diff options
Diffstat (limited to 'arch')
84 files changed, 1283 insertions, 433 deletions
diff --git a/arch/Kconfig b/arch/Kconfig index a62965d057f6..550cce4dd648 100644 --- a/arch/Kconfig +++ b/arch/Kconfig @@ -313,4 +313,7 @@ config HAVE_IRQ_TIME_ACCOUNTING Archs need to ensure they use a high enough resolution clock to support irq time accounting and then call enable_sched_clock_irqtime(). +config HAVE_ARCH_TRANSPARENT_HUGEPAGE + bool + source "kernel/gcov/Kconfig" diff --git a/arch/alpha/kernel/pci-sysfs.c b/arch/alpha/kernel/pci-sysfs.c index 53649c7d0068..b51f7b4818cd 100644 --- a/arch/alpha/kernel/pci-sysfs.c +++ b/arch/alpha/kernel/pci-sysfs.c @@ -26,7 +26,7 @@ static int hose_mmap_page_range(struct pci_controller *hose, base = sparse ? hose->sparse_io_base : hose->dense_io_base; vma->vm_pgoff += base >> PAGE_SHIFT; - vma->vm_flags |= (VM_IO | VM_RESERVED); + vma->vm_flags |= VM_IO | VM_DONTEXPAND | VM_DONTDUMP; return io_remap_pfn_range(vma, vma->vm_start, vma->vm_pgoff, vma->vm_end - vma->vm_start, diff --git a/arch/arm/Kconfig b/arch/arm/Kconfig index 6d2f7f5c0036..2867a7742306 100644 --- a/arch/arm/Kconfig +++ b/arch/arm/Kconfig @@ -25,6 +25,7 @@ config ARM select HAVE_FUNCTION_GRAPH_TRACER if (!THUMB2_KERNEL) select ARCH_BINFMT_ELF_RANDOMIZE_PIE select HAVE_GENERIC_DMA_COHERENT + select HAVE_DEBUG_KMEMLEAK select HAVE_KERNEL_GZIP select HAVE_KERNEL_LZO select HAVE_KERNEL_LZMA @@ -39,6 +40,7 @@ config ARM select HARDIRQS_SW_RESEND select GENERIC_IRQ_PROBE select GENERIC_IRQ_SHOW + select HAVE_UID16 select ARCH_WANT_IPC_PARSE_VERSION select HARDIRQS_SW_RESEND select CPU_PM if (SUSPEND || CPU_IDLE) diff --git a/arch/arm/mm/fault-armv.c b/arch/arm/mm/fault-armv.c index 7599e2625c7d..2a5907b5c8d2 100644 --- a/arch/arm/mm/fault-armv.c +++ b/arch/arm/mm/fault-armv.c @@ -134,7 +134,6 @@ make_coherent(struct address_space *mapping, struct vm_area_struct *vma, { struct mm_struct *mm = vma->vm_mm; struct vm_area_struct *mpnt; - struct prio_tree_iter iter; unsigned long offset; pgoff_t pgoff; int aliases = 0; @@ -147,7 +146,7 @@ make_coherent(struct address_space *mapping, struct vm_area_struct *vma, * cache coherency. */ flush_dcache_mmap_lock(mapping); - vma_prio_tree_foreach(mpnt, &iter, &mapping->i_mmap, pgoff, pgoff) { + vma_interval_tree_foreach(mpnt, &mapping->i_mmap, pgoff, pgoff) { /* * If this VMA is not in our MM, we can ignore it. * Note that we intentionally mask out the VMA diff --git a/arch/arm/mm/fault.c b/arch/arm/mm/fault.c index c3bd83450227..5dbf13f954f6 100644 --- a/arch/arm/mm/fault.c +++ b/arch/arm/mm/fault.c @@ -336,6 +336,7 @@ retry: /* Clear FAULT_FLAG_ALLOW_RETRY to avoid any risk * of starvation. */ flags &= ~FAULT_FLAG_ALLOW_RETRY; + flags |= FAULT_FLAG_TRIED; goto retry; } } diff --git a/arch/arm/mm/flush.c b/arch/arm/mm/flush.c index 40ca11ed6e5f..1c8f7f564175 100644 --- a/arch/arm/mm/flush.c +++ b/arch/arm/mm/flush.c @@ -196,7 +196,6 @@ static void __flush_dcache_aliases(struct address_space *mapping, struct page *p { struct mm_struct *mm = current->active_mm; struct vm_area_struct *mpnt; - struct prio_tree_iter iter; pgoff_t pgoff; /* @@ -208,7 +207,7 @@ static void __flush_dcache_aliases(struct address_space *mapping, struct page *p pgoff = page->index << (PAGE_CACHE_SHIFT - PAGE_SHIFT); flush_dcache_mmap_lock(mapping); - vma_prio_tree_foreach(mpnt, &iter, &mapping->i_mmap, pgoff, pgoff) { + vma_interval_tree_foreach(mpnt, &mapping->i_mmap, pgoff, pgoff) { unsigned long offset; /* diff --git a/arch/arm64/Kconfig b/arch/arm64/Kconfig index 767ba5685454..7ff68c946073 100644 --- a/arch/arm64/Kconfig +++ b/arch/arm64/Kconfig @@ -10,6 +10,8 @@ config ARM64 select GENERIC_TIME_VSYSCALL select HARDIRQS_SW_RESEND select HAVE_ARCH_TRACEHOOK + select HAVE_DEBUG_BUGVERBOSE + select HAVE_DEBUG_KMEMLEAK select HAVE_DMA_API_DEBUG select HAVE_DMA_ATTRS select HAVE_GENERIC_DMA_COHERENT @@ -26,6 +28,7 @@ config ARM64 select PERF_USE_VMALLOC select RTC_LIB select SPARSE_IRQ + select SYSCTL_EXCEPTION_TRACE help ARM 64-bit (AArch64) Linux support. @@ -193,6 +196,7 @@ config COMPAT bool "Kernel support for 32-bit EL0" depends on !ARM64_64K_PAGES select COMPAT_BINFMT_ELF + select HAVE_UID16 help This option enables support for a 32-bit EL0 running under a 64-bit kernel at EL1. AArch32-specific components such as system calls, diff --git a/arch/avr32/mm/fault.c b/arch/avr32/mm/fault.c index b92e60958617..b2f2d2d66849 100644 --- a/arch/avr32/mm/fault.c +++ b/arch/avr32/mm/fault.c @@ -152,6 +152,7 @@ good_area: tsk->min_flt++; if (fault & VM_FAULT_RETRY) { flags &= ~FAULT_FLAG_ALLOW_RETRY; + flags |= FAULT_FLAG_TRIED; /* * No need to up_read(&mm->mmap_sem) as we would have diff --git a/arch/blackfin/Kconfig b/arch/blackfin/Kconfig index 99224c4eb86b..ccd9193932b2 100644 --- a/arch/blackfin/Kconfig +++ b/arch/blackfin/Kconfig @@ -33,6 +33,7 @@ config BLACKFIN select HAVE_PERF_EVENTS select ARCH_HAVE_CUSTOM_GPIO_H select ARCH_WANT_OPTIONAL_GPIOLIB + select HAVE_UID16 select ARCH_WANT_IPC_PARSE_VERSION select HAVE_GENERIC_HARDIRQS select GENERIC_ATOMIC64 diff --git a/arch/cris/Kconfig b/arch/cris/Kconfig index 72bd5ae50a89..a118163b04ee 100644 --- a/arch/cris/Kconfig +++ b/arch/cris/Kconfig @@ -42,6 +42,7 @@ config CRIS select HAVE_IDE select GENERIC_ATOMIC64 select HAVE_GENERIC_HARDIRQS + select HAVE_UID16 select ARCH_WANT_IPC_PARSE_VERSION select GENERIC_IRQ_SHOW select GENERIC_IOMAP diff --git a/arch/cris/mm/fault.c b/arch/cris/mm/fault.c index 45fd542cf173..73312ab6c696 100644 --- a/arch/cris/mm/fault.c +++ b/arch/cris/mm/fault.c @@ -186,6 +186,7 @@ retry: tsk->min_flt++; if (fault & VM_FAULT_RETRY) { flags &= ~FAULT_FLAG_ALLOW_RETRY; + flags |= FAULT_FLAG_TRIED; /* * No need to up_read(&mm->mmap_sem) as we would diff --git a/arch/frv/Kconfig b/arch/frv/Kconfig index 971c0a19facb..9d262645f667 100644 --- a/arch/frv/Kconfig +++ b/arch/frv/Kconfig @@ -5,8 +5,10 @@ config FRV select HAVE_ARCH_TRACEHOOK select HAVE_IRQ_WORK select HAVE_PERF_EVENTS + select HAVE_UID16 select HAVE_GENERIC_HARDIRQS select GENERIC_IRQ_SHOW + select HAVE_DEBUG_BUGVERBOSE select ARCH_HAVE_NMI_SAFE_CMPXCHG select GENERIC_CPU_DEVICES select ARCH_WANT_IPC_PARSE_VERSION diff --git a/arch/h8300/Kconfig b/arch/h8300/Kconfig index 5e8a0d9a09ce..90462eb23d02 100644 --- a/arch/h8300/Kconfig +++ b/arch/h8300/Kconfig @@ -3,6 +3,7 @@ config H8300 default y select HAVE_IDE select HAVE_GENERIC_HARDIRQS + select HAVE_UID16 select ARCH_WANT_IPC_PARSE_VERSION select GENERIC_IRQ_SHOW select GENERIC_CPU_DEVICES diff --git a/arch/hexagon/mm/vm_fault.c b/arch/hexagon/mm/vm_fault.c index 06695cc4fe58..513b74cb397e 100644 --- a/arch/hexagon/mm/vm_fault.c +++ b/arch/hexagon/mm/vm_fault.c @@ -113,6 +113,7 @@ good_area: current->min_flt++; if (fault & VM_FAULT_RETRY) { flags &= ~FAULT_FLAG_ALLOW_RETRY; + flags |= FAULT_FLAG_TRIED; goto retry; } } diff --git a/arch/ia64/include/asm/hugetlb.h b/arch/ia64/include/asm/hugetlb.h index da55c63728e0..94eaa5bd5d0c 100644 --- a/arch/ia64/include/asm/hugetlb.h +++ b/arch/ia64/include/asm/hugetlb.h @@ -77,4 +77,8 @@ static inline void arch_release_hugepage(struct page *page) { } +static inline void arch_clear_hugepage_flags(struct page *page) +{ +} + #endif /* _ASM_IA64_HUGETLB_H */ diff --git a/arch/ia64/kernel/perfmon.c b/arch/ia64/kernel/perfmon.c index f388b4e18a37..ea39eba61ef5 100644 --- a/arch/ia64/kernel/perfmon.c +++ b/arch/ia64/kernel/perfmon.c @@ -2307,7 +2307,7 @@ pfm_smpl_buffer_alloc(struct task_struct *task, struct file *filp, pfm_context_t */ vma->vm_mm = mm; vma->vm_file = get_file(filp); - vma->vm_flags = VM_READ| VM_MAYREAD |VM_RESERVED; + vma->vm_flags = VM_READ|VM_MAYREAD|VM_DONTEXPAND|VM_DONTDUMP; vma->vm_page_prot = PAGE_READONLY; /* XXX may need to change */ /* diff --git a/arch/ia64/mm/fault.c b/arch/ia64/mm/fault.c index 8443daf4f515..6cf0341f978e 100644 --- a/arch/ia64/mm/fault.c +++ b/arch/ia64/mm/fault.c @@ -184,6 +184,7 @@ retry: current->min_flt++; if (fault & VM_FAULT_RETRY) { flags &= ~FAULT_FLAG_ALLOW_RETRY; + flags |= FAULT_FLAG_TRIED; /* No need to up_read(&mm->mmap_sem) as we would * have already released it in __lock_page_or_retry diff --git a/arch/ia64/mm/init.c b/arch/ia64/mm/init.c index 0eab454867a2..acd5b68e8871 100644 --- a/arch/ia64/mm/init.c +++ b/arch/ia64/mm/init.c @@ -138,7 +138,8 @@ ia64_init_addr_space (void) vma->vm_mm = current->mm; vma->vm_end = PAGE_SIZE; vma->vm_page_prot = __pgprot(pgprot_val(PAGE_READONLY) | _PAGE_MA_NAT); - vma->vm_flags = VM_READ | VM_MAYREAD | VM_IO | VM_RESERVED; + vma->vm_flags = VM_READ | VM_MAYREAD | VM_IO | + VM_DONTEXPAND | VM_DONTDUMP; down_write(¤t->mm->mmap_sem); if (insert_vm_struct(current->mm, vma)) { up_write(¤t->mm->mmap_sem); @@ -636,6 +637,7 @@ mem_init (void) high_memory = __va(max_low_pfn * PAGE_SIZE); + reset_zone_present_pages(); for_each_online_pgdat(pgdat) if (pgdat->bdata->node_bootmem_map) totalram_pages += free_all_bootmem_node(pgdat); diff --git a/arch/m32r/Kconfig b/arch/m32r/Kconfig index 49498bbb9616..e875fc3ce9cb 100644 --- a/arch/m32r/Kconfig +++ b/arch/m32r/Kconfig @@ -8,6 +8,7 @@ config M32R select HAVE_KERNEL_BZIP2 select HAVE_KERNEL_LZMA select ARCH_WANT_IPC_PARSE_VERSION + select HAVE_DEBUG_BUGVERBOSE select HAVE_GENERIC_HARDIRQS select GENERIC_IRQ_PROBE select GENERIC_IRQ_SHOW diff --git a/arch/m68k/Kconfig b/arch/m68k/Kconfig index b22df9410dce..dae1e7e16a37 100644 --- a/arch/m68k/Kconfig +++ b/arch/m68k/Kconfig @@ -3,9 +3,11 @@ config M68K default y select HAVE_IDE select HAVE_AOUT if MMU + select HAVE_DEBUG_BUGVERBOSE select HAVE_GENERIC_HARDIRQS select GENERIC_IRQ_SHOW select GENERIC_ATOMIC64 + select HAVE_UID16 select ARCH_HAVE_NMI_SAFE_CMPXCHG if RMW_INSNS select GENERIC_CPU_DEVICES select GENERIC_STRNCPY_FROM_USER if MMU diff --git a/arch/m68k/mm/fault.c b/arch/m68k/mm/fault.c index aeebbb7b30f0..a563727806bf 100644 --- a/arch/m68k/mm/fault.c +++ b/arch/m68k/mm/fault.c @@ -170,6 +170,7 @@ good_area: /* Clear FAULT_FLAG_ALLOW_RETRY to avoid any risk * of starvation. */ flags &= ~FAULT_FLAG_ALLOW_RETRY; + flags |= FAULT_FLAG_TRIED; /* * No need to up_read(&mm->mmap_sem) as we would diff --git a/arch/microblaze/Kconfig b/arch/microblaze/Kconfig index 6133bed2b855..53fd94ab60f0 100644 --- a/arch/microblaze/Kconfig +++ b/arch/microblaze/Kconfig @@ -16,6 +16,7 @@ config MICROBLAZE select OF select OF_EARLY_FLATTREE select ARCH_WANT_IPC_PARSE_VERSION + select HAVE_DEBUG_KMEMLEAK select IRQ_DOMAIN select HAVE_GENERIC_HARDIRQS select GENERIC_IRQ_PROBE diff --git a/arch/microblaze/include/asm/atomic.h b/arch/microblaze/include/asm/atomic.h index 472d8bf726df..42ac382a09da 100644 --- a/arch/microblaze/include/asm/atomic.h +++ b/arch/microblaze/include/asm/atomic.h @@ -22,5 +22,6 @@ static inline int atomic_dec_if_positive(atomic_t *v) return res; } +#define atomic_dec_if_positive atomic_dec_if_positive #endif /* _ASM_MICROBLAZE_ATOMIC_H */ diff --git a/arch/microblaze/mm/fault.c b/arch/microblaze/mm/fault.c index eb365d6795fa..714b35a9c4f7 100644 --- a/arch/microblaze/mm/fault.c +++ b/arch/microblaze/mm/fault.c @@ -233,6 +233,7 @@ good_area: current->min_flt++; if (fault & VM_FAULT_RETRY) { flags &= ~FAULT_FLAG_ALLOW_RETRY; + flags |= FAULT_FLAG_TRIED; /* * No need to up_read(&mm->mmap_sem) as we would diff --git a/arch/mips/Kconfig b/arch/mips/Kconfig index 4cd538b42a3f..35453eaeffb5 100644 --- a/arch/mips/Kconfig +++ b/arch/mips/Kconfig @@ -17,6 +17,7 @@ config MIPS select HAVE_FUNCTION_GRAPH_TRACER select HAVE_KPROBES select HAVE_KRETPROBES + select HAVE_DEBUG_KMEMLEAK select ARCH_BINFMT_ELF_RANDOMIZE_PIE select RTC_LIB if !MACH_LOONGSON select GENERIC_ATOMIC64 if !64BIT diff --git a/arch/mips/include/asm/hugetlb.h b/arch/mips/include/asm/hugetlb.h index 58d36889f09b..bd94946a18f3 100644 --- a/arch/mips/include/asm/hugetlb.h +++ b/arch/mips/include/asm/hugetlb.h @@ -112,4 +112,8 @@ static inline void arch_release_hugepage(struct page *page) { } +static inline void arch_clear_hugepage_flags(struct page *page) +{ +} + #endif /* __ASM_HUGETLB_H */ diff --git a/arch/mips/mm/fault.c b/arch/mips/mm/fault.c index 7a19957735e9..ddcec1e1a0cd 100644 --- a/arch/mips/mm/fault.c +++ b/arch/mips/mm/fault.c @@ -171,6 +171,7 @@ good_area: } if (fault & VM_FAULT_RETRY) { flags &= ~FAULT_FLAG_ALLOW_RETRY; + flags |= FAULT_FLAG_TRIED; /* * No need to up_read(&mm->mmap_sem) as we would diff --git a/arch/openrisc/mm/fault.c b/arch/openrisc/mm/fault.c index 40f850e9766c..e2bfafce66c5 100644 --- a/arch/openrisc/mm/fault.c +++ b/arch/openrisc/mm/fault.c @@ -183,6 +183,7 @@ good_area: tsk->min_flt++; if (fault & VM_FAULT_RETRY) { flags &= ~FAULT_FLAG_ALLOW_RETRY; + flags |= FAULT_FLAG_TRIED; /* No need to up_read(&mm->mmap_sem) as we would * have already released it in __lock_page_or_retry diff --git a/arch/parisc/kernel/cache.c b/arch/parisc/kernel/cache.c index 9d181890a7e3..48e16dc20102 100644 --- a/arch/parisc/kernel/cache.c +++ b/arch/parisc/kernel/cache.c @@ -276,7 +276,6 @@ void flush_dcache_page(struct page *page) { struct address_space *mapping = page_mapping(page); struct vm_area_struct *mpnt; - struct prio_tree_iter iter; unsigned long offset; unsigned long addr, old_addr = 0; pgoff_t pgoff; @@ -299,7 +298,7 @@ void flush_dcache_page(struct page *page) * to flush one address here for them all to become coherent */ flush_dcache_mmap_lock(mapping); - vma_prio_tree_foreach(mpnt, &iter, &mapping->i_mmap, pgoff, pgoff) { + vma_interval_tree_foreach(mpnt, &mapping->i_mmap, pgoff, pgoff) { offset = (pgoff - mpnt->vm_pgoff) << PAGE_SHIFT; addr = mpnt->vm_start + offset; diff --git a/arch/powerpc/Kconfig b/arch/powerpc/Kconfig index 4ce0be32d153..df7edb887a04 100644 --- a/arch/powerpc/Kconfig +++ b/arch/powerpc/Kconfig @@ -99,6 +99,7 @@ config PPC select HAVE_DYNAMIC_FTRACE select HAVE_FUNCTION_TRACER select HAVE_FUNCTION_GRAPH_TRACER + select SYSCTL_EXCEPTION_TRACE select ARCH_WANT_OPTIONAL_GPIOLIB select HAVE_IDE select HAVE_IOREMAP_PROT @@ -113,6 +114,7 @@ config PPC select HAVE_DMA_API_DEBUG select USE_GENERIC_SMP_HELPERS if SMP select HAVE_OPROFILE + select HAVE_DEBUG_KMEMLEAK select HAVE_SYSCALL_WRAPPERS if PPC64 select GENERIC_ATOMIC64 if PPC32 select ARCH_HAS_ATOMIC64_DEC_IF_POSITIVE diff --git a/arch/powerpc/include/asm/atomic.h b/arch/powerpc/include/asm/atomic.h index da29032ae38f..e3b1d41c89be 100644 --- a/arch/powerpc/include/asm/atomic.h +++ b/arch/powerpc/include/asm/atomic.h @@ -268,6 +268,7 @@ static __inline__ int atomic_dec_if_positive(atomic_t *v) return t; } +#define atomic_dec_if_positive atomic_dec_if_positive #define smp_mb__before_atomic_dec() smp_mb() #define smp_mb__after_atomic_dec() smp_mb() diff --git a/arch/powerpc/include/asm/hugetlb.h b/arch/powerpc/include/asm/hugetlb.h index dfdb95bc59a5..62e11a32c4c2 100644 --- a/arch/powerpc/include/asm/hugetlb.h +++ b/arch/powerpc/include/asm/hugetlb.h @@ -151,6 +151,10 @@ static inline void arch_release_hugepage(struct page *page) { } +static inline void arch_clear_hugepage_flags(struct page *page) +{ +} + #else /* ! CONFIG_HUGETLB_PAGE */ static inline void flush_hugetlb_page(struct vm_area_struct *vma, unsigned long vmaddr) diff --git a/arch/powerpc/kvm/book3s_hv.c b/arch/powerpc/kvm/book3s_hv.c index 83e929e66f9d..721d4603a235 100644 --- a/arch/powerpc/kvm/book3s_hv.c +++ b/arch/powerpc/kvm/book3s_hv.c @@ -1183,7 +1183,7 @@ static const struct vm_operations_struct kvm_rma_vm_ops = { static int kvm_rma_mmap(struct file *file, struct vm_area_struct *vma) { - vma->vm_flags |= VM_RESERVED; + vma->vm_flags |= VM_DONTEXPAND | VM_DONTDUMP; vma->vm_ops = &kvm_rma_vm_ops; return 0; } diff --git a/arch/powerpc/mm/fault.c b/arch/powerpc/mm/fault.c index 5495ebe983a2..0a6b28336eb0 100644 --- a/arch/powerpc/mm/fault.c +++ b/arch/powerpc/mm/fault.c @@ -451,6 +451,7 @@ good_area: /* Clear FAULT_FLAG_ALLOW_RETRY to avoid any risk * of starvation. */ flags &= ~FAULT_FLAG_ALLOW_RETRY; + flags |= FAULT_FLAG_TRIED; goto retry; } } diff --git a/arch/powerpc/oprofile/cell/spu_task_sync.c b/arch/powerpc/oprofile/cell/spu_task_sync.c index 642fca137ccb..28f1af2db1f5 100644 --- a/arch/powerpc/oprofile/cell/spu_task_sync.c +++ b/arch/powerpc/oprofile/cell/spu_task_sync.c @@ -304,7 +304,7 @@ static inline unsigned long fast_get_dcookie(struct path *path) return cookie; } -/* Look up the dcookie for the task's first VM_EXECUTABLE mapping, +/* Look up the dcookie for the task's mm->exe_file, * which corresponds loosely to "application name". Also, determine * the offset for the SPU ELF object. If computed offset is * non-zero, it implies an embedded SPU object; otherwise, it's a @@ -321,7 +321,6 @@ get_exec_dcookie_and_offset(struct spu *spu, unsigned int *offsetp, { unsigned long app_cookie = 0; unsigned int my_offset = 0; - struct file *app = NULL; struct vm_area_struct *vma; struct mm_struct *mm = spu->mm; @@ -330,16 +329,10 @@ get_exec_dcookie_and_offset(struct spu *spu, unsigned int *offsetp, down_read(&mm->mmap_sem); - for (vma = mm->mmap; vma; vma = vma->vm_next) { - if (!vma->vm_file) - continue; - if (!(vma->vm_flags & VM_EXECUTABLE)) - continue; - app_cookie = fast_get_dcookie(&vma->vm_file->f_path); + if (mm->exe_file) { + app_cookie = fast_get_dcookie(&mm->exe_file->f_path); pr_debug("got dcookie for %s\n", - vma->vm_file->f_dentry->d_name.name); - app = vma->vm_file; - break; + mm->exe_file->f_dentry->d_name.name); } for (vma = mm->mmap; vma; vma = vma->vm_next) { diff --git a/arch/powerpc/platforms/pseries/hotplug-memory.c b/arch/powerpc/platforms/pseries/hotplug-memory.c index 11d8e0544ac0..dc0a035e63bb 100644 --- a/arch/powerpc/platforms/pseries/hotplug-memory.c +++ b/arch/powerpc/platforms/pseries/hotplug-memory.c @@ -77,7 +77,8 @@ static int pseries_remove_memblock(unsigned long base, unsigned int memblock_siz { unsigned long start, start_pfn; struct zone *zone; - int ret; + int i, ret; + int sections_to_remove; start_pfn = base >> PAGE_SHIFT; @@ -97,9 +98,13 @@ static int pseries_remove_memblock(unsigned long base, unsigned int memblock_siz * to sysfs "state" file and we can't remove sysfs entries * while writing to it. So we have to defer it to here. */ - ret = __remove_pages(zone, start_pfn, memblock_size >> PAGE_SHIFT); - if (ret) - return ret; + sections_to_remove = (memblock_size >> PAGE_SHIFT) / PAGES_PER_SECTION; + for (i = 0; i < sections_to_remove; i++) { + unsigned long pfn = start_pfn + i * PAGES_PER_SECTION; + ret = __remove_pages(zone, start_pfn, PAGES_PER_SECTION); + if (ret) + return ret; + } /* * Update memory regions for memory remove diff --git a/arch/s390/Kconfig b/arch/s390/Kconfig index c8af429991d9..ceff7aef2477 100644 --- a/arch/s390/Kconfig +++ b/arch/s390/Kconfig @@ -68,6 +68,7 @@ config S390 select HAVE_FTRACE_MCOUNT_RECORD select HAVE_C_RECORDMCOUNT select HAVE_SYSCALL_TRACEPOINTS + select SYSCTL_EXCEPTION_TRACE select HAVE_DYNAMIC_FTRACE select HAVE_FUNCTION_GRAPH_TRACER select HAVE_REGS_AND_STACK_ACCESS_API @@ -80,6 +81,7 @@ config S390 select HAVE_IRQ_WORK select HAVE_PERF_EVENTS select ARCH_HAVE_NMI_SAFE_CMPXCHG + select HAVE_DEBUG_KMEMLEAK select HAVE_KERNEL_GZIP select HAVE_KERNEL_BZIP2 select HAVE_KERNEL_LZMA @@ -126,6 +128,7 @@ config S390 select ARCH_INLINE_WRITE_UNLOCK_BH select ARCH_INLINE_WRITE_UNLOCK_IRQ select ARCH_INLINE_WRITE_UNLOCK_IRQRESTORE + select HAVE_UID16 if 32BIT select ARCH_WANT_IPC_PARSE_VERSION select GENERIC_SMP_IDLE_THREAD select GENERIC_TIME_VSYSCALL diff --git a/arch/s390/include/asm/hugetlb.h b/arch/s390/include/asm/hugetlb.h index 2d6e6e380564..593753ee07f3 100644 --- a/arch/s390/include/asm/hugetlb.h +++ b/arch/s390/include/asm/hugetlb.h @@ -33,6 +33,7 @@ static inline int prepare_hugepage_range(struct file *file, } #define hugetlb_prefault_arch_hook(mm) do { } while (0) +#define arch_clear_hugepage_flags(page) do { } while (0) int arch_prepare_hugepage(struct page *page); void arch_release_hugepage(struct page *page); @@ -77,23 +78,6 @@ static inline void __pmd_csp(pmd_t *pmdp) " csp %1,%3" : "=m" (*pmdp) : "d" (reg2), "d" (reg3), "d" (reg4), "m" (*pmdp) : "cc"); - pmd_val(*pmdp) = _SEGMENT_ENTRY_INV | _SEGMENT_ENTRY; -} - -static inline void __pmd_idte(unsigned long address, pmd_t *pmdp) -{ - unsigned long sto = (unsigned long) pmdp - - pmd_index(address) * sizeof(pmd_t); - - if (!(pmd_val(*pmdp) & _SEGMENT_ENTRY_INV)) { - asm volatile( - " .insn rrf,0xb98e0000,%2,%3,0,0" - : "=m" (*pmdp) - : "m" (*pmdp), "a" (sto), - "a" ((address & HPAGE_MASK)) - ); - } - pmd_val(*pmdp) = _SEGMENT_ENTRY_INV | _SEGMENT_ENTRY; } static inline void huge_ptep_invalidate(struct mm_struct *mm, @@ -105,6 +89,7 @@ static inline void huge_ptep_invalidate(struct mm_struct *mm, __pmd_idte(address, pmdp); else __pmd_csp(pmdp); + pmd_val(*pmdp) = _SEGMENT_ENTRY_INV | _SEGMENT_ENTRY; } static inline pte_t huge_ptep_get_and_clear(struct mm_struct *mm, diff --git a/arch/s390/include/asm/pgtable.h b/arch/s390/include/asm/pgtable.h index 6bd7d7483017..979fe3dc0788 100644 --- a/arch/s390/include/asm/pgtable.h +++ b/arch/s390/include/asm/pgtable.h @@ -42,6 +42,7 @@ extern void fault_init(void); * tables contain all the necessary information. */ #define update_mmu_cache(vma, address, ptep) do { } while (0) +#define update_mmu_cache_pmd(vma, address, ptep) do { } while (0) /* * ZERO_PAGE is a global shared page that is always zero; used @@ -347,6 +348,12 @@ extern struct page *vmemmap; #define _SEGMENT_ENTRY_LARGE 0x400 /* STE-format control, large page */ #define _SEGMENT_ENTRY_CO 0x100 /* change-recording override */ +#define _SEGMENT_ENTRY_SPLIT_BIT 0 /* THP splitting bit number */ +#define _SEGMENT_ENTRY_SPLIT (1UL << _SEGMENT_ENTRY_SPLIT_BIT) + +/* Set of bits not changed in pmd_modify */ +#define _SEGMENT_CHG_MASK (_SEGMENT_ENTRY_ORIGIN | _SEGMENT_ENTRY_LARGE \ + | _SEGMENT_ENTRY_SPLIT | _SEGMENT_ENTRY_CO) /* Page status table bits for virtualization */ #define RCP_ACC_BITS 0xf000000000000000UL @@ -506,6 +513,30 @@ static inline int pmd_bad(pmd_t pmd) return (pmd_val(pmd) & mask) != _SEGMENT_ENTRY; } +#define __HAVE_ARCH_PMDP_SPLITTING_FLUSH +extern void pmdp_splitting_flush(struct vm_area_struct *vma, + unsigned long addr, pmd_t *pmdp); + +#define __HAVE_ARCH_PMDP_SET_ACCESS_FLAGS +extern int pmdp_set_access_flags(struct vm_area_struct *vma, + unsigned long address, pmd_t *pmdp, + pmd_t entry, int dirty); + +#define __HAVE_ARCH_PMDP_CLEAR_YOUNG_FLUSH +extern int pmdp_clear_flush_young(struct vm_area_struct *vma, + unsigned long address, pmd_t *pmdp); + +#define __HAVE_ARCH_PMD_WRITE +static inline int pmd_write(pmd_t pmd) +{ + return (pmd_val(pmd) & _SEGMENT_ENTRY_RO) == 0; +} + +static inline int pmd_young(pmd_t pmd) +{ + return 0; +} + static inline int pte_none(pte_t pte) { return (pte_val(pte) & _PAGE_INVALID) && !(pte_val(pte) & _PAGE_SWT); @@ -1159,6 +1190,185 @@ static inline pmd_t *pmd_offset(pud_t *pud, unsigned long address) #define pte_offset_map(pmd, address) pte_offset_kernel(pmd, address) #define pte_unmap(pte) do { } while (0) +static inline void __pmd_idte(unsigned long address, pmd_t *pmdp) +{ + unsigned long sto = (unsigned long) pmdp - + pmd_index(address) * sizeof(pmd_t); + + if (!(pmd_val(*pmdp) & _SEGMENT_ENTRY_INV)) { + asm volatile( + " .insn rrf,0xb98e0000,%2,%3,0,0" + : "=m" (*pmdp) + : "m" (*pmdp), "a" (sto), + "a" ((address & HPAGE_MASK)) + : "cc" + ); + } +} + +#ifdef CONFIG_TRANSPARENT_HUGEPAGE +#define __HAVE_ARCH_PGTABLE_DEPOSIT +extern void pgtable_trans_huge_deposit(struct mm_struct *mm, pgtable_t pgtable); + +#define __HAVE_ARCH_PGTABLE_WITHDRAW +extern pgtable_t pgtable_trans_huge_withdraw(struct mm_struct *mm); + +static inline int pmd_trans_splitting(pmd_t pmd) +{ + return pmd_val(pmd) & _SEGMENT_ENTRY_SPLIT; +} + +static inline void set_pmd_at(struct mm_struct *mm, unsigned long addr, + pmd_t *pmdp, pmd_t entry) +{ + *pmdp = entry; +} + +static inline unsigned long massage_pgprot_pmd(pgprot_t pgprot) +{ + unsigned long pgprot_pmd = 0; + + if (pgprot_val(pgprot) & _PAGE_INVALID) { + if (pgprot_val(pgprot) & _PAGE_SWT) + pgprot_pmd |= _HPAGE_TYPE_NONE; + pgprot_pmd |= _SEGMENT_ENTRY_INV; + } + if (pgprot_val(pgprot) & _PAGE_RO) + pgprot_pmd |= _SEGMENT_ENTRY_RO; + return pgprot_pmd; +} + +static inline pmd_t pmd_modify(pmd_t pmd, pgprot_t newprot) +{ + pmd_val(pmd) &= _SEGMENT_CHG_MASK; + pmd_val(pmd) |= massage_pgprot_pmd(newprot); + return pmd; +} + +static inline pmd_t pmd_mkhuge(pmd_t pmd) +{ + pmd_val(pmd) |= _SEGMENT_ENTRY_LARGE; + return pmd; +} + +static inline pmd_t pmd_mkwrite(pmd_t pmd) +{ + pmd_val(pmd) &= ~_SEGMENT_ENTRY_RO; + return pmd; +} + +static inline pmd_t pmd_wrprotect(pmd_t pmd) +{ + pmd_val(pmd) |= _SEGMENT_ENTRY_RO; + return pmd; +} + +static inline pmd_t pmd_mkdirty(pmd_t pmd) +{ + /* No dirty bit in the segment table entry. */ + return pmd; +} + +static inline pmd_t pmd_mkold(pmd_t pmd) +{ + /* No referenced bit in the segment table entry. */ + return pmd; +} + +static inline pmd_t pmd_mkyoung(pmd_t pmd) +{ + /* No referenced bit in the segment table entry. */ + return pmd; +} + +#define __HAVE_ARCH_PMDP_TEST_AND_CLEAR_YOUNG +static inline int pmdp_test_and_clear_young(struct vm_area_struct *vma, + unsigned long address, pmd_t *pmdp) +{ + unsigned long pmd_addr = pmd_val(*pmdp) & HPAGE_MASK; + long tmp, rc; + int counter; + + rc = 0; + if (MACHINE_HAS_RRBM) { + counter = PTRS_PER_PTE >> 6; + asm volatile( + "0: .insn rre,0xb9ae0000,%0,%3\n" /* rrbm */ + " ogr %1,%0\n" + " la %3,0(%4,%3)\n" + " brct %2,0b\n" + : "=&d" (tmp), "+&d" (rc), "+d" (counter), + "+a" (pmd_addr) + : "a" (64 * 4096UL) : "cc"); + rc = !!rc; + } else { + counter = PTRS_PER_PTE; + asm volatile( + "0: rrbe 0,%2\n" + " la %2,0(%3,%2)\n" + " brc 12,1f\n" + " lhi %0,1\n" + "1: brct %1,0b\n" + : "+d" (rc), "+d" (counter), "+a" (pmd_addr) + : "a" (4096UL) : "cc"); + } + return rc; +} + +#define __HAVE_ARCH_PMDP_GET_AND_CLEAR +static inline pmd_t pmdp_get_and_clear(struct mm_struct *mm, + unsigned long address, pmd_t *pmdp) +{ + pmd_t pmd = *pmdp; + + __pmd_idte(address, pmdp); + pmd_clear(pmdp); + return pmd; +} + +#define __HAVE_ARCH_PMDP_CLEAR_FLUSH +static inline pmd_t pmdp_clear_flush(struct vm_area_struct *vma, + unsigned long address, pmd_t *pmdp) +{ + return pmdp_get_and_clear(vma->vm_mm, address, pmdp); +} + +#define __HAVE_ARCH_PMDP_INVALIDATE +static inline void pmdp_invalidate(struct vm_area_struct *vma, + unsigned long address, pmd_t *pmdp) +{ + __pmd_idte(address, pmdp); +} + +static inline pmd_t mk_pmd_phys(unsigned long physpage, pgprot_t pgprot) +{ + pmd_t __pmd; + pmd_val(__pmd) = physpage + massage_pgprot_pmd(pgprot); + return __pmd; +} + +#define pfn_pmd(pfn, pgprot) mk_pmd_phys(__pa((pfn) << PAGE_SHIFT), (pgprot)) +#define mk_pmd(page, pgprot) pfn_pmd(page_to_pfn(page), (pgprot)) + +static inline int pmd_trans_huge(pmd_t pmd) +{ + return pmd_val(pmd) & _SEGMENT_ENTRY_LARGE; +} + +static inline int has_transparent_hugepage(void) +{ + return MACHINE_HAS_HPAGE ? 1 : 0; +} + +static inline unsigned long pmd_pfn(pmd_t pmd) +{ + if (pmd_trans_huge(pmd)) + return pmd_val(pmd) >> HPAGE_SHIFT; + else + return pmd_val(pmd) >> PAGE_SHIFT; +} +#endif /* CONFIG_TRANSPARENT_HUGEPAGE */ + /* * 31 bit swap entry format: * A page-table entry has some bits we have to treat in a special way. diff --git a/arch/s390/include/asm/setup.h b/arch/s390/include/asm/setup.h index 87b47ca954f1..8cfd731a18d8 100644 --- a/arch/s390/include/asm/setup.h +++ b/arch/s390/include/asm/setup.h @@ -81,6 +81,7 @@ extern unsigned int s390_user_mode; #define MACHINE_FLAG_SPP (1UL << 13) #define MACHINE_FLAG_TOPOLOGY (1UL << 14) #define MACHINE_FLAG_TE (1UL << 15) +#define MACHINE_FLAG_RRBM (1UL << 16) #define MACHINE_IS_VM (S390_lowcore.machine_flags & MACHINE_FLAG_VM) #define MACHINE_IS_KVM (S390_lowcore.machine_flags & MACHINE_FLAG_KVM) @@ -99,7 +100,8 @@ extern unsigned int s390_user_mode; #define MACHINE_HAS_PFMF (0) #define MACHINE_HAS_SPP (0) #define MACHINE_HAS_TOPOLOGY (0) -#define MACHINE_HAS_TE (0) +#define MACHINE_HAS_TE (0) +#define MACHINE_HAS_RRBM (0) #else /* CONFIG_64BIT */ #define MACHINE_HAS_IEEE (1) #define MACHINE_HAS_CSP (1) @@ -112,6 +114,7 @@ extern unsigned int s390_user_mode; #define MACHINE_HAS_SPP (S390_lowcore.machine_flags & MACHINE_FLAG_SPP) #define MACHINE_HAS_TOPOLOGY (S390_lowcore.machine_flags & MACHINE_FLAG_TOPOLOGY) #define MACHINE_HAS_TE (S390_lowcore.machine_flags & MACHINE_FLAG_TE) +#define MACHINE_HAS_RRBM (S390_lowcore.machine_flags & MACHINE_FLAG_RRBM) #endif /* CONFIG_64BIT */ #define ZFCPDUMP_HSA_SIZE (32UL<<20) diff --git a/arch/s390/include/asm/tlb.h b/arch/s390/include/asm/tlb.h index 06e5acbc84bd..b75d7d686684 100644 --- a/arch/s390/include/asm/tlb.h +++ b/arch/s390/include/asm/tlb.h @@ -137,6 +137,7 @@ static inline void pud_free_tlb(struct mmu_gather *tlb, pud_t *pud, #define tlb_start_vma(tlb, vma) do { } while (0) #define tlb_end_vma(tlb, vma) do { } while (0) #define tlb_remove_tlb_entry(tlb, ptep, addr) do { } while (0) +#define tlb_remove_pmd_tlb_entry(tlb, pmdp, addr) do { } while (0) #define tlb_migrate_finish(mm) do { } while (0) #endif /* _S390_TLB_H */ diff --git a/arch/s390/kernel/early.c b/arch/s390/kernel/early.c index 7f4717675c19..00d114445068 100644 --- a/arch/s390/kernel/early.c +++ b/arch/s390/kernel/early.c @@ -388,6 +388,8 @@ static __init void detect_machine_facilities(void) S390_lowcore.machine_flags |= MACHINE_FLAG_SPP; if (test_facility(50) && test_facility(73)) S390_lowcore.machine_flags |= MACHINE_FLAG_TE; + if (test_facility(66)) + S390_lowcore.machine_flags |= MACHINE_FLAG_RRBM; #endif } diff --git a/arch/s390/mm/fault.c b/arch/s390/mm/fault.c index ac9122ca1152..04ad4001a289 100644 --- a/arch/s390/mm/fault.c +++ b/arch/s390/mm/fault.c @@ -367,6 +367,7 @@ retry: /* Clear FAULT_FLAG_ALLOW_RETRY to avoid any risk * of starvation. */ flags &= ~FAULT_FLAG_ALLOW_RETRY; + flags |= FAULT_FLAG_TRIED; down_read(&mm->mmap_sem); goto retry; } diff --git a/arch/s390/mm/gup.c b/arch/s390/mm/gup.c index eeaf8023851f..60acb93a4680 100644 --- a/arch/s390/mm/gup.c +++ b/arch/s390/mm/gup.c @@ -115,7 +115,16 @@ static inline int gup_pmd_range(pud_t *pudp, pud_t pud, unsigned long addr, pmd = *pmdp; barrier(); next = pmd_addr_end(addr, end); - if (pmd_none(pmd)) + /* + * The pmd_trans_splitting() check below explains why + * pmdp_splitting_flush() has to serialize with + * smp_call_function() against our disabled IRQs, to stop + * this gup-fast code from running while we set the + * splitting bit in the pmd. Returning zero will take + * the slow path that will call wait_split_huge_page() + * if the pmd is still in splitting state. + */ + if (pmd_none(pmd) || pmd_trans_splitting(pmd)) return 0; if (unlikely(pmd_huge(pmd))) { if (!gup_huge_pmd(pmdp, pmd, addr, next, diff --git a/arch/s390/mm/pgtable.c b/arch/s390/mm/pgtable.c index b402991e43d7..c8188a18af05 100644 --- a/arch/s390/mm/pgtable.c +++ b/arch/s390/mm/pgtable.c @@ -787,6 +787,30 @@ void tlb_remove_table(struct mmu_gather *tlb, void *table) tlb_table_flush(tlb); } +#ifdef CONFIG_TRANSPARENT_HUGEPAGE +void thp_split_vma(struct vm_area_struct *vma) +{ + unsigned long addr; + struct page *page; + + for (addr = vma->vm_start; addr < vma->vm_end; addr += PAGE_SIZE) { + page = follow_page(vma, addr, FOLL_SPLIT); + } +} + +void thp_split_mm(struct mm_struct *mm) +{ + struct vm_area_struct *vma = mm->mmap; + + while (vma != NULL) { + thp_split_vma(vma); + vma->vm_flags &= ~VM_HUGEPAGE; + vma->vm_flags |= VM_NOHUGEPAGE; + vma = vma->vm_next; + } +} +#endif /* CONFIG_TRANSPARENT_HUGEPAGE */ + /* * switch on pgstes for its userspace process (for kvm) */ @@ -824,6 +848,12 @@ int s390_enable_sie(void) if (!mm) return -ENOMEM; +#ifdef CONFIG_TRANSPARENT_HUGEPAGE + /* split thp mappings and disable thp for future mappings */ + thp_split_mm(mm); + mm->def_flags |= VM_NOHUGEPAGE; +#endif + /* Now lets check again if something happened */ task_lock(tsk); if (!tsk->mm || atomic_read(&tsk->mm->mm_users) > 1 || @@ -866,3 +896,81 @@ bool kernel_page_present(struct page *page) return cc == 0; } #endif /* CONFIG_HIBERNATION && CONFIG_DEBUG_PAGEALLOC */ + +#ifdef CONFIG_TRANSPARENT_HUGEPAGE +int pmdp_clear_flush_young(struct vm_area_struct *vma, unsigned long address, + pmd_t *pmdp) +{ + VM_BUG_ON(address & ~HPAGE_PMD_MASK); + /* No need to flush TLB + * On s390 reference bits are in storage key and never in TLB */ + return pmdp_test_and_clear_young(vma, address, pmdp); +} + +int pmdp_set_access_flags(struct vm_area_struct *vma, + unsigned long address, pmd_t *pmdp, + pmd_t entry, int dirty) +{ + VM_BUG_ON(address & ~HPAGE_PMD_MASK); + + if (pmd_same(*pmdp, entry)) + return 0; + pmdp_invalidate(vma, address, pmdp); + set_pmd_at(vma->vm_mm, address, pmdp, entry); + return 1; +} + +static void pmdp_splitting_flush_sync(void *arg) +{ + /* Simply deliver the interrupt */ +} + +void pmdp_splitting_flush(struct vm_area_struct *vma, unsigned long address, + pmd_t *pmdp) +{ + VM_BUG_ON(address & ~HPAGE_PMD_MASK); + if (!test_and_set_bit(_SEGMENT_ENTRY_SPLIT_BIT, + (unsigned long *) pmdp)) { + /* need to serialize against gup-fast (IRQ disabled) */ + smp_call_function(pmdp_splitting_flush_sync, NULL, 1); + } +} + +void pgtable_trans_huge_deposit(struct mm_struct *mm, pgtable_t pgtable) +{ + struct list_head *lh = (struct list_head *) pgtable; + + assert_spin_locked(&mm->page_table_lock); + + /* FIFO */ + if (!mm->pmd_huge_pte) + INIT_LIST_HEAD(lh); + else + list_add(lh, (struct list_head *) mm->pmd_huge_pte); + mm->pmd_huge_pte = pgtable; +} + +pgtable_t pgtable_trans_huge_withdraw(struct mm_struct *mm) +{ + struct list_head *lh; + pgtable_t pgtable; + pte_t *ptep; + + assert_spin_locked(&mm->page_table_lock); + + /* FIFO */ + pgtable = mm->pmd_huge_pte; + lh = (struct list_head *) pgtable; + if (list_empty(lh)) + mm->pmd_huge_pte = NULL; + else { + mm->pmd_huge_pte = (pgtable_t) lh->next; + list_del(lh); + } + ptep = (pte_t *) pgtable; + pte_val(*ptep) = _PAGE_TYPE_EMPTY; + ptep++; + pte_val(*ptep) = _PAGE_TYPE_EMPTY; + return pgtable; +} +#endif /* CONFIG_TRANSPARENT_HUGEPAGE */ diff --git a/arch/sh/Kconfig b/arch/sh/Kconfig index 36f5141e8041..3b3e27a3ff2c 100644 --- a/arch/sh/Kconfig +++ b/arch/sh/Kconfig @@ -13,14 +13,17 @@ config SUPERH select HAVE_DMA_ATTRS select HAVE_IRQ_WORK select HAVE_PERF_EVENTS + select HAVE_DEBUG_BUGVERBOSE select ARCH_HAVE_CUSTOM_GPIO_H select ARCH_HAVE_NMI_SAFE_CMPXCHG if (GUSA_RB || CPU_SH4A) select PERF_USE_VMALLOC + select HAVE_DEBUG_KMEMLEAK select HAVE_KERNEL_GZIP select HAVE_KERNEL_BZIP2 select HAVE_KERNEL_LZMA select HAVE_KERNEL_XZ select HAVE_KERNEL_LZO + select HAVE_UID16 select ARCH_WANT_IPC_PARSE_VERSION select HAVE_SYSCALL_TRACEPOINTS select HAVE_REGS_AND_STACK_ACCESS_API diff --git a/arch/sh/include/asm/hugetlb.h b/arch/sh/include/asm/hugetlb.h index 967068fb79ac..b3808c7d67b2 100644 --- a/arch/sh/include/asm/hugetlb.h +++ b/arch/sh/include/asm/hugetlb.h @@ -1,6 +1,7 @@ #ifndef _ASM_SH_HUGETLB_H #define _ASM_SH_HUGETLB_H +#include <asm/cacheflush.h> #include <asm/page.h> @@ -89,4 +90,9 @@ static inline void arch_release_hugepage(struct page *page) { } +static inline void arch_clear_hugepage_flags(struct page *page) +{ + clear_bit(PG_dcache_clean, &page->flags); +} + #endif /* _ASM_SH_HUGETLB_H */ diff --git a/arch/sh/mm/fault.c b/arch/sh/mm/fault.c index 3bdc1ad9a341..cbbdcad8fcb3 100644 --- a/arch/sh/mm/fault.c +++ b/arch/sh/mm/fault.c @@ -504,6 +504,7 @@ good_area: } if (fault & VM_FAULT_RETRY) { flags &= ~FAULT_FLAG_ALLOW_RETRY; + flags |= FAULT_FLAG_TRIED; /* * No need to up_read(&mm->mmap_sem) as we would diff --git a/arch/sparc/Kconfig b/arch/sparc/Kconfig index 67f1f6f5f4e1..91c780c973ba 100644 --- a/arch/sparc/Kconfig +++ b/arch/sparc/Kconfig @@ -18,6 +18,7 @@ config SPARC select HAVE_OPROFILE select HAVE_ARCH_KGDB if !SMP || SPARC64 select HAVE_ARCH_TRACEHOOK + select SYSCTL_EXCEPTION_TRACE select ARCH_WANT_OPTIONAL_GPIOLIB select RTC_CLASS select RTC_DRV_M48T59 @@ -32,6 +33,7 @@ config SPARC select GENERIC_PCI_IOMAP select HAVE_NMI_WATCHDOG if SPARC64 select HAVE_BPF_JIT + select HAVE_DEBUG_BUGVERBOSE select GENERIC_SMP_IDLE_THREAD select GENERIC_CMOS_UPDATE select GENERIC_CLOCKEVENTS @@ -42,6 +44,7 @@ config SPARC32 def_bool !64BIT select GENERIC_ATOMIC64 select CLZ_TAB + select HAVE_UID16 config SPARC64 def_bool 64BIT @@ -59,6 +62,7 @@ config SPARC64 select HAVE_DYNAMIC_FTRACE select HAVE_FTRACE_MCOUNT_RECORD select HAVE_SYSCALL_TRACEPOINTS + select HAVE_DEBUG_KMEMLEAK select RTC_DRV_CMOS select RTC_DRV_BQ4802 select RTC_DRV_SUN4V @@ -226,25 +230,6 @@ config EARLYFB help Say Y here to enable a faster early framebuffer boot console. -choice - prompt "Kernel page size" if SPARC64 - default SPARC64_PAGE_SIZE_8KB - -config SPARC64_PAGE_SIZE_8KB - bool "8KB" - help - This lets you select the page size of the kernel. - - 8KB and 64KB work quite well, since SPARC ELF sections - provide for up to 64KB alignment. - - If you don't know what to do, choose 8KB. - -config SPARC64_PAGE_SIZE_64KB - bool "64KB" - -endchoice - config SECCOMP bool "Enable seccomp to safely compute untrusted bytecode" depends on SPARC64 && PROC_FS @@ -316,23 +301,6 @@ config GENERIC_LOCKBREAK default y depends on SPARC64 && SMP && PREEMPT -choice - prompt "SPARC64 Huge TLB Page Size" - depends on SPARC64 && HUGETLB_PAGE - default HUGETLB_PAGE_SIZE_4MB - -config HUGETLB_PAGE_SIZE_4MB - bool "4MB" - -config HUGETLB_PAGE_SIZE_512K - bool "512K" - -config HUGETLB_PAGE_SIZE_64K - depends on !SPARC64_PAGE_SIZE_64KB - bool "64K" - -endchoice - config NUMA bool "NUMA support" depends on SPARC64 && SMP @@ -571,6 +539,7 @@ config COMPAT depends on SPARC64 default y select COMPAT_BINFMT_ELF + select HAVE_UID16 select ARCH_WANT_OLD_COMPAT_IPC config SYSVIPC_COMPAT diff --git a/arch/sparc/include/asm/hugetlb.h b/arch/sparc/include/asm/hugetlb.h index 177061064ee6..8c5eed6d267f 100644 --- a/arch/sparc/include/asm/hugetlb.h +++ b/arch/sparc/include/asm/hugetlb.h @@ -10,7 +10,10 @@ void set_huge_pte_at(struct mm_struct *mm, unsigned long addr, pte_t huge_ptep_get_and_clear(struct mm_struct *mm, unsigned long addr, pte_t *ptep); -void hugetlb_prefault_arch_hook(struct mm_struct *mm); +static inline void hugetlb_prefault_arch_hook(struct mm_struct *mm) +{ + hugetlb_setup(mm); +} static inline int is_hugepage_only_range(struct mm_struct *mm, unsigned long addr, @@ -82,4 +85,8 @@ static inline void arch_release_hugepage(struct page *page) { } +static inline void arch_clear_hugepage_flags(struct page *page) +{ +} + #endif /* _ASM_SPARC64_HUGETLB_H */ diff --git a/arch/sparc/include/asm/mmu_64.h b/arch/sparc/include/asm/mmu_64.h index 9067dc500535..76092c4dd277 100644 --- a/arch/sparc/include/asm/mmu_64.h +++ b/arch/sparc/include/asm/mmu_64.h @@ -30,22 +30,8 @@ #define CTX_PGSZ_MASK ((CTX_PGSZ_BITS << CTX_PGSZ0_SHIFT) | \ (CTX_PGSZ_BITS << CTX_PGSZ1_SHIFT)) -#if defined(CONFIG_SPARC64_PAGE_SIZE_8KB) #define CTX_PGSZ_BASE CTX_PGSZ_8KB -#elif defined(CONFIG_SPARC64_PAGE_SIZE_64KB) -#define CTX_PGSZ_BASE CTX_PGSZ_64KB -#else -#error No page size specified in kernel configuration -#endif - -#if defined(CONFIG_HUGETLB_PAGE_SIZE_4MB) -#define CTX_PGSZ_HUGE CTX_PGSZ_4MB -#elif defined(CONFIG_HUGETLB_PAGE_SIZE_512K) -#define CTX_PGSZ_HUGE CTX_PGSZ_512KB -#elif defined(CONFIG_HUGETLB_PAGE_SIZE_64K) -#define CTX_PGSZ_HUGE CTX_PGSZ_64KB -#endif - +#define CTX_PGSZ_HUGE CTX_PGSZ_4MB #define CTX_PGSZ_KERN CTX_PGSZ_4MB /* Thus, when running on UltraSPARC-III+ and later, we use the following @@ -96,7 +82,7 @@ struct tsb_config { #define MM_TSB_BASE 0 -#ifdef CONFIG_HUGETLB_PAGE +#if defined(CONFIG_HUGETLB_PAGE) || defined(CONFIG_TRANSPARENT_HUGEPAGE) #define MM_TSB_HUGE 1 #define MM_NUM_TSBS 2 #else @@ -107,6 +93,7 @@ typedef struct { spinlock_t lock; unsigned long sparc64_ctx_val; unsigned long huge_pte_count; + struct page *pgtable_page; struct tsb_config tsb_block[MM_NUM_TSBS]; struct hv_tsb_descr tsb_descr[MM_NUM_TSBS]; } mm_context_t; diff --git a/arch/sparc/include/asm/mmu_context_64.h b/arch/sparc/include/asm/mmu_context_64.h index a97fd085cebe..9191ca62ed9c 100644 --- a/arch/sparc/include/asm/mmu_context_64.h +++ b/arch/sparc/include/asm/mmu_context_64.h @@ -36,7 +36,7 @@ static inline void tsb_context_switch(struct mm_struct *mm) { __tsb_context_switch(__pa(mm->pgd), &mm->context.tsb_block[0], -#ifdef CONFIG_HUGETLB_PAGE +#if defined(CONFIG_HUGETLB_PAGE) || defined(CONFIG_TRANSPARENT_HUGEPAGE) (mm->context.tsb_block[1].tsb ? &mm->context.tsb_block[1] : NULL) diff --git a/arch/sparc/include/asm/page_64.h b/arch/sparc/include/asm/page_64.h index f0d09b401036..4b39f74d6ca0 100644 --- a/arch/sparc/include/asm/page_64.h +++ b/arch/sparc/include/asm/page_64.h @@ -3,13 +3,7 @@ #include <linux/const.h> -#if defined(CONFIG_SPARC64_PAGE_SIZE_8KB) #define PAGE_SHIFT 13 -#elif defined(CONFIG_SPARC64_PAGE_SIZE_64KB) -#define PAGE_SHIFT 16 -#else -#error No page size specified in kernel configuration -#endif #define PAGE_SIZE (_AC(1,UL) << PAGE_SHIFT) #define PAGE_MASK (~(PAGE_SIZE-1)) @@ -21,15 +15,9 @@ #define DCACHE_ALIASING_POSSIBLE #endif -#if defined(CONFIG_HUGETLB_PAGE_SIZE_4MB) #define HPAGE_SHIFT 22 -#elif defined(CONFIG_HUGETLB_PAGE_SIZE_512K) -#define HPAGE_SHIFT 19 -#elif defined(CONFIG_HUGETLB_PAGE_SIZE_64K) -#define HPAGE_SHIFT 16 -#endif -#ifdef CONFIG_HUGETLB_PAGE +#if defined(CONFIG_HUGETLB_PAGE) || defined(CONFIG_TRANSPARENT_HUGEPAGE) #define HPAGE_SIZE (_AC(1,UL) << HPAGE_SHIFT) #define HPAGE_MASK (~(HPAGE_SIZE - 1UL)) #define HUGETLB_PAGE_ORDER (HPAGE_SHIFT - PAGE_SHIFT) @@ -38,6 +26,11 @@ #ifndef __ASSEMBLY__ +#if defined(CONFIG_HUGETLB_PAGE) || defined(CONFIG_TRANSPARENT_HUGEPAGE) +struct mm_struct; +extern void hugetlb_setup(struct mm_struct *mm); +#endif + #define WANT_PAGE_VIRTUAL extern void _clear_page(void *page); @@ -98,7 +91,7 @@ typedef unsigned long pgprot_t; #endif /* (STRICT_MM_TYPECHECKS) */ -typedef struct page *pgtable_t; +typedef pte_t *pgtable_t; #define TASK_UNMAPPED_BASE (test_thread_flag(TIF_32BIT) ? \ (_AC(0x0000000070000000,UL)) : \ diff --git a/arch/sparc/include/asm/pgalloc_64.h b/arch/sparc/include/asm/pgalloc_64.h index 40b2d7a7023d..bcfe063bce23 100644 --- a/arch/sparc/include/asm/pgalloc_64.h +++ b/arch/sparc/include/asm/pgalloc_64.h @@ -38,51 +38,20 @@ static inline void pmd_free(struct mm_struct *mm, pmd_t *pmd) kmem_cache_free(pgtable_cache, pmd); } -static inline pte_t *pte_alloc_one_kernel(struct mm_struct *mm, - unsigned long address) -{ - return (pte_t *)__get_free_page(GFP_KERNEL | __GFP_REPEAT | __GFP_ZERO); -} - -static inline pgtable_t pte_alloc_one(struct mm_struct *mm, - unsigned long address) -{ - struct page *page; - pte_t *pte; - - pte = pte_alloc_one_kernel(mm, address); - if (!pte) - return NULL; - page = virt_to_page(pte); - pgtable_page_ctor(page); - return page; -} - -static inline void pte_free_kernel(struct mm_struct *mm, pte_t *pte) -{ - free_page((unsigned long)pte); -} - -static inline void pte_free(struct mm_struct *mm, pgtable_t ptepage) -{ - pgtable_page_dtor(ptepage); - __free_page(ptepage); -} +extern pte_t *pte_alloc_one_kernel(struct mm_struct *mm, + unsigned long address); +extern pgtable_t pte_alloc_one(struct mm_struct *mm, + unsigned long address); +extern void pte_free_kernel(struct mm_struct *mm, pte_t *pte); +extern void pte_free(struct mm_struct *mm, pgtable_t ptepage); -#define pmd_populate_kernel(MM, PMD, PTE) pmd_set(PMD, PTE) -#define pmd_populate(MM,PMD,PTE_PAGE) \ - pmd_populate_kernel(MM,PMD,page_address(PTE_PAGE)) -#define pmd_pgtable(pmd) pmd_page(pmd) +#define pmd_populate_kernel(MM, PMD, PTE) pmd_set(MM, PMD, PTE) +#define pmd_populate(MM, PMD, PTE) pmd_set(MM, PMD, PTE) +#define pmd_pgtable(PMD) ((pte_t *)__pmd_page(PMD)) #define check_pgt_cache() do { } while (0) -static inline void pgtable_free(void *table, bool is_page) -{ - if (is_page) - free_page((unsigned long)table); - else - kmem_cache_free(pgtable_cache, table); -} +extern void pgtable_free(void *table, bool is_page); #ifdef CONFIG_SMP @@ -113,11 +82,10 @@ static inline void pgtable_free_tlb(struct mmu_gather *tlb, void *table, bool is } #endif /* !CONFIG_SMP */ -static inline void __pte_free_tlb(struct mmu_gather *tlb, struct page *ptepage, +static inline void __pte_free_tlb(struct mmu_gather *tlb, pte_t *pte, unsigned long address) { - pgtable_page_dtor(ptepage); - pgtable_free_tlb(tlb, page_address(ptepage), true); + pgtable_free_tlb(tlb, pte, true); } #define __pmd_free_tlb(tlb, pmd, addr) \ diff --git a/arch/sparc/include/asm/pgtable_64.h b/arch/sparc/include/asm/pgtable_64.h index 61210db139fb..95515f1e7cef 100644 --- a/arch/sparc/include/asm/pgtable_64.h +++ b/arch/sparc/include/asm/pgtable_64.h @@ -45,40 +45,59 @@ #define vmemmap ((struct page *)VMEMMAP_BASE) -/* XXX All of this needs to be rethought so we can take advantage - * XXX cheetah's full 64-bit virtual address space, ie. no more hole - * XXX in the middle like on spitfire. -DaveM - */ -/* - * Given a virtual address, the lowest PAGE_SHIFT bits determine offset - * into the page; the next higher PAGE_SHIFT-3 bits determine the pte# - * in the proper pagetable (the -3 is from the 8 byte ptes, and each page - * table is a single page long). The next higher PMD_BITS determine pmd# - * in the proper pmdtable (where we must have PMD_BITS <= (PAGE_SHIFT-2) - * since the pmd entries are 4 bytes, and each pmd page is a single page - * long). Finally, the higher few bits determine pgde#. - */ - /* PMD_SHIFT determines the size of the area a second-level page * table can map */ -#define PMD_SHIFT (PAGE_SHIFT + (PAGE_SHIFT-3)) +#define PMD_SHIFT (PAGE_SHIFT + (PAGE_SHIFT-4)) #define PMD_SIZE (_AC(1,UL) << PMD_SHIFT) #define PMD_MASK (~(PMD_SIZE-1)) #define PMD_BITS (PAGE_SHIFT - 2) /* PGDIR_SHIFT determines what a third-level page table entry can map */ -#define PGDIR_SHIFT (PAGE_SHIFT + (PAGE_SHIFT-3) + PMD_BITS) +#define PGDIR_SHIFT (PAGE_SHIFT + (PAGE_SHIFT-4) + PMD_BITS) #define PGDIR_SIZE (_AC(1,UL) << PGDIR_SHIFT) #define PGDIR_MASK (~(PGDIR_SIZE-1)) #define PGDIR_BITS (PAGE_SHIFT - 2) +#if (PGDIR_SHIFT + PGDIR_BITS) != 44 +#error Page table parameters do not cover virtual address space properly. +#endif + +#if (PMD_SHIFT != HPAGE_SHIFT) +#error PMD_SHIFT must equal HPAGE_SHIFT for transparent huge pages. +#endif + +/* PMDs point to PTE tables which are 4K aligned. */ +#define PMD_PADDR _AC(0xfffffffe,UL) +#define PMD_PADDR_SHIFT _AC(11,UL) + +#ifdef CONFIG_TRANSPARENT_HUGEPAGE +#define PMD_ISHUGE _AC(0x00000001,UL) + +/* This is the PMD layout when PMD_ISHUGE is set. With 4MB huge + * pages, this frees up a bunch of bits in the layout that we can + * use for the protection settings and software metadata. + */ +#define PMD_HUGE_PADDR _AC(0xfffff800,UL) +#define PMD_HUGE_PROTBITS _AC(0x000007ff,UL) +#define PMD_HUGE_PRESENT _AC(0x00000400,UL) +#define PMD_HUGE_WRITE _AC(0x00000200,UL) +#define PMD_HUGE_DIRTY _AC(0x00000100,UL) +#define PMD_HUGE_ACCESSED _AC(0x00000080,UL) +#define PMD_HUGE_EXEC _AC(0x00000040,UL) +#define PMD_HUGE_SPLITTING _AC(0x00000020,UL) +#endif + +/* PGDs point to PMD tables which are 8K aligned. */ +#define PGD_PADDR _AC(0xfffffffc,UL) +#define PGD_PADDR_SHIFT _AC(11,UL) + #ifndef __ASSEMBLY__ #include <linux/sched.h> /* Entries per page directory level. */ -#define PTRS_PER_PTE (1UL << (PAGE_SHIFT-3)) +#define PTRS_PER_PTE (1UL << (PAGE_SHIFT-4)) #define PTRS_PER_PMD (1UL << PMD_BITS) #define PTRS_PER_PGD (1UL << PGDIR_BITS) @@ -160,26 +179,11 @@ #define _PAGE_SZ8K_4V _AC(0x0000000000000000,UL) /* 8K Page */ #define _PAGE_SZALL_4V _AC(0x0000000000000007,UL) /* All pgsz bits */ -#if PAGE_SHIFT == 13 #define _PAGE_SZBITS_4U _PAGE_SZ8K_4U #define _PAGE_SZBITS_4V _PAGE_SZ8K_4V -#elif PAGE_SHIFT == 16 -#define _PAGE_SZBITS_4U _PAGE_SZ64K_4U -#define _PAGE_SZBITS_4V _PAGE_SZ64K_4V -#else -#error Wrong PAGE_SHIFT specified -#endif -#if defined(CONFIG_HUGETLB_PAGE_SIZE_4MB) #define _PAGE_SZHUGE_4U _PAGE_SZ4MB_4U #define _PAGE_SZHUGE_4V _PAGE_SZ4MB_4V -#elif defined(CONFIG_HUGETLB_PAGE_SIZE_512K) -#define _PAGE_SZHUGE_4U _PAGE_SZ512K_4U -#define _PAGE_SZHUGE_4V _PAGE_SZ512K_4V -#elif defined(CONFIG_HUGETLB_PAGE_SIZE_64K) -#define _PAGE_SZHUGE_4U _PAGE_SZ64K_4U -#define _PAGE_SZHUGE_4V _PAGE_SZ64K_4V -#endif /* These are actually filled in at boot time by sun4{u,v}_pgprot_init() */ #define __P000 __pgprot(0) @@ -218,7 +222,6 @@ extern unsigned long _PAGE_CACHE; extern unsigned long pg_iobits; extern unsigned long _PAGE_ALL_SZ_BITS; -extern unsigned long _PAGE_SZBITS; extern struct page *mem_map_zero; #define ZERO_PAGE(vaddr) (mem_map_zero) @@ -231,25 +234,25 @@ extern struct page *mem_map_zero; static inline pte_t pfn_pte(unsigned long pfn, pgprot_t prot) { unsigned long paddr = pfn << PAGE_SHIFT; - unsigned long sz_bits; - - sz_bits = 0UL; - if (_PAGE_SZBITS_4U != 0UL || _PAGE_SZBITS_4V != 0UL) { - __asm__ __volatile__( - "\n661: sethi %%uhi(%1), %0\n" - " sllx %0, 32, %0\n" - " .section .sun4v_2insn_patch, \"ax\"\n" - " .word 661b\n" - " mov %2, %0\n" - " nop\n" - " .previous\n" - : "=r" (sz_bits) - : "i" (_PAGE_SZBITS_4U), "i" (_PAGE_SZBITS_4V)); - } - return __pte(paddr | sz_bits | pgprot_val(prot)); + + BUILD_BUG_ON(_PAGE_SZBITS_4U != 0UL || _PAGE_SZBITS_4V != 0UL); + return __pte(paddr | pgprot_val(prot)); } #define mk_pte(page, pgprot) pfn_pte(page_to_pfn(page), (pgprot)) +#ifdef CONFIG_TRANSPARENT_HUGEPAGE +extern pmd_t pfn_pmd(unsigned long page_nr, pgprot_t pgprot); +#define mk_pmd(page, pgprot) pfn_pmd(page_to_pfn(page), (pgprot)) + +extern pmd_t pmd_modify(pmd_t pmd, pgprot_t newprot); + +static inline pmd_t pmd_mkhuge(pmd_t pmd) +{ + /* Do nothing, mk_pmd() does this part. */ + return pmd; +} +#endif + /* This one can be done with two shifts. */ static inline unsigned long pte_pfn(pte_t pte) { @@ -286,6 +289,7 @@ static inline pte_t pte_modify(pte_t pte, pgprot_t prot) * Note: We encode this into 3 sun4v 2-insn patch sequences. */ + BUILD_BUG_ON(_PAGE_SZBITS_4U != 0UL || _PAGE_SZBITS_4V != 0UL); __asm__ __volatile__( "\n661: sethi %%uhi(%2), %1\n" " sethi %%hi(%2), %0\n" @@ -307,10 +311,10 @@ static inline pte_t pte_modify(pte_t pte, pgprot_t prot) : "=r" (mask), "=r" (tmp) : "i" (_PAGE_PADDR_4U | _PAGE_MODIFIED_4U | _PAGE_ACCESSED_4U | _PAGE_CP_4U | _PAGE_CV_4U | _PAGE_E_4U | _PAGE_PRESENT_4U | - _PAGE_SZBITS_4U | _PAGE_SPECIAL), + _PAGE_SPECIAL), "i" (_PAGE_PADDR_4V | _PAGE_MODIFIED_4V | _PAGE_ACCESSED_4V | _PAGE_CP_4V | _PAGE_CV_4V | _PAGE_E_4V | _PAGE_PRESENT_4V | - _PAGE_SZBITS_4V | _PAGE_SPECIAL)); + _PAGE_SPECIAL)); return __pte((pte_val(pte) & mask) | (pgprot_val(prot) & ~mask)); } @@ -618,19 +622,130 @@ static inline unsigned long pte_special(pte_t pte) return pte_val(pte) & _PAGE_SPECIAL; } -#define pmd_set(pmdp, ptep) \ - (pmd_val(*(pmdp)) = (__pa((unsigned long) (ptep)) >> 11UL)) +#ifdef CONFIG_TRANSPARENT_HUGEPAGE +static inline int pmd_young(pmd_t pmd) +{ + return pmd_val(pmd) & PMD_HUGE_ACCESSED; +} + +static inline int pmd_write(pmd_t pmd) +{ + return pmd_val(pmd) & PMD_HUGE_WRITE; +} + +static inline unsigned long pmd_pfn(pmd_t pmd) +{ + unsigned long val = pmd_val(pmd) & PMD_HUGE_PADDR; + + return val >> (PAGE_SHIFT - PMD_PADDR_SHIFT); +} + +static inline int pmd_large(pmd_t pmd) +{ + return (pmd_val(pmd) & (PMD_ISHUGE | PMD_HUGE_PRESENT)) == + (PMD_ISHUGE | PMD_HUGE_PRESENT); +} + +static inline int pmd_trans_splitting(pmd_t pmd) +{ + return (pmd_val(pmd) & (PMD_ISHUGE|PMD_HUGE_SPLITTING)) == + (PMD_ISHUGE|PMD_HUGE_SPLITTING); +} + +static inline int pmd_trans_huge(pmd_t pmd) +{ + return pmd_val(pmd) & PMD_ISHUGE; +} + +#define has_transparent_hugepage() 1 + +static inline pmd_t pmd_mkold(pmd_t pmd) +{ + pmd_val(pmd) &= ~PMD_HUGE_ACCESSED; + return pmd; +} + +static inline pmd_t pmd_wrprotect(pmd_t pmd) +{ + pmd_val(pmd) &= ~PMD_HUGE_WRITE; + return pmd; +} + +static inline pmd_t pmd_mkdirty(pmd_t pmd) +{ + pmd_val(pmd) |= PMD_HUGE_DIRTY; + return pmd; +} + +static inline pmd_t pmd_mkyoung(pmd_t pmd) +{ + pmd_val(pmd) |= PMD_HUGE_ACCESSED; + return pmd; +} + +static inline pmd_t pmd_mkwrite(pmd_t pmd) +{ + pmd_val(pmd) |= PMD_HUGE_WRITE; + return pmd; +} + +static inline pmd_t pmd_mknotpresent(pmd_t pmd) +{ + pmd_val(pmd) &= ~PMD_HUGE_PRESENT; + return pmd; +} + +static inline pmd_t pmd_mksplitting(pmd_t pmd) +{ + pmd_val(pmd) |= PMD_HUGE_SPLITTING; + return pmd; +} + +extern pgprot_t pmd_pgprot(pmd_t entry); +#endif + +static inline int pmd_present(pmd_t pmd) +{ + return pmd_val(pmd) != 0U; +} + +#define pmd_none(pmd) (!pmd_val(pmd)) + +#ifdef CONFIG_TRANSPARENT_HUGEPAGE +extern void set_pmd_at(struct mm_struct *mm, unsigned long addr, + pmd_t *pmdp, pmd_t pmd); +#else +static inline void set_pmd_at(struct mm_struct *mm, unsigned long addr, + pmd_t *pmdp, pmd_t pmd) +{ + *pmdp = pmd; +} +#endif + +static inline void pmd_set(struct mm_struct *mm, pmd_t *pmdp, pte_t *ptep) +{ + unsigned long val = __pa((unsigned long) (ptep)) >> PMD_PADDR_SHIFT; + + pmd_val(*pmdp) = val; +} + #define pud_set(pudp, pmdp) \ - (pud_val(*(pudp)) = (__pa((unsigned long) (pmdp)) >> 11UL)) -#define __pmd_page(pmd) \ - ((unsigned long) __va((((unsigned long)pmd_val(pmd))<<11UL))) + (pud_val(*(pudp)) = (__pa((unsigned long) (pmdp)) >> PGD_PADDR_SHIFT)) +static inline unsigned long __pmd_page(pmd_t pmd) +{ + unsigned long paddr = (unsigned long) pmd_val(pmd); +#ifdef CONFIG_TRANSPARENT_HUGEPAGE + if (pmd_val(pmd) & PMD_ISHUGE) + paddr &= PMD_HUGE_PADDR; +#endif + paddr <<= PMD_PADDR_SHIFT; + return ((unsigned long) __va(paddr)); +} #define pmd_page(pmd) virt_to_page((void *)__pmd_page(pmd)) #define pud_page_vaddr(pud) \ - ((unsigned long) __va((((unsigned long)pud_val(pud))<<11UL))) + ((unsigned long) __va((((unsigned long)pud_val(pud))<<PGD_PADDR_SHIFT))) #define pud_page(pud) virt_to_page((void *)pud_page_vaddr(pud)) -#define pmd_none(pmd) (!pmd_val(pmd)) #define pmd_bad(pmd) (0) -#define pmd_present(pmd) (pmd_val(pmd) != 0U) #define pmd_clear(pmdp) (pmd_val(*(pmdp)) = 0U) #define pud_none(pud) (!pud_val(pud)) #define pud_bad(pud) (0) @@ -664,6 +779,16 @@ static inline unsigned long pte_special(pte_t pte) extern void tlb_batch_add(struct mm_struct *mm, unsigned long vaddr, pte_t *ptep, pte_t orig, int fullmm); +#define __HAVE_ARCH_PMDP_GET_AND_CLEAR +static inline pmd_t pmdp_get_and_clear(struct mm_struct *mm, + unsigned long addr, + pmd_t *pmdp) +{ + pmd_t pmd = *pmdp; + set_pmd_at(mm, addr, pmdp, __pmd(0U)); + return pmd; +} + static inline void __set_pte_at(struct mm_struct *mm, unsigned long addr, pte_t *ptep, pte_t pte, int fullmm) { @@ -719,6 +844,16 @@ extern void mmu_info(struct seq_file *); struct vm_area_struct; extern void update_mmu_cache(struct vm_area_struct *, unsigned long, pte_t *); +#ifdef CONFIG_TRANSPARENT_HUGEPAGE +extern void update_mmu_cache_pmd(struct vm_area_struct *vma, unsigned long addr, + pmd_t *pmd); + +#define __HAVE_ARCH_PGTABLE_DEPOSIT +extern void pgtable_trans_huge_deposit(struct mm_struct *mm, pgtable_t pgtable); + +#define __HAVE_ARCH_PGTABLE_WITHDRAW +extern pgtable_t pgtable_trans_huge_withdraw(struct mm_struct *mm); +#endif /* Encode and de-code a swap entry */ #define __swp_type(entry) (((entry).val >> PAGE_SHIFT) & 0xffUL) diff --git a/arch/sparc/include/asm/tsb.h b/arch/sparc/include/asm/tsb.h index 1a8afd1ad04f..b4c258de4443 100644 --- a/arch/sparc/include/asm/tsb.h +++ b/arch/sparc/include/asm/tsb.h @@ -147,20 +147,96 @@ extern struct tsb_phys_patch_entry __tsb_phys_patch, __tsb_phys_patch_end; brz,pn REG1, FAIL_LABEL; \ sllx VADDR, 64 - (PMD_SHIFT + PMD_BITS), REG2; \ srlx REG2, 64 - PAGE_SHIFT, REG2; \ - sllx REG1, 11, REG1; \ + sllx REG1, PGD_PADDR_SHIFT, REG1; \ andn REG2, 0x3, REG2; \ lduwa [REG1 + REG2] ASI_PHYS_USE_EC, REG1; \ brz,pn REG1, FAIL_LABEL; \ sllx VADDR, 64 - PMD_SHIFT, REG2; \ - srlx REG2, 64 - PAGE_SHIFT, REG2; \ - sllx REG1, 11, REG1; \ + srlx REG2, 64 - (PAGE_SHIFT - 1), REG2; \ + sllx REG1, PMD_PADDR_SHIFT, REG1; \ andn REG2, 0x7, REG2; \ add REG1, REG2, REG1; - /* Do a user page table walk in MMU globals. Leaves physical PTE - * pointer in REG1. Jumps to FAIL_LABEL on early page table walk - * termination. Physical base of page tables is in PHYS_PGD which - * will not be modified. + /* This macro exists only to make the PMD translator below easier + * to read. It hides the ELF section switch for the sun4v code + * patching. + */ +#define OR_PTE_BIT(REG, NAME) \ +661: or REG, _PAGE_##NAME##_4U, REG; \ + .section .sun4v_1insn_patch, "ax"; \ + .word 661b; \ + or REG, _PAGE_##NAME##_4V, REG; \ + .previous; + + /* Load into REG the PTE value for VALID, CACHE, and SZHUGE. */ +#define BUILD_PTE_VALID_SZHUGE_CACHE(REG) \ +661: sethi %uhi(_PAGE_VALID|_PAGE_SZHUGE_4U), REG; \ + .section .sun4v_1insn_patch, "ax"; \ + .word 661b; \ + sethi %uhi(_PAGE_VALID), REG; \ + .previous; \ + sllx REG, 32, REG; \ +661: or REG, _PAGE_CP_4U|_PAGE_CV_4U, REG; \ + .section .sun4v_1insn_patch, "ax"; \ + .word 661b; \ + or REG, _PAGE_CP_4V|_PAGE_CV_4V|_PAGE_SZHUGE_4V, REG; \ + .previous; + + /* PMD has been loaded into REG1, interpret the value, seeing + * if it is a HUGE PMD or a normal one. If it is not valid + * then jump to FAIL_LABEL. If it is a HUGE PMD, and it + * translates to a valid PTE, branch to PTE_LABEL. + * + * We translate the PMD by hand, one bit at a time, + * constructing the huge PTE. + * + * So we construct the PTE in REG2 as follows: + * + * 1) Extract the PMD PFN from REG1 and place it into REG2. + * + * 2) Translate PMD protection bits in REG1 into REG2, one bit + * at a time using andcc tests on REG1 and OR's into REG2. + * + * Only two bits to be concerned with here, EXEC and WRITE. + * Now REG1 is freed up and we can use it as a temporary. + * + * 3) Construct the VALID, CACHE, and page size PTE bits in + * REG1, OR with REG2 to form final PTE. + */ +#ifdef CONFIG_TRANSPARENT_HUGEPAGE +#define USER_PGTABLE_CHECK_PMD_HUGE(VADDR, REG1, REG2, FAIL_LABEL, PTE_LABEL) \ + brz,pn REG1, FAIL_LABEL; \ + andcc REG1, PMD_ISHUGE, %g0; \ + be,pt %xcc, 700f; \ + and REG1, PMD_HUGE_PRESENT|PMD_HUGE_ACCESSED, REG2; \ + cmp REG2, PMD_HUGE_PRESENT|PMD_HUGE_ACCESSED; \ + bne,pn %xcc, FAIL_LABEL; \ + andn REG1, PMD_HUGE_PROTBITS, REG2; \ + sllx REG2, PMD_PADDR_SHIFT, REG2; \ + /* REG2 now holds PFN << PAGE_SHIFT */ \ + andcc REG1, PMD_HUGE_EXEC, %g0; \ + bne,a,pt %xcc, 1f; \ + OR_PTE_BIT(REG2, EXEC); \ +1: andcc REG1, PMD_HUGE_WRITE, %g0; \ + bne,a,pt %xcc, 1f; \ + OR_PTE_BIT(REG2, W); \ + /* REG1 can now be clobbered, build final PTE */ \ +1: BUILD_PTE_VALID_SZHUGE_CACHE(REG1); \ + ba,pt %xcc, PTE_LABEL; \ + or REG1, REG2, REG1; \ +700: +#else +#define USER_PGTABLE_CHECK_PMD_HUGE(VADDR, REG1, REG2, FAIL_LABEL, PTE_LABEL) \ + brz,pn REG1, FAIL_LABEL; \ + nop; +#endif + + /* Do a user page table walk in MMU globals. Leaves final, + * valid, PTE value in REG1. Jumps to FAIL_LABEL on early + * page table walk termination or if the PTE is not valid. + * + * Physical base of page tables is in PHYS_PGD which will not + * be modified. * * VADDR will not be clobbered, but REG1 and REG2 will. */ @@ -172,15 +248,19 @@ extern struct tsb_phys_patch_entry __tsb_phys_patch, __tsb_phys_patch_end; brz,pn REG1, FAIL_LABEL; \ sllx VADDR, 64 - (PMD_SHIFT + PMD_BITS), REG2; \ srlx REG2, 64 - PAGE_SHIFT, REG2; \ - sllx REG1, 11, REG1; \ + sllx REG1, PGD_PADDR_SHIFT, REG1; \ andn REG2, 0x3, REG2; \ lduwa [REG1 + REG2] ASI_PHYS_USE_EC, REG1; \ - brz,pn REG1, FAIL_LABEL; \ - sllx VADDR, 64 - PMD_SHIFT, REG2; \ - srlx REG2, 64 - PAGE_SHIFT, REG2; \ - sllx REG1, 11, REG1; \ + USER_PGTABLE_CHECK_PMD_HUGE(VADDR, REG1, REG2, FAIL_LABEL, 800f) \ + sllx VADDR, 64 - PMD_SHIFT, REG2; \ + srlx REG2, 64 - (PAGE_SHIFT - 1), REG2; \ + sllx REG1, PMD_PADDR_SHIFT, REG1; \ andn REG2, 0x7, REG2; \ - add REG1, REG2, REG1; + add REG1, REG2, REG1; \ + ldxa [REG1] ASI_PHYS_USE_EC, REG1; \ + brgez,pn REG1, FAIL_LABEL; \ + nop; \ +800: /* Lookup a OBP mapping on VADDR in the prom_trans[] table at TL>0. * If no entry is found, FAIL_LABEL will be branched to. On success diff --git a/arch/sparc/kernel/pci.c b/arch/sparc/kernel/pci.c index acc8c838ff72..75b31bcdeadf 100644 --- a/arch/sparc/kernel/pci.c +++ b/arch/sparc/kernel/pci.c @@ -779,7 +779,7 @@ static int __pci_mmap_make_offset(struct pci_dev *pdev, static void __pci_mmap_set_flags(struct pci_dev *dev, struct vm_area_struct *vma, enum pci_mmap_state mmap_state) { - vma->vm_flags |= (VM_IO | VM_RESERVED); + vma->vm_flags |= VM_IO | VM_DONTEXPAND | VM_DONTDUMP; } /* Set vm_page_prot of VMA, as appropriate for this architecture, for a pci diff --git a/arch/sparc/kernel/sun4v_tlb_miss.S b/arch/sparc/kernel/sun4v_tlb_miss.S index e1fbf8c75787..bde867fd71e8 100644 --- a/arch/sparc/kernel/sun4v_tlb_miss.S +++ b/arch/sparc/kernel/sun4v_tlb_miss.S @@ -176,7 +176,7 @@ sun4v_tsb_miss_common: sub %g2, TRAP_PER_CPU_FAULT_INFO, %g2 -#ifdef CONFIG_HUGETLB_PAGE +#if defined(CONFIG_HUGETLB_PAGE) || defined(CONFIG_TRANSPARENT_HUGEPAGE) mov SCRATCHPAD_UTSBREG2, %g5 ldxa [%g5] ASI_SCRATCHPAD, %g5 cmp %g5, -1 diff --git a/arch/sparc/kernel/tsb.S b/arch/sparc/kernel/tsb.S index db15d123f054..d4bdc7a62375 100644 --- a/arch/sparc/kernel/tsb.S +++ b/arch/sparc/kernel/tsb.S @@ -49,7 +49,7 @@ tsb_miss_page_table_walk: /* Before committing to a full page table walk, * check the huge page TSB. */ -#ifdef CONFIG_HUGETLB_PAGE +#if defined(CONFIG_HUGETLB_PAGE) || defined(CONFIG_TRANSPARENT_HUGEPAGE) 661: ldx [%g7 + TRAP_PER_CPU_TSB_HUGE], %g5 nop @@ -110,12 +110,9 @@ tsb_miss_page_table_walk: tsb_miss_page_table_walk_sun4v_fastpath: USER_PGTABLE_WALK_TL1(%g4, %g7, %g5, %g2, tsb_do_fault) - /* Load and check PTE. */ - ldxa [%g5] ASI_PHYS_USE_EC, %g5 - brgez,pn %g5, tsb_do_fault - nop + /* Valid PTE is now in %g5. */ -#ifdef CONFIG_HUGETLB_PAGE +#if defined(CONFIG_HUGETLB_PAGE) || defined(CONFIG_TRANSPARENT_HUGEPAGE) 661: sethi %uhi(_PAGE_SZALL_4U), %g7 sllx %g7, 32, %g7 .section .sun4v_2insn_patch, "ax" diff --git a/arch/sparc/mm/fault_32.c b/arch/sparc/mm/fault_32.c index 77ac917be152..e98bfda205a2 100644 --- a/arch/sparc/mm/fault_32.c +++ b/arch/sparc/mm/fault_32.c @@ -265,6 +265,7 @@ good_area: } if (fault & VM_FAULT_RETRY) { flags &= ~FAULT_FLAG_ALLOW_RETRY; + flags |= FAULT_FLAG_TRIED; /* No need to up_read(&mm->mmap_sem) as we would * have already released it in __lock_page_or_retry diff --git a/arch/sparc/mm/fault_64.c b/arch/sparc/mm/fault_64.c index 1fe0429b6314..2976dba1ebaf 100644 --- a/arch/sparc/mm/fault_64.c +++ b/arch/sparc/mm/fault_64.c @@ -452,6 +452,7 @@ good_area: } if (fault & VM_FAULT_RETRY) { flags &= ~FAULT_FLAG_ALLOW_RETRY; + flags |= FAULT_FLAG_TRIED; /* No need to up_read(&mm->mmap_sem) as we would * have already released it in __lock_page_or_retry @@ -464,13 +465,13 @@ good_area: up_read(&mm->mmap_sem); mm_rss = get_mm_rss(mm); -#ifdef CONFIG_HUGETLB_PAGE +#if defined(CONFIG_HUGETLB_PAGE) || defined(CONFIG_TRANSPARENT_HUGEPAGE) mm_rss -= (mm->context.huge_pte_count * (HPAGE_SIZE / PAGE_SIZE)); #endif if (unlikely(mm_rss > mm->context.tsb_block[MM_TSB_BASE].tsb_rss_limit)) tsb_grow(mm, MM_TSB_BASE, mm_rss); -#ifdef CONFIG_HUGETLB_PAGE +#if defined(CONFIG_HUGETLB_PAGE) || defined(CONFIG_TRANSPARENT_HUGEPAGE) mm_rss = mm->context.huge_pte_count; if (unlikely(mm_rss > mm->context.tsb_block[MM_TSB_HUGE].tsb_rss_limit)) diff --git a/arch/sparc/mm/hugetlbpage.c b/arch/sparc/mm/hugetlbpage.c index 07e14535375c..f76f83d5ac63 100644 --- a/arch/sparc/mm/hugetlbpage.c +++ b/arch/sparc/mm/hugetlbpage.c @@ -303,53 +303,3 @@ struct page *follow_huge_pmd(struct mm_struct *mm, unsigned long address, { return NULL; } - -static void context_reload(void *__data) -{ - struct mm_struct *mm = __data; - - if (mm == current->mm) - load_secondary_context(mm); -} - -void hugetlb_prefault_arch_hook(struct mm_struct *mm) -{ - struct tsb_config *tp = &mm->context.tsb_block[MM_TSB_HUGE]; - - if (likely(tp->tsb != NULL)) - return; - - tsb_grow(mm, MM_TSB_HUGE, 0); - tsb_context_switch(mm); - smp_tsb_sync(mm); - - /* On UltraSPARC-III+ and later, configure the second half of - * the Data-TLB for huge pages. - */ - if (tlb_type == cheetah_plus) { - unsigned long ctx; - - spin_lock(&ctx_alloc_lock); - ctx = mm->context.sparc64_ctx_val; - ctx &= ~CTX_PGSZ_MASK; - ctx |= CTX_PGSZ_BASE << CTX_PGSZ0_SHIFT; - ctx |= CTX_PGSZ_HUGE << CTX_PGSZ1_SHIFT; - - if (ctx != mm->context.sparc64_ctx_val) { - /* When changing the page size fields, we - * must perform a context flush so that no - * stale entries match. This flush must - * occur with the original context register - * settings. - */ - do_flush_tlb_mm(mm); - - /* Reload the context register of all processors - * also executing in this address space. - */ - mm->context.sparc64_ctx_val = ctx; - on_each_cpu(context_reload, mm, 0); - } - spin_unlock(&ctx_alloc_lock); - } -} diff --git a/arch/sparc/mm/init_64.c b/arch/sparc/mm/init_64.c index 7a9b788c6ced..9e28a118e6a4 100644 --- a/arch/sparc/mm/init_64.c +++ b/arch/sparc/mm/init_64.c @@ -276,7 +276,6 @@ static inline void tsb_insert(struct tsb *ent, unsigned long tag, unsigned long } unsigned long _PAGE_ALL_SZ_BITS __read_mostly; -unsigned long _PAGE_SZBITS __read_mostly; static void flush_dcache(unsigned long pfn) { @@ -307,12 +306,24 @@ static void flush_dcache(unsigned long pfn) } } +/* mm->context.lock must be held */ +static void __update_mmu_tsb_insert(struct mm_struct *mm, unsigned long tsb_index, + unsigned long tsb_hash_shift, unsigned long address, + unsigned long tte) +{ + struct tsb *tsb = mm->context.tsb_block[tsb_index].tsb; + unsigned long tag; + + tsb += ((address >> tsb_hash_shift) & + (mm->context.tsb_block[tsb_index].tsb_nentries - 1UL)); + tag = (address >> 22UL); + tsb_insert(tsb, tag, tte); +} + void update_mmu_cache(struct vm_area_struct *vma, unsigned long address, pte_t *ptep) { + unsigned long tsb_index, tsb_hash_shift, flags; struct mm_struct *mm; - struct tsb *tsb; - unsigned long tag, flags; - unsigned long tsb_index, tsb_hash_shift; pte_t pte = *ptep; if (tlb_type != hypervisor) { @@ -329,7 +340,7 @@ void update_mmu_cache(struct vm_area_struct *vma, unsigned long address, pte_t * spin_lock_irqsave(&mm->context.lock, flags); -#ifdef CONFIG_HUGETLB_PAGE +#if defined(CONFIG_HUGETLB_PAGE) || defined(CONFIG_TRANSPARENT_HUGEPAGE) if (mm->context.tsb_block[MM_TSB_HUGE].tsb != NULL) { if ((tlb_type == hypervisor && (pte_val(pte) & _PAGE_SZALL_4V) == _PAGE_SZHUGE_4V) || @@ -341,11 +352,8 @@ void update_mmu_cache(struct vm_area_struct *vma, unsigned long address, pte_t * } #endif - tsb = mm->context.tsb_block[tsb_index].tsb; - tsb += ((address >> tsb_hash_shift) & - (mm->context.tsb_block[tsb_index].tsb_nentries - 1UL)); - tag = (address >> 22UL); - tsb_insert(tsb, tag, pte_val(pte)); + __update_mmu_tsb_insert(mm, tsb_index, tsb_hash_shift, + address, pte_val(pte)); spin_unlock_irqrestore(&mm->context.lock, flags); } @@ -2275,8 +2283,7 @@ static void __init sun4u_pgprot_init(void) __ACCESS_BITS_4U | _PAGE_E_4U); #ifdef CONFIG_DEBUG_PAGEALLOC - kern_linear_pte_xor[0] = (_PAGE_VALID | _PAGE_SZBITS_4U) ^ - 0xfffff80000000000UL; + kern_linear_pte_xor[0] = _PAGE_VALID ^ 0xfffff80000000000UL; #else kern_linear_pte_xor[0] = (_PAGE_VALID | _PAGE_SZ4MB_4U) ^ 0xfffff80000000000UL; @@ -2287,7 +2294,6 @@ static void __init sun4u_pgprot_init(void) for (i = 1; i < 4; i++) kern_linear_pte_xor[i] = kern_linear_pte_xor[0]; - _PAGE_SZBITS = _PAGE_SZBITS_4U; _PAGE_ALL_SZ_BITS = (_PAGE_SZ4MB_4U | _PAGE_SZ512K_4U | _PAGE_SZ64K_4U | _PAGE_SZ8K_4U | _PAGE_SZ32MB_4U | _PAGE_SZ256MB_4U); @@ -2324,8 +2330,7 @@ static void __init sun4v_pgprot_init(void) _PAGE_CACHE = _PAGE_CACHE_4V; #ifdef CONFIG_DEBUG_PAGEALLOC - kern_linear_pte_xor[0] = (_PAGE_VALID | _PAGE_SZBITS_4V) ^ - 0xfffff80000000000UL; + kern_linear_pte_xor[0] = _PAGE_VALID ^ 0xfffff80000000000UL; #else kern_linear_pte_xor[0] = (_PAGE_VALID | _PAGE_SZ4MB_4V) ^ 0xfffff80000000000UL; @@ -2339,7 +2344,6 @@ static void __init sun4v_pgprot_init(void) pg_iobits = (_PAGE_VALID | _PAGE_PRESENT_4V | __DIRTY_BITS_4V | __ACCESS_BITS_4V | _PAGE_E_4V); - _PAGE_SZBITS = _PAGE_SZBITS_4V; _PAGE_ALL_SZ_BITS = (_PAGE_SZ16GB_4V | _PAGE_SZ2GB_4V | _PAGE_SZ256MB_4V | _PAGE_SZ32MB_4V | _PAGE_SZ4MB_4V | _PAGE_SZ512K_4V | @@ -2472,3 +2476,281 @@ void __flush_tlb_all(void) __asm__ __volatile__("wrpr %0, 0, %%pstate" : : "r" (pstate)); } + +static pte_t *get_from_cache(struct mm_struct *mm) +{ + struct page *page; + pte_t *ret; + + spin_lock(&mm->page_table_lock); + page = mm->context.pgtable_page; + ret = NULL; + if (page) { + void *p = page_address(page); + + mm->context.pgtable_page = NULL; + + ret = (pte_t *) (p + (PAGE_SIZE / 2)); + } + spin_unlock(&mm->page_table_lock); + + return ret; +} + +static struct page *__alloc_for_cache(struct mm_struct *mm) +{ + struct page *page = alloc_page(GFP_KERNEL | __GFP_NOTRACK | + __GFP_REPEAT | __GFP_ZERO); + + if (page) { + spin_lock(&mm->page_table_lock); + if (!mm->context.pgtable_page) { + atomic_set(&page->_count, 2); + mm->context.pgtable_page = page; + } + spin_unlock(&mm->page_table_lock); + } + return page; +} + +pte_t *pte_alloc_one_kernel(struct mm_struct *mm, + unsigned long address) +{ + struct page *page; + pte_t *pte; + + pte = get_from_cache(mm); + if (pte) + return pte; + + page = __alloc_for_cache(mm); + if (page) + pte = (pte_t *) page_address(page); + + return pte; +} + +pgtable_t pte_alloc_one(struct mm_struct *mm, + unsigned long address) +{ + struct page *page; + pte_t *pte; + + pte = get_from_cache(mm); + if (pte) + return pte; + + page = __alloc_for_cache(mm); + if (page) { + pgtable_page_ctor(page); + pte = (pte_t *) page_address(page); + } + + return pte; +} + +void pte_free_kernel(struct mm_struct *mm, pte_t *pte) +{ + struct page *page = virt_to_page(pte); + if (put_page_testzero(page)) + free_hot_cold_page(page, 0); +} + +static void __pte_free(pgtable_t pte) +{ + struct page *page = virt_to_page(pte); + if (put_page_testzero(page)) { + pgtable_page_dtor(page); + free_hot_cold_page(page, 0); + } +} + +void pte_free(struct mm_struct *mm, pgtable_t pte) +{ + __pte_free(pte); +} + +void pgtable_free(void *table, bool is_page) +{ + if (is_page) + __pte_free(table); + else + kmem_cache_free(pgtable_cache, table); +} + +#ifdef CONFIG_TRANSPARENT_HUGEPAGE +static pmd_t pmd_set_protbits(pmd_t pmd, pgprot_t pgprot, bool for_modify) +{ + if (pgprot_val(pgprot) & _PAGE_VALID) + pmd_val(pmd) |= PMD_HUGE_PRESENT; + if (tlb_type == hypervisor) { + if (pgprot_val(pgprot) & _PAGE_WRITE_4V) + pmd_val(pmd) |= PMD_HUGE_WRITE; + if (pgprot_val(pgprot) & _PAGE_EXEC_4V) + pmd_val(pmd) |= PMD_HUGE_EXEC; + + if (!for_modify) { + if (pgprot_val(pgprot) & _PAGE_ACCESSED_4V) + pmd_val(pmd) |= PMD_HUGE_ACCESSED; + if (pgprot_val(pgprot) & _PAGE_MODIFIED_4V) + pmd_val(pmd) |= PMD_HUGE_DIRTY; + } + } else { + if (pgprot_val(pgprot) & _PAGE_WRITE_4U) + pmd_val(pmd) |= PMD_HUGE_WRITE; + if (pgprot_val(pgprot) & _PAGE_EXEC_4U) + pmd_val(pmd) |= PMD_HUGE_EXEC; + + if (!for_modify) { + if (pgprot_val(pgprot) & _PAGE_ACCESSED_4U) + pmd_val(pmd) |= PMD_HUGE_ACCESSED; + if (pgprot_val(pgprot) & _PAGE_MODIFIED_4U) + pmd_val(pmd) |= PMD_HUGE_DIRTY; + } + } + + return pmd; +} + +pmd_t pfn_pmd(unsigned long page_nr, pgprot_t pgprot) +{ + pmd_t pmd; + + pmd_val(pmd) = (page_nr << ((PAGE_SHIFT - PMD_PADDR_SHIFT))); + pmd_val(pmd) |= PMD_ISHUGE; + pmd = pmd_set_protbits(pmd, pgprot, false); + return pmd; +} + +pmd_t pmd_modify(pmd_t pmd, pgprot_t newprot) +{ + pmd_val(pmd) &= ~(PMD_HUGE_PRESENT | + PMD_HUGE_WRITE | + PMD_HUGE_EXEC); + pmd = pmd_set_protbits(pmd, newprot, true); + return pmd; +} + +pgprot_t pmd_pgprot(pmd_t entry) +{ + unsigned long pte = 0; + + if (pmd_val(entry) & PMD_HUGE_PRESENT) + pte |= _PAGE_VALID; + + if (tlb_type == hypervisor) { + if (pmd_val(entry) & PMD_HUGE_PRESENT) + pte |= _PAGE_PRESENT_4V; + if (pmd_val(entry) & PMD_HUGE_EXEC) + pte |= _PAGE_EXEC_4V; + if (pmd_val(entry) & PMD_HUGE_WRITE) + pte |= _PAGE_W_4V; + if (pmd_val(entry) & PMD_HUGE_ACCESSED) + pte |= _PAGE_ACCESSED_4V; + if (pmd_val(entry) & PMD_HUGE_DIRTY) + pte |= _PAGE_MODIFIED_4V; + pte |= _PAGE_CP_4V|_PAGE_CV_4V; + } else { + if (pmd_val(entry) & PMD_HUGE_PRESENT) + pte |= _PAGE_PRESENT_4U; + if (pmd_val(entry) & PMD_HUGE_EXEC) + pte |= _PAGE_EXEC_4U; + if (pmd_val(entry) & PMD_HUGE_WRITE) + pte |= _PAGE_W_4U; + if (pmd_val(entry) & PMD_HUGE_ACCESSED) + pte |= _PAGE_ACCESSED_4U; + if (pmd_val(entry) & PMD_HUGE_DIRTY) + pte |= _PAGE_MODIFIED_4U; + pte |= _PAGE_CP_4U|_PAGE_CV_4U; + } + + return __pgprot(pte); +} + +void update_mmu_cache_pmd(struct vm_area_struct *vma, unsigned long addr, + pmd_t *pmd) +{ + unsigned long pte, flags; + struct mm_struct *mm; + pmd_t entry = *pmd; + pgprot_t prot; + + if (!pmd_large(entry) || !pmd_young(entry)) + return; + + pte = (pmd_val(entry) & ~PMD_HUGE_PROTBITS); + pte <<= PMD_PADDR_SHIFT; + pte |= _PAGE_VALID; + + prot = pmd_pgprot(entry); + + if (tlb_type == hypervisor) + pgprot_val(prot) |= _PAGE_SZHUGE_4V; + else + pgprot_val(prot) |= _PAGE_SZHUGE_4U; + + pte |= pgprot_val(prot); + + mm = vma->vm_mm; + + spin_lock_irqsave(&mm->context.lock, flags); + + if (mm->context.tsb_block[MM_TSB_HUGE].tsb != NULL) + __update_mmu_tsb_insert(mm, MM_TSB_HUGE, HPAGE_SHIFT, + addr, pte); + + spin_unlock_irqrestore(&mm->context.lock, flags); +} +#endif /* CONFIG_TRANSPARENT_HUGEPAGE */ + +#if defined(CONFIG_HUGETLB_PAGE) || defined(CONFIG_TRANSPARENT_HUGEPAGE) +static void context_reload(void *__data) +{ + struct mm_struct *mm = __data; + + if (mm == current->mm) + load_secondary_context(mm); +} + +void hugetlb_setup(struct mm_struct *mm) +{ + struct tsb_config *tp = &mm->context.tsb_block[MM_TSB_HUGE]; + + if (likely(tp->tsb != NULL)) + return; + + tsb_grow(mm, MM_TSB_HUGE, 0); + tsb_context_switch(mm); + smp_tsb_sync(mm); + + /* On UltraSPARC-III+ and later, configure the second half of + * the Data-TLB for huge pages. + */ + if (tlb_type == cheetah_plus) { + unsigned long ctx; + + spin_lock(&ctx_alloc_lock); + ctx = mm->context.sparc64_ctx_val; + ctx &= ~CTX_PGSZ_MASK; + ctx |= CTX_PGSZ_BASE << CTX_PGSZ0_SHIFT; + ctx |= CTX_PGSZ_HUGE << CTX_PGSZ1_SHIFT; + + if (ctx != mm->context.sparc64_ctx_val) { + /* When changing the page size fields, we + * must perform a context flush so that no + * stale entries match. This flush must + * occur with the original context register + * settings. + */ + do_flush_tlb_mm(mm); + + /* Reload the context register of all processors + * also executing in this address space. + */ + mm->context.sparc64_ctx_val = ctx; + on_each_cpu(context_reload, mm, 0); + } + spin_unlock(&ctx_alloc_lock); + } +} +#endif diff --git a/arch/sparc/mm/tlb.c b/arch/sparc/mm/tlb.c index b1f279cd00bf..3e8fec391fe0 100644 --- a/arch/sparc/mm/tlb.c +++ b/arch/sparc/mm/tlb.c @@ -43,16 +43,37 @@ void flush_tlb_pending(void) put_cpu_var(tlb_batch); } -void tlb_batch_add(struct mm_struct *mm, unsigned long vaddr, - pte_t *ptep, pte_t orig, int fullmm) +static void tlb_batch_add_one(struct mm_struct *mm, unsigned long vaddr, + bool exec) { struct tlb_batch *tb = &get_cpu_var(tlb_batch); unsigned long nr; vaddr &= PAGE_MASK; - if (pte_exec(orig)) + if (exec) vaddr |= 0x1UL; + nr = tb->tlb_nr; + + if (unlikely(nr != 0 && mm != tb->mm)) { + flush_tlb_pending(); + nr = 0; + } + + if (nr == 0) + tb->mm = mm; + + tb->vaddrs[nr] = vaddr; + tb->tlb_nr = ++nr; + if (nr >= TLB_BATCH_NR) + flush_tlb_pending(); + + put_cpu_var(tlb_batch); +} + +void tlb_batch_add(struct mm_struct *mm, unsigned long vaddr, + pte_t *ptep, pte_t orig, int fullmm) +{ if (tlb_type != hypervisor && pte_dirty(orig)) { unsigned long paddr, pfn = pte_pfn(orig); @@ -77,26 +98,91 @@ void tlb_batch_add(struct mm_struct *mm, unsigned long vaddr, } no_cache_flush: + if (!fullmm) + tlb_batch_add_one(mm, vaddr, pte_exec(orig)); +} + +#ifdef CONFIG_TRANSPARENT_HUGEPAGE +static void tlb_batch_pmd_scan(struct mm_struct *mm, unsigned long vaddr, + pmd_t pmd, bool exec) +{ + unsigned long end; + pte_t *pte; + + pte = pte_offset_map(&pmd, vaddr); + end = vaddr + HPAGE_SIZE; + while (vaddr < end) { + if (pte_val(*pte) & _PAGE_VALID) + tlb_batch_add_one(mm, vaddr, exec); + pte++; + vaddr += PAGE_SIZE; + } + pte_unmap(pte); +} - if (fullmm) { - put_cpu_var(tlb_batch); +void set_pmd_at(struct mm_struct *mm, unsigned long addr, + pmd_t *pmdp, pmd_t pmd) +{ + pmd_t orig = *pmdp; + + *pmdp = pmd; + + if (mm == &init_mm) return; + + if ((pmd_val(pmd) ^ pmd_val(orig)) & PMD_ISHUGE) { + if (pmd_val(pmd) & PMD_ISHUGE) + mm->context.huge_pte_count++; + else + mm->context.huge_pte_count--; + if (mm->context.huge_pte_count == 1) + hugetlb_setup(mm); } - nr = tb->tlb_nr; + if (!pmd_none(orig)) { + bool exec = ((pmd_val(orig) & PMD_HUGE_EXEC) != 0); - if (unlikely(nr != 0 && mm != tb->mm)) { - flush_tlb_pending(); - nr = 0; + addr &= HPAGE_MASK; + if (pmd_val(orig) & PMD_ISHUGE) + tlb_batch_add_one(mm, addr, exec); + else + tlb_batch_pmd_scan(mm, addr, orig, exec); } +} - if (nr == 0) - tb->mm = mm; +void pgtable_trans_huge_deposit(struct mm_struct *mm, pgtable_t pgtable) +{ + struct list_head *lh = (struct list_head *) pgtable; - tb->vaddrs[nr] = vaddr; - tb->tlb_nr = ++nr; - if (nr >= TLB_BATCH_NR) - flush_tlb_pending(); + assert_spin_locked(&mm->page_table_lock); - put_cpu_var(tlb_batch); + /* FIFO */ + if (!mm->pmd_huge_pte) + INIT_LIST_HEAD(lh); + else + list_add(lh, (struct list_head *) mm->pmd_huge_pte); + mm->pmd_huge_pte = pgtable; +} + +pgtable_t pgtable_trans_huge_withdraw(struct mm_struct *mm) +{ + struct list_head *lh; + pgtable_t pgtable; + + assert_spin_locked(&mm->page_table_lock); + + /* FIFO */ + pgtable = mm->pmd_huge_pte; + lh = (struct list_head *) pgtable; + if (list_empty(lh)) + mm->pmd_huge_pte = NULL; + else { + mm->pmd_huge_pte = (pgtable_t) lh->next; + list_del(lh); + } + pte_val(pgtable[0]) = 0; + pte_val(pgtable[1]) = 0; + + return pgtable; } +#endif /* CONFIG_TRANSPARENT_HUGEPAGE */ diff --git a/arch/sparc/mm/tsb.c b/arch/sparc/mm/tsb.c index c52add79b83d..7f6474347491 100644 --- a/arch/sparc/mm/tsb.c +++ b/arch/sparc/mm/tsb.c @@ -78,7 +78,7 @@ void flush_tsb_user(struct tlb_batch *tb) base = __pa(base); __flush_tsb_one(tb, PAGE_SHIFT, base, nentries); -#ifdef CONFIG_HUGETLB_PAGE +#if defined(CONFIG_HUGETLB_PAGE) || defined(CONFIG_TRANSPARENT_HUGEPAGE) if (mm->context.tsb_block[MM_TSB_HUGE].tsb) { base = (unsigned long) mm->context.tsb_block[MM_TSB_HUGE].tsb; nentries = mm->context.tsb_block[MM_TSB_HUGE].tsb_nentries; @@ -90,29 +90,12 @@ void flush_tsb_user(struct tlb_batch *tb) spin_unlock_irqrestore(&mm->context.lock, flags); } -#if defined(CONFIG_SPARC64_PAGE_SIZE_8KB) #define HV_PGSZ_IDX_BASE HV_PGSZ_IDX_8K #define HV_PGSZ_MASK_BASE HV_PGSZ_MASK_8K -#elif defined(CONFIG_SPARC64_PAGE_SIZE_64KB) -#define HV_PGSZ_IDX_BASE HV_PGSZ_IDX_64K -#define HV_PGSZ_MASK_BASE HV_PGSZ_MASK_64K -#else -#error Broken base page size setting... -#endif -#ifdef CONFIG_HUGETLB_PAGE -#if defined(CONFIG_HUGETLB_PAGE_SIZE_64K) -#define HV_PGSZ_IDX_HUGE HV_PGSZ_IDX_64K -#define HV_PGSZ_MASK_HUGE HV_PGSZ_MASK_64K -#elif defined(CONFIG_HUGETLB_PAGE_SIZE_512K) -#define HV_PGSZ_IDX_HUGE HV_PGSZ_IDX_512K -#define HV_PGSZ_MASK_HUGE HV_PGSZ_MASK_512K -#elif defined(CONFIG_HUGETLB_PAGE_SIZE_4MB) +#if defined(CONFIG_HUGETLB_PAGE) || defined(CONFIG_TRANSPARENT_HUGEPAGE) #define HV_PGSZ_IDX_HUGE HV_PGSZ_IDX_4MB #define HV_PGSZ_MASK_HUGE HV_PGSZ_MASK_4MB -#else -#error Broken huge page size setting... -#endif #endif static void setup_tsb_params(struct mm_struct *mm, unsigned long tsb_idx, unsigned long tsb_bytes) @@ -207,7 +190,7 @@ static void setup_tsb_params(struct mm_struct *mm, unsigned long tsb_idx, unsign case MM_TSB_BASE: hp->pgsz_idx = HV_PGSZ_IDX_BASE; break; -#ifdef CONFIG_HUGETLB_PAGE +#if defined(CONFIG_HUGETLB_PAGE) || defined(CONFIG_TRANSPARENT_HUGEPAGE) case MM_TSB_HUGE: hp->pgsz_idx = HV_PGSZ_IDX_HUGE; break; @@ -222,7 +205,7 @@ static void setup_tsb_params(struct mm_struct *mm, unsigned long tsb_idx, unsign case MM_TSB_BASE: hp->pgsz_mask = HV_PGSZ_MASK_BASE; break; -#ifdef CONFIG_HUGETLB_PAGE +#if defined(CONFIG_HUGETLB_PAGE) || defined(CONFIG_TRANSPARENT_HUGEPAGE) case MM_TSB_HUGE: hp->pgsz_mask = HV_PGSZ_MASK_HUGE; break; @@ -444,7 +427,7 @@ retry_tsb_alloc: int init_new_context(struct task_struct *tsk, struct mm_struct *mm) { -#ifdef CONFIG_HUGETLB_PAGE +#if defined(CONFIG_HUGETLB_PAGE) || defined(CONFIG_TRANSPARENT_HUGEPAGE) unsigned long huge_pte_count; #endif unsigned int i; @@ -453,7 +436,7 @@ int init_new_context(struct task_struct *tsk, struct mm_struct *mm) mm->context.sparc64_ctx_val = 0UL; -#ifdef CONFIG_HUGETLB_PAGE +#if defined(CONFIG_HUGETLB_PAGE) || defined(CONFIG_TRANSPARENT_HUGEPAGE) /* We reset it to zero because the fork() page copying * will re-increment the counters as the parent PTEs are * copied into the child address space. @@ -462,6 +445,8 @@ int init_new_context(struct task_struct *tsk, struct mm_struct *mm) mm->context.huge_pte_count = 0; #endif + mm->context.pgtable_page = NULL; + /* copy_mm() copies over the parent's mm_struct before calling * us, so we need to zero out the TSB pointer or else tsb_grow() * will be confused and think there is an older TSB to free up. @@ -474,7 +459,7 @@ int init_new_context(struct task_struct *tsk, struct mm_struct *mm) */ tsb_grow(mm, MM_TSB_BASE, get_mm_rss(mm)); -#ifdef CONFIG_HUGETLB_PAGE +#if defined(CONFIG_HUGETLB_PAGE) || defined(CONFIG_TRANSPARENT_HUGEPAGE) if (unlikely(huge_pte_count)) tsb_grow(mm, MM_TSB_HUGE, huge_pte_count); #endif @@ -500,10 +485,17 @@ static void tsb_destroy_one(struct tsb_config *tp) void destroy_context(struct mm_struct *mm) { unsigned long flags, i; + struct page *page; for (i = 0; i < MM_NUM_TSBS; i++) tsb_destroy_one(&mm->context.tsb_block[i]); + page = mm->context.pgtable_page; + if (page && put_page_testzero(page)) { + pgtable_page_dtor(page); + free_hot_cold_page(page, 0); + } + spin_lock_irqsave(&ctx_alloc_lock, flags); if (CTX_VALID(mm->context)) { diff --git a/arch/tile/Kconfig b/arch/tile/Kconfig index c9a3c1fe7297..dc46490adca0 100644 --- a/arch/tile/Kconfig +++ b/arch/tile/Kconfig @@ -7,12 +7,15 @@ config TILE select HAVE_DMA_API_DEBUG select HAVE_KVM if !TILEGX select GENERIC_FIND_FIRST_BIT + select SYSCTL_EXCEPTION_TRACE select USE_GENERIC_SMP_HELPERS select CC_OPTIMIZE_FOR_SIZE + select HAVE_DEBUG_KMEMLEAK select HAVE_GENERIC_HARDIRQS select GENERIC_IRQ_PROBE select GENERIC_PENDING_IRQ if SMP select GENERIC_IRQ_SHOW + select HAVE_DEBUG_BUGVERBOSE select HAVE_SYSCALL_WRAPPERS if TILEGX select SYS_HYPERVISOR select ARCH_HAVE_NMI_SAFE_CMPXCHG diff --git a/arch/tile/include/asm/hugetlb.h b/arch/tile/include/asm/hugetlb.h index b2042380a5aa..0f885af2b621 100644 --- a/arch/tile/include/asm/hugetlb.h +++ b/arch/tile/include/asm/hugetlb.h @@ -106,6 +106,10 @@ static inline void arch_release_hugepage(struct page *page) { } +static inline void arch_clear_hugepage_flags(struct page *page) +{ +} + #ifdef CONFIG_HUGETLB_SUPER_PAGES static inline pte_t arch_make_huge_pte(pte_t entry, struct vm_area_struct *vma, struct page *page, int writable) diff --git a/arch/tile/mm/elf.c b/arch/tile/mm/elf.c index 758b6038c2b7..3cfa98bf9125 100644 --- a/arch/tile/mm/elf.c +++ b/arch/tile/mm/elf.c @@ -36,19 +36,14 @@ static void sim_notify_exec(const char *binary_name) } while (c); } -static int notify_exec(void) +static int notify_exec(struct mm_struct *mm) { int retval = 0; /* failure */ - struct vm_area_struct *vma = current->mm->mmap; - while (vma) { - if ((vma->vm_flags & VM_EXECUTABLE) && vma->vm_file) - break; - vma = vma->vm_next; - } - if (vma) { + + if (mm->exe_file) { char *buf = (char *) __get_free_page(GFP_KERNEL); if (buf) { - char *path = d_path(&vma->vm_file->f_path, + char *path = d_path(&mm->exe_file->f_path, buf, PAGE_SIZE); if (!IS_ERR(path)) { sim_notify_exec(path); @@ -106,16 +101,16 @@ int arch_setup_additional_pages(struct linux_binprm *bprm, unsigned long vdso_base; int retval = 0; + down_write(&mm->mmap_sem); + /* * Notify the simulator that an exec just occurred. * If we can't find the filename of the mapping, just use * whatever was passed as the linux_binprm filename. */ - if (!notify_exec()) + if (!notify_exec(mm)) sim_notify_exec(bprm->filename); - down_write(&mm->mmap_sem); - /* * MAYWRITE to allow gdb to COW and set breakpoints */ diff --git a/arch/tile/mm/fault.c b/arch/tile/mm/fault.c index 84ce7abbf5af..fe811fa5f1b9 100644 --- a/arch/tile/mm/fault.c +++ b/arch/tile/mm/fault.c @@ -454,6 +454,7 @@ good_area: tsk->min_flt++; if (fault & VM_FAULT_RETRY) { flags &= ~FAULT_FLAG_ALLOW_RETRY; + flags |= FAULT_FLAG_TRIED; /* * No need to up_read(&mm->mmap_sem) as we would diff --git a/arch/um/Kconfig.common b/arch/um/Kconfig.common index cb837c223922..648121b037d5 100644 --- a/arch/um/Kconfig.common +++ b/arch/um/Kconfig.common @@ -7,6 +7,7 @@ config UML bool default y select HAVE_GENERIC_HARDIRQS + select HAVE_UID16 select GENERIC_IRQ_SHOW select GENERIC_CPU_DEVICES select GENERIC_IO diff --git a/arch/um/kernel/trap.c b/arch/um/kernel/trap.c index 0353b98ae35a..0f00e9c82080 100644 --- a/arch/um/kernel/trap.c +++ b/arch/um/kernel/trap.c @@ -89,6 +89,7 @@ good_area: current->min_flt++; if (fault & VM_FAULT_RETRY) { flags &= ~FAULT_FLAG_ALLOW_RETRY; + flags |= FAULT_FLAG_TRIED; goto retry; } diff --git a/arch/unicore32/kernel/process.c b/arch/unicore32/kernel/process.c index b6f0458c3143..b008586dad75 100644 --- a/arch/unicore32/kernel/process.c +++ b/arch/unicore32/kernel/process.c @@ -380,7 +380,7 @@ int vectors_user_mapping(void) return install_special_mapping(mm, 0xffff0000, PAGE_SIZE, VM_READ | VM_EXEC | VM_MAYREAD | VM_MAYEXEC | - VM_RESERVED, + VM_DONTEXPAND | VM_DONTDUMP, NULL); } diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig index b72777ff32a9..1ae94bcae5d9 100644 --- a/arch/x86/Kconfig +++ b/arch/x86/Kconfig @@ -10,6 +10,7 @@ config X86_32 def_bool y depends on !64BIT select CLKSRC_I8253 + select HAVE_UID16 config X86_64 def_bool y @@ -46,6 +47,7 @@ config X86 select HAVE_FUNCTION_GRAPH_FP_TEST select HAVE_FUNCTION_TRACE_MCOUNT_TEST select HAVE_SYSCALL_TRACEPOINTS + select SYSCTL_EXCEPTION_TRACE select HAVE_KVM select HAVE_ARCH_KGDB select HAVE_ARCH_TRACEHOOK @@ -65,6 +67,7 @@ config X86 select HAVE_PERF_EVENTS_NMI select HAVE_PERF_REGS select HAVE_PERF_USER_STACK_DUMP + select HAVE_DEBUG_KMEMLEAK select ANON_INODES select HAVE_ALIGNED_STRUCT_PAGE if SLUB && !M386 select HAVE_CMPXCHG_LOCAL if !M386 @@ -85,6 +88,7 @@ config X86 select IRQ_FORCED_THREADING select USE_GENERIC_SMP_HELPERS if SMP select HAVE_BPF_JIT if X86_64 + select HAVE_ARCH_TRANSPARENT_HUGEPAGE select CLKEVT_I8253 select ARCH_HAVE_NMI_SAFE_CMPXCHG select GENERIC_IOMAP @@ -2168,6 +2172,7 @@ config IA32_EMULATION bool "IA32 Emulation" depends on X86_64 select COMPAT_BINFMT_ELF + select HAVE_UID16 ---help--- Include code to run legacy 32-bit programs under a 64-bit kernel. You should likely turn this on, unless you're diff --git a/arch/x86/include/asm/atomic.h b/arch/x86/include/asm/atomic.h index 250b8774c158..b6c3b821acf6 100644 --- a/arch/x86/include/asm/atomic.h +++ b/arch/x86/include/asm/atomic.h @@ -240,30 +240,6 @@ static inline int __atomic_add_unless(atomic_t *v, int a, int u) return c; } - -/* - * atomic_dec_if_positive - decrement by 1 if old value positive - * @v: pointer of type atomic_t - * - * The function returns the old value of *v minus 1, even if - * the atomic variable, v, was not decremented. - */ -static inline int atomic_dec_if_positive(atomic_t *v) -{ - int c, old, dec; - c = atomic_read(v); - for (;;) { - dec = c - 1; - if (unlikely(dec < 0)) - break; - old = atomic_cmpxchg((v), c, dec); - if (likely(old == c)) - break; - c = old; - } - return dec; -} - /** * atomic_inc_short - increment of a short integer * @v: pointer to type int diff --git a/arch/x86/include/asm/hugetlb.h b/arch/x86/include/asm/hugetlb.h index 439a9acc132d..bdd35dbd0605 100644 --- a/arch/x86/include/asm/hugetlb.h +++ b/arch/x86/include/asm/hugetlb.h @@ -90,4 +90,8 @@ static inline void arch_release_hugepage(struct page *page) { } +static inline void arch_clear_hugepage_flags(struct page *page) +{ +} + #endif /* _ASM_X86_HUGETLB_H */ diff --git a/arch/x86/include/asm/pgtable.h b/arch/x86/include/asm/pgtable.h index fc9948465293..a1f780d45f76 100644 --- a/arch/x86/include/asm/pgtable.h +++ b/arch/x86/include/asm/pgtable.h @@ -146,8 +146,7 @@ static inline unsigned long pmd_pfn(pmd_t pmd) static inline int pmd_large(pmd_t pte) { - return (pmd_flags(pte) & (_PAGE_PSE | _PAGE_PRESENT)) == - (_PAGE_PSE | _PAGE_PRESENT); + return pmd_flags(pte) & _PAGE_PSE; } #ifdef CONFIG_TRANSPARENT_HUGEPAGE @@ -415,7 +414,13 @@ static inline int pte_hidden(pte_t pte) static inline int pmd_present(pmd_t pmd) { - return pmd_flags(pmd) & _PAGE_PRESENT; + /* + * Checking for _PAGE_PSE is needed too because + * split_huge_page will temporarily clear the present bit (but + * the _PAGE_PSE flag will remain set at all times while the + * _PAGE_PRESENT bit is clear). + */ + return pmd_flags(pmd) & (_PAGE_PRESENT | _PAGE_PROTNONE | _PAGE_PSE); } static inline int pmd_none(pmd_t pmd) diff --git a/arch/x86/include/asm/pgtable_32.h b/arch/x86/include/asm/pgtable_32.h index 0c92113c4cb6..8faa215a503e 100644 --- a/arch/x86/include/asm/pgtable_32.h +++ b/arch/x86/include/asm/pgtable_32.h @@ -71,6 +71,7 @@ do { \ * tables contain all the necessary information. */ #define update_mmu_cache(vma, address, ptep) do { } while (0) +#define update_mmu_cache_pmd(vma, address, pmd) do { } while (0) #endif /* !__ASSEMBLY__ */ diff --git a/arch/x86/include/asm/pgtable_64.h b/arch/x86/include/asm/pgtable_64.h index 8251be02301e..47356f9df82e 100644 --- a/arch/x86/include/asm/pgtable_64.h +++ b/arch/x86/include/asm/pgtable_64.h @@ -143,6 +143,7 @@ static inline int pgd_large(pgd_t pgd) { return 0; } #define pte_unmap(pte) ((void)(pte))/* NOP */ #define update_mmu_cache(vma, address, ptep) do { } while (0) +#define update_mmu_cache_pmd(vma, address, pmd) do { } while (0) /* Encode and de-code a swap entry */ #if _PAGE_BIT_FILE < _PAGE_BIT_PROTNONE diff --git a/arch/x86/mm/fault.c b/arch/x86/mm/fault.c index a530b230e7d7..8e13ecb41bee 100644 --- a/arch/x86/mm/fault.c +++ b/arch/x86/mm/fault.c @@ -1220,6 +1220,7 @@ good_area: /* Clear FAULT_FLAG_ALLOW_RETRY to avoid any risk * of starvation. */ flags &= ~FAULT_FLAG_ALLOW_RETRY; + flags |= FAULT_FLAG_TRIED; goto retry; } } diff --git a/arch/x86/mm/hugetlbpage.c b/arch/x86/mm/hugetlbpage.c index b91e48512425..937bff5cdaa7 100644 --- a/arch/x86/mm/hugetlbpage.c +++ b/arch/x86/mm/hugetlbpage.c @@ -71,7 +71,6 @@ huge_pmd_share(struct mm_struct *mm, unsigned long addr, pud_t *pud) struct address_space *mapping = vma->vm_file->f_mapping; pgoff_t idx = ((addr - vma->vm_start) >> PAGE_SHIFT) + vma->vm_pgoff; - struct prio_tree_iter iter; struct vm_area_struct *svma; unsigned long saddr; pte_t *spte = NULL; @@ -81,7 +80,7 @@ huge_pmd_share(struct mm_struct *mm, unsigned long addr, pud_t *pud) return (pte_t *)pmd_alloc(mm, pud, addr); mutex_lock(&mapping->i_mmap_mutex); - vma_prio_tree_foreach(svma, &iter, &mapping->i_mmap, idx, idx) { + vma_interval_tree_foreach(svma, &mapping->i_mmap, idx, idx) { if (svma == vma) continue; diff --git a/arch/x86/mm/pat.c b/arch/x86/mm/pat.c index 3d68ef6d2266..0eb572eda406 100644 --- a/arch/x86/mm/pat.c +++ b/arch/x86/mm/pat.c @@ -664,20 +664,20 @@ static void free_pfn_range(u64 paddr, unsigned long size) } /* - * track_pfn_vma_copy is called when vma that is covering the pfnmap gets + * track_pfn_copy is called when vma that is covering the pfnmap gets * copied through copy_page_range(). * * If the vma has a linear pfn mapping for the entire range, we get the prot * from pte and reserve the entire vma range with single reserve_pfn_range call. */ -int track_pfn_vma_copy(struct vm_area_struct *vma) +int track_pfn_copy(struct vm_area_struct *vma) { resource_size_t paddr; unsigned long prot; unsigned long vma_size = vma->vm_end - vma->vm_start; pgprot_t pgprot; - if (is_linear_pfn_mapping(vma)) { + if (vma->vm_flags & VM_PAT) { /* * reserve the whole chunk covered by vma. We need the * starting address and protection from pte. @@ -694,31 +694,59 @@ int track_pfn_vma_copy(struct vm_area_struct *vma) } /* - * track_pfn_vma_new is called when a _new_ pfn mapping is being established - * for physical range indicated by pfn and size. - * * prot is passed in as a parameter for the new mapping. If the vma has a * linear pfn mapping for the entire range reserve the entire vma range with * single reserve_pfn_range call. */ -int track_pfn_vma_new(struct vm_area_struct *vma, pgprot_t *prot, - unsigned long pfn, unsigned long size) +int track_pfn_remap(struct vm_area_struct *vma, pgprot_t *prot, + unsigned long pfn, unsigned long addr, unsigned long size) { + resource_size_t paddr = (resource_size_t)pfn << PAGE_SHIFT; unsigned long flags; - resource_size_t paddr; - unsigned long vma_size = vma->vm_end - vma->vm_start; - if (is_linear_pfn_mapping(vma)) { - /* reserve the whole chunk starting from vm_pgoff */ - paddr = (resource_size_t)vma->vm_pgoff << PAGE_SHIFT; - return reserve_pfn_range(paddr, vma_size, prot, 0); + /* reserve the whole chunk starting from paddr */ + if (addr == vma->vm_start && size == (vma->vm_end - vma->vm_start)) { + int ret; + + ret = reserve_pfn_range(paddr, size, prot, 0); + if (!ret) + vma->vm_flags |= VM_PAT; + return ret; } if (!pat_enabled) return 0; - /* for vm_insert_pfn and friends, we set prot based on lookup */ - flags = lookup_memtype(pfn << PAGE_SHIFT); + /* + * For anything smaller than the vma size we set prot based on the + * lookup. + */ + flags = lookup_memtype(paddr); + + /* Check memtype for the remaining pages */ + while (size > PAGE_SIZE) { + size -= PAGE_SIZE; + paddr += PAGE_SIZE; + if (flags != lookup_memtype(paddr)) + return -EINVAL; + } + + *prot = __pgprot((pgprot_val(vma->vm_page_prot) & (~_PAGE_CACHE_MASK)) | + flags); + + return 0; +} + +int track_pfn_insert(struct vm_area_struct *vma, pgprot_t *prot, + unsigned long pfn) +{ + unsigned long flags; + + if (!pat_enabled) + return 0; + + /* Set prot based on lookup */ + flags = lookup_memtype((resource_size_t)pfn << PAGE_SHIFT); *prot = __pgprot((pgprot_val(vma->vm_page_prot) & (~_PAGE_CACHE_MASK)) | flags); @@ -726,22 +754,31 @@ int track_pfn_vma_new(struct vm_area_struct *vma, pgprot_t *prot, } /* - * untrack_pfn_vma is called while unmapping a pfnmap for a region. + * untrack_pfn is called while unmapping a pfnmap for a region. * untrack can be called for a specific region indicated by pfn and size or - * can be for the entire vma (in which case size can be zero). + * can be for the entire vma (in which case pfn, size are zero). */ -void untrack_pfn_vma(struct vm_area_struct *vma, unsigned long pfn, - unsigned long size) +void untrack_pfn(struct vm_area_struct *vma, unsigned long pfn, + unsigned long size) { resource_size_t paddr; - unsigned long vma_size = vma->vm_end - vma->vm_start; + unsigned long prot; - if (is_linear_pfn_mapping(vma)) { - /* free the whole chunk starting from vm_pgoff */ - paddr = (resource_size_t)vma->vm_pgoff << PAGE_SHIFT; - free_pfn_range(paddr, vma_size); + if (!(vma->vm_flags & VM_PAT)) return; + + /* free the chunk starting from pfn or the whole chunk */ + paddr = (resource_size_t)pfn << PAGE_SHIFT; + if (!paddr && !size) { + if (follow_phys(vma, vma->vm_start, 0, &prot, &paddr)) { + WARN_ON_ONCE(1); + return; + } + + size = vma->vm_end - vma->vm_start; } + free_pfn_range(paddr, size); + vma->vm_flags &= ~VM_PAT; } pgprot_t pgprot_writecombine(pgprot_t prot) diff --git a/arch/x86/mm/pat_rbtree.c b/arch/x86/mm/pat_rbtree.c index 8acaddd0fb21..415f6c4ced36 100644 --- a/arch/x86/mm/pat_rbtree.c +++ b/arch/x86/mm/pat_rbtree.c @@ -12,7 +12,7 @@ #include <linux/debugfs.h> #include <linux/kernel.h> #include <linux/module.h> -#include <linux/rbtree.h> +#include <linux/rbtree_augmented.h> #include <linux/sched.h> #include <linux/gfp.h> @@ -54,29 +54,24 @@ static u64 get_subtree_max_end(struct rb_node *node) return ret; } -/* Update 'subtree_max_end' for a node, based on node and its children */ -static void memtype_rb_augment_cb(struct rb_node *node, void *__unused) +static u64 compute_subtree_max_end(struct memtype *data) { - struct memtype *data; - u64 max_end, child_max_end; - - if (!node) - return; + u64 max_end = data->end, child_max_end; - data = container_of(node, struct memtype, rb); - max_end = data->end; - - child_max_end = get_subtree_max_end(node->rb_right); + child_max_end = get_subtree_max_end(data->rb.rb_right); if (child_max_end > max_end) max_end = child_max_end; - child_max_end = get_subtree_max_end(node->rb_left); + child_max_end = get_subtree_max_end(data->rb.rb_left); if (child_max_end > max_end) max_end = child_max_end; - data->subtree_max_end = max_end; + return max_end; } +RB_DECLARE_CALLBACKS(static, memtype_rb_augment_cb, struct memtype, rb, + u64, subtree_max_end, compute_subtree_max_end) + /* Find the first (lowest start addr) overlapping range from rb tree */ static struct memtype *memtype_rb_lowest_match(struct rb_root *root, u64 start, u64 end) @@ -179,15 +174,17 @@ static void memtype_rb_insert(struct rb_root *root, struct memtype *newdata) struct memtype *data = container_of(*node, struct memtype, rb); parent = *node; + if (data->subtree_max_end < newdata->end) + data->subtree_max_end = newdata->end; if (newdata->start <= data->start) node = &((*node)->rb_left); else if (newdata->start > data->start) node = &((*node)->rb_right); } + newdata->subtree_max_end = newdata->end; rb_link_node(&newdata->rb, parent, node); - rb_insert_color(&newdata->rb, root); - rb_augment_insert(&newdata->rb, memtype_rb_augment_cb, NULL); + rb_insert_augmented(&newdata->rb, root, &memtype_rb_augment_cb); } int rbt_memtype_check_insert(struct memtype *new, unsigned long *ret_type) @@ -209,16 +206,13 @@ int rbt_memtype_check_insert(struct memtype *new, unsigned long *ret_type) struct memtype *rbt_memtype_erase(u64 start, u64 end) { - struct rb_node *deepest; struct memtype *data; data = memtype_rb_exact_match(&memtype_rbroot, start, end); if (!data) goto out; - deepest = rb_augment_erase_begin(&data->rb); - rb_erase(&data->rb, &memtype_rbroot); - rb_augment_erase_end(deepest, memtype_rb_augment_cb, NULL); + rb_erase_augmented(&data->rb, &memtype_rbroot, &memtype_rb_augment_cb); out: return data; } diff --git a/arch/x86/xen/mmu.c b/arch/x86/xen/mmu.c index 5a16824cc2b3..fd28d86fe3d2 100644 --- a/arch/x86/xen/mmu.c +++ b/arch/x86/xen/mmu.c @@ -2451,8 +2451,7 @@ int xen_remap_domain_mfn_range(struct vm_area_struct *vma, prot = __pgprot(pgprot_val(prot) | _PAGE_IOMAP); - BUG_ON(!((vma->vm_flags & (VM_PFNMAP | VM_RESERVED | VM_IO)) == - (VM_PFNMAP | VM_RESERVED | VM_IO))); + BUG_ON(!((vma->vm_flags & (VM_PFNMAP | VM_IO)) == (VM_PFNMAP | VM_IO))); rmd.mfn = mfn; rmd.prot = prot; diff --git a/arch/xtensa/mm/fault.c b/arch/xtensa/mm/fault.c index 5a74c53bc69c..2c2f710ed1dc 100644 --- a/arch/xtensa/mm/fault.c +++ b/arch/xtensa/mm/fault.c @@ -126,6 +126,7 @@ good_area: current->min_flt++; if (fault & VM_FAULT_RETRY) { flags &= ~FAULT_FLAG_ALLOW_RETRY; + flags |= FAULT_FLAG_TRIED; /* No need to up_read(&mm->mmap_sem) as we would * have already released it in __lock_page_or_retry |