLength = 0, backwardMatchLength = 0, + bestMatchLength = 0, mLength; + BYTE const* const split = candidates[n].split; + U32 const checksum = candidates[n].checksum; + U32 const hash = candidates[n].hash; + ldmEntry_t* const bucket = candidates[n].bucket; + ldmEntry_t const* cur; + ldmEntry_t const* bestEntry = NULL; + ldmEntry_t newEntry; + + newEntry.offset = (U32)(split - base); + newEntry.checksum = checksum; + + /* If a split point would generate a sequence overlapping with + * the previous one, we merely register it in the hash table and + * move on */ + if (split < anchor) { + ZSTD_ldm_insertEntry(ldmState, hash, newEntry, *params); + continue; + } + + for (cur = bucket; cur < bucket + entsPerBucket; cur++) { + size_t curForwardMatchLength, curBackwardMatchLength, + curTotalMatchLength; + if (cur->checksum != checksum || cur->offset <= lowestIndex) { + continue; + } + if (extDict) { + BYTE const* const curMatchBase = + cur->offset < dictLimit ? dictBase : base; + BYTE const* const pMatch = curMatchBase + cur->offset; + BYTE const* const matchEnd = + cur->offset < dictLimit ? dictEnd : iend; + BYTE const* const lowMatchPtr = + cur->offset < dictLimit ? dictStart : lowPrefixPtr; + curForwardMatchLength = + ZSTD_count_2segments(split, pMatch, iend, matchEnd, lowPrefixPtr); + if (curForwardMatchLength < minMatchLength) { + continue; + } + curBackwardMatchLength = ZSTD_ldm_countBackwardsMatch_2segments( + split, anchor, pMatch, lowMatchPtr, dictStart, dictEnd); + } else { /* !extDict */ + BYTE const* const pMatch = base + cur->offset; + curForwardMatchLength = ZSTD_count(split, pMatch, iend); + if (curForwardMatchLength < minMatchLength) { + continue; + } + curBackwardMatchLength = + ZSTD_ldm_countBackwardsMatch(split, anchor, pMatch, lowPrefixPtr); + } + curTotalMatchLength = curForwardMatchLength + curBackwardMatchLength; + + if (curTotalMatchLength > bestMatchLength) { + bestMatchLength = curTotalMatchLength; + forwardMatchLength = curForwardMatchLength; + backwardMatchLength = curBackwardMatchLength; + bestEntry = cur; + } + } + + /* No match found -- insert an entry into the hash table + * and process the next candidate match */ + if (bestEntry == NULL) { + ZSTD_ldm_insertEntry(ldmState, hash, newEntry, *params); + continue; + } + + /* Match found */ + mLength = forwardMatchLength + backwardMatchLength; + { + U32 const offset = (U32)(split - base) - bestEntry->offset; + rawSeq* const seq = rawSeqStore->seq + rawSeqStore->size; + + /* Out of sequence storage */ + if (rawSeqStore->size == rawSeqStore->capacity) + return ERROR(dstSize_tooSmall); + seq->litLength = (U32)(split - backwardMatchLength - anchor); + seq->matchLength = (U32)mLength; + seq->offset = offset; + rawSeqStore->size++; + } + + /* Insert the current entry into the hash table --- it must be + * done after the previous block to avoid clobbering bestEntry */ + ZSTD_ldm_insertEntry(ldmState, hash, newEntry, *params); + + anchor = split + forwardMatchLength; + } + + ip += hashed; + } + + return iend - anchor; +} + +/*! ZSTD_ldm_reduceTable() : + * reduce table indexes by `reducerValue` */ +static void ZSTD_ldm_reduceTable(ldmEntry_t* const table, U32 const size, + U32 const reducerValue) +{ + U32 u; + for (u = 0; u < size; u++) { + if (table[u].offset < reducerValue) table[u].offset = 0; + else table[u].offset -= reducerValue; + } +} + +size_t ZSTD_ldm_generateSequences( + ldmState_t* ldmState, rawSeqStore_t* sequences, + ldmParams_t const* params, void const* src, size_t srcSize) +{ + U32 const maxDist = 1U << params->windowLog; + BYTE const* const istart = (BYTE const*)src; + BYTE const* const iend = istart + srcSize; + size_t const kMaxChunkSize = 1 << 20; + size_t const nbChunks = (srcSize / kMaxChunkSize) + ((srcSize % kMaxChunkSize) != 0); + size_t chunk; + size_t leftoverSize = 0; + + assert(ZSTD_CHUNKSIZE_MAX >= kMaxChunkSize); + /* Check that ZSTD_window_update() has been called for this chunk prior + * to passing it to this function. + */ + assert(ldmState->window.nextSrc >= (BYTE const*)src + srcSize); + /* The input could be very large (in zstdmt), so it must be broken up into + * chunks to enforce the maximum distance and handle overflow correction. + */ + assert(sequences->pos <= sequences->size); + assert(sequences->size <= sequences->capacity); + for (chunk = 0; chunk < nbChunks && sequences->size < sequences->capacity; ++chunk) { + BYTE const* const chunkStart = istart + chunk * kMaxChunkSize; + size_t const remaining = (size_t)(iend - chunkStart); + BYTE const *const chunkEnd = + (remaining < kMaxChunkSize) ? iend : chunkStart + kMaxChunkSize; + size_t const chunkSize = chunkEnd - chunkStart; + size_t newLeftoverSize; + size_t const prevSize = sequences->size; + + assert(chunkStart < iend); + /* 1. Perform overflow correction if necessary. */ + if (ZSTD_window_needOverflowCorrection(ldmState->window, chunkEnd)) { + U32 const ldmHSize = 1U << params->hashLog; + U32 const correction = ZSTD_window_correctOverflow( + &ldmState->window, /* cycleLog */ 0, maxDist, chunkStart); + ZSTD_ldm_reduceTable(ldmState->hashTable, ldmHSize, correction); + /* invalidate dictionaries on overflow correction */ + ldmState->loadedDictEnd = 0; + } + /* 2. We enforce the maximum offset allowed. + * + * kMaxChunkSize should be small enough that we don't lose too much of + * the window through early invalidation. + * TODO: * Test the chunk size. + * * Try invalidation after the sequence generation and test the + * the offset against maxDist directly. + * + * NOTE: Because of dictionaries + sequence splitting we MUST make sure + * that any offset used is valid at the END of the sequence, since it may + * be split into two sequences. This condition holds when using + * ZSTD_window_enforceMaxDist(), but if we move to checking offsets + * against maxDist directly, we'll have to carefully handle that case. + */ + ZSTD_window_enforceMaxDist(&ldmState->window, chunkEnd, maxDist, &ldmState->loadedDictEnd, NULL); + /* 3. Generate the sequences for the chunk, and get newLeftoverSize. */ + newLeftoverSize = ZSTD_ldm_generateSequences_internal( + ldmState, sequences, params, chunkStart, chunkSize); + if (ZSTD_isError(newLeftoverSize)) + return newLeftoverSize; + /* 4. We add the leftover literals from previous iterations to the first + * newly generated sequence, or add the `newLeftoverSize` if none are + * generated. + */ + /* Prepend the leftover literals from the last call */ + if (prevSize < sequences->size) { + sequences->seq[prevSize].litLength += (U32)leftoverSize; + leftoverSize = newLeftoverSize; + } else { + assert(newLeftoverSize == chunkSize); + leftoverSize += chunkSize; + } + } + return 0; +} + +void ZSTD_ldm_skipSequences(rawSeqStore_t* rawSeqStore, size_t srcSize, U32 const minMatch) { + while (srcSize > 0 && rawSeqStore->pos < rawSeqStore->size) { + rawSeq* seq = rawSeqStore->seq + rawSeqStore->pos; + if (srcSize <= seq->litLength) { + /* Skip past srcSize literals */ + seq->litLength -= (U32)srcSize; + return; + } + srcSize -= seq->litLength; + seq->litLength = 0; + if (srcSize < seq->matchLength) { + /* Skip past the first srcSize of the match */ + seq->matchLength -= (U32)srcSize; + if (seq->matchLength < minMatch) { + /* The match is too short, omit it */ + if (rawSeqStore->pos + 1 < rawSeqStore->size) { + seq[1].litLength += seq[0].matchLength; + } + rawSeqStore->pos++; + } + return; + } + srcSize -= seq->matchLength; + seq->matchLength = 0; + rawSeqStore->pos++; + } +} + +/* + * If the sequence length is longer than remaining then the sequence is split + * between this block and the next. + * + * Returns the current sequence to handle, or if the rest of the block should + * be literals, it returns a sequence with offset == 0. + */ +static rawSeq maybeSplitSequence(rawSeqStore_t* rawSeqStore, + U32 const remaining, U32 const minMatch) +{ + rawSeq sequence = rawSeqStore->seq[rawSeqStore->pos]; + assert(sequence.offset > 0); + /* Likely: No partial sequence */ + if (remaining >= sequence.litLength + sequence.matchLength) { + rawSeqStore->pos++; + return sequence; + } + /* Cut the sequence short (offset == 0 ==> rest is literals). */ + if (remaining <= sequence.litLength) { + sequence.offset = 0; + } else if (remaining < sequence.litLength + sequence.matchLength) { + sequence.matchLength = remaining - sequence.litLength; + if (sequence.matchLength < minMatch) { + sequence.offset = 0; + } + } + /* Skip past `remaining` bytes for the future sequences. */ + ZSTD_ldm_skipSequences(rawSeqStore, remaining, minMatch); + return sequence; +} + +void ZSTD_ldm_skipRawSeqStoreBytes(rawSeqStore_t* rawSeqStore, size_t nbBytes) { + U32 currPos = (U32)(rawSeqStore->posInSequence + nbBytes); + while (currPos && rawSeqStore->pos < rawSeqStore->size) { + rawSeq currSeq = rawSeqStore->seq[rawSeqStore->pos]; + if (currPos >= currSeq.litLength + currSeq.matchLength) { + currPos -= currSeq.litLength + currSeq.matchLength; + rawSeqStore->pos++; + } else { + rawSeqStore->posInSequence = currPos; + break; + } + } + if (currPos == 0 || rawSeqStore->pos == rawSeqStore->size) { + rawSeqStore->posInSequence = 0; + } +} + +size_t ZSTD_ldm_blockCompress(rawSeqStore_t* rawSeqStore, + ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM], + void const* src, size_t srcSize) +{ + const ZSTD_compressionParameters* const cParams = &ms->cParams; + unsigned const minMatch = cParams->minMatch; + ZSTD_blockCompressor const blockCompressor = + ZSTD_selectBlockCompressor(cParams->strategy, ZSTD_matchState_dictMode(ms)); + /* Input bounds */ + BYTE const* const istart = (BYTE const*)src; + BYTE const* const iend = istart + srcSize; + /* Input positions */ + BYTE const* ip = istart; + + DEBUGLOG(5, "ZSTD_ldm_blockCompress: srcSize=%zu", srcSize); + /* If using opt parser, use LDMs only as candidates rather than always accepting them */ + if (cParams->strategy >= ZSTD_btopt) { + size_t lastLLSize; + ms->ldmSeqStore = rawSeqStore; + lastLLSize = blockCompressor(ms, seqStore, rep, src, srcSize); + ZSTD_ldm_skipRawSeqStoreBytes(rawSeqStore, srcSize); + return lastLLSize; + } + + assert(rawSeqStore->pos <= rawSeqStore->size); + assert(rawSeqStore->size <= rawSeqStore->capacity); + /* Loop through each sequence and apply the block compressor to the literals */ + while (rawSeqStore->pos < rawSeqStore->size && ip < iend) { + /* maybeSplitSequence updates rawSeqStore->pos */ + rawSeq const sequence = maybeSplitSequence(rawSeqStore, + (U32)(iend - ip), minMatch); + int i; + /* End signal */ + if (sequence.offset == 0) + break; + + assert(ip + sequence.litLength + sequence.matchLength <= iend); + + /* Fill tables for block compressor */ + ZSTD_ldm_limitTableUpdate(ms, ip); + ZSTD_ldm_fillFastTables(ms, ip); + /* Run the block compressor */ + DEBUGLOG(5, "pos %u : calling block compressor on segment of size %u", (unsigned)(ip-istart), sequence.litLength); + { + size_t const newLitLength = + blockCompressor(ms, seqStore, rep, ip, sequence.litLength); + ip += sequence.litLength; + /* Update the repcodes */ + for (i = ZSTD_REP_NUM - 1; i > 0; i--) + rep[i] = rep[i-1]; + rep[0] = sequence.offset; + /* Store the sequence */ + ZSTD_storeSeq(seqStore, newLitLength, ip - newLitLength, iend, + sequence.offset + ZSTD_REP_MOVE, + sequence.matchLength - MINMATCH); + ip += sequence.matchLength; + } + } + /* Fill the tables for the block compressor */ + ZSTD_ldm_limitTableUpdate(ms, ip); + ZSTD_ldm_fillFastTables(ms, ip); + /* Compress the last literals */ + return blockCompressor(ms, seqStore, rep, ip, iend - ip); +} diff --git a/lib/zstd/compress/zstd_ldm.h b/lib/zstd/compress/zstd_ldm.h new file mode 100644 index 000000000000..25b25270b72e --- /dev/null +++ b/lib/zstd/compress/zstd_ldm.h @@ -0,0 +1,110 @@ +/* + * Copyright (c) Yann Collet, Facebook, Inc. + * All rights reserved. + * + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). + * You may select, at your option, one of the above-listed licenses. + */ + +#ifndef ZSTD_LDM_H +#define ZSTD_LDM_H + + +#include "zstd_compress_internal.h" /* ldmParams_t, U32 */ +#include /* ZSTD_CCtx, size_t */ + +/*-************************************* +* Long distance matching +***************************************/ + +#define ZSTD_LDM_DEFAULT_WINDOW_LOG ZSTD_WINDOWLOG_LIMIT_DEFAULT + +void ZSTD_ldm_fillHashTable( + ldmState_t* state, const BYTE* ip, + const BYTE* iend, ldmParams_t const* params); + +/* + * ZSTD_ldm_generateSequences(): + * + * Generates the sequences using the long distance match finder. + * Generates long range matching sequences in `sequences`, which parse a prefix + * of the source. `sequences` must be large enough to store every sequence, + * which can be checked with `ZSTD_ldm_getMaxNbSeq()`. + * @returns 0 or an error code. + * + * NOTE: The user must have called ZSTD_window_update() for all of the input + * they have, even if they pass it to ZSTD_ldm_generateSequences() in chunks. + * NOTE: This function returns an error if it runs out of space to store + * sequences. + */ +size_t ZSTD_ldm_generateSequences( + ldmState_t* ldms, rawSeqStore_t* sequences, + ldmParams_t const* params, void const* src, size_t srcSize); + +/* + * ZSTD_ldm_blockCompress(): + * + * Compresses a block using the predefined sequences, along with a secondary + * block compressor. The literals section of every sequence is passed to the + * secondary block compressor, and those sequences are interspersed with the + * predefined sequences. Returns the length of the last literals. + * Updates `rawSeqStore.pos` to indicate how many sequences have been consumed. + * `rawSeqStore.seq` may also be updated to split the last sequence between two + * blocks. + * @return The length of the last literals. + * + * NOTE: The source must be at most the maximum block size, but the predefined + * sequences can be any size, and may be longer than the block. In the case that + * they are longer than the block, the last sequences may need to be split into + * two. We handle that case correctly, and update `rawSeqStore` appropriately. + * NOTE: This function does not return any errors. + */ +size_t ZSTD_ldm_blockCompress(rawSeqStore_t* rawSeqStore, + ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM], + void const* src, size_t srcSize); + +/* + * ZSTD_ldm_skipSequences(): + * + * Skip past `srcSize` bytes worth of sequences in `rawSeqStore`. + * Avoids emitting matches less than `minMatch` bytes. + * Must be called for data that is not passed to ZSTD_ldm_blockCompress(). + */ +void ZSTD_ldm_skipSequences(rawSeqStore_t* rawSeqStore, size_t srcSize, + U32 const minMatch); + +/* ZSTD_ldm_skipRawSeqStoreBytes(): + * Moves forward in rawSeqStore by nbBytes, updating fields 'pos' and 'posInSequence'. + * Not to be used in conjunction with ZSTD_ldm_skipSequences(). + * Must be called for data with is not passed to ZSTD_ldm_blockCompress(). + */ +void ZSTD_ldm_skipRawSeqStoreBytes(rawSeqStore_t* rawSeqStore, size_t nbBytes); + +/* ZSTD_ldm_getTableSize() : + * Estimate the space needed for long distance matching tables or 0 if LDM is + * disabled. + */ +size_t ZSTD_ldm_getTableSize(ldmParams_t params); + +/* ZSTD_ldm_getSeqSpace() : + * Return an upper bound on the number of sequences that can be produced by + * the long distance matcher, or 0 if LDM is disabled. + */ +size_t ZSTD_ldm_getMaxNbSeq(ldmParams_t params, size_t maxChunkSize); + +/* ZSTD_ldm_adjustParameters() : + * If the params->hashRateLog is not set, set it to its default value based on + * windowLog and params->hashLog. + * + * Ensures that params->bucketSizeLog is <= params->hashLog (setting it to + * params->hashLog if it is not). + * + * Ensures that the minMatchLength >= targetLength during optimal parsing. + */ +void ZSTD_ldm_adjustParameters(ldmParams_t* params, + ZSTD_compressionParameters const* cParams); + + +#endif /* ZSTD_FAST_H */ diff --git a/lib/zstd/compress/zstd_ldm_geartab.h b/lib/zstd/compress/zstd_ldm_geartab.h new file mode 100644 index 000000000000..e5c24d856b0a --- /dev/null +++ b/lib/zstd/compress/zstd_ldm_geartab.h @@ -0,0 +1,103 @@ +/* + * Copyright (c) Yann Collet, Facebook, Inc. + * All rights reserved. + * + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). + * You may select, at your option, one of the above-listed licenses. + */ + +#ifndef ZSTD_LDM_GEARTAB_H +#define ZSTD_LDM_GEARTAB_H + +static U64 ZSTD_ldm_gearTab[256] = { + 0xf5b8f72c5f77775c, 0x84935f266b7ac412, 0xb647ada9ca730ccc, + 0xb065bb4b114fb1de, 0x34584e7e8c3a9fd0, 0x4e97e17c6ae26b05, + 0x3a03d743bc99a604, 0xcecd042422c4044f, 0x76de76c58524259e, + 0x9c8528f65badeaca, 0x86563706e2097529, 0x2902475fa375d889, + 0xafb32a9739a5ebe6, 0xce2714da3883e639, 0x21eaf821722e69e, + 0x37b628620b628, 0x49a8d455d88caf5, 0x8556d711e6958140, + 0x4f7ae74fc605c1f, 0x829f0c3468bd3a20, 0x4ffdc885c625179e, + 0x8473de048a3daf1b, 0x51008822b05646b2, 0x69d75d12b2d1cc5f, + 0x8c9d4a19159154bc, 0xc3cc10f4abbd4003, 0xd06ddc1cecb97391, + 0xbe48e6e7ed80302e, 0x3481db31cee03547, 0xacc3f67cdaa1d210, + 0x65cb771d8c7f96cc, 0x8eb27177055723dd, 0xc789950d44cd94be, + 0x934feadc3700b12b, 0x5e485f11edbdf182, 0x1e2e2a46fd64767a, + 0x2969ca71d82efa7c, 0x9d46e9935ebbba2e, 0xe056b67e05e6822b, + 0x94d73f55739d03a0, 0xcd7010bdb69b5a03, 0x455ef9fcd79b82f4, + 0x869cb54a8749c161, 0x38d1a4fa6185d225, 0xb475166f94bbe9bb, + 0xa4143548720959f1, 0x7aed4780ba6b26ba, 0xd0ce264439e02312, + 0x84366d746078d508, 0xa8ce973c72ed17be, 0x21c323a29a430b01, + 0x9962d617e3af80ee, 0xab0ce91d9c8cf75b, 0x530e8ee6d19a4dbc, + 0x2ef68c0cf53f5d72, 0xc03a681640a85506, 0x496e4e9f9c310967, + 0x78580472b59b14a0, 0x273824c23b388577, 0x66bf923ad45cb553, + 0x47ae1a5a2492ba86, 0x35e304569e229659, 0x4765182a46870b6f, + 0x6cbab625e9099412, 0xddac9a2e598522c1, 0x7172086e666624f2, + 0xdf5003ca503b7837, 0x88c0c1db78563d09, 0x58d51865acfc289d, + 0x177671aec65224f1, 0xfb79d8a241e967d7, 0x2be1e101cad9a49a, + 0x6625682f6e29186b, 0x399553457ac06e50, 0x35dffb4c23abb74, + 0x429db2591f54aade, 0xc52802a8037d1009, 0x6acb27381f0b25f3, + 0xf45e2551ee4f823b, 0x8b0ea2d99580c2f7, 0x3bed519cbcb4e1e1, + 0xff452823dbb010a, 0x9d42ed614f3dd267, 0x5b9313c06257c57b, + 0xa114b8008b5e1442, 0xc1fe311c11c13d4b, 0x66e8763ea34c5568, + 0x8b982af1c262f05d, 0xee8876faaa75fbb7, 0x8a62a4d0d172bb2a, + 0xc13d94a3b7449a97, 0x6dbbba9dc15d037c, 0xc786101f1d92e0f1, + 0xd78681a907a0b79b, 0xf61aaf2962c9abb9, 0x2cfd16fcd3cb7ad9, + 0x868c5b6744624d21, 0x25e650899c74ddd7, 0xba042af4a7c37463, + 0x4eb1a539465a3eca, 0xbe09dbf03b05d5ca, 0x774e5a362b5472ba, + 0x47a1221229d183cd, 0x504b0ca18ef5a2df, 0xdffbdfbde2456eb9, + 0x46cd2b2fbee34634, 0xf2aef8fe819d98c3, 0x357f5276d4599d61, + 0x24a5483879c453e3, 0x88026889192b4b9, 0x28da96671782dbec, + 0x4ef37c40588e9aaa, 0x8837b90651bc9fb3, 0xc164f741d3f0e5d6, + 0xbc135a0a704b70ba, 0x69cd868f7622ada, 0xbc37ba89e0b9c0ab, + 0x47c14a01323552f6, 0x4f00794bacee98bb, 0x7107de7d637a69d5, + 0x88af793bb6f2255e, 0xf3c6466b8799b598, 0xc288c616aa7f3b59, + 0x81ca63cf42fca3fd, 0x88d85ace36a2674b, 0xd056bd3792389e7, + 0xe55c396c4e9dd32d, 0xbefb504571e6c0a6, 0x96ab32115e91e8cc, + 0xbf8acb18de8f38d1, 0x66dae58801672606, 0x833b6017872317fb, + 0xb87c16f2d1c92864, 0xdb766a74e58b669c, 0x89659f85c61417be, + 0xc8daad856011ea0c, 0x76a4b565b6fe7eae, 0xa469d085f6237312, + 0xaaf0365683a3e96c, 0x4dbb746f8424f7b8, 0x638755af4e4acc1, + 0x3d7807f5bde64486, 0x17be6d8f5bbb7639, 0x903f0cd44dc35dc, + 0x67b672eafdf1196c, 0xa676ff93ed4c82f1, 0x521d1004c5053d9d, + 0x37ba9ad09ccc9202, 0x84e54d297aacfb51, 0xa0b4b776a143445, + 0x820d471e20b348e, 0x1874383cb83d46dc, 0x97edeec7a1efe11c, + 0xb330e50b1bdc42aa, 0x1dd91955ce70e032, 0xa514cdb88f2939d5, + 0x2791233fd90db9d3, 0x7b670a4cc50f7a9b, 0x77c07d2a05c6dfa5, + 0xe3778b6646d0a6fa, 0xb39c8eda47b56749, 0x933ed448addbef28, + 0xaf846af6ab7d0bf4, 0xe5af208eb666e49, 0x5e6622f73534cd6a, + 0x297daeca42ef5b6e, 0x862daef3d35539a6, 0xe68722498f8e1ea9, + 0x981c53093dc0d572, 0xfa09b0bfbf86fbf5, 0x30b1e96166219f15, + 0x70e7d466bdc4fb83, 0x5a66736e35f2a8e9, 0xcddb59d2b7c1baef, + 0xd6c7d247d26d8996, 0xea4e39eac8de1ba3, 0x539c8bb19fa3aff2, + 0x9f90e4c5fd508d8, 0xa34e5956fbaf3385, 0x2e2f8e151d3ef375, + 0x173691e9b83faec1, 0xb85a8d56bf016379, 0x8382381267408ae3, + 0xb90f901bbdc0096d, 0x7c6ad32933bcec65, 0x76bb5e2f2c8ad595, + 0x390f851a6cf46d28, 0xc3e6064da1c2da72, 0xc52a0c101cfa5389, + 0xd78eaf84a3fbc530, 0x3781b9e2288b997e, 0x73c2f6dea83d05c4, + 0x4228e364c5b5ed7, 0x9d7a3edf0da43911, 0x8edcfeda24686756, + 0x5e7667a7b7a9b3a1, 0x4c4f389fa143791d, 0xb08bc1023da7cddc, + 0x7ab4be3ae529b1cc, 0x754e6132dbe74ff9, 0x71635442a839df45, + 0x2f6fb1643fbe52de, 0x961e0a42cf7a8177, 0xf3b45d83d89ef2ea, + 0xee3de4cf4a6e3e9b, 0xcd6848542c3295e7, 0xe4cee1664c78662f, + 0x9947548b474c68c4, 0x25d73777a5ed8b0b, 0xc915b1d636b7fc, + 0x21c2ba75d9b0d2da, 0x5f6b5dcf608a64a1, 0xdcf333255ff9570c, + 0x633b922418ced4ee, 0xc136dde0b004b34a, 0x58cc83b05d4b2f5a, + 0x5eb424dda28e42d2, 0x62df47369739cd98, 0xb4e0b42485e4ce17, + 0x16e1f0c1f9a8d1e7, 0x8ec3916707560ebf, 0x62ba6e2df2cc9db3, + 0xcbf9f4ff77d83a16, 0x78d9d7d07d2bbcc4, 0xef554ce1e02c41f4, + 0x8d7581127eccf94d, 0xa9b53336cb3c8a05, 0x38c42c0bf45c4f91, + 0x640893cdf4488863, 0x80ec34bc575ea568, 0x39f324f5b48eaa40, + 0xe9d9ed1f8eff527f, 0x9224fc058cc5a214, 0xbaba00b04cfe7741, + 0x309a9f120fcf52af, 0xa558f3ec65626212, 0x424bec8b7adabe2f, + 0x41622513a6aea433, 0xb88da2d5324ca798, 0xd287733b245528a4, + 0x9a44697e6d68aec3, 0x7b1093be2f49bb28, 0x50bbec632e3d8aad, + 0x6cd90723e1ea8283, 0x897b9e7431b02bf3, 0x219efdcb338a7047, + 0x3b0311f0a27c0656, 0xdb17bf91c0db96e7, 0x8cd4fd6b4e85a5b2, + 0xfab071054ba6409d, 0x40d6fe831fa9dfd9, 0xaf358debad7d791e, + 0xeb8d0e25a65e3e58, 0xbbcbd3df14e08580, 0xcf751f27ecdab2b, + 0x2b4da14f2613d8f4 +}; + +#endif /* ZSTD_LDM_GEARTAB_H */ diff --git a/lib/zstd/compress/zstd_opt.c b/lib/zstd/compress/zstd_opt.c new file mode 100644 index 000000000000..04337050fe9a --- /dev/null +++ b/lib/zstd/compress/zstd_opt.c @@ -0,0 +1,1346 @@ +/* + * Copyright (c) Przemyslaw Skibinski, Yann Collet, Facebook, Inc. + * All rights reserved. + * + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). + * You may select, at your option, one of the above-listed licenses. + */ + +#include "zstd_compress_internal.h" +#include "hist.h" +#include "zstd_opt.h" + + +#define ZSTD_LITFREQ_ADD 2 /* scaling factor for litFreq, so that frequencies adapt faster to new stats */ +#define ZSTD_FREQ_DIV 4 /* log factor when using previous stats to init next stats */ +#define ZSTD_MAX_PRICE (1<<30) + +#define ZSTD_PREDEF_THRESHOLD 1024 /* if srcSize < ZSTD_PREDEF_THRESHOLD, symbols' cost is assumed static, directly determined by pre-defined distributions */ + + +/*-************************************* +* Price functions for optimal parser +***************************************/ + +#if 0 /* approximation at bit level */ +# define BITCOST_ACCURACY 0 +# define BITCOST_MULTIPLIER (1 << BITCOST_ACCURACY) +# define WEIGHT(stat) ((void)opt, ZSTD_bitWeight(stat)) +#elif 0 /* fractional bit accuracy */ +# define BITCOST_ACCURACY 8 +# define BITCOST_MULTIPLIER (1 << BITCOST_ACCURACY) +# define WEIGHT(stat,opt) ((void)opt, ZSTD_fracWeight(stat)) +#else /* opt==approx, ultra==accurate */ +# define BITCOST_ACCURACY 8 +# define BITCOST_MULTIPLIER (1 << BITCOST_ACCURACY) +# define WEIGHT(stat,opt) (opt ? ZSTD_fracWeight(stat) : ZSTD_bitWeight(stat)) +#endif + +MEM_STATIC U32 ZSTD_bitWeight(U32 stat) +{ + return (ZSTD_highbit32(stat+1) * BITCOST_MULTIPLIER); +} + +MEM_STATIC U32 ZSTD_fracWeight(U32 rawStat) +{ + U32 const stat = rawStat + 1; + U32 const hb = ZSTD_highbit32(stat); + U32 const BWeight = hb * BITCOST_MULTIPLIER; + U32 const FWeight = (stat << BITCOST_ACCURACY) >> hb; + U32 const weight = BWeight + FWeight; + assert(hb + BITCOST_ACCURACY < 31); + return weight; +} + +#if (DEBUGLEVEL>=2) +/* debugging function, + * @return price in bytes as fractional value + * for debug messages only */ +MEM_STATIC double ZSTD_fCost(U32 price) +{ + return (double)price / (BITCOST_MULTIPLIER*8); +} +#endif + +static int ZSTD_compressedLiterals(optState_t const* const optPtr) +{ + return optPtr->literalCompressionMode != ZSTD_lcm_uncompressed; +} + +static void ZSTD_setBasePrices(optState_t* optPtr, int optLevel) +{ + if (ZSTD_compressedLiterals(optPtr)) + optPtr->litSumBasePrice = WEIGHT(optPtr->litSum, optLevel); + optPtr->litLengthSumBasePrice = WEIGHT(optPtr->litLengthSum, optLevel); + optPtr->matchLengthSumBasePrice = WEIGHT(optPtr->matchLengthSum, optLevel); + optPtr->offCodeSumBasePrice = WEIGHT(optPtr->offCodeSum, optLevel); +} + + +/* ZSTD_downscaleStat() : + * reduce all elements in table by a factor 2^(ZSTD_FREQ_DIV+malus) + * return the resulting sum of elements */ +static U32 ZSTD_downscaleStat(unsigned* table, U32 lastEltIndex, int malus) +{ + U32 s, sum=0; + DEBUGLOG(5, "ZSTD_downscaleStat (nbElts=%u)", (unsigned)lastEltIndex+1); + assert(ZSTD_FREQ_DIV+malus > 0 && ZSTD_FREQ_DIV+malus < 31); + for (s=0; s> (ZSTD_FREQ_DIV+malus)); + sum += table[s]; + } + return sum; +} + +/* ZSTD_rescaleFreqs() : + * if first block (detected by optPtr->litLengthSum == 0) : init statistics + * take hints from dictionary if there is one + * or init from zero, using src for literals stats, or flat 1 for match symbols + * otherwise downscale existing stats, to be used as seed for next block. + */ +static void +ZSTD_rescaleFreqs(optState_t* const optPtr, + const BYTE* const src, size_t const srcSize, + int const optLevel) +{ + int const compressedLiterals = ZSTD_compressedLiterals(optPtr); + DEBUGLOG(5, "ZSTD_rescaleFreqs (srcSize=%u)", (unsigned)srcSize); + optPtr->priceType = zop_dynamic; + + if (optPtr->litLengthSum == 0) { /* first block : init */ + if (srcSize <= ZSTD_PREDEF_THRESHOLD) { /* heuristic */ + DEBUGLOG(5, "(srcSize <= ZSTD_PREDEF_THRESHOLD) => zop_predef"); + optPtr->priceType = zop_predef; + } + + assert(optPtr->symbolCosts != NULL); + if (optPtr->symbolCosts->huf.repeatMode == HUF_repeat_valid) { + /* huffman table presumed generated by dictionary */ + optPtr->priceType = zop_dynamic; + + if (compressedLiterals) { + unsigned lit; + assert(optPtr->litFreq != NULL); + optPtr->litSum = 0; + for (lit=0; lit<=MaxLit; lit++) { + U32 const scaleLog = 11; /* scale to 2K */ + U32 const bitCost = HUF_getNbBits(optPtr->symbolCosts->huf.CTable, lit); + assert(bitCost <= scaleLog); + optPtr->litFreq[lit] = bitCost ? 1 << (scaleLog-bitCost) : 1 /*minimum to calculate cost*/; + optPtr->litSum += optPtr->litFreq[lit]; + } } + + { unsigned ll; + FSE_CState_t llstate; + FSE_initCState(&llstate, optPtr->symbolCosts->fse.litlengthCTable); + optPtr->litLengthSum = 0; + for (ll=0; ll<=MaxLL; ll++) { + U32 const scaleLog = 10; /* scale to 1K */ + U32 const bitCost = FSE_getMaxNbBits(llstate.symbolTT, ll); + assert(bitCost < scaleLog); + optPtr->litLengthFreq[ll] = bitCost ? 1 << (scaleLog-bitCost) : 1 /*minimum to calculate cost*/; + optPtr->litLengthSum += optPtr->litLengthFreq[ll]; + } } + + { unsigned ml; + FSE_CState_t mlstate; + FSE_initCState(&mlstate, optPtr->symbolCosts->fse.matchlengthCTable); + optPtr->matchLengthSum = 0; + for (ml=0; ml<=MaxML; ml++) { + U32 const scaleLog = 10; + U32 const bitCost = FSE_getMaxNbBits(mlstate.symbolTT, ml); + assert(bitCost < scaleLog); + optPtr->matchLengthFreq[ml] = bitCost ? 1 << (scaleLog-bitCost) : 1 /*minimum to calculate cost*/; + optPtr->matchLengthSum += optPtr->matchLengthFreq[ml]; + } } + + { unsigned of; + FSE_CState_t ofstate; + FSE_initCState(&ofstate, optPtr->symbolCosts->fse.offcodeCTable); + optPtr->offCodeSum = 0; + for (of=0; of<=MaxOff; of++) { + U32 const scaleLog = 10; + U32 const bitCost = FSE_getMaxNbBits(ofstate.symbolTT, of); + assert(bitCost < scaleLog); + optPtr->offCodeFreq[of] = bitCost ? 1 << (scaleLog-bitCost) : 1 /*minimum to calculate cost*/; + optPtr->offCodeSum += optPtr->offCodeFreq[of]; + } } + + } else { /* not a dictionary */ + + assert(optPtr->litFreq != NULL); + if (compressedLiterals) { + unsigned lit = MaxLit; + HIST_count_simple(optPtr->litFreq, &lit, src, srcSize); /* use raw first block to init statistics */ + optPtr->litSum = ZSTD_downscaleStat(optPtr->litFreq, MaxLit, 1); + } + + { unsigned ll; + for (ll=0; ll<=MaxLL; ll++) + optPtr->litLengthFreq[ll] = 1; + } + optPtr->litLengthSum = MaxLL+1; + + { unsigned ml; + for (ml=0; ml<=MaxML; ml++) + optPtr->matchLengthFreq[ml] = 1; + } + optPtr->matchLengthSum = MaxML+1; + + { unsigned of; + for (of=0; of<=MaxOff; of++) + optPtr->offCodeFreq[of] = 1; + } + optPtr->offCodeSum = MaxOff+1; + + } + + } else { /* new block : re-use previous statistics, scaled down */ + + if (compressedLiterals) + optPtr->litSum = ZSTD_downscaleStat(optPtr->litFreq, MaxLit, 1); + optPtr->litLengthSum = ZSTD_downscaleStat(optPtr->litLengthFreq, MaxLL, 0); + optPtr->matchLengthSum = ZSTD_downscaleStat(optPtr->matchLengthFreq, MaxML, 0); + optPtr->offCodeSum = ZSTD_downscaleStat(optPtr->offCodeFreq, MaxOff, 0); + } + + ZSTD_setBasePrices(optPtr, optLevel); +} + +/* ZSTD_rawLiteralsCost() : + * price of literals (only) in specified segment (which length can be 0). + * does not include price of literalLength symbol */ +static U32 ZSTD_rawLiteralsCost(const BYTE* const literals, U32 const litLength, + const optState_t* const optPtr, + int optLevel) +{ + if (litLength == 0) return 0; + + if (!ZSTD_compressedLiterals(optPtr)) + return (litLength << 3) * BITCOST_MULTIPLIER; /* Uncompressed - 8 bytes per literal. */ + + if (optPtr->priceType == zop_predef) + return (litLength*6) * BITCOST_MULTIPLIER; /* 6 bit per literal - no statistic used */ + + /* dynamic statistics */ + { U32 price = litLength * optPtr->litSumBasePrice; + U32 u; + for (u=0; u < litLength; u++) { + assert(WEIGHT(optPtr->litFreq[literals[u]], optLevel) <= optPtr->litSumBasePrice); /* literal cost should never be negative */ + price -= WEIGHT(optPtr->litFreq[literals[u]], optLevel); + } + return price; + } +} + +/* ZSTD_litLengthPrice() : + * cost of literalLength symbol */ +static U32 ZSTD_litLengthPrice(U32 const litLength, const optState_t* const optPtr, int optLevel) +{ + if (optPtr->priceType == zop_predef) return WEIGHT(litLength, optLevel); + + /* dynamic statistics */ + { U32 const llCode = ZSTD_LLcode(litLength); + return (LL_bits[llCode] * BITCOST_MULTIPLIER) + + optPtr->litLengthSumBasePrice + - WEIGHT(optPtr->litLengthFreq[llCode], optLevel); + } +} + +/* ZSTD_getMatchPrice() : + * Provides the cost of the match part (offset + matchLength) of a sequence + * Must be combined with ZSTD_fullLiteralsCost() to get the full cost of a sequence. + * optLevel: when <2, favors small offset for decompression speed (improved cache efficiency) */ +FORCE_INLINE_TEMPLATE U32 +ZSTD_getMatchPrice(U32 const offset, + U32 const matchLength, + const optState_t* const optPtr, + int const optLevel) +{ + U32 price; + U32 const offCode = ZSTD_highbit32(offset+1); + U32 const mlBase = matchLength - MINMATCH; + assert(matchLength >= MINMATCH); + + if (optPtr->priceType == zop_predef) /* fixed scheme, do not use statistics */ + return WEIGHT(mlBase, optLevel) + ((16 + offCode) * BITCOST_MULTIPLIER); + + /* dynamic statistics */ + price = (offCode * BITCOST_MULTIPLIER) + (optPtr->offCodeSumBasePrice - WEIGHT(optPtr->offCodeFreq[offCode], optLevel)); + if ((optLevel<2) /*static*/ && offCode >= 20) + price += (offCode-19)*2 * BITCOST_MULTIPLIER; /* handicap for long distance offsets, favor decompression speed */ + + /* match Length */ + { U32 const mlCode = ZSTD_MLcode(mlBase); + price += (ML_bits[mlCode] * BITCOST_MULTIPLIER) + (optPtr->matchLengthSumBasePrice - WEIGHT(optPtr->matchLengthFreq[mlCode], optLevel)); + } + + price += BITCOST_MULTIPLIER / 5; /* heuristic : make matches a bit more costly to favor less sequences -> faster decompression speed */ + + DEBUGLOG(8, "ZSTD_getMatchPrice(ml:%u) = %u", matchLength, price); + return price; +} + +/* ZSTD_updateStats() : + * assumption : literals + litLengtn <= iend */ +static void ZSTD_updateStats(optState_t* const optPtr, + U32 litLength, const BYTE* literals, + U32 offsetCode, U32 matchLength) +{ + /* literals */ + if (ZSTD_compressedLiterals(optPtr)) { + U32 u; + for (u=0; u < litLength; u++) + optPtr->litFreq[literals[u]] += ZSTD_LITFREQ_ADD; + optPtr->litSum += litLength*ZSTD_LITFREQ_ADD; + } + + /* literal Length */ + { U32 const llCode = ZSTD_LLcode(litLength); + optPtr->litLengthFreq[llCode]++; + optPtr->litLengthSum++; + } + + /* match offset code (0-2=>repCode; 3+=>offset+2) */ + { U32 const offCode = ZSTD_highbit32(offsetCode+1); + assert(offCode <= MaxOff); + optPtr->offCodeFreq[offCode]++; + optPtr->offCodeSum++; + } + + /* match Length */ + { U32 const mlBase = matchLength - MINMATCH; + U32 const mlCode = ZSTD_MLcode(mlBase); + optPtr->matchLengthFreq[mlCode]++; + optPtr->matchLengthSum++; + } +} + + +/* ZSTD_readMINMATCH() : + * function safe only for comparisons + * assumption : memPtr must be at least 4 bytes before end of buffer */ +MEM_STATIC U32 ZSTD_readMINMATCH(const void* memPtr, U32 length) +{ + switch (length) + { + default : + case 4 : return MEM_read32(memPtr); + case 3 : if (MEM_isLittleEndian()) + return MEM_read32(memPtr)<<8; + else + return MEM_read32(memPtr)>>8; + } +} + + +/* Update hashTable3 up to ip (excluded) + Assumption : always within prefix (i.e. not within extDict) */ +static U32 ZSTD_insertAndFindFirstIndexHash3 (ZSTD_matchState_t* ms, + U32* nextToUpdate3, + const BYTE* const ip) +{ + U32* const hashTable3 = ms->hashTable3; + U32 const hashLog3 = ms->hashLog3; + const BYTE* const base = ms->window.base; + U32 idx = *nextToUpdate3; + U32 const target = (U32)(ip - base); + size_t const hash3 = ZSTD_hash3Ptr(ip, hashLog3); + assert(hashLog3 > 0); + + while(idx < target) { + hashTable3[ZSTD_hash3Ptr(base+idx, hashLog3)] = idx; + idx++; + } + + *nextToUpdate3 = target; + return hashTable3[hash3]; +} + + +/*-************************************* +* Binary Tree search +***************************************/ +/* ZSTD_insertBt1() : add one or multiple positions to tree. + * ip : assumed <= iend-8 . + * @return : nb of positions added */ +static U32 ZSTD_insertBt1( + ZSTD_matchState_t* ms, + const BYTE* const ip, const BYTE* const iend, + U32 const mls, const int extDict) +{ + const ZSTD_compressionParameters* const cParams = &ms->cParams; + U32* const hashTable = ms->hashTable; + U32 const hashLog = cParams->hashLog; + size_t const h = ZSTD_hashPtr(ip, hashLog, mls); + U32* const bt = ms->chainTable; + U32 const btLog = cParams->chainLog - 1; + U32 const btMask = (1 << btLog) - 1; + U32 matchIndex = hashTable[h]; + size_t commonLengthSmaller=0, commonLengthLarger=0; + const BYTE* const base = ms->window.base; + const BYTE* const dictBase = ms->window.dictBase; + const U32 dictLimit = ms->window.dictLimit; + const BYTE* const dictEnd = dictBase + dictLimit; + const BYTE* const prefixStart = base + dictLimit; + const BYTE* match; + const U32 curr = (U32)(ip-base); + const U32 btLow = btMask >= curr ? 0 : curr - btMask; + U32* smallerPtr = bt + 2*(curr&btMask); + U32* largerPtr = smallerPtr + 1; + U32 dummy32; /* to be nullified at the end */ + U32 const windowLow = ms->window.lowLimit; + U32 matchEndIdx = curr+8+1; + size_t bestLength = 8; + U32 nbCompares = 1U << cParams->searchLog; +#ifdef ZSTD_C_PREDICT + U32 predictedSmall = *(bt + 2*((curr-1)&btMask) + 0); + U32 predictedLarge = *(bt + 2*((curr-1)&btMask) + 1); + predictedSmall += (predictedSmall>0); + predictedLarge += (predictedLarge>0); +#endif /* ZSTD_C_PREDICT */ + + DEBUGLOG(8, "ZSTD_insertBt1 (%u)", curr); + + assert(ip <= iend-8); /* required for h calculation */ + hashTable[h] = curr; /* Update Hash Table */ + + assert(windowLow > 0); + for (; nbCompares && (matchIndex >= windowLow); --nbCompares) { + U32* const nextPtr = bt + 2*(matchIndex & btMask); + size_t matchLength = MIN(commonLengthSmaller, commonLengthLarger); /* guaranteed minimum nb of common bytes */ + assert(matchIndex < curr); + +#ifdef ZSTD_C_PREDICT /* note : can create issues when hlog small <= 11 */ + const U32* predictPtr = bt + 2*((matchIndex-1) & btMask); /* written this way, as bt is a roll buffer */ + if (matchIndex == predictedSmall) { + /* no need to check length, result known */ + *smallerPtr = matchIndex; + if (matchIndex <= btLow) { smallerPtr=&dummy32; break; } /* beyond tree size, stop the search */ + smallerPtr = nextPtr+1; /* new "smaller" => larger of match */ + matchIndex = nextPtr[1]; /* new matchIndex larger than previous (closer to current) */ + predictedSmall = predictPtr[1] + (predictPtr[1]>0); + continue; + } + if (matchIndex == predictedLarge) { + *largerPtr = matchIndex; + if (matchIndex <= btLow) { largerPtr=&dummy32; break; } /* beyond tree size, stop the search */ + largerPtr = nextPtr; + matchIndex = nextPtr[0]; + predictedLarge = predictPtr[0] + (predictPtr[0]>0); + continue; + } +#endif + + if (!extDict || (matchIndex+matchLength >= dictLimit)) { + assert(matchIndex+matchLength >= dictLimit); /* might be wrong if actually extDict */ + match = base + matchIndex; + matchLength += ZSTD_count(ip+matchLength, match+matchLength, iend); + } else { + match = dictBase + matchIndex; + matchLength += ZSTD_count_2segments(ip+matchLength, match+matchLength, iend, dictEnd, prefixStart); + if (matchIndex+matchLength >= dictLimit) + match = base + matchIndex; /* to prepare for next usage of match[matchLength] */ + } + + if (matchLength > bestLength) { + bestLength = matchLength; + if (matchLength > matchEndIdx - matchIndex) + matchEndIdx = matchIndex + (U32)matchLength; + } + + if (ip+matchLength == iend) { /* equal : no way to know if inf or sup */ + break; /* drop , to guarantee consistency ; miss a bit of compression, but other solutions can corrupt tree */ + } + + if (match[matchLength] < ip[matchLength]) { /* necessarily within buffer */ + /* match is smaller than current */ + *smallerPtr = matchIndex; /* update smaller idx */ + commonLengthSmaller = matchLength; /* all smaller will now have at least this guaranteed common length */ + if (matchIndex <= btLow) { smallerPtr=&dummy32; break; } /* beyond tree size, stop searching */ + smallerPtr = nextPtr+1; /* new "candidate" => larger than match, which was smaller than target */ + matchIndex = nextPtr[1]; /* new matchIndex, larger than previous and closer to current */ + } else { + /* match is larger than current */ + *largerPtr = matchIndex; + commonLengthLarger = matchLength; + if (matchIndex <= btLow) { largerPtr=&dummy32; break; } /* beyond tree size, stop searching */ + largerPtr = nextPtr; + matchIndex = nextPtr[0]; + } } + + *smallerPtr = *largerPtr = 0; + { U32 positions = 0; + if (bestLength > 384) positions = MIN(192, (U32)(bestLength - 384)); /* speed optimization */ + assert(matchEndIdx > curr + 8); + return MAX(positions, matchEndIdx - (curr + 8)); + } +} + +FORCE_INLINE_TEMPLATE +void ZSTD_updateTree_internal( + ZSTD_matchState_t* ms, + const BYTE* const ip, const BYTE* const iend, + const U32 mls, const ZSTD_dictMode_e dictMode) +{ + const BYTE* const base = ms->window.base; + U32 const target = (U32)(ip - base); + U32 idx = ms->nextToUpdate; + DEBUGLOG(6, "ZSTD_updateTree_internal, from %u to %u (dictMode:%u)", + idx, target, dictMode); + + while(idx < target) { + U32 const forward = ZSTD_insertBt1(ms, base+idx, iend, mls, dictMode == ZSTD_extDict); + assert(idx < (U32)(idx + forward)); + idx += forward; + } + assert((size_t)(ip - base) <= (size_t)(U32)(-1)); + assert((size_t)(iend - base) <= (size_t)(U32)(-1)); + ms->nextToUpdate = target; +} + +void ZSTD_updateTree(ZSTD_matchState_t* ms, const BYTE* ip, const BYTE* iend) { + ZSTD_updateTree_internal(ms, ip, iend, ms->cParams.minMatch, ZSTD_noDict); +} + +FORCE_INLINE_TEMPLATE +U32 ZSTD_insertBtAndGetAllMatches ( + ZSTD_match_t* matches, /* store result (found matches) in this table (presumed large enough) */ + ZSTD_matchState_t* ms, + U32* nextToUpdate3, + const BYTE* const ip, const BYTE* const iLimit, const ZSTD_dictMode_e dictMode, + const U32 rep[ZSTD_REP_NUM], + U32 const ll0, /* tells if associated literal length is 0 or not. This value must be 0 or 1 */ + const U32 lengthToBeat, + U32 const mls /* template */) +{ + const ZSTD_compressionParameters* const cParams = &ms->cParams; + U32 const sufficient_len = MIN(cParams->targetLength, ZSTD_OPT_NUM -1); + const BYTE* const base = ms->window.base; + U32 const curr = (U32)(ip-base); + U32 const hashLog = cParams->hashLog; + U32 const minMatch = (mls==3) ? 3 : 4; + U32* const hashTable = ms->hashTable; + size_t const h = ZSTD_hashPtr(ip, hashLog, mls); + U32 matchIndex = hashTable[h]; + U32* const bt = ms->chainTable; + U32 const btLog = cParams->chainLog - 1; + U32 const btMask= (1U << btLog) - 1; + size_t commonLengthSmaller=0, commonLengthLarger=0; + const BYTE* const dictBase = ms->window.dictBase; + U32 const dictLimit = ms->window.dictLimit; + const BYTE* const dictEnd = dictBase + dictLimit; + const BYTE* const prefixStart = base + dictLimit; + U32 const btLow = (btMask >= curr) ? 0 : curr - btMask; + U32 const windowLow = ZSTD_getLowestMatchIndex(ms, curr, cParams->windowLog); + U32 const matchLow = windowLow ? windowLow : 1; + U32* smallerPtr = bt + 2*(curr&btMask); + U32* largerPtr = bt + 2*(curr&btMask) + 1; + U32 matchEndIdx = curr+8+1; /* farthest referenced position of any match => detects repetitive patterns */ + U32 dummy32; /* to be nullified at the end */ + U32 mnum = 0; + U32 nbCompares = 1U << cParams->searchLog; + + const ZSTD_matchState_t* dms = dictMode == ZSTD_dictMatchState ? ms->dictMatchState : NULL; + const ZSTD_compressionParameters* const dmsCParams = + dictMode == ZSTD_dictMatchState ? &dms->cParams : NULL; + const BYTE* const dmsBase = dictMode == ZSTD_dictMatchState ? dms->window.base : NULL; + const BYTE* const dmsEnd = dictMode == ZSTD_dictMatchState ? dms->window.nextSrc : NULL; + U32 const dmsHighLimit = dictMode == ZSTD_dictMatchState ? (U32)(dmsEnd - dmsBase) : 0; + U32 const dmsLowLimit = dictMode == ZSTD_dictMatchState ? dms->window.lowLimit : 0; + U32 const dmsIndexDelta = dictMode == ZSTD_dictMatchState ? windowLow - dmsHighLimit : 0; + U32 const dmsHashLog = dictMode == ZSTD_dictMatchState ? dmsCParams->hashLog : hashLog; + U32 const dmsBtLog = dictMode == ZSTD_dictMatchState ? dmsCParams->chainLog - 1 : btLog; + U32 const dmsBtMask = dictMode == ZSTD_dictMatchState ? (1U << dmsBtLog) - 1 : 0; + U32 const dmsBtLow = dictMode == ZSTD_dictMatchState && dmsBtMask < dmsHighLimit - dmsLowLimit ? dmsHighLimit - dmsBtMask : dmsLowLimit; + + size_t bestLength = lengthToBeat-1; + DEBUGLOG(8, "ZSTD_insertBtAndGetAllMatches: current=%u", curr); + + /* check repCode */ + assert(ll0 <= 1); /* necessarily 1 or 0 */ + { U32 const lastR = ZSTD_REP_NUM + ll0; + U32 repCode; + for (repCode = ll0; repCode < lastR; repCode++) { + U32 const repOffset = (repCode==ZSTD_REP_NUM) ? (rep[0] - 1) : rep[repCode]; + U32 const repIndex = curr - repOffset; + U32 repLen = 0; + assert(curr >= dictLimit); + if (repOffset-1 /* intentional overflow, discards 0 and -1 */ < curr-dictLimit) { /* equivalent to `curr > repIndex >= dictLimit` */ + /* We must validate the repcode offset because when we're using a dictionary the + * valid offset range shrinks when the dictionary goes out of bounds. + */ + if ((repIndex >= windowLow) & (ZSTD_readMINMATCH(ip, minMatch) == ZSTD_readMINMATCH(ip - repOffset, minMatch))) { + repLen = (U32)ZSTD_count(ip+minMatch, ip+minMatch-repOffset, iLimit) + minMatch; + } + } else { /* repIndex < dictLimit || repIndex >= curr */ + const BYTE* const repMatch = dictMode == ZSTD_dictMatchState ? + dmsBase + repIndex - dmsIndexDelta : + dictBase + repIndex; + assert(curr >= windowLow); + if ( dictMode == ZSTD_extDict + && ( ((repOffset-1) /*intentional overflow*/ < curr - windowLow) /* equivalent to `curr > repIndex >= windowLow` */ + & (((U32)((dictLimit-1) - repIndex) >= 3) ) /* intentional overflow : do not test positions overlapping 2 memory segments */) + && (ZSTD_readMINMATCH(ip, minMatch) == ZSTD_readMINMATCH(repMatch, minMatch)) ) { + repLen = (U32)ZSTD_count_2segments(ip+minMatch, repMatch+minMatch, iLimit, dictEnd, prefixStart) + minMatch; + } + if (dictMode == ZSTD_dictMatchState + && ( ((repOffset-1) /*intentional overflow*/ < curr - (dmsLowLimit + dmsIndexDelta)) /* equivalent to `curr > repIndex >= dmsLowLimit` */ + & ((U32)((dictLimit-1) - repIndex) >= 3) ) /* intentional overflow : do not test positions overlapping 2 memory segments */ + && (ZSTD_readMINMATCH(ip, minMatch) == ZSTD_readMINMATCH(repMatch, minMatch)) ) { + repLen = (U32)ZSTD_count_2segments(ip+minMatch, repMatch+minMatch, iLimit, dmsEnd, prefixStart) + minMatch; + } } + /* save longer solution */ + if (repLen > bestLength) { + DEBUGLOG(8, "found repCode %u (ll0:%u, offset:%u) of length %u", + repCode, ll0, repOffset, repLen); + bestLength = repLen; + matches[mnum].off = repCode - ll0; + matches[mnum].len = (U32)repLen; + mnum++; + if ( (repLen > sufficient_len) + | (ip+repLen == iLimit) ) { /* best possible */ + return mnum; + } } } } + + /* HC3 match finder */ + if ((mls == 3) /*static*/ && (bestLength < mls)) { + U32 const matchIndex3 = ZSTD_insertAndFindFirstIndexHash3(ms, nextToUpdate3, ip); + if ((matchIndex3 >= matchLow) + & (curr - matchIndex3 < (1<<18)) /*heuristic : longer distance likely too expensive*/ ) { + size_t mlen; + if ((dictMode == ZSTD_noDict) /*static*/ || (dictMode == ZSTD_dictMatchState) /*static*/ || (matchIndex3 >= dictLimit)) { + const BYTE* const match = base + matchIndex3; + mlen = ZSTD_count(ip, match, iLimit); + } else { + const BYTE* const match = dictBase + matchIndex3; + mlen = ZSTD_count_2segments(ip, match, iLimit, dictEnd, prefixStart); + } + + /* save best solution */ + if (mlen >= mls /* == 3 > bestLength */) { + DEBUGLOG(8, "found small match with hlog3, of length %u", + (U32)mlen); + bestLength = mlen; + assert(curr > matchIndex3); + assert(mnum==0); /* no prior solution */ + matches[0].off = (curr - matchIndex3) + ZSTD_REP_MOVE; + matches[0].len = (U32)mlen; + mnum = 1; + if ( (mlen > sufficient_len) | + (ip+mlen == iLimit) ) { /* best possible length */ + ms->nextToUpdate = curr+1; /* skip insertion */ + return 1; + } } } + /* no dictMatchState lookup: dicts don't have a populated HC3 table */ + } + + hashTable[h] = curr; /* Update Hash Table */ + + for (; nbCompares && (matchIndex >= matchLow); --nbCompares) { + U32* const nextPtr = bt + 2*(matchIndex & btMask); + const BYTE* match; + size_t matchLength = MIN(commonLengthSmaller, commonLengthLarger); /* guaranteed minimum nb of common bytes */ + assert(curr > matchIndex); + + if ((dictMode == ZSTD_noDict) || (dictMode == ZSTD_dictMatchState) || (matchIndex+matchLength >= dictLimit)) { + assert(matchIndex+matchLength >= dictLimit); /* ensure the condition is correct when !extDict */ + match = base + matchIndex; + if (matchIndex >= dictLimit) assert(memcmp(match, ip, matchLength) == 0); /* ensure early section of match is equal as expected */ + matchLength += ZSTD_count(ip+matchLength, match+matchLength, iLimit); + } else { + match = dictBase + matchIndex; + assert(memcmp(match, ip, matchLength) == 0); /* ensure early section of match is equal as expected */ + matchLength += ZSTD_count_2segments(ip+matchLength, match+matchLength, iLimit, dictEnd, prefixStart); + if (matchIndex+matchLength >= dictLimit) + match = base + matchIndex; /* prepare for match[matchLength] read */ + } + + if (matchLength > bestLength) { + DEBUGLOG(8, "found match of length %u at distance %u (offCode=%u)", + (U32)matchLength, curr - matchIndex, curr - matchIndex + ZSTD_REP_MOVE); + assert(matchEndIdx > matchIndex); + if (matchLength > matchEndIdx - matchIndex) + matchEndIdx = matchIndex + (U32)matchLength; + bestLength = matchLength; + matches[mnum].off = (curr - matchIndex) + ZSTD_REP_MOVE; + matches[mnum].len = (U32)matchLength; + mnum++; + if ( (matchLength > ZSTD_OPT_NUM) + | (ip+matchLength == iLimit) /* equal : no way to know if inf or sup */) { + if (dictMode == ZSTD_dictMatchState) nbCompares = 0; /* break should also skip searching dms */ + break; /* drop, to preserve bt consistency (miss a little bit of compression) */ + } + } + + if (match[matchLength] < ip[matchLength]) { + /* match smaller than current */ + *smallerPtr = matchIndex; /* update smaller idx */ + commonLengthSmaller = matchLength; /* all smaller will now have at least this guaranteed common length */ + if (matchIndex <= btLow) { smallerPtr=&dummy32; break; } /* beyond tree size, stop the search */ + smallerPtr = nextPtr+1; /* new candidate => larger than match, which was smaller than current */ + matchIndex = nextPtr[1]; /* new matchIndex, larger than previous, closer to current */ + } else { + *largerPtr = matchIndex; + commonLengthLarger = matchLength; + if (matchIndex <= btLow) { largerPtr=&dummy32; break; } /* beyond tree size, stop the search */ + largerPtr = nextPtr; + matchIndex = nextPtr[0]; + } } + + *smallerPtr = *largerPtr = 0; + + assert(nbCompares <= (1U << ZSTD_SEARCHLOG_MAX)); /* Check we haven't underflowed. */ + if (dictMode == ZSTD_dictMatchState && nbCompares) { + size_t const dmsH = ZSTD_hashPtr(ip, dmsHashLog, mls); + U32 dictMatchIndex = dms->hashTable[dmsH]; + const U32* const dmsBt = dms->chainTable; + commonLengthSmaller = commonLengthLarger = 0; + for (; nbCompares && (dictMatchIndex > dmsLowLimit); --nbCompares) { + const U32* const nextPtr = dmsBt + 2*(dictMatchIndex & dmsBtMask); + size_t matchLength = MIN(commonLengthSmaller, commonLengthLarger); /* guaranteed minimum nb of common bytes */ + const BYTE* match = dmsBase + dictMatchIndex; + matchLength += ZSTD_count_2segments(ip+matchLength, match+matchLength, iLimit, dmsEnd, prefixStart); + if (dictMatchIndex+matchLength >= dmsHighLimit) + match = base + dictMatchIndex + dmsIndexDelta; /* to prepare for next usage of match[matchLength] */ + + if (matchLength > bestLength) { + matchIndex = dictMatchIndex + dmsIndexDelta; + DEBUGLOG(8, "found dms match of length %u at distance %u (offCode=%u)", + (U32)matchLength, curr - matchIndex, curr - matchIndex + ZSTD_REP_MOVE); + if (matchLength > matchEndIdx - matchIndex) + matchEndIdx = matchIndex + (U32)matchLength; + bestLength = matchLength; + matches[mnum].off = (curr - matchIndex) + ZSTD_REP_MOVE; + matches[mnum].len = (U32)matchLength; + mnum++; + if ( (matchLength > ZSTD_OPT_NUM) + | (ip+matchLength == iLimit) /* equal : no way to know if inf or sup */) { + break; /* drop, to guarantee consistency (miss a little bit of compression) */ + } + } + + if (dictMatchIndex <= dmsBtLow) { break; } /* beyond tree size, stop the search */ + if (match[matchLength] < ip[matchLength]) { + commonLengthSmaller = matchLength; /* all smaller will now have at least this guaranteed common length */ + dictMatchIndex = nextPtr[1]; /* new matchIndex larger than previous (closer to current) */ + } else { + /* match is larger than current */ + commonLengthLarger = matchLength; + dictMatchIndex = nextPtr[0]; + } + } + } + + assert(matchEndIdx > curr+8); + ms->nextToUpdate = matchEndIdx - 8; /* skip repetitive patterns */ + return mnum; +} + + +FORCE_INLINE_TEMPLATE U32 ZSTD_BtGetAllMatches ( + ZSTD_match_t* matches, /* store result (match found, increasing size) in this table */ + ZSTD_matchState_t* ms, + U32* nextToUpdate3, + const BYTE* ip, const BYTE* const iHighLimit, const ZSTD_dictMode_e dictMode, + const U32 rep[ZSTD_REP_NUM], + U32 const ll0, + U32 const lengthToBeat) +{ + const ZSTD_compressionParameters* const cParams = &ms->cParams; + U32 const matchLengthSearch = cParams->minMatch; + DEBUGLOG(8, "ZSTD_BtGetAllMatches"); + if (ip < ms->window.base + ms->nextToUpdate) return 0; /* skipped area */ + ZSTD_updateTree_internal(ms, ip, iHighLimit, matchLengthSearch, dictMode); + switch(matchLengthSearch) + { + case 3 : return ZSTD_insertBtAndGetAllMatches(matches, ms, nextToUpdate3, ip, iHighLimit, dictMode, rep, ll0, lengthToBeat, 3); + default : + case 4 : return ZSTD_insertBtAndGetAllMatches(matches, ms, nextToUpdate3, ip, iHighLimit, dictMode, rep, ll0, lengthToBeat, 4); + case 5 : return ZSTD_insertBtAndGetAllMatches(matches, ms, nextToUpdate3, ip, iHighLimit, dictMode, rep, ll0, lengthToBeat, 5); + case 7 : + case 6 : return ZSTD_insertBtAndGetAllMatches(matches, ms, nextToUpdate3, ip, iHighLimit, dictMode, rep, ll0, lengthToBeat, 6); + } +} + +/* *********************** +* LDM helper functions * +*************************/ + +/* Struct containing info needed to make decision about ldm inclusion */ +typedef struct { + rawSeqStore_t seqStore; /* External match candidates store for this block */ + U32 startPosInBlock; /* Start position of the current match candidate */ + U32 endPosInBlock; /* End position of the current match candidate */ + U32 offset; /* Offset of the match candidate */ +} ZSTD_optLdm_t; + +/* ZSTD_optLdm_skipRawSeqStoreBytes(): + * Moves forward in rawSeqStore by nbBytes, which will update the fields 'pos' and 'posInSequence'. + */ +static void ZSTD_optLdm_skipRawSeqStoreBytes(rawSeqStore_t* rawSeqStore, size_t nbBytes) { + U32 currPos = (U32)(rawSeqStore->posInSequence + nbBytes); + while (currPos && rawSeqStore->pos < rawSeqStore->size) { + rawSeq currSeq = rawSeqStore->seq[rawSeqStore->pos]; + if (currPos >= currSeq.litLength + currSeq.matchLength) { + currPos -= currSeq.litLength + currSeq.matchLength; + rawSeqStore->pos++; + } else { + rawSeqStore->posInSequence = currPos; + break; + } + } + if (currPos == 0 || rawSeqStore->pos == rawSeqStore->size) { + rawSeqStore->posInSequence = 0; + } +} + +/* ZSTD_opt_getNextMatchAndUpdateSeqStore(): + * Calculates the beginning and end of the next match in the current block. + * Updates 'pos' and 'posInSequence' of the ldmSeqStore. + */ +static void ZSTD_opt_getNextMatchAndUpdateSeqStore(ZSTD_optLdm_t* optLdm, U32 currPosInBlock, + U32 blockBytesRemaining) { + rawSeq currSeq; + U32 currBlockEndPos; + U32 literalsBytesRemaining; + U32 matchBytesRemaining; + + /* Setting match end position to MAX to ensure we never use an LDM during this block */ + if (optLdm->seqStore.size == 0 || optLdm->seqStore.pos >= optLdm->seqStore.size) { + optLdm->startPosInBlock = UINT_MAX; + optLdm->endPosInBlock = UINT_MAX; + return; + } + /* Calculate appropriate bytes left in matchLength and litLength after adjusting + based on ldmSeqStore->posInSequence */ + currSeq = optLdm->seqStore.seq[optLdm->seqStore.pos]; + assert(optLdm->seqStore.posInSequence <= currSeq.litLength + currSeq.matchLength); + currBlockEndPos = currPosInBlock + blockBytesRemaining; + literalsBytesRemaining = (optLdm->seqStore.posInSequence < currSeq.litLength) ? + currSeq.litLength - (U32)optLdm->seqStore.posInSequence : + 0; + matchBytesRemaining = (literalsBytesRemaining == 0) ? + currSeq.matchLength - ((U32)optLdm->seqStore.posInSequence - currSeq.litLength) : + currSeq.matchLength; + + /* If there are more literal bytes than bytes remaining in block, no ldm is possible */ + if (literalsBytesRemaining >= blockBytesRemaining) { + optLdm->startPosInBlock = UINT_MAX; + optLdm->endPosInBlock = UINT_MAX; + ZSTD_optLdm_skipRawSeqStoreBytes(&optLdm->seqStore, blockBytesRemaining); + return; + } + + /* Matches may be < MINMATCH by this process. In that case, we will reject them + when we are deciding whether or not to add the ldm */ + optLdm->startPosInBlock = currPosInBlock + literalsBytesRemaining; + optLdm->endPosInBlock = optLdm->startPosInBlock + matchBytesRemaining; + optLdm->offset = currSeq.offset; + + if (optLdm->endPosInBlock > currBlockEndPos) { + /* Match ends after the block ends, we can't use the whole match */ + optLdm->endPosInBlock = currBlockEndPos; + ZSTD_optLdm_skipRawSeqStoreBytes(&optLdm->seqStore, currBlockEndPos - currPosInBlock); + } else { + /* Consume nb of bytes equal to size of sequence left */ + ZSTD_optLdm_skipRawSeqStoreBytes(&optLdm->seqStore, literalsBytesRemaining + matchBytesRemaining); + } +} + +/* ZSTD_optLdm_maybeAddMatch(): + * Adds a match if it's long enough, based on it's 'matchStartPosInBlock' + * and 'matchEndPosInBlock', into 'matches'. Maintains the correct ordering of 'matches' + */ +static void ZSTD_optLdm_maybeAddMatch(ZSTD_match_t* matches, U32* nbMatches, + ZSTD_optLdm_t* optLdm, U32 currPosInBlock) { + U32 posDiff = currPosInBlock - optLdm->startPosInBlock; + /* Note: ZSTD_match_t actually contains offCode and matchLength (before subtracting MINMATCH) */ + U32 candidateMatchLength = optLdm->endPosInBlock - optLdm->startPosInBlock - posDiff; + U32 candidateOffCode = optLdm->offset + ZSTD_REP_MOVE; + + /* Ensure that current block position is not outside of the match */ + if (currPosInBlock < optLdm->startPosInBlock + || currPosInBlock >= optLdm->endPosInBlock + || candidateMatchLength < MINMATCH) { + return; + } + + if (*nbMatches == 0 || ((candidateMatchLength > matches[*nbMatches-1].len) && *nbMatches < ZSTD_OPT_NUM)) { + DEBUGLOG(6, "ZSTD_optLdm_maybeAddMatch(): Adding ldm candidate match (offCode: %u matchLength %u) at block position=%u", + candidateOffCode, candidateMatchLength, currPosInBlock); + matches[*nbMatches].len = candidateMatchLength; + matches[*nbMatches].off = candidateOffCode; + (*nbMatches)++; + } +} + +/* ZSTD_optLdm_processMatchCandidate(): + * Wrapper function to update ldm seq store and call ldm functions as necessary. + */ +static void ZSTD_optLdm_processMatchCandidate(ZSTD_optLdm_t* optLdm, ZSTD_match_t* matches, U32* nbMatches, + U32 currPosInBlock, U32 remainingBytes) { + if (optLdm->seqStore.size == 0 || optLdm->seqStore.pos >= optLdm->seqStore.size) { + return; + } + + if (currPosInBlock >= optLdm->endPosInBlock) { + if (currPosInBlock > optLdm->endPosInBlock) { + /* The position at which ZSTD_optLdm_processMatchCandidate() is called is not necessarily + * at the end of a match from the ldm seq store, and will often be some bytes + * over beyond matchEndPosInBlock. As such, we need to correct for these "overshoots" + */ + U32 posOvershoot = currPosInBlock - optLdm->endPosInBlock; + ZSTD_optLdm_skipRawSeqStoreBytes(&optLdm->seqStore, posOvershoot); + } + ZSTD_opt_getNextMatchAndUpdateSeqStore(optLdm, currPosInBlock, remainingBytes); + } + ZSTD_optLdm_maybeAddMatch(matches, nbMatches, optLdm, currPosInBlock); +} + +/*-******************************* +* Optimal parser +*********************************/ + + +static U32 ZSTD_totalLen(ZSTD_optimal_t sol) +{ + return sol.litlen + sol.mlen; +} + +#if 0 /* debug */ + +static void +listStats(const U32* table, int lastEltID) +{ + int const nbElts = lastEltID + 1; + int enb; + for (enb=0; enb < nbElts; enb++) { + (void)table; + /* RAWLOG(2, "%3i:%3i, ", enb, table[enb]); */ + RAWLOG(2, "%4i,", table[enb]); + } + RAWLOG(2, " \n"); +} + +#endif + +FORCE_INLINE_TEMPLATE size_t +ZSTD_compressBlock_opt_generic(ZSTD_matchState_t* ms, + seqStore_t* seqStore, + U32 rep[ZSTD_REP_NUM], + const void* src, size_t srcSize, + const int optLevel, + const ZSTD_dictMode_e dictMode) +{ + optState_t* const optStatePtr = &ms->opt; + const BYTE* const istart = (const BYTE*)src; + const BYTE* ip = istart; + const BYTE* anchor = istart; + const BYTE* const iend = istart + srcSize; + const BYTE* const ilimit = iend - 8; + const BYTE* const base = ms->window.base; + const BYTE* const prefixStart = base + ms->window.dictLimit; + const ZSTD_compressionParameters* const cParams = &ms->cParams; + + U32 const sufficient_len = MIN(cParams->targetLength, ZSTD_OPT_NUM -1); + U32 const minMatch = (cParams->minMatch == 3) ? 3 : 4; + U32 nextToUpdate3 = ms->nextToUpdate; + + ZSTD_optimal_t* const opt = optStatePtr->priceTable; + ZSTD_match_t* const matches = optStatePtr->matchTable; + ZSTD_optimal_t lastSequence; + ZSTD_optLdm_t optLdm; + + optLdm.seqStore = ms->ldmSeqStore ? *ms->ldmSeqStore : kNullRawSeqStore; + optLdm.endPosInBlock = optLdm.startPosInBlock = optLdm.offset = 0; + ZSTD_opt_getNextMatchAndUpdateSeqStore(&optLdm, (U32)(ip-istart), (U32)(iend-ip)); + + /* init */ + DEBUGLOG(5, "ZSTD_compressBlock_opt_generic: current=%u, prefix=%u, nextToUpdate=%u", + (U32)(ip - base), ms->window.dictLimit, ms->nextToUpdate); + assert(optLevel <= 2); + ZSTD_rescaleFreqs(optStatePtr, (const BYTE*)src, srcSize, optLevel); + ip += (ip==prefixStart); + + /* Match Loop */ + while (ip < ilimit) { + U32 cur, last_pos = 0; + + /* find first match */ + { U32 const litlen = (U32)(ip - anchor); + U32 const ll0 = !litlen; + U32 nbMatches = ZSTD_BtGetAllMatches(matches, ms, &nextToUpdate3, ip, iend, dictMode, rep, ll0, minMatch); + ZSTD_optLdm_processMatchCandidate(&optLdm, matches, &nbMatches, + (U32)(ip-istart), (U32)(iend - ip)); + if (!nbMatches) { ip++; continue; } + + /* initialize opt[0] */ + { U32 i ; for (i=0; i immediate encoding */ + { U32 const maxML = matches[nbMatches-1].len; + U32 const maxOffset = matches[nbMatches-1].off; + DEBUGLOG(6, "found %u matches of maxLength=%u and maxOffCode=%u at cPos=%u => start new series", + nbMatches, maxML, maxOffset, (U32)(ip-prefixStart)); + + if (maxML > sufficient_len) { + lastSequence.litlen = litlen; + lastSequence.mlen = maxML; + lastSequence.off = maxOffset; + DEBUGLOG(6, "large match (%u>%u), immediate encoding", + maxML, sufficient_len); + cur = 0; + last_pos = ZSTD_totalLen(lastSequence); + goto _shortestPath; + } } + + /* set prices for first matches starting position == 0 */ + { U32 const literalsPrice = opt[0].price + ZSTD_litLengthPrice(0, optStatePtr, optLevel); + U32 pos; + U32 matchNb; + for (pos = 1; pos < minMatch; pos++) { + opt[pos].price = ZSTD_MAX_PRICE; /* mlen, litlen and price will be fixed during forward scanning */ + } + for (matchNb = 0; matchNb < nbMatches; matchNb++) { + U32 const offset = matches[matchNb].off; + U32 const end = matches[matchNb].len; + for ( ; pos <= end ; pos++ ) { + U32 const matchPrice = ZSTD_getMatchPrice(offset, pos, optStatePtr, optLevel); + U32 const sequencePrice = literalsPrice + matchPrice; + DEBUGLOG(7, "rPos:%u => set initial price : %.2f", + pos, ZSTD_fCost(sequencePrice)); + opt[pos].mlen = pos; + opt[pos].off = offset; + opt[pos].litlen = litlen; + opt[pos].price = sequencePrice; + } } + last_pos = pos-1; + } + } + + /* check further positions */ + for (cur = 1; cur <= last_pos; cur++) { + const BYTE* const inr = ip + cur; + assert(cur < ZSTD_OPT_NUM); + DEBUGLOG(7, "cPos:%zi==rPos:%u", inr-istart, cur) + + /* Fix current position with one literal if cheaper */ + { U32 const litlen = (opt[cur-1].mlen == 0) ? opt[cur-1].litlen + 1 : 1; + int const price = opt[cur-1].price + + ZSTD_rawLiteralsCost(ip+cur-1, 1, optStatePtr, optLevel) + + ZSTD_litLengthPrice(litlen, optStatePtr, optLevel) + - ZSTD_litLengthPrice(litlen-1, optStatePtr, optLevel); + assert(price < 1000000000); /* overflow check */ + if (price <= opt[cur].price) { + DEBUGLOG(7, "cPos:%zi==rPos:%u : better price (%.2f<=%.2f) using literal (ll==%u) (hist:%u,%u,%u)", + inr-istart, cur, ZSTD_fCost(price), ZSTD_fCost(opt[cur].price), litlen, + opt[cur-1].rep[0], opt[cur-1].rep[1], opt[cur-1].rep[2]); + opt[cur].mlen = 0; + opt[cur].off = 0; + opt[cur].litlen = litlen; + opt[cur].price = price; + } else { + DEBUGLOG(7, "cPos:%zi==rPos:%u : literal would cost more (%.2f>%.2f) (hist:%u,%u,%u)", + inr-istart, cur, ZSTD_fCost(price), ZSTD_fCost(opt[cur].price), + opt[cur].rep[0], opt[cur].rep[1], opt[cur].rep[2]); + } + } + + /* Set the repcodes of the current position. We must do it here + * because we rely on the repcodes of the 2nd to last sequence being + * correct to set the next chunks repcodes during the backward + * traversal. + */ + ZSTD_STATIC_ASSERT(sizeof(opt[cur].rep) == sizeof(repcodes_t)); + assert(cur >= opt[cur].mlen); + if (opt[cur].mlen != 0) { + U32 const prev = cur - opt[cur].mlen; + repcodes_t newReps = ZSTD_updateRep(opt[prev].rep, opt[cur].off, opt[cur].litlen==0); + ZSTD_memcpy(opt[cur].rep, &newReps, sizeof(repcodes_t)); + } else { + ZSTD_memcpy(opt[cur].rep, opt[cur - 1].rep, sizeof(repcodes_t)); + } + + /* last match must start at a minimum distance of 8 from oend */ + if (inr > ilimit) continue; + + if (cur == last_pos) break; + + if ( (optLevel==0) /*static_test*/ + && (opt[cur+1].price <= opt[cur].price + (BITCOST_MULTIPLIER/2)) ) { + DEBUGLOG(7, "move to next rPos:%u : price is <=", cur+1); + continue; /* skip unpromising positions; about ~+6% speed, -0.01 ratio */ + } + + { U32 const ll0 = (opt[cur].mlen != 0); + U32 const litlen = (opt[cur].mlen == 0) ? opt[cur].litlen : 0; + U32 const previousPrice = opt[cur].price; + U32 const basePrice = previousPrice + ZSTD_litLengthPrice(0, optStatePtr, optLevel); + U32 nbMatches = ZSTD_BtGetAllMatches(matches, ms, &nextToUpdate3, inr, iend, dictMode, opt[cur].rep, ll0, minMatch); + U32 matchNb; + + ZSTD_optLdm_processMatchCandidate(&optLdm, matches, &nbMatches, + (U32)(inr-istart), (U32)(iend-inr)); + + if (!nbMatches) { + DEBUGLOG(7, "rPos:%u : no match found", cur); + continue; + } + + { U32 const maxML = matches[nbMatches-1].len; + DEBUGLOG(7, "cPos:%zi==rPos:%u, found %u matches, of maxLength=%u", + inr-istart, cur, nbMatches, maxML); + + if ( (maxML > sufficient_len) + || (cur + maxML >= ZSTD_OPT_NUM) ) { + lastSequence.mlen = maxML; + lastSequence.off = matches[nbMatches-1].off; + lastSequence.litlen = litlen; + cur -= (opt[cur].mlen==0) ? opt[cur].litlen : 0; /* last sequence is actually only literals, fix cur to last match - note : may underflow, in which case, it's first sequence, and it's okay */ + last_pos = cur + ZSTD_totalLen(lastSequence); + if (cur > ZSTD_OPT_NUM) cur = 0; /* underflow => first match */ + goto _shortestPath; + } } + + /* set prices using matches found at position == cur */ + for (matchNb = 0; matchNb < nbMatches; matchNb++) { + U32 const offset = matches[matchNb].off; + U32 const lastML = matches[matchNb].len; + U32 const startML = (matchNb>0) ? matches[matchNb-1].len+1 : minMatch; + U32 mlen; + + DEBUGLOG(7, "testing match %u => offCode=%4u, mlen=%2u, llen=%2u", + matchNb, matches[matchNb].off, lastML, litlen); + + for (mlen = lastML; mlen >= startML; mlen--) { /* scan downward */ + U32 const pos = cur + mlen; + int const price = basePrice + ZSTD_getMatchPrice(offset, mlen, optStatePtr, optLevel); + + if ((pos > last_pos) || (price < opt[pos].price)) { + DEBUGLOG(7, "rPos:%u (ml=%2u) => new better price (%.2f<%.2f)", + pos, mlen, ZSTD_fCost(price), ZSTD_fCost(opt[pos].price)); + while (last_pos < pos) { opt[last_pos+1].price = ZSTD_MAX_PRICE; last_pos++; } /* fill empty positions */ + opt[pos].mlen = mlen; + opt[pos].off = offset; + opt[pos].litlen = litlen; + opt[pos].price = price; + } else { + DEBUGLOG(7, "rPos:%u (ml=%2u) => new price is worse (%.2f>=%.2f)", + pos, mlen, ZSTD_fCost(price), ZSTD_fCost(opt[pos].price)); + if (optLevel==0) break; /* early update abort; gets ~+10% speed for about -0.01 ratio loss */ + } + } } } + } /* for (cur = 1; cur <= last_pos; cur++) */ + + lastSequence = opt[last_pos]; + cur = last_pos > ZSTD_totalLen(lastSequence) ? last_pos - ZSTD_totalLen(lastSequence) : 0; /* single sequence, and it starts before `ip` */ + assert(cur < ZSTD_OPT_NUM); /* control overflow*/ + +_shortestPath: /* cur, last_pos, best_mlen, best_off have to be set */ + assert(opt[0].mlen == 0); + + /* Set the next chunk's repcodes based on the repcodes of the beginning + * of the last match, and the last sequence. This avoids us having to + * update them while traversing the sequences. + */ + if (lastSequence.mlen != 0) { + repcodes_t reps = ZSTD_updateRep(opt[cur].rep, lastSequence.off, lastSequence.litlen==0); + ZSTD_memcpy(rep, &reps, sizeof(reps)); + } else { + ZSTD_memcpy(rep, opt[cur].rep, sizeof(repcodes_t)); + } + + { U32 const storeEnd = cur + 1; + U32 storeStart = storeEnd; + U32 seqPos = cur; + + DEBUGLOG(6, "start reverse traversal (last_pos:%u, cur:%u)", + last_pos, cur); (void)last_pos; + assert(storeEnd < ZSTD_OPT_NUM); + DEBUGLOG(6, "last sequence copied into pos=%u (llen=%u,mlen=%u,ofc=%u)", + storeEnd, lastSequence.litlen, lastSequence.mlen, lastSequence.off); + opt[storeEnd] = lastSequence; + while (seqPos > 0) { + U32 const backDist = ZSTD_totalLen(opt[seqPos]); + storeStart--; + DEBUGLOG(6, "sequence from rPos=%u copied into pos=%u (llen=%u,mlen=%u,ofc=%u)", + seqPos, storeStart, opt[seqPos].litlen, opt[seqPos].mlen, opt[seqPos].off); + opt[storeStart] = opt[seqPos]; + seqPos = (seqPos > backDist) ? seqPos - backDist : 0; + } + + /* save sequences */ + DEBUGLOG(6, "sending selected sequences into seqStore") + { U32 storePos; + for (storePos=storeStart; storePos <= storeEnd; storePos++) { + U32 const llen = opt[storePos].litlen; + U32 const mlen = opt[storePos].mlen; + U32 const offCode = opt[storePos].off; + U32 const advance = llen + mlen; + DEBUGLOG(6, "considering seq starting at %zi, llen=%u, mlen=%u", + anchor - istart, (unsigned)llen, (unsigned)mlen); + + if (mlen==0) { /* only literals => must be last "sequence", actually starting a new stream of sequences */ + assert(storePos == storeEnd); /* must be last sequence */ + ip = anchor + llen; /* last "sequence" is a bunch of literals => don't progress anchor */ + continue; /* will finish */ + } + + assert(anchor + llen <= iend); + ZSTD_updateStats(optStatePtr, llen, anchor, offCode, mlen); + ZSTD_storeSeq(seqStore, llen, anchor, iend, offCode, mlen-MINMATCH); + anchor += advance; + ip = anchor; + } } + ZSTD_setBasePrices(optStatePtr, optLevel); + } + } /* while (ip < ilimit) */ + + /* Return the last literals size */ + return (size_t)(iend - anchor); +} + + +size_t ZSTD_compressBlock_btopt( + ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM], + const void* src, size_t srcSize) +{ + DEBUGLOG(5, "ZSTD_compressBlock_btopt"); + return ZSTD_compressBlock_opt_generic(ms, seqStore, rep, src, srcSize, 0 /*optLevel*/, ZSTD_noDict); +} + + +/* used in 2-pass strategy */ +static U32 ZSTD_upscaleStat(unsigned* table, U32 lastEltIndex, int bonus) +{ + U32 s, sum=0; + assert(ZSTD_FREQ_DIV+bonus >= 0); + for (s=0; slitSum = ZSTD_upscaleStat(optPtr->litFreq, MaxLit, 0); + optPtr->litLengthSum = ZSTD_upscaleStat(optPtr->litLengthFreq, MaxLL, 0); + optPtr->matchLengthSum = ZSTD_upscaleStat(optPtr->matchLengthFreq, MaxML, 0); + optPtr->offCodeSum = ZSTD_upscaleStat(optPtr->offCodeFreq, MaxOff, 0); +} + +/* ZSTD_initStats_ultra(): + * make a first compression pass, just to seed stats with more accurate starting values. + * only works on first block, with no dictionary and no ldm. + * this function cannot error, hence its contract must be respected. + */ +static void +ZSTD_initStats_ultra(ZSTD_matchState_t* ms, + seqStore_t* seqStore, + U32 rep[ZSTD_REP_NUM], + const void* src, size_t srcSize) +{ + U32 tmpRep[ZSTD_REP_NUM]; /* updated rep codes will sink here */ + ZSTD_memcpy(tmpRep, rep, sizeof(tmpRep)); + + DEBUGLOG(4, "ZSTD_initStats_ultra (srcSize=%zu)", srcSize); + assert(ms->opt.litLengthSum == 0); /* first block */ + assert(seqStore->sequences == seqStore->sequencesStart); /* no ldm */ + assert(ms->window.dictLimit == ms->window.lowLimit); /* no dictionary */ + assert(ms->window.dictLimit - ms->nextToUpdate <= 1); /* no prefix (note: intentional overflow, defined as 2-complement) */ + + ZSTD_compressBlock_opt_generic(ms, seqStore, tmpRep, src, srcSize, 2 /*optLevel*/, ZSTD_noDict); /* generate stats into ms->opt*/ + + /* invalidate first scan from history */ + ZSTD_resetSeqStore(seqStore); + ms->window.base -= srcSize; + ms->window.dictLimit += (U32)srcSize; + ms->window.lowLimit = ms->window.dictLimit; + ms->nextToUpdate = ms->window.dictLimit; + + /* re-inforce weight of collected statistics */ + ZSTD_upscaleStats(&ms->opt); +} + +size_t ZSTD_compressBlock_btultra( + ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM], + const void* src, size_t srcSize) +{ + DEBUGLOG(5, "ZSTD_compressBlock_btultra (srcSize=%zu)", srcSize); + return ZSTD_compressBlock_opt_generic(ms, seqStore, rep, src, srcSize, 2 /*optLevel*/, ZSTD_noDict); +} + +size_t ZSTD_compressBlock_btultra2( + ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM], + const void* src, size_t srcSize) +{ + U32 const curr = (U32)((const BYTE*)src - ms->window.base); + DEBUGLOG(5, "ZSTD_compressBlock_btultra2 (srcSize=%zu)", srcSize); + + /* 2-pass strategy: + * this strategy makes a first pass over first block to collect statistics + * and seed next round's statistics with it. + * After 1st pass, function forgets everything, and starts a new block. + * Consequently, this can only work if no data has been previously loaded in tables, + * aka, no dictionary, no prefix, no ldm preprocessing. + * The compression ratio gain is generally small (~0.5% on first block), + * the cost is 2x cpu time on first block. */ + assert(srcSize <= ZSTD_BLOCKSIZE_MAX); + if ( (ms->opt.litLengthSum==0) /* first block */ + && (seqStore->sequences == seqStore->sequencesStart) /* no ldm */ + && (ms->window.dictLimit == ms->window.lowLimit) /* no dictionary */ + && (curr == ms->window.dictLimit) /* start of frame, nothing already loaded nor skipped */ + && (srcSize > ZSTD_PREDEF_THRESHOLD) + ) { + ZSTD_initStats_ultra(ms, seqStore, rep, src, srcSize); + } + + return ZSTD_compressBlock_opt_generic(ms, seqStore, rep, src, srcSize, 2 /*optLevel*/, ZSTD_noDict); +} + +size_t ZSTD_compressBlock_btopt_dictMatchState( + ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM], + const void* src, size_t srcSize) +{ + return ZSTD_compressBlock_opt_generic(ms, seqStore, rep, src, srcSize, 0 /*optLevel*/, ZSTD_dictMatchState); +} + +size_t ZSTD_compressBlock_btultra_dictMatchState( + ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM], + const void* src, size_t srcSize) +{ + return ZSTD_compressBlock_opt_generic(ms, seqStore, rep, src, srcSize, 2 /*optLevel*/, ZSTD_dictMatchState); +} + +size_t ZSTD_compressBlock_btopt_extDict( + ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM], + const void* src, size_t srcSize) +{ + return ZSTD_compressBlock_opt_generic(ms, seqStore, rep, src, srcSize, 0 /*optLevel*/, ZSTD_extDict); +} + +size_t ZSTD_compressBlock_btultra_extDict( + ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM], + const void* src, size_t srcSize) +{ + return ZSTD_compressBlock_opt_generic(ms, seqStore, rep, src, srcSize, 2 /*optLevel*/, ZSTD_extDict); +} + +/* note : no btultra2 variant for extDict nor dictMatchState, + * because btultra2 is not meant to work with dictionaries + * and is only specific for the first block (no prefix) */ diff --git a/lib/zstd/compress/zstd_opt.h b/lib/zstd/compress/zstd_opt.h new file mode 100644 index 000000000000..22b862858ba7 --- /dev/null +++ b/lib/zstd/compress/zstd_opt.h @@ -0,0 +1,50 @@ +/* + * Copyright (c) Yann Collet, Facebook, Inc. + * All rights reserved. + * + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). + * You may select, at your option, one of the above-listed licenses. + */ + +#ifndef ZSTD_OPT_H +#define ZSTD_OPT_H + + +#include "zstd_compress_internal.h" + +/* used in ZSTD_loadDictionaryContent() */ +void ZSTD_updateTree(ZSTD_matchState_t* ms, const BYTE* ip, const BYTE* iend); + +size_t ZSTD_compressBlock_btopt( + ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM], + void const* src, size_t srcSize); +size_t ZSTD_compressBlock_btultra( + ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM], + void const* src, size_t srcSize); +size_t ZSTD_compressBlock_btultra2( + ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM], + void const* src, size_t srcSize); + + +size_t ZSTD_compressBlock_btopt_dictMatchState( + ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM], + void const* src, size_t srcSize); +size_t ZSTD_compressBlock_btultra_dictMatchState( + ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM], + void const* src, size_t srcSize); + +size_t ZSTD_compressBlock_btopt_extDict( + ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM], + void const* src, size_t srcSize); +size_t ZSTD_compressBlock_btultra_extDict( + ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM], + void const* src, size_t srcSize); + + /* note : no btultra2 variant for extDict nor dictMatchState, + * because btultra2 is not meant to work with dictionaries + * and is only specific for the first block (no prefix) */ + + +#endif /* ZSTD_OPT_H */ diff --git a/lib/zstd/decompress.c b/lib/zstd/decompress.c deleted file mode 100644 index 02e92c2cbf4f..000000000000 --- a/lib/zstd/decompress.c +++ /dev/null @@ -1,2571 +0,0 @@ -/** - * Copyright (c) 2016-present, Yann Collet, Facebook, Inc. - * All rights reserved. - * - * This source code is licensed under the BSD-style license found in the - * LICENSE file in the root directory of https://github.com/facebook/zstd. - * An additional grant of patent rights can be found in the PATENTS file in the - * same directory. - * - * This program is free software; you can redistribute it and/or modify it under - * the terms of the GNU General Public License version 2 as published by the - * Free Software Foundation. This program is dual-licensed; you may select - * either version 2 of the GNU General Public License ("GPL") or BSD license - * ("BSD"). - */ - -/* *************************************************************** -* Tuning parameters -*****************************************************************/ -/*! -* MAXWINDOWSIZE_DEFAULT : -* maximum window size accepted by DStream, by default. -* Frames requiring more memory will be rejected. -*/ -#ifndef ZSTD_MAXWINDOWSIZE_DEFAULT -#define ZSTD_MAXWINDOWSIZE_DEFAULT ((1 << ZSTD_WINDOWLOG_MAX) + 1) /* defined within zstd.h */ -#endif - -/*-******************************************************* -* Dependencies -*********************************************************/ -#include "fse.h" -#include "huf.h" -#include "mem.h" /* low level memory routines */ -#include "zstd_internal.h" -#include -#include -#include /* memcpy, memmove, memset */ - -#define ZSTD_PREFETCH(ptr) __builtin_prefetch(ptr, 0, 0) - -/*-************************************* -* Macros -***************************************/ -#define ZSTD_isError ERR_isError /* for inlining */ -#define FSE_isError ERR_isError -#define HUF_isError ERR_isError - -/*_******************************************************* -* Memory operations -**********************************************************/ -static void ZSTD_copy4(void *dst, const void *src) { memcpy(dst, src, 4); } - -/*-************************************************************* -* Context management -***************************************************************/ -typedef enum { - ZSTDds_getFrameHeaderSize, - ZSTDds_decodeFrameHeader, - ZSTDds_decodeBlockHeader, - ZSTDds_decompressBlock, - ZSTDds_decompressLastBlock, - ZSTDds_checkChecksum, - ZSTDds_decodeSkippableHeader, - ZSTDds_skipFrame -} ZSTD_dStage; - -typedef struct { - FSE_DTable LLTable[FSE_DTABLE_SIZE_U32(LLFSELog)]; - FSE_DTable OFTable[FSE_DTABLE_SIZE_U32(OffFSELog)]; - FSE_DTable MLTable[FSE_DTABLE_SIZE_U32(MLFSELog)]; - HUF_DTable hufTable[HUF_DTABLE_SIZE(HufLog)]; /* can accommodate HUF_decompress4X */ - U64 workspace[HUF_DECOMPRESS_WORKSPACE_SIZE_U32 / 2]; - U32 rep[ZSTD_REP_NUM]; -} ZSTD_entropyTables_t; - -struct ZSTD_DCtx_s { - const FSE_DTable *LLTptr; - const FSE_DTable *MLTptr; - const FSE_DTable *OFTptr; - const HUF_DTable *HUFptr; - ZSTD_entropyTables_t entropy; - const void *previousDstEnd; /* detect continuity */ - const void *base; /* start of curr segment */ - const void *vBase; /* virtual start of previous segment if it was just before curr one */ - const void *dictEnd; /* end of previous segment */ - size_t expected; - ZSTD_frameParams fParams; - blockType_e bType; /* used in ZSTD_decompressContinue(), to transfer blockType between header decoding and block decoding stages */ - ZSTD_dStage stage; - U32 litEntropy; - U32 fseEntropy; - struct xxh64_state xxhState; - size_t headerSize; - U32 dictID; - const BYTE *litPtr; - ZSTD_customMem customMem; - size_t litSize; - size_t rleSize; - BYTE litBuffer[ZSTD_BLOCKSIZE_ABSOLUTEMAX + WILDCOPY_OVERLENGTH]; - BYTE headerBuffer[ZSTD_FRAMEHEADERSIZE_MAX]; -}; /* typedef'd to ZSTD_DCtx within "zstd.h" */ - -size_t ZSTD_DCtxWorkspaceBound(void) { return ZSTD_ALIGN(sizeof(ZSTD_stack)) + ZSTD_ALIGN(sizeof(ZSTD_DCtx)); } - -size_t ZSTD_decompressBegin(ZSTD_DCtx *dctx) -{ - dctx->expected = ZSTD_frameHeaderSize_prefix; - dctx->stage = ZSTDds_getFrameHeaderSize; - dctx->previousDstEnd = NULL; - dctx->base = NULL; - dctx->vBase = NULL; - dctx->dictEnd = NULL; - dctx->entropy.hufTable[0] = (HUF_DTable)((HufLog)*0x1000001); /* cover both little and big endian */ - dctx->litEntropy = dctx->fseEntropy = 0; - dctx->dictID = 0; - ZSTD_STATIC_ASSERT(sizeof(dctx->entropy.rep) == sizeof(repStartValue)); - memcpy(dctx->entropy.rep, repStartValue, sizeof(repStartValue)); /* initial repcodes */ - dctx->LLTptr = dctx->entropy.LLTable; - dctx->MLTptr = dctx->entropy.MLTable; - dctx->OFTptr = dctx->entropy.OFTable; - dctx->HUFptr = dctx->entropy.hufTable; - return 0; -} - -ZSTD_DCtx *ZSTD_createDCtx_advanced(ZSTD_customMem customMem) -{ - ZSTD_DCtx *dctx; - - if (!customMem.customAlloc || !customMem.customFree) - return NULL; - - dctx = (ZSTD_DCtx *)ZSTD_malloc(sizeof(ZSTD_DCtx), customMem); - if (!dctx) - return NULL; - memcpy(&dctx->customMem, &customMem, sizeof(customMem)); - ZSTD_decompressBegin(dctx); - return dctx; -} - -ZSTD_DCtx *ZSTD_initDCtx(void *workspace, size_t workspaceSize) -{ - ZSTD_customMem const stackMem = ZSTD_initStack(workspace, workspaceSize); - return ZSTD_createDCtx_advanced(stackMem); -} - -size_t ZSTD_freeDCtx(ZSTD_DCtx *dctx) -{ - if (dctx == NULL) - return 0; /* support free on NULL */ - ZSTD_free(dctx, dctx->customMem); - return 0; /* reserved as a potential error code in the future */ -} - -void ZSTD_copyDCtx(ZSTD_DCtx *dstDCtx, const ZSTD_DCtx *srcDCtx) -{ - size_t const workSpaceSize = (ZSTD_BLOCKSIZE_ABSOLUTEMAX + WILDCOPY_OVERLENGTH) + ZSTD_frameHeaderSize_max; - memcpy(dstDCtx, srcDCtx, sizeof(ZSTD_DCtx) - workSpaceSize); /* no need to copy workspace */ -} - -static void ZSTD_refDDict(ZSTD_DCtx *dstDCtx, const ZSTD_DDict *ddict); - -/*-************************************************************* -* Decompression section -***************************************************************/ - -/*! ZSTD_isFrame() : - * Tells if the content of `buffer` starts with a valid Frame Identifier. - * Note : Frame Identifier is 4 bytes. If `size < 4`, @return will always be 0. - * Note 2 : Legacy Frame Identifiers are considered valid only if Legacy Support is enabled. - * Note 3 : Skippable Frame Identifiers are considered valid. */ -unsigned ZSTD_isFrame(const void *buffer, size_t size) -{ - if (size < 4) - return 0; - { - U32 const magic = ZSTD_readLE32(buffer); - if (magic == ZSTD_MAGICNUMBER) - return 1; - if ((magic & 0xFFFFFFF0U) == ZSTD_MAGIC_SKIPPABLE_START) - return 1; - } - return 0; -} - -/** ZSTD_frameHeaderSize() : -* srcSize must be >= ZSTD_frameHeaderSize_prefix. -* @return : size of the Frame Header */ -static size_t ZSTD_frameHeaderSize(const void *src, size_t srcSize) -{ - if (srcSize < ZSTD_frameHeaderSize_prefix) - return ERROR(srcSize_wrong); - { - BYTE const fhd = ((const BYTE *)src)[4]; - U32 const dictID = fhd & 3; - U32 const singleSegment = (fhd >> 5) & 1; - U32 const fcsId = fhd >> 6; - return ZSTD_frameHeaderSize_prefix + !singleSegment + ZSTD_did_fieldSize[dictID] + ZSTD_fcs_fieldSize[fcsId] + (singleSegment && !fcsId); - } -} - -/** ZSTD_getFrameParams() : -* decode Frame Header, or require larger `srcSize`. -* @return : 0, `fparamsPtr` is correctly filled, -* >0, `srcSize` is too small, result is expected `srcSize`, -* or an error code, which can be tested using ZSTD_isError() */ -size_t ZSTD_getFrameParams(ZSTD_frameParams *fparamsPtr, const void *src, size_t srcSize) -{ - const BYTE *ip = (const BYTE *)src; - - if (srcSize < ZSTD_frameHeaderSize_prefix) - return ZSTD_frameHeaderSize_prefix; - if (ZSTD_readLE32(src) != ZSTD_MAGICNUMBER) { - if ((ZSTD_readLE32(src) & 0xFFFFFFF0U) == ZSTD_MAGIC_SKIPPABLE_START) { - if (srcSize < ZSTD_skippableHeaderSize) - return ZSTD_skippableHeaderSize; /* magic number + skippable frame length */ - memset(fparamsPtr, 0, sizeof(*fparamsPtr)); - fparamsPtr->frameContentSize = ZSTD_readLE32((const char *)src + 4); - fparamsPtr->windowSize = 0; /* windowSize==0 means a frame is skippable */ - return 0; - } - return ERROR(prefix_unknown); - } - - /* ensure there is enough `srcSize` to fully read/decode frame header */ - { - size_t const fhsize = ZSTD_frameHeaderSize(src, srcSize); - if (srcSize < fhsize) - return fhsize; - } - - { - BYTE const fhdByte = ip[4]; - size_t pos = 5; - U32 const dictIDSizeCode = fhdByte & 3; - U32 const checksumFlag = (fhdByte >> 2) & 1; - U32 const singleSegment = (fhdByte >> 5) & 1; - U32 const fcsID = fhdByte >> 6; - U32 const windowSizeMax = 1U << ZSTD_WINDOWLOG_MAX; - U32 windowSize = 0; - U32 dictID = 0; - U64 frameContentSize = 0; - if ((fhdByte & 0x08) != 0) - return ERROR(frameParameter_unsupported); /* reserved bits, which must be zero */ - if (!singleSegment) { - BYTE const wlByte = ip[pos++]; - U32 const windowLog = (wlByte >> 3) + ZSTD_WINDOWLOG_ABSOLUTEMIN; - if (windowLog > ZSTD_WINDOWLOG_MAX) - return ERROR(frameParameter_windowTooLarge); /* avoids issue with 1 << windowLog */ - windowSize = (1U << windowLog); - windowSize += (windowSize >> 3) * (wlByte & 7); - } - - switch (dictIDSizeCode) { - default: /* impossible */ - case 0: break; - case 1: - dictID = ip[pos]; - pos++; - break; - case 2: - dictID = ZSTD_readLE16(ip + pos); - pos += 2; - break; - case 3: - dictID = ZSTD_readLE32(ip + pos); - pos += 4; - break; - } - switch (fcsID) { - default: /* impossible */ - case 0: - if (singleSegment) - frameContentSize = ip[pos]; - break; - case 1: frameContentSize = ZSTD_readLE16(ip + pos) + 256; break; - case 2: frameContentSize = ZSTD_readLE32(ip + pos); break; - case 3: frameContentSize = ZSTD_readLE64(ip + pos); break; - } - if (!windowSize) - windowSize = (U32)frameContentSize; - if (windowSize > windowSizeMax) - return ERROR(frameParameter_windowTooLarge); - fparamsPtr->frameContentSize = frameContentSize; - fparamsPtr->windowSize = windowSize; - fparamsPtr->dictID = dictID; - fparamsPtr->checksumFlag = checksumFlag; - } - return 0; -} - -/** ZSTD_getFrameContentSize() : -* compatible with legacy mode -* @return : decompressed size of the single frame pointed to be `src` if known, otherwise -* - ZSTD_CONTENTSIZE_UNKNOWN if the size cannot be determined -* - ZSTD_CONTENTSIZE_ERROR if an error occurred (e.g. invalid magic number, srcSize too small) */ -unsigned long long ZSTD_getFrameContentSize(const void *src, size_t srcSize) -{ - { - ZSTD_frameParams fParams; - if (ZSTD_getFrameParams(&fParams, src, srcSize) != 0) - return ZSTD_CONTENTSIZE_ERROR; - if (fParams.windowSize == 0) { - /* Either skippable or empty frame, size == 0 either way */ - return 0; - } else if (fParams.frameContentSize != 0) { - return fParams.frameContentSize; - } else { - return ZSTD_CONTENTSIZE_UNKNOWN; - } - } -} - -/** ZSTD_findDecompressedSize() : - * compatible with legacy mode - * `srcSize` must be the exact length of some number of ZSTD compressed and/or - * skippable frames - * @return : decompressed size of the frames contained */ -unsigned long long ZSTD_findDecompressedSize(const void *src, size_t srcSize) -{ - { - unsigned long long totalDstSize = 0; - while (srcSize >= ZSTD_frameHeaderSize_prefix) { - const U32 magicNumber = ZSTD_readLE32(src); - - if ((magicNumber & 0xFFFFFFF0U) == ZSTD_MAGIC_SKIPPABLE_START) { - size_t skippableSize; - if (srcSize < ZSTD_skippableHeaderSize) - return ERROR(srcSize_wrong); - skippableSize = ZSTD_readLE32((const BYTE *)src + 4) + ZSTD_skippableHeaderSize; - if (srcSize < skippableSize) { - return ZSTD_CONTENTSIZE_ERROR; - } - - src = (const BYTE *)src + skippableSize; - srcSize -= skippableSize; - continue; - } - - { - unsigned long long const ret = ZSTD_getFrameContentSize(src, srcSize); - if (ret >= ZSTD_CONTENTSIZE_ERROR) - return ret; - - /* check for overflow */ - if (totalDstSize + ret < totalDstSize) - return ZSTD_CONTENTSIZE_ERROR; - totalDstSize += ret; - } - { - size_t const frameSrcSize = ZSTD_findFrameCompressedSize(src, srcSize); - if (ZSTD_isError(frameSrcSize)) { - return ZSTD_CONTENTSIZE_ERROR; - } - - src = (const BYTE *)src + frameSrcSize; - srcSize -= frameSrcSize; - } - } - - if (srcSize) { - return ZSTD_CONTENTSIZE_ERROR; - } - - return totalDstSize; - } -} - -/** ZSTD_decodeFrameHeader() : -* `headerSize` must be the size provided by ZSTD_frameHeaderSize(). -* @return : 0 if success, or an error code, which can be tested using ZSTD_isError() */ -static size_t ZSTD_decodeFrameHeader(ZSTD_DCtx *dctx, const void *src, size_t headerSize) -{ - size_t const result = ZSTD_getFrameParams(&(dctx->fParams), src, headerSize); - if (ZSTD_isError(result)) - return result; /* invalid header */ - if (result > 0) - return ERROR(srcSize_wrong); /* headerSize too small */ - if (dctx->fParams.dictID && (dctx->dictID != dctx->fParams.dictID)) - return ERROR(dictionary_wrong); - if (dctx->fParams.checksumFlag) - xxh64_reset(&dctx->xxhState, 0); - return 0; -} - -typedef struct { - blockType_e blockType; - U32 lastBlock; - U32 origSize; -} blockProperties_t; - -/*! ZSTD_getcBlockSize() : -* Provides the size of compressed block from block header `src` */ -size_t ZSTD_getcBlockSize(const void *src, size_t srcSize, blockProperties_t *bpPtr) -{ - if (srcSize < ZSTD_blockHeaderSize) - return ERROR(srcSize_wrong); - { - U32 const cBlockHeader = ZSTD_readLE24(src); - U32 const cSize = cBlockHeader >> 3; - bpPtr->lastBlock = cBlockHeader & 1; - bpPtr->blockType = (blockType_e)((cBlockHeader >> 1) & 3); - bpPtr->origSize = cSize; /* only useful for RLE */ - if (bpPtr->blockType == bt_rle) - return 1; - if (bpPtr->blockType == bt_reserved) - return ERROR(corruption_detected); - return cSize; - } -} - -static size_t ZSTD_copyRawBlock(void *dst, size_t dstCapacity, const void *src, size_t srcSize) -{ - if (srcSize > dstCapacity) - return ERROR(dstSize_tooSmall); - memcpy(dst, src, srcSize); - return srcSize; -} - -static size_t ZSTD_setRleBlock(void *dst, size_t dstCapacity, const void *src, size_t srcSize, size_t regenSize) -{ - if (srcSize != 1) - return ERROR(srcSize_wrong); - if (regenSize > dstCapacity) - return ERROR(dstSize_tooSmall); - memset(dst, *(const BYTE *)src, regenSize); - return regenSize; -} - -/*! ZSTD_decodeLiteralsBlock() : - @return : nb of bytes read from src (< srcSize ) */ -size_t ZSTD_decodeLiteralsBlock(ZSTD_DCtx *dctx, const void *src, size_t srcSize) /* note : srcSize < BLOCKSIZE */ -{ - if (srcSize < MIN_CBLOCK_SIZE) - return ERROR(corruption_detected); - - { - const BYTE *const istart = (const BYTE *)src; - symbolEncodingType_e const litEncType = (symbolEncodingType_e)(istart[0] & 3); - - switch (litEncType) { - case set_repeat: - if (dctx->litEntropy == 0) - return ERROR(dictionary_corrupted); - fallthrough; - case set_compressed: - if (srcSize < 5) - return ERROR(corruption_detected); /* srcSize >= MIN_CBLOCK_SIZE == 3; here we need up to 5 for case 3 */ - { - size_t lhSize, litSize, litCSize; - U32 singleStream = 0; - U32 const lhlCode = (istart[0] >> 2) & 3; - U32 const lhc = ZSTD_readLE32(istart); - switch (lhlCode) { - case 0: - case 1: - default: /* note : default is impossible, since lhlCode into [0..3] */ - /* 2 - 2 - 10 - 10 */ - singleStream = !lhlCode; - lhSize = 3; - litSize = (lhc >> 4) & 0x3FF; - litCSize = (lhc >> 14) & 0x3FF; - break; - case 2: - /* 2 - 2 - 14 - 14 */ - lhSize = 4; - litSize = (lhc >> 4) & 0x3FFF; - litCSize = lhc >> 18; - break; - case 3: - /* 2 - 2 - 18 - 18 */ - lhSize = 5; - litSize = (lhc >> 4) & 0x3FFFF; - litCSize = (lhc >> 22) + (istart[4] << 10); - break; - } - if (litSize > ZSTD_BLOCKSIZE_ABSOLUTEMAX) - return ERROR(corruption_detected); - if (litCSize + lhSize > srcSize) - return ERROR(corruption_detected); - - if (HUF_isError( - (litEncType == set_repeat) - ? (singleStream ? HUF_decompress1X_usingDTable(dctx->litBuffer, litSize, istart + lhSize, litCSize, dctx->HUFptr) - : HUF_decompress4X_usingDTable(dctx->litBuffer, litSize, istart + lhSize, litCSize, dctx->HUFptr)) - : (singleStream - ? HUF_decompress1X2_DCtx_wksp(dctx->entropy.hufTable, dctx->litBuffer, litSize, istart + lhSize, litCSize, - dctx->entropy.workspace, sizeof(dctx->entropy.workspace)) - : HUF_decompress4X_hufOnly_wksp(dctx->entropy.hufTable, dctx->litBuffer, litSize, istart + lhSize, litCSize, - dctx->entropy.workspace, sizeof(dctx->entropy.workspace))))) - return ERROR(corruption_detected); - - dctx->litPtr = dctx->litBuffer; - dctx->litSize = litSize; - dctx->litEntropy = 1; - if (litEncType == set_compressed) - dctx->HUFptr = dctx->entropy.hufTable; - memset(dctx->litBuffer + dctx->litSize, 0, WILDCOPY_OVERLENGTH); - return litCSize + lhSize; - } - - case set_basic: { - size_t litSize, lhSize; - U32 const lhlCode = ((istart[0]) >> 2) & 3; - switch (lhlCode) { - case 0: - case 2: - default: /* note : default is impossible, since lhlCode into [0..3] */ - lhSize = 1; - litSize = istart[0] >> 3; - break; - case 1: - lhSize = 2; - litSize = ZSTD_readLE16(istart) >> 4; - break; - case 3: - lhSize = 3; - litSize = ZSTD_readLE24(istart) >> 4; - break; - } - - if (lhSize + litSize + WILDCOPY_OVERLENGTH > srcSize) { /* risk reading beyond src buffer with wildcopy */ - if (litSize + lhSize > srcSize) - return ERROR(corruption_detected); - memcpy(dctx->litBuffer, istart + lhSize, litSize); - dctx->litPtr = dctx->litBuffer; - dctx->litSize = litSize; - memset(dctx->litBuffer + dctx->litSize, 0, WILDCOPY_OVERLENGTH); - return lhSize + litSize; - } - /* direct reference into compressed stream */ - dctx->litPtr = istart + lhSize; - dctx->litSize = litSize; - return lhSize + litSize; - } - - case set_rle: { - U32 const lhlCode = ((istart[0]) >> 2) & 3; - size_t litSize, lhSize; - switch (lhlCode) { - case 0: - case 2: - default: /* note : default is impossible, since lhlCode into [0..3] */ - lhSize = 1; - litSize = istart[0] >> 3; - break; - case 1: - lhSize = 2; - litSize = ZSTD_readLE16(istart) >> 4; - break; - case 3: - lhSize = 3; - litSize = ZSTD_readLE24(istart) >> 4; - if (srcSize < 4) - return ERROR(corruption_detected); /* srcSize >= MIN_CBLOCK_SIZE == 3; here we need lhSize+1 = 4 */ - break; - } - if (litSize > ZSTD_BLOCKSIZE_ABSOLUTEMAX) - return ERROR(corruption_detected); - memset(dctx->litBuffer, istart[lhSize], litSize + WILDCOPY_OVERLENGTH); - dctx->litPtr = dctx->litBuffer; - dctx->litSize = litSize; - return lhSize + 1; - } - default: - return ERROR(corruption_detected); /* impossible */ - } - } -} - -typedef union { - FSE_decode_t realData; - U32 alignedBy4; -} FSE_decode_t4; - -static const FSE_decode_t4 LL_defaultDTable[(1 << LL_DEFAULTNORMLOG) + 1] = { - {{LL_DEFAULTNORMLOG, 1, 1}}, /* header : tableLog, fastMode, fastMode */ - {{0, 0, 4}}, /* 0 : base, symbol, bits */ - {{16, 0, 4}}, - {{32, 1, 5}}, - {{0, 3, 5}}, - {{0, 4, 5}}, - {{0, 6, 5}}, - {{0, 7, 5}}, - {{0, 9, 5}}, - {{0, 10, 5}}, - {{0, 12, 5}}, - {{0, 14, 6}}, - {{0, 16, 5}}, - {{0, 18, 5}}, - {{0, 19, 5}}, - {{0, 21, 5}}, - {{0, 22, 5}}, - {{0, 24, 5}}, - {{32, 25, 5}}, - {{0, 26, 5}}, - {{0, 27, 6}}, - {{0, 29, 6}}, - {{0, 31, 6}}, - {{32, 0, 4}}, - {{0, 1, 4}}, - {{0, 2, 5}}, - {{32, 4, 5}}, - {{0, 5, 5}}, - {{32, 7, 5}}, - {{0, 8, 5}}, - {{32, 10, 5}}, - {{0, 11, 5}}, - {{0, 13, 6}}, - {{32, 16, 5}}, - {{0, 17, 5}}, - {{32, 19, 5}}, - {{0, 20, 5}}, - {{32, 22, 5}}, - {{0, 23, 5}}, - {{0, 25, 4}}, - {{16, 25, 4}}, - {{32, 26, 5}}, - {{0, 28, 6}}, - {{0, 30, 6}}, - {{48, 0, 4}}, - {{16, 1, 4}}, - {{32, 2, 5}}, - {{32, 3, 5}}, - {{32, 5, 5}}, - {{32, 6, 5}}, - {{32, 8, 5}}, - {{32, 9, 5}}, - {{32, 11, 5}}, - {{32, 12, 5}}, - {{0, 15, 6}}, - {{32, 17, 5}}, - {{32, 18, 5}}, - {{32, 20, 5}}, - {{32, 21, 5}}, - {{32, 23, 5}}, - {{32, 24, 5}}, - {{0, 35, 6}}, - {{0, 34, 6}}, - {{0, 33, 6}}, - {{0, 32, 6}}, -}; /* LL_defaultDTable */ - -static const FSE_decode_t4 ML_defaultDTable[(1 << ML_DEFAULTNORMLOG) + 1] = { - {{ML_DEFAULTNORMLOG, 1, 1}}, /* header : tableLog, fastMode, fastMode */ - {{0, 0, 6}}, /* 0 : base, symbol, bits */ - {{0, 1, 4}}, - {{32, 2, 5}}, - {{0, 3, 5}}, - {{0, 5, 5}}, - {{0, 6, 5}}, - {{0, 8, 5}}, - {{0, 10, 6}}, - {{0, 13, 6}}, - {{0, 16, 6}}, - {{0, 19, 6}}, - {{0, 22, 6}}, - {{0, 25, 6}}, - {{0, 28, 6}}, - {{0, 31, 6}}, - {{0, 33, 6}}, - {{0, 35, 6}}, - {{0, 37, 6}}, - {{0, 39, 6}}, - {{0, 41, 6}}, - {{0, 43, 6}}, - {{0, 45, 6}}, - {{16, 1, 4}}, - {{0, 2, 4}}, - {{32, 3, 5}}, - {{0, 4, 5}}, - {{32, 6, 5}}, - {{0, 7, 5}}, - {{0, 9, 6}}, - {{0, 12, 6}}, - {{0, 15, 6}}, - {{0, 18, 6}}, - {{0, 21, 6}}, - {{0, 24, 6}}, - {{0, 27, 6}}, - {{0, 30, 6}}, - {{0, 32, 6}}, - {{0, 34, 6}}, - {{0, 36, 6}}, - {{0, 38, 6}}, - {{0, 40, 6}}, - {{0, 42, 6}}, - {{0, 44, 6}}, - {{32, 1, 4}}, - {{48, 1, 4}}, - {{16, 2, 4}}, - {{32, 4, 5}}, - {{32, 5, 5}}, - {{32, 7, 5}}, - {{32, 8, 5}}, - {{0, 11, 6}}, - {{0, 14, 6}}, - {{0, 17, 6}}, - {{0, 20, 6}}, - {{0, 23, 6}}, - {{0, 26, 6}}, - {{0, 29, 6}}, - {{0, 52, 6}}, - {{0, 51, 6}}, - {{0, 50, 6}}, - {{0, 49, 6}}, - {{0, 48, 6}}, - {{0, 47, 6}}, - {{0, 46, 6}}, -}; /* ML_defaultDTable */ - -static const FSE_decode_t4 OF_defaultDTable[(1 << OF_DEFAULTNORMLOG) + 1] = { - {{OF_DEFAULTNORMLOG, 1, 1}}, /* header : tableLog, fastMode, fastMode */ - {{0, 0, 5}}, /* 0 : base, symbol, bits */ - {{0, 6, 4}}, - {{0, 9, 5}}, - {{0, 15, 5}}, - {{0, 21, 5}}, - {{0, 3, 5}}, - {{0, 7, 4}}, - {{0, 12, 5}}, - {{0, 18, 5}}, - {{0, 23, 5}}, - {{0, 5, 5}}, - {{0, 8, 4}}, - {{0, 14, 5}}, - {{0, 20, 5}}, - {{0, 2, 5}}, - {{16, 7, 4}}, - {{0, 11, 5}}, - {{0, 17, 5}}, - {{0, 22, 5}}, - {{0, 4, 5}}, - {{16, 8, 4}}, - {{0, 13, 5}}, - {{0, 19, 5}}, - {{0, 1, 5}}, - {{16, 6, 4}}, - {{0, 10, 5}}, - {{0, 16, 5}}, - {{0, 28, 5}}, - {{0, 27, 5}}, - {{0, 26, 5}}, - {{0, 25, 5}}, - {{0, 24, 5}}, -}; /* OF_defaultDTable */ - -/*! ZSTD_buildSeqTable() : - @return : nb bytes read from src, - or an error code if it fails, testable with ZSTD_isError() -*/ -static size_t ZSTD_buildSeqTable(FSE_DTable *DTableSpace, const FSE_DTable **DTablePtr, symbolEncodingType_e type, U32 max, U32 maxLog, const void *src, - size_t srcSize, const FSE_decode_t4 *defaultTable, U32 flagRepeatTable, void *workspace, size_t workspaceSize) -{ - const void *const tmpPtr = defaultTable; /* bypass strict aliasing */ - switch (type) { - case set_rle: - if (!srcSize) - return ERROR(srcSize_wrong); - if ((*(const BYTE *)src) > max) - return ERROR(corruption_detected); - FSE_buildDTable_rle(DTableSpace, *(const BYTE *)src); - *DTablePtr = DTableSpace; - return 1; - case set_basic: *DTablePtr = (const FSE_DTable *)tmpPtr; return 0; - case set_repeat: - if (!flagRepeatTable) - return ERROR(corruption_detected); - return 0; - default: /* impossible */ - case set_compressed: { - U32 tableLog; - S16 *norm = (S16 *)workspace; - size_t const spaceUsed32 = ALIGN(sizeof(S16) * (MaxSeq + 1), sizeof(U32)) >> 2; - - if ((spaceUsed32 << 2) > workspaceSize) - return ERROR(GENERIC); - workspace = (U32 *)workspace + spaceUsed32; - workspaceSize -= (spaceUsed32 << 2); - { - size_t const headerSize = FSE_readNCount(norm, &max, &tableLog, src, srcSize); - if (FSE_isError(headerSize)) - return ERROR(corruption_detected); - if (tableLog > maxLog) - return ERROR(corruption_detected); - FSE_buildDTable_wksp(DTableSpace, norm, max, tableLog, workspace, workspaceSize); - *DTablePtr = DTableSpace; - return headerSize; - } - } - } -} - -size_t ZSTD_decodeSeqHeaders(ZSTD_DCtx *dctx, int *nbSeqPtr, const void *src, size_t srcSize) -{ - const BYTE *const istart = (const BYTE *const)src; - const BYTE *const iend = istart + srcSize; - const BYTE *ip = istart; - - /* check */ - if (srcSize < MIN_SEQUENCES_SIZE) - return ERROR(srcSize_wrong); - - /* SeqHead */ - { - int nbSeq = *ip++; - if (!nbSeq) { - *nbSeqPtr = 0; - return 1; - } - if (nbSeq > 0x7F) { - if (nbSeq == 0xFF) { - if (ip + 2 > iend) - return ERROR(srcSize_wrong); - nbSeq = ZSTD_readLE16(ip) + LONGNBSEQ, ip += 2; - } else { - if (ip >= iend) - return ERROR(srcSize_wrong); - nbSeq = ((nbSeq - 0x80) << 8) + *ip++; - } - } - *nbSeqPtr = nbSeq; - } - - /* FSE table descriptors */ - if (ip + 4 > iend) - return ERROR(srcSize_wrong); /* minimum possible size */ - { - symbolEncodingType_e const LLtype = (symbolEncodingType_e)(*ip >> 6); - symbolEncodingType_e const OFtype = (symbolEncodingType_e)((*ip >> 4) & 3); - symbolEncodingType_e const MLtype = (symbolEncodingType_e)((*ip >> 2) & 3); - ip++; - - /* Build DTables */ - { - size_t const llhSize = ZSTD_buildSeqTable(dctx->entropy.LLTable, &dctx->LLTptr, LLtype, MaxLL, LLFSELog, ip, iend - ip, - LL_defaultDTable, dctx->fseEntropy, dctx->entropy.workspace, sizeof(dctx->entropy.workspace)); - if (ZSTD_isError(llhSize)) - return ERROR(corruption_detected); - ip += llhSize; - } - { - size_t const ofhSize = ZSTD_buildSeqTable(dctx->entropy.OFTable, &dctx->OFTptr, OFtype, MaxOff, OffFSELog, ip, iend - ip, - OF_defaultDTable, dctx->fseEntropy, dctx->entropy.workspace, sizeof(dctx->entropy.workspace)); - if (ZSTD_isError(ofhSize)) - return ERROR(corruption_detected); - ip += ofhSize; - } - { - size_t const mlhSize = ZSTD_buildSeqTable(dctx->entropy.MLTable, &dctx->MLTptr, MLtype, MaxML, MLFSELog, ip, iend - ip, - ML_defaultDTable, dctx->fseEntropy, dctx->entropy.workspace, sizeof(dctx->entropy.workspace)); - if (ZSTD_isError(mlhSize)) - return ERROR(corruption_detected); - ip += mlhSize; - } - } - - return ip - istart; -} - -typedef struct { - size_t litLength; - size_t matchLength; - size_t offset; - const BYTE *match; -} seq_t; - -typedef struct { - BIT_DStream_t DStream; - FSE_DState_t stateLL; - FSE_DState_t stateOffb; - FSE_DState_t stateML; - size_t prevOffset[ZSTD_REP_NUM]; - const BYTE *base; - size_t pos; - uPtrDiff gotoDict; -} seqState_t; - -FORCE_NOINLINE -size_t ZSTD_execSequenceLast7(BYTE *op, BYTE *const oend, seq_t sequence, const BYTE **litPtr, const BYTE *const litLimit, const BYTE *const base, - const BYTE *const vBase, const BYTE *const dictEnd) -{ - BYTE *const oLitEnd = op + sequence.litLength; - size_t const sequenceLength = sequence.litLength + sequence.matchLength; - BYTE *const oMatchEnd = op + sequenceLength; /* risk : address space overflow (32-bits) */ - BYTE *const oend_w = oend - WILDCOPY_OVERLENGTH; - const BYTE *const iLitEnd = *litPtr + sequence.litLength; - const BYTE *match = oLitEnd - sequence.offset; - - /* check */ - if (oMatchEnd > oend) - return ERROR(dstSize_tooSmall); /* last match must start at a minimum distance of WILDCOPY_OVERLENGTH from oend */ - if (iLitEnd > litLimit) - return ERROR(corruption_detected); /* over-read beyond lit buffer */ - if (oLitEnd <= oend_w) - return ERROR(GENERIC); /* Precondition */ - - /* copy literals */ - if (op < oend_w) { - ZSTD_wildcopy(op, *litPtr, oend_w - op); - *litPtr += oend_w - op; - op = oend_w; - } - while (op < oLitEnd) - *op++ = *(*litPtr)++; - - /* copy Match */ - if (sequence.offset > (size_t)(oLitEnd - base)) { - /* offset beyond prefix */ - if (sequence.offset > (size_t)(oLitEnd - vBase)) - return ERROR(corruption_detected); - match = dictEnd - (base - match); - if (match + sequence.matchLength <= dictEnd) { - memmove(oLitEnd, match, sequence.matchLength); - return sequenceLength; - } - /* span extDict & currPrefixSegment */ - { - size_t const length1 = dictEnd - match; - memmove(oLitEnd, match, length1); - op = oLitEnd + length1; - sequence.matchLength -= length1; - match = base; - } - } - while (op < oMatchEnd) - *op++ = *match++; - return sequenceLength; -} - -static seq_t ZSTD_decodeSequence(seqState_t *seqState) -{ - seq_t seq; - - U32 const llCode = FSE_peekSymbol(&seqState->stateLL); - U32 const mlCode = FSE_peekSymbol(&seqState->stateML); - U32 const ofCode = FSE_peekSymbol(&seqState->stateOffb); /* <= maxOff, by table construction */ - - U32 const llBits = LL_bits[llCode]; - U32 const mlBits = ML_bits[mlCode]; - U32 const ofBits = ofCode; - U32 const totalBits = llBits + mlBits + ofBits; - - static const U32 LL_base[MaxLL + 1] = {0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 18, - 20, 22, 24, 28, 32, 40, 48, 64, 0x80, 0x100, 0x200, 0x400, 0x800, 0x1000, 0x2000, 0x4000, 0x8000, 0x10000}; - - static const U32 ML_base[MaxML + 1] = {3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, - 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 37, 39, 41, - 43, 47, 51, 59, 67, 83, 99, 0x83, 0x103, 0x203, 0x403, 0x803, 0x1003, 0x2003, 0x4003, 0x8003, 0x10003}; - - static const U32 OF_base[MaxOff + 1] = {0, 1, 1, 5, 0xD, 0x1D, 0x3D, 0x7D, 0xFD, 0x1FD, - 0x3FD, 0x7FD, 0xFFD, 0x1FFD, 0x3FFD, 0x7FFD, 0xFFFD, 0x1FFFD, 0x3FFFD, 0x7FFFD, - 0xFFFFD, 0x1FFFFD, 0x3FFFFD, 0x7FFFFD, 0xFFFFFD, 0x1FFFFFD, 0x3FFFFFD, 0x7FFFFFD, 0xFFFFFFD}; - - /* sequence */ - { - size_t offset; - if (!ofCode) - offset = 0; - else { - offset = OF_base[ofCode] + BIT_readBitsFast(&seqState->DStream, ofBits); /* <= (ZSTD_WINDOWLOG_MAX-1) bits */ - if (ZSTD_32bits()) - BIT_reloadDStream(&seqState->DStream); - } - - if (ofCode <= 1) { - offset += (llCode == 0); - if (offset) { - size_t temp = (offset == 3) ? seqState->prevOffset[0] - 1 : seqState->prevOffset[offset]; - temp += !temp; /* 0 is not valid; input is corrupted; force offset to 1 */ - if (offset != 1) - seqState->prevOffset[2] = seqState->prevOffset[1]; - seqState->prevOffset[1] = seqState->prevOffset[0]; - seqState->prevOffset[0] = offset = temp; - } else { - offset = seqState->prevOffset[0]; - } - } else { - seqState->prevOffset[2] = seqState->prevOffset[1]; - seqState->prevOffset[1] = seqState->prevOffset[0]; - seqState->prevOffset[0] = offset; - } - seq.offset = offset; - } - - seq.matchLength = ML_base[mlCode] + ((mlCode > 31) ? BIT_readBitsFast(&seqState->DStream, mlBits) : 0); /* <= 16 bits */ - if (ZSTD_32bits() && (mlBits + llBits > 24)) - BIT_reloadDStream(&seqState->DStream); - - seq.litLength = LL_base[llCode] + ((llCode > 15) ? BIT_readBitsFast(&seqState->DStream, llBits) : 0); /* <= 16 bits */ - if (ZSTD_32bits() || (totalBits > 64 - 7 - (LLFSELog + MLFSELog + OffFSELog))) - BIT_reloadDStream(&seqState->DStream); - - /* ANS state update */ - FSE_updateState(&seqState->stateLL, &seqState->DStream); /* <= 9 bits */ - FSE_updateState(&seqState->stateML, &seqState->DStream); /* <= 9 bits */ - if (ZSTD_32bits()) - BIT_reloadDStream(&seqState->DStream); /* <= 18 bits */ - FSE_updateState(&seqState->stateOffb, &seqState->DStream); /* <= 8 bits */ - - seq.match = NULL; - - return seq; -} - -FORCE_INLINE -size_t ZSTD_execSequence(BYTE *op, BYTE *const oend, seq_t sequence, const BYTE **litPtr, const BYTE *const litLimit, const BYTE *const base, - const BYTE *const vBase, const BYTE *const dictEnd) -{ - BYTE *const oLitEnd = op + sequence.litLength; - size_t const sequenceLength = sequence.litLength + sequence.matchLength; - BYTE *const oMatchEnd = op + sequenceLength; /* risk : address space overflow (32-bits) */ - BYTE *const oend_w = oend - WILDCOPY_OVERLENGTH; - const BYTE *const iLitEnd = *litPtr + sequence.litLength; - const BYTE *match = oLitEnd - sequence.offset; - - /* check */ - if (oMatchEnd > oend) - return ERROR(dstSize_tooSmall); /* last match must start at a minimum distance of WILDCOPY_OVERLENGTH from oend */ - if (iLitEnd > litLimit) - return ERROR(corruption_detected); /* over-read beyond lit buffer */ - if (oLitEnd > oend_w) - return ZSTD_execSequenceLast7(op, oend, sequence, litPtr, litLimit, base, vBase, dictEnd); - - /* copy Literals */ - ZSTD_copy8(op, *litPtr); - if (sequence.litLength > 8) - ZSTD_wildcopy(op + 8, (*litPtr) + 8, - sequence.litLength - 8); /* note : since oLitEnd <= oend-WILDCOPY_OVERLENGTH, no risk of overwrite beyond oend */ - op = oLitEnd; - *litPtr = iLitEnd; /* update for next sequence */ - - /* copy Match */ - if (sequence.offset > (size_t)(oLitEnd - base)) { - /* offset beyond prefix */ - if (sequence.offset > (size_t)(oLitEnd - vBase)) - return ERROR(corruption_detected); - match = dictEnd + (match - base); - if (match + sequence.matchLength <= dictEnd) { - memmove(oLitEnd, match, sequence.matchLength); - return sequenceLength; - } - /* span extDict & currPrefixSegment */ - { - size_t const length1 = dictEnd - match; - memmove(oLitEnd, match, length1); - op = oLitEnd + length1; - sequence.matchLength -= length1; - match = base; - if (op > oend_w || sequence.matchLength < MINMATCH) { - U32 i; - for (i = 0; i < sequence.matchLength; ++i) - op[i] = match[i]; - return sequenceLength; - } - } - } - /* Requirement: op <= oend_w && sequence.matchLength >= MINMATCH */ - - /* match within prefix */ - if (sequence.offset < 8) { - /* close range match, overlap */ - static const U32 dec32table[] = {0, 1, 2, 1, 4, 4, 4, 4}; /* added */ - static const int dec64table[] = {8, 8, 8, 7, 8, 9, 10, 11}; /* subtracted */ - int const sub2 = dec64table[sequence.offset]; - op[0] = match[0]; - op[1] = match[1]; - op[2] = match[2]; - op[3] = match[3]; - match += dec32table[sequence.offset]; - ZSTD_copy4(op + 4, match); - match -= sub2; - } else { - ZSTD_copy8(op, match); - } - op += 8; - match += 8; - - if (oMatchEnd > oend - (16 - MINMATCH)) { - if (op < oend_w) { - ZSTD_wildcopy(op, match, oend_w - op); - match += oend_w - op; - op = oend_w; - } - while (op < oMatchEnd) - *op++ = *match++; - } else { - ZSTD_wildcopy(op, match, (ptrdiff_t)sequence.matchLength - 8); /* works even if matchLength < 8 */ - } - return sequenceLength; -} - -static size_t ZSTD_decompressSequences(ZSTD_DCtx *dctx, void *dst, size_t maxDstSize, const void *seqStart, size_t seqSize) -{ - const BYTE *ip = (const BYTE *)seqStart; - const BYTE *const iend = ip + seqSize; - BYTE *const ostart = (BYTE * const)dst; - BYTE *const oend = ostart + maxDstSize; - BYTE *op = ostart; - const BYTE *litPtr = dctx->litPtr; - const BYTE *const litEnd = litPtr + dctx->litSize; - const BYTE *const base = (const BYTE *)(dctx->base); - const BYTE *const vBase = (const BYTE *)(dctx->vBase); - const BYTE *const dictEnd = (const BYTE *)(dctx->dictEnd); - int nbSeq; - - /* Build Decoding Tables */ - { - size_t const seqHSize = ZSTD_decodeSeqHeaders(dctx, &nbSeq, ip, seqSize); - if (ZSTD_isError(seqHSize)) - return seqHSize; - ip += seqHSize; - } - - /* Regen sequences */ - if (nbSeq) { - seqState_t seqState; - dctx->fseEntropy = 1; - { - U32 i; - for (i = 0; i < ZSTD_REP_NUM; i++) - seqState.prevOffset[i] = dctx->entropy.rep[i]; - } - CHECK_E(BIT_initDStream(&seqState.DStream, ip, iend - ip), corruption_detected); - FSE_initDState(&seqState.stateLL, &seqState.DStream, dctx->LLTptr); - FSE_initDState(&seqState.stateOffb, &seqState.DStream, dctx->OFTptr); - FSE_initDState(&seqState.stateML, &seqState.DStream, dctx->MLTptr); - - for (; (BIT_reloadDStream(&(seqState.DStream)) <= BIT_DStream_completed) && nbSeq;) { - nbSeq--; - { - seq_t const sequence = ZSTD_decodeSequence(&seqState); - size_t const oneSeqSize = ZSTD_execSequence(op, oend, sequence, &litPtr, litEnd, base, vBase, dictEnd); - if (ZSTD_isError(oneSeqSize)) - return oneSeqSize; - op += oneSeqSize; - } - } - - /* check if reached exact end */ - if (nbSeq) - return ERROR(corruption_detected); - /* save reps for next block */ - { - U32 i; - for (i = 0; i < ZSTD_REP_NUM; i++) - dctx->entropy.rep[i] = (U32)(seqState.prevOffset[i]); - } - } - - /* last literal segment */ - { - size_t const lastLLSize = litEnd - litPtr; - if (lastLLSize > (size_t)(oend - op)) - return ERROR(dstSize_tooSmall); - memcpy(op, litPtr, lastLLSize); - op += lastLLSize; - } - - return op - ostart; -} - -FORCE_INLINE seq_t ZSTD_decodeSequenceLong_generic(seqState_t *seqState, int const longOffsets) -{ - seq_t seq; - - U32 const llCode = FSE_peekSymbol(&seqState->stateLL); - U32 const mlCode = FSE_peekSymbol(&seqState->stateML); - U32 const ofCode = FSE_peekSymbol(&seqState->stateOffb); /* <= maxOff, by table construction */ - - U32 const llBits = LL_bits[llCode]; - U32 const mlBits = ML_bits[mlCode]; - U32 const ofBits = ofCode; - U32 const totalBits = llBits + mlBits + ofBits; - - static const U32 LL_base[MaxLL + 1] = {0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 18, - 20, 22, 24, 28, 32, 40, 48, 64, 0x80, 0x100, 0x200, 0x400, 0x800, 0x1000, 0x2000, 0x4000, 0x8000, 0x10000}; - - static const U32 ML_base[MaxML + 1] = {3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, - 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 37, 39, 41, - 43, 47, 51, 59, 67, 83, 99, 0x83, 0x103, 0x203, 0x403, 0x803, 0x1003, 0x2003, 0x4003, 0x8003, 0x10003}; - - static const U32 OF_base[MaxOff + 1] = {0, 1, 1, 5, 0xD, 0x1D, 0x3D, 0x7D, 0xFD, 0x1FD, - 0x3FD, 0x7FD, 0xFFD, 0x1FFD, 0x3FFD, 0x7FFD, 0xFFFD, 0x1FFFD, 0x3FFFD, 0x7FFFD, - 0xFFFFD, 0x1FFFFD, 0x3FFFFD, 0x7FFFFD, 0xFFFFFD, 0x1FFFFFD, 0x3FFFFFD, 0x7FFFFFD, 0xFFFFFFD}; - - /* sequence */ - { - size_t offset; - if (!ofCode) - offset = 0; - else { - if (longOffsets) { - int const extraBits = ofBits - MIN(ofBits, STREAM_ACCUMULATOR_MIN); - offset = OF_base[ofCode] + (BIT_readBitsFast(&seqState->DStream, ofBits - extraBits) << extraBits); - if (ZSTD_32bits() || extraBits) - BIT_reloadDStream(&seqState->DStream); - if (extraBits) - offset += BIT_readBitsFast(&seqState->DStream, extraBits); - } else { - offset = OF_base[ofCode] + BIT_readBitsFast(&seqState->DStream, ofBits); /* <= (ZSTD_WINDOWLOG_MAX-1) bits */ - if (ZSTD_32bits()) - BIT_reloadDStream(&seqState->DStream); - } - } - - if (ofCode <= 1) { - offset += (llCode == 0); - if (offset) { - size_t temp = (offset == 3) ? seqState->prevOffset[0] - 1 : seqState->prevOffset[offset]; - temp += !temp; /* 0 is not valid; input is corrupted; force offset to 1 */ - if (offset != 1) - seqState->prevOffset[2] = seqState->prevOffset[1]; - seqState->prevOffset[1] = seqState->prevOffset[0]; - seqState->prevOffset[0] = offset = temp; - } else { - offset = seqState->prevOffset[0]; - } - } else { - seqState->prevOffset[2] = seqState->prevOffset[1]; - seqState->prevOffset[1] = seqState->prevOffset[0]; - seqState->prevOffset[0] = offset; - } - seq.offset = offset; - } - - seq.matchLength = ML_base[mlCode] + ((mlCode > 31) ? BIT_readBitsFast(&seqState->DStream, mlBits) : 0); /* <= 16 bits */ - if (ZSTD_32bits() && (mlBits + llBits > 24)) - BIT_reloadDStream(&seqState->DStream); - - seq.litLength = LL_base[llCode] + ((llCode > 15) ? BIT_readBitsFast(&seqState->DStream, llBits) : 0); /* <= 16 bits */ - if (ZSTD_32bits() || (totalBits > 64 - 7 - (LLFSELog + MLFSELog + OffFSELog))) - BIT_reloadDStream(&seqState->DStream); - - { - size_t const pos = seqState->pos + seq.litLength; - seq.match = seqState->base + pos - seq.offset; /* single memory segment */ - if (seq.offset > pos) - seq.match += seqState->gotoDict; /* separate memory segment */ - seqState->pos = pos + seq.matchLength; - } - - /* ANS state update */ - FSE_updateState(&seqState->stateLL, &seqState->DStream); /* <= 9 bits */ - FSE_updateState(&seqState->stateML, &seqState->DStream); /* <= 9 bits */ - if (ZSTD_32bits()) - BIT_reloadDStream(&seqState->DStream); /* <= 18 bits */ - FSE_updateState(&seqState->stateOffb, &seqState->DStream); /* <= 8 bits */ - - return seq; -} - -static seq_t ZSTD_decodeSequenceLong(seqState_t *seqState, unsigned const windowSize) -{ - if (ZSTD_highbit32(windowSize) > STREAM_ACCUMULATOR_MIN) { - return ZSTD_decodeSequenceLong_generic(seqState, 1); - } else { - return ZSTD_decodeSequenceLong_generic(seqState, 0); - } -} - -FORCE_INLINE -size_t ZSTD_execSequenceLong(BYTE *op, BYTE *const oend, seq_t sequence, const BYTE **litPtr, const BYTE *const litLimit, const BYTE *const base, - const BYTE *const vBase, const BYTE *const dictEnd) -{ - BYTE *const oLitEnd = op + sequence.litLength; - size_t const sequenceLength = sequence.litLength + sequence.matchLength; - BYTE *const oMatchEnd = op + sequenceLength; /* risk : address space overflow (32-bits) */ - BYTE *const oend_w = oend - WILDCOPY_OVERLENGTH; - const BYTE *const iLitEnd = *litPtr + sequence.litLength; - const BYTE *match = sequence.match; - - /* check */ - if (oMatchEnd > oend) - return ERROR(dstSize_tooSmall); /* last match must start at a minimum distance of WILDCOPY_OVERLENGTH from oend */ - if (iLitEnd > litLimit) - return ERROR(corruption_detected); /* over-read beyond lit buffer */ - if (oLitEnd > oend_w) - return ZSTD_execSequenceLast7(op, oend, sequence, litPtr, litLimit, base, vBase, dictEnd); - - /* copy Literals */ - ZSTD_copy8(op, *litPtr); - if (sequence.litLength > 8) - ZSTD_wildcopy(op + 8, (*litPtr) + 8, - sequence.litLength - 8); /* note : since oLitEnd <= oend-WILDCOPY_OVERLENGTH, no risk of overwrite beyond oend */ - op = oLitEnd; - *litPtr = iLitEnd; /* update for next sequence */ - - /* copy Match */ - if (sequence.offset > (size_t)(oLitEnd - base)) { - /* offset beyond prefix */ - if (sequence.offset > (size_t)(oLitEnd - vBase)) - return ERROR(corruption_detected); - if (match + sequence.matchLength <= dictEnd) { - memmove(oLitEnd, match, sequence.matchLength); - return sequenceLength; - } - /* span extDict & currPrefixSegment */ - { - size_t const length1 = dictEnd - match; - memmove(oLitEnd, match, length1); - op = oLitEnd + length1; - sequence.matchLength -= length1; - match = base; - if (op > oend_w || sequence.matchLength < MINMATCH) { - U32 i; - for (i = 0; i < sequence.matchLength; ++i) - op[i] = match[i]; - return sequenceLength; - } - } - } - /* Requirement: op <= oend_w && sequence.matchLength >= MINMATCH */ - - /* match within prefix */ - if (sequence.offset < 8) { - /* close range match, overlap */ - static const U32 dec32table[] = {0, 1, 2, 1, 4, 4, 4, 4}; /* added */ - static const int dec64table[] = {8, 8, 8, 7, 8, 9, 10, 11}; /* subtracted */ - int const sub2 = dec64table[sequence.offset]; - op[0] = match[0]; - op[1] = match[1]; - op[2] = match[2]; - op[3] = match[3]; - match += dec32table[sequence.offset]; - ZSTD_copy4(op + 4, match); - match -= sub2; - } else { - ZSTD_copy8(op, match); - } - op += 8; - match += 8; - - if (oMatchEnd > oend - (16 - MINMATCH)) { - if (op < oend_w) { - ZSTD_wildcopy(op, match, oend_w - op); - match += oend_w - op; - op = oend_w; - } - while (op < oMatchEnd) - *op++ = *match++; - } else { - ZSTD_wildcopy(op, match, (ptrdiff_t)sequence.matchLength - 8); /* works even if matchLength < 8 */ - } - return sequenceLength; -} - -static size_t ZSTD_decompressSequencesLong(ZSTD_DCtx *dctx, void *dst, size_t maxDstSize, const void *seqStart, size_t seqSize) -{ - const BYTE *ip = (const BYTE *)seqStart; - const BYTE *const iend = ip + seqSize; - BYTE *const ostart = (BYTE * const)dst; - BYTE *const oend = ostart + maxDstSize; - BYTE *op = ostart; - const BYTE *litPtr = dctx->litPtr; - const BYTE *const litEnd = litPtr + dctx->litSize; - const BYTE *const base = (const BYTE *)(dctx->base); - const BYTE *const vBase = (const BYTE *)(dctx->vBase); - const BYTE *const dictEnd = (const BYTE *)(dctx->dictEnd); - unsigned const windowSize = dctx->fParams.windowSize; - int nbSeq; - - /* Build Decoding Tables */ - { - size_t const seqHSize = ZSTD_decodeSeqHeaders(dctx, &nbSeq, ip, seqSize); - if (ZSTD_isError(seqHSize)) - return seqHSize; - ip += seqHSize; - } - - /* Regen sequences */ - if (nbSeq) { -#define STORED_SEQS 4 -#define STOSEQ_MASK (STORED_SEQS - 1) -#define ADVANCED_SEQS 4 - seq_t *sequences = (seq_t *)dctx->entropy.workspace; - int const seqAdvance = MIN(nbSeq, ADVANCED_SEQS); - seqState_t seqState; - int seqNb; - ZSTD_STATIC_ASSERT(sizeof(dctx->entropy.workspace) >= sizeof(seq_t) * STORED_SEQS); - dctx->fseEntropy = 1; - { - U32 i; - for (i = 0; i < ZSTD_REP_NUM; i++) - seqState.prevOffset[i] = dctx->entropy.rep[i]; - } - seqState.base = base; - seqState.pos = (size_t)(op - base); - seqState.gotoDict = (uPtrDiff)dictEnd - (uPtrDiff)base; /* cast to avoid undefined behaviour */ - CHECK_E(BIT_initDStream(&seqState.DStream, ip, iend - ip), corruption_detected); - FSE_initDState(&seqState.stateLL, &seqState.DStream, dctx->LLTptr); - FSE_initDState(&seqState.stateOffb, &seqState.DStream, dctx->OFTptr); - FSE_initDState(&seqState.stateML, &seqState.DStream, dctx->MLTptr); - - /* prepare in advance */ - for (seqNb = 0; (BIT_reloadDStream(&seqState.DStream) <= BIT_DStream_completed) && seqNb < seqAdvance; seqNb++) { - sequences[seqNb] = ZSTD_decodeSequenceLong(&seqState, windowSize); - } - if (seqNb < seqAdvance) - return ERROR(corruption_detected); - - /* decode and decompress */ - for (; (BIT_reloadDStream(&(seqState.DStream)) <= BIT_DStream_completed) && seqNb < nbSeq; seqNb++) { - seq_t const sequence = ZSTD_decodeSequenceLong(&seqState, windowSize); - size_t const oneSeqSize = - ZSTD_execSequenceLong(op, oend, sequences[(seqNb - ADVANCED_SEQS) & STOSEQ_MASK], &litPtr, litEnd, base, vBase, dictEnd); - if (ZSTD_isError(oneSeqSize)) - return oneSeqSize; - ZSTD_PREFETCH(sequence.match); - sequences[seqNb & STOSEQ_MASK] = sequence; - op += oneSeqSize; - } - if (seqNb < nbSeq) - return ERROR(corruption_detected); - - /* finish queue */ - seqNb -= seqAdvance; - for (; seqNb < nbSeq; seqNb++) { - size_t const oneSeqSize = ZSTD_execSequenceLong(op, oend, sequences[seqNb & STOSEQ_MASK], &litPtr, litEnd, base, vBase, dictEnd); - if (ZSTD_isError(oneSeqSize)) - return oneSeqSize; - op += oneSeqSize; - } - - /* save reps for next block */ - { - U32 i; - for (i = 0; i < ZSTD_REP_NUM; i++) - dctx->entropy.rep[i] = (U32)(seqState.prevOffset[i]); - } - } - - /* last literal segment */ - { - size_t const lastLLSize = litEnd - litPtr; - if (lastLLSize > (size_t)(oend - op)) - return ERROR(dstSize_tooSmall); - memcpy(op, litPtr, lastLLSize); - op += lastLLSize; - } - - return op - ostart; -} - -static size_t ZSTD_decompressBlock_internal(ZSTD_DCtx *dctx, void *dst, size_t dstCapacity, const void *src, size_t srcSize) -{ /* blockType == blockCompressed */ - const BYTE *ip = (const BYTE *)src; - - if (srcSize >= ZSTD_BLOCKSIZE_ABSOLUTEMAX) - return ERROR(srcSize_wrong); - - /* Decode literals section */ - { - size_t const litCSize = ZSTD_decodeLiteralsBlock(dctx, src, srcSize); - if (ZSTD_isError(litCSize)) - return litCSize; - ip += litCSize; - srcSize -= litCSize; - } - if (sizeof(size_t) > 4) /* do not enable prefetching on 32-bits x86, as it's performance detrimental */ - /* likely because of register pressure */ - /* if that's the correct cause, then 32-bits ARM should be affected differently */ - /* it would be good to test this on ARM real hardware, to see if prefetch version improves speed */ - if (dctx->fParams.windowSize > (1 << 23)) - return ZSTD_decompressSequencesLong(dctx, dst, dstCapacity, ip, srcSize); - return ZSTD_decompressSequences(dctx, dst, dstCapacity, ip, srcSize); -} - -static void ZSTD_checkContinuity(ZSTD_DCtx *dctx, const void *dst) -{ - if (dst != dctx->previousDstEnd) { /* not contiguous */ - dctx->dictEnd = dctx->previousDstEnd; - dctx->vBase = (const char *)dst - ((const char *)(dctx->previousDstEnd) - (const char *)(dctx->base)); - dctx->base = dst; - dctx->previousDstEnd = dst; - } -} - -size_t ZSTD_decompressBlock(ZSTD_DCtx *dctx, void *dst, size_t dstCapacity, const void *src, size_t srcSize) -{ - size_t dSize; - ZSTD_checkContinuity(dctx, dst); - dSize = ZSTD_decompressBlock_internal(dctx, dst, dstCapacity, src, srcSize); - dctx->previousDstEnd = (char *)dst + dSize; - return dSize; -} - -/** ZSTD_insertBlock() : - insert `src` block into `dctx` history. Useful to track uncompressed blocks. */ -size_t ZSTD_insertBlock(ZSTD_DCtx *dctx, const void *blockStart, size_t blockSize) -{ - ZSTD_checkContinuity(dctx, blockStart); - dctx->previousDstEnd = (const char *)blockStart + blockSize; - return blockSize; -} - -size_t ZSTD_generateNxBytes(void *dst, size_t dstCapacity, BYTE byte, size_t length) -{ - if (length > dstCapacity) - return ERROR(dstSize_tooSmall); - memset(dst, byte, length); - return length; -} - -/** ZSTD_findFrameCompressedSize() : - * compatible with legacy mode - * `src` must point to the start of a ZSTD frame, ZSTD legacy frame, or skippable frame - * `srcSize` must be at least as large as the frame contained - * @return : the compressed size of the frame starting at `src` */ -size_t ZSTD_findFrameCompressedSize(const void *src, size_t srcSize) -{ - if (srcSize >= ZSTD_skippableHeaderSize && (ZSTD_readLE32(src) & 0xFFFFFFF0U) == ZSTD_MAGIC_SKIPPABLE_START) { - return ZSTD_skippableHeaderSize + ZSTD_readLE32((const BYTE *)src + 4); - } else { - const BYTE *ip = (const BYTE *)src; - const BYTE *const ipstart = ip; - size_t remainingSize = srcSize; - ZSTD_frameParams fParams; - - size_t const headerSize = ZSTD_frameHeaderSize(ip, remainingSize); - if (ZSTD_isError(headerSize)) - return headerSize; - - /* Frame Header */ - { - size_t const ret = ZSTD_getFrameParams(&fParams, ip, remainingSize); - if (ZSTD_isError(ret)) - return ret; - if (ret > 0) - return ERROR(srcSize_wrong); - } - - ip += headerSize; - remainingSize -= headerSize; - - /* Loop on each block */ - while (1) { - blockProperties_t blockProperties; - size_t const cBlockSize = ZSTD_getcBlockSize(ip, remainingSize, &blockProperties); - if (ZSTD_isError(cBlockSize)) - return cBlockSize; - - if (ZSTD_blockHeaderSize + cBlockSize > remainingSize) - return ERROR(srcSize_wrong); - - ip += ZSTD_blockHeaderSize + cBlockSize; - remainingSize -= ZSTD_blockHeaderSize + cBlockSize; - - if (blockProperties.lastBlock) - break; - } - - if (fParams.checksumFlag) { /* Frame content checksum */ - if (remainingSize < 4) - return ERROR(srcSize_wrong); - ip += 4; - remainingSize -= 4; - } - - return ip - ipstart; - } -} - -/*! ZSTD_decompressFrame() : -* @dctx must be properly initialized */ -static size_t ZSTD_decompressFrame(ZSTD_DCtx *dctx, void *dst, size_t dstCapacity, const void **srcPtr, size_t *srcSizePtr) -{ - const BYTE *ip = (const BYTE *)(*srcPtr); - BYTE *const ostart = (BYTE * const)dst; - BYTE *const oend = ostart + dstCapacity; - BYTE *op = ostart; - size_t remainingSize = *srcSizePtr; - - /* check */ - if (remainingSize < ZSTD_frameHeaderSize_min + ZSTD_blockHeaderSize) - return ERROR(srcSize_wrong); - - /* Frame Header */ - { - size_t const frameHeaderSize = ZSTD_frameHeaderSize(ip, ZSTD_frameHeaderSize_prefix); - if (ZSTD_isError(frameHeaderSize)) - return frameHeaderSize; - if (remainingSize < frameHeaderSize + ZSTD_blockHeaderSize) - return ERROR(srcSize_wrong); - CHECK_F(ZSTD_decodeFrameHeader(dctx, ip, frameHeaderSize)); - ip += frameHeaderSize; - remainingSize -= frameHeaderSize; - } - - /* Loop on each block */ - while (1) { - size_t decodedSize; - blockProperties_t blockProperties; - size_t const cBlockSize = ZSTD_getcBlockSize(ip, remainingSize, &blockProperties); - if (ZSTD_isError(cBlockSize)) - return cBlockSize; - - ip += ZSTD_blockHeaderSize; - remainingSize -= ZSTD_blockHeaderSize; - if (cBlockSize > remainingSize) - return ERROR(srcSize_wrong); - - switch (blockProperties.blockType) { - case bt_compressed: decodedSize = ZSTD_decompressBlock_internal(dctx, op, oend - op, ip, cBlockSize); break; - case bt_raw: decodedSize = ZSTD_copyRawBlock(op, oend - op, ip, cBlockSize); break; - case bt_rle: decodedSize = ZSTD_generateNxBytes(op, oend - op, *ip, blockProperties.origSize); break; - case bt_reserved: - default: return ERROR(corruption_detected); - } - - if (ZSTD_isError(decodedSize)) - return decodedSize; - if (dctx->fParams.checksumFlag) - xxh64_update(&dctx->xxhState, op, decodedSize); - op += decodedSize; - ip += cBlockSize; - remainingSize -= cBlockSize; - if (blockProperties.lastBlock) - break; - } - - if (dctx->fParams.checksumFlag) { /* Frame content checksum verification */ - U32 const checkCalc = (U32)xxh64_digest(&dctx->xxhState); - U32 checkRead; - if (remainingSize < 4) - return ERROR(checksum_wrong); - checkRead = ZSTD_readLE32(ip); - if (checkRead != checkCalc) - return ERROR(checksum_wrong); - ip += 4; - remainingSize -= 4; - } - - /* Allow caller to get size read */ - *srcPtr = ip; - *srcSizePtr = remainingSize; - return op - ostart; -} - -static const void *ZSTD_DDictDictContent(const ZSTD_DDict *ddict); -static size_t ZSTD_DDictDictSize(const ZSTD_DDict *ddict); - -static size_t ZSTD_decompressMultiFrame(ZSTD_DCtx *dctx, void *dst, size_t dstCapacity, const void *src, size_t srcSize, const void *dict, size_t dictSize, - const ZSTD_DDict *ddict) -{ - void *const dststart = dst; - - if (ddict) { - if (dict) { - /* programmer error, these two cases should be mutually exclusive */ - return ERROR(GENERIC); - } - - dict = ZSTD_DDictDictContent(ddict); - dictSize = ZSTD_DDictDictSize(ddict); - } - - while (srcSize >= ZSTD_frameHeaderSize_prefix) { - U32 magicNumber; - - magicNumber = ZSTD_readLE32(src); - if (magicNumber != ZSTD_MAGICNUMBER) { - if ((magicNumber & 0xFFFFFFF0U) == ZSTD_MAGIC_SKIPPABLE_START) { - size_t skippableSize; - if (srcSize < ZSTD_skippableHeaderSize) - return ERROR(srcSize_wrong); - skippableSize = ZSTD_readLE32((const BYTE *)src + 4) + ZSTD_skippableHeaderSize; - if (srcSize < skippableSize) { - return ERROR(srcSize_wrong); - } - - src = (const BYTE *)src + skippableSize; - srcSize -= skippableSize; - continue; - } else { - return ERROR(prefix_unknown); - } - } - - if (ddict) { - /* we were called from ZSTD_decompress_usingDDict */ - ZSTD_refDDict(dctx, ddict); - } else { - /* this will initialize correctly with no dict if dict == NULL, so - * use this in all cases but ddict */ - CHECK_F(ZSTD_decompressBegin_usingDict(dctx, dict, dictSize)); - } - ZSTD_checkContinuity(dctx, dst); - - { - const size_t res = ZSTD_decompressFrame(dctx, dst, dstCapacity, &src, &srcSize); - if (ZSTD_isError(res)) - return res; - /* don't need to bounds check this, ZSTD_decompressFrame will have - * already */ - dst = (BYTE *)dst + res; - dstCapacity -= res; - } - } - - if (srcSize) - return ERROR(srcSize_wrong); /* input not entirely consumed */ - - return (BYTE *)dst - (BYTE *)dststart; -} - -size_t ZSTD_decompress_usingDict(ZSTD_DCtx *dctx, void *dst, size_t dstCapacity, const void *src, size_t srcSize, const void *dict, size_t dictSize) -{ - return ZSTD_decompressMultiFrame(dctx, dst, dstCapacity, src, srcSize, dict, dictSize, NULL); -} - -size_t ZSTD_decompressDCtx(ZSTD_DCtx *dctx, void *dst, size_t dstCapacity, const void *src, size_t srcSize) -{ - return ZSTD_decompress_usingDict(dctx, dst, dstCapacity, src, srcSize, NULL, 0); -} - -/*-************************************** -* Advanced Streaming Decompression API -* Bufferless and synchronous -****************************************/ -size_t ZSTD_nextSrcSizeToDecompress(ZSTD_DCtx *dctx) { return dctx->expected; } - -ZSTD_nextInputType_e ZSTD_nextInputType(ZSTD_DCtx *dctx) -{ - switch (dctx->stage) { - default: /* should not happen */ - case ZSTDds_getFrameHeaderSize: - case ZSTDds_decodeFrameHeader: return ZSTDnit_frameHeader; - case ZSTDds_decodeBlockHeader: return ZSTDnit_blockHeader; - case ZSTDds_decompressBlock: return ZSTDnit_block; - case ZSTDds_decompressLastBlock: return ZSTDnit_lastBlock; - case ZSTDds_checkChecksum: return ZSTDnit_checksum; - case ZSTDds_decodeSkippableHeader: - case ZSTDds_skipFrame: return ZSTDnit_skippableFrame; - } -} - -int ZSTD_isSkipFrame(ZSTD_DCtx *dctx) { return dctx->stage == ZSTDds_skipFrame; } /* for zbuff */ - -/** ZSTD_decompressContinue() : -* @return : nb of bytes generated into `dst` (necessarily <= `dstCapacity) -* or an error code, which can be tested using ZSTD_isError() */ -size_t ZSTD_decompressContinue(ZSTD_DCtx *dctx, void *dst, size_t dstCapacity, const void *src, size_t srcSize) -{ - /* Sanity check */ - if (srcSize != dctx->expected) - return ERROR(srcSize_wrong); - if (dstCapacity) - ZSTD_checkContinuity(dctx, dst); - - switch (dctx->stage) { - case ZSTDds_getFrameHeaderSize: - if (srcSize != ZSTD_frameHeaderSize_prefix) - return ERROR(srcSize_wrong); /* impossible */ - if ((ZSTD_readLE32(src) & 0xFFFFFFF0U) == ZSTD_MAGIC_SKIPPABLE_START) { /* skippable frame */ - memcpy(dctx->headerBuffer, src, ZSTD_frameHeaderSize_prefix); - dctx->expected = ZSTD_skippableHeaderSize - ZSTD_frameHeaderSize_prefix; /* magic number + skippable frame length */ - dctx->stage = ZSTDds_decodeSkippableHeader; - return 0; - } - dctx->headerSize = ZSTD_frameHeaderSize(src, ZSTD_frameHeaderSize_prefix); - if (ZSTD_isError(dctx->headerSize)) - return dctx->headerSize; - memcpy(dctx->headerBuffer, src, ZSTD_frameHeaderSize_prefix); - if (dctx->headerSize > ZSTD_frameHeaderSize_prefix) { - dctx->expected = dctx->headerSize - ZSTD_frameHeaderSize_prefix; - dctx->stage = ZSTDds_decodeFrameHeader; - return 0; - } - dctx->expected = 0; /* not necessary to copy more */ - fallthrough; - - case ZSTDds_decodeFrameHeader: - memcpy(dctx->headerBuffer + ZSTD_frameHeaderSize_prefix, src, dctx->expected); - CHECK_F(ZSTD_decodeFrameHeader(dctx, dctx->headerBuffer, dctx->headerSize)); - dctx->expected = ZSTD_blockHeaderSize; - dctx->stage = ZSTDds_decodeBlockHeader; - return 0; - - case ZSTDds_decodeBlockHeader: { - blockProperties_t bp; - size_t const cBlockSize = ZSTD_getcBlockSize(src, ZSTD_blockHeaderSize, &bp); - if (ZSTD_isError(cBlockSize)) - return cBlockSize; - dctx->expected = cBlockSize; - dctx->bType = bp.blockType; - dctx->rleSize = bp.origSize; - if (cBlockSize) { - dctx->stage = bp.lastBlock ? ZSTDds_decompressLastBlock : ZSTDds_decompressBlock; - return 0; - } - /* empty block */ - if (bp.lastBlock) { - if (dctx->fParams.checksumFlag) { - dctx->expected = 4; - dctx->stage = ZSTDds_checkChecksum; - } else { - dctx->expected = 0; /* end of frame */ - dctx->stage = ZSTDds_getFrameHeaderSize; - } - } else { - dctx->expected = 3; /* go directly to next header */ - dctx->stage = ZSTDds_decodeBlockHeader; - } - return 0; - } - case ZSTDds_decompressLastBlock: - case ZSTDds_decompressBlock: { - size_t rSize; - switch (dctx->bType) { - case bt_compressed: rSize = ZSTD_decompressBlock_internal(dctx, dst, dstCapacity, src, srcSize); break; - case bt_raw: rSize = ZSTD_copyRawBlock(dst, dstCapacity, src, srcSize); break; - case bt_rle: rSize = ZSTD_setRleBlock(dst, dstCapacity, src, srcSize, dctx->rleSize); break; - case bt_reserved: /* should never happen */ - default: return ERROR(corruption_detected); - } - if (ZSTD_isError(rSize)) - return rSize; - if (dctx->fParams.checksumFlag) - xxh64_update(&dctx->xxhState, dst, rSize); - - if (dctx->stage == ZSTDds_decompressLastBlock) { /* end of frame */ - if (dctx->fParams.checksumFlag) { /* another round for frame checksum */ - dctx->expected = 4; - dctx->stage = ZSTDds_checkChecksum; - } else { - dctx->expected = 0; /* ends here */ - dctx->stage = ZSTDds_getFrameHeaderSize; - } - } else { - dctx->stage = ZSTDds_decodeBlockHeader; - dctx->expected = ZSTD_blockHeaderSize; - dctx->previousDstEnd = (char *)dst + rSize; - } - return rSize; - } - case ZSTDds_checkChecksum: { - U32 const h32 = (U32)xxh64_digest(&dctx->xxhState); - U32 const check32 = ZSTD_readLE32(src); /* srcSize == 4, guaranteed by dctx->expected */ - if (check32 != h32) - return ERROR(checksum_wrong); - dctx->expected = 0; - dctx->stage = ZSTDds_getFrameHeaderSize; - return 0; - } - case ZSTDds_decodeSkippableHeader: { - memcpy(dctx->headerBuffer + ZSTD_frameHeaderSize_prefix, src, dctx->expected); - dctx->expected = ZSTD_readLE32(dctx->headerBuffer + 4); - dctx->stage = ZSTDds_skipFrame; - return 0; - } - case ZSTDds_skipFrame: { - dctx->expected = 0; - dctx->stage = ZSTDds_getFrameHeaderSize; - return 0; - } - default: - return ERROR(GENERIC); /* impossible */ - } -} - -static size_t ZSTD_refDictContent(ZSTD_DCtx *dctx, const void *dict, size_t dictSize) -{ - dctx->dictEnd = dctx->previousDstEnd; - dctx->vBase = (const char *)dict - ((const char *)(dctx->previousDstEnd) - (const char *)(dctx->base)); - dctx->base = dict; - dctx->previousDstEnd = (const char *)dict + dictSize; - return 0; -} - -/* ZSTD_loadEntropy() : - * dict : must point at beginning of a valid zstd dictionary - * @return : size of entropy tables read */ -static size_t ZSTD_loadEntropy(ZSTD_entropyTables_t *entropy, const void *const dict, size_t const dictSize) -{ - const BYTE *dictPtr = (const BYTE *)dict; - const BYTE *const dictEnd = dictPtr + dictSize; - - if (dictSize <= 8) - return ERROR(dictionary_corrupted); - dictPtr += 8; /* skip header = magic + dictID */ - - { - size_t const hSize = HUF_readDTableX4_wksp(entropy->hufTable, dictPtr, dictEnd - dictPtr, entropy->workspace, sizeof(entropy->workspace)); - if (HUF_isError(hSize)) - return ERROR(dictionary_corrupted); - dictPtr += hSize; - } - - { - short offcodeNCount[MaxOff + 1]; - U32 offcodeMaxValue = MaxOff, offcodeLog; - size_t const offcodeHeaderSize = FSE_readNCount(offcodeNCount, &offcodeMaxValue, &offcodeLog, dictPtr, dictEnd - dictPtr); - if (FSE_isError(offcodeHeaderSize)) - return ERROR(dictionary_corrupted); - if (offcodeLog > OffFSELog) - return ERROR(dictionary_corrupted); - CHECK_E(FSE_buildDTable_wksp(entropy->OFTable, offcodeNCount, offcodeMaxValue, offcodeLog, entropy->workspace, sizeof(entropy->workspace)), dictionary_corrupted); - dictPtr += offcodeHeaderSize; - } - - { - short matchlengthNCount[MaxML + 1]; - unsigned matchlengthMaxValue = MaxML, matchlengthLog; - size_t const matchlengthHeaderSize = FSE_readNCount(matchlengthNCount, &matchlengthMaxValue, &matchlengthLog, dictPtr, dictEnd - dictPtr); - if (FSE_isError(matchlengthHeaderSize)) - return ERROR(dictionary_corrupted); - if (matchlengthLog > MLFSELog) - return ERROR(dictionary_corrupted); - CHECK_E(FSE_buildDTable_wksp(entropy->MLTable, matchlengthNCount, matchlengthMaxValue, matchlengthLog, entropy->workspace, sizeof(entropy->workspace)), dictionary_corrupted); - dictPtr += matchlengthHeaderSize; - } - - { - short litlengthNCount[MaxLL + 1]; - unsigned litlengthMaxValue = MaxLL, litlengthLog; - size_t const litlengthHeaderSize = FSE_readNCount(litlengthNCount, &litlengthMaxValue, &litlengthLog, dictPtr, dictEnd - dictPtr); - if (FSE_isError(litlengthHeaderSize)) - return ERROR(dictionary_corrupted); - if (litlengthLog > LLFSELog) - return ERROR(dictionary_corrupted); - CHECK_E(FSE_buildDTable_wksp(entropy->LLTable, litlengthNCount, litlengthMaxValue, litlengthLog, entropy->workspace, sizeof(entropy->workspace)), dictionary_corrupted); - dictPtr += litlengthHeaderSize; - } - - if (dictPtr + 12 > dictEnd) - return ERROR(dictionary_corrupted); - { - int i; - size_t const dictContentSize = (size_t)(dictEnd - (dictPtr + 12)); - for (i = 0; i < 3; i++) { - U32 const rep = ZSTD_readLE32(dictPtr); - dictPtr += 4; - if (rep == 0 || rep >= dictContentSize) - return ERROR(dictionary_corrupted); - entropy->rep[i] = rep; - } - } - - return dictPtr - (const BYTE *)dict; -} - -static size_t ZSTD_decompress_insertDictionary(ZSTD_DCtx *dctx, const void *dict, size_t dictSize) -{ - if (dictSize < 8) - return ZSTD_refDictContent(dctx, dict, dictSize); - { - U32 const magic = ZSTD_readLE32(dict); - if (magic != ZSTD_DICT_MAGIC) { - return ZSTD_refDictContent(dctx, dict, dictSize); /* pure content mode */ - } - } - dctx->dictID = ZSTD_readLE32((const char *)dict + 4); - - /* load entropy tables */ - { - size_t const eSize = ZSTD_loadEntropy(&dctx->entropy, dict, dictSize); - if (ZSTD_isError(eSize)) - return ERROR(dictionary_corrupted); - dict = (const char *)dict + eSize; - dictSize -= eSize; - } - dctx->litEntropy = dctx->fseEntropy = 1; - - /* reference dictionary content */ - return ZSTD_refDictContent(dctx, dict, dictSize); -} - -size_t ZSTD_decompressBegin_usingDict(ZSTD_DCtx *dctx, const void *dict, size_t dictSize) -{ - CHECK_F(ZSTD_decompressBegin(dctx)); - if (dict && dictSize) - CHECK_E(ZSTD_decompress_insertDictionary(dctx, dict, dictSize), dictionary_corrupted); - return 0; -} - -/* ====== ZSTD_DDict ====== */ - -struct ZSTD_DDict_s { - void *dictBuffer; - const void *dictContent; - size_t dictSize; - ZSTD_entropyTables_t entropy; - U32 dictID; - U32 entropyPresent; - ZSTD_customMem cMem; -}; /* typedef'd to ZSTD_DDict within "zstd.h" */ - -size_t ZSTD_DDictWorkspaceBound(void) { return ZSTD_ALIGN(sizeof(ZSTD_stack)) + ZSTD_ALIGN(sizeof(ZSTD_DDict)); } - -static const void *ZSTD_DDictDictContent(const ZSTD_DDict *ddict) { return ddict->dictContent; } - -static size_t ZSTD_DDictDictSize(const ZSTD_DDict *ddict) { return ddict->dictSize; } - -static void ZSTD_refDDict(ZSTD_DCtx *dstDCtx, const ZSTD_DDict *ddict) -{ - ZSTD_decompressBegin(dstDCtx); /* init */ - if (ddict) { /* support refDDict on NULL */ - dstDCtx->dictID = ddict->dictID; - dstDCtx->base = ddict->dictContent; - dstDCtx->vBase = ddict->dictContent; - dstDCtx->dictEnd = (const BYTE *)ddict->dictContent + ddict->dictSize; - dstDCtx->previousDstEnd = dstDCtx->dictEnd; - if (ddict->entropyPresent) { - dstDCtx->litEntropy = 1; - dstDCtx->fseEntropy = 1; - dstDCtx->LLTptr = ddict->entropy.LLTable; - dstDCtx->MLTptr = ddict->entropy.MLTable; - dstDCtx->OFTptr = ddict->entropy.OFTable; - dstDCtx->HUFptr = ddict->entropy.hufTable; - dstDCtx->entropy.rep[0] = ddict->entropy.rep[0]; - dstDCtx->entropy.rep[1] = ddict->entropy.rep[1]; - dstDCtx->entropy.rep[2] = ddict->entropy.rep[2]; - } else { - dstDCtx->litEntropy = 0; - dstDCtx->fseEntropy = 0; - } - } -} - -static size_t ZSTD_loadEntropy_inDDict(ZSTD_DDict *ddict) -{ - ddict->dictID = 0; - ddict->entropyPresent = 0; - if (ddict->dictSize < 8) - return 0; - { - U32 const magic = ZSTD_readLE32(ddict->dictContent); - if (magic != ZSTD_DICT_MAGIC) - return 0; /* pure content mode */ - } - ddict->dictID = ZSTD_readLE32((const char *)ddict->dictContent + 4); - - /* load entropy tables */ - CHECK_E(ZSTD_loadEntropy(&ddict->entropy, ddict->dictContent, ddict->dictSize), dictionary_corrupted); - ddict->entropyPresent = 1; - return 0; -} - -static ZSTD_DDict *ZSTD_createDDict_advanced(const void *dict, size_t dictSize, unsigned byReference, ZSTD_customMem customMem) -{ - if (!customMem.customAlloc || !customMem.customFree) - return NULL; - - { - ZSTD_DDict *const ddict = (ZSTD_DDict *)ZSTD_malloc(sizeof(ZSTD_DDict), customMem); - if (!ddict) - return NULL; - ddict->cMem = customMem; - - if ((byReference) || (!dict) || (!dictSize)) { - ddict->dictBuffer = NULL; - ddict->dictContent = dict; - } else { - void *const internalBuffer = ZSTD_malloc(dictSize, customMem); - if (!internalBuffer) { - ZSTD_freeDDict(ddict); - return NULL; - } - memcpy(internalBuffer, dict, dictSize); - ddict->dictBuffer = internalBuffer; - ddict->dictContent = internalBuffer; - } - ddict->dictSize = dictSize; - ddict->entropy.hufTable[0] = (HUF_DTable)((HufLog)*0x1000001); /* cover both little and big endian */ - /* parse dictionary content */ - { - size_t const errorCode = ZSTD_loadEntropy_inDDict(ddict); - if (ZSTD_isError(errorCode)) { - ZSTD_freeDDict(ddict); - return NULL; - } - } - - return ddict; - } -} - -/*! ZSTD_initDDict() : -* Create a digested dictionary, to start decompression without startup delay. -* `dict` content is copied inside DDict. -* Consequently, `dict` can be released after `ZSTD_DDict` creation */ -ZSTD_DDict *ZSTD_initDDict(const void *dict, size_t dictSize, void *workspace, size_t workspaceSize) -{ - ZSTD_customMem const stackMem = ZSTD_initStack(workspace, workspaceSize); - return ZSTD_createDDict_advanced(dict, dictSize, 1, stackMem); -} - -size_t ZSTD_freeDDict(ZSTD_DDict *ddict) -{ - if (ddict == NULL) - return 0; /* support free on NULL */ - { - ZSTD_customMem const cMem = ddict->cMem; - ZSTD_free(ddict->dictBuffer, cMem); - ZSTD_free(ddict, cMem); - return 0; - } -} - -/*! ZSTD_getDictID_fromDict() : - * Provides the dictID stored within dictionary. - * if @return == 0, the dictionary is not conformant with Zstandard specification. - * It can still be loaded, but as a content-only dictionary. */ -unsigned ZSTD_getDictID_fromDict(const void *dict, size_t dictSize) -{ - if (dictSize < 8) - return 0; - if (ZSTD_readLE32(dict) != ZSTD_DICT_MAGIC) - return 0; - return ZSTD_readLE32((const char *)dict + 4); -} - -/*! ZSTD_getDictID_fromDDict() : - * Provides the dictID of the dictionary loaded into `ddict`. - * If @return == 0, the dictionary is not conformant to Zstandard specification, or empty. - * Non-conformant dictionaries can still be loaded, but as content-only dictionaries. */ -unsigned ZSTD_getDictID_fromDDict(const ZSTD_DDict *ddict) -{ - if (ddict == NULL) - return 0; - return ZSTD_getDictID_fromDict(ddict->dictContent, ddict->dictSize); -} - -/*! ZSTD_getDictID_fromFrame() : - * Provides the dictID required to decompressed the frame stored within `src`. - * If @return == 0, the dictID could not be decoded. - * This could for one of the following reasons : - * - The frame does not require a dictionary to be decoded (most common case). - * - The frame was built with dictID intentionally removed. Whatever dictionary is necessary is a hidden information. - * Note : this use case also happens when using a non-conformant dictionary. - * - `srcSize` is too small, and as a result, the frame header could not be decoded (only possible if `srcSize < ZSTD_FRAMEHEADERSIZE_MAX`). - * - This is not a Zstandard frame. - * When identifying the exact failure cause, it's possible to used ZSTD_getFrameParams(), which will provide a more precise error code. */ -unsigned ZSTD_getDictID_fromFrame(const void *src, size_t srcSize) -{ - ZSTD_frameParams zfp = {0, 0, 0, 0}; - size_t const hError = ZSTD_getFrameParams(&zfp, src, srcSize); - if (ZSTD_isError(hError)) - return 0; - return zfp.dictID; -} - -/*! ZSTD_decompress_usingDDict() : -* Decompression using a pre-digested Dictionary -* Use dictionary without significant overhead. */ -size_t ZSTD_decompress_usingDDict(ZSTD_DCtx *dctx, void *dst, size_t dstCapacity, const void *src, size_t srcSize, const ZSTD_DDict *ddict) -{ - /* pass content and size in case legacy frames are encountered */ - return ZSTD_decompressMultiFrame(dctx, dst, dstCapacity, src, srcSize, NULL, 0, ddict); -} - -/*===================================== -* Streaming decompression -*====================================*/ - -typedef enum { zdss_init, zdss_loadHeader, zdss_read, zdss_load, zdss_flush } ZSTD_dStreamStage; - -/* *** Resource management *** */ -struct ZSTD_DStream_s { - ZSTD_DCtx *dctx; - ZSTD_DDict *ddictLocal; - const ZSTD_DDict *ddict; - ZSTD_frameParams fParams; - ZSTD_dStreamStage stage; - char *inBuff; - size_t inBuffSize; - size_t inPos; - size_t maxWindowSize; - char *outBuff; - size_t outBuffSize; - size_t outStart; - size_t outEnd; - size_t blockSize; - BYTE headerBuffer[ZSTD_FRAMEHEADERSIZE_MAX]; /* tmp buffer to store frame header */ - size_t lhSize; - ZSTD_customMem customMem; - void *legacyContext; - U32 previousLegacyVersion; - U32 legacyVersion; - U32 hostageByte; -}; /* typedef'd to ZSTD_DStream within "zstd.h" */ - -size_t ZSTD_DStreamWorkspaceBound(size_t maxWindowSize) -{ - size_t const blockSize = MIN(maxWindowSize, ZSTD_BLOCKSIZE_ABSOLUTEMAX); - size_t const inBuffSize = blockSize; - size_t const outBuffSize = maxWindowSize + blockSize + WILDCOPY_OVERLENGTH * 2; - return ZSTD_DCtxWorkspaceBound() + ZSTD_ALIGN(sizeof(ZSTD_DStream)) + ZSTD_ALIGN(inBuffSize) + ZSTD_ALIGN(outBuffSize); -} - -static ZSTD_DStream *ZSTD_createDStream_advanced(ZSTD_customMem customMem) -{ - ZSTD_DStream *zds; - - if (!customMem.customAlloc || !customMem.customFree) - return NULL; - - zds = (ZSTD_DStream *)ZSTD_malloc(sizeof(ZSTD_DStream), customMem); - if (zds == NULL) - return NULL; - memset(zds, 0, sizeof(ZSTD_DStream)); - memcpy(&zds->customMem, &customMem, sizeof(ZSTD_customMem)); - zds->dctx = ZSTD_createDCtx_advanced(customMem); - if (zds->dctx == NULL) { - ZSTD_freeDStream(zds); - return NULL; - } - zds->stage = zdss_init; - zds->maxWindowSize = ZSTD_MAXWINDOWSIZE_DEFAULT; - return zds; -} - -ZSTD_DStream *ZSTD_initDStream(size_t maxWindowSize, void *workspace, size_t workspaceSize) -{ - ZSTD_customMem const stackMem = ZSTD_initStack(workspace, workspaceSize); - ZSTD_DStream *zds = ZSTD_createDStream_advanced(stackMem); - if (!zds) { - return NULL; - } - - zds->maxWindowSize = maxWindowSize; - zds->stage = zdss_loadHeader; - zds->lhSize = zds->inPos = zds->outStart = zds->outEnd = 0; - ZSTD_freeDDict(zds->ddictLocal); - zds->ddictLocal = NULL; - zds->ddict = zds->ddictLocal; - zds->legacyVersion = 0; - zds->hostageByte = 0; - - { - size_t const blockSize = MIN(zds->maxWindowSize, ZSTD_BLOCKSIZE_ABSOLUTEMAX); - size_t const neededOutSize = zds->maxWindowSize + blockSize + WILDCOPY_OVERLENGTH * 2; - - zds->inBuff = (char *)ZSTD_malloc(blockSize, zds->customMem); - zds->inBuffSize = blockSize; - zds->outBuff = (char *)ZSTD_malloc(neededOutSize, zds->customMem); - zds->outBuffSize = neededOutSize; - if (zds->inBuff == NULL || zds->outBuff == NULL) { - ZSTD_freeDStream(zds); - return NULL; - } - } - return zds; -} - -ZSTD_DStream *ZSTD_initDStream_usingDDict(size_t maxWindowSize, const ZSTD_DDict *ddict, void *workspace, size_t workspaceSize) -{ - ZSTD_DStream *zds = ZSTD_initDStream(maxWindowSize, workspace, workspaceSize); - if (zds) { - zds->ddict = ddict; - } - return zds; -} - -size_t ZSTD_freeDStream(ZSTD_DStream *zds) -{ - if (zds == NULL) - return 0; /* support free on null */ - { - ZSTD_customMem const cMem = zds->customMem; - ZSTD_freeDCtx(zds->dctx); - zds->dctx = NULL; - ZSTD_freeDDict(zds->ddictLocal); - zds->ddictLocal = NULL; - ZSTD_free(zds->inBuff, cMem); - zds->inBuff = NULL; - ZSTD_free(zds->outBuff, cMem); - zds->outBuff = NULL; - ZSTD_free(zds, cMem); - return 0; - } -} - -/* *** Initialization *** */ - -size_t ZSTD_DStreamInSize(void) { return ZSTD_BLOCKSIZE_ABSOLUTEMAX + ZSTD_blockHeaderSize; } -size_t ZSTD_DStreamOutSize(void) { return ZSTD_BLOCKSIZE_ABSOLUTEMAX; } - -size_t ZSTD_resetDStream(ZSTD_DStream *zds) -{ - zds->stage = zdss_loadHeader; - zds->lhSize = zds->inPos = zds->outStart = zds->outEnd = 0; - zds->legacyVersion = 0; - zds->hostageByte = 0; - return ZSTD_frameHeaderSize_prefix; -} - -/* ***** Decompression ***** */ - -ZSTD_STATIC size_t ZSTD_limitCopy(void *dst, size_t dstCapacity, const void *src, size_t srcSize) -{ - size_t const length = MIN(dstCapacity, srcSize); - memcpy(dst, src, length); - return length; -} - -size_t ZSTD_decompressStream(ZSTD_DStream *zds, ZSTD_outBuffer *output, ZSTD_inBuffer *input) -{ - const char *const istart = (const char *)(input->src) + input->pos; - const char *const iend = (const char *)(input->src) + input->size; - const char *ip = istart; - char *const ostart = (char *)(output->dst) + output->pos; - char *const oend = (char *)(output->dst) + output->size; - char *op = ostart; - U32 someMoreWork = 1; - - while (someMoreWork) { - switch (zds->stage) { - case zdss_init: - ZSTD_resetDStream(zds); /* transparent reset on starting decoding a new frame */ - fallthrough; - - case zdss_loadHeader: { - size_t const hSize = ZSTD_getFrameParams(&zds->fParams, zds->headerBuffer, zds->lhSize); - if (ZSTD_isError(hSize)) - return hSize; - if (hSize != 0) { /* need more input */ - size_t const toLoad = hSize - zds->lhSize; /* if hSize!=0, hSize > zds->lhSize */ - if (toLoad > (size_t)(iend - ip)) { /* not enough input to load full header */ - memcpy(zds->headerBuffer + zds->lhSize, ip, iend - ip); - zds->lhSize += iend - ip; - input->pos = input->size; - return (MAX(ZSTD_frameHeaderSize_min, hSize) - zds->lhSize) + - ZSTD_blockHeaderSize; /* remaining header bytes + next block header */ - } - memcpy(zds->headerBuffer + zds->lhSize, ip, toLoad); - zds->lhSize = hSize; - ip += toLoad; - break; - } - - /* check for single-pass mode opportunity */ - if (zds->fParams.frameContentSize && zds->fParams.windowSize /* skippable frame if == 0 */ - && (U64)(size_t)(oend - op) >= zds->fParams.frameContentSize) { - size_t const cSize = ZSTD_findFrameCompressedSize(istart, iend - istart); - if (cSize <= (size_t)(iend - istart)) { - size_t const decompressedSize = ZSTD_decompress_usingDDict(zds->dctx, op, oend - op, istart, cSize, zds->ddict); - if (ZSTD_isError(decompressedSize)) - return decompressedSize; - ip = istart + cSize; - op += decompressedSize; - zds->dctx->expected = 0; - zds->stage = zdss_init; - someMoreWork = 0; - break; - } - } - - /* Consume header */ - ZSTD_refDDict(zds->dctx, zds->ddict); - { - size_t const h1Size = ZSTD_nextSrcSizeToDecompress(zds->dctx); /* == ZSTD_frameHeaderSize_prefix */ - CHECK_F(ZSTD_decompressContinue(zds->dctx, NULL, 0, zds->headerBuffer, h1Size)); - { - size_t const h2Size = ZSTD_nextSrcSizeToDecompress(zds->dctx); - CHECK_F(ZSTD_decompressContinue(zds->dctx, NULL, 0, zds->headerBuffer + h1Size, h2Size)); - } - } - - zds->fParams.windowSize = MAX(zds->fParams.windowSize, 1U << ZSTD_WINDOWLOG_ABSOLUTEMIN); - if (zds->fParams.windowSize > zds->maxWindowSize) - return ERROR(frameParameter_windowTooLarge); - - /* Buffers are preallocated, but double check */ - { - size_t const blockSize = MIN(zds->maxWindowSize, ZSTD_BLOCKSIZE_ABSOLUTEMAX); - size_t const neededOutSize = zds->maxWindowSize + blockSize + WILDCOPY_OVERLENGTH * 2; - if (zds->inBuffSize < blockSize) { - return ERROR(GENERIC); - } - if (zds->outBuffSize < neededOutSize) { - return ERROR(GENERIC); - } - zds->blockSize = blockSize; - } - zds->stage = zdss_read; - } - fallthrough; - - case zdss_read: { - size_t const neededInSize = ZSTD_nextSrcSizeToDecompress(zds->dctx); - if (neededInSize == 0) { /* end of frame */ - zds->stage = zdss_init; - someMoreWork = 0; - break; - } - if ((size_t)(iend - ip) >= neededInSize) { /* decode directly from src */ - const int isSkipFrame = ZSTD_isSkipFrame(zds->dctx); - size_t const decodedSize = ZSTD_decompressContinue(zds->dctx, zds->outBuff + zds->outStart, - (isSkipFrame ? 0 : zds->outBuffSize - zds->outStart), ip, neededInSize); - if (ZSTD_isError(decodedSize)) - return decodedSize; - ip += neededInSize; - if (!decodedSize && !isSkipFrame) - break; /* this was just a header */ - zds->outEnd = zds->outStart + decodedSize; - zds->stage = zdss_flush; - break; - } - if (ip == iend) { - someMoreWork = 0; - break; - } /* no more input */ - zds->stage = zdss_load; - /* pass-through */ - } - fallthrough; - - case zdss_load: { - size_t const neededInSize = ZSTD_nextSrcSizeToDecompress(zds->dctx); - size_t const toLoad = neededInSize - zds->inPos; /* should always be <= remaining space within inBuff */ - size_t loadedSize; - if (toLoad > zds->inBuffSize - zds->inPos) - return ERROR(corruption_detected); /* should never happen */ - loadedSize = ZSTD_limitCopy(zds->inBuff + zds->inPos, toLoad, ip, iend - ip); - ip += loadedSize; - zds->inPos += loadedSize; - if (loadedSize < toLoad) { - someMoreWork = 0; - break; - } /* not enough input, wait for more */ - - /* decode loaded input */ - { - const int isSkipFrame = ZSTD_isSkipFrame(zds->dctx); - size_t const decodedSize = ZSTD_decompressContinue(zds->dctx, zds->outBuff + zds->outStart, zds->outBuffSize - zds->outStart, - zds->inBuff, neededInSize); - if (ZSTD_isError(decodedSize)) - return decodedSize; - zds->inPos = 0; /* input is consumed */ - if (!decodedSize && !isSkipFrame) { - zds->stage = zdss_read; - break; - } /* this was just a header */ - zds->outEnd = zds->outStart + decodedSize; - zds->stage = zdss_flush; - /* pass-through */ - } - } - fallthrough; - - case zdss_flush: { - size_t const toFlushSize = zds->outEnd - zds->outStart; - size_t const flushedSize = ZSTD_limitCopy(op, oend - op, zds->outBuff + zds->outStart, toFlushSize); - op += flushedSize; - zds->outStart += flushedSize; - if (flushedSize == toFlushSize) { /* flush completed */ - zds->stage = zdss_read; - if (zds->outStart + zds->blockSize > zds->outBuffSize) - zds->outStart = zds->outEnd = 0; - break; - } - /* cannot complete flush */ - someMoreWork = 0; - break; - } - default: - return ERROR(GENERIC); /* impossible */ - } - } - - /* result */ - input->pos += (size_t)(ip - istart); - output->pos += (size_t)(op - ostart); - { - size_t nextSrcSizeHint = ZSTD_nextSrcSizeToDecompress(zds->dctx); - if (!nextSrcSizeHint) { /* frame fully decoded */ - if (zds->outEnd == zds->outStart) { /* output fully flushed */ - if (zds->hostageByte) { - if (input->pos >= input->size) { - zds->stage = zdss_read; - return 1; - } /* can't release hostage (not present) */ - input->pos++; /* release hostage */ - } - return 0; - } - if (!zds->hostageByte) { /* output not fully flushed; keep last byte as hostage; will be released when all output is flushed */ - input->pos--; /* note : pos > 0, otherwise, impossible to finish reading last block */ - zds->hostageByte = 1; - } - return 1; - } - nextSrcSizeHint += ZSTD_blockHeaderSize * (ZSTD_nextInputType(zds->dctx) == ZSTDnit_block); /* preload header of next block */ - if (zds->inPos > nextSrcSizeHint) - return ERROR(GENERIC); /* should never happen */ - nextSrcSizeHint -= zds->inPos; /* already loaded*/ - return nextSrcSizeHint; - } -} - -unsigned int zstd_is_error(size_t code) -{ - return ZSTD_isError(code); -} -EXPORT_SYMBOL(zstd_is_error); - -zstd_error_code zstd_get_error_code(size_t code) -{ - return ZSTD_getErrorCode(code); -} -EXPORT_SYMBOL(zstd_get_error_code); - -const char *zstd_get_error_name(size_t code) -{ - /* Real implementation in zstd-1.4.6. */ - return "GENERIC"; -} -EXPORT_SYMBOL(zstd_get_error_name); - -size_t zstd_dctx_workspace_bound(void) -{ - return ZSTD_DCtxWorkspaceBound(); -} -EXPORT_SYMBOL(zstd_dctx_workspace_bound); - -zstd_dctx *zstd_init_dctx(void *workspace, size_t workspace_size) -{ - return ZSTD_initDCtx(workspace, workspace_size); -} -EXPORT_SYMBOL(zstd_init_dctx); - -size_t zstd_decompress_dctx(zstd_dctx *dctx, void *dst, size_t dst_capacity, - const void *src, size_t src_size) -{ - return ZSTD_decompressDCtx(dctx, dst, dst_capacity, src, src_size); -} -EXPORT_SYMBOL(zstd_decompress_dctx); - -size_t zstd_dstream_workspace_bound(size_t max_window_size) -{ - return ZSTD_DStreamWorkspaceBound(max_window_size); -} -EXPORT_SYMBOL(zstd_dstream_workspace_bound); - -zstd_dstream *zstd_init_dstream(size_t max_window_size, void *workspace, - size_t workspace_size) -{ - return ZSTD_initDStream(max_window_size, workspace, workspace_size); -} -EXPORT_SYMBOL(zstd_init_dstream); - -size_t zstd_reset_dstream(zstd_dstream *dstream) -{ - return ZSTD_resetDStream(dstream); -} -EXPORT_SYMBOL(zstd_reset_dstream); - -size_t zstd_decompress_stream(zstd_dstream *dstream, zstd_out_buffer *output, - zstd_in_buffer *input) -{ - return ZSTD_decompressStream(dstream, output, input); -} -EXPORT_SYMBOL(zstd_decompress_stream); - -size_t zstd_find_frame_compressed_size(const void *src, size_t src_size) -{ - return ZSTD_findFrameCompressedSize(src, src_size); -} -EXPORT_SYMBOL(zstd_find_frame_compressed_size); - -size_t zstd_get_frame_header(zstd_frame_header *header, const void *src, - size_t src_size) -{ - return ZSTD_getFrameParams(header, src, src_size); -} -EXPORT_SYMBOL(zstd_get_frame_header); - -MODULE_LICENSE("Dual BSD/GPL"); -MODULE_DESCRIPTION("Zstd Decompressor"); diff --git a/lib/zstd/decompress/huf_decompress.c b/lib/zstd/decompress/huf_decompress.c new file mode 100644 index 000000000000..05570ed5f8be --- /dev/null +++ b/lib/zstd/decompress/huf_decompress.c @@ -0,0 +1,1206 @@ +/* ****************************************************************** + * huff0 huffman decoder, + * part of Finite State Entropy library + * Copyright (c) Yann Collet, Facebook, Inc. + * + * You can contact the author at : + * - FSE+HUF source repository : https://github.com/Cyan4973/FiniteStateEntropy + * + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). + * You may select, at your option, one of the above-listed licenses. +****************************************************************** */ + +/* ************************************************************** +* Dependencies +****************************************************************/ +#include "../common/zstd_deps.h" /* ZSTD_memcpy, ZSTD_memset */ +#include "../common/compiler.h" +#include "../common/bitstream.h" /* BIT_* */ +#include "../common/fse.h" /* to compress headers */ +#define HUF_STATIC_LINKING_ONLY +#include "../common/huf.h" +#include "../common/error_private.h" + +/* ************************************************************** +* Macros +****************************************************************/ + +/* These two optional macros force the use one way or another of the two + * Huffman decompression implementations. You can't force in both directions + * at the same time. + */ +#if defined(HUF_FORCE_DECOMPRESS_X1) && \ + defined(HUF_FORCE_DECOMPRESS_X2) +#error "Cannot force the use of the X1 and X2 decoders at the same time!" +#endif + + +/* ************************************************************** +* Error Management +****************************************************************/ +#define HUF_isError ERR_isError + + +/* ************************************************************** +* Byte alignment for workSpace management +****************************************************************/ +#define HUF_ALIGN(x, a) HUF_ALIGN_MASK((x), (a) - 1) +#define HUF_ALIGN_MASK(x, mask) (((x) + (mask)) & ~(mask)) + + +/* ************************************************************** +* BMI2 Variant Wrappers +****************************************************************/ +#if DYNAMIC_BMI2 + +#define HUF_DGEN(fn) \ + \ + static size_t fn##_default( \ + void* dst, size_t dstSize, \ + const void* cSrc, size_t cSrcSize, \ + const HUF_DTable* DTable) \ + { \ + return fn##_body(dst, dstSize, cSrc, cSrcSize, DTable); \ + } \ + \ + static TARGET_ATTRIBUTE("bmi2") size_t fn##_bmi2( \ + void* dst, size_t dstSize, \ + const void* cSrc, size_t cSrcSize, \ + const HUF_DTable* DTable) \ + { \ + return fn##_body(dst, dstSize, cSrc, cSrcSize, DTable); \ + } \ + \ + static size_t fn(void* dst, size_t dstSize, void const* cSrc, \ + size_t cSrcSize, HUF_DTable const* DTable, int bmi2) \ + { \ + if (bmi2) { \ + return fn##_bmi2(dst, dstSize, cSrc, cSrcSize, DTable); \ + } \ + return fn##_default(dst, dstSize, cSrc, cSrcSize, DTable); \ + } + +#else + +#define HUF_DGEN(fn) \ + static size_t fn(void* dst, size_t dstSize, void const* cSrc, \ + size_t cSrcSize, HUF_DTable const* DTable, int bmi2) \ + { \ + (void)bmi2; \ + return fn##_body(dst, dstSize, cSrc, cSrcSize, DTable); \ + } + +#endif + + +/*-***************************/ +/* generic DTableDesc */ +/*-***************************/ +typedef struct { BYTE maxTableLog; BYTE tableType; BYTE tableLog; BYTE reserved; } DTableDesc; + +static DTableDesc HUF_getDTableDesc(const HUF_DTable* table) +{ + DTableDesc dtd; + ZSTD_memcpy(&dtd, table, sizeof(dtd)); + return dtd; +} + + +#ifndef HUF_FORCE_DECOMPRESS_X2 + +/*-***************************/ +/* single-symbol decoding */ +/*-***************************/ +typedef struct { BYTE byte; BYTE nbBits; } HUF_DEltX1; /* single-symbol decoding */ + +/* + * Packs 4 HUF_DEltX1 structs into a U64. This is used to lay down 4 entries at + * a time. + */ +static U64 HUF_DEltX1_set4(BYTE symbol, BYTE nbBits) { + U64 D4; + if (MEM_isLittleEndian()) { + D4 = symbol + (nbBits << 8); + } else { + D4 = (symbol << 8) + nbBits; + } + D4 *= 0x0001000100010001ULL; + return D4; +} + +typedef struct { + U32 rankVal[HUF_TABLELOG_ABSOLUTEMAX + 1]; + U32 rankStart[HUF_TABLELOG_ABSOLUTEMAX + 1]; + U32 statsWksp[HUF_READ_STATS_WORKSPACE_SIZE_U32]; + BYTE symbols[HUF_SYMBOLVALUE_MAX + 1]; + BYTE huffWeight[HUF_SYMBOLVALUE_MAX + 1]; +} HUF_ReadDTableX1_Workspace; + + +size_t HUF_readDTableX1_wksp(HUF_DTable* DTable, const void* src, size_t srcSize, void* workSpace, size_t wkspSize) +{ + return HUF_readDTableX1_wksp_bmi2(DTable, src, srcSize, workSpace, wkspSize, /* bmi2 */ 0); +} + +size_t HUF_readDTableX1_wksp_bmi2(HUF_DTable* DTable, const void* src, size_t srcSize, void* workSpace, size_t wkspSize, int bmi2) +{ + U32 tableLog = 0; + U32 nbSymbols = 0; + size_t iSize; + void* const dtPtr = DTable + 1; + HUF_DEltX1* const dt = (HUF_DEltX1*)dtPtr; + HUF_ReadDTableX1_Workspace* wksp = (HUF_ReadDTableX1_Workspace*)workSpace; + + DEBUG_STATIC_ASSERT(HUF_DECOMPRESS_WORKSPACE_SIZE >= sizeof(*wksp)); + if (sizeof(*wksp) > wkspSize) return ERROR(tableLog_tooLarge); + + DEBUG_STATIC_ASSERT(sizeof(DTableDesc) == sizeof(HUF_DTable)); + /* ZSTD_memset(huffWeight, 0, sizeof(huffWeight)); */ /* is not necessary, even though some analyzer complain ... */ + + iSize = HUF_readStats_wksp(wksp->huffWeight, HUF_SYMBOLVALUE_MAX + 1, wksp->rankVal, &nbSymbols, &tableLog, src, srcSize, wksp->statsWksp, sizeof(wksp->statsWksp), bmi2); + if (HUF_isError(iSize)) return iSize; + + /* Table header */ + { DTableDesc dtd = HUF_getDTableDesc(DTable); + if (tableLog > (U32)(dtd.maxTableLog+1)) return ERROR(tableLog_tooLarge); /* DTable too small, Huffman tree cannot fit in */ + dtd.tableType = 0; + dtd.tableLog = (BYTE)tableLog; + ZSTD_memcpy(DTable, &dtd, sizeof(dtd)); + } + + /* Compute symbols and rankStart given rankVal: + * + * rankVal already contains the number of values of each weight. + * + * symbols contains the symbols ordered by weight. First are the rankVal[0] + * weight 0 symbols, followed by the rankVal[1] weight 1 symbols, and so on. + * symbols[0] is filled (but unused) to avoid a branch. + * + * rankStart contains the offset where each rank belongs in the DTable. + * rankStart[0] is not filled because there are no entries in the table for + * weight 0. + */ + { + int n; + int nextRankStart = 0; + int const unroll = 4; + int const nLimit = (int)nbSymbols - unroll + 1; + for (n=0; n<(int)tableLog+1; n++) { + U32 const curr = nextRankStart; + nextRankStart += wksp->rankVal[n]; + wksp->rankStart[n] = curr; + } + for (n=0; n < nLimit; n += unroll) { + int u; + for (u=0; u < unroll; ++u) { + size_t const w = wksp->huffWeight[n+u]; + wksp->symbols[wksp->rankStart[w]++] = (BYTE)(n+u); + } + } + for (; n < (int)nbSymbols; ++n) { + size_t const w = wksp->huffWeight[n]; + wksp->symbols[wksp->rankStart[w]++] = (BYTE)n; + } + } + + /* fill DTable + * We fill all entries of each weight in order. + * That way length is a constant for each iteration of the outter loop. + * We can switch based on the length to a different inner loop which is + * optimized for that particular case. + */ + { + U32 w; + int symbol=wksp->rankVal[0]; + int rankStart=0; + for (w=1; wrankVal[w]; + int const length = (1 << w) >> 1; + int uStart = rankStart; + BYTE const nbBits = (BYTE)(tableLog + 1 - w); + int s; + int u; + switch (length) { + case 1: + for (s=0; ssymbols[symbol + s]; + D.nbBits = nbBits; + dt[uStart] = D; + uStart += 1; + } + break; + case 2: + for (s=0; ssymbols[symbol + s]; + D.nbBits = nbBits; + dt[uStart+0] = D; + dt[uStart+1] = D; + uStart += 2; + } + break; + case 4: + for (s=0; ssymbols[symbol + s], nbBits); + MEM_write64(dt + uStart, D4); + uStart += 4; + } + break; + case 8: + for (s=0; ssymbols[symbol + s], nbBits); + MEM_write64(dt + uStart, D4); + MEM_write64(dt + uStart + 4, D4); + uStart += 8; + } + break; + default: + for (s=0; ssymbols[symbol + s], nbBits); + for (u=0; u < length; u += 16) { + MEM_write64(dt + uStart + u + 0, D4); + MEM_write64(dt + uStart + u + 4, D4); + MEM_write64(dt + uStart + u + 8, D4); + MEM_write64(dt + uStart + u + 12, D4); + } + assert(u == length); + uStart += length; + } + break; + } + symbol += symbolCount; + rankStart += symbolCount * length; + } + } + return iSize; +} + +FORCE_INLINE_TEMPLATE BYTE +HUF_decodeSymbolX1(BIT_DStream_t* Dstream, const HUF_DEltX1* dt, const U32 dtLog) +{ + size_t const val = BIT_lookBitsFast(Dstream, dtLog); /* note : dtLog >= 1 */ + BYTE const c = dt[val].byte; + BIT_skipBits(Dstream, dt[val].nbBits); + return c; +} + +#define HUF_DECODE_SYMBOLX1_0(ptr, DStreamPtr) \ + *ptr++ = HUF_decodeSymbolX1(DStreamPtr, dt, dtLog) + +#define HUF_DECODE_SYMBOLX1_1(ptr, DStreamPtr) \ + if (MEM_64bits() || (HUF_TABLELOG_MAX<=12)) \ + HUF_DECODE_SYMBOLX1_0(ptr, DStreamPtr) + +#define HUF_DECODE_SYMBOLX1_2(ptr, DStreamPtr) \ + if (MEM_64bits()) \ + HUF_DECODE_SYMBOLX1_0(ptr, DStreamPtr) + +HINT_INLINE size_t +HUF_decodeStreamX1(BYTE* p, BIT_DStream_t* const bitDPtr, BYTE* const pEnd, const HUF_DEltX1* const dt, const U32 dtLog) +{ + BYTE* const pStart = p; + + /* up to 4 symbols at a time */ + while ((BIT_reloadDStream(bitDPtr) == BIT_DStream_unfinished) & (p < pEnd-3)) { + HUF_DECODE_SYMBOLX1_2(p, bitDPtr); + HUF_DECODE_SYMBOLX1_1(p, bitDPtr); + HUF_DECODE_SYMBOLX1_2(p, bitDPtr); + HUF_DECODE_SYMBOLX1_0(p, bitDPtr); + } + + /* [0-3] symbols remaining */ + if (MEM_32bits()) + while ((BIT_reloadDStream(bitDPtr) == BIT_DStream_unfinished) & (p < pEnd)) + HUF_DECODE_SYMBOLX1_0(p, bitDPtr); + + /* no more data to retrieve from bitstream, no need to reload */ + while (p < pEnd) + HUF_DECODE_SYMBOLX1_0(p, bitDPtr); + + return pEnd-pStart; +} + +FORCE_INLINE_TEMPLATE size_t +HUF_decompress1X1_usingDTable_internal_body( + void* dst, size_t dstSize, + const void* cSrc, size_t cSrcSize, + const HUF_DTable* DTable) +{ + BYTE* op = (BYTE*)dst; + BYTE* const oend = op + dstSize; + const void* dtPtr = DTable + 1; + const HUF_DEltX1* const dt = (const HUF_DEltX1*)dtPtr; + BIT_DStream_t bitD; + DTableDesc const dtd = HUF_getDTableDesc(DTable); + U32 const dtLog = dtd.tableLog; + + CHECK_F( BIT_initDStream(&bitD, cSrc, cSrcSize) ); + + HUF_decodeStreamX1(op, &bitD, oend, dt, dtLog); + + if (!BIT_endOfDStream(&bitD)) return ERROR(corruption_detected); + + return dstSize; +} + +FORCE_INLINE_TEMPLATE size_t +HUF_decompress4X1_usingDTable_internal_body( + void* dst, size_t dstSize, + const void* cSrc, size_t cSrcSize, + const HUF_DTable* DTable) +{ + /* Check */ + if (cSrcSize < 10) return ERROR(corruption_detected); /* strict minimum : jump table + 1 byte per stream */ + + { const BYTE* const istart = (const BYTE*) cSrc; + BYTE* const ostart = (BYTE*) dst; + BYTE* const oend = ostart + dstSize; + BYTE* const olimit = oend - 3; + const void* const dtPtr = DTable + 1; + const HUF_DEltX1* const dt = (const HUF_DEltX1*)dtPtr; + + /* Init */ + BIT_DStream_t bitD1; + BIT_DStream_t bitD2; + BIT_DStream_t bitD3; + BIT_DStream_t bitD4; + size_t const length1 = MEM_readLE16(istart); + size_t const length2 = MEM_readLE16(istart+2); + size_t const length3 = MEM_readLE16(istart+4); + size_t const length4 = cSrcSize - (length1 + length2 + length3 + 6); + const BYTE* const istart1 = istart + 6; /* jumpTable */ + const BYTE* const istart2 = istart1 + length1; + const BYTE* const istart3 = istart2 + length2; + const BYTE* const istart4 = istart3 + length3; + const size_t segmentSize = (dstSize+3) / 4; + BYTE* const opStart2 = ostart + segmentSize; + BYTE* const opStart3 = opStart2 + segmentSize; + BYTE* const opStart4 = opStart3 + segmentSize; + BYTE* op1 = ostart; + BYTE* op2 = opStart2; + BYTE* op3 = opStart3; + BYTE* op4 = opStart4; + DTableDesc const dtd = HUF_getDTableDesc(DTable); + U32 const dtLog = dtd.tableLog; + U32 endSignal = 1; + + if (length4 > cSrcSize) return ERROR(corruption_detected); /* overflow */ + CHECK_F( BIT_initDStream(&bitD1, istart1, length1) ); + CHECK_F( BIT_initDStream(&bitD2, istart2, length2) ); + CHECK_F( BIT_initDStream(&bitD3, istart3, length3) ); + CHECK_F( BIT_initDStream(&bitD4, istart4, length4) ); + + /* up to 16 symbols per loop (4 symbols per stream) in 64-bit mode */ + for ( ; (endSignal) & (op4 < olimit) ; ) { + HUF_DECODE_SYMBOLX1_2(op1, &bitD1); + HUF_DECODE_SYMBOLX1_2(op2, &bitD2); + HUF_DECODE_SYMBOLX1_2(op3, &bitD3); + HUF_DECODE_SYMBOLX1_2(op4, &bitD4); + HUF_DECODE_SYMBOLX1_1(op1, &bitD1); + HUF_DECODE_SYMBOLX1_1(op2, &bitD2); + HUF_DECODE_SYMBOLX1_1(op3, &bitD3); + HUF_DECODE_SYMBOLX1_1(op4, &bitD4); + HUF_DECODE_SYMBOLX1_2(op1, &bitD1); + HUF_DECODE_SYMBOLX1_2(op2, &bitD2); + HUF_DECODE_SYMBOLX1_2(op3, &bitD3); + HUF_DECODE_SYMBOLX1_2(op4, &bitD4); + HUF_DECODE_SYMBOLX1_0(op1, &bitD1); + HUF_DECODE_SYMBOLX1_0(op2, &bitD2); + HUF_DECODE_SYMBOLX1_0(op3, &bitD3); + HUF_DECODE_SYMBOLX1_0(op4, &bitD4); + endSignal &= BIT_reloadDStreamFast(&bitD1) == BIT_DStream_unfinished; + endSignal &= BIT_reloadDStreamFast(&bitD2) == BIT_DStream_unfinished; + endSignal &= BIT_reloadDStreamFast(&bitD3) == BIT_DStream_unfinished; + endSignal &= BIT_reloadDStreamFast(&bitD4) == BIT_DStream_unfinished; + } + + /* check corruption */ + /* note : should not be necessary : op# advance in lock step, and we control op4. + * but curiously, binary generated by gcc 7.2 & 7.3 with -mbmi2 runs faster when >=1 test is present */ + if (op1 > opStart2) return ERROR(corruption_detected); + if (op2 > opStart3) return ERROR(corruption_detected); + if (op3 > opStart4) return ERROR(corruption_detected); + /* note : op4 supposed already verified within main loop */ + + /* finish bitStreams one by one */ + HUF_decodeStreamX1(op1, &bitD1, opStart2, dt, dtLog); + HUF_decodeStreamX1(op2, &bitD2, opStart3, dt, dtLog); + HUF_decodeStreamX1(op3, &bitD3, opStart4, dt, dtLog); + HUF_decodeStreamX1(op4, &bitD4, oend, dt, dtLog); + + /* check */ + { U32 const endCheck = BIT_endOfDStream(&bitD1) & BIT_endOfDStream(&bitD2) & BIT_endOfDStream(&bitD3) & BIT_endOfDStream(&bitD4); + if (!endCheck) return ERROR(corruption_detected); } + + /* decoded size */ + return dstSize; + } +} + + +typedef size_t (*HUF_decompress_usingDTable_t)(void *dst, size_t dstSize, + const void *cSrc, + size_t cSrcSize, + const HUF_DTable *DTable); + +HUF_DGEN(HUF_decompress1X1_usingDTable_internal) +HUF_DGEN(HUF_decompress4X1_usingDTable_internal) + + + +size_t HUF_decompress1X1_usingDTable( + void* dst, size_t dstSize, + const void* cSrc, size_t cSrcSize, + const HUF_DTable* DTable) +{ + DTableDesc dtd = HUF_getDTableDesc(DTable); + if (dtd.tableType != 0) return ERROR(GENERIC); + return HUF_decompress1X1_usingDTable_internal(dst, dstSize, cSrc, cSrcSize, DTable, /* bmi2 */ 0); +} + +size_t HUF_decompress1X1_DCtx_wksp(HUF_DTable* DCtx, void* dst, size_t dstSize, + const void* cSrc, size_t cSrcSize, + void* workSpace, size_t wkspSize) +{ + const BYTE* ip = (const BYTE*) cSrc; + + size_t const hSize = HUF_readDTableX1_wksp(DCtx, cSrc, cSrcSize, workSpace, wkspSize); + if (HUF_isError(hSize)) return hSize; + if (hSize >= cSrcSize) return ERROR(srcSize_wrong); + ip += hSize; cSrcSize -= hSize; + + return HUF_decompress1X1_usingDTable_internal(dst, dstSize, ip, cSrcSize, DCtx, /* bmi2 */ 0); +} + + +size_t HUF_decompress4X1_usingDTable( + void* dst, size_t dstSize, + const void* cSrc, size_t cSrcSize, + const HUF_DTable* DTable) +{ + DTableDesc dtd = HUF_getDTableDesc(DTable); + if (dtd.tableType != 0) return ERROR(GENERIC); + return HUF_decompress4X1_usingDTable_internal(dst, dstSize, cSrc, cSrcSize, DTable, /* bmi2 */ 0); +} + +static size_t HUF_decompress4X1_DCtx_wksp_bmi2(HUF_DTable* dctx, void* dst, size_t dstSize, + const void* cSrc, size_t cSrcSize, + void* workSpace, size_t wkspSize, int bmi2) +{ + const BYTE* ip = (const BYTE*) cSrc; + + size_t const hSize = HUF_readDTableX1_wksp_bmi2(dctx, cSrc, cSrcSize, workSpace, wkspSize, bmi2); + if (HUF_isError(hSize)) return hSize; + if (hSize >= cSrcSize) return ERROR(srcSize_wrong); + ip += hSize; cSrcSize -= hSize; + + return HUF_decompress4X1_usingDTable_internal(dst, dstSize, ip, cSrcSize, dctx, bmi2); +} + +size_t HUF_decompress4X1_DCtx_wksp(HUF_DTable* dctx, void* dst, size_t dstSize, + const void* cSrc, size_t cSrcSize, + void* workSpace, size_t wkspSize) +{ + return HUF_decompress4X1_DCtx_wksp_bmi2(dctx, dst, dstSize, cSrc, cSrcSize, workSpace, wkspSize, 0); +} + + +#endif /* HUF_FORCE_DECOMPRESS_X2 */ + + +#ifndef HUF_FORCE_DECOMPRESS_X1 + +/* *************************/ +/* double-symbols decoding */ +/* *************************/ + +typedef struct { U16 sequence; BYTE nbBits; BYTE length; } HUF_DEltX2; /* double-symbols decoding */ +typedef struct { BYTE symbol; BYTE weight; } sortedSymbol_t; +typedef U32 rankValCol_t[HUF_TABLELOG_MAX + 1]; +typedef rankValCol_t rankVal_t[HUF_TABLELOG_MAX]; + + +/* HUF_fillDTableX2Level2() : + * `rankValOrigin` must be a table of at least (HUF_TABLELOG_MAX + 1) U32 */ +static void HUF_fillDTableX2Level2(HUF_DEltX2* DTable, U32 sizeLog, const U32 consumed, + const U32* rankValOrigin, const int minWeight, + const sortedSymbol_t* sortedSymbols, const U32 sortedListSize, + U32 nbBitsBaseline, U16 baseSeq, U32* wksp, size_t wkspSize) +{ + HUF_DEltX2 DElt; + U32* rankVal = wksp; + + assert(wkspSize >= HUF_TABLELOG_MAX + 1); + (void)wkspSize; + /* get pre-calculated rankVal */ + ZSTD_memcpy(rankVal, rankValOrigin, sizeof(U32) * (HUF_TABLELOG_MAX + 1)); + + /* fill skipped values */ + if (minWeight>1) { + U32 i, skipSize = rankVal[minWeight]; + MEM_writeLE16(&(DElt.sequence), baseSeq); + DElt.nbBits = (BYTE)(consumed); + DElt.length = 1; + for (i = 0; i < skipSize; i++) + DTable[i] = DElt; + } + + /* fill DTable */ + { U32 s; for (s=0; s= 1 */ + + rankVal[weight] += length; + } } +} + + +static void HUF_fillDTableX2(HUF_DEltX2* DTable, const U32 targetLog, + const sortedSymbol_t* sortedList, const U32 sortedListSize, + const U32* rankStart, rankVal_t rankValOrigin, const U32 maxWeight, + const U32 nbBitsBaseline, U32* wksp, size_t wkspSize) +{ + U32* rankVal = wksp; + const int scaleLog = nbBitsBaseline - targetLog; /* note : targetLog >= srcLog, hence scaleLog <= 1 */ + const U32 minBits = nbBitsBaseline - maxWeight; + U32 s; + + assert(wkspSize >= HUF_TABLELOG_MAX + 1); + wksp += HUF_TABLELOG_MAX + 1; + wkspSize -= HUF_TABLELOG_MAX + 1; + + ZSTD_memcpy(rankVal, rankValOrigin, sizeof(U32) * (HUF_TABLELOG_MAX + 1)); + + /* fill DTable */ + for (s=0; s= minBits) { /* enough room for a second symbol */ + U32 sortedRank; + int minWeight = nbBits + scaleLog; + if (minWeight < 1) minWeight = 1; + sortedRank = rankStart[minWeight]; + HUF_fillDTableX2Level2(DTable+start, targetLog-nbBits, nbBits, + rankValOrigin[nbBits], minWeight, + sortedList+sortedRank, sortedListSize-sortedRank, + nbBitsBaseline, symbol, wksp, wkspSize); + } else { + HUF_DEltX2 DElt; + MEM_writeLE16(&(DElt.sequence), symbol); + DElt.nbBits = (BYTE)(nbBits); + DElt.length = 1; + { U32 const end = start + length; + U32 u; + for (u = start; u < end; u++) DTable[u] = DElt; + } } + rankVal[weight] += length; + } +} + +typedef struct { + rankValCol_t rankVal[HUF_TABLELOG_MAX]; + U32 rankStats[HUF_TABLELOG_MAX + 1]; + U32 rankStart0[HUF_TABLELOG_MAX + 2]; + sortedSymbol_t sortedSymbol[HUF_SYMBOLVALUE_MAX + 1]; + BYTE weightList[HUF_SYMBOLVALUE_MAX + 1]; + U32 calleeWksp[HUF_READ_STATS_WORKSPACE_SIZE_U32]; +} HUF_ReadDTableX2_Workspace; + +size_t HUF_readDTableX2_wksp(HUF_DTable* DTable, + const void* src, size_t srcSize, + void* workSpace, size_t wkspSize) +{ + U32 tableLog, maxW, sizeOfSort, nbSymbols; + DTableDesc dtd = HUF_getDTableDesc(DTable); + U32 const maxTableLog = dtd.maxTableLog; + size_t iSize; + void* dtPtr = DTable+1; /* force compiler to avoid strict-aliasing */ + HUF_DEltX2* const dt = (HUF_DEltX2*)dtPtr; + U32 *rankStart; + + HUF_ReadDTableX2_Workspace* const wksp = (HUF_ReadDTableX2_Workspace*)workSpace; + + if (sizeof(*wksp) > wkspSize) return ERROR(GENERIC); + + rankStart = wksp->rankStart0 + 1; + ZSTD_memset(wksp->rankStats, 0, sizeof(wksp->rankStats)); + ZSTD_memset(wksp->rankStart0, 0, sizeof(wksp->rankStart0)); + + DEBUG_STATIC_ASSERT(sizeof(HUF_DEltX2) == sizeof(HUF_DTable)); /* if compiler fails here, assertion is wrong */ + if (maxTableLog > HUF_TABLELOG_MAX) return ERROR(tableLog_tooLarge); + /* ZSTD_memset(weightList, 0, sizeof(weightList)); */ /* is not necessary, even though some analyzer complain ... */ + + iSize = HUF_readStats_wksp(wksp->weightList, HUF_SYMBOLVALUE_MAX + 1, wksp->rankStats, &nbSymbols, &tableLog, src, srcSize, wksp->calleeWksp, sizeof(wksp->calleeWksp), /* bmi2 */ 0); + if (HUF_isError(iSize)) return iSize; + + /* check result */ + if (tableLog > maxTableLog) return ERROR(tableLog_tooLarge); /* DTable can't fit code depth */ + + /* find maxWeight */ + for (maxW = tableLog; wksp->rankStats[maxW]==0; maxW--) {} /* necessarily finds a solution before 0 */ + + /* Get start index of each weight */ + { U32 w, nextRankStart = 0; + for (w=1; wrankStats[w]; + rankStart[w] = curr; + } + rankStart[0] = nextRankStart; /* put all 0w symbols at the end of sorted list*/ + sizeOfSort = nextRankStart; + } + + /* sort symbols by weight */ + { U32 s; + for (s=0; sweightList[s]; + U32 const r = rankStart[w]++; + wksp->sortedSymbol[r].symbol = (BYTE)s; + wksp->sortedSymbol[r].weight = (BYTE)w; + } + rankStart[0] = 0; /* forget 0w symbols; this is beginning of weight(1) */ + } + + /* Build rankVal */ + { U32* const rankVal0 = wksp->rankVal[0]; + { int const rescale = (maxTableLog-tableLog) - 1; /* tableLog <= maxTableLog */ + U32 nextRankVal = 0; + U32 w; + for (w=1; wrankStats[w] << (w+rescale); + rankVal0[w] = curr; + } } + { U32 const minBits = tableLog+1 - maxW; + U32 consumed; + for (consumed = minBits; consumed < maxTableLog - minBits + 1; consumed++) { + U32* const rankValPtr = wksp->rankVal[consumed]; + U32 w; + for (w = 1; w < maxW+1; w++) { + rankValPtr[w] = rankVal0[w] >> consumed; + } } } } + + HUF_fillDTableX2(dt, maxTableLog, + wksp->sortedSymbol, sizeOfSort, + wksp->rankStart0, wksp->rankVal, maxW, + tableLog+1, + wksp->calleeWksp, sizeof(wksp->calleeWksp) / sizeof(U32)); + + dtd.tableLog = (BYTE)maxTableLog; + dtd.tableType = 1; + ZSTD_memcpy(DTable, &dtd, sizeof(dtd)); + return iSize; +} + + +FORCE_INLINE_TEMPLATE U32 +HUF_decodeSymbolX2(void* op, BIT_DStream_t* DStream, const HUF_DEltX2* dt, const U32 dtLog) +{ + size_t const val = BIT_lookBitsFast(DStream, dtLog); /* note : dtLog >= 1 */ + ZSTD_memcpy(op, dt+val, 2); + BIT_skipBits(DStream, dt[val].nbBits); + return dt[val].length; +} + +FORCE_INLINE_TEMPLATE U32 +HUF_decodeLastSymbolX2(void* op, BIT_DStream_t* DStream, const HUF_DEltX2* dt, const U32 dtLog) +{ + size_t const val = BIT_lookBitsFast(DStream, dtLog); /* note : dtLog >= 1 */ + ZSTD_memcpy(op, dt+val, 1); + if (dt[val].length==1) BIT_skipBits(DStream, dt[val].nbBits); + else { + if (DStream->bitsConsumed < (sizeof(DStream->bitContainer)*8)) { + BIT_skipBits(DStream, dt[val].nbBits); + if (DStream->bitsConsumed > (sizeof(DStream->bitContainer)*8)) + /* ugly hack; works only because it's the last symbol. Note : can't easily extract nbBits from just this symbol */ + DStream->bitsConsumed = (sizeof(DStream->bitContainer)*8); + } } + return 1; +} + +#define HUF_DECODE_SYMBOLX2_0(ptr, DStreamPtr) \ + ptr += HUF_decodeSymbolX2(ptr, DStreamPtr, dt, dtLog) + +#define HUF_DECODE_SYMBOLX2_1(ptr, DStreamPtr) \ + if (MEM_64bits() || (HUF_TABLELOG_MAX<=12)) \ + ptr += HUF_decodeSymbolX2(ptr, DStreamPtr, dt, dtLog) + +#define HUF_DECODE_SYMBOLX2_2(ptr, DStreamPtr) \ + if (MEM_64bits()) \ + ptr += HUF_decodeSymbolX2(ptr, DStreamPtr, dt, dtLog) + +HINT_INLINE size_t +HUF_decodeStreamX2(BYTE* p, BIT_DStream_t* bitDPtr, BYTE* const pEnd, + const HUF_DEltX2* const dt, const U32 dtLog) +{ + BYTE* const pStart = p; + + /* up to 8 symbols at a time */ + while ((BIT_reloadDStream(bitDPtr) == BIT_DStream_unfinished) & (p < pEnd-(sizeof(bitDPtr->bitContainer)-1))) { + HUF_DECODE_SYMBOLX2_2(p, bitDPtr); + HUF_DECODE_SYMBOLX2_1(p, bitDPtr); + HUF_DECODE_SYMBOLX2_2(p, bitDPtr); + HUF_DECODE_SYMBOLX2_0(p, bitDPtr); + } + + /* closer to end : up to 2 symbols at a time */ + while ((BIT_reloadDStream(bitDPtr) == BIT_DStream_unfinished) & (p <= pEnd-2)) + HUF_DECODE_SYMBOLX2_0(p, bitDPtr); + + while (p <= pEnd-2) + HUF_DECODE_SYMBOLX2_0(p, bitDPtr); /* no need to reload : reached the end of DStream */ + + if (p < pEnd) + p += HUF_decodeLastSymbolX2(p, bitDPtr, dt, dtLog); + + return p-pStart; +} + +FORCE_INLINE_TEMPLATE size_t +HUF_decompress1X2_usingDTable_internal_body( + void* dst, size_t dstSize, + const void* cSrc, size_t cSrcSize, + const HUF_DTable* DTable) +{ + BIT_DStream_t bitD; + + /* Init */ + CHECK_F( BIT_initDStream(&bitD, cSrc, cSrcSize) ); + + /* decode */ + { BYTE* const ostart = (BYTE*) dst; + BYTE* const oend = ostart + dstSize; + const void* const dtPtr = DTable+1; /* force compiler to not use strict-aliasing */ + const HUF_DEltX2* const dt = (const HUF_DEltX2*)dtPtr; + DTableDesc const dtd = HUF_getDTableDesc(DTable); + HUF_decodeStreamX2(ostart, &bitD, oend, dt, dtd.tableLog); + } + + /* check */ + if (!BIT_endOfDStream(&bitD)) return ERROR(corruption_detected); + + /* decoded size */ + return dstSize; +} + +FORCE_INLINE_TEMPLATE size_t +HUF_decompress4X2_usingDTable_internal_body( + void* dst, size_t dstSize, + const void* cSrc, size_t cSrcSize, + const HUF_DTable* DTable) +{ + if (cSrcSize < 10) return ERROR(corruption_detected); /* strict minimum : jump table + 1 byte per stream */ + + { const BYTE* const istart = (const BYTE*) cSrc; + BYTE* const ostart = (BYTE*) dst; + BYTE* const oend = ostart + dstSize; + BYTE* const olimit = oend - (sizeof(size_t)-1); + const void* const dtPtr = DTable+1; + const HUF_DEltX2* const dt = (const HUF_DEltX2*)dtPtr; + + /* Init */ + BIT_DStream_t bitD1; + BIT_DStream_t bitD2; + BIT_DStream_t bitD3; + BIT_DStream_t bitD4; + size_t const length1 = MEM_readLE16(istart); + size_t const length2 = MEM_readLE16(istart+2); + size_t const length3 = MEM_readLE16(istart+4); + size_t const length4 = cSrcSize - (length1 + length2 + length3 + 6); + const BYTE* const istart1 = istart + 6; /* jumpTable */ + const BYTE* const istart2 = istart1 + length1; + const BYTE* const istart3 = istart2 + length2; + const BYTE* const istart4 = istart3 + length3; + size_t const segmentSize = (dstSize+3) / 4; + BYTE* const opStart2 = ostart + segmentSize; + BYTE* const opStart3 = opStart2 + segmentSize; + BYTE* const opStart4 = opStart3 + segmentSize; + BYTE* op1 = ostart; + BYTE* op2 = opStart2; + BYTE* op3 = opStart3; + BYTE* op4 = opStart4; + U32 endSignal = 1; + DTableDesc const dtd = HUF_getDTableDesc(DTable); + U32 const dtLog = dtd.tableLog; + + if (length4 > cSrcSize) return ERROR(corruption_detected); /* overflow */ + CHECK_F( BIT_initDStream(&bitD1, istart1, length1) ); + CHECK_F( BIT_initDStream(&bitD2, istart2, length2) ); + CHECK_F( BIT_initDStream(&bitD3, istart3, length3) ); + CHECK_F( BIT_initDStream(&bitD4, istart4, length4) ); + + /* 16-32 symbols per loop (4-8 symbols per stream) */ + for ( ; (endSignal) & (op4 < olimit); ) { +#if defined(__clang__) && (defined(__x86_64__) || defined(__i386__)) + HUF_DECODE_SYMBOLX2_2(op1, &bitD1); + HUF_DECODE_SYMBOLX2_1(op1, &bitD1); + HUF_DECODE_SYMBOLX2_2(op1, &bitD1); + HUF_DECODE_SYMBOLX2_0(op1, &bitD1); + HUF_DECODE_SYMBOLX2_2(op2, &bitD2); + HUF_DECODE_SYMBOLX2_1(op2, &bitD2); + HUF_DECODE_SYMBOLX2_2(op2, &bitD2); + HUF_DECODE_SYMBOLX2_0(op2, &bitD2); + endSignal &= BIT_reloadDStreamFast(&bitD1) == BIT_DStream_unfinished; + endSignal &= BIT_reloadDStreamFast(&bitD2) == BIT_DStream_unfinished; + HUF_DECODE_SYMBOLX2_2(op3, &bitD3); + HUF_DECODE_SYMBOLX2_1(op3, &bitD3); + HUF_DECODE_SYMBOLX2_2(op3, &bitD3); + HUF_DECODE_SYMBOLX2_0(op3, &bitD3); + HUF_DECODE_SYMBOLX2_2(op4, &bitD4); + HUF_DECODE_SYMBOLX2_1(op4, &bitD4); + HUF_DECODE_SYMBOLX2_2(op4, &bitD4); + HUF_DECODE_SYMBOLX2_0(op4, &bitD4); + endSignal &= BIT_reloadDStreamFast(&bitD3) == BIT_DStream_unfinished; + endSignal &= BIT_reloadDStreamFast(&bitD4) == BIT_DStream_unfinished; +#else + HUF_DECODE_SYMBOLX2_2(op1, &bitD1); + HUF_DECODE_SYMBOLX2_2(op2, &bitD2); + HUF_DECODE_SYMBOLX2_2(op3, &bitD3); + HUF_DECODE_SYMBOLX2_2(op4, &bitD4); + HUF_DECODE_SYMBOLX2_1(op1, &bitD1); + HUF_DECODE_SYMBOLX2_1(op2, &bitD2); + HUF_DECODE_SYMBOLX2_1(op3, &bitD3); + HUF_DECODE_SYMBOLX2_1(op4, &bitD4); + HUF_DECODE_SYMBOLX2_2(op1, &bitD1); + HUF_DECODE_SYMBOLX2_2(op2, &bitD2); + HUF_DECODE_SYMBOLX2_2(op3, &bitD3); + HUF_DECODE_SYMBOLX2_2(op4, &bitD4); + HUF_DECODE_SYMBOLX2_0(op1, &bitD1); + HUF_DECODE_SYMBOLX2_0(op2, &bitD2); + HUF_DECODE_SYMBOLX2_0(op3, &bitD3); + HUF_DECODE_SYMBOLX2_0(op4, &bitD4); + endSignal = (U32)LIKELY( + (BIT_reloadDStreamFast(&bitD1) == BIT_DStream_unfinished) + & (BIT_reloadDStreamFast(&bitD2) == BIT_DStream_unfinished) + & (BIT_reloadDStreamFast(&bitD3) == BIT_DStream_unfinished) + & (BIT_reloadDStreamFast(&bitD4) == BIT_DStream_unfinished)); +#endif + } + + /* check corruption */ + if (op1 > opStart2) return ERROR(corruption_detected); + if (op2 > opStart3) return ERROR(corruption_detected); + if (op3 > opStart4) return ERROR(corruption_detected); + /* note : op4 already verified within main loop */ + + /* finish bitStreams one by one */ + HUF_decodeStreamX2(op1, &bitD1, opStart2, dt, dtLog); + HUF_decodeStreamX2(op2, &bitD2, opStart3, dt, dtLog); + HUF_decodeStreamX2(op3, &bitD3, opStart4, dt, dtLog); + HUF_decodeStreamX2(op4, &bitD4, oend, dt, dtLog); + + /* check */ + { U32 const endCheck = BIT_endOfDStream(&bitD1) & BIT_endOfDStream(&bitD2) & BIT_endOfDStream(&bitD3) & BIT_endOfDStream(&bitD4); + if (!endCheck) return ERROR(corruption_detected); } + + /* decoded size */ + return dstSize; + } +} + +HUF_DGEN(HUF_decompress1X2_usingDTable_internal) +HUF_DGEN(HUF_decompress4X2_usingDTable_internal) + +size_t HUF_decompress1X2_usingDTable( + void* dst, size_t dstSize, + const void* cSrc, size_t cSrcSize, + const HUF_DTable* DTable) +{ + DTableDesc dtd = HUF_getDTableDesc(DTable); + if (dtd.tableType != 1) return ERROR(GENERIC); + return HUF_decompress1X2_usingDTable_internal(dst, dstSize, cSrc, cSrcSize, DTable, /* bmi2 */ 0); +} + +size_t HUF_decompress1X2_DCtx_wksp(HUF_DTable* DCtx, void* dst, size_t dstSize, + const void* cSrc, size_t cSrcSize, + void* workSpace, size_t wkspSize) +{ + const BYTE* ip = (const BYTE*) cSrc; + + size_t const hSize = HUF_readDTableX2_wksp(DCtx, cSrc, cSrcSize, + workSpace, wkspSize); + if (HUF_isError(hSize)) return hSize; + if (hSize >= cSrcSize) return ERROR(srcSize_wrong); + ip += hSize; cSrcSize -= hSize; + + return HUF_decompress1X2_usingDTable_internal(dst, dstSize, ip, cSrcSize, DCtx, /* bmi2 */ 0); +} + + +size_t HUF_decompress4X2_usingDTable( + void* dst, size_t dstSize, + const void* cSrc, size_t cSrcSize, + const HUF_DTable* DTable) +{ + DTableDesc dtd = HUF_getDTableDesc(DTable); + if (dtd.tableType != 1) return ERROR(GENERIC); + return HUF_decompress4X2_usingDTable_internal(dst, dstSize, cSrc, cSrcSize, DTable, /* bmi2 */ 0); +} + +static size_t HUF_decompress4X2_DCtx_wksp_bmi2(HUF_DTable* dctx, void* dst, size_t dstSize, + const void* cSrc, size_t cSrcSize, + void* workSpace, size_t wkspSize, int bmi2) +{ + const BYTE* ip = (const BYTE*) cSrc; + + size_t hSize = HUF_readDTableX2_wksp(dctx, cSrc, cSrcSize, + workSpace, wkspSize); + if (HUF_isError(hSize)) return hSize; + if (hSize >= cSrcSize) return ERROR(srcSize_wrong); + ip += hSize; cSrcSize -= hSize; + + return HUF_decompress4X2_usingDTable_internal(dst, dstSize, ip, cSrcSize, dctx, bmi2); +} + +size_t HUF_decompress4X2_DCtx_wksp(HUF_DTable* dctx, void* dst, size_t dstSize, + const void* cSrc, size_t cSrcSize, + void* workSpace, size_t wkspSize) +{ + return HUF_decompress4X2_DCtx_wksp_bmi2(dctx, dst, dstSize, cSrc, cSrcSize, workSpace, wkspSize, /* bmi2 */ 0); +} + + +#endif /* HUF_FORCE_DECOMPRESS_X1 */ + + +/* ***********************************/ +/* Universal decompression selectors */ +/* ***********************************/ + +size_t HUF_decompress1X_usingDTable(void* dst, size_t maxDstSize, + const void* cSrc, size_t cSrcSize, + const HUF_DTable* DTable) +{ + DTableDesc const dtd = HUF_getDTableDesc(DTable); +#if defined(HUF_FORCE_DECOMPRESS_X1) + (void)dtd; + assert(dtd.tableType == 0); + return HUF_decompress1X1_usingDTable_internal(dst, maxDstSize, cSrc, cSrcSize, DTable, /* bmi2 */ 0); +#elif defined(HUF_FORCE_DECOMPRESS_X2) + (void)dtd; + assert(dtd.tableType == 1); + return HUF_decompress1X2_usingDTable_internal(dst, maxDstSize, cSrc, cSrcSize, DTable, /* bmi2 */ 0); +#else + return dtd.tableType ? HUF_decompress1X2_usingDTable_internal(dst, maxDstSize, cSrc, cSrcSize, DTable, /* bmi2 */ 0) : + HUF_decompress1X1_usingDTable_internal(dst, maxDstSize, cSrc, cSrcSize, DTable, /* bmi2 */ 0); +#endif +} + +size_t HUF_decompress4X_usingDTable(void* dst, size_t maxDstSize, + const void* cSrc, size_t cSrcSize, + const HUF_DTable* DTable) +{ + DTableDesc const dtd = HUF_getDTableDesc(DTable); +#if defined(HUF_FORCE_DECOMPRESS_X1) + (void)dtd; + assert(dtd.tableType == 0); + return HUF_decompress4X1_usingDTable_internal(dst, maxDstSize, cSrc, cSrcSize, DTable, /* bmi2 */ 0); +#elif defined(HUF_FORCE_DECOMPRESS_X2) + (void)dtd; + assert(dtd.tableType == 1); + return HUF_decompress4X2_usingDTable_internal(dst, maxDstSize, cSrc, cSrcSize, DTable, /* bmi2 */ 0); +#else + return dtd.tableType ? HUF_decompress4X2_usingDTable_internal(dst, maxDstSize, cSrc, cSrcSize, DTable, /* bmi2 */ 0) : + HUF_decompress4X1_usingDTable_internal(dst, maxDstSize, cSrc, cSrcSize, DTable, /* bmi2 */ 0); +#endif +} + + +#if !defined(HUF_FORCE_DECOMPRESS_X1) && !defined(HUF_FORCE_DECOMPRESS_X2) +typedef struct { U32 tableTime; U32 decode256Time; } algo_time_t; +static const algo_time_t algoTime[16 /* Quantization */][3 /* single, double, quad */] = +{ + /* single, double, quad */ + {{0,0}, {1,1}, {2,2}}, /* Q==0 : impossible */ + {{0,0}, {1,1}, {2,2}}, /* Q==1 : impossible */ + {{ 38,130}, {1313, 74}, {2151, 38}}, /* Q == 2 : 12-18% */ + {{ 448,128}, {1353, 74}, {2238, 41}}, /* Q == 3 : 18-25% */ + {{ 556,128}, {1353, 74}, {2238, 47}}, /* Q == 4 : 25-32% */ + {{ 714,128}, {1418, 74}, {2436, 53}}, /* Q == 5 : 32-38% */ + {{ 883,128}, {1437, 74}, {2464, 61}}, /* Q == 6 : 38-44% */ + {{ 897,128}, {1515, 75}, {2622, 68}}, /* Q == 7 : 44-50% */ + {{ 926,128}, {1613, 75}, {2730, 75}}, /* Q == 8 : 50-56% */ + {{ 947,128}, {1729, 77}, {3359, 77}}, /* Q == 9 : 56-62% */ + {{1107,128}, {2083, 81}, {4006, 84}}, /* Q ==10 : 62-69% */ + {{1177,128}, {2379, 87}, {4785, 88}}, /* Q ==11 : 69-75% */ + {{1242,128}, {2415, 93}, {5155, 84}}, /* Q ==12 : 75-81% */ + {{1349,128}, {2644,106}, {5260,106}}, /* Q ==13 : 81-87% */ + {{1455,128}, {2422,124}, {4174,124}}, /* Q ==14 : 87-93% */ + {{ 722,128}, {1891,145}, {1936,146}}, /* Q ==15 : 93-99% */ +}; +#endif + +/* HUF_selectDecoder() : + * Tells which decoder is likely to decode faster, + * based on a set of pre-computed metrics. + * @return : 0==HUF_decompress4X1, 1==HUF_decompress4X2 . + * Assumption : 0 < dstSize <= 128 KB */ +U32 HUF_selectDecoder (size_t dstSize, size_t cSrcSize) +{ + assert(dstSize > 0); + assert(dstSize <= 128*1024); +#if defined(HUF_FORCE_DECOMPRESS_X1) + (void)dstSize; + (void)cSrcSize; + return 0; +#elif defined(HUF_FORCE_DECOMPRESS_X2) + (void)dstSize; + (void)cSrcSize; + return 1; +#else + /* decoder timing evaluation */ + { U32 const Q = (cSrcSize >= dstSize) ? 15 : (U32)(cSrcSize * 16 / dstSize); /* Q < 16 */ + U32 const D256 = (U32)(dstSize >> 8); + U32 const DTime0 = algoTime[Q][0].tableTime + (algoTime[Q][0].decode256Time * D256); + U32 DTime1 = algoTime[Q][1].tableTime + (algoTime[Q][1].decode256Time * D256); + DTime1 += DTime1 >> 3; /* advantage to algorithm using less memory, to reduce cache eviction */ + return DTime1 < DTime0; + } +#endif +} + + +size_t HUF_decompress4X_hufOnly_wksp(HUF_DTable* dctx, void* dst, + size_t dstSize, const void* cSrc, + size_t cSrcSize, void* workSpace, + size_t wkspSize) +{ + /* validation checks */ + if (dstSize == 0) return ERROR(dstSize_tooSmall); + if (cSrcSize == 0) return ERROR(corruption_detected); + + { U32 const algoNb = HUF_selectDecoder(dstSize, cSrcSize); +#if defined(HUF_FORCE_DECOMPRESS_X1) + (void)algoNb; + assert(algoNb == 0); + return HUF_decompress4X1_DCtx_wksp(dctx, dst, dstSize, cSrc, cSrcSize, workSpace, wkspSize); +#elif defined(HUF_FORCE_DECOMPRESS_X2) + (void)algoNb; + assert(algoNb == 1); + return HUF_decompress4X2_DCtx_wksp(dctx, dst, dstSize, cSrc, cSrcSize, workSpace, wkspSize); +#else + return algoNb ? HUF_decompress4X2_DCtx_wksp(dctx, dst, dstSize, cSrc, + cSrcSize, workSpace, wkspSize): + HUF_decompress4X1_DCtx_wksp(dctx, dst, dstSize, cSrc, cSrcSize, workSpace, wkspSize); +#endif + } +} + +size_t HUF_decompress1X_DCtx_wksp(HUF_DTable* dctx, void* dst, size_t dstSize, + const void* cSrc, size_t cSrcSize, + void* workSpace, size_t wkspSize) +{ + /* validation checks */ + if (dstSize == 0) return ERROR(dstSize_tooSmall); + if (cSrcSize > dstSize) return ERROR(corruption_detected); /* invalid */ + if (cSrcSize == dstSize) { ZSTD_memcpy(dst, cSrc, dstSize); return dstSize; } /* not compressed */ + if (cSrcSize == 1) { ZSTD_memset(dst, *(const BYTE*)cSrc, dstSize); return dstSize; } /* RLE */ + + { U32 const algoNb = HUF_selectDecoder(dstSize, cSrcSize); +#if defined(HUF_FORCE_DECOMPRESS_X1) + (void)algoNb; + assert(algoNb == 0); + return HUF_decompress1X1_DCtx_wksp(dctx, dst, dstSize, cSrc, + cSrcSize, workSpace, wkspSize); +#elif defined(HUF_FORCE_DECOMPRESS_X2) + (void)algoNb; + assert(algoNb == 1); + return HUF_decompress1X2_DCtx_wksp(dctx, dst, dstSize, cSrc, + cSrcSize, workSpace, wkspSize); +#else + return algoNb ? HUF_decompress1X2_DCtx_wksp(dctx, dst, dstSize, cSrc, + cSrcSize, workSpace, wkspSize): + HUF_decompress1X1_DCtx_wksp(dctx, dst, dstSize, cSrc, + cSrcSize, workSpace, wkspSize); +#endif + } +} + + +size_t HUF_decompress1X_usingDTable_bmi2(void* dst, size_t maxDstSize, const void* cSrc, size_t cSrcSize, const HUF_DTable* DTable, int bmi2) +{ + DTableDesc const dtd = HUF_getDTableDesc(DTable); +#if defined(HUF_FORCE_DECOMPRESS_X1) + (void)dtd; + assert(dtd.tableType == 0); + return HUF_decompress1X1_usingDTable_internal(dst, maxDstSize, cSrc, cSrcSize, DTable, bmi2); +#elif defined(HUF_FORCE_DECOMPRESS_X2) + (void)dtd; + assert(dtd.tableType == 1); + return HUF_decompress1X2_usingDTable_internal(dst, maxDstSize, cSrc, cSrcSize, DTable, bmi2); +#else + return dtd.tableType ? HUF_decompress1X2_usingDTable_internal(dst, maxDstSize, cSrc, cSrcSize, DTable, bmi2) : + HUF_decompress1X1_usingDTable_internal(dst, maxDstSize, cSrc, cSrcSize, DTable, bmi2); +#endif +} + +#ifndef HUF_FORCE_DECOMPRESS_X2 +size_t HUF_decompress1X1_DCtx_wksp_bmi2(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize, void* workSpace, size_t wkspSize, int bmi2) +{ + const BYTE* ip = (const BYTE*) cSrc; + + size_t const hSize = HUF_readDTableX1_wksp_bmi2(dctx, cSrc, cSrcSize, workSpace, wkspSize, bmi2); + if (HUF_isError(hSize)) return hSize; + if (hSize >= cSrcSize) return ERROR(srcSize_wrong); + ip += hSize; cSrcSize -= hSize; + + return HUF_decompress1X1_usingDTable_internal(dst, dstSize, ip, cSrcSize, dctx, bmi2); +} +#endif + +size_t HUF_decompress4X_usingDTable_bmi2(void* dst, size_t maxDstSize, const void* cSrc, size_t cSrcSize, const HUF_DTable* DTable, int bmi2) +{ + DTableDesc const dtd = HUF_getDTableDesc(DTable); +#if defined(HUF_FORCE_DECOMPRESS_X1) + (void)dtd; + assert(dtd.tableType == 0); + return HUF_decompress4X1_usingDTable_internal(dst, maxDstSize, cSrc, cSrcSize, DTable, bmi2); +#elif defined(HUF_FORCE_DECOMPRESS_X2) + (void)dtd; + assert(dtd.tableType == 1); + return HUF_decompress4X2_usingDTable_internal(dst, maxDstSize, cSrc, cSrcSize, DTable, bmi2); +#else + return dtd.tableType ? HUF_decompress4X2_usingDTable_internal(dst, maxDstSize, cSrc, cSrcSize, DTable, bmi2) : + HUF_decompress4X1_usingDTable_internal(dst, maxDstSize, cSrc, cSrcSize, DTable, bmi2); +#endif +} + +size_t HUF_decompress4X_hufOnly_wksp_bmi2(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize, void* workSpace, size_t wkspSize, int bmi2) +{ + /* validation checks */ + if (dstSize == 0) return ERROR(dstSize_tooSmall); + if (cSrcSize == 0) return ERROR(corruption_detected); + + { U32 const algoNb = HUF_selectDecoder(dstSize, cSrcSize); +#if defined(HUF_FORCE_DECOMPRESS_X1) + (void)algoNb; + assert(algoNb == 0); + return HUF_decompress4X1_DCtx_wksp_bmi2(dctx, dst, dstSize, cSrc, cSrcSize, workSpace, wkspSize, bmi2); +#elif defined(HUF_FORCE_DECOMPRESS_X2) + (void)algoNb; + assert(algoNb == 1); + return HUF_decompress4X2_DCtx_wksp_bmi2(dctx, dst, dstSize, cSrc, cSrcSize, workSpace, wkspSize, bmi2); +#else + return algoNb ? HUF_decompress4X2_DCtx_wksp_bmi2(dctx, dst, dstSize, cSrc, cSrcSize, workSpace, wkspSize, bmi2) : + HUF_decompress4X1_DCtx_wksp_bmi2(dctx, dst, dstSize, cSrc, cSrcSize, workSpace, wkspSize, bmi2); +#endif + } +} + diff --git a/lib/zstd/decompress/zstd_ddict.c b/lib/zstd/decompress/zstd_ddict.c new file mode 100644 index 000000000000..dbbc7919de53 --- /dev/null +++ b/lib/zstd/decompress/zstd_ddict.c @@ -0,0 +1,241 @@ +/* + * Copyright (c) Yann Collet, Facebook, Inc. + * All rights reserved. + * + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). + * You may select, at your option, one of the above-listed licenses. + */ + +/* zstd_ddict.c : + * concentrates all logic that needs to know the internals of ZSTD_DDict object */ + +/*-******************************************************* +* Dependencies +*********************************************************/ +#include "../common/zstd_deps.h" /* ZSTD_memcpy, ZSTD_memmove, ZSTD_memset */ +#include "../common/cpu.h" /* bmi2 */ +#include "../common/mem.h" /* low level memory routines */ +#define FSE_STATIC_LINKING_ONLY +#include "../common/fse.h" +#define HUF_STATIC_LINKING_ONLY +#include "../common/huf.h" +#include "zstd_decompress_internal.h" +#include "zstd_ddict.h" + + + + +/*-******************************************************* +* Types +*********************************************************/ +struct ZSTD_DDict_s { + void* dictBuffer; + const void* dictContent; + size_t dictSize; + ZSTD_entropyDTables_t entropy; + U32 dictID; + U32 entropyPresent; + ZSTD_customMem cMem; +}; /* typedef'd to ZSTD_DDict within "zstd.h" */ + +const void* ZSTD_DDict_dictContent(const ZSTD_DDict* ddict) +{ + assert(ddict != NULL); + return ddict->dictContent; +} + +size_t ZSTD_DDict_dictSize(const ZSTD_DDict* ddict) +{ + assert(ddict != NULL); + return ddict->dictSize; +} + +void ZSTD_copyDDictParameters(ZSTD_DCtx* dctx, const ZSTD_DDict* ddict) +{ + DEBUGLOG(4, "ZSTD_copyDDictParameters"); + assert(dctx != NULL); + assert(ddict != NULL); + dctx->dictID = ddict->dictID; + dctx->prefixStart = ddict->dictContent; + dctx->virtualStart = ddict->dictContent; + dctx->dictEnd = (const BYTE*)ddict->dictContent + ddict->dictSize; + dctx->previousDstEnd = dctx->dictEnd; +#ifdef FUZZING_BUILD_MODE_UNSAFE_FOR_PRODUCTION + dctx->dictContentBeginForFuzzing = dctx->prefixStart; + dctx->dictContentEndForFuzzing = dctx->previousDstEnd; +#endif + if (ddict->entropyPresent) { + dctx->litEntropy = 1; + dctx->fseEntropy = 1; + dctx->LLTptr = ddict->entropy.LLTable; + dctx->MLTptr = ddict->entropy.MLTable; + dctx->OFTptr = ddict->entropy.OFTable; + dctx->HUFptr = ddict->entropy.hufTable; + dctx->entropy.rep[0] = ddict->entropy.rep[0]; + dctx->entropy.rep[1] = ddict->entropy.rep[1]; + dctx->entropy.rep[2] = ddict->entropy.rep[2]; + } else { + dctx->litEntropy = 0; + dctx->fseEntropy = 0; + } +} + + +static size_t +ZSTD_loadEntropy_intoDDict(ZSTD_DDict* ddict, + ZSTD_dictContentType_e dictContentType) +{ + ddict->dictID = 0; + ddict->entropyPresent = 0; + if (dictContentType == ZSTD_dct_rawContent) return 0; + + if (ddict->dictSize < 8) { + if (dictContentType == ZSTD_dct_fullDict) + return ERROR(dictionary_corrupted); /* only accept specified dictionaries */ + return 0; /* pure content mode */ + } + { U32 const magic = MEM_readLE32(ddict->dictContent); + if (magic != ZSTD_MAGIC_DICTIONARY) { + if (dictContentType == ZSTD_dct_fullDict) + return ERROR(dictionary_corrupted); /* only accept specified dictionaries */ + return 0; /* pure content mode */ + } + } + ddict->dictID = MEM_readLE32((const char*)ddict->dictContent + ZSTD_FRAMEIDSIZE); + + /* load entropy tables */ + RETURN_ERROR_IF(ZSTD_isError(ZSTD_loadDEntropy( + &ddict->entropy, ddict->dictContent, ddict->dictSize)), + dictionary_corrupted, ""); + ddict->entropyPresent = 1; + return 0; +} + + +static size_t ZSTD_initDDict_internal(ZSTD_DDict* ddict, + const void* dict, size_t dictSize, + ZSTD_dictLoadMethod_e dictLoadMethod, + ZSTD_dictContentType_e dictContentType) +{ + if ((dictLoadMethod == ZSTD_dlm_byRef) || (!dict) || (!dictSize)) { + ddict->dictBuffer = NULL; + ddict->dictContent = dict; + if (!dict) dictSize = 0; + } else { + void* const internalBuffer = ZSTD_customMalloc(dictSize, ddict->cMem); + ddict->dictBuffer = internalBuffer; + ddict->dictContent = internalBuffer; + if (!internalBuffer) return ERROR(memory_allocation); + ZSTD_memcpy(internalBuffer, dict, dictSize); + } + ddict->dictSize = dictSize; + ddict->entropy.hufTable[0] = (HUF_DTable)((HufLog)*0x1000001); /* cover both little and big endian */ + + /* parse dictionary content */ + FORWARD_IF_ERROR( ZSTD_loadEntropy_intoDDict(ddict, dictContentType) , ""); + + return 0; +} + +ZSTD_DDict* ZSTD_createDDict_advanced(const void* dict, size_t dictSize, + ZSTD_dictLoadMethod_e dictLoadMethod, + ZSTD_dictContentType_e dictContentType, + ZSTD_customMem customMem) +{ + if ((!customMem.customAlloc) ^ (!customMem.customFree)) return NULL; + + { ZSTD_DDict* const ddict = (ZSTD_DDict*) ZSTD_customMalloc(sizeof(ZSTD_DDict), customMem); + if (ddict == NULL) return NULL; + ddict->cMem = customMem; + { size_t const initResult = ZSTD_initDDict_internal(ddict, + dict, dictSize, + dictLoadMethod, dictContentType); + if (ZSTD_isError(initResult)) { + ZSTD_freeDDict(ddict); + return NULL; + } } + return ddict; + } +} + +/*! ZSTD_createDDict() : +* Create a digested dictionary, to start decompression without startup delay. +* `dict` content is copied inside DDict. +* Consequently, `dict` can be released after `ZSTD_DDict` creation */ +ZSTD_DDict* ZSTD_createDDict(const void* dict, size_t dictSize) +{ + ZSTD_customMem const allocator = { NULL, NULL, NULL }; + return ZSTD_createDDict_advanced(dict, dictSize, ZSTD_dlm_byCopy, ZSTD_dct_auto, allocator); +} + +/*! ZSTD_createDDict_byReference() : + * Create a digested dictionary, to start decompression without startup delay. + * Dictionary content is simply referenced, it will be accessed during decompression. + * Warning : dictBuffer must outlive DDict (DDict must be freed before dictBuffer) */ +ZSTD_DDict* ZSTD_createDDict_byReference(const void* dictBuffer, size_t dictSize) +{ + ZSTD_customMem const allocator = { NULL, NULL, NULL }; + return ZSTD_createDDict_advanced(dictBuffer, dictSize, ZSTD_dlm_byRef, ZSTD_dct_auto, allocator); +} + + +const ZSTD_DDict* ZSTD_initStaticDDict( + void* sBuffer, size_t sBufferSize, + const void* dict, size_t dictSize, + ZSTD_dictLoadMethod_e dictLoadMethod, + ZSTD_dictContentType_e dictContentType) +{ + size_t const neededSpace = sizeof(ZSTD_DDict) + + (dictLoadMethod == ZSTD_dlm_byRef ? 0 : dictSize); + ZSTD_DDict* const ddict = (ZSTD_DDict*)sBuffer; + assert(sBuffer != NULL); + assert(dict != NULL); + if ((size_t)sBuffer & 7) return NULL; /* 8-aligned */ + if (sBufferSize < neededSpace) return NULL; + if (dictLoadMethod == ZSTD_dlm_byCopy) { + ZSTD_memcpy(ddict+1, dict, dictSize); /* local copy */ + dict = ddict+1; + } + if (ZSTD_isError( ZSTD_initDDict_internal(ddict, + dict, dictSize, + ZSTD_dlm_byRef, dictContentType) )) + return NULL; + return ddict; +} + + +size_t ZSTD_freeDDict(ZSTD_DDict* ddict) +{ + if (ddict==NULL) return 0; /* support free on NULL */ + { ZSTD_customMem const cMem = ddict->cMem; + ZSTD_customFree(ddict->dictBuffer, cMem); + ZSTD_customFree(ddict, cMem); + return 0; + } +} + +/*! ZSTD_estimateDDictSize() : + * Estimate amount of memory that will be needed to create a dictionary for decompression. + * Note : dictionary created by reference using ZSTD_dlm_byRef are smaller */ +size_t ZSTD_estimateDDictSize(size_t dictSize, ZSTD_dictLoadMethod_e dictLoadMethod) +{ + return sizeof(ZSTD_DDict) + (dictLoadMethod == ZSTD_dlm_byRef ? 0 : dictSize); +} + +size_t ZSTD_sizeof_DDict(const ZSTD_DDict* ddict) +{ + if (ddict==NULL) return 0; /* support sizeof on NULL */ + return sizeof(*ddict) + (ddict->dictBuffer ? ddict->dictSize : 0) ; +} + +/*! ZSTD_getDictID_fromDDict() : + * Provides the dictID of the dictionary loaded into `ddict`. + * If @return == 0, the dictionary is not conformant to Zstandard specification, or empty. + * Non-conformant dictionaries can still be loaded, but as content-only dictionaries. */ +unsigned ZSTD_getDictID_fromDDict(const ZSTD_DDict* ddict) +{ + if (ddict==NULL) return 0; + return ZSTD_getDictID_fromDict(ddict->dictContent, ddict->dictSize); +} diff --git a/lib/zstd/decompress/zstd_ddict.h b/lib/zstd/decompress/zstd_ddict.h new file mode 100644 index 000000000000..8c1a79d666f8 --- /dev/null +++ b/lib/zstd/decompress/zstd_ddict.h @@ -0,0 +1,44 @@ +/* + * Copyright (c) Yann Collet, Facebook, Inc. + * All rights reserved. + * + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). + * You may select, at your option, one of the above-listed licenses. + */ + + +#ifndef ZSTD_DDICT_H +#define ZSTD_DDICT_H + +/*-******************************************************* + * Dependencies + *********************************************************/ +#include "../common/zstd_deps.h" /* size_t */ +#include /* ZSTD_DDict, and several public functions */ + + +/*-******************************************************* + * Interface + *********************************************************/ + +/* note: several prototypes are already published in `zstd.h` : + * ZSTD_createDDict() + * ZSTD_createDDict_byReference() + * ZSTD_createDDict_advanced() + * ZSTD_freeDDict() + * ZSTD_initStaticDDict() + * ZSTD_sizeof_DDict() + * ZSTD_estimateDDictSize() + * ZSTD_getDictID_fromDict() + */ + +const void* ZSTD_DDict_dictContent(const ZSTD_DDict* ddict); +size_t ZSTD_DDict_dictSize(const ZSTD_DDict* ddict); + +void ZSTD_copyDDictParameters(ZSTD_DCtx* dctx, const ZSTD_DDict* ddict); + + + +#endif /* ZSTD_DDICT_H */ diff --git a/lib/zstd/decompress/zstd_decompress.c b/lib/zstd/decompress/zstd_decompress.c new file mode 100644 index 000000000000..b4d81d84479a --- /dev/null +++ b/lib/zstd/decompress/zstd_decompress.c @@ -0,0 +1,2085 @@ +/* + * Copyright (c) Yann Collet, Facebook, Inc. + * All rights reserved. + * + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). + * You may select, at your option, one of the above-listed licenses. + */ + + +/* *************************************************************** +* Tuning parameters +*****************************************************************/ +/*! + * HEAPMODE : + * Select how default decompression function ZSTD_decompress() allocates its context, + * on stack (0), or into heap (1, default; requires malloc()). + * Note that functions with explicit context such as ZSTD_decompressDCtx() are unaffected. + */ +#ifndef ZSTD_HEAPMODE +# define ZSTD_HEAPMODE 1 +#endif + +/*! +* LEGACY_SUPPORT : +* if set to 1+, ZSTD_decompress() can decode older formats (v0.1+) +*/ + +/*! + * MAXWINDOWSIZE_DEFAULT : + * maximum window size accepted by DStream __by default__. + * Frames requiring more memory will be rejected. + * It's possible to set a different limit using ZSTD_DCtx_setMaxWindowSize(). + */ +#ifndef ZSTD_MAXWINDOWSIZE_DEFAULT +# define ZSTD_MAXWINDOWSIZE_DEFAULT (((U32)1 << ZSTD_WINDOWLOG_LIMIT_DEFAULT) + 1) +#endif + +/*! + * NO_FORWARD_PROGRESS_MAX : + * maximum allowed nb of calls to ZSTD_decompressStream() + * without any forward progress + * (defined as: no byte read from input, and no byte flushed to output) + * before triggering an error. + */ +#ifndef ZSTD_NO_FORWARD_PROGRESS_MAX +# define ZSTD_NO_FORWARD_PROGRESS_MAX 16 +#endif + + +/*-******************************************************* +* Dependencies +*********************************************************/ +#include "../common/zstd_deps.h" /* ZSTD_memcpy, ZSTD_memmove, ZSTD_memset */ +#include "../common/cpu.h" /* bmi2 */ +#include "../common/mem.h" /* low level memory routines */ +#define FSE_STATIC_LINKING_ONLY +#include "../common/fse.h" +#define HUF_STATIC_LINKING_ONLY +#include "../common/huf.h" +#include /* xxh64_reset, xxh64_update, xxh64_digest, XXH64 */ +#include "../common/zstd_internal.h" /* blockProperties_t */ +#include "zstd_decompress_internal.h" /* ZSTD_DCtx */ +#include "zstd_ddict.h" /* ZSTD_DDictDictContent */ +#include "zstd_decompress_block.h" /* ZSTD_decompressBlock_internal */ + + + + +/* *********************************** + * Multiple DDicts Hashset internals * + *************************************/ + +#define DDICT_HASHSET_MAX_LOAD_FACTOR_COUNT_MULT 4 +#define DDICT_HASHSET_MAX_LOAD_FACTOR_SIZE_MULT 3 /* These two constants represent SIZE_MULT/COUNT_MULT load factor without using a float. + * Currently, that means a 0.75 load factor. + * So, if count * COUNT_MULT / size * SIZE_MULT != 0, then we've exceeded + * the load factor of the ddict hash set. + */ + +#define DDICT_HASHSET_TABLE_BASE_SIZE 64 +#define DDICT_HASHSET_RESIZE_FACTOR 2 + +/* Hash function to determine starting position of dict insertion within the table + * Returns an index between [0, hashSet->ddictPtrTableSize] + */ +static size_t ZSTD_DDictHashSet_getIndex(const ZSTD_DDictHashSet* hashSet, U32 dictID) { + const U64 hash = xxh64(&dictID, sizeof(U32), 0); + /* DDict ptr table size is a multiple of 2, use size - 1 as mask to get index within [0, hashSet->ddictPtrTableSize) */ + return hash & (hashSet->ddictPtrTableSize - 1); +} + +/* Adds DDict to a hashset without resizing it. + * If inserting a DDict with a dictID that already exists in the set, replaces the one in the set. + * Returns 0 if successful, or a zstd error code if something went wrong. + */ +static size_t ZSTD_DDictHashSet_emplaceDDict(ZSTD_DDictHashSet* hashSet, const ZSTD_DDict* ddict) { + const U32 dictID = ZSTD_getDictID_fromDDict(ddict); + size_t idx = ZSTD_DDictHashSet_getIndex(hashSet, dictID); + const size_t idxRangeMask = hashSet->ddictPtrTableSize - 1; + RETURN_ERROR_IF(hashSet->ddictPtrCount == hashSet->ddictPtrTableSize, GENERIC, "Hash set is full!"); + DEBUGLOG(4, "Hashed index: for dictID: %u is %zu", dictID, idx); + while (hashSet->ddictPtrTable[idx] != NULL) { + /* Replace existing ddict if inserting ddict with same dictID */ + if (ZSTD_getDictID_fromDDict(hashSet->ddictPtrTable[idx]) == dictID) { + DEBUGLOG(4, "DictID already exists, replacing rather than adding"); + hashSet->ddictPtrTable[idx] = ddict; + return 0; + } + idx &= idxRangeMask; + idx++; + } + DEBUGLOG(4, "Final idx after probing for dictID %u is: %zu", dictID, idx); + hashSet->ddictPtrTable[idx] = ddict; + hashSet->ddictPtrCount++; + return 0; +} + +/* Expands hash table by factor of DDICT_HASHSET_RESIZE_FACTOR and + * rehashes all values, allocates new table, frees old table. + * Returns 0 on success, otherwise a zstd error code. + */ +static size_t ZSTD_DDictHashSet_expand(ZSTD_DDictHashSet* hashSet, ZSTD_customMem customMem) { + size_t newTableSize = hashSet->ddictPtrTableSize * DDICT_HASHSET_RESIZE_FACTOR; + const ZSTD_DDict** newTable = (const ZSTD_DDict**)ZSTD_customCalloc(sizeof(ZSTD_DDict*) * newTableSize, customMem); + const ZSTD_DDict** oldTable = hashSet->ddictPtrTable; + size_t oldTableSize = hashSet->ddictPtrTableSize; + size_t i; + + DEBUGLOG(4, "Expanding DDict hash table! Old size: %zu new size: %zu", oldTableSize, newTableSize); + RETURN_ERROR_IF(!newTable, memory_allocation, "Expanded hashset allocation failed!"); + hashSet->ddictPtrTable = newTable; + hashSet->ddictPtrTableSize = newTableSize; + hashSet->ddictPtrCount = 0; + for (i = 0; i < oldTableSize; ++i) { + if (oldTable[i] != NULL) { + FORWARD_IF_ERROR(ZSTD_DDictHashSet_emplaceDDict(hashSet, oldTable[i]), ""); + } + } + ZSTD_customFree((void*)oldTable, customMem); + DEBUGLOG(4, "Finished re-hash"); + return 0; +} + +/* Fetches a DDict with the given dictID + * Returns the ZSTD_DDict* with the requested dictID. If it doesn't exist, then returns NULL. + */ +static const ZSTD_DDict* ZSTD_DDictHashSet_getDDict(ZSTD_DDictHashSet* hashSet, U32 dictID) { + size_t idx = ZSTD_DDictHashSet_getIndex(hashSet, dictID); + const size_t idxRangeMask = hashSet->ddictPtrTableSize - 1; + DEBUGLOG(4, "Hashed index: for dictID: %u is %zu", dictID, idx); + for (;;) { + size_t currDictID = ZSTD_getDictID_fromDDict(hashSet->ddictPtrTable[idx]); + if (currDictID == dictID || currDictID == 0) { + /* currDictID == 0 implies a NULL ddict entry */ + break; + } else { + idx &= idxRangeMask; /* Goes to start of table when we reach the end */ + idx++; + } + } + DEBUGLOG(4, "Final idx after probing for dictID %u is: %zu", dictID, idx); + return hashSet->ddictPtrTable[idx]; +} + +/* Allocates space for and returns a ddict hash set + * The hash set's ZSTD_DDict* table has all values automatically set to NULL to begin with. + * Returns NULL if allocation failed. + */ +static ZSTD_DDictHashSet* ZSTD_createDDictHashSet(ZSTD_customMem customMem) { + ZSTD_DDictHashSet* ret = (ZSTD_DDictHashSet*)ZSTD_customMalloc(sizeof(ZSTD_DDictHashSet), customMem); + DEBUGLOG(4, "Allocating new hash set"); + if (!ret) + return NULL; + ret->ddictPtrTable = (const ZSTD_DDict**)ZSTD_customCalloc(DDICT_HASHSET_TABLE_BASE_SIZE * sizeof(ZSTD_DDict*), customMem); + if (!ret->ddictPtrTable) { + ZSTD_customFree(ret, customMem); + return NULL; + } + ret->ddictPtrTableSize = DDICT_HASHSET_TABLE_BASE_SIZE; + ret->ddictPtrCount = 0; + return ret; +} + +/* Frees the table of ZSTD_DDict* within a hashset, then frees the hashset itself. + * Note: The ZSTD_DDict* within the table are NOT freed. + */ +static void ZSTD_freeDDictHashSet(ZSTD_DDictHashSet* hashSet, ZSTD_customMem customMem) { + DEBUGLOG(4, "Freeing ddict hash set"); + if (hashSet && hashSet->ddictPtrTable) { + ZSTD_customFree((void*)hashSet->ddictPtrTable, customMem); + } + if (hashSet) { + ZSTD_customFree(hashSet, customMem); + } +} + +/* Public function: Adds a DDict into the ZSTD_DDictHashSet, possibly triggering a resize of the hash set. + * Returns 0 on success, or a ZSTD error. + */ +static size_t ZSTD_DDictHashSet_addDDict(ZSTD_DDictHashSet* hashSet, const ZSTD_DDict* ddict, ZSTD_customMem customMem) { + DEBUGLOG(4, "Adding dict ID: %u to hashset with - Count: %zu Tablesize: %zu", ZSTD_getDictID_fromDDict(ddict), hashSet->ddictPtrCount, hashSet->ddictPtrTableSize); + if (hashSet->ddictPtrCount * DDICT_HASHSET_MAX_LOAD_FACTOR_COUNT_MULT / hashSet->ddictPtrTableSize * DDICT_HASHSET_MAX_LOAD_FACTOR_SIZE_MULT != 0) { + FORWARD_IF_ERROR(ZSTD_DDictHashSet_expand(hashSet, customMem), ""); + } + FORWARD_IF_ERROR(ZSTD_DDictHashSet_emplaceDDict(hashSet, ddict), ""); + return 0; +} + +/*-************************************************************* +* Context management +***************************************************************/ +size_t ZSTD_sizeof_DCtx (const ZSTD_DCtx* dctx) +{ + if (dctx==NULL) return 0; /* support sizeof NULL */ + return sizeof(*dctx) + + ZSTD_sizeof_DDict(dctx->ddictLocal) + + dctx->inBuffSize + dctx->outBuffSize; +} + +size_t ZSTD_estimateDCtxSize(void) { return sizeof(ZSTD_DCtx); } + + +static size_t ZSTD_startingInputLength(ZSTD_format_e format) +{ + size_t const startingInputLength = ZSTD_FRAMEHEADERSIZE_PREFIX(format); + /* only supports formats ZSTD_f_zstd1 and ZSTD_f_zstd1_magicless */ + assert( (format == ZSTD_f_zstd1) || (format == ZSTD_f_zstd1_magicless) ); + return startingInputLength; +} + +static void ZSTD_DCtx_resetParameters(ZSTD_DCtx* dctx) +{ + assert(dctx->streamStage == zdss_init); + dctx->format = ZSTD_f_zstd1; + dctx->maxWindowSize = ZSTD_MAXWINDOWSIZE_DEFAULT; + dctx->outBufferMode = ZSTD_bm_buffered; + dctx->forceIgnoreChecksum = ZSTD_d_validateChecksum; + dctx->refMultipleDDicts = ZSTD_rmd_refSingleDDict; +} + +static void ZSTD_initDCtx_internal(ZSTD_DCtx* dctx) +{ + dctx->staticSize = 0; + dctx->ddict = NULL; + dctx->ddictLocal = NULL; + dctx->dictEnd = NULL; + dctx->ddictIsCold = 0; + dctx->dictUses = ZSTD_dont_use; + dctx->inBuff = NULL; + dctx->inBuffSize = 0; + dctx->outBuffSize = 0; + dctx->streamStage = zdss_init; + dctx->legacyContext = NULL; + dctx->previousLegacyVersion = 0; + dctx->noForwardProgress = 0; + dctx->oversizedDuration = 0; + dctx->bmi2 = ZSTD_cpuid_bmi2(ZSTD_cpuid()); + dctx->ddictSet = NULL; + ZSTD_DCtx_resetParameters(dctx); +#ifdef FUZZING_BUILD_MODE_UNSAFE_FOR_PRODUCTION + dctx->dictContentEndForFuzzing = NULL; +#endif +} + +ZSTD_DCtx* ZSTD_initStaticDCtx(void *workspace, size_t workspaceSize) +{ + ZSTD_DCtx* const dctx = (ZSTD_DCtx*) workspace; + + if ((size_t)workspace & 7) return NULL; /* 8-aligned */ + if (workspaceSize < sizeof(ZSTD_DCtx)) return NULL; /* minimum size */ + + ZSTD_initDCtx_internal(dctx); + dctx->staticSize = workspaceSize; + dctx->inBuff = (char*)(dctx+1); + return dctx; +} + +ZSTD_DCtx* ZSTD_createDCtx_advanced(ZSTD_customMem customMem) +{ + if ((!customMem.customAlloc) ^ (!customMem.customFree)) return NULL; + + { ZSTD_DCtx* const dctx = (ZSTD_DCtx*)ZSTD_customMalloc(sizeof(*dctx), customMem); + if (!dctx) return NULL; + dctx->customMem = customMem; + ZSTD_initDCtx_internal(dctx); + return dctx; + } +} + +ZSTD_DCtx* ZSTD_createDCtx(void) +{ + DEBUGLOG(3, "ZSTD_createDCtx"); + return ZSTD_createDCtx_advanced(ZSTD_defaultCMem); +} + +static void ZSTD_clearDict(ZSTD_DCtx* dctx) +{ + ZSTD_freeDDict(dctx->ddictLocal); + dctx->ddictLocal = NULL; + dctx->ddict = NULL; + dctx->dictUses = ZSTD_dont_use; +} + +size_t ZSTD_freeDCtx(ZSTD_DCtx* dctx) +{ + if (dctx==NULL) return 0; /* support free on NULL */ + RETURN_ERROR_IF(dctx->staticSize, memory_allocation, "not compatible with static DCtx"); + { ZSTD_customMem const cMem = dctx->customMem; + ZSTD_clearDict(dctx); + ZSTD_customFree(dctx->inBuff, cMem); + dctx->inBuff = NULL; + if (dctx->ddictSet) { + ZSTD_freeDDictHashSet(dctx->ddictSet, cMem); + dctx->ddictSet = NULL; + } + ZSTD_customFree(dctx, cMem); + return 0; + } +} + +/* no longer useful */ +void ZSTD_copyDCtx(ZSTD_DCtx* dstDCtx, const ZSTD_DCtx* srcDCtx) +{ + size_t const toCopy = (size_t)((char*)(&dstDCtx->inBuff) - (char*)dstDCtx); + ZSTD_memcpy(dstDCtx, srcDCtx, toCopy); /* no need to copy workspace */ +} + +/* Given a dctx with a digested frame params, re-selects the correct ZSTD_DDict based on + * the requested dict ID from the frame. If there exists a reference to the correct ZSTD_DDict, then + * accordingly sets the ddict to be used to decompress the frame. + * + * If no DDict is found, then no action is taken, and the ZSTD_DCtx::ddict remains as-is. + * + * ZSTD_d_refMultipleDDicts must be enabled for this function to be called. + */ +static void ZSTD_DCtx_selectFrameDDict(ZSTD_DCtx* dctx) { + assert(dctx->refMultipleDDicts && dctx->ddictSet); + DEBUGLOG(4, "Adjusting DDict based on requested dict ID from frame"); + if (dctx->ddict) { + const ZSTD_DDict* frameDDict = ZSTD_DDictHashSet_getDDict(dctx->ddictSet, dctx->fParams.dictID); + if (frameDDict) { + DEBUGLOG(4, "DDict found!"); + ZSTD_clearDict(dctx); + dctx->dictID = dctx->fParams.dictID; + dctx->ddict = frameDDict; + dctx->dictUses = ZSTD_use_indefinitely; + } + } +} + + +/*-************************************************************* + * Frame header decoding + ***************************************************************/ + +/*! ZSTD_isFrame() : + * Tells if the content of `buffer` starts with a valid Frame Identifier. + * Note : Frame Identifier is 4 bytes. If `size < 4`, @return will always be 0. + * Note 2 : Legacy Frame Identifiers are considered valid only if Legacy Support is enabled. + * Note 3 : Skippable Frame Identifiers are considered valid. */ +unsigned ZSTD_isFrame(const void* buffer, size_t size) +{ + if (size < ZSTD_FRAMEIDSIZE) return 0; + { U32 const magic = MEM_readLE32(buffer); + if (magic == ZSTD_MAGICNUMBER) return 1; + if ((magic & ZSTD_MAGIC_SKIPPABLE_MASK) == ZSTD_MAGIC_SKIPPABLE_START) return 1; + } + return 0; +} + +/* ZSTD_frameHeaderSize_internal() : + * srcSize must be large enough to reach header size fields. + * note : only works for formats ZSTD_f_zstd1 and ZSTD_f_zstd1_magicless. + * @return : size of the Frame Header + * or an error code, which can be tested with ZSTD_isError() */ +static size_t ZSTD_frameHeaderSize_internal(const void* src, size_t srcSize, ZSTD_format_e format) +{ + size_t const minInputSize = ZSTD_startingInputLength(format); + RETURN_ERROR_IF(srcSize < minInputSize, srcSize_wrong, ""); + + { BYTE const fhd = ((const BYTE*)src)[minInputSize-1]; + U32 const dictID= fhd & 3; + U32 const singleSegment = (fhd >> 5) & 1; + U32 const fcsId = fhd >> 6; + return minInputSize + !singleSegment + + ZSTD_did_fieldSize[dictID] + ZSTD_fcs_fieldSize[fcsId] + + (singleSegment && !fcsId); + } +} + +/* ZSTD_frameHeaderSize() : + * srcSize must be >= ZSTD_frameHeaderSize_prefix. + * @return : size of the Frame Header, + * or an error code (if srcSize is too small) */ +size_t ZSTD_frameHeaderSize(const void* src, size_t srcSize) +{ + return ZSTD_frameHeaderSize_internal(src, srcSize, ZSTD_f_zstd1); +} + + +/* ZSTD_getFrameHeader_advanced() : + * decode Frame Header, or require larger `srcSize`. + * note : only works for formats ZSTD_f_zstd1 and ZSTD_f_zstd1_magicless + * @return : 0, `zfhPtr` is correctly filled, + * >0, `srcSize` is too small, value is wanted `srcSize` amount, + * or an error code, which can be tested using ZSTD_isError() */ +size_t ZSTD_getFrameHeader_advanced(ZSTD_frameHeader* zfhPtr, const void* src, size_t srcSize, ZSTD_format_e format) +{ + const BYTE* ip = (const BYTE*)src; + size_t const minInputSize = ZSTD_startingInputLength(format); + + ZSTD_memset(zfhPtr, 0, sizeof(*zfhPtr)); /* not strictly necessary, but static analyzer do not understand that zfhPtr is only going to be read only if return value is zero, since they are 2 different signals */ + if (srcSize < minInputSize) return minInputSize; + RETURN_ERROR_IF(src==NULL, GENERIC, "invalid parameter"); + + if ( (format != ZSTD_f_zstd1_magicless) + && (MEM_readLE32(src) != ZSTD_MAGICNUMBER) ) { + if ((MEM_readLE32(src) & ZSTD_MAGIC_SKIPPABLE_MASK) == ZSTD_MAGIC_SKIPPABLE_START) { + /* skippable frame */ + if (srcSize < ZSTD_SKIPPABLEHEADERSIZE) + return ZSTD_SKIPPABLEHEADERSIZE; /* magic number + frame length */ + ZSTD_memset(zfhPtr, 0, sizeof(*zfhPtr)); + zfhPtr->frameContentSize = MEM_readLE32((const char *)src + ZSTD_FRAMEIDSIZE); + zfhPtr->frameType = ZSTD_skippableFrame; + return 0; + } + RETURN_ERROR(prefix_unknown, ""); + } + + /* ensure there is enough `srcSize` to fully read/decode frame header */ + { size_t const fhsize = ZSTD_frameHeaderSize_internal(src, srcSize, format); + if (srcSize < fhsize) return fhsize; + zfhPtr->headerSize = (U32)fhsize; + } + + { BYTE const fhdByte = ip[minInputSize-1]; + size_t pos = minInputSize; + U32 const dictIDSizeCode = fhdByte&3; + U32 const checksumFlag = (fhdByte>>2)&1; + U32 const singleSegment = (fhdByte>>5)&1; + U32 const fcsID = fhdByte>>6; + U64 windowSize = 0; + U32 dictID = 0; + U64 frameContentSize = ZSTD_CONTENTSIZE_UNKNOWN; + RETURN_ERROR_IF((fhdByte & 0x08) != 0, frameParameter_unsupported, + "reserved bits, must be zero"); + + if (!singleSegment) { + BYTE const wlByte = ip[pos++]; + U32 const windowLog = (wlByte >> 3) + ZSTD_WINDOWLOG_ABSOLUTEMIN; + RETURN_ERROR_IF(windowLog > ZSTD_WINDOWLOG_MAX, frameParameter_windowTooLarge, ""); + windowSize = (1ULL << windowLog); + windowSize += (windowSize >> 3) * (wlByte&7); + } + switch(dictIDSizeCode) + { + default: + assert(0); /* impossible */ + ZSTD_FALLTHROUGH; + case 0 : break; + case 1 : dictID = ip[pos]; pos++; break; + case 2 : dictID = MEM_readLE16(ip+pos); pos+=2; break; + case 3 : dictID = MEM_readLE32(ip+pos); pos+=4; break; + } + switch(fcsID) + { + default: + assert(0); /* impossible */ + ZSTD_FALLTHROUGH; + case 0 : if (singleSegment) frameContentSize = ip[pos]; break; + case 1 : frameContentSize = MEM_readLE16(ip+pos)+256; break; + case 2 : frameContentSize = MEM_readLE32(ip+pos); break; + case 3 : frameContentSize = MEM_readLE64(ip+pos); break; + } + if (singleSegment) windowSize = frameContentSize; + + zfhPtr->frameType = ZSTD_frame; + zfhPtr->frameContentSize = frameContentSize; + zfhPtr->windowSize = windowSize; + zfhPtr->blockSizeMax = (unsigned) MIN(windowSize, ZSTD_BLOCKSIZE_MAX); + zfhPtr->dictID = dictID; + zfhPtr->checksumFlag = checksumFlag; + } + return 0; +} + +/* ZSTD_getFrameHeader() : + * decode Frame Header, or require larger `srcSize`. + * note : this function does not consume input, it only reads it. + * @return : 0, `zfhPtr` is correctly filled, + * >0, `srcSize` is too small, value is wanted `srcSize` amount, + * or an error code, which can be tested using ZSTD_isError() */ +size_t ZSTD_getFrameHeader(ZSTD_frameHeader* zfhPtr, const void* src, size_t srcSize) +{ + return ZSTD_getFrameHeader_advanced(zfhPtr, src, srcSize, ZSTD_f_zstd1); +} + + +/* ZSTD_getFrameContentSize() : + * compatible with legacy mode + * @return : decompressed size of the single frame pointed to be `src` if known, otherwise + * - ZSTD_CONTENTSIZE_UNKNOWN if the size cannot be determined + * - ZSTD_CONTENTSIZE_ERROR if an error occurred (e.g. invalid magic number, srcSize too small) */ +unsigned long long ZSTD_getFrameContentSize(const void *src, size_t srcSize) +{ + { ZSTD_frameHeader zfh; + if (ZSTD_getFrameHeader(&zfh, src, srcSize) != 0) + return ZSTD_CONTENTSIZE_ERROR; + if (zfh.frameType == ZSTD_skippableFrame) { + return 0; + } else { + return zfh.frameContentSize; + } } +} + +static size_t readSkippableFrameSize(void const* src, size_t srcSize) +{ + size_t const skippableHeaderSize = ZSTD_SKIPPABLEHEADERSIZE; + U32 sizeU32; + + RETURN_ERROR_IF(srcSize < ZSTD_SKIPPABLEHEADERSIZE, srcSize_wrong, ""); + + sizeU32 = MEM_readLE32((BYTE const*)src + ZSTD_FRAMEIDSIZE); + RETURN_ERROR_IF((U32)(sizeU32 + ZSTD_SKIPPABLEHEADERSIZE) < sizeU32, + frameParameter_unsupported, ""); + { + size_t const skippableSize = skippableHeaderSize + sizeU32; + RETURN_ERROR_IF(skippableSize > srcSize, srcSize_wrong, ""); + return skippableSize; + } +} + +/* ZSTD_findDecompressedSize() : + * compatible with legacy mode + * `srcSize` must be the exact length of some number of ZSTD compressed and/or + * skippable frames + * @return : decompressed size of the frames contained */ +unsigned long long ZSTD_findDecompressedSize(const void* src, size_t srcSize) +{ + unsigned long long totalDstSize = 0; + + while (srcSize >= ZSTD_startingInputLength(ZSTD_f_zstd1)) { + U32 const magicNumber = MEM_readLE32(src); + + if ((magicNumber & ZSTD_MAGIC_SKIPPABLE_MASK) == ZSTD_MAGIC_SKIPPABLE_START) { + size_t const skippableSize = readSkippableFrameSize(src, srcSize); + if (ZSTD_isError(skippableSize)) { + return ZSTD_CONTENTSIZE_ERROR; + } + assert(skippableSize <= srcSize); + + src = (const BYTE *)src + skippableSize; + srcSize -= skippableSize; + continue; + } + + { unsigned long long const ret = ZSTD_getFrameContentSize(src, srcSize); + if (ret >= ZSTD_CONTENTSIZE_ERROR) return ret; + + /* check for overflow */ + if (totalDstSize + ret < totalDstSize) return ZSTD_CONTENTSIZE_ERROR; + totalDstSize += ret; + } + { size_t const frameSrcSize = ZSTD_findFrameCompressedSize(src, srcSize); + if (ZSTD_isError(frameSrcSize)) { + return ZSTD_CONTENTSIZE_ERROR; + } + + src = (const BYTE *)src + frameSrcSize; + srcSize -= frameSrcSize; + } + } /* while (srcSize >= ZSTD_frameHeaderSize_prefix) */ + + if (srcSize) return ZSTD_CONTENTSIZE_ERROR; + + return totalDstSize; +} + +/* ZSTD_getDecompressedSize() : + * compatible with legacy mode + * @return : decompressed size if known, 0 otherwise + note : 0 can mean any of the following : + - frame content is empty + - decompressed size field is not present in frame header + - frame header unknown / not supported + - frame header not complete (`srcSize` too small) */ +unsigned long long ZSTD_getDecompressedSize(const void* src, size_t srcSize) +{ + unsigned long long const ret = ZSTD_getFrameContentSize(src, srcSize); + ZSTD_STATIC_ASSERT(ZSTD_CONTENTSIZE_ERROR < ZSTD_CONTENTSIZE_UNKNOWN); + return (ret >= ZSTD_CONTENTSIZE_ERROR) ? 0 : ret; +} + + +/* ZSTD_decodeFrameHeader() : + * `headerSize` must be the size provided by ZSTD_frameHeaderSize(). + * If multiple DDict references are enabled, also will choose the correct DDict to use. + * @return : 0 if success, or an error code, which can be tested using ZSTD_isError() */ +static size_t ZSTD_decodeFrameHeader(ZSTD_DCtx* dctx, const void* src, size_t headerSize) +{ + size_t const result = ZSTD_getFrameHeader_advanced(&(dctx->fParams), src, headerSize, dctx->format); + if (ZSTD_isError(result)) return result; /* invalid header */ + RETURN_ERROR_IF(result>0, srcSize_wrong, "headerSize too small"); + + /* Reference DDict requested by frame if dctx references multiple ddicts */ + if (dctx->refMultipleDDicts == ZSTD_rmd_refMultipleDDicts && dctx->ddictSet) { + ZSTD_DCtx_selectFrameDDict(dctx); + } + +#ifndef FUZZING_BUILD_MODE_UNSAFE_FOR_PRODUCTION + /* Skip the dictID check in fuzzing mode, because it makes the search + * harder. + */ + RETURN_ERROR_IF(dctx->fParams.dictID && (dctx->dictID != dctx->fParams.dictID), + dictionary_wrong, ""); +#endif + dctx->validateChecksum = (dctx->fParams.checksumFlag && !dctx->forceIgnoreChecksum) ? 1 : 0; + if (dctx->validateChecksum) xxh64_reset(&dctx->xxhState, 0); + dctx->processedCSize += headerSize; + return 0; +} + +static ZSTD_frameSizeInfo ZSTD_errorFrameSizeInfo(size_t ret) +{ + ZSTD_frameSizeInfo frameSizeInfo; + frameSizeInfo.compressedSize = ret; + frameSizeInfo.decompressedBound = ZSTD_CONTENTSIZE_ERROR; + return frameSizeInfo; +} + +static ZSTD_frameSizeInfo ZSTD_findFrameSizeInfo(const void* src, size_t srcSize) +{ + ZSTD_frameSizeInfo frameSizeInfo; + ZSTD_memset(&frameSizeInfo, 0, sizeof(ZSTD_frameSizeInfo)); + + + if ((srcSize >= ZSTD_SKIPPABLEHEADERSIZE) + && (MEM_readLE32(src) & ZSTD_MAGIC_SKIPPABLE_MASK) == ZSTD_MAGIC_SKIPPABLE_START) { + frameSizeInfo.compressedSize = readSkippableFrameSize(src, srcSize); + assert(ZSTD_isError(frameSizeInfo.compressedSize) || + frameSizeInfo.compressedSize <= srcSize); + return frameSizeInfo; + } else { + const BYTE* ip = (const BYTE*)src; + const BYTE* const ipstart = ip; + size_t remainingSize = srcSize; + size_t nbBlocks = 0; + ZSTD_frameHeader zfh; + + /* Extract Frame Header */ + { size_t const ret = ZSTD_getFrameHeader(&zfh, src, srcSize); + if (ZSTD_isError(ret)) + return ZSTD_errorFrameSizeInfo(ret); + if (ret > 0) + return ZSTD_errorFrameSizeInfo(ERROR(srcSize_wrong)); + } + + ip += zfh.headerSize; + remainingSize -= zfh.headerSize; + + /* Iterate over each block */ + while (1) { + blockProperties_t blockProperties; + size_t const cBlockSize = ZSTD_getcBlockSize(ip, remainingSize, &blockProperties); + if (ZSTD_isError(cBlockSize)) + return ZSTD_errorFrameSizeInfo(cBlockSize); + + if (ZSTD_blockHeaderSize + cBlockSize > remainingSize) + return ZSTD_errorFrameSizeInfo(ERROR(srcSize_wrong)); + + ip += ZSTD_blockHeaderSize + cBlockSize; + remainingSize -= ZSTD_blockHeaderSize + cBlockSize; + nbBlocks++; + + if (blockProperties.lastBlock) break; + } + + /* Final frame content checksum */ + if (zfh.checksumFlag) { + if (remainingSize < 4) + return ZSTD_errorFrameSizeInfo(ERROR(srcSize_wrong)); + ip += 4; + } + + frameSizeInfo.compressedSize = (size_t)(ip - ipstart); + frameSizeInfo.decompressedBound = (zfh.frameContentSize != ZSTD_CONTENTSIZE_UNKNOWN) + ? zfh.frameContentSize + : nbBlocks * zfh.blockSizeMax; + return frameSizeInfo; + } +} + +/* ZSTD_findFrameCompressedSize() : + * compatible with legacy mode + * `src` must point to the start of a ZSTD frame, ZSTD legacy frame, or skippable frame + * `srcSize` must be at least as large as the frame contained + * @return : the compressed size of the frame starting at `src` */ +size_t ZSTD_findFrameCompressedSize(const void *src, size_t srcSize) +{ + ZSTD_frameSizeInfo const frameSizeInfo = ZSTD_findFrameSizeInfo(src, srcSize); + return frameSizeInfo.compressedSize; +} + +/* ZSTD_decompressBound() : + * compatible with legacy mode + * `src` must point to the start of a ZSTD frame or a skippeable frame + * `srcSize` must be at least as large as the frame contained + * @return : the maximum decompressed size of the compressed source + */ +unsigned long long ZSTD_decompressBound(const void* src, size_t srcSize) +{ + unsigned long long bound = 0; + /* Iterate over each frame */ + while (srcSize > 0) { + ZSTD_frameSizeInfo const frameSizeInfo = ZSTD_findFrameSizeInfo(src, srcSize); + size_t const compressedSize = frameSizeInfo.compressedSize; + unsigned long long const decompressedBound = frameSizeInfo.decompressedBound; + if (ZSTD_isError(compressedSize) || decompressedBound == ZSTD_CONTENTSIZE_ERROR) + return ZSTD_CONTENTSIZE_ERROR; + assert(srcSize >= compressedSize); + src = (const BYTE*)src + compressedSize; + srcSize -= compressedSize; + bound += decompressedBound; + } + return bound; +} + + +/*-************************************************************* + * Frame decoding + ***************************************************************/ + +/* ZSTD_insertBlock() : + * insert `src` block into `dctx` history. Useful to track uncompressed blocks. */ +size_t ZSTD_insertBlock(ZSTD_DCtx* dctx, const void* blockStart, size_t blockSize) +{ + DEBUGLOG(5, "ZSTD_insertBlock: %u bytes", (unsigned)blockSize); + ZSTD_checkContinuity(dctx, blockStart, blockSize); + dctx->previousDstEnd = (const char*)blockStart + blockSize; + return blockSize; +} + + +static size_t ZSTD_copyRawBlock(void* dst, size_t dstCapacity, + const void* src, size_t srcSize) +{ + DEBUGLOG(5, "ZSTD_copyRawBlock"); + RETURN_ERROR_IF(srcSize > dstCapacity, dstSize_tooSmall, ""); + if (dst == NULL) { + if (srcSize == 0) return 0; + RETURN_ERROR(dstBuffer_null, ""); + } + ZSTD_memcpy(dst, src, srcSize); + return srcSize; +} + +static size_t ZSTD_setRleBlock(void* dst, size_t dstCapacity, + BYTE b, + size_t regenSize) +{ + RETURN_ERROR_IF(regenSize > dstCapacity, dstSize_tooSmall, ""); + if (dst == NULL) { + if (regenSize == 0) return 0; + RETURN_ERROR(dstBuffer_null, ""); + } + ZSTD_memset(dst, b, regenSize); + return regenSize; +} + +static void ZSTD_DCtx_trace_end(ZSTD_DCtx const* dctx, U64 uncompressedSize, U64 compressedSize, unsigned streaming) +{ + (void)dctx; + (void)uncompressedSize; + (void)compressedSize; + (void)streaming; +} + + +/*! ZSTD_decompressFrame() : + * @dctx must be properly initialized + * will update *srcPtr and *srcSizePtr, + * to make *srcPtr progress by one frame. */ +static size_t ZSTD_decompressFrame(ZSTD_DCtx* dctx, + void* dst, size_t dstCapacity, + const void** srcPtr, size_t *srcSizePtr) +{ + const BYTE* const istart = (const BYTE*)(*srcPtr); + const BYTE* ip = istart; + BYTE* const ostart = (BYTE*)dst; + BYTE* const oend = dstCapacity != 0 ? ostart + dstCapacity : ostart; + BYTE* op = ostart; + size_t remainingSrcSize = *srcSizePtr; + + DEBUGLOG(4, "ZSTD_decompressFrame (srcSize:%i)", (int)*srcSizePtr); + + /* check */ + RETURN_ERROR_IF( + remainingSrcSize < ZSTD_FRAMEHEADERSIZE_MIN(dctx->format)+ZSTD_blockHeaderSize, + srcSize_wrong, ""); + + /* Frame Header */ + { size_t const frameHeaderSize = ZSTD_frameHeaderSize_internal( + ip, ZSTD_FRAMEHEADERSIZE_PREFIX(dctx->format), dctx->format); + if (ZSTD_isError(frameHeaderSize)) return frameHeaderSize; + RETURN_ERROR_IF(remainingSrcSize < frameHeaderSize+ZSTD_blockHeaderSize, + srcSize_wrong, ""); + FORWARD_IF_ERROR( ZSTD_decodeFrameHeader(dctx, ip, frameHeaderSize) , ""); + ip += frameHeaderSize; remainingSrcSize -= frameHeaderSize; + } + + /* Loop on each block */ + while (1) { + size_t decodedSize; + blockProperties_t blockProperties; + size_t const cBlockSize = ZSTD_getcBlockSize(ip, remainingSrcSize, &blockProperties); + if (ZSTD_isError(cBlockSize)) return cBlockSize; + + ip += ZSTD_blockHeaderSize; + remainingSrcSize -= ZSTD_blockHeaderSize; + RETURN_ERROR_IF(cBlockSize > remainingSrcSize, srcSize_wrong, ""); + + switch(blockProperties.blockType) + { + case bt_compressed: + decodedSize = ZSTD_decompressBlock_internal(dctx, op, (size_t)(oend-op), ip, cBlockSize, /* frame */ 1); + break; + case bt_raw : + decodedSize = ZSTD_copyRawBlock(op, (size_t)(oend-op), ip, cBlockSize); + break; + case bt_rle : + decodedSize = ZSTD_setRleBlock(op, (size_t)(oend-op), *ip, blockProperties.origSize); + break; + case bt_reserved : + default: + RETURN_ERROR(corruption_detected, "invalid block type"); + } + + if (ZSTD_isError(decodedSize)) return decodedSize; + if (dctx->validateChecksum) + xxh64_update(&dctx->xxhState, op, decodedSize); + if (decodedSize != 0) + op += decodedSize; + assert(ip != NULL); + ip += cBlockSize; + remainingSrcSize -= cBlockSize; + if (blockProperties.lastBlock) break; + } + + if (dctx->fParams.frameContentSize != ZSTD_CONTENTSIZE_UNKNOWN) { + RETURN_ERROR_IF((U64)(op-ostart) != dctx->fParams.frameContentSize, + corruption_detected, ""); + } + if (dctx->fParams.checksumFlag) { /* Frame content checksum verification */ + RETURN_ERROR_IF(remainingSrcSize<4, checksum_wrong, ""); + if (!dctx->forceIgnoreChecksum) { + U32 const checkCalc = (U32)xxh64_digest(&dctx->xxhState); + U32 checkRead; + checkRead = MEM_readLE32(ip); + RETURN_ERROR_IF(checkRead != checkCalc, checksum_wrong, ""); + } + ip += 4; + remainingSrcSize -= 4; + } + ZSTD_DCtx_trace_end(dctx, (U64)(op-ostart), (U64)(ip-istart), /* streaming */ 0); + /* Allow caller to get size read */ + *srcPtr = ip; + *srcSizePtr = remainingSrcSize; + return (size_t)(op-ostart); +} + +static size_t ZSTD_decompressMultiFrame(ZSTD_DCtx* dctx, + void* dst, size_t dstCapacity, + const void* src, size_t srcSize, + const void* dict, size_t dictSize, + const ZSTD_DDict* ddict) +{ + void* const dststart = dst; + int moreThan1Frame = 0; + + DEBUGLOG(5, "ZSTD_decompressMultiFrame"); + assert(dict==NULL || ddict==NULL); /* either dict or ddict set, not both */ + + if (ddict) { + dict = ZSTD_DDict_dictContent(ddict); + dictSize = ZSTD_DDict_dictSize(ddict); + } + + while (srcSize >= ZSTD_startingInputLength(dctx->format)) { + + + { U32 const magicNumber = MEM_readLE32(src); + DEBUGLOG(4, "reading magic number %08X (expecting %08X)", + (unsigned)magicNumber, ZSTD_MAGICNUMBER); + if ((magicNumber & ZSTD_MAGIC_SKIPPABLE_MASK) == ZSTD_MAGIC_SKIPPABLE_START) { + size_t const skippableSize = readSkippableFrameSize(src, srcSize); + FORWARD_IF_ERROR(skippableSize, "readSkippableFrameSize failed"); + assert(skippableSize <= srcSize); + + src = (const BYTE *)src + skippableSize; + srcSize -= skippableSize; + continue; + } } + + if (ddict) { + /* we were called from ZSTD_decompress_usingDDict */ + FORWARD_IF_ERROR(ZSTD_decompressBegin_usingDDict(dctx, ddict), ""); + } else { + /* this will initialize correctly with no dict if dict == NULL, so + * use this in all cases but ddict */ + FORWARD_IF_ERROR(ZSTD_decompressBegin_usingDict(dctx, dict, dictSize), ""); + } + ZSTD_checkContinuity(dctx, dst, dstCapacity); + + { const size_t res = ZSTD_decompressFrame(dctx, dst, dstCapacity, + &src, &srcSize); + RETURN_ERROR_IF( + (ZSTD_getErrorCode(res) == ZSTD_error_prefix_unknown) + && (moreThan1Frame==1), + srcSize_wrong, + "At least one frame successfully completed, " + "but following bytes are garbage: " + "it's more likely to be a srcSize error, " + "specifying more input bytes than size of frame(s). " + "Note: one could be unlucky, it might be a corruption error instead, " + "happening right at the place where we expect zstd magic bytes. " + "But this is _much_ less likely than a srcSize field error."); + if (ZSTD_isError(res)) return res; + assert(res <= dstCapacity); + if (res != 0) + dst = (BYTE*)dst + res; + dstCapacity -= res; + } + moreThan1Frame = 1; + } /* while (srcSize >= ZSTD_frameHeaderSize_prefix) */ + + RETURN_ERROR_IF(srcSize, srcSize_wrong, "input not entirely consumed"); + + return (size_t)((BYTE*)dst - (BYTE*)dststart); +} + +size_t ZSTD_decompress_usingDict(ZSTD_DCtx* dctx, + void* dst, size_t dstCapacity, + const void* src, size_t srcSize, + const void* dict, size_t dictSize) +{ + return ZSTD_decompressMultiFrame(dctx, dst, dstCapacity, src, srcSize, dict, dictSize, NULL); +} + + +static ZSTD_DDict const* ZSTD_getDDict(ZSTD_DCtx* dctx) +{ + switch (dctx->dictUses) { + default: + assert(0 /* Impossible */); + ZSTD_FALLTHROUGH; + case ZSTD_dont_use: + ZSTD_clearDict(dctx); + return NULL; + case ZSTD_use_indefinitely: + return dctx->ddict; + case ZSTD_use_once: + dctx->dictUses = ZSTD_dont_use; + return dctx->ddict; + } +} + +size_t ZSTD_decompressDCtx(ZSTD_DCtx* dctx, void* dst, size_t dstCapacity, const void* src, size_t srcSize) +{ + return ZSTD_decompress_usingDDict(dctx, dst, dstCapacity, src, srcSize, ZSTD_getDDict(dctx)); +} + + +size_t ZSTD_decompress(void* dst, size_t dstCapacity, const void* src, size_t srcSize) +{ +#if defined(ZSTD_HEAPMODE) && (ZSTD_HEAPMODE>=1) + size_t regenSize; + ZSTD_DCtx* const dctx = ZSTD_createDCtx(); + RETURN_ERROR_IF(dctx==NULL, memory_allocation, "NULL pointer!"); + regenSize = ZSTD_decompressDCtx(dctx, dst, dstCapacity, src, srcSize); + ZSTD_freeDCtx(dctx); + return regenSize; +#else /* stack mode */ + ZSTD_DCtx dctx; + ZSTD_initDCtx_internal(&dctx); + return ZSTD_decompressDCtx(&dctx, dst, dstCapacity, src, srcSize); +#endif +} + + +/*-************************************** +* Advanced Streaming Decompression API +* Bufferless and synchronous +****************************************/ +size_t ZSTD_nextSrcSizeToDecompress(ZSTD_DCtx* dctx) { return dctx->expected; } + +/* + * Similar to ZSTD_nextSrcSizeToDecompress(), but when when a block input can be streamed, + * we allow taking a partial block as the input. Currently only raw uncompressed blocks can + * be streamed. + * + * For blocks that can be streamed, this allows us to reduce the latency until we produce + * output, and avoid copying the input. + * + * @param inputSize - The total amount of input that the caller currently has. + */ +static size_t ZSTD_nextSrcSizeToDecompressWithInputSize(ZSTD_DCtx* dctx, size_t inputSize) { + if (!(dctx->stage == ZSTDds_decompressBlock || dctx->stage == ZSTDds_decompressLastBlock)) + return dctx->expected; + if (dctx->bType != bt_raw) + return dctx->expected; + return MIN(MAX(inputSize, 1), dctx->expected); +} + +ZSTD_nextInputType_e ZSTD_nextInputType(ZSTD_DCtx* dctx) { + switch(dctx->stage) + { + default: /* should not happen */ + assert(0); + ZSTD_FALLTHROUGH; + case ZSTDds_getFrameHeaderSize: + ZSTD_FALLTHROUGH; + case ZSTDds_decodeFrameHeader: + return ZSTDnit_frameHeader; + case ZSTDds_decodeBlockHeader: + return ZSTDnit_blockHeader; + case ZSTDds_decompressBlock: + return ZSTDnit_block; + case ZSTDds_decompressLastBlock: + return ZSTDnit_lastBlock; + case ZSTDds_checkChecksum: + return ZSTDnit_checksum; + case ZSTDds_decodeSkippableHeader: + ZSTD_FALLTHROUGH; + case ZSTDds_skipFrame: + return ZSTDnit_skippableFrame; + } +} + +static int ZSTD_isSkipFrame(ZSTD_DCtx* dctx) { return dctx->stage == ZSTDds_skipFrame; } + +/* ZSTD_decompressContinue() : + * srcSize : must be the exact nb of bytes expected (see ZSTD_nextSrcSizeToDecompress()) + * @return : nb of bytes generated into `dst` (necessarily <= `dstCapacity) + * or an error code, which can be tested using ZSTD_isError() */ +size_t ZSTD_decompressContinue(ZSTD_DCtx* dctx, void* dst, size_t dstCapacity, const void* src, size_t srcSize) +{ + DEBUGLOG(5, "ZSTD_decompressContinue (srcSize:%u)", (unsigned)srcSize); + /* Sanity check */ + RETURN_ERROR_IF(srcSize != ZSTD_nextSrcSizeToDecompressWithInputSize(dctx, srcSize), srcSize_wrong, "not allowed"); + ZSTD_checkContinuity(dctx, dst, dstCapacity); + + dctx->processedCSize += srcSize; + + switch (dctx->stage) + { + case ZSTDds_getFrameHeaderSize : + assert(src != NULL); + if (dctx->format == ZSTD_f_zstd1) { /* allows header */ + assert(srcSize >= ZSTD_FRAMEIDSIZE); /* to read skippable magic number */ + if ((MEM_readLE32(src) & ZSTD_MAGIC_SKIPPABLE_MASK) == ZSTD_MAGIC_SKIPPABLE_START) { /* skippable frame */ + ZSTD_memcpy(dctx->headerBuffer, src, srcSize); + dctx->expected = ZSTD_SKIPPABLEHEADERSIZE - srcSize; /* remaining to load to get full skippable frame header */ + dctx->stage = ZSTDds_decodeSkippableHeader; + return 0; + } } + dctx->headerSize = ZSTD_frameHeaderSize_internal(src, srcSize, dctx->format); + if (ZSTD_isError(dctx->headerSize)) return dctx->headerSize; + ZSTD_memcpy(dctx->headerBuffer, src, srcSize); + dctx->expected = dctx->headerSize - srcSize; + dctx->stage = ZSTDds_decodeFrameHeader; + return 0; + + case ZSTDds_decodeFrameHeader: + assert(src != NULL); + ZSTD_memcpy(dctx->headerBuffer + (dctx->headerSize - srcSize), src, srcSize); + FORWARD_IF_ERROR(ZSTD_decodeFrameHeader(dctx, dctx->headerBuffer, dctx->headerSize), ""); + dctx->expected = ZSTD_blockHeaderSize; + dctx->stage = ZSTDds_decodeBlockHeader; + return 0; + + case ZSTDds_decodeBlockHeader: + { blockProperties_t bp; + size_t const cBlockSize = ZSTD_getcBlockSize(src, ZSTD_blockHeaderSize, &bp); + if (ZSTD_isError(cBlockSize)) return cBlockSize; + RETURN_ERROR_IF(cBlockSize > dctx->fParams.blockSizeMax, corruption_detected, "Block Size Exceeds Maximum"); + dctx->expected = cBlockSize; + dctx->bType = bp.blockType; + dctx->rleSize = bp.origSize; + if (cBlockSize) { + dctx->stage = bp.lastBlock ? ZSTDds_decompressLastBlock : ZSTDds_decompressBlock; + return 0; + } + /* empty block */ + if (bp.lastBlock) { + if (dctx->fParams.checksumFlag) { + dctx->expected = 4; + dctx->stage = ZSTDds_checkChecksum; + } else { + dctx->expected = 0; /* end of frame */ + dctx->stage = ZSTDds_getFrameHeaderSize; + } + } else { + dctx->expected = ZSTD_blockHeaderSize; /* jump to next header */ + dctx->stage = ZSTDds_decodeBlockHeader; + } + return 0; + } + + case ZSTDds_decompressLastBlock: + case ZSTDds_decompressBlock: + DEBUGLOG(5, "ZSTD_decompressContinue: case ZSTDds_decompressBlock"); + { size_t rSize; + switch(dctx->bType) + { + case bt_compressed: + DEBUGLOG(5, "ZSTD_decompressContinue: case bt_compressed"); + rSize = ZSTD_decompressBlock_internal(dctx, dst, dstCapacity, src, srcSize, /* frame */ 1); + dctx->expected = 0; /* Streaming not supported */ + break; + case bt_raw : + assert(srcSize <= dctx->expected); + rSize = ZSTD_copyRawBlock(dst, dstCapacity, src, srcSize); + FORWARD_IF_ERROR(rSize, "ZSTD_copyRawBlock failed"); + assert(rSize == srcSize); + dctx->expected -= rSize; + break; + case bt_rle : + rSize = ZSTD_setRleBlock(dst, dstCapacity, *(const BYTE*)src, dctx->rleSize); + dctx->expected = 0; /* Streaming not supported */ + break; + case bt_reserved : /* should never happen */ + default: + RETURN_ERROR(corruption_detected, "invalid block type"); + } + FORWARD_IF_ERROR(rSize, ""); + RETURN_ERROR_IF(rSize > dctx->fParams.blockSizeMax, corruption_detected, "Decompressed Block Size Exceeds Maximum"); + DEBUGLOG(5, "ZSTD_decompressContinue: decoded size from block : %u", (unsigned)rSize); + dctx->decodedSize += rSize; + if (dctx->validateChecksum) xxh64_update(&dctx->xxhState, dst, rSize); + dctx->previousDstEnd = (char*)dst + rSize; + + /* Stay on the same stage until we are finished streaming the block. */ + if (dctx->expected > 0) { + return rSize; + } + + if (dctx->stage == ZSTDds_decompressLastBlock) { /* end of frame */ + DEBUGLOG(4, "ZSTD_decompressContinue: decoded size from frame : %u", (unsigned)dctx->decodedSize); + RETURN_ERROR_IF( + dctx->fParams.frameContentSize != ZSTD_CONTENTSIZE_UNKNOWN + && dctx->decodedSize != dctx->fParams.frameContentSize, + corruption_detected, ""); + if (dctx->fParams.checksumFlag) { /* another round for frame checksum */ + dctx->expected = 4; + dctx->stage = ZSTDds_checkChecksum; + } else { + ZSTD_DCtx_trace_end(dctx, dctx->decodedSize, dctx->processedCSize, /* streaming */ 1); + dctx->expected = 0; /* ends here */ + dctx->stage = ZSTDds_getFrameHeaderSize; + } + } else { + dctx->stage = ZSTDds_decodeBlockHeader; + dctx->expected = ZSTD_blockHeaderSize; + } + return rSize; + } + + case ZSTDds_checkChecksum: + assert(srcSize == 4); /* guaranteed by dctx->expected */ + { + if (dctx->validateChecksum) { + U32 const h32 = (U32)xxh64_digest(&dctx->xxhState); + U32 const check32 = MEM_readLE32(src); + DEBUGLOG(4, "ZSTD_decompressContinue: checksum : calculated %08X :: %08X read", (unsigned)h32, (unsigned)check32); + RETURN_ERROR_IF(check32 != h32, checksum_wrong, ""); + } + ZSTD_DCtx_trace_end(dctx, dctx->decodedSize, dctx->processedCSize, /* streaming */ 1); + dctx->expected = 0; + dctx->stage = ZSTDds_getFrameHeaderSize; + return 0; + } + + case ZSTDds_decodeSkippableHeader: + assert(src != NULL); + assert(srcSize <= ZSTD_SKIPPABLEHEADERSIZE); + ZSTD_memcpy(dctx->headerBuffer + (ZSTD_SKIPPABLEHEADERSIZE - srcSize), src, srcSize); /* complete skippable header */ + dctx->expected = MEM_readLE32(dctx->headerBuffer + ZSTD_FRAMEIDSIZE); /* note : dctx->expected can grow seriously large, beyond local buffer size */ + dctx->stage = ZSTDds_skipFrame; + return 0; + + case ZSTDds_skipFrame: + dctx->expected = 0; + dctx->stage = ZSTDds_getFrameHeaderSize; + return 0; + + default: + assert(0); /* impossible */ + RETURN_ERROR(GENERIC, "impossible to reach"); /* some compiler require default to do something */ + } +} + + +static size_t ZSTD_refDictContent(ZSTD_DCtx* dctx, const void* dict, size_t dictSize) +{ + dctx->dictEnd = dctx->previousDstEnd; + dctx->virtualStart = (const char*)dict - ((const char*)(dctx->previousDstEnd) - (const char*)(dctx->prefixStart)); + dctx->prefixStart = dict; + dctx->previousDstEnd = (const char*)dict + dictSize; +#ifdef FUZZING_BUILD_MODE_UNSAFE_FOR_PRODUCTION + dctx->dictContentBeginForFuzzing = dctx->prefixStart; + dctx->dictContentEndForFuzzing = dctx->previousDstEnd; +#endif + return 0; +} + +/*! ZSTD_loadDEntropy() : + * dict : must point at beginning of a valid zstd dictionary. + * @return : size of entropy tables read */ +size_t +ZSTD_loadDEntropy(ZSTD_entropyDTables_t* entropy, + const void* const dict, size_t const dictSize) +{ + const BYTE* dictPtr = (const BYTE*)dict; + const BYTE* const dictEnd = dictPtr + dictSize; + + RETURN_ERROR_IF(dictSize <= 8, dictionary_corrupted, "dict is too small"); + assert(MEM_readLE32(dict) == ZSTD_MAGIC_DICTIONARY); /* dict must be valid */ + dictPtr += 8; /* skip header = magic + dictID */ + + ZSTD_STATIC_ASSERT(offsetof(ZSTD_entropyDTables_t, OFTable) == offsetof(ZSTD_entropyDTables_t, LLTable) + sizeof(entropy->LLTable)); + ZSTD_STATIC_ASSERT(offsetof(ZSTD_entropyDTables_t, MLTable) == offsetof(ZSTD_entropyDTables_t, OFTable) + sizeof(entropy->OFTable)); + ZSTD_STATIC_ASSERT(sizeof(entropy->LLTable) + sizeof(entropy->OFTable) + sizeof(entropy->MLTable) >= HUF_DECOMPRESS_WORKSPACE_SIZE); + { void* const workspace = &entropy->LLTable; /* use fse tables as temporary workspace; implies fse tables are grouped together */ + size_t const workspaceSize = sizeof(entropy->LLTable) + sizeof(entropy->OFTable) + sizeof(entropy->MLTable); +#ifdef HUF_FORCE_DECOMPRESS_X1 + /* in minimal huffman, we always use X1 variants */ + size_t const hSize = HUF_readDTableX1_wksp(entropy->hufTable, + dictPtr, dictEnd - dictPtr, + workspace, workspaceSize); +#else + size_t const hSize = HUF_readDTableX2_wksp(entropy->hufTable, + dictPtr, (size_t)(dictEnd - dictPtr), + workspace, workspaceSize); +#endif + RETURN_ERROR_IF(HUF_isError(hSize), dictionary_corrupted, ""); + dictPtr += hSize; + } + + { short offcodeNCount[MaxOff+1]; + unsigned offcodeMaxValue = MaxOff, offcodeLog; + size_t const offcodeHeaderSize = FSE_readNCount(offcodeNCount, &offcodeMaxValue, &offcodeLog, dictPtr, (size_t)(dictEnd-dictPtr)); + RETURN_ERROR_IF(FSE_isError(offcodeHeaderSize), dictionary_corrupted, ""); + RETURN_ERROR_IF(offcodeMaxValue > MaxOff, dictionary_corrupted, ""); + RETURN_ERROR_IF(offcodeLog > OffFSELog, dictionary_corrupted, ""); + ZSTD_buildFSETable( entropy->OFTable, + offcodeNCount, offcodeMaxValue, + OF_base, OF_bits, + offcodeLog, + entropy->workspace, sizeof(entropy->workspace), + /* bmi2 */0); + dictPtr += offcodeHeaderSize; + } + + { short matchlengthNCount[MaxML+1]; + unsigned matchlengthMaxValue = MaxML, matchlengthLog; + size_t const matchlengthHeaderSize = FSE_readNCount(matchlengthNCount, &matchlengthMaxValue, &matchlengthLog, dictPtr, (size_t)(dictEnd-dictPtr)); + RETURN_ERROR_IF(FSE_isError(matchlengthHeaderSize), dictionary_corrupted, ""); + RETURN_ERROR_IF(matchlengthMaxValue > MaxML, dictionary_corrupted, ""); + RETURN_ERROR_IF(matchlengthLog > MLFSELog, dictionary_corrupted, ""); + ZSTD_buildFSETable( entropy->MLTable, + matchlengthNCount, matchlengthMaxValue, + ML_base, ML_bits, + matchlengthLog, + entropy->workspace, sizeof(entropy->workspace), + /* bmi2 */ 0); + dictPtr += matchlengthHeaderSize; + } + + { short litlengthNCount[MaxLL+1]; + unsigned litlengthMaxValue = MaxLL, litlengthLog; + size_t const litlengthHeaderSize = FSE_readNCount(litlengthNCount, &litlengthMaxValue, &litlengthLog, dictPtr, (size_t)(dictEnd-dictPtr)); + RETURN_ERROR_IF(FSE_isError(litlengthHeaderSize), dictionary_corrupted, ""); + RETURN_ERROR_IF(litlengthMaxValue > MaxLL, dictionary_corrupted, ""); + RETURN_ERROR_IF(litlengthLog > LLFSELog, dictionary_corrupted, ""); + ZSTD_buildFSETable( entropy->LLTable, + litlengthNCount, litlengthMaxValue, + LL_base, LL_bits, + litlengthLog, + entropy->workspace, sizeof(entropy->workspace), + /* bmi2 */ 0); + dictPtr += litlengthHeaderSize; + } + + RETURN_ERROR_IF(dictPtr+12 > dictEnd, dictionary_corrupted, ""); + { int i; + size_t const dictContentSize = (size_t)(dictEnd - (dictPtr+12)); + for (i=0; i<3; i++) { + U32 const rep = MEM_readLE32(dictPtr); dictPtr += 4; + RETURN_ERROR_IF(rep==0 || rep > dictContentSize, + dictionary_corrupted, ""); + entropy->rep[i] = rep; + } } + + return (size_t)(dictPtr - (const BYTE*)dict); +} + +static size_t ZSTD_decompress_insertDictionary(ZSTD_DCtx* dctx, const void* dict, size_t dictSize) +{ + if (dictSize < 8) return ZSTD_refDictContent(dctx, dict, dictSize); + { U32 const magic = MEM_readLE32(dict); + if (magic != ZSTD_MAGIC_DICTIONARY) { + return ZSTD_refDictContent(dctx, dict, dictSize); /* pure content mode */ + } } + dctx->dictID = MEM_readLE32((const char*)dict + ZSTD_FRAMEIDSIZE); + + /* load entropy tables */ + { size_t const eSize = ZSTD_loadDEntropy(&dctx->entropy, dict, dictSize); + RETURN_ERROR_IF(ZSTD_isError(eSize), dictionary_corrupted, ""); + dict = (const char*)dict + eSize; + dictSize -= eSize; + } + dctx->litEntropy = dctx->fseEntropy = 1; + + /* reference dictionary content */ + return ZSTD_refDictContent(dctx, dict, dictSize); +} + +size_t ZSTD_decompressBegin(ZSTD_DCtx* dctx) +{ + assert(dctx != NULL); + dctx->expected = ZSTD_startingInputLength(dctx->format); /* dctx->format must be properly set */ + dctx->stage = ZSTDds_getFrameHeaderSize; + dctx->processedCSize = 0; + dctx->decodedSize = 0; + dctx->previousDstEnd = NULL; + dctx->prefixStart = NULL; + dctx->virtualStart = NULL; + dctx->dictEnd = NULL; + dctx->entropy.hufTable[0] = (HUF_DTable)((HufLog)*0x1000001); /* cover both little and big endian */ + dctx->litEntropy = dctx->fseEntropy = 0; + dctx->dictID = 0; + dctx->bType = bt_reserved; + ZSTD_STATIC_ASSERT(sizeof(dctx->entropy.rep) == sizeof(repStartValue)); + ZSTD_memcpy(dctx->entropy.rep, repStartValue, sizeof(repStartValue)); /* initial repcodes */ + dctx->LLTptr = dctx->entropy.LLTable; + dctx->MLTptr = dctx->entropy.MLTable; + dctx->OFTptr = dctx->entropy.OFTable; + dctx->HUFptr = dctx->entropy.hufTable; + return 0; +} + +size_t ZSTD_decompressBegin_usingDict(ZSTD_DCtx* dctx, const void* dict, size_t dictSize) +{ + FORWARD_IF_ERROR( ZSTD_decompressBegin(dctx) , ""); + if (dict && dictSize) + RETURN_ERROR_IF( + ZSTD_isError(ZSTD_decompress_insertDictionary(dctx, dict, dictSize)), + dictionary_corrupted, ""); + return 0; +} + + +/* ====== ZSTD_DDict ====== */ + +size_t ZSTD_decompressBegin_usingDDict(ZSTD_DCtx* dctx, const ZSTD_DDict* ddict) +{ + DEBUGLOG(4, "ZSTD_decompressBegin_usingDDict"); + assert(dctx != NULL); + if (ddict) { + const char* const dictStart = (const char*)ZSTD_DDict_dictContent(ddict); + size_t const dictSize = ZSTD_DDict_dictSize(ddict); + const void* const dictEnd = dictStart + dictSize; + dctx->ddictIsCold = (dctx->dictEnd != dictEnd); + DEBUGLOG(4, "DDict is %s", + dctx->ddictIsCold ? "~cold~" : "hot!"); + } + FORWARD_IF_ERROR( ZSTD_decompressBegin(dctx) , ""); + if (ddict) { /* NULL ddict is equivalent to no dictionary */ + ZSTD_copyDDictParameters(dctx, ddict); + } + return 0; +} + +/*! ZSTD_getDictID_fromDict() : + * Provides the dictID stored within dictionary. + * if @return == 0, the dictionary is not conformant with Zstandard specification. + * It can still be loaded, but as a content-only dictionary. */ +unsigned ZSTD_getDictID_fromDict(const void* dict, size_t dictSize) +{ + if (dictSize < 8) return 0; + if (MEM_readLE32(dict) != ZSTD_MAGIC_DICTIONARY) return 0; + return MEM_readLE32((const char*)dict + ZSTD_FRAMEIDSIZE); +} + +/*! ZSTD_getDictID_fromFrame() : + * Provides the dictID required to decompress frame stored within `src`. + * If @return == 0, the dictID could not be decoded. + * This could for one of the following reasons : + * - The frame does not require a dictionary (most common case). + * - The frame was built with dictID intentionally removed. + * Needed dictionary is a hidden information. + * Note : this use case also happens when using a non-conformant dictionary. + * - `srcSize` is too small, and as a result, frame header could not be decoded. + * Note : possible if `srcSize < ZSTD_FRAMEHEADERSIZE_MAX`. + * - This is not a Zstandard frame. + * When identifying the exact failure cause, it's possible to use + * ZSTD_getFrameHeader(), which will provide a more precise error code. */ +unsigned ZSTD_getDictID_fromFrame(const void* src, size_t srcSize) +{ + ZSTD_frameHeader zfp = { 0, 0, 0, ZSTD_frame, 0, 0, 0 }; + size_t const hError = ZSTD_getFrameHeader(&zfp, src, srcSize); + if (ZSTD_isError(hError)) return 0; + return zfp.dictID; +} + + +/*! ZSTD_decompress_usingDDict() : +* Decompression using a pre-digested Dictionary +* Use dictionary without significant overhead. */ +size_t ZSTD_decompress_usingDDict(ZSTD_DCtx* dctx, + void* dst, size_t dstCapacity, + const void* src, size_t srcSize, + const ZSTD_DDict* ddict) +{ + /* pass content and size in case legacy frames are encountered */ + return ZSTD_decompressMultiFrame(dctx, dst, dstCapacity, src, srcSize, + NULL, 0, + ddict); +} + + +/*===================================== +* Streaming decompression +*====================================*/ + +ZSTD_DStream* ZSTD_createDStream(void) +{ + DEBUGLOG(3, "ZSTD_createDStream"); + return ZSTD_createDStream_advanced(ZSTD_defaultCMem); +} + +ZSTD_DStream* ZSTD_initStaticDStream(void *workspace, size_t workspaceSize) +{ + return ZSTD_initStaticDCtx(workspace, workspaceSize); +} + +ZSTD_DStream* ZSTD_createDStream_advanced(ZSTD_customMem customMem) +{ + return ZSTD_createDCtx_advanced(customMem); +} + +size_t ZSTD_freeDStream(ZSTD_DStream* zds) +{ + return ZSTD_freeDCtx(zds); +} + + +/* *** Initialization *** */ + +size_t ZSTD_DStreamInSize(void) { return ZSTD_BLOCKSIZE_MAX + ZSTD_blockHeaderSize; } +size_t ZSTD_DStreamOutSize(void) { return ZSTD_BLOCKSIZE_MAX; } + +size_t ZSTD_DCtx_loadDictionary_advanced(ZSTD_DCtx* dctx, + const void* dict, size_t dictSize, + ZSTD_dictLoadMethod_e dictLoadMethod, + ZSTD_dictContentType_e dictContentType) +{ + RETURN_ERROR_IF(dctx->streamStage != zdss_init, stage_wrong, ""); + ZSTD_clearDict(dctx); + if (dict && dictSize != 0) { + dctx->ddictLocal = ZSTD_createDDict_advanced(dict, dictSize, dictLoadMethod, dictContentType, dctx->customMem); + RETURN_ERROR_IF(dctx->ddictLocal == NULL, memory_allocation, "NULL pointer!"); + dctx->ddict = dctx->ddictLocal; + dctx->dictUses = ZSTD_use_indefinitely; + } + return 0; +} + +size_t ZSTD_DCtx_loadDictionary_byReference(ZSTD_DCtx* dctx, const void* dict, size_t dictSize) +{ + return ZSTD_DCtx_loadDictionary_advanced(dctx, dict, dictSize, ZSTD_dlm_byRef, ZSTD_dct_auto); +} + +size_t ZSTD_DCtx_loadDictionary(ZSTD_DCtx* dctx, const void* dict, size_t dictSize) +{ + return ZSTD_DCtx_loadDictionary_advanced(dctx, dict, dictSize, ZSTD_dlm_byCopy, ZSTD_dct_auto); +} + +size_t ZSTD_DCtx_refPrefix_advanced(ZSTD_DCtx* dctx, const void* prefix, size_t prefixSize, ZSTD_dictContentType_e dictContentType) +{ + FORWARD_IF_ERROR(ZSTD_DCtx_loadDictionary_advanced(dctx, prefix, prefixSize, ZSTD_dlm_byRef, dictContentType), ""); + dctx->dictUses = ZSTD_use_once; + return 0; +} + +size_t ZSTD_DCtx_refPrefix(ZSTD_DCtx* dctx, const void* prefix, size_t prefixSize) +{ + return ZSTD_DCtx_refPrefix_advanced(dctx, prefix, prefixSize, ZSTD_dct_rawContent); +} + + +/* ZSTD_initDStream_usingDict() : + * return : expected size, aka ZSTD_startingInputLength(). + * this function cannot fail */ +size_t ZSTD_initDStream_usingDict(ZSTD_DStream* zds, const void* dict, size_t dictSize) +{ + DEBUGLOG(4, "ZSTD_initDStream_usingDict"); + FORWARD_IF_ERROR( ZSTD_DCtx_reset(zds, ZSTD_reset_session_only) , ""); + FORWARD_IF_ERROR( ZSTD_DCtx_loadDictionary(zds, dict, dictSize) , ""); + return ZSTD_startingInputLength(zds->format); +} + +/* note : this variant can't fail */ +size_t ZSTD_initDStream(ZSTD_DStream* zds) +{ + DEBUGLOG(4, "ZSTD_initDStream"); + return ZSTD_initDStream_usingDDict(zds, NULL); +} + +/* ZSTD_initDStream_usingDDict() : + * ddict will just be referenced, and must outlive decompression session + * this function cannot fail */ +size_t ZSTD_initDStream_usingDDict(ZSTD_DStream* dctx, const ZSTD_DDict* ddict) +{ + FORWARD_IF_ERROR( ZSTD_DCtx_reset(dctx, ZSTD_reset_session_only) , ""); + FORWARD_IF_ERROR( ZSTD_DCtx_refDDict(dctx, ddict) , ""); + return ZSTD_startingInputLength(dctx->format); +} + +/* ZSTD_resetDStream() : + * return : expected size, aka ZSTD_startingInputLength(). + * this function cannot fail */ +size_t ZSTD_resetDStream(ZSTD_DStream* dctx) +{ + FORWARD_IF_ERROR(ZSTD_DCtx_reset(dctx, ZSTD_reset_session_only), ""); + return ZSTD_startingInputLength(dctx->format); +} + + +size_t ZSTD_DCtx_refDDict(ZSTD_DCtx* dctx, const ZSTD_DDict* ddict) +{ + RETURN_ERROR_IF(dctx->streamStage != zdss_init, stage_wrong, ""); + ZSTD_clearDict(dctx); + if (ddict) { + dctx->ddict = ddict; + dctx->dictUses = ZSTD_use_indefinitely; + if (dctx->refMultipleDDicts == ZSTD_rmd_refMultipleDDicts) { + if (dctx->ddictSet == NULL) { + dctx->ddictSet = ZSTD_createDDictHashSet(dctx->customMem); + if (!dctx->ddictSet) { + RETURN_ERROR(memory_allocation, "Failed to allocate memory for hash set!"); + } + } + assert(!dctx->staticSize); /* Impossible: ddictSet cannot have been allocated if static dctx */ + FORWARD_IF_ERROR(ZSTD_DDictHashSet_addDDict(dctx->ddictSet, ddict, dctx->customMem), ""); + } + } + return 0; +} + +/* ZSTD_DCtx_setMaxWindowSize() : + * note : no direct equivalence in ZSTD_DCtx_setParameter, + * since this version sets windowSize, and the other sets windowLog */ +size_t ZSTD_DCtx_setMaxWindowSize(ZSTD_DCtx* dctx, size_t maxWindowSize) +{ + ZSTD_bounds const bounds = ZSTD_dParam_getBounds(ZSTD_d_windowLogMax); + size_t const min = (size_t)1 << bounds.lowerBound; + size_t const max = (size_t)1 << bounds.upperBound; + RETURN_ERROR_IF(dctx->streamStage != zdss_init, stage_wrong, ""); + RETURN_ERROR_IF(maxWindowSize < min, parameter_outOfBound, ""); + RETURN_ERROR_IF(maxWindowSize > max, parameter_outOfBound, ""); + dctx->maxWindowSize = maxWindowSize; + return 0; +} + +size_t ZSTD_DCtx_setFormat(ZSTD_DCtx* dctx, ZSTD_format_e format) +{ + return ZSTD_DCtx_setParameter(dctx, ZSTD_d_format, (int)format); +} + +ZSTD_bounds ZSTD_dParam_getBounds(ZSTD_dParameter dParam) +{ + ZSTD_bounds bounds = { 0, 0, 0 }; + switch(dParam) { + case ZSTD_d_windowLogMax: + bounds.lowerBound = ZSTD_WINDOWLOG_ABSOLUTEMIN; + bounds.upperBound = ZSTD_WINDOWLOG_MAX; + return bounds; + case ZSTD_d_format: + bounds.lowerBound = (int)ZSTD_f_zstd1; + bounds.upperBound = (int)ZSTD_f_zstd1_magicless; + ZSTD_STATIC_ASSERT(ZSTD_f_zstd1 < ZSTD_f_zstd1_magicless); + return bounds; + case ZSTD_d_stableOutBuffer: + bounds.lowerBound = (int)ZSTD_bm_buffered; + bounds.upperBound = (int)ZSTD_bm_stable; + return bounds; + case ZSTD_d_forceIgnoreChecksum: + bounds.lowerBound = (int)ZSTD_d_validateChecksum; + bounds.upperBound = (int)ZSTD_d_ignoreChecksum; + return bounds; + case ZSTD_d_refMultipleDDicts: + bounds.lowerBound = (int)ZSTD_rmd_refSingleDDict; + bounds.upperBound = (int)ZSTD_rmd_refMultipleDDicts; + return bounds; + default:; + } + bounds.error = ERROR(parameter_unsupported); + return bounds; +} + +/* ZSTD_dParam_withinBounds: + * @return 1 if value is within dParam bounds, + * 0 otherwise */ +static int ZSTD_dParam_withinBounds(ZSTD_dParameter dParam, int value) +{ + ZSTD_bounds const bounds = ZSTD_dParam_getBounds(dParam); + if (ZSTD_isError(bounds.error)) return 0; + if (value < bounds.lowerBound) return 0; + if (value > bounds.upperBound) return 0; + return 1; +} + +#define CHECK_DBOUNDS(p,v) { \ + RETURN_ERROR_IF(!ZSTD_dParam_withinBounds(p, v), parameter_outOfBound, ""); \ +} + +size_t ZSTD_DCtx_getParameter(ZSTD_DCtx* dctx, ZSTD_dParameter param, int* value) +{ + switch (param) { + case ZSTD_d_windowLogMax: + *value = (int)ZSTD_highbit32((U32)dctx->maxWindowSize); + return 0; + case ZSTD_d_format: + *value = (int)dctx->format; + return 0; + case ZSTD_d_stableOutBuffer: + *value = (int)dctx->outBufferMode; + return 0; + case ZSTD_d_forceIgnoreChecksum: + *value = (int)dctx->forceIgnoreChecksum; + return 0; + case ZSTD_d_refMultipleDDicts: + *value = (int)dctx->refMultipleDDicts; + return 0; + default:; + } + RETURN_ERROR(parameter_unsupported, ""); +} + +size_t ZSTD_DCtx_setParameter(ZSTD_DCtx* dctx, ZSTD_dParameter dParam, int value) +{ + RETURN_ERROR_IF(dctx->streamStage != zdss_init, stage_wrong, ""); + switch(dParam) { + case ZSTD_d_windowLogMax: + if (value == 0) value = ZSTD_WINDOWLOG_LIMIT_DEFAULT; + CHECK_DBOUNDS(ZSTD_d_windowLogMax, value); + dctx->maxWindowSize = ((size_t)1) << value; + return 0; + case ZSTD_d_format: + CHECK_DBOUNDS(ZSTD_d_format, value); + dctx->format = (ZSTD_format_e)value; + return 0; + case ZSTD_d_stableOutBuffer: + CHECK_DBOUNDS(ZSTD_d_stableOutBuffer, value); + dctx->outBufferMode = (ZSTD_bufferMode_e)value; + return 0; + case ZSTD_d_forceIgnoreChecksum: + CHECK_DBOUNDS(ZSTD_d_forceIgnoreChecksum, value); + dctx->forceIgnoreChecksum = (ZSTD_forceIgnoreChecksum_e)value; + return 0; + case ZSTD_d_refMultipleDDicts: + CHECK_DBOUNDS(ZSTD_d_refMultipleDDicts, value); + if (dctx->staticSize != 0) { + RETURN_ERROR(parameter_unsupported, "Static dctx does not support multiple DDicts!"); + } + dctx->refMultipleDDicts = (ZSTD_refMultipleDDicts_e)value; + return 0; + default:; + } + RETURN_ERROR(parameter_unsupported, ""); +} + +size_t ZSTD_DCtx_reset(ZSTD_DCtx* dctx, ZSTD_ResetDirective reset) +{ + if ( (reset == ZSTD_reset_session_only) + || (reset == ZSTD_reset_session_and_parameters) ) { + dctx->streamStage = zdss_init; + dctx->noForwardProgress = 0; + } + if ( (reset == ZSTD_reset_parameters) + || (reset == ZSTD_reset_session_and_parameters) ) { + RETURN_ERROR_IF(dctx->streamStage != zdss_init, stage_wrong, ""); + ZSTD_clearDict(dctx); + ZSTD_DCtx_resetParameters(dctx); + } + return 0; +} + + +size_t ZSTD_sizeof_DStream(const ZSTD_DStream* dctx) +{ + return ZSTD_sizeof_DCtx(dctx); +} + +size_t ZSTD_decodingBufferSize_min(unsigned long long windowSize, unsigned long long frameContentSize) +{ + size_t const blockSize = (size_t) MIN(windowSize, ZSTD_BLOCKSIZE_MAX); + unsigned long long const neededRBSize = windowSize + blockSize + (WILDCOPY_OVERLENGTH * 2); + unsigned long long const neededSize = MIN(frameContentSize, neededRBSize); + size_t const minRBSize = (size_t) neededSize; + RETURN_ERROR_IF((unsigned long long)minRBSize != neededSize, + frameParameter_windowTooLarge, ""); + return minRBSize; +} + +size_t ZSTD_estimateDStreamSize(size_t windowSize) +{ + size_t const blockSize = MIN(windowSize, ZSTD_BLOCKSIZE_MAX); + size_t const inBuffSize = blockSize; /* no block can be larger */ + size_t const outBuffSize = ZSTD_decodingBufferSize_min(windowSize, ZSTD_CONTENTSIZE_UNKNOWN); + return ZSTD_estimateDCtxSize() + inBuffSize + outBuffSize; +} + +size_t ZSTD_estimateDStreamSize_fromFrame(const void* src, size_t srcSize) +{ + U32 const windowSizeMax = 1U << ZSTD_WINDOWLOG_MAX; /* note : should be user-selectable, but requires an additional parameter (or a dctx) */ + ZSTD_frameHeader zfh; + size_t const err = ZSTD_getFrameHeader(&zfh, src, srcSize); + if (ZSTD_isError(err)) return err; + RETURN_ERROR_IF(err>0, srcSize_wrong, ""); + RETURN_ERROR_IF(zfh.windowSize > windowSizeMax, + frameParameter_windowTooLarge, ""); + return ZSTD_estimateDStreamSize((size_t)zfh.windowSize); +} + + +/* ***** Decompression ***** */ + +static int ZSTD_DCtx_isOverflow(ZSTD_DStream* zds, size_t const neededInBuffSize, size_t const neededOutBuffSize) +{ + return (zds->inBuffSize + zds->outBuffSize) >= (neededInBuffSize + neededOutBuffSize) * ZSTD_WORKSPACETOOLARGE_FACTOR; +} + +static void ZSTD_DCtx_updateOversizedDuration(ZSTD_DStream* zds, size_t const neededInBuffSize, size_t const neededOutBuffSize) +{ + if (ZSTD_DCtx_isOverflow(zds, neededInBuffSize, neededOutBuffSize)) + zds->oversizedDuration++; + else + zds->oversizedDuration = 0; +} + +static int ZSTD_DCtx_isOversizedTooLong(ZSTD_DStream* zds) +{ + return zds->oversizedDuration >= ZSTD_WORKSPACETOOLARGE_MAXDURATION; +} + +/* Checks that the output buffer hasn't changed if ZSTD_obm_stable is used. */ +static size_t ZSTD_checkOutBuffer(ZSTD_DStream const* zds, ZSTD_outBuffer const* output) +{ + ZSTD_outBuffer const expect = zds->expectedOutBuffer; + /* No requirement when ZSTD_obm_stable is not enabled. */ + if (zds->outBufferMode != ZSTD_bm_stable) + return 0; + /* Any buffer is allowed in zdss_init, this must be the same for every other call until + * the context is reset. + */ + if (zds->streamStage == zdss_init) + return 0; + /* The buffer must match our expectation exactly. */ + if (expect.dst == output->dst && expect.pos == output->pos && expect.size == output->size) + return 0; + RETURN_ERROR(dstBuffer_wrong, "ZSTD_d_stableOutBuffer enabled but output differs!"); +} + +/* Calls ZSTD_decompressContinue() with the right parameters for ZSTD_decompressStream() + * and updates the stage and the output buffer state. This call is extracted so it can be + * used both when reading directly from the ZSTD_inBuffer, and in buffered input mode. + * NOTE: You must break after calling this function since the streamStage is modified. + */ +static size_t ZSTD_decompressContinueStream( + ZSTD_DStream* zds, char** op, char* oend, + void const* src, size_t srcSize) { + int const isSkipFrame = ZSTD_isSkipFrame(zds); + if (zds->outBufferMode == ZSTD_bm_buffered) { + size_t const dstSize = isSkipFrame ? 0 : zds->outBuffSize - zds->outStart; + size_t const decodedSize = ZSTD_decompressContinue(zds, + zds->outBuff + zds->outStart, dstSize, src, srcSize); + FORWARD_IF_ERROR(decodedSize, ""); + if (!decodedSize && !isSkipFrame) { + zds->streamStage = zdss_read; + } else { + zds->outEnd = zds->outStart + decodedSize; + zds->streamStage = zdss_flush; + } + } else { + /* Write directly into the output buffer */ + size_t const dstSize = isSkipFrame ? 0 : (size_t)(oend - *op); + size_t const decodedSize = ZSTD_decompressContinue(zds, *op, dstSize, src, srcSize); + FORWARD_IF_ERROR(decodedSize, ""); + *op += decodedSize; + /* Flushing is not needed. */ + zds->streamStage = zdss_read; + assert(*op <= oend); + assert(zds->outBufferMode == ZSTD_bm_stable); + } + return 0; +} + +size_t ZSTD_decompressStream(ZSTD_DStream* zds, ZSTD_outBuffer* output, ZSTD_inBuffer* input) +{ + const char* const src = (const char*)input->src; + const char* const istart = input->pos != 0 ? src + input->pos : src; + const char* const iend = input->size != 0 ? src + input->size : src; + const char* ip = istart; + char* const dst = (char*)output->dst; + char* const ostart = output->pos != 0 ? dst + output->pos : dst; + char* const oend = output->size != 0 ? dst + output->size : dst; + char* op = ostart; + U32 someMoreWork = 1; + + DEBUGLOG(5, "ZSTD_decompressStream"); + RETURN_ERROR_IF( + input->pos > input->size, + srcSize_wrong, + "forbidden. in: pos: %u vs size: %u", + (U32)input->pos, (U32)input->size); + RETURN_ERROR_IF( + output->pos > output->size, + dstSize_tooSmall, + "forbidden. out: pos: %u vs size: %u", + (U32)output->pos, (U32)output->size); + DEBUGLOG(5, "input size : %u", (U32)(input->size - input->pos)); + FORWARD_IF_ERROR(ZSTD_checkOutBuffer(zds, output), ""); + + while (someMoreWork) { + switch(zds->streamStage) + { + case zdss_init : + DEBUGLOG(5, "stage zdss_init => transparent reset "); + zds->streamStage = zdss_loadHeader; + zds->lhSize = zds->inPos = zds->outStart = zds->outEnd = 0; + zds->legacyVersion = 0; + zds->hostageByte = 0; + zds->expectedOutBuffer = *output; + ZSTD_FALLTHROUGH; + + case zdss_loadHeader : + DEBUGLOG(5, "stage zdss_loadHeader (srcSize : %u)", (U32)(iend - ip)); + { size_t const hSize = ZSTD_getFrameHeader_advanced(&zds->fParams, zds->headerBuffer, zds->lhSize, zds->format); + if (zds->refMultipleDDicts && zds->ddictSet) { + ZSTD_DCtx_selectFrameDDict(zds); + } + DEBUGLOG(5, "header size : %u", (U32)hSize); + if (ZSTD_isError(hSize)) { + return hSize; /* error */ + } + if (hSize != 0) { /* need more input */ + size_t const toLoad = hSize - zds->lhSize; /* if hSize!=0, hSize > zds->lhSize */ + size_t const remainingInput = (size_t)(iend-ip); + assert(iend >= ip); + if (toLoad > remainingInput) { /* not enough input to load full header */ + if (remainingInput > 0) { + ZSTD_memcpy(zds->headerBuffer + zds->lhSize, ip, remainingInput); + zds->lhSize += remainingInput; + } + input->pos = input->size; + return (MAX((size_t)ZSTD_FRAMEHEADERSIZE_MIN(zds->format), hSize) - zds->lhSize) + ZSTD_blockHeaderSize; /* remaining header bytes + next block header */ + } + assert(ip != NULL); + ZSTD_memcpy(zds->headerBuffer + zds->lhSize, ip, toLoad); zds->lhSize = hSize; ip += toLoad; + break; + } } + + /* check for single-pass mode opportunity */ + if (zds->fParams.frameContentSize != ZSTD_CONTENTSIZE_UNKNOWN + && zds->fParams.frameType != ZSTD_skippableFrame + && (U64)(size_t)(oend-op) >= zds->fParams.frameContentSize) { + size_t const cSize = ZSTD_findFrameCompressedSize(istart, (size_t)(iend-istart)); + if (cSize <= (size_t)(iend-istart)) { + /* shortcut : using single-pass mode */ + size_t const decompressedSize = ZSTD_decompress_usingDDict(zds, op, (size_t)(oend-op), istart, cSize, ZSTD_getDDict(zds)); + if (ZSTD_isError(decompressedSize)) return decompressedSize; + DEBUGLOG(4, "shortcut to single-pass ZSTD_decompress_usingDDict()") + ip = istart + cSize; + op += decompressedSize; + zds->expected = 0; + zds->streamStage = zdss_init; + someMoreWork = 0; + break; + } } + + /* Check output buffer is large enough for ZSTD_odm_stable. */ + if (zds->outBufferMode == ZSTD_bm_stable + && zds->fParams.frameType != ZSTD_skippableFrame + && zds->fParams.frameContentSize != ZSTD_CONTENTSIZE_UNKNOWN + && (U64)(size_t)(oend-op) < zds->fParams.frameContentSize) { + RETURN_ERROR(dstSize_tooSmall, "ZSTD_obm_stable passed but ZSTD_outBuffer is too small"); + } + + /* Consume header (see ZSTDds_decodeFrameHeader) */ + DEBUGLOG(4, "Consume header"); + FORWARD_IF_ERROR(ZSTD_decompressBegin_usingDDict(zds, ZSTD_getDDict(zds)), ""); + + if ((MEM_readLE32(zds->headerBuffer) & ZSTD_MAGIC_SKIPPABLE_MASK) == ZSTD_MAGIC_SKIPPABLE_START) { /* skippable frame */ + zds->expected = MEM_readLE32(zds->headerBuffer + ZSTD_FRAMEIDSIZE); + zds->stage = ZSTDds_skipFrame; + } else { + FORWARD_IF_ERROR(ZSTD_decodeFrameHeader(zds, zds->headerBuffer, zds->lhSize), ""); + zds->expected = ZSTD_blockHeaderSize; + zds->stage = ZSTDds_decodeBlockHeader; + } + + /* control buffer memory usage */ + DEBUGLOG(4, "Control max memory usage (%u KB <= max %u KB)", + (U32)(zds->fParams.windowSize >>10), + (U32)(zds->maxWindowSize >> 10) ); + zds->fParams.windowSize = MAX(zds->fParams.windowSize, 1U << ZSTD_WINDOWLOG_ABSOLUTEMIN); + RETURN_ERROR_IF(zds->fParams.windowSize > zds->maxWindowSize, + frameParameter_windowTooLarge, ""); + + /* Adapt buffer sizes to frame header instructions */ + { size_t const neededInBuffSize = MAX(zds->fParams.blockSizeMax, 4 /* frame checksum */); + size_t const neededOutBuffSize = zds->outBufferMode == ZSTD_bm_buffered + ? ZSTD_decodingBufferSize_min(zds->fParams.windowSize, zds->fParams.frameContentSize) + : 0; + + ZSTD_DCtx_updateOversizedDuration(zds, neededInBuffSize, neededOutBuffSize); + + { int const tooSmall = (zds->inBuffSize < neededInBuffSize) || (zds->outBuffSize < neededOutBuffSize); + int const tooLarge = ZSTD_DCtx_isOversizedTooLong(zds); + + if (tooSmall || tooLarge) { + size_t const bufferSize = neededInBuffSize + neededOutBuffSize; + DEBUGLOG(4, "inBuff : from %u to %u", + (U32)zds->inBuffSize, (U32)neededInBuffSize); + DEBUGLOG(4, "outBuff : from %u to %u", + (U32)zds->outBuffSize, (U32)neededOutBuffSize); + if (zds->staticSize) { /* static DCtx */ + DEBUGLOG(4, "staticSize : %u", (U32)zds->staticSize); + assert(zds->staticSize >= sizeof(ZSTD_DCtx)); /* controlled at init */ + RETURN_ERROR_IF( + bufferSize > zds->staticSize - sizeof(ZSTD_DCtx), + memory_allocation, ""); + } else { + ZSTD_customFree(zds->inBuff, zds->customMem); + zds->inBuffSize = 0; + zds->outBuffSize = 0; + zds->inBuff = (char*)ZSTD_customMalloc(bufferSize, zds->customMem); + RETURN_ERROR_IF(zds->inBuff == NULL, memory_allocation, ""); + } + zds->inBuffSize = neededInBuffSize; + zds->outBuff = zds->inBuff + zds->inBuffSize; + zds->outBuffSize = neededOutBuffSize; + } } } + zds->streamStage = zdss_read; + ZSTD_FALLTHROUGH; + + case zdss_read: + DEBUGLOG(5, "stage zdss_read"); + { size_t const neededInSize = ZSTD_nextSrcSizeToDecompressWithInputSize(zds, (size_t)(iend - ip)); + DEBUGLOG(5, "neededInSize = %u", (U32)neededInSize); + if (neededInSize==0) { /* end of frame */ + zds->streamStage = zdss_init; + someMoreWork = 0; + break; + } + if ((size_t)(iend-ip) >= neededInSize) { /* decode directly from src */ + FORWARD_IF_ERROR(ZSTD_decompressContinueStream(zds, &op, oend, ip, neededInSize), ""); + ip += neededInSize; + /* Function modifies the stage so we must break */ + break; + } } + if (ip==iend) { someMoreWork = 0; break; } /* no more input */ + zds->streamStage = zdss_load; + ZSTD_FALLTHROUGH; + + case zdss_load: + { size_t const neededInSize = ZSTD_nextSrcSizeToDecompress(zds); + size_t const toLoad = neededInSize - zds->inPos; + int const isSkipFrame = ZSTD_isSkipFrame(zds); + size_t loadedSize; + /* At this point we shouldn't be decompressing a block that we can stream. */ + assert(neededInSize == ZSTD_nextSrcSizeToDecompressWithInputSize(zds, iend - ip)); + if (isSkipFrame) { + loadedSize = MIN(toLoad, (size_t)(iend-ip)); + } else { + RETURN_ERROR_IF(toLoad > zds->inBuffSize - zds->inPos, + corruption_detected, + "should never happen"); + loadedSize = ZSTD_limitCopy(zds->inBuff + zds->inPos, toLoad, ip, (size_t)(iend-ip)); + } + ip += loadedSize; + zds->inPos += loadedSize; + if (loadedSize < toLoad) { someMoreWork = 0; break; } /* not enough input, wait for more */ + + /* decode loaded input */ + zds->inPos = 0; /* input is consumed */ + FORWARD_IF_ERROR(ZSTD_decompressContinueStream(zds, &op, oend, zds->inBuff, neededInSize), ""); + /* Function modifies the stage so we must break */ + break; + } + case zdss_flush: + { size_t const toFlushSize = zds->outEnd - zds->outStart; + size_t const flushedSize = ZSTD_limitCopy(op, (size_t)(oend-op), zds->outBuff + zds->outStart, toFlushSize); + op += flushedSize; + zds->outStart += flushedSize; + if (flushedSize == toFlushSize) { /* flush completed */ + zds->streamStage = zdss_read; + if ( (zds->outBuffSize < zds->fParams.frameContentSize) + && (zds->outStart + zds->fParams.blockSizeMax > zds->outBuffSize) ) { + DEBUGLOG(5, "restart filling outBuff from beginning (left:%i, needed:%u)", + (int)(zds->outBuffSize - zds->outStart), + (U32)zds->fParams.blockSizeMax); + zds->outStart = zds->outEnd = 0; + } + break; + } } + /* cannot complete flush */ + someMoreWork = 0; + break; + + default: + assert(0); /* impossible */ + RETURN_ERROR(GENERIC, "impossible to reach"); /* some compiler require default to do something */ + } } + + /* result */ + input->pos = (size_t)(ip - (const char*)(input->src)); + output->pos = (size_t)(op - (char*)(output->dst)); + + /* Update the expected output buffer for ZSTD_obm_stable. */ + zds->expectedOutBuffer = *output; + + if ((ip==istart) && (op==ostart)) { /* no forward progress */ + zds->noForwardProgress ++; + if (zds->noForwardProgress >= ZSTD_NO_FORWARD_PROGRESS_MAX) { + RETURN_ERROR_IF(op==oend, dstSize_tooSmall, ""); + RETURN_ERROR_IF(ip==iend, srcSize_wrong, ""); + assert(0); + } + } else { + zds->noForwardProgress = 0; + } + { size_t nextSrcSizeHint = ZSTD_nextSrcSizeToDecompress(zds); + if (!nextSrcSizeHint) { /* frame fully decoded */ + if (zds->outEnd == zds->outStart) { /* output fully flushed */ + if (zds->hostageByte) { + if (input->pos >= input->size) { + /* can't release hostage (not present) */ + zds->streamStage = zdss_read; + return 1; + } + input->pos++; /* release hostage */ + } /* zds->hostageByte */ + return 0; + } /* zds->outEnd == zds->outStart */ + if (!zds->hostageByte) { /* output not fully flushed; keep last byte as hostage; will be released when all output is flushed */ + input->pos--; /* note : pos > 0, otherwise, impossible to finish reading last block */ + zds->hostageByte=1; + } + return 1; + } /* nextSrcSizeHint==0 */ + nextSrcSizeHint += ZSTD_blockHeaderSize * (ZSTD_nextInputType(zds) == ZSTDnit_block); /* preload header of next block */ + assert(zds->inPos <= nextSrcSizeHint); + nextSrcSizeHint -= zds->inPos; /* part already loaded*/ + return nextSrcSizeHint; + } +} + +size_t ZSTD_decompressStream_simpleArgs ( + ZSTD_DCtx* dctx, + void* dst, size_t dstCapacity, size_t* dstPos, + const void* src, size_t srcSize, size_t* srcPos) +{ + ZSTD_outBuffer output = { dst, dstCapacity, *dstPos }; + ZSTD_inBuffer input = { src, srcSize, *srcPos }; + /* ZSTD_compress_generic() will check validity of dstPos and srcPos */ + size_t const cErr = ZSTD_decompressStream(dctx, &output, &input); + *dstPos = output.pos; + *srcPos = input.pos; + return cErr; +} diff --git a/lib/zstd/decompress/zstd_decompress_block.c b/lib/zstd/decompress/zstd_decompress_block.c new file mode 100644 index 000000000000..2d101d9a842e --- /dev/null +++ b/lib/zstd/decompress/zstd_decompress_block.c @@ -0,0 +1,1540 @@ +/* + * Copyright (c) Yann Collet, Facebook, Inc. + * All rights reserved. + * + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). + * You may select, at your option, one of the above-listed licenses. + */ + +/* zstd_decompress_block : + * this module takes care of decompressing _compressed_ block */ + +/*-******************************************************* +* Dependencies +*********************************************************/ +#include "../common/zstd_deps.h" /* ZSTD_memcpy, ZSTD_memmove, ZSTD_memset */ +#include "../common/compiler.h" /* prefetch */ +#include "../common/cpu.h" /* bmi2 */ +#include "../common/mem.h" /* low level memory routines */ +#define FSE_STATIC_LINKING_ONLY +#include "../common/fse.h" +#define HUF_STATIC_LINKING_ONLY +#include "../common/huf.h" +#include "../common/zstd_internal.h" +#include "zstd_decompress_internal.h" /* ZSTD_DCtx */ +#include "zstd_ddict.h" /* ZSTD_DDictDictContent */ +#include "zstd_decompress_block.h" + +/*_******************************************************* +* Macros +**********************************************************/ + +/* These two optional macros force the use one way or another of the two + * ZSTD_decompressSequences implementations. You can't force in both directions + * at the same time. + */ +#if defined(ZSTD_FORCE_DECOMPRESS_SEQUENCES_SHORT) && \ + defined(ZSTD_FORCE_DECOMPRESS_SEQUENCES_LONG) +#error "Cannot force the use of the short and the long ZSTD_decompressSequences variants!" +#endif + + +/*_******************************************************* +* Memory operations +**********************************************************/ +static void ZSTD_copy4(void* dst, const void* src) { ZSTD_memcpy(dst, src, 4); } + + +/*-************************************************************* + * Block decoding + ***************************************************************/ + +/*! ZSTD_getcBlockSize() : + * Provides the size of compressed block from block header `src` */ +size_t ZSTD_getcBlockSize(const void* src, size_t srcSize, + blockProperties_t* bpPtr) +{ + RETURN_ERROR_IF(srcSize < ZSTD_blockHeaderSize, srcSize_wrong, ""); + + { U32 const cBlockHeader = MEM_readLE24(src); + U32 const cSize = cBlockHeader >> 3; + bpPtr->lastBlock = cBlockHeader & 1; + bpPtr->blockType = (blockType_e)((cBlockHeader >> 1) & 3); + bpPtr->origSize = cSize; /* only useful for RLE */ + if (bpPtr->blockType == bt_rle) return 1; + RETURN_ERROR_IF(bpPtr->blockType == bt_reserved, corruption_detected, ""); + return cSize; + } +} + + +/* Hidden declaration for fullbench */ +size_t ZSTD_decodeLiteralsBlock(ZSTD_DCtx* dctx, + const void* src, size_t srcSize); +/*! ZSTD_decodeLiteralsBlock() : + * @return : nb of bytes read from src (< srcSize ) + * note : symbol not declared but exposed for fullbench */ +size_t ZSTD_decodeLiteralsBlock(ZSTD_DCtx* dctx, + const void* src, size_t srcSize) /* note : srcSize < BLOCKSIZE */ +{ + DEBUGLOG(5, "ZSTD_decodeLiteralsBlock"); + RETURN_ERROR_IF(srcSize < MIN_CBLOCK_SIZE, corruption_detected, ""); + + { const BYTE* const istart = (const BYTE*) src; + symbolEncodingType_e const litEncType = (symbolEncodingType_e)(istart[0] & 3); + + switch(litEncType) + { + case set_repeat: + DEBUGLOG(5, "set_repeat flag : re-using stats from previous compressed literals block"); + RETURN_ERROR_IF(dctx->litEntropy==0, dictionary_corrupted, ""); + ZSTD_FALLTHROUGH; + + case set_compressed: + RETURN_ERROR_IF(srcSize < 5, corruption_detected, "srcSize >= MIN_CBLOCK_SIZE == 3; here we need up to 5 for case 3"); + { size_t lhSize, litSize, litCSize; + U32 singleStream=0; + U32 const lhlCode = (istart[0] >> 2) & 3; + U32 const lhc = MEM_readLE32(istart); + size_t hufSuccess; + switch(lhlCode) + { + case 0: case 1: default: /* note : default is impossible, since lhlCode into [0..3] */ + /* 2 - 2 - 10 - 10 */ + singleStream = !lhlCode; + lhSize = 3; + litSize = (lhc >> 4) & 0x3FF; + litCSize = (lhc >> 14) & 0x3FF; + break; + case 2: + /* 2 - 2 - 14 - 14 */ + lhSize = 4; + litSize = (lhc >> 4) & 0x3FFF; + litCSize = lhc >> 18; + break; + case 3: + /* 2 - 2 - 18 - 18 */ + lhSize = 5; + litSize = (lhc >> 4) & 0x3FFFF; + litCSize = (lhc >> 22) + ((size_t)istart[4] << 10); + break; + } + RETURN_ERROR_IF(litSize > ZSTD_BLOCKSIZE_MAX, corruption_detected, ""); + RETURN_ERROR_IF(litCSize + lhSize > srcSize, corruption_detected, ""); + + /* prefetch huffman table if cold */ + if (dctx->ddictIsCold && (litSize > 768 /* heuristic */)) { + PREFETCH_AREA(dctx->HUFptr, sizeof(dctx->entropy.hufTable)); + } + + if (litEncType==set_repeat) { + if (singleStream) { + hufSuccess = HUF_decompress1X_usingDTable_bmi2( + dctx->litBuffer, litSize, istart+lhSize, litCSize, + dctx->HUFptr, dctx->bmi2); + } else { + hufSuccess = HUF_decompress4X_usingDTable_bmi2( + dctx->litBuffer, litSize, istart+lhSize, litCSize, + dctx->HUFptr, dctx->bmi2); + } + } else { + if (singleStream) { +#if defined(HUF_FORCE_DECOMPRESS_X2) + hufSuccess = HUF_decompress1X_DCtx_wksp( + dctx->entropy.hufTable, dctx->litBuffer, litSize, + istart+lhSize, litCSize, dctx->workspace, + sizeof(dctx->workspace)); +#else + hufSuccess = HUF_decompress1X1_DCtx_wksp_bmi2( + dctx->entropy.hufTable, dctx->litBuffer, litSize, + istart+lhSize, litCSize, dctx->workspace, + sizeof(dctx->workspace), dctx->bmi2); +#endif + } else { + hufSuccess = HUF_decompress4X_hufOnly_wksp_bmi2( + dctx->entropy.hufTable, dctx->litBuffer, litSize, + istart+lhSize, litCSize, dctx->workspace, + sizeof(dctx->workspace), dctx->bmi2); + } + } + + RETURN_ERROR_IF(HUF_isError(hufSuccess), corruption_detected, ""); + + dctx->litPtr = dctx->litBuffer; + dctx->litSize = litSize; + dctx->litEntropy = 1; + if (litEncType==set_compressed) dctx->HUFptr = dctx->entropy.hufTable; + ZSTD_memset(dctx->litBuffer + dctx->litSize, 0, WILDCOPY_OVERLENGTH); + return litCSize + lhSize; + } + + case set_basic: + { size_t litSize, lhSize; + U32 const lhlCode = ((istart[0]) >> 2) & 3; + switch(lhlCode) + { + case 0: case 2: default: /* note : default is impossible, since lhlCode into [0..3] */ + lhSize = 1; + litSize = istart[0] >> 3; + break; + case 1: + lhSize = 2; + litSize = MEM_readLE16(istart) >> 4; + break; + case 3: + lhSize = 3; + litSize = MEM_readLE24(istart) >> 4; + break; + } + + if (lhSize+litSize+WILDCOPY_OVERLENGTH > srcSize) { /* risk reading beyond src buffer with wildcopy */ + RETURN_ERROR_IF(litSize+lhSize > srcSize, corruption_detected, ""); + ZSTD_memcpy(dctx->litBuffer, istart+lhSize, litSize); + dctx->litPtr = dctx->litBuffer; + dctx->litSize = litSize; + ZSTD_memset(dctx->litBuffer + dctx->litSize, 0, WILDCOPY_OVERLENGTH); + return lhSize+litSize; + } + /* direct reference into compressed stream */ + dctx->litPtr = istart+lhSize; + dctx->litSize = litSize; + return lhSize+litSize; + } + + case set_rle: + { U32 const lhlCode = ((istart[0]) >> 2) & 3; + size_t litSize, lhSize; + switch(lhlCode) + { + case 0: case 2: default: /* note : default is impossible, since lhlCode into [0..3] */ + lhSize = 1; + litSize = istart[0] >> 3; + break; + case 1: + lhSize = 2; + litSize = MEM_readLE16(istart) >> 4; + break; + case 3: + lhSize = 3; + litSize = MEM_readLE24(istart) >> 4; + RETURN_ERROR_IF(srcSize<4, corruption_detected, "srcSize >= MIN_CBLOCK_SIZE == 3; here we need lhSize+1 = 4"); + break; + } + RETURN_ERROR_IF(litSize > ZSTD_BLOCKSIZE_MAX, corruption_detected, ""); + ZSTD_memset(dctx->litBuffer, istart[lhSize], litSize + WILDCOPY_OVERLENGTH); + dctx->litPtr = dctx->litBuffer; + dctx->litSize = litSize; + return lhSize+1; + } + default: + RETURN_ERROR(corruption_detected, "impossible"); + } + } +} + +/* Default FSE distribution tables. + * These are pre-calculated FSE decoding tables using default distributions as defined in specification : + * https://github.com/facebook/zstd/blob/release/doc/zstd_compression_format.md#default-distributions + * They were generated programmatically with following method : + * - start from default distributions, present in /lib/common/zstd_internal.h + * - generate tables normally, using ZSTD_buildFSETable() + * - printout the content of tables + * - pretify output, report below, test with fuzzer to ensure it's correct */ + +/* Default FSE distribution table for Literal Lengths */ +static const ZSTD_seqSymbol LL_defaultDTable[(1<tableLog = 0; + DTableH->fastMode = 0; + + cell->nbBits = 0; + cell->nextState = 0; + assert(nbAddBits < 255); + cell->nbAdditionalBits = (BYTE)nbAddBits; + cell->baseValue = baseValue; +} + + +/* ZSTD_buildFSETable() : + * generate FSE decoding table for one symbol (ll, ml or off) + * cannot fail if input is valid => + * all inputs are presumed validated at this stage */ +FORCE_INLINE_TEMPLATE +void ZSTD_buildFSETable_body(ZSTD_seqSymbol* dt, + const short* normalizedCounter, unsigned maxSymbolValue, + const U32* baseValue, const U32* nbAdditionalBits, + unsigned tableLog, void* wksp, size_t wkspSize) +{ + ZSTD_seqSymbol* const tableDecode = dt+1; + U32 const maxSV1 = maxSymbolValue + 1; + U32 const tableSize = 1 << tableLog; + + U16* symbolNext = (U16*)wksp; + BYTE* spread = (BYTE*)(symbolNext + MaxSeq + 1); + U32 highThreshold = tableSize - 1; + + + /* Sanity Checks */ + assert(maxSymbolValue <= MaxSeq); + assert(tableLog <= MaxFSELog); + assert(wkspSize >= ZSTD_BUILD_FSE_TABLE_WKSP_SIZE); + (void)wkspSize; + /* Init, lay down lowprob symbols */ + { ZSTD_seqSymbol_header DTableH; + DTableH.tableLog = tableLog; + DTableH.fastMode = 1; + { S16 const largeLimit= (S16)(1 << (tableLog-1)); + U32 s; + for (s=0; s= largeLimit) DTableH.fastMode=0; + assert(normalizedCounter[s]>=0); + symbolNext[s] = (U16)normalizedCounter[s]; + } } } + ZSTD_memcpy(dt, &DTableH, sizeof(DTableH)); + } + + /* Spread symbols */ + assert(tableSize <= 512); + /* Specialized symbol spreading for the case when there are + * no low probability (-1 count) symbols. When compressing + * small blocks we avoid low probability symbols to hit this + * case, since header decoding speed matters more. + */ + if (highThreshold == tableSize - 1) { + size_t const tableMask = tableSize-1; + size_t const step = FSE_TABLESTEP(tableSize); + /* First lay down the symbols in order. + * We use a uint64_t to lay down 8 bytes at a time. This reduces branch + * misses since small blocks generally have small table logs, so nearly + * all symbols have counts <= 8. We ensure we have 8 bytes at the end of + * our buffer to handle the over-write. + */ + { + U64 const add = 0x0101010101010101ull; + size_t pos = 0; + U64 sv = 0; + U32 s; + for (s=0; s highThreshold) position = (position + step) & tableMask; /* lowprob area */ + } } + assert(position == 0); /* position must reach all cells once, otherwise normalizedCounter is incorrect */ + } + + /* Build Decoding table */ + { + U32 u; + for (u=0; u max, corruption_detected, ""); + { U32 const symbol = *(const BYTE*)src; + U32 const baseline = baseValue[symbol]; + U32 const nbBits = nbAdditionalBits[symbol]; + ZSTD_buildSeqTable_rle(DTableSpace, baseline, nbBits); + } + *DTablePtr = DTableSpace; + return 1; + case set_basic : + *DTablePtr = defaultTable; + return 0; + case set_repeat: + RETURN_ERROR_IF(!flagRepeatTable, corruption_detected, ""); + /* prefetch FSE table if used */ + if (ddictIsCold && (nbSeq > 24 /* heuristic */)) { + const void* const pStart = *DTablePtr; + size_t const pSize = sizeof(ZSTD_seqSymbol) * (SEQSYMBOL_TABLE_SIZE(maxLog)); + PREFETCH_AREA(pStart, pSize); + } + return 0; + case set_compressed : + { unsigned tableLog; + S16 norm[MaxSeq+1]; + size_t const headerSize = FSE_readNCount(norm, &max, &tableLog, src, srcSize); + RETURN_ERROR_IF(FSE_isError(headerSize), corruption_detected, ""); + RETURN_ERROR_IF(tableLog > maxLog, corruption_detected, ""); + ZSTD_buildFSETable(DTableSpace, norm, max, baseValue, nbAdditionalBits, tableLog, wksp, wkspSize, bmi2); + *DTablePtr = DTableSpace; + return headerSize; + } + default : + assert(0); + RETURN_ERROR(GENERIC, "impossible"); + } +} + +size_t ZSTD_decodeSeqHeaders(ZSTD_DCtx* dctx, int* nbSeqPtr, + const void* src, size_t srcSize) +{ + const BYTE* const istart = (const BYTE*)src; + const BYTE* const iend = istart + srcSize; + const BYTE* ip = istart; + int nbSeq; + DEBUGLOG(5, "ZSTD_decodeSeqHeaders"); + + /* check */ + RETURN_ERROR_IF(srcSize < MIN_SEQUENCES_SIZE, srcSize_wrong, ""); + + /* SeqHead */ + nbSeq = *ip++; + if (!nbSeq) { + *nbSeqPtr=0; + RETURN_ERROR_IF(srcSize != 1, srcSize_wrong, ""); + return 1; + } + if (nbSeq > 0x7F) { + if (nbSeq == 0xFF) { + RETURN_ERROR_IF(ip+2 > iend, srcSize_wrong, ""); + nbSeq = MEM_readLE16(ip) + LONGNBSEQ; + ip+=2; + } else { + RETURN_ERROR_IF(ip >= iend, srcSize_wrong, ""); + nbSeq = ((nbSeq-0x80)<<8) + *ip++; + } + } + *nbSeqPtr = nbSeq; + + /* FSE table descriptors */ + RETURN_ERROR_IF(ip+1 > iend, srcSize_wrong, ""); /* minimum possible size: 1 byte for symbol encoding types */ + { symbolEncodingType_e const LLtype = (symbolEncodingType_e)(*ip >> 6); + symbolEncodingType_e const OFtype = (symbolEncodingType_e)((*ip >> 4) & 3); + symbolEncodingType_e const MLtype = (symbolEncodingType_e)((*ip >> 2) & 3); + ip++; + + /* Build DTables */ + { size_t const llhSize = ZSTD_buildSeqTable(dctx->entropy.LLTable, &dctx->LLTptr, + LLtype, MaxLL, LLFSELog, + ip, iend-ip, + LL_base, LL_bits, + LL_defaultDTable, dctx->fseEntropy, + dctx->ddictIsCold, nbSeq, + dctx->workspace, sizeof(dctx->workspace), + dctx->bmi2); + RETURN_ERROR_IF(ZSTD_isError(llhSize), corruption_detected, "ZSTD_buildSeqTable failed"); + ip += llhSize; + } + + { size_t const ofhSize = ZSTD_buildSeqTable(dctx->entropy.OFTable, &dctx->OFTptr, + OFtype, MaxOff, OffFSELog, + ip, iend-ip, + OF_base, OF_bits, + OF_defaultDTable, dctx->fseEntropy, + dctx->ddictIsCold, nbSeq, + dctx->workspace, sizeof(dctx->workspace), + dctx->bmi2); + RETURN_ERROR_IF(ZSTD_isError(ofhSize), corruption_detected, "ZSTD_buildSeqTable failed"); + ip += ofhSize; + } + + { size_t const mlhSize = ZSTD_buildSeqTable(dctx->entropy.MLTable, &dctx->MLTptr, + MLtype, MaxML, MLFSELog, + ip, iend-ip, + ML_base, ML_bits, + ML_defaultDTable, dctx->fseEntropy, + dctx->ddictIsCold, nbSeq, + dctx->workspace, sizeof(dctx->workspace), + dctx->bmi2); + RETURN_ERROR_IF(ZSTD_isError(mlhSize), corruption_detected, "ZSTD_buildSeqTable failed"); + ip += mlhSize; + } + } + + return ip-istart; +} + + +typedef struct { + size_t litLength; + size_t matchLength; + size_t offset; + const BYTE* match; +} seq_t; + +typedef struct { + size_t state; + const ZSTD_seqSymbol* table; +} ZSTD_fseState; + +typedef struct { + BIT_DStream_t DStream; + ZSTD_fseState stateLL; + ZSTD_fseState stateOffb; + ZSTD_fseState stateML; + size_t prevOffset[ZSTD_REP_NUM]; + const BYTE* prefixStart; + const BYTE* dictEnd; + size_t pos; +} seqState_t; + +/*! ZSTD_overlapCopy8() : + * Copies 8 bytes from ip to op and updates op and ip where ip <= op. + * If the offset is < 8 then the offset is spread to at least 8 bytes. + * + * Precondition: *ip <= *op + * Postcondition: *op - *op >= 8 + */ +HINT_INLINE void ZSTD_overlapCopy8(BYTE** op, BYTE const** ip, size_t offset) { + assert(*ip <= *op); + if (offset < 8) { + /* close range match, overlap */ + static const U32 dec32table[] = { 0, 1, 2, 1, 4, 4, 4, 4 }; /* added */ + static const int dec64table[] = { 8, 8, 8, 7, 8, 9,10,11 }; /* subtracted */ + int const sub2 = dec64table[offset]; + (*op)[0] = (*ip)[0]; + (*op)[1] = (*ip)[1]; + (*op)[2] = (*ip)[2]; + (*op)[3] = (*ip)[3]; + *ip += dec32table[offset]; + ZSTD_copy4(*op+4, *ip); + *ip -= sub2; + } else { + ZSTD_copy8(*op, *ip); + } + *ip += 8; + *op += 8; + assert(*op - *ip >= 8); +} + +/*! ZSTD_safecopy() : + * Specialized version of memcpy() that is allowed to READ up to WILDCOPY_OVERLENGTH past the input buffer + * and write up to 16 bytes past oend_w (op >= oend_w is allowed). + * This function is only called in the uncommon case where the sequence is near the end of the block. It + * should be fast for a single long sequence, but can be slow for several short sequences. + * + * @param ovtype controls the overlap detection + * - ZSTD_no_overlap: The source and destination are guaranteed to be at least WILDCOPY_VECLEN bytes apart. + * - ZSTD_overlap_src_before_dst: The src and dst may overlap and may be any distance apart. + * The src buffer must be before the dst buffer. + */ +static void ZSTD_safecopy(BYTE* op, BYTE* const oend_w, BYTE const* ip, ptrdiff_t length, ZSTD_overlap_e ovtype) { + ptrdiff_t const diff = op - ip; + BYTE* const oend = op + length; + + assert((ovtype == ZSTD_no_overlap && (diff <= -8 || diff >= 8 || op >= oend_w)) || + (ovtype == ZSTD_overlap_src_before_dst && diff >= 0)); + + if (length < 8) { + /* Handle short lengths. */ + while (op < oend) *op++ = *ip++; + return; + } + if (ovtype == ZSTD_overlap_src_before_dst) { + /* Copy 8 bytes and ensure the offset >= 8 when there can be overlap. */ + assert(length >= 8); + ZSTD_overlapCopy8(&op, &ip, diff); + assert(op - ip >= 8); + assert(op <= oend); + } + + if (oend <= oend_w) { + /* No risk of overwrite. */ + ZSTD_wildcopy(op, ip, length, ovtype); + return; + } + if (op <= oend_w) { + /* Wildcopy until we get close to the end. */ + assert(oend > oend_w); + ZSTD_wildcopy(op, ip, oend_w - op, ovtype); + ip += oend_w - op; + op = oend_w; + } + /* Handle the leftovers. */ + while (op < oend) *op++ = *ip++; +} + +/* ZSTD_execSequenceEnd(): + * This version handles cases that are near the end of the output buffer. It requires + * more careful checks to make sure there is no overflow. By separating out these hard + * and unlikely cases, we can speed up the common cases. + * + * NOTE: This function needs to be fast for a single long sequence, but doesn't need + * to be optimized for many small sequences, since those fall into ZSTD_execSequence(). + */ +FORCE_NOINLINE +size_t ZSTD_execSequenceEnd(BYTE* op, + BYTE* const oend, seq_t sequence, + const BYTE** litPtr, const BYTE* const litLimit, + const BYTE* const prefixStart, const BYTE* const virtualStart, const BYTE* const dictEnd) +{ + BYTE* const oLitEnd = op + sequence.litLength; + size_t const sequenceLength = sequence.litLength + sequence.matchLength; + const BYTE* const iLitEnd = *litPtr + sequence.litLength; + const BYTE* match = oLitEnd - sequence.offset; + BYTE* const oend_w = oend - WILDCOPY_OVERLENGTH; + + /* bounds checks : careful of address space overflow in 32-bit mode */ + RETURN_ERROR_IF(sequenceLength > (size_t)(oend - op), dstSize_tooSmall, "last match must fit within dstBuffer"); + RETURN_ERROR_IF(sequence.litLength > (size_t)(litLimit - *litPtr), corruption_detected, "try to read beyond literal buffer"); + assert(op < op + sequenceLength); + assert(oLitEnd < op + sequenceLength); + + /* copy literals */ + ZSTD_safecopy(op, oend_w, *litPtr, sequence.litLength, ZSTD_no_overlap); + op = oLitEnd; + *litPtr = iLitEnd; + + /* copy Match */ + if (sequence.offset > (size_t)(oLitEnd - prefixStart)) { + /* offset beyond prefix */ + RETURN_ERROR_IF(sequence.offset > (size_t)(oLitEnd - virtualStart), corruption_detected, ""); + match = dictEnd - (prefixStart-match); + if (match + sequence.matchLength <= dictEnd) { + ZSTD_memmove(oLitEnd, match, sequence.matchLength); + return sequenceLength; + } + /* span extDict & currentPrefixSegment */ + { size_t const length1 = dictEnd - match; + ZSTD_memmove(oLitEnd, match, length1); + op = oLitEnd + length1; + sequence.matchLength -= length1; + match = prefixStart; + } } + ZSTD_safecopy(op, oend_w, match, sequence.matchLength, ZSTD_overlap_src_before_dst); + return sequenceLength; +} + +HINT_INLINE +size_t ZSTD_execSequence(BYTE* op, + BYTE* const oend, seq_t sequence, + const BYTE** litPtr, const BYTE* const litLimit, + const BYTE* const prefixStart, const BYTE* const virtualStart, const BYTE* const dictEnd) +{ + BYTE* const oLitEnd = op + sequence.litLength; + size_t const sequenceLength = sequence.litLength + sequence.matchLength; + BYTE* const oMatchEnd = op + sequenceLength; /* risk : address space overflow (32-bits) */ + BYTE* const oend_w = oend - WILDCOPY_OVERLENGTH; /* risk : address space underflow on oend=NULL */ + const BYTE* const iLitEnd = *litPtr + sequence.litLength; + const BYTE* match = oLitEnd - sequence.offset; + + assert(op != NULL /* Precondition */); + assert(oend_w < oend /* No underflow */); + /* Handle edge cases in a slow path: + * - Read beyond end of literals + * - Match end is within WILDCOPY_OVERLIMIT of oend + * - 32-bit mode and the match length overflows + */ + if (UNLIKELY( + iLitEnd > litLimit || + oMatchEnd > oend_w || + (MEM_32bits() && (size_t)(oend - op) < sequenceLength + WILDCOPY_OVERLENGTH))) + return ZSTD_execSequenceEnd(op, oend, sequence, litPtr, litLimit, prefixStart, virtualStart, dictEnd); + + /* Assumptions (everything else goes into ZSTD_execSequenceEnd()) */ + assert(op <= oLitEnd /* No overflow */); + assert(oLitEnd < oMatchEnd /* Non-zero match & no overflow */); + assert(oMatchEnd <= oend /* No underflow */); + assert(iLitEnd <= litLimit /* Literal length is in bounds */); + assert(oLitEnd <= oend_w /* Can wildcopy literals */); + assert(oMatchEnd <= oend_w /* Can wildcopy matches */); + + /* Copy Literals: + * Split out litLength <= 16 since it is nearly always true. +1.6% on gcc-9. + * We likely don't need the full 32-byte wildcopy. + */ + assert(WILDCOPY_OVERLENGTH >= 16); + ZSTD_copy16(op, (*litPtr)); + if (UNLIKELY(sequence.litLength > 16)) { + ZSTD_wildcopy(op+16, (*litPtr)+16, sequence.litLength-16, ZSTD_no_overlap); + } + op = oLitEnd; + *litPtr = iLitEnd; /* update for next sequence */ + + /* Copy Match */ + if (sequence.offset > (size_t)(oLitEnd - prefixStart)) { + /* offset beyond prefix -> go into extDict */ + RETURN_ERROR_IF(UNLIKELY(sequence.offset > (size_t)(oLitEnd - virtualStart)), corruption_detected, ""); + match = dictEnd + (match - prefixStart); + if (match + sequence.matchLength <= dictEnd) { + ZSTD_memmove(oLitEnd, match, sequence.matchLength); + return sequenceLength; + } + /* span extDict & currentPrefixSegment */ + { size_t const length1 = dictEnd - match; + ZSTD_memmove(oLitEnd, match, length1); + op = oLitEnd + length1; + sequence.matchLength -= length1; + match = prefixStart; + } } + /* Match within prefix of 1 or more bytes */ + assert(op <= oMatchEnd); + assert(oMatchEnd <= oend_w); + assert(match >= prefixStart); + assert(sequence.matchLength >= 1); + + /* Nearly all offsets are >= WILDCOPY_VECLEN bytes, which means we can use wildcopy + * without overlap checking. + */ + if (LIKELY(sequence.offset >= WILDCOPY_VECLEN)) { + /* We bet on a full wildcopy for matches, since we expect matches to be + * longer than literals (in general). In silesia, ~10% of matches are longer + * than 16 bytes. + */ + ZSTD_wildcopy(op, match, (ptrdiff_t)sequence.matchLength, ZSTD_no_overlap); + return sequenceLength; + } + assert(sequence.offset < WILDCOPY_VECLEN); + + /* Copy 8 bytes and spread the offset to be >= 8. */ + ZSTD_overlapCopy8(&op, &match, sequence.offset); + + /* If the match length is > 8 bytes, then continue with the wildcopy. */ + if (sequence.matchLength > 8) { + assert(op < oMatchEnd); + ZSTD_wildcopy(op, match, (ptrdiff_t)sequence.matchLength-8, ZSTD_overlap_src_before_dst); + } + return sequenceLength; +} + +static void +ZSTD_initFseState(ZSTD_fseState* DStatePtr, BIT_DStream_t* bitD, const ZSTD_seqSymbol* dt) +{ + const void* ptr = dt; + const ZSTD_seqSymbol_header* const DTableH = (const ZSTD_seqSymbol_header*)ptr; + DStatePtr->state = BIT_readBits(bitD, DTableH->tableLog); + DEBUGLOG(6, "ZSTD_initFseState : val=%u using %u bits", + (U32)DStatePtr->state, DTableH->tableLog); + BIT_reloadDStream(bitD); + DStatePtr->table = dt + 1; +} + +FORCE_INLINE_TEMPLATE void +ZSTD_updateFseState(ZSTD_fseState* DStatePtr, BIT_DStream_t* bitD) +{ + ZSTD_seqSymbol const DInfo = DStatePtr->table[DStatePtr->state]; + U32 const nbBits = DInfo.nbBits; + size_t const lowBits = BIT_readBits(bitD, nbBits); + DStatePtr->state = DInfo.nextState + lowBits; +} + +FORCE_INLINE_TEMPLATE void +ZSTD_updateFseStateWithDInfo(ZSTD_fseState* DStatePtr, BIT_DStream_t* bitD, ZSTD_seqSymbol const DInfo) +{ + U32 const nbBits = DInfo.nbBits; + size_t const lowBits = BIT_readBits(bitD, nbBits); + DStatePtr->state = DInfo.nextState + lowBits; +} + +/* We need to add at most (ZSTD_WINDOWLOG_MAX_32 - 1) bits to read the maximum + * offset bits. But we can only read at most (STREAM_ACCUMULATOR_MIN_32 - 1) + * bits before reloading. This value is the maximum number of bytes we read + * after reloading when we are decoding long offsets. + */ +#define LONG_OFFSETS_MAX_EXTRA_BITS_32 \ + (ZSTD_WINDOWLOG_MAX_32 > STREAM_ACCUMULATOR_MIN_32 \ + ? ZSTD_WINDOWLOG_MAX_32 - STREAM_ACCUMULATOR_MIN_32 \ + : 0) + +typedef enum { ZSTD_lo_isRegularOffset, ZSTD_lo_isLongOffset=1 } ZSTD_longOffset_e; +typedef enum { ZSTD_p_noPrefetch=0, ZSTD_p_prefetch=1 } ZSTD_prefetch_e; + +FORCE_INLINE_TEMPLATE seq_t +ZSTD_decodeSequence(seqState_t* seqState, const ZSTD_longOffset_e longOffsets, const ZSTD_prefetch_e prefetch) +{ + seq_t seq; + ZSTD_seqSymbol const llDInfo = seqState->stateLL.table[seqState->stateLL.state]; + ZSTD_seqSymbol const mlDInfo = seqState->stateML.table[seqState->stateML.state]; + ZSTD_seqSymbol const ofDInfo = seqState->stateOffb.table[seqState->stateOffb.state]; + U32 const llBase = llDInfo.baseValue; + U32 const mlBase = mlDInfo.baseValue; + U32 const ofBase = ofDInfo.baseValue; + BYTE const llBits = llDInfo.nbAdditionalBits; + BYTE const mlBits = mlDInfo.nbAdditionalBits; + BYTE const ofBits = ofDInfo.nbAdditionalBits; + BYTE const totalBits = llBits+mlBits+ofBits; + + /* sequence */ + { size_t offset; + if (ofBits > 1) { + ZSTD_STATIC_ASSERT(ZSTD_lo_isLongOffset == 1); + ZSTD_STATIC_ASSERT(LONG_OFFSETS_MAX_EXTRA_BITS_32 == 5); + assert(ofBits <= MaxOff); + if (MEM_32bits() && longOffsets && (ofBits >= STREAM_ACCUMULATOR_MIN_32)) { + U32 const extraBits = ofBits - MIN(ofBits, 32 - seqState->DStream.bitsConsumed); + offset = ofBase + (BIT_readBitsFast(&seqState->DStream, ofBits - extraBits) << extraBits); + BIT_reloadDStream(&seqState->DStream); + if (extraBits) offset += BIT_readBitsFast(&seqState->DStream, extraBits); + assert(extraBits <= LONG_OFFSETS_MAX_EXTRA_BITS_32); /* to avoid another reload */ + } else { + offset = ofBase + BIT_readBitsFast(&seqState->DStream, ofBits/*>0*/); /* <= (ZSTD_WINDOWLOG_MAX-1) bits */ + if (MEM_32bits()) BIT_reloadDStream(&seqState->DStream); + } + seqState->prevOffset[2] = seqState->prevOffset[1]; + seqState->prevOffset[1] = seqState->prevOffset[0]; + seqState->prevOffset[0] = offset; + } else { + U32 const ll0 = (llBase == 0); + if (LIKELY((ofBits == 0))) { + if (LIKELY(!ll0)) + offset = seqState->prevOffset[0]; + else { + offset = seqState->prevOffset[1]; + seqState->prevOffset[1] = seqState->prevOffset[0]; + seqState->prevOffset[0] = offset; + } + } else { + offset = ofBase + ll0 + BIT_readBitsFast(&seqState->DStream, 1); + { size_t temp = (offset==3) ? seqState->prevOffset[0] - 1 : seqState->prevOffset[offset]; + temp += !temp; /* 0 is not valid; input is corrupted; force offset to 1 */ + if (offset != 1) seqState->prevOffset[2] = seqState->prevOffset[1]; + seqState->prevOffset[1] = seqState->prevOffset[0]; + seqState->prevOffset[0] = offset = temp; + } } } + seq.offset = offset; + } + + seq.matchLength = mlBase; + if (mlBits > 0) + seq.matchLength += BIT_readBitsFast(&seqState->DStream, mlBits/*>0*/); + + if (MEM_32bits() && (mlBits+llBits >= STREAM_ACCUMULATOR_MIN_32-LONG_OFFSETS_MAX_EXTRA_BITS_32)) + BIT_reloadDStream(&seqState->DStream); + if (MEM_64bits() && UNLIKELY(totalBits >= STREAM_ACCUMULATOR_MIN_64-(LLFSELog+MLFSELog+OffFSELog))) + BIT_reloadDStream(&seqState->DStream); + /* Ensure there are enough bits to read the rest of data in 64-bit mode. */ + ZSTD_STATIC_ASSERT(16+LLFSELog+MLFSELog+OffFSELog < STREAM_ACCUMULATOR_MIN_64); + + seq.litLength = llBase; + if (llBits > 0) + seq.litLength += BIT_readBitsFast(&seqState->DStream, llBits/*>0*/); + + if (MEM_32bits()) + BIT_reloadDStream(&seqState->DStream); + + DEBUGLOG(6, "seq: litL=%u, matchL=%u, offset=%u", + (U32)seq.litLength, (U32)seq.matchLength, (U32)seq.offset); + + if (prefetch == ZSTD_p_prefetch) { + size_t const pos = seqState->pos + seq.litLength; + const BYTE* const matchBase = (seq.offset > pos) ? seqState->dictEnd : seqState->prefixStart; + seq.match = matchBase + pos - seq.offset; /* note : this operation can overflow when seq.offset is really too large, which can only happen when input is corrupted. + * No consequence though : no memory access will occur, offset is only used for prefetching */ + seqState->pos = pos + seq.matchLength; + } + + /* ANS state update + * gcc-9.0.0 does 2.5% worse with ZSTD_updateFseStateWithDInfo(). + * clang-9.2.0 does 7% worse with ZSTD_updateFseState(). + * Naturally it seems like ZSTD_updateFseStateWithDInfo() should be the + * better option, so it is the default for other compilers. But, if you + * measure that it is worse, please put up a pull request. + */ + { +#if !defined(__clang__) + const int kUseUpdateFseState = 1; +#else + const int kUseUpdateFseState = 0; +#endif + if (kUseUpdateFseState) { + ZSTD_updateFseState(&seqState->stateLL, &seqState->DStream); /* <= 9 bits */ + ZSTD_updateFseState(&seqState->stateML, &seqState->DStream); /* <= 9 bits */ + if (MEM_32bits()) BIT_reloadDStream(&seqState->DStream); /* <= 18 bits */ + ZSTD_updateFseState(&seqState->stateOffb, &seqState->DStream); /* <= 8 bits */ + } else { + ZSTD_updateFseStateWithDInfo(&seqState->stateLL, &seqState->DStream, llDInfo); /* <= 9 bits */ + ZSTD_updateFseStateWithDInfo(&seqState->stateML, &seqState->DStream, mlDInfo); /* <= 9 bits */ + if (MEM_32bits()) BIT_reloadDStream(&seqState->DStream); /* <= 18 bits */ + ZSTD_updateFseStateWithDInfo(&seqState->stateOffb, &seqState->DStream, ofDInfo); /* <= 8 bits */ + } + } + + return seq; +} + +#ifdef FUZZING_BUILD_MODE_UNSAFE_FOR_PRODUCTION +MEM_STATIC int ZSTD_dictionaryIsActive(ZSTD_DCtx const* dctx, BYTE const* prefixStart, BYTE const* oLitEnd) +{ + size_t const windowSize = dctx->fParams.windowSize; + /* No dictionary used. */ + if (dctx->dictContentEndForFuzzing == NULL) return 0; + /* Dictionary is our prefix. */ + if (prefixStart == dctx->dictContentBeginForFuzzing) return 1; + /* Dictionary is not our ext-dict. */ + if (dctx->dictEnd != dctx->dictContentEndForFuzzing) return 0; + /* Dictionary is not within our window size. */ + if ((size_t)(oLitEnd - prefixStart) >= windowSize) return 0; + /* Dictionary is active. */ + return 1; +} + +MEM_STATIC void ZSTD_assertValidSequence( + ZSTD_DCtx const* dctx, + BYTE const* op, BYTE const* oend, + seq_t const seq, + BYTE const* prefixStart, BYTE const* virtualStart) +{ +#if DEBUGLEVEL >= 1 + size_t const windowSize = dctx->fParams.windowSize; + size_t const sequenceSize = seq.litLength + seq.matchLength; + BYTE const* const oLitEnd = op + seq.litLength; + DEBUGLOG(6, "Checking sequence: litL=%u matchL=%u offset=%u", + (U32)seq.litLength, (U32)seq.matchLength, (U32)seq.offset); + assert(op <= oend); + assert((size_t)(oend - op) >= sequenceSize); + assert(sequenceSize <= ZSTD_BLOCKSIZE_MAX); + if (ZSTD_dictionaryIsActive(dctx, prefixStart, oLitEnd)) { + size_t const dictSize = (size_t)((char const*)dctx->dictContentEndForFuzzing - (char const*)dctx->dictContentBeginForFuzzing); + /* Offset must be within the dictionary. */ + assert(seq.offset <= (size_t)(oLitEnd - virtualStart)); + assert(seq.offset <= windowSize + dictSize); + } else { + /* Offset must be within our window. */ + assert(seq.offset <= windowSize); + } +#else + (void)dctx, (void)op, (void)oend, (void)seq, (void)prefixStart, (void)virtualStart; +#endif +} +#endif + +#ifndef ZSTD_FORCE_DECOMPRESS_SEQUENCES_LONG +FORCE_INLINE_TEMPLATE size_t +DONT_VECTORIZE +ZSTD_decompressSequences_body( ZSTD_DCtx* dctx, + void* dst, size_t maxDstSize, + const void* seqStart, size_t seqSize, int nbSeq, + const ZSTD_longOffset_e isLongOffset, + const int frame) +{ + const BYTE* ip = (const BYTE*)seqStart; + const BYTE* const iend = ip + seqSize; + BYTE* const ostart = (BYTE*)dst; + BYTE* const oend = ostart + maxDstSize; + BYTE* op = ostart; + const BYTE* litPtr = dctx->litPtr; + const BYTE* const litEnd = litPtr + dctx->litSize; + const BYTE* const prefixStart = (const BYTE*) (dctx->prefixStart); + const BYTE* const vBase = (const BYTE*) (dctx->virtualStart); + const BYTE* const dictEnd = (const BYTE*) (dctx->dictEnd); + DEBUGLOG(5, "ZSTD_decompressSequences_body"); + (void)frame; + + /* Regen sequences */ + if (nbSeq) { + seqState_t seqState; + size_t error = 0; + dctx->fseEntropy = 1; + { U32 i; for (i=0; ientropy.rep[i]; } + RETURN_ERROR_IF( + ERR_isError(BIT_initDStream(&seqState.DStream, ip, iend-ip)), + corruption_detected, ""); + ZSTD_initFseState(&seqState.stateLL, &seqState.DStream, dctx->LLTptr); + ZSTD_initFseState(&seqState.stateOffb, &seqState.DStream, dctx->OFTptr); + ZSTD_initFseState(&seqState.stateML, &seqState.DStream, dctx->MLTptr); + assert(dst != NULL); + + ZSTD_STATIC_ASSERT( + BIT_DStream_unfinished < BIT_DStream_completed && + BIT_DStream_endOfBuffer < BIT_DStream_completed && + BIT_DStream_completed < BIT_DStream_overflow); + +#if defined(__x86_64__) + /* Align the decompression loop to 32 + 16 bytes. + * + * zstd compiled with gcc-9 on an Intel i9-9900k shows 10% decompression + * speed swings based on the alignment of the decompression loop. This + * performance swing is caused by parts of the decompression loop falling + * out of the DSB. The entire decompression loop should fit in the DSB, + * when it can't we get much worse performance. You can measure if you've + * hit the good case or the bad case with this perf command for some + * compressed file test.zst: + * + * perf stat -e cycles -e instructions -e idq.all_dsb_cycles_any_uops \ + * -e idq.all_mite_cycles_any_uops -- ./zstd -tq test.zst + * + * If you see most cycles served out of the MITE you've hit the bad case. + * If you see most cycles served out of the DSB you've hit the good case. + * If it is pretty even then you may be in an okay case. + * + * I've been able to reproduce this issue on the following CPUs: + * - Kabylake: Macbook Pro (15-inch, 2019) 2.4 GHz Intel Core i9 + * Use Instruments->Counters to get DSB/MITE cycles. + * I never got performance swings, but I was able to + * go from the good case of mostly DSB to half of the + * cycles served from MITE. + * - Coffeelake: Intel i9-9900k + * + * I haven't been able to reproduce the instability or DSB misses on any + * of the following CPUS: + * - Haswell + * - Broadwell: Intel(R) Xeon(R) CPU E5-2680 v4 @ 2.40GH + * - Skylake + * + * If you are seeing performance stability this script can help test. + * It tests on 4 commits in zstd where I saw performance change. + * + * https://gist.github.com/terrelln/9889fc06a423fd5ca6e99351564473f4 + */ + __asm__(".p2align 5"); + __asm__("nop"); + __asm__(".p2align 4"); +#endif + for ( ; ; ) { + seq_t const sequence = ZSTD_decodeSequence(&seqState, isLongOffset, ZSTD_p_noPrefetch); + size_t const oneSeqSize = ZSTD_execSequence(op, oend, sequence, &litPtr, litEnd, prefixStart, vBase, dictEnd); +#if defined(FUZZING_BUILD_MODE_UNSAFE_FOR_PRODUCTION) && defined(FUZZING_ASSERT_VALID_SEQUENCE) + assert(!ZSTD_isError(oneSeqSize)); + if (frame) ZSTD_assertValidSequence(dctx, op, oend, sequence, prefixStart, vBase); +#endif + DEBUGLOG(6, "regenerated sequence size : %u", (U32)oneSeqSize); + BIT_reloadDStream(&(seqState.DStream)); + op += oneSeqSize; + /* gcc and clang both don't like early returns in this loop. + * Instead break and check for an error at the end of the loop. + */ + if (UNLIKELY(ZSTD_isError(oneSeqSize))) { + error = oneSeqSize; + break; + } + if (UNLIKELY(!--nbSeq)) break; + } + + /* check if reached exact end */ + DEBUGLOG(5, "ZSTD_decompressSequences_body: after decode loop, remaining nbSeq : %i", nbSeq); + if (ZSTD_isError(error)) return error; + RETURN_ERROR_IF(nbSeq, corruption_detected, ""); + RETURN_ERROR_IF(BIT_reloadDStream(&seqState.DStream) < BIT_DStream_completed, corruption_detected, ""); + /* save reps for next block */ + { U32 i; for (i=0; ientropy.rep[i] = (U32)(seqState.prevOffset[i]); } + } + + /* last literal segment */ + { size_t const lastLLSize = litEnd - litPtr; + RETURN_ERROR_IF(lastLLSize > (size_t)(oend-op), dstSize_tooSmall, ""); + if (op != NULL) { + ZSTD_memcpy(op, litPtr, lastLLSize); + op += lastLLSize; + } + } + + return op-ostart; +} + +static size_t +ZSTD_decompressSequences_default(ZSTD_DCtx* dctx, + void* dst, size_t maxDstSize, + const void* seqStart, size_t seqSize, int nbSeq, + const ZSTD_longOffset_e isLongOffset, + const int frame) +{ + return ZSTD_decompressSequences_body(dctx, dst, maxDstSize, seqStart, seqSize, nbSeq, isLongOffset, frame); +} +#endif /* ZSTD_FORCE_DECOMPRESS_SEQUENCES_LONG */ + +#ifndef ZSTD_FORCE_DECOMPRESS_SEQUENCES_SHORT +FORCE_INLINE_TEMPLATE size_t +ZSTD_decompressSequencesLong_body( + ZSTD_DCtx* dctx, + void* dst, size_t maxDstSize, + const void* seqStart, size_t seqSize, int nbSeq, + const ZSTD_longOffset_e isLongOffset, + const int frame) +{ + const BYTE* ip = (const BYTE*)seqStart; + const BYTE* const iend = ip + seqSize; + BYTE* const ostart = (BYTE*)dst; + BYTE* const oend = ostart + maxDstSize; + BYTE* op = ostart; + const BYTE* litPtr = dctx->litPtr; + const BYTE* const litEnd = litPtr + dctx->litSize; + const BYTE* const prefixStart = (const BYTE*) (dctx->prefixStart); + const BYTE* const dictStart = (const BYTE*) (dctx->virtualStart); + const BYTE* const dictEnd = (const BYTE*) (dctx->dictEnd); + (void)frame; + + /* Regen sequences */ + if (nbSeq) { +#define STORED_SEQS 4 +#define STORED_SEQS_MASK (STORED_SEQS-1) +#define ADVANCED_SEQS 4 + seq_t sequences[STORED_SEQS]; + int const seqAdvance = MIN(nbSeq, ADVANCED_SEQS); + seqState_t seqState; + int seqNb; + dctx->fseEntropy = 1; + { int i; for (i=0; ientropy.rep[i]; } + seqState.prefixStart = prefixStart; + seqState.pos = (size_t)(op-prefixStart); + seqState.dictEnd = dictEnd; + assert(dst != NULL); + assert(iend >= ip); + RETURN_ERROR_IF( + ERR_isError(BIT_initDStream(&seqState.DStream, ip, iend-ip)), + corruption_detected, ""); + ZSTD_initFseState(&seqState.stateLL, &seqState.DStream, dctx->LLTptr); + ZSTD_initFseState(&seqState.stateOffb, &seqState.DStream, dctx->OFTptr); + ZSTD_initFseState(&seqState.stateML, &seqState.DStream, dctx->MLTptr); + + /* prepare in advance */ + for (seqNb=0; (BIT_reloadDStream(&seqState.DStream) <= BIT_DStream_completed) && (seqNbentropy.rep[i] = (U32)(seqState.prevOffset[i]); } + } + + /* last literal segment */ + { size_t const lastLLSize = litEnd - litPtr; + RETURN_ERROR_IF(lastLLSize > (size_t)(oend-op), dstSize_tooSmall, ""); + if (op != NULL) { + ZSTD_memcpy(op, litPtr, lastLLSize); + op += lastLLSize; + } + } + + return op-ostart; +} + +static size_t +ZSTD_decompressSequencesLong_default(ZSTD_DCtx* dctx, + void* dst, size_t maxDstSize, + const void* seqStart, size_t seqSize, int nbSeq, + const ZSTD_longOffset_e isLongOffset, + const int frame) +{ + return ZSTD_decompressSequencesLong_body(dctx, dst, maxDstSize, seqStart, seqSize, nbSeq, isLongOffset, frame); +} +#endif /* ZSTD_FORCE_DECOMPRESS_SEQUENCES_SHORT */ + + + +#if DYNAMIC_BMI2 + +#ifndef ZSTD_FORCE_DECOMPRESS_SEQUENCES_LONG +static TARGET_ATTRIBUTE("bmi2") size_t +DONT_VECTORIZE +ZSTD_decompressSequences_bmi2(ZSTD_DCtx* dctx, + void* dst, size_t maxDstSize, + const void* seqStart, size_t seqSize, int nbSeq, + const ZSTD_longOffset_e isLongOffset, + const int frame) +{ + return ZSTD_decompressSequences_body(dctx, dst, maxDstSize, seqStart, seqSize, nbSeq, isLongOffset, frame); +} +#endif /* ZSTD_FORCE_DECOMPRESS_SEQUENCES_LONG */ + +#ifndef ZSTD_FORCE_DECOMPRESS_SEQUENCES_SHORT +static TARGET_ATTRIBUTE("bmi2") size_t +ZSTD_decompressSequencesLong_bmi2(ZSTD_DCtx* dctx, + void* dst, size_t maxDstSize, + const void* seqStart, size_t seqSize, int nbSeq, + const ZSTD_longOffset_e isLongOffset, + const int frame) +{ + return ZSTD_decompressSequencesLong_body(dctx, dst, maxDstSize, seqStart, seqSize, nbSeq, isLongOffset, frame); +} +#endif /* ZSTD_FORCE_DECOMPRESS_SEQUENCES_SHORT */ + +#endif /* DYNAMIC_BMI2 */ + +typedef size_t (*ZSTD_decompressSequences_t)( + ZSTD_DCtx* dctx, + void* dst, size_t maxDstSize, + const void* seqStart, size_t seqSize, int nbSeq, + const ZSTD_longOffset_e isLongOffset, + const int frame); + +#ifndef ZSTD_FORCE_DECOMPRESS_SEQUENCES_LONG +static size_t +ZSTD_decompressSequences(ZSTD_DCtx* dctx, void* dst, size_t maxDstSize, + const void* seqStart, size_t seqSize, int nbSeq, + const ZSTD_longOffset_e isLongOffset, + const int frame) +{ + DEBUGLOG(5, "ZSTD_decompressSequences"); +#if DYNAMIC_BMI2 + if (dctx->bmi2) { + return ZSTD_decompressSequences_bmi2(dctx, dst, maxDstSize, seqStart, seqSize, nbSeq, isLongOffset, frame); + } +#endif + return ZSTD_decompressSequences_default(dctx, dst, maxDstSize, seqStart, seqSize, nbSeq, isLongOffset, frame); +} +#endif /* ZSTD_FORCE_DECOMPRESS_SEQUENCES_LONG */ + + +#ifndef ZSTD_FORCE_DECOMPRESS_SEQUENCES_SHORT +/* ZSTD_decompressSequencesLong() : + * decompression function triggered when a minimum share of offsets is considered "long", + * aka out of cache. + * note : "long" definition seems overloaded here, sometimes meaning "wider than bitstream register", and sometimes meaning "farther than memory cache distance". + * This function will try to mitigate main memory latency through the use of prefetching */ +static size_t +ZSTD_decompressSequencesLong(ZSTD_DCtx* dctx, + void* dst, size_t maxDstSize, + const void* seqStart, size_t seqSize, int nbSeq, + const ZSTD_longOffset_e isLongOffset, + const int frame) +{ + DEBUGLOG(5, "ZSTD_decompressSequencesLong"); +#if DYNAMIC_BMI2 + if (dctx->bmi2) { + return ZSTD_decompressSequencesLong_bmi2(dctx, dst, maxDstSize, seqStart, seqSize, nbSeq, isLongOffset, frame); + } +#endif + return ZSTD_decompressSequencesLong_default(dctx, dst, maxDstSize, seqStart, seqSize, nbSeq, isLongOffset, frame); +} +#endif /* ZSTD_FORCE_DECOMPRESS_SEQUENCES_SHORT */ + + + +#if !defined(ZSTD_FORCE_DECOMPRESS_SEQUENCES_SHORT) && \ + !defined(ZSTD_FORCE_DECOMPRESS_SEQUENCES_LONG) +/* ZSTD_getLongOffsetsShare() : + * condition : offTable must be valid + * @return : "share" of long offsets (arbitrarily defined as > (1<<23)) + * compared to maximum possible of (1< 22) total += 1; + } + + assert(tableLog <= OffFSELog); + total <<= (OffFSELog - tableLog); /* scale to OffFSELog */ + + return total; +} +#endif + +size_t +ZSTD_decompressBlock_internal(ZSTD_DCtx* dctx, + void* dst, size_t dstCapacity, + const void* src, size_t srcSize, const int frame) +{ /* blockType == blockCompressed */ + const BYTE* ip = (const BYTE*)src; + /* isLongOffset must be true if there are long offsets. + * Offsets are long if they are larger than 2^STREAM_ACCUMULATOR_MIN. + * We don't expect that to be the case in 64-bit mode. + * In block mode, window size is not known, so we have to be conservative. + * (note: but it could be evaluated from current-lowLimit) + */ + ZSTD_longOffset_e const isLongOffset = (ZSTD_longOffset_e)(MEM_32bits() && (!frame || (dctx->fParams.windowSize > (1ULL << STREAM_ACCUMULATOR_MIN)))); + DEBUGLOG(5, "ZSTD_decompressBlock_internal (size : %u)", (U32)srcSize); + + RETURN_ERROR_IF(srcSize >= ZSTD_BLOCKSIZE_MAX, srcSize_wrong, ""); + + /* Decode literals section */ + { size_t const litCSize = ZSTD_decodeLiteralsBlock(dctx, src, srcSize); + DEBUGLOG(5, "ZSTD_decodeLiteralsBlock : %u", (U32)litCSize); + if (ZSTD_isError(litCSize)) return litCSize; + ip += litCSize; + srcSize -= litCSize; + } + + /* Build Decoding Tables */ + { + /* These macros control at build-time which decompressor implementation + * we use. If neither is defined, we do some inspection and dispatch at + * runtime. + */ +#if !defined(ZSTD_FORCE_DECOMPRESS_SEQUENCES_SHORT) && \ + !defined(ZSTD_FORCE_DECOMPRESS_SEQUENCES_LONG) + int usePrefetchDecoder = dctx->ddictIsCold; +#endif + int nbSeq; + size_t const seqHSize = ZSTD_decodeSeqHeaders(dctx, &nbSeq, ip, srcSize); + if (ZSTD_isError(seqHSize)) return seqHSize; + ip += seqHSize; + srcSize -= seqHSize; + + RETURN_ERROR_IF(dst == NULL && nbSeq > 0, dstSize_tooSmall, "NULL not handled"); + +#if !defined(ZSTD_FORCE_DECOMPRESS_SEQUENCES_SHORT) && \ + !defined(ZSTD_FORCE_DECOMPRESS_SEQUENCES_LONG) + if ( !usePrefetchDecoder + && (!frame || (dctx->fParams.windowSize > (1<<24))) + && (nbSeq>ADVANCED_SEQS) ) { /* could probably use a larger nbSeq limit */ + U32 const shareLongOffsets = ZSTD_getLongOffsetsShare(dctx->OFTptr); + U32 const minShare = MEM_64bits() ? 7 : 20; /* heuristic values, correspond to 2.73% and 7.81% */ + usePrefetchDecoder = (shareLongOffsets >= minShare); + } +#endif + + dctx->ddictIsCold = 0; + +#if !defined(ZSTD_FORCE_DECOMPRESS_SEQUENCES_SHORT) && \ + !defined(ZSTD_FORCE_DECOMPRESS_SEQUENCES_LONG) + if (usePrefetchDecoder) +#endif +#ifndef ZSTD_FORCE_DECOMPRESS_SEQUENCES_SHORT + return ZSTD_decompressSequencesLong(dctx, dst, dstCapacity, ip, srcSize, nbSeq, isLongOffset, frame); +#endif + +#ifndef ZSTD_FORCE_DECOMPRESS_SEQUENCES_LONG + /* else */ + return ZSTD_decompressSequences(dctx, dst, dstCapacity, ip, srcSize, nbSeq, isLongOffset, frame); +#endif + } +} + + +void ZSTD_checkContinuity(ZSTD_DCtx* dctx, const void* dst, size_t dstSize) +{ + if (dst != dctx->previousDstEnd && dstSize > 0) { /* not contiguous */ + dctx->dictEnd = dctx->previousDstEnd; + dctx->virtualStart = (const char*)dst - ((const char*)(dctx->previousDstEnd) - (const char*)(dctx->prefixStart)); + dctx->prefixStart = dst; + dctx->previousDstEnd = dst; + } +} + + +size_t ZSTD_decompressBlock(ZSTD_DCtx* dctx, + void* dst, size_t dstCapacity, + const void* src, size_t srcSize) +{ + size_t dSize; + ZSTD_checkContinuity(dctx, dst, dstCapacity); + dSize = ZSTD_decompressBlock_internal(dctx, dst, dstCapacity, src, srcSize, /* frame */ 0); + dctx->previousDstEnd = (char*)dst + dSize; + return dSize; +} diff --git a/lib/zstd/decompress/zstd_decompress_block.h b/lib/zstd/decompress/zstd_decompress_block.h new file mode 100644 index 000000000000..e7f5f6689459 --- /dev/null +++ b/lib/zstd/decompress/zstd_decompress_block.h @@ -0,0 +1,62 @@ +/* + * Copyright (c) Yann Collet, Facebook, Inc. + * All rights reserved. + * + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). + * You may select, at your option, one of the above-listed licenses. + */ + + +#ifndef ZSTD_DEC_BLOCK_H +#define ZSTD_DEC_BLOCK_H + +/*-******************************************************* + * Dependencies + *********************************************************/ +#include "../common/zstd_deps.h" /* size_t */ +#include /* DCtx, and some public functions */ +#include "../common/zstd_internal.h" /* blockProperties_t, and some public functions */ +#include "zstd_decompress_internal.h" /* ZSTD_seqSymbol */ + + +/* === Prototypes === */ + +/* note: prototypes already published within `zstd.h` : + * ZSTD_decompressBlock() + */ + +/* note: prototypes already published within `zstd_internal.h` : + * ZSTD_getcBlockSize() + * ZSTD_decodeSeqHeaders() + */ + + +/* ZSTD_decompressBlock_internal() : + * decompress block, starting at `src`, + * into destination buffer `dst`. + * @return : decompressed block size, + * or an error code (which can be tested using ZSTD_isError()) + */ +size_t ZSTD_decompressBlock_internal(ZSTD_DCtx* dctx, + void* dst, size_t dstCapacity, + const void* src, size_t srcSize, const int frame); + +/* ZSTD_buildFSETable() : + * generate FSE decoding table for one symbol (ll, ml or off) + * this function must be called with valid parameters only + * (dt is large enough, normalizedCounter distribution total is a power of 2, max is within range, etc.) + * in which case it cannot fail. + * The workspace must be 4-byte aligned and at least ZSTD_BUILD_FSE_TABLE_WKSP_SIZE bytes, which is + * defined in zstd_decompress_internal.h. + * Internal use only. + */ +void ZSTD_buildFSETable(ZSTD_seqSymbol* dt, + const short* normalizedCounter, unsigned maxSymbolValue, + const U32* baseValue, const U32* nbAdditionalBits, + unsigned tableLog, void* wksp, size_t wkspSize, + int bmi2); + + +#endif /* ZSTD_DEC_BLOCK_H */ diff --git a/lib/zstd/decompress/zstd_decompress_internal.h b/lib/zstd/decompress/zstd_decompress_internal.h new file mode 100644 index 000000000000..4b9052f68755 --- /dev/null +++ b/lib/zstd/decompress/zstd_decompress_internal.h @@ -0,0 +1,202 @@ +/* + * Copyright (c) Yann Collet, Facebook, Inc. + * All rights reserved. + * + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). + * You may select, at your option, one of the above-listed licenses. + */ + + +/* zstd_decompress_internal: + * objects and definitions shared within lib/decompress modules */ + + #ifndef ZSTD_DECOMPRESS_INTERNAL_H + #define ZSTD_DECOMPRESS_INTERNAL_H + + +/*-******************************************************* + * Dependencies + *********************************************************/ +#include "../common/mem.h" /* BYTE, U16, U32 */ +#include "../common/zstd_internal.h" /* ZSTD_seqSymbol */ + + + +/*-******************************************************* + * Constants + *********************************************************/ +static UNUSED_ATTR const U32 LL_base[MaxLL+1] = { + 0, 1, 2, 3, 4, 5, 6, 7, + 8, 9, 10, 11, 12, 13, 14, 15, + 16, 18, 20, 22, 24, 28, 32, 40, + 48, 64, 0x80, 0x100, 0x200, 0x400, 0x800, 0x1000, + 0x2000, 0x4000, 0x8000, 0x10000 }; + +static UNUSED_ATTR const U32 OF_base[MaxOff+1] = { + 0, 1, 1, 5, 0xD, 0x1D, 0x3D, 0x7D, + 0xFD, 0x1FD, 0x3FD, 0x7FD, 0xFFD, 0x1FFD, 0x3FFD, 0x7FFD, + 0xFFFD, 0x1FFFD, 0x3FFFD, 0x7FFFD, 0xFFFFD, 0x1FFFFD, 0x3FFFFD, 0x7FFFFD, + 0xFFFFFD, 0x1FFFFFD, 0x3FFFFFD, 0x7FFFFFD, 0xFFFFFFD, 0x1FFFFFFD, 0x3FFFFFFD, 0x7FFFFFFD }; + +static UNUSED_ATTR const U32 OF_bits[MaxOff+1] = { + 0, 1, 2, 3, 4, 5, 6, 7, + 8, 9, 10, 11, 12, 13, 14, 15, + 16, 17, 18, 19, 20, 21, 22, 23, + 24, 25, 26, 27, 28, 29, 30, 31 }; + +static UNUSED_ATTR const U32 ML_base[MaxML+1] = { + 3, 4, 5, 6, 7, 8, 9, 10, + 11, 12, 13, 14, 15, 16, 17, 18, + 19, 20, 21, 22, 23, 24, 25, 26, + 27, 28, 29, 30, 31, 32, 33, 34, + 35, 37, 39, 41, 43, 47, 51, 59, + 67, 83, 99, 0x83, 0x103, 0x203, 0x403, 0x803, + 0x1003, 0x2003, 0x4003, 0x8003, 0x10003 }; + + +/*-******************************************************* + * Decompression types + *********************************************************/ + typedef struct { + U32 fastMode; + U32 tableLog; + } ZSTD_seqSymbol_header; + + typedef struct { + U16 nextState; + BYTE nbAdditionalBits; + BYTE nbBits; + U32 baseValue; + } ZSTD_seqSymbol; + + #define SEQSYMBOL_TABLE_SIZE(log) (1 + (1 << (log))) + +#define ZSTD_BUILD_FSE_TABLE_WKSP_SIZE (sizeof(S16) * (MaxSeq + 1) + (1u << MaxFSELog) + sizeof(U64)) +#define ZSTD_BUILD_FSE_TABLE_WKSP_SIZE_U32 ((ZSTD_BUILD_FSE_TABLE_WKSP_SIZE + sizeof(U32) - 1) / sizeof(U32)) + +typedef struct { + ZSTD_seqSymbol LLTable[SEQSYMBOL_TABLE_SIZE(LLFSELog)]; /* Note : Space reserved for FSE Tables */ + ZSTD_seqSymbol OFTable[SEQSYMBOL_TABLE_SIZE(OffFSELog)]; /* is also used as temporary workspace while building hufTable during DDict creation */ + ZSTD_seqSymbol MLTable[SEQSYMBOL_TABLE_SIZE(MLFSELog)]; /* and therefore must be at least HUF_DECOMPRESS_WORKSPACE_SIZE large */ + HUF_DTable hufTable[HUF_DTABLE_SIZE(HufLog)]; /* can accommodate HUF_decompress4X */ + U32 rep[ZSTD_REP_NUM]; + U32 workspace[ZSTD_BUILD_FSE_TABLE_WKSP_SIZE_U32]; +} ZSTD_entropyDTables_t; + +typedef enum { ZSTDds_getFrameHeaderSize, ZSTDds_decodeFrameHeader, + ZSTDds_decodeBlockHeader, ZSTDds_decompressBlock, + ZSTDds_decompressLastBlock, ZSTDds_checkChecksum, + ZSTDds_decodeSkippableHeader, ZSTDds_skipFrame } ZSTD_dStage; + +typedef enum { zdss_init=0, zdss_loadHeader, + zdss_read, zdss_load, zdss_flush } ZSTD_dStreamStage; + +typedef enum { + ZSTD_use_indefinitely = -1, /* Use the dictionary indefinitely */ + ZSTD_dont_use = 0, /* Do not use the dictionary (if one exists free it) */ + ZSTD_use_once = 1 /* Use the dictionary once and set to ZSTD_dont_use */ +} ZSTD_dictUses_e; + +/* Hashset for storing references to multiple ZSTD_DDict within ZSTD_DCtx */ +typedef struct { + const ZSTD_DDict** ddictPtrTable; + size_t ddictPtrTableSize; + size_t ddictPtrCount; +} ZSTD_DDictHashSet; + +struct ZSTD_DCtx_s +{ + const ZSTD_seqSymbol* LLTptr; + const ZSTD_seqSymbol* MLTptr; + const ZSTD_seqSymbol* OFTptr; + const HUF_DTable* HUFptr; + ZSTD_entropyDTables_t entropy; + U32 workspace[HUF_DECOMPRESS_WORKSPACE_SIZE_U32]; /* space needed when building huffman tables */ + const void* previousDstEnd; /* detect continuity */ + const void* prefixStart; /* start of current segment */ + const void* virtualStart; /* virtual start of previous segment if it was just before current one */ + const void* dictEnd; /* end of previous segment */ + size_t expected; + ZSTD_frameHeader fParams; + U64 processedCSize; + U64 decodedSize; + blockType_e bType; /* used in ZSTD_decompressContinue(), store blockType between block header decoding and block decompression stages */ + ZSTD_dStage stage; + U32 litEntropy; + U32 fseEntropy; + struct xxh64_state xxhState; + size_t headerSize; + ZSTD_format_e format; + ZSTD_forceIgnoreChecksum_e forceIgnoreChecksum; /* User specified: if == 1, will ignore checksums in compressed frame. Default == 0 */ + U32 validateChecksum; /* if == 1, will validate checksum. Is == 1 if (fParams.checksumFlag == 1) and (forceIgnoreChecksum == 0). */ + const BYTE* litPtr; + ZSTD_customMem customMem; + size_t litSize; + size_t rleSize; + size_t staticSize; + int bmi2; /* == 1 if the CPU supports BMI2 and 0 otherwise. CPU support is determined dynamically once per context lifetime. */ + + /* dictionary */ + ZSTD_DDict* ddictLocal; + const ZSTD_DDict* ddict; /* set by ZSTD_initDStream_usingDDict(), or ZSTD_DCtx_refDDict() */ + U32 dictID; + int ddictIsCold; /* if == 1 : dictionary is "new" for working context, and presumed "cold" (not in cpu cache) */ + ZSTD_dictUses_e dictUses; + ZSTD_DDictHashSet* ddictSet; /* Hash set for multiple ddicts */ + ZSTD_refMultipleDDicts_e refMultipleDDicts; /* User specified: if == 1, will allow references to multiple DDicts. Default == 0 (disabled) */ + + /* streaming */ + ZSTD_dStreamStage streamStage; + char* inBuff; + size_t inBuffSize; + size_t inPos; + size_t maxWindowSize; + char* outBuff; + size_t outBuffSize; + size_t outStart; + size_t outEnd; + size_t lhSize; + void* legacyContext; + U32 previousLegacyVersion; + U32 legacyVersion; + U32 hostageByte; + int noForwardProgress; + ZSTD_bufferMode_e outBufferMode; + ZSTD_outBuffer expectedOutBuffer; + + /* workspace */ + BYTE litBuffer[ZSTD_BLOCKSIZE_MAX + WILDCOPY_OVERLENGTH]; + BYTE headerBuffer[ZSTD_FRAMEHEADERSIZE_MAX]; + + size_t oversizedDuration; + +#ifdef FUZZING_BUILD_MODE_UNSAFE_FOR_PRODUCTION + void const* dictContentBeginForFuzzing; + void const* dictContentEndForFuzzing; +#endif + + /* Tracing */ +}; /* typedef'd to ZSTD_DCtx within "zstd.h" */ + + +/*-******************************************************* + * Shared internal functions + *********************************************************/ + +/*! ZSTD_loadDEntropy() : + * dict : must point at beginning of a valid zstd dictionary. + * @return : size of dictionary header (size of magic number + dict ID + entropy tables) */ +size_t ZSTD_loadDEntropy(ZSTD_entropyDTables_t* entropy, + const void* const dict, size_t const dictSize); + +/*! ZSTD_checkContinuity() : + * check if next `dst` follows previous position, where decompression ended. + * If yes, do nothing (continue on current segment). + * If not, classify previous segment as "external dictionary", and start a new segment. + * This function cannot fail. */ +void ZSTD_checkContinuity(ZSTD_DCtx* dctx, const void* dst, size_t dstSize); + + +#endif /* ZSTD_DECOMPRESS_INTERNAL_H */ diff --git a/lib/zstd/decompress_sources.h b/lib/zstd/decompress_sources.h index 9ba367b44149..0fbec508f285 100644 --- a/lib/zstd/decompress_sources.h +++ b/lib/zstd/decompress_sources.h @@ -16,8 +16,13 @@ * decompression. */ -#include "entropy_common.c" -#include "fse_decompress.c" -#include "huf_decompress.c" -#include "zstd_common.c" -#include "decompress.c" +#include "common/debug.c" +#include "common/entropy_common.c" +#include "common/error_private.c" +#include "common/fse_decompress.c" +#include "common/zstd_common.c" +#include "decompress/huf_decompress.c" +#include "decompress/zstd_ddict.c" +#include "decompress/zstd_decompress.c" +#include "decompress/zstd_decompress_block.c" +#include "zstd_decompress_module.c" diff --git a/lib/zstd/entropy_common.c b/lib/zstd/entropy_common.c deleted file mode 100644 index 2b0a643c32c4..000000000000 --- a/lib/zstd/entropy_common.c +++ /dev/null @@ -1,243 +0,0 @@ -/* - * Common functions of New Generation Entropy library - * Copyright (C) 2016, Yann Collet. - * - * BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php) - * - * Redistribution and use in source and binary forms, with or without - * modification, are permitted provided that the following conditions are - * met: - * - * * Redistributions of source code must retain the above copyright - * notice, this list of conditions and the following disclaimer. - * * Redistributions in binary form must reproduce the above - * copyright notice, this list of conditions and the following disclaimer - * in the documentation and/or other materials provided with the - * distribution. - * - * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS - * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT - * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR - * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT - * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, - * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT - * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, - * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY - * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT - * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE - * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. - * - * This program is free software; you can redistribute it and/or modify it under - * the terms of the GNU General Public License version 2 as published by the - * Free Software Foundation. This program is dual-licensed; you may select - * either version 2 of the GNU General Public License ("GPL") or BSD license - * ("BSD"). - * - * You can contact the author at : - * - Source repository : https://github.com/Cyan4973/FiniteStateEntropy - */ - -/* ************************************* -* Dependencies -***************************************/ -#include "error_private.h" /* ERR_*, ERROR */ -#include "fse.h" -#include "huf.h" -#include "mem.h" - -/*=== Version ===*/ -unsigned FSE_versionNumber(void) { return FSE_VERSION_NUMBER; } - -/*=== Error Management ===*/ -unsigned FSE_isError(size_t code) { return ERR_isError(code); } - -unsigned HUF_isError(size_t code) { return ERR_isError(code); } - -/*-************************************************************** -* FSE NCount encoding-decoding -****************************************************************/ -size_t FSE_readNCount(short *normalizedCounter, unsigned *maxSVPtr, unsigned *tableLogPtr, const void *headerBuffer, size_t hbSize) -{ - const BYTE *const istart = (const BYTE *)headerBuffer; - const BYTE *const iend = istart + hbSize; - const BYTE *ip = istart; - int nbBits; - int remaining; - int threshold; - U32 bitStream; - int bitCount; - unsigned charnum = 0; - int previous0 = 0; - - if (hbSize < 4) - return ERROR(srcSize_wrong); - bitStream = ZSTD_readLE32(ip); - nbBits = (bitStream & 0xF) + FSE_MIN_TABLELOG; /* extract tableLog */ - if (nbBits > FSE_TABLELOG_ABSOLUTE_MAX) - return ERROR(tableLog_tooLarge); - bitStream >>= 4; - bitCount = 4; - *tableLogPtr = nbBits; - remaining = (1 << nbBits) + 1; - threshold = 1 << nbBits; - nbBits++; - - while ((remaining > 1) & (charnum <= *maxSVPtr)) { - if (previous0) { - unsigned n0 = charnum; - while ((bitStream & 0xFFFF) == 0xFFFF) { - n0 += 24; - if (ip < iend - 5) { - ip += 2; - bitStream = ZSTD_readLE32(ip) >> bitCount; - } else { - bitStream >>= 16; - bitCount += 16; - } - } - while ((bitStream & 3) == 3) { - n0 += 3; - bitStream >>= 2; - bitCount += 2; - } - n0 += bitStream & 3; - bitCount += 2; - if (n0 > *maxSVPtr) - return ERROR(maxSymbolValue_tooSmall); - while (charnum < n0) - normalizedCounter[charnum++] = 0; - if ((ip <= iend - 7) || (ip + (bitCount >> 3) <= iend - 4)) { - ip += bitCount >> 3; - bitCount &= 7; - bitStream = ZSTD_readLE32(ip) >> bitCount; - } else { - bitStream >>= 2; - } - } - { - int const max = (2 * threshold - 1) - remaining; - int count; - - if ((bitStream & (threshold - 1)) < (U32)max) { - count = bitStream & (threshold - 1); - bitCount += nbBits - 1; - } else { - count = bitStream & (2 * threshold - 1); - if (count >= threshold) - count -= max; - bitCount += nbBits; - } - - count--; /* extra accuracy */ - remaining -= count < 0 ? -count : count; /* -1 means +1 */ - normalizedCounter[charnum++] = (short)count; - previous0 = !count; - while (remaining < threshold) { - nbBits--; - threshold >>= 1; - } - - if ((ip <= iend - 7) || (ip + (bitCount >> 3) <= iend - 4)) { - ip += bitCount >> 3; - bitCount &= 7; - } else { - bitCount -= (int)(8 * (iend - 4 - ip)); - ip = iend - 4; - } - bitStream = ZSTD_readLE32(ip) >> (bitCount & 31); - } - } /* while ((remaining>1) & (charnum<=*maxSVPtr)) */ - if (remaining != 1) - return ERROR(corruption_detected); - if (bitCount > 32) - return ERROR(corruption_detected); - *maxSVPtr = charnum - 1; - - ip += (bitCount + 7) >> 3; - return ip - istart; -} - -/*! HUF_readStats() : - Read compact Huffman tree, saved by HUF_writeCTable(). - `huffWeight` is destination buffer. - `rankStats` is assumed to be a table of at least HUF_TABLELOG_MAX U32. - @return : size read from `src` , or an error Code . - Note : Needed by HUF_readCTable() and HUF_readDTableX?() . -*/ -size_t HUF_readStats_wksp(BYTE *huffWeight, size_t hwSize, U32 *rankStats, U32 *nbSymbolsPtr, U32 *tableLogPtr, const void *src, size_t srcSize, void *workspace, size_t workspaceSize) -{ - U32 weightTotal; - const BYTE *ip = (const BYTE *)src; - size_t iSize; - size_t oSize; - - if (!srcSize) - return ERROR(srcSize_wrong); - iSize = ip[0]; - /* memset(huffWeight, 0, hwSize); */ /* is not necessary, even though some analyzer complain ... */ - - if (iSize >= 128) { /* special header */ - oSize = iSize - 127; - iSize = ((oSize + 1) / 2); - if (iSize + 1 > srcSize) - return ERROR(srcSize_wrong); - if (oSize >= hwSize) - return ERROR(corruption_detected); - ip += 1; - { - U32 n; - for (n = 0; n < oSize; n += 2) { - huffWeight[n] = ip[n / 2] >> 4; - huffWeight[n + 1] = ip[n / 2] & 15; - } - } - } else { /* header compressed with FSE (normal case) */ - if (iSize + 1 > srcSize) - return ERROR(srcSize_wrong); - oSize = FSE_decompress_wksp(huffWeight, hwSize - 1, ip + 1, iSize, 6, workspace, workspaceSize); /* max (hwSize-1) values decoded, as last one is implied */ - if (FSE_isError(oSize)) - return oSize; - } - - /* collect weight stats */ - memset(rankStats, 0, (HUF_TABLELOG_MAX + 1) * sizeof(U32)); - weightTotal = 0; - { - U32 n; - for (n = 0; n < oSize; n++) { - if (huffWeight[n] >= HUF_TABLELOG_MAX) - return ERROR(corruption_detected); - rankStats[huffWeight[n]]++; - weightTotal += (1 << huffWeight[n]) >> 1; - } - } - if (weightTotal == 0) - return ERROR(corruption_detected); - - /* get last non-null symbol weight (implied, total must be 2^n) */ - { - U32 const tableLog = BIT_highbit32(weightTotal) + 1; - if (tableLog > HUF_TABLELOG_MAX) - return ERROR(corruption_detected); - *tableLogPtr = tableLog; - /* determine last weight */ - { - U32 const total = 1 << tableLog; - U32 const rest = total - weightTotal; - U32 const verif = 1 << BIT_highbit32(rest); - U32 const lastWeight = BIT_highbit32(rest) + 1; - if (verif != rest) - return ERROR(corruption_detected); /* last value must be a clean power of 2 */ - huffWeight[oSize] = (BYTE)lastWeight; - rankStats[lastWeight]++; - } - } - - /* check tree construction validity */ - if ((rankStats[1] < 2) || (rankStats[1] & 1)) - return ERROR(corruption_detected); /* by construction : at least 2 elts of rank 1, must be even */ - - /* results */ - *nbSymbolsPtr = (U32)(oSize + 1); - return iSize + 1; -} diff --git a/lib/zstd/error_private.h b/lib/zstd/error_private.h deleted file mode 100644 index 1a60b31f706c..000000000000 --- a/lib/zstd/error_private.h +++ /dev/null @@ -1,53 +0,0 @@ -/** - * Copyright (c) 2016-present, Yann Collet, Facebook, Inc. - * All rights reserved. - * - * This source code is licensed under the BSD-style license found in the - * LICENSE file in the root directory of https://github.com/facebook/zstd. - * An additional grant of patent rights can be found in the PATENTS file in the - * same directory. - * - * This program is free software; you can redistribute it and/or modify it under - * the terms of the GNU General Public License version 2 as published by the - * Free Software Foundation. This program is dual-licensed; you may select - * either version 2 of the GNU General Public License ("GPL") or BSD license - * ("BSD"). - */ - -/* Note : this module is expected to remain private, do not expose it */ - -#ifndef ERROR_H_MODULE -#define ERROR_H_MODULE - -/* **************************************** -* Dependencies -******************************************/ -#include /* size_t */ -#include /* enum list */ - -/* **************************************** -* Compiler-specific -******************************************/ -#define ERR_STATIC static __attribute__((unused)) - -/*-**************************************** -* Customization (error_public.h) -******************************************/ -typedef ZSTD_ErrorCode ERR_enum; -#define PREFIX(name) ZSTD_error_##name - -/*-**************************************** -* Error codes handling -******************************************/ -#define ERROR(name) ((size_t)-PREFIX(name)) - -ERR_STATIC unsigned ERR_isError(size_t code) { return (code > ERROR(maxCode)); } - -ERR_STATIC ERR_enum ERR_getErrorCode(size_t code) -{ - if (!ERR_isError(code)) - return (ERR_enum)0; - return (ERR_enum)(0 - code); -} - -#endif /* ERROR_H_MODULE */ diff --git a/lib/zstd/fse.h b/lib/zstd/fse.h deleted file mode 100644 index 7460ab04b191..000000000000 --- a/lib/zstd/fse.h +++ /dev/null @@ -1,575 +0,0 @@ -/* - * FSE : Finite State Entropy codec - * Public Prototypes declaration - * Copyright (C) 2013-2016, Yann Collet. - * - * BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php) - * - * Redistribution and use in source and binary forms, with or without - * modification, are permitted provided that the following conditions are - * met: - * - * * Redistributions of source code must retain the above copyright - * notice, this list of conditions and the following disclaimer. - * * Redistributions in binary form must reproduce the above - * copyright notice, this list of conditions and the following disclaimer - * in the documentation and/or other materials provided with the - * distribution. - * - * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS - * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT - * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR - * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT - * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, - * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT - * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, - * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY - * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT - * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE - * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. - * - * This program is free software; you can redistribute it and/or modify it under - * the terms of the GNU General Public License version 2 as published by the - * Free Software Foundation. This program is dual-licensed; you may select - * either version 2 of the GNU General Public License ("GPL") or BSD license - * ("BSD"). - * - * You can contact the author at : - * - Source repository : https://github.com/Cyan4973/FiniteStateEntropy - */ -#ifndef FSE_H -#define FSE_H - -/*-***************************************** -* Dependencies -******************************************/ -#include /* size_t, ptrdiff_t */ - -/*-***************************************** -* FSE_PUBLIC_API : control library symbols visibility -******************************************/ -#define FSE_PUBLIC_API - -/*------ Version ------*/ -#define FSE_VERSION_MAJOR 0 -#define FSE_VERSION_MINOR 9 -#define FSE_VERSION_RELEASE 0 - -#define FSE_LIB_VERSION FSE_VERSION_MAJOR.FSE_VERSION_MINOR.FSE_VERSION_RELEASE -#define FSE_QUOTE(str) #str -#define FSE_EXPAND_AND_QUOTE(str) FSE_QUOTE(str) -#define FSE_VERSION_STRING FSE_EXPAND_AND_QUOTE(FSE_LIB_VERSION) - -#define FSE_VERSION_NUMBER (FSE_VERSION_MAJOR * 100 * 100 + FSE_VERSION_MINOR * 100 + FSE_VERSION_RELEASE) -FSE_PUBLIC_API unsigned FSE_versionNumber(void); /**< library version number; to be used when checking dll version */ - -/*-***************************************** -* Tool functions -******************************************/ -FSE_PUBLIC_API size_t FSE_compressBound(size_t size); /* maximum compressed size */ - -/* Error Management */ -FSE_PUBLIC_API unsigned FSE_isError(size_t code); /* tells if a return value is an error code */ - -/*-***************************************** -* FSE detailed API -******************************************/ -/*! -FSE_compress() does the following: -1. count symbol occurrence from source[] into table count[] -2. normalize counters so that sum(count[]) == Power_of_2 (2^tableLog) -3. save normalized counters to memory buffer using writeNCount() -4. build encoding table 'CTable' from normalized counters -5. encode the data stream using encoding table 'CTable' - -FSE_decompress() does the following: -1. read normalized counters with readNCount() -2. build decoding table 'DTable' from normalized counters -3. decode the data stream using decoding table 'DTable' - -The following API allows targeting specific sub-functions for advanced tasks. -For example, it's possible to compress several blocks using the same 'CTable', -or to save and provide normalized distribution using external method. -*/ - -/* *** COMPRESSION *** */ -/*! FSE_optimalTableLog(): - dynamically downsize 'tableLog' when conditions are met. - It saves CPU time, by using smaller tables, while preserving or even improving compression ratio. - @return : recommended tableLog (necessarily <= 'maxTableLog') */ -FSE_PUBLIC_API unsigned FSE_optimalTableLog(unsigned maxTableLog, size_t srcSize, unsigned maxSymbolValue); - -/*! FSE_normalizeCount(): - normalize counts so that sum(count[]) == Power_of_2 (2^tableLog) - 'normalizedCounter' is a table of short, of minimum size (maxSymbolValue+1). - @return : tableLog, - or an errorCode, which can be tested using FSE_isError() */ -FSE_PUBLIC_API size_t FSE_normalizeCount(short *normalizedCounter, unsigned tableLog, const unsigned *count, size_t srcSize, unsigned maxSymbolValue); - -/*! FSE_NCountWriteBound(): - Provides the maximum possible size of an FSE normalized table, given 'maxSymbolValue' and 'tableLog'. - Typically useful for allocation purpose. */ -FSE_PUBLIC_API size_t FSE_NCountWriteBound(unsigned maxSymbolValue, unsigned tableLog); - -/*! FSE_writeNCount(): - Compactly save 'normalizedCounter' into 'buffer'. - @return : size of the compressed table, - or an errorCode, which can be tested using FSE_isError(). */ -FSE_PUBLIC_API size_t FSE_writeNCount(void *buffer, size_t bufferSize, const short *normalizedCounter, unsigned maxSymbolValue, unsigned tableLog); - -/*! Constructor and Destructor of FSE_CTable. - Note that FSE_CTable size depends on 'tableLog' and 'maxSymbolValue' */ -typedef unsigned FSE_CTable; /* don't allocate that. It's only meant to be more restrictive than void* */ - -/*! FSE_compress_usingCTable(): - Compress `src` using `ct` into `dst` which must be already allocated. - @return : size of compressed data (<= `dstCapacity`), - or 0 if compressed data could not fit into `dst`, - or an errorCode, which can be tested using FSE_isError() */ -FSE_PUBLIC_API size_t FSE_compress_usingCTable(void *dst, size_t dstCapacity, const void *src, size_t srcSize, const FSE_CTable *ct); - -/*! -Tutorial : ----------- -The first step is to count all symbols. FSE_count() does this job very fast. -Result will be saved into 'count', a table of unsigned int, which must be already allocated, and have 'maxSymbolValuePtr[0]+1' cells. -'src' is a table of bytes of size 'srcSize'. All values within 'src' MUST be <= maxSymbolValuePtr[0] -maxSymbolValuePtr[0] will be updated, with its real value (necessarily <= original value) -FSE_count() will return the number of occurrence of the most frequent symbol. -This can be used to know if there is a single symbol within 'src', and to quickly evaluate its compressibility. -If there is an error, the function will return an ErrorCode (which can be tested using FSE_isError()). - -The next step is to normalize the frequencies. -FSE_normalizeCount() will ensure that sum of frequencies is == 2 ^'tableLog'. -It also guarantees a minimum of 1 to any Symbol with frequency >= 1. -You can use 'tableLog'==0 to mean "use default tableLog value". -If you are unsure of which tableLog value to use, you can ask FSE_optimalTableLog(), -which will provide the optimal valid tableLog given sourceSize, maxSymbolValue, and a user-defined maximum (0 means "default"). - -The result of FSE_normalizeCount() will be saved into a table, -called 'normalizedCounter', which is a table of signed short. -'normalizedCounter' must be already allocated, and have at least 'maxSymbolValue+1' cells. -The return value is tableLog if everything proceeded as expected. -It is 0 if there is a single symbol within distribution. -If there is an error (ex: invalid tableLog value), the function will return an ErrorCode (which can be tested using FSE_isError()). - -'normalizedCounter' can be saved in a compact manner to a memory area using FSE_writeNCount(). -'buffer' must be already allocated. -For guaranteed success, buffer size must be at least FSE_headerBound(). -The result of the function is the number of bytes written into 'buffer'. -If there is an error, the function will return an ErrorCode (which can be tested using FSE_isError(); ex : buffer size too small). - -'normalizedCounter' can then be used to create the compression table 'CTable'. -The space required by 'CTable' must be already allocated, using FSE_createCTable(). -You can then use FSE_buildCTable() to fill 'CTable'. -If there is an error, both functions will return an ErrorCode (which can be tested using FSE_isError()). - -'CTable' can then be used to compress 'src', with FSE_compress_usingCTable(). -Similar to FSE_count(), the convention is that 'src' is assumed to be a table of char of size 'srcSize' -The function returns the size of compressed data (without header), necessarily <= `dstCapacity`. -If it returns '0', compressed data could not fit into 'dst'. -If there is an error, the function will return an ErrorCode (which can be tested using FSE_isError()). -*/ - -/* *** DECOMPRESSION *** */ - -/*! FSE_readNCount(): - Read compactly saved 'normalizedCounter' from 'rBuffer'. - @return : size read from 'rBuffer', - or an errorCode, which can be tested using FSE_isError(). - maxSymbolValuePtr[0] and tableLogPtr[0] will also be updated with their respective values */ -FSE_PUBLIC_API size_t FSE_readNCount(short *normalizedCounter, unsigned *maxSymbolValuePtr, unsigned *tableLogPtr, const void *rBuffer, size_t rBuffSize); - -/*! Constructor and Destructor of FSE_DTable. - Note that its size depends on 'tableLog' */ -typedef unsigned FSE_DTable; /* don't allocate that. It's just a way to be more restrictive than void* */ - -/*! FSE_buildDTable(): - Builds 'dt', which must be already allocated, using FSE_createDTable(). - return : 0, or an errorCode, which can be tested using FSE_isError() */ -FSE_PUBLIC_API size_t FSE_buildDTable_wksp(FSE_DTable *dt, const short *normalizedCounter, unsigned maxSymbolValue, unsigned tableLog, void *workspace, size_t workspaceSize); - -/*! FSE_decompress_usingDTable(): - Decompress compressed source `cSrc` of size `cSrcSize` using `dt` - into `dst` which must be already allocated. - @return : size of regenerated data (necessarily <= `dstCapacity`), - or an errorCode, which can be tested using FSE_isError() */ -FSE_PUBLIC_API size_t FSE_decompress_usingDTable(void *dst, size_t dstCapacity, const void *cSrc, size_t cSrcSize, const FSE_DTable *dt); - -/*! -Tutorial : ----------- -(Note : these functions only decompress FSE-compressed blocks. - If block is uncompressed, use memcpy() instead - If block is a single repeated byte, use memset() instead ) - -The first step is to obtain the normalized frequencies of symbols. -This can be performed by FSE_readNCount() if it was saved using FSE_writeNCount(). -'normalizedCounter' must be already allocated, and have at least 'maxSymbolValuePtr[0]+1' cells of signed short. -In practice, that means it's necessary to know 'maxSymbolValue' beforehand, -or size the table to handle worst case situations (typically 256). -FSE_readNCount() will provide 'tableLog' and 'maxSymbolValue'. -The result of FSE_readNCount() is the number of bytes read from 'rBuffer'. -Note that 'rBufferSize' must be at least 4 bytes, even if useful information is less than that. -If there is an error, the function will return an error code, which can be tested using FSE_isError(). - -The next step is to build the decompression tables 'FSE_DTable' from 'normalizedCounter'. -This is performed by the function FSE_buildDTable(). -The space required by 'FSE_DTable' must be already allocated using FSE_createDTable(). -If there is an error, the function will return an error code, which can be tested using FSE_isError(). - -`FSE_DTable` can then be used to decompress `cSrc`, with FSE_decompress_usingDTable(). -`cSrcSize` must be strictly correct, otherwise decompression will fail. -FSE_decompress_usingDTable() result will tell how many bytes were regenerated (<=`dstCapacity`). -If there is an error, the function will return an error code, which can be tested using FSE_isError(). (ex: dst buffer too small) -*/ - -/* *** Dependency *** */ -#include "bitstream.h" - -/* ***************************************** -* Static allocation -*******************************************/ -/* FSE buffer bounds */ -#define FSE_NCOUNTBOUND 512 -#define FSE_BLOCKBOUND(size) (size + (size >> 7)) -#define FSE_COMPRESSBOUND(size) (FSE_NCOUNTBOUND + FSE_BLOCKBOUND(size)) /* Macro version, useful for static allocation */ - -/* It is possible to statically allocate FSE CTable/DTable as a table of FSE_CTable/FSE_DTable using below macros */ -#define FSE_CTABLE_SIZE_U32(maxTableLog, maxSymbolValue) (1 + (1 << (maxTableLog - 1)) + ((maxSymbolValue + 1) * 2)) -#define FSE_DTABLE_SIZE_U32(maxTableLog) (1 + (1 << maxTableLog)) - -/* ***************************************** -* FSE advanced API -*******************************************/ -/* FSE_count_wksp() : - * Same as FSE_count(), but using an externally provided scratch buffer. - * `workSpace` size must be table of >= `1024` unsigned - */ -size_t FSE_count_wksp(unsigned *count, unsigned *maxSymbolValuePtr, const void *source, size_t sourceSize, unsigned *workSpace); - -/* FSE_countFast_wksp() : - * Same as FSE_countFast(), but using an externally provided scratch buffer. - * `workSpace` must be a table of minimum `1024` unsigned - */ -size_t FSE_countFast_wksp(unsigned *count, unsigned *maxSymbolValuePtr, const void *src, size_t srcSize, unsigned *workSpace); - -/*! FSE_count_simple - * Same as FSE_countFast(), but does not use any additional memory (not even on stack). - * This function is unsafe, and will segfault if any value within `src` is `> *maxSymbolValuePtr` (presuming it's also the size of `count`). -*/ -size_t FSE_count_simple(unsigned *count, unsigned *maxSymbolValuePtr, const void *src, size_t srcSize); - -unsigned FSE_optimalTableLog_internal(unsigned maxTableLog, size_t srcSize, unsigned maxSymbolValue, unsigned minus); -/**< same as FSE_optimalTableLog(), which used `minus==2` */ - -size_t FSE_buildCTable_raw(FSE_CTable *ct, unsigned nbBits); -/**< build a fake FSE_CTable, designed for a flat distribution, where each symbol uses nbBits */ - -size_t FSE_buildCTable_rle(FSE_CTable *ct, unsigned char symbolValue); -/**< build a fake FSE_CTable, designed to compress always the same symbolValue */ - -/* FSE_buildCTable_wksp() : - * Same as FSE_buildCTable(), but using an externally allocated scratch buffer (`workSpace`). - * `wkspSize` must be >= `(1<= BIT_DStream_completed - -When it's done, verify decompression is fully completed, by checking both DStream and the relevant states. -Checking if DStream has reached its end is performed by : - BIT_endOfDStream(&DStream); -Check also the states. There might be some symbols left there, if some high probability ones (>50%) are possible. - FSE_endOfDState(&DState); -*/ - -/* ***************************************** -* FSE unsafe API -*******************************************/ -static unsigned char FSE_decodeSymbolFast(FSE_DState_t *DStatePtr, BIT_DStream_t *bitD); -/* faster, but works only if nbBits is always >= 1 (otherwise, result will be corrupted) */ - -/* ***************************************** -* Implementation of inlined functions -*******************************************/ -typedef struct { - int deltaFindState; - U32 deltaNbBits; -} FSE_symbolCompressionTransform; /* total 8 bytes */ - -ZSTD_STATIC void FSE_initCState(FSE_CState_t *statePtr, const FSE_CTable *ct) -{ - const void *ptr = ct; - const U16 *u16ptr = (const U16 *)ptr; - const U32 tableLog = ZSTD_read16(ptr); - statePtr->value = (ptrdiff_t)1 << tableLog; - statePtr->stateTable = u16ptr + 2; - statePtr->symbolTT = ((const U32 *)ct + 1 + (tableLog ? (1 << (tableLog - 1)) : 1)); - statePtr->stateLog = tableLog; -} - -/*! FSE_initCState2() : -* Same as FSE_initCState(), but the first symbol to include (which will be the last to be read) -* uses the smallest state value possible, saving the cost of this symbol */ -ZSTD_STATIC void FSE_initCState2(FSE_CState_t *statePtr, const FSE_CTable *ct, U32 symbol) -{ - FSE_initCState(statePtr, ct); - { - const FSE_symbolCompressionTransform symbolTT = ((const FSE_symbolCompressionTransform *)(statePtr->symbolTT))[symbol]; - const U16 *stateTable = (const U16 *)(statePtr->stateTable); - U32 nbBitsOut = (U32)((symbolTT.deltaNbBits + (1 << 15)) >> 16); - statePtr->value = (nbBitsOut << 16) - symbolTT.deltaNbBits; - statePtr->value = stateTable[(statePtr->value >> nbBitsOut) + symbolTT.deltaFindState]; - } -} - -ZSTD_STATIC void FSE_encodeSymbol(BIT_CStream_t *bitC, FSE_CState_t *statePtr, U32 symbol) -{ - const FSE_symbolCompressionTransform symbolTT = ((const FSE_symbolCompressionTransform *)(statePtr->symbolTT))[symbol]; - const U16 *const stateTable = (const U16 *)(statePtr->stateTable); - U32 nbBitsOut = (U32)((statePtr->value + symbolTT.deltaNbBits) >> 16); - BIT_addBits(bitC, statePtr->value, nbBitsOut); - statePtr->value = stateTable[(statePtr->value >> nbBitsOut) + symbolTT.deltaFindState]; -} - -ZSTD_STATIC void FSE_flushCState(BIT_CStream_t *bitC, const FSE_CState_t *statePtr) -{ - BIT_addBits(bitC, statePtr->value, statePtr->stateLog); - BIT_flushBits(bitC); -} - -/* ====== Decompression ====== */ - -typedef struct { - U16 tableLog; - U16 fastMode; -} FSE_DTableHeader; /* sizeof U32 */ - -typedef struct { - unsigned short newState; - unsigned char symbol; - unsigned char nbBits; -} FSE_decode_t; /* size == U32 */ - -ZSTD_STATIC void FSE_initDState(FSE_DState_t *DStatePtr, BIT_DStream_t *bitD, const FSE_DTable *dt) -{ - const void *ptr = dt; - const FSE_DTableHeader *const DTableH = (const FSE_DTableHeader *)ptr; - DStatePtr->state = BIT_readBits(bitD, DTableH->tableLog); - BIT_reloadDStream(bitD); - DStatePtr->table = dt + 1; -} - -ZSTD_STATIC BYTE FSE_peekSymbol(const FSE_DState_t *DStatePtr) -{ - FSE_decode_t const DInfo = ((const FSE_decode_t *)(DStatePtr->table))[DStatePtr->state]; - return DInfo.symbol; -} - -ZSTD_STATIC void FSE_updateState(FSE_DState_t *DStatePtr, BIT_DStream_t *bitD) -{ - FSE_decode_t const DInfo = ((const FSE_decode_t *)(DStatePtr->table))[DStatePtr->state]; - U32 const nbBits = DInfo.nbBits; - size_t const lowBits = BIT_readBits(bitD, nbBits); - DStatePtr->state = DInfo.newState + lowBits; -} - -ZSTD_STATIC BYTE FSE_decodeSymbol(FSE_DState_t *DStatePtr, BIT_DStream_t *bitD) -{ - FSE_decode_t const DInfo = ((const FSE_decode_t *)(DStatePtr->table))[DStatePtr->state]; - U32 const nbBits = DInfo.nbBits; - BYTE const symbol = DInfo.symbol; - size_t const lowBits = BIT_readBits(bitD, nbBits); - - DStatePtr->state = DInfo.newState + lowBits; - return symbol; -} - -/*! FSE_decodeSymbolFast() : - unsafe, only works if no symbol has a probability > 50% */ -ZSTD_STATIC BYTE FSE_decodeSymbolFast(FSE_DState_t *DStatePtr, BIT_DStream_t *bitD) -{ - FSE_decode_t const DInfo = ((const FSE_decode_t *)(DStatePtr->table))[DStatePtr->state]; - U32 const nbBits = DInfo.nbBits; - BYTE const symbol = DInfo.symbol; - size_t const lowBits = BIT_readBitsFast(bitD, nbBits); - - DStatePtr->state = DInfo.newState + lowBits; - return symbol; -} - -ZSTD_STATIC unsigned FSE_endOfDState(const FSE_DState_t *DStatePtr) { return DStatePtr->state == 0; } - -/* ************************************************************** -* Tuning parameters -****************************************************************/ -/*!MEMORY_USAGE : -* Memory usage formula : N->2^N Bytes (examples : 10 -> 1KB; 12 -> 4KB ; 16 -> 64KB; 20 -> 1MB; etc.) -* Increasing memory usage improves compression ratio -* Reduced memory usage can improve speed, due to cache effect -* Recommended max value is 14, for 16KB, which nicely fits into Intel x86 L1 cache */ -#ifndef FSE_MAX_MEMORY_USAGE -#define FSE_MAX_MEMORY_USAGE 14 -#endif -#ifndef FSE_DEFAULT_MEMORY_USAGE -#define FSE_DEFAULT_MEMORY_USAGE 13 -#endif - -/*!FSE_MAX_SYMBOL_VALUE : -* Maximum symbol value authorized. -* Required for proper stack allocation */ -#ifndef FSE_MAX_SYMBOL_VALUE -#define FSE_MAX_SYMBOL_VALUE 255 -#endif - -/* ************************************************************** -* template functions type & suffix -****************************************************************/ -#define FSE_FUNCTION_TYPE BYTE -#define FSE_FUNCTION_EXTENSION -#define FSE_DECODE_TYPE FSE_decode_t - -/* *************************************************************** -* Constants -*****************************************************************/ -#define FSE_MAX_TABLELOG (FSE_MAX_MEMORY_USAGE - 2) -#define FSE_MAX_TABLESIZE (1U << FSE_MAX_TABLELOG) -#define FSE_MAXTABLESIZE_MASK (FSE_MAX_TABLESIZE - 1) -#define FSE_DEFAULT_TABLELOG (FSE_DEFAULT_MEMORY_USAGE - 2) -#define FSE_MIN_TABLELOG 5 - -#define FSE_TABLELOG_ABSOLUTE_MAX 15 -#if FSE_MAX_TABLELOG > FSE_TABLELOG_ABSOLUTE_MAX -#error "FSE_MAX_TABLELOG > FSE_TABLELOG_ABSOLUTE_MAX is not supported" -#endif - -#define FSE_TABLESTEP(tableSize) ((tableSize >> 1) + (tableSize >> 3) + 3) - -#endif /* FSE_H */ diff --git a/lib/zstd/fse_compress.c b/lib/zstd/fse_compress.c deleted file mode 100644 index ef3d1741d532..000000000000 --- a/lib/zstd/fse_compress.c +++ /dev/null @@ -1,795 +0,0 @@ -/* - * FSE : Finite State Entropy encoder - * Copyright (C) 2013-2015, Yann Collet. - * - * BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php) - * - * Redistribution and use in source and binary forms, with or without - * modification, are permitted provided that the following conditions are - * met: - * - * * Redistributions of source code must retain the above copyright - * notice, this list of conditions and the following disclaimer. - * * Redistributions in binary form must reproduce the above - * copyright notice, this list of conditions and the following disclaimer - * in the documentation and/or other materials provided with the - * distribution. - * - * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS - * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT - * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR - * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT - * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, - * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT - * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, - * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY - * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT - * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE - * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. - * - * This program is free software; you can redistribute it and/or modify it under - * the terms of the GNU General Public License version 2 as published by the - * Free Software Foundation. This program is dual-licensed; you may select - * either version 2 of the GNU General Public License ("GPL") or BSD license - * ("BSD"). - * - * You can contact the author at : - * - Source repository : https://github.com/Cyan4973/FiniteStateEntropy - */ - -/* ************************************************************** -* Compiler specifics -****************************************************************/ -#define FORCE_INLINE static __always_inline - -/* ************************************************************** -* Includes -****************************************************************/ -#include "bitstream.h" -#include "fse.h" -#include -#include -#include -#include /* memcpy, memset */ - -/* ************************************************************** -* Error Management -****************************************************************/ -#define FSE_STATIC_ASSERT(c) \ - { \ - enum { FSE_static_assert = 1 / (int)(!!(c)) }; \ - } /* use only *after* variable declarations */ - -/* ************************************************************** -* Templates -****************************************************************/ -/* - designed to be included - for type-specific functions (template emulation in C) - Objective is to write these functions only once, for improved maintenance -*/ - -/* safety checks */ -#ifndef FSE_FUNCTION_EXTENSION -#error "FSE_FUNCTION_EXTENSION must be defined" -#endif -#ifndef FSE_FUNCTION_TYPE -#error "FSE_FUNCTION_TYPE must be defined" -#endif - -/* Function names */ -#define FSE_CAT(X, Y) X##Y -#define FSE_FUNCTION_NAME(X, Y) FSE_CAT(X, Y) -#define FSE_TYPE_NAME(X, Y) FSE_CAT(X, Y) - -/* Function templates */ - -/* FSE_buildCTable_wksp() : - * Same as FSE_buildCTable(), but using an externally allocated scratch buffer (`workSpace`). - * wkspSize should be sized to handle worst case situation, which is `1<> 1 : 1); - FSE_symbolCompressionTransform *const symbolTT = (FSE_symbolCompressionTransform *)(FSCT); - U32 const step = FSE_TABLESTEP(tableSize); - U32 highThreshold = tableSize - 1; - - U32 *cumul; - FSE_FUNCTION_TYPE *tableSymbol; - size_t spaceUsed32 = 0; - - cumul = (U32 *)workspace + spaceUsed32; - spaceUsed32 += FSE_MAX_SYMBOL_VALUE + 2; - tableSymbol = (FSE_FUNCTION_TYPE *)((U32 *)workspace + spaceUsed32); - spaceUsed32 += ALIGN(sizeof(FSE_FUNCTION_TYPE) * ((size_t)1 << tableLog), sizeof(U32)) >> 2; - - if ((spaceUsed32 << 2) > workspaceSize) - return ERROR(tableLog_tooLarge); - workspace = (U32 *)workspace + spaceUsed32; - workspaceSize -= (spaceUsed32 << 2); - - /* CTable header */ - tableU16[-2] = (U16)tableLog; - tableU16[-1] = (U16)maxSymbolValue; - - /* For explanations on how to distribute symbol values over the table : - * http://fastcompression.blogspot.fr/2014/02/fse-distributing-symbol-values.html */ - - /* symbol start positions */ - { - U32 u; - cumul[0] = 0; - for (u = 1; u <= maxSymbolValue + 1; u++) { - if (normalizedCounter[u - 1] == -1) { /* Low proba symbol */ - cumul[u] = cumul[u - 1] + 1; - tableSymbol[highThreshold--] = (FSE_FUNCTION_TYPE)(u - 1); - } else { - cumul[u] = cumul[u - 1] + normalizedCounter[u - 1]; - } - } - cumul[maxSymbolValue + 1] = tableSize + 1; - } - - /* Spread symbols */ - { - U32 position = 0; - U32 symbol; - for (symbol = 0; symbol <= maxSymbolValue; symbol++) { - int nbOccurences; - for (nbOccurences = 0; nbOccurences < normalizedCounter[symbol]; nbOccurences++) { - tableSymbol[position] = (FSE_FUNCTION_TYPE)symbol; - position = (position + step) & tableMask; - while (position > highThreshold) - position = (position + step) & tableMask; /* Low proba area */ - } - } - - if (position != 0) - return ERROR(GENERIC); /* Must have gone through all positions */ - } - - /* Build table */ - { - U32 u; - for (u = 0; u < tableSize; u++) { - FSE_FUNCTION_TYPE s = tableSymbol[u]; /* note : static analyzer may not understand tableSymbol is properly initialized */ - tableU16[cumul[s]++] = (U16)(tableSize + u); /* TableU16 : sorted by symbol order; gives next state value */ - } - } - - /* Build Symbol Transformation Table */ - { - unsigned total = 0; - unsigned s; - for (s = 0; s <= maxSymbolValue; s++) { - switch (normalizedCounter[s]) { - case 0: break; - - case -1: - case 1: - symbolTT[s].deltaNbBits = (tableLog << 16) - (1 << tableLog); - symbolTT[s].deltaFindState = total - 1; - total++; - break; - default: { - U32 const maxBitsOut = tableLog - BIT_highbit32(normalizedCounter[s] - 1); - U32 const minStatePlus = normalizedCounter[s] << maxBitsOut; - symbolTT[s].deltaNbBits = (maxBitsOut << 16) - minStatePlus; - symbolTT[s].deltaFindState = total - normalizedCounter[s]; - total += normalizedCounter[s]; - } - } - } - } - - return 0; -} - -/*-************************************************************** -* FSE NCount encoding-decoding -****************************************************************/ -size_t FSE_NCountWriteBound(unsigned maxSymbolValue, unsigned tableLog) -{ - size_t const maxHeaderSize = (((maxSymbolValue + 1) * tableLog) >> 3) + 3; - return maxSymbolValue ? maxHeaderSize : FSE_NCOUNTBOUND; /* maxSymbolValue==0 ? use default */ -} - -static size_t FSE_writeNCount_generic(void *header, size_t headerBufferSize, const short *normalizedCounter, unsigned maxSymbolValue, unsigned tableLog, - unsigned writeIsSafe) -{ - BYTE *const ostart = (BYTE *)header; - BYTE *out = ostart; - BYTE *const oend = ostart + headerBufferSize; - int nbBits; - const int tableSize = 1 << tableLog; - int remaining; - int threshold; - U32 bitStream; - int bitCount; - unsigned charnum = 0; - int previous0 = 0; - - bitStream = 0; - bitCount = 0; - /* Table Size */ - bitStream += (tableLog - FSE_MIN_TABLELOG) << bitCount; - bitCount += 4; - - /* Init */ - remaining = tableSize + 1; /* +1 for extra accuracy */ - threshold = tableSize; - nbBits = tableLog + 1; - - while (remaining > 1) { /* stops at 1 */ - if (previous0) { - unsigned start = charnum; - while (!normalizedCounter[charnum]) - charnum++; - while (charnum >= start + 24) { - start += 24; - bitStream += 0xFFFFU << bitCount; - if ((!writeIsSafe) && (out > oend - 2)) - return ERROR(dstSize_tooSmall); /* Buffer overflow */ - out[0] = (BYTE)bitStream; - out[1] = (BYTE)(bitStream >> 8); - out += 2; - bitStream >>= 16; - } - while (charnum >= start + 3) { - start += 3; - bitStream += 3 << bitCount; - bitCount += 2; - } - bitStream += (charnum - start) << bitCount; - bitCount += 2; - if (bitCount > 16) { - if ((!writeIsSafe) && (out > oend - 2)) - return ERROR(dstSize_tooSmall); /* Buffer overflow */ - out[0] = (BYTE)bitStream; - out[1] = (BYTE)(bitStream >> 8); - out += 2; - bitStream >>= 16; - bitCount -= 16; - } - } - { - int count = normalizedCounter[charnum++]; - int const max = (2 * threshold - 1) - remaining; - remaining -= count < 0 ? -count : count; - count++; /* +1 for extra accuracy */ - if (count >= threshold) - count += max; /* [0..max[ [max..threshold[ (...) [threshold+max 2*threshold[ */ - bitStream += count << bitCount; - bitCount += nbBits; - bitCount -= (count < max); - previous0 = (count == 1); - if (remaining < 1) - return ERROR(GENERIC); - while (remaining < threshold) - nbBits--, threshold >>= 1; - } - if (bitCount > 16) { - if ((!writeIsSafe) && (out > oend - 2)) - return ERROR(dstSize_tooSmall); /* Buffer overflow */ - out[0] = (BYTE)bitStream; - out[1] = (BYTE)(bitStream >> 8); - out += 2; - bitStream >>= 16; - bitCount -= 16; - } - } - - /* flush remaining bitStream */ - if ((!writeIsSafe) && (out > oend - 2)) - return ERROR(dstSize_tooSmall); /* Buffer overflow */ - out[0] = (BYTE)bitStream; - out[1] = (BYTE)(bitStream >> 8); - out += (bitCount + 7) / 8; - - if (charnum > maxSymbolValue + 1) - return ERROR(GENERIC); - - return (out - ostart); -} - -size_t FSE_writeNCount(void *buffer, size_t bufferSize, const short *normalizedCounter, unsigned maxSymbolValue, unsigned tableLog) -{ - if (tableLog > FSE_MAX_TABLELOG) - return ERROR(tableLog_tooLarge); /* Unsupported */ - if (tableLog < FSE_MIN_TABLELOG) - return ERROR(GENERIC); /* Unsupported */ - - if (bufferSize < FSE_NCountWriteBound(maxSymbolValue, tableLog)) - return FSE_writeNCount_generic(buffer, bufferSize, normalizedCounter, maxSymbolValue, tableLog, 0); - - return FSE_writeNCount_generic(buffer, bufferSize, normalizedCounter, maxSymbolValue, tableLog, 1); -} - -/*-************************************************************** -* Counting histogram -****************************************************************/ -/*! FSE_count_simple - This function counts byte values within `src`, and store the histogram into table `count`. - It doesn't use any additional memory. - But this function is unsafe : it doesn't check that all values within `src` can fit into `count`. - For this reason, prefer using a table `count` with 256 elements. - @return : count of most numerous element -*/ -size_t FSE_count_simple(unsigned *count, unsigned *maxSymbolValuePtr, const void *src, size_t srcSize) -{ - const BYTE *ip = (const BYTE *)src; - const BYTE *const end = ip + srcSize; - unsigned maxSymbolValue = *maxSymbolValuePtr; - unsigned max = 0; - - memset(count, 0, (maxSymbolValue + 1) * sizeof(*count)); - if (srcSize == 0) { - *maxSymbolValuePtr = 0; - return 0; - } - - while (ip < end) - count[*ip++]++; - - while (!count[maxSymbolValue]) - maxSymbolValue--; - *maxSymbolValuePtr = maxSymbolValue; - - { - U32 s; - for (s = 0; s <= maxSymbolValue; s++) - if (count[s] > max) - max = count[s]; - } - - return (size_t)max; -} - -/* FSE_count_parallel_wksp() : - * Same as FSE_count_parallel(), but using an externally provided scratch buffer. - * `workSpace` size must be a minimum of `1024 * sizeof(unsigned)`` */ -static size_t FSE_count_parallel_wksp(unsigned *count, unsigned *maxSymbolValuePtr, const void *source, size_t sourceSize, unsigned checkMax, - unsigned *const workSpace) -{ - const BYTE *ip = (const BYTE *)source; - const BYTE *const iend = ip + sourceSize; - unsigned maxSymbolValue = *maxSymbolValuePtr; - unsigned max = 0; - U32 *const Counting1 = workSpace; - U32 *const Counting2 = Counting1 + 256; - U32 *const Counting3 = Counting2 + 256; - U32 *const Counting4 = Counting3 + 256; - - memset(Counting1, 0, 4 * 256 * sizeof(unsigned)); - - /* safety checks */ - if (!sourceSize) { - memset(count, 0, maxSymbolValue + 1); - *maxSymbolValuePtr = 0; - return 0; - } - if (!maxSymbolValue) - maxSymbolValue = 255; /* 0 == default */ - - /* by stripes of 16 bytes */ - { - U32 cached = ZSTD_read32(ip); - ip += 4; - while (ip < iend - 15) { - U32 c = cached; - cached = ZSTD_read32(ip); - ip += 4; - Counting1[(BYTE)c]++; - Counting2[(BYTE)(c >> 8)]++; - Counting3[(BYTE)(c >> 16)]++; - Counting4[c >> 24]++; - c = cached; - cached = ZSTD_read32(ip); - ip += 4; - Counting1[(BYTE)c]++; - Counting2[(BYTE)(c >> 8)]++; - Counting3[(BYTE)(c >> 16)]++; - Counting4[c >> 24]++; - c = cached; - cached = ZSTD_read32(ip); - ip += 4; - Counting1[(BYTE)c]++; - Counting2[(BYTE)(c >> 8)]++; - Counting3[(BYTE)(c >> 16)]++; - Counting4[c >> 24]++; - c = cached; - cached = ZSTD_read32(ip); - ip += 4; - Counting1[(BYTE)c]++; - Counting2[(BYTE)(c >> 8)]++; - Counting3[(BYTE)(c >> 16)]++; - Counting4[c >> 24]++; - } - ip -= 4; - } - - /* finish last symbols */ - while (ip < iend) - Counting1[*ip++]++; - - if (checkMax) { /* verify stats will fit into destination table */ - U32 s; - for (s = 255; s > maxSymbolValue; s--) { - Counting1[s] += Counting2[s] + Counting3[s] + Counting4[s]; - if (Counting1[s]) - return ERROR(maxSymbolValue_tooSmall); - } - } - - { - U32 s; - for (s = 0; s <= maxSymbolValue; s++) { - count[s] = Counting1[s] + Counting2[s] + Counting3[s] + Counting4[s]; - if (count[s] > max) - max = count[s]; - } - } - - while (!count[maxSymbolValue]) - maxSymbolValue--; - *maxSymbolValuePtr = maxSymbolValue; - return (size_t)max; -} - -/* FSE_countFast_wksp() : - * Same as FSE_countFast(), but using an externally provided scratch buffer. - * `workSpace` size must be table of >= `1024` unsigned */ -size_t FSE_countFast_wksp(unsigned *count, unsigned *maxSymbolValuePtr, const void *source, size_t sourceSize, unsigned *workSpace) -{ - if (sourceSize < 1500) - return FSE_count_simple(count, maxSymbolValuePtr, source, sourceSize); - return FSE_count_parallel_wksp(count, maxSymbolValuePtr, source, sourceSize, 0, workSpace); -} - -/* FSE_count_wksp() : - * Same as FSE_count(), but using an externally provided scratch buffer. - * `workSpace` size must be table of >= `1024` unsigned */ -size_t FSE_count_wksp(unsigned *count, unsigned *maxSymbolValuePtr, const void *source, size_t sourceSize, unsigned *workSpace) -{ - if (*maxSymbolValuePtr < 255) - return FSE_count_parallel_wksp(count, maxSymbolValuePtr, source, sourceSize, 1, workSpace); - *maxSymbolValuePtr = 255; - return FSE_countFast_wksp(count, maxSymbolValuePtr, source, sourceSize, workSpace); -} - -/*-************************************************************** -* FSE Compression Code -****************************************************************/ -/*! FSE_sizeof_CTable() : - FSE_CTable is a variable size structure which contains : - `U16 tableLog;` - `U16 maxSymbolValue;` - `U16 nextStateNumber[1 << tableLog];` // This size is variable - `FSE_symbolCompressionTransform symbolTT[maxSymbolValue+1];` // This size is variable -Allocation is manual (C standard does not support variable-size structures). -*/ -size_t FSE_sizeof_CTable(unsigned maxSymbolValue, unsigned tableLog) -{ - if (tableLog > FSE_MAX_TABLELOG) - return ERROR(tableLog_tooLarge); - return FSE_CTABLE_SIZE_U32(tableLog, maxSymbolValue) * sizeof(U32); -} - -/* provides the minimum logSize to safely represent a distribution */ -static unsigned FSE_minTableLog(size_t srcSize, unsigned maxSymbolValue) -{ - U32 minBitsSrc = BIT_highbit32((U32)(srcSize - 1)) + 1; - U32 minBitsSymbols = BIT_highbit32(maxSymbolValue) + 2; - U32 minBits = minBitsSrc < minBitsSymbols ? minBitsSrc : minBitsSymbols; - return minBits; -} - -unsigned FSE_optimalTableLog_internal(unsigned maxTableLog, size_t srcSize, unsigned maxSymbolValue, unsigned minus) -{ - U32 maxBitsSrc = BIT_highbit32((U32)(srcSize - 1)) - minus; - U32 tableLog = maxTableLog; - U32 minBits = FSE_minTableLog(srcSize, maxSymbolValue); - if (tableLog == 0) - tableLog = FSE_DEFAULT_TABLELOG; - if (maxBitsSrc < tableLog) - tableLog = maxBitsSrc; /* Accuracy can be reduced */ - if (minBits > tableLog) - tableLog = minBits; /* Need a minimum to safely represent all symbol values */ - if (tableLog < FSE_MIN_TABLELOG) - tableLog = FSE_MIN_TABLELOG; - if (tableLog > FSE_MAX_TABLELOG) - tableLog = FSE_MAX_TABLELOG; - return tableLog; -} - -unsigned FSE_optimalTableLog(unsigned maxTableLog, size_t srcSize, unsigned maxSymbolValue) -{ - return FSE_optimalTableLog_internal(maxTableLog, srcSize, maxSymbolValue, 2); -} - -/* Secondary normalization method. - To be used when primary method fails. */ - -static size_t FSE_normalizeM2(short *norm, U32 tableLog, const unsigned *count, size_t total, U32 maxSymbolValue) -{ - short const NOT_YET_ASSIGNED = -2; - U32 s; - U32 distributed = 0; - U32 ToDistribute; - - /* Init */ - U32 const lowThreshold = (U32)(total >> tableLog); - U32 lowOne = (U32)((total * 3) >> (tableLog + 1)); - - for (s = 0; s <= maxSymbolValue; s++) { - if (count[s] == 0) { - norm[s] = 0; - continue; - } - if (count[s] <= lowThreshold) { - norm[s] = -1; - distributed++; - total -= count[s]; - continue; - } - if (count[s] <= lowOne) { - norm[s] = 1; - distributed++; - total -= count[s]; - continue; - } - - norm[s] = NOT_YET_ASSIGNED; - } - ToDistribute = (1 << tableLog) - distributed; - - if ((total / ToDistribute) > lowOne) { - /* risk of rounding to zero */ - lowOne = (U32)((total * 3) / (ToDistribute * 2)); - for (s = 0; s <= maxSymbolValue; s++) { - if ((norm[s] == NOT_YET_ASSIGNED) && (count[s] <= lowOne)) { - norm[s] = 1; - distributed++; - total -= count[s]; - continue; - } - } - ToDistribute = (1 << tableLog) - distributed; - } - - if (distributed == maxSymbolValue + 1) { - /* all values are pretty poor; - probably incompressible data (should have already been detected); - find max, then give all remaining points to max */ - U32 maxV = 0, maxC = 0; - for (s = 0; s <= maxSymbolValue; s++) - if (count[s] > maxC) - maxV = s, maxC = count[s]; - norm[maxV] += (short)ToDistribute; - return 0; - } - - if (total == 0) { - /* all of the symbols were low enough for the lowOne or lowThreshold */ - for (s = 0; ToDistribute > 0; s = (s + 1) % (maxSymbolValue + 1)) - if (norm[s] > 0) - ToDistribute--, norm[s]++; - return 0; - } - - { - U64 const vStepLog = 62 - tableLog; - U64 const mid = (1ULL << (vStepLog - 1)) - 1; - U64 const rStep = div_u64((((U64)1 << vStepLog) * ToDistribute) + mid, (U32)total); /* scale on remaining */ - U64 tmpTotal = mid; - for (s = 0; s <= maxSymbolValue; s++) { - if (norm[s] == NOT_YET_ASSIGNED) { - U64 const end = tmpTotal + (count[s] * rStep); - U32 const sStart = (U32)(tmpTotal >> vStepLog); - U32 const sEnd = (U32)(end >> vStepLog); - U32 const weight = sEnd - sStart; - if (weight < 1) - return ERROR(GENERIC); - norm[s] = (short)weight; - tmpTotal = end; - } - } - } - - return 0; -} - -size_t FSE_normalizeCount(short *normalizedCounter, unsigned tableLog, const unsigned *count, size_t total, unsigned maxSymbolValue) -{ - /* Sanity checks */ - if (tableLog == 0) - tableLog = FSE_DEFAULT_TABLELOG; - if (tableLog < FSE_MIN_TABLELOG) - return ERROR(GENERIC); /* Unsupported size */ - if (tableLog > FSE_MAX_TABLELOG) - return ERROR(tableLog_tooLarge); /* Unsupported size */ - if (tableLog < FSE_minTableLog(total, maxSymbolValue)) - return ERROR(GENERIC); /* Too small tableLog, compression potentially impossible */ - - { - U32 const rtbTable[] = {0, 473195, 504333, 520860, 550000, 700000, 750000, 830000}; - U64 const scale = 62 - tableLog; - U64 const step = div_u64((U64)1 << 62, (U32)total); /* <== here, one division ! */ - U64 const vStep = 1ULL << (scale - 20); - int stillToDistribute = 1 << tableLog; - unsigned s; - unsigned largest = 0; - short largestP = 0; - U32 lowThreshold = (U32)(total >> tableLog); - - for (s = 0; s <= maxSymbolValue; s++) { - if (count[s] == total) - return 0; /* rle special case */ - if (count[s] == 0) { - normalizedCounter[s] = 0; - continue; - } - if (count[s] <= lowThreshold) { - normalizedCounter[s] = -1; - stillToDistribute--; - } else { - short proba = (short)((count[s] * step) >> scale); - if (proba < 8) { - U64 restToBeat = vStep * rtbTable[proba]; - proba += (count[s] * step) - ((U64)proba << scale) > restToBeat; - } - if (proba > largestP) - largestP = proba, largest = s; - normalizedCounter[s] = proba; - stillToDistribute -= proba; - } - } - if (-stillToDistribute >= (normalizedCounter[largest] >> 1)) { - /* corner case, need another normalization method */ - size_t const errorCode = FSE_normalizeM2(normalizedCounter, tableLog, count, total, maxSymbolValue); - if (FSE_isError(errorCode)) - return errorCode; - } else - normalizedCounter[largest] += (short)stillToDistribute; - } - - return tableLog; -} - -/* fake FSE_CTable, for raw (uncompressed) input */ -size_t FSE_buildCTable_raw(FSE_CTable *ct, unsigned nbBits) -{ - const unsigned tableSize = 1 << nbBits; - const unsigned tableMask = tableSize - 1; - const unsigned maxSymbolValue = tableMask; - void *const ptr = ct; - U16 *const tableU16 = ((U16 *)ptr) + 2; - void *const FSCT = ((U32 *)ptr) + 1 /* header */ + (tableSize >> 1); /* assumption : tableLog >= 1 */ - FSE_symbolCompressionTransform *const symbolTT = (FSE_symbolCompressionTransform *)(FSCT); - unsigned s; - - /* Sanity checks */ - if (nbBits < 1) - return ERROR(GENERIC); /* min size */ - - /* header */ - tableU16[-2] = (U16)nbBits; - tableU16[-1] = (U16)maxSymbolValue; - - /* Build table */ - for (s = 0; s < tableSize; s++) - tableU16[s] = (U16)(tableSize + s); - - /* Build Symbol Transformation Table */ - { - const U32 deltaNbBits = (nbBits << 16) - (1 << nbBits); - for (s = 0; s <= maxSymbolValue; s++) { - symbolTT[s].deltaNbBits = deltaNbBits; - symbolTT[s].deltaFindState = s - 1; - } - } - - return 0; -} - -/* fake FSE_CTable, for rle input (always same symbol) */ -size_t FSE_buildCTable_rle(FSE_CTable *ct, BYTE symbolValue) -{ - void *ptr = ct; - U16 *tableU16 = ((U16 *)ptr) + 2; - void *FSCTptr = (U32 *)ptr + 2; - FSE_symbolCompressionTransform *symbolTT = (FSE_symbolCompressionTransform *)FSCTptr; - - /* header */ - tableU16[-2] = (U16)0; - tableU16[-1] = (U16)symbolValue; - - /* Build table */ - tableU16[0] = 0; - tableU16[1] = 0; /* just in case */ - - /* Build Symbol Transformation Table */ - symbolTT[symbolValue].deltaNbBits = 0; - symbolTT[symbolValue].deltaFindState = 0; - - return 0; -} - -static size_t FSE_compress_usingCTable_generic(void *dst, size_t dstSize, const void *src, size_t srcSize, const FSE_CTable *ct, const unsigned fast) -{ - const BYTE *const istart = (const BYTE *)src; - const BYTE *const iend = istart + srcSize; - const BYTE *ip = iend; - - BIT_CStream_t bitC; - FSE_CState_t CState1, CState2; - - /* init */ - if (srcSize <= 2) - return 0; - { - size_t const initError = BIT_initCStream(&bitC, dst, dstSize); - if (FSE_isError(initError)) - return 0; /* not enough space available to write a bitstream */ - } - -#define FSE_FLUSHBITS(s) (fast ? BIT_flushBitsFast(s) : BIT_flushBits(s)) - - if (srcSize & 1) { - FSE_initCState2(&CState1, ct, *--ip); - FSE_initCState2(&CState2, ct, *--ip); - FSE_encodeSymbol(&bitC, &CState1, *--ip); - FSE_FLUSHBITS(&bitC); - } else { - FSE_initCState2(&CState2, ct, *--ip); - FSE_initCState2(&CState1, ct, *--ip); - } - - /* join to mod 4 */ - srcSize -= 2; - if ((sizeof(bitC.bitContainer) * 8 > FSE_MAX_TABLELOG * 4 + 7) && (srcSize & 2)) { /* test bit 2 */ - FSE_encodeSymbol(&bitC, &CState2, *--ip); - FSE_encodeSymbol(&bitC, &CState1, *--ip); - FSE_FLUSHBITS(&bitC); - } - - /* 2 or 4 encoding per loop */ - while (ip > istart) { - - FSE_encodeSymbol(&bitC, &CState2, *--ip); - - if (sizeof(bitC.bitContainer) * 8 < FSE_MAX_TABLELOG * 2 + 7) /* this test must be static */ - FSE_FLUSHBITS(&bitC); - - FSE_encodeSymbol(&bitC, &CState1, *--ip); - - if (sizeof(bitC.bitContainer) * 8 > FSE_MAX_TABLELOG * 4 + 7) { /* this test must be static */ - FSE_encodeSymbol(&bitC, &CState2, *--ip); - FSE_encodeSymbol(&bitC, &CState1, *--ip); - } - - FSE_FLUSHBITS(&bitC); - } - - FSE_flushCState(&bitC, &CState2); - FSE_flushCState(&bitC, &CState1); - return BIT_closeCStream(&bitC); -} - -size_t FSE_compress_usingCTable(void *dst, size_t dstSize, const void *src, size_t srcSize, const FSE_CTable *ct) -{ - unsigned const fast = (dstSize >= FSE_BLOCKBOUND(srcSize)); - - if (fast) - return FSE_compress_usingCTable_generic(dst, dstSize, src, srcSize, ct, 1); - else - return FSE_compress_usingCTable_generic(dst, dstSize, src, srcSize, ct, 0); -} - -size_t FSE_compressBound(size_t size) { return FSE_COMPRESSBOUND(size); } diff --git a/lib/zstd/fse_decompress.c b/lib/zstd/fse_decompress.c deleted file mode 100644 index 0b353530fb3f..000000000000 --- a/lib/zstd/fse_decompress.c +++ /dev/null @@ -1,325 +0,0 @@ -/* - * FSE : Finite State Entropy decoder - * Copyright (C) 2013-2015, Yann Collet. - * - * BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php) - * - * Redistribution and use in source and binary forms, with or without - * modification, are permitted provided that the following conditions are - * met: - * - * * Redistributions of source code must retain the above copyright - * notice, this list of conditions and the following disclaimer. - * * Redistributions in binary form must reproduce the above - * copyright notice, this list of conditions and the following disclaimer - * in the documentation and/or other materials provided with the - * distribution. - * - * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS - * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT - * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR - * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT - * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, - * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT - * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, - * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY - * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT - * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE - * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. - * - * This program is free software; you can redistribute it and/or modify it under - * the terms of the GNU General Public License version 2 as published by the - * Free Software Foundation. This program is dual-licensed; you may select - * either version 2 of the GNU General Public License ("GPL") or BSD license - * ("BSD"). - * - * You can contact the author at : - * - Source repository : https://github.com/Cyan4973/FiniteStateEntropy - */ - -/* ************************************************************** -* Compiler specifics -****************************************************************/ -#define FORCE_INLINE static __always_inline - -/* ************************************************************** -* Includes -****************************************************************/ -#include "bitstream.h" -#include "fse.h" -#include "zstd_internal.h" -#include -#include -#include /* memcpy, memset */ - -/* ************************************************************** -* Error Management -****************************************************************/ -#define FSE_isError ERR_isError -#define FSE_STATIC_ASSERT(c) \ - { \ - enum { FSE_static_assert = 1 / (int)(!!(c)) }; \ - } /* use only *after* variable declarations */ - -/* ************************************************************** -* Templates -****************************************************************/ -/* - designed to be included - for type-specific functions (template emulation in C) - Objective is to write these functions only once, for improved maintenance -*/ - -/* safety checks */ -#ifndef FSE_FUNCTION_EXTENSION -#error "FSE_FUNCTION_EXTENSION must be defined" -#endif -#ifndef FSE_FUNCTION_TYPE -#error "FSE_FUNCTION_TYPE must be defined" -#endif - -/* Function names */ -#define FSE_CAT(X, Y) X##Y -#define FSE_FUNCTION_NAME(X, Y) FSE_CAT(X, Y) -#define FSE_TYPE_NAME(X, Y) FSE_CAT(X, Y) - -/* Function templates */ - -size_t FSE_buildDTable_wksp(FSE_DTable *dt, const short *normalizedCounter, unsigned maxSymbolValue, unsigned tableLog, void *workspace, size_t workspaceSize) -{ - void *const tdPtr = dt + 1; /* because *dt is unsigned, 32-bits aligned on 32-bits */ - FSE_DECODE_TYPE *const tableDecode = (FSE_DECODE_TYPE *)(tdPtr); - U16 *symbolNext = (U16 *)workspace; - - U32 const maxSV1 = maxSymbolValue + 1; - U32 const tableSize = 1 << tableLog; - U32 highThreshold = tableSize - 1; - - /* Sanity Checks */ - if (workspaceSize < sizeof(U16) * (FSE_MAX_SYMBOL_VALUE + 1)) - return ERROR(tableLog_tooLarge); - if (maxSymbolValue > FSE_MAX_SYMBOL_VALUE) - return ERROR(maxSymbolValue_tooLarge); - if (tableLog > FSE_MAX_TABLELOG) - return ERROR(tableLog_tooLarge); - - /* Init, lay down lowprob symbols */ - { - FSE_DTableHeader DTableH; - DTableH.tableLog = (U16)tableLog; - DTableH.fastMode = 1; - { - S16 const largeLimit = (S16)(1 << (tableLog - 1)); - U32 s; - for (s = 0; s < maxSV1; s++) { - if (normalizedCounter[s] == -1) { - tableDecode[highThreshold--].symbol = (FSE_FUNCTION_TYPE)s; - symbolNext[s] = 1; - } else { - if (normalizedCounter[s] >= largeLimit) - DTableH.fastMode = 0; - symbolNext[s] = normalizedCounter[s]; - } - } - } - memcpy(dt, &DTableH, sizeof(DTableH)); - } - - /* Spread symbols */ - { - U32 const tableMask = tableSize - 1; - U32 const step = FSE_TABLESTEP(tableSize); - U32 s, position = 0; - for (s = 0; s < maxSV1; s++) { - int i; - for (i = 0; i < normalizedCounter[s]; i++) { - tableDecode[position].symbol = (FSE_FUNCTION_TYPE)s; - position = (position + step) & tableMask; - while (position > highThreshold) - position = (position + step) & tableMask; /* lowprob area */ - } - } - if (position != 0) - return ERROR(GENERIC); /* position must reach all cells once, otherwise normalizedCounter is incorrect */ - } - - /* Build Decoding table */ - { - U32 u; - for (u = 0; u < tableSize; u++) { - FSE_FUNCTION_TYPE const symbol = (FSE_FUNCTION_TYPE)(tableDecode[u].symbol); - U16 nextState = symbolNext[symbol]++; - tableDecode[u].nbBits = (BYTE)(tableLog - BIT_highbit32((U32)nextState)); - tableDecode[u].newState = (U16)((nextState << tableDecode[u].nbBits) - tableSize); - } - } - - return 0; -} - -/*-******************************************************* -* Decompression (Byte symbols) -*********************************************************/ -size_t FSE_buildDTable_rle(FSE_DTable *dt, BYTE symbolValue) -{ - void *ptr = dt; - FSE_DTableHeader *const DTableH = (FSE_DTableHeader *)ptr; - void *dPtr = dt + 1; - FSE_decode_t *const cell = (FSE_decode_t *)dPtr; - - DTableH->tableLog = 0; - DTableH->fastMode = 0; - - cell->newState = 0; - cell->symbol = symbolValue; - cell->nbBits = 0; - - return 0; -} - -size_t FSE_buildDTable_raw(FSE_DTable *dt, unsigned nbBits) -{ - void *ptr = dt; - FSE_DTableHeader *const DTableH = (FSE_DTableHeader *)ptr; - void *dPtr = dt + 1; - FSE_decode_t *const dinfo = (FSE_decode_t *)dPtr; - const unsigned tableSize = 1 << nbBits; - const unsigned tableMask = tableSize - 1; - const unsigned maxSV1 = tableMask + 1; - unsigned s; - - /* Sanity checks */ - if (nbBits < 1) - return ERROR(GENERIC); /* min size */ - - /* Build Decoding Table */ - DTableH->tableLog = (U16)nbBits; - DTableH->fastMode = 1; - for (s = 0; s < maxSV1; s++) { - dinfo[s].newState = 0; - dinfo[s].symbol = (BYTE)s; - dinfo[s].nbBits = (BYTE)nbBits; - } - - return 0; -} - -FORCE_INLINE size_t FSE_decompress_usingDTable_generic(void *dst, size_t maxDstSize, const void *cSrc, size_t cSrcSize, const FSE_DTable *dt, - const unsigned fast) -{ - BYTE *const ostart = (BYTE *)dst; - BYTE *op = ostart; - BYTE *const omax = op + maxDstSize; - BYTE *const olimit = omax - 3; - - BIT_DStream_t bitD; - FSE_DState_t state1; - FSE_DState_t state2; - - /* Init */ - CHECK_F(BIT_initDStream(&bitD, cSrc, cSrcSize)); - - FSE_initDState(&state1, &bitD, dt); - FSE_initDState(&state2, &bitD, dt); - -#define FSE_GETSYMBOL(statePtr) fast ? FSE_decodeSymbolFast(statePtr, &bitD) : FSE_decodeSymbol(statePtr, &bitD) - - /* 4 symbols per loop */ - for (; (BIT_reloadDStream(&bitD) == BIT_DStream_unfinished) & (op < olimit); op += 4) { - op[0] = FSE_GETSYMBOL(&state1); - - if (FSE_MAX_TABLELOG * 2 + 7 > sizeof(bitD.bitContainer) * 8) /* This test must be static */ - BIT_reloadDStream(&bitD); - - op[1] = FSE_GETSYMBOL(&state2); - - if (FSE_MAX_TABLELOG * 4 + 7 > sizeof(bitD.bitContainer) * 8) /* This test must be static */ - { - if (BIT_reloadDStream(&bitD) > BIT_DStream_unfinished) { - op += 2; - break; - } - } - - op[2] = FSE_GETSYMBOL(&state1); - - if (FSE_MAX_TABLELOG * 2 + 7 > sizeof(bitD.bitContainer) * 8) /* This test must be static */ - BIT_reloadDStream(&bitD); - - op[3] = FSE_GETSYMBOL(&state2); - } - - /* tail */ - /* note : BIT_reloadDStream(&bitD) >= FSE_DStream_partiallyFilled; Ends at exactly BIT_DStream_completed */ - while (1) { - if (op > (omax - 2)) - return ERROR(dstSize_tooSmall); - *op++ = FSE_GETSYMBOL(&state1); - if (BIT_reloadDStream(&bitD) == BIT_DStream_overflow) { - *op++ = FSE_GETSYMBOL(&state2); - break; - } - - if (op > (omax - 2)) - return ERROR(dstSize_tooSmall); - *op++ = FSE_GETSYMBOL(&state2); - if (BIT_reloadDStream(&bitD) == BIT_DStream_overflow) { - *op++ = FSE_GETSYMBOL(&state1); - break; - } - } - - return op - ostart; -} - -size_t FSE_decompress_usingDTable(void *dst, size_t originalSize, const void *cSrc, size_t cSrcSize, const FSE_DTable *dt) -{ - const void *ptr = dt; - const FSE_DTableHeader *DTableH = (const FSE_DTableHeader *)ptr; - const U32 fastMode = DTableH->fastMode; - - /* select fast mode (static) */ - if (fastMode) - return FSE_decompress_usingDTable_generic(dst, originalSize, cSrc, cSrcSize, dt, 1); - return FSE_decompress_usingDTable_generic(dst, originalSize, cSrc, cSrcSize, dt, 0); -} - -size_t FSE_decompress_wksp(void *dst, size_t dstCapacity, const void *cSrc, size_t cSrcSize, unsigned maxLog, void *workspace, size_t workspaceSize) -{ - const BYTE *const istart = (const BYTE *)cSrc; - const BYTE *ip = istart; - unsigned tableLog; - unsigned maxSymbolValue = FSE_MAX_SYMBOL_VALUE; - size_t NCountLength; - - FSE_DTable *dt; - short *counting; - size_t spaceUsed32 = 0; - - FSE_STATIC_ASSERT(sizeof(FSE_DTable) == sizeof(U32)); - - dt = (FSE_DTable *)((U32 *)workspace + spaceUsed32); - spaceUsed32 += FSE_DTABLE_SIZE_U32(maxLog); - counting = (short *)((U32 *)workspace + spaceUsed32); - spaceUsed32 += ALIGN(sizeof(short) * (FSE_MAX_SYMBOL_VALUE + 1), sizeof(U32)) >> 2; - - if ((spaceUsed32 << 2) > workspaceSize) - return ERROR(tableLog_tooLarge); - workspace = (U32 *)workspace + spaceUsed32; - workspaceSize -= (spaceUsed32 << 2); - - /* normal FSE decoding mode */ - NCountLength = FSE_readNCount(counting, &maxSymbolValue, &tableLog, istart, cSrcSize); - if (FSE_isError(NCountLength)) - return NCountLength; - // if (NCountLength >= cSrcSize) return ERROR(srcSize_wrong); /* too small input size; supposed to be already checked in NCountLength, only remaining - // case : NCountLength==cSrcSize */ - if (tableLog > maxLog) - return ERROR(tableLog_tooLarge); - ip += NCountLength; - cSrcSize -= NCountLength; - - CHECK_F(FSE_buildDTable_wksp(dt, counting, maxSymbolValue, tableLog, workspace, workspaceSize)); - - return FSE_decompress_usingDTable(dst, dstCapacity, ip, cSrcSize, dt); /* always return, even if it is an error code */ -} diff --git a/lib/zstd/huf.h b/lib/zstd/huf.h deleted file mode 100644 index 923218d12e28..000000000000 --- a/lib/zstd/huf.h +++ /dev/null @@ -1,212 +0,0 @@ -/* - * Huffman coder, part of New Generation Entropy library - * header file - * Copyright (C) 2013-2016, Yann Collet. - * - * BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php) - * - * Redistribution and use in source and binary forms, with or without - * modification, are permitted provided that the following conditions are - * met: - * - * * Redistributions of source code must retain the above copyright - * notice, this list of conditions and the following disclaimer. - * * Redistributions in binary form must reproduce the above - * copyright notice, this list of conditions and the following disclaimer - * in the documentation and/or other materials provided with the - * distribution. - * - * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS - * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT - * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR - * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT - * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, - * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT - * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, - * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY - * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT - * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE - * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. - * - * This program is free software; you can redistribute it and/or modify it under - * the terms of the GNU General Public License version 2 as published by the - * Free Software Foundation. This program is dual-licensed; you may select - * either version 2 of the GNU General Public License ("GPL") or BSD license - * ("BSD"). - * - * You can contact the author at : - * - Source repository : https://github.com/Cyan4973/FiniteStateEntropy - */ -#ifndef HUF_H_298734234 -#define HUF_H_298734234 - -/* *** Dependencies *** */ -#include /* size_t */ - -/* *** Tool functions *** */ -#define HUF_BLOCKSIZE_MAX (128 * 1024) /**< maximum input size for a single block compressed with HUF_compress */ -size_t HUF_compressBound(size_t size); /**< maximum compressed size (worst case) */ - -/* Error Management */ -unsigned HUF_isError(size_t code); /**< tells if a return value is an error code */ - -/* *** Advanced function *** */ - -/** HUF_compress4X_wksp() : -* Same as HUF_compress2(), but uses externally allocated `workSpace`, which must be a table of >= 1024 unsigned */ -size_t HUF_compress4X_wksp(void *dst, size_t dstSize, const void *src, size_t srcSize, unsigned maxSymbolValue, unsigned tableLog, void *workSpace, - size_t wkspSize); /**< `workSpace` must be a table of at least HUF_COMPRESS_WORKSPACE_SIZE_U32 unsigned */ - -/* *** Dependencies *** */ -#include "mem.h" /* U32 */ - -/* *** Constants *** */ -#define HUF_TABLELOG_MAX 12 /* max configured tableLog (for static allocation); can be modified up to HUF_ABSOLUTEMAX_TABLELOG */ -#define HUF_TABLELOG_DEFAULT 11 /* tableLog by default, when not specified */ -#define HUF_SYMBOLVALUE_MAX 255 - -#define HUF_TABLELOG_ABSOLUTEMAX 15 /* absolute limit of HUF_MAX_TABLELOG. Beyond that value, code does not work */ -#if (HUF_TABLELOG_MAX > HUF_TABLELOG_ABSOLUTEMAX) -#error "HUF_TABLELOG_MAX is too large !" -#endif - -/* **************************************** -* Static allocation -******************************************/ -/* HUF buffer bounds */ -#define HUF_CTABLEBOUND 129 -#define HUF_BLOCKBOUND(size) (size + (size >> 8) + 8) /* only true if incompressible pre-filtered with fast heuristic */ -#define HUF_COMPRESSBOUND(size) (HUF_CTABLEBOUND + HUF_BLOCKBOUND(size)) /* Macro version, useful for static allocation */ - -/* static allocation of HUF's Compression Table */ -#define HUF_CREATE_STATIC_CTABLE(name, maxSymbolValue) \ - U32 name##hb[maxSymbolValue + 1]; \ - void *name##hv = &(name##hb); \ - HUF_CElt *name = (HUF_CElt *)(name##hv) /* no final ; */ - -/* static allocation of HUF's DTable */ -typedef U32 HUF_DTable; -#define HUF_DTABLE_SIZE(maxTableLog) (1 + (1 << (maxTableLog))) -#define HUF_CREATE_STATIC_DTABLEX2(DTable, maxTableLog) HUF_DTable DTable[HUF_DTABLE_SIZE((maxTableLog)-1)] = {((U32)((maxTableLog)-1) * 0x01000001)} -#define HUF_CREATE_STATIC_DTABLEX4(DTable, maxTableLog) HUF_DTable DTable[HUF_DTABLE_SIZE(maxTableLog)] = {((U32)(maxTableLog)*0x01000001)} - -/* The workspace must have alignment at least 4 and be at least this large */ -#define HUF_COMPRESS_WORKSPACE_SIZE (6 << 10) -#define HUF_COMPRESS_WORKSPACE_SIZE_U32 (HUF_COMPRESS_WORKSPACE_SIZE / sizeof(U32)) - -/* The workspace must have alignment at least 4 and be at least this large */ -#define HUF_DECOMPRESS_WORKSPACE_SIZE (3 << 10) -#define HUF_DECOMPRESS_WORKSPACE_SIZE_U32 (HUF_DECOMPRESS_WORKSPACE_SIZE / sizeof(U32)) - -/* **************************************** -* Advanced decompression functions -******************************************/ -size_t HUF_decompress4X_DCtx_wksp(HUF_DTable *dctx, void *dst, size_t dstSize, const void *cSrc, size_t cSrcSize, void *workspace, size_t workspaceSize); /**< decodes RLE and uncompressed */ -size_t HUF_decompress4X_hufOnly_wksp(HUF_DTable *dctx, void *dst, size_t dstSize, const void *cSrc, size_t cSrcSize, void *workspace, - size_t workspaceSize); /**< considers RLE and uncompressed as errors */ -size_t HUF_decompress4X2_DCtx_wksp(HUF_DTable *dctx, void *dst, size_t dstSize, const void *cSrc, size_t cSrcSize, void *workspace, - size_t workspaceSize); /**< single-symbol decoder */ -size_t HUF_decompress4X4_DCtx_wksp(HUF_DTable *dctx, void *dst, size_t dstSize, const void *cSrc, size_t cSrcSize, void *workspace, - size_t workspaceSize); /**< double-symbols decoder */ - -/* **************************************** -* HUF detailed API -******************************************/ -/*! -HUF_compress() does the following: -1. count symbol occurrence from source[] into table count[] using FSE_count() -2. (optional) refine tableLog using HUF_optimalTableLog() -3. build Huffman table from count using HUF_buildCTable() -4. save Huffman table to memory buffer using HUF_writeCTable_wksp() -5. encode the data stream using HUF_compress4X_usingCTable() - -The following API allows targeting specific sub-functions for advanced tasks. -For example, it's possible to compress several blocks using the same 'CTable', -or to save and regenerate 'CTable' using external methods. -*/ -/* FSE_count() : find it within "fse.h" */ -unsigned HUF_optimalTableLog(unsigned maxTableLog, size_t srcSize, unsigned maxSymbolValue); -typedef struct HUF_CElt_s HUF_CElt; /* incomplete type */ -size_t HUF_writeCTable_wksp(void *dst, size_t maxDstSize, const HUF_CElt *CTable, unsigned maxSymbolValue, unsigned huffLog, void *workspace, size_t workspaceSize); -size_t HUF_compress4X_usingCTable(void *dst, size_t dstSize, const void *src, size_t srcSize, const HUF_CElt *CTable); - -typedef enum { - HUF_repeat_none, /**< Cannot use the previous table */ - HUF_repeat_check, /**< Can use the previous table but it must be checked. Note : The previous table must have been constructed by HUF_compress{1, - 4}X_repeat */ - HUF_repeat_valid /**< Can use the previous table and it is assumed to be valid */ -} HUF_repeat; -/** HUF_compress4X_repeat() : -* Same as HUF_compress4X_wksp(), but considers using hufTable if *repeat != HUF_repeat_none. -* If it uses hufTable it does not modify hufTable or repeat. -* If it doesn't, it sets *repeat = HUF_repeat_none, and it sets hufTable to the table used. -* If preferRepeat then the old table will always be used if valid. */ -size_t HUF_compress4X_repeat(void *dst, size_t dstSize, const void *src, size_t srcSize, unsigned maxSymbolValue, unsigned tableLog, void *workSpace, - size_t wkspSize, HUF_CElt *hufTable, HUF_repeat *repeat, - int preferRepeat); /**< `workSpace` must be a table of at least HUF_COMPRESS_WORKSPACE_SIZE_U32 unsigned */ - -/** HUF_buildCTable_wksp() : - * Same as HUF_buildCTable(), but using externally allocated scratch buffer. - * `workSpace` must be aligned on 4-bytes boundaries, and be at least as large as a table of 1024 unsigned. - */ -size_t HUF_buildCTable_wksp(HUF_CElt *tree, const U32 *count, U32 maxSymbolValue, U32 maxNbBits, void *workSpace, size_t wkspSize); - -/*! HUF_readStats() : - Read compact Huffman tree, saved by HUF_writeCTable(). - `huffWeight` is destination buffer. - @return : size read from `src` , or an error Code . - Note : Needed by HUF_readCTable() and HUF_readDTableXn() . */ -size_t HUF_readStats_wksp(BYTE *huffWeight, size_t hwSize, U32 *rankStats, U32 *nbSymbolsPtr, U32 *tableLogPtr, const void *src, size_t srcSize, - void *workspace, size_t workspaceSize); - -/** HUF_readCTable() : -* Loading a CTable saved with HUF_writeCTable() */ -size_t HUF_readCTable_wksp(HUF_CElt *CTable, unsigned maxSymbolValue, const void *src, size_t srcSize, void *workspace, size_t workspaceSize); - -/* -HUF_decompress() does the following: -1. select the decompression algorithm (X2, X4) based on pre-computed heuristics -2. build Huffman table from save, using HUF_readDTableXn() -3. decode 1 or 4 segments in parallel using HUF_decompressSXn_usingDTable -*/ - -/** HUF_selectDecoder() : -* Tells which decoder is likely to decode faster, -* based on a set of pre-determined metrics. -* @return : 0==HUF_decompress4X2, 1==HUF_decompress4X4 . -* Assumption : 0 < cSrcSize < dstSize <= 128 KB */ -U32 HUF_selectDecoder(size_t dstSize, size_t cSrcSize); - -size_t HUF_readDTableX2_wksp(HUF_DTable *DTable, const void *src, size_t srcSize, void *workspace, size_t workspaceSize); -size_t HUF_readDTableX4_wksp(HUF_DTable *DTable, const void *src, size_t srcSize, void *workspace, size_t workspaceSize); - -size_t HUF_decompress4X_usingDTable(void *dst, size_t maxDstSize, const void *cSrc, size_t cSrcSize, const HUF_DTable *DTable); -size_t HUF_decompress4X2_usingDTable(void *dst, size_t maxDstSize, const void *cSrc, size_t cSrcSize, const HUF_DTable *DTable); -size_t HUF_decompress4X4_usingDTable(void *dst, size_t maxDstSize, const void *cSrc, size_t cSrcSize, const HUF_DTable *DTable); - -/* single stream variants */ - -size_t HUF_compress1X_wksp(void *dst, size_t dstSize, const void *src, size_t srcSize, unsigned maxSymbolValue, unsigned tableLog, void *workSpace, - size_t wkspSize); /**< `workSpace` must be a table of at least HUF_COMPRESS_WORKSPACE_SIZE_U32 unsigned */ -size_t HUF_compress1X_usingCTable(void *dst, size_t dstSize, const void *src, size_t srcSize, const HUF_CElt *CTable); -/** HUF_compress1X_repeat() : -* Same as HUF_compress1X_wksp(), but considers using hufTable if *repeat != HUF_repeat_none. -* If it uses hufTable it does not modify hufTable or repeat. -* If it doesn't, it sets *repeat = HUF_repeat_none, and it sets hufTable to the table used. -* If preferRepeat then the old table will always be used if valid. */ -size_t HUF_compress1X_repeat(void *dst, size_t dstSize, const void *src, size_t srcSize, unsigned maxSymbolValue, unsigned tableLog, void *workSpace, - size_t wkspSize, HUF_CElt *hufTable, HUF_repeat *repeat, - int preferRepeat); /**< `workSpace` must be a table of at least HUF_COMPRESS_WORKSPACE_SIZE_U32 unsigned */ - -size_t HUF_decompress1X_DCtx_wksp(HUF_DTable *dctx, void *dst, size_t dstSize, const void *cSrc, size_t cSrcSize, void *workspace, size_t workspaceSize); -size_t HUF_decompress1X2_DCtx_wksp(HUF_DTable *dctx, void *dst, size_t dstSize, const void *cSrc, size_t cSrcSize, void *workspace, - size_t workspaceSize); /**< single-symbol decoder */ -size_t HUF_decompress1X4_DCtx_wksp(HUF_DTable *dctx, void *dst, size_t dstSize, const void *cSrc, size_t cSrcSize, void *workspace, - size_t workspaceSize); /**< double-symbols decoder */ - -size_t HUF_decompress1X_usingDTable(void *dst, size_t maxDstSize, const void *cSrc, size_t cSrcSize, - const HUF_DTable *DTable); /**< automatic selection of sing or double symbol decoder, based on DTable */ -size_t HUF_decompress1X2_usingDTable(void *dst, size_t maxDstSize, const void *cSrc, size_t cSrcSize, const HUF_DTable *DTable); -size_t HUF_decompress1X4_usingDTable(void *dst, size_t maxDstSize, const void *cSrc, size_t cSrcSize, const HUF_DTable *DTable); - -#endif /* HUF_H_298734234 */ diff --git a/lib/zstd/huf_compress.c b/lib/zstd/huf_compress.c deleted file mode 100644 index fd32838c185f..000000000000 --- a/lib/zstd/huf_compress.c +++ /dev/null @@ -1,773 +0,0 @@ -/* - * Huffman encoder, part of New Generation Entropy library - * Copyright (C) 2013-2016, Yann Collet. - * - * BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php) - * - * Redistribution and use in source and binary forms, with or without - * modification, are permitted provided that the following conditions are - * met: - * - * * Redistributions of source code must retain the above copyright - * notice, this list of conditions and the following disclaimer. - * * Redistributions in binary form must reproduce the above - * copyright notice, this list of conditions and the following disclaimer - * in the documentation and/or other materials provided with the - * distribution. - * - * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS - * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT - * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR - * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT - * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, - * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT - * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, - * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY - * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT - * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE - * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. - * - * This program is free software; you can redistribute it and/or modify it under - * the terms of the GNU General Public License version 2 as published by the - * Free Software Foundation. This program is dual-licensed; you may select - * either version 2 of the GNU General Public License ("GPL") or BSD license - * ("BSD"). - * - * You can contact the author at : - * - Source repository : https://github.com/Cyan4973/FiniteStateEntropy - */ - -/* ************************************************************** -* Includes -****************************************************************/ -#include "bitstream.h" -#include "fse.h" /* header compression */ -#include "huf.h" -#include -#include /* memcpy, memset */ - -/* ************************************************************** -* Error Management -****************************************************************/ -#define HUF_STATIC_ASSERT(c) \ - { \ - enum { HUF_static_assert = 1 / (int)(!!(c)) }; \ - } /* use only *after* variable declarations */ -#define CHECK_V_F(e, f) \ - size_t const e = f; \ - if (ERR_isError(e)) \ - return f -#define CHECK_F(f) \ - { \ - CHECK_V_F(_var_err__, f); \ - } - -/* ************************************************************** -* Utils -****************************************************************/ -unsigned HUF_optimalTableLog(unsigned maxTableLog, size_t srcSize, unsigned maxSymbolValue) -{ - return FSE_optimalTableLog_internal(maxTableLog, srcSize, maxSymbolValue, 1); -} - -/* ******************************************************* -* HUF : Huffman block compression -*********************************************************/ -/* HUF_compressWeights() : - * Same as FSE_compress(), but dedicated to huff0's weights compression. - * The use case needs much less stack memory. - * Note : all elements within weightTable are supposed to be <= HUF_TABLELOG_MAX. - */ -#define MAX_FSE_TABLELOG_FOR_HUFF_HEADER 6 -size_t HUF_compressWeights_wksp(void *dst, size_t dstSize, const void *weightTable, size_t wtSize, void *workspace, size_t workspaceSize) -{ - BYTE *const ostart = (BYTE *)dst; - BYTE *op = ostart; - BYTE *const oend = ostart + dstSize; - - U32 maxSymbolValue = HUF_TABLELOG_MAX; - U32 tableLog = MAX_FSE_TABLELOG_FOR_HUFF_HEADER; - - FSE_CTable *CTable; - U32 *count; - S16 *norm; - size_t spaceUsed32 = 0; - - HUF_STATIC_ASSERT(sizeof(FSE_CTable) == sizeof(U32)); - - CTable = (FSE_CTable *)((U32 *)workspace + spaceUsed32); - spaceUsed32 += FSE_CTABLE_SIZE_U32(MAX_FSE_TABLELOG_FOR_HUFF_HEADER, HUF_TABLELOG_MAX); - count = (U32 *)workspace + spaceUsed32; - spaceUsed32 += HUF_TABLELOG_MAX + 1; - norm = (S16 *)((U32 *)workspace + spaceUsed32); - spaceUsed32 += ALIGN(sizeof(S16) * (HUF_TABLELOG_MAX + 1), sizeof(U32)) >> 2; - - if ((spaceUsed32 << 2) > workspaceSize) - return ERROR(tableLog_tooLarge); - workspace = (U32 *)workspace + spaceUsed32; - workspaceSize -= (spaceUsed32 << 2); - - /* init conditions */ - if (wtSize <= 1) - return 0; /* Not compressible */ - - /* Scan input and build symbol stats */ - { - CHECK_V_F(maxCount, FSE_count_simple(count, &maxSymbolValue, weightTable, wtSize)); - if (maxCount == wtSize) - return 1; /* only a single symbol in src : rle */ - if (maxCount == 1) - return 0; /* each symbol present maximum once => not compressible */ - } - - tableLog = FSE_optimalTableLog(tableLog, wtSize, maxSymbolValue); - CHECK_F(FSE_normalizeCount(norm, tableLog, count, wtSize, maxSymbolValue)); - - /* Write table description header */ - { - CHECK_V_F(hSize, FSE_writeNCount(op, oend - op, norm, maxSymbolValue, tableLog)); - op += hSize; - } - - /* Compress */ - CHECK_F(FSE_buildCTable_wksp(CTable, norm, maxSymbolValue, tableLog, workspace, workspaceSize)); - { - CHECK_V_F(cSize, FSE_compress_usingCTable(op, oend - op, weightTable, wtSize, CTable)); - if (cSize == 0) - return 0; /* not enough space for compressed data */ - op += cSize; - } - - return op - ostart; -} - -struct HUF_CElt_s { - U16 val; - BYTE nbBits; -}; /* typedef'd to HUF_CElt within "huf.h" */ - -/*! HUF_writeCTable_wksp() : - `CTable` : Huffman tree to save, using huf representation. - @return : size of saved CTable */ -size_t HUF_writeCTable_wksp(void *dst, size_t maxDstSize, const HUF_CElt *CTable, U32 maxSymbolValue, U32 huffLog, void *workspace, size_t workspaceSize) -{ - BYTE *op = (BYTE *)dst; - U32 n; - - BYTE *bitsToWeight; - BYTE *huffWeight; - size_t spaceUsed32 = 0; - - bitsToWeight = (BYTE *)((U32 *)workspace + spaceUsed32); - spaceUsed32 += ALIGN(HUF_TABLELOG_MAX + 1, sizeof(U32)) >> 2; - huffWeight = (BYTE *)((U32 *)workspace + spaceUsed32); - spaceUsed32 += ALIGN(HUF_SYMBOLVALUE_MAX, sizeof(U32)) >> 2; - - if ((spaceUsed32 << 2) > workspaceSize) - return ERROR(tableLog_tooLarge); - workspace = (U32 *)workspace + spaceUsed32; - workspaceSize -= (spaceUsed32 << 2); - - /* check conditions */ - if (maxSymbolValue > HUF_SYMBOLVALUE_MAX) - return ERROR(maxSymbolValue_tooLarge); - - /* convert to weight */ - bitsToWeight[0] = 0; - for (n = 1; n < huffLog + 1; n++) - bitsToWeight[n] = (BYTE)(huffLog + 1 - n); - for (n = 0; n < maxSymbolValue; n++) - huffWeight[n] = bitsToWeight[CTable[n].nbBits]; - - /* attempt weights compression by FSE */ - { - CHECK_V_F(hSize, HUF_compressWeights_wksp(op + 1, maxDstSize - 1, huffWeight, maxSymbolValue, workspace, workspaceSize)); - if ((hSize > 1) & (hSize < maxSymbolValue / 2)) { /* FSE compressed */ - op[0] = (BYTE)hSize; - return hSize + 1; - } - } - - /* write raw values as 4-bits (max : 15) */ - if (maxSymbolValue > (256 - 128)) - return ERROR(GENERIC); /* should not happen : likely means source cannot be compressed */ - if (((maxSymbolValue + 1) / 2) + 1 > maxDstSize) - return ERROR(dstSize_tooSmall); /* not enough space within dst buffer */ - op[0] = (BYTE)(128 /*special case*/ + (maxSymbolValue - 1)); - huffWeight[maxSymbolValue] = 0; /* to be sure it doesn't cause msan issue in final combination */ - for (n = 0; n < maxSymbolValue; n += 2) - op[(n / 2) + 1] = (BYTE)((huffWeight[n] << 4) + huffWeight[n + 1]); - return ((maxSymbolValue + 1) / 2) + 1; -} - -size_t HUF_readCTable_wksp(HUF_CElt *CTable, U32 maxSymbolValue, const void *src, size_t srcSize, void *workspace, size_t workspaceSize) -{ - U32 *rankVal; - BYTE *huffWeight; - U32 tableLog = 0; - U32 nbSymbols = 0; - size_t readSize; - size_t spaceUsed32 = 0; - - rankVal = (U32 *)workspace + spaceUsed32; - spaceUsed32 += HUF_TABLELOG_ABSOLUTEMAX + 1; - huffWeight = (BYTE *)((U32 *)workspace + spaceUsed32); - spaceUsed32 += ALIGN(HUF_SYMBOLVALUE_MAX + 1, sizeof(U32)) >> 2; - - if ((spaceUsed32 << 2) > workspaceSize) - return ERROR(tableLog_tooLarge); - workspace = (U32 *)workspace + spaceUsed32; - workspaceSize -= (spaceUsed32 << 2); - - /* get symbol weights */ - readSize = HUF_readStats_wksp(huffWeight, HUF_SYMBOLVALUE_MAX + 1, rankVal, &nbSymbols, &tableLog, src, srcSize, workspace, workspaceSize); - if (ERR_isError(readSize)) - return readSize; - - /* check result */ - if (tableLog > HUF_TABLELOG_MAX) - return ERROR(tableLog_tooLarge); - if (nbSymbols > maxSymbolValue + 1) - return ERROR(maxSymbolValue_tooSmall); - - /* Prepare base value per rank */ - { - U32 n, nextRankStart = 0; - for (n = 1; n <= tableLog; n++) { - U32 curr = nextRankStart; - nextRankStart += (rankVal[n] << (n - 1)); - rankVal[n] = curr; - } - } - - /* fill nbBits */ - { - U32 n; - for (n = 0; n < nbSymbols; n++) { - const U32 w = huffWeight[n]; - CTable[n].nbBits = (BYTE)(tableLog + 1 - w); - } - } - - /* fill val */ - { - U16 nbPerRank[HUF_TABLELOG_MAX + 2] = {0}; /* support w=0=>n=tableLog+1 */ - U16 valPerRank[HUF_TABLELOG_MAX + 2] = {0}; - { - U32 n; - for (n = 0; n < nbSymbols; n++) - nbPerRank[CTable[n].nbBits]++; - } - /* determine stating value per rank */ - valPerRank[tableLog + 1] = 0; /* for w==0 */ - { - U16 min = 0; - U32 n; - for (n = tableLog; n > 0; n--) { /* start at n=tablelog <-> w=1 */ - valPerRank[n] = min; /* get starting value within each rank */ - min += nbPerRank[n]; - min >>= 1; - } - } - /* assign value within rank, symbol order */ - { - U32 n; - for (n = 0; n <= maxSymbolValue; n++) - CTable[n].val = valPerRank[CTable[n].nbBits]++; - } - } - - return readSize; -} - -typedef struct nodeElt_s { - U32 count; - U16 parent; - BYTE byte; - BYTE nbBits; -} nodeElt; - -static U32 HUF_setMaxHeight(nodeElt *huffNode, U32 lastNonNull, U32 maxNbBits) -{ - const U32 largestBits = huffNode[lastNonNull].nbBits; - if (largestBits <= maxNbBits) - return largestBits; /* early exit : no elt > maxNbBits */ - - /* there are several too large elements (at least >= 2) */ - { - int totalCost = 0; - const U32 baseCost = 1 << (largestBits - maxNbBits); - U32 n = lastNonNull; - - while (huffNode[n].nbBits > maxNbBits) { - totalCost += baseCost - (1 << (largestBits - huffNode[n].nbBits)); - huffNode[n].nbBits = (BYTE)maxNbBits; - n--; - } /* n stops at huffNode[n].nbBits <= maxNbBits */ - while (huffNode[n].nbBits == maxNbBits) - n--; /* n end at index of smallest symbol using < maxNbBits */ - - /* renorm totalCost */ - totalCost >>= (largestBits - maxNbBits); /* note : totalCost is necessarily a multiple of baseCost */ - - /* repay normalized cost */ - { - U32 const noSymbol = 0xF0F0F0F0; - U32 rankLast[HUF_TABLELOG_MAX + 2]; - int pos; - - /* Get pos of last (smallest) symbol per rank */ - memset(rankLast, 0xF0, sizeof(rankLast)); - { - U32 currNbBits = maxNbBits; - for (pos = n; pos >= 0; pos--) { - if (huffNode[pos].nbBits >= currNbBits) - continue; - currNbBits = huffNode[pos].nbBits; /* < maxNbBits */ - rankLast[maxNbBits - currNbBits] = pos; - } - } - - while (totalCost > 0) { - U32 nBitsToDecrease = BIT_highbit32(totalCost) + 1; - for (; nBitsToDecrease > 1; nBitsToDecrease--) { - U32 highPos = rankLast[nBitsToDecrease]; - U32 lowPos = rankLast[nBitsToDecrease - 1]; - if (highPos == noSymbol) - continue; - if (lowPos == noSymbol) - break; - { - U32 const highTotal = huffNode[highPos].count; - U32 const lowTotal = 2 * huffNode[lowPos].count; - if (highTotal <= lowTotal) - break; - } - } - /* only triggered when no more rank 1 symbol left => find closest one (note : there is necessarily at least one !) */ - /* HUF_MAX_TABLELOG test just to please gcc 5+; but it should not be necessary */ - while ((nBitsToDecrease <= HUF_TABLELOG_MAX) && (rankLast[nBitsToDecrease] == noSymbol)) - nBitsToDecrease++; - totalCost -= 1 << (nBitsToDecrease - 1); - if (rankLast[nBitsToDecrease - 1] == noSymbol) - rankLast[nBitsToDecrease - 1] = rankLast[nBitsToDecrease]; /* this rank is no longer empty */ - huffNode[rankLast[nBitsToDecrease]].nbBits++; - if (rankLast[nBitsToDecrease] == 0) /* special case, reached largest symbol */ - rankLast[nBitsToDecrease] = noSymbol; - else { - rankLast[nBitsToDecrease]--; - if (huffNode[rankLast[nBitsToDecrease]].nbBits != maxNbBits - nBitsToDecrease) - rankLast[nBitsToDecrease] = noSymbol; /* this rank is now empty */ - } - } /* while (totalCost > 0) */ - - while (totalCost < 0) { /* Sometimes, cost correction overshoot */ - if (rankLast[1] == noSymbol) { /* special case : no rank 1 symbol (using maxNbBits-1); let's create one from largest rank 0 - (using maxNbBits) */ - while (huffNode[n].nbBits == maxNbBits) - n--; - huffNode[n + 1].nbBits--; - rankLast[1] = n + 1; - totalCost++; - continue; - } - huffNode[rankLast[1] + 1].nbBits--; - rankLast[1]++; - totalCost++; - } - } - } /* there are several too large elements (at least >= 2) */ - - return maxNbBits; -} - -typedef struct { - U32 base; - U32 curr; -} rankPos; - -static void HUF_sort(nodeElt *huffNode, const U32 *count, U32 maxSymbolValue) -{ - rankPos rank[32]; - U32 n; - - memset(rank, 0, sizeof(rank)); - for (n = 0; n <= maxSymbolValue; n++) { - U32 r = BIT_highbit32(count[n] + 1); - rank[r].base++; - } - for (n = 30; n > 0; n--) - rank[n - 1].base += rank[n].base; - for (n = 0; n < 32; n++) - rank[n].curr = rank[n].base; - for (n = 0; n <= maxSymbolValue; n++) { - U32 const c = count[n]; - U32 const r = BIT_highbit32(c + 1) + 1; - U32 pos = rank[r].curr++; - while ((pos > rank[r].base) && (c > huffNode[pos - 1].count)) - huffNode[pos] = huffNode[pos - 1], pos--; - huffNode[pos].count = c; - huffNode[pos].byte = (BYTE)n; - } -} - -/** HUF_buildCTable_wksp() : - * Same as HUF_buildCTable(), but using externally allocated scratch buffer. - * `workSpace` must be aligned on 4-bytes boundaries, and be at least as large as a table of 1024 unsigned. - */ -#define STARTNODE (HUF_SYMBOLVALUE_MAX + 1) -typedef nodeElt huffNodeTable[2 * HUF_SYMBOLVALUE_MAX + 1 + 1]; -size_t HUF_buildCTable_wksp(HUF_CElt *tree, const U32 *count, U32 maxSymbolValue, U32 maxNbBits, void *workSpace, size_t wkspSize) -{ - nodeElt *const huffNode0 = (nodeElt *)workSpace; - nodeElt *const huffNode = huffNode0 + 1; - U32 n, nonNullRank; - int lowS, lowN; - U16 nodeNb = STARTNODE; - U32 nodeRoot; - - /* safety checks */ - if (wkspSize < sizeof(huffNodeTable)) - return ERROR(GENERIC); /* workSpace is not large enough */ - if (maxNbBits == 0) - maxNbBits = HUF_TABLELOG_DEFAULT; - if (maxSymbolValue > HUF_SYMBOLVALUE_MAX) - return ERROR(GENERIC); - memset(huffNode0, 0, sizeof(huffNodeTable)); - - /* sort, decreasing order */ - HUF_sort(huffNode, count, maxSymbolValue); - - /* init for parents */ - nonNullRank = maxSymbolValue; - while (huffNode[nonNullRank].count == 0) - nonNullRank--; - lowS = nonNullRank; - nodeRoot = nodeNb + lowS - 1; - lowN = nodeNb; - huffNode[nodeNb].count = huffNode[lowS].count + huffNode[lowS - 1].count; - huffNode[lowS].parent = huffNode[lowS - 1].parent = nodeNb; - nodeNb++; - lowS -= 2; - for (n = nodeNb; n <= nodeRoot; n++) - huffNode[n].count = (U32)(1U << 30); - huffNode0[0].count = (U32)(1U << 31); /* fake entry, strong barrier */ - - /* create parents */ - while (nodeNb <= nodeRoot) { - U32 n1 = (huffNode[lowS].count < huffNode[lowN].count) ? lowS-- : lowN++; - U32 n2 = (huffNode[lowS].count < huffNode[lowN].count) ? lowS-- : lowN++; - huffNode[nodeNb].count = huffNode[n1].count + huffNode[n2].count; - huffNode[n1].parent = huffNode[n2].parent = nodeNb; - nodeNb++; - } - - /* distribute weights (unlimited tree height) */ - huffNode[nodeRoot].nbBits = 0; - for (n = nodeRoot - 1; n >= STARTNODE; n--) - huffNode[n].nbBits = huffNode[huffNode[n].parent].nbBits + 1; - for (n = 0; n <= nonNullRank; n++) - huffNode[n].nbBits = huffNode[huffNode[n].parent].nbBits + 1; - - /* enforce maxTableLog */ - maxNbBits = HUF_setMaxHeight(huffNode, nonNullRank, maxNbBits); - - /* fill result into tree (val, nbBits) */ - { - U16 nbPerRank[HUF_TABLELOG_MAX + 1] = {0}; - U16 valPerRank[HUF_TABLELOG_MAX + 1] = {0}; - if (maxNbBits > HUF_TABLELOG_MAX) - return ERROR(GENERIC); /* check fit into table */ - for (n = 0; n <= nonNullRank; n++) - nbPerRank[huffNode[n].nbBits]++; - /* determine stating value per rank */ - { - U16 min = 0; - for (n = maxNbBits; n > 0; n--) { - valPerRank[n] = min; /* get starting value within each rank */ - min += nbPerRank[n]; - min >>= 1; - } - } - for (n = 0; n <= maxSymbolValue; n++) - tree[huffNode[n].byte].nbBits = huffNode[n].nbBits; /* push nbBits per symbol, symbol order */ - for (n = 0; n <= maxSymbolValue; n++) - tree[n].val = valPerRank[tree[n].nbBits]++; /* assign value within rank, symbol order */ - } - - return maxNbBits; -} - -static size_t HUF_estimateCompressedSize(HUF_CElt *CTable, const unsigned *count, unsigned maxSymbolValue) -{ - size_t nbBits = 0; - int s; - for (s = 0; s <= (int)maxSymbolValue; ++s) { - nbBits += CTable[s].nbBits * count[s]; - } - return nbBits >> 3; -} - -static int HUF_validateCTable(const HUF_CElt *CTable, const unsigned *count, unsigned maxSymbolValue) -{ - int bad = 0; - int s; - for (s = 0; s <= (int)maxSymbolValue; ++s) { - bad |= (count[s] != 0) & (CTable[s].nbBits == 0); - } - return !bad; -} - -static void HUF_encodeSymbol(BIT_CStream_t *bitCPtr, U32 symbol, const HUF_CElt *CTable) -{ - BIT_addBitsFast(bitCPtr, CTable[symbol].val, CTable[symbol].nbBits); -} - -size_t HUF_compressBound(size_t size) { return HUF_COMPRESSBOUND(size); } - -#define HUF_FLUSHBITS(s) BIT_flushBits(s) - -#define HUF_FLUSHBITS_1(stream) \ - if (sizeof((stream)->bitContainer) * 8 < HUF_TABLELOG_MAX * 2 + 7) \ - HUF_FLUSHBITS(stream) - -#define HUF_FLUSHBITS_2(stream) \ - if (sizeof((stream)->bitContainer) * 8 < HUF_TABLELOG_MAX * 4 + 7) \ - HUF_FLUSHBITS(stream) - -size_t HUF_compress1X_usingCTable(void *dst, size_t dstSize, const void *src, size_t srcSize, const HUF_CElt *CTable) -{ - const BYTE *ip = (const BYTE *)src; - BYTE *const ostart = (BYTE *)dst; - BYTE *const oend = ostart + dstSize; - BYTE *op = ostart; - size_t n; - BIT_CStream_t bitC; - - /* init */ - if (dstSize < 8) - return 0; /* not enough space to compress */ - { - size_t const initErr = BIT_initCStream(&bitC, op, oend - op); - if (HUF_isError(initErr)) - return 0; - } - - n = srcSize & ~3; /* join to mod 4 */ - switch (srcSize & 3) { - case 3: HUF_encodeSymbol(&bitC, ip[n + 2], CTable); HUF_FLUSHBITS_2(&bitC); - fallthrough; - case 2: HUF_encodeSymbol(&bitC, ip[n + 1], CTable); HUF_FLUSHBITS_1(&bitC); - fallthrough; - case 1: HUF_encodeSymbol(&bitC, ip[n + 0], CTable); HUF_FLUSHBITS(&bitC); - fallthrough; - case 0: - default:; - } - - for (; n > 0; n -= 4) { /* note : n&3==0 at this stage */ - HUF_encodeSymbol(&bitC, ip[n - 1], CTable); - HUF_FLUSHBITS_1(&bitC); - HUF_encodeSymbol(&bitC, ip[n - 2], CTable); - HUF_FLUSHBITS_2(&bitC); - HUF_encodeSymbol(&bitC, ip[n - 3], CTable); - HUF_FLUSHBITS_1(&bitC); - HUF_encodeSymbol(&bitC, ip[n - 4], CTable); - HUF_FLUSHBITS(&bitC); - } - - return BIT_closeCStream(&bitC); -} - -size_t HUF_compress4X_usingCTable(void *dst, size_t dstSize, const void *src, size_t srcSize, const HUF_CElt *CTable) -{ - size_t const segmentSize = (srcSize + 3) / 4; /* first 3 segments */ - const BYTE *ip = (const BYTE *)src; - const BYTE *const iend = ip + srcSize; - BYTE *const ostart = (BYTE *)dst; - BYTE *const oend = ostart + dstSize; - BYTE *op = ostart; - - if (dstSize < 6 + 1 + 1 + 1 + 8) - return 0; /* minimum space to compress successfully */ - if (srcSize < 12) - return 0; /* no saving possible : too small input */ - op += 6; /* jumpTable */ - - { - CHECK_V_F(cSize, HUF_compress1X_usingCTable(op, oend - op, ip, segmentSize, CTable)); - if (cSize == 0) - return 0; - ZSTD_writeLE16(ostart, (U16)cSize); - op += cSize; - } - - ip += segmentSize; - { - CHECK_V_F(cSize, HUF_compress1X_usingCTable(op, oend - op, ip, segmentSize, CTable)); - if (cSize == 0) - return 0; - ZSTD_writeLE16(ostart + 2, (U16)cSize); - op += cSize; - } - - ip += segmentSize; - { - CHECK_V_F(cSize, HUF_compress1X_usingCTable(op, oend - op, ip, segmentSize, CTable)); - if (cSize == 0) - return 0; - ZSTD_writeLE16(ostart + 4, (U16)cSize); - op += cSize; - } - - ip += segmentSize; - { - CHECK_V_F(cSize, HUF_compress1X_usingCTable(op, oend - op, ip, iend - ip, CTable)); - if (cSize == 0) - return 0; - op += cSize; - } - - return op - ostart; -} - -static size_t HUF_compressCTable_internal(BYTE *const ostart, BYTE *op, BYTE *const oend, const void *src, size_t srcSize, unsigned singleStream, - const HUF_CElt *CTable) -{ - size_t const cSize = - singleStream ? HUF_compress1X_usingCTable(op, oend - op, src, srcSize, CTable) : HUF_compress4X_usingCTable(op, oend - op, src, srcSize, CTable); - if (HUF_isError(cSize)) { - return cSize; - } - if (cSize == 0) { - return 0; - } /* uncompressible */ - op += cSize; - /* check compressibility */ - if ((size_t)(op - ostart) >= srcSize - 1) { - return 0; - } - return op - ostart; -} - -/* `workSpace` must a table of at least 1024 unsigned */ -static size_t HUF_compress_internal(void *dst, size_t dstSize, const void *src, size_t srcSize, unsigned maxSymbolValue, unsigned huffLog, - unsigned singleStream, void *workSpace, size_t wkspSize, HUF_CElt *oldHufTable, HUF_repeat *repeat, int preferRepeat) -{ - BYTE *const ostart = (BYTE *)dst; - BYTE *const oend = ostart + dstSize; - BYTE *op = ostart; - - U32 *count; - size_t const countSize = sizeof(U32) * (HUF_SYMBOLVALUE_MAX + 1); - HUF_CElt *CTable; - size_t const CTableSize = sizeof(HUF_CElt) * (HUF_SYMBOLVALUE_MAX + 1); - - /* checks & inits */ - if (wkspSize < sizeof(huffNodeTable) + countSize + CTableSize) - return ERROR(GENERIC); - if (!srcSize) - return 0; /* Uncompressed (note : 1 means rle, so first byte must be correct) */ - if (!dstSize) - return 0; /* cannot fit within dst budget */ - if (srcSize > HUF_BLOCKSIZE_MAX) - return ERROR(srcSize_wrong); /* curr block size limit */ - if (huffLog > HUF_TABLELOG_MAX) - return ERROR(tableLog_tooLarge); - if (!maxSymbolValue) - maxSymbolValue = HUF_SYMBOLVALUE_MAX; - if (!huffLog) - huffLog = HUF_TABLELOG_DEFAULT; - - count = (U32 *)workSpace; - workSpace = (BYTE *)workSpace + countSize; - wkspSize -= countSize; - CTable = (HUF_CElt *)workSpace; - workSpace = (BYTE *)workSpace + CTableSize; - wkspSize -= CTableSize; - - /* Heuristic : If we don't need to check the validity of the old table use the old table for small inputs */ - if (preferRepeat && repeat && *repeat == HUF_repeat_valid) { - return HUF_compressCTable_internal(ostart, op, oend, src, srcSize, singleStream, oldHufTable); - } - - /* Scan input and build symbol stats */ - { - CHECK_V_F(largest, FSE_count_wksp(count, &maxSymbolValue, (const BYTE *)src, srcSize, (U32 *)workSpace)); - if (largest == srcSize) { - *ostart = ((const BYTE *)src)[0]; - return 1; - } /* single symbol, rle */ - if (largest <= (srcSize >> 7) + 1) - return 0; /* Fast heuristic : not compressible enough */ - } - - /* Check validity of previous table */ - if (repeat && *repeat == HUF_repeat_check && !HUF_validateCTable(oldHufTable, count, maxSymbolValue)) { - *repeat = HUF_repeat_none; - } - /* Heuristic : use existing table for small inputs */ - if (preferRepeat && repeat && *repeat != HUF_repeat_none) { - return HUF_compressCTable_internal(ostart, op, oend, src, srcSize, singleStream, oldHufTable); - } - - /* Build Huffman Tree */ - huffLog = HUF_optimalTableLog(huffLog, srcSize, maxSymbolValue); - { - CHECK_V_F(maxBits, HUF_buildCTable_wksp(CTable, count, maxSymbolValue, huffLog, workSpace, wkspSize)); - huffLog = (U32)maxBits; - /* Zero the unused symbols so we can check it for validity */ - memset(CTable + maxSymbolValue + 1, 0, CTableSize - (maxSymbolValue + 1) * sizeof(HUF_CElt)); - } - - /* Write table description header */ - { - CHECK_V_F(hSize, HUF_writeCTable_wksp(op, dstSize, CTable, maxSymbolValue, huffLog, workSpace, wkspSize)); - /* Check if using the previous table will be beneficial */ - if (repeat && *repeat != HUF_repeat_none) { - size_t const oldSize = HUF_estimateCompressedSize(oldHufTable, count, maxSymbolValue); - size_t const newSize = HUF_estimateCompressedSize(CTable, count, maxSymbolValue); - if (oldSize <= hSize + newSize || hSize + 12 >= srcSize) { - return HUF_compressCTable_internal(ostart, op, oend, src, srcSize, singleStream, oldHufTable); - } - } - /* Use the new table */ - if (hSize + 12ul >= srcSize) { - return 0; - } - op += hSize; - if (repeat) { - *repeat = HUF_repeat_none; - } - if (oldHufTable) { - memcpy(oldHufTable, CTable, CTableSize); - } /* Save the new table */ - } - return HUF_compressCTable_internal(ostart, op, oend, src, srcSize, singleStream, CTable); -} - -size_t HUF_compress1X_wksp(void *dst, size_t dstSize, const void *src, size_t srcSize, unsigned maxSymbolValue, unsigned huffLog, void *workSpace, - size_t wkspSize) -{ - return HUF_compress_internal(dst, dstSize, src, srcSize, maxSymbolValue, huffLog, 1 /* single stream */, workSpace, wkspSize, NULL, NULL, 0); -} - -size_t HUF_compress1X_repeat(void *dst, size_t dstSize, const void *src, size_t srcSize, unsigned maxSymbolValue, unsigned huffLog, void *workSpace, - size_t wkspSize, HUF_CElt *hufTable, HUF_repeat *repeat, int preferRepeat) -{ - return HUF_compress_internal(dst, dstSize, src, srcSize, maxSymbolValue, huffLog, 1 /* single stream */, workSpace, wkspSize, hufTable, repeat, - preferRepeat); -} - -size_t HUF_compress4X_wksp(void *dst, size_t dstSize, const void *src, size_t srcSize, unsigned maxSymbolValue, unsigned huffLog, void *workSpace, - size_t wkspSize) -{ - return HUF_compress_internal(dst, dstSize, src, srcSize, maxSymbolValue, huffLog, 0 /* 4 streams */, workSpace, wkspSize, NULL, NULL, 0); -} - -size_t HUF_compress4X_repeat(void *dst, size_t dstSize, const void *src, size_t srcSize, unsigned maxSymbolValue, unsigned huffLog, void *workSpace, - size_t wkspSize, HUF_CElt *hufTable, HUF_repeat *repeat, int preferRepeat) -{ - return HUF_compress_internal(dst, dstSize, src, srcSize, maxSymbolValue, huffLog, 0 /* 4 streams */, workSpace, wkspSize, hufTable, repeat, - preferRepeat); -} diff --git a/lib/zstd/huf_decompress.c b/lib/zstd/huf_decompress.c deleted file mode 100644 index 6526482047dc..000000000000 --- a/lib/zstd/huf_decompress.c +++ /dev/null @@ -1,960 +0,0 @@ -/* - * Huffman decoder, part of New Generation Entropy library - * Copyright (C) 2013-2016, Yann Collet. - * - * BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php) - * - * Redistribution and use in source and binary forms, with or without - * modification, are permitted provided that the following conditions are - * met: - * - * * Redistributions of source code must retain the above copyright - * notice, this list of conditions and the following disclaimer. - * * Redistributions in binary form must reproduce the above - * copyright notice, this list of conditions and the following disclaimer - * in the documentation and/or other materials provided with the - * distribution. - * - * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS - * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT - * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR - * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT - * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, - * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT - * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, - * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY - * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT - * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE - * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. - * - * This program is free software; you can redistribute it and/or modify it under - * the terms of the GNU General Public License version 2 as published by the - * Free Software Foundation. This program is dual-licensed; you may select - * either version 2 of the GNU General Public License ("GPL") or BSD license - * ("BSD"). - * - * You can contact the author at : - * - Source repository : https://github.com/Cyan4973/FiniteStateEntropy - */ - -/* ************************************************************** -* Compiler specifics -****************************************************************/ -#define FORCE_INLINE static __always_inline - -/* ************************************************************** -* Dependencies -****************************************************************/ -#include "bitstream.h" /* BIT_* */ -#include "fse.h" /* header compression */ -#include "huf.h" -#include -#include -#include /* memcpy, memset */ - -/* ************************************************************** -* Error Management -****************************************************************/ -#define HUF_STATIC_ASSERT(c) \ - { \ - enum { HUF_static_assert = 1 / (int)(!!(c)) }; \ - } /* use only *after* variable declarations */ - -/*-***************************/ -/* generic DTableDesc */ -/*-***************************/ - -typedef struct { - BYTE maxTableLog; - BYTE tableType; - BYTE tableLog; - BYTE reserved; -} DTableDesc; - -static DTableDesc HUF_getDTableDesc(const HUF_DTable *table) -{ - DTableDesc dtd; - memcpy(&dtd, table, sizeof(dtd)); - return dtd; -} - -/*-***************************/ -/* single-symbol decoding */ -/*-***************************/ - -typedef struct { - BYTE byte; - BYTE nbBits; -} HUF_DEltX2; /* single-symbol decoding */ - -size_t HUF_readDTableX2_wksp(HUF_DTable *DTable, const void *src, size_t srcSize, void *workspace, size_t workspaceSize) -{ - U32 tableLog = 0; - U32 nbSymbols = 0; - size_t iSize; - void *const dtPtr = DTable + 1; - HUF_DEltX2 *const dt = (HUF_DEltX2 *)dtPtr; - - U32 *rankVal; - BYTE *huffWeight; - size_t spaceUsed32 = 0; - - rankVal = (U32 *)workspace + spaceUsed32; - spaceUsed32 += HUF_TABLELOG_ABSOLUTEMAX + 1; - huffWeight = (BYTE *)((U32 *)workspace + spaceUsed32); - spaceUsed32 += ALIGN(HUF_SYMBOLVALUE_MAX + 1, sizeof(U32)) >> 2; - - if ((spaceUsed32 << 2) > workspaceSize) - return ERROR(tableLog_tooLarge); - workspace = (U32 *)workspace + spaceUsed32; - workspaceSize -= (spaceUsed32 << 2); - - HUF_STATIC_ASSERT(sizeof(DTableDesc) == sizeof(HUF_DTable)); - /* memset(huffWeight, 0, sizeof(huffWeight)); */ /* is not necessary, even though some analyzer complain ... */ - - iSize = HUF_readStats_wksp(huffWeight, HUF_SYMBOLVALUE_MAX + 1, rankVal, &nbSymbols, &tableLog, src, srcSize, workspace, workspaceSize); - if (HUF_isError(iSize)) - return iSize; - - /* Table header */ - { - DTableDesc dtd = HUF_getDTableDesc(DTable); - if (tableLog > (U32)(dtd.maxTableLog + 1)) - return ERROR(tableLog_tooLarge); /* DTable too small, Huffman tree cannot fit in */ - dtd.tableType = 0; - dtd.tableLog = (BYTE)tableLog; - memcpy(DTable, &dtd, sizeof(dtd)); - } - - /* Calculate starting value for each rank */ - { - U32 n, nextRankStart = 0; - for (n = 1; n < tableLog + 1; n++) { - U32 const curr = nextRankStart; - nextRankStart += (rankVal[n] << (n - 1)); - rankVal[n] = curr; - } - } - - /* fill DTable */ - { - U32 n; - for (n = 0; n < nbSymbols; n++) { - U32 const w = huffWeight[n]; - U32 const length = (1 << w) >> 1; - U32 u; - HUF_DEltX2 D; - D.byte = (BYTE)n; - D.nbBits = (BYTE)(tableLog + 1 - w); - for (u = rankVal[w]; u < rankVal[w] + length; u++) - dt[u] = D; - rankVal[w] += length; - } - } - - return iSize; -} - -static BYTE HUF_decodeSymbolX2(BIT_DStream_t *Dstream, const HUF_DEltX2 *dt, const U32 dtLog) -{ - size_t const val = BIT_lookBitsFast(Dstream, dtLog); /* note : dtLog >= 1 */ - BYTE const c = dt[val].byte; - BIT_skipBits(Dstream, dt[val].nbBits); - return c; -} - -#define HUF_DECODE_SYMBOLX2_0(ptr, DStreamPtr) *ptr++ = HUF_decodeSymbolX2(DStreamPtr, dt, dtLog) - -#define HUF_DECODE_SYMBOLX2_1(ptr, DStreamPtr) \ - if (ZSTD_64bits() || (HUF_TABLELOG_MAX <= 12)) \ - HUF_DECODE_SYMBOLX2_0(ptr, DStreamPtr) - -#define HUF_DECODE_SYMBOLX2_2(ptr, DStreamPtr) \ - if (ZSTD_64bits()) \ - HUF_DECODE_SYMBOLX2_0(ptr, DStreamPtr) - -FORCE_INLINE size_t HUF_decodeStreamX2(BYTE *p, BIT_DStream_t *const bitDPtr, BYTE *const pEnd, const HUF_DEltX2 *const dt, const U32 dtLog) -{ - BYTE *const pStart = p; - - /* up to 4 symbols at a time */ - while ((BIT_reloadDStream(bitDPtr) == BIT_DStream_unfinished) && (p <= pEnd - 4)) { - HUF_DECODE_SYMBOLX2_2(p, bitDPtr); - HUF_DECODE_SYMBOLX2_1(p, bitDPtr); - HUF_DECODE_SYMBOLX2_2(p, bitDPtr); - HUF_DECODE_SYMBOLX2_0(p, bitDPtr); - } - - /* closer to the end */ - while ((BIT_reloadDStream(bitDPtr) == BIT_DStream_unfinished) && (p < pEnd)) - HUF_DECODE_SYMBOLX2_0(p, bitDPtr); - - /* no more data to retrieve from bitstream, hence no need to reload */ - while (p < pEnd) - HUF_DECODE_SYMBOLX2_0(p, bitDPtr); - - return pEnd - pStart; -} - -static size_t HUF_decompress1X2_usingDTable_internal(void *dst, size_t dstSize, const void *cSrc, size_t cSrcSize, const HUF_DTable *DTable) -{ - BYTE *op = (BYTE *)dst; - BYTE *const oend = op + dstSize; - const void *dtPtr = DTable + 1; - const HUF_DEltX2 *const dt = (const HUF_DEltX2 *)dtPtr; - BIT_DStream_t bitD; - DTableDesc const dtd = HUF_getDTableDesc(DTable); - U32 const dtLog = dtd.tableLog; - - { - size_t const errorCode = BIT_initDStream(&bitD, cSrc, cSrcSize); - if (HUF_isError(errorCode)) - return errorCode; - } - - HUF_decodeStreamX2(op, &bitD, oend, dt, dtLog); - - /* check */ - if (!BIT_endOfDStream(&bitD)) - return ERROR(corruption_detected); - - return dstSize; -} - -size_t HUF_decompress1X2_usingDTable(void *dst, size_t dstSize, const void *cSrc, size_t cSrcSize, const HUF_DTable *DTable) -{ - DTableDesc dtd = HUF_getDTableDesc(DTable); - if (dtd.tableType != 0) - return ERROR(GENERIC); - return HUF_decompress1X2_usingDTable_internal(dst, dstSize, cSrc, cSrcSize, DTable); -} - -size_t HUF_decompress1X2_DCtx_wksp(HUF_DTable *DCtx, void *dst, size_t dstSize, const void *cSrc, size_t cSrcSize, void *workspace, size_t workspaceSize) -{ - const BYTE *ip = (const BYTE *)cSrc; - - size_t const hSize = HUF_readDTableX2_wksp(DCtx, cSrc, cSrcSize, workspace, workspaceSize); - if (HUF_isError(hSize)) - return hSize; - if (hSize >= cSrcSize) - return ERROR(srcSize_wrong); - ip += hSize; - cSrcSize -= hSize; - - return HUF_decompress1X2_usingDTable_internal(dst, dstSize, ip, cSrcSize, DCtx); -} - -static size_t HUF_decompress4X2_usingDTable_internal(void *dst, size_t dstSize, const void *cSrc, size_t cSrcSize, const HUF_DTable *DTable) -{ - /* Check */ - if (cSrcSize < 10) - return ERROR(corruption_detected); /* strict minimum : jump table + 1 byte per stream */ - - { - const BYTE *const istart = (const BYTE *)cSrc; - BYTE *const ostart = (BYTE *)dst; - BYTE *const oend = ostart + dstSize; - const void *const dtPtr = DTable + 1; - const HUF_DEltX2 *const dt = (const HUF_DEltX2 *)dtPtr; - - /* Init */ - BIT_DStream_t bitD1; - BIT_DStream_t bitD2; - BIT_DStream_t bitD3; - BIT_DStream_t bitD4; - size_t const length1 = ZSTD_readLE16(istart); - size_t const length2 = ZSTD_readLE16(istart + 2); - size_t const length3 = ZSTD_readLE16(istart + 4); - size_t const length4 = cSrcSize - (length1 + length2 + length3 + 6); - const BYTE *const istart1 = istart + 6; /* jumpTable */ - const BYTE *const istart2 = istart1 + length1; - const BYTE *const istart3 = istart2 + length2; - const BYTE *const istart4 = istart3 + length3; - const size_t segmentSize = (dstSize + 3) / 4; - BYTE *const opStart2 = ostart + segmentSize; - BYTE *const opStart3 = opStart2 + segmentSize; - BYTE *const opStart4 = opStart3 + segmentSize; - BYTE *op1 = ostart; - BYTE *op2 = opStart2; - BYTE *op3 = opStart3; - BYTE *op4 = opStart4; - U32 endSignal; - DTableDesc const dtd = HUF_getDTableDesc(DTable); - U32 const dtLog = dtd.tableLog; - - if (length4 > cSrcSize) - return ERROR(corruption_detected); /* overflow */ - { - size_t const errorCode = BIT_initDStream(&bitD1, istart1, length1); - if (HUF_isError(errorCode)) - return errorCode; - } - { - size_t const errorCode = BIT_initDStream(&bitD2, istart2, length2); - if (HUF_isError(errorCode)) - return errorCode; - } - { - size_t const errorCode = BIT_initDStream(&bitD3, istart3, length3); - if (HUF_isError(errorCode)) - return errorCode; - } - { - size_t const errorCode = BIT_initDStream(&bitD4, istart4, length4); - if (HUF_isError(errorCode)) - return errorCode; - } - - /* 16-32 symbols per loop (4-8 symbols per stream) */ - endSignal = BIT_reloadDStream(&bitD1) | BIT_reloadDStream(&bitD2) | BIT_reloadDStream(&bitD3) | BIT_reloadDStream(&bitD4); - for (; (endSignal == BIT_DStream_unfinished) && (op4 < (oend - 7));) { - HUF_DECODE_SYMBOLX2_2(op1, &bitD1); - HUF_DECODE_SYMBOLX2_2(op2, &bitD2); - HUF_DECODE_SYMBOLX2_2(op3, &bitD3); - HUF_DECODE_SYMBOLX2_2(op4, &bitD4); - HUF_DECODE_SYMBOLX2_1(op1, &bitD1); - HUF_DECODE_SYMBOLX2_1(op2, &bitD2); - HUF_DECODE_SYMBOLX2_1(op3, &bitD3); - HUF_DECODE_SYMBOLX2_1(op4, &bitD4); - HUF_DECODE_SYMBOLX2_2(op1, &bitD1); - HUF_DECODE_SYMBOLX2_2(op2, &bitD2); - HUF_DECODE_SYMBOLX2_2(op3, &bitD3); - HUF_DECODE_SYMBOLX2_2(op4, &bitD4); - HUF_DECODE_SYMBOLX2_0(op1, &bitD1); - HUF_DECODE_SYMBOLX2_0(op2, &bitD2); - HUF_DECODE_SYMBOLX2_0(op3, &bitD3); - HUF_DECODE_SYMBOLX2_0(op4, &bitD4); - endSignal = BIT_reloadDStream(&bitD1) | BIT_reloadDStream(&bitD2) | BIT_reloadDStream(&bitD3) | BIT_reloadDStream(&bitD4); - } - - /* check corruption */ - if (op1 > opStart2) - return ERROR(corruption_detected); - if (op2 > opStart3) - return ERROR(corruption_detected); - if (op3 > opStart4) - return ERROR(corruption_detected); - /* note : op4 supposed already verified within main loop */ - - /* finish bitStreams one by one */ - HUF_decodeStreamX2(op1, &bitD1, opStart2, dt, dtLog); - HUF_decodeStreamX2(op2, &bitD2, opStart3, dt, dtLog); - HUF_decodeStreamX2(op3, &bitD3, opStart4, dt, dtLog); - HUF_decodeStreamX2(op4, &bitD4, oend, dt, dtLog); - - /* check */ - endSignal = BIT_endOfDStream(&bitD1) & BIT_endOfDStream(&bitD2) & BIT_endOfDStream(&bitD3) & BIT_endOfDStream(&bitD4); - if (!endSignal) - return ERROR(corruption_detected); - - /* decoded size */ - return dstSize; - } -} - -size_t HUF_decompress4X2_usingDTable(void *dst, size_t dstSize, const void *cSrc, size_t cSrcSize, const HUF_DTable *DTable) -{ - DTableDesc dtd = HUF_getDTableDesc(DTable); - if (dtd.tableType != 0) - return ERROR(GENERIC); - return HUF_decompress4X2_usingDTable_internal(dst, dstSize, cSrc, cSrcSize, DTable); -} - -size_t HUF_decompress4X2_DCtx_wksp(HUF_DTable *dctx, void *dst, size_t dstSize, const void *cSrc, size_t cSrcSize, void *workspace, size_t workspaceSize) -{ - const BYTE *ip = (const BYTE *)cSrc; - - size_t const hSize = HUF_readDTableX2_wksp(dctx, cSrc, cSrcSize, workspace, workspaceSize); - if (HUF_isError(hSize)) - return hSize; - if (hSize >= cSrcSize) - return ERROR(srcSize_wrong); - ip += hSize; - cSrcSize -= hSize; - - return HUF_decompress4X2_usingDTable_internal(dst, dstSize, ip, cSrcSize, dctx); -} - -/* *************************/ -/* double-symbols decoding */ -/* *************************/ -typedef struct { - U16 sequence; - BYTE nbBits; - BYTE length; -} HUF_DEltX4; /* double-symbols decoding */ - -typedef struct { - BYTE symbol; - BYTE weight; -} sortedSymbol_t; - -/* HUF_fillDTableX4Level2() : - * `rankValOrigin` must be a table of at least (HUF_TABLELOG_MAX + 1) U32 */ -static void HUF_fillDTableX4Level2(HUF_DEltX4 *DTable, U32 sizeLog, const U32 consumed, const U32 *rankValOrigin, const int minWeight, - const sortedSymbol_t *sortedSymbols, const U32 sortedListSize, U32 nbBitsBaseline, U16 baseSeq) -{ - HUF_DEltX4 DElt; - U32 rankVal[HUF_TABLELOG_MAX + 1]; - - /* get pre-calculated rankVal */ - memcpy(rankVal, rankValOrigin, sizeof(rankVal)); - - /* fill skipped values */ - if (minWeight > 1) { - U32 i, skipSize = rankVal[minWeight]; - ZSTD_writeLE16(&(DElt.sequence), baseSeq); - DElt.nbBits = (BYTE)(consumed); - DElt.length = 1; - for (i = 0; i < skipSize; i++) - DTable[i] = DElt; - } - - /* fill DTable */ - { - U32 s; - for (s = 0; s < sortedListSize; s++) { /* note : sortedSymbols already skipped */ - const U32 symbol = sortedSymbols[s].symbol; - const U32 weight = sortedSymbols[s].weight; - const U32 nbBits = nbBitsBaseline - weight; - const U32 length = 1 << (sizeLog - nbBits); - const U32 start = rankVal[weight]; - U32 i = start; - const U32 end = start + length; - - ZSTD_writeLE16(&(DElt.sequence), (U16)(baseSeq + (symbol << 8))); - DElt.nbBits = (BYTE)(nbBits + consumed); - DElt.length = 2; - do { - DTable[i++] = DElt; - } while (i < end); /* since length >= 1 */ - - rankVal[weight] += length; - } - } -} - -typedef U32 rankVal_t[HUF_TABLELOG_MAX][HUF_TABLELOG_MAX + 1]; -typedef U32 rankValCol_t[HUF_TABLELOG_MAX + 1]; - -static void HUF_fillDTableX4(HUF_DEltX4 *DTable, const U32 targetLog, const sortedSymbol_t *sortedList, const U32 sortedListSize, const U32 *rankStart, - rankVal_t rankValOrigin, const U32 maxWeight, const U32 nbBitsBaseline) -{ - U32 rankVal[HUF_TABLELOG_MAX + 1]; - const int scaleLog = nbBitsBaseline - targetLog; /* note : targetLog >= srcLog, hence scaleLog <= 1 */ - const U32 minBits = nbBitsBaseline - maxWeight; - U32 s; - - memcpy(rankVal, rankValOrigin, sizeof(rankVal)); - - /* fill DTable */ - for (s = 0; s < sortedListSize; s++) { - const U16 symbol = sortedList[s].symbol; - const U32 weight = sortedList[s].weight; - const U32 nbBits = nbBitsBaseline - weight; - const U32 start = rankVal[weight]; - const U32 length = 1 << (targetLog - nbBits); - - if (targetLog - nbBits >= minBits) { /* enough room for a second symbol */ - U32 sortedRank; - int minWeight = nbBits + scaleLog; - if (minWeight < 1) - minWeight = 1; - sortedRank = rankStart[minWeight]; - HUF_fillDTableX4Level2(DTable + start, targetLog - nbBits, nbBits, rankValOrigin[nbBits], minWeight, sortedList + sortedRank, - sortedListSize - sortedRank, nbBitsBaseline, symbol); - } else { - HUF_DEltX4 DElt; - ZSTD_writeLE16(&(DElt.sequence), symbol); - DElt.nbBits = (BYTE)(nbBits); - DElt.length = 1; - { - U32 const end = start + length; - U32 u; - for (u = start; u < end; u++) - DTable[u] = DElt; - } - } - rankVal[weight] += length; - } -} - -size_t HUF_readDTableX4_wksp(HUF_DTable *DTable, const void *src, size_t srcSize, void *workspace, size_t workspaceSize) -{ - U32 tableLog, maxW, sizeOfSort, nbSymbols; - DTableDesc dtd = HUF_getDTableDesc(DTable); - U32 const maxTableLog = dtd.maxTableLog; - size_t iSize; - void *dtPtr = DTable + 1; /* force compiler to avoid strict-aliasing */ - HUF_DEltX4 *const dt = (HUF_DEltX4 *)dtPtr; - U32 *rankStart; - - rankValCol_t *rankVal; - U32 *rankStats; - U32 *rankStart0; - sortedSymbol_t *sortedSymbol; - BYTE *weightList; - size_t spaceUsed32 = 0; - - HUF_STATIC_ASSERT((sizeof(rankValCol_t) & 3) == 0); - - rankVal = (rankValCol_t *)((U32 *)workspace + spaceUsed32); - spaceUsed32 += (sizeof(rankValCol_t) * HUF_TABLELOG_MAX) >> 2; - rankStats = (U32 *)workspace + spaceUsed32; - spaceUsed32 += HUF_TABLELOG_MAX + 1; - rankStart0 = (U32 *)workspace + spaceUsed32; - spaceUsed32 += HUF_TABLELOG_MAX + 2; - sortedSymbol = (sortedSymbol_t *)((U32 *)workspace + spaceUsed32); - spaceUsed32 += ALIGN(sizeof(sortedSymbol_t) * (HUF_SYMBOLVALUE_MAX + 1), sizeof(U32)) >> 2; - weightList = (BYTE *)((U32 *)workspace + spaceUsed32); - spaceUsed32 += ALIGN(HUF_SYMBOLVALUE_MAX + 1, sizeof(U32)) >> 2; - - if ((spaceUsed32 << 2) > workspaceSize) - return ERROR(tableLog_tooLarge); - workspace = (U32 *)workspace + spaceUsed32; - workspaceSize -= (spaceUsed32 << 2); - - rankStart = rankStart0 + 1; - memset(rankStats, 0, sizeof(U32) * (2 * HUF_TABLELOG_MAX + 2 + 1)); - - HUF_STATIC_ASSERT(sizeof(HUF_DEltX4) == sizeof(HUF_DTable)); /* if compiler fails here, assertion is wrong */ - if (maxTableLog > HUF_TABLELOG_MAX) - return ERROR(tableLog_tooLarge); - /* memset(weightList, 0, sizeof(weightList)); */ /* is not necessary, even though some analyzer complain ... */ - - iSize = HUF_readStats_wksp(weightList, HUF_SYMBOLVALUE_MAX + 1, rankStats, &nbSymbols, &tableLog, src, srcSize, workspace, workspaceSize); - if (HUF_isError(iSize)) - return iSize; - - /* check result */ - if (tableLog > maxTableLog) - return ERROR(tableLog_tooLarge); /* DTable can't fit code depth */ - - /* find maxWeight */ - for (maxW = tableLog; rankStats[maxW] == 0; maxW--) { - } /* necessarily finds a solution before 0 */ - - /* Get start index of each weight */ - { - U32 w, nextRankStart = 0; - for (w = 1; w < maxW + 1; w++) { - U32 curr = nextRankStart; - nextRankStart += rankStats[w]; - rankStart[w] = curr; - } - rankStart[0] = nextRankStart; /* put all 0w symbols at the end of sorted list*/ - sizeOfSort = nextRankStart; - } - - /* sort symbols by weight */ - { - U32 s; - for (s = 0; s < nbSymbols; s++) { - U32 const w = weightList[s]; - U32 const r = rankStart[w]++; - sortedSymbol[r].symbol = (BYTE)s; - sortedSymbol[r].weight = (BYTE)w; - } - rankStart[0] = 0; /* forget 0w symbols; this is beginning of weight(1) */ - } - - /* Build rankVal */ - { - U32 *const rankVal0 = rankVal[0]; - { - int const rescale = (maxTableLog - tableLog) - 1; /* tableLog <= maxTableLog */ - U32 nextRankVal = 0; - U32 w; - for (w = 1; w < maxW + 1; w++) { - U32 curr = nextRankVal; - nextRankVal += rankStats[w] << (w + rescale); - rankVal0[w] = curr; - } - } - { - U32 const minBits = tableLog + 1 - maxW; - U32 consumed; - for (consumed = minBits; consumed < maxTableLog - minBits + 1; consumed++) { - U32 *const rankValPtr = rankVal[consumed]; - U32 w; - for (w = 1; w < maxW + 1; w++) { - rankValPtr[w] = rankVal0[w] >> consumed; - } - } - } - } - - HUF_fillDTableX4(dt, maxTableLog, sortedSymbol, sizeOfSort, rankStart0, rankVal, maxW, tableLog + 1); - - dtd.tableLog = (BYTE)maxTableLog; - dtd.tableType = 1; - memcpy(DTable, &dtd, sizeof(dtd)); - return iSize; -} - -static U32 HUF_decodeSymbolX4(void *op, BIT_DStream_t *DStream, const HUF_DEltX4 *dt, const U32 dtLog) -{ - size_t const val = BIT_lookBitsFast(DStream, dtLog); /* note : dtLog >= 1 */ - memcpy(op, dt + val, 2); - BIT_skipBits(DStream, dt[val].nbBits); - return dt[val].length; -} - -static U32 HUF_decodeLastSymbolX4(void *op, BIT_DStream_t *DStream, const HUF_DEltX4 *dt, const U32 dtLog) -{ - size_t const val = BIT_lookBitsFast(DStream, dtLog); /* note : dtLog >= 1 */ - memcpy(op, dt + val, 1); - if (dt[val].length == 1) - BIT_skipBits(DStream, dt[val].nbBits); - else { - if (DStream->bitsConsumed < (sizeof(DStream->bitContainer) * 8)) { - BIT_skipBits(DStream, dt[val].nbBits); - if (DStream->bitsConsumed > (sizeof(DStream->bitContainer) * 8)) - /* ugly hack; works only because it's the last symbol. Note : can't easily extract nbBits from just this symbol */ - DStream->bitsConsumed = (sizeof(DStream->bitContainer) * 8); - } - } - return 1; -} - -#define HUF_DECODE_SYMBOLX4_0(ptr, DStreamPtr) ptr += HUF_decodeSymbolX4(ptr, DStreamPtr, dt, dtLog) - -#define HUF_DECODE_SYMBOLX4_1(ptr, DStreamPtr) \ - if (ZSTD_64bits() || (HUF_TABLELOG_MAX <= 12)) \ - ptr += HUF_decodeSymbolX4(ptr, DStreamPtr, dt, dtLog) - -#define HUF_DECODE_SYMBOLX4_2(ptr, DStreamPtr) \ - if (ZSTD_64bits()) \ - ptr += HUF_decodeSymbolX4(ptr, DStreamPtr, dt, dtLog) - -FORCE_INLINE size_t HUF_decodeStreamX4(BYTE *p, BIT_DStream_t *bitDPtr, BYTE *const pEnd, const HUF_DEltX4 *const dt, const U32 dtLog) -{ - BYTE *const pStart = p; - - /* up to 8 symbols at a time */ - while ((BIT_reloadDStream(bitDPtr) == BIT_DStream_unfinished) & (p < pEnd - (sizeof(bitDPtr->bitContainer) - 1))) { - HUF_DECODE_SYMBOLX4_2(p, bitDPtr); - HUF_DECODE_SYMBOLX4_1(p, bitDPtr); - HUF_DECODE_SYMBOLX4_2(p, bitDPtr); - HUF_DECODE_SYMBOLX4_0(p, bitDPtr); - } - - /* closer to end : up to 2 symbols at a time */ - while ((BIT_reloadDStream(bitDPtr) == BIT_DStream_unfinished) & (p <= pEnd - 2)) - HUF_DECODE_SYMBOLX4_0(p, bitDPtr); - - while (p <= pEnd - 2) - HUF_DECODE_SYMBOLX4_0(p, bitDPtr); /* no need to reload : reached the end of DStream */ - - if (p < pEnd) - p += HUF_decodeLastSymbolX4(p, bitDPtr, dt, dtLog); - - return p - pStart; -} - -static size_t HUF_decompress1X4_usingDTable_internal(void *dst, size_t dstSize, const void *cSrc, size_t cSrcSize, const HUF_DTable *DTable) -{ - BIT_DStream_t bitD; - - /* Init */ - { - size_t const errorCode = BIT_initDStream(&bitD, cSrc, cSrcSize); - if (HUF_isError(errorCode)) - return errorCode; - } - - /* decode */ - { - BYTE *const ostart = (BYTE *)dst; - BYTE *const oend = ostart + dstSize; - const void *const dtPtr = DTable + 1; /* force compiler to not use strict-aliasing */ - const HUF_DEltX4 *const dt = (const HUF_DEltX4 *)dtPtr; - DTableDesc const dtd = HUF_getDTableDesc(DTable); - HUF_decodeStreamX4(ostart, &bitD, oend, dt, dtd.tableLog); - } - - /* check */ - if (!BIT_endOfDStream(&bitD)) - return ERROR(corruption_detected); - - /* decoded size */ - return dstSize; -} - -size_t HUF_decompress1X4_usingDTable(void *dst, size_t dstSize, const void *cSrc, size_t cSrcSize, const HUF_DTable *DTable) -{ - DTableDesc dtd = HUF_getDTableDesc(DTable); - if (dtd.tableType != 1) - return ERROR(GENERIC); - return HUF_decompress1X4_usingDTable_internal(dst, dstSize, cSrc, cSrcSize, DTable); -} - -size_t HUF_decompress1X4_DCtx_wksp(HUF_DTable *DCtx, void *dst, size_t dstSize, const void *cSrc, size_t cSrcSize, void *workspace, size_t workspaceSize) -{ - const BYTE *ip = (const BYTE *)cSrc; - - size_t const hSize = HUF_readDTableX4_wksp(DCtx, cSrc, cSrcSize, workspace, workspaceSize); - if (HUF_isError(hSize)) - return hSize; - if (hSize >= cSrcSize) - return ERROR(srcSize_wrong); - ip += hSize; - cSrcSize -= hSize; - - return HUF_decompress1X4_usingDTable_internal(dst, dstSize, ip, cSrcSize, DCtx); -} - -static size_t HUF_decompress4X4_usingDTable_internal(void *dst, size_t dstSize, const void *cSrc, size_t cSrcSize, const HUF_DTable *DTable) -{ - if (cSrcSize < 10) - return ERROR(corruption_detected); /* strict minimum : jump table + 1 byte per stream */ - - { - const BYTE *const istart = (const BYTE *)cSrc; - BYTE *const ostart = (BYTE *)dst; - BYTE *const oend = ostart + dstSize; - const void *const dtPtr = DTable + 1; - const HUF_DEltX4 *const dt = (const HUF_DEltX4 *)dtPtr; - - /* Init */ - BIT_DStream_t bitD1; - BIT_DStream_t bitD2; - BIT_DStream_t bitD3; - BIT_DStream_t bitD4; - size_t const length1 = ZSTD_readLE16(istart); - size_t const length2 = ZSTD_readLE16(istart + 2); - size_t const length3 = ZSTD_readLE16(istart + 4); - size_t const length4 = cSrcSize - (length1 + length2 + length3 + 6); - const BYTE *const istart1 = istart + 6; /* jumpTable */ - const BYTE *const istart2 = istart1 + length1; - const BYTE *const istart3 = istart2 + length2; - const BYTE *const istart4 = istart3 + length3; - size_t const segmentSize = (dstSize + 3) / 4; - BYTE *const opStart2 = ostart + segmentSize; - BYTE *const opStart3 = opStart2 + segmentSize; - BYTE *const opStart4 = opStart3 + segmentSize; - BYTE *op1 = ostart; - BYTE *op2 = opStart2; - BYTE *op3 = opStart3; - BYTE *op4 = opStart4; - U32 endSignal; - DTableDesc const dtd = HUF_getDTableDesc(DTable); - U32 const dtLog = dtd.tableLog; - - if (length4 > cSrcSize) - return ERROR(corruption_detected); /* overflow */ - { - size_t const errorCode = BIT_initDStream(&bitD1, istart1, length1); - if (HUF_isError(errorCode)) - return errorCode; - } - { - size_t const errorCode = BIT_initDStream(&bitD2, istart2, length2); - if (HUF_isError(errorCode)) - return errorCode; - } - { - size_t const errorCode = BIT_initDStream(&bitD3, istart3, length3); - if (HUF_isError(errorCode)) - return errorCode; - } - { - size_t const errorCode = BIT_initDStream(&bitD4, istart4, length4); - if (HUF_isError(errorCode)) - return errorCode; - } - - /* 16-32 symbols per loop (4-8 symbols per stream) */ - endSignal = BIT_reloadDStream(&bitD1) | BIT_reloadDStream(&bitD2) | BIT_reloadDStream(&bitD3) | BIT_reloadDStream(&bitD4); - for (; (endSignal == BIT_DStream_unfinished) & (op4 < (oend - (sizeof(bitD4.bitContainer) - 1)));) { - HUF_DECODE_SYMBOLX4_2(op1, &bitD1); - HUF_DECODE_SYMBOLX4_2(op2, &bitD2); - HUF_DECODE_SYMBOLX4_2(op3, &bitD3); - HUF_DECODE_SYMBOLX4_2(op4, &bitD4); - HUF_DECODE_SYMBOLX4_1(op1, &bitD1); - HUF_DECODE_SYMBOLX4_1(op2, &bitD2); - HUF_DECODE_SYMBOLX4_1(op3, &bitD3); - HUF_DECODE_SYMBOLX4_1(op4, &bitD4); - HUF_DECODE_SYMBOLX4_2(op1, &bitD1); - HUF_DECODE_SYMBOLX4_2(op2, &bitD2); - HUF_DECODE_SYMBOLX4_2(op3, &bitD3); - HUF_DECODE_SYMBOLX4_2(op4, &bitD4); - HUF_DECODE_SYMBOLX4_0(op1, &bitD1); - HUF_DECODE_SYMBOLX4_0(op2, &bitD2); - HUF_DECODE_SYMBOLX4_0(op3, &bitD3); - HUF_DECODE_SYMBOLX4_0(op4, &bitD4); - - endSignal = BIT_reloadDStream(&bitD1) | BIT_reloadDStream(&bitD2) | BIT_reloadDStream(&bitD3) | BIT_reloadDStream(&bitD4); - } - - /* check corruption */ - if (op1 > opStart2) - return ERROR(corruption_detected); - if (op2 > opStart3) - return ERROR(corruption_detected); - if (op3 > opStart4) - return ERROR(corruption_detected); - /* note : op4 already verified within main loop */ - - /* finish bitStreams one by one */ - HUF_decodeStreamX4(op1, &bitD1, opStart2, dt, dtLog); - HUF_decodeStreamX4(op2, &bitD2, opStart3, dt, dtLog); - HUF_decodeStreamX4(op3, &bitD3, opStart4, dt, dtLog); - HUF_decodeStreamX4(op4, &bitD4, oend, dt, dtLog); - - /* check */ - { - U32 const endCheck = BIT_endOfDStream(&bitD1) & BIT_endOfDStream(&bitD2) & BIT_endOfDStream(&bitD3) & BIT_endOfDStream(&bitD4); - if (!endCheck) - return ERROR(corruption_detected); - } - - /* decoded size */ - return dstSize; - } -} - -size_t HUF_decompress4X4_usingDTable(void *dst, size_t dstSize, const void *cSrc, size_t cSrcSize, const HUF_DTable *DTable) -{ - DTableDesc dtd = HUF_getDTableDesc(DTable); - if (dtd.tableType != 1) - return ERROR(GENERIC); - return HUF_decompress4X4_usingDTable_internal(dst, dstSize, cSrc, cSrcSize, DTable); -} - -size_t HUF_decompress4X4_DCtx_wksp(HUF_DTable *dctx, void *dst, size_t dstSize, const void *cSrc, size_t cSrcSize, void *workspace, size_t workspaceSize) -{ - const BYTE *ip = (const BYTE *)cSrc; - - size_t hSize = HUF_readDTableX4_wksp(dctx, cSrc, cSrcSize, workspace, workspaceSize); - if (HUF_isError(hSize)) - return hSize; - if (hSize >= cSrcSize) - return ERROR(srcSize_wrong); - ip += hSize; - cSrcSize -= hSize; - - return HUF_decompress4X4_usingDTable_internal(dst, dstSize, ip, cSrcSize, dctx); -} - -/* ********************************/ -/* Generic decompression selector */ -/* ********************************/ - -size_t HUF_decompress1X_usingDTable(void *dst, size_t maxDstSize, const void *cSrc, size_t cSrcSize, const HUF_DTable *DTable) -{ - DTableDesc const dtd = HUF_getDTableDesc(DTable); - return dtd.tableType ? HUF_decompress1X4_usingDTable_internal(dst, maxDstSize, cSrc, cSrcSize, DTable) - : HUF_decompress1X2_usingDTable_internal(dst, maxDstSize, cSrc, cSrcSize, DTable); -} - -size_t HUF_decompress4X_usingDTable(void *dst, size_t maxDstSize, const void *cSrc, size_t cSrcSize, const HUF_DTable *DTable) -{ - DTableDesc const dtd = HUF_getDTableDesc(DTable); - return dtd.tableType ? HUF_decompress4X4_usingDTable_internal(dst, maxDstSize, cSrc, cSrcSize, DTable) - : HUF_decompress4X2_usingDTable_internal(dst, maxDstSize, cSrc, cSrcSize, DTable); -} - -typedef struct { - U32 tableTime; - U32 decode256Time; -} algo_time_t; -static const algo_time_t algoTime[16 /* Quantization */][3 /* single, double, quad */] = { - /* single, double, quad */ - {{0, 0}, {1, 1}, {2, 2}}, /* Q==0 : impossible */ - {{0, 0}, {1, 1}, {2, 2}}, /* Q==1 : impossible */ - {{38, 130}, {1313, 74}, {2151, 38}}, /* Q == 2 : 12-18% */ - {{448, 128}, {1353, 74}, {2238, 41}}, /* Q == 3 : 18-25% */ - {{556, 128}, {1353, 74}, {2238, 47}}, /* Q == 4 : 25-32% */ - {{714, 128}, {1418, 74}, {2436, 53}}, /* Q == 5 : 32-38% */ - {{883, 128}, {1437, 74}, {2464, 61}}, /* Q == 6 : 38-44% */ - {{897, 128}, {1515, 75}, {2622, 68}}, /* Q == 7 : 44-50% */ - {{926, 128}, {1613, 75}, {2730, 75}}, /* Q == 8 : 50-56% */ - {{947, 128}, {1729, 77}, {3359, 77}}, /* Q == 9 : 56-62% */ - {{1107, 128}, {2083, 81}, {4006, 84}}, /* Q ==10 : 62-69% */ - {{1177, 128}, {2379, 87}, {4785, 88}}, /* Q ==11 : 69-75% */ - {{1242, 128}, {2415, 93}, {5155, 84}}, /* Q ==12 : 75-81% */ - {{1349, 128}, {2644, 106}, {5260, 106}}, /* Q ==13 : 81-87% */ - {{1455, 128}, {2422, 124}, {4174, 124}}, /* Q ==14 : 87-93% */ - {{722, 128}, {1891, 145}, {1936, 146}}, /* Q ==15 : 93-99% */ -}; - -/** HUF_selectDecoder() : -* Tells which decoder is likely to decode faster, -* based on a set of pre-determined metrics. -* @return : 0==HUF_decompress4X2, 1==HUF_decompress4X4 . -* Assumption : 0 < cSrcSize < dstSize <= 128 KB */ -U32 HUF_selectDecoder(size_t dstSize, size_t cSrcSize) -{ - /* decoder timing evaluation */ - U32 const Q = (U32)(cSrcSize * 16 / dstSize); /* Q < 16 since dstSize > cSrcSize */ - U32 const D256 = (U32)(dstSize >> 8); - U32 const DTime0 = algoTime[Q][0].tableTime + (algoTime[Q][0].decode256Time * D256); - U32 DTime1 = algoTime[Q][1].tableTime + (algoTime[Q][1].decode256Time * D256); - DTime1 += DTime1 >> 3; /* advantage to algorithm using less memory, for cache eviction */ - - return DTime1 < DTime0; -} - -typedef size_t (*decompressionAlgo)(void *dst, size_t dstSize, const void *cSrc, size_t cSrcSize); - -size_t HUF_decompress4X_DCtx_wksp(HUF_DTable *dctx, void *dst, size_t dstSize, const void *cSrc, size_t cSrcSize, void *workspace, size_t workspaceSize) -{ - /* validation checks */ - if (dstSize == 0) - return ERROR(dstSize_tooSmall); - if (cSrcSize > dstSize) - return ERROR(corruption_detected); /* invalid */ - if (cSrcSize == dstSize) { - memcpy(dst, cSrc, dstSize); - return dstSize; - } /* not compressed */ - if (cSrcSize == 1) { - memset(dst, *(const BYTE *)cSrc, dstSize); - return dstSize; - } /* RLE */ - - { - U32 const algoNb = HUF_selectDecoder(dstSize, cSrcSize); - return algoNb ? HUF_decompress4X4_DCtx_wksp(dctx, dst, dstSize, cSrc, cSrcSize, workspace, workspaceSize) - : HUF_decompress4X2_DCtx_wksp(dctx, dst, dstSize, cSrc, cSrcSize, workspace, workspaceSize); - } -} - -size_t HUF_decompress4X_hufOnly_wksp(HUF_DTable *dctx, void *dst, size_t dstSize, const void *cSrc, size_t cSrcSize, void *workspace, size_t workspaceSize) -{ - /* validation checks */ - if (dstSize == 0) - return ERROR(dstSize_tooSmall); - if ((cSrcSize >= dstSize) || (cSrcSize <= 1)) - return ERROR(corruption_detected); /* invalid */ - - { - U32 const algoNb = HUF_selectDecoder(dstSize, cSrcSize); - return algoNb ? HUF_decompress4X4_DCtx_wksp(dctx, dst, dstSize, cSrc, cSrcSize, workspace, workspaceSize) - : HUF_decompress4X2_DCtx_wksp(dctx, dst, dstSize, cSrc, cSrcSize, workspace, workspaceSize); - } -} - -size_t HUF_decompress1X_DCtx_wksp(HUF_DTable *dctx, void *dst, size_t dstSize, const void *cSrc, size_t cSrcSize, void *workspace, size_t workspaceSize) -{ - /* validation checks */ - if (dstSize == 0) - return ERROR(dstSize_tooSmall); - if (cSrcSize > dstSize) - return ERROR(corruption_detected); /* invalid */ - if (cSrcSize == dstSize) { - memcpy(dst, cSrc, dstSize); - return dstSize; - } /* not compressed */ - if (cSrcSize == 1) { - memset(dst, *(const BYTE *)cSrc, dstSize); - return dstSize; - } /* RLE */ - - { - U32 const algoNb = HUF_selectDecoder(dstSize, cSrcSize); - return algoNb ? HUF_decompress1X4_DCtx_wksp(dctx, dst, dstSize, cSrc, cSrcSize, workspace, workspaceSize) - : HUF_decompress1X2_DCtx_wksp(dctx, dst, dstSize, cSrc, cSrcSize, workspace, workspaceSize); - } -} diff --git a/lib/zstd/mem.h b/lib/zstd/mem.h deleted file mode 100644 index 93d7a2c377fe..000000000000 --- a/lib/zstd/mem.h +++ /dev/null @@ -1,151 +0,0 @@ -/** - * Copyright (c) 2016-present, Yann Collet, Facebook, Inc. - * All rights reserved. - * - * This source code is licensed under the BSD-style license found in the - * LICENSE file in the root directory of https://github.com/facebook/zstd. - * An additional grant of patent rights can be found in the PATENTS file in the - * same directory. - * - * This program is free software; you can redistribute it and/or modify it under - * the terms of the GNU General Public License version 2 as published by the - * Free Software Foundation. This program is dual-licensed; you may select - * either version 2 of the GNU General Public License ("GPL") or BSD license - * ("BSD"). - */ - -#ifndef MEM_H_MODULE -#define MEM_H_MODULE - -/*-**************************************** -* Dependencies -******************************************/ -#include -#include /* memcpy */ -#include /* size_t, ptrdiff_t */ - -/*-**************************************** -* Compiler specifics -******************************************/ -#define ZSTD_STATIC static inline - -/*-************************************************************** -* Basic Types -*****************************************************************/ -typedef uint8_t BYTE; -typedef uint16_t U16; -typedef int16_t S16; -typedef uint32_t U32; -typedef int32_t S32; -typedef uint64_t U64; -typedef int64_t S64; -typedef ptrdiff_t iPtrDiff; -typedef uintptr_t uPtrDiff; - -/*-************************************************************** -* Memory I/O -*****************************************************************/ -ZSTD_STATIC unsigned ZSTD_32bits(void) { return sizeof(size_t) == 4; } -ZSTD_STATIC unsigned ZSTD_64bits(void) { return sizeof(size_t) == 8; } - -#if defined(__LITTLE_ENDIAN) -#define ZSTD_LITTLE_ENDIAN 1 -#else -#define ZSTD_LITTLE_ENDIAN 0 -#endif - -ZSTD_STATIC unsigned ZSTD_isLittleEndian(void) { return ZSTD_LITTLE_ENDIAN; } - -ZSTD_STATIC U16 ZSTD_read16(const void *memPtr) { return get_unaligned((const U16 *)memPtr); } - -ZSTD_STATIC U32 ZSTD_read32(const void *memPtr) { return get_unaligned((const U32 *)memPtr); } - -ZSTD_STATIC U64 ZSTD_read64(const void *memPtr) { return get_unaligned((const U64 *)memPtr); } - -ZSTD_STATIC size_t ZSTD_readST(const void *memPtr) { return get_unaligned((const size_t *)memPtr); } - -ZSTD_STATIC void ZSTD_write16(void *memPtr, U16 value) { put_unaligned(value, (U16 *)memPtr); } - -ZSTD_STATIC void ZSTD_write32(void *memPtr, U32 value) { put_unaligned(value, (U32 *)memPtr); } - -ZSTD_STATIC void ZSTD_write64(void *memPtr, U64 value) { put_unaligned(value, (U64 *)memPtr); } - -/*=== Little endian r/w ===*/ - -ZSTD_STATIC U16 ZSTD_readLE16(const void *memPtr) { return get_unaligned_le16(memPtr); } - -ZSTD_STATIC void ZSTD_writeLE16(void *memPtr, U16 val) { put_unaligned_le16(val, memPtr); } - -ZSTD_STATIC U32 ZSTD_readLE24(const void *memPtr) { return ZSTD_readLE16(memPtr) + (((const BYTE *)memPtr)[2] << 16); } - -ZSTD_STATIC void ZSTD_writeLE24(void *memPtr, U32 val) -{ - ZSTD_writeLE16(memPtr, (U16)val); - ((BYTE *)memPtr)[2] = (BYTE)(val >> 16); -} - -ZSTD_STATIC U32 ZSTD_readLE32(const void *memPtr) { return get_unaligned_le32(memPtr); } - -ZSTD_STATIC void ZSTD_writeLE32(void *memPtr, U32 val32) { put_unaligned_le32(val32, memPtr); } - -ZSTD_STATIC U64 ZSTD_readLE64(const void *memPtr) { return get_unaligned_le64(memPtr); } - -ZSTD_STATIC void ZSTD_writeLE64(void *memPtr, U64 val64) { put_unaligned_le64(val64, memPtr); } - -ZSTD_STATIC size_t ZSTD_readLEST(const void *memPtr) -{ - if (ZSTD_32bits()) - return (size_t)ZSTD_readLE32(memPtr); - else - return (size_t)ZSTD_readLE64(memPtr); -} - -ZSTD_STATIC void ZSTD_writeLEST(void *memPtr, size_t val) -{ - if (ZSTD_32bits()) - ZSTD_writeLE32(memPtr, (U32)val); - else - ZSTD_writeLE64(memPtr, (U64)val); -} - -/*=== Big endian r/w ===*/ - -ZSTD_STATIC U32 ZSTD_readBE32(const void *memPtr) { return get_unaligned_be32(memPtr); } - -ZSTD_STATIC void ZSTD_writeBE32(void *memPtr, U32 val32) { put_unaligned_be32(val32, memPtr); } - -ZSTD_STATIC U64 ZSTD_readBE64(const void *memPtr) { return get_unaligned_be64(memPtr); } - -ZSTD_STATIC void ZSTD_writeBE64(void *memPtr, U64 val64) { put_unaligned_be64(val64, memPtr); } - -ZSTD_STATIC size_t ZSTD_readBEST(const void *memPtr) -{ - if (ZSTD_32bits()) - return (size_t)ZSTD_readBE32(memPtr); - else - return (size_t)ZSTD_readBE64(memPtr); -} - -ZSTD_STATIC void ZSTD_writeBEST(void *memPtr, size_t val) -{ - if (ZSTD_32bits()) - ZSTD_writeBE32(memPtr, (U32)val); - else - ZSTD_writeBE64(memPtr, (U64)val); -} - -/* function safe only for comparisons */ -ZSTD_STATIC U32 ZSTD_readMINMATCH(const void *memPtr, U32 length) -{ - switch (length) { - default: - case 4: return ZSTD_read32(memPtr); - case 3: - if (ZSTD_isLittleEndian()) - return ZSTD_read32(memPtr) << 8; - else - return ZSTD_read32(memPtr) >> 8; - } -} - -#endif /* MEM_H_MODULE */ diff --git a/lib/zstd/zstd_common.c b/lib/zstd/zstd_common.c deleted file mode 100644 index a282624ee155..000000000000 --- a/lib/zstd/zstd_common.c +++ /dev/null @@ -1,75 +0,0 @@ -/** - * Copyright (c) 2016-present, Yann Collet, Facebook, Inc. - * All rights reserved. - * - * This source code is licensed under the BSD-style license found in the - * LICENSE file in the root directory of https://github.com/facebook/zstd. - * An additional grant of patent rights can be found in the PATENTS file in the - * same directory. - * - * This program is free software; you can redistribute it and/or modify it under - * the terms of the GNU General Public License version 2 as published by the - * Free Software Foundation. This program is dual-licensed; you may select - * either version 2 of the GNU General Public License ("GPL") or BSD license - * ("BSD"). - */ - -/*-************************************* -* Dependencies -***************************************/ -#include "error_private.h" -#include "zstd_internal.h" /* declaration of ZSTD_isError, ZSTD_getErrorName, ZSTD_getErrorCode, ZSTD_getErrorString, ZSTD_versionNumber */ -#include - -/*=************************************************************** -* Custom allocator -****************************************************************/ - -#define stack_push(stack, size) \ - ({ \ - void *const ptr = ZSTD_PTR_ALIGN((stack)->ptr); \ - (stack)->ptr = (char *)ptr + (size); \ - (stack)->ptr <= (stack)->end ? ptr : NULL; \ - }) - -ZSTD_customMem ZSTD_initStack(void *workspace, size_t workspaceSize) -{ - ZSTD_customMem stackMem = {ZSTD_stackAlloc, ZSTD_stackFree, workspace}; - ZSTD_stack *stack = (ZSTD_stack *)workspace; - /* Verify preconditions */ - if (!workspace || workspaceSize < sizeof(ZSTD_stack) || workspace != ZSTD_PTR_ALIGN(workspace)) { - ZSTD_customMem error = {NULL, NULL, NULL}; - return error; - } - /* Initialize the stack */ - stack->ptr = workspace; - stack->end = (char *)workspace + workspaceSize; - stack_push(stack, sizeof(ZSTD_stack)); - return stackMem; -} - -void *ZSTD_stackAllocAll(void *opaque, size_t *size) -{ - ZSTD_stack *stack = (ZSTD_stack *)opaque; - *size = (BYTE const *)stack->end - (BYTE *)ZSTD_PTR_ALIGN(stack->ptr); - return stack_push(stack, *size); -} - -void *ZSTD_stackAlloc(void *opaque, size_t size) -{ - ZSTD_stack *stack = (ZSTD_stack *)opaque; - return stack_push(stack, size); -} -void ZSTD_stackFree(void *opaque, void *address) -{ - (void)opaque; - (void)address; -} - -void *ZSTD_malloc(size_t size, ZSTD_customMem customMem) { return customMem.customAlloc(customMem.opaque, size); } - -void ZSTD_free(void *ptr, ZSTD_customMem customMem) -{ - if (ptr != NULL) - customMem.customFree(customMem.opaque, ptr); -} diff --git a/lib/zstd/zstd_compress_module.c b/lib/zstd/zstd_compress_module.c new file mode 100644 index 000000000000..65548a4bb934 --- /dev/null +++ b/lib/zstd/zstd_compress_module.c @@ -0,0 +1,160 @@ +// SPDX-License-Identifier: GPL-2.0+ OR BSD-3-Clause +/* + * Copyright (c) Facebook, Inc. + * All rights reserved. + * + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). + * You may select, at your option, one of the above-listed licenses. + */ + +#include +#include +#include +#include + +#include "common/zstd_deps.h" +#include "common/zstd_internal.h" + +#define ZSTD_FORWARD_IF_ERR(ret) \ + do { \ + size_t const __ret = (ret); \ + if (ZSTD_isError(__ret)) \ + return __ret; \ + } while (0) + +static size_t zstd_cctx_init(zstd_cctx *cctx, const zstd_parameters *parameters, + unsigned long long pledged_src_size) +{ + ZSTD_FORWARD_IF_ERR(ZSTD_CCtx_reset( + cctx, ZSTD_reset_session_and_parameters)); + ZSTD_FORWARD_IF_ERR(ZSTD_CCtx_setPledgedSrcSize( + cctx, pledged_src_size)); + ZSTD_FORWARD_IF_ERR(ZSTD_CCtx_setParameter( + cctx, ZSTD_c_windowLog, parameters->cParams.windowLog)); + ZSTD_FORWARD_IF_ERR(ZSTD_CCtx_setParameter( + cctx, ZSTD_c_hashLog, parameters->cParams.hashLog)); + ZSTD_FORWARD_IF_ERR(ZSTD_CCtx_setParameter( + cctx, ZSTD_c_chainLog, parameters->cParams.chainLog)); + ZSTD_FORWARD_IF_ERR(ZSTD_CCtx_setParameter( + cctx, ZSTD_c_searchLog, parameters->cParams.searchLog)); + ZSTD_FORWARD_IF_ERR(ZSTD_CCtx_setParameter( + cctx, ZSTD_c_minMatch, parameters->cParams.minMatch)); + ZSTD_FORWARD_IF_ERR(ZSTD_CCtx_setParameter( + cctx, ZSTD_c_targetLength, parameters->cParams.targetLength)); + ZSTD_FORWARD_IF_ERR(ZSTD_CCtx_setParameter( + cctx, ZSTD_c_strategy, parameters->cParams.strategy)); + ZSTD_FORWARD_IF_ERR(ZSTD_CCtx_setParameter( + cctx, ZSTD_c_contentSizeFlag, parameters->fParams.contentSizeFlag)); + ZSTD_FORWARD_IF_ERR(ZSTD_CCtx_setParameter( + cctx, ZSTD_c_checksumFlag, parameters->fParams.checksumFlag)); + ZSTD_FORWARD_IF_ERR(ZSTD_CCtx_setParameter( + cctx, ZSTD_c_dictIDFlag, !parameters->fParams.noDictIDFlag)); + return 0; +} + +int zstd_min_clevel(void) +{ + return ZSTD_minCLevel(); +} +EXPORT_SYMBOL(zstd_min_clevel); + +int zstd_max_clevel(void) +{ + return ZSTD_maxCLevel(); +} +EXPORT_SYMBOL(zstd_max_clevel); + +size_t zstd_compress_bound(size_t src_size) +{ + return ZSTD_compressBound(src_size); +} +EXPORT_SYMBOL(zstd_compress_bound); + +zstd_parameters zstd_get_params(int level, + unsigned long long estimated_src_size) +{ + return ZSTD_getParams(level, estimated_src_size, 0); +} +EXPORT_SYMBOL(zstd_get_params); + +size_t zstd_cctx_workspace_bound(const zstd_compression_parameters *cparams) +{ + return ZSTD_estimateCCtxSize_usingCParams(*cparams); +} +EXPORT_SYMBOL(zstd_cctx_workspace_bound); + +zstd_cctx *zstd_init_cctx(void *workspace, size_t workspace_size) +{ + if (workspace == NULL) + return NULL; + return ZSTD_initStaticCCtx(workspace, workspace_size); +} +EXPORT_SYMBOL(zstd_init_cctx); + +size_t zstd_compress_cctx(zstd_cctx *cctx, void *dst, size_t dst_capacity, + const void *src, size_t src_size, const zstd_parameters *parameters) +{ + ZSTD_FORWARD_IF_ERR(zstd_cctx_init(cctx, parameters, src_size)); + return ZSTD_compress2(cctx, dst, dst_capacity, src, src_size); +} +EXPORT_SYMBOL(zstd_compress_cctx); + +size_t zstd_cstream_workspace_bound(const zstd_compression_parameters *cparams) +{ + return ZSTD_estimateCStreamSize_usingCParams(*cparams); +} +EXPORT_SYMBOL(zstd_cstream_workspace_bound); + +zstd_cstream *zstd_init_cstream(const zstd_parameters *parameters, + unsigned long long pledged_src_size, void *workspace, size_t workspace_size) +{ + zstd_cstream *cstream; + + if (workspace == NULL) + return NULL; + + cstream = ZSTD_initStaticCStream(workspace, workspace_size); + if (cstream == NULL) + return NULL; + + /* 0 means unknown in linux zstd API but means 0 in new zstd API */ + if (pledged_src_size == 0) + pledged_src_size = ZSTD_CONTENTSIZE_UNKNOWN; + + if (ZSTD_isError(zstd_cctx_init(cstream, parameters, pledged_src_size))) + return NULL; + + return cstream; +} +EXPORT_SYMBOL(zstd_init_cstream); + +size_t zstd_reset_cstream(zstd_cstream *cstream, + unsigned long long pledged_src_size) +{ + return ZSTD_resetCStream(cstream, pledged_src_size); +} +EXPORT_SYMBOL(zstd_reset_cstream); + +size_t zstd_compress_stream(zstd_cstream *cstream, zstd_out_buffer *output, + zstd_in_buffer *input) +{ + return ZSTD_compressStream(cstream, output, input); +} +EXPORT_SYMBOL(zstd_compress_stream); + +size_t zstd_flush_stream(zstd_cstream *cstream, zstd_out_buffer *output) +{ + return ZSTD_flushStream(cstream, output); +} +EXPORT_SYMBOL(zstd_flush_stream); + +size_t zstd_end_stream(zstd_cstream *cstream, zstd_out_buffer *output) +{ + return ZSTD_endStream(cstream, output); +} +EXPORT_SYMBOL(zstd_end_stream); + +MODULE_LICENSE("Dual BSD/GPL"); +MODULE_DESCRIPTION("Zstd Compressor"); diff --git a/lib/zstd/zstd_decompress_module.c b/lib/zstd/zstd_decompress_module.c new file mode 100644 index 000000000000..f4ed952ed485 --- /dev/null +++ b/lib/zstd/zstd_decompress_module.c @@ -0,0 +1,105 @@ +// SPDX-License-Identifier: GPL-2.0+ OR BSD-3-Clause +/* + * Copyright (c) Facebook, Inc. + * All rights reserved. + * + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). + * You may select, at your option, one of the above-listed licenses. + */ + +#include +#include +#include +#include + +#include "common/zstd_deps.h" + +/* Common symbols. zstd_compress must depend on zstd_decompress. */ + +unsigned int zstd_is_error(size_t code) +{ + return ZSTD_isError(code); +} +EXPORT_SYMBOL(zstd_is_error); + +zstd_error_code zstd_get_error_code(size_t code) +{ + return ZSTD_getErrorCode(code); +} +EXPORT_SYMBOL(zstd_get_error_code); + +const char *zstd_get_error_name(size_t code) +{ + return ZSTD_getErrorName(code); +} +EXPORT_SYMBOL(zstd_get_error_name); + +/* Decompression symbols. */ + +size_t zstd_dctx_workspace_bound(void) +{ + return ZSTD_estimateDCtxSize(); +} +EXPORT_SYMBOL(zstd_dctx_workspace_bound); + +zstd_dctx *zstd_init_dctx(void *workspace, size_t workspace_size) +{ + if (workspace == NULL) + return NULL; + return ZSTD_initStaticDCtx(workspace, workspace_size); +} +EXPORT_SYMBOL(zstd_init_dctx); + +size_t zstd_decompress_dctx(zstd_dctx *dctx, void *dst, size_t dst_capacity, + const void *src, size_t src_size) +{ + return ZSTD_decompressDCtx(dctx, dst, dst_capacity, src, src_size); +} +EXPORT_SYMBOL(zstd_decompress_dctx); + +size_t zstd_dstream_workspace_bound(size_t max_window_size) +{ + return ZSTD_estimateDStreamSize(max_window_size); +} +EXPORT_SYMBOL(zstd_dstream_workspace_bound); + +zstd_dstream *zstd_init_dstream(size_t max_window_size, void *workspace, + size_t workspace_size) +{ + if (workspace == NULL) + return NULL; + (void)max_window_size; + return ZSTD_initStaticDStream(workspace, workspace_size); +} +EXPORT_SYMBOL(zstd_init_dstream); + +size_t zstd_reset_dstream(zstd_dstream *dstream) +{ + return ZSTD_resetDStream(dstream); +} +EXPORT_SYMBOL(zstd_reset_dstream); + +size_t zstd_decompress_stream(zstd_dstream *dstream, zstd_out_buffer *output, + zstd_in_buffer *input) +{ + return ZSTD_decompressStream(dstream, output, input); +} +EXPORT_SYMBOL(zstd_decompress_stream); + +size_t zstd_find_frame_compressed_size(const void *src, size_t src_size) +{ + return ZSTD_findFrameCompressedSize(src, src_size); +} +EXPORT_SYMBOL(zstd_find_frame_compressed_size); + +size_t zstd_get_frame_header(zstd_frame_header *header, const void *src, + size_t src_size) +{ + return ZSTD_getFrameHeader(header, src, src_size); +} +EXPORT_SYMBOL(zstd_get_frame_header); + +MODULE_LICENSE("Dual BSD/GPL"); +MODULE_DESCRIPTION("Zstd Decompressor"); diff --git a/lib/zstd/zstd_internal.h b/lib/zstd/zstd_internal.h deleted file mode 100644 index dac753397f86..000000000000 --- a/lib/zstd/zstd_internal.h +++ /dev/null @@ -1,273 +0,0 @@ -/** - * Copyright (c) 2016-present, Yann Collet, Facebook, Inc. - * All rights reserved. - * - * This source code is licensed under the BSD-style license found in the - * LICENSE file in the root directory of https://github.com/facebook/zstd. - * An additional grant of patent rights can be found in the PATENTS file in the - * same directory. - * - * This program is free software; you can redistribute it and/or modify it under - * the terms of the GNU General Public License version 2 as published by the - * Free Software Foundation. This program is dual-licensed; you may select - * either version 2 of the GNU General Public License ("GPL") or BSD license - * ("BSD"). - */ - -#ifndef ZSTD_CCOMMON_H_MODULE -#define ZSTD_CCOMMON_H_MODULE - -/*-******************************************************* -* Compiler specifics -*********************************************************/ -#define FORCE_INLINE static __always_inline -#define FORCE_NOINLINE static noinline - -/*-************************************* -* Dependencies -***************************************/ -#include "error_private.h" -#include "mem.h" -#include -#include -#include -#include - -/*-************************************* -* shared macros -***************************************/ -#define MIN(a, b) ((a) < (b) ? (a) : (b)) -#define MAX(a, b) ((a) > (b) ? (a) : (b)) -#define CHECK_F(f) \ - { \ - size_t const errcod = f; \ - if (ERR_isError(errcod)) \ - return errcod; \ - } /* check and Forward error code */ -#define CHECK_E(f, e) \ - { \ - size_t const errcod = f; \ - if (ERR_isError(errcod)) \ - return ERROR(e); \ - } /* check and send Error code */ -#define ZSTD_STATIC_ASSERT(c) \ - { \ - enum { ZSTD_static_assert = 1 / (int)(!!(c)) }; \ - } - -/*-************************************* -* Common constants -***************************************/ -#define ZSTD_OPT_NUM (1 << 12) -#define ZSTD_DICT_MAGIC 0xEC30A437 /* v0.7+ */ - -#define ZSTD_REP_NUM 3 /* number of repcodes */ -#define ZSTD_REP_CHECK (ZSTD_REP_NUM) /* number of repcodes to check by the optimal parser */ -#define ZSTD_REP_MOVE (ZSTD_REP_NUM - 1) -#define ZSTD_REP_MOVE_OPT (ZSTD_REP_NUM) -static const U32 repStartValue[ZSTD_REP_NUM] = {1, 4, 8}; - -#define KB *(1 << 10) -#define MB *(1 << 20) -#define GB *(1U << 30) - -#define BIT7 128 -#define BIT6 64 -#define BIT5 32 -#define BIT4 16 -#define BIT1 2 -#define BIT0 1 - -#define ZSTD_WINDOWLOG_ABSOLUTEMIN 10 -static const size_t ZSTD_fcs_fieldSize[4] = {0, 2, 4, 8}; -static const size_t ZSTD_did_fieldSize[4] = {0, 1, 2, 4}; - -#define ZSTD_BLOCKHEADERSIZE 3 /* C standard doesn't allow `static const` variable to be init using another `static const` variable */ -static const size_t ZSTD_blockHeaderSize = ZSTD_BLOCKHEADERSIZE; -typedef enum { bt_raw, bt_rle, bt_compressed, bt_reserved } blockType_e; - -#define MIN_SEQUENCES_SIZE 1 /* nbSeq==0 */ -#define MIN_CBLOCK_SIZE (1 /*litCSize*/ + 1 /* RLE or RAW */ + MIN_SEQUENCES_SIZE /* nbSeq==0 */) /* for a non-null block */ - -#define HufLog 12 -typedef enum { set_basic, set_rle, set_compressed, set_repeat } symbolEncodingType_e; - -#define LONGNBSEQ 0x7F00 - -#define MINMATCH 3 -#define EQUAL_READ32 4 - -#define Litbits 8 -#define MaxLit ((1 << Litbits) - 1) -#define MaxML 52 -#define MaxLL 35 -#define MaxOff 28 -#define MaxSeq MAX(MaxLL, MaxML) /* Assumption : MaxOff < MaxLL,MaxML */ -#define MLFSELog 9 -#define LLFSELog 9 -#define OffFSELog 8 - -static const U32 LL_bits[MaxLL + 1] = {0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 2, 2, 3, 3, 4, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16}; -static const S16 LL_defaultNorm[MaxLL + 1] = {4, 3, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 1, 1, 1, 2, 2, 2, 2, 2, 2, 2, 2, 2, 3, 2, 1, 1, 1, 1, 1, -1, -1, -1, -1}; -#define LL_DEFAULTNORMLOG 6 /* for static allocation */ -static const U32 LL_defaultNormLog = LL_DEFAULTNORMLOG; - -static const U32 ML_bits[MaxML + 1] = {0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, - 0, 0, 0, 0, 0, 1, 1, 1, 1, 2, 2, 3, 3, 4, 4, 5, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16}; -static const S16 ML_defaultNorm[MaxML + 1] = {1, 4, 3, 2, 2, 2, 2, 2, 2, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, - 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, -1, -1, -1, -1, -1, -1, -1}; -#define ML_DEFAULTNORMLOG 6 /* for static allocation */ -static const U32 ML_defaultNormLog = ML_DEFAULTNORMLOG; - -static const S16 OF_defaultNorm[MaxOff + 1] = {1, 1, 1, 1, 1, 1, 2, 2, 2, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, -1, -1, -1, -1, -1}; -#define OF_DEFAULTNORMLOG 5 /* for static allocation */ -static const U32 OF_defaultNormLog = OF_DEFAULTNORMLOG; - -/*-******************************************* -* Shared functions to include for inlining -*********************************************/ -ZSTD_STATIC void ZSTD_copy8(void *dst, const void *src) { - /* - * zstd relies heavily on gcc being able to analyze and inline this - * memcpy() call, since it is called in a tight loop. Preboot mode - * is compiled in freestanding mode, which stops gcc from analyzing - * memcpy(). Use __builtin_memcpy() to tell gcc to analyze this as a - * regular memcpy(). - */ - __builtin_memcpy(dst, src, 8); -} -/*! ZSTD_wildcopy() : -* custom version of memcpy(), can copy up to 7 bytes too many (8 bytes if length==0) */ -#define WILDCOPY_OVERLENGTH 8 -ZSTD_STATIC void ZSTD_wildcopy(void *dst, const void *src, ptrdiff_t length) -{ - const BYTE* ip = (const BYTE*)src; - BYTE* op = (BYTE*)dst; - BYTE* const oend = op + length; -#if defined(GCC_VERSION) && GCC_VERSION >= 70000 && GCC_VERSION < 70200 - /* - * Work around https://gcc.gnu.org/bugzilla/show_bug.cgi?id=81388. - * Avoid the bad case where the loop only runs once by handling the - * special case separately. This doesn't trigger the bug because it - * doesn't involve pointer/integer overflow. - */ - if (length <= 8) - return ZSTD_copy8(dst, src); -#endif - do { - ZSTD_copy8(op, ip); - op += 8; - ip += 8; - } while (op < oend); -} - -/*-******************************************* -* Private interfaces -*********************************************/ -typedef struct ZSTD_stats_s ZSTD_stats_t; - -typedef struct { - U32 off; - U32 len; -} ZSTD_match_t; - -typedef struct { - U32 price; - U32 off; - U32 mlen; - U32 litlen; - U32 rep[ZSTD_REP_NUM]; -} ZSTD_optimal_t; - -typedef struct seqDef_s { - U32 offset; - U16 litLength; - U16 matchLength; -} seqDef; - -typedef struct { - seqDef *sequencesStart; - seqDef *sequences; - BYTE *litStart; - BYTE *lit; - BYTE *llCode; - BYTE *mlCode; - BYTE *ofCode; - U32 longLengthID; /* 0 == no longLength; 1 == Lit.longLength; 2 == Match.longLength; */ - U32 longLengthPos; - /* opt */ - ZSTD_optimal_t *priceTable; - ZSTD_match_t *matchTable; - U32 *matchLengthFreq; - U32 *litLengthFreq; - U32 *litFreq; - U32 *offCodeFreq; - U32 matchLengthSum; - U32 matchSum; - U32 litLengthSum; - U32 litSum; - U32 offCodeSum; - U32 log2matchLengthSum; - U32 log2matchSum; - U32 log2litLengthSum; - U32 log2litSum; - U32 log2offCodeSum; - U32 factor; - U32 staticPrices; - U32 cachedPrice; - U32 cachedLitLength; - const BYTE *cachedLiterals; -} seqStore_t; - -const seqStore_t *ZSTD_getSeqStore(const ZSTD_CCtx *ctx); -void ZSTD_seqToCodes(const seqStore_t *seqStorePtr); -int ZSTD_isSkipFrame(ZSTD_DCtx *dctx); - -/*= Custom memory allocation functions */ -typedef void *(*ZSTD_allocFunction)(void *opaque, size_t size); -typedef void (*ZSTD_freeFunction)(void *opaque, void *address); -typedef struct { - ZSTD_allocFunction customAlloc; - ZSTD_freeFunction customFree; - void *opaque; -} ZSTD_customMem; - -void *ZSTD_malloc(size_t size, ZSTD_customMem customMem); -void ZSTD_free(void *ptr, ZSTD_customMem customMem); - -/*====== stack allocation ======*/ - -typedef struct { - void *ptr; - const void *end; -} ZSTD_stack; - -#define ZSTD_ALIGN(x) ALIGN(x, sizeof(size_t)) -#define ZSTD_PTR_ALIGN(p) PTR_ALIGN(p, sizeof(size_t)) - -ZSTD_customMem ZSTD_initStack(void *workspace, size_t workspaceSize); - -void *ZSTD_stackAllocAll(void *opaque, size_t *size); -void *ZSTD_stackAlloc(void *opaque, size_t size); -void ZSTD_stackFree(void *opaque, void *address); - -/*====== common function ======*/ - -ZSTD_STATIC U32 ZSTD_highbit32(U32 val) { return 31 - __builtin_clz(val); } - -/* hidden functions */ - -/* ZSTD_invalidateRepCodes() : - * ensures next compression will not use repcodes from previous block. - * Note : only works with regular variant; - * do not use with extDict variant ! */ -void ZSTD_invalidateRepCodes(ZSTD_CCtx *cctx); - -size_t ZSTD_freeCCtx(ZSTD_CCtx *cctx); -size_t ZSTD_freeDCtx(ZSTD_DCtx *dctx); -size_t ZSTD_freeCDict(ZSTD_CDict *cdict); -size_t ZSTD_freeDDict(ZSTD_DDict *cdict); -size_t ZSTD_freeCStream(ZSTD_CStream *zcs); -size_t ZSTD_freeDStream(ZSTD_DStream *zds); - -#endif /* ZSTD_CCOMMON_H_MODULE */ diff --git a/lib/zstd/zstd_opt.h b/lib/zstd/zstd_opt.h deleted file mode 100644 index 55e1b4cba808..000000000000 --- a/lib/zstd/zstd_opt.h +++ /dev/null @@ -1,1014 +0,0 @@ -/** - * Copyright (c) 2016-present, Przemyslaw Skibinski, Yann Collet, Facebook, Inc. - * All rights reserved. - * - * This source code is licensed under the BSD-style license found in the - * LICENSE file in the root directory of https://github.com/facebook/zstd. - * An additional grant of patent rights can be found in the PATENTS file in the - * same directory. - * - * This program is free software; you can redistribute it and/or modify it under - * the terms of the GNU General Public License version 2 as published by the - * Free Software Foundation. This program is dual-licensed; you may select - * either version 2 of the GNU General Public License ("GPL") or BSD license - * ("BSD"). - */ - -/* Note : this file is intended to be included within zstd_compress.c */ - -#ifndef ZSTD_OPT_H_91842398743 -#define ZSTD_OPT_H_91842398743 - -#define ZSTD_LITFREQ_ADD 2 -#define ZSTD_FREQ_DIV 4 -#define ZSTD_MAX_PRICE (1 << 30) - -/*-************************************* -* Price functions for optimal parser -***************************************/ -FORCE_INLINE void ZSTD_setLog2Prices(seqStore_t *ssPtr) -{ - ssPtr->log2matchLengthSum = ZSTD_highbit32(ssPtr->matchLengthSum + 1); - ssPtr->log2litLengthSum = ZSTD_highbit32(ssPtr->litLengthSum + 1); - ssPtr->log2litSum = ZSTD_highbit32(ssPtr->litSum + 1); - ssPtr->log2offCodeSum = ZSTD_highbit32(ssPtr->offCodeSum + 1); - ssPtr->factor = 1 + ((ssPtr->litSum >> 5) / ssPtr->litLengthSum) + ((ssPtr->litSum << 1) / (ssPtr->litSum + ssPtr->matchSum)); -} - -ZSTD_STATIC void ZSTD_rescaleFreqs(seqStore_t *ssPtr, const BYTE *src, size_t srcSize) -{ - unsigned u; - - ssPtr->cachedLiterals = NULL; - ssPtr->cachedPrice = ssPtr->cachedLitLength = 0; - ssPtr->staticPrices = 0; - - if (ssPtr->litLengthSum == 0) { - if (srcSize <= 1024) - ssPtr->staticPrices = 1; - - for (u = 0; u <= MaxLit; u++) - ssPtr->litFreq[u] = 0; - for (u = 0; u < srcSize; u++) - ssPtr->litFreq[src[u]]++; - - ssPtr->litSum = 0; - ssPtr->litLengthSum = MaxLL + 1; - ssPtr->matchLengthSum = MaxML + 1; - ssPtr->offCodeSum = (MaxOff + 1); - ssPtr->matchSum = (ZSTD_LITFREQ_ADD << Litbits); - - for (u = 0; u <= MaxLit; u++) { - ssPtr->litFreq[u] = 1 + (ssPtr->litFreq[u] >> ZSTD_FREQ_DIV); - ssPtr->litSum += ssPtr->litFreq[u]; - } - for (u = 0; u <= MaxLL; u++) - ssPtr->litLengthFreq[u] = 1; - for (u = 0; u <= MaxML; u++) - ssPtr->matchLengthFreq[u] = 1; - for (u = 0; u <= MaxOff; u++) - ssPtr->offCodeFreq[u] = 1; - } else { - ssPtr->matchLengthSum = 0; - ssPtr->litLengthSum = 0; - ssPtr->offCodeSum = 0; - ssPtr->matchSum = 0; - ssPtr->litSum = 0; - - for (u = 0; u <= MaxLit; u++) { - ssPtr->litFreq[u] = 1 + (ssPtr->litFreq[u] >> (ZSTD_FREQ_DIV + 1)); - ssPtr->litSum += ssPtr->litFreq[u]; - } - for (u = 0; u <= MaxLL; u++) { - ssPtr->litLengthFreq[u] = 1 + (ssPtr->litLengthFreq[u] >> (ZSTD_FREQ_DIV + 1)); - ssPtr->litLengthSum += ssPtr->litLengthFreq[u]; - } - for (u = 0; u <= MaxML; u++) { - ssPtr->matchLengthFreq[u] = 1 + (ssPtr->matchLengthFreq[u] >> ZSTD_FREQ_DIV); - ssPtr->matchLengthSum += ssPtr->matchLengthFreq[u]; - ssPtr->matchSum += ssPtr->matchLengthFreq[u] * (u + 3); - } - ssPtr->matchSum *= ZSTD_LITFREQ_ADD; - for (u = 0; u <= MaxOff; u++) { - ssPtr->offCodeFreq[u] = 1 + (ssPtr->offCodeFreq[u] >> ZSTD_FREQ_DIV); - ssPtr->offCodeSum += ssPtr->offCodeFreq[u]; - } - } - - ZSTD_setLog2Prices(ssPtr); -} - -FORCE_INLINE U32 ZSTD_getLiteralPrice(seqStore_t *ssPtr, U32 litLength, const BYTE *literals) -{ - U32 price, u; - - if (ssPtr->staticPrices) - return ZSTD_highbit32((U32)litLength + 1) + (litLength * 6); - - if (litLength == 0) - return ssPtr->log2litLengthSum - ZSTD_highbit32(ssPtr->litLengthFreq[0] + 1); - - /* literals */ - if (ssPtr->cachedLiterals == literals) { - U32 const additional = litLength - ssPtr->cachedLitLength; - const BYTE *literals2 = ssPtr->cachedLiterals + ssPtr->cachedLitLength; - price = ssPtr->cachedPrice + additional * ssPtr->log2litSum; - for (u = 0; u < additional; u++) - price -= ZSTD_highbit32(ssPtr->litFreq[literals2[u]] + 1); - ssPtr->cachedPrice = price; - ssPtr->cachedLitLength = litLength; - } else { - price = litLength * ssPtr->log2litSum; - for (u = 0; u < litLength; u++) - price -= ZSTD_highbit32(ssPtr->litFreq[literals[u]] + 1); - - if (litLength >= 12) { - ssPtr->cachedLiterals = literals; - ssPtr->cachedPrice = price; - ssPtr->cachedLitLength = litLength; - } - } - - /* literal Length */ - { - const BYTE LL_deltaCode = 19; - const BYTE llCode = (litLength > 63) ? (BYTE)ZSTD_highbit32(litLength) + LL_deltaCode : LL_Code[litLength]; - price += LL_bits[llCode] + ssPtr->log2litLengthSum - ZSTD_highbit32(ssPtr->litLengthFreq[llCode] + 1); - } - - return price; -} - -FORCE_INLINE U32 ZSTD_getPrice(seqStore_t *seqStorePtr, U32 litLength, const BYTE *literals, U32 offset, U32 matchLength, const int ultra) -{ - /* offset */ - U32 price; - BYTE const offCode = (BYTE)ZSTD_highbit32(offset + 1); - - if (seqStorePtr->staticPrices) - return ZSTD_getLiteralPrice(seqStorePtr, litLength, literals) + ZSTD_highbit32((U32)matchLength + 1) + 16 + offCode; - - price = offCode + seqStorePtr->log2offCodeSum - ZSTD_highbit32(seqStorePtr->offCodeFreq[offCode] + 1); - if (!ultra && offCode >= 20) - price += (offCode - 19) * 2; - - /* match Length */ - { - const BYTE ML_deltaCode = 36; - const BYTE mlCode = (matchLength > 127) ? (BYTE)ZSTD_highbit32(matchLength) + ML_deltaCode : ML_Code[matchLength]; - price += ML_bits[mlCode] + seqStorePtr->log2matchLengthSum - ZSTD_highbit32(seqStorePtr->matchLengthFreq[mlCode] + 1); - } - - return price + ZSTD_getLiteralPrice(seqStorePtr, litLength, literals) + seqStorePtr->factor; -} - -ZSTD_STATIC void ZSTD_updatePrice(seqStore_t *seqStorePtr, U32 litLength, const BYTE *literals, U32 offset, U32 matchLength) -{ - U32 u; - - /* literals */ - seqStorePtr->litSum += litLength * ZSTD_LITFREQ_ADD; - for (u = 0; u < litLength; u++) - seqStorePtr->litFreq[literals[u]] += ZSTD_LITFREQ_ADD; - - /* literal Length */ - { - const BYTE LL_deltaCode = 19; - const BYTE llCode = (litLength > 63) ? (BYTE)ZSTD_highbit32(litLength) + LL_deltaCode : LL_Code[litLength]; - seqStorePtr->litLengthFreq[llCode]++; - seqStorePtr->litLengthSum++; - } - - /* match offset */ - { - BYTE const offCode = (BYTE)ZSTD_highbit32(offset + 1); - seqStorePtr->offCodeSum++; - seqStorePtr->offCodeFreq[offCode]++; - } - - /* match Length */ - { - const BYTE ML_deltaCode = 36; - const BYTE mlCode = (matchLength > 127) ? (BYTE)ZSTD_highbit32(matchLength) + ML_deltaCode : ML_Code[matchLength]; - seqStorePtr->matchLengthFreq[mlCode]++; - seqStorePtr->matchLengthSum++; - } - - ZSTD_setLog2Prices(seqStorePtr); -} - -#define SET_PRICE(pos, mlen_, offset_, litlen_, price_) \ - { \ - while (last_pos < pos) { \ - opt[last_pos + 1].price = ZSTD_MAX_PRICE; \ - last_pos++; \ - } \ - opt[pos].mlen = mlen_; \ - opt[pos].off = offset_; \ - opt[pos].litlen = litlen_; \ - opt[pos].price = price_; \ - } - -/* Update hashTable3 up to ip (excluded) - Assumption : always within prefix (i.e. not within extDict) */ -FORCE_INLINE -U32 ZSTD_insertAndFindFirstIndexHash3(ZSTD_CCtx *zc, const BYTE *ip) -{ - U32 *const hashTable3 = zc->hashTable3; - U32 const hashLog3 = zc->hashLog3; - const BYTE *const base = zc->base; - U32 idx = zc->nextToUpdate3; - const U32 target = zc->nextToUpdate3 = (U32)(ip - base); - const size_t hash3 = ZSTD_hash3Ptr(ip, hashLog3); - - while (idx < target) { - hashTable3[ZSTD_hash3Ptr(base + idx, hashLog3)] = idx; - idx++; - } - - return hashTable3[hash3]; -} - -/*-************************************* -* Binary Tree search -***************************************/ -static U32 ZSTD_insertBtAndGetAllMatches(ZSTD_CCtx *zc, const BYTE *const ip, const BYTE *const iLimit, U32 nbCompares, const U32 mls, U32 extDict, - ZSTD_match_t *matches, const U32 minMatchLen) -{ - const BYTE *const base = zc->base; - const U32 curr = (U32)(ip - base); - const U32 hashLog = zc->params.cParams.hashLog; - const size_t h = ZSTD_hashPtr(ip, hashLog, mls); - U32 *const hashTable = zc->hashTable; - U32 matchIndex = hashTable[h]; - U32 *const bt = zc->chainTable; - const U32 btLog = zc->params.cParams.chainLog - 1; - const U32 btMask = (1U << btLog) - 1; - size_t commonLengthSmaller = 0, commonLengthLarger = 0; - const BYTE *const dictBase = zc->dictBase; - const U32 dictLimit = zc->dictLimit; - const BYTE *const dictEnd = dictBase + dictLimit; - const BYTE *const prefixStart = base + dictLimit; - const U32 btLow = btMask >= curr ? 0 : curr - btMask; - const U32 windowLow = zc->lowLimit; - U32 *smallerPtr = bt + 2 * (curr & btMask); - U32 *largerPtr = bt + 2 * (curr & btMask) + 1; - U32 matchEndIdx = curr + 8; - U32 dummy32; /* to be nullified at the end */ - U32 mnum = 0; - - const U32 minMatch = (mls == 3) ? 3 : 4; - size_t bestLength = minMatchLen - 1; - - if (minMatch == 3) { /* HC3 match finder */ - U32 const matchIndex3 = ZSTD_insertAndFindFirstIndexHash3(zc, ip); - if (matchIndex3 > windowLow && (curr - matchIndex3 < (1 << 18))) { - const BYTE *match; - size_t currMl = 0; - if ((!extDict) || matchIndex3 >= dictLimit) { - match = base + matchIndex3; - if (match[bestLength] == ip[bestLength]) - currMl = ZSTD_count(ip, match, iLimit); - } else { - match = dictBase + matchIndex3; - if (ZSTD_readMINMATCH(match, MINMATCH) == - ZSTD_readMINMATCH(ip, MINMATCH)) /* assumption : matchIndex3 <= dictLimit-4 (by table construction) */ - currMl = ZSTD_count_2segments(ip + MINMATCH, match + MINMATCH, iLimit, dictEnd, prefixStart) + MINMATCH; - } - - /* save best solution */ - if (currMl > bestLength) { - bestLength = currMl; - matches[mnum].off = ZSTD_REP_MOVE_OPT + curr - matchIndex3; - matches[mnum].len = (U32)currMl; - mnum++; - if (currMl > ZSTD_OPT_NUM) - goto update; - if (ip + currMl == iLimit) - goto update; /* best possible, and avoid read overflow*/ - } - } - } - - hashTable[h] = curr; /* Update Hash Table */ - - while (nbCompares-- && (matchIndex > windowLow)) { - U32 *nextPtr = bt + 2 * (matchIndex & btMask); - size_t matchLength = MIN(commonLengthSmaller, commonLengthLarger); /* guaranteed minimum nb of common bytes */ - const BYTE *match; - - if ((!extDict) || (matchIndex + matchLength >= dictLimit)) { - match = base + matchIndex; - if (match[matchLength] == ip[matchLength]) { - matchLength += ZSTD_count(ip + matchLength + 1, match + matchLength + 1, iLimit) + 1; - } - } else { - match = dictBase + matchIndex; - matchLength += ZSTD_count_2segments(ip + matchLength, match + matchLength, iLimit, dictEnd, prefixStart); - if (matchIndex + matchLength >= dictLimit) - match = base + matchIndex; /* to prepare for next usage of match[matchLength] */ - } - - if (matchLength > bestLength) { - if (matchLength > matchEndIdx - matchIndex) - matchEndIdx = matchIndex + (U32)matchLength; - bestLength = matchLength; - matches[mnum].off = ZSTD_REP_MOVE_OPT + curr - matchIndex; - matches[mnum].len = (U32)matchLength; - mnum++; - if (matchLength > ZSTD_OPT_NUM) - break; - if (ip + matchLength == iLimit) /* equal : no way to know if inf or sup */ - break; /* drop, to guarantee consistency (miss a little bit of compression) */ - } - - if (match[matchLength] < ip[matchLength]) { - /* match is smaller than curr */ - *smallerPtr = matchIndex; /* update smaller idx */ - commonLengthSmaller = matchLength; /* all smaller will now have at least this guaranteed common length */ - if (matchIndex <= btLow) { - smallerPtr = &dummy32; - break; - } /* beyond tree size, stop the search */ - smallerPtr = nextPtr + 1; /* new "smaller" => larger of match */ - matchIndex = nextPtr[1]; /* new matchIndex larger than previous (closer to curr) */ - } else { - /* match is larger than curr */ - *largerPtr = matchIndex; - commonLengthLarger = matchLength; - if (matchIndex <= btLow) { - largerPtr = &dummy32; - break; - } /* beyond tree size, stop the search */ - largerPtr = nextPtr; - matchIndex = nextPtr[0]; - } - } - - *smallerPtr = *largerPtr = 0; - -update: - zc->nextToUpdate = (matchEndIdx > curr + 8) ? matchEndIdx - 8 : curr + 1; - return mnum; -} - -/** Tree updater, providing best match */ -static U32 ZSTD_BtGetAllMatches(ZSTD_CCtx *zc, const BYTE *const ip, const BYTE *const iLimit, const U32 maxNbAttempts, const U32 mls, ZSTD_match_t *matches, - const U32 minMatchLen) -{ - if (ip < zc->base + zc->nextToUpdate) - return 0; /* skipped area */ - ZSTD_updateTree(zc, ip, iLimit, maxNbAttempts, mls); - return ZSTD_insertBtAndGetAllMatches(zc, ip, iLimit, maxNbAttempts, mls, 0, matches, minMatchLen); -} - -static U32 ZSTD_BtGetAllMatches_selectMLS(ZSTD_CCtx *zc, /* Index table will be updated */ - const BYTE *ip, const BYTE *const iHighLimit, const U32 maxNbAttempts, const U32 matchLengthSearch, - ZSTD_match_t *matches, const U32 minMatchLen) -{ - switch (matchLengthSearch) { - case 3: return ZSTD_BtGetAllMatches(zc, ip, iHighLimit, maxNbAttempts, 3, matches, minMatchLen); - default: - case 4: return ZSTD_BtGetAllMatches(zc, ip, iHighLimit, maxNbAttempts, 4, matches, minMatchLen); - case 5: return ZSTD_BtGetAllMatches(zc, ip, iHighLimit, maxNbAttempts, 5, matches, minMatchLen); - case 7: - case 6: return ZSTD_BtGetAllMatches(zc, ip, iHighLimit, maxNbAttempts, 6, matches, minMatchLen); - } -} - -/** Tree updater, providing best match */ -static U32 ZSTD_BtGetAllMatches_extDict(ZSTD_CCtx *zc, const BYTE *const ip, const BYTE *const iLimit, const U32 maxNbAttempts, const U32 mls, - ZSTD_match_t *matches, const U32 minMatchLen) -{ - if (ip < zc->base + zc->nextToUpdate) - return 0; /* skipped area */ - ZSTD_updateTree_extDict(zc, ip, iLimit, maxNbAttempts, mls); - return ZSTD_insertBtAndGetAllMatches(zc, ip, iLimit, maxNbAttempts, mls, 1, matches, minMatchLen); -} - -static U32 ZSTD_BtGetAllMatches_selectMLS_extDict(ZSTD_CCtx *zc, /* Index table will be updated */ - const BYTE *ip, const BYTE *const iHighLimit, const U32 maxNbAttempts, const U32 matchLengthSearch, - ZSTD_match_t *matches, const U32 minMatchLen) -{ - switch (matchLengthSearch) { - case 3: return ZSTD_BtGetAllMatches_extDict(zc, ip, iHighLimit, maxNbAttempts, 3, matches, minMatchLen); - default: - case 4: return ZSTD_BtGetAllMatches_extDict(zc, ip, iHighLimit, maxNbAttempts, 4, matches, minMatchLen); - case 5: return ZSTD_BtGetAllMatches_extDict(zc, ip, iHighLimit, maxNbAttempts, 5, matches, minMatchLen); - case 7: - case 6: return ZSTD_BtGetAllMatches_extDict(zc, ip, iHighLimit, maxNbAttempts, 6, matches, minMatchLen); - } -} - -/*-******************************* -* Optimal parser -*********************************/ -FORCE_INLINE -void ZSTD_compressBlock_opt_generic(ZSTD_CCtx *ctx, const void *src, size_t srcSize, const int ultra) -{ - seqStore_t *seqStorePtr = &(ctx->seqStore); - const BYTE *const istart = (const BYTE *)src; - const BYTE *ip = istart; - const BYTE *anchor = istart; - const BYTE *const iend = istart + srcSize; - const BYTE *const ilimit = iend - 8; - const BYTE *const base = ctx->base; - const BYTE *const prefixStart = base + ctx->dictLimit; - - const U32 maxSearches = 1U << ctx->params.cParams.searchLog; - const U32 sufficient_len = ctx->params.cParams.targetLength; - const U32 mls = ctx->params.cParams.searchLength; - const U32 minMatch = (ctx->params.cParams.searchLength == 3) ? 3 : 4; - - ZSTD_optimal_t *opt = seqStorePtr->priceTable; - ZSTD_match_t *matches = seqStorePtr->matchTable; - const BYTE *inr; - U32 offset, rep[ZSTD_REP_NUM]; - - /* init */ - ctx->nextToUpdate3 = ctx->nextToUpdate; - ZSTD_rescaleFreqs(seqStorePtr, (const BYTE *)src, srcSize); - ip += (ip == prefixStart); - { - U32 i; - for (i = 0; i < ZSTD_REP_NUM; i++) - rep[i] = ctx->rep[i]; - } - - /* Match Loop */ - while (ip < ilimit) { - U32 cur, match_num, last_pos, litlen, price; - U32 u, mlen, best_mlen, best_off, litLength; - memset(opt, 0, sizeof(ZSTD_optimal_t)); - last_pos = 0; - litlen = (U32)(ip - anchor); - - /* check repCode */ - { - U32 i, last_i = ZSTD_REP_CHECK + (ip == anchor); - for (i = (ip == anchor); i < last_i; i++) { - const S32 repCur = (i == ZSTD_REP_MOVE_OPT) ? (rep[0] - 1) : rep[i]; - if ((repCur > 0) && (repCur < (S32)(ip - prefixStart)) && - (ZSTD_readMINMATCH(ip, minMatch) == ZSTD_readMINMATCH(ip - repCur, minMatch))) { - mlen = (U32)ZSTD_count(ip + minMatch, ip + minMatch - repCur, iend) + minMatch; - if (mlen > sufficient_len || mlen >= ZSTD_OPT_NUM) { - best_mlen = mlen; - best_off = i; - cur = 0; - last_pos = 1; - goto _storeSequence; - } - best_off = i - (ip == anchor); - do { - price = ZSTD_getPrice(seqStorePtr, litlen, anchor, best_off, mlen - MINMATCH, ultra); - if (mlen > last_pos || price < opt[mlen].price) - SET_PRICE(mlen, mlen, i, litlen, price); /* note : macro modifies last_pos */ - mlen--; - } while (mlen >= minMatch); - } - } - } - - match_num = ZSTD_BtGetAllMatches_selectMLS(ctx, ip, iend, maxSearches, mls, matches, minMatch); - - if (!last_pos && !match_num) { - ip++; - continue; - } - - if (match_num && (matches[match_num - 1].len > sufficient_len || matches[match_num - 1].len >= ZSTD_OPT_NUM)) { - best_mlen = matches[match_num - 1].len; - best_off = matches[match_num - 1].off; - cur = 0; - last_pos = 1; - goto _storeSequence; - } - - /* set prices using matches at position = 0 */ - best_mlen = (last_pos) ? last_pos : minMatch; - for (u = 0; u < match_num; u++) { - mlen = (u > 0) ? matches[u - 1].len + 1 : best_mlen; - best_mlen = matches[u].len; - while (mlen <= best_mlen) { - price = ZSTD_getPrice(seqStorePtr, litlen, anchor, matches[u].off - 1, mlen - MINMATCH, ultra); - if (mlen > last_pos || price < opt[mlen].price) - SET_PRICE(mlen, mlen, matches[u].off, litlen, price); /* note : macro modifies last_pos */ - mlen++; - } - } - - if (last_pos < minMatch) { - ip++; - continue; - } - - /* initialize opt[0] */ - { - U32 i; - for (i = 0; i < ZSTD_REP_NUM; i++) - opt[0].rep[i] = rep[i]; - } - opt[0].mlen = 1; - opt[0].litlen = litlen; - - /* check further positions */ - for (cur = 1; cur <= last_pos; cur++) { - inr = ip + cur; - - if (opt[cur - 1].mlen == 1) { - litlen = opt[cur - 1].litlen + 1; - if (cur > litlen) { - price = opt[cur - litlen].price + ZSTD_getLiteralPrice(seqStorePtr, litlen, inr - litlen); - } else - price = ZSTD_getLiteralPrice(seqStorePtr, litlen, anchor); - } else { - litlen = 1; - price = opt[cur - 1].price + ZSTD_getLiteralPrice(seqStorePtr, litlen, inr - 1); - } - - if (cur > last_pos || price <= opt[cur].price) - SET_PRICE(cur, 1, 0, litlen, price); - - if (cur == last_pos) - break; - - if (inr > ilimit) /* last match must start at a minimum distance of 8 from oend */ - continue; - - mlen = opt[cur].mlen; - if (opt[cur].off > ZSTD_REP_MOVE_OPT) { - opt[cur].rep[2] = opt[cur - mlen].rep[1]; - opt[cur].rep[1] = opt[cur - mlen].rep[0]; - opt[cur].rep[0] = opt[cur].off - ZSTD_REP_MOVE_OPT; - } else { - opt[cur].rep[2] = (opt[cur].off > 1) ? opt[cur - mlen].rep[1] : opt[cur - mlen].rep[2]; - opt[cur].rep[1] = (opt[cur].off > 0) ? opt[cur - mlen].rep[0] : opt[cur - mlen].rep[1]; - opt[cur].rep[0] = - ((opt[cur].off == ZSTD_REP_MOVE_OPT) && (mlen != 1)) ? (opt[cur - mlen].rep[0] - 1) : (opt[cur - mlen].rep[opt[cur].off]); - } - - best_mlen = minMatch; - { - U32 i, last_i = ZSTD_REP_CHECK + (mlen != 1); - for (i = (opt[cur].mlen != 1); i < last_i; i++) { /* check rep */ - const S32 repCur = (i == ZSTD_REP_MOVE_OPT) ? (opt[cur].rep[0] - 1) : opt[cur].rep[i]; - if ((repCur > 0) && (repCur < (S32)(inr - prefixStart)) && - (ZSTD_readMINMATCH(inr, minMatch) == ZSTD_readMINMATCH(inr - repCur, minMatch))) { - mlen = (U32)ZSTD_count(inr + minMatch, inr + minMatch - repCur, iend) + minMatch; - - if (mlen > sufficient_len || cur + mlen >= ZSTD_OPT_NUM) { - best_mlen = mlen; - best_off = i; - last_pos = cur + 1; - goto _storeSequence; - } - - best_off = i - (opt[cur].mlen != 1); - if (mlen > best_mlen) - best_mlen = mlen; - - do { - if (opt[cur].mlen == 1) { - litlen = opt[cur].litlen; - if (cur > litlen) { - price = opt[cur - litlen].price + ZSTD_getPrice(seqStorePtr, litlen, inr - litlen, - best_off, mlen - MINMATCH, ultra); - } else - price = ZSTD_getPrice(seqStorePtr, litlen, anchor, best_off, mlen - MINMATCH, ultra); - } else { - litlen = 0; - price = opt[cur].price + ZSTD_getPrice(seqStorePtr, 0, NULL, best_off, mlen - MINMATCH, ultra); - } - - if (cur + mlen > last_pos || price <= opt[cur + mlen].price) - SET_PRICE(cur + mlen, mlen, i, litlen, price); - mlen--; - } while (mlen >= minMatch); - } - } - } - - match_num = ZSTD_BtGetAllMatches_selectMLS(ctx, inr, iend, maxSearches, mls, matches, best_mlen); - - if (match_num > 0 && (matches[match_num - 1].len > sufficient_len || cur + matches[match_num - 1].len >= ZSTD_OPT_NUM)) { - best_mlen = matches[match_num - 1].len; - best_off = matches[match_num - 1].off; - last_pos = cur + 1; - goto _storeSequence; - } - - /* set prices using matches at position = cur */ - for (u = 0; u < match_num; u++) { - mlen = (u > 0) ? matches[u - 1].len + 1 : best_mlen; - best_mlen = matches[u].len; - - while (mlen <= best_mlen) { - if (opt[cur].mlen == 1) { - litlen = opt[cur].litlen; - if (cur > litlen) - price = opt[cur - litlen].price + ZSTD_getPrice(seqStorePtr, litlen, ip + cur - litlen, - matches[u].off - 1, mlen - MINMATCH, ultra); - else - price = ZSTD_getPrice(seqStorePtr, litlen, anchor, matches[u].off - 1, mlen - MINMATCH, ultra); - } else { - litlen = 0; - price = opt[cur].price + ZSTD_getPrice(seqStorePtr, 0, NULL, matches[u].off - 1, mlen - MINMATCH, ultra); - } - - if (cur + mlen > last_pos || (price < opt[cur + mlen].price)) - SET_PRICE(cur + mlen, mlen, matches[u].off, litlen, price); - - mlen++; - } - } - } - - best_mlen = opt[last_pos].mlen; - best_off = opt[last_pos].off; - cur = last_pos - best_mlen; - - /* store sequence */ -_storeSequence: /* cur, last_pos, best_mlen, best_off have to be set */ - opt[0].mlen = 1; - - while (1) { - mlen = opt[cur].mlen; - offset = opt[cur].off; - opt[cur].mlen = best_mlen; - opt[cur].off = best_off; - best_mlen = mlen; - best_off = offset; - if (mlen > cur) - break; - cur -= mlen; - } - - for (u = 0; u <= last_pos;) { - u += opt[u].mlen; - } - - for (cur = 0; cur < last_pos;) { - mlen = opt[cur].mlen; - if (mlen == 1) { - ip++; - cur++; - continue; - } - offset = opt[cur].off; - cur += mlen; - litLength = (U32)(ip - anchor); - - if (offset > ZSTD_REP_MOVE_OPT) { - rep[2] = rep[1]; - rep[1] = rep[0]; - rep[0] = offset - ZSTD_REP_MOVE_OPT; - offset--; - } else { - if (offset != 0) { - best_off = (offset == ZSTD_REP_MOVE_OPT) ? (rep[0] - 1) : (rep[offset]); - if (offset != 1) - rep[2] = rep[1]; - rep[1] = rep[0]; - rep[0] = best_off; - } - if (litLength == 0) - offset--; - } - - ZSTD_updatePrice(seqStorePtr, litLength, anchor, offset, mlen - MINMATCH); - ZSTD_storeSeq(seqStorePtr, litLength, anchor, offset, mlen - MINMATCH); - anchor = ip = ip + mlen; - } - } /* for (cur=0; cur < last_pos; ) */ - - /* Save reps for next block */ - { - int i; - for (i = 0; i < ZSTD_REP_NUM; i++) - ctx->repToConfirm[i] = rep[i]; - } - - /* Last Literals */ - { - size_t const lastLLSize = iend - anchor; - memcpy(seqStorePtr->lit, anchor, lastLLSize); - seqStorePtr->lit += lastLLSize; - } -} - -FORCE_INLINE -void ZSTD_compressBlock_opt_extDict_generic(ZSTD_CCtx *ctx, const void *src, size_t srcSize, const int ultra) -{ - seqStore_t *seqStorePtr = &(ctx->seqStore); - const BYTE *const istart = (const BYTE *)src; - const BYTE *ip = istart; - const BYTE *anchor = istart; - const BYTE *const iend = istart + srcSize; - const BYTE *const ilimit = iend - 8; - const BYTE *const base = ctx->base; - const U32 lowestIndex = ctx->lowLimit; - const U32 dictLimit = ctx->dictLimit; - const BYTE *const prefixStart = base + dictLimit; - const BYTE *const dictBase = ctx->dictBase; - const BYTE *const dictEnd = dictBase + dictLimit; - - const U32 maxSearches = 1U << ctx->params.cParams.searchLog; - const U32 sufficient_len = ctx->params.cParams.targetLength; - const U32 mls = ctx->params.cParams.searchLength; - const U32 minMatch = (ctx->params.cParams.searchLength == 3) ? 3 : 4; - - ZSTD_optimal_t *opt = seqStorePtr->priceTable; - ZSTD_match_t *matches = seqStorePtr->matchTable; - const BYTE *inr; - - /* init */ - U32 offset, rep[ZSTD_REP_NUM]; - { - U32 i; - for (i = 0; i < ZSTD_REP_NUM; i++) - rep[i] = ctx->rep[i]; - } - - ctx->nextToUpdate3 = ctx->nextToUpdate; - ZSTD_rescaleFreqs(seqStorePtr, (const BYTE *)src, srcSize); - ip += (ip == prefixStart); - - /* Match Loop */ - while (ip < ilimit) { - U32 cur, match_num, last_pos, litlen, price; - U32 u, mlen, best_mlen, best_off, litLength; - U32 curr = (U32)(ip - base); - memset(opt, 0, sizeof(ZSTD_optimal_t)); - last_pos = 0; - opt[0].litlen = (U32)(ip - anchor); - - /* check repCode */ - { - U32 i, last_i = ZSTD_REP_CHECK + (ip == anchor); - for (i = (ip == anchor); i < last_i; i++) { - const S32 repCur = (i == ZSTD_REP_MOVE_OPT) ? (rep[0] - 1) : rep[i]; - const U32 repIndex = (U32)(curr - repCur); - const BYTE *const repBase = repIndex < dictLimit ? dictBase : base; - const BYTE *const repMatch = repBase + repIndex; - if ((repCur > 0 && repCur <= (S32)curr) && - (((U32)((dictLimit - 1) - repIndex) >= 3) & (repIndex > lowestIndex)) /* intentional overflow */ - && (ZSTD_readMINMATCH(ip, minMatch) == ZSTD_readMINMATCH(repMatch, minMatch))) { - /* repcode detected we should take it */ - const BYTE *const repEnd = repIndex < dictLimit ? dictEnd : iend; - mlen = (U32)ZSTD_count_2segments(ip + minMatch, repMatch + minMatch, iend, repEnd, prefixStart) + minMatch; - - if (mlen > sufficient_len || mlen >= ZSTD_OPT_NUM) { - best_mlen = mlen; - best_off = i; - cur = 0; - last_pos = 1; - goto _storeSequence; - } - - best_off = i - (ip == anchor); - litlen = opt[0].litlen; - do { - price = ZSTD_getPrice(seqStorePtr, litlen, anchor, best_off, mlen - MINMATCH, ultra); - if (mlen > last_pos || price < opt[mlen].price) - SET_PRICE(mlen, mlen, i, litlen, price); /* note : macro modifies last_pos */ - mlen--; - } while (mlen >= minMatch); - } - } - } - - match_num = ZSTD_BtGetAllMatches_selectMLS_extDict(ctx, ip, iend, maxSearches, mls, matches, minMatch); /* first search (depth 0) */ - - if (!last_pos && !match_num) { - ip++; - continue; - } - - { - U32 i; - for (i = 0; i < ZSTD_REP_NUM; i++) - opt[0].rep[i] = rep[i]; - } - opt[0].mlen = 1; - - if (match_num && (matches[match_num - 1].len > sufficient_len || matches[match_num - 1].len >= ZSTD_OPT_NUM)) { - best_mlen = matches[match_num - 1].len; - best_off = matches[match_num - 1].off; - cur = 0; - last_pos = 1; - goto _storeSequence; - } - - best_mlen = (last_pos) ? last_pos : minMatch; - - /* set prices using matches at position = 0 */ - for (u = 0; u < match_num; u++) { - mlen = (u > 0) ? matches[u - 1].len + 1 : best_mlen; - best_mlen = matches[u].len; - litlen = opt[0].litlen; - while (mlen <= best_mlen) { - price = ZSTD_getPrice(seqStorePtr, litlen, anchor, matches[u].off - 1, mlen - MINMATCH, ultra); - if (mlen > last_pos || price < opt[mlen].price) - SET_PRICE(mlen, mlen, matches[u].off, litlen, price); - mlen++; - } - } - - if (last_pos < minMatch) { - ip++; - continue; - } - - /* check further positions */ - for (cur = 1; cur <= last_pos; cur++) { - inr = ip + cur; - - if (opt[cur - 1].mlen == 1) { - litlen = opt[cur - 1].litlen + 1; - if (cur > litlen) { - price = opt[cur - litlen].price + ZSTD_getLiteralPrice(seqStorePtr, litlen, inr - litlen); - } else - price = ZSTD_getLiteralPrice(seqStorePtr, litlen, anchor); - } else { - litlen = 1; - price = opt[cur - 1].price + ZSTD_getLiteralPrice(seqStorePtr, litlen, inr - 1); - } - - if (cur > last_pos || price <= opt[cur].price) - SET_PRICE(cur, 1, 0, litlen, price); - - if (cur == last_pos) - break; - - if (inr > ilimit) /* last match must start at a minimum distance of 8 from oend */ - continue; - - mlen = opt[cur].mlen; - if (opt[cur].off > ZSTD_REP_MOVE_OPT) { - opt[cur].rep[2] = opt[cur - mlen].rep[1]; - opt[cur].rep[1] = opt[cur - mlen].rep[0]; - opt[cur].rep[0] = opt[cur].off - ZSTD_REP_MOVE_OPT; - } else { - opt[cur].rep[2] = (opt[cur].off > 1) ? opt[cur - mlen].rep[1] : opt[cur - mlen].rep[2]; - opt[cur].rep[1] = (opt[cur].off > 0) ? opt[cur - mlen].rep[0] : opt[cur - mlen].rep[1]; - opt[cur].rep[0] = - ((opt[cur].off == ZSTD_REP_MOVE_OPT) && (mlen != 1)) ? (opt[cur - mlen].rep[0] - 1) : (opt[cur - mlen].rep[opt[cur].off]); - } - - best_mlen = minMatch; - { - U32 i, last_i = ZSTD_REP_CHECK + (mlen != 1); - for (i = (mlen != 1); i < last_i; i++) { - const S32 repCur = (i == ZSTD_REP_MOVE_OPT) ? (opt[cur].rep[0] - 1) : opt[cur].rep[i]; - const U32 repIndex = (U32)(curr + cur - repCur); - const BYTE *const repBase = repIndex < dictLimit ? dictBase : base; - const BYTE *const repMatch = repBase + repIndex; - if ((repCur > 0 && repCur <= (S32)(curr + cur)) && - (((U32)((dictLimit - 1) - repIndex) >= 3) & (repIndex > lowestIndex)) /* intentional overflow */ - && (ZSTD_readMINMATCH(inr, minMatch) == ZSTD_readMINMATCH(repMatch, minMatch))) { - /* repcode detected */ - const BYTE *const repEnd = repIndex < dictLimit ? dictEnd : iend; - mlen = (U32)ZSTD_count_2segments(inr + minMatch, repMatch + minMatch, iend, repEnd, prefixStart) + minMatch; - - if (mlen > sufficient_len || cur + mlen >= ZSTD_OPT_NUM) { - best_mlen = mlen; - best_off = i; - last_pos = cur + 1; - goto _storeSequence; - } - - best_off = i - (opt[cur].mlen != 1); - if (mlen > best_mlen) - best_mlen = mlen; - - do { - if (opt[cur].mlen == 1) { - litlen = opt[cur].litlen; - if (cur > litlen) { - price = opt[cur - litlen].price + ZSTD_getPrice(seqStorePtr, litlen, inr - litlen, - best_off, mlen - MINMATCH, ultra); - } else - price = ZSTD_getPrice(seqStorePtr, litlen, anchor, best_off, mlen - MINMATCH, ultra); - } else { - litlen = 0; - price = opt[cur].price + ZSTD_getPrice(seqStorePtr, 0, NULL, best_off, mlen - MINMATCH, ultra); - } - - if (cur + mlen > last_pos || price <= opt[cur + mlen].price) - SET_PRICE(cur + mlen, mlen, i, litlen, price); - mlen--; - } while (mlen >= minMatch); - } - } - } - - match_num = ZSTD_BtGetAllMatches_selectMLS_extDict(ctx, inr, iend, maxSearches, mls, matches, minMatch); - - if (match_num > 0 && (matches[match_num - 1].len > sufficient_len || cur + matches[match_num - 1].len >= ZSTD_OPT_NUM)) { - best_mlen = matches[match_num - 1].len; - best_off = matches[match_num - 1].off; - last_pos = cur + 1; - goto _storeSequence; - } - - /* set prices using matches at position = cur */ - for (u = 0; u < match_num; u++) { - mlen = (u > 0) ? matches[u - 1].len + 1 : best_mlen; - best_mlen = matches[u].len; - - while (mlen <= best_mlen) { - if (opt[cur].mlen == 1) { - litlen = opt[cur].litlen; - if (cur > litlen) - price = opt[cur - litlen].price + ZSTD_getPrice(seqStorePtr, litlen, ip + cur - litlen, - matches[u].off - 1, mlen - MINMATCH, ultra); - else - price = ZSTD_getPrice(seqStorePtr, litlen, anchor, matches[u].off - 1, mlen - MINMATCH, ultra); - } else { - litlen = 0; - price = opt[cur].price + ZSTD_getPrice(seqStorePtr, 0, NULL, matches[u].off - 1, mlen - MINMATCH, ultra); - } - - if (cur + mlen > last_pos || (price < opt[cur + mlen].price)) - SET_PRICE(cur + mlen, mlen, matches[u].off, litlen, price); - - mlen++; - } - } - } /* for (cur = 1; cur <= last_pos; cur++) */ - - best_mlen = opt[last_pos].mlen; - best_off = opt[last_pos].off; - cur = last_pos - best_mlen; - - /* store sequence */ -_storeSequence: /* cur, last_pos, best_mlen, best_off have to be set */ - opt[0].mlen = 1; - - while (1) { - mlen = opt[cur].mlen; - offset = opt[cur].off; - opt[cur].mlen = best_mlen; - opt[cur].off = best_off; - best_mlen = mlen; - best_off = offset; - if (mlen > cur) - break; - cur -= mlen; - } - - for (u = 0; u <= last_pos;) { - u += opt[u].mlen; - } - - for (cur = 0; cur < last_pos;) { - mlen = opt[cur].mlen; - if (mlen == 1) { - ip++; - cur++; - continue; - } - offset = opt[cur].off; - cur += mlen; - litLength = (U32)(ip - anchor); - - if (offset > ZSTD_REP_MOVE_OPT) { - rep[2] = rep[1]; - rep[1] = rep[0]; - rep[0] = offset - ZSTD_REP_MOVE_OPT; - offset--; - } else { - if (offset != 0) { - best_off = (offset == ZSTD_REP_MOVE_OPT) ? (rep[0] - 1) : (rep[offset]); - if (offset != 1) - rep[2] = rep[1]; - rep[1] = rep[0]; - rep[0] = best_off; - } - - if (litLength == 0) - offset--; - } - - ZSTD_updatePrice(seqStorePtr, litLength, anchor, offset, mlen - MINMATCH); - ZSTD_storeSeq(seqStorePtr, litLength, anchor, offset, mlen - MINMATCH); - anchor = ip = ip + mlen; - } - } /* for (cur=0; cur < last_pos; ) */ - - /* Save reps for next block */ - { - int i; - for (i = 0; i < ZSTD_REP_NUM; i++) - ctx->repToConfirm[i] = rep[i]; - } - - /* Last Literals */ - { - size_t lastLLSize = iend - anchor; - memcpy(seqStorePtr->lit, anchor, lastLLSize); - seqStorePtr->lit += lastLLSize; - } -} - -#endif /* ZSTD_OPT_H_91842398743 */ -- cgit v1.2.3 From 0a8ea235837cc39f27c45689930aa97ae91d5953 Mon Sep 17 00:00:00 2001 From: Nathan Chancellor Date: Thu, 21 Oct 2021 13:23:53 -0700 Subject: lib: zstd: Add cast to silence clang's -Wbitwise-instead-of-logical A new warning in clang warns that there is an instance where boolean expressions are being used with bitwise operators instead of logical ones: lib/zstd/decompress/huf_decompress.c:890:25: warning: use of bitwise '&' with boolean operands [-Wbitwise-instead-of-logical] (BIT_reloadDStreamFast(&bitD1) == BIT_DStream_unfinished) ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ zstd does this frequently to help with performance, as logical operators have branches whereas bitwise ones do not. To fix this warning in other cases, the expressions were placed on separate lines with the '&=' operator; however, this particular instance was moved away from that so that it could be surrounded by LIKELY, which is a macro for __builtin_expect(), to help with a performance regression, according to upstream zstd pull #1973. Aside from switching to logical operators, which is likely undesirable in this instance, or disabling the warning outright, the solution is casting one of the expressions to an integer type to make it clear to clang that the author knows what they are doing. Add a cast to U32 to silence the warning. The first U32 cast is to silence an instance of -Wshorten-64-to-32 because __builtin_expect() returns long so it cannot be moved. Link: https://github.com/ClangBuiltLinux/linux/issues/1486 Link: https://github.com/facebook/zstd/pull/1973 Reported-by: Nick Desaulniers Signed-off-by: Nathan Chancellor Signed-off-by: Nick Terrell --- lib/zstd/decompress/huf_decompress.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'lib') diff --git a/lib/zstd/decompress/huf_decompress.c b/lib/zstd/decompress/huf_decompress.c index 05570ed5f8be..5105e59ac04a 100644 --- a/lib/zstd/decompress/huf_decompress.c +++ b/lib/zstd/decompress/huf_decompress.c @@ -886,7 +886,7 @@ HUF_decompress4X2_usingDTable_internal_body( HUF_DECODE_SYMBOLX2_0(op2, &bitD2); HUF_DECODE_SYMBOLX2_0(op3, &bitD3); HUF_DECODE_SYMBOLX2_0(op4, &bitD4); - endSignal = (U32)LIKELY( + endSignal = (U32)LIKELY((U32) (BIT_reloadDStreamFast(&bitD1) == BIT_DStream_unfinished) & (BIT_reloadDStreamFast(&bitD2) == BIT_DStream_unfinished) & (BIT_reloadDStreamFast(&bitD3) == BIT_DStream_unfinished) -- cgit v1.2.3 From 4d4712c1a4ac331574fb4542e4adcfd24fdb07d0 Mon Sep 17 00:00:00 2001 From: Imran Khan Date: Mon, 8 Nov 2021 18:33:09 -0800 Subject: lib, stackdepot: check stackdepot handle before accessing slabs Patch series "lib, stackdepot: check stackdepot handle before accessing slabs", v2. PATCH-1: Checks validity of a stackdepot handle before proceeding to access stackdepot slab/objects. PATCH-2: Adds a helper in stackdepot, to allow users to print stack entries just by specifying the stackdepot handle. It also changes such users to use this new interface. PATCH-3: Adds a helper in stackdepot, to allow users to print stack entries into buffers just by specifying the stackdepot handle and destination buffer. It also changes such users to use this new interface. This patch (of 3): stack_depot_save allocates slabs that will be used for storing objects in future.If this slab allocation fails we may get to a situation where space allocation for a new stack_record fails, causing stack_depot_save to return 0 as handle. If user of this handle ends up invoking stack_depot_fetch with this handle value, current implementation of stack_depot_fetch will end up using slab from wrong index. To avoid this check handle value at the beginning. Link: https://lkml.kernel.org/r/20210915175321.3472770-1-imran.f.khan@oracle.com Link: https://lkml.kernel.org/r/20210915014806.3206938-1-imran.f.khan@oracle.com Link: https://lkml.kernel.org/r/20210915014806.3206938-2-imran.f.khan@oracle.com Signed-off-by: Imran Khan Suggested-by: Vlastimil Babka Acked-by: Vlastimil Babka Cc: Geert Uytterhoeven Cc: Andrey Ryabinin Cc: Alexander Potapenko Cc: Andrey Konovalov Cc: Dmitry Vyukov Cc: Maarten Lankhorst Cc: Maxime Ripard Cc: Thomas Zimmermann Cc: David Airlie Cc: Daniel Vetter Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- lib/stackdepot.c | 3 +++ 1 file changed, 3 insertions(+) (limited to 'lib') diff --git a/lib/stackdepot.c b/lib/stackdepot.c index 09485dc5bd12..f034f095c627 100644 --- a/lib/stackdepot.c +++ b/lib/stackdepot.c @@ -231,6 +231,9 @@ unsigned int stack_depot_fetch(depot_stack_handle_t handle, struct stack_record *stack; *entries = NULL; + if (!handle) + return 0; + if (parts.slabindex > depot_index) { WARN(1, "slab index %d out of bounds (%d) for stack id %08x\n", parts.slabindex, depot_index, handle); -- cgit v1.2.3 From 505be48165fa2f2cb795659b4fc61f35563d105e Mon Sep 17 00:00:00 2001 From: Imran Khan Date: Mon, 8 Nov 2021 18:33:12 -0800 Subject: lib, stackdepot: add helper to print stack entries To print a stack entries, users of stackdepot, first use stack_depot_fetch to get a list of stack entries and then use stack_trace_print to print this list. Provide a helper in stackdepot to print stack entries based on stackdepot handle. Also change above mentioned users to use this helper. Link: https://lkml.kernel.org/r/20210915014806.3206938-3-imran.f.khan@oracle.com Signed-off-by: Imran Khan Suggested-by: Vlastimil Babka Acked-by: Vlastimil Babka Reviewed-by: Alexander Potapenko Cc: Andrey Konovalov Cc: Andrey Ryabinin Cc: Daniel Vetter Cc: David Airlie Cc: Dmitry Vyukov Cc: Geert Uytterhoeven Cc: Maarten Lankhorst Cc: Maxime Ripard Cc: Thomas Zimmermann Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/linux/stackdepot.h | 2 ++ lib/stackdepot.c | 18 ++++++++++++++++++ mm/kasan/report.c | 15 +++------------ mm/page_owner.c | 13 ++++--------- 4 files changed, 27 insertions(+), 21 deletions(-) (limited to 'lib') diff --git a/include/linux/stackdepot.h b/include/linux/stackdepot.h index d29860966bc9..8ab37500fcba 100644 --- a/include/linux/stackdepot.h +++ b/include/linux/stackdepot.h @@ -25,6 +25,8 @@ depot_stack_handle_t stack_depot_save(unsigned long *entries, unsigned int stack_depot_fetch(depot_stack_handle_t handle, unsigned long **entries); +void stack_depot_print(depot_stack_handle_t stack); + #ifdef CONFIG_STACKDEPOT int stack_depot_init(void); #else diff --git a/lib/stackdepot.c b/lib/stackdepot.c index f034f095c627..4e1f2982d0fa 100644 --- a/lib/stackdepot.c +++ b/lib/stackdepot.c @@ -213,6 +213,24 @@ static inline struct stack_record *find_stack(struct stack_record *bucket, return NULL; } +/** + * stack_depot_print - print stack entries from a depot + * + * @stack: Stack depot handle which was returned from + * stack_depot_save(). + * + */ +void stack_depot_print(depot_stack_handle_t stack) +{ + unsigned long *entries; + unsigned int nr_entries; + + nr_entries = stack_depot_fetch(stack, &entries); + if (nr_entries > 0) + stack_trace_print(entries, nr_entries, 0); +} +EXPORT_SYMBOL_GPL(stack_depot_print); + /** * stack_depot_fetch - Fetch stack entries from a depot * diff --git a/mm/kasan/report.c b/mm/kasan/report.c index 884a950c7026..3239fd8f8747 100644 --- a/mm/kasan/report.c +++ b/mm/kasan/report.c @@ -132,20 +132,11 @@ static void end_report(unsigned long *flags, unsigned long addr) kasan_enable_current(); } -static void print_stack(depot_stack_handle_t stack) -{ - unsigned long *entries; - unsigned int nr_entries; - - nr_entries = stack_depot_fetch(stack, &entries); - stack_trace_print(entries, nr_entries, 0); -} - static void print_track(struct kasan_track *track, const char *prefix) { pr_err("%s by task %u:\n", prefix, track->pid); if (track->stack) { - print_stack(track->stack); + stack_depot_print(track->stack); } else { pr_err("(stack is not available)\n"); } @@ -214,12 +205,12 @@ static void describe_object_stacks(struct kmem_cache *cache, void *object, return; if (alloc_meta->aux_stack[0]) { pr_err("Last potentially related work creation:\n"); - print_stack(alloc_meta->aux_stack[0]); + stack_depot_print(alloc_meta->aux_stack[0]); pr_err("\n"); } if (alloc_meta->aux_stack[1]) { pr_err("Second to last potentially related work creation:\n"); - print_stack(alloc_meta->aux_stack[1]); + stack_depot_print(alloc_meta->aux_stack[1]); pr_err("\n"); } #endif diff --git a/mm/page_owner.c b/mm/page_owner.c index 62402d22539b..eff29be1218b 100644 --- a/mm/page_owner.c +++ b/mm/page_owner.c @@ -394,8 +394,6 @@ void __dump_page_owner(const struct page *page) struct page_ext *page_ext = lookup_page_ext(page); struct page_owner *page_owner; depot_stack_handle_t handle; - unsigned long *entries; - unsigned int nr_entries; gfp_t gfp_mask; int mt; @@ -423,20 +421,17 @@ void __dump_page_owner(const struct page *page) page_owner->pid, page_owner->ts_nsec, page_owner->free_ts_nsec); handle = READ_ONCE(page_owner->handle); - if (!handle) { + if (!handle) pr_alert("page_owner allocation stack trace missing\n"); - } else { - nr_entries = stack_depot_fetch(handle, &entries); - stack_trace_print(entries, nr_entries, 0); - } + else + stack_depot_print(handle); handle = READ_ONCE(page_owner->free_handle); if (!handle) { pr_alert("page_owner free stack trace missing\n"); } else { - nr_entries = stack_depot_fetch(handle, &entries); pr_alert("page last free stack trace:\n"); - stack_trace_print(entries, nr_entries, 0); + stack_depot_print(handle); } if (page_owner->last_migrate_reason != -1) -- cgit v1.2.3 From 0f68d45ef41abb618a9ca33996348ae73800a106 Mon Sep 17 00:00:00 2001 From: Imran Khan Date: Mon, 8 Nov 2021 18:33:16 -0800 Subject: lib, stackdepot: add helper to print stack entries into buffer To print stack entries into a buffer, users of stackdepot, first get a list of stack entries using stack_depot_fetch and then print this list into a buffer using stack_trace_snprint. Provide a helper in stackdepot for this purpose. Also change above mentioned users to use this helper. [imran.f.khan@oracle.com: fix build error] Link: https://lkml.kernel.org/r/20210915175321.3472770-4-imran.f.khan@oracle.com [imran.f.khan@oracle.com: export stack_depot_snprint() to modules] Link: https://lkml.kernel.org/r/20210916133535.3592491-4-imran.f.khan@oracle.com Link: https://lkml.kernel.org/r/20210915014806.3206938-4-imran.f.khan@oracle.com Signed-off-by: Imran Khan Suggested-by: Vlastimil Babka Acked-by: Vlastimil Babka Acked-by: Jani Nikula [i915] Cc: Alexander Potapenko Cc: Andrey Konovalov Cc: Andrey Ryabinin Cc: Daniel Vetter Cc: David Airlie Cc: Dmitry Vyukov Cc: Geert Uytterhoeven Cc: Maarten Lankhorst Cc: Maxime Ripard Cc: Thomas Zimmermann Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- drivers/gpu/drm/drm_dp_mst_topology.c | 5 +---- drivers/gpu/drm/drm_mm.c | 5 +---- drivers/gpu/drm/i915/i915_vma.c | 5 +---- drivers/gpu/drm/i915/intel_runtime_pm.c | 20 +++++--------------- include/linux/stackdepot.h | 3 +++ lib/stackdepot.c | 25 +++++++++++++++++++++++++ mm/page_owner.c | 5 +---- 7 files changed, 37 insertions(+), 31 deletions(-) (limited to 'lib') diff --git a/drivers/gpu/drm/drm_dp_mst_topology.c b/drivers/gpu/drm/drm_dp_mst_topology.c index 86d13d6bc463..2d1adab9e360 100644 --- a/drivers/gpu/drm/drm_dp_mst_topology.c +++ b/drivers/gpu/drm/drm_dp_mst_topology.c @@ -1668,13 +1668,10 @@ __dump_topology_ref_history(struct drm_dp_mst_topology_ref_history *history, for (i = 0; i < history->len; i++) { const struct drm_dp_mst_topology_ref_entry *entry = &history->entries[i]; - ulong *entries; - uint nr_entries; u64 ts_nsec = entry->ts_nsec; u32 rem_nsec = do_div(ts_nsec, 1000000000); - nr_entries = stack_depot_fetch(entry->backtrace, &entries); - stack_trace_snprint(buf, PAGE_SIZE, entries, nr_entries, 4); + stack_depot_snprint(entry->backtrace, buf, PAGE_SIZE, 4); drm_printf(&p, " %d %ss (last at %5llu.%06u):\n%s", entry->count, diff --git a/drivers/gpu/drm/drm_mm.c b/drivers/gpu/drm/drm_mm.c index 93d48a6f04ab..7d1c578388d3 100644 --- a/drivers/gpu/drm/drm_mm.c +++ b/drivers/gpu/drm/drm_mm.c @@ -118,8 +118,6 @@ static noinline void save_stack(struct drm_mm_node *node) static void show_leaks(struct drm_mm *mm) { struct drm_mm_node *node; - unsigned long *entries; - unsigned int nr_entries; char *buf; buf = kmalloc(BUFSZ, GFP_KERNEL); @@ -133,8 +131,7 @@ static void show_leaks(struct drm_mm *mm) continue; } - nr_entries = stack_depot_fetch(node->stack, &entries); - stack_trace_snprint(buf, BUFSZ, entries, nr_entries, 0); + stack_depot_snprint(node->stack, buf, BUFSZ, 0); DRM_ERROR("node [%08llx + %08llx]: inserted at\n%s", node->start, node->size, buf); } diff --git a/drivers/gpu/drm/i915/i915_vma.c b/drivers/gpu/drm/i915/i915_vma.c index 4b7fc4647e46..f2d9ed375109 100644 --- a/drivers/gpu/drm/i915/i915_vma.c +++ b/drivers/gpu/drm/i915/i915_vma.c @@ -56,8 +56,6 @@ void i915_vma_free(struct i915_vma *vma) static void vma_print_allocator(struct i915_vma *vma, const char *reason) { - unsigned long *entries; - unsigned int nr_entries; char buf[512]; if (!vma->node.stack) { @@ -66,8 +64,7 @@ static void vma_print_allocator(struct i915_vma *vma, const char *reason) return; } - nr_entries = stack_depot_fetch(vma->node.stack, &entries); - stack_trace_snprint(buf, sizeof(buf), entries, nr_entries, 0); + stack_depot_snprint(vma->node.stack, buf, sizeof(buf), 0); DRM_DEBUG_DRIVER("vma.node [%08llx + %08llx] %s: inserted at %s\n", vma->node.start, vma->node.size, reason, buf); } diff --git a/drivers/gpu/drm/i915/intel_runtime_pm.c b/drivers/gpu/drm/i915/intel_runtime_pm.c index eaf7688f517d..0d85f3c5c526 100644 --- a/drivers/gpu/drm/i915/intel_runtime_pm.c +++ b/drivers/gpu/drm/i915/intel_runtime_pm.c @@ -65,16 +65,6 @@ static noinline depot_stack_handle_t __save_depot_stack(void) return stack_depot_save(entries, n, GFP_NOWAIT | __GFP_NOWARN); } -static void __print_depot_stack(depot_stack_handle_t stack, - char *buf, int sz, int indent) -{ - unsigned long *entries; - unsigned int nr_entries; - - nr_entries = stack_depot_fetch(stack, &entries); - stack_trace_snprint(buf, sz, entries, nr_entries, indent); -} - static void init_intel_runtime_pm_wakeref(struct intel_runtime_pm *rpm) { spin_lock_init(&rpm->debug.lock); @@ -146,12 +136,12 @@ static void untrack_intel_runtime_pm_wakeref(struct intel_runtime_pm *rpm, if (!buf) return; - __print_depot_stack(stack, buf, PAGE_SIZE, 2); + stack_depot_snprint(stack, buf, PAGE_SIZE, 2); DRM_DEBUG_DRIVER("wakeref %x from\n%s", stack, buf); stack = READ_ONCE(rpm->debug.last_release); if (stack) { - __print_depot_stack(stack, buf, PAGE_SIZE, 2); + stack_depot_snprint(stack, buf, PAGE_SIZE, 2); DRM_DEBUG_DRIVER("wakeref last released at\n%s", buf); } @@ -183,12 +173,12 @@ __print_intel_runtime_pm_wakeref(struct drm_printer *p, return; if (dbg->last_acquire) { - __print_depot_stack(dbg->last_acquire, buf, PAGE_SIZE, 2); + stack_depot_snprint(dbg->last_acquire, buf, PAGE_SIZE, 2); drm_printf(p, "Wakeref last acquired:\n%s", buf); } if (dbg->last_release) { - __print_depot_stack(dbg->last_release, buf, PAGE_SIZE, 2); + stack_depot_snprint(dbg->last_release, buf, PAGE_SIZE, 2); drm_printf(p, "Wakeref last released:\n%s", buf); } @@ -203,7 +193,7 @@ __print_intel_runtime_pm_wakeref(struct drm_printer *p, rep = 1; while (i + 1 < dbg->count && dbg->owners[i + 1] == stack) rep++, i++; - __print_depot_stack(stack, buf, PAGE_SIZE, 2); + stack_depot_snprint(stack, buf, PAGE_SIZE, 2); drm_printf(p, "Wakeref x%lu taken at:\n%s", rep, buf); } diff --git a/include/linux/stackdepot.h b/include/linux/stackdepot.h index 8ab37500fcba..c34b55a6e554 100644 --- a/include/linux/stackdepot.h +++ b/include/linux/stackdepot.h @@ -25,6 +25,9 @@ depot_stack_handle_t stack_depot_save(unsigned long *entries, unsigned int stack_depot_fetch(depot_stack_handle_t handle, unsigned long **entries); +int stack_depot_snprint(depot_stack_handle_t handle, char *buf, size_t size, + int spaces); + void stack_depot_print(depot_stack_handle_t stack); #ifdef CONFIG_STACKDEPOT diff --git a/lib/stackdepot.c b/lib/stackdepot.c index 4e1f2982d0fa..b437ae79aca1 100644 --- a/lib/stackdepot.c +++ b/lib/stackdepot.c @@ -213,6 +213,31 @@ static inline struct stack_record *find_stack(struct stack_record *bucket, return NULL; } +/** + * stack_depot_snprint - print stack entries from a depot into a buffer + * + * @handle: Stack depot handle which was returned from + * stack_depot_save(). + * @buf: Pointer to the print buffer + * + * @size: Size of the print buffer + * + * @spaces: Number of leading spaces to print + * + * Return: Number of bytes printed. + */ +int stack_depot_snprint(depot_stack_handle_t handle, char *buf, size_t size, + int spaces) +{ + unsigned long *entries; + unsigned int nr_entries; + + nr_entries = stack_depot_fetch(handle, &entries); + return nr_entries ? stack_trace_snprint(buf, size, entries, nr_entries, + spaces) : 0; +} +EXPORT_SYMBOL_GPL(stack_depot_snprint); + /** * stack_depot_print - print stack entries from a depot * diff --git a/mm/page_owner.c b/mm/page_owner.c index eff29be1218b..1653040d1133 100644 --- a/mm/page_owner.c +++ b/mm/page_owner.c @@ -329,8 +329,6 @@ print_page_owner(char __user *buf, size_t count, unsigned long pfn, depot_stack_handle_t handle) { int ret, pageblock_mt, page_mt; - unsigned long *entries; - unsigned int nr_entries; char *kbuf; count = min_t(size_t, count, PAGE_SIZE); @@ -361,8 +359,7 @@ print_page_owner(char __user *buf, size_t count, unsigned long pfn, if (ret >= count) goto err; - nr_entries = stack_depot_fetch(handle, &entries); - ret += stack_trace_snprint(kbuf + ret, count - ret, entries, nr_entries, 0); + ret += stack_depot_snprint(handle, kbuf + ret, count - ret, 0); if (ret >= count) goto err; -- cgit v1.2.3 From 839b395eb9c13ae56ea5fc3ca9802734a72293f0 Mon Sep 17 00:00:00 2001 From: Alexey Dobriyan Date: Mon, 8 Nov 2021 18:33:22 -0800 Subject: lib: uninline simple_strntoull() as well Codegen become bloated again after simple_strntoull() introduction add/remove: 0/0 grow/shrink: 0/4 up/down: 0/-224 (-224) Function old new delta simple_strtoul 5 2 -3 simple_strtol 23 20 -3 simple_strtoull 119 15 -104 simple_strtoll 155 41 -114 Link: https://lkml.kernel.org/r/YVmlB9yY4lvbNKYt@localhost.localdomain Signed-off-by: Alexey Dobriyan Cc: Richard Fitzgerald Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- lib/vsprintf.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) (limited to 'lib') diff --git a/lib/vsprintf.c b/lib/vsprintf.c index d7ad44f2c8f5..25d79ed53226 100644 --- a/lib/vsprintf.c +++ b/lib/vsprintf.c @@ -53,8 +53,7 @@ #include #include "kstrtox.h" -static unsigned long long simple_strntoull(const char *startp, size_t max_chars, - char **endp, unsigned int base) +static noinline unsigned long long simple_strntoull(const char *startp, size_t max_chars, char **endp, unsigned int base) { const char *cp; unsigned long long result = 0ULL; -- cgit v1.2.3 From 723aca2085166bb7213bf6af1729ddfd94c25a3e Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Mon, 8 Nov 2021 18:33:25 -0800 Subject: mm/scatterlist: replace the !preemptible warning in sg_miter_stop() sg_miter_stop() checks for disabled preemption before unmapping a page via kunmap_atomic(). The kernel doc mentions under context that preemption must be disabled if SG_MITER_ATOMIC is set. There is no active requirement for the caller to have preemption disabled before invoking sg_mitter_stop(). The sg_mitter_*() implementation itself has no such requirement. In fact, preemption is disabled by kmap_atomic() as part of sg_miter_next() and remains disabled as long as there is an active SG_MITER_ATOMIC mapping. This is a consequence of kmap_atomic() and not a requirement for sg_mitter_*() itself. The user chooses SG_MITER_ATOMIC because it uses the API in a context where blocking is not possible or blocking is possible but he chooses a lower weight mapping which is not available on all CPUs and so it might need less overhead to setup at a price that now preemption will be disabled. The kmap_atomic() implementation on PREEMPT_RT does not disable preemption. It simply disables CPU migration to ensure that the task remains on the same CPU while the caller remains preemptible. This in turn triggers the warning in sg_miter_stop() because preemption is allowed. The PREEMPT_RT and !PREEMPT_RT implementation of kmap_atomic() disable pagefaults as a requirement. It is sufficient to check for this instead of disabled preemption. Check for disabled pagefault handler in the SG_MITER_ATOMIC case. Remove the "preemption disabled" part from the kernel doc as the sg_milter*() implementation does not care. [bigeasy@linutronix.de: commit description] Link: https://lkml.kernel.org/r/20211015211409.cqopacv3pxdwn2ty@linutronix.de Signed-off-by: Thomas Gleixner Signed-off-by: Sebastian Andrzej Siewior Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- lib/scatterlist.c | 11 ++++------- 1 file changed, 4 insertions(+), 7 deletions(-) (limited to 'lib') diff --git a/lib/scatterlist.c b/lib/scatterlist.c index abb3432ed744..d5e82e4a57ad 100644 --- a/lib/scatterlist.c +++ b/lib/scatterlist.c @@ -828,8 +828,7 @@ static bool sg_miter_get_next_page(struct sg_mapping_iter *miter) * stops @miter. * * Context: - * Don't care if @miter is stopped, or not proceeded yet. - * Otherwise, preemption disabled if the SG_MITER_ATOMIC is set. + * Don't care. * * Returns: * true if @miter contains the valid mapping. false if end of sg @@ -865,8 +864,7 @@ EXPORT_SYMBOL(sg_miter_skip); * @miter->addr and @miter->length point to the current mapping. * * Context: - * Preemption disabled if SG_MITER_ATOMIC. Preemption must stay disabled - * till @miter is stopped. May sleep if !SG_MITER_ATOMIC. + * May sleep if !SG_MITER_ATOMIC. * * Returns: * true if @miter contains the next mapping. false if end of sg @@ -906,8 +904,7 @@ EXPORT_SYMBOL(sg_miter_next); * need to be released during iteration. * * Context: - * Preemption disabled if the SG_MITER_ATOMIC is set. Don't care - * otherwise. + * Don't care otherwise. */ void sg_miter_stop(struct sg_mapping_iter *miter) { @@ -922,7 +919,7 @@ void sg_miter_stop(struct sg_mapping_iter *miter) flush_dcache_page(miter->page); if (miter->__flags & SG_MITER_ATOMIC) { - WARN_ON_ONCE(preemptible()); + WARN_ON_ONCE(!pagefault_disabled()); kunmap_atomic(miter->addr); } else kunmap(miter->page); -- cgit v1.2.3 From 5d5e4522a7f404d1a96fd6c703989d32a9c9568d Mon Sep 17 00:00:00 2001 From: Nicholas Piggin Date: Sun, 7 Nov 2021 14:51:16 +1000 Subject: printk: restore flushing of NMI buffers on remote CPUs after NMI backtraces printk from NMI context relies on irq work being raised on the local CPU to print to console. This can be a problem if the NMI was raised by a lockup detector to print lockup stack and regs, because the CPU may not enable irqs (because it is locked up). Introduce printk_trigger_flush() that can be called another CPU to try to get those messages to the console, call that where printk_safe_flush was previously called. Fixes: 93d102f094be ("printk: remove safe buffers") Cc: stable@vger.kernel.org # 5.15 Signed-off-by: Nicholas Piggin Reviewed-by: Petr Mladek Reviewed-by: John Ogness Signed-off-by: Petr Mladek Link: https://lore.kernel.org/r/20211107045116.1754411-1-npiggin@gmail.com --- arch/powerpc/kernel/watchdog.c | 6 ++++++ include/linux/printk.h | 4 ++++ kernel/printk/printk.c | 5 +++++ lib/nmi_backtrace.c | 6 ++++++ 4 files changed, 21 insertions(+) (limited to 'lib') diff --git a/arch/powerpc/kernel/watchdog.c b/arch/powerpc/kernel/watchdog.c index dc17d8903d4f..6b7a83d5e03e 100644 --- a/arch/powerpc/kernel/watchdog.c +++ b/arch/powerpc/kernel/watchdog.c @@ -186,6 +186,12 @@ static void watchdog_smp_panic(int cpu, u64 tb) if (sysctl_hardlockup_all_cpu_backtrace) trigger_allbutself_cpu_backtrace(); + /* + * Force flush any remote buffers that might be stuck in IRQ context + * and therefore could not run their irq_work. + */ + printk_trigger_flush(); + if (hardlockup_panic) nmi_panic(NULL, "Hard LOCKUP"); diff --git a/include/linux/printk.h b/include/linux/printk.h index a1379df43251..596ad6fa0336 100644 --- a/include/linux/printk.h +++ b/include/linux/printk.h @@ -206,6 +206,7 @@ void dump_stack_print_info(const char *log_lvl); void show_regs_print_info(const char *log_lvl); extern asmlinkage void dump_stack_lvl(const char *log_lvl) __cold; extern asmlinkage void dump_stack(void) __cold; +void printk_trigger_flush(void); #else static inline __printf(1, 0) int vprintk(const char *s, va_list args) @@ -282,6 +283,9 @@ static inline void dump_stack_lvl(const char *log_lvl) static inline void dump_stack(void) { } +static inline void printk_trigger_flush(void) +{ +} #endif #ifdef CONFIG_SMP diff --git a/kernel/printk/printk.c b/kernel/printk/printk.c index 65fffa6368c9..eabe23b0a982 100644 --- a/kernel/printk/printk.c +++ b/kernel/printk/printk.c @@ -3261,6 +3261,11 @@ void defer_console_output(void) preempt_enable(); } +void printk_trigger_flush(void) +{ + defer_console_output(); +} + int vprintk_deferred(const char *fmt, va_list args) { int r; diff --git a/lib/nmi_backtrace.c b/lib/nmi_backtrace.c index f9e89001b52e..199ab201d501 100644 --- a/lib/nmi_backtrace.c +++ b/lib/nmi_backtrace.c @@ -75,6 +75,12 @@ void nmi_trigger_cpumask_backtrace(const cpumask_t *mask, touch_softlockup_watchdog(); } + /* + * Force flush any remote buffers that might be stuck in IRQ context + * and therefore could not run their irq_work. + */ + printk_trigger_flush(); + clear_bit_unlock(0, &backtrace_flag); put_cpu(); } -- cgit v1.2.3 From ab09243aa95a72bac5c71e852773de34116f8d0f Mon Sep 17 00:00:00 2001 From: Alistair Popple Date: Wed, 10 Nov 2021 20:32:40 -0800 Subject: mm/migrate.c: remove MIGRATE_PFN_LOCKED MIGRATE_PFN_LOCKED is used to indicate to migrate_vma_prepare() that a source page was already locked during migrate_vma_collect(). If it wasn't then the a second attempt is made to lock the page. However if the first attempt failed it's unlikely a second attempt will succeed, and the retry adds complexity. So clean this up by removing the retry and MIGRATE_PFN_LOCKED flag. Destination pages are also meant to have the MIGRATE_PFN_LOCKED flag set, but nothing actually checks that. Link: https://lkml.kernel.org/r/20211025041608.289017-1-apopple@nvidia.com Signed-off-by: Alistair Popple Reviewed-by: Ralph Campbell Acked-by: Felix Kuehling Cc: Alex Deucher Cc: Jerome Glisse Cc: John Hubbard Cc: Zi Yan Cc: Christoph Hellwig Cc: Ben Skeggs Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- Documentation/vm/hmm.rst | 2 +- arch/powerpc/kvm/book3s_hv_uvmem.c | 4 +- drivers/gpu/drm/amd/amdkfd/kfd_migrate.c | 2 - drivers/gpu/drm/nouveau/nouveau_dmem.c | 4 +- include/linux/migrate.h | 1 - lib/test_hmm.c | 5 +- mm/migrate.c | 145 ++++++------------------------- 7 files changed, 35 insertions(+), 128 deletions(-) (limited to 'lib') diff --git a/Documentation/vm/hmm.rst b/Documentation/vm/hmm.rst index a14c2938e7af..f2a59ed82ed3 100644 --- a/Documentation/vm/hmm.rst +++ b/Documentation/vm/hmm.rst @@ -360,7 +360,7 @@ between device driver specific code and shared common code: system memory page, locks the page with ``lock_page()``, and fills in the ``dst`` array entry with:: - dst[i] = migrate_pfn(page_to_pfn(dpage)) | MIGRATE_PFN_LOCKED; + dst[i] = migrate_pfn(page_to_pfn(dpage)); Now that the driver knows that this page is being migrated, it can invalidate device private MMU mappings and copy device private memory diff --git a/arch/powerpc/kvm/book3s_hv_uvmem.c b/arch/powerpc/kvm/book3s_hv_uvmem.c index a7061ee3b157..28c436df9935 100644 --- a/arch/powerpc/kvm/book3s_hv_uvmem.c +++ b/arch/powerpc/kvm/book3s_hv_uvmem.c @@ -560,7 +560,7 @@ static int __kvmppc_svm_page_out(struct vm_area_struct *vma, gpa, 0, page_shift); if (ret == U_SUCCESS) - *mig.dst = migrate_pfn(pfn) | MIGRATE_PFN_LOCKED; + *mig.dst = migrate_pfn(pfn); else { unlock_page(dpage); __free_page(dpage); @@ -774,7 +774,7 @@ static int kvmppc_svm_page_in(struct vm_area_struct *vma, } } - *mig.dst = migrate_pfn(page_to_pfn(dpage)) | MIGRATE_PFN_LOCKED; + *mig.dst = migrate_pfn(page_to_pfn(dpage)); migrate_vma_pages(&mig); out_finalize: migrate_vma_finalize(&mig); diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_migrate.c b/drivers/gpu/drm/amd/amdkfd/kfd_migrate.c index 6d8634e40b3b..d43bfd8b35ae 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_migrate.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_migrate.c @@ -317,7 +317,6 @@ svm_migrate_copy_to_vram(struct amdgpu_device *adev, struct svm_range *prange, migrate->dst[i] = svm_migrate_addr_to_pfn(adev, dst[i]); svm_migrate_get_vram_page(prange, migrate->dst[i]); migrate->dst[i] = migrate_pfn(migrate->dst[i]); - migrate->dst[i] |= MIGRATE_PFN_LOCKED; src[i] = dma_map_page(dev, spage, 0, PAGE_SIZE, DMA_TO_DEVICE); r = dma_mapping_error(dev, src[i]); @@ -610,7 +609,6 @@ svm_migrate_copy_to_ram(struct amdgpu_device *adev, struct svm_range *prange, dst[i] >> PAGE_SHIFT, page_to_pfn(dpage)); migrate->dst[i] = migrate_pfn(page_to_pfn(dpage)); - migrate->dst[i] |= MIGRATE_PFN_LOCKED; j++; } diff --git a/drivers/gpu/drm/nouveau/nouveau_dmem.c b/drivers/gpu/drm/nouveau/nouveau_dmem.c index 92987daa5e17..3828aafd3ac4 100644 --- a/drivers/gpu/drm/nouveau/nouveau_dmem.c +++ b/drivers/gpu/drm/nouveau/nouveau_dmem.c @@ -166,7 +166,7 @@ static vm_fault_t nouveau_dmem_fault_copy_one(struct nouveau_drm *drm, goto error_dma_unmap; mutex_unlock(&svmm->mutex); - args->dst[0] = migrate_pfn(page_to_pfn(dpage)) | MIGRATE_PFN_LOCKED; + args->dst[0] = migrate_pfn(page_to_pfn(dpage)); return 0; error_dma_unmap: @@ -602,7 +602,7 @@ static unsigned long nouveau_dmem_migrate_copy_one(struct nouveau_drm *drm, ((paddr >> PAGE_SHIFT) << NVIF_VMM_PFNMAP_V0_ADDR_SHIFT); if (src & MIGRATE_PFN_WRITE) *pfn |= NVIF_VMM_PFNMAP_V0_W; - return migrate_pfn(page_to_pfn(dpage)) | MIGRATE_PFN_LOCKED; + return migrate_pfn(page_to_pfn(dpage)); out_dma_unmap: dma_unmap_page(dev, *dma_addr, PAGE_SIZE, DMA_BIDIRECTIONAL); diff --git a/include/linux/migrate.h b/include/linux/migrate.h index eeb818c4fc78..4850cc5bf813 100644 --- a/include/linux/migrate.h +++ b/include/linux/migrate.h @@ -110,7 +110,6 @@ static inline int migrate_misplaced_page(struct page *page, */ #define MIGRATE_PFN_VALID (1UL << 0) #define MIGRATE_PFN_MIGRATE (1UL << 1) -#define MIGRATE_PFN_LOCKED (1UL << 2) #define MIGRATE_PFN_WRITE (1UL << 3) #define MIGRATE_PFN_SHIFT 6 diff --git a/lib/test_hmm.c b/lib/test_hmm.c index c259842f6d44..e2ce8f9b7605 100644 --- a/lib/test_hmm.c +++ b/lib/test_hmm.c @@ -613,8 +613,7 @@ static void dmirror_migrate_alloc_and_copy(struct migrate_vma *args, */ rpage->zone_device_data = dmirror; - *dst = migrate_pfn(page_to_pfn(dpage)) | - MIGRATE_PFN_LOCKED; + *dst = migrate_pfn(page_to_pfn(dpage)); if ((*src & MIGRATE_PFN_WRITE) || (!spage && args->vma->vm_flags & VM_WRITE)) *dst |= MIGRATE_PFN_WRITE; @@ -1137,7 +1136,7 @@ static vm_fault_t dmirror_devmem_fault_alloc_and_copy(struct migrate_vma *args, lock_page(dpage); xa_erase(&dmirror->pt, addr >> PAGE_SHIFT); copy_highpage(dpage, spage); - *dst = migrate_pfn(page_to_pfn(dpage)) | MIGRATE_PFN_LOCKED; + *dst = migrate_pfn(page_to_pfn(dpage)); if (*src & MIGRATE_PFN_WRITE) *dst |= MIGRATE_PFN_WRITE; } diff --git a/mm/migrate.c b/mm/migrate.c index 43dd88c7fcdc..cf25b00f03c8 100644 --- a/mm/migrate.c +++ b/mm/migrate.c @@ -2362,7 +2362,6 @@ again: * can't be dropped from it). */ get_page(page); - migrate->cpages++; /* * Optimize for the common case where page is only mapped once @@ -2372,7 +2371,7 @@ again: if (trylock_page(page)) { pte_t swp_pte; - mpfn |= MIGRATE_PFN_LOCKED; + migrate->cpages++; ptep_get_and_clear(mm, addr, ptep); /* Setup special migration page table entry */ @@ -2406,6 +2405,9 @@ again: if (pte_present(pte)) unmapped++; + } else { + put_page(page); + mpfn = 0; } next: @@ -2510,15 +2512,17 @@ static bool migrate_vma_check_page(struct page *page) } /* - * migrate_vma_prepare() - lock pages and isolate them from the lru + * migrate_vma_unmap() - replace page mapping with special migration pte entry * @migrate: migrate struct containing all migration information * - * This locks pages that have been collected by migrate_vma_collect(). Once each - * page is locked it is isolated from the lru (for non-device pages). Finally, - * the ref taken by migrate_vma_collect() is dropped, as locked pages cannot be - * migrated by concurrent kernel threads. + * Isolate pages from the LRU and replace mappings (CPU page table pte) with a + * special migration pte entry and check if it has been pinned. Pinned pages are + * restored because we cannot migrate them. + * + * This is the last step before we call the device driver callback to allocate + * destination memory and copy contents of original page over to new page. */ -static void migrate_vma_prepare(struct migrate_vma *migrate) +static void migrate_vma_unmap(struct migrate_vma *migrate) { const unsigned long npages = migrate->npages; const unsigned long start = migrate->start; @@ -2527,32 +2531,12 @@ static void migrate_vma_prepare(struct migrate_vma *migrate) lru_add_drain(); - for (i = 0; (i < npages) && migrate->cpages; i++) { + for (i = 0; i < npages; i++) { struct page *page = migrate_pfn_to_page(migrate->src[i]); - bool remap = true; if (!page) continue; - if (!(migrate->src[i] & MIGRATE_PFN_LOCKED)) { - /* - * Because we are migrating several pages there can be - * a deadlock between 2 concurrent migration where each - * are waiting on each other page lock. - * - * Make migrate_vma() a best effort thing and backoff - * for any page we can not lock right away. - */ - if (!trylock_page(page)) { - migrate->src[i] = 0; - migrate->cpages--; - put_page(page); - continue; - } - remap = false; - migrate->src[i] |= MIGRATE_PFN_LOCKED; - } - /* ZONE_DEVICE pages are not on LRU */ if (!is_zone_device_page(page)) { if (!PageLRU(page) && allow_drain) { @@ -2562,16 +2546,9 @@ static void migrate_vma_prepare(struct migrate_vma *migrate) } if (isolate_lru_page(page)) { - if (remap) { - migrate->src[i] &= ~MIGRATE_PFN_MIGRATE; - migrate->cpages--; - restore++; - } else { - migrate->src[i] = 0; - unlock_page(page); - migrate->cpages--; - put_page(page); - } + migrate->src[i] &= ~MIGRATE_PFN_MIGRATE; + migrate->cpages--; + restore++; continue; } @@ -2579,80 +2556,20 @@ static void migrate_vma_prepare(struct migrate_vma *migrate) put_page(page); } - if (!migrate_vma_check_page(page)) { - if (remap) { - migrate->src[i] &= ~MIGRATE_PFN_MIGRATE; - migrate->cpages--; - restore++; - - if (!is_zone_device_page(page)) { - get_page(page); - putback_lru_page(page); - } - } else { - migrate->src[i] = 0; - unlock_page(page); - migrate->cpages--; + if (page_mapped(page)) + try_to_migrate(page, 0); - if (!is_zone_device_page(page)) - putback_lru_page(page); - else - put_page(page); + if (page_mapped(page) || !migrate_vma_check_page(page)) { + if (!is_zone_device_page(page)) { + get_page(page); + putback_lru_page(page); } - } - } - - for (i = 0, addr = start; i < npages && restore; i++, addr += PAGE_SIZE) { - struct page *page = migrate_pfn_to_page(migrate->src[i]); - - if (!page || (migrate->src[i] & MIGRATE_PFN_MIGRATE)) - continue; - remove_migration_pte(page, migrate->vma, addr, page); - - migrate->src[i] = 0; - unlock_page(page); - put_page(page); - restore--; - } -} - -/* - * migrate_vma_unmap() - replace page mapping with special migration pte entry - * @migrate: migrate struct containing all migration information - * - * Replace page mapping (CPU page table pte) with a special migration pte entry - * and check again if it has been pinned. Pinned pages are restored because we - * cannot migrate them. - * - * This is the last step before we call the device driver callback to allocate - * destination memory and copy contents of original page over to new page. - */ -static void migrate_vma_unmap(struct migrate_vma *migrate) -{ - const unsigned long npages = migrate->npages; - const unsigned long start = migrate->start; - unsigned long addr, i, restore = 0; - - for (i = 0; i < npages; i++) { - struct page *page = migrate_pfn_to_page(migrate->src[i]); - - if (!page || !(migrate->src[i] & MIGRATE_PFN_MIGRATE)) + migrate->src[i] &= ~MIGRATE_PFN_MIGRATE; + migrate->cpages--; + restore++; continue; - - if (page_mapped(page)) { - try_to_migrate(page, 0); - if (page_mapped(page)) - goto restore; } - - if (migrate_vma_check_page(page)) - continue; - -restore: - migrate->src[i] &= ~MIGRATE_PFN_MIGRATE; - migrate->cpages--; - restore++; } for (addr = start, i = 0; i < npages && restore; addr += PAGE_SIZE, i++) { @@ -2665,12 +2582,8 @@ restore: migrate->src[i] = 0; unlock_page(page); + put_page(page); restore--; - - if (is_zone_device_page(page)) - put_page(page); - else - putback_lru_page(page); } } @@ -2693,8 +2606,8 @@ restore: * it for all those entries (ie with MIGRATE_PFN_VALID and MIGRATE_PFN_MIGRATE * flag set). Once these are allocated and copied, the caller must update each * corresponding entry in the dst array with the pfn value of the destination - * page and with the MIGRATE_PFN_VALID and MIGRATE_PFN_LOCKED flags set - * (destination pages must have their struct pages locked, via lock_page()). + * page and with MIGRATE_PFN_VALID. Destination pages must be locked via + * lock_page(). * * Note that the caller does not have to migrate all the pages that are marked * with MIGRATE_PFN_MIGRATE flag in src array unless this is a migration from @@ -2763,8 +2676,6 @@ int migrate_vma_setup(struct migrate_vma *args) migrate_vma_collect(args); - if (args->cpages) - migrate_vma_prepare(args); if (args->cpages) migrate_vma_unmap(args); -- cgit v1.2.3 From ebf7f6f0a6cdcc17a3da52b81e4b3a98c4005028 Mon Sep 17 00:00:00 2001 From: Tiezhu Yang Date: Fri, 5 Nov 2021 09:30:00 +0800 Subject: bpf: Change value of MAX_TAIL_CALL_CNT from 32 to 33 MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit In the current code, the actual max tail call count is 33 which is greater than MAX_TAIL_CALL_CNT (defined as 32). The actual limit is not consistent with the meaning of MAX_TAIL_CALL_CNT and thus confusing at first glance. We can see the historical evolution from commit 04fd61ab36ec ("bpf: allow bpf programs to tail-call other bpf programs") and commit f9dabe016b63 ("bpf: Undo off-by-one in interpreter tail call count limit"). In order to avoid changing existing behavior, the actual limit is 33 now, this is reasonable. After commit 874be05f525e ("bpf, tests: Add tail call test suite"), we can see there exists failed testcase. On all archs when CONFIG_BPF_JIT_ALWAYS_ON is not set: # echo 0 > /proc/sys/net/core/bpf_jit_enable # modprobe test_bpf # dmesg | grep -w FAIL Tail call error path, max count reached jited:0 ret 34 != 33 FAIL On some archs: # echo 1 > /proc/sys/net/core/bpf_jit_enable # modprobe test_bpf # dmesg | grep -w FAIL Tail call error path, max count reached jited:1 ret 34 != 33 FAIL Although the above failed testcase has been fixed in commit 18935a72eb25 ("bpf/tests: Fix error in tail call limit tests"), it would still be good to change the value of MAX_TAIL_CALL_CNT from 32 to 33 to make the code more readable. The 32-bit x86 JIT was using a limit of 32, just fix the wrong comments and limit to 33 tail calls as the constant MAX_TAIL_CALL_CNT updated. For the mips64 JIT, use "ori" instead of "addiu" as suggested by Johan Almbladh. For the riscv JIT, use RV_REG_TCC directly to save one register move as suggested by Björn Töpel. For the other implementations, no function changes, it does not change the current limit 33, the new value of MAX_TAIL_CALL_CNT can reflect the actual max tail call count, the related tail call testcases in test_bpf module and selftests can work well for the interpreter and the JIT. Here are the test results on x86_64: # uname -m x86_64 # echo 0 > /proc/sys/net/core/bpf_jit_enable # modprobe test_bpf test_suite=test_tail_calls # dmesg | tail -1 test_bpf: test_tail_calls: Summary: 8 PASSED, 0 FAILED, [0/8 JIT'ed] # rmmod test_bpf # echo 1 > /proc/sys/net/core/bpf_jit_enable # modprobe test_bpf test_suite=test_tail_calls # dmesg | tail -1 test_bpf: test_tail_calls: Summary: 8 PASSED, 0 FAILED, [8/8 JIT'ed] # rmmod test_bpf # ./test_progs -t tailcalls #142 tailcalls:OK Summary: 1/11 PASSED, 0 SKIPPED, 0 FAILED Signed-off-by: Tiezhu Yang Signed-off-by: Daniel Borkmann Tested-by: Johan Almbladh Tested-by: Ilya Leoshkevich Acked-by: Björn Töpel Acked-by: Johan Almbladh Acked-by: Ilya Leoshkevich Link: https://lore.kernel.org/bpf/1636075800-3264-1-git-send-email-yangtiezhu@loongson.cn --- arch/arm/net/bpf_jit_32.c | 5 +++-- arch/arm64/net/bpf_jit_comp.c | 5 +++-- arch/mips/net/bpf_jit_comp32.c | 3 +-- arch/mips/net/bpf_jit_comp64.c | 2 +- arch/powerpc/net/bpf_jit_comp32.c | 4 ++-- arch/powerpc/net/bpf_jit_comp64.c | 4 ++-- arch/riscv/net/bpf_jit_comp32.c | 6 ++---- arch/riscv/net/bpf_jit_comp64.c | 7 +++---- arch/s390/net/bpf_jit_comp.c | 6 +++--- arch/sparc/net/bpf_jit_comp_64.c | 2 +- arch/x86/net/bpf_jit_comp.c | 10 +++++----- arch/x86/net/bpf_jit_comp32.c | 4 ++-- include/linux/bpf.h | 2 +- include/uapi/linux/bpf.h | 2 +- kernel/bpf/core.c | 3 ++- lib/test_bpf.c | 4 ++-- tools/include/uapi/linux/bpf.h | 2 +- 17 files changed, 35 insertions(+), 36 deletions(-) (limited to 'lib') diff --git a/arch/arm/net/bpf_jit_32.c b/arch/arm/net/bpf_jit_32.c index eeb6dc0ecf46..e59b41e9ab0c 100644 --- a/arch/arm/net/bpf_jit_32.c +++ b/arch/arm/net/bpf_jit_32.c @@ -1199,7 +1199,8 @@ static int emit_bpf_tail_call(struct jit_ctx *ctx) /* tmp2[0] = array, tmp2[1] = index */ - /* if (tail_call_cnt > MAX_TAIL_CALL_CNT) + /* + * if (tail_call_cnt >= MAX_TAIL_CALL_CNT) * goto out; * tail_call_cnt++; */ @@ -1208,7 +1209,7 @@ static int emit_bpf_tail_call(struct jit_ctx *ctx) tc = arm_bpf_get_reg64(tcc, tmp, ctx); emit(ARM_CMP_I(tc[0], hi), ctx); _emit(ARM_COND_EQ, ARM_CMP_I(tc[1], lo), ctx); - _emit(ARM_COND_HI, ARM_B(jmp_offset), ctx); + _emit(ARM_COND_CS, ARM_B(jmp_offset), ctx); emit(ARM_ADDS_I(tc[1], tc[1], 1), ctx); emit(ARM_ADC_I(tc[0], tc[0], 0), ctx); arm_bpf_put_reg64(tcc, tmp, ctx); diff --git a/arch/arm64/net/bpf_jit_comp.c b/arch/arm64/net/bpf_jit_comp.c index 86c9dc0681cc..07c12c42b751 100644 --- a/arch/arm64/net/bpf_jit_comp.c +++ b/arch/arm64/net/bpf_jit_comp.c @@ -287,13 +287,14 @@ static int emit_bpf_tail_call(struct jit_ctx *ctx) emit(A64_CMP(0, r3, tmp), ctx); emit(A64_B_(A64_COND_CS, jmp_offset), ctx); - /* if (tail_call_cnt > MAX_TAIL_CALL_CNT) + /* + * if (tail_call_cnt >= MAX_TAIL_CALL_CNT) * goto out; * tail_call_cnt++; */ emit_a64_mov_i64(tmp, MAX_TAIL_CALL_CNT, ctx); emit(A64_CMP(1, tcc, tmp), ctx); - emit(A64_B_(A64_COND_HI, jmp_offset), ctx); + emit(A64_B_(A64_COND_CS, jmp_offset), ctx); emit(A64_ADD_I(1, tcc, tcc, 1), ctx); /* prog = array->ptrs[index]; diff --git a/arch/mips/net/bpf_jit_comp32.c b/arch/mips/net/bpf_jit_comp32.c index bd996ede12f8..044b11b65bca 100644 --- a/arch/mips/net/bpf_jit_comp32.c +++ b/arch/mips/net/bpf_jit_comp32.c @@ -1381,8 +1381,7 @@ void build_prologue(struct jit_context *ctx) * 16-byte area in the parent's stack frame. On a tail call, the * calling function jumps into the prologue after these instructions. */ - emit(ctx, ori, MIPS_R_T9, MIPS_R_ZERO, - min(MAX_TAIL_CALL_CNT + 1, 0xffff)); + emit(ctx, ori, MIPS_R_T9, MIPS_R_ZERO, min(MAX_TAIL_CALL_CNT, 0xffff)); emit(ctx, sw, MIPS_R_T9, 0, MIPS_R_SP); /* diff --git a/arch/mips/net/bpf_jit_comp64.c b/arch/mips/net/bpf_jit_comp64.c index 815ade724227..6475828ffb36 100644 --- a/arch/mips/net/bpf_jit_comp64.c +++ b/arch/mips/net/bpf_jit_comp64.c @@ -552,7 +552,7 @@ void build_prologue(struct jit_context *ctx) * On a tail call, the calling function jumps into the prologue * after this instruction. */ - emit(ctx, addiu, tc, MIPS_R_ZERO, min(MAX_TAIL_CALL_CNT + 1, 0xffff)); + emit(ctx, ori, tc, MIPS_R_ZERO, min(MAX_TAIL_CALL_CNT, 0xffff)); /* === Entry-point for tail calls === */ diff --git a/arch/powerpc/net/bpf_jit_comp32.c b/arch/powerpc/net/bpf_jit_comp32.c index 0da31d41d413..8a4faa05f9e4 100644 --- a/arch/powerpc/net/bpf_jit_comp32.c +++ b/arch/powerpc/net/bpf_jit_comp32.c @@ -221,13 +221,13 @@ static int bpf_jit_emit_tail_call(u32 *image, struct codegen_context *ctx, u32 o PPC_BCC(COND_GE, out); /* - * if (tail_call_cnt > MAX_TAIL_CALL_CNT) + * if (tail_call_cnt >= MAX_TAIL_CALL_CNT) * goto out; */ EMIT(PPC_RAW_CMPLWI(_R0, MAX_TAIL_CALL_CNT)); /* tail_call_cnt++; */ EMIT(PPC_RAW_ADDIC(_R0, _R0, 1)); - PPC_BCC(COND_GT, out); + PPC_BCC(COND_GE, out); /* prog = array->ptrs[index]; */ EMIT(PPC_RAW_RLWINM(_R3, b2p_index, 2, 0, 29)); diff --git a/arch/powerpc/net/bpf_jit_comp64.c b/arch/powerpc/net/bpf_jit_comp64.c index 8b5157ccfeba..8571aafcc9e1 100644 --- a/arch/powerpc/net/bpf_jit_comp64.c +++ b/arch/powerpc/net/bpf_jit_comp64.c @@ -228,12 +228,12 @@ static int bpf_jit_emit_tail_call(u32 *image, struct codegen_context *ctx, u32 o PPC_BCC(COND_GE, out); /* - * if (tail_call_cnt > MAX_TAIL_CALL_CNT) + * if (tail_call_cnt >= MAX_TAIL_CALL_CNT) * goto out; */ PPC_BPF_LL(b2p[TMP_REG_1], 1, bpf_jit_stack_tailcallcnt(ctx)); EMIT(PPC_RAW_CMPLWI(b2p[TMP_REG_1], MAX_TAIL_CALL_CNT)); - PPC_BCC(COND_GT, out); + PPC_BCC(COND_GE, out); /* * tail_call_cnt++; diff --git a/arch/riscv/net/bpf_jit_comp32.c b/arch/riscv/net/bpf_jit_comp32.c index e6497424cbf6..529a83b85c1c 100644 --- a/arch/riscv/net/bpf_jit_comp32.c +++ b/arch/riscv/net/bpf_jit_comp32.c @@ -799,11 +799,10 @@ static int emit_bpf_tail_call(int insn, struct rv_jit_context *ctx) emit_bcc(BPF_JGE, lo(idx_reg), RV_REG_T1, off, ctx); /* - * temp_tcc = tcc - 1; - * if (tcc < 0) + * if (--tcc < 0) * goto out; */ - emit(rv_addi(RV_REG_T1, RV_REG_TCC, -1), ctx); + emit(rv_addi(RV_REG_TCC, RV_REG_TCC, -1), ctx); off = ninsns_rvoff(tc_ninsn - (ctx->ninsns - start_insn)); emit_bcc(BPF_JSLT, RV_REG_TCC, RV_REG_ZERO, off, ctx); @@ -829,7 +828,6 @@ static int emit_bpf_tail_call(int insn, struct rv_jit_context *ctx) if (is_12b_check(off, insn)) return -1; emit(rv_lw(RV_REG_T0, off, RV_REG_T0), ctx); - emit(rv_addi(RV_REG_TCC, RV_REG_T1, 0), ctx); /* Epilogue jumps to *(t0 + 4). */ __build_epilogue(true, ctx); return 0; diff --git a/arch/riscv/net/bpf_jit_comp64.c b/arch/riscv/net/bpf_jit_comp64.c index f2a779c7e225..603630b6f3c5 100644 --- a/arch/riscv/net/bpf_jit_comp64.c +++ b/arch/riscv/net/bpf_jit_comp64.c @@ -327,12 +327,12 @@ static int emit_bpf_tail_call(int insn, struct rv_jit_context *ctx) off = ninsns_rvoff(tc_ninsn - (ctx->ninsns - start_insn)); emit_branch(BPF_JGE, RV_REG_A2, RV_REG_T1, off, ctx); - /* if (TCC-- < 0) + /* if (--TCC < 0) * goto out; */ - emit_addi(RV_REG_T1, tcc, -1, ctx); + emit_addi(RV_REG_TCC, tcc, -1, ctx); off = ninsns_rvoff(tc_ninsn - (ctx->ninsns - start_insn)); - emit_branch(BPF_JSLT, tcc, RV_REG_ZERO, off, ctx); + emit_branch(BPF_JSLT, RV_REG_TCC, RV_REG_ZERO, off, ctx); /* prog = array->ptrs[index]; * if (!prog) @@ -352,7 +352,6 @@ static int emit_bpf_tail_call(int insn, struct rv_jit_context *ctx) if (is_12b_check(off, insn)) return -1; emit_ld(RV_REG_T3, off, RV_REG_T2, ctx); - emit_mv(RV_REG_TCC, RV_REG_T1, ctx); __build_epilogue(true, ctx); return 0; } diff --git a/arch/s390/net/bpf_jit_comp.c b/arch/s390/net/bpf_jit_comp.c index 233cc9bcd652..9ff2bd83aad7 100644 --- a/arch/s390/net/bpf_jit_comp.c +++ b/arch/s390/net/bpf_jit_comp.c @@ -1369,7 +1369,7 @@ static noinline int bpf_jit_insn(struct bpf_jit *jit, struct bpf_prog *fp, jit->prg); /* - * if (tail_call_cnt++ > MAX_TAIL_CALL_CNT) + * if (tail_call_cnt++ >= MAX_TAIL_CALL_CNT) * goto out; */ @@ -1381,9 +1381,9 @@ static noinline int bpf_jit_insn(struct bpf_jit *jit, struct bpf_prog *fp, EMIT4_IMM(0xa7080000, REG_W0, 1); /* laal %w1,%w0,off(%r15) */ EMIT6_DISP_LH(0xeb000000, 0x00fa, REG_W1, REG_W0, REG_15, off); - /* clij %w1,MAX_TAIL_CALL_CNT,0x2,out */ + /* clij %w1,MAX_TAIL_CALL_CNT-1,0x2,out */ patch_2_clij = jit->prg; - EMIT6_PCREL_RIEC(0xec000000, 0x007f, REG_W1, MAX_TAIL_CALL_CNT, + EMIT6_PCREL_RIEC(0xec000000, 0x007f, REG_W1, MAX_TAIL_CALL_CNT - 1, 2, jit->prg); /* diff --git a/arch/sparc/net/bpf_jit_comp_64.c b/arch/sparc/net/bpf_jit_comp_64.c index 9a2f20cbd48b..0bfe1c72a0c9 100644 --- a/arch/sparc/net/bpf_jit_comp_64.c +++ b/arch/sparc/net/bpf_jit_comp_64.c @@ -867,7 +867,7 @@ static void emit_tail_call(struct jit_ctx *ctx) emit(LD32 | IMMED | RS1(SP) | S13(off) | RD(tmp), ctx); emit_cmpi(tmp, MAX_TAIL_CALL_CNT, ctx); #define OFFSET2 13 - emit_branch(BGU, ctx->idx, ctx->idx + OFFSET2, ctx); + emit_branch(BGEU, ctx->idx, ctx->idx + OFFSET2, ctx); emit_nop(ctx); emit_alu_K(ADD, tmp, 1, ctx); diff --git a/arch/x86/net/bpf_jit_comp.c b/arch/x86/net/bpf_jit_comp.c index 726700fabca6..631847907786 100644 --- a/arch/x86/net/bpf_jit_comp.c +++ b/arch/x86/net/bpf_jit_comp.c @@ -412,7 +412,7 @@ static void emit_indirect_jump(u8 **pprog, int reg, u8 *ip) * ... bpf_tail_call(void *ctx, struct bpf_array *array, u64 index) ... * if (index >= array->map.max_entries) * goto out; - * if (++tail_call_cnt > MAX_TAIL_CALL_CNT) + * if (tail_call_cnt++ >= MAX_TAIL_CALL_CNT) * goto out; * prog = array->ptrs[index]; * if (prog == NULL) @@ -446,14 +446,14 @@ static void emit_bpf_tail_call_indirect(u8 **pprog, bool *callee_regs_used, EMIT2(X86_JBE, offset); /* jbe out */ /* - * if (tail_call_cnt > MAX_TAIL_CALL_CNT) + * if (tail_call_cnt++ >= MAX_TAIL_CALL_CNT) * goto out; */ EMIT2_off32(0x8B, 0x85, tcc_off); /* mov eax, dword ptr [rbp - tcc_off] */ EMIT3(0x83, 0xF8, MAX_TAIL_CALL_CNT); /* cmp eax, MAX_TAIL_CALL_CNT */ offset = ctx->tail_call_indirect_label - (prog + 2 - start); - EMIT2(X86_JA, offset); /* ja out */ + EMIT2(X86_JAE, offset); /* jae out */ EMIT3(0x83, 0xC0, 0x01); /* add eax, 1 */ EMIT2_off32(0x89, 0x85, tcc_off); /* mov dword ptr [rbp - tcc_off], eax */ @@ -504,14 +504,14 @@ static void emit_bpf_tail_call_direct(struct bpf_jit_poke_descriptor *poke, int offset; /* - * if (tail_call_cnt > MAX_TAIL_CALL_CNT) + * if (tail_call_cnt++ >= MAX_TAIL_CALL_CNT) * goto out; */ EMIT2_off32(0x8B, 0x85, tcc_off); /* mov eax, dword ptr [rbp - tcc_off] */ EMIT3(0x83, 0xF8, MAX_TAIL_CALL_CNT); /* cmp eax, MAX_TAIL_CALL_CNT */ offset = ctx->tail_call_direct_label - (prog + 2 - start); - EMIT2(X86_JA, offset); /* ja out */ + EMIT2(X86_JAE, offset); /* jae out */ EMIT3(0x83, 0xC0, 0x01); /* add eax, 1 */ EMIT2_off32(0x89, 0x85, tcc_off); /* mov dword ptr [rbp - tcc_off], eax */ diff --git a/arch/x86/net/bpf_jit_comp32.c b/arch/x86/net/bpf_jit_comp32.c index da9b7cfa4632..429a89c5468b 100644 --- a/arch/x86/net/bpf_jit_comp32.c +++ b/arch/x86/net/bpf_jit_comp32.c @@ -1323,7 +1323,7 @@ static void emit_bpf_tail_call(u8 **pprog, u8 *ip) EMIT2(IA32_JBE, jmp_label(jmp_label1, 2)); /* - * if (tail_call_cnt > MAX_TAIL_CALL_CNT) + * if (tail_call_cnt++ >= MAX_TAIL_CALL_CNT) * goto out; */ lo = (u32)MAX_TAIL_CALL_CNT; @@ -1337,7 +1337,7 @@ static void emit_bpf_tail_call(u8 **pprog, u8 *ip) /* cmp ecx,lo */ EMIT3(0x83, add_1reg(0xF8, IA32_ECX), lo); - /* ja out */ + /* jae out */ EMIT2(IA32_JAE, jmp_label(jmp_label1, 2)); /* add eax,0x1 */ diff --git a/include/linux/bpf.h b/include/linux/bpf.h index 56098c866704..cc7a0c36e7df 100644 --- a/include/linux/bpf.h +++ b/include/linux/bpf.h @@ -1081,7 +1081,7 @@ struct bpf_array { }; #define BPF_COMPLEXITY_LIMIT_INSNS 1000000 /* yes. 1M insns */ -#define MAX_TAIL_CALL_CNT 32 +#define MAX_TAIL_CALL_CNT 33 #define BPF_F_ACCESS_MASK (BPF_F_RDONLY | \ BPF_F_RDONLY_PROG | \ diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h index 6297eafdc40f..a69e4b04ffeb 100644 --- a/include/uapi/linux/bpf.h +++ b/include/uapi/linux/bpf.h @@ -1744,7 +1744,7 @@ union bpf_attr { * if the maximum number of tail calls has been reached for this * chain of programs. This limit is defined in the kernel by the * macro **MAX_TAIL_CALL_CNT** (not accessible to user space), - * which is currently set to 32. + * which is currently set to 33. * Return * 0 on success, or a negative error in case of failure. * diff --git a/kernel/bpf/core.c b/kernel/bpf/core.c index 2405e39d800f..b52dc845ecea 100644 --- a/kernel/bpf/core.c +++ b/kernel/bpf/core.c @@ -1574,7 +1574,8 @@ select_insn: if (unlikely(index >= array->map.max_entries)) goto out; - if (unlikely(tail_call_cnt > MAX_TAIL_CALL_CNT)) + + if (unlikely(tail_call_cnt >= MAX_TAIL_CALL_CNT)) goto out; tail_call_cnt++; diff --git a/lib/test_bpf.c b/lib/test_bpf.c index adae39567264..0c5cb2d6436a 100644 --- a/lib/test_bpf.c +++ b/lib/test_bpf.c @@ -14683,7 +14683,7 @@ static struct tail_call_test tail_call_tests[] = { BPF_EXIT_INSN(), }, .flags = FLAG_NEED_STATE | FLAG_RESULT_IN_STATE, - .result = (MAX_TAIL_CALL_CNT + 1 + 1) * MAX_TESTRUNS, + .result = (MAX_TAIL_CALL_CNT + 1) * MAX_TESTRUNS, }, { "Tail call count preserved across function calls", @@ -14705,7 +14705,7 @@ static struct tail_call_test tail_call_tests[] = { }, .stack_depth = 8, .flags = FLAG_NEED_STATE | FLAG_RESULT_IN_STATE, - .result = (MAX_TAIL_CALL_CNT + 1 + 1) * MAX_TESTRUNS, + .result = (MAX_TAIL_CALL_CNT + 1) * MAX_TESTRUNS, }, { "Tail call error path, NULL target", diff --git a/tools/include/uapi/linux/bpf.h b/tools/include/uapi/linux/bpf.h index 6297eafdc40f..a69e4b04ffeb 100644 --- a/tools/include/uapi/linux/bpf.h +++ b/tools/include/uapi/linux/bpf.h @@ -1744,7 +1744,7 @@ union bpf_attr { * if the maximum number of tail calls has been reached for this * chain of programs. This limit is defined in the kernel by the * macro **MAX_TAIL_CALL_CNT** (not accessible to user space), - * which is currently set to 32. + * which is currently set to 33. * Return * 0 on success, or a negative error in case of failure. * -- cgit v1.2.3 From 418e0a3551bbef5b221705b0e5b8412cdc0afd39 Mon Sep 17 00:00:00 2001 From: Andy Shevchenko Date: Fri, 5 Nov 2021 14:42:24 +0200 Subject: lib/string_helpers: Introduce kasprintf_strarray() We have a few users already that basically want to have array of sequential strings to be allocated and filled. Provide a helper for them (basically adjusted version from gpio-mockup.c). Signed-off-by: Andy Shevchenko Reviewed-by: Linus Walleij --- include/linux/string_helpers.h | 1 + lib/string_helpers.c | 33 +++++++++++++++++++++++++++++++++ 2 files changed, 34 insertions(+) (limited to 'lib') diff --git a/include/linux/string_helpers.h b/include/linux/string_helpers.h index 4ba39e1403b2..f67a94013c87 100644 --- a/include/linux/string_helpers.h +++ b/include/linux/string_helpers.h @@ -100,6 +100,7 @@ char *kstrdup_quotable(const char *src, gfp_t gfp); char *kstrdup_quotable_cmdline(struct task_struct *task, gfp_t gfp); char *kstrdup_quotable_file(struct file *file, gfp_t gfp); +char **kasprintf_strarray(gfp_t gfp, const char *prefix, size_t n); void kfree_strarray(char **array, size_t n); #endif diff --git a/lib/string_helpers.c b/lib/string_helpers.c index d5d008f5b1d9..9758997c465e 100644 --- a/lib/string_helpers.c +++ b/lib/string_helpers.c @@ -674,6 +674,39 @@ char *kstrdup_quotable_file(struct file *file, gfp_t gfp) } EXPORT_SYMBOL_GPL(kstrdup_quotable_file); +/** + * kasprintf_strarray - allocate and fill array of sequential strings + * @gfp: flags for the slab allocator + * @prefix: prefix to be used + * @n: amount of lines to be allocated and filled + * + * Allocates and fills @n strings using pattern "%s-%zu", where prefix + * is provided by caller. The caller is responsible to free them with + * kfree_strarray() after use. + * + * Returns array of strings or NULL when memory can't be allocated. + */ +char **kasprintf_strarray(gfp_t gfp, const char *prefix, size_t n) +{ + char **names; + size_t i; + + names = kcalloc(n + 1, sizeof(char *), gfp); + if (!names) + return NULL; + + for (i = 0; i < n; i++) { + names[i] = kasprintf(gfp, "%s-%zu", prefix, i); + if (!names[i]) { + kfree_strarray(names, i); + return NULL; + } + } + + return names; +} +EXPORT_SYMBOL_GPL(kasprintf_strarray); + /** * kfree_strarray - free a number of dynamically allocated strings contained * in an array and the array itself -- cgit v1.2.3 From acdb89b6c87a2d7b5c48a82756e6f5c6f599f60a Mon Sep 17 00:00:00 2001 From: Andy Shevchenko Date: Fri, 5 Nov 2021 14:42:25 +0200 Subject: lib/string_helpers: Introduce managed variant of kasprintf_strarray() Some of the users want to have easy way to allocate array of strings that will be automatically cleaned when associated device is gone. Introduce managed variant of kasprintf_strarray() for such use cases. Signed-off-by: Andy Shevchenko --- include/linux/string_helpers.h | 3 +++ lib/string_helpers.c | 31 +++++++++++++++++++++++++++++++ 2 files changed, 34 insertions(+) (limited to 'lib') diff --git a/include/linux/string_helpers.h b/include/linux/string_helpers.h index f67a94013c87..7a22921c9db7 100644 --- a/include/linux/string_helpers.h +++ b/include/linux/string_helpers.h @@ -7,6 +7,7 @@ #include #include +struct device; struct file; struct task_struct; @@ -103,4 +104,6 @@ char *kstrdup_quotable_file(struct file *file, gfp_t gfp); char **kasprintf_strarray(gfp_t gfp, const char *prefix, size_t n); void kfree_strarray(char **array, size_t n); +char **devm_kasprintf_strarray(struct device *dev, const char *prefix, size_t n); + #endif diff --git a/lib/string_helpers.c b/lib/string_helpers.c index 9758997c465e..90f9f1b7afec 100644 --- a/lib/string_helpers.c +++ b/lib/string_helpers.c @@ -10,6 +10,7 @@ #include #include #include +#include #include #include #include @@ -730,6 +731,36 @@ void kfree_strarray(char **array, size_t n) } EXPORT_SYMBOL_GPL(kfree_strarray); +struct strarray { + char **array; + size_t n; +}; + +static void devm_kfree_strarray(struct device *dev, void *res) +{ + struct strarray *array = res; + + kfree_strarray(array->array, array->n); +} + +char **devm_kasprintf_strarray(struct device *dev, const char *prefix, size_t n) +{ + struct strarray *ptr; + + ptr = devres_alloc(devm_kfree_strarray, sizeof(*ptr), GFP_KERNEL); + if (!ptr) + return ERR_PTR(-ENOMEM); + + ptr->array = kasprintf_strarray(GFP_KERNEL, prefix, n); + if (!ptr->array) { + devres_free(ptr); + return ERR_PTR(-ENOMEM); + } + + return ptr->array; +} +EXPORT_SYMBOL_GPL(devm_kasprintf_strarray); + /** * strscpy_pad() - Copy a C-string into a sized buffer * @dest: Where to copy the string to -- cgit v1.2.3 From ae8d67b2117f1ec6c8170d6e1af8ded17392bd2c Mon Sep 17 00:00:00 2001 From: Nick Terrell Date: Mon, 15 Nov 2021 19:08:19 -0800 Subject: lib: zstd: Fix unused variable warning The variable `litLengthSum` is only used by an `assert()`, so when asserts are disabled the compiler doesn't see any usage and warns. This issue is already fixed upstream by PR #2838 [0]. It was reported by the Kernel test robot in [1]. Another approach would be to change zstd's disabled `assert()` definition to use the argument in a disabled branch, instead of ignoring the argument. I've avoided this approach because there are some small changes necessary to get zstd to build, and I would want to thoroughly re-test for performance, since that is slightly changing the code in every function in zstd. It seems like a trivial change, but some functions are pretty sensitive to small changes. However, I think it is a valid approach that I would like to see upstream take, so I've opened Issue #2868 to attempt this upstream. Lastly, I've chosen not to use __maybe_unused because all code in lib/zstd/ must eventually be upstreamed. Upstream zstd can't use __maybe_unused because it isn't portable across all compilers. [0] https://github.com/facebook/zstd/pull/2838 [1] https://lore.kernel.org/linux-mm/202111120312.833wII4i-lkp@intel.com/T/ [2] https://github.com/facebook/zstd/issues/2868 Link: https://lore.kernel.org/r/20211117014949.1169186-2-nickrterrell@gmail.com/ Link: https://lore.kernel.org/r/20211117201459.1194876-2-nickrterrell@gmail.com/ Reported-by: kernel test robot Signed-off-by: Nick Terrell --- lib/zstd/compress/zstd_compress_superblock.c | 2 ++ 1 file changed, 2 insertions(+) (limited to 'lib') diff --git a/lib/zstd/compress/zstd_compress_superblock.c b/lib/zstd/compress/zstd_compress_superblock.c index ee03e0aedb03..b0610b255653 100644 --- a/lib/zstd/compress/zstd_compress_superblock.c +++ b/lib/zstd/compress/zstd_compress_superblock.c @@ -411,6 +411,8 @@ static size_t ZSTD_seqDecompressedSize(seqStore_t const* seqStore, const seqDef* const seqDef* sp = sstart; size_t matchLengthSum = 0; size_t litLengthSum = 0; + /* Only used by assert(), suppress unused variable warnings in production. */ + (void)litLengthSum; while (send-sp > 0) { ZSTD_sequenceLength const seqLen = ZSTD_getSequenceLength(seqStore, sp); litLengthSum += seqLen.litLength; -- cgit v1.2.3 From 1974990cca43a6ba708a70b15862113eb9c2f399 Mon Sep 17 00:00:00 2001 From: Nick Terrell Date: Mon, 15 Nov 2021 20:33:08 -0800 Subject: lib: zstd: Don't inline functions in zstd_opt.c `zstd_opt.c` contains the match finder for the highest compression levels. These levels are already very slow, and are unlikely to be used in the kernel. If they are used, they shouldn't be used in latency sensitive workloads, so slowing them down shouldn't be a big deal. This saves 188 KB of the 288 KB regression reported by Geert Uytterhoeven [0]. I've also opened an issue upstream [1] so that we can properly tackle the code size issue in `zstd_opt.c` for all users, and can hopefully remove this hack in the next zstd version we import. Bloat-o-meter output on x86-64: ``` > ../scripts/bloat-o-meter vmlinux.old vmlinux add/remove: 6/5 grow/shrink: 1/9 up/down: 16673/-209939 (-193266) Function old new delta ZSTD_compressBlock_opt_generic.constprop - 7559 +7559 ZSTD_insertBtAndGetAllMatches - 6304 +6304 ZSTD_insertBt1 - 1731 +1731 ZSTD_storeSeq - 693 +693 ZSTD_BtGetAllMatches - 255 +255 ZSTD_updateRep - 128 +128 ZSTD_updateTree 96 99 +3 ZSTD_insertAndFindFirstIndexHash3 81 - -81 ZSTD_setBasePrices.constprop 98 - -98 ZSTD_litLengthPrice.constprop 138 - -138 ZSTD_count 362 181 -181 ZSTD_count_2segments 1407 938 -469 ZSTD_insertBt1.constprop 2689 - -2689 ZSTD_compressBlock_btultra2 19990 423 -19567 ZSTD_compressBlock_btultra 19633 15 -19618 ZSTD_initStats_ultra 19825 - -19825 ZSTD_compressBlock_btopt 20374 12 -20362 ZSTD_compressBlock_btopt_extDict 29984 12 -29972 ZSTD_compressBlock_btultra_extDict 30718 15 -30703 ZSTD_compressBlock_btopt_dictMatchState 32689 12 -32677 ZSTD_compressBlock_btultra_dictMatchState 33574 15 -33559 Total: Before=6611828, After=6418562, chg -2.92% ``` [0] https://lkml.org/lkml/2021/11/14/189 [1] https://github.com/facebook/zstd/issues/2862 Link: https://lore.kernel.org/r/20211117014949.1169186-3-nickrterrell@gmail.com/ Link: https://lore.kernel.org/r/20211117201459.1194876-3-nickrterrell@gmail.com/ Reported-by: Geert Uytterhoeven Tested-by: Geert Uytterhoeven Reviewed-by: Geert Uytterhoeven Signed-off-by: Nick Terrell --- lib/zstd/common/compiler.h | 7 +++++++ lib/zstd/compress/zstd_opt.c | 12 ++++++++++++ 2 files changed, 19 insertions(+) (limited to 'lib') diff --git a/lib/zstd/common/compiler.h b/lib/zstd/common/compiler.h index a1a051e4bce6..f5a9c70a228a 100644 --- a/lib/zstd/common/compiler.h +++ b/lib/zstd/common/compiler.h @@ -16,6 +16,7 @@ *********************************************************/ /* force inlining */ +#if !defined(ZSTD_NO_INLINE) #if (defined(__GNUC__) && !defined(__STRICT_ANSI__)) || defined(__cplusplus) || defined(__STDC_VERSION__) && __STDC_VERSION__ >= 199901L /* C99 */ # define INLINE_KEYWORD inline #else @@ -24,6 +25,12 @@ #define FORCE_INLINE_ATTR __attribute__((always_inline)) +#else + +#define INLINE_KEYWORD +#define FORCE_INLINE_ATTR + +#endif /* On MSVC qsort requires that functions passed into it use the __cdecl calling conversion(CC). diff --git a/lib/zstd/compress/zstd_opt.c b/lib/zstd/compress/zstd_opt.c index 04337050fe9a..dfc55e3e8119 100644 --- a/lib/zstd/compress/zstd_opt.c +++ b/lib/zstd/compress/zstd_opt.c @@ -8,6 +8,18 @@ * You may select, at your option, one of the above-listed licenses. */ +/* + * Disable inlining for the optimal parser for the kernel build. + * It is unlikely to be used in the kernel, and where it is used + * latency shouldn't matter because it is very slow to begin with. + * We prefer a ~180KB binary size win over faster optimal parsing. + * + * TODO(https://github.com/facebook/zstd/issues/2862): + * Improve the code size of the optimal parser in general, so we + * don't need this hack for the kernel build. + */ +#define ZSTD_NO_INLINE 1 + #include "zstd_compress_internal.h" #include "hist.h" #include "zstd_opt.h" -- cgit v1.2.3 From 7416cdc9b9c10968c57b1f73be5d48b3ecdaf3c8 Mon Sep 17 00:00:00 2001 From: Nick Terrell Date: Tue, 16 Nov 2021 15:11:39 -0800 Subject: lib: zstd: Don't add -O3 to cflags After the update to zstd-1.4.10 passing -O3 is no longer necessary to get good performance from zstd. Using the default optimization level -O2 is sufficient to get good performance. I've measured no significant change to compression speed, and a ~1% decompression speed loss, which is acceptable. This fixes the reported parisc -Wframe-larger-than=1536 errors [0]. The gcc-8-hppa-linux-gnu compiler performed very poorly with -O3, generating stacks that are ~3KB. With -O2 these same functions generate stacks in the < 100B, completely fixing the problem. Function size deltas are listed below: ZSTD_compressBlock_fast_extDict_generic: 3800 -> 68 ZSTD_compressBlock_fast: 2216 -> 40 ZSTD_compressBlock_fast_dictMatchState: 1848 -> 64 ZSTD_compressBlock_doubleFast_extDict_generic: 3744 -> 76 ZSTD_fillDoubleHashTable: 3252 -> 0 ZSTD_compressBlock_doubleFast: 5856 -> 36 ZSTD_compressBlock_doubleFast_dictMatchState: 5380 -> 84 ZSTD_copmressBlock_lazy2: 2420 -> 72 Additionally, this improves the reported code bloat [1]. With gcc-11 bloat-o-meter shows an 80KB code size improvement: ``` > ../scripts/bloat-o-meter vmlinux.old vmlinux add/remove: 31/8 grow/shrink: 24/155 up/down: 25734/-107924 (-82190) Total: Before=6418562, After=6336372, chg -1.28% ``` Compared to before the zstd-1.4.10 update we see a total code size regression of 105KB, down from 374KB at v5.16-rc1: ``` > ../scripts/bloat-o-meter vmlinux.old vmlinux add/remove: 292/62 grow/shrink: 56/88 up/down: 235009/-127487 (107522) Total: Before=6228850, After=6336372, chg +1.73% ``` [0] https://lkml.org/lkml/2021/11/15/710 [1] https://lkml.org/lkml/2021/11/14/189 Link: https://lore.kernel.org/r/20211117014949.1169186-4-nickrterrell@gmail.com/ Link: https://lore.kernel.org/r/20211117201459.1194876-4-nickrterrell@gmail.com/ Reported-by: Geert Uytterhoeven Tested-by: Geert Uytterhoeven Reviewed-by: Geert Uytterhoeven Signed-off-by: Nick Terrell --- lib/zstd/Makefile | 2 -- 1 file changed, 2 deletions(-) (limited to 'lib') diff --git a/lib/zstd/Makefile b/lib/zstd/Makefile index 65218ec5b8f2..fc45339fc3a3 100644 --- a/lib/zstd/Makefile +++ b/lib/zstd/Makefile @@ -11,8 +11,6 @@ obj-$(CONFIG_ZSTD_COMPRESS) += zstd_compress.o obj-$(CONFIG_ZSTD_DECOMPRESS) += zstd_decompress.o -ccflags-y += -O3 - zstd_compress-y := \ zstd_compress_module.o \ common/debug.o \ -- cgit v1.2.3 From cab71f7495f7aa639ca4b8508f4c3e426e9cb2f7 Mon Sep 17 00:00:00 2001 From: Kees Cook Date: Fri, 19 Nov 2021 16:43:46 -0800 Subject: kasan: test: silence intentional read overflow warnings As done in commit d73dad4eb5ad ("kasan: test: bypass __alloc_size checks") for __write_overflow warnings, also silence some more cases that trip the __read_overflow warnings seen in 5.16-rc1[1]: In file included from include/linux/string.h:253, from include/linux/bitmap.h:10, from include/linux/cpumask.h:12, from include/linux/mm_types_task.h:14, from include/linux/mm_types.h:5, from include/linux/page-flags.h:13, from arch/arm64/include/asm/mte.h:14, from arch/arm64/include/asm/pgtable.h:12, from include/linux/pgtable.h:6, from include/linux/kasan.h:29, from lib/test_kasan.c:10: In function 'memcmp', inlined from 'kasan_memcmp' at lib/test_kasan.c:897:2: include/linux/fortify-string.h:263:25: error: call to '__read_overflow' declared with attribute error: detected read beyond size of object (1st parameter) 263 | __read_overflow(); | ^~~~~~~~~~~~~~~~~ In function 'memchr', inlined from 'kasan_memchr' at lib/test_kasan.c:872:2: include/linux/fortify-string.h:277:17: error: call to '__read_overflow' declared with attribute error: detected read beyond size of object (1st parameter) 277 | __read_overflow(); | ^~~~~~~~~~~~~~~~~ [1] http://kisskb.ellerman.id.au/kisskb/buildresult/14660585/log/ Link: https://lkml.kernel.org/r/20211116004111.3171781-1-keescook@chromium.org Fixes: d73dad4eb5ad ("kasan: test: bypass __alloc_size checks") Signed-off-by: Kees Cook Reviewed-by: Andrey Konovalov Acked-by: Marco Elver Cc: Andrey Ryabinin Cc: Alexander Potapenko Cc: Dmitry Vyukov Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- lib/test_kasan.c | 2 ++ 1 file changed, 2 insertions(+) (limited to 'lib') diff --git a/lib/test_kasan.c b/lib/test_kasan.c index 67ed689a0b1b..0643573f8686 100644 --- a/lib/test_kasan.c +++ b/lib/test_kasan.c @@ -869,6 +869,7 @@ static void kasan_memchr(struct kunit *test) ptr = kmalloc(size, GFP_KERNEL | __GFP_ZERO); KUNIT_ASSERT_NOT_ERR_OR_NULL(test, ptr); + OPTIMIZER_HIDE_VAR(size); KUNIT_EXPECT_KASAN_FAIL(test, kasan_ptr_result = memchr(ptr, '1', size + 1)); @@ -894,6 +895,7 @@ static void kasan_memcmp(struct kunit *test) KUNIT_ASSERT_NOT_ERR_OR_NULL(test, ptr); memset(arr, 0, sizeof(arr)); + OPTIMIZER_HIDE_VAR(size); KUNIT_EXPECT_KASAN_FAIL(test, kasan_int_result = memcmp(ptr, arr, size+1)); kfree(ptr); -- cgit v1.2.3 From 8d192bec534bd5b778135769a12e5f04580771f7 Mon Sep 17 00:00:00 2001 From: Helge Deller Date: Fri, 19 Nov 2021 22:31:03 +0100 Subject: parisc: Increase FRAME_WARN to 2048 bytes on parisc PA-RISC uses a much bigger frame size for functions than other architectures. So increase it to 2048 for 32- and 64-bit kernels. This fixes e.g. a warning in lib/xxhash.c. Reported-by: kernel test robot Signed-off-by: Helge Deller --- lib/Kconfig.debug | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) (limited to 'lib') diff --git a/lib/Kconfig.debug b/lib/Kconfig.debug index 9ef7ce18b4f5..5c12bde10996 100644 --- a/lib/Kconfig.debug +++ b/lib/Kconfig.debug @@ -346,8 +346,9 @@ config FRAME_WARN int "Warn for stack frames larger than" range 0 8192 default 2048 if GCC_PLUGIN_LATENT_ENTROPY - default 1536 if (!64BIT && (PARISC || XTENSA)) - default 1024 if (!64BIT && !PARISC) + default 2048 if PARISC + default 1536 if (!64BIT && XTENSA) + default 1024 if !64BIT default 2048 if 64BIT help Tell gcc to warn at build time for stack frames larger than this. -- cgit v1.2.3 From f7e5b9bfa6c8820407b64eabc1f29c9a87e8993d Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Mon, 29 Nov 2021 10:39:29 -0500 Subject: siphash: use _unaligned version by default On ARM v6 and later, we define CONFIG_HAVE_EFFICIENT_UNALIGNED_ACCESS because the ordinary load/store instructions (ldr, ldrh, ldrb) can tolerate any misalignment of the memory address. However, load/store double and load/store multiple instructions (ldrd, ldm) may still only be used on memory addresses that are 32-bit aligned, and so we have to use the CONFIG_HAVE_EFFICIENT_UNALIGNED_ACCESS macro with care, or we may end up with a severe performance hit due to alignment traps that require fixups by the kernel. Testing shows that this currently happens with clang-13 but not gcc-11. In theory, any compiler version can produce this bug or other problems, as we are dealing with undefined behavior in C99 even on architectures that support this in hardware, see also https://gcc.gnu.org/bugzilla/show_bug.cgi?id=100363. Fortunately, the get_unaligned() accessors do the right thing: when building for ARMv6 or later, the compiler will emit unaligned accesses using the ordinary load/store instructions (but avoid the ones that require 32-bit alignment). When building for older ARM, those accessors will emit the appropriate sequence of ldrb/mov/orr instructions. And on architectures that can truly tolerate any kind of misalignment, the get_unaligned() accessors resolve to the leXX_to_cpup accessors that operate on aligned addresses. Since the compiler will in fact emit ldrd or ldm instructions when building this code for ARM v6 or later, the solution is to use the unaligned accessors unconditionally on architectures where this is known to be fast. The _aligned version of the hash function is however still needed to get the best performance on architectures that cannot do any unaligned access in hardware. This new version avoids the undefined behavior and should produce the fastest hash on all architectures we support. Link: https://lore.kernel.org/linux-arm-kernel/20181008211554.5355-4-ard.biesheuvel@linaro.org/ Link: https://lore.kernel.org/linux-crypto/CAK8P3a2KfmmGDbVHULWevB0hv71P2oi2ZCHEAqT=8dQfa0=cqQ@mail.gmail.com/ Reported-by: Ard Biesheuvel Fixes: 2c956a60778c ("siphash: add cryptographically secure PRF") Signed-off-by: Arnd Bergmann Reviewed-by: Jason A. Donenfeld Acked-by: Ard Biesheuvel Signed-off-by: Jason A. Donenfeld Signed-off-by: Jakub Kicinski --- include/linux/siphash.h | 14 ++++---------- lib/siphash.c | 12 ++++++------ 2 files changed, 10 insertions(+), 16 deletions(-) (limited to 'lib') diff --git a/include/linux/siphash.h b/include/linux/siphash.h index bf21591a9e5e..0cda61855d90 100644 --- a/include/linux/siphash.h +++ b/include/linux/siphash.h @@ -27,9 +27,7 @@ static inline bool siphash_key_is_zero(const siphash_key_t *key) } u64 __siphash_aligned(const void *data, size_t len, const siphash_key_t *key); -#ifndef CONFIG_HAVE_EFFICIENT_UNALIGNED_ACCESS u64 __siphash_unaligned(const void *data, size_t len, const siphash_key_t *key); -#endif u64 siphash_1u64(const u64 a, const siphash_key_t *key); u64 siphash_2u64(const u64 a, const u64 b, const siphash_key_t *key); @@ -82,10 +80,9 @@ static inline u64 ___siphash_aligned(const __le64 *data, size_t len, static inline u64 siphash(const void *data, size_t len, const siphash_key_t *key) { -#ifndef CONFIG_HAVE_EFFICIENT_UNALIGNED_ACCESS - if (!IS_ALIGNED((unsigned long)data, SIPHASH_ALIGNMENT)) + if (IS_ENABLED(CONFIG_HAVE_EFFICIENT_UNALIGNED_ACCESS) || + !IS_ALIGNED((unsigned long)data, SIPHASH_ALIGNMENT)) return __siphash_unaligned(data, len, key); -#endif return ___siphash_aligned(data, len, key); } @@ -96,10 +93,8 @@ typedef struct { u32 __hsiphash_aligned(const void *data, size_t len, const hsiphash_key_t *key); -#ifndef CONFIG_HAVE_EFFICIENT_UNALIGNED_ACCESS u32 __hsiphash_unaligned(const void *data, size_t len, const hsiphash_key_t *key); -#endif u32 hsiphash_1u32(const u32 a, const hsiphash_key_t *key); u32 hsiphash_2u32(const u32 a, const u32 b, const hsiphash_key_t *key); @@ -135,10 +130,9 @@ static inline u32 ___hsiphash_aligned(const __le32 *data, size_t len, static inline u32 hsiphash(const void *data, size_t len, const hsiphash_key_t *key) { -#ifndef CONFIG_HAVE_EFFICIENT_UNALIGNED_ACCESS - if (!IS_ALIGNED((unsigned long)data, HSIPHASH_ALIGNMENT)) + if (IS_ENABLED(CONFIG_HAVE_EFFICIENT_UNALIGNED_ACCESS) || + !IS_ALIGNED((unsigned long)data, HSIPHASH_ALIGNMENT)) return __hsiphash_unaligned(data, len, key); -#endif return ___hsiphash_aligned(data, len, key); } diff --git a/lib/siphash.c b/lib/siphash.c index a90112ee72a1..72b9068ab57b 100644 --- a/lib/siphash.c +++ b/lib/siphash.c @@ -49,6 +49,7 @@ SIPROUND; \ return (v0 ^ v1) ^ (v2 ^ v3); +#ifndef CONFIG_HAVE_EFFICIENT_UNALIGNED_ACCESS u64 __siphash_aligned(const void *data, size_t len, const siphash_key_t *key) { const u8 *end = data + len - (len % sizeof(u64)); @@ -80,8 +81,8 @@ u64 __siphash_aligned(const void *data, size_t len, const siphash_key_t *key) POSTAMBLE } EXPORT_SYMBOL(__siphash_aligned); +#endif -#ifndef CONFIG_HAVE_EFFICIENT_UNALIGNED_ACCESS u64 __siphash_unaligned(const void *data, size_t len, const siphash_key_t *key) { const u8 *end = data + len - (len % sizeof(u64)); @@ -113,7 +114,6 @@ u64 __siphash_unaligned(const void *data, size_t len, const siphash_key_t *key) POSTAMBLE } EXPORT_SYMBOL(__siphash_unaligned); -#endif /** * siphash_1u64 - compute 64-bit siphash PRF value of a u64 @@ -250,6 +250,7 @@ EXPORT_SYMBOL(siphash_3u32); HSIPROUND; \ return (v0 ^ v1) ^ (v2 ^ v3); +#ifndef CONFIG_HAVE_EFFICIENT_UNALIGNED_ACCESS u32 __hsiphash_aligned(const void *data, size_t len, const hsiphash_key_t *key) { const u8 *end = data + len - (len % sizeof(u64)); @@ -280,8 +281,8 @@ u32 __hsiphash_aligned(const void *data, size_t len, const hsiphash_key_t *key) HPOSTAMBLE } EXPORT_SYMBOL(__hsiphash_aligned); +#endif -#ifndef CONFIG_HAVE_EFFICIENT_UNALIGNED_ACCESS u32 __hsiphash_unaligned(const void *data, size_t len, const hsiphash_key_t *key) { @@ -313,7 +314,6 @@ u32 __hsiphash_unaligned(const void *data, size_t len, HPOSTAMBLE } EXPORT_SYMBOL(__hsiphash_unaligned); -#endif /** * hsiphash_1u32 - compute 64-bit hsiphash PRF value of a u32 @@ -418,6 +418,7 @@ EXPORT_SYMBOL(hsiphash_4u32); HSIPROUND; \ return v1 ^ v3; +#ifndef CONFIG_HAVE_EFFICIENT_UNALIGNED_ACCESS u32 __hsiphash_aligned(const void *data, size_t len, const hsiphash_key_t *key) { const u8 *end = data + len - (len % sizeof(u32)); @@ -438,8 +439,8 @@ u32 __hsiphash_aligned(const void *data, size_t len, const hsiphash_key_t *key) HPOSTAMBLE } EXPORT_SYMBOL(__hsiphash_aligned); +#endif -#ifndef CONFIG_HAVE_EFFICIENT_UNALIGNED_ACCESS u32 __hsiphash_unaligned(const void *data, size_t len, const hsiphash_key_t *key) { @@ -461,7 +462,6 @@ u32 __hsiphash_unaligned(const void *data, size_t len, HPOSTAMBLE } EXPORT_SYMBOL(__hsiphash_unaligned); -#endif /** * hsiphash_1u32 - compute 32-bit hsiphash PRF value of a u32 -- cgit v1.2.3 From d9847eb8be3d895b2b5f514fdf3885d47a0b92a2 Mon Sep 17 00:00:00 2001 From: Kumar Kartikeya Dwivedi Date: Mon, 22 Nov 2021 20:17:40 +0530 Subject: bpf: Make CONFIG_DEBUG_INFO_BTF depend upon CONFIG_BPF_SYSCALL Vinicius Costa Gomes reported [0] that build fails when CONFIG_DEBUG_INFO_BTF is enabled and CONFIG_BPF_SYSCALL is disabled. This leads to btf.c not being compiled, and then no symbol being present in vmlinux for the declarations in btf.h. Since BTF is not useful without enabling BPF subsystem, disallow this combination. However, theoretically disabling both now could still fail, as the symbol for kfunc_btf_id_list variables is not available. This isn't a problem as the compiler usually optimizes the whole register/unregister call, but at lower optimization levels it can fail the build in linking stage. Fix that by adding dummy variables so that modules taking address of them still work, but the whole thing is a noop. [0]: https://lore.kernel.org/bpf/20211110205418.332403-1-vinicius.gomes@intel.com Fixes: 14f267d95fe4 ("bpf: btf: Introduce helpers for dynamic BTF set registration") Reported-by: Vinicius Costa Gomes Signed-off-by: Kumar Kartikeya Dwivedi Signed-off-by: Andrii Nakryiko Acked-by: Song Liu Link: https://lore.kernel.org/bpf/20211122144742.477787-2-memxor@gmail.com --- include/linux/btf.h | 14 ++++++++++---- kernel/bpf/btf.c | 9 ++------- lib/Kconfig.debug | 1 + 3 files changed, 13 insertions(+), 11 deletions(-) (limited to 'lib') diff --git a/include/linux/btf.h b/include/linux/btf.h index 203eef993d76..0e1b6281fd8f 100644 --- a/include/linux/btf.h +++ b/include/linux/btf.h @@ -245,7 +245,10 @@ struct kfunc_btf_id_set { struct module *owner; }; -struct kfunc_btf_id_list; +struct kfunc_btf_id_list { + struct list_head list; + struct mutex mutex; +}; #ifdef CONFIG_DEBUG_INFO_BTF_MODULES void register_kfunc_btf_id_set(struct kfunc_btf_id_list *l, @@ -254,6 +257,9 @@ void unregister_kfunc_btf_id_set(struct kfunc_btf_id_list *l, struct kfunc_btf_id_set *s); bool bpf_check_mod_kfunc_call(struct kfunc_btf_id_list *klist, u32 kfunc_id, struct module *owner); + +extern struct kfunc_btf_id_list bpf_tcp_ca_kfunc_list; +extern struct kfunc_btf_id_list prog_test_kfunc_list; #else static inline void register_kfunc_btf_id_set(struct kfunc_btf_id_list *l, struct kfunc_btf_id_set *s) @@ -268,13 +274,13 @@ static inline bool bpf_check_mod_kfunc_call(struct kfunc_btf_id_list *klist, { return false; } + +static struct kfunc_btf_id_list bpf_tcp_ca_kfunc_list __maybe_unused; +static struct kfunc_btf_id_list prog_test_kfunc_list __maybe_unused; #endif #define DEFINE_KFUNC_BTF_ID_SET(set, name) \ struct kfunc_btf_id_set name = { LIST_HEAD_INIT(name.list), (set), \ THIS_MODULE } -extern struct kfunc_btf_id_list bpf_tcp_ca_kfunc_list; -extern struct kfunc_btf_id_list prog_test_kfunc_list; - #endif diff --git a/kernel/bpf/btf.c b/kernel/bpf/btf.c index dbc3ad07e21b..ea3df9867cec 100644 --- a/kernel/bpf/btf.c +++ b/kernel/bpf/btf.c @@ -6346,11 +6346,6 @@ BTF_ID_LIST_GLOBAL_SINGLE(btf_task_struct_ids, struct, task_struct) /* BTF ID set registration API for modules */ -struct kfunc_btf_id_list { - struct list_head list; - struct mutex mutex; -}; - #ifdef CONFIG_DEBUG_INFO_BTF_MODULES void register_kfunc_btf_id_set(struct kfunc_btf_id_list *l, @@ -6389,8 +6384,6 @@ bool bpf_check_mod_kfunc_call(struct kfunc_btf_id_list *klist, u32 kfunc_id, return false; } -#endif - #define DEFINE_KFUNC_BTF_ID_LIST(name) \ struct kfunc_btf_id_list name = { LIST_HEAD_INIT(name.list), \ __MUTEX_INITIALIZER(name.mutex) }; \ @@ -6398,3 +6391,5 @@ bool bpf_check_mod_kfunc_call(struct kfunc_btf_id_list *klist, u32 kfunc_id, DEFINE_KFUNC_BTF_ID_LIST(bpf_tcp_ca_kfunc_list); DEFINE_KFUNC_BTF_ID_LIST(prog_test_kfunc_list); + +#endif diff --git a/lib/Kconfig.debug b/lib/Kconfig.debug index 9ef7ce18b4f5..596bb5e4790c 100644 --- a/lib/Kconfig.debug +++ b/lib/Kconfig.debug @@ -316,6 +316,7 @@ config DEBUG_INFO_BTF bool "Generate BTF typeinfo" depends on !DEBUG_INFO_SPLIT && !DEBUG_INFO_REDUCED depends on !GCC_PLUGIN_RANDSTRUCT || COMPILE_TEST + depends on BPF_SYSCALL help Generate deduplicated BTF type information from DWARF debug info. Turning this on expects presence of pahole tool, which will convert -- cgit v1.2.3 From fc78dd08e64011865799764d5b641bf823f84c66 Mon Sep 17 00:00:00 2001 From: Sebastian Andrzej Siewior Date: Mon, 29 Nov 2021 18:46:49 +0100 Subject: lockdep/selftests: Avoid using local_lock_{acquire|release}(). The local_lock related functions local_lock_acquire() local_lock_release() are part of the internal implementation and should be avoided. Define the lock as DEFINE_PER_CPU so the normal local_lock() function can be used. Signed-off-by: Sebastian Andrzej Siewior Signed-off-by: Peter Zijlstra (Intel) Link: https://lore.kernel.org/r/20211129174654.668506-7-bigeasy@linutronix.de --- lib/locking-selftest.c | 32 ++++++++++++++++---------------- 1 file changed, 16 insertions(+), 16 deletions(-) (limited to 'lib') diff --git a/lib/locking-selftest.c b/lib/locking-selftest.c index 71652e1c397c..4d614c74e6ec 100644 --- a/lib/locking-selftest.c +++ b/lib/locking-selftest.c @@ -139,7 +139,7 @@ static DEFINE_RT_MUTEX(rtmutex_Z2); #endif -static local_lock_t local_A = INIT_LOCAL_LOCK(local_A); +static DEFINE_PER_CPU(local_lock_t, local_A); /* * non-inlined runtime initializers, to let separate locks share @@ -1320,7 +1320,7 @@ GENERATE_PERMUTATIONS_3_EVENTS(irq_read_recursion3_soft_wlock) # define I_MUTEX(x) lockdep_reset_lock(&mutex_##x.dep_map) # define I_RWSEM(x) lockdep_reset_lock(&rwsem_##x.dep_map) # define I_WW(x) lockdep_reset_lock(&x.dep_map) -# define I_LOCAL_LOCK(x) lockdep_reset_lock(&local_##x.dep_map) +# define I_LOCAL_LOCK(x) lockdep_reset_lock(this_cpu_ptr(&local_##x.dep_map)) #ifdef CONFIG_RT_MUTEXES # define I_RTMUTEX(x) lockdep_reset_lock(&rtmutex_##x.dep_map) #endif @@ -1380,7 +1380,7 @@ static void reset_locks(void) init_shared_classes(); raw_spin_lock_init(&raw_lock_A); raw_spin_lock_init(&raw_lock_B); - local_lock_init(&local_A); + local_lock_init(this_cpu_ptr(&local_A)); ww_mutex_init(&o, &ww_lockdep); ww_mutex_init(&o2, &ww_lockdep); ww_mutex_init(&o3, &ww_lockdep); memset(&t, 0, sizeof(t)); memset(&t2, 0, sizeof(t2)); @@ -2646,8 +2646,8 @@ static void wait_context_tests(void) static void local_lock_2(void) { - local_lock_acquire(&local_A); /* IRQ-ON */ - local_lock_release(&local_A); + local_lock(&local_A); /* IRQ-ON */ + local_unlock(&local_A); HARDIRQ_ENTER(); spin_lock(&lock_A); /* IN-IRQ */ @@ -2656,18 +2656,18 @@ static void local_lock_2(void) HARDIRQ_DISABLE(); spin_lock(&lock_A); - local_lock_acquire(&local_A); /* IN-IRQ <-> IRQ-ON cycle, false */ - local_lock_release(&local_A); + local_lock(&local_A); /* IN-IRQ <-> IRQ-ON cycle, false */ + local_unlock(&local_A); spin_unlock(&lock_A); HARDIRQ_ENABLE(); } static void local_lock_3A(void) { - local_lock_acquire(&local_A); /* IRQ-ON */ + local_lock(&local_A); /* IRQ-ON */ spin_lock(&lock_B); /* IRQ-ON */ spin_unlock(&lock_B); - local_lock_release(&local_A); + local_unlock(&local_A); HARDIRQ_ENTER(); spin_lock(&lock_A); /* IN-IRQ */ @@ -2676,18 +2676,18 @@ static void local_lock_3A(void) HARDIRQ_DISABLE(); spin_lock(&lock_A); - local_lock_acquire(&local_A); /* IN-IRQ <-> IRQ-ON cycle only if we count local_lock(), false */ - local_lock_release(&local_A); + local_lock(&local_A); /* IN-IRQ <-> IRQ-ON cycle only if we count local_lock(), false */ + local_unlock(&local_A); spin_unlock(&lock_A); HARDIRQ_ENABLE(); } static void local_lock_3B(void) { - local_lock_acquire(&local_A); /* IRQ-ON */ + local_lock(&local_A); /* IRQ-ON */ spin_lock(&lock_B); /* IRQ-ON */ spin_unlock(&lock_B); - local_lock_release(&local_A); + local_unlock(&local_A); HARDIRQ_ENTER(); spin_lock(&lock_A); /* IN-IRQ */ @@ -2696,8 +2696,8 @@ static void local_lock_3B(void) HARDIRQ_DISABLE(); spin_lock(&lock_A); - local_lock_acquire(&local_A); /* IN-IRQ <-> IRQ-ON cycle only if we count local_lock(), false */ - local_lock_release(&local_A); + local_lock(&local_A); /* IN-IRQ <-> IRQ-ON cycle only if we count local_lock(), false */ + local_unlock(&local_A); spin_unlock(&lock_A); HARDIRQ_ENABLE(); @@ -2812,7 +2812,7 @@ void locking_selftest(void) printk("------------------------\n"); printk("| Locking API testsuite:\n"); printk("----------------------------------------------------------------------------\n"); - printk(" | spin |wlock |rlock |mutex | wsem | rsem |\n"); + printk(" | spin |wlock |rlock |mutex | wsem | rsem |rtmutex\n"); printk(" --------------------------------------------------------------------------\n"); init_shared_classes(); -- cgit v1.2.3 From 512bf713cb4c8a42ae76e5ba1a78e70a768af301 Mon Sep 17 00:00:00 2001 From: Sebastian Andrzej Siewior Date: Mon, 29 Nov 2021 18:46:50 +0100 Subject: lockdep/selftests: Unbalanced migrate_disable() & rcu_read_lock(). The tests with unbalanced lock() + unlock() operation leave a modified preemption counter behind which is then reset to its original value after the test. The spin_lock() function on PREEMPT_RT does not include a preempt_disable() statement but migrate_disable() and read_rcu_lock(). As a consequence both counter never get back to their original value and the system explodes later after the selftest. In the double-unlock case on PREEMPT_RT, the migrate_disable() and RCU code will trigger a warning which should be avoided. These counter should not be decremented below their initial value. Save both counters and bring them back to their original value after the test. In the double-unlock case, increment both counter in advance to they become balanced after the double unlock. Signed-off-by: Sebastian Andrzej Siewior Signed-off-by: Peter Zijlstra (Intel) Link: https://lore.kernel.org/r/20211129174654.668506-8-bigeasy@linutronix.de --- lib/locking-selftest.c | 26 +++++++++++++++++++++++++- 1 file changed, 25 insertions(+), 1 deletion(-) (limited to 'lib') diff --git a/lib/locking-selftest.c b/lib/locking-selftest.c index 4d614c74e6ec..417056ba28e1 100644 --- a/lib/locking-selftest.c +++ b/lib/locking-selftest.c @@ -712,12 +712,18 @@ GENERATE_TESTCASE(ABCDBCDA_rtmutex); #undef E +#ifdef CONFIG_PREEMPT_RT +# define RT_PREPARE_DBL_UNLOCK() { migrate_disable(); rcu_read_lock(); } +#else +# define RT_PREPARE_DBL_UNLOCK() +#endif /* * Double unlock: */ #define E() \ \ LOCK(A); \ + RT_PREPARE_DBL_UNLOCK(); \ UNLOCK(A); \ UNLOCK(A); /* fail */ @@ -1398,7 +1404,13 @@ static int unexpected_testcase_failures; static void dotest(void (*testcase_fn)(void), int expected, int lockclass_mask) { - unsigned long saved_preempt_count = preempt_count(); + int saved_preempt_count = preempt_count(); +#ifdef CONFIG_PREEMPT_RT +#ifdef CONFIG_SMP + int saved_mgd_count = current->migration_disabled; +#endif + int saved_rcu_count = current->rcu_read_lock_nesting; +#endif WARN_ON(irqs_disabled()); @@ -1432,6 +1444,18 @@ static void dotest(void (*testcase_fn)(void), int expected, int lockclass_mask) * count, so restore it: */ preempt_count_set(saved_preempt_count); + +#ifdef CONFIG_PREEMPT_RT +#ifdef CONFIG_SMP + while (current->migration_disabled > saved_mgd_count) + migrate_enable(); +#endif + + while (current->rcu_read_lock_nesting > saved_rcu_count) + rcu_read_unlock(); + WARN_ON_ONCE(current->rcu_read_lock_nesting < saved_rcu_count); +#endif + #ifdef CONFIG_TRACE_IRQFLAGS if (softirq_count()) current->softirqs_enabled = 0; -- cgit v1.2.3 From a529f8db897625859b640b565325463e5d5ff01e Mon Sep 17 00:00:00 2001 From: Sebastian Andrzej Siewior Date: Mon, 29 Nov 2021 18:46:51 +0100 Subject: lockdep/selftests: Skip the softirq related tests on PREEMPT_RT The softirq context on PREEMPT_RT is different compared to !PREEMPT_RT. As such lockdep_softirq_enter() is a nop and the all the "softirq safe" tests fail on PREEMPT_RT because there is no difference. Skip the softirq context tests on PREEMPT_RT. Signed-off-by: Sebastian Andrzej Siewior Signed-off-by: Peter Zijlstra (Intel) Link: https://lore.kernel.org/r/20211129174654.668506-9-bigeasy@linutronix.de --- lib/locking-selftest.c | 38 +++++++++++++++++++++++++++++++------- 1 file changed, 31 insertions(+), 7 deletions(-) (limited to 'lib') diff --git a/lib/locking-selftest.c b/lib/locking-selftest.c index 417056ba28e1..9031f50905ae 100644 --- a/lib/locking-selftest.c +++ b/lib/locking-selftest.c @@ -26,6 +26,12 @@ #include #include +#ifdef CONFIG_PREEMPT_RT +# define NON_RT(...) +#else +# define NON_RT(...) __VA_ARGS__ +#endif + /* * Change this to 1 if you want to see the failure printouts: */ @@ -808,6 +814,7 @@ GENERATE_PERMUTATIONS_2_EVENTS(irqsafe1_hard_rlock) #include "locking-selftest-wlock-hardirq.h" GENERATE_PERMUTATIONS_2_EVENTS(irqsafe1_hard_wlock) +#ifndef CONFIG_PREEMPT_RT #include "locking-selftest-spin-softirq.h" GENERATE_PERMUTATIONS_2_EVENTS(irqsafe1_soft_spin) @@ -816,10 +823,12 @@ GENERATE_PERMUTATIONS_2_EVENTS(irqsafe1_soft_rlock) #include "locking-selftest-wlock-softirq.h" GENERATE_PERMUTATIONS_2_EVENTS(irqsafe1_soft_wlock) +#endif #undef E1 #undef E2 +#ifndef CONFIG_PREEMPT_RT /* * Enabling hardirqs with a softirq-safe lock held: */ @@ -852,6 +861,8 @@ GENERATE_PERMUTATIONS_2_EVENTS(irqsafe2A_rlock) #undef E1 #undef E2 +#endif + /* * Enabling irqs with an irq-safe lock held: */ @@ -881,6 +892,7 @@ GENERATE_PERMUTATIONS_2_EVENTS(irqsafe2B_hard_rlock) #include "locking-selftest-wlock-hardirq.h" GENERATE_PERMUTATIONS_2_EVENTS(irqsafe2B_hard_wlock) +#ifndef CONFIG_PREEMPT_RT #include "locking-selftest-spin-softirq.h" GENERATE_PERMUTATIONS_2_EVENTS(irqsafe2B_soft_spin) @@ -889,6 +901,7 @@ GENERATE_PERMUTATIONS_2_EVENTS(irqsafe2B_soft_rlock) #include "locking-selftest-wlock-softirq.h" GENERATE_PERMUTATIONS_2_EVENTS(irqsafe2B_soft_wlock) +#endif #undef E1 #undef E2 @@ -927,6 +940,7 @@ GENERATE_PERMUTATIONS_3_EVENTS(irqsafe3_hard_rlock) #include "locking-selftest-wlock-hardirq.h" GENERATE_PERMUTATIONS_3_EVENTS(irqsafe3_hard_wlock) +#ifndef CONFIG_PREEMPT_RT #include "locking-selftest-spin-softirq.h" GENERATE_PERMUTATIONS_3_EVENTS(irqsafe3_soft_spin) @@ -935,6 +949,7 @@ GENERATE_PERMUTATIONS_3_EVENTS(irqsafe3_soft_rlock) #include "locking-selftest-wlock-softirq.h" GENERATE_PERMUTATIONS_3_EVENTS(irqsafe3_soft_wlock) +#endif #undef E1 #undef E2 @@ -975,6 +990,7 @@ GENERATE_PERMUTATIONS_3_EVENTS(irqsafe4_hard_rlock) #include "locking-selftest-wlock-hardirq.h" GENERATE_PERMUTATIONS_3_EVENTS(irqsafe4_hard_wlock) +#ifndef CONFIG_PREEMPT_RT #include "locking-selftest-spin-softirq.h" GENERATE_PERMUTATIONS_3_EVENTS(irqsafe4_soft_spin) @@ -983,6 +999,7 @@ GENERATE_PERMUTATIONS_3_EVENTS(irqsafe4_soft_rlock) #include "locking-selftest-wlock-softirq.h" GENERATE_PERMUTATIONS_3_EVENTS(irqsafe4_soft_wlock) +#endif #undef E1 #undef E2 @@ -1037,6 +1054,7 @@ GENERATE_PERMUTATIONS_3_EVENTS(irq_inversion_hard_rlock) #include "locking-selftest-wlock-hardirq.h" GENERATE_PERMUTATIONS_3_EVENTS(irq_inversion_hard_wlock) +#ifndef CONFIG_PREEMPT_RT #include "locking-selftest-spin-softirq.h" GENERATE_PERMUTATIONS_3_EVENTS(irq_inversion_soft_spin) @@ -1045,6 +1063,7 @@ GENERATE_PERMUTATIONS_3_EVENTS(irq_inversion_soft_rlock) #include "locking-selftest-wlock-softirq.h" GENERATE_PERMUTATIONS_3_EVENTS(irq_inversion_soft_wlock) +#endif #undef E1 #undef E2 @@ -1212,12 +1231,14 @@ GENERATE_PERMUTATIONS_3_EVENTS(irq_read_recursion_hard_rlock) #include "locking-selftest-wlock.h" GENERATE_PERMUTATIONS_3_EVENTS(irq_read_recursion_hard_wlock) +#ifndef CONFIG_PREEMPT_RT #include "locking-selftest-softirq.h" #include "locking-selftest-rlock.h" GENERATE_PERMUTATIONS_3_EVENTS(irq_read_recursion_soft_rlock) #include "locking-selftest-wlock.h" GENERATE_PERMUTATIONS_3_EVENTS(irq_read_recursion_soft_wlock) +#endif #undef E1 #undef E2 @@ -1258,12 +1279,14 @@ GENERATE_PERMUTATIONS_3_EVENTS(irq_read_recursion2_hard_rlock) #include "locking-selftest-wlock.h" GENERATE_PERMUTATIONS_3_EVENTS(irq_read_recursion2_hard_wlock) +#ifndef CONFIG_PREEMPT_RT #include "locking-selftest-softirq.h" #include "locking-selftest-rlock.h" GENERATE_PERMUTATIONS_3_EVENTS(irq_read_recursion2_soft_rlock) #include "locking-selftest-wlock.h" GENERATE_PERMUTATIONS_3_EVENTS(irq_read_recursion2_soft_wlock) +#endif #undef E1 #undef E2 @@ -1312,12 +1335,14 @@ GENERATE_PERMUTATIONS_3_EVENTS(irq_read_recursion3_hard_rlock) #include "locking-selftest-wlock.h" GENERATE_PERMUTATIONS_3_EVENTS(irq_read_recursion3_hard_wlock) +#ifndef CONFIG_PREEMPT_RT #include "locking-selftest-softirq.h" #include "locking-selftest-rlock.h" GENERATE_PERMUTATIONS_3_EVENTS(irq_read_recursion3_soft_rlock) #include "locking-selftest-wlock.h" GENERATE_PERMUTATIONS_3_EVENTS(irq_read_recursion3_soft_wlock) +#endif #ifdef CONFIG_DEBUG_LOCK_ALLOC # define I_SPINLOCK(x) lockdep_reset_lock(&lock_##x.dep_map) @@ -1523,7 +1548,7 @@ static inline void print_testname(const char *testname) #define DO_TESTCASE_2x2RW(desc, name, nr) \ DO_TESTCASE_2RW("hard-"desc, name##_hard, nr) \ - DO_TESTCASE_2RW("soft-"desc, name##_soft, nr) \ + NON_RT(DO_TESTCASE_2RW("soft-"desc, name##_soft, nr)) \ #define DO_TESTCASE_6x2x2RW(desc, name) \ DO_TESTCASE_2x2RW(desc, name, 123); \ @@ -1571,19 +1596,19 @@ static inline void print_testname(const char *testname) #define DO_TESTCASE_2I(desc, name, nr) \ DO_TESTCASE_1("hard-"desc, name##_hard, nr); \ - DO_TESTCASE_1("soft-"desc, name##_soft, nr); + NON_RT(DO_TESTCASE_1("soft-"desc, name##_soft, nr)); #define DO_TESTCASE_2IB(desc, name, nr) \ DO_TESTCASE_1B("hard-"desc, name##_hard, nr); \ - DO_TESTCASE_1B("soft-"desc, name##_soft, nr); + NON_RT(DO_TESTCASE_1B("soft-"desc, name##_soft, nr)); #define DO_TESTCASE_6I(desc, name, nr) \ DO_TESTCASE_3("hard-"desc, name##_hard, nr); \ - DO_TESTCASE_3("soft-"desc, name##_soft, nr); + NON_RT(DO_TESTCASE_3("soft-"desc, name##_soft, nr)); #define DO_TESTCASE_6IRW(desc, name, nr) \ DO_TESTCASE_3RW("hard-"desc, name##_hard, nr); \ - DO_TESTCASE_3RW("soft-"desc, name##_soft, nr); + NON_RT(DO_TESTCASE_3RW("soft-"desc, name##_soft, nr)); #define DO_TESTCASE_2x3(desc, name) \ DO_TESTCASE_3(desc, name, 12); \ @@ -2909,12 +2934,11 @@ void locking_selftest(void) DO_TESTCASE_6x1RR("rlock W1R2/R2R3/W3W1", W1R2_R2R3_W3W1); printk(" --------------------------------------------------------------------------\n"); - /* * irq-context testcases: */ DO_TESTCASE_2x6("irqs-on + irq-safe-A", irqsafe1); - DO_TESTCASE_2x3("sirq-safe-A => hirqs-on", irqsafe2A); + NON_RT(DO_TESTCASE_2x3("sirq-safe-A => hirqs-on", irqsafe2A)); DO_TESTCASE_2x6("safe-A + irqs-on", irqsafe2B); DO_TESTCASE_6x6("safe-A + unsafe-B #1", irqsafe3); DO_TESTCASE_6x6("safe-A + unsafe-B #2", irqsafe4); -- cgit v1.2.3 From 9a75bd0c52df6cff44735f73dfb9d00e67969fc5 Mon Sep 17 00:00:00 2001 From: Sebastian Andrzej Siewior Date: Mon, 29 Nov 2021 18:46:52 +0100 Subject: lockdep/selftests: Adapt ww-tests for PREEMPT_RT The ww-mutex selftest operates directly on ww_mutex::base and assumes its type is struct mutex. This isn't true on PREEMPT_RT which turns the mutex into a rtmutex. Add a ww_mutex_base_ abstraction which maps to the relevant mutex_ or rt_mutex_ function. Change the CONFIG_DEBUG_MUTEXES ifdef to DEBUG_WW_MUTEXES. The latter is true for the MUTEX and RTMUTEX implementation of WW-MUTEX. The assignment is required in order to pass the tests. Signed-off-by: Sebastian Andrzej Siewior Signed-off-by: Peter Zijlstra (Intel) Link: https://lore.kernel.org/r/20211129174654.668506-10-bigeasy@linutronix.de --- lib/locking-selftest.c | 76 ++++++++++++++++++++++++++++++-------------------- 1 file changed, 46 insertions(+), 30 deletions(-) (limited to 'lib') diff --git a/lib/locking-selftest.c b/lib/locking-selftest.c index 9031f50905ae..8d24279fad05 100644 --- a/lib/locking-selftest.c +++ b/lib/locking-selftest.c @@ -1700,6 +1700,22 @@ static void ww_test_fail_acquire(void) #endif } +#ifdef CONFIG_PREEMPT_RT +#define ww_mutex_base_lock(b) rt_mutex_lock(b) +#define ww_mutex_base_trylock(b) rt_mutex_trylock(b) +#define ww_mutex_base_lock_nest_lock(b, b2) rt_mutex_lock_nest_lock(b, b2) +#define ww_mutex_base_lock_interruptible(b) rt_mutex_lock_interruptible(b) +#define ww_mutex_base_lock_killable(b) rt_mutex_lock_killable(b) +#define ww_mutex_base_unlock(b) rt_mutex_unlock(b) +#else +#define ww_mutex_base_lock(b) mutex_lock(b) +#define ww_mutex_base_trylock(b) mutex_trylock(b) +#define ww_mutex_base_lock_nest_lock(b, b2) mutex_lock_nest_lock(b, b2) +#define ww_mutex_base_lock_interruptible(b) mutex_lock_interruptible(b) +#define ww_mutex_base_lock_killable(b) mutex_lock_killable(b) +#define ww_mutex_base_unlock(b) mutex_unlock(b) +#endif + static void ww_test_normal(void) { int ret; @@ -1714,50 +1730,50 @@ static void ww_test_normal(void) /* mutex_lock (and indirectly, mutex_lock_nested) */ o.ctx = (void *)~0UL; - mutex_lock(&o.base); - mutex_unlock(&o.base); + ww_mutex_base_lock(&o.base); + ww_mutex_base_unlock(&o.base); WARN_ON(o.ctx != (void *)~0UL); /* mutex_lock_interruptible (and *_nested) */ o.ctx = (void *)~0UL; - ret = mutex_lock_interruptible(&o.base); + ret = ww_mutex_base_lock_interruptible(&o.base); if (!ret) - mutex_unlock(&o.base); + ww_mutex_base_unlock(&o.base); else WARN_ON(1); WARN_ON(o.ctx != (void *)~0UL); /* mutex_lock_killable (and *_nested) */ o.ctx = (void *)~0UL; - ret = mutex_lock_killable(&o.base); + ret = ww_mutex_base_lock_killable(&o.base); if (!ret) - mutex_unlock(&o.base); + ww_mutex_base_unlock(&o.base); else WARN_ON(1); WARN_ON(o.ctx != (void *)~0UL); /* trylock, succeeding */ o.ctx = (void *)~0UL; - ret = mutex_trylock(&o.base); + ret = ww_mutex_base_trylock(&o.base); WARN_ON(!ret); if (ret) - mutex_unlock(&o.base); + ww_mutex_base_unlock(&o.base); else WARN_ON(1); WARN_ON(o.ctx != (void *)~0UL); /* trylock, failing */ o.ctx = (void *)~0UL; - mutex_lock(&o.base); - ret = mutex_trylock(&o.base); + ww_mutex_base_lock(&o.base); + ret = ww_mutex_base_trylock(&o.base); WARN_ON(ret); - mutex_unlock(&o.base); + ww_mutex_base_unlock(&o.base); WARN_ON(o.ctx != (void *)~0UL); /* nest_lock */ o.ctx = (void *)~0UL; - mutex_lock_nest_lock(&o.base, &t); - mutex_unlock(&o.base); + ww_mutex_base_lock_nest_lock(&o.base, &t); + ww_mutex_base_unlock(&o.base); WARN_ON(o.ctx != (void *)~0UL); } @@ -1770,7 +1786,7 @@ static void ww_test_two_contexts(void) static void ww_test_diff_class(void) { WWAI(&t); -#ifdef CONFIG_DEBUG_MUTEXES +#ifdef DEBUG_WW_MUTEXES t.ww_class = NULL; #endif WWL(&o, &t); @@ -1834,7 +1850,7 @@ static void ww_test_edeadlk_normal(void) { int ret; - mutex_lock(&o2.base); + ww_mutex_base_lock(&o2.base); o2.ctx = &t2; mutex_release(&o2.base.dep_map, _THIS_IP_); @@ -1850,7 +1866,7 @@ static void ww_test_edeadlk_normal(void) o2.ctx = NULL; mutex_acquire(&o2.base.dep_map, 0, 1, _THIS_IP_); - mutex_unlock(&o2.base); + ww_mutex_base_unlock(&o2.base); WWU(&o); WWL(&o2, &t); @@ -1860,7 +1876,7 @@ static void ww_test_edeadlk_normal_slow(void) { int ret; - mutex_lock(&o2.base); + ww_mutex_base_lock(&o2.base); mutex_release(&o2.base.dep_map, _THIS_IP_); o2.ctx = &t2; @@ -1876,7 +1892,7 @@ static void ww_test_edeadlk_normal_slow(void) o2.ctx = NULL; mutex_acquire(&o2.base.dep_map, 0, 1, _THIS_IP_); - mutex_unlock(&o2.base); + ww_mutex_base_unlock(&o2.base); WWU(&o); ww_mutex_lock_slow(&o2, &t); @@ -1886,7 +1902,7 @@ static void ww_test_edeadlk_no_unlock(void) { int ret; - mutex_lock(&o2.base); + ww_mutex_base_lock(&o2.base); o2.ctx = &t2; mutex_release(&o2.base.dep_map, _THIS_IP_); @@ -1902,7 +1918,7 @@ static void ww_test_edeadlk_no_unlock(void) o2.ctx = NULL; mutex_acquire(&o2.base.dep_map, 0, 1, _THIS_IP_); - mutex_unlock(&o2.base); + ww_mutex_base_unlock(&o2.base); WWL(&o2, &t); } @@ -1911,7 +1927,7 @@ static void ww_test_edeadlk_no_unlock_slow(void) { int ret; - mutex_lock(&o2.base); + ww_mutex_base_lock(&o2.base); mutex_release(&o2.base.dep_map, _THIS_IP_); o2.ctx = &t2; @@ -1927,7 +1943,7 @@ static void ww_test_edeadlk_no_unlock_slow(void) o2.ctx = NULL; mutex_acquire(&o2.base.dep_map, 0, 1, _THIS_IP_); - mutex_unlock(&o2.base); + ww_mutex_base_unlock(&o2.base); ww_mutex_lock_slow(&o2, &t); } @@ -1936,7 +1952,7 @@ static void ww_test_edeadlk_acquire_more(void) { int ret; - mutex_lock(&o2.base); + ww_mutex_base_lock(&o2.base); mutex_release(&o2.base.dep_map, _THIS_IP_); o2.ctx = &t2; @@ -1957,7 +1973,7 @@ static void ww_test_edeadlk_acquire_more_slow(void) { int ret; - mutex_lock(&o2.base); + ww_mutex_base_lock(&o2.base); mutex_release(&o2.base.dep_map, _THIS_IP_); o2.ctx = &t2; @@ -1978,11 +1994,11 @@ static void ww_test_edeadlk_acquire_more_edeadlk(void) { int ret; - mutex_lock(&o2.base); + ww_mutex_base_lock(&o2.base); mutex_release(&o2.base.dep_map, _THIS_IP_); o2.ctx = &t2; - mutex_lock(&o3.base); + ww_mutex_base_lock(&o3.base); mutex_release(&o3.base.dep_map, _THIS_IP_); o3.ctx = &t2; @@ -2004,11 +2020,11 @@ static void ww_test_edeadlk_acquire_more_edeadlk_slow(void) { int ret; - mutex_lock(&o2.base); + ww_mutex_base_lock(&o2.base); mutex_release(&o2.base.dep_map, _THIS_IP_); o2.ctx = &t2; - mutex_lock(&o3.base); + ww_mutex_base_lock(&o3.base); mutex_release(&o3.base.dep_map, _THIS_IP_); o3.ctx = &t2; @@ -2029,7 +2045,7 @@ static void ww_test_edeadlk_acquire_wrong(void) { int ret; - mutex_lock(&o2.base); + ww_mutex_base_lock(&o2.base); mutex_release(&o2.base.dep_map, _THIS_IP_); o2.ctx = &t2; @@ -2054,7 +2070,7 @@ static void ww_test_edeadlk_acquire_wrong_slow(void) { int ret; - mutex_lock(&o2.base); + ww_mutex_base_lock(&o2.base); mutex_release(&o2.base.dep_map, _THIS_IP_); o2.ctx = &t2; -- cgit v1.2.3 From 52e68cd60ddf11802f5135921aba77c0833909a8 Mon Sep 17 00:00:00 2001 From: Christophe JAILLET Date: Sat, 27 Nov 2021 15:27:35 +0100 Subject: vsprintf: Use non-atomic bitmap API when applicable The 'set' bitmap is local to this function. No concurrent access to it is possible. So prefer the non-atomic '__[set|clear]_bit()' function to save a few cycles. Signed-off-by: Christophe JAILLET Reviewed-by: Petr Mladek Signed-off-by: Petr Mladek Link: https://lore.kernel.org/r/1abf81a5e509d372393bd22041eed4ebc07ef9f7.1638023178.git.christophe.jaillet@wanadoo.fr --- lib/vsprintf.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'lib') diff --git a/lib/vsprintf.c b/lib/vsprintf.c index 58d5e567f836..53d6081f9e8b 100644 --- a/lib/vsprintf.c +++ b/lib/vsprintf.c @@ -3564,7 +3564,7 @@ int vsscanf(const char *buf, const char *fmt, va_list args) ++fmt; for ( ; *fmt && *fmt != ']'; ++fmt, ++len) - set_bit((u8)*fmt, set); + __set_bit((u8)*fmt, set); /* no ']' or no character set found */ if (!*fmt || !len) @@ -3574,7 +3574,7 @@ int vsscanf(const char *buf, const char *fmt, va_list args) if (negate) { bitmap_complement(set, set, 256); /* exclude null '\0' byte */ - clear_bit(0, set); + __clear_bit(0, set); } /* match must be non-empty */ -- cgit v1.2.3 From 4e66934eaadc83b27ada8d42b60894018f3bfabf Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Sat, 4 Dec 2021 20:21:55 -0800 Subject: lib: add reference counting tracking infrastructure It can be hard to track where references are taken and released. In networking, we have annoying issues at device or netns dismantles, and we had various proposals to ease root causing them. This patch adds new infrastructure pairing refcount increases and decreases. This will self document code, because programmers will have to associate increments/decrements. This is controled by CONFIG_REF_TRACKER which can be selected by users of this feature. This adds both cpu and memory costs, and thus should probably be used with care. Signed-off-by: Eric Dumazet Reviewed-by: Dmitry Vyukov Signed-off-by: Jakub Kicinski --- include/linux/ref_tracker.h | 73 +++++++++++++++++++++++ lib/Kconfig | 5 ++ lib/Makefile | 2 + lib/ref_tracker.c | 140 ++++++++++++++++++++++++++++++++++++++++++++ 4 files changed, 220 insertions(+) create mode 100644 include/linux/ref_tracker.h create mode 100644 lib/ref_tracker.c (limited to 'lib') diff --git a/include/linux/ref_tracker.h b/include/linux/ref_tracker.h new file mode 100644 index 000000000000..c11c9db5825c --- /dev/null +++ b/include/linux/ref_tracker.h @@ -0,0 +1,73 @@ +// SPDX-License-Identifier: GPL-2.0-or-later +#ifndef _LINUX_REF_TRACKER_H +#define _LINUX_REF_TRACKER_H +#include +#include +#include + +struct ref_tracker; + +struct ref_tracker_dir { +#ifdef CONFIG_REF_TRACKER + spinlock_t lock; + unsigned int quarantine_avail; + refcount_t untracked; + struct list_head list; /* List of active trackers */ + struct list_head quarantine; /* List of dead trackers */ +#endif +}; + +#ifdef CONFIG_REF_TRACKER +static inline void ref_tracker_dir_init(struct ref_tracker_dir *dir, + unsigned int quarantine_count) +{ + INIT_LIST_HEAD(&dir->list); + INIT_LIST_HEAD(&dir->quarantine); + spin_lock_init(&dir->lock); + dir->quarantine_avail = quarantine_count; + refcount_set(&dir->untracked, 1); +} + +void ref_tracker_dir_exit(struct ref_tracker_dir *dir); + +void ref_tracker_dir_print(struct ref_tracker_dir *dir, + unsigned int display_limit); + +int ref_tracker_alloc(struct ref_tracker_dir *dir, + struct ref_tracker **trackerp, gfp_t gfp); + +int ref_tracker_free(struct ref_tracker_dir *dir, + struct ref_tracker **trackerp); + +#else /* CONFIG_REF_TRACKER */ + +static inline void ref_tracker_dir_init(struct ref_tracker_dir *dir, + unsigned int quarantine_count) +{ +} + +static inline void ref_tracker_dir_exit(struct ref_tracker_dir *dir) +{ +} + +static inline void ref_tracker_dir_print(struct ref_tracker_dir *dir, + unsigned int display_limit) +{ +} + +static inline int ref_tracker_alloc(struct ref_tracker_dir *dir, + struct ref_tracker **trackerp, + gfp_t gfp) +{ + return 0; +} + +static inline int ref_tracker_free(struct ref_tracker_dir *dir, + struct ref_tracker **trackerp) +{ + return 0; +} + +#endif + +#endif /* _LINUX_REF_TRACKER_H */ diff --git a/lib/Kconfig b/lib/Kconfig index 5e7165e6a346..655b0e43f260 100644 --- a/lib/Kconfig +++ b/lib/Kconfig @@ -680,6 +680,11 @@ config STACK_HASH_ORDER Select the hash size as a power of 2 for the stackdepot hash table. Choose a lower value to reduce the memory impact. +config REF_TRACKER + bool + depends on STACKTRACE_SUPPORT + select STACKDEPOT + config SBITMAP bool diff --git a/lib/Makefile b/lib/Makefile index 364c23f15578..c1fd9243ddb9 100644 --- a/lib/Makefile +++ b/lib/Makefile @@ -270,6 +270,8 @@ obj-$(CONFIG_STACKDEPOT) += stackdepot.o KASAN_SANITIZE_stackdepot.o := n KCOV_INSTRUMENT_stackdepot.o := n +obj-$(CONFIG_REF_TRACKER) += ref_tracker.o + libfdt_files = fdt.o fdt_ro.o fdt_wip.o fdt_rw.o fdt_sw.o fdt_strerror.o \ fdt_empty_tree.o fdt_addresses.o $(foreach file, $(libfdt_files), \ diff --git a/lib/ref_tracker.c b/lib/ref_tracker.c new file mode 100644 index 000000000000..0ae2e66dcf0f --- /dev/null +++ b/lib/ref_tracker.c @@ -0,0 +1,140 @@ +// SPDX-License-Identifier: GPL-2.0-or-later +#include +#include +#include +#include +#include + +#define REF_TRACKER_STACK_ENTRIES 16 + +struct ref_tracker { + struct list_head head; /* anchor into dir->list or dir->quarantine */ + bool dead; + depot_stack_handle_t alloc_stack_handle; + depot_stack_handle_t free_stack_handle; +}; + +void ref_tracker_dir_exit(struct ref_tracker_dir *dir) +{ + struct ref_tracker *tracker, *n; + unsigned long flags; + bool leak = false; + + spin_lock_irqsave(&dir->lock, flags); + list_for_each_entry_safe(tracker, n, &dir->quarantine, head) { + list_del(&tracker->head); + kfree(tracker); + dir->quarantine_avail++; + } + list_for_each_entry_safe(tracker, n, &dir->list, head) { + pr_err("leaked reference.\n"); + if (tracker->alloc_stack_handle) + stack_depot_print(tracker->alloc_stack_handle); + leak = true; + list_del(&tracker->head); + kfree(tracker); + } + spin_unlock_irqrestore(&dir->lock, flags); + WARN_ON_ONCE(leak); + WARN_ON_ONCE(refcount_read(&dir->untracked) != 1); +} +EXPORT_SYMBOL(ref_tracker_dir_exit); + +void ref_tracker_dir_print(struct ref_tracker_dir *dir, + unsigned int display_limit) +{ + struct ref_tracker *tracker; + unsigned long flags; + unsigned int i = 0; + + spin_lock_irqsave(&dir->lock, flags); + list_for_each_entry(tracker, &dir->list, head) { + if (i < display_limit) { + pr_err("leaked reference.\n"); + if (tracker->alloc_stack_handle) + stack_depot_print(tracker->alloc_stack_handle); + i++; + } else { + break; + } + } + spin_unlock_irqrestore(&dir->lock, flags); +} +EXPORT_SYMBOL(ref_tracker_dir_print); + +int ref_tracker_alloc(struct ref_tracker_dir *dir, + struct ref_tracker **trackerp, + gfp_t gfp) +{ + unsigned long entries[REF_TRACKER_STACK_ENTRIES]; + struct ref_tracker *tracker; + unsigned int nr_entries; + unsigned long flags; + + *trackerp = tracker = kzalloc(sizeof(*tracker), gfp | __GFP_NOFAIL); + if (unlikely(!tracker)) { + pr_err_once("memory allocation failure, unreliable refcount tracker.\n"); + refcount_inc(&dir->untracked); + return -ENOMEM; + } + nr_entries = stack_trace_save(entries, ARRAY_SIZE(entries), 1); + nr_entries = filter_irq_stacks(entries, nr_entries); + tracker->alloc_stack_handle = stack_depot_save(entries, nr_entries, gfp); + + spin_lock_irqsave(&dir->lock, flags); + list_add(&tracker->head, &dir->list); + spin_unlock_irqrestore(&dir->lock, flags); + return 0; +} +EXPORT_SYMBOL_GPL(ref_tracker_alloc); + +int ref_tracker_free(struct ref_tracker_dir *dir, + struct ref_tracker **trackerp) +{ + unsigned long entries[REF_TRACKER_STACK_ENTRIES]; + struct ref_tracker *tracker = *trackerp; + depot_stack_handle_t stack_handle; + unsigned int nr_entries; + unsigned long flags; + + if (!tracker) { + refcount_dec(&dir->untracked); + return -EEXIST; + } + nr_entries = stack_trace_save(entries, ARRAY_SIZE(entries), 1); + nr_entries = filter_irq_stacks(entries, nr_entries); + stack_handle = stack_depot_save(entries, nr_entries, GFP_ATOMIC); + + spin_lock_irqsave(&dir->lock, flags); + if (tracker->dead) { + pr_err("reference already released.\n"); + if (tracker->alloc_stack_handle) { + pr_err("allocated in:\n"); + stack_depot_print(tracker->alloc_stack_handle); + } + if (tracker->free_stack_handle) { + pr_err("freed in:\n"); + stack_depot_print(tracker->free_stack_handle); + } + spin_unlock_irqrestore(&dir->lock, flags); + WARN_ON_ONCE(1); + return -EINVAL; + } + tracker->dead = true; + + tracker->free_stack_handle = stack_handle; + + list_move_tail(&tracker->head, &dir->quarantine); + if (!dir->quarantine_avail) { + tracker = list_first_entry(&dir->quarantine, struct ref_tracker, head); + list_del(&tracker->head); + } else { + dir->quarantine_avail--; + tracker = NULL; + } + spin_unlock_irqrestore(&dir->lock, flags); + + kfree(tracker); + return 0; +} +EXPORT_SYMBOL_GPL(ref_tracker_free); -- cgit v1.2.3 From 914a7b5000d08f1487e0efa52f27c4b7ea75b893 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Sat, 4 Dec 2021 20:21:56 -0800 Subject: lib: add tests for reference tracker This module uses reference tracker, forcing two issues. 1) Double free of a tracker 2) leak of two trackers, one being allocated from softirq context. "modprobe test_ref_tracker" would emit the following traces. (Use scripts/decode_stacktrace.sh if necessary) [ 171.648681] reference already released. [ 171.653213] allocated in: [ 171.656523] alloctest_ref_tracker_alloc2+0x1c/0x20 [test_ref_tracker] [ 171.656526] init_module+0x86/0x1000 [test_ref_tracker] [ 171.656528] do_one_initcall+0x9c/0x220 [ 171.656532] do_init_module+0x60/0x240 [ 171.656536] load_module+0x32b5/0x3610 [ 171.656538] __do_sys_init_module+0x148/0x1a0 [ 171.656540] __x64_sys_init_module+0x1d/0x20 [ 171.656542] do_syscall_64+0x4a/0xb0 [ 171.656546] entry_SYSCALL_64_after_hwframe+0x44/0xae [ 171.656549] freed in: [ 171.659520] alloctest_ref_tracker_free+0x13/0x20 [test_ref_tracker] [ 171.659522] init_module+0xec/0x1000 [test_ref_tracker] [ 171.659523] do_one_initcall+0x9c/0x220 [ 171.659525] do_init_module+0x60/0x240 [ 171.659527] load_module+0x32b5/0x3610 [ 171.659529] __do_sys_init_module+0x148/0x1a0 [ 171.659532] __x64_sys_init_module+0x1d/0x20 [ 171.659534] do_syscall_64+0x4a/0xb0 [ 171.659536] entry_SYSCALL_64_after_hwframe+0x44/0xae [ 171.659575] ------------[ cut here ]------------ [ 171.659576] WARNING: CPU: 5 PID: 13016 at lib/ref_tracker.c:112 ref_tracker_free+0x224/0x270 [ 171.659581] Modules linked in: test_ref_tracker(+) [ 171.659591] CPU: 5 PID: 13016 Comm: modprobe Tainted: G S 5.16.0-smp-DEV #290 [ 171.659595] RIP: 0010:ref_tracker_free+0x224/0x270 [ 171.659599] Code: 5e 41 5f 5d c3 48 c7 c7 04 9c 74 a6 31 c0 e8 62 ee 67 00 83 7b 14 00 75 1a 83 7b 18 00 75 30 4c 89 ff 4c 89 f6 e8 9c 00 69 00 <0f> 0b bb ea ff ff ff eb ae 48 c7 c7 3a 0a 77 a6 31 c0 e8 34 ee 67 [ 171.659601] RSP: 0018:ffff89058ba0bbd0 EFLAGS: 00010286 [ 171.659603] RAX: 0000000000000029 RBX: ffff890586b19780 RCX: 08895bff57c7d100 [ 171.659604] RDX: c0000000ffff7fff RSI: 0000000000000282 RDI: ffffffffc0407000 [ 171.659606] RBP: ffff89058ba0bc88 R08: 0000000000000000 R09: ffffffffa6f342e0 [ 171.659607] R10: 00000000ffff7fff R11: 0000000000000000 R12: 000000008f000000 [ 171.659608] R13: 0000000000000014 R14: 0000000000000282 R15: ffffffffc0407000 [ 171.659609] FS: 00007f97ea29d740(0000) GS:ffff8923ff940000(0000) knlGS:0000000000000000 [ 171.659611] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 171.659613] CR2: 00007f97ea299000 CR3: 0000000186b4a004 CR4: 00000000001706e0 [ 171.659614] Call Trace: [ 171.659615] [ 171.659631] ? alloctest_ref_tracker_free+0x13/0x20 [test_ref_tracker] [ 171.659633] ? init_module+0x105/0x1000 [test_ref_tracker] [ 171.659636] ? do_one_initcall+0x9c/0x220 [ 171.659638] ? do_init_module+0x60/0x240 [ 171.659641] ? load_module+0x32b5/0x3610 [ 171.659644] ? __do_sys_init_module+0x148/0x1a0 [ 171.659646] ? __x64_sys_init_module+0x1d/0x20 [ 171.659649] ? do_syscall_64+0x4a/0xb0 [ 171.659652] ? entry_SYSCALL_64_after_hwframe+0x44/0xae [ 171.659656] ? 0xffffffffc040a000 [ 171.659658] alloctest_ref_tracker_free+0x13/0x20 [test_ref_tracker] [ 171.659660] init_module+0x105/0x1000 [test_ref_tracker] [ 171.659663] do_one_initcall+0x9c/0x220 [ 171.659666] do_init_module+0x60/0x240 [ 171.659669] load_module+0x32b5/0x3610 [ 171.659672] __do_sys_init_module+0x148/0x1a0 [ 171.659676] __x64_sys_init_module+0x1d/0x20 [ 171.659678] do_syscall_64+0x4a/0xb0 [ 171.659694] ? exc_page_fault+0x6e/0x140 [ 171.659696] entry_SYSCALL_64_after_hwframe+0x44/0xae [ 171.659698] RIP: 0033:0x7f97ea3dbe7a [ 171.659700] Code: 48 8b 0d 61 8d 06 00 f7 d8 64 89 01 48 83 c8 ff c3 cc cc cc cc cc cc cc cc cc cc cc cc cc cc cc 49 89 ca b8 af 00 00 00 0f 05 <48> 3d 01 f0 ff ff 73 01 c3 48 8b 0d 2e 8d 06 00 f7 d8 64 89 01 48 [ 171.659701] RSP: 002b:00007ffea67ce608 EFLAGS: 00000246 ORIG_RAX: 00000000000000af [ 171.659703] RAX: ffffffffffffffda RBX: 0000000000000000 RCX: 00007f97ea3dbe7a [ 171.659704] RDX: 00000000013a0ba0 RSI: 0000000000002808 RDI: 00007f97ea299000 [ 171.659705] RBP: 00007ffea67ce670 R08: 0000000000000003 R09: 0000000000000000 [ 171.659706] R10: 0000000000000000 R11: 0000000000000246 R12: 00000000013a1048 [ 171.659707] R13: 00000000013a0ba0 R14: 0000000001399930 R15: 00000000013a1030 [ 171.659709] [ 171.659710] ---[ end trace f5dbd6afa41e60a9 ]--- [ 171.659712] leaked reference. [ 171.663393] alloctest_ref_tracker_alloc0+0x1c/0x20 [test_ref_tracker] [ 171.663395] test_ref_tracker_timer_func+0x9/0x20 [test_ref_tracker] [ 171.663397] call_timer_fn+0x31/0x140 [ 171.663401] expire_timers+0x46/0x110 [ 171.663403] __run_timers+0x16f/0x1b0 [ 171.663404] run_timer_softirq+0x1d/0x40 [ 171.663406] __do_softirq+0x148/0x2d3 [ 171.663408] leaked reference. [ 171.667101] alloctest_ref_tracker_alloc1+0x1c/0x20 [test_ref_tracker] [ 171.667103] init_module+0x81/0x1000 [test_ref_tracker] [ 171.667104] do_one_initcall+0x9c/0x220 [ 171.667106] do_init_module+0x60/0x240 [ 171.667108] load_module+0x32b5/0x3610 [ 171.667111] __do_sys_init_module+0x148/0x1a0 [ 171.667113] __x64_sys_init_module+0x1d/0x20 [ 171.667115] do_syscall_64+0x4a/0xb0 [ 171.667117] entry_SYSCALL_64_after_hwframe+0x44/0xae [ 171.667131] ------------[ cut here ]------------ [ 171.667132] WARNING: CPU: 5 PID: 13016 at lib/ref_tracker.c:30 ref_tracker_dir_exit+0x104/0x130 [ 171.667136] Modules linked in: test_ref_tracker(+) [ 171.667144] CPU: 5 PID: 13016 Comm: modprobe Tainted: G S W 5.16.0-smp-DEV #290 [ 171.667147] RIP: 0010:ref_tracker_dir_exit+0x104/0x130 [ 171.667150] Code: 01 00 00 00 00 ad de 48 89 03 4c 89 63 08 48 89 df e8 20 a0 d5 ff 4c 89 f3 4d 39 ee 75 a8 4c 89 ff 48 8b 75 d0 e8 7c 05 69 00 <0f> 0b eb 0c 4c 89 ff 48 8b 75 d0 e8 6c 05 69 00 41 8b 47 08 83 f8 [ 171.667151] RSP: 0018:ffff89058ba0bc68 EFLAGS: 00010286 [ 171.667154] RAX: 08895bff57c7d100 RBX: ffffffffc0407010 RCX: 000000000000003b [ 171.667156] RDX: 000000000000003c RSI: 0000000000000282 RDI: ffffffffc0407000 [ 171.667157] RBP: ffff89058ba0bc98 R08: 0000000000000000 R09: ffffffffa6f342e0 [ 171.667159] R10: 00000000ffff7fff R11: 0000000000000000 R12: dead000000000122 [ 171.667160] R13: ffffffffc0407010 R14: ffffffffc0407010 R15: ffffffffc0407000 [ 171.667162] FS: 00007f97ea29d740(0000) GS:ffff8923ff940000(0000) knlGS:0000000000000000 [ 171.667164] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 171.667166] CR2: 00007f97ea299000 CR3: 0000000186b4a004 CR4: 00000000001706e0 [ 171.667169] Call Trace: [ 171.667170] [ 171.667171] ? 0xffffffffc040a000 [ 171.667173] init_module+0x126/0x1000 [test_ref_tracker] [ 171.667175] do_one_initcall+0x9c/0x220 [ 171.667179] do_init_module+0x60/0x240 [ 171.667182] load_module+0x32b5/0x3610 [ 171.667186] __do_sys_init_module+0x148/0x1a0 [ 171.667189] __x64_sys_init_module+0x1d/0x20 [ 171.667192] do_syscall_64+0x4a/0xb0 [ 171.667194] ? exc_page_fault+0x6e/0x140 [ 171.667196] entry_SYSCALL_64_after_hwframe+0x44/0xae [ 171.667199] RIP: 0033:0x7f97ea3dbe7a [ 171.667200] Code: 48 8b 0d 61 8d 06 00 f7 d8 64 89 01 48 83 c8 ff c3 cc cc cc cc cc cc cc cc cc cc cc cc cc cc cc 49 89 ca b8 af 00 00 00 0f 05 <48> 3d 01 f0 ff ff 73 01 c3 48 8b 0d 2e 8d 06 00 f7 d8 64 89 01 48 [ 171.667201] RSP: 002b:00007ffea67ce608 EFLAGS: 00000246 ORIG_RAX: 00000000000000af [ 171.667203] RAX: ffffffffffffffda RBX: 0000000000000000 RCX: 00007f97ea3dbe7a [ 171.667204] RDX: 00000000013a0ba0 RSI: 0000000000002808 RDI: 00007f97ea299000 [ 171.667205] RBP: 00007ffea67ce670 R08: 0000000000000003 R09: 0000000000000000 [ 171.667206] R10: 0000000000000000 R11: 0000000000000246 R12: 00000000013a1048 [ 171.667207] R13: 00000000013a0ba0 R14: 0000000001399930 R15: 00000000013a1030 [ 171.667209] [ 171.667210] ---[ end trace f5dbd6afa41e60aa ]--- Signed-off-by: Eric Dumazet Signed-off-by: Jakub Kicinski --- lib/Kconfig.debug | 10 +++++ lib/Makefile | 2 +- lib/test_ref_tracker.c | 115 +++++++++++++++++++++++++++++++++++++++++++++++++ 3 files changed, 126 insertions(+), 1 deletion(-) create mode 100644 lib/test_ref_tracker.c (limited to 'lib') diff --git a/lib/Kconfig.debug b/lib/Kconfig.debug index 5c12bde10996..633c2c5cb45b 100644 --- a/lib/Kconfig.debug +++ b/lib/Kconfig.debug @@ -2106,6 +2106,16 @@ config BACKTRACE_SELF_TEST Say N if you are unsure. +config TEST_REF_TRACKER + tristate "Self test for reference tracker" + depends on DEBUG_KERNEL && STACKTRACE_SUPPORT + select REF_TRACKER + help + This option provides a kernel module performing tests + using reference tracker infrastructure. + + Say N if you are unsure. + config RBTREE_TEST tristate "Red-Black tree test" depends on DEBUG_KERNEL diff --git a/lib/Makefile b/lib/Makefile index c1fd9243ddb9..b213a7bbf3fd 100644 --- a/lib/Makefile +++ b/lib/Makefile @@ -101,7 +101,7 @@ obj-$(CONFIG_TEST_LOCKUP) += test_lockup.o obj-$(CONFIG_TEST_HMM) += test_hmm.o obj-$(CONFIG_TEST_FREE_PAGES) += test_free_pages.o obj-$(CONFIG_KPROBES_SANITY_TEST) += test_kprobes.o - +obj-$(CONFIG_TEST_REF_TRACKER) += test_ref_tracker.o # # CFLAGS for compiling floating point code inside the kernel. x86/Makefile turns # off the generation of FPU/SSE* instructions for kernel proper but FPU_FLAGS diff --git a/lib/test_ref_tracker.c b/lib/test_ref_tracker.c new file mode 100644 index 000000000000..19d7dec70cc6 --- /dev/null +++ b/lib/test_ref_tracker.c @@ -0,0 +1,115 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* + * Referrence tracker self test. + * + * Copyright (c) 2021 Eric Dumazet + */ +#include +#include +#include +#include +#include +#include + +static struct ref_tracker_dir ref_dir; +static struct ref_tracker *tracker[20]; + +#define TRT_ALLOC(X) static noinline void \ + alloctest_ref_tracker_alloc##X(struct ref_tracker_dir *dir, \ + struct ref_tracker **trackerp) \ + { \ + ref_tracker_alloc(dir, trackerp, GFP_KERNEL); \ + } + +TRT_ALLOC(1) +TRT_ALLOC(2) +TRT_ALLOC(3) +TRT_ALLOC(4) +TRT_ALLOC(5) +TRT_ALLOC(6) +TRT_ALLOC(7) +TRT_ALLOC(8) +TRT_ALLOC(9) +TRT_ALLOC(10) +TRT_ALLOC(11) +TRT_ALLOC(12) +TRT_ALLOC(13) +TRT_ALLOC(14) +TRT_ALLOC(15) +TRT_ALLOC(16) +TRT_ALLOC(17) +TRT_ALLOC(18) +TRT_ALLOC(19) + +#undef TRT_ALLOC + +static noinline void +alloctest_ref_tracker_free(struct ref_tracker_dir *dir, + struct ref_tracker **trackerp) +{ + ref_tracker_free(dir, trackerp); +} + + +static struct timer_list test_ref_tracker_timer; +static atomic_t test_ref_timer_done = ATOMIC_INIT(0); + +static void test_ref_tracker_timer_func(struct timer_list *t) +{ + ref_tracker_alloc(&ref_dir, &tracker[0], GFP_ATOMIC); + atomic_set(&test_ref_timer_done, 1); +} + +static int __init test_ref_tracker_init(void) +{ + int i; + + ref_tracker_dir_init(&ref_dir, 100); + + timer_setup(&test_ref_tracker_timer, test_ref_tracker_timer_func, 0); + mod_timer(&test_ref_tracker_timer, jiffies + 1); + + alloctest_ref_tracker_alloc1(&ref_dir, &tracker[1]); + alloctest_ref_tracker_alloc2(&ref_dir, &tracker[2]); + alloctest_ref_tracker_alloc3(&ref_dir, &tracker[3]); + alloctest_ref_tracker_alloc4(&ref_dir, &tracker[4]); + alloctest_ref_tracker_alloc5(&ref_dir, &tracker[5]); + alloctest_ref_tracker_alloc6(&ref_dir, &tracker[6]); + alloctest_ref_tracker_alloc7(&ref_dir, &tracker[7]); + alloctest_ref_tracker_alloc8(&ref_dir, &tracker[8]); + alloctest_ref_tracker_alloc9(&ref_dir, &tracker[9]); + alloctest_ref_tracker_alloc10(&ref_dir, &tracker[10]); + alloctest_ref_tracker_alloc11(&ref_dir, &tracker[11]); + alloctest_ref_tracker_alloc12(&ref_dir, &tracker[12]); + alloctest_ref_tracker_alloc13(&ref_dir, &tracker[13]); + alloctest_ref_tracker_alloc14(&ref_dir, &tracker[14]); + alloctest_ref_tracker_alloc15(&ref_dir, &tracker[15]); + alloctest_ref_tracker_alloc16(&ref_dir, &tracker[16]); + alloctest_ref_tracker_alloc17(&ref_dir, &tracker[17]); + alloctest_ref_tracker_alloc18(&ref_dir, &tracker[18]); + alloctest_ref_tracker_alloc19(&ref_dir, &tracker[19]); + + /* free all trackers but first 0 and 1. */ + for (i = 2; i < ARRAY_SIZE(tracker); i++) + alloctest_ref_tracker_free(&ref_dir, &tracker[i]); + + /* Attempt to free an already freed tracker. */ + alloctest_ref_tracker_free(&ref_dir, &tracker[2]); + + while (!atomic_read(&test_ref_timer_done)) + msleep(1); + + /* This should warn about tracker[0] & tracker[1] being not freed. */ + ref_tracker_dir_exit(&ref_dir); + + return 0; +} + +static void __exit test_ref_tracker_exit(void) +{ +} + +module_init(test_ref_tracker_init); +module_exit(test_ref_tracker_exit); + +MODULE_LICENSE("GPL v2"); -- cgit v1.2.3 From 4d92b95ff2f95f13df9bad0b5a25a9f60e72758d Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Sat, 4 Dec 2021 20:21:57 -0800 Subject: net: add net device refcount tracker infrastructure net device are refcounted. Over the years we had numerous bugs caused by imbalanced dev_hold() and dev_put() calls. The general idea is to be able to precisely pair each decrement with a corresponding prior increment. Both share a cookie, basically a pointer to private data storing stack traces. This patch adds dev_hold_track() and dev_put_track(). To use these helpers, each data structure owning a refcount should also use a "netdevice_tracker" to pair the hold and put. netdevice_tracker dev_tracker; ... dev_hold_track(dev, &dev_tracker, GFP_ATOMIC); ... dev_put_track(dev, &dev_tracker); Whenever a leak happens, we will get precise stack traces of the point dev_hold_track() happened, at device dismantle phase. We will also get a stack trace if too many dev_put_track() for the same netdevice_tracker are attempted. This is guarded by CONFIG_NET_DEV_REFCNT_TRACKER option. Signed-off-by: Eric Dumazet Signed-off-by: Jakub Kicinski --- include/linux/netdevice.h | 45 +++++++++++++++++++++++++++++++++++++++++++++ lib/Kconfig.debug | 5 +++++ net/Kconfig.debug | 10 ++++++++++ net/core/dev.c | 3 +++ 4 files changed, 63 insertions(+) create mode 100644 net/Kconfig.debug (limited to 'lib') diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h index 65117f01d5f2..143d60ed0047 100644 --- a/include/linux/netdevice.h +++ b/include/linux/netdevice.h @@ -48,6 +48,7 @@ #include #include #include +#include struct netpoll_info; struct device; @@ -300,6 +301,12 @@ enum netdev_state_t { }; +#ifdef CONFIG_NET_DEV_REFCNT_TRACKER +typedef struct ref_tracker *netdevice_tracker; +#else +typedef struct {} netdevice_tracker; +#endif + struct gro_list { struct list_head list; int count; @@ -1865,6 +1872,7 @@ enum netdev_ml_priv_type { * @proto_down_reason: reason a netdev interface is held down * @pcpu_refcnt: Number of references to this device * @dev_refcnt: Number of references to this device + * @refcnt_tracker: Tracker directory for tracked references to this device * @todo_list: Delayed register/unregister * @link_watch_list: XXX: need comments on this one * @@ -2178,6 +2186,7 @@ struct net_device { #else refcount_t dev_refcnt; #endif + struct ref_tracker_dir refcnt_tracker; struct list_head link_watch_list; @@ -3805,6 +3814,7 @@ void netdev_run_todo(void); * @dev: network device * * Release reference to device to allow it to be freed. + * Try using dev_put_track() instead. */ static inline void dev_put(struct net_device *dev) { @@ -3822,6 +3832,7 @@ static inline void dev_put(struct net_device *dev) * @dev: network device * * Hold reference to device to keep it from being freed. + * Try using dev_hold_track() instead. */ static inline void dev_hold(struct net_device *dev) { @@ -3834,6 +3845,40 @@ static inline void dev_hold(struct net_device *dev) } } +static inline void netdev_tracker_alloc(struct net_device *dev, + netdevice_tracker *tracker, gfp_t gfp) +{ +#ifdef CONFIG_NET_DEV_REFCNT_TRACKER + ref_tracker_alloc(&dev->refcnt_tracker, tracker, gfp); +#endif +} + +static inline void netdev_tracker_free(struct net_device *dev, + netdevice_tracker *tracker) +{ +#ifdef CONFIG_NET_DEV_REFCNT_TRACKER + ref_tracker_free(&dev->refcnt_tracker, tracker); +#endif +} + +static inline void dev_hold_track(struct net_device *dev, + netdevice_tracker *tracker, gfp_t gfp) +{ + if (dev) { + dev_hold(dev); + netdev_tracker_alloc(dev, tracker, gfp); + } +} + +static inline void dev_put_track(struct net_device *dev, + netdevice_tracker *tracker) +{ + if (dev) { + netdev_tracker_free(dev, tracker); + dev_put(dev); + } +} + /* Carrier loss detection, dial on demand. The functions netif_carrier_on * and _off may be called from IRQ context, but it is caller * who is responsible for serialization of these calls. diff --git a/lib/Kconfig.debug b/lib/Kconfig.debug index 633c2c5cb45b..6504b97f8dfd 100644 --- a/lib/Kconfig.debug +++ b/lib/Kconfig.debug @@ -598,6 +598,11 @@ config DEBUG_MISC Say Y here if you need to enable miscellaneous debug code that should be under a more specific debug option but isn't. +menu "Networking Debugging" + +source "net/Kconfig.debug" + +endmenu # "Networking Debugging" menu "Memory Debugging" diff --git a/net/Kconfig.debug b/net/Kconfig.debug new file mode 100644 index 000000000000..fb5c70e01cb3 --- /dev/null +++ b/net/Kconfig.debug @@ -0,0 +1,10 @@ +# SPDX-License-Identifier: GPL-2.0-only + +config NET_DEV_REFCNT_TRACKER + bool "Enable net device refcount tracking" + depends on DEBUG_KERNEL && STACKTRACE_SUPPORT + select REF_TRACKER + default n + help + Enable debugging feature to track device references. + This adds memory and cpu costs. diff --git a/net/core/dev.c b/net/core/dev.c index aba8acc1238c..1740d6cfe86b 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -9864,6 +9864,7 @@ static void netdev_wait_allrefs(struct net_device *dev) netdev_unregister_timeout_secs * HZ)) { pr_emerg("unregister_netdevice: waiting for %s to become free. Usage count = %d\n", dev->name, refcnt); + ref_tracker_dir_print(&dev->refcnt_tracker, 10); warning_time = jiffies; } } @@ -10154,6 +10155,7 @@ struct net_device *alloc_netdev_mqs(int sizeof_priv, const char *name, dev = PTR_ALIGN(p, NETDEV_ALIGN); dev->padded = (char *)dev - (char *)p; + ref_tracker_dir_init(&dev->refcnt_tracker, 128); #ifdef CONFIG_PCPU_DEV_REFCNT dev->pcpu_refcnt = alloc_percpu(int); if (!dev->pcpu_refcnt) @@ -10270,6 +10272,7 @@ void free_netdev(struct net_device *dev) list_for_each_entry_safe(p, n, &dev->napi_list, dev_list) netif_napi_del(p); + ref_tracker_dir_exit(&dev->refcnt_tracker); #ifdef CONFIG_PCPU_DEV_REFCNT free_percpu(dev->pcpu_refcnt); dev->pcpu_refcnt = NULL; -- cgit v1.2.3 From 69562e4983d93e2791c0bf128b07462afbd7f4dc Mon Sep 17 00:00:00 2001 From: Marco Elver Date: Thu, 5 Aug 2021 14:57:45 +0200 Subject: kcsan: Add core support for a subset of weak memory modeling Add support for modeling a subset of weak memory, which will enable detection of a subset of data races due to missing memory barriers. KCSAN's approach to detecting missing memory barriers is based on modeling access reordering, and enabled if `CONFIG_KCSAN_WEAK_MEMORY=y`, which depends on `CONFIG_KCSAN_STRICT=y`. The feature can be enabled or disabled at boot and runtime via the `kcsan.weak_memory` boot parameter. Each memory access for which a watchpoint is set up, is also selected for simulated reordering within the scope of its function (at most 1 in-flight access). We are limited to modeling the effects of "buffering" (delaying the access), since the runtime cannot "prefetch" accesses (therefore no acquire modeling). Once an access has been selected for reordering, it is checked along every other access until the end of the function scope. If an appropriate memory barrier is encountered, the access will no longer be considered for reordering. When the result of a memory operation should be ordered by a barrier, KCSAN can then detect data races where the conflict only occurs as a result of a missing barrier due to reordering accesses. Suggested-by: Dmitry Vyukov Signed-off-by: Marco Elver Signed-off-by: Paul E. McKenney --- include/linux/kcsan-checks.h | 10 ++- include/linux/kcsan.h | 10 ++- include/linux/sched.h | 3 + kernel/kcsan/core.c | 202 +++++++++++++++++++++++++++++++++++++++---- lib/Kconfig.kcsan | 20 +++++ scripts/Makefile.kcsan | 9 +- 6 files changed, 235 insertions(+), 19 deletions(-) (limited to 'lib') diff --git a/include/linux/kcsan-checks.h b/include/linux/kcsan-checks.h index 5f5965246877..a1c6a89fde71 100644 --- a/include/linux/kcsan-checks.h +++ b/include/linux/kcsan-checks.h @@ -99,7 +99,15 @@ void kcsan_set_access_mask(unsigned long mask); /* Scoped access information. */ struct kcsan_scoped_access { - struct list_head list; + union { + struct list_head list; /* scoped_accesses list */ + /* + * Not an entry in scoped_accesses list; stack depth from where + * the access was initialized. + */ + int stack_depth; + }; + /* Access information. */ const volatile void *ptr; size_t size; diff --git a/include/linux/kcsan.h b/include/linux/kcsan.h index 13cef3458fed..c07c71f5ba4f 100644 --- a/include/linux/kcsan.h +++ b/include/linux/kcsan.h @@ -49,8 +49,16 @@ struct kcsan_ctx { */ unsigned long access_mask; - /* List of scoped accesses. */ + /* List of scoped accesses; likely to be empty. */ struct list_head scoped_accesses; + +#ifdef CONFIG_KCSAN_WEAK_MEMORY + /* + * Scoped access for modeling access reordering to detect missing memory + * barriers; only keep 1 to keep fast-path complexity manageable. + */ + struct kcsan_scoped_access reorder_access; +#endif }; /** diff --git a/include/linux/sched.h b/include/linux/sched.h index 78c351e35fec..0cd40b010487 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -1339,6 +1339,9 @@ struct task_struct { #ifdef CONFIG_TRACE_IRQFLAGS struct irqtrace_events kcsan_save_irqtrace; #endif +#ifdef CONFIG_KCSAN_WEAK_MEMORY + int kcsan_stack_depth; +#endif #endif #if IS_ENABLED(CONFIG_KUNIT) diff --git a/kernel/kcsan/core.c b/kernel/kcsan/core.c index bd359f8ee63a..481f8a524089 100644 --- a/kernel/kcsan/core.c +++ b/kernel/kcsan/core.c @@ -40,6 +40,13 @@ module_param_named(udelay_interrupt, kcsan_udelay_interrupt, uint, 0644); module_param_named(skip_watch, kcsan_skip_watch, long, 0644); module_param_named(interrupt_watcher, kcsan_interrupt_watcher, bool, 0444); +#ifdef CONFIG_KCSAN_WEAK_MEMORY +static bool kcsan_weak_memory = true; +module_param_named(weak_memory, kcsan_weak_memory, bool, 0644); +#else +#define kcsan_weak_memory false +#endif + bool kcsan_enabled; /* Per-CPU kcsan_ctx for interrupts */ @@ -351,6 +358,67 @@ void kcsan_restore_irqtrace(struct task_struct *task) #endif } +static __always_inline int get_kcsan_stack_depth(void) +{ +#ifdef CONFIG_KCSAN_WEAK_MEMORY + return current->kcsan_stack_depth; +#else + BUILD_BUG(); + return 0; +#endif +} + +static __always_inline void add_kcsan_stack_depth(int val) +{ +#ifdef CONFIG_KCSAN_WEAK_MEMORY + current->kcsan_stack_depth += val; +#else + BUILD_BUG(); +#endif +} + +static __always_inline struct kcsan_scoped_access *get_reorder_access(struct kcsan_ctx *ctx) +{ +#ifdef CONFIG_KCSAN_WEAK_MEMORY + return ctx->disable_scoped ? NULL : &ctx->reorder_access; +#else + return NULL; +#endif +} + +static __always_inline bool +find_reorder_access(struct kcsan_ctx *ctx, const volatile void *ptr, size_t size, + int type, unsigned long ip) +{ + struct kcsan_scoped_access *reorder_access = get_reorder_access(ctx); + + if (!reorder_access) + return false; + + /* + * Note: If accesses are repeated while reorder_access is identical, + * never matches the new access, because !(type & KCSAN_ACCESS_SCOPED). + */ + return reorder_access->ptr == ptr && reorder_access->size == size && + reorder_access->type == type && reorder_access->ip == ip; +} + +static inline void +set_reorder_access(struct kcsan_ctx *ctx, const volatile void *ptr, size_t size, + int type, unsigned long ip) +{ + struct kcsan_scoped_access *reorder_access = get_reorder_access(ctx); + + if (!reorder_access || !kcsan_weak_memory) + return; + + reorder_access->ptr = ptr; + reorder_access->size = size; + reorder_access->type = type | KCSAN_ACCESS_SCOPED; + reorder_access->ip = ip; + reorder_access->stack_depth = get_kcsan_stack_depth(); +} + /* * Pull everything together: check_access() below contains the performance * critical operations; the fast-path (including check_access) functions should @@ -389,8 +457,10 @@ static noinline void kcsan_found_watchpoint(const volatile void *ptr, * The access_mask check relies on value-change comparison. To avoid * reporting a race where e.g. the writer set up the watchpoint, but the * reader has access_mask!=0, we have to ignore the found watchpoint. + * + * reorder_access is never created from an access with access_mask set. */ - if (ctx->access_mask) + if (ctx->access_mask && !find_reorder_access(ctx, ptr, size, type, ip)) return; /* @@ -440,11 +510,13 @@ kcsan_setup_watchpoint(const volatile void *ptr, size_t size, int type, unsigned const bool is_assert = (type & KCSAN_ACCESS_ASSERT) != 0; atomic_long_t *watchpoint; u64 old, new, diff; - unsigned long access_mask; enum kcsan_value_change value_change = KCSAN_VALUE_CHANGE_MAYBE; + bool interrupt_watcher = kcsan_interrupt_watcher; unsigned long ua_flags = user_access_save(); struct kcsan_ctx *ctx = get_ctx(); + unsigned long access_mask = ctx->access_mask; unsigned long irq_flags = 0; + bool is_reorder_access; /* * Always reset kcsan_skip counter in slow-path to avoid underflow; see @@ -467,6 +539,17 @@ kcsan_setup_watchpoint(const volatile void *ptr, size_t size, int type, unsigned goto out; } + /* + * The local CPU cannot observe reordering of its own accesses, and + * therefore we need to take care of 2 cases to avoid false positives: + * + * 1. Races of the reordered access with interrupts. To avoid, if + * the current access is reorder_access, disable interrupts. + * 2. Avoid races of scoped accesses from nested interrupts (below). + */ + is_reorder_access = find_reorder_access(ctx, ptr, size, type, ip); + if (is_reorder_access) + interrupt_watcher = false; /* * Avoid races of scoped accesses from nested interrupts (or scheduler). * Assume setting up a watchpoint for a non-scoped (normal) access that @@ -482,7 +565,7 @@ kcsan_setup_watchpoint(const volatile void *ptr, size_t size, int type, unsigned * information is lost if dirtied by KCSAN. */ kcsan_save_irqtrace(current); - if (!kcsan_interrupt_watcher) + if (!interrupt_watcher) local_irq_save(irq_flags); watchpoint = insert_watchpoint((unsigned long)ptr, size, is_write); @@ -503,7 +586,7 @@ kcsan_setup_watchpoint(const volatile void *ptr, size_t size, int type, unsigned * Read the current value, to later check and infer a race if the data * was modified via a non-instrumented access, e.g. from a device. */ - old = read_instrumented_memory(ptr, size); + old = is_reorder_access ? 0 : read_instrumented_memory(ptr, size); /* * Delay this thread, to increase probability of observing a racy @@ -515,8 +598,17 @@ kcsan_setup_watchpoint(const volatile void *ptr, size_t size, int type, unsigned * Re-read value, and check if it is as expected; if not, we infer a * racy access. */ - access_mask = ctx->access_mask; - new = read_instrumented_memory(ptr, size); + if (!is_reorder_access) { + new = read_instrumented_memory(ptr, size); + } else { + /* + * Reordered accesses cannot be used for value change detection, + * because the memory location may no longer be accessible and + * could result in a fault. + */ + new = 0; + access_mask = 0; + } diff = old ^ new; if (access_mask) @@ -585,11 +677,20 @@ kcsan_setup_watchpoint(const volatile void *ptr, size_t size, int type, unsigned */ remove_watchpoint(watchpoint); atomic_long_dec(&kcsan_counters[KCSAN_COUNTER_USED_WATCHPOINTS]); + out_unlock: - if (!kcsan_interrupt_watcher) + if (!interrupt_watcher) local_irq_restore(irq_flags); kcsan_restore_irqtrace(current); ctx->disable_scoped--; + + /* + * Reordered accesses cannot be used for value change detection, + * therefore never consider for reordering if access_mask is set. + * ASSERT_EXCLUSIVE are not real accesses, ignore them as well. + */ + if (!access_mask && !is_assert) + set_reorder_access(ctx, ptr, size, type, ip); out: user_access_restore(ua_flags); } @@ -597,7 +698,6 @@ out: static __always_inline void check_access(const volatile void *ptr, size_t size, int type, unsigned long ip) { - const bool is_write = (type & KCSAN_ACCESS_WRITE) != 0; atomic_long_t *watchpoint; long encoded_watchpoint; @@ -608,12 +708,14 @@ check_access(const volatile void *ptr, size_t size, int type, unsigned long ip) if (unlikely(size == 0)) return; +again: /* * Avoid user_access_save in fast-path: find_watchpoint is safe without * user_access_save, as the address that ptr points to is only used to * check if a watchpoint exists; ptr is never dereferenced. */ - watchpoint = find_watchpoint((unsigned long)ptr, size, !is_write, + watchpoint = find_watchpoint((unsigned long)ptr, size, + !(type & KCSAN_ACCESS_WRITE), &encoded_watchpoint); /* * It is safe to check kcsan_is_enabled() after find_watchpoint in the @@ -627,9 +729,42 @@ check_access(const volatile void *ptr, size_t size, int type, unsigned long ip) else { struct kcsan_ctx *ctx = get_ctx(); /* Call only once in fast-path. */ - if (unlikely(should_watch(ctx, ptr, size, type))) + if (unlikely(should_watch(ctx, ptr, size, type))) { kcsan_setup_watchpoint(ptr, size, type, ip); - else if (unlikely(ctx->scoped_accesses.prev)) + return; + } + + if (!(type & KCSAN_ACCESS_SCOPED)) { + struct kcsan_scoped_access *reorder_access = get_reorder_access(ctx); + + if (reorder_access) { + /* + * reorder_access check: simulates reordering of + * the access after subsequent operations. + */ + ptr = reorder_access->ptr; + type = reorder_access->type; + ip = reorder_access->ip; + /* + * Upon a nested interrupt, this context's + * reorder_access can be modified (shared ctx). + * We know that upon return, reorder_access is + * always invalidated by setting size to 0 via + * __tsan_func_exit(). Therefore we must read + * and check size after the other fields. + */ + barrier(); + size = READ_ONCE(reorder_access->size); + if (size) + goto again; + } + } + + /* + * Always checked last, right before returning from runtime; + * if reorder_access is valid, checked after it was checked. + */ + if (unlikely(ctx->scoped_accesses.prev)) kcsan_check_scoped_accesses(); } } @@ -916,19 +1051,56 @@ DEFINE_TSAN_VOLATILE_READ_WRITE(8); DEFINE_TSAN_VOLATILE_READ_WRITE(16); /* - * The below are not required by KCSAN, but can still be emitted by the - * compiler. + * Function entry and exit are used to determine the validty of reorder_access. + * Reordering of the access ends at the end of the function scope where the + * access happened. This is done for two reasons: + * + * 1. Artificially limits the scope where missing barriers are detected. + * This minimizes false positives due to uninstrumented functions that + * contain the required barriers but were missed. + * + * 2. Simplifies generating the stack trace of the access. */ void __tsan_func_entry(void *call_pc); -void __tsan_func_entry(void *call_pc) +noinline void __tsan_func_entry(void *call_pc) { + if (!IS_ENABLED(CONFIG_KCSAN_WEAK_MEMORY)) + return; + + add_kcsan_stack_depth(1); } EXPORT_SYMBOL(__tsan_func_entry); + void __tsan_func_exit(void); -void __tsan_func_exit(void) +noinline void __tsan_func_exit(void) { + struct kcsan_scoped_access *reorder_access; + + if (!IS_ENABLED(CONFIG_KCSAN_WEAK_MEMORY)) + return; + + reorder_access = get_reorder_access(get_ctx()); + if (!reorder_access) + goto out; + + if (get_kcsan_stack_depth() <= reorder_access->stack_depth) { + /* + * Access check to catch cases where write without a barrier + * (supposed release) was last access in function: because + * instrumentation is inserted before the real access, a data + * race due to the write giving up a c-s would only be caught if + * we do the conflicting access after. + */ + check_access(reorder_access->ptr, reorder_access->size, + reorder_access->type, reorder_access->ip); + reorder_access->size = 0; + reorder_access->stack_depth = INT_MIN; + } +out: + add_kcsan_stack_depth(-1); } EXPORT_SYMBOL(__tsan_func_exit); + void __tsan_init(void); void __tsan_init(void) { diff --git a/lib/Kconfig.kcsan b/lib/Kconfig.kcsan index e0a93ffdef30..e4394ea8068b 100644 --- a/lib/Kconfig.kcsan +++ b/lib/Kconfig.kcsan @@ -191,6 +191,26 @@ config KCSAN_STRICT closely aligns with the rules defined by the Linux-kernel memory consistency model (LKMM). +config KCSAN_WEAK_MEMORY + bool "Enable weak memory modeling to detect missing memory barriers" + default y + depends on KCSAN_STRICT + # We can either let objtool nop __tsan_func_{entry,exit}() and builtin + # atomics instrumentation in .noinstr.text, or use a compiler that can + # implement __no_kcsan to really remove all instrumentation. + depends on STACK_VALIDATION || CC_IS_GCC + help + Enable support for modeling a subset of weak memory, which allows + detecting a subset of data races due to missing memory barriers. + + Depends on KCSAN_STRICT, because the options strenghtening certain + plain accesses by default (depending on !KCSAN_STRICT) reduce the + ability to detect any data races invoving reordered accesses, in + particular reordered writes. + + Weak memory modeling relies on additional instrumentation and may + affect performance. + config KCSAN_REPORT_VALUE_CHANGE_ONLY bool "Only report races where watcher observed a data value change" default y diff --git a/scripts/Makefile.kcsan b/scripts/Makefile.kcsan index 37cb504c77e1..4c7f0d282e42 100644 --- a/scripts/Makefile.kcsan +++ b/scripts/Makefile.kcsan @@ -9,7 +9,12 @@ endif # Keep most options here optional, to allow enabling more compilers if absence # of some options does not break KCSAN nor causes false positive reports. -export CFLAGS_KCSAN := -fsanitize=thread \ - $(call cc-option,$(call cc-param,tsan-instrument-func-entry-exit=0) -fno-optimize-sibling-calls) \ +kcsan-cflags := -fsanitize=thread -fno-optimize-sibling-calls \ $(call cc-option,$(call cc-param,tsan-compound-read-before-write=1),$(call cc-option,$(call cc-param,tsan-instrument-read-before-write=1))) \ $(call cc-param,tsan-distinguish-volatile=1) + +ifndef CONFIG_KCSAN_WEAK_MEMORY +kcsan-cflags += $(call cc-option,$(call cc-param,tsan-instrument-func-entry-exit=0)) +endif + +export CFLAGS_KCSAN := $(kcsan-cflags) -- cgit v1.2.3 From bd3d5bd1a0ad386475ea7a3de8a91e7d8a600536 Mon Sep 17 00:00:00 2001 From: Marco Elver Date: Tue, 30 Nov 2021 12:44:33 +0100 Subject: kcsan: Support WEAK_MEMORY with Clang where no objtool support exists Clang and GCC behave a little differently when it comes to the __no_sanitize_thread attribute, which has valid reasons, and depending on context either one could be right. Traditionally, user space ThreadSanitizer [1] still expects instrumented builtin atomics (to avoid false positives) and __tsan_func_{entry,exit} (to generate meaningful stack traces), even if the function has the attribute no_sanitize("thread"). [1] https://clang.llvm.org/docs/ThreadSanitizer.html#attribute-no-sanitize-thread GCC doesn't follow the same policy (for better or worse), and removes all kinds of instrumentation if no_sanitize is added. Arguably, since this may be a problem for user space ThreadSanitizer, we expect this may change in future. Since KCSAN != ThreadSanitizer, the likelihood of false positives even without barrier instrumentation everywhere, is much lower by design. At least for Clang, however, to fully remove all sanitizer instrumentation, we must add the disable_sanitizer_instrumentation attribute, which is available since Clang 14.0. Signed-off-by: Marco Elver Signed-off-by: Paul E. McKenney --- include/linux/compiler_types.h | 13 ++++++++++++- lib/Kconfig.kcsan | 2 +- 2 files changed, 13 insertions(+), 2 deletions(-) (limited to 'lib') diff --git a/include/linux/compiler_types.h b/include/linux/compiler_types.h index 1d32f4c03c9e..3c1795fdb568 100644 --- a/include/linux/compiler_types.h +++ b/include/linux/compiler_types.h @@ -198,9 +198,20 @@ struct ftrace_likely_data { # define __no_kasan_or_inline __always_inline #endif -#define __no_kcsan __no_sanitize_thread #ifdef __SANITIZE_THREAD__ +/* + * Clang still emits instrumentation for __tsan_func_{entry,exit}() and builtin + * atomics even with __no_sanitize_thread (to avoid false positives in userspace + * ThreadSanitizer). The kernel's requirements are stricter and we really do not + * want any instrumentation with __no_kcsan. + * + * Therefore we add __disable_sanitizer_instrumentation where available to + * disable all instrumentation. See Kconfig.kcsan where this is mandatory. + */ +# define __no_kcsan __no_sanitize_thread __disable_sanitizer_instrumentation # define __no_sanitize_or_inline __no_kcsan notrace __maybe_unused +#else +# define __no_kcsan #endif #ifndef __no_sanitize_or_inline diff --git a/lib/Kconfig.kcsan b/lib/Kconfig.kcsan index e4394ea8068b..63b70b8c5551 100644 --- a/lib/Kconfig.kcsan +++ b/lib/Kconfig.kcsan @@ -198,7 +198,7 @@ config KCSAN_WEAK_MEMORY # We can either let objtool nop __tsan_func_{entry,exit}() and builtin # atomics instrumentation in .noinstr.text, or use a compiler that can # implement __no_kcsan to really remove all instrumentation. - depends on STACK_VALIDATION || CC_IS_GCC + depends on STACK_VALIDATION || CC_IS_GCC || CLANG_VERSION >= 140000 help Enable support for modeling a subset of weak memory, which allows detecting a subset of data races due to missing memory barriers. -- cgit v1.2.3 From 5fb6e8cf53b005d287d4c2d137a415ff7d025a81 Mon Sep 17 00:00:00 2001 From: Mark Rutland Date: Fri, 26 Nov 2021 11:59:23 +0000 Subject: locking/atomic: atomic64: Remove unusable atomic ops The generic atomic64 implementation provides: * atomic64_and_return() * atomic64_or_return() * atomic64_xor_return() ... but none of these exist in the standard atomic64 API as described by scripts/atomic/atomics.tbl, and none of these have prototypes exposed by . The lkp kernel test robot noted this results in warnings when building with W=1: lib/atomic64.c:82:5: warning: no previous prototype for 'generic_atomic64_and_return' [-Wmissing-prototypes] lib/atomic64.c:82:5: warning: no previous prototype for 'generic_atomic64_or_return' [-Wmissing-prototypes] lib/atomic64.c:82:5: warning: no previous prototype for 'generic_atomic64_xor_return' [-Wmissing-prototypes] This appears to have been a thinko in commit: 28aa2bda2211f432 ("locking/atomic: Implement atomic{,64,_long}_fetch_{add,sub,and,andnot,or,xor}{,_relaxed,_acquire,_release}()") ... where we grouped add/sub separately from and/ox/xor, so that we could avoid implementing _return forms for the latter group, but forgot to remove ATOMIC64_OP_RETURN() for that group. This doesn't cause any functional problem, but it's pointless to build code which cannot be used. Remove the unusable code. This does not affect add/sub, for which _return forms will still be built. Reported-by: kernel test robot Signed-off-by: Mark Rutland Signed-off-by: Peter Zijlstra (Intel) Acked-by: Boqun Feng Link: https://lore.kernel.org/r/20211126115923.41489-1-mark.rutland@arm.com --- lib/atomic64.c | 2 -- 1 file changed, 2 deletions(-) (limited to 'lib') diff --git a/lib/atomic64.c b/lib/atomic64.c index 3df653994177..caf895789a1e 100644 --- a/lib/atomic64.c +++ b/lib/atomic64.c @@ -118,7 +118,6 @@ ATOMIC64_OPS(sub, -=) #undef ATOMIC64_OPS #define ATOMIC64_OPS(op, c_op) \ ATOMIC64_OP(op, c_op) \ - ATOMIC64_OP_RETURN(op, c_op) \ ATOMIC64_FETCH_OP(op, c_op) ATOMIC64_OPS(and, &=) @@ -127,7 +126,6 @@ ATOMIC64_OPS(xor, ^=) #undef ATOMIC64_OPS #undef ATOMIC64_FETCH_OP -#undef ATOMIC64_OP_RETURN #undef ATOMIC64_OP s64 generic_atomic64_dec_if_positive(atomic64_t *v) -- cgit v1.2.3 From cead18552660702a4a46f58e65188fe5f36e9dfe Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Mon, 22 Nov 2021 11:15:19 -0600 Subject: exit: Rename complete_and_exit to kthread_complete_and_exit Update complete_and_exit to call kthread_exit instead of do_exit. Change the name to reflect this change in functionality. All of the users of complete_and_exit are causing the current kthread to exit so this change makes it clear what is happening. Move the implementation of kthread_complete_and_exit from kernel/exit.c to to kernel/kthread.c. As this function is kthread specific it makes most sense to live with the kthread functions. There are no functional change. Signed-off-by: "Eric W. Biederman" --- drivers/net/wireless/rsi/rsi_91x_coex.c | 2 +- drivers/net/wireless/rsi/rsi_91x_main.c | 2 +- drivers/net/wireless/rsi/rsi_91x_sdio_ops.c | 2 +- drivers/net/wireless/rsi/rsi_91x_usb_ops.c | 2 +- drivers/pnp/pnpbios/core.c | 6 +++--- drivers/staging/rts5208/rtsx.c | 16 ++++++++-------- drivers/usb/atm/usbatm.c | 2 +- drivers/usb/gadget/function/f_mass_storage.c | 2 +- fs/jffs2/background.c | 2 +- include/linux/kernel.h | 1 - include/linux/kthread.h | 1 + kernel/exit.c | 9 --------- kernel/kthread.c | 21 +++++++++++++++++++++ lib/kunit/try-catch.c | 4 ++-- tools/objtool/check.c | 2 +- 15 files changed, 43 insertions(+), 31 deletions(-) (limited to 'lib') diff --git a/drivers/net/wireless/rsi/rsi_91x_coex.c b/drivers/net/wireless/rsi/rsi_91x_coex.c index a0c5d02ae88c..8a3d86897ea8 100644 --- a/drivers/net/wireless/rsi/rsi_91x_coex.c +++ b/drivers/net/wireless/rsi/rsi_91x_coex.c @@ -63,7 +63,7 @@ static void rsi_coex_scheduler_thread(struct rsi_common *common) rsi_coex_sched_tx_pkts(coex_cb); } while (atomic_read(&coex_cb->coex_tx_thread.thread_done) == 0); - complete_and_exit(&coex_cb->coex_tx_thread.completion, 0); + kthread_complete_and_exit(&coex_cb->coex_tx_thread.completion, 0); } int rsi_coex_recv_pkt(struct rsi_common *common, u8 *msg) diff --git a/drivers/net/wireless/rsi/rsi_91x_main.c b/drivers/net/wireless/rsi/rsi_91x_main.c index f1bf71e6c608..c7f5cec5e446 100644 --- a/drivers/net/wireless/rsi/rsi_91x_main.c +++ b/drivers/net/wireless/rsi/rsi_91x_main.c @@ -260,7 +260,7 @@ static void rsi_tx_scheduler_thread(struct rsi_common *common) if (common->init_done) rsi_core_qos_processor(common); } while (atomic_read(&common->tx_thread.thread_done) == 0); - complete_and_exit(&common->tx_thread.completion, 0); + kthread_complete_and_exit(&common->tx_thread.completion, 0); } #ifdef CONFIG_RSI_COEX diff --git a/drivers/net/wireless/rsi/rsi_91x_sdio_ops.c b/drivers/net/wireless/rsi/rsi_91x_sdio_ops.c index 8ace1874e5cb..b2b47a0abcbf 100644 --- a/drivers/net/wireless/rsi/rsi_91x_sdio_ops.c +++ b/drivers/net/wireless/rsi/rsi_91x_sdio_ops.c @@ -75,7 +75,7 @@ void rsi_sdio_rx_thread(struct rsi_common *common) rsi_dbg(INFO_ZONE, "%s: Terminated SDIO RX thread\n", __func__); atomic_inc(&sdev->rx_thread.thread_done); - complete_and_exit(&sdev->rx_thread.completion, 0); + kthread_complete_and_exit(&sdev->rx_thread.completion, 0); } /** diff --git a/drivers/net/wireless/rsi/rsi_91x_usb_ops.c b/drivers/net/wireless/rsi/rsi_91x_usb_ops.c index 4ffcdde1acb1..5130b0e72adc 100644 --- a/drivers/net/wireless/rsi/rsi_91x_usb_ops.c +++ b/drivers/net/wireless/rsi/rsi_91x_usb_ops.c @@ -56,6 +56,6 @@ void rsi_usb_rx_thread(struct rsi_common *common) out: rsi_dbg(INFO_ZONE, "%s: Terminated thread\n", __func__); skb_queue_purge(&dev->rx_q); - complete_and_exit(&dev->rx_thread.completion, 0); + kthread_complete_and_exit(&dev->rx_thread.completion, 0); } diff --git a/drivers/pnp/pnpbios/core.c b/drivers/pnp/pnpbios/core.c index 669ef4700c1a..f7e86ae9f72f 100644 --- a/drivers/pnp/pnpbios/core.c +++ b/drivers/pnp/pnpbios/core.c @@ -160,7 +160,7 @@ static int pnp_dock_thread(void *unused) * No dock to manage */ case PNP_FUNCTION_NOT_SUPPORTED: - complete_and_exit(&unload_sem, 0); + kthread_complete_and_exit(&unload_sem, 0); case PNP_SYSTEM_NOT_DOCKED: d = 0; break; @@ -170,7 +170,7 @@ static int pnp_dock_thread(void *unused) default: pnpbios_print_status("pnp_dock_thread", status); printk(KERN_WARNING "PnPBIOS: disabling dock monitoring.\n"); - complete_and_exit(&unload_sem, 0); + kthread_complete_and_exit(&unload_sem, 0); } if (d != docked) { if (pnp_dock_event(d, &now) == 0) { @@ -183,7 +183,7 @@ static int pnp_dock_thread(void *unused) } } } - complete_and_exit(&unload_sem, 0); + kthread_complete_and_exit(&unload_sem, 0); } static int pnpbios_get_resources(struct pnp_dev *dev) diff --git a/drivers/staging/rts5208/rtsx.c b/drivers/staging/rts5208/rtsx.c index 91fcf85e150a..5a58dac76c88 100644 --- a/drivers/staging/rts5208/rtsx.c +++ b/drivers/staging/rts5208/rtsx.c @@ -450,13 +450,13 @@ skip_for_abort: * after the down() -- that's necessary for the thread-shutdown * case. * - * complete_and_exit() goes even further than this -- it is safe in - * the case that the thread of the caller is going away (not just - * the structure) -- this is necessary for the module-remove case. - * This is important in preemption kernels, which transfer the flow - * of execution immediately upon a complete(). + * kthread_complete_and_exit() goes even further than this -- + * it is safe in the case that the thread of the caller is going away + * (not just the structure) -- this is necessary for the module-remove + * case. This is important in preemption kernels, which transfer the + * flow of execution immediately upon a complete(). */ - complete_and_exit(&dev->control_exit, 0); + kthread_complete_and_exit(&dev->control_exit, 0); } static int rtsx_polling_thread(void *__dev) @@ -501,7 +501,7 @@ static int rtsx_polling_thread(void *__dev) mutex_unlock(&dev->dev_mutex); } - complete_and_exit(&dev->polling_exit, 0); + kthread_complete_and_exit(&dev->polling_exit, 0); } /* @@ -682,7 +682,7 @@ static int rtsx_scan_thread(void *__dev) /* Should we unbind if no devices were detected? */ } - complete_and_exit(&dev->scanning_done, 0); + kthread_complete_and_exit(&dev->scanning_done, 0); } static void rtsx_init_options(struct rtsx_chip *chip) diff --git a/drivers/usb/atm/usbatm.c b/drivers/usb/atm/usbatm.c index da17be1ef64e..e3a49d837609 100644 --- a/drivers/usb/atm/usbatm.c +++ b/drivers/usb/atm/usbatm.c @@ -969,7 +969,7 @@ static int usbatm_do_heavy_init(void *arg) instance->thread = NULL; mutex_unlock(&instance->serialize); - complete_and_exit(&instance->thread_exited, ret); + kthread_complete_and_exit(&instance->thread_exited, ret); } static int usbatm_heavy_init(struct usbatm_data *instance) diff --git a/drivers/usb/gadget/function/f_mass_storage.c b/drivers/usb/gadget/function/f_mass_storage.c index 752439690fda..46dd11dcb3a8 100644 --- a/drivers/usb/gadget/function/f_mass_storage.c +++ b/drivers/usb/gadget/function/f_mass_storage.c @@ -2547,7 +2547,7 @@ static int fsg_main_thread(void *common_) up_write(&common->filesem); /* Let fsg_unbind() know the thread has exited */ - complete_and_exit(&common->thread_notifier, 0); + kthread_complete_and_exit(&common->thread_notifier, 0); } diff --git a/fs/jffs2/background.c b/fs/jffs2/background.c index 2b4d5013dc5d..6da92ecaf66d 100644 --- a/fs/jffs2/background.c +++ b/fs/jffs2/background.c @@ -161,5 +161,5 @@ static int jffs2_garbage_collect_thread(void *_c) spin_lock(&c->erase_completion_lock); c->gc_task = NULL; spin_unlock(&c->erase_completion_lock); - complete_and_exit(&c->gc_thread_exit, 0); + kthread_complete_and_exit(&c->gc_thread_exit, 0); } diff --git a/include/linux/kernel.h b/include/linux/kernel.h index 77755ac3e189..055eb203c00e 100644 --- a/include/linux/kernel.h +++ b/include/linux/kernel.h @@ -187,7 +187,6 @@ static inline void might_fault(void) { } #endif void do_exit(long error_code) __noreturn; -void complete_and_exit(struct completion *, long) __noreturn; extern int num_to_str(char *buf, int size, unsigned long long num, unsigned int width); diff --git a/include/linux/kthread.h b/include/linux/kthread.h index 22c43d419687..d86a7e3b9a52 100644 --- a/include/linux/kthread.h +++ b/include/linux/kthread.h @@ -71,6 +71,7 @@ int kthread_park(struct task_struct *k); void kthread_unpark(struct task_struct *k); void kthread_parkme(void); void kthread_exit(long result) __noreturn; +void kthread_complete_and_exit(struct completion *, long) __noreturn; int kthreadd(void *unused); extern struct task_struct *kthreadd_task; diff --git a/kernel/exit.c b/kernel/exit.c index 57afac845a0a..6c4b04531f17 100644 --- a/kernel/exit.c +++ b/kernel/exit.c @@ -891,15 +891,6 @@ void __noreturn make_task_dead(int signr) do_exit(signr); } -void complete_and_exit(struct completion *comp, long code) -{ - if (comp) - complete(comp); - - do_exit(code); -} -EXPORT_SYMBOL(complete_and_exit); - SYSCALL_DEFINE1(exit, int, error_code) { do_exit((error_code&0xff)<<8); diff --git a/kernel/kthread.c b/kernel/kthread.c index 77b7c3f23f18..4388d6694a7f 100644 --- a/kernel/kthread.c +++ b/kernel/kthread.c @@ -283,6 +283,27 @@ void __noreturn kthread_exit(long result) do_exit(result); } +/** + * kthread_complete_and exit - Exit the current kthread. + * @comp: Completion to complete + * @code: The integer value to return to kthread_stop(). + * + * If present complete @comp and the reuturn code to kthread_stop(). + * + * A kernel thread whose module may be removed after the completion of + * @comp can use this function exit safely. + * + * Does not return. + */ +void __noreturn kthread_complete_and_exit(struct completion *comp, long code) +{ + if (comp) + complete(comp); + + kthread_exit(code); +} +EXPORT_SYMBOL(kthread_complete_and_exit); + static int kthread(void *_create) { static const struct sched_param param = { .sched_priority = 0 }; diff --git a/lib/kunit/try-catch.c b/lib/kunit/try-catch.c index 0dd434e40487..be38a2c5ecc2 100644 --- a/lib/kunit/try-catch.c +++ b/lib/kunit/try-catch.c @@ -17,7 +17,7 @@ void __noreturn kunit_try_catch_throw(struct kunit_try_catch *try_catch) { try_catch->try_result = -EFAULT; - complete_and_exit(try_catch->try_completion, -EFAULT); + kthread_complete_and_exit(try_catch->try_completion, -EFAULT); } EXPORT_SYMBOL_GPL(kunit_try_catch_throw); @@ -27,7 +27,7 @@ static int kunit_generic_run_threadfn_adapter(void *data) try_catch->try(try_catch->context); - complete_and_exit(try_catch->try_completion, 0); + kthread_complete_and_exit(try_catch->try_completion, 0); } static unsigned long kunit_test_timeout(void) diff --git a/tools/objtool/check.c b/tools/objtool/check.c index 120e9598c11a..282273a1ffa5 100644 --- a/tools/objtool/check.c +++ b/tools/objtool/check.c @@ -171,7 +171,7 @@ static bool __dead_end_function(struct objtool_file *file, struct symbol *func, "kthread_exit", "make_task_dead", "__module_put_and_kthread_exit", - "complete_and_exit", + "kthread_complete_and_exit", "__reiserfs_panic", "lbug_with_loc", "fortify_panic", -- cgit v1.2.3 From 37dbb4c7c7442dbfc9b651e4ddd4afe30b26afc9 Mon Sep 17 00:00:00 2001 From: David Gow Date: Tue, 2 Nov 2021 00:30:13 -0700 Subject: kunit: Don't crash if no parameters are generated It's possible that a parameterised test could end up with zero parameters. At the moment, the test function will nevertheless be called with NULL as the parameter. Instead, don't try to run the test code, and just mark the test as SKIPped. Reported-by: Daniel Latypov Signed-off-by: David Gow Reviewed-by: Daniel Latypov Reviewed-by: Brendan Higgins Signed-off-by: Shuah Khan --- lib/kunit/test.c | 18 ++++++++++-------- 1 file changed, 10 insertions(+), 8 deletions(-) (limited to 'lib') diff --git a/lib/kunit/test.c b/lib/kunit/test.c index 3bd741e50a2d..f96498ede2cc 100644 --- a/lib/kunit/test.c +++ b/lib/kunit/test.c @@ -504,16 +504,18 @@ int kunit_run_tests(struct kunit_suite *suite) struct kunit_result_stats param_stats = { 0 }; test_case->status = KUNIT_SKIPPED; - if (test_case->generate_params) { + if (!test_case->generate_params) { + /* Non-parameterised test. */ + kunit_run_case_catch_errors(suite, test_case, &test); + kunit_update_stats(¶m_stats, test.status); + } else { /* Get initial param. */ param_desc[0] = '\0'; test.param_value = test_case->generate_params(NULL, param_desc); - } - do { - kunit_run_case_catch_errors(suite, test_case, &test); + while (test.param_value) { + kunit_run_case_catch_errors(suite, test_case, &test); - if (test_case->generate_params) { if (param_desc[0] == '\0') { snprintf(param_desc, sizeof(param_desc), "param-%d", test.param_index); @@ -530,11 +532,11 @@ int kunit_run_tests(struct kunit_suite *suite) param_desc[0] = '\0'; test.param_value = test_case->generate_params(test.param_value, param_desc); test.param_index++; - } - kunit_update_stats(¶m_stats, test.status); + kunit_update_stats(¶m_stats, test.status); + } + } - } while (test.param_value); kunit_print_test_stats(&test, param_stats); -- cgit v1.2.3 From 44b7da5fcd4c99de1ec5cc783cdd605398246280 Mon Sep 17 00:00:00 2001 From: David Gow Date: Tue, 2 Nov 2021 00:30:14 -0700 Subject: kunit: Report test parameter results as (K)TAP subtests Currently, the results for individial parameters in a parameterised test are simply output as (K)TAP diagnostic lines. As kunit_tool now supports nested subtests, report each parameter as its own subtest. For example, here's what the output now looks like: # Subtest: inode_test_xtimestamp_decoding ok 1 - 1901-12-13 Lower bound of 32bit < 0 timestamp, no extra bits ok 2 - 1969-12-31 Upper bound of 32bit < 0 timestamp, no extra bits ok 3 - 1970-01-01 Lower bound of 32bit >=0 timestamp, no extra bits ok 4 - 2038-01-19 Upper bound of 32bit >=0 timestamp, no extra bits ok 5 - 2038-01-19 Lower bound of 32bit <0 timestamp, lo extra sec bit on ok 6 - 2106-02-07 Upper bound of 32bit <0 timestamp, lo extra sec bit on ok 7 - 2106-02-07 Lower bound of 32bit >=0 timestamp, lo extra sec bit on ok 8 - 2174-02-25 Upper bound of 32bit >=0 timestamp, lo extra sec bit on ok 9 - 2174-02-25 Lower bound of 32bit <0 timestamp, hi extra sec bit on ok 10 - 2242-03-16 Upper bound of 32bit <0 timestamp, hi extra sec bit on ok 11 - 2242-03-16 Lower bound of 32bit >=0 timestamp, hi extra sec bit on ok 12 - 2310-04-04 Upper bound of 32bit >=0 timestamp, hi extra sec bit on ok 13 - 2310-04-04 Upper bound of 32bit>=0 timestamp, hi extra sec bit 1. 1 ns ok 14 - 2378-04-22 Lower bound of 32bit>= timestamp. Extra sec bits 1. Max ns ok 15 - 2378-04-22 Lower bound of 32bit >=0 timestamp. All extra sec bits on ok 16 - 2446-05-10 Upper bound of 32bit >=0 timestamp. All extra sec bits on # inode_test_xtimestamp_decoding: pass:16 fail:0 skip:0 total:16 ok 1 - inode_test_xtimestamp_decoding Signed-off-by: David Gow Reviewed-by: Daniel Latypov Reviewed-by: Brendan Higgins Signed-off-by: Shuah Khan --- lib/kunit/test.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) (limited to 'lib') diff --git a/lib/kunit/test.c b/lib/kunit/test.c index f96498ede2cc..c7ed4aabec04 100644 --- a/lib/kunit/test.c +++ b/lib/kunit/test.c @@ -512,6 +512,8 @@ int kunit_run_tests(struct kunit_suite *suite) /* Get initial param. */ param_desc[0] = '\0'; test.param_value = test_case->generate_params(NULL, param_desc); + kunit_log(KERN_INFO, &test, KUNIT_SUBTEST_INDENT KUNIT_SUBTEST_INDENT + "# Subtest: %s", test_case->name); while (test.param_value) { kunit_run_case_catch_errors(suite, test_case, &test); @@ -522,9 +524,8 @@ int kunit_run_tests(struct kunit_suite *suite) } kunit_log(KERN_INFO, &test, - KUNIT_SUBTEST_INDENT - "# %s: %s %d - %s", - test_case->name, + KUNIT_SUBTEST_INDENT KUNIT_SUBTEST_INDENT + "%s %d - %s", kunit_status_to_ok_not_ok(test.status), test.param_index + 1, param_desc); -- cgit v1.2.3 From 4e84139e14af5ea60772cc4f33d7059aec76e0eb Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Wed, 15 Sep 2021 20:30:21 +0200 Subject: lib/logic_iomem: Fix 32-bit build On a 32-bit build, the (unsigned long long) casts throw warnings (or errors) due to being to a different integer size. Cast to uintptr_t first (with the __force for sparse) and then further to get the consistent print on 32 and 64-bit. Fixes: ca2e334232b6 ("lib: add iomem emulation (logic_iomem)") Reported-by: Al Viro Signed-off-by: Johannes Berg Signed-off-by: Richard Weinberger --- lib/logic_iomem.c | 13 +++++++------ 1 file changed, 7 insertions(+), 6 deletions(-) (limited to 'lib') diff --git a/lib/logic_iomem.c b/lib/logic_iomem.c index 9bdfde0c0f86..54fa601f3300 100644 --- a/lib/logic_iomem.c +++ b/lib/logic_iomem.c @@ -79,7 +79,7 @@ static void __iomem *real_ioremap(phys_addr_t offset, size_t size) static void real_iounmap(void __iomem *addr) { WARN(1, "invalid iounmap for addr 0x%llx\n", - (unsigned long long __force)addr); + (unsigned long long)(uintptr_t __force)addr); } #endif /* CONFIG_LOGIC_IOMEM_FALLBACK */ @@ -173,7 +173,7 @@ EXPORT_SYMBOL(iounmap); static u##sz real_raw_read ## op(const volatile void __iomem *addr) \ { \ WARN(1, "Invalid read" #op " at address %llx\n", \ - (unsigned long long __force)addr); \ + (unsigned long long)(uintptr_t __force)addr); \ return (u ## sz)~0ULL; \ } \ \ @@ -181,7 +181,8 @@ static void real_raw_write ## op(u ## sz val, \ volatile void __iomem *addr) \ { \ WARN(1, "Invalid writeq" #op " of 0x%llx at address %llx\n", \ - (unsigned long long)val, (unsigned long long __force)addr);\ + (unsigned long long)val, \ + (unsigned long long)(uintptr_t __force)addr);\ } \ MAKE_FALLBACK(b, 8); @@ -194,14 +195,14 @@ MAKE_FALLBACK(q, 64); static void real_memset_io(volatile void __iomem *addr, int value, size_t size) { WARN(1, "Invalid memset_io at address 0x%llx\n", - (unsigned long long __force)addr); + (unsigned long long)(uintptr_t __force)addr); } static void real_memcpy_fromio(void *buffer, const volatile void __iomem *addr, size_t size) { WARN(1, "Invalid memcpy_fromio at address 0x%llx\n", - (unsigned long long __force)addr); + (unsigned long long)(uintptr_t __force)addr); memset(buffer, 0xff, size); } @@ -210,7 +211,7 @@ static void real_memcpy_toio(volatile void __iomem *addr, const void *buffer, size_t size) { WARN(1, "Invalid memcpy_toio at address 0x%llx\n", - (unsigned long long __force)addr); + (unsigned long long)(uintptr_t __force)addr); } #endif /* CONFIG_LOGIC_IOMEM_FALLBACK */ -- cgit v1.2.3 From 4e8a5edac5010820e7c5303fc96f5a262e096bb6 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Wed, 15 Sep 2021 20:30:22 +0200 Subject: lib/logic_iomem: Fix operation on 32-bit On 32-bit, the first entry might be at 0/NULL, but that's strange and leads to issues, e.g. where we check "if (ret)". Use a IOREMAP_BIAS/IOREMAP_MASK of 0x80000000UL to avoid this. This then requires reducing the number of areas (via MAX_AREAS), but we still have 128 areas, which is enough. Fixes: ca2e334232b6 ("lib: add iomem emulation (logic_iomem)") Signed-off-by: Johannes Berg Signed-off-by: Richard Weinberger --- lib/logic_iomem.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) (limited to 'lib') diff --git a/lib/logic_iomem.c b/lib/logic_iomem.c index 54fa601f3300..549b22d4bcde 100644 --- a/lib/logic_iomem.c +++ b/lib/logic_iomem.c @@ -21,15 +21,15 @@ struct logic_iomem_area { #define AREA_SHIFT 24 #define MAX_AREA_SIZE (1 << AREA_SHIFT) -#define MAX_AREAS ((1ULL<<32) / MAX_AREA_SIZE) +#define MAX_AREAS ((1U << 31) / MAX_AREA_SIZE) #define AREA_BITS ((MAX_AREAS - 1) << AREA_SHIFT) #define AREA_MASK (MAX_AREA_SIZE - 1) #ifdef CONFIG_64BIT #define IOREMAP_BIAS 0xDEAD000000000000UL #define IOREMAP_MASK 0xFFFFFFFF00000000UL #else -#define IOREMAP_BIAS 0 -#define IOREMAP_MASK 0 +#define IOREMAP_BIAS 0x80000000UL +#define IOREMAP_MASK 0x80000000UL #endif static DEFINE_MUTEX(regions_mtx); -- cgit v1.2.3 From 5f174ec3c1d62013f86db6597249174d8cb227b2 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Mon, 20 Sep 2021 21:32:49 +0000 Subject: logic_io instance of iounmap() needs volatile on argument ... same as the rest of implementations Signed-off-by: Al Viro Signed-off-by: Richard Weinberger --- include/asm-generic/logic_io.h | 2 +- lib/logic_iomem.c | 4 ++-- 2 files changed, 3 insertions(+), 3 deletions(-) (limited to 'lib') diff --git a/include/asm-generic/logic_io.h b/include/asm-generic/logic_io.h index a53116b8c57e..8a59b6e567df 100644 --- a/include/asm-generic/logic_io.h +++ b/include/asm-generic/logic_io.h @@ -34,7 +34,7 @@ void __iomem *ioremap(phys_addr_t offset, size_t size); #define iounmap iounmap -void iounmap(void __iomem *addr); +void iounmap(void volatile __iomem *addr); #define __raw_readb __raw_readb u8 __raw_readb(const volatile void __iomem *addr); diff --git a/lib/logic_iomem.c b/lib/logic_iomem.c index 549b22d4bcde..8c3365f26e51 100644 --- a/lib/logic_iomem.c +++ b/lib/logic_iomem.c @@ -76,7 +76,7 @@ static void __iomem *real_ioremap(phys_addr_t offset, size_t size) return NULL; } -static void real_iounmap(void __iomem *addr) +static void real_iounmap(volatile void __iomem *addr) { WARN(1, "invalid iounmap for addr 0x%llx\n", (unsigned long long)(uintptr_t __force)addr); @@ -149,7 +149,7 @@ get_area(const volatile void __iomem *addr) return NULL; } -void iounmap(void __iomem *addr) +void iounmap(volatile void __iomem *addr) { struct logic_iomem_area *area = get_area(addr); -- cgit v1.2.3 From 7c63f26cb518f57aa5feaef7e0abe2697155df35 Mon Sep 17 00:00:00 2001 From: Christophe JAILLET Date: Thu, 23 Dec 2021 22:33:42 +0100 Subject: lib: objagg: Use the bitmap API when applicable Use 'bitmap_zalloc()' to simplify code, improve the semantic and reduce some open-coded arithmetic in allocator arguments. Also change the corresponding 'kfree()' into 'bitmap_free()' to keep consistency. Signed-off-by: Christophe JAILLET Link: https://lore.kernel.org/r/f9541b085ec68e573004e1be200c11c9c901181a.1640295165.git.christophe.jaillet@wanadoo.fr Signed-off-by: Jakub Kicinski --- lib/objagg.c | 7 ++----- 1 file changed, 2 insertions(+), 5 deletions(-) (limited to 'lib') diff --git a/lib/objagg.c b/lib/objagg.c index 5e1676ccdadd..1e248629ed64 100644 --- a/lib/objagg.c +++ b/lib/objagg.c @@ -781,7 +781,6 @@ static struct objagg_tmp_graph *objagg_tmp_graph_create(struct objagg *objagg) struct objagg_tmp_node *node; struct objagg_tmp_node *pnode; struct objagg_obj *objagg_obj; - size_t alloc_size; int i, j; graph = kzalloc(sizeof(*graph), GFP_KERNEL); @@ -793,9 +792,7 @@ static struct objagg_tmp_graph *objagg_tmp_graph_create(struct objagg *objagg) goto err_nodes_alloc; graph->nodes_count = nodes_count; - alloc_size = BITS_TO_LONGS(nodes_count * nodes_count) * - sizeof(unsigned long); - graph->edges = kzalloc(alloc_size, GFP_KERNEL); + graph->edges = bitmap_zalloc(nodes_count * nodes_count, GFP_KERNEL); if (!graph->edges) goto err_edges_alloc; @@ -833,7 +830,7 @@ err_nodes_alloc: static void objagg_tmp_graph_destroy(struct objagg_tmp_graph *graph) { - kfree(graph->edges); + bitmap_free(graph->edges); kfree(graph->nodes); kfree(graph); } -- cgit v1.2.3 From ee6d3dd4ed48ab24b74bab3c3977b8218518247d Mon Sep 17 00:00:00 2001 From: Wedson Almeida Filho Date: Fri, 24 Dec 2021 23:13:45 +0000 Subject: driver core: make kobj_type constant. This way instances of kobj_type (which contain function pointers) can be stored in .rodata, which means that they cannot be [easily/accidentally] modified at runtime. Signed-off-by: Wedson Almeida Filho Link: https://lore.kernel.org/r/20211224231345.777370-1-wedsonaf@google.com Signed-off-by: Greg Kroah-Hartman --- Documentation/core-api/kobject.rst | 4 ++-- drivers/base/bus.c | 2 +- drivers/base/core.c | 2 +- include/linux/kobject.h | 8 ++++---- kernel/params.c | 2 +- lib/kobject.c | 8 ++++---- 6 files changed, 13 insertions(+), 13 deletions(-) (limited to 'lib') diff --git a/Documentation/core-api/kobject.rst b/Documentation/core-api/kobject.rst index 2739f8b72575..d3b5bf9f643a 100644 --- a/Documentation/core-api/kobject.rst +++ b/Documentation/core-api/kobject.rst @@ -118,7 +118,7 @@ Initialization of kobjects Code which creates a kobject must, of course, initialize that object. Some of the internal fields are setup with a (mandatory) call to kobject_init():: - void kobject_init(struct kobject *kobj, struct kobj_type *ktype); + void kobject_init(struct kobject *kobj, const struct kobj_type *ktype); The ktype is required for a kobject to be created properly, as every kobject must have an associated kobj_type. After calling kobject_init(), to @@ -156,7 +156,7 @@ kobject_name():: There is a helper function to both initialize and add the kobject to the kernel at the same time, called surprisingly enough kobject_init_and_add():: - int kobject_init_and_add(struct kobject *kobj, struct kobj_type *ktype, + int kobject_init_and_add(struct kobject *kobj, const struct kobj_type *ktype, struct kobject *parent, const char *fmt, ...); The arguments are the same as the individual kobject_init() and diff --git a/drivers/base/bus.c b/drivers/base/bus.c index bdc98c5713d5..a64454f5f8c0 100644 --- a/drivers/base/bus.c +++ b/drivers/base/bus.c @@ -165,7 +165,7 @@ static struct kobj_type bus_ktype = { static int bus_uevent_filter(struct kset *kset, struct kobject *kobj) { - struct kobj_type *ktype = get_ktype(kobj); + const struct kobj_type *ktype = get_ktype(kobj); if (ktype == &bus_ktype) return 1; diff --git a/drivers/base/core.c b/drivers/base/core.c index fd034d742447..d712ea11066b 100644 --- a/drivers/base/core.c +++ b/drivers/base/core.c @@ -2263,7 +2263,7 @@ static struct kobj_type device_ktype = { static int dev_uevent_filter(struct kset *kset, struct kobject *kobj) { - struct kobj_type *ktype = get_ktype(kobj); + const struct kobj_type *ktype = get_ktype(kobj); if (ktype == &device_ktype) { struct device *dev = kobj_to_dev(kobj); diff --git a/include/linux/kobject.h b/include/linux/kobject.h index c740062b4b1a..683172b2e094 100644 --- a/include/linux/kobject.h +++ b/include/linux/kobject.h @@ -66,7 +66,7 @@ struct kobject { struct list_head entry; struct kobject *parent; struct kset *kset; - struct kobj_type *ktype; + const struct kobj_type *ktype; struct kernfs_node *sd; /* sysfs directory entry */ struct kref kref; #ifdef CONFIG_DEBUG_KOBJECT_RELEASE @@ -90,13 +90,13 @@ static inline const char *kobject_name(const struct kobject *kobj) return kobj->name; } -extern void kobject_init(struct kobject *kobj, struct kobj_type *ktype); +extern void kobject_init(struct kobject *kobj, const struct kobj_type *ktype); extern __printf(3, 4) __must_check int kobject_add(struct kobject *kobj, struct kobject *parent, const char *fmt, ...); extern __printf(4, 5) __must_check int kobject_init_and_add(struct kobject *kobj, - struct kobj_type *ktype, struct kobject *parent, + const struct kobj_type *ktype, struct kobject *parent, const char *fmt, ...); extern void kobject_del(struct kobject *kobj); @@ -217,7 +217,7 @@ static inline void kset_put(struct kset *k) kobject_put(&k->kobj); } -static inline struct kobj_type *get_ktype(struct kobject *kobj) +static inline const struct kobj_type *get_ktype(struct kobject *kobj) { return kobj->ktype; } diff --git a/kernel/params.c b/kernel/params.c index 8299bd764e42..9b90e3c4d3c0 100644 --- a/kernel/params.c +++ b/kernel/params.c @@ -928,7 +928,7 @@ static const struct sysfs_ops module_sysfs_ops = { static int uevent_filter(struct kset *kset, struct kobject *kobj) { - struct kobj_type *ktype = get_ktype(kobj); + const struct kobj_type *ktype = get_ktype(kobj); if (ktype == &module_ktype) return 1; diff --git a/lib/kobject.c b/lib/kobject.c index 4a56f519139d..56fa037501b5 100644 --- a/lib/kobject.c +++ b/lib/kobject.c @@ -65,7 +65,7 @@ void kobject_get_ownership(struct kobject *kobj, kuid_t *uid, kgid_t *gid) */ static int populate_dir(struct kobject *kobj) { - struct kobj_type *t = get_ktype(kobj); + const struct kobj_type *t = get_ktype(kobj); struct attribute *attr; int error = 0; int i; @@ -346,7 +346,7 @@ EXPORT_SYMBOL(kobject_set_name); * to kobject_put(), not by a call to kfree directly to ensure that all of * the memory is cleaned up properly. */ -void kobject_init(struct kobject *kobj, struct kobj_type *ktype) +void kobject_init(struct kobject *kobj, const struct kobj_type *ktype) { char *err_str; @@ -461,7 +461,7 @@ EXPORT_SYMBOL(kobject_add); * same type of error handling after a call to kobject_add() and kobject * lifetime rules are the same here. */ -int kobject_init_and_add(struct kobject *kobj, struct kobj_type *ktype, +int kobject_init_and_add(struct kobject *kobj, const struct kobj_type *ktype, struct kobject *parent, const char *fmt, ...) { va_list args; @@ -679,7 +679,7 @@ EXPORT_SYMBOL(kobject_get_unless_zero); static void kobject_cleanup(struct kobject *kobj) { struct kobject *parent = kobj->parent; - struct kobj_type *t = get_ktype(kobj); + const struct kobj_type *t = get_ktype(kobj); const char *name = kobj->name; pr_debug("kobject: '%s' (%p): %s, parent %p\n", -- cgit v1.2.3 From cf6299b6101903c31bddb0065804b2121ed510c7 Mon Sep 17 00:00:00 2001 From: Greg Kroah-Hartman Date: Mon, 27 Dec 2021 17:39:24 +0100 Subject: kobject: remove kset from struct kset_uevent_ops callbacks There is no need to pass the pointer to the kset in the struct kset_uevent_ops callbacks as no one uses it, so just remove that pointer entirely. Reviewed-by: Rafael J. Wysocki Reviewed-by: Wedson Almeida Filho Link: https://lore.kernel.org/r/20211227163924.3970661-1-gregkh@linuxfoundation.org Signed-off-by: Greg Kroah-Hartman --- Documentation/core-api/kobject.rst | 7 +++---- Documentation/translations/zh_CN/core-api/kobject.rst | 7 +++---- drivers/base/bus.c | 2 +- drivers/base/core.c | 11 +++++------ drivers/dma-buf/dma-buf-sysfs-stats.c | 2 +- fs/dlm/lockspace.c | 3 +-- fs/gfs2/sys.c | 3 +-- include/linux/kobject.h | 7 +++---- kernel/params.c | 2 +- lib/kobject_uevent.c | 6 +++--- 10 files changed, 22 insertions(+), 28 deletions(-) (limited to 'lib') diff --git a/Documentation/core-api/kobject.rst b/Documentation/core-api/kobject.rst index d3b5bf9f643a..3d6e3107315d 100644 --- a/Documentation/core-api/kobject.rst +++ b/Documentation/core-api/kobject.rst @@ -373,10 +373,9 @@ If a kset wishes to control the uevent operations of the kobjects associated with it, it can use the struct kset_uevent_ops to handle it:: struct kset_uevent_ops { - int (* const filter)(struct kset *kset, struct kobject *kobj); - const char *(* const name)(struct kset *kset, struct kobject *kobj); - int (* const uevent)(struct kset *kset, struct kobject *kobj, - struct kobj_uevent_env *env); + int (* const filter)(struct kobject *kobj); + const char *(* const name)(struct kobject *kobj); + int (* const uevent)(struct kobject *kobj, struct kobj_uevent_env *env); }; diff --git a/Documentation/translations/zh_CN/core-api/kobject.rst b/Documentation/translations/zh_CN/core-api/kobject.rst index b7c37794cc7f..95634083dca0 100644 --- a/Documentation/translations/zh_CN/core-api/kobject.rst +++ b/Documentation/translations/zh_CN/core-api/kobject.rst @@ -325,10 +325,9 @@ ksets 结构体kset_uevent_ops来处理它:: struct kset_uevent_ops { - int (* const filter)(struct kset *kset, struct kobject *kobj); - const char *(* const name)(struct kset *kset, struct kobject *kobj); - int (* const uevent)(struct kset *kset, struct kobject *kobj, - struct kobj_uevent_env *env); + int (* const filter)(struct kobject *kobj); + const char *(* const name)(struct kobject *kobj); + int (* const uevent)(struct kobject *kobj, struct kobj_uevent_env *env); }; diff --git a/drivers/base/bus.c b/drivers/base/bus.c index a64454f5f8c0..97936ec49bde 100644 --- a/drivers/base/bus.c +++ b/drivers/base/bus.c @@ -163,7 +163,7 @@ static struct kobj_type bus_ktype = { .release = bus_release, }; -static int bus_uevent_filter(struct kset *kset, struct kobject *kobj) +static int bus_uevent_filter(struct kobject *kobj) { const struct kobj_type *ktype = get_ktype(kobj); diff --git a/drivers/base/core.c b/drivers/base/core.c index d712ea11066b..60d703ebd123 100644 --- a/drivers/base/core.c +++ b/drivers/base/core.c @@ -2261,7 +2261,7 @@ static struct kobj_type device_ktype = { }; -static int dev_uevent_filter(struct kset *kset, struct kobject *kobj) +static int dev_uevent_filter(struct kobject *kobj) { const struct kobj_type *ktype = get_ktype(kobj); @@ -2275,7 +2275,7 @@ static int dev_uevent_filter(struct kset *kset, struct kobject *kobj) return 0; } -static const char *dev_uevent_name(struct kset *kset, struct kobject *kobj) +static const char *dev_uevent_name(struct kobject *kobj) { struct device *dev = kobj_to_dev(kobj); @@ -2286,8 +2286,7 @@ static const char *dev_uevent_name(struct kset *kset, struct kobject *kobj) return NULL; } -static int dev_uevent(struct kset *kset, struct kobject *kobj, - struct kobj_uevent_env *env) +static int dev_uevent(struct kobject *kobj, struct kobj_uevent_env *env) { struct device *dev = kobj_to_dev(kobj); int retval = 0; @@ -2382,7 +2381,7 @@ static ssize_t uevent_show(struct device *dev, struct device_attribute *attr, /* respect filter */ if (kset->uevent_ops && kset->uevent_ops->filter) - if (!kset->uevent_ops->filter(kset, &dev->kobj)) + if (!kset->uevent_ops->filter(&dev->kobj)) goto out; env = kzalloc(sizeof(struct kobj_uevent_env), GFP_KERNEL); @@ -2390,7 +2389,7 @@ static ssize_t uevent_show(struct device *dev, struct device_attribute *attr, return -ENOMEM; /* let the kset specific function add its keys */ - retval = kset->uevent_ops->uevent(kset, &dev->kobj, env); + retval = kset->uevent_ops->uevent(&dev->kobj, env); if (retval) goto out; diff --git a/drivers/dma-buf/dma-buf-sysfs-stats.c b/drivers/dma-buf/dma-buf-sysfs-stats.c index 053baadcada9..2bba0babcb62 100644 --- a/drivers/dma-buf/dma-buf-sysfs-stats.c +++ b/drivers/dma-buf/dma-buf-sysfs-stats.c @@ -132,7 +132,7 @@ void dma_buf_stats_teardown(struct dma_buf *dmabuf) /* Statistics files do not need to send uevents. */ -static int dmabuf_sysfs_uevent_filter(struct kset *kset, struct kobject *kobj) +static int dmabuf_sysfs_uevent_filter(struct kobject *kobj) { return 0; } diff --git a/fs/dlm/lockspace.c b/fs/dlm/lockspace.c index 10eddfa6c3d7..0bbb346cb892 100644 --- a/fs/dlm/lockspace.c +++ b/fs/dlm/lockspace.c @@ -216,8 +216,7 @@ static int do_uevent(struct dlm_ls *ls, int in) return ls->ls_uevent_result; } -static int dlm_uevent(struct kset *kset, struct kobject *kobj, - struct kobj_uevent_env *env) +static int dlm_uevent(struct kobject *kobj, struct kobj_uevent_env *env) { struct dlm_ls *ls = container_of(kobj, struct dlm_ls, ls_kobj); diff --git a/fs/gfs2/sys.c b/fs/gfs2/sys.c index c0a34d9ddee4..a6002b2d146d 100644 --- a/fs/gfs2/sys.c +++ b/fs/gfs2/sys.c @@ -767,8 +767,7 @@ void gfs2_sys_fs_del(struct gfs2_sbd *sdp) wait_for_completion(&sdp->sd_kobj_unregister); } -static int gfs2_uevent(struct kset *kset, struct kobject *kobj, - struct kobj_uevent_env *env) +static int gfs2_uevent(struct kobject *kobj, struct kobj_uevent_env *env) { struct gfs2_sbd *sdp = container_of(kobj, struct gfs2_sbd, sd_kobj); struct super_block *s = sdp->sd_vfs; diff --git a/include/linux/kobject.h b/include/linux/kobject.h index 683172b2e094..ad90b49824dc 100644 --- a/include/linux/kobject.h +++ b/include/linux/kobject.h @@ -153,10 +153,9 @@ struct kobj_uevent_env { }; struct kset_uevent_ops { - int (* const filter)(struct kset *kset, struct kobject *kobj); - const char *(* const name)(struct kset *kset, struct kobject *kobj); - int (* const uevent)(struct kset *kset, struct kobject *kobj, - struct kobj_uevent_env *env); + int (* const filter)(struct kobject *kobj); + const char *(* const name)(struct kobject *kobj); + int (* const uevent)(struct kobject *kobj, struct kobj_uevent_env *env); }; struct kobj_attribute { diff --git a/kernel/params.c b/kernel/params.c index 9b90e3c4d3c0..5b92310425c5 100644 --- a/kernel/params.c +++ b/kernel/params.c @@ -926,7 +926,7 @@ static const struct sysfs_ops module_sysfs_ops = { .store = module_attr_store, }; -static int uevent_filter(struct kset *kset, struct kobject *kobj) +static int uevent_filter(struct kobject *kobj) { const struct kobj_type *ktype = get_ktype(kobj); diff --git a/lib/kobject_uevent.c b/lib/kobject_uevent.c index c87d5b6a8a55..7c44b7ae4c5c 100644 --- a/lib/kobject_uevent.c +++ b/lib/kobject_uevent.c @@ -501,7 +501,7 @@ int kobject_uevent_env(struct kobject *kobj, enum kobject_action action, } /* skip the event, if the filter returns zero. */ if (uevent_ops && uevent_ops->filter) - if (!uevent_ops->filter(kset, kobj)) { + if (!uevent_ops->filter(kobj)) { pr_debug("kobject: '%s' (%p): %s: filter function " "caused the event to drop!\n", kobject_name(kobj), kobj, __func__); @@ -510,7 +510,7 @@ int kobject_uevent_env(struct kobject *kobj, enum kobject_action action, /* originating subsystem */ if (uevent_ops && uevent_ops->name) - subsystem = uevent_ops->name(kset, kobj); + subsystem = uevent_ops->name(kobj); else subsystem = kobject_name(&kset->kobj); if (!subsystem) { @@ -554,7 +554,7 @@ int kobject_uevent_env(struct kobject *kobj, enum kobject_action action, /* let the kset specific function add its stuff */ if (uevent_ops && uevent_ops->uevent) { - retval = uevent_ops->uevent(kset, kobj, env); + retval = uevent_ops->uevent(kobj, env); if (retval) { pr_debug("kobject: '%s' (%p): %s: uevent() returned " "%d\n", kobject_name(kobj), kobj, -- cgit v1.2.3 From 821979f5098b05dd1cc83075369304ed65dfef4e Mon Sep 17 00:00:00 2001 From: "Matthew Wilcox (Oracle)" Date: Sun, 28 Nov 2021 19:18:27 -0500 Subject: iov_iter: Convert iter_xarray to use folios Take advantage of how kmap_local_folio() works to simplify the loop. Signed-off-by: Matthew Wilcox (Oracle) Reviewed-by: Christoph Hellwig Reviewed-by: William Kucharski --- lib/iov_iter.c | 30 ++++++++++++++---------------- 1 file changed, 14 insertions(+), 16 deletions(-) (limited to 'lib') diff --git a/lib/iov_iter.c b/lib/iov_iter.c index 66a740e6e153..b0e0acdf96c1 100644 --- a/lib/iov_iter.c +++ b/lib/iov_iter.c @@ -69,42 +69,40 @@ #define iterate_xarray(i, n, base, len, __off, STEP) { \ __label__ __out; \ size_t __off = 0; \ - struct page *head = NULL; \ + struct folio *folio; \ loff_t start = i->xarray_start + i->iov_offset; \ - unsigned offset = start % PAGE_SIZE; \ pgoff_t index = start / PAGE_SIZE; \ - int j; \ - \ XA_STATE(xas, i->xarray, index); \ \ + len = PAGE_SIZE - offset_in_page(start); \ rcu_read_lock(); \ - xas_for_each(&xas, head, ULONG_MAX) { \ + xas_for_each(&xas, folio, ULONG_MAX) { \ unsigned left; \ - if (xas_retry(&xas, head)) \ + size_t offset; \ + if (xas_retry(&xas, folio)) \ continue; \ - if (WARN_ON(xa_is_value(head))) \ + if (WARN_ON(xa_is_value(folio))) \ break; \ - if (WARN_ON(PageHuge(head))) \ + if (WARN_ON(folio_test_hugetlb(folio))) \ break; \ - for (j = (head->index < index) ? index - head->index : 0; \ - j < thp_nr_pages(head); j++) { \ - void *kaddr = kmap_local_page(head + j); \ - base = kaddr + offset; \ - len = PAGE_SIZE - offset; \ + offset = offset_in_folio(folio, start + __off); \ + while (offset < folio_size(folio)) { \ + base = kmap_local_folio(folio, offset); \ len = min(n, len); \ left = (STEP); \ - kunmap_local(kaddr); \ + kunmap_local(base); \ len -= left; \ __off += len; \ n -= len; \ if (left || n == 0) \ goto __out; \ - offset = 0; \ + offset += len; \ + len = PAGE_SIZE; \ } \ } \ __out: \ rcu_read_unlock(); \ - i->iov_offset += __off; \ + i->iov_offset += __off; \ n = __off; \ } -- cgit v1.2.3 From 38640c480939d56cc8b03d58642fc5261761a697 Mon Sep 17 00:00:00 2001 From: Dirk Müller Date: Wed, 5 Jan 2022 17:38:46 +0100 Subject: lib/raid6: skip benchmark of non-chosen xor_syndrome functions MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit In commit fe5cbc6e06c7 ("md/raid6 algorithms: delta syndrome functions") a xor_syndrome() benchmarking was added also to the raid6_choose_gen() function. However, the results of that benchmarking were intentionally discarded and did not influence the choice. It picked the xor_syndrome() variant related to the best performing gen_syndrome(). Reduce runtime of raid6_choose_gen() without modifying its outcome by only benchmarking the xor_syndrome() of the best gen_syndrome() variant. For a HZ=250 x86_64 system with avx2 and without avx512 this removes 5 out of 6 xor() benchmarks, saving 340ms of raid6 initialization time. Signed-off-by: Dirk Müller Signed-off-by: Song Liu --- lib/raid6/algos.c | 76 +++++++++++++++++++++++++++---------------------------- 1 file changed, 37 insertions(+), 39 deletions(-) (limited to 'lib') diff --git a/lib/raid6/algos.c b/lib/raid6/algos.c index 6d5e5000fdd7..9b7e8a837b27 100644 --- a/lib/raid6/algos.c +++ b/lib/raid6/algos.c @@ -145,12 +145,12 @@ static inline const struct raid6_recov_calls *raid6_choose_recov(void) static inline const struct raid6_calls *raid6_choose_gen( void *(*const dptrs)[RAID6_TEST_DISKS], const int disks) { - unsigned long perf, bestgenperf, bestxorperf, j0, j1; + unsigned long perf, bestgenperf, j0, j1; int start = (disks>>1)-1, stop = disks-3; /* work on the second half of the disks */ const struct raid6_calls *const *algo; const struct raid6_calls *best; - for (bestgenperf = 0, bestxorperf = 0, best = NULL, algo = raid6_algos; *algo; algo++) { + for (bestgenperf = 0, best = NULL, algo = raid6_algos; *algo; algo++) { if (!best || (*algo)->prefer >= best->prefer) { if ((*algo)->valid && !(*algo)->valid()) continue; @@ -180,50 +180,48 @@ static inline const struct raid6_calls *raid6_choose_gen( pr_info("raid6: %-8s gen() %5ld MB/s\n", (*algo)->name, (perf * HZ * (disks-2)) >> (20 - PAGE_SHIFT + RAID6_TIME_JIFFIES_LG2)); + } + } - if (!(*algo)->xor_syndrome) - continue; + if (!best) { + pr_err("raid6: Yikes! No algorithm found!\n"); + goto out; + } - perf = 0; + raid6_call = *best; - preempt_disable(); - j0 = jiffies; - while ((j1 = jiffies) == j0) - cpu_relax(); - while (time_before(jiffies, - j1 + (1<xor_syndrome(disks, start, stop, - PAGE_SIZE, *dptrs); - perf++; - } - preempt_enable(); - - if (best == *algo) - bestxorperf = perf; + if (!IS_ENABLED(CONFIG_RAID6_PQ_BENCHMARK)) { + pr_info("raid6: skipped pq benchmark and selected %s\n", + best->name); + goto out; + } - pr_info("raid6: %-8s xor() %5ld MB/s\n", (*algo)->name, - (perf * HZ * (disks-2)) >> - (20 - PAGE_SHIFT + RAID6_TIME_JIFFIES_LG2 + 1)); + pr_info("raid6: using algorithm %s gen() %ld MB/s\n", + best->name, + (bestgenperf * HZ * (disks - 2)) >> + (20 - PAGE_SHIFT + RAID6_TIME_JIFFIES_LG2)); + + if (best->xor_syndrome) { + perf = 0; + + preempt_disable(); + j0 = jiffies; + while ((j1 = jiffies) == j0) + cpu_relax(); + while (time_before(jiffies, + j1 + (1 << RAID6_TIME_JIFFIES_LG2))) { + best->xor_syndrome(disks, start, stop, + PAGE_SIZE, *dptrs); + perf++; } - } + preempt_enable(); - if (best) { - if (IS_ENABLED(CONFIG_RAID6_PQ_BENCHMARK)) { - pr_info("raid6: using algorithm %s gen() %ld MB/s\n", - best->name, - (bestgenperf * HZ * (disks-2)) >> - (20 - PAGE_SHIFT+RAID6_TIME_JIFFIES_LG2)); - if (best->xor_syndrome) - pr_info("raid6: .... xor() %ld MB/s, rmw enabled\n", - (bestxorperf * HZ * (disks-2)) >> - (20 - PAGE_SHIFT + RAID6_TIME_JIFFIES_LG2 + 1)); - } else - pr_info("raid6: skip pq benchmark and using algorithm %s\n", - best->name); - raid6_call = *best; - } else - pr_err("raid6: Yikes! No algorithm found!\n"); + pr_info("raid6: .... xor() %ld MB/s, rmw enabled\n", + (perf * HZ * (disks - 2)) >> + (20 - PAGE_SHIFT + RAID6_TIME_JIFFIES_LG2 + 1)); + } +out: return best; } -- cgit v1.2.3 From 36dacddbf0bdba86cd00f066b4d724157eeb63f1 Mon Sep 17 00:00:00 2001 From: Dirk Müller Date: Wed, 5 Jan 2022 17:38:47 +0100 Subject: lib/raid6: Use strict priority ranking for pq gen() benchmarking MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit On x86_64, currently 3 variants of AVX512, 3 variants of AVX2 and 3 variants of SSE2 are benchmarked on initialization, taking between 144-153 jiffies. Testing across a hardware pool of various generations of intel cpus I could not find a single case where SSE2 won over AVX2 or AVX512. There are cases where AVX2 wins over AVX512 however. Change "prefer" into an integer priority field (similar to how recov selection works) to have more than one ranking level available, which is backwards compatible with existing behavior. Give AVX2/512 variants higher priority over SSE2 in order to skip SSE testing when AVX is available. in a AVX2/x86_64/HZ=250 case this saves in the order of 200ms of initialization time. Signed-off-by: Dirk Müller Acked-by: Paul Menzel Signed-off-by: Song Liu --- include/linux/raid/pq.h | 2 +- lib/raid6/algos.c | 2 +- lib/raid6/avx2.c | 8 ++++---- lib/raid6/avx512.c | 6 +++--- 4 files changed, 9 insertions(+), 9 deletions(-) (limited to 'lib') diff --git a/include/linux/raid/pq.h b/include/linux/raid/pq.h index 154e954b711d..d6e5a1feb947 100644 --- a/include/linux/raid/pq.h +++ b/include/linux/raid/pq.h @@ -81,7 +81,7 @@ struct raid6_calls { void (*xor_syndrome)(int, int, int, size_t, void **); int (*valid)(void); /* Returns 1 if this routine set is usable */ const char *name; /* Name of this routine set */ - int prefer; /* Has special performance attribute */ + int priority; /* Relative priority ranking if non-zero */ }; /* Selected algorithm */ diff --git a/lib/raid6/algos.c b/lib/raid6/algos.c index 9b7e8a837b27..39b74221f4a7 100644 --- a/lib/raid6/algos.c +++ b/lib/raid6/algos.c @@ -151,7 +151,7 @@ static inline const struct raid6_calls *raid6_choose_gen( const struct raid6_calls *best; for (bestgenperf = 0, best = NULL, algo = raid6_algos; *algo; algo++) { - if (!best || (*algo)->prefer >= best->prefer) { + if (!best || (*algo)->priority >= best->priority) { if ((*algo)->valid && !(*algo)->valid()) continue; diff --git a/lib/raid6/avx2.c b/lib/raid6/avx2.c index f299476e1d76..059024234dce 100644 --- a/lib/raid6/avx2.c +++ b/lib/raid6/avx2.c @@ -132,7 +132,7 @@ const struct raid6_calls raid6_avx2x1 = { raid6_avx21_xor_syndrome, raid6_have_avx2, "avx2x1", - 1 /* Has cache hints */ + .priority = 2 /* Prefer AVX2 over priority 1 (SSE2 and others) */ }; /* @@ -262,7 +262,7 @@ const struct raid6_calls raid6_avx2x2 = { raid6_avx22_xor_syndrome, raid6_have_avx2, "avx2x2", - 1 /* Has cache hints */ + .priority = 2 /* Prefer AVX2 over priority 1 (SSE2 and others) */ }; #ifdef CONFIG_X86_64 @@ -465,6 +465,6 @@ const struct raid6_calls raid6_avx2x4 = { raid6_avx24_xor_syndrome, raid6_have_avx2, "avx2x4", - 1 /* Has cache hints */ + .priority = 2 /* Prefer AVX2 over priority 1 (SSE2 and others) */ }; -#endif +#endif /* CONFIG_X86_64 */ diff --git a/lib/raid6/avx512.c b/lib/raid6/avx512.c index bb684d144ee2..9c3e822e1adf 100644 --- a/lib/raid6/avx512.c +++ b/lib/raid6/avx512.c @@ -162,7 +162,7 @@ const struct raid6_calls raid6_avx512x1 = { raid6_avx5121_xor_syndrome, raid6_have_avx512, "avx512x1", - 1 /* Has cache hints */ + .priority = 2 /* Prefer AVX512 over priority 1 (SSE2 and others) */ }; /* @@ -319,7 +319,7 @@ const struct raid6_calls raid6_avx512x2 = { raid6_avx5122_xor_syndrome, raid6_have_avx512, "avx512x2", - 1 /* Has cache hints */ + .priority = 2 /* Prefer AVX512 over priority 1 (SSE2 and others) */ }; #ifdef CONFIG_X86_64 @@ -557,7 +557,7 @@ const struct raid6_calls raid6_avx512x4 = { raid6_avx5124_xor_syndrome, raid6_have_avx512, "avx512x4", - 1 /* Has cache hints */ + .priority = 2 /* Prefer AVX512 over priority 1 (SSE2 and others) */ }; #endif -- cgit v1.2.3 From 6048fdcc5f269c7f31d774c295ce59081b36e6f9 Mon Sep 17 00:00:00 2001 From: "Jason A. Donenfeld" Date: Wed, 22 Dec 2021 14:56:58 +0100 Subject: lib/crypto: blake2s: include as built-in In preparation for using blake2s in the RNG, we change the way that it is wired-in to the build system. Instead of using ifdefs to select the right symbol, we use weak symbols. And because ARM doesn't need the generic implementation, we make the generic one default only if an arch library doesn't need it already, and then have arch libraries that do need it opt-in. So that the arch libraries can remain tristate rather than bool, we then split the shash part from the glue code. Acked-by: Herbert Xu Acked-by: Ard Biesheuvel Acked-by: Greg Kroah-Hartman Cc: Masahiro Yamada Cc: linux-kbuild@vger.kernel.org Cc: linux-crypto@vger.kernel.org Signed-off-by: Jason A. Donenfeld --- arch/arm/crypto/Makefile | 4 +- arch/arm/crypto/blake2s-core.S | 8 ++-- arch/arm/crypto/blake2s-glue.c | 73 +------------------------------------ arch/arm/crypto/blake2s-shash.c | 75 ++++++++++++++++++++++++++++++++++++++ arch/x86/crypto/Makefile | 4 +- arch/x86/crypto/blake2s-glue.c | 68 +++------------------------------- arch/x86/crypto/blake2s-shash.c | 77 +++++++++++++++++++++++++++++++++++++++ crypto/Kconfig | 3 +- drivers/net/Kconfig | 1 - include/crypto/internal/blake2s.h | 6 +-- lib/crypto/Kconfig | 23 +++--------- lib/crypto/Makefile | 9 ++--- lib/crypto/blake2s-generic.c | 6 ++- lib/crypto/blake2s.c | 6 --- 14 files changed, 189 insertions(+), 174 deletions(-) create mode 100644 arch/arm/crypto/blake2s-shash.c create mode 100644 arch/x86/crypto/blake2s-shash.c (limited to 'lib') diff --git a/arch/arm/crypto/Makefile b/arch/arm/crypto/Makefile index eafa898ba6a7..0274f81cc8ea 100644 --- a/arch/arm/crypto/Makefile +++ b/arch/arm/crypto/Makefile @@ -10,6 +10,7 @@ obj-$(CONFIG_CRYPTO_SHA1_ARM_NEON) += sha1-arm-neon.o obj-$(CONFIG_CRYPTO_SHA256_ARM) += sha256-arm.o obj-$(CONFIG_CRYPTO_SHA512_ARM) += sha512-arm.o obj-$(CONFIG_CRYPTO_BLAKE2S_ARM) += blake2s-arm.o +obj-$(if $(CONFIG_CRYPTO_BLAKE2S_ARM),y) += libblake2s-arm.o obj-$(CONFIG_CRYPTO_BLAKE2B_NEON) += blake2b-neon.o obj-$(CONFIG_CRYPTO_CHACHA20_NEON) += chacha-neon.o obj-$(CONFIG_CRYPTO_POLY1305_ARM) += poly1305-arm.o @@ -31,7 +32,8 @@ sha256-arm-neon-$(CONFIG_KERNEL_MODE_NEON) := sha256_neon_glue.o sha256-arm-y := sha256-core.o sha256_glue.o $(sha256-arm-neon-y) sha512-arm-neon-$(CONFIG_KERNEL_MODE_NEON) := sha512-neon-glue.o sha512-arm-y := sha512-core.o sha512-glue.o $(sha512-arm-neon-y) -blake2s-arm-y := blake2s-core.o blake2s-glue.o +blake2s-arm-y := blake2s-shash.o +libblake2s-arm-y:= blake2s-core.o blake2s-glue.o blake2b-neon-y := blake2b-neon-core.o blake2b-neon-glue.o sha1-arm-ce-y := sha1-ce-core.o sha1-ce-glue.o sha2-arm-ce-y := sha2-ce-core.o sha2-ce-glue.o diff --git a/arch/arm/crypto/blake2s-core.S b/arch/arm/crypto/blake2s-core.S index 86345751bbf3..df40e46601f1 100644 --- a/arch/arm/crypto/blake2s-core.S +++ b/arch/arm/crypto/blake2s-core.S @@ -167,8 +167,8 @@ .endm // -// void blake2s_compress_arch(struct blake2s_state *state, -// const u8 *block, size_t nblocks, u32 inc); +// void blake2s_compress(struct blake2s_state *state, +// const u8 *block, size_t nblocks, u32 inc); // // Only the first three fields of struct blake2s_state are used: // u32 h[8]; (inout) @@ -176,7 +176,7 @@ // u32 f[2]; (in) // .align 5 -ENTRY(blake2s_compress_arch) +ENTRY(blake2s_compress) push {r0-r2,r4-r11,lr} // keep this an even number .Lnext_block: @@ -303,4 +303,4 @@ ENTRY(blake2s_compress_arch) str r3, [r12], #4 bne 1b b .Lcopy_block_done -ENDPROC(blake2s_compress_arch) +ENDPROC(blake2s_compress) diff --git a/arch/arm/crypto/blake2s-glue.c b/arch/arm/crypto/blake2s-glue.c index f2cc1e5fc9ec..0238a70d9581 100644 --- a/arch/arm/crypto/blake2s-glue.c +++ b/arch/arm/crypto/blake2s-glue.c @@ -1,78 +1,7 @@ // SPDX-License-Identifier: GPL-2.0-or-later -/* - * BLAKE2s digest algorithm, ARM scalar implementation - * - * Copyright 2020 Google LLC - */ #include -#include - #include /* defined in blake2s-core.S */ -EXPORT_SYMBOL(blake2s_compress_arch); - -static int crypto_blake2s_update_arm(struct shash_desc *desc, - const u8 *in, unsigned int inlen) -{ - return crypto_blake2s_update(desc, in, inlen, blake2s_compress_arch); -} - -static int crypto_blake2s_final_arm(struct shash_desc *desc, u8 *out) -{ - return crypto_blake2s_final(desc, out, blake2s_compress_arch); -} - -#define BLAKE2S_ALG(name, driver_name, digest_size) \ - { \ - .base.cra_name = name, \ - .base.cra_driver_name = driver_name, \ - .base.cra_priority = 200, \ - .base.cra_flags = CRYPTO_ALG_OPTIONAL_KEY, \ - .base.cra_blocksize = BLAKE2S_BLOCK_SIZE, \ - .base.cra_ctxsize = sizeof(struct blake2s_tfm_ctx), \ - .base.cra_module = THIS_MODULE, \ - .digestsize = digest_size, \ - .setkey = crypto_blake2s_setkey, \ - .init = crypto_blake2s_init, \ - .update = crypto_blake2s_update_arm, \ - .final = crypto_blake2s_final_arm, \ - .descsize = sizeof(struct blake2s_state), \ - } - -static struct shash_alg blake2s_arm_algs[] = { - BLAKE2S_ALG("blake2s-128", "blake2s-128-arm", BLAKE2S_128_HASH_SIZE), - BLAKE2S_ALG("blake2s-160", "blake2s-160-arm", BLAKE2S_160_HASH_SIZE), - BLAKE2S_ALG("blake2s-224", "blake2s-224-arm", BLAKE2S_224_HASH_SIZE), - BLAKE2S_ALG("blake2s-256", "blake2s-256-arm", BLAKE2S_256_HASH_SIZE), -}; - -static int __init blake2s_arm_mod_init(void) -{ - return IS_REACHABLE(CONFIG_CRYPTO_HASH) ? - crypto_register_shashes(blake2s_arm_algs, - ARRAY_SIZE(blake2s_arm_algs)) : 0; -} - -static void __exit blake2s_arm_mod_exit(void) -{ - if (IS_REACHABLE(CONFIG_CRYPTO_HASH)) - crypto_unregister_shashes(blake2s_arm_algs, - ARRAY_SIZE(blake2s_arm_algs)); -} - -module_init(blake2s_arm_mod_init); -module_exit(blake2s_arm_mod_exit); - -MODULE_DESCRIPTION("BLAKE2s digest algorithm, ARM scalar implementation"); -MODULE_LICENSE("GPL"); -MODULE_AUTHOR("Eric Biggers "); -MODULE_ALIAS_CRYPTO("blake2s-128"); -MODULE_ALIAS_CRYPTO("blake2s-128-arm"); -MODULE_ALIAS_CRYPTO("blake2s-160"); -MODULE_ALIAS_CRYPTO("blake2s-160-arm"); -MODULE_ALIAS_CRYPTO("blake2s-224"); -MODULE_ALIAS_CRYPTO("blake2s-224-arm"); -MODULE_ALIAS_CRYPTO("blake2s-256"); -MODULE_ALIAS_CRYPTO("blake2s-256-arm"); +EXPORT_SYMBOL(blake2s_compress); diff --git a/arch/arm/crypto/blake2s-shash.c b/arch/arm/crypto/blake2s-shash.c new file mode 100644 index 000000000000..17c1c3bfe2f5 --- /dev/null +++ b/arch/arm/crypto/blake2s-shash.c @@ -0,0 +1,75 @@ +// SPDX-License-Identifier: GPL-2.0-or-later +/* + * BLAKE2s digest algorithm, ARM scalar implementation + * + * Copyright 2020 Google LLC + */ + +#include +#include + +#include + +static int crypto_blake2s_update_arm(struct shash_desc *desc, + const u8 *in, unsigned int inlen) +{ + return crypto_blake2s_update(desc, in, inlen, blake2s_compress); +} + +static int crypto_blake2s_final_arm(struct shash_desc *desc, u8 *out) +{ + return crypto_blake2s_final(desc, out, blake2s_compress); +} + +#define BLAKE2S_ALG(name, driver_name, digest_size) \ + { \ + .base.cra_name = name, \ + .base.cra_driver_name = driver_name, \ + .base.cra_priority = 200, \ + .base.cra_flags = CRYPTO_ALG_OPTIONAL_KEY, \ + .base.cra_blocksize = BLAKE2S_BLOCK_SIZE, \ + .base.cra_ctxsize = sizeof(struct blake2s_tfm_ctx), \ + .base.cra_module = THIS_MODULE, \ + .digestsize = digest_size, \ + .setkey = crypto_blake2s_setkey, \ + .init = crypto_blake2s_init, \ + .update = crypto_blake2s_update_arm, \ + .final = crypto_blake2s_final_arm, \ + .descsize = sizeof(struct blake2s_state), \ + } + +static struct shash_alg blake2s_arm_algs[] = { + BLAKE2S_ALG("blake2s-128", "blake2s-128-arm", BLAKE2S_128_HASH_SIZE), + BLAKE2S_ALG("blake2s-160", "blake2s-160-arm", BLAKE2S_160_HASH_SIZE), + BLAKE2S_ALG("blake2s-224", "blake2s-224-arm", BLAKE2S_224_HASH_SIZE), + BLAKE2S_ALG("blake2s-256", "blake2s-256-arm", BLAKE2S_256_HASH_SIZE), +}; + +static int __init blake2s_arm_mod_init(void) +{ + return IS_REACHABLE(CONFIG_CRYPTO_HASH) ? + crypto_register_shashes(blake2s_arm_algs, + ARRAY_SIZE(blake2s_arm_algs)) : 0; +} + +static void __exit blake2s_arm_mod_exit(void) +{ + if (IS_REACHABLE(CONFIG_CRYPTO_HASH)) + crypto_unregister_shashes(blake2s_arm_algs, + ARRAY_SIZE(blake2s_arm_algs)); +} + +module_init(blake2s_arm_mod_init); +module_exit(blake2s_arm_mod_exit); + +MODULE_DESCRIPTION("BLAKE2s digest algorithm, ARM scalar implementation"); +MODULE_LICENSE("GPL"); +MODULE_AUTHOR("Eric Biggers "); +MODULE_ALIAS_CRYPTO("blake2s-128"); +MODULE_ALIAS_CRYPTO("blake2s-128-arm"); +MODULE_ALIAS_CRYPTO("blake2s-160"); +MODULE_ALIAS_CRYPTO("blake2s-160-arm"); +MODULE_ALIAS_CRYPTO("blake2s-224"); +MODULE_ALIAS_CRYPTO("blake2s-224-arm"); +MODULE_ALIAS_CRYPTO("blake2s-256"); +MODULE_ALIAS_CRYPTO("blake2s-256-arm"); diff --git a/arch/x86/crypto/Makefile b/arch/x86/crypto/Makefile index f307c93fc90a..c3af959648e6 100644 --- a/arch/x86/crypto/Makefile +++ b/arch/x86/crypto/Makefile @@ -62,7 +62,9 @@ obj-$(CONFIG_CRYPTO_SHA512_SSSE3) += sha512-ssse3.o sha512-ssse3-y := sha512-ssse3-asm.o sha512-avx-asm.o sha512-avx2-asm.o sha512_ssse3_glue.o obj-$(CONFIG_CRYPTO_BLAKE2S_X86) += blake2s-x86_64.o -blake2s-x86_64-y := blake2s-core.o blake2s-glue.o +blake2s-x86_64-y := blake2s-shash.o +obj-$(if $(CONFIG_CRYPTO_BLAKE2S_X86),y) += libblake2s-x86_64.o +libblake2s-x86_64-y := blake2s-core.o blake2s-glue.o obj-$(CONFIG_CRYPTO_GHASH_CLMUL_NI_INTEL) += ghash-clmulni-intel.o ghash-clmulni-intel-y := ghash-clmulni-intel_asm.o ghash-clmulni-intel_glue.o diff --git a/arch/x86/crypto/blake2s-glue.c b/arch/x86/crypto/blake2s-glue.c index a40365ab301e..69853c13e8fb 100644 --- a/arch/x86/crypto/blake2s-glue.c +++ b/arch/x86/crypto/blake2s-glue.c @@ -5,7 +5,6 @@ #include #include -#include #include #include @@ -28,9 +27,8 @@ asmlinkage void blake2s_compress_avx512(struct blake2s_state *state, static __ro_after_init DEFINE_STATIC_KEY_FALSE(blake2s_use_ssse3); static __ro_after_init DEFINE_STATIC_KEY_FALSE(blake2s_use_avx512); -void blake2s_compress_arch(struct blake2s_state *state, - const u8 *block, size_t nblocks, - const u32 inc) +void blake2s_compress(struct blake2s_state *state, const u8 *block, + size_t nblocks, const u32 inc) { /* SIMD disables preemption, so relax after processing each page. */ BUILD_BUG_ON(SZ_4K / BLAKE2S_BLOCK_SIZE < 8); @@ -56,49 +54,12 @@ void blake2s_compress_arch(struct blake2s_state *state, block += blocks * BLAKE2S_BLOCK_SIZE; } while (nblocks); } -EXPORT_SYMBOL(blake2s_compress_arch); - -static int crypto_blake2s_update_x86(struct shash_desc *desc, - const u8 *in, unsigned int inlen) -{ - return crypto_blake2s_update(desc, in, inlen, blake2s_compress_arch); -} - -static int crypto_blake2s_final_x86(struct shash_desc *desc, u8 *out) -{ - return crypto_blake2s_final(desc, out, blake2s_compress_arch); -} - -#define BLAKE2S_ALG(name, driver_name, digest_size) \ - { \ - .base.cra_name = name, \ - .base.cra_driver_name = driver_name, \ - .base.cra_priority = 200, \ - .base.cra_flags = CRYPTO_ALG_OPTIONAL_KEY, \ - .base.cra_blocksize = BLAKE2S_BLOCK_SIZE, \ - .base.cra_ctxsize = sizeof(struct blake2s_tfm_ctx), \ - .base.cra_module = THIS_MODULE, \ - .digestsize = digest_size, \ - .setkey = crypto_blake2s_setkey, \ - .init = crypto_blake2s_init, \ - .update = crypto_blake2s_update_x86, \ - .final = crypto_blake2s_final_x86, \ - .descsize = sizeof(struct blake2s_state), \ - } - -static struct shash_alg blake2s_algs[] = { - BLAKE2S_ALG("blake2s-128", "blake2s-128-x86", BLAKE2S_128_HASH_SIZE), - BLAKE2S_ALG("blake2s-160", "blake2s-160-x86", BLAKE2S_160_HASH_SIZE), - BLAKE2S_ALG("blake2s-224", "blake2s-224-x86", BLAKE2S_224_HASH_SIZE), - BLAKE2S_ALG("blake2s-256", "blake2s-256-x86", BLAKE2S_256_HASH_SIZE), -}; +EXPORT_SYMBOL(blake2s_compress); static int __init blake2s_mod_init(void) { - if (!boot_cpu_has(X86_FEATURE_SSSE3)) - return 0; - - static_branch_enable(&blake2s_use_ssse3); + if (boot_cpu_has(X86_FEATURE_SSSE3)) + static_branch_enable(&blake2s_use_ssse3); if (IS_ENABLED(CONFIG_AS_AVX512) && boot_cpu_has(X86_FEATURE_AVX) && @@ -109,26 +70,9 @@ static int __init blake2s_mod_init(void) XFEATURE_MASK_AVX512, NULL)) static_branch_enable(&blake2s_use_avx512); - return IS_REACHABLE(CONFIG_CRYPTO_HASH) ? - crypto_register_shashes(blake2s_algs, - ARRAY_SIZE(blake2s_algs)) : 0; -} - -static void __exit blake2s_mod_exit(void) -{ - if (IS_REACHABLE(CONFIG_CRYPTO_HASH) && boot_cpu_has(X86_FEATURE_SSSE3)) - crypto_unregister_shashes(blake2s_algs, ARRAY_SIZE(blake2s_algs)); + return 0; } module_init(blake2s_mod_init); -module_exit(blake2s_mod_exit); -MODULE_ALIAS_CRYPTO("blake2s-128"); -MODULE_ALIAS_CRYPTO("blake2s-128-x86"); -MODULE_ALIAS_CRYPTO("blake2s-160"); -MODULE_ALIAS_CRYPTO("blake2s-160-x86"); -MODULE_ALIAS_CRYPTO("blake2s-224"); -MODULE_ALIAS_CRYPTO("blake2s-224-x86"); -MODULE_ALIAS_CRYPTO("blake2s-256"); -MODULE_ALIAS_CRYPTO("blake2s-256-x86"); MODULE_LICENSE("GPL v2"); diff --git a/arch/x86/crypto/blake2s-shash.c b/arch/x86/crypto/blake2s-shash.c new file mode 100644 index 000000000000..f9e2fecdb761 --- /dev/null +++ b/arch/x86/crypto/blake2s-shash.c @@ -0,0 +1,77 @@ +// SPDX-License-Identifier: GPL-2.0 OR MIT +/* + * Copyright (C) 2015-2019 Jason A. Donenfeld . All Rights Reserved. + */ + +#include +#include +#include + +#include +#include +#include +#include + +#include +#include + +static int crypto_blake2s_update_x86(struct shash_desc *desc, + const u8 *in, unsigned int inlen) +{ + return crypto_blake2s_update(desc, in, inlen, blake2s_compress); +} + +static int crypto_blake2s_final_x86(struct shash_desc *desc, u8 *out) +{ + return crypto_blake2s_final(desc, out, blake2s_compress); +} + +#define BLAKE2S_ALG(name, driver_name, digest_size) \ + { \ + .base.cra_name = name, \ + .base.cra_driver_name = driver_name, \ + .base.cra_priority = 200, \ + .base.cra_flags = CRYPTO_ALG_OPTIONAL_KEY, \ + .base.cra_blocksize = BLAKE2S_BLOCK_SIZE, \ + .base.cra_ctxsize = sizeof(struct blake2s_tfm_ctx), \ + .base.cra_module = THIS_MODULE, \ + .digestsize = digest_size, \ + .setkey = crypto_blake2s_setkey, \ + .init = crypto_blake2s_init, \ + .update = crypto_blake2s_update_x86, \ + .final = crypto_blake2s_final_x86, \ + .descsize = sizeof(struct blake2s_state), \ + } + +static struct shash_alg blake2s_algs[] = { + BLAKE2S_ALG("blake2s-128", "blake2s-128-x86", BLAKE2S_128_HASH_SIZE), + BLAKE2S_ALG("blake2s-160", "blake2s-160-x86", BLAKE2S_160_HASH_SIZE), + BLAKE2S_ALG("blake2s-224", "blake2s-224-x86", BLAKE2S_224_HASH_SIZE), + BLAKE2S_ALG("blake2s-256", "blake2s-256-x86", BLAKE2S_256_HASH_SIZE), +}; + +static int __init blake2s_mod_init(void) +{ + if (IS_REACHABLE(CONFIG_CRYPTO_HASH) && boot_cpu_has(X86_FEATURE_SSSE3)) + return crypto_register_shashes(blake2s_algs, ARRAY_SIZE(blake2s_algs)); + return 0; +} + +static void __exit blake2s_mod_exit(void) +{ + if (IS_REACHABLE(CONFIG_CRYPTO_HASH) && boot_cpu_has(X86_FEATURE_SSSE3)) + crypto_unregister_shashes(blake2s_algs, ARRAY_SIZE(blake2s_algs)); +} + +module_init(blake2s_mod_init); +module_exit(blake2s_mod_exit); + +MODULE_ALIAS_CRYPTO("blake2s-128"); +MODULE_ALIAS_CRYPTO("blake2s-128-x86"); +MODULE_ALIAS_CRYPTO("blake2s-160"); +MODULE_ALIAS_CRYPTO("blake2s-160-x86"); +MODULE_ALIAS_CRYPTO("blake2s-224"); +MODULE_ALIAS_CRYPTO("blake2s-224-x86"); +MODULE_ALIAS_CRYPTO("blake2s-256"); +MODULE_ALIAS_CRYPTO("blake2s-256-x86"); +MODULE_LICENSE("GPL v2"); diff --git a/crypto/Kconfig b/crypto/Kconfig index 285f82647d2b..55718de56137 100644 --- a/crypto/Kconfig +++ b/crypto/Kconfig @@ -1919,9 +1919,10 @@ config CRYPTO_STATS config CRYPTO_HASH_INFO bool -source "lib/crypto/Kconfig" source "drivers/crypto/Kconfig" source "crypto/asymmetric_keys/Kconfig" source "certs/Kconfig" endif # if CRYPTO + +source "lib/crypto/Kconfig" diff --git a/drivers/net/Kconfig b/drivers/net/Kconfig index 6cccc3dc00bc..b2a4f998c180 100644 --- a/drivers/net/Kconfig +++ b/drivers/net/Kconfig @@ -81,7 +81,6 @@ config WIREGUARD select CRYPTO select CRYPTO_LIB_CURVE25519 select CRYPTO_LIB_CHACHA20POLY1305 - select CRYPTO_LIB_BLAKE2S select CRYPTO_CHACHA20_X86_64 if X86 && 64BIT select CRYPTO_POLY1305_X86_64 if X86 && 64BIT select CRYPTO_BLAKE2S_X86 if X86 && 64BIT diff --git a/include/crypto/internal/blake2s.h b/include/crypto/internal/blake2s.h index 8e50d487500f..d39cfa0d333e 100644 --- a/include/crypto/internal/blake2s.h +++ b/include/crypto/internal/blake2s.h @@ -11,11 +11,11 @@ #include #include -void blake2s_compress_generic(struct blake2s_state *state,const u8 *block, +void blake2s_compress_generic(struct blake2s_state *state, const u8 *block, size_t nblocks, const u32 inc); -void blake2s_compress_arch(struct blake2s_state *state,const u8 *block, - size_t nblocks, const u32 inc); +void blake2s_compress(struct blake2s_state *state, const u8 *block, + size_t nblocks, const u32 inc); bool blake2s_selftest(void); diff --git a/lib/crypto/Kconfig b/lib/crypto/Kconfig index 545ccbddf6a1..8620f38e117c 100644 --- a/lib/crypto/Kconfig +++ b/lib/crypto/Kconfig @@ -1,7 +1,5 @@ # SPDX-License-Identifier: GPL-2.0 -comment "Crypto library routines" - config CRYPTO_LIB_AES tristate @@ -9,14 +7,14 @@ config CRYPTO_LIB_ARC4 tristate config CRYPTO_ARCH_HAVE_LIB_BLAKE2S - tristate + bool help Declares whether the architecture provides an arch-specific accelerated implementation of the Blake2s library interface, either builtin or as a module. config CRYPTO_LIB_BLAKE2S_GENERIC - tristate + def_bool !CRYPTO_ARCH_HAVE_LIB_BLAKE2S help This symbol can be depended upon by arch implementations of the Blake2s library interface that require the generic code as a @@ -24,15 +22,6 @@ config CRYPTO_LIB_BLAKE2S_GENERIC implementation is enabled, this implementation serves the users of CRYPTO_LIB_BLAKE2S. -config CRYPTO_LIB_BLAKE2S - tristate "BLAKE2s hash function library" - depends on CRYPTO_ARCH_HAVE_LIB_BLAKE2S || !CRYPTO_ARCH_HAVE_LIB_BLAKE2S - select CRYPTO_LIB_BLAKE2S_GENERIC if CRYPTO_ARCH_HAVE_LIB_BLAKE2S=n - help - Enable the Blake2s library interface. This interface may be fulfilled - by either the generic implementation or an arch-specific one, if one - is available and enabled. - config CRYPTO_ARCH_HAVE_LIB_CHACHA tristate help @@ -51,7 +40,7 @@ config CRYPTO_LIB_CHACHA_GENERIC of CRYPTO_LIB_CHACHA. config CRYPTO_LIB_CHACHA - tristate "ChaCha library interface" + tristate depends on CRYPTO_ARCH_HAVE_LIB_CHACHA || !CRYPTO_ARCH_HAVE_LIB_CHACHA select CRYPTO_LIB_CHACHA_GENERIC if CRYPTO_ARCH_HAVE_LIB_CHACHA=n help @@ -76,7 +65,7 @@ config CRYPTO_LIB_CURVE25519_GENERIC of CRYPTO_LIB_CURVE25519. config CRYPTO_LIB_CURVE25519 - tristate "Curve25519 scalar multiplication library" + tristate depends on CRYPTO_ARCH_HAVE_LIB_CURVE25519 || !CRYPTO_ARCH_HAVE_LIB_CURVE25519 select CRYPTO_LIB_CURVE25519_GENERIC if CRYPTO_ARCH_HAVE_LIB_CURVE25519=n help @@ -111,7 +100,7 @@ config CRYPTO_LIB_POLY1305_GENERIC of CRYPTO_LIB_POLY1305. config CRYPTO_LIB_POLY1305 - tristate "Poly1305 library interface" + tristate depends on CRYPTO_ARCH_HAVE_LIB_POLY1305 || !CRYPTO_ARCH_HAVE_LIB_POLY1305 select CRYPTO_LIB_POLY1305_GENERIC if CRYPTO_ARCH_HAVE_LIB_POLY1305=n help @@ -120,7 +109,7 @@ config CRYPTO_LIB_POLY1305 is available and enabled. config CRYPTO_LIB_CHACHA20POLY1305 - tristate "ChaCha20-Poly1305 AEAD support (8-byte nonce library version)" + tristate depends on CRYPTO_ARCH_HAVE_LIB_CHACHA || !CRYPTO_ARCH_HAVE_LIB_CHACHA depends on CRYPTO_ARCH_HAVE_LIB_POLY1305 || !CRYPTO_ARCH_HAVE_LIB_POLY1305 select CRYPTO_LIB_CHACHA diff --git a/lib/crypto/Makefile b/lib/crypto/Makefile index 73205ed269ba..ed43a41f2dcc 100644 --- a/lib/crypto/Makefile +++ b/lib/crypto/Makefile @@ -10,11 +10,10 @@ libaes-y := aes.o obj-$(CONFIG_CRYPTO_LIB_ARC4) += libarc4.o libarc4-y := arc4.o -obj-$(CONFIG_CRYPTO_LIB_BLAKE2S_GENERIC) += libblake2s-generic.o -libblake2s-generic-y += blake2s-generic.o - -obj-$(CONFIG_CRYPTO_LIB_BLAKE2S) += libblake2s.o -libblake2s-y += blake2s.o +# blake2s is used by the /dev/random driver which is always builtin +obj-y += libblake2s.o +libblake2s-y := blake2s.o +libblake2s-$(CONFIG_CRYPTO_LIB_BLAKE2S_GENERIC) += blake2s-generic.o obj-$(CONFIG_CRYPTO_LIB_CHACHA20POLY1305) += libchacha20poly1305.o libchacha20poly1305-y += chacha20poly1305.o diff --git a/lib/crypto/blake2s-generic.c b/lib/crypto/blake2s-generic.c index 04ff8df24513..75ccb3e633e6 100644 --- a/lib/crypto/blake2s-generic.c +++ b/lib/crypto/blake2s-generic.c @@ -37,7 +37,11 @@ static inline void blake2s_increment_counter(struct blake2s_state *state, state->t[1] += (state->t[0] < inc); } -void blake2s_compress_generic(struct blake2s_state *state,const u8 *block, +void blake2s_compress(struct blake2s_state *state, const u8 *block, + size_t nblocks, const u32 inc) + __weak __alias(blake2s_compress_generic); + +void blake2s_compress_generic(struct blake2s_state *state, const u8 *block, size_t nblocks, const u32 inc) { u32 m[16]; diff --git a/lib/crypto/blake2s.c b/lib/crypto/blake2s.c index 4055aa593ec4..93f2ae051370 100644 --- a/lib/crypto/blake2s.c +++ b/lib/crypto/blake2s.c @@ -16,12 +16,6 @@ #include #include -#if IS_ENABLED(CONFIG_CRYPTO_ARCH_HAVE_LIB_BLAKE2S) -# define blake2s_compress blake2s_compress_arch -#else -# define blake2s_compress blake2s_compress_generic -#endif - void blake2s_update(struct blake2s_state *state, const u8 *in, size_t inlen) { __blake2s_update(state, in, inlen, blake2s_compress); -- cgit v1.2.3 From dd827abe296fe4249b2f8c9b95f72f814ea8348c Mon Sep 17 00:00:00 2001 From: Zizhuang Deng Date: Thu, 30 Dec 2021 15:03:31 +0800 Subject: lib/mpi: Add the return value check of kcalloc() Add the return value check of kcalloc() to avoid potential NULL ptr dereference. Fixes: a8ea8bdd9df9 ("lib/mpi: Extend the MPI library") Signed-off-by: Zizhuang Deng Reviewed-by: Tianjia Zhang Signed-off-by: Herbert Xu --- lib/mpi/mpi-mod.c | 2 ++ 1 file changed, 2 insertions(+) (limited to 'lib') diff --git a/lib/mpi/mpi-mod.c b/lib/mpi/mpi-mod.c index 47bc59edd4ff..54fcc01564d9 100644 --- a/lib/mpi/mpi-mod.c +++ b/lib/mpi/mpi-mod.c @@ -40,6 +40,8 @@ mpi_barrett_t mpi_barrett_init(MPI m, int copy) mpi_normalize(m); ctx = kcalloc(1, sizeof(*ctx), GFP_KERNEL); + if (!ctx) + return NULL; if (copy) { ctx->m = mpi_copy(m); -- cgit v1.2.3 From 25a8de7f8d970ffa7263bd9d32a08138cd949f17 Mon Sep 17 00:00:00 2001 From: "Matthew Wilcox (Oracle)" Date: Fri, 27 Aug 2021 07:21:49 -0400 Subject: XArray: Add xas_advance() Add a new helper function to help iterate over multi-index entries. Signed-off-by: Matthew Wilcox (Oracle) Reviewed-by: Christoph Hellwig Reviewed-by: William Kucharski --- include/linux/xarray.h | 18 ++++++++++++++++++ lib/xarray.c | 6 +++--- 2 files changed, 21 insertions(+), 3 deletions(-) (limited to 'lib') diff --git a/include/linux/xarray.h b/include/linux/xarray.h index a91e3d90df8a..d6d5da6ed735 100644 --- a/include/linux/xarray.h +++ b/include/linux/xarray.h @@ -1580,6 +1580,24 @@ static inline void xas_set(struct xa_state *xas, unsigned long index) xas->xa_node = XAS_RESTART; } +/** + * xas_advance() - Skip over sibling entries. + * @xas: XArray operation state. + * @index: Index of last sibling entry. + * + * Move the operation state to refer to the last sibling entry. + * This is useful for loops that normally want to see sibling + * entries but sometimes want to skip them. Use xas_set() if you + * want to move to an index which is not part of this entry. + */ +static inline void xas_advance(struct xa_state *xas, unsigned long index) +{ + unsigned char shift = xas_is_node(xas) ? xas->xa_node->shift : 0; + + xas->xa_index = index; + xas->xa_offset = (index >> shift) & XA_CHUNK_MASK; +} + /** * xas_set_order() - Set up XArray operation state for a multislot entry. * @xas: XArray operation state. diff --git a/lib/xarray.c b/lib/xarray.c index f5d8f54907b4..6f47f6375808 100644 --- a/lib/xarray.c +++ b/lib/xarray.c @@ -157,7 +157,7 @@ static void xas_move_index(struct xa_state *xas, unsigned long offset) xas->xa_index += offset << shift; } -static void xas_advance(struct xa_state *xas) +static void xas_next_offset(struct xa_state *xas) { xas->xa_offset++; xas_move_index(xas, xas->xa_offset); @@ -1250,7 +1250,7 @@ void *xas_find(struct xa_state *xas, unsigned long max) xas->xa_offset = ((xas->xa_index - 1) & XA_CHUNK_MASK) + 1; } - xas_advance(xas); + xas_next_offset(xas); while (xas->xa_node && (xas->xa_index <= max)) { if (unlikely(xas->xa_offset == XA_CHUNK_SIZE)) { @@ -1268,7 +1268,7 @@ void *xas_find(struct xa_state *xas, unsigned long max) if (entry && !xa_is_sibling(entry)) return entry; - xas_advance(xas); + xas_next_offset(xas); } if (!xas->xa_node) -- cgit v1.2.3 From d99a8af48a3de727173415ccb17f6b6ba60d5573 Mon Sep 17 00:00:00 2001 From: Colin Ian King Date: Thu, 30 Dec 2021 13:45:57 +0000 Subject: lib: remove redundant assignment to variable ret Variable ret is being assigned a value that is never read. If the for-loop is entered then ret is immediately re-assigned a new value. If the for-loop is not executed ret is never read. The assignment is redundant and can be removed. Signed-off-by: Colin Ian King Reviewed-by: Jarkko Sakkinen Signed-off-by: Jarkko Sakkinen --- lib/asn1_encoder.c | 2 -- 1 file changed, 2 deletions(-) (limited to 'lib') diff --git a/lib/asn1_encoder.c b/lib/asn1_encoder.c index 27bbe891714f..0fd3c454a468 100644 --- a/lib/asn1_encoder.c +++ b/lib/asn1_encoder.c @@ -164,8 +164,6 @@ asn1_encode_oid(unsigned char *data, const unsigned char *end_data, data_len -= 3; - ret = 0; - for (i = 2; i < oid_len; i++) { ret = asn1_encode_oid_digit(&d, &data_len, oid[i]); if (ret < 0) -- cgit v1.2.3 From c12837d1bb31032bead9060dec99ef310d5b9fb7 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Wed, 12 Jan 2022 03:14:45 -0800 Subject: ref_tracker: use __GFP_NOFAIL more carefully syzbot was able to trigger this warning from new_slab() /* * All existing users of the __GFP_NOFAIL are blockable, so warn * of any new users that actually require GFP_NOWAIT */ if (WARN_ON_ONCE(!can_direct_reclaim)) goto fail; Indeed, we should use __GFP_NOFAIL if direct reclaim is possible. Hopefully in the future we will be able to use SLAB_NOFAILSLAB option so that syzbot can benefit from full ref_tracker even in the presence of memory fault injections. WARNING: CPU: 0 PID: 13 at mm/page_alloc.c:5081 __alloc_pages_slowpath.constprop.0+0x1b7b/0x20d0 mm/page_alloc.c:5081 mm/page_alloc.c:5081 Modules linked in: CPU: 0 PID: 13 Comm: ksoftirqd/0 Not tainted 5.16.0-rc5-syzkaller #0 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011 RIP: 0010:__alloc_pages_slowpath.constprop.0+0x1b7b/0x20d0 mm/page_alloc.c:5081 mm/page_alloc.c:5081 Code: 90 08 00 00 48 81 c7 d8 04 00 00 48 89 f8 48 c1 e8 03 42 80 3c 30 00 0f 84 f0 ea ff ff e8 3d 82 09 00 e9 e6 ea ff ff 4d 89 fd <0f> 0b 48 b8 00 00 00 00 00 fc ff df 48 8b 54 24 30 48 c1 ea 03 80 RSP: 0018:ffffc90000d272b8 EFLAGS: 00010246 RAX: 0000000000000000 RBX: ffff88813fffc300 RCX: 0000000000000000 RDX: 0000000000000000 RSI: 0000000000000002 RDI: ffff88813fffc348 RBP: ffff88813fffc300 R08: 00000000000013dc R09: 00000000000013c8 R10: 0000000000000000 R11: 0000000000000000 R12: 0000000000000000 R13: ffffc90000d274e8 R14: dffffc0000000000 R15: ffffc90000d274e8 FS: 0000000000000000(0000) GS:ffff8880b9c00000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 00007ffefe6000f8 CR3: 000000001d21e000 CR4: 00000000003506f0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 Call Trace: __alloc_pages+0x412/0x500 mm/page_alloc.c:5382 mm/page_alloc.c:5382 alloc_pages+0x1a7/0x300 mm/mempolicy.c:2191 mm/mempolicy.c:2191 alloc_slab_page mm/slub.c:1793 [inline] allocate_slab mm/slub.c:1938 [inline] alloc_slab_page mm/slub.c:1793 [inline] mm/slub.c:1993 allocate_slab mm/slub.c:1938 [inline] mm/slub.c:1993 new_slab+0x349/0x4a0 mm/slub.c:1993 mm/slub.c:1993 ___slab_alloc+0x918/0xfe0 mm/slub.c:3022 mm/slub.c:3022 __slab_alloc.constprop.0+0x4d/0xa0 mm/slub.c:3109 mm/slub.c:3109 slab_alloc_node mm/slub.c:3200 [inline] slab_alloc mm/slub.c:3242 [inline] slab_alloc_node mm/slub.c:3200 [inline] mm/slub.c:3259 slab_alloc mm/slub.c:3242 [inline] mm/slub.c:3259 kmem_cache_alloc_trace+0x289/0x2c0 mm/slub.c:3259 mm/slub.c:3259 kmalloc include/linux/slab.h:590 [inline] kzalloc include/linux/slab.h:724 [inline] kmalloc include/linux/slab.h:590 [inline] lib/ref_tracker.c:74 kzalloc include/linux/slab.h:724 [inline] lib/ref_tracker.c:74 ref_tracker_alloc+0xe1/0x430 lib/ref_tracker.c:74 lib/ref_tracker.c:74 netdev_tracker_alloc include/linux/netdevice.h:3855 [inline] dev_hold_track include/linux/netdevice.h:3872 [inline] netdev_tracker_alloc include/linux/netdevice.h:3855 [inline] net/core/dst.c:52 dev_hold_track include/linux/netdevice.h:3872 [inline] net/core/dst.c:52 dst_init+0xe0/0x520 net/core/dst.c:52 net/core/dst.c:52 dst_alloc+0x16b/0x1f0 net/core/dst.c:96 net/core/dst.c:96 rt_dst_alloc+0x73/0x450 net/ipv4/route.c:1614 net/ipv4/route.c:1614 ip_route_input_mc net/ipv4/route.c:1720 [inline] ip_route_input_mc net/ipv4/route.c:1720 [inline] net/ipv4/route.c:2465 ip_route_input_rcu.part.0+0x4fe/0xcc0 net/ipv4/route.c:2465 net/ipv4/route.c:2465 ip_route_input_rcu net/ipv4/route.c:2420 [inline] ip_route_input_rcu net/ipv4/route.c:2420 [inline] net/ipv4/route.c:2416 ip_route_input_noref+0x1b8/0x2a0 net/ipv4/route.c:2416 net/ipv4/route.c:2416 ip_rcv_finish_core.constprop.0+0x288/0x1e90 net/ipv4/ip_input.c:354 net/ipv4/ip_input.c:354 ip_rcv_finish+0x135/0x2f0 net/ipv4/ip_input.c:427 net/ipv4/ip_input.c:427 NF_HOOK include/linux/netfilter.h:307 [inline] NF_HOOK include/linux/netfilter.h:301 [inline] NF_HOOK include/linux/netfilter.h:307 [inline] net/ipv4/ip_input.c:540 NF_HOOK include/linux/netfilter.h:301 [inline] net/ipv4/ip_input.c:540 ip_rcv+0xaa/0xd0 net/ipv4/ip_input.c:540 net/ipv4/ip_input.c:540 __netif_receive_skb_one_core+0x114/0x180 net/core/dev.c:5350 net/core/dev.c:5350 __netif_receive_skb+0x24/0x1b0 net/core/dev.c:5464 net/core/dev.c:5464 process_backlog+0x2a5/0x6c0 net/core/dev.c:5796 net/core/dev.c:5796 __napi_poll+0xaf/0x440 net/core/dev.c:6364 net/core/dev.c:6364 napi_poll net/core/dev.c:6431 [inline] napi_poll net/core/dev.c:6431 [inline] net/core/dev.c:6518 net_rx_action+0x801/0xb40 net/core/dev.c:6518 net/core/dev.c:6518 __do_softirq+0x29b/0x9c2 kernel/softirq.c:558 kernel/softirq.c:558 run_ksoftirqd kernel/softirq.c:921 [inline] run_ksoftirqd kernel/softirq.c:921 [inline] kernel/softirq.c:913 run_ksoftirqd+0x2d/0x60 kernel/softirq.c:913 kernel/softirq.c:913 smpboot_thread_fn+0x645/0x9c0 kernel/smpboot.c:164 kernel/smpboot.c:164 kthread+0x405/0x4f0 kernel/kthread.c:327 kernel/kthread.c:327 ret_from_fork+0x1f/0x30 arch/x86/entry/entry_64.S:295 arch/x86/entry/entry_64.S:295 Fixes: 4e66934eaadc ("lib: add reference counting tracking infrastructure") Signed-off-by: Eric Dumazet Cc: Dmitry Vyukov Reported-by: syzbot Signed-off-by: David S. Miller --- lib/ref_tracker.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) (limited to 'lib') diff --git a/lib/ref_tracker.c b/lib/ref_tracker.c index 0ae2e66dcf0f..a6789c0c626b 100644 --- a/lib/ref_tracker.c +++ b/lib/ref_tracker.c @@ -69,9 +69,12 @@ int ref_tracker_alloc(struct ref_tracker_dir *dir, unsigned long entries[REF_TRACKER_STACK_ENTRIES]; struct ref_tracker *tracker; unsigned int nr_entries; + gfp_t gfp_mask = gfp; unsigned long flags; - *trackerp = tracker = kzalloc(sizeof(*tracker), gfp | __GFP_NOFAIL); + if (gfp & __GFP_DIRECT_RECLAIM) + gfp_mask |= __GFP_NOFAIL; + *trackerp = tracker = kzalloc(sizeof(*tracker), gfp_mask); if (unlikely(!tracker)) { pr_err_once("memory allocation failure, unreliable refcount tracker.\n"); refcount_inc(&dir->untracked); -- cgit v1.2.3 From 180dccb0dba4f5e84a4a70c1be1d34cbb6528b32 Mon Sep 17 00:00:00 2001 From: Laibin Qiu Date: Thu, 13 Jan 2022 10:55:36 +0800 Subject: blk-mq: fix tag_get wait task can't be awakened In case of shared tags, there might be more than one hctx which allocates from the same tags, and each hctx is limited to allocate at most: hctx_max_depth = max((bt->sb.depth + users - 1) / users, 4U); tag idle detection is lazy, and may be delayed for 30sec, so there could be just one real active hctx(queue) but all others are actually idle and still accounted as active because of the lazy idle detection. Then if wake_batch is > hctx_max_depth, driver tag allocation may wait forever on this real active hctx. Fix this by recalculating wake_batch when inc or dec active_queues. Fixes: 0d2602ca30e41 ("blk-mq: improve support for shared tags maps") Suggested-by: Ming Lei Suggested-by: John Garry Signed-off-by: Laibin Qiu Reviewed-by: Andy Shevchenko Link: https://lore.kernel.org/r/20220113025536.1479653-1-qiulaibin@huawei.com Signed-off-by: Jens Axboe --- block/blk-mq-tag.c | 40 +++++++++++++++++++++++++++++++++------- include/linux/sbitmap.h | 11 +++++++++++ lib/sbitmap.c | 25 ++++++++++++++++++++++--- 3 files changed, 66 insertions(+), 10 deletions(-) (limited to 'lib') diff --git a/block/blk-mq-tag.c b/block/blk-mq-tag.c index e55a6834c9a6..845f74e8dd7b 100644 --- a/block/blk-mq-tag.c +++ b/block/blk-mq-tag.c @@ -16,6 +16,21 @@ #include "blk-mq-sched.h" #include "blk-mq-tag.h" +/* + * Recalculate wakeup batch when tag is shared by hctx. + */ +static void blk_mq_update_wake_batch(struct blk_mq_tags *tags, + unsigned int users) +{ + if (!users) + return; + + sbitmap_queue_recalculate_wake_batch(&tags->bitmap_tags, + users); + sbitmap_queue_recalculate_wake_batch(&tags->breserved_tags, + users); +} + /* * If a previously inactive queue goes active, bump the active user count. * We need to do this before try to allocate driver tag, then even if fail @@ -24,18 +39,26 @@ */ bool __blk_mq_tag_busy(struct blk_mq_hw_ctx *hctx) { + unsigned int users; + if (blk_mq_is_shared_tags(hctx->flags)) { struct request_queue *q = hctx->queue; - if (!test_bit(QUEUE_FLAG_HCTX_ACTIVE, &q->queue_flags) && - !test_and_set_bit(QUEUE_FLAG_HCTX_ACTIVE, &q->queue_flags)) - atomic_inc(&hctx->tags->active_queues); + if (test_bit(QUEUE_FLAG_HCTX_ACTIVE, &q->queue_flags) || + test_and_set_bit(QUEUE_FLAG_HCTX_ACTIVE, &q->queue_flags)) { + return true; + } } else { - if (!test_bit(BLK_MQ_S_TAG_ACTIVE, &hctx->state) && - !test_and_set_bit(BLK_MQ_S_TAG_ACTIVE, &hctx->state)) - atomic_inc(&hctx->tags->active_queues); + if (test_bit(BLK_MQ_S_TAG_ACTIVE, &hctx->state) || + test_and_set_bit(BLK_MQ_S_TAG_ACTIVE, &hctx->state)) { + return true; + } } + users = atomic_inc_return(&hctx->tags->active_queues); + + blk_mq_update_wake_batch(hctx->tags, users); + return true; } @@ -56,6 +79,7 @@ void blk_mq_tag_wakeup_all(struct blk_mq_tags *tags, bool include_reserve) void __blk_mq_tag_idle(struct blk_mq_hw_ctx *hctx) { struct blk_mq_tags *tags = hctx->tags; + unsigned int users; if (blk_mq_is_shared_tags(hctx->flags)) { struct request_queue *q = hctx->queue; @@ -68,7 +92,9 @@ void __blk_mq_tag_idle(struct blk_mq_hw_ctx *hctx) return; } - atomic_dec(&tags->active_queues); + users = atomic_dec_return(&tags->active_queues); + + blk_mq_update_wake_batch(tags, users); blk_mq_tag_wakeup_all(tags, false); } diff --git a/include/linux/sbitmap.h b/include/linux/sbitmap.h index fc0357a6e19b..95df357ec009 100644 --- a/include/linux/sbitmap.h +++ b/include/linux/sbitmap.h @@ -415,6 +415,17 @@ static inline void sbitmap_queue_free(struct sbitmap_queue *sbq) sbitmap_free(&sbq->sb); } +/** + * sbitmap_queue_recalculate_wake_batch() - Recalculate wake batch + * @sbq: Bitmap queue to recalculate wake batch. + * @users: Number of shares. + * + * Like sbitmap_queue_update_wake_batch(), this will calculate wake batch + * by depth. This interface is for HCTX shared tags or queue shared tags. + */ +void sbitmap_queue_recalculate_wake_batch(struct sbitmap_queue *sbq, + unsigned int users); + /** * sbitmap_queue_resize() - Resize a &struct sbitmap_queue. * @sbq: Bitmap queue to resize. diff --git a/lib/sbitmap.c b/lib/sbitmap.c index 2709ab825499..6220fa67fb7e 100644 --- a/lib/sbitmap.c +++ b/lib/sbitmap.c @@ -457,10 +457,9 @@ int sbitmap_queue_init_node(struct sbitmap_queue *sbq, unsigned int depth, } EXPORT_SYMBOL_GPL(sbitmap_queue_init_node); -static void sbitmap_queue_update_wake_batch(struct sbitmap_queue *sbq, - unsigned int depth) +static inline void __sbitmap_queue_update_wake_batch(struct sbitmap_queue *sbq, + unsigned int wake_batch) { - unsigned int wake_batch = sbq_calc_wake_batch(sbq, depth); int i; if (sbq->wake_batch != wake_batch) { @@ -476,6 +475,26 @@ static void sbitmap_queue_update_wake_batch(struct sbitmap_queue *sbq, } } +static void sbitmap_queue_update_wake_batch(struct sbitmap_queue *sbq, + unsigned int depth) +{ + unsigned int wake_batch; + + wake_batch = sbq_calc_wake_batch(sbq, depth); + __sbitmap_queue_update_wake_batch(sbq, wake_batch); +} + +void sbitmap_queue_recalculate_wake_batch(struct sbitmap_queue *sbq, + unsigned int users) +{ + unsigned int wake_batch; + + wake_batch = clamp_val((sbq->sb.depth + users - 1) / + users, 4, SBQ_WAKE_BATCH); + __sbitmap_queue_update_wake_batch(sbq, wake_batch); +} +EXPORT_SYMBOL_GPL(sbitmap_queue_recalculate_wake_batch); + void sbitmap_queue_resize(struct sbitmap_queue *sbq, unsigned int depth) { sbitmap_queue_update_wake_batch(sbq, depth); -- cgit v1.2.3 From e5f4728767d2ec9e3eb122c74e224242d21ee650 Mon Sep 17 00:00:00 2001 From: Marco Elver Date: Fri, 14 Jan 2022 14:04:51 -0800 Subject: kasan: test: add globals left-out-of-bounds test Add a test checking that KASAN generic can also detect out-of-bounds accesses to the left of globals. Unfortunately it seems that GCC doesn't catch this (tested GCC 10, 11). The main difference between GCC's globals redzoning and Clang's is that GCC relies on using increased alignment to producing padding, where Clang's redzoning implementation actually adds real data after the global and doesn't rely on alignment to produce padding. I believe this is the main reason why GCC can't reliably catch globals out-of-bounds in this case. Given this is now a known issue, to avoid failing the whole test suite, skip this test case with GCC. Link: https://lkml.kernel.org/r/20211117130714.135656-1-elver@google.com Signed-off-by: Marco Elver Reported-by: Kaiwan N Billimoria Reviewed-by: Andrey Konovalov Cc: Alexander Potapenko Cc: Andrey Ryabinin Cc: Dmitry Vyukov Cc: Kaiwan N Billimoria Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- lib/test_kasan.c | 19 +++++++++++++++++-- 1 file changed, 17 insertions(+), 2 deletions(-) (limited to 'lib') diff --git a/lib/test_kasan.c b/lib/test_kasan.c index 0643573f8686..818e763b5b87 100644 --- a/lib/test_kasan.c +++ b/lib/test_kasan.c @@ -700,7 +700,7 @@ static void kmem_cache_bulk(struct kunit *test) static char global_array[10]; -static void kasan_global_oob(struct kunit *test) +static void kasan_global_oob_right(struct kunit *test) { /* * Deliberate out-of-bounds access. To prevent CONFIG_UBSAN_LOCAL_BOUNDS @@ -723,6 +723,20 @@ static void kasan_global_oob(struct kunit *test) KUNIT_EXPECT_KASAN_FAIL(test, *(volatile char *)p); } +static void kasan_global_oob_left(struct kunit *test) +{ + char *volatile array = global_array; + char *p = array - 3; + + /* + * GCC is known to fail this test, skip it. + * See https://bugzilla.kernel.org/show_bug.cgi?id=215051. + */ + KASAN_TEST_NEEDS_CONFIG_ON(test, CONFIG_CC_IS_CLANG); + KASAN_TEST_NEEDS_CONFIG_ON(test, CONFIG_KASAN_GENERIC); + KUNIT_EXPECT_KASAN_FAIL(test, *(volatile char *)p); +} + /* Check that ksize() makes the whole object accessible. */ static void ksize_unpoisons_memory(struct kunit *test) { @@ -1162,7 +1176,8 @@ static struct kunit_case kasan_kunit_test_cases[] = { KUNIT_CASE(kmem_cache_oob), KUNIT_CASE(kmem_cache_accounted), KUNIT_CASE(kmem_cache_bulk), - KUNIT_CASE(kasan_global_oob), + KUNIT_CASE(kasan_global_oob_right), + KUNIT_CASE(kasan_global_oob_left), KUNIT_CASE(kasan_stack_oob), KUNIT_CASE(kasan_alloca_oob_left), KUNIT_CASE(kasan_alloca_oob_right), -- cgit v1.2.3 From f98f966cd75002a71caec1b6d209da5762c0efac Mon Sep 17 00:00:00 2001 From: Marco Elver Date: Fri, 14 Jan 2022 14:04:57 -0800 Subject: kasan: test: add test case for double-kmem_cache_destroy() Add a test case for double-kmem_cache_destroy() detection. Link: https://lkml.kernel.org/r/20211119142219.1519617-2-elver@google.com Signed-off-by: Marco Elver Reviewed-by: Andrey Konovalov Cc: Andrey Ryabinin Cc: Alexander Potapenko Cc: Dmitry Vyukov Cc: Christoph Lameter Cc: Pekka Enberg Cc: David Rientjes Cc: Joonsoo Kim Cc: Vlastimil Babka Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- lib/test_kasan.c | 11 +++++++++++ 1 file changed, 11 insertions(+) (limited to 'lib') diff --git a/lib/test_kasan.c b/lib/test_kasan.c index 818e763b5b87..847cdbefab46 100644 --- a/lib/test_kasan.c +++ b/lib/test_kasan.c @@ -866,6 +866,16 @@ static void kmem_cache_invalid_free(struct kunit *test) kmem_cache_destroy(cache); } +static void kmem_cache_double_destroy(struct kunit *test) +{ + struct kmem_cache *cache; + + cache = kmem_cache_create("test_cache", 200, 0, 0, NULL); + KUNIT_ASSERT_NOT_ERR_OR_NULL(test, cache); + kmem_cache_destroy(cache); + KUNIT_EXPECT_KASAN_FAIL(test, kmem_cache_destroy(cache)); +} + static void kasan_memchr(struct kunit *test) { char *ptr; @@ -1185,6 +1195,7 @@ static struct kunit_case kasan_kunit_test_cases[] = { KUNIT_CASE(ksize_uaf), KUNIT_CASE(kmem_cache_double_free), KUNIT_CASE(kmem_cache_invalid_free), + KUNIT_CASE(kmem_cache_double_destroy), KUNIT_CASE(kasan_memchr), KUNIT_CASE(kasan_memcmp), KUNIT_CASE(kasan_strings), -- cgit v1.2.3 From 87c01d57fa23de82fff593a7d070933d08755801 Mon Sep 17 00:00:00 2001 From: Alistair Popple Date: Fri, 14 Jan 2022 14:09:31 -0800 Subject: mm/hmm.c: allow VM_MIXEDMAP to work with hmm_range_fault hmm_range_fault() can be used instead of get_user_pages() for devices which allow faulting however unlike get_user_pages() it will return an error when used on a VM_MIXEDMAP range. To make hmm_range_fault() more closely match get_user_pages() remove this restriction. This requires dealing with the !ARCH_HAS_PTE_SPECIAL case in hmm_vma_handle_pte(). Rather than replicating the logic of vm_normal_page() call it directly and do a check for the zero pfn similar to what get_user_pages() currently does. Also add a test to hmm selftest to verify functionality. Link: https://lkml.kernel.org/r/20211104012001.2555676-1-apopple@nvidia.com Fixes: da4c3c735ea4 ("mm/hmm/mirror: helper to snapshot CPU page table") Signed-off-by: Alistair Popple Reviewed-by: Jason Gunthorpe Cc: Jerome Glisse Cc: John Hubbard Cc: Zi Yan Cc: Ralph Campbell Cc: Felix Kuehling Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- lib/test_hmm.c | 24 +++++++++++++++++++ mm/hmm.c | 5 ++-- tools/testing/selftests/vm/hmm-tests.c | 42 ++++++++++++++++++++++++++++++++++ 3 files changed, 69 insertions(+), 2 deletions(-) (limited to 'lib') diff --git a/lib/test_hmm.c b/lib/test_hmm.c index e2ce8f9b7605..767538089a62 100644 --- a/lib/test_hmm.c +++ b/lib/test_hmm.c @@ -1086,9 +1086,33 @@ static long dmirror_fops_unlocked_ioctl(struct file *filp, return 0; } +static int dmirror_fops_mmap(struct file *file, struct vm_area_struct *vma) +{ + unsigned long addr; + + for (addr = vma->vm_start; addr < vma->vm_end; addr += PAGE_SIZE) { + struct page *page; + int ret; + + page = alloc_page(GFP_KERNEL | __GFP_ZERO); + if (!page) + return -ENOMEM; + + ret = vm_insert_page(vma, addr, page); + if (ret) { + __free_page(page); + return ret; + } + put_page(page); + } + + return 0; +} + static const struct file_operations dmirror_fops = { .open = dmirror_fops_open, .release = dmirror_fops_release, + .mmap = dmirror_fops_mmap, .unlocked_ioctl = dmirror_fops_unlocked_ioctl, .llseek = default_llseek, .owner = THIS_MODULE, diff --git a/mm/hmm.c b/mm/hmm.c index 842e26599238..bd56641c79d4 100644 --- a/mm/hmm.c +++ b/mm/hmm.c @@ -300,7 +300,8 @@ static int hmm_vma_handle_pte(struct mm_walk *walk, unsigned long addr, * Since each architecture defines a struct page for the zero page, just * fall through and treat it like a normal page. */ - if (pte_special(pte) && !pte_devmap(pte) && + if (!vm_normal_page(walk->vma, addr, pte) && + !pte_devmap(pte) && !is_zero_pfn(pte_pfn(pte))) { if (hmm_pte_need_fault(hmm_vma_walk, pfn_req_flags, 0)) { pte_unmap(ptep); @@ -518,7 +519,7 @@ static int hmm_vma_walk_test(unsigned long start, unsigned long end, struct hmm_range *range = hmm_vma_walk->range; struct vm_area_struct *vma = walk->vma; - if (!(vma->vm_flags & (VM_IO | VM_PFNMAP | VM_MIXEDMAP)) && + if (!(vma->vm_flags & (VM_IO | VM_PFNMAP)) && vma->vm_flags & VM_READ) return 0; diff --git a/tools/testing/selftests/vm/hmm-tests.c b/tools/testing/selftests/vm/hmm-tests.c index 864f126ffd78..203323967b50 100644 --- a/tools/testing/selftests/vm/hmm-tests.c +++ b/tools/testing/selftests/vm/hmm-tests.c @@ -1248,6 +1248,48 @@ TEST_F(hmm, anon_teardown) } } +/* + * Test memory snapshot without faulting in pages accessed by the device. + */ +TEST_F(hmm, mixedmap) +{ + struct hmm_buffer *buffer; + unsigned long npages; + unsigned long size; + unsigned char *m; + int ret; + + npages = 1; + size = npages << self->page_shift; + + buffer = malloc(sizeof(*buffer)); + ASSERT_NE(buffer, NULL); + + buffer->fd = -1; + buffer->size = size; + buffer->mirror = malloc(npages); + ASSERT_NE(buffer->mirror, NULL); + + + /* Reserve a range of addresses. */ + buffer->ptr = mmap(NULL, size, + PROT_READ | PROT_WRITE, + MAP_PRIVATE, + self->fd, 0); + ASSERT_NE(buffer->ptr, MAP_FAILED); + + /* Simulate a device snapshotting CPU pagetables. */ + ret = hmm_dmirror_cmd(self->fd, HMM_DMIRROR_SNAPSHOT, buffer, npages); + ASSERT_EQ(ret, 0); + ASSERT_EQ(buffer->cpages, npages); + + /* Check what the device saw. */ + m = buffer->mirror; + ASSERT_EQ(m[0], HMM_DMIRROR_PROT_READ); + + hmm_buffer_free(buffer); +} + /* * Test memory snapshot without faulting in pages accessed by the device. */ -- cgit v1.2.3 From c126a53c276048125b4a950072bab37ad0fea120 Mon Sep 17 00:00:00 2001 From: Yury Norov Date: Sat, 14 Aug 2021 14:17:00 -0700 Subject: arch: remove GENERIC_FIND_FIRST_BIT entirely In 5.12 cycle we enabled GENERIC_FIND_FIRST_BIT config option for ARM64 and MIPS. It increased performance and shrunk .text size; and so far I didn't receive any negative feedback on the change. https://lore.kernel.org/linux-arch/20210225135700.1381396-1-yury.norov@gmail.com/ Now I think it's a good time to switch all architectures to use find_{first,last}_bit() unconditionally, and so remove corresponding config option. The patch does't introduce functioal changes for arc, arm, arm64, mips, m68k, s390 and x86, for other architectures I expect improvement both in performance and .text size. Signed-off-by: Yury Norov Tested-by: Alexander Lobakin (mips) Reviewed-by: Alexander Lobakin (mips) Reviewed-by: Andy Shevchenko Acked-by: Will Deacon Tested-by: Wolfram Sang --- arch/arc/Kconfig | 1 - arch/arm64/Kconfig | 1 - arch/mips/Kconfig | 1 - arch/s390/Kconfig | 1 - arch/x86/Kconfig | 1 - arch/x86/um/Kconfig | 1 - include/linux/find.h | 13 ------------- lib/Kconfig | 3 --- 8 files changed, 22 deletions(-) (limited to 'lib') diff --git a/arch/arc/Kconfig b/arch/arc/Kconfig index b4ae6058902a..4bec4b0b6ce1 100644 --- a/arch/arc/Kconfig +++ b/arch/arc/Kconfig @@ -20,7 +20,6 @@ config ARC select COMMON_CLK select DMA_DIRECT_REMAP select GENERIC_ATOMIC64 if !ISA_ARCV2 || !(ARC_HAS_LL64 && ARC_HAS_LLSC) - select GENERIC_FIND_FIRST_BIT # for now, we don't need GENERIC_IRQ_PROBE, CONFIG_GENERIC_IRQ_CHIP select GENERIC_IRQ_SHOW select GENERIC_PCI_IOMAP diff --git a/arch/arm64/Kconfig b/arch/arm64/Kconfig index c4207cf9bb17..517d26c8002d 100644 --- a/arch/arm64/Kconfig +++ b/arch/arm64/Kconfig @@ -120,7 +120,6 @@ config ARM64 select GENERIC_CPU_AUTOPROBE select GENERIC_CPU_VULNERABILITIES select GENERIC_EARLY_IOREMAP - select GENERIC_FIND_FIRST_BIT select GENERIC_IDLE_POLL_SETUP select GENERIC_IRQ_IPI select GENERIC_IRQ_PROBE diff --git a/arch/mips/Kconfig b/arch/mips/Kconfig index 0215dc1529e9..00951bfdbab0 100644 --- a/arch/mips/Kconfig +++ b/arch/mips/Kconfig @@ -32,7 +32,6 @@ config MIPS select GENERIC_ATOMIC64 if !64BIT select GENERIC_CMOS_UPDATE select GENERIC_CPU_AUTOPROBE - select GENERIC_FIND_FIRST_BIT select GENERIC_GETTIMEOFDAY select GENERIC_IOMAP select GENERIC_IRQ_PROBE diff --git a/arch/s390/Kconfig b/arch/s390/Kconfig index 2a5bb4f29cfe..4f80f1c95468 100644 --- a/arch/s390/Kconfig +++ b/arch/s390/Kconfig @@ -127,7 +127,6 @@ config S390 select GENERIC_CPU_AUTOPROBE select GENERIC_CPU_VULNERABILITIES select GENERIC_ENTRY - select GENERIC_FIND_FIRST_BIT select GENERIC_GETTIMEOFDAY select GENERIC_PTDUMP select GENERIC_SMP_IDLE_THREAD diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig index 5c2ccb85f2ef..60484b39257c 100644 --- a/arch/x86/Kconfig +++ b/arch/x86/Kconfig @@ -136,7 +136,6 @@ config X86 select GENERIC_CPU_VULNERABILITIES select GENERIC_EARLY_IOREMAP select GENERIC_ENTRY - select GENERIC_FIND_FIRST_BIT select GENERIC_IOMAP select GENERIC_IRQ_EFFECTIVE_AFF_MASK if SMP select GENERIC_IRQ_MATRIX_ALLOCATOR if X86_LOCAL_APIC diff --git a/arch/x86/um/Kconfig b/arch/x86/um/Kconfig index 95d26a69088b..40d6a06e41c8 100644 --- a/arch/x86/um/Kconfig +++ b/arch/x86/um/Kconfig @@ -8,7 +8,6 @@ endmenu config UML_X86 def_bool y - select GENERIC_FIND_FIRST_BIT config 64BIT bool "64-bit kernel" if "$(SUBARCH)" = "x86" diff --git a/include/linux/find.h b/include/linux/find.h index c5410c243e04..ea57f7f38c49 100644 --- a/include/linux/find.h +++ b/include/linux/find.h @@ -101,8 +101,6 @@ unsigned long find_next_zero_bit(const unsigned long *addr, unsigned long size, } #endif -#ifdef CONFIG_GENERIC_FIND_FIRST_BIT - #ifndef find_first_bit /** * find_first_bit - find the first set bit in a memory region @@ -147,17 +145,6 @@ unsigned long find_first_zero_bit(const unsigned long *addr, unsigned long size) } #endif -#else /* CONFIG_GENERIC_FIND_FIRST_BIT */ - -#ifndef find_first_bit -#define find_first_bit(addr, size) find_next_bit((addr), (size), 0) -#endif -#ifndef find_first_zero_bit -#define find_first_zero_bit(addr, size) find_next_zero_bit((addr), (size), 0) -#endif - -#endif /* CONFIG_GENERIC_FIND_FIRST_BIT */ - #ifndef find_last_bit /** * find_last_bit - find the last set bit in a memory region diff --git a/lib/Kconfig b/lib/Kconfig index 5e7165e6a346..6a6ae5312fa0 100644 --- a/lib/Kconfig +++ b/lib/Kconfig @@ -65,9 +65,6 @@ config GENERIC_STRNLEN_USER config GENERIC_NET_UTILS bool -config GENERIC_FIND_FIRST_BIT - bool - source "lib/math/Kconfig" config NO_GENERIC_PCI_IOPORT_MAP -- cgit v1.2.3 From f68edc9297bf3f7c94abb54b9b0b053607f7587b Mon Sep 17 00:00:00 2001 From: Yury Norov Date: Sat, 14 Aug 2021 14:17:01 -0700 Subject: lib: add find_first_and_bit() Currently find_first_and_bit() is an alias to find_next_and_bit(). However, it is widely used in cpumask, so it worth to optimize it. This patch adds its own implementation for find_first_and_bit(). On x86_64 find_bit_benchmark says: Before (#define find_first_and_bit(...) find_next_and_bit(..., 0): Start testing find_bit() with random-filled bitmap [ 140.291468] find_first_and_bit: 46890919 ns, 32671 iterations Start testing find_bit() with sparse bitmap [ 140.295028] find_first_and_bit: 7103 ns, 1 iterations After: Start testing find_bit() with random-filled bitmap [ 162.574907] find_first_and_bit: 25045813 ns, 32846 iterations Start testing find_bit() with sparse bitmap [ 162.578458] find_first_and_bit: 4900 ns, 1 iterations (Thanks to Alexey Klimov for thorough testing.) Signed-off-by: Yury Norov Tested-by: Wolfram Sang Tested-by: Alexey Klimov --- include/linux/find.h | 27 +++++++++++++++++++++++++++ lib/find_bit.c | 21 +++++++++++++++++++++ lib/find_bit_benchmark.c | 21 +++++++++++++++++++++ 3 files changed, 69 insertions(+) (limited to 'lib') diff --git a/include/linux/find.h b/include/linux/find.h index ea57f7f38c49..6048f8c97418 100644 --- a/include/linux/find.h +++ b/include/linux/find.h @@ -12,6 +12,8 @@ extern unsigned long _find_next_bit(const unsigned long *addr1, const unsigned long *addr2, unsigned long nbits, unsigned long start, unsigned long invert, unsigned long le); extern unsigned long _find_first_bit(const unsigned long *addr, unsigned long size); +extern unsigned long _find_first_and_bit(const unsigned long *addr1, + const unsigned long *addr2, unsigned long size); extern unsigned long _find_first_zero_bit(const unsigned long *addr, unsigned long size); extern unsigned long _find_last_bit(const unsigned long *addr, unsigned long size); @@ -123,6 +125,31 @@ unsigned long find_first_bit(const unsigned long *addr, unsigned long size) } #endif +#ifndef find_first_and_bit +/** + * find_first_and_bit - find the first set bit in both memory regions + * @addr1: The first address to base the search on + * @addr2: The second address to base the search on + * @size: The bitmap size in bits + * + * Returns the bit number for the next set bit + * If no bits are set, returns @size. + */ +static inline +unsigned long find_first_and_bit(const unsigned long *addr1, + const unsigned long *addr2, + unsigned long size) +{ + if (small_const_nbits(size)) { + unsigned long val = *addr1 & *addr2 & GENMASK(size - 1, 0); + + return val ? __ffs(val) : size; + } + + return _find_first_and_bit(addr1, addr2, size); +} +#endif + #ifndef find_first_zero_bit /** * find_first_zero_bit - find the first cleared bit in a memory region diff --git a/lib/find_bit.c b/lib/find_bit.c index 0f8e2e369b1d..1b8e4b2a9cba 100644 --- a/lib/find_bit.c +++ b/lib/find_bit.c @@ -89,6 +89,27 @@ unsigned long _find_first_bit(const unsigned long *addr, unsigned long size) EXPORT_SYMBOL(_find_first_bit); #endif +#ifndef find_first_and_bit +/* + * Find the first set bit in two memory regions. + */ +unsigned long _find_first_and_bit(const unsigned long *addr1, + const unsigned long *addr2, + unsigned long size) +{ + unsigned long idx, val; + + for (idx = 0; idx * BITS_PER_LONG < size; idx++) { + val = addr1[idx] & addr2[idx]; + if (val) + return min(idx * BITS_PER_LONG + __ffs(val), size); + } + + return size; +} +EXPORT_SYMBOL(_find_first_and_bit); +#endif + #ifndef find_first_zero_bit /* * Find the first cleared bit in a memory region. diff --git a/lib/find_bit_benchmark.c b/lib/find_bit_benchmark.c index 5637c5711db9..db904b57d4b8 100644 --- a/lib/find_bit_benchmark.c +++ b/lib/find_bit_benchmark.c @@ -49,6 +49,25 @@ static int __init test_find_first_bit(void *bitmap, unsigned long len) return 0; } +static int __init test_find_first_and_bit(void *bitmap, const void *bitmap2, unsigned long len) +{ + static DECLARE_BITMAP(cp, BITMAP_LEN) __initdata; + unsigned long i, cnt; + ktime_t time; + + bitmap_copy(cp, bitmap, BITMAP_LEN); + + time = ktime_get(); + for (cnt = i = 0; i < len; cnt++) { + i = find_first_and_bit(cp, bitmap2, len); + __clear_bit(i, cp); + } + time = ktime_get() - time; + pr_err("find_first_and_bit: %18llu ns, %6ld iterations\n", time, cnt); + + return 0; +} + static int __init test_find_next_bit(const void *bitmap, unsigned long len) { unsigned long i, cnt; @@ -129,6 +148,7 @@ static int __init find_bit_test(void) * traverse only part of bitmap to avoid soft lockup. */ test_find_first_bit(bitmap, BITMAP_LEN / 10); + test_find_first_and_bit(bitmap, bitmap2, BITMAP_LEN / 2); test_find_next_and_bit(bitmap, bitmap2, BITMAP_LEN); pr_err("\nStart testing find_bit() with sparse bitmap\n"); @@ -145,6 +165,7 @@ static int __init find_bit_test(void) test_find_next_zero_bit(bitmap, BITMAP_LEN); test_find_last_bit(bitmap, BITMAP_LEN); test_find_first_bit(bitmap, BITMAP_LEN); + test_find_first_and_bit(bitmap, bitmap2, BITMAP_LEN); test_find_next_and_bit(bitmap, bitmap2, BITMAP_LEN); /* -- cgit v1.2.3 From b5c7e7ec7d3418af2544452b45cc67297c857a86 Mon Sep 17 00:00:00 2001 From: Yury Norov Date: Sat, 14 Aug 2021 14:17:03 -0700 Subject: all: replace find_next{,_zero}_bit with find_first{,_zero}_bit where appropriate find_first{,_zero}_bit is a more effective analogue of 'next' version if start == 0. This patch replaces 'next' with 'first' where things look trivial. Signed-off-by: Yury Norov Tested-by: Wolfram Sang --- arch/powerpc/platforms/pasemi/dma_lib.c | 4 ++-- arch/s390/kvm/kvm-s390.c | 2 +- drivers/block/rnbd/rnbd-clt.c | 2 +- drivers/dma/ti/edma.c | 2 +- drivers/iio/adc/ad7124.c | 2 +- drivers/infiniband/hw/irdma/hw.c | 16 ++++++++-------- drivers/media/cec/core/cec-core.c | 2 +- drivers/media/mc/mc-devnode.c | 2 +- drivers/pci/controller/dwc/pci-dra7xx.c | 2 +- drivers/scsi/lpfc/lpfc_sli.c | 10 +++++----- drivers/soc/ti/k3-ringacc.c | 4 ++-- drivers/tty/n_tty.c | 2 +- drivers/virt/acrn/ioreq.c | 3 +-- fs/f2fs/segment.c | 8 ++++---- fs/ocfs2/cluster/heartbeat.c | 2 +- fs/ocfs2/dlm/dlmdomain.c | 4 ++-- fs/ocfs2/dlm/dlmmaster.c | 18 +++++++++--------- fs/ocfs2/dlm/dlmrecovery.c | 2 +- fs/ocfs2/dlm/dlmthread.c | 2 +- lib/genalloc.c | 2 +- net/ncsi/ncsi-manage.c | 4 ++-- 21 files changed, 47 insertions(+), 48 deletions(-) (limited to 'lib') diff --git a/arch/powerpc/platforms/pasemi/dma_lib.c b/arch/powerpc/platforms/pasemi/dma_lib.c index 270fa3c0d372..26427311fc72 100644 --- a/arch/powerpc/platforms/pasemi/dma_lib.c +++ b/arch/powerpc/platforms/pasemi/dma_lib.c @@ -375,7 +375,7 @@ int pasemi_dma_alloc_flag(void) int bit; retry: - bit = find_next_bit(flags_free, MAX_FLAGS, 0); + bit = find_first_bit(flags_free, MAX_FLAGS); if (bit >= MAX_FLAGS) return -ENOSPC; if (!test_and_clear_bit(bit, flags_free)) @@ -440,7 +440,7 @@ int pasemi_dma_alloc_fun(void) int bit; retry: - bit = find_next_bit(fun_free, MAX_FLAGS, 0); + bit = find_first_bit(fun_free, MAX_FLAGS); if (bit >= MAX_FLAGS) return -ENOSPC; if (!test_and_clear_bit(bit, fun_free)) diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c index 14a18ba5ff2c..9aba96d621b9 100644 --- a/arch/s390/kvm/kvm-s390.c +++ b/arch/s390/kvm/kvm-s390.c @@ -2021,7 +2021,7 @@ static unsigned long kvm_s390_next_dirty_cmma(struct kvm_memslots *slots, while ((slotidx > 0) && (ofs >= ms->npages)) { slotidx--; ms = slots->memslots + slotidx; - ofs = find_next_bit(kvm_second_dirty_bitmap(ms), ms->npages, 0); + ofs = find_first_bit(kvm_second_dirty_bitmap(ms), ms->npages); } return ms->base_gfn + ofs; } diff --git a/drivers/block/rnbd/rnbd-clt.c b/drivers/block/rnbd/rnbd-clt.c index 2df0657cdf00..cef1058ec2fd 100644 --- a/drivers/block/rnbd/rnbd-clt.c +++ b/drivers/block/rnbd/rnbd-clt.c @@ -196,7 +196,7 @@ rnbd_get_cpu_qlist(struct rnbd_clt_session *sess, int cpu) return per_cpu_ptr(sess->cpu_queues, bit); } else if (cpu != 0) { /* Search from 0 to cpu */ - bit = find_next_bit(sess->cpu_queues_bm, cpu, 0); + bit = find_first_bit(sess->cpu_queues_bm, cpu); if (bit < cpu) return per_cpu_ptr(sess->cpu_queues, bit); } diff --git a/drivers/dma/ti/edma.c b/drivers/dma/ti/edma.c index 35d81bd857f1..caa4050ecc02 100644 --- a/drivers/dma/ti/edma.c +++ b/drivers/dma/ti/edma.c @@ -1681,7 +1681,7 @@ static irqreturn_t dma_ccerr_handler(int irq, void *data) dev_dbg(ecc->dev, "EMR%d 0x%08x\n", j, val); emr = val; - for (i = find_next_bit(&emr, 32, 0); i < 32; + for (i = find_first_bit(&emr, 32); i < 32; i = find_next_bit(&emr, 32, i + 1)) { int k = (j << 5) + i; diff --git a/drivers/iio/adc/ad7124.c b/drivers/iio/adc/ad7124.c index e45c600fccc0..bc2cfa5f9592 100644 --- a/drivers/iio/adc/ad7124.c +++ b/drivers/iio/adc/ad7124.c @@ -347,7 +347,7 @@ static int ad7124_find_free_config_slot(struct ad7124_state *st) { unsigned int free_cfg_slot; - free_cfg_slot = find_next_zero_bit(&st->cfg_slots_status, AD7124_MAX_CONFIGS, 0); + free_cfg_slot = find_first_zero_bit(&st->cfg_slots_status, AD7124_MAX_CONFIGS); if (free_cfg_slot == AD7124_MAX_CONFIGS) return -1; diff --git a/drivers/infiniband/hw/irdma/hw.c b/drivers/infiniband/hw/irdma/hw.c index b4c657f5f2f9..e2dbd6b1ffcd 100644 --- a/drivers/infiniband/hw/irdma/hw.c +++ b/drivers/infiniband/hw/irdma/hw.c @@ -1709,14 +1709,14 @@ clean_msixtbl: */ static void irdma_get_used_rsrc(struct irdma_device *iwdev) { - iwdev->rf->used_pds = find_next_zero_bit(iwdev->rf->allocated_pds, - iwdev->rf->max_pd, 0); - iwdev->rf->used_qps = find_next_zero_bit(iwdev->rf->allocated_qps, - iwdev->rf->max_qp, 0); - iwdev->rf->used_cqs = find_next_zero_bit(iwdev->rf->allocated_cqs, - iwdev->rf->max_cq, 0); - iwdev->rf->used_mrs = find_next_zero_bit(iwdev->rf->allocated_mrs, - iwdev->rf->max_mr, 0); + iwdev->rf->used_pds = find_first_zero_bit(iwdev->rf->allocated_pds, + iwdev->rf->max_pd); + iwdev->rf->used_qps = find_first_zero_bit(iwdev->rf->allocated_qps, + iwdev->rf->max_qp); + iwdev->rf->used_cqs = find_first_zero_bit(iwdev->rf->allocated_cqs, + iwdev->rf->max_cq); + iwdev->rf->used_mrs = find_first_zero_bit(iwdev->rf->allocated_mrs, + iwdev->rf->max_mr); } void irdma_ctrl_deinit_hw(struct irdma_pci_f *rf) diff --git a/drivers/media/cec/core/cec-core.c b/drivers/media/cec/core/cec-core.c index 551689d371a7..7322e7cd9753 100644 --- a/drivers/media/cec/core/cec-core.c +++ b/drivers/media/cec/core/cec-core.c @@ -106,7 +106,7 @@ static int __must_check cec_devnode_register(struct cec_devnode *devnode, /* Part 1: Find a free minor number */ mutex_lock(&cec_devnode_lock); - minor = find_next_zero_bit(cec_devnode_nums, CEC_NUM_DEVICES, 0); + minor = find_first_zero_bit(cec_devnode_nums, CEC_NUM_DEVICES); if (minor == CEC_NUM_DEVICES) { mutex_unlock(&cec_devnode_lock); pr_err("could not get a free minor\n"); diff --git a/drivers/media/mc/mc-devnode.c b/drivers/media/mc/mc-devnode.c index f11382afe23b..680fbb3a9340 100644 --- a/drivers/media/mc/mc-devnode.c +++ b/drivers/media/mc/mc-devnode.c @@ -217,7 +217,7 @@ int __must_check media_devnode_register(struct media_device *mdev, /* Part 1: Find a free minor number */ mutex_lock(&media_devnode_lock); - minor = find_next_zero_bit(media_devnode_nums, MEDIA_NUM_DEVICES, 0); + minor = find_first_zero_bit(media_devnode_nums, MEDIA_NUM_DEVICES); if (minor == MEDIA_NUM_DEVICES) { mutex_unlock(&media_devnode_lock); pr_err("could not get a free minor\n"); diff --git a/drivers/pci/controller/dwc/pci-dra7xx.c b/drivers/pci/controller/dwc/pci-dra7xx.c index a4221f6f3629..279a6fa56584 100644 --- a/drivers/pci/controller/dwc/pci-dra7xx.c +++ b/drivers/pci/controller/dwc/pci-dra7xx.c @@ -213,7 +213,7 @@ static int dra7xx_pcie_handle_msi(struct pcie_port *pp, int index) if (!val) return 0; - pos = find_next_bit(&val, MAX_MSI_IRQS_PER_CTRL, 0); + pos = find_first_bit(&val, MAX_MSI_IRQS_PER_CTRL); while (pos != MAX_MSI_IRQS_PER_CTRL) { generic_handle_domain_irq(pp->irq_domain, (index * MAX_MSI_IRQS_PER_CTRL) + pos); diff --git a/drivers/scsi/lpfc/lpfc_sli.c b/drivers/scsi/lpfc/lpfc_sli.c index 5dedb3de271d..77dfe293bf23 100644 --- a/drivers/scsi/lpfc/lpfc_sli.c +++ b/drivers/scsi/lpfc/lpfc_sli.c @@ -17990,8 +17990,8 @@ lpfc_sli4_alloc_xri(struct lpfc_hba *phba) * the driver starts at 0 each time. */ spin_lock_irq(&phba->hbalock); - xri = find_next_zero_bit(phba->sli4_hba.xri_bmask, - phba->sli4_hba.max_cfg_param.max_xri, 0); + xri = find_first_zero_bit(phba->sli4_hba.xri_bmask, + phba->sli4_hba.max_cfg_param.max_xri); if (xri >= phba->sli4_hba.max_cfg_param.max_xri) { spin_unlock_irq(&phba->hbalock); return NO_XRI; @@ -19668,7 +19668,7 @@ lpfc_sli4_alloc_rpi(struct lpfc_hba *phba) max_rpi = phba->sli4_hba.max_cfg_param.max_rpi; rpi_limit = phba->sli4_hba.next_rpi; - rpi = find_next_zero_bit(phba->sli4_hba.rpi_bmask, rpi_limit, 0); + rpi = find_first_zero_bit(phba->sli4_hba.rpi_bmask, rpi_limit); if (rpi >= rpi_limit) rpi = LPFC_RPI_ALLOC_ERROR; else { @@ -20311,8 +20311,8 @@ next_priority: * have been tested so that we can detect when we should * change the priority level. */ - next_fcf_index = find_next_bit(phba->fcf.fcf_rr_bmask, - LPFC_SLI4_FCF_TBL_INDX_MAX, 0); + next_fcf_index = find_first_bit(phba->fcf.fcf_rr_bmask, + LPFC_SLI4_FCF_TBL_INDX_MAX); } diff --git a/drivers/soc/ti/k3-ringacc.c b/drivers/soc/ti/k3-ringacc.c index 312ba0f98ad7..573be88f8191 100644 --- a/drivers/soc/ti/k3-ringacc.c +++ b/drivers/soc/ti/k3-ringacc.c @@ -358,8 +358,8 @@ struct k3_ring *k3_ringacc_request_ring(struct k3_ringacc *ringacc, goto out; if (flags & K3_RINGACC_RING_USE_PROXY) { - proxy_id = find_next_zero_bit(ringacc->proxy_inuse, - ringacc->num_proxies, 0); + proxy_id = find_first_zero_bit(ringacc->proxy_inuse, + ringacc->num_proxies); if (proxy_id == ringacc->num_proxies) goto error; } diff --git a/drivers/tty/n_tty.c b/drivers/tty/n_tty.c index 5be6d02dc690..9fc2319a394d 100644 --- a/drivers/tty/n_tty.c +++ b/drivers/tty/n_tty.c @@ -1975,7 +1975,7 @@ static bool canon_copy_from_read_buf(struct tty_struct *tty, more = n - (size - tail); if (eol == N_TTY_BUF_SIZE && more) { /* scan wrapped without finding set bit */ - eol = find_next_bit(ldata->read_flags, more, 0); + eol = find_first_bit(ldata->read_flags, more); found = eol != more; } else found = eol != size; diff --git a/drivers/virt/acrn/ioreq.c b/drivers/virt/acrn/ioreq.c index 80b2e3f0e276..5ff1c53740c0 100644 --- a/drivers/virt/acrn/ioreq.c +++ b/drivers/virt/acrn/ioreq.c @@ -246,8 +246,7 @@ void acrn_ioreq_request_clear(struct acrn_vm *vm) spin_lock_bh(&vm->ioreq_clients_lock); client = vm->default_client; if (client) { - vcpu = find_next_bit(client->ioreqs_map, - ACRN_IO_REQUEST_MAX, 0); + vcpu = find_first_bit(client->ioreqs_map, ACRN_IO_REQUEST_MAX); while (vcpu < ACRN_IO_REQUEST_MAX) { acrn_ioreq_complete_request(client, vcpu, NULL); vcpu = find_next_bit(client->ioreqs_map, diff --git a/fs/f2fs/segment.c b/fs/f2fs/segment.c index df9ed75f0b7a..913552c98171 100644 --- a/fs/f2fs/segment.c +++ b/fs/f2fs/segment.c @@ -2558,8 +2558,8 @@ find_other_zone: secno = find_next_zero_bit(free_i->free_secmap, MAIN_SECS(sbi), hint); if (secno >= MAIN_SECS(sbi)) { if (dir == ALLOC_RIGHT) { - secno = find_next_zero_bit(free_i->free_secmap, - MAIN_SECS(sbi), 0); + secno = find_first_zero_bit(free_i->free_secmap, + MAIN_SECS(sbi)); f2fs_bug_on(sbi, secno >= MAIN_SECS(sbi)); } else { go_left = 1; @@ -2574,8 +2574,8 @@ find_other_zone: left_start--; continue; } - left_start = find_next_zero_bit(free_i->free_secmap, - MAIN_SECS(sbi), 0); + left_start = find_first_zero_bit(free_i->free_secmap, + MAIN_SECS(sbi)); f2fs_bug_on(sbi, left_start >= MAIN_SECS(sbi)); break; } diff --git a/fs/ocfs2/cluster/heartbeat.c b/fs/ocfs2/cluster/heartbeat.c index f89ffcbd585f..a17be1618bf7 100644 --- a/fs/ocfs2/cluster/heartbeat.c +++ b/fs/ocfs2/cluster/heartbeat.c @@ -379,7 +379,7 @@ static void o2hb_nego_timeout(struct work_struct *work) o2hb_fill_node_map(live_node_bitmap, sizeof(live_node_bitmap)); /* lowest node as master node to make negotiate decision. */ - master_node = find_next_bit(live_node_bitmap, O2NM_MAX_NODES, 0); + master_node = find_first_bit(live_node_bitmap, O2NM_MAX_NODES); if (master_node == o2nm_this_node()) { if (!test_bit(master_node, reg->hr_nego_node_bitmap)) { diff --git a/fs/ocfs2/dlm/dlmdomain.c b/fs/ocfs2/dlm/dlmdomain.c index 9f90fc9551e1..c4eccd499db8 100644 --- a/fs/ocfs2/dlm/dlmdomain.c +++ b/fs/ocfs2/dlm/dlmdomain.c @@ -1045,7 +1045,7 @@ static int dlm_send_regions(struct dlm_ctxt *dlm, unsigned long *node_map) int status, ret = 0, i; char *p; - if (find_next_bit(node_map, O2NM_MAX_NODES, 0) >= O2NM_MAX_NODES) + if (find_first_bit(node_map, O2NM_MAX_NODES) >= O2NM_MAX_NODES) goto bail; qr = kzalloc(sizeof(struct dlm_query_region), GFP_KERNEL); @@ -1217,7 +1217,7 @@ static int dlm_send_nodeinfo(struct dlm_ctxt *dlm, unsigned long *node_map) struct o2nm_node *node; int ret = 0, status, count, i; - if (find_next_bit(node_map, O2NM_MAX_NODES, 0) >= O2NM_MAX_NODES) + if (find_first_bit(node_map, O2NM_MAX_NODES) >= O2NM_MAX_NODES) goto bail; qn = kzalloc(sizeof(struct dlm_query_nodeinfo), GFP_KERNEL); diff --git a/fs/ocfs2/dlm/dlmmaster.c b/fs/ocfs2/dlm/dlmmaster.c index 9b88219febb5..227da5b1b6ab 100644 --- a/fs/ocfs2/dlm/dlmmaster.c +++ b/fs/ocfs2/dlm/dlmmaster.c @@ -861,7 +861,7 @@ lookup: * to see if there are any nodes that still need to be * considered. these will not appear in the mle nodemap * but they might own this lockres. wait on them. */ - bit = find_next_bit(dlm->recovery_map, O2NM_MAX_NODES, 0); + bit = find_first_bit(dlm->recovery_map, O2NM_MAX_NODES); if (bit < O2NM_MAX_NODES) { mlog(0, "%s: res %.*s, At least one node (%d) " "to recover before lock mastery can begin\n", @@ -912,7 +912,7 @@ redo_request: dlm_wait_for_recovery(dlm); spin_lock(&dlm->spinlock); - bit = find_next_bit(dlm->recovery_map, O2NM_MAX_NODES, 0); + bit = find_first_bit(dlm->recovery_map, O2NM_MAX_NODES); if (bit < O2NM_MAX_NODES) { mlog(0, "%s: res %.*s, At least one node (%d) " "to recover before lock mastery can begin\n", @@ -1079,7 +1079,7 @@ recheck: sleep = 1; /* have all nodes responded? */ if (voting_done && !*blocked) { - bit = find_next_bit(mle->maybe_map, O2NM_MAX_NODES, 0); + bit = find_first_bit(mle->maybe_map, O2NM_MAX_NODES); if (dlm->node_num <= bit) { /* my node number is lowest. * now tell other nodes that I am @@ -1234,8 +1234,8 @@ static int dlm_restart_lock_mastery(struct dlm_ctxt *dlm, } else { mlog(ML_ERROR, "node down! %d\n", node); if (blocked) { - int lowest = find_next_bit(mle->maybe_map, - O2NM_MAX_NODES, 0); + int lowest = find_first_bit(mle->maybe_map, + O2NM_MAX_NODES); /* act like it was never there */ clear_bit(node, mle->maybe_map); @@ -1795,7 +1795,7 @@ int dlm_assert_master_handler(struct o2net_msg *msg, u32 len, void *data, "MLE for it! (%.*s)\n", assert->node_idx, namelen, name); } else { - int bit = find_next_bit (mle->maybe_map, O2NM_MAX_NODES, 0); + int bit = find_first_bit(mle->maybe_map, O2NM_MAX_NODES); if (bit >= O2NM_MAX_NODES) { /* not necessarily an error, though less likely. * could be master just re-asserting. */ @@ -2521,7 +2521,7 @@ static int dlm_is_lockres_migratable(struct dlm_ctxt *dlm, } if (!nonlocal) { - node_ref = find_next_bit(res->refmap, O2NM_MAX_NODES, 0); + node_ref = find_first_bit(res->refmap, O2NM_MAX_NODES); if (node_ref >= O2NM_MAX_NODES) return 0; } @@ -3303,7 +3303,7 @@ static void dlm_clean_block_mle(struct dlm_ctxt *dlm, BUG_ON(mle->type != DLM_MLE_BLOCK); spin_lock(&mle->spinlock); - bit = find_next_bit(mle->maybe_map, O2NM_MAX_NODES, 0); + bit = find_first_bit(mle->maybe_map, O2NM_MAX_NODES); if (bit != dead_node) { mlog(0, "mle found, but dead node %u would not have been " "master\n", dead_node); @@ -3542,7 +3542,7 @@ void dlm_force_free_mles(struct dlm_ctxt *dlm) spin_lock(&dlm->master_lock); BUG_ON(dlm->dlm_state != DLM_CTXT_LEAVING); - BUG_ON((find_next_bit(dlm->domain_map, O2NM_MAX_NODES, 0) < O2NM_MAX_NODES)); + BUG_ON((find_first_bit(dlm->domain_map, O2NM_MAX_NODES) < O2NM_MAX_NODES)); for (i = 0; i < DLM_HASH_BUCKETS; i++) { bucket = dlm_master_hash(dlm, i); diff --git a/fs/ocfs2/dlm/dlmrecovery.c b/fs/ocfs2/dlm/dlmrecovery.c index 5cd5f7511dac..52ad342fec3e 100644 --- a/fs/ocfs2/dlm/dlmrecovery.c +++ b/fs/ocfs2/dlm/dlmrecovery.c @@ -451,7 +451,7 @@ static int dlm_do_recovery(struct dlm_ctxt *dlm) if (dlm->reco.dead_node == O2NM_INVALID_NODE_NUM) { int bit; - bit = find_next_bit (dlm->recovery_map, O2NM_MAX_NODES, 0); + bit = find_first_bit(dlm->recovery_map, O2NM_MAX_NODES); if (bit >= O2NM_MAX_NODES || bit < 0) dlm_set_reco_dead_node(dlm, O2NM_INVALID_NODE_NUM); else diff --git a/fs/ocfs2/dlm/dlmthread.c b/fs/ocfs2/dlm/dlmthread.c index c350bd4df770..eedf07ca23ca 100644 --- a/fs/ocfs2/dlm/dlmthread.c +++ b/fs/ocfs2/dlm/dlmthread.c @@ -92,7 +92,7 @@ int __dlm_lockres_unused(struct dlm_lock_resource *res) return 0; /* Another node has this resource with this node as the master */ - bit = find_next_bit(res->refmap, O2NM_MAX_NODES, 0); + bit = find_first_bit(res->refmap, O2NM_MAX_NODES); if (bit < O2NM_MAX_NODES) return 0; diff --git a/lib/genalloc.c b/lib/genalloc.c index 9a57257988c7..00fc50d0a640 100644 --- a/lib/genalloc.c +++ b/lib/genalloc.c @@ -251,7 +251,7 @@ void gen_pool_destroy(struct gen_pool *pool) list_del(&chunk->next_chunk); end_bit = chunk_size(chunk) >> order; - bit = find_next_bit(chunk->bits, end_bit, 0); + bit = find_first_bit(chunk->bits, end_bit); BUG_ON(bit < end_bit); vfree(chunk); diff --git a/net/ncsi/ncsi-manage.c b/net/ncsi/ncsi-manage.c index 7121ce2a47c0..78814417d753 100644 --- a/net/ncsi/ncsi-manage.c +++ b/net/ncsi/ncsi-manage.c @@ -608,7 +608,7 @@ static int clear_one_vid(struct ncsi_dev_priv *ndp, struct ncsi_channel *nc, bitmap = &ncf->bitmap; spin_lock_irqsave(&nc->lock, flags); - index = find_next_bit(bitmap, ncf->n_vids, 0); + index = find_first_bit(bitmap, ncf->n_vids); if (index >= ncf->n_vids) { spin_unlock_irqrestore(&nc->lock, flags); return -1; @@ -667,7 +667,7 @@ static int set_one_vid(struct ncsi_dev_priv *ndp, struct ncsi_channel *nc, return -1; } - index = find_next_zero_bit(bitmap, ncf->n_vids, 0); + index = find_first_zero_bit(bitmap, ncf->n_vids); if (index < 0 || index >= ncf->n_vids) { netdev_err(ndp->ndev.dev, "Channel %u already has all VLAN filters set\n", -- cgit v1.2.3 From db7313005e9c2d4e80888dd18d4a83926b920e8c Mon Sep 17 00:00:00 2001 From: Yury Norov Date: Sat, 14 Aug 2021 14:17:12 -0700 Subject: lib: bitmap: add performance test for bitmap_print_to_pagebuf Functional tests for bitmap_print_to_pagebuf() are provided in lib/test_printf.c. This patch adds performance test for a case of fully set bitmap. Signed-off-by: Yury Norov Tested-by: Wolfram Sang --- lib/test_bitmap.c | 37 +++++++++++++++++++++++++++++++++++++ 1 file changed, 37 insertions(+) (limited to 'lib') diff --git a/lib/test_bitmap.c b/lib/test_bitmap.c index d33fa5a61b95..0c82f07f74fc 100644 --- a/lib/test_bitmap.c +++ b/lib/test_bitmap.c @@ -446,6 +446,42 @@ static void __init test_bitmap_parselist(void) } } +static void __init test_bitmap_printlist(void) +{ + unsigned long *bmap = kmalloc(PAGE_SIZE, GFP_KERNEL); + char *buf = kmalloc(PAGE_SIZE, GFP_KERNEL); + char expected[256]; + int ret, slen; + ktime_t time; + + if (!buf || !bmap) + goto out; + + memset(bmap, -1, PAGE_SIZE); + slen = snprintf(expected, 256, "0-%ld", PAGE_SIZE * 8 - 1); + if (slen < 0) + goto out; + + time = ktime_get(); + ret = bitmap_print_to_pagebuf(true, buf, bmap, PAGE_SIZE * 8); + time = ktime_get() - time; + + if (ret != slen + 1) { + pr_err("bitmap_print_to_pagebuf: result is %d, expected %d\n", ret, slen); + goto out; + } + + if (strncmp(buf, expected, slen)) { + pr_err("bitmap_print_to_pagebuf: result is %s, expected %s\n", buf, expected); + goto out; + } + + pr_err("bitmap_print_to_pagebuf: input is '%s', Time: %llu\n", buf, time); +out: + kfree(buf); + kfree(bmap); +} + static const unsigned long parse_test[] __initconst = { BITMAP_FROM_U64(0), BITMAP_FROM_U64(1), @@ -818,6 +854,7 @@ static void __init selftest(void) test_bitmap_arr32(); test_bitmap_parse(); test_bitmap_parselist(); + test_bitmap_printlist(); test_mem_optimisations(); test_for_each_set_clump8(); test_bitmap_cut(); -- cgit v1.2.3 From 15325b4f768f2b27b5765489eeab6ec0d6b5e902 Mon Sep 17 00:00:00 2001 From: Yury Norov Date: Sat, 14 Aug 2021 14:17:13 -0700 Subject: vsprintf: rework bitmap_list_string bitmap_list_string() is very ineffective when printing bitmaps with long ranges of set bits because it calls find_next_bit for each bit in the bitmap. We can do better by detecting ranges of set bits. In my environment, before/after is 943008/31008 ns. Signed-off-by: Yury Norov Tested-by: Wolfram Sang --- lib/vsprintf.c | 24 +++++++----------------- 1 file changed, 7 insertions(+), 17 deletions(-) (limited to 'lib') diff --git a/lib/vsprintf.c b/lib/vsprintf.c index 58d5e567f836..d4b4e481045b 100644 --- a/lib/vsprintf.c +++ b/lib/vsprintf.c @@ -1241,20 +1241,13 @@ char *bitmap_list_string(char *buf, char *end, unsigned long *bitmap, struct printf_spec spec, const char *fmt) { int nr_bits = max_t(int, spec.field_width, 0); - /* current bit is 'cur', most recently seen range is [rbot, rtop] */ - int cur, rbot, rtop; bool first = true; + int rbot, rtop; if (check_pointer(&buf, end, bitmap, spec)) return buf; - rbot = cur = find_first_bit(bitmap, nr_bits); - while (cur < nr_bits) { - rtop = cur; - cur = find_next_bit(bitmap, nr_bits, cur + 1); - if (cur < nr_bits && cur <= rtop + 1) - continue; - + for_each_set_bitrange(rbot, rtop, bitmap, nr_bits) { if (!first) { if (buf < end) *buf = ','; @@ -1263,15 +1256,12 @@ char *bitmap_list_string(char *buf, char *end, unsigned long *bitmap, first = false; buf = number(buf, end, rbot, default_dec_spec); - if (rbot < rtop) { - if (buf < end) - *buf = '-'; - buf++; - - buf = number(buf, end, rtop, default_dec_spec); - } + if (rtop == rbot + 1) + continue; - rbot = cur; + if (buf < end) + *buf = '-'; + buf = number(++buf, end, rtop - 1, default_dec_spec); } return buf; } -- cgit v1.2.3 From e56e18985596617ae426ed5997fb2e737cffb58b Mon Sep 17 00:00:00 2001 From: "Justin M. Forbes" Date: Wed, 12 Jan 2022 08:01:38 -0600 Subject: lib/crypto: add prompts back to crypto libraries Commit 6048fdcc5f269 ("lib/crypto: blake2s: include as built-in") took away a number of prompt texts from other crypto libraries. This makes values flip from built-in to module when oldconfig runs, and causes problems when these crypto libs need to be built in for thingslike BIG_KEYS. Fixes: 6048fdcc5f269 ("lib/crypto: blake2s: include as built-in") Cc: Herbert Xu Cc: linux-crypto@vger.kernel.org Signed-off-by: Justin M. Forbes [Jason: - moved menu into submenu of lib/ instead of root menu - fixed chacha sub-dependencies for CONFIG_CRYPTO] Signed-off-by: Jason A. Donenfeld --- crypto/Kconfig | 2 -- lib/Kconfig | 2 ++ lib/crypto/Kconfig | 17 ++++++++++++----- 3 files changed, 14 insertions(+), 7 deletions(-) (limited to 'lib') diff --git a/crypto/Kconfig b/crypto/Kconfig index 94bfa32cc6a1..442765219c37 100644 --- a/crypto/Kconfig +++ b/crypto/Kconfig @@ -1928,5 +1928,3 @@ source "crypto/asymmetric_keys/Kconfig" source "certs/Kconfig" endif # if CRYPTO - -source "lib/crypto/Kconfig" diff --git a/lib/Kconfig b/lib/Kconfig index 655b0e43f260..c20b68ad2bc3 100644 --- a/lib/Kconfig +++ b/lib/Kconfig @@ -122,6 +122,8 @@ config INDIRECT_IOMEM_FALLBACK mmio accesses when the IO memory address is not a registered emulated region. +source "lib/crypto/Kconfig" + config CRC_CCITT tristate "CRC-CCITT functions" help diff --git a/lib/crypto/Kconfig b/lib/crypto/Kconfig index 8620f38e117c..e8e525650cf2 100644 --- a/lib/crypto/Kconfig +++ b/lib/crypto/Kconfig @@ -1,5 +1,7 @@ # SPDX-License-Identifier: GPL-2.0 +menu "Crypto library routines" + config CRYPTO_LIB_AES tristate @@ -31,7 +33,7 @@ config CRYPTO_ARCH_HAVE_LIB_CHACHA config CRYPTO_LIB_CHACHA_GENERIC tristate - select CRYPTO_ALGAPI + select XOR_BLOCKS help This symbol can be depended upon by arch implementations of the ChaCha library interface that require the generic code as a @@ -40,7 +42,8 @@ config CRYPTO_LIB_CHACHA_GENERIC of CRYPTO_LIB_CHACHA. config CRYPTO_LIB_CHACHA - tristate + tristate "ChaCha library interface" + depends on CRYPTO depends on CRYPTO_ARCH_HAVE_LIB_CHACHA || !CRYPTO_ARCH_HAVE_LIB_CHACHA select CRYPTO_LIB_CHACHA_GENERIC if CRYPTO_ARCH_HAVE_LIB_CHACHA=n help @@ -65,7 +68,7 @@ config CRYPTO_LIB_CURVE25519_GENERIC of CRYPTO_LIB_CURVE25519. config CRYPTO_LIB_CURVE25519 - tristate + tristate "Curve25519 scalar multiplication library" depends on CRYPTO_ARCH_HAVE_LIB_CURVE25519 || !CRYPTO_ARCH_HAVE_LIB_CURVE25519 select CRYPTO_LIB_CURVE25519_GENERIC if CRYPTO_ARCH_HAVE_LIB_CURVE25519=n help @@ -100,7 +103,7 @@ config CRYPTO_LIB_POLY1305_GENERIC of CRYPTO_LIB_POLY1305. config CRYPTO_LIB_POLY1305 - tristate + tristate "Poly1305 library interface" depends on CRYPTO_ARCH_HAVE_LIB_POLY1305 || !CRYPTO_ARCH_HAVE_LIB_POLY1305 select CRYPTO_LIB_POLY1305_GENERIC if CRYPTO_ARCH_HAVE_LIB_POLY1305=n help @@ -109,14 +112,18 @@ config CRYPTO_LIB_POLY1305 is available and enabled. config CRYPTO_LIB_CHACHA20POLY1305 - tristate + tristate "ChaCha20-Poly1305 AEAD support (8-byte nonce library version)" depends on CRYPTO_ARCH_HAVE_LIB_CHACHA || !CRYPTO_ARCH_HAVE_LIB_CHACHA depends on CRYPTO_ARCH_HAVE_LIB_POLY1305 || !CRYPTO_ARCH_HAVE_LIB_POLY1305 + depends on CRYPTO select CRYPTO_LIB_CHACHA select CRYPTO_LIB_POLY1305 + select CRYPTO_ALGAPI config CRYPTO_LIB_SHA256 tristate config CRYPTO_LIB_SM4 tristate + +endmenu -- cgit v1.2.3 From d8d83d8ab0a453e17e68b3a3bed1f940c34b8646 Mon Sep 17 00:00:00 2001 From: "Jason A. Donenfeld" Date: Tue, 11 Jan 2022 14:37:41 +0100 Subject: lib/crypto: blake2s: move hmac construction into wireguard Basically nobody should use blake2s in an HMAC construction; it already has a keyed variant. But unfortunately for historical reasons, Noise, used by WireGuard, uses HKDF quite strictly, which means we have to use this. Because this really shouldn't be used by others, this commit moves it into wireguard's noise.c locally, so that kernels that aren't using WireGuard don't get this superfluous code baked in. On m68k systems, this shaves off ~314 bytes. Cc: Herbert Xu Tested-by: Geert Uytterhoeven Acked-by: Ard Biesheuvel Signed-off-by: Jason A. Donenfeld --- drivers/net/wireguard/noise.c | 45 +++++++++++++++++++++++++++++++++++++------ include/crypto/blake2s.h | 3 --- lib/crypto/blake2s-selftest.c | 31 ----------------------------- lib/crypto/blake2s.c | 37 ----------------------------------- 4 files changed, 39 insertions(+), 77 deletions(-) (limited to 'lib') diff --git a/drivers/net/wireguard/noise.c b/drivers/net/wireguard/noise.c index c0cfd9b36c0b..720952b92e78 100644 --- a/drivers/net/wireguard/noise.c +++ b/drivers/net/wireguard/noise.c @@ -302,6 +302,41 @@ void wg_noise_set_static_identity_private_key( static_identity->static_public, private_key); } +static void hmac(u8 *out, const u8 *in, const u8 *key, const size_t inlen, const size_t keylen) +{ + struct blake2s_state state; + u8 x_key[BLAKE2S_BLOCK_SIZE] __aligned(__alignof__(u32)) = { 0 }; + u8 i_hash[BLAKE2S_HASH_SIZE] __aligned(__alignof__(u32)); + int i; + + if (keylen > BLAKE2S_BLOCK_SIZE) { + blake2s_init(&state, BLAKE2S_HASH_SIZE); + blake2s_update(&state, key, keylen); + blake2s_final(&state, x_key); + } else + memcpy(x_key, key, keylen); + + for (i = 0; i < BLAKE2S_BLOCK_SIZE; ++i) + x_key[i] ^= 0x36; + + blake2s_init(&state, BLAKE2S_HASH_SIZE); + blake2s_update(&state, x_key, BLAKE2S_BLOCK_SIZE); + blake2s_update(&state, in, inlen); + blake2s_final(&state, i_hash); + + for (i = 0; i < BLAKE2S_BLOCK_SIZE; ++i) + x_key[i] ^= 0x5c ^ 0x36; + + blake2s_init(&state, BLAKE2S_HASH_SIZE); + blake2s_update(&state, x_key, BLAKE2S_BLOCK_SIZE); + blake2s_update(&state, i_hash, BLAKE2S_HASH_SIZE); + blake2s_final(&state, i_hash); + + memcpy(out, i_hash, BLAKE2S_HASH_SIZE); + memzero_explicit(x_key, BLAKE2S_BLOCK_SIZE); + memzero_explicit(i_hash, BLAKE2S_HASH_SIZE); +} + /* This is Hugo Krawczyk's HKDF: * - https://eprint.iacr.org/2010/264.pdf * - https://tools.ietf.org/html/rfc5869 @@ -322,14 +357,14 @@ static void kdf(u8 *first_dst, u8 *second_dst, u8 *third_dst, const u8 *data, ((third_len || third_dst) && (!second_len || !second_dst)))); /* Extract entropy from data into secret */ - blake2s256_hmac(secret, data, chaining_key, data_len, NOISE_HASH_LEN); + hmac(secret, data, chaining_key, data_len, NOISE_HASH_LEN); if (!first_dst || !first_len) goto out; /* Expand first key: key = secret, data = 0x1 */ output[0] = 1; - blake2s256_hmac(output, output, secret, 1, BLAKE2S_HASH_SIZE); + hmac(output, output, secret, 1, BLAKE2S_HASH_SIZE); memcpy(first_dst, output, first_len); if (!second_dst || !second_len) @@ -337,8 +372,7 @@ static void kdf(u8 *first_dst, u8 *second_dst, u8 *third_dst, const u8 *data, /* Expand second key: key = secret, data = first-key || 0x2 */ output[BLAKE2S_HASH_SIZE] = 2; - blake2s256_hmac(output, output, secret, BLAKE2S_HASH_SIZE + 1, - BLAKE2S_HASH_SIZE); + hmac(output, output, secret, BLAKE2S_HASH_SIZE + 1, BLAKE2S_HASH_SIZE); memcpy(second_dst, output, second_len); if (!third_dst || !third_len) @@ -346,8 +380,7 @@ static void kdf(u8 *first_dst, u8 *second_dst, u8 *third_dst, const u8 *data, /* Expand third key: key = secret, data = second-key || 0x3 */ output[BLAKE2S_HASH_SIZE] = 3; - blake2s256_hmac(output, output, secret, BLAKE2S_HASH_SIZE + 1, - BLAKE2S_HASH_SIZE); + hmac(output, output, secret, BLAKE2S_HASH_SIZE + 1, BLAKE2S_HASH_SIZE); memcpy(third_dst, output, third_len); out: diff --git a/include/crypto/blake2s.h b/include/crypto/blake2s.h index df3c6c2f9553..f9ffd39194eb 100644 --- a/include/crypto/blake2s.h +++ b/include/crypto/blake2s.h @@ -101,7 +101,4 @@ static inline void blake2s(u8 *out, const u8 *in, const u8 *key, blake2s_final(&state, out); } -void blake2s256_hmac(u8 *out, const u8 *in, const u8 *key, const size_t inlen, - const size_t keylen); - #endif /* _CRYPTO_BLAKE2S_H */ diff --git a/lib/crypto/blake2s-selftest.c b/lib/crypto/blake2s-selftest.c index 5d9ea53be973..409e4b728770 100644 --- a/lib/crypto/blake2s-selftest.c +++ b/lib/crypto/blake2s-selftest.c @@ -15,7 +15,6 @@ * #include * * #include - * #include * * #define BLAKE2S_TESTVEC_COUNT 256 * @@ -58,16 +57,6 @@ * } * printf("};\n\n"); * - * printf("static const u8 blake2s_hmac_testvecs[][BLAKE2S_HASH_SIZE] __initconst = {\n"); - * - * HMAC(EVP_blake2s256(), key, sizeof(key), buf, sizeof(buf), hash, NULL); - * print_vec(hash, BLAKE2S_OUTBYTES); - * - * HMAC(EVP_blake2s256(), buf, sizeof(buf), key, sizeof(key), hash, NULL); - * print_vec(hash, BLAKE2S_OUTBYTES); - * - * printf("};\n"); - * * return 0; *} */ @@ -554,15 +543,6 @@ static const u8 blake2s_testvecs[][BLAKE2S_HASH_SIZE] __initconst = { 0xd6, 0x98, 0x6b, 0x07, 0x10, 0x65, 0x52, 0x65, }, }; -static const u8 blake2s_hmac_testvecs[][BLAKE2S_HASH_SIZE] __initconst = { - { 0xce, 0xe1, 0x57, 0x69, 0x82, 0xdc, 0xbf, 0x43, 0xad, 0x56, 0x4c, 0x70, - 0xed, 0x68, 0x16, 0x96, 0xcf, 0xa4, 0x73, 0xe8, 0xe8, 0xfc, 0x32, 0x79, - 0x08, 0x0a, 0x75, 0x82, 0xda, 0x3f, 0x05, 0x11, }, - { 0x77, 0x2f, 0x0c, 0x71, 0x41, 0xf4, 0x4b, 0x2b, 0xb3, 0xc6, 0xb6, 0xf9, - 0x60, 0xde, 0xe4, 0x52, 0x38, 0x66, 0xe8, 0xbf, 0x9b, 0x96, 0xc4, 0x9f, - 0x60, 0xd9, 0x24, 0x37, 0x99, 0xd6, 0xec, 0x31, }, -}; - bool __init blake2s_selftest(void) { u8 key[BLAKE2S_KEY_SIZE]; @@ -607,16 +587,5 @@ bool __init blake2s_selftest(void) } } - if (success) { - blake2s256_hmac(hash, buf, key, sizeof(buf), sizeof(key)); - success &= !memcmp(hash, blake2s_hmac_testvecs[0], BLAKE2S_HASH_SIZE); - - blake2s256_hmac(hash, key, buf, sizeof(key), sizeof(buf)); - success &= !memcmp(hash, blake2s_hmac_testvecs[1], BLAKE2S_HASH_SIZE); - - if (!success) - pr_err("blake2s256_hmac self-test: FAIL\n"); - } - return success; } diff --git a/lib/crypto/blake2s.c b/lib/crypto/blake2s.c index 93f2ae051370..9364f79937b8 100644 --- a/lib/crypto/blake2s.c +++ b/lib/crypto/blake2s.c @@ -30,43 +30,6 @@ void blake2s_final(struct blake2s_state *state, u8 *out) } EXPORT_SYMBOL(blake2s_final); -void blake2s256_hmac(u8 *out, const u8 *in, const u8 *key, const size_t inlen, - const size_t keylen) -{ - struct blake2s_state state; - u8 x_key[BLAKE2S_BLOCK_SIZE] __aligned(__alignof__(u32)) = { 0 }; - u8 i_hash[BLAKE2S_HASH_SIZE] __aligned(__alignof__(u32)); - int i; - - if (keylen > BLAKE2S_BLOCK_SIZE) { - blake2s_init(&state, BLAKE2S_HASH_SIZE); - blake2s_update(&state, key, keylen); - blake2s_final(&state, x_key); - } else - memcpy(x_key, key, keylen); - - for (i = 0; i < BLAKE2S_BLOCK_SIZE; ++i) - x_key[i] ^= 0x36; - - blake2s_init(&state, BLAKE2S_HASH_SIZE); - blake2s_update(&state, x_key, BLAKE2S_BLOCK_SIZE); - blake2s_update(&state, in, inlen); - blake2s_final(&state, i_hash); - - for (i = 0; i < BLAKE2S_BLOCK_SIZE; ++i) - x_key[i] ^= 0x5c ^ 0x36; - - blake2s_init(&state, BLAKE2S_HASH_SIZE); - blake2s_update(&state, x_key, BLAKE2S_BLOCK_SIZE); - blake2s_update(&state, i_hash, BLAKE2S_HASH_SIZE); - blake2s_final(&state, i_hash); - - memcpy(out, i_hash, BLAKE2S_HASH_SIZE); - memzero_explicit(x_key, BLAKE2S_BLOCK_SIZE); - memzero_explicit(i_hash, BLAKE2S_HASH_SIZE); -} -EXPORT_SYMBOL(blake2s256_hmac); - static int __init blake2s_mod_init(void) { if (!IS_ENABLED(CONFIG_CRYPTO_MANAGER_DISABLE_TESTS) && -- cgit v1.2.3 From 9a1536b093bb5bf60689021275fd24d513bb8db0 Mon Sep 17 00:00:00 2001 From: "Jason A. Donenfeld" Date: Tue, 11 Jan 2022 18:58:43 +0100 Subject: lib/crypto: sha1: re-roll loops to reduce code size With SHA-1 no longer being used for anything performance oriented, and also soon to be phased out entirely, we can make up for the space added by unrolled BLAKE2s by simply re-rolling SHA-1. Since SHA-1 is so much more complex, re-rolling it more or less takes care of the code size added by BLAKE2s. And eventually, hopefully we'll see SHA-1 removed entirely from most small kernel builds. Cc: Herbert Xu Cc: Ard Biesheuvel Tested-by: Geert Uytterhoeven Signed-off-by: Jason A. Donenfeld --- lib/sha1.c | 95 +++++++++----------------------------------------------------- 1 file changed, 14 insertions(+), 81 deletions(-) (limited to 'lib') diff --git a/lib/sha1.c b/lib/sha1.c index 9bd1935a1472..0494766fc574 100644 --- a/lib/sha1.c +++ b/lib/sha1.c @@ -9,6 +9,7 @@ #include #include #include +#include #include #include @@ -55,7 +56,8 @@ #define SHA_ROUND(t, input, fn, constant, A, B, C, D, E) do { \ __u32 TEMP = input(t); setW(t, TEMP); \ E += TEMP + rol32(A,5) + (fn) + (constant); \ - B = ror32(B, 2); } while (0) + B = ror32(B, 2); \ + TEMP = E; E = D; D = C; C = B; B = A; A = TEMP; } while (0) #define T_0_15(t, A, B, C, D, E) SHA_ROUND(t, SHA_SRC, (((C^D)&B)^D) , 0x5a827999, A, B, C, D, E ) #define T_16_19(t, A, B, C, D, E) SHA_ROUND(t, SHA_MIX, (((C^D)&B)^D) , 0x5a827999, A, B, C, D, E ) @@ -84,6 +86,7 @@ void sha1_transform(__u32 *digest, const char *data, __u32 *array) { __u32 A, B, C, D, E; + unsigned int i = 0; A = digest[0]; B = digest[1]; @@ -92,94 +95,24 @@ void sha1_transform(__u32 *digest, const char *data, __u32 *array) E = digest[4]; /* Round 1 - iterations 0-16 take their input from 'data' */ - T_0_15( 0, A, B, C, D, E); - T_0_15( 1, E, A, B, C, D); - T_0_15( 2, D, E, A, B, C); - T_0_15( 3, C, D, E, A, B); - T_0_15( 4, B, C, D, E, A); - T_0_15( 5, A, B, C, D, E); - T_0_15( 6, E, A, B, C, D); - T_0_15( 7, D, E, A, B, C); - T_0_15( 8, C, D, E, A, B); - T_0_15( 9, B, C, D, E, A); - T_0_15(10, A, B, C, D, E); - T_0_15(11, E, A, B, C, D); - T_0_15(12, D, E, A, B, C); - T_0_15(13, C, D, E, A, B); - T_0_15(14, B, C, D, E, A); - T_0_15(15, A, B, C, D, E); + for (; i < 16; ++i) + T_0_15(i, A, B, C, D, E); /* Round 1 - tail. Input from 512-bit mixing array */ - T_16_19(16, E, A, B, C, D); - T_16_19(17, D, E, A, B, C); - T_16_19(18, C, D, E, A, B); - T_16_19(19, B, C, D, E, A); + for (; i < 20; ++i) + T_16_19(i, A, B, C, D, E); /* Round 2 */ - T_20_39(20, A, B, C, D, E); - T_20_39(21, E, A, B, C, D); - T_20_39(22, D, E, A, B, C); - T_20_39(23, C, D, E, A, B); - T_20_39(24, B, C, D, E, A); - T_20_39(25, A, B, C, D, E); - T_20_39(26, E, A, B, C, D); - T_20_39(27, D, E, A, B, C); - T_20_39(28, C, D, E, A, B); - T_20_39(29, B, C, D, E, A); - T_20_39(30, A, B, C, D, E); - T_20_39(31, E, A, B, C, D); - T_20_39(32, D, E, A, B, C); - T_20_39(33, C, D, E, A, B); - T_20_39(34, B, C, D, E, A); - T_20_39(35, A, B, C, D, E); - T_20_39(36, E, A, B, C, D); - T_20_39(37, D, E, A, B, C); - T_20_39(38, C, D, E, A, B); - T_20_39(39, B, C, D, E, A); + for (; i < 40; ++i) + T_20_39(i, A, B, C, D, E); /* Round 3 */ - T_40_59(40, A, B, C, D, E); - T_40_59(41, E, A, B, C, D); - T_40_59(42, D, E, A, B, C); - T_40_59(43, C, D, E, A, B); - T_40_59(44, B, C, D, E, A); - T_40_59(45, A, B, C, D, E); - T_40_59(46, E, A, B, C, D); - T_40_59(47, D, E, A, B, C); - T_40_59(48, C, D, E, A, B); - T_40_59(49, B, C, D, E, A); - T_40_59(50, A, B, C, D, E); - T_40_59(51, E, A, B, C, D); - T_40_59(52, D, E, A, B, C); - T_40_59(53, C, D, E, A, B); - T_40_59(54, B, C, D, E, A); - T_40_59(55, A, B, C, D, E); - T_40_59(56, E, A, B, C, D); - T_40_59(57, D, E, A, B, C); - T_40_59(58, C, D, E, A, B); - T_40_59(59, B, C, D, E, A); + for (; i < 60; ++i) + T_40_59(i, A, B, C, D, E); /* Round 4 */ - T_60_79(60, A, B, C, D, E); - T_60_79(61, E, A, B, C, D); - T_60_79(62, D, E, A, B, C); - T_60_79(63, C, D, E, A, B); - T_60_79(64, B, C, D, E, A); - T_60_79(65, A, B, C, D, E); - T_60_79(66, E, A, B, C, D); - T_60_79(67, D, E, A, B, C); - T_60_79(68, C, D, E, A, B); - T_60_79(69, B, C, D, E, A); - T_60_79(70, A, B, C, D, E); - T_60_79(71, E, A, B, C, D); - T_60_79(72, D, E, A, B, C); - T_60_79(73, C, D, E, A, B); - T_60_79(74, B, C, D, E, A); - T_60_79(75, A, B, C, D, E); - T_60_79(76, E, A, B, C, D); - T_60_79(77, D, E, A, B, C); - T_60_79(78, C, D, E, A, B); - T_60_79(79, B, C, D, E, A); + for (; i < 80; ++i) + T_60_79(i, A, B, C, D, E); digest[0] += A; digest[1] += B; -- cgit v1.2.3 From 22c033989c3eb9731ad0c497dfab4231b8e367d6 Mon Sep 17 00:00:00 2001 From: Andy Shevchenko Date: Wed, 19 Jan 2022 18:08:12 -0800 Subject: include/linux/unaligned: replace kernel.h with the necessary inclusions When kernel.h is used in the headers it adds a lot into dependency hell, especially when there are circular dependencies are involved. Replace kernel.h inclusion with the list of what is really being used. The rest of the changes are induced by the above and may not be split. Link: https://lkml.kernel.org/r/20211209123823.20425-1-andriy.shevchenko@linux.intel.com Signed-off-by: Andy Shevchenko Acked-by: Arend van Spriel [brcmfmac] Acked-by: Kalle Valo Cc: Arend van Spriel Cc: Franky Lin Cc: Hante Meuleman Cc: Chi-hsien Lin Cc: Wright Feng Cc: Chung-hsien Hsu Cc: Kalle Valo Cc: David S. Miller Cc: Jakub Kicinski Cc: Heikki Krogerus Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- drivers/net/wireless/broadcom/brcm80211/brcmfmac/xtlv.c | 2 ++ include/linux/unaligned/packed_struct.h | 2 +- lib/lz4/lz4defs.h | 2 ++ 3 files changed, 5 insertions(+), 1 deletion(-) (limited to 'lib') diff --git a/drivers/net/wireless/broadcom/brcm80211/brcmfmac/xtlv.c b/drivers/net/wireless/broadcom/brcm80211/brcmfmac/xtlv.c index 2f3c451148db..2f8908074303 100644 --- a/drivers/net/wireless/broadcom/brcm80211/brcmfmac/xtlv.c +++ b/drivers/net/wireless/broadcom/brcm80211/brcmfmac/xtlv.c @@ -4,6 +4,8 @@ */ #include + +#include #include #include diff --git a/include/linux/unaligned/packed_struct.h b/include/linux/unaligned/packed_struct.h index c0d817de4df2..f4c8eaf4d012 100644 --- a/include/linux/unaligned/packed_struct.h +++ b/include/linux/unaligned/packed_struct.h @@ -1,7 +1,7 @@ #ifndef _LINUX_UNALIGNED_PACKED_STRUCT_H #define _LINUX_UNALIGNED_PACKED_STRUCT_H -#include +#include struct __una_u16 { u16 x; } __packed; struct __una_u32 { u32 x; } __packed; diff --git a/lib/lz4/lz4defs.h b/lib/lz4/lz4defs.h index 673bd206aa98..330aa539b46e 100644 --- a/lib/lz4/lz4defs.h +++ b/lib/lz4/lz4defs.h @@ -36,6 +36,8 @@ */ #include + +#include #include /* memset, memcpy */ #define FORCE_INLINE __always_inline -- cgit v1.2.3 From 70ac69928e9717a313a4c72647ebe80663e397a3 Mon Sep 17 00:00:00 2001 From: Alexey Dobriyan Date: Wed, 19 Jan 2022 18:08:53 -0800 Subject: kstrtox: uninline everything I've made a mistake of looking into lib/kstrtox.o code generation. The only function remotely performance critical is _parse_integer() (via /proc/*/map_files/*), everything else is not. Uninline everything, shrink lib/kstrtox.o by ~20 % ! Space savings on x86_64: add/remove: 0/0 grow/shrink: 0/23 up/down: 0/-1269 (-1269 !!!) Function old new delta kstrtoull 16 13 -3 kstrtouint 59 48 -11 kstrtou8 60 49 -11 kstrtou16 61 50 -11 _kstrtoul 46 35 -11 kstrtoull_from_user 95 83 -12 kstrtoul_from_user 95 83 -12 kstrtoll 93 80 -13 kstrtouint_from_user 124 83 -41 kstrtou8_from_user 125 83 -42 kstrtou16_from_user 126 83 -43 kstrtos8 101 50 -51 kstrtos16 102 51 -51 kstrtoint 100 49 -51 _kstrtol 93 35 -58 kstrtobool_from_user 156 75 -81 kstrtoll_from_user 165 83 -82 kstrtol_from_user 165 83 -82 kstrtoint_from_user 172 83 -89 kstrtos8_from_user 173 83 -90 kstrtos16_from_user 174 83 -91 _parse_integer 136 10 -126 _kstrtoull 308 101 -207 Total: Before=3421236, After=3419967, chg -0.04% Link: https://lkml.kernel.org/r/YZDsFDhHst4m2Pnt@localhost.localdomain Signed-off-by: Alexey Dobriyan Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- lib/kstrtox.c | 12 ++++++++++++ 1 file changed, 12 insertions(+) (limited to 'lib') diff --git a/lib/kstrtox.c b/lib/kstrtox.c index 059b8b00dc53..886510d248e5 100644 --- a/lib/kstrtox.c +++ b/lib/kstrtox.c @@ -22,6 +22,7 @@ #include "kstrtox.h" +noinline const char *_parse_integer_fixup_radix(const char *s, unsigned int *base) { if (*base == 0) { @@ -47,6 +48,7 @@ const char *_parse_integer_fixup_radix(const char *s, unsigned int *base) * * Don't you dare use this function. */ +noinline unsigned int _parse_integer_limit(const char *s, unsigned int base, unsigned long long *p, size_t max_chars) { @@ -85,6 +87,7 @@ unsigned int _parse_integer_limit(const char *s, unsigned int base, unsigned lon return rv; } +noinline unsigned int _parse_integer(const char *s, unsigned int base, unsigned long long *p) { return _parse_integer_limit(s, base, p, INT_MAX); @@ -125,6 +128,7 @@ static int _kstrtoull(const char *s, unsigned int base, unsigned long long *res) * Returns 0 on success, -ERANGE on overflow and -EINVAL on parsing error. * Preferred over simple_strtoull(). Return code must be checked. */ +noinline int kstrtoull(const char *s, unsigned int base, unsigned long long *res) { if (s[0] == '+') @@ -148,6 +152,7 @@ EXPORT_SYMBOL(kstrtoull); * Returns 0 on success, -ERANGE on overflow and -EINVAL on parsing error. * Preferred over simple_strtoll(). Return code must be checked. */ +noinline int kstrtoll(const char *s, unsigned int base, long long *res) { unsigned long long tmp; @@ -219,6 +224,7 @@ EXPORT_SYMBOL(_kstrtol); * Returns 0 on success, -ERANGE on overflow and -EINVAL on parsing error. * Preferred over simple_strtoul(). Return code must be checked. */ +noinline int kstrtouint(const char *s, unsigned int base, unsigned int *res) { unsigned long long tmp; @@ -249,6 +255,7 @@ EXPORT_SYMBOL(kstrtouint); * Returns 0 on success, -ERANGE on overflow and -EINVAL on parsing error. * Preferred over simple_strtol(). Return code must be checked. */ +noinline int kstrtoint(const char *s, unsigned int base, int *res) { long long tmp; @@ -264,6 +271,7 @@ int kstrtoint(const char *s, unsigned int base, int *res) } EXPORT_SYMBOL(kstrtoint); +noinline int kstrtou16(const char *s, unsigned int base, u16 *res) { unsigned long long tmp; @@ -279,6 +287,7 @@ int kstrtou16(const char *s, unsigned int base, u16 *res) } EXPORT_SYMBOL(kstrtou16); +noinline int kstrtos16(const char *s, unsigned int base, s16 *res) { long long tmp; @@ -294,6 +303,7 @@ int kstrtos16(const char *s, unsigned int base, s16 *res) } EXPORT_SYMBOL(kstrtos16); +noinline int kstrtou8(const char *s, unsigned int base, u8 *res) { unsigned long long tmp; @@ -309,6 +319,7 @@ int kstrtou8(const char *s, unsigned int base, u8 *res) } EXPORT_SYMBOL(kstrtou8); +noinline int kstrtos8(const char *s, unsigned int base, s8 *res) { long long tmp; @@ -333,6 +344,7 @@ EXPORT_SYMBOL(kstrtos8); * [oO][NnFf] for "on" and "off". Otherwise it will return -EINVAL. Value * pointed to by res is updated upon finding a match. */ +noinline int kstrtobool(const char *s, bool *res) { if (!s) -- cgit v1.2.3 From a31f9336ed48317d61c2299d595ed14294ffe5f9 Mon Sep 17 00:00:00 2001 From: Zhen Lei Date: Wed, 19 Jan 2022 18:08:59 -0800 Subject: lib/list_debug.c: print more list debugging context in __list_del_entry_valid() Currently, the entry->prev and entry->next are considered to be valid as long as they are not LIST_POISON{1|2}. However, the memory may be corrupted. The prev->next is invalid probably because 'prev' is invalid, not because prev->next's content is illegal. Unfortunately, the printk and its subfunctions will modify the registers that hold the 'prev' and 'next', and we don't see this valuable information in the BUG context. So print the contents of 'entry->prev' and 'entry->next'. Here's an example: list_del corruption. prev->next should be c0ecbf74, but was c08410dc kernel BUG at lib/list_debug.c:53! ... ... PC is at __list_del_entry_valid+0x58/0x98 LR is at __list_del_entry_valid+0x58/0x98 psr: 60000093 sp : c0ecbf30 ip : 00000000 fp : 00000001 r10: c08410d0 r9 : 00000001 r8 : c0825e0c r7 : 20000013 r6 : c08410d0 r5 : c0ecbf74 r4 : c0ecbf74 r3 : c0825d08 r2 : 00000000 r1 : df7ce6f4 r0 : 00000044 ... ... Stack: (0xc0ecbf30 to 0xc0ecc000) bf20: c0ecbf74 c0164fd0 c0ecbf70 c0165170 bf40: c0eca000 c0840c00 c0840c00 c0824500 c0825e0c c0189bbc c088f404 60000013 bf60: 60000013 c0e85100 000004ec 00000000 c0ebcdc0 c0ecbf74 c0ecbf74 c0825d08 bf80: c0e807c0 c018965c 00000000 c013f2a0 c0e807c0 c013f154 00000000 00000000 bfa0: 00000000 00000000 00000000 c01001b0 00000000 00000000 00000000 00000000 bfc0: 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000 bfe0: 00000000 00000000 00000000 00000000 00000013 00000000 00000000 00000000 (__list_del_entry_valid) from (__list_del_entry+0xc/0x20) (__list_del_entry) from (finish_swait+0x60/0x7c) (finish_swait) from (rcu_gp_kthread+0x560/0xa20) (rcu_gp_kthread) from (kthread+0x14c/0x15c) (kthread) from (ret_from_fork+0x14/0x24) At first, I thought prev->next was overwritten. Later, I carefully analyzed the RCU code and the disassembly code. The error occurred when deleting a node from the list rcu_state.gp_wq. The System.map shows that the address of rcu_state is c0840c00. Then I use gdb to obtain the offset of rcu_state.gp_wq.task_list. (gdb) p &((struct rcu_state *)0)->gp_wq.task_list $1 = (struct list_head *) 0x4dc Again: list_del corruption. prev->next should be c0ecbf74, but was c08410dc c08410dc = c0840c00 + 0x4dc = &rcu_state.gp_wq.task_list Because rcu_state.gp_wq has at most one node, so I can guess that "prev = &rcu_state.gp_wq.task_list". But for other scenes, maybe I wasn't so lucky, I cannot figure out the value of 'prev'. Link: https://lkml.kernel.org/r/20211207025835.1909-1-thunder.leizhen@huawei.com Signed-off-by: Zhen Lei Cc: "Paul E . McKenney" Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- lib/list_debug.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) (limited to 'lib') diff --git a/lib/list_debug.c b/lib/list_debug.c index 5d5424b51b74..9daa3fb9d1cd 100644 --- a/lib/list_debug.c +++ b/lib/list_debug.c @@ -49,11 +49,11 @@ bool __list_del_entry_valid(struct list_head *entry) "list_del corruption, %px->prev is LIST_POISON2 (%px)\n", entry, LIST_POISON2) || CHECK_DATA_CORRUPTION(prev->next != entry, - "list_del corruption. prev->next should be %px, but was %px\n", - entry, prev->next) || + "list_del corruption. prev->next should be %px, but was %px. (prev=%px)\n", + entry, prev->next, prev) || CHECK_DATA_CORRUPTION(next->prev != entry, - "list_del corruption. next->prev should be %px, but was %px\n", - entry, next->prev)) + "list_del corruption. next->prev should be %px, but was %px. (next=%px)\n", + entry, next->prev, next)) return false; return true; -- cgit v1.2.3 From fd0a1462405b087377e59b84e119fe7e2d08499a Mon Sep 17 00:00:00 2001 From: Isabella Basso Date: Wed, 19 Jan 2022 18:09:02 -0800 Subject: hash.h: remove unused define directive MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Patch series "test_hash.c: refactor into KUnit", v3. We refactored the lib/test_hash.c file into KUnit as part of the student group LKCAMP [1] introductory hackathon for kernel development. This test was pointed to our group by Daniel Latypov [2], so its full conversion into a pure KUnit test was our goal in this patch series, but we ran into many problems relating to it not being split as unit tests, which complicated matters a bit, as the reasoning behind the original tests is quite cryptic for those unfamiliar with hash implementations. Some interesting developments we'd like to highlight are: - In patch 1/5 we noticed that there was an unused define directive that could be removed. - In patch 4/5 we noticed how stringhash and hash tests are all under the lib/test_hash.c file, which might cause some confusion, and we also broke those kernel config entries up. Overall KUnit developments have been made in the other patches in this series: In patches 2/5, 3/5 and 5/5 we refactored the lib/test_hash.c file so as to make it more compatible with the KUnit style, whilst preserving the original idea of the maintainer who designed it (i.e. George Spelvin), which might be undesirable for unit tests, but we assume it is enough for a first patch. This patch (of 5): Currently, there exist hash_32() and __hash_32() functions, which were introduced in a patch [1] targeting architecture specific optimizations. These functions can be overridden on a per-architecture basis to achieve such optimizations. They must set their corresponding define directive (HAVE_ARCH_HASH_32 and HAVE_ARCH__HASH_32, respectively) so that header files can deal with these overrides properly. As the supported 32-bit architectures that have their own hash function implementation (i.e. m68k, Microblaze, H8/300, pa-risc) have only been making use of the (more general) __hash_32() function (which only lacks a right shift operation when compared to the hash_32() function), remove the define directive corresponding to the arch-specific hash_32() implementation. [1] https://lore.kernel.org/lkml/20160525073311.5600.qmail@ns.sciencehorizons.net/ [akpm@linux-foundation.org: hash_32_generic() becomes hash_32()] Link: https://lkml.kernel.org/r/20211208183711.390454-1-isabbasso@riseup.net Link: https://lkml.kernel.org/r/20211208183711.390454-2-isabbasso@riseup.net Reviewed-by: David Gow Tested-by: David Gow Co-developed-by: Augusto Durães Camargo Signed-off-by: Augusto Durães Camargo Co-developed-by: Enzo Ferreira Signed-off-by: Enzo Ferreira Signed-off-by: Isabella Basso Cc: Geert Uytterhoeven Cc: Brendan Higgins Cc: Daniel Latypov Cc: Shuah Khan Cc: Rodrigo Siqueira Cc: kernel test robot Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- drivers/infiniband/sw/rxe/rxe_qp.c | 3 +-- include/linux/hash.h | 5 +---- lib/test_hash.c | 24 +----------------------- tools/include/linux/hash.h | 5 +---- 4 files changed, 4 insertions(+), 33 deletions(-) (limited to 'lib') diff --git a/drivers/infiniband/sw/rxe/rxe_qp.c b/drivers/infiniband/sw/rxe/rxe_qp.c index 54b8711321c1..44c9ea601bff 100644 --- a/drivers/infiniband/sw/rxe/rxe_qp.c +++ b/drivers/infiniband/sw/rxe/rxe_qp.c @@ -217,8 +217,7 @@ static int rxe_qp_init_req(struct rxe_dev *rxe, struct rxe_qp *qp, * the port number must be in the Dynamic Ports range * (0xc000 - 0xffff). */ - qp->src_port = RXE_ROCE_V2_SPORT + - (hash_32_generic(qp_num(qp), 14) & 0x3fff); + qp->src_port = RXE_ROCE_V2_SPORT + (hash_32(qp_num(qp), 14) & 0x3fff); qp->sq.max_wr = init->cap.max_send_wr; /* These caps are limited by rxe_qp_chk_cap() done by the caller */ diff --git a/include/linux/hash.h b/include/linux/hash.h index ad6fa21d977b..38edaa08f862 100644 --- a/include/linux/hash.h +++ b/include/linux/hash.h @@ -62,10 +62,7 @@ static inline u32 __hash_32_generic(u32 val) return val * GOLDEN_RATIO_32; } -#ifndef HAVE_ARCH_HASH_32 -#define hash_32 hash_32_generic -#endif -static inline u32 hash_32_generic(u32 val, unsigned int bits) +static inline u32 hash_32(u32 val, unsigned int bits) { /* High bits are more random, so use them. */ return __hash_32(val) >> (32 - bits); diff --git a/lib/test_hash.c b/lib/test_hash.c index 0ee40b4a56dd..d4b0cfdb0377 100644 --- a/lib/test_hash.c +++ b/lib/test_hash.c @@ -94,22 +94,7 @@ test_int_hash(unsigned long long h64, u32 hash_or[2][33]) pr_err("hash_32(%#x, %d) = %#x > %#x", h0, k, h1, m); return false; } -#ifdef HAVE_ARCH_HASH_32 - h2 = hash_32_generic(h0, k); -#if HAVE_ARCH_HASH_32 == 1 - if (h1 != h2) { - pr_err("hash_32(%#x, %d) = %#x != hash_32_generic() " - " = %#x", h0, k, h1, h2); - return false; - } -#else - if (h2 > m) { - pr_err("hash_32_generic(%#x, %d) = %#x > %#x", - h0, k, h1, m); - return false; - } -#endif -#endif + /* Test hash_64 */ hash_or[1][k] |= h1 = hash_64(h64, k); if (h1 > m) { @@ -227,13 +212,6 @@ test_hash_init(void) #else pr_info("__hash_32() has no arch implementation to test."); #endif -#ifdef HAVE_ARCH_HASH_32 -#if HAVE_ARCH_HASH_32 != 1 - pr_info("hash_32() is arch-specific; not compared to generic."); -#endif -#else - pr_info("hash_32() has no arch implementation to test."); -#endif #ifdef HAVE_ARCH_HASH_64 #if HAVE_ARCH_HASH_64 != 1 pr_info("hash_64() is arch-specific; not compared to generic."); diff --git a/tools/include/linux/hash.h b/tools/include/linux/hash.h index ad6fa21d977b..38edaa08f862 100644 --- a/tools/include/linux/hash.h +++ b/tools/include/linux/hash.h @@ -62,10 +62,7 @@ static inline u32 __hash_32_generic(u32 val) return val * GOLDEN_RATIO_32; } -#ifndef HAVE_ARCH_HASH_32 -#define hash_32 hash_32_generic -#endif -static inline u32 hash_32_generic(u32 val, unsigned int bits) +static inline u32 hash_32(u32 val, unsigned int bits) { /* High bits are more random, so use them. */ return __hash_32(val) >> (32 - bits); -- cgit v1.2.3 From ae7880676bc8019ff61e49126c558ad7c4b6fa21 Mon Sep 17 00:00:00 2001 From: Isabella Basso Date: Wed, 19 Jan 2022 18:09:05 -0800 Subject: test_hash.c: split test_int_hash into arch-specific functions MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Split the test_int_hash function to keep its mainloop separate from arch-specific chunks, which are only compiled as needed. This aims at improving readability. Link: https://lkml.kernel.org/r/20211208183711.390454-3-isabbasso@riseup.net Reviewed-by: David Gow Tested-by: David Gow Signed-off-by: Isabella Basso Cc: Augusto Durães Camargo Cc: Brendan Higgins Cc: Daniel Latypov Cc: Enzo Ferreira Cc: Geert Uytterhoeven Cc: kernel test robot Cc: Rodrigo Siqueira Cc: Shuah Khan Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- lib/test_hash.c | 91 +++++++++++++++++++++++++++++++++++++++------------------ 1 file changed, 62 insertions(+), 29 deletions(-) (limited to 'lib') diff --git a/lib/test_hash.c b/lib/test_hash.c index d4b0cfdb0377..2b4fe4976cc4 100644 --- a/lib/test_hash.c +++ b/lib/test_hash.c @@ -56,6 +56,58 @@ fill_buf(char *buf, size_t len, u32 seed) } } +/* Holds most testing variables for the int test. */ +struct test_hash_params { + /* Pointer to integer to be hashed. */ + unsigned long long *h64; + /* Low 32-bits of integer to be hashed. */ + u32 h0; + /* Arch-specific hash result. */ + u32 h1; + /* Generic hash result. */ + u32 h2; + /* ORed hashes of given size (in bits). */ + u32 (*hash_or)[33]; +}; + +#ifdef HAVE_ARCH__HASH_32 +static bool __init +test_int__hash_32(struct test_hash_params *params) +{ + params->hash_or[1][0] |= params->h2 = __hash_32_generic(params->h0); +#if HAVE_ARCH__HASH_32 == 1 + if (params->h1 != params->h2) { + pr_err("__hash_32(%#x) = %#x != __hash_32_generic() = %#x", + params->h0, params->h1, params->h2); + return false; + } +#endif + return true; +} +#endif + +#ifdef HAVE_ARCH_HASH_64 +static bool __init +test_int_hash_64(struct test_hash_params *params, u32 const *m, int *k) +{ + params->h2 = hash_64_generic(*params->h64, *k); +#if HAVE_ARCH_HASH_64 == 1 + if (params->h1 != params->h2) { + pr_err("hash_64(%#llx, %d) = %#x != hash_64_generic() = %#x", + *params->h64, *k, params->h1, params->h2); + return false; + } +#else + if (params->h2 > *m) { + pr_err("hash_64_generic(%#llx, %d) = %#x > %#x", + *params->h64, *k, params->h1, *m); + return false; + } +#endif + return true; +} +#endif + /* * Test the various integer hash functions. h64 (or its low-order bits) * is the integer to hash. hash_or accumulates the OR of the hash values, @@ -69,19 +121,13 @@ static bool __init test_int_hash(unsigned long long h64, u32 hash_or[2][33]) { int k; - u32 h0 = (u32)h64, h1, h2; + struct test_hash_params params = { &h64, (u32)h64, 0, 0, hash_or }; /* Test __hash32 */ - hash_or[0][0] |= h1 = __hash_32(h0); + hash_or[0][0] |= params.h1 = __hash_32(params.h0); #ifdef HAVE_ARCH__HASH_32 - hash_or[1][0] |= h2 = __hash_32_generic(h0); -#if HAVE_ARCH__HASH_32 == 1 - if (h1 != h2) { - pr_err("__hash_32(%#x) = %#x != __hash_32_generic() = %#x", - h0, h1, h2); + if (!test_int__hash_32(¶ms)) return false; - } -#endif #endif /* Test k = 1..32 bits */ @@ -89,37 +135,24 @@ test_int_hash(unsigned long long h64, u32 hash_or[2][33]) u32 const m = ((u32)2 << (k-1)) - 1; /* Low k bits set */ /* Test hash_32 */ - hash_or[0][k] |= h1 = hash_32(h0, k); - if (h1 > m) { - pr_err("hash_32(%#x, %d) = %#x > %#x", h0, k, h1, m); + hash_or[0][k] |= params.h1 = hash_32(params.h0, k); + if (params.h1 > m) { + pr_err("hash_32(%#x, %d) = %#x > %#x", params.h0, k, params.h1, m); return false; } /* Test hash_64 */ - hash_or[1][k] |= h1 = hash_64(h64, k); - if (h1 > m) { - pr_err("hash_64(%#llx, %d) = %#x > %#x", h64, k, h1, m); + hash_or[1][k] |= params.h1 = hash_64(h64, k); + if (params.h1 > m) { + pr_err("hash_64(%#llx, %d) = %#x > %#x", h64, k, params.h1, m); return false; } #ifdef HAVE_ARCH_HASH_64 - h2 = hash_64_generic(h64, k); -#if HAVE_ARCH_HASH_64 == 1 - if (h1 != h2) { - pr_err("hash_64(%#llx, %d) = %#x != hash_64_generic() " - "= %#x", h64, k, h1, h2); + if (!test_int_hash_64(¶ms, &m, &k)) return false; - } -#else - if (h2 > m) { - pr_err("hash_64_generic(%#llx, %d) = %#x > %#x", - h64, k, h1, m); - return false; - } -#endif #endif } - (void)h2; /* Suppress unused variable warning */ return true; } -- cgit v1.2.3 From 5427d3d772a77a4d67fece057064832ec5cfa078 Mon Sep 17 00:00:00 2001 From: Isabella Basso Date: Wed, 19 Jan 2022 18:09:09 -0800 Subject: test_hash.c: split test_hash_init MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Split up test_hash_init so that it calls each test more explicitly insofar it is possible without rewriting the entire file. This aims at improving readability. Split tests performed on string_or as they don't interfere with those performed in hash_or. Also separate pr_info calls about skipped tests as they're not part of the tests themselves, but only warn about (un)defined arch-specific hash functions. Link: https://lkml.kernel.org/r/20211208183711.390454-4-isabbasso@riseup.net Reviewed-by: David Gow Tested-by: David Gow Signed-off-by: Isabella Basso Cc: Augusto Durães Camargo Cc: Brendan Higgins Cc: Daniel Latypov Cc: Enzo Ferreira Cc: Geert Uytterhoeven Cc: kernel test robot Cc: Rodrigo Siqueira Cc: Shuah Khan Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- lib/test_hash.c | 66 ++++++++++++++++++++++++++++++++++++++++++++++----------- 1 file changed, 54 insertions(+), 12 deletions(-) (limited to 'lib') diff --git a/lib/test_hash.c b/lib/test_hash.c index 2b4fe4976cc4..032849a48da7 100644 --- a/lib/test_hash.c +++ b/lib/test_hash.c @@ -158,11 +158,39 @@ test_int_hash(unsigned long long h64, u32 hash_or[2][33]) #define SIZE 256 /* Run time is cubic in SIZE */ -static int __init -test_hash_init(void) +static int __init test_string_or(void) { char buf[SIZE+1]; - u32 string_or = 0, hash_or[2][33] = { { 0, } }; + u32 string_or = 0; + int i, j; + + fill_buf(buf, SIZE, 1); + + /* Test every possible non-empty substring in the buffer. */ + for (j = SIZE; j > 0; --j) { + buf[j] = '\0'; + + for (i = 0; i <= j; i++) { + u32 h0 = full_name_hash(buf+i, buf+i, j-i); + + string_or |= h0; + } /* i */ + } /* j */ + + /* The OR of all the hash values should cover all the bits */ + if (~string_or) { + pr_err("OR of all string hash results = %#x != %#x", + string_or, -1u); + return -EINVAL; + } + + return 0; +} + +static int __init test_hash_or(void) +{ + char buf[SIZE+1]; + u32 hash_or[2][33] = { { 0, } }; unsigned tests = 0; unsigned long long h64 = 0; int i, j; @@ -192,7 +220,6 @@ test_hash_init(void) return -EINVAL; } - string_or |= h0; h64 = h64 << 32 | h0; /* For use with hash_64 */ if (!test_int_hash(h64, hash_or)) return -EINVAL; @@ -200,12 +227,6 @@ test_hash_init(void) } /* i */ } /* j */ - /* The OR of all the hash values should cover all the bits */ - if (~string_or) { - pr_err("OR of all string hash results = %#x != %#x", - string_or, -1u); - return -EINVAL; - } if (~hash_or[0][0]) { pr_err("OR of all __hash_32 results = %#x != %#x", hash_or[0][0], -1u); @@ -237,6 +258,13 @@ test_hash_init(void) } } + pr_notice("%u tests passed.", tests); + + return 0; +} + +static void __init notice_skipped_tests(void) +{ /* Issue notices about skipped tests. */ #ifdef HAVE_ARCH__HASH_32 #if HAVE_ARCH__HASH_32 != 1 @@ -252,10 +280,24 @@ test_hash_init(void) #else pr_info("hash_64() has no arch implementation to test."); #endif +} - pr_notice("%u tests passed.", tests); +static int __init +test_hash_init(void) +{ + int ret; - return 0; + ret = test_string_or(); + if (ret < 0) + return ret; + + ret = test_hash_or(); + if (ret < 0) + return ret; + + notice_skipped_tests(); + + return ret; } static void __exit test_hash_exit(void) -- cgit v1.2.3 From 88168bf35c5260013daab4bddf944cd557cb6f08 Mon Sep 17 00:00:00 2001 From: Isabella Basso Date: Wed, 19 Jan 2022 18:09:12 -0800 Subject: lib/Kconfig.debug: properly split hash test kernel entries MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Split TEST_HASH so that each entry only has one file. Note that there's no stringhash test file, but actually tests are performed in lib/test_hash.c. Link: https://lkml.kernel.org/r/20211208183711.390454-5-isabbasso@riseup.net Reviewed-by: David Gow Tested-by: David Gow Signed-off-by: Isabella Basso Cc: Augusto Durães Camargo Cc: Brendan Higgins Cc: Daniel Latypov Cc: Enzo Ferreira Cc: Geert Uytterhoeven Cc: kernel test robot Cc: Rodrigo Siqueira Cc: Shuah Khan Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- lib/Kconfig.debug | 14 +++++++++++--- lib/Makefile | 3 ++- 2 files changed, 13 insertions(+), 4 deletions(-) (limited to 'lib') diff --git a/lib/Kconfig.debug b/lib/Kconfig.debug index 5e14e32056ad..f27de2050ca0 100644 --- a/lib/Kconfig.debug +++ b/lib/Kconfig.debug @@ -2210,9 +2210,17 @@ config TEST_RHASHTABLE config TEST_HASH tristate "Perform selftest on hash functions" help - Enable this option to test the kernel's integer (), - string (), and siphash () - hash functions on boot (or module load). + Enable this option to test the kernel's integer (), and + string () hash functions on boot (or module load). + + This is intended to help people writing architecture-specific + optimized versions. If unsure, say N. + +config TEST_SIPHASH + tristate "Perform selftest on siphash functions" + help + Enable this option to test the kernel's siphash () hash + functions on boot (or module load). This is intended to help people writing architecture-specific optimized versions. If unsure, say N. diff --git a/lib/Makefile b/lib/Makefile index 364c23f15578..f3a2a251471d 100644 --- a/lib/Makefile +++ b/lib/Makefile @@ -61,7 +61,8 @@ obj-$(CONFIG_TEST_FIRMWARE) += test_firmware.o obj-$(CONFIG_TEST_BITOPS) += test_bitops.o CFLAGS_test_bitops.o += -Werror obj-$(CONFIG_TEST_SYSCTL) += test_sysctl.o -obj-$(CONFIG_TEST_HASH) += test_hash.o test_siphash.o +obj-$(CONFIG_TEST_SIPHASH) += test_siphash.o +obj-$(CONFIG_TEST_HASH) += test_hash.o obj-$(CONFIG_TEST_IDA) += test_ida.o obj-$(CONFIG_KASAN_KUNIT_TEST) += test_kasan.o CFLAGS_test_kasan.o += -fno-builtin -- cgit v1.2.3 From 0acc968f352336a459f27ba1f23745a174933c9c Mon Sep 17 00:00:00 2001 From: Isabella Basso Date: Wed, 19 Jan 2022 18:09:15 -0800 Subject: test_hash.c: refactor into kunit MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Use KUnit framework to make tests more easily integrable with CIs. Even though these tests are not yet properly written as unit tests this change should help in debugging. Also remove kernel messages (i.e. through pr_info) as KUnit handles all debugging output and let it handle module init and exit details. Link: https://lkml.kernel.org/r/20211208183711.390454-6-isabbasso@riseup.net Reviewed-by: David Gow Reported-by: kernel test robot Tested-by: David Gow Co-developed-by: Augusto Durães Camargo Signed-off-by: Augusto Durães Camargo Co-developed-by: Enzo Ferreira Signed-off-by: Enzo Ferreira Signed-off-by: Isabella Basso Cc: Brendan Higgins Cc: Daniel Latypov Cc: Geert Uytterhoeven Cc: Rodrigo Siqueira Cc: Shuah Khan Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- lib/Kconfig.debug | 28 +++++--- lib/Makefile | 2 +- lib/test_hash.c | 194 +++++++++++++++++------------------------------------- 3 files changed, 81 insertions(+), 143 deletions(-) (limited to 'lib') diff --git a/lib/Kconfig.debug b/lib/Kconfig.debug index f27de2050ca0..a789da4a19a1 100644 --- a/lib/Kconfig.debug +++ b/lib/Kconfig.debug @@ -2207,15 +2207,6 @@ config TEST_RHASHTABLE If unsure, say N. -config TEST_HASH - tristate "Perform selftest on hash functions" - help - Enable this option to test the kernel's integer (), and - string () hash functions on boot (or module load). - - This is intended to help people writing architecture-specific - optimized versions. If unsure, say N. - config TEST_SIPHASH tristate "Perform selftest on siphash functions" help @@ -2364,6 +2355,25 @@ config BITFIELD_KUNIT If unsure, say N. +config HASH_KUNIT_TEST + tristate "KUnit Test for integer hash functions" if !KUNIT_ALL_TESTS + depends on KUNIT + default KUNIT_ALL_TESTS + help + Enable this option to test the kernel's string (), and + integer () hash functions on boot. + + KUnit tests run during boot and output the results to the debug log + in TAP format (https://testanything.org/). Only useful for kernel devs + running the KUnit test harness, and not intended for inclusion into a + production build. + + For more information on KUnit and unit tests in general please refer + to the KUnit documentation in Documentation/dev-tools/kunit/. + + This is intended to help people writing architecture-specific + optimized versions. If unsure, say N. + config RESOURCE_KUNIT_TEST tristate "KUnit test for resource API" depends on KUNIT diff --git a/lib/Makefile b/lib/Makefile index f3a2a251471d..511c27827701 100644 --- a/lib/Makefile +++ b/lib/Makefile @@ -62,7 +62,7 @@ obj-$(CONFIG_TEST_BITOPS) += test_bitops.o CFLAGS_test_bitops.o += -Werror obj-$(CONFIG_TEST_SYSCTL) += test_sysctl.o obj-$(CONFIG_TEST_SIPHASH) += test_siphash.o -obj-$(CONFIG_TEST_HASH) += test_hash.o +obj-$(CONFIG_HASH_KUNIT_TEST) += test_hash.o obj-$(CONFIG_TEST_IDA) += test_ida.o obj-$(CONFIG_KASAN_KUNIT_TEST) += test_kasan.o CFLAGS_test_kasan.o += -fno-builtin diff --git a/lib/test_hash.c b/lib/test_hash.c index 032849a48da7..bb25fda34794 100644 --- a/lib/test_hash.c +++ b/lib/test_hash.c @@ -14,17 +14,15 @@ * and hash_64(). */ -#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt "\n" - #include #include #include #include #include -#include +#include /* 32-bit XORSHIFT generator. Seed must not be zero. */ -static u32 __init __attribute_const__ +static u32 __attribute_const__ xorshift(u32 seed) { seed ^= seed << 13; @@ -34,7 +32,7 @@ xorshift(u32 seed) } /* Given a non-zero x, returns a non-zero byte. */ -static u8 __init __attribute_const__ +static u8 __attribute_const__ mod255(u32 x) { x = (x & 0xffff) + (x >> 16); /* 1 <= x <= 0x1fffe */ @@ -45,8 +43,7 @@ mod255(u32 x) } /* Fill the buffer with non-zero bytes. */ -static void __init -fill_buf(char *buf, size_t len, u32 seed) +static void fill_buf(char *buf, size_t len, u32 seed) { size_t i; @@ -71,40 +68,32 @@ struct test_hash_params { }; #ifdef HAVE_ARCH__HASH_32 -static bool __init -test_int__hash_32(struct test_hash_params *params) +static void +test_int__hash_32(struct kunit *test, struct test_hash_params *params) { params->hash_or[1][0] |= params->h2 = __hash_32_generic(params->h0); #if HAVE_ARCH__HASH_32 == 1 - if (params->h1 != params->h2) { - pr_err("__hash_32(%#x) = %#x != __hash_32_generic() = %#x", - params->h0, params->h1, params->h2); - return false; - } + KUNIT_EXPECT_EQ_MSG(test, params->h1, params->h2, + "__hash_32(%#x) = %#x != __hash_32_generic() = %#x", + params->h0, params->h1, params->h2); #endif - return true; } #endif #ifdef HAVE_ARCH_HASH_64 -static bool __init -test_int_hash_64(struct test_hash_params *params, u32 const *m, int *k) +static void +test_int_hash_64(struct kunit *test, struct test_hash_params *params, u32 const *m, int *k) { params->h2 = hash_64_generic(*params->h64, *k); #if HAVE_ARCH_HASH_64 == 1 - if (params->h1 != params->h2) { - pr_err("hash_64(%#llx, %d) = %#x != hash_64_generic() = %#x", - *params->h64, *k, params->h1, params->h2); - return false; - } + KUNIT_EXPECT_EQ_MSG(test, params->h1, params->h2, + "hash_64(%#llx, %d) = %#x != hash_64_generic() = %#x", + *params->h64, *k, params->h1, params->h2); #else - if (params->h2 > *m) { - pr_err("hash_64_generic(%#llx, %d) = %#x > %#x", - *params->h64, *k, params->h1, *m); - return false; - } + KUNIT_EXPECT_LE_MSG(test, params->h1, params->h2, + "hash_64_generic(%#llx, %d) = %#x > %#x", + *params->h64, *k, params->h1, *m); #endif - return true; } #endif @@ -117,8 +106,8 @@ test_int_hash_64(struct test_hash_params *params, u32 const *m, int *k) * inline, the code being tested is actually in the module, and you can * recompile and re-test the module without rebooting. */ -static bool __init -test_int_hash(unsigned long long h64, u32 hash_or[2][33]) +static void +test_int_hash(struct kunit *test, unsigned long long h64, u32 hash_or[2][33]) { int k; struct test_hash_params params = { &h64, (u32)h64, 0, 0, hash_or }; @@ -126,8 +115,7 @@ test_int_hash(unsigned long long h64, u32 hash_or[2][33]) /* Test __hash32 */ hash_or[0][0] |= params.h1 = __hash_32(params.h0); #ifdef HAVE_ARCH__HASH_32 - if (!test_int__hash_32(¶ms)) - return false; + test_int__hash_32(test, ¶ms); #endif /* Test k = 1..32 bits */ @@ -136,29 +124,24 @@ test_int_hash(unsigned long long h64, u32 hash_or[2][33]) /* Test hash_32 */ hash_or[0][k] |= params.h1 = hash_32(params.h0, k); - if (params.h1 > m) { - pr_err("hash_32(%#x, %d) = %#x > %#x", params.h0, k, params.h1, m); - return false; - } + KUNIT_EXPECT_LE_MSG(test, params.h1, m, + "hash_32(%#x, %d) = %#x > %#x", + params.h0, k, params.h1, m); /* Test hash_64 */ hash_or[1][k] |= params.h1 = hash_64(h64, k); - if (params.h1 > m) { - pr_err("hash_64(%#llx, %d) = %#x > %#x", h64, k, params.h1, m); - return false; - } + KUNIT_EXPECT_LE_MSG(test, params.h1, m, + "hash_64(%#llx, %d) = %#x > %#x", + h64, k, params.h1, m); #ifdef HAVE_ARCH_HASH_64 - if (!test_int_hash_64(¶ms, &m, &k)) - return false; + test_int_hash_64(test, ¶ms, &m, &k); #endif } - - return true; } #define SIZE 256 /* Run time is cubic in SIZE */ -static int __init test_string_or(void) +static void test_string_or(struct kunit *test) { char buf[SIZE+1]; u32 string_or = 0; @@ -178,20 +161,15 @@ static int __init test_string_or(void) } /* j */ /* The OR of all the hash values should cover all the bits */ - if (~string_or) { - pr_err("OR of all string hash results = %#x != %#x", - string_or, -1u); - return -EINVAL; - } - - return 0; + KUNIT_EXPECT_EQ_MSG(test, string_or, -1u, + "OR of all string hash results = %#x != %#x", + string_or, -1u); } -static int __init test_hash_or(void) +static void test_hash_or(struct kunit *test) { char buf[SIZE+1]; u32 hash_or[2][33] = { { 0, } }; - unsigned tests = 0; unsigned long long h64 = 0; int i, j; @@ -206,39 +184,27 @@ static int __init test_hash_or(void) u32 h0 = full_name_hash(buf+i, buf+i, j-i); /* Check that hashlen_string gets the length right */ - if (hashlen_len(hashlen) != j-i) { - pr_err("hashlen_string(%d..%d) returned length" - " %u, expected %d", - i, j, hashlen_len(hashlen), j-i); - return -EINVAL; - } + KUNIT_EXPECT_EQ_MSG(test, hashlen_len(hashlen), j-i, + "hashlen_string(%d..%d) returned length %u, expected %d", + i, j, hashlen_len(hashlen), j-i); /* Check that the hashes match */ - if (hashlen_hash(hashlen) != h0) { - pr_err("hashlen_string(%d..%d) = %08x != " - "full_name_hash() = %08x", - i, j, hashlen_hash(hashlen), h0); - return -EINVAL; - } + KUNIT_EXPECT_EQ_MSG(test, hashlen_hash(hashlen), h0, + "hashlen_string(%d..%d) = %08x != full_name_hash() = %08x", + i, j, hashlen_hash(hashlen), h0); h64 = h64 << 32 | h0; /* For use with hash_64 */ - if (!test_int_hash(h64, hash_or)) - return -EINVAL; - tests++; + test_int_hash(test, h64, hash_or); } /* i */ } /* j */ - if (~hash_or[0][0]) { - pr_err("OR of all __hash_32 results = %#x != %#x", - hash_or[0][0], -1u); - return -EINVAL; - } + KUNIT_EXPECT_EQ_MSG(test, hash_or[0][0], -1u, + "OR of all __hash_32 results = %#x != %#x", + hash_or[0][0], -1u); #ifdef HAVE_ARCH__HASH_32 #if HAVE_ARCH__HASH_32 != 1 /* Test is pointless if results match */ - if (~hash_or[1][0]) { - pr_err("OR of all __hash_32_generic results = %#x != %#x", - hash_or[1][0], -1u); - return -EINVAL; - } + KUNIT_EXPECT_EQ_MSG(test, hash_or[1][0], -1u, + "OR of all __hash_32_generic results = %#x != %#x", + hash_or[1][0], -1u); #endif #endif @@ -246,65 +212,27 @@ static int __init test_hash_or(void) for (i = 1; i <= 32; i++) { u32 const m = ((u32)2 << (i-1)) - 1; /* Low i bits set */ - if (hash_or[0][i] != m) { - pr_err("OR of all hash_32(%d) results = %#x " - "(%#x expected)", i, hash_or[0][i], m); - return -EINVAL; - } - if (hash_or[1][i] != m) { - pr_err("OR of all hash_64(%d) results = %#x " - "(%#x expected)", i, hash_or[1][i], m); - return -EINVAL; - } + KUNIT_EXPECT_EQ_MSG(test, hash_or[0][i], m, + "OR of all hash_32(%d) results = %#x (%#x expected)", + i, hash_or[0][i], m); + KUNIT_EXPECT_EQ_MSG(test, hash_or[1][i], m, + "OR of all hash_64(%d) results = %#x (%#x expected)", + i, hash_or[1][i], m); } - - pr_notice("%u tests passed.", tests); - - return 0; } -static void __init notice_skipped_tests(void) -{ - /* Issue notices about skipped tests. */ -#ifdef HAVE_ARCH__HASH_32 -#if HAVE_ARCH__HASH_32 != 1 - pr_info("__hash_32() is arch-specific; not compared to generic."); -#endif -#else - pr_info("__hash_32() has no arch implementation to test."); -#endif -#ifdef HAVE_ARCH_HASH_64 -#if HAVE_ARCH_HASH_64 != 1 - pr_info("hash_64() is arch-specific; not compared to generic."); -#endif -#else - pr_info("hash_64() has no arch implementation to test."); -#endif -} - -static int __init -test_hash_init(void) -{ - int ret; - - ret = test_string_or(); - if (ret < 0) - return ret; - - ret = test_hash_or(); - if (ret < 0) - return ret; - - notice_skipped_tests(); +static struct kunit_case hash_test_cases[] __refdata = { + KUNIT_CASE(test_string_or), + KUNIT_CASE(test_hash_or), + {} +}; - return ret; -} +static struct kunit_suite hash_test_suite = { + .name = "hash", + .test_cases = hash_test_cases, +}; -static void __exit test_hash_exit(void) -{ -} -module_init(test_hash_init); /* Does everything */ -module_exit(test_hash_exit); /* Does nothing */ +kunit_test_suite(hash_test_suite); MODULE_LICENSE("GPL"); -- cgit v1.2.3 From e073e5ef90298d2d6e5e7f04b545a0815e92110c Mon Sep 17 00:00:00 2001 From: Andrey Konovalov Date: Wed, 19 Jan 2022 18:09:28 -0800 Subject: lib/test_meminit: destroy cache in kmem_cache_alloc_bulk() test Make do_kmem_cache_size_bulk() destroy the cache it creates. Link: https://lkml.kernel.org/r/aced20a94bf04159a139f0846e41d38a1537debb.1640018297.git.andreyknvl@google.com Fixes: 03a9349ac0e0 ("lib/test_meminit: add a kmem_cache_alloc_bulk() test") Signed-off-by: Andrey Konovalov Reviewed-by: Marco Elver Cc: Alexander Potapenko Cc: Dmitry Vyukov Cc: Andrey Ryabinin Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- lib/test_meminit.c | 1 + 1 file changed, 1 insertion(+) (limited to 'lib') diff --git a/lib/test_meminit.c b/lib/test_meminit.c index e4f706a404b3..3ca717f11397 100644 --- a/lib/test_meminit.c +++ b/lib/test_meminit.c @@ -337,6 +337,7 @@ static int __init do_kmem_cache_size_bulk(int size, int *total_failures) if (num) kmem_cache_free_bulk(c, num, objects); } + kmem_cache_destroy(c); *total_failures += fail; return 1; } -- cgit v1.2.3 From bbd2e05fad3e692ff2495895975bd0fce02bdbae Mon Sep 17 00:00:00 2001 From: Nathan Chancellor Date: Wed, 19 Jan 2022 18:10:28 -0800 Subject: lib/Kconfig.debug: make TEST_KMOD depend on PAGE_SIZE_LESS_THAN_256KB Commit b05fbcc36be1 ("btrfs: disable build on platforms having page size 256K") disabled btrfs for configurations that used a 256kB page size. However, it did not fully solve the problem because CONFIG_TEST_KMOD selects CONFIG_BTRFS, which does not account for the dependency. This results in a Kconfig warning and the failed BUILD_BUG_ON error returning. WARNING: unmet direct dependencies detected for BTRFS_FS Depends on [n]: BLOCK [=y] && !PPC_256K_PAGES && !PAGE_SIZE_256KB [=y] Selected by [m]: - TEST_KMOD [=m] && RUNTIME_TESTING_MENU [=y] && m && MODULES [=y] && NETDEVICES [=y] && NET_CORE [=y] && INET [=y] && BLOCK [=y] To resolve this, add CONFIG_PAGE_SIZE_LESS_THAN_256KB as a dependency of CONFIG_TEST_KMOD so there is no more invalid configuration or build errors. Link: https://lkml.kernel.org/r/20211129230141.228085-4-nathan@kernel.org Fixes: b05fbcc36be1 ("btrfs: disable build on platforms having page size 256K") Signed-off-by: Nathan Chancellor Reported-by: kernel test robot Cc: Chris Mason Cc: David Sterba Cc: Josef Bacik Cc: Luis Chamberlain Cc: Nick Desaulniers Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- lib/Kconfig.debug | 1 + 1 file changed, 1 insertion(+) (limited to 'lib') diff --git a/lib/Kconfig.debug b/lib/Kconfig.debug index a789da4a19a1..666e070feeb6 100644 --- a/lib/Kconfig.debug +++ b/lib/Kconfig.debug @@ -2505,6 +2505,7 @@ config TEST_KMOD depends on m depends on NETDEVICES && NET_CORE && INET # for TUN depends on BLOCK + depends on PAGE_SIZE_LESS_THAN_256KB # for BTRFS select TEST_LKM select XFS_FS select TUN -- cgit v1.2.3 From bece04b5b41dd7730dd06aec0d6b15c53d1fbb5a Mon Sep 17 00:00:00 2001 From: Marco Elver Date: Wed, 19 Jan 2022 18:10:31 -0800 Subject: kcov: fix generic Kconfig dependencies if ARCH_WANTS_NO_INSTR Until recent versions of GCC and Clang, it was not possible to disable KCOV instrumentation via a function attribute. The relevant function attribute was introduced in 540540d06e9d9 ("kcov: add __no_sanitize_coverage to fix noinstr for all architectures"). x86 was the first architecture to want a working noinstr, and at the time no compiler support for the attribute existed yet. Therefore, commit 0f1441b44e823 ("objtool: Fix noinstr vs KCOV") introduced the ability to NOP __sanitizer_cov_*() calls in .noinstr.text. However, this doesn't work for other architectures like arm64 and s390 that want a working noinstr per ARCH_WANTS_NO_INSTR. At the time of 0f1441b44e823, we didn't yet have ARCH_WANTS_NO_INSTR, but now we can move the Kconfig dependency checks to the generic KCOV option. KCOV will be available if: - architecture does not care about noinstr, OR - we have objtool support (like on x86), OR - GCC is 12.0 or newer, OR - Clang is 13.0 or newer. Link: https://lkml.kernel.org/r/20211201152604.3984495-1-elver@google.com Signed-off-by: Marco Elver Reviewed-by: Nathan Chancellor Acked-by: Peter Zijlstra (Intel) Cc: Mark Rutland Cc: Thomas Gleixner Cc: Ingo Molnar Cc: Borislav Petkov Cc: Dave Hansen Cc: H. Peter Anvin Cc: Nick Desaulniers Cc: Dmitry Vyukov Cc: Andrey Konovalov Cc: Catalin Marinas Cc: Will Deacon Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/x86/Kconfig | 2 +- lib/Kconfig.debug | 2 ++ 2 files changed, 3 insertions(+), 1 deletion(-) (limited to 'lib') diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig index 1275bab8be2c..f79a063d8ea3 100644 --- a/arch/x86/Kconfig +++ b/arch/x86/Kconfig @@ -78,7 +78,7 @@ config X86 select ARCH_HAS_FILTER_PGPROT select ARCH_HAS_FORTIFY_SOURCE select ARCH_HAS_GCOV_PROFILE_ALL - select ARCH_HAS_KCOV if X86_64 && STACK_VALIDATION + select ARCH_HAS_KCOV if X86_64 select ARCH_HAS_MEM_ENCRYPT select ARCH_HAS_MEMBARRIER_SYNC_CORE select ARCH_HAS_NON_OVERLAPPING_ADDRESS_SPACE diff --git a/lib/Kconfig.debug b/lib/Kconfig.debug index 666e070feeb6..1a5d168c388b 100644 --- a/lib/Kconfig.debug +++ b/lib/Kconfig.debug @@ -1979,6 +1979,8 @@ config KCOV bool "Code coverage for fuzzing" depends on ARCH_HAS_KCOV depends on CC_HAS_SANCOV_TRACE_PC || GCC_PLUGINS + depends on !ARCH_WANTS_NO_INSTR || STACK_VALIDATION || \ + GCC_VERSION >= 120000 || CLANG_VERSION >= 130000 select DEBUG_FS select GCC_PLUGIN_SANCOV if !CC_HAS_SANCOV_TRACE_PC help -- cgit v1.2.3 From 69d0db01e210e07fe915e5da91b54a867cda040f Mon Sep 17 00:00:00 2001 From: Kees Cook Date: Wed, 19 Jan 2022 18:10:35 -0800 Subject: ubsan: remove CONFIG_UBSAN_OBJECT_SIZE The object-size sanitizer is redundant to -Warray-bounds, and inappropriately performs its checks at run-time when all information needed for the evaluation is available at compile-time, making it quite difficult to use: https://bugzilla.kernel.org/show_bug.cgi?id=214861 With -Warray-bounds almost enabled globally, it doesn't make sense to keep this around. Link: https://lkml.kernel.org/r/20211203235346.110809-1-keescook@chromium.org Signed-off-by: Kees Cook Reviewed-by: Marco Elver Cc: Masahiro Yamada Cc: Michal Marek Cc: Nick Desaulniers Cc: Nathan Chancellor Cc: Andrey Ryabinin Cc: "Peter Zijlstra (Intel)" Cc: Stephen Rothwell Cc: Arnd Bergmann Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- lib/Kconfig.ubsan | 13 ------------- lib/test_ubsan.c | 22 ---------------------- scripts/Makefile.ubsan | 1 - 3 files changed, 36 deletions(-) (limited to 'lib') diff --git a/lib/Kconfig.ubsan b/lib/Kconfig.ubsan index e5372a13511d..236c5cefc4cc 100644 --- a/lib/Kconfig.ubsan +++ b/lib/Kconfig.ubsan @@ -112,19 +112,6 @@ config UBSAN_UNREACHABLE This option enables -fsanitize=unreachable which checks for control flow reaching an expected-to-be-unreachable position. -config UBSAN_OBJECT_SIZE - bool "Perform checking for accesses beyond the end of objects" - default UBSAN - # gcc hugely expands stack usage with -fsanitize=object-size - # https://lore.kernel.org/lkml/CAHk-=wjPasyJrDuwDnpHJS2TuQfExwe=px-SzLeN8GFMAQJPmQ@mail.gmail.com/ - depends on !CC_IS_GCC - depends on $(cc-option,-fsanitize=object-size) - help - This option enables -fsanitize=object-size which checks for accesses - beyond the end of objects where the optimizer can determine both the - object being operated on and its size, usually seen with bad downcasts, - or access to struct members from NULL pointers. - config UBSAN_BOOL bool "Perform checking for non-boolean values used as boolean" default UBSAN diff --git a/lib/test_ubsan.c b/lib/test_ubsan.c index 7e7bbd0f3fd2..2062be1f2e80 100644 --- a/lib/test_ubsan.c +++ b/lib/test_ubsan.c @@ -79,15 +79,6 @@ static void test_ubsan_load_invalid_value(void) eval2 = eval; } -static void test_ubsan_null_ptr_deref(void) -{ - volatile int *ptr = NULL; - int val; - - UBSAN_TEST(CONFIG_UBSAN_OBJECT_SIZE); - val = *ptr; -} - static void test_ubsan_misaligned_access(void) { volatile char arr[5] __aligned(4) = {1, 2, 3, 4, 5}; @@ -98,29 +89,16 @@ static void test_ubsan_misaligned_access(void) *ptr = val; } -static void test_ubsan_object_size_mismatch(void) -{ - /* "((aligned(8)))" helps this not into be misaligned for ptr-access. */ - volatile int val __aligned(8) = 4; - volatile long long *ptr, val2; - - UBSAN_TEST(CONFIG_UBSAN_OBJECT_SIZE); - ptr = (long long *)&val; - val2 = *ptr; -} - static const test_ubsan_fp test_ubsan_array[] = { test_ubsan_shift_out_of_bounds, test_ubsan_out_of_bounds, test_ubsan_load_invalid_value, test_ubsan_misaligned_access, - test_ubsan_object_size_mismatch, }; /* Excluded because they Oops the module. */ static const test_ubsan_fp skip_ubsan_array[] = { test_ubsan_divrem_overflow, - test_ubsan_null_ptr_deref, }; static int __init test_ubsan_init(void) diff --git a/scripts/Makefile.ubsan b/scripts/Makefile.ubsan index 9e2092fd5206..7099c603ff0a 100644 --- a/scripts/Makefile.ubsan +++ b/scripts/Makefile.ubsan @@ -8,7 +8,6 @@ ubsan-cflags-$(CONFIG_UBSAN_LOCAL_BOUNDS) += -fsanitize=local-bounds ubsan-cflags-$(CONFIG_UBSAN_SHIFT) += -fsanitize=shift ubsan-cflags-$(CONFIG_UBSAN_DIV_ZERO) += -fsanitize=integer-divide-by-zero ubsan-cflags-$(CONFIG_UBSAN_UNREACHABLE) += -fsanitize=unreachable -ubsan-cflags-$(CONFIG_UBSAN_OBJECT_SIZE) += -fsanitize=object-size ubsan-cflags-$(CONFIG_UBSAN_BOOL) += -fsanitize=bool ubsan-cflags-$(CONFIG_UBSAN_ENUM) += -fsanitize=enum ubsan-cflags-$(CONFIG_UBSAN_TRAP) += -fsanitize-undefined-trap-on-error -- cgit v1.2.3 From b1e78ef3be2533973953a35a56739fda7325875c Mon Sep 17 00:00:00 2001 From: Colin Ian King Date: Wed, 19 Jan 2022 18:10:38 -0800 Subject: lib: remove redundant assignment to variable ret The variable ret is being assigned a value that is never read. If the for-loop is entered then ret is immediately re-assigned a new value. If the for-loop is not executed ret is never read. The assignment is redundant and can be removed. Link: https://lkml.kernel.org/r/20211230134557.83633-1-colin.i.king@gmail.com Signed-off-by: Colin Ian King Reviewed-by: Jarkko Sakkinen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- lib/asn1_encoder.c | 2 -- 1 file changed, 2 deletions(-) (limited to 'lib') diff --git a/lib/asn1_encoder.c b/lib/asn1_encoder.c index 27bbe891714f..0fd3c454a468 100644 --- a/lib/asn1_encoder.c +++ b/lib/asn1_encoder.c @@ -164,8 +164,6 @@ asn1_encode_oid(unsigned char *data, const unsigned char *end_data, data_len -= 3; - ret = 0; - for (i = 2; i < oid_len; i++) { ret = asn1_encode_oid_digit(&d, &data_len, oid[i]); if (ret < 0) -- cgit v1.2.3 From 04bc883c986d9c8a64fc1f1cc2cbc328c2b2a496 Mon Sep 17 00:00:00 2001 From: Luis Chamberlain Date: Fri, 21 Jan 2022 22:11:54 -0800 Subject: test_sysctl: simplify subdirectory registration with register_sysctl() There is no need to user boiler plate code to specify a set of base directories we're going to stuff sysctls under. Simplify this by using register_sysctl() and specifying the directory path directly. // pycocci sysctl-subdir-register-sysctl-simplify.cocci lib/test_sysctl.c @c1@ expression E1; identifier subdir, sysctls; @@ static struct ctl_table subdir[] = { { .procname = E1, .maxlen = 0, .mode = 0555, .child = sysctls, }, { } }; @c2@ identifier c1.subdir; expression E2; identifier base; @@ static struct ctl_table base[] = { { .procname = E2, .maxlen = 0, .mode = 0555, .child = subdir, }, { } }; @c3@ identifier c2.base; identifier header; @@ header = register_sysctl_table(base); @r1 depends on c1 && c2 && c3@ expression c1.E1; identifier c1.subdir, c1.sysctls; @@ -static struct ctl_table subdir[] = { - { - .procname = E1, - .maxlen = 0, - .mode = 0555, - .child = sysctls, - }, - { } -}; @r2 depends on c1 && c2 && c3@ identifier c1.subdir; expression c2.E2; identifier c2.base; @@ -static struct ctl_table base[] = { - { - .procname = E2, - .maxlen = 0, - .mode = 0555, - .child = subdir, - }, - { } -}; @initialize:python@ @@ def make_my_fresh_expression(s1, s2): return '"' + s1.strip('"') + "/" + s2.strip('"') + '"' @r3 depends on c1 && c2 && c3@ expression c1.E1; identifier c1.sysctls; expression c2.E2; identifier c2.base; identifier c3.header; fresh identifier E3 = script:python(E2, E1) { make_my_fresh_expression(E2, E1) }; @@ header = -register_sysctl_table(base); +register_sysctl(E3, sysctls); Generated-by: Coccinelle SmPL Link: https://lkml.kernel.org/r/20211123202422.819032-6-mcgrof@kernel.org Signed-off-by: Luis Chamberlain Cc: Al Viro Cc: Amir Goldstein Cc: Andy Shevchenko Cc: Antti Palosaari Cc: Arnd Bergmann Cc: Benjamin Herrenschmidt Cc: Benjamin LaHaise Cc: Clemens Ladisch Cc: David Airlie Cc: "Eric W. Biederman" Cc: Greg Kroah-Hartman Cc: Iurii Zaikin Cc: Jani Nikula Cc: Jan Kara Cc: Joel Becker Cc: Joonas Lahtinen Cc: Joseph Qi Cc: Julia Lawall Cc: Kees Cook Cc: Lukas Middendorf Cc: Mark Fasheh Cc: Paul Turner Cc: Peter Zijlstra Cc: Petr Mladek Cc: Phillip Potter Cc: Qing Wang Cc: Rodrigo Vivi Cc: Sebastian Reichel Cc: Sergey Senozhatsky Cc: Stephen Kitt Cc: Tetsuo Handa Cc: Xiaoming Ni Cc: Douglas Gilbert Cc: James E.J. Bottomley Cc: Jani Nikula Cc: John Ogness Cc: Martin K. Petersen Cc: "Rafael J. Wysocki" Cc: Steven Rostedt (VMware) Cc: Suren Baghdasaryan Cc: "Theodore Ts'o" Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- lib/test_sysctl.c | 22 +--------------------- 1 file changed, 1 insertion(+), 21 deletions(-) (limited to 'lib') diff --git a/lib/test_sysctl.c b/lib/test_sysctl.c index 3750323973f4..a5a3d6c27e1f 100644 --- a/lib/test_sysctl.c +++ b/lib/test_sysctl.c @@ -128,26 +128,6 @@ static struct ctl_table test_table[] = { { } }; -static struct ctl_table test_sysctl_table[] = { - { - .procname = "test_sysctl", - .maxlen = 0, - .mode = 0555, - .child = test_table, - }, - { } -}; - -static struct ctl_table test_sysctl_root_table[] = { - { - .procname = "debug", - .maxlen = 0, - .mode = 0555, - .child = test_sysctl_table, - }, - { } -}; - static struct ctl_table_header *test_sysctl_header; static int __init test_sysctl_init(void) @@ -155,7 +135,7 @@ static int __init test_sysctl_init(void) test_data.bitmap_0001 = kzalloc(SYSCTL_TEST_BITMAP_SIZE/8, GFP_KERNEL); if (!test_data.bitmap_0001) return -ENOMEM; - test_sysctl_header = register_sysctl_table(test_sysctl_root_table); + test_sysctl_header = register_sysctl("debug/test_sysctl", test_table); if (!test_sysctl_header) { kfree(test_data.bitmap_0001); return -ENOMEM; -- cgit v1.2.3 From 2dba5eb1c73b6ba2988ced07250edeac0f8cbf5a Mon Sep 17 00:00:00 2001 From: Vlastimil Babka Date: Fri, 21 Jan 2022 22:14:27 -0800 Subject: lib/stackdepot: allow optional init and stack_table allocation by kvmalloc() Currently, enabling CONFIG_STACKDEPOT means its stack_table will be allocated from memblock, even if stack depot ends up not actually used. The default size of stack_table is 4MB on 32-bit, 8MB on 64-bit. This is fine for use-cases such as KASAN which is also a config option and has overhead on its own. But it's an issue for functionality that has to be actually enabled on boot (page_owner) or depends on hardware (GPU drivers) and thus the memory might be wasted. This was raised as an issue [1] when attempting to add stackdepot support for SLUB's debug object tracking functionality. It's common to build kernels with CONFIG_SLUB_DEBUG and enable slub_debug on boot only when needed, or create only specific kmem caches with debugging for testing purposes. It would thus be more efficient if stackdepot's table was allocated only when actually going to be used. This patch thus makes the allocation (and whole stack_depot_init() call) optional: - Add a CONFIG_STACKDEPOT_ALWAYS_INIT flag to keep using the current well-defined point of allocation as part of mem_init(). Make CONFIG_KASAN select this flag. - Other users have to call stack_depot_init() as part of their own init when it's determined that stack depot will actually be used. This may depend on both config and runtime conditions. Convert current users which are page_owner and several in the DRM subsystem. Same will be done for SLUB later. - Because the init might now be called after the boot-time memblock allocation has given all memory to the buddy allocator, change stack_depot_init() to allocate stack_table with kvmalloc() when memblock is no longer available. Also handle allocation failure by disabling stackdepot (could have theoretically happened even with memblock allocation previously), and don't unnecessarily align the memblock allocation to its own size anymore. [1] https://lore.kernel.org/all/CAMuHMdW=eoVzM1Re5FVoEN87nKfiLmM2+Ah7eNu2KXEhCvbZyA@mail.gmail.com/ Link: https://lkml.kernel.org/r/20211013073005.11351-1-vbabka@suse.cz Signed-off-by: Vlastimil Babka Acked-by: Dmitry Vyukov Reviewed-by: Marco Elver # stackdepot Cc: Marco Elver Cc: Vijayanand Jitta Cc: Maarten Lankhorst Cc: Maxime Ripard Cc: Thomas Zimmermann Cc: David Airlie Cc: Daniel Vetter Cc: Andrey Ryabinin Cc: Alexander Potapenko Cc: Andrey Konovalov Cc: Dmitry Vyukov Cc: Geert Uytterhoeven Cc: Oliver Glitta Cc: Imran Khan From: Colin Ian King Subject: lib/stackdepot: fix spelling mistake and grammar in pr_err message There is a spelling mistake of the work allocation so fix this and re-phrase the message to make it easier to read. Link: https://lkml.kernel.org/r/20211015104159.11282-1-colin.king@canonical.com Signed-off-by: Colin Ian King Cc: Vlastimil Babka From: Vlastimil Babka Subject: lib/stackdepot: allow optional init and stack_table allocation by kvmalloc() - fixup On FLATMEM, we call page_ext_init_flatmem_late() just before kmem_cache_init() which means stack_depot_init() (called by page owner init) will not recognize properly it should use kvmalloc() and not memblock_alloc(). memblock_alloc() will also not issue a warning and return a block memory that can be invalid and cause kernel page fault when saving stacks, as reported by the kernel test robot [1]. Fix this by moving page_ext_init_flatmem_late() below kmem_cache_init() so that slab_is_available() is true during stack_depot_init(). SPARSEMEM doesn't have this issue, as it doesn't do page_ext_init_flatmem_late(), but a different page_ext_init() even later in the boot process. Thanks to Mike Rapoport for pointing out the FLATMEM init ordering issue. While at it, also actually resolve a checkpatch warning in stack_depot_init() from DRM CI, which was supposed to be in the original patch already. [1] https://lore.kernel.org/all/20211014085450.GC18719@xsang-OptiPlex-9020/ Link: https://lkml.kernel.org/r/6abd9213-19a9-6d58-cedc-2414386d2d81@suse.cz Signed-off-by: Vlastimil Babka Reported-by: kernel test robot Cc: Mike Rapoport Cc: Stephen Rothwell From: Vlastimil Babka Subject: lib/stackdepot: allow optional init and stack_table allocation by kvmalloc() - fixup3 Due to cd06ab2fd48f ("drm/locking: add backtrace for locking contended locks without backoff") landing recently to -next adding a new stack depot user in drivers/gpu/drm/drm_modeset_lock.c we need to add an appropriate call to stack_depot_init() there as well. Link: https://lkml.kernel.org/r/2a692365-cfa1-64f2-34e0-8aa5674dce5e@suse.cz Signed-off-by: Vlastimil Babka Cc: Jani Nikula Cc: Naresh Kamboju Cc: Marco Elver Cc: Vijayanand Jitta Cc: Maarten Lankhorst Cc: Maxime Ripard Cc: Thomas Zimmermann Cc: David Airlie Cc: Daniel Vetter Cc: Andrey Ryabinin Cc: Alexander Potapenko Cc: Andrey Konovalov Cc: Dmitry Vyukov Cc: Geert Uytterhoeven Cc: Oliver Glitta Cc: Imran Khan Cc: Stephen Rothwell From: Vlastimil Babka Subject: lib/stackdepot: allow optional init and stack_table allocation by kvmalloc() - fixup4 Due to 4e66934eaadc ("lib: add reference counting tracking infrastructure") landing recently to net-next adding a new stack depot user in lib/ref_tracker.c we need to add an appropriate call to stack_depot_init() there as well. Link: https://lkml.kernel.org/r/45c1b738-1a2f-5b5f-2f6d-86fab206d01c@suse.cz Signed-off-by: Vlastimil Babka Reviewed-by: Eric Dumazet Cc: Jiri Slab Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- drivers/gpu/drm/drm_dp_mst_topology.c | 1 + drivers/gpu/drm/drm_mm.c | 4 ++++ drivers/gpu/drm/drm_modeset_lock.c | 9 +++++++++ drivers/gpu/drm/i915/intel_runtime_pm.c | 3 +++ include/linux/ref_tracker.h | 2 ++ include/linux/stackdepot.h | 25 ++++++++++++++++--------- init/main.c | 9 ++++++--- lib/Kconfig | 4 ++++ lib/Kconfig.kasan | 2 +- lib/stackdepot.c | 33 ++++++++++++++++++++++++++++----- mm/page_owner.c | 2 ++ 11 files changed, 76 insertions(+), 18 deletions(-) (limited to 'lib') diff --git a/drivers/gpu/drm/drm_dp_mst_topology.c b/drivers/gpu/drm/drm_dp_mst_topology.c index f3d79eda94bb..8b3822142fed 100644 --- a/drivers/gpu/drm/drm_dp_mst_topology.c +++ b/drivers/gpu/drm/drm_dp_mst_topology.c @@ -5511,6 +5511,7 @@ int drm_dp_mst_topology_mgr_init(struct drm_dp_mst_topology_mgr *mgr, mutex_init(&mgr->probe_lock); #if IS_ENABLED(CONFIG_DRM_DEBUG_DP_MST_TOPOLOGY_REFS) mutex_init(&mgr->topology_ref_history_lock); + stack_depot_init(); #endif INIT_LIST_HEAD(&mgr->tx_msg_downq); INIT_LIST_HEAD(&mgr->destroy_port_list); diff --git a/drivers/gpu/drm/drm_mm.c b/drivers/gpu/drm/drm_mm.c index 7d1c578388d3..8257f9d4f619 100644 --- a/drivers/gpu/drm/drm_mm.c +++ b/drivers/gpu/drm/drm_mm.c @@ -980,6 +980,10 @@ void drm_mm_init(struct drm_mm *mm, u64 start, u64 size) add_hole(&mm->head_node); mm->scan_active = 0; + +#ifdef CONFIG_DRM_DEBUG_MM + stack_depot_init(); +#endif } EXPORT_SYMBOL(drm_mm_init); diff --git a/drivers/gpu/drm/drm_modeset_lock.c b/drivers/gpu/drm/drm_modeset_lock.c index c97323365675..918065982db4 100644 --- a/drivers/gpu/drm/drm_modeset_lock.c +++ b/drivers/gpu/drm/drm_modeset_lock.c @@ -107,6 +107,11 @@ static void __drm_stack_depot_print(depot_stack_handle_t stack_depot) kfree(buf); } + +static void __drm_stack_depot_init(void) +{ + stack_depot_init(); +} #else /* CONFIG_DRM_DEBUG_MODESET_LOCK */ static depot_stack_handle_t __drm_stack_depot_save(void) { @@ -115,6 +120,9 @@ static depot_stack_handle_t __drm_stack_depot_save(void) static void __drm_stack_depot_print(depot_stack_handle_t stack_depot) { } +static void __drm_stack_depot_init(void) +{ +} #endif /* CONFIG_DRM_DEBUG_MODESET_LOCK */ /** @@ -359,6 +367,7 @@ void drm_modeset_lock_init(struct drm_modeset_lock *lock) { ww_mutex_init(&lock->mutex, &crtc_ww_class); INIT_LIST_HEAD(&lock->head); + __drm_stack_depot_init(); } EXPORT_SYMBOL(drm_modeset_lock_init); diff --git a/drivers/gpu/drm/i915/intel_runtime_pm.c b/drivers/gpu/drm/i915/intel_runtime_pm.c index 22dab36afcb6..53f1ccb78849 100644 --- a/drivers/gpu/drm/i915/intel_runtime_pm.c +++ b/drivers/gpu/drm/i915/intel_runtime_pm.c @@ -68,6 +68,9 @@ static noinline depot_stack_handle_t __save_depot_stack(void) static void init_intel_runtime_pm_wakeref(struct intel_runtime_pm *rpm) { spin_lock_init(&rpm->debug.lock); + + if (rpm->available) + stack_depot_init(); } static noinline depot_stack_handle_t diff --git a/include/linux/ref_tracker.h b/include/linux/ref_tracker.h index c11c9db5825c..60f3453be23e 100644 --- a/include/linux/ref_tracker.h +++ b/include/linux/ref_tracker.h @@ -4,6 +4,7 @@ #include #include #include +#include struct ref_tracker; @@ -26,6 +27,7 @@ static inline void ref_tracker_dir_init(struct ref_tracker_dir *dir, spin_lock_init(&dir->lock); dir->quarantine_avail = quarantine_count; refcount_set(&dir->untracked, 1); + stack_depot_init(); } void ref_tracker_dir_exit(struct ref_tracker_dir *dir); diff --git a/include/linux/stackdepot.h b/include/linux/stackdepot.h index c34b55a6e554..17f992fe6355 100644 --- a/include/linux/stackdepot.h +++ b/include/linux/stackdepot.h @@ -19,6 +19,22 @@ depot_stack_handle_t __stack_depot_save(unsigned long *entries, unsigned int nr_entries, gfp_t gfp_flags, bool can_alloc); +/* + * Every user of stack depot has to call this during its own init when it's + * decided that it will be calling stack_depot_save() later. + * + * The alternative is to select STACKDEPOT_ALWAYS_INIT to have stack depot + * enabled as part of mm_init(), for subsystems where it's known at compile time + * that stack depot will be used. + */ +int stack_depot_init(void); + +#ifdef CONFIG_STACKDEPOT_ALWAYS_INIT +static inline int stack_depot_early_init(void) { return stack_depot_init(); } +#else +static inline int stack_depot_early_init(void) { return 0; } +#endif + depot_stack_handle_t stack_depot_save(unsigned long *entries, unsigned int nr_entries, gfp_t gfp_flags); @@ -30,13 +46,4 @@ int stack_depot_snprint(depot_stack_handle_t handle, char *buf, size_t size, void stack_depot_print(depot_stack_handle_t stack); -#ifdef CONFIG_STACKDEPOT -int stack_depot_init(void); -#else -static inline int stack_depot_init(void) -{ - return 0; -} -#endif /* CONFIG_STACKDEPOT */ - #endif diff --git a/init/main.c b/init/main.c index bb984ed79de0..65fa2e41a9c0 100644 --- a/init/main.c +++ b/init/main.c @@ -834,12 +834,15 @@ static void __init mm_init(void) init_mem_debugging_and_hardening(); kfence_alloc_pool(); report_meminit(); - stack_depot_init(); + stack_depot_early_init(); mem_init(); mem_init_print_info(); - /* page_owner must be initialized after buddy is ready */ - page_ext_init_flatmem_late(); kmem_cache_init(); + /* + * page_owner must be initialized after buddy is ready, and also after + * slab is ready so that stack_depot_init() works properly + */ + page_ext_init_flatmem_late(); kmemleak_init(); pgtable_init(); debug_objects_mem_init(); diff --git a/lib/Kconfig b/lib/Kconfig index c20b68ad2bc3..51c368a50b16 100644 --- a/lib/Kconfig +++ b/lib/Kconfig @@ -673,6 +673,10 @@ config STACKDEPOT bool select STACKTRACE +config STACKDEPOT_ALWAYS_INIT + bool + select STACKDEPOT + config STACK_HASH_ORDER int "stack depot hash size (12 => 4KB, 20 => 1024KB)" range 12 20 diff --git a/lib/Kconfig.kasan b/lib/Kconfig.kasan index cdc842d090db..879757b6dd14 100644 --- a/lib/Kconfig.kasan +++ b/lib/Kconfig.kasan @@ -38,7 +38,7 @@ menuconfig KASAN CC_HAS_WORKING_NOSANITIZE_ADDRESS) || \ HAVE_ARCH_KASAN_HW_TAGS depends on (SLUB && SYSFS) || (SLAB && !DEBUG_SLAB) - select STACKDEPOT + select STACKDEPOT_ALWAYS_INIT help Enables KASAN (KernelAddressSANitizer) - runtime memory debugger, designed to find out-of-bounds accesses and use-after-free bugs. diff --git a/lib/stackdepot.c b/lib/stackdepot.c index b437ae79aca1..00ccb106f1a8 100644 --- a/lib/stackdepot.c +++ b/lib/stackdepot.c @@ -23,6 +23,7 @@ #include #include #include +#include #include #include #include @@ -161,18 +162,40 @@ static int __init is_stack_depot_disabled(char *str) } early_param("stack_depot_disable", is_stack_depot_disabled); -int __init stack_depot_init(void) +/* + * __ref because of memblock_alloc(), which will not be actually called after + * the __init code is gone, because at that point slab_is_available() is true + */ +__ref int stack_depot_init(void) { - if (!stack_depot_disable) { + static DEFINE_MUTEX(stack_depot_init_mutex); + + mutex_lock(&stack_depot_init_mutex); + if (!stack_depot_disable && !stack_table) { size_t size = (STACK_HASH_SIZE * sizeof(struct stack_record *)); int i; - stack_table = memblock_alloc(size, size); - for (i = 0; i < STACK_HASH_SIZE; i++) - stack_table[i] = NULL; + if (slab_is_available()) { + pr_info("Stack Depot allocating hash table with kvmalloc\n"); + stack_table = kvmalloc(size, GFP_KERNEL); + } else { + pr_info("Stack Depot allocating hash table with memblock_alloc\n"); + stack_table = memblock_alloc(size, SMP_CACHE_BYTES); + } + if (stack_table) { + for (i = 0; i < STACK_HASH_SIZE; i++) + stack_table[i] = NULL; + } else { + pr_err("Stack Depot hash table allocation failed, disabling\n"); + stack_depot_disable = true; + mutex_unlock(&stack_depot_init_mutex); + return -ENOMEM; + } } + mutex_unlock(&stack_depot_init_mutex); return 0; } +EXPORT_SYMBOL_GPL(stack_depot_init); /* Calculate hash for a stack */ static inline u32 hash_stack(unsigned long *entries, unsigned int size) diff --git a/mm/page_owner.c b/mm/page_owner.c index 5eea061bb1e5..99e360df9465 100644 --- a/mm/page_owner.c +++ b/mm/page_owner.c @@ -80,6 +80,8 @@ static __init void init_page_owner(void) if (!page_owner_enabled) return; + stack_depot_init(); + register_dummy_stack(); register_failure_stack(); register_early_stack(); -- cgit v1.2.3 From e940066089490efde86abc519593be84362f4e53 Mon Sep 17 00:00:00 2001 From: Marco Elver Date: Fri, 21 Jan 2022 22:14:31 -0800 Subject: lib/stackdepot: always do filter_irq_stacks() in stack_depot_save() The non-interrupt portion of interrupt stack traces before interrupt entry is usually arbitrary. Therefore, saving stack traces of interrupts (that include entries before interrupt entry) to stack depot leads to unbounded stackdepot growth. As such, use of filter_irq_stacks() is a requirement to ensure stackdepot can efficiently deduplicate interrupt stacks. Looking through all current users of stack_depot_save(), none (except KASAN) pass the stack trace through filter_irq_stacks() before passing it on to stack_depot_save(). Rather than adding filter_irq_stacks() to all current users of stack_depot_save(), it became clear that stack_depot_save() should simply do filter_irq_stacks(). Link: https://lkml.kernel.org/r/20211130095727.2378739-1-elver@google.com Signed-off-by: Marco Elver Reviewed-by: Alexander Potapenko Acked-by: Vlastimil Babka Reviewed-by: Andrey Konovalov Cc: Andrey Ryabinin Cc: Dmitry Vyukov Cc: Vijayanand Jitta Cc: "Gustavo A. R. Silva" Cc: Imran Khan Cc: Chris Wilson Cc: Jani Nikula Cc: Mika Kuoppala Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- lib/stackdepot.c | 13 +++++++++++++ mm/kasan/common.c | 1 - 2 files changed, 13 insertions(+), 1 deletion(-) (limited to 'lib') diff --git a/lib/stackdepot.c b/lib/stackdepot.c index 00ccb106f1a8..bf5ba9af0500 100644 --- a/lib/stackdepot.c +++ b/lib/stackdepot.c @@ -328,6 +328,9 @@ EXPORT_SYMBOL_GPL(stack_depot_fetch); * (allocates using GFP flags of @alloc_flags). If @can_alloc is %false, avoids * any allocations and will fail if no space is left to store the stack trace. * + * If the stack trace in @entries is from an interrupt, only the portion up to + * interrupt entry is saved. + * * Context: Any context, but setting @can_alloc to %false is required if * alloc_pages() cannot be used from the current context. Currently * this is the case from contexts where neither %GFP_ATOMIC nor @@ -346,6 +349,16 @@ depot_stack_handle_t __stack_depot_save(unsigned long *entries, unsigned long flags; u32 hash; + /* + * If this stack trace is from an interrupt, including anything before + * interrupt entry usually leads to unbounded stackdepot growth. + * + * Because use of filter_irq_stacks() is a requirement to ensure + * stackdepot can efficiently deduplicate interrupt stacks, always + * filter_irq_stacks() to simplify all callers' use of stackdepot. + */ + nr_entries = filter_irq_stacks(entries, nr_entries); + if (unlikely(nr_entries == 0) || stack_depot_disable) goto fast_exit; diff --git a/mm/kasan/common.c b/mm/kasan/common.c index 7c06db78a76c..92196562687b 100644 --- a/mm/kasan/common.c +++ b/mm/kasan/common.c @@ -36,7 +36,6 @@ depot_stack_handle_t kasan_save_stack(gfp_t flags, bool can_alloc) unsigned int nr_entries; nr_entries = stack_trace_save(entries, ARRAY_SIZE(entries), 0); - nr_entries = filter_irq_stacks(entries, nr_entries); return __stack_depot_save(entries, nr_entries, flags, can_alloc); } -- cgit v1.2.3 From 10825410b956dc1ed8c5fbc8bbedaffdadde7f20 Mon Sep 17 00:00:00 2001 From: Laibin Qiu Date: Thu, 27 Jan 2022 18:00:47 +0800 Subject: blk-mq: Fix wrong wakeup batch configuration which will cause hang Commit 180dccb0dba4f ("blk-mq: fix tag_get wait task can't be awakened") will recalculate wake_batch when incrementing or decrementing active_queues to avoid wake_batch > hctx_max_depth. At the same time, in order to not affect performance as much as possible, the minimum wakeup batch is set to 4. But when the QD is small (such as QD=1), if inc or dec active_queues increases wakeup batch, that can lead to a hang: Fix this problem with the following strategies: QD : >= 32 | < 32 --------------------------------- wakeup batch: 8~4 | 3~1 Fixes: 180dccb0dba4f ("blk-mq: fix tag_get wait task can't be awakened") Link: https://lore.kernel.org/linux-block/78cafe94-a787-e006-8851-69906f0c2128@huawei.com/T/#t Reported-by: Alex Xu (Hello71) Signed-off-by: Laibin Qiu Tested-by: Alex Xu (Hello71) Link: https://lore.kernel.org/r/20220127100047.1763746-1-qiulaibin@huawei.com Signed-off-by: Jens Axboe --- lib/sbitmap.c | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-) (limited to 'lib') diff --git a/lib/sbitmap.c b/lib/sbitmap.c index 6220fa67fb7e..09d293c30fd2 100644 --- a/lib/sbitmap.c +++ b/lib/sbitmap.c @@ -488,9 +488,13 @@ void sbitmap_queue_recalculate_wake_batch(struct sbitmap_queue *sbq, unsigned int users) { unsigned int wake_batch; + unsigned int min_batch; + unsigned int depth = (sbq->sb.depth + users - 1) / users; - wake_batch = clamp_val((sbq->sb.depth + users - 1) / - users, 4, SBQ_WAKE_BATCH); + min_batch = sbq->sb.depth >= (4 * SBQ_WAIT_QUEUES) ? 4 : 1; + + wake_batch = clamp_val(depth / SBQ_WAIT_QUEUES, + min_batch, SBQ_WAKE_BATCH); __sbitmap_queue_update_wake_batch(sbq, wake_batch); } EXPORT_SYMBOL_GPL(sbitmap_queue_recalculate_wake_batch); -- cgit v1.2.3 From 09c6304e38e440b93a9ebf3f3cf75cd6cb529f91 Mon Sep 17 00:00:00 2001 From: Marco Elver Date: Sat, 29 Jan 2022 13:41:11 -0800 Subject: kasan: test: fix compatibility with FORTIFY_SOURCE With CONFIG_FORTIFY_SOURCE enabled, string functions will also perform dynamic checks using __builtin_object_size(ptr), which when failed will panic the kernel. Because the KASAN test deliberately performs out-of-bounds operations, the kernel panics with FORTIFY_SOURCE, for example: | kernel BUG at lib/string_helpers.c:910! | invalid opcode: 0000 [#1] PREEMPT SMP KASAN PTI | CPU: 1 PID: 137 Comm: kunit_try_catch Tainted: G B 5.16.0-rc3+ #3 | Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.14.0-2 04/01/2014 | RIP: 0010:fortify_panic+0x19/0x1b | ... | Call Trace: | kmalloc_oob_in_memset.cold+0x16/0x16 | ... Fix it by also hiding `ptr` from the optimizer, which will ensure that __builtin_object_size() does not return a valid size, preventing fortified string functions from panicking. Link: https://lkml.kernel.org/r/20220124160744.1244685-1-elver@google.com Signed-off-by: Marco Elver Reported-by: Nico Pache Reviewed-by: Nico Pache Reviewed-by: Andrey Konovalov Reviewed-by: Kees Cook Cc: Andrey Ryabinin Cc: Alexander Potapenko Cc: Dmitry Vyukov Cc: Brendan Higgins Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- lib/test_kasan.c | 5 +++++ 1 file changed, 5 insertions(+) (limited to 'lib') diff --git a/lib/test_kasan.c b/lib/test_kasan.c index 847cdbefab46..26a5c9007653 100644 --- a/lib/test_kasan.c +++ b/lib/test_kasan.c @@ -492,6 +492,7 @@ static void kmalloc_oob_in_memset(struct kunit *test) ptr = kmalloc(size, GFP_KERNEL); KUNIT_ASSERT_NOT_ERR_OR_NULL(test, ptr); + OPTIMIZER_HIDE_VAR(ptr); OPTIMIZER_HIDE_VAR(size); KUNIT_EXPECT_KASAN_FAIL(test, memset(ptr, 0, size + KASAN_GRANULE_SIZE)); @@ -515,6 +516,7 @@ static void kmalloc_memmove_negative_size(struct kunit *test) KUNIT_ASSERT_NOT_ERR_OR_NULL(test, ptr); memset((char *)ptr, 0, 64); + OPTIMIZER_HIDE_VAR(ptr); OPTIMIZER_HIDE_VAR(invalid_size); KUNIT_EXPECT_KASAN_FAIL(test, memmove((char *)ptr, (char *)ptr + 4, invalid_size)); @@ -531,6 +533,7 @@ static void kmalloc_memmove_invalid_size(struct kunit *test) KUNIT_ASSERT_NOT_ERR_OR_NULL(test, ptr); memset((char *)ptr, 0, 64); + OPTIMIZER_HIDE_VAR(ptr); KUNIT_EXPECT_KASAN_FAIL(test, memmove((char *)ptr, (char *)ptr + 4, invalid_size)); kfree(ptr); @@ -893,6 +896,7 @@ static void kasan_memchr(struct kunit *test) ptr = kmalloc(size, GFP_KERNEL | __GFP_ZERO); KUNIT_ASSERT_NOT_ERR_OR_NULL(test, ptr); + OPTIMIZER_HIDE_VAR(ptr); OPTIMIZER_HIDE_VAR(size); KUNIT_EXPECT_KASAN_FAIL(test, kasan_ptr_result = memchr(ptr, '1', size + 1)); @@ -919,6 +923,7 @@ static void kasan_memcmp(struct kunit *test) KUNIT_ASSERT_NOT_ERR_OR_NULL(test, ptr); memset(arr, 0, sizeof(arr)); + OPTIMIZER_HIDE_VAR(ptr); OPTIMIZER_HIDE_VAR(size); KUNIT_EXPECT_KASAN_FAIL(test, kasan_int_result = memcmp(ptr, arr, size+1)); -- cgit v1.2.3 From d2a02e3c8bb6b347818518edff5a4b40ff52d6d8 Mon Sep 17 00:00:00 2001 From: "Jason A. Donenfeld" Date: Wed, 19 Jan 2022 14:35:06 +0100 Subject: lib/crypto: blake2s: avoid indirect calls to compression function for Clang CFI blake2s_compress_generic is weakly aliased by blake2s_compress. The current harness for function selection uses a function pointer, which is ordinarily inlined and resolved at compile time. But when Clang's CFI is enabled, CFI still triggers when making an indirect call via a weak symbol. This seems like a bug in Clang's CFI, as though it's bucketing weak symbols and strong symbols differently. It also only seems to trigger when "full LTO" mode is used, rather than "thin LTO". [ 0.000000][ T0] Kernel panic - not syncing: CFI failure (target: blake2s_compress_generic+0x0/0x1444) [ 0.000000][ T0] CPU: 0 PID: 0 Comm: swapper/0 Not tainted 5.16.0-mainline-06981-g076c855b846e #1 [ 0.000000][ T0] Hardware name: MT6873 (DT) [ 0.000000][ T0] Call trace: [ 0.000000][ T0] dump_backtrace+0xfc/0x1dc [ 0.000000][ T0] dump_stack_lvl+0xa8/0x11c [ 0.000000][ T0] panic+0x194/0x464 [ 0.000000][ T0] __cfi_check_fail+0x54/0x58 [ 0.000000][ T0] __cfi_slowpath_diag+0x354/0x4b0 [ 0.000000][ T0] blake2s_update+0x14c/0x178 [ 0.000000][ T0] _extract_entropy+0xf4/0x29c [ 0.000000][ T0] crng_initialize_primary+0x24/0x94 [ 0.000000][ T0] rand_initialize+0x2c/0x6c [ 0.000000][ T0] start_kernel+0x2f8/0x65c [ 0.000000][ T0] __primary_switched+0xc4/0x7be4 [ 0.000000][ T0] Rebooting in 5 seconds.. Nonetheless, the function pointer method isn't so terrific anyway, so this patch replaces it with a simple boolean, which also gets inlined away. This successfully works around the Clang bug. In general, I'm not too keen on all of the indirection involved here; it clearly does more harm than good. Hopefully the whole thing can get cleaned up down the road when lib/crypto is overhauled more comprehensively. But for now, we go with a simple bandaid. Fixes: 6048fdcc5f26 ("lib/crypto: blake2s: include as built-in") Link: https://github.com/ClangBuiltLinux/linux/issues/1567 Reported-by: Miles Chen Tested-by: Miles Chen Tested-by: Nathan Chancellor Tested-by: John Stultz Acked-by: Nick Desaulniers Reviewed-by: Eric Biggers Signed-off-by: Jason A. Donenfeld --- arch/arm/crypto/blake2s-shash.c | 4 ++-- arch/x86/crypto/blake2s-shash.c | 4 ++-- crypto/blake2s_generic.c | 4 ++-- include/crypto/internal/blake2s.h | 40 ++++++++++++++++++++++++--------------- lib/crypto/blake2s.c | 4 ++-- 5 files changed, 33 insertions(+), 23 deletions(-) (limited to 'lib') diff --git a/arch/arm/crypto/blake2s-shash.c b/arch/arm/crypto/blake2s-shash.c index 17c1c3bfe2f5..763c73beea2d 100644 --- a/arch/arm/crypto/blake2s-shash.c +++ b/arch/arm/crypto/blake2s-shash.c @@ -13,12 +13,12 @@ static int crypto_blake2s_update_arm(struct shash_desc *desc, const u8 *in, unsigned int inlen) { - return crypto_blake2s_update(desc, in, inlen, blake2s_compress); + return crypto_blake2s_update(desc, in, inlen, false); } static int crypto_blake2s_final_arm(struct shash_desc *desc, u8 *out) { - return crypto_blake2s_final(desc, out, blake2s_compress); + return crypto_blake2s_final(desc, out, false); } #define BLAKE2S_ALG(name, driver_name, digest_size) \ diff --git a/arch/x86/crypto/blake2s-shash.c b/arch/x86/crypto/blake2s-shash.c index f9e2fecdb761..59ae28abe35c 100644 --- a/arch/x86/crypto/blake2s-shash.c +++ b/arch/x86/crypto/blake2s-shash.c @@ -18,12 +18,12 @@ static int crypto_blake2s_update_x86(struct shash_desc *desc, const u8 *in, unsigned int inlen) { - return crypto_blake2s_update(desc, in, inlen, blake2s_compress); + return crypto_blake2s_update(desc, in, inlen, false); } static int crypto_blake2s_final_x86(struct shash_desc *desc, u8 *out) { - return crypto_blake2s_final(desc, out, blake2s_compress); + return crypto_blake2s_final(desc, out, false); } #define BLAKE2S_ALG(name, driver_name, digest_size) \ diff --git a/crypto/blake2s_generic.c b/crypto/blake2s_generic.c index 72fe480f9bd6..5f96a21f8788 100644 --- a/crypto/blake2s_generic.c +++ b/crypto/blake2s_generic.c @@ -15,12 +15,12 @@ static int crypto_blake2s_update_generic(struct shash_desc *desc, const u8 *in, unsigned int inlen) { - return crypto_blake2s_update(desc, in, inlen, blake2s_compress_generic); + return crypto_blake2s_update(desc, in, inlen, true); } static int crypto_blake2s_final_generic(struct shash_desc *desc, u8 *out) { - return crypto_blake2s_final(desc, out, blake2s_compress_generic); + return crypto_blake2s_final(desc, out, true); } #define BLAKE2S_ALG(name, driver_name, digest_size) \ diff --git a/include/crypto/internal/blake2s.h b/include/crypto/internal/blake2s.h index d39cfa0d333e..52363eee2b20 100644 --- a/include/crypto/internal/blake2s.h +++ b/include/crypto/internal/blake2s.h @@ -24,14 +24,11 @@ static inline void blake2s_set_lastblock(struct blake2s_state *state) state->f[0] = -1; } -typedef void (*blake2s_compress_t)(struct blake2s_state *state, - const u8 *block, size_t nblocks, u32 inc); - /* Helper functions for BLAKE2s shared by the library and shash APIs */ -static inline void __blake2s_update(struct blake2s_state *state, - const u8 *in, size_t inlen, - blake2s_compress_t compress) +static __always_inline void +__blake2s_update(struct blake2s_state *state, const u8 *in, size_t inlen, + bool force_generic) { const size_t fill = BLAKE2S_BLOCK_SIZE - state->buflen; @@ -39,7 +36,12 @@ static inline void __blake2s_update(struct blake2s_state *state, return; if (inlen > fill) { memcpy(state->buf + state->buflen, in, fill); - (*compress)(state, state->buf, 1, BLAKE2S_BLOCK_SIZE); + if (force_generic) + blake2s_compress_generic(state, state->buf, 1, + BLAKE2S_BLOCK_SIZE); + else + blake2s_compress(state, state->buf, 1, + BLAKE2S_BLOCK_SIZE); state->buflen = 0; in += fill; inlen -= fill; @@ -47,7 +49,12 @@ static inline void __blake2s_update(struct blake2s_state *state, if (inlen > BLAKE2S_BLOCK_SIZE) { const size_t nblocks = DIV_ROUND_UP(inlen, BLAKE2S_BLOCK_SIZE); /* Hash one less (full) block than strictly possible */ - (*compress)(state, in, nblocks - 1, BLAKE2S_BLOCK_SIZE); + if (force_generic) + blake2s_compress_generic(state, in, nblocks - 1, + BLAKE2S_BLOCK_SIZE); + else + blake2s_compress(state, in, nblocks - 1, + BLAKE2S_BLOCK_SIZE); in += BLAKE2S_BLOCK_SIZE * (nblocks - 1); inlen -= BLAKE2S_BLOCK_SIZE * (nblocks - 1); } @@ -55,13 +62,16 @@ static inline void __blake2s_update(struct blake2s_state *state, state->buflen += inlen; } -static inline void __blake2s_final(struct blake2s_state *state, u8 *out, - blake2s_compress_t compress) +static __always_inline void +__blake2s_final(struct blake2s_state *state, u8 *out, bool force_generic) { blake2s_set_lastblock(state); memset(state->buf + state->buflen, 0, BLAKE2S_BLOCK_SIZE - state->buflen); /* Padding */ - (*compress)(state, state->buf, 1, state->buflen); + if (force_generic) + blake2s_compress_generic(state, state->buf, 1, state->buflen); + else + blake2s_compress(state, state->buf, 1, state->buflen); cpu_to_le32_array(state->h, ARRAY_SIZE(state->h)); memcpy(out, state->h, state->outlen); } @@ -99,20 +109,20 @@ static inline int crypto_blake2s_init(struct shash_desc *desc) static inline int crypto_blake2s_update(struct shash_desc *desc, const u8 *in, unsigned int inlen, - blake2s_compress_t compress) + bool force_generic) { struct blake2s_state *state = shash_desc_ctx(desc); - __blake2s_update(state, in, inlen, compress); + __blake2s_update(state, in, inlen, force_generic); return 0; } static inline int crypto_blake2s_final(struct shash_desc *desc, u8 *out, - blake2s_compress_t compress) + bool force_generic) { struct blake2s_state *state = shash_desc_ctx(desc); - __blake2s_final(state, out, compress); + __blake2s_final(state, out, force_generic); return 0; } diff --git a/lib/crypto/blake2s.c b/lib/crypto/blake2s.c index 9364f79937b8..c71c09621c09 100644 --- a/lib/crypto/blake2s.c +++ b/lib/crypto/blake2s.c @@ -18,14 +18,14 @@ void blake2s_update(struct blake2s_state *state, const u8 *in, size_t inlen) { - __blake2s_update(state, in, inlen, blake2s_compress); + __blake2s_update(state, in, inlen, false); } EXPORT_SYMBOL(blake2s_update); void blake2s_final(struct blake2s_state *state, u8 *out) { WARN_ON(IS_ENABLED(DEBUG) && !out); - __blake2s_final(state, out, blake2s_compress); + __blake2s_final(state, out, false); memzero_explicit(state, sizeof(*state)); } EXPORT_SYMBOL(blake2s_final); -- cgit v1.2.3