BMC/Intel-BMC/linux.git/arch/riscv/kernel, branch dev-5.15-intel

uaccess: fix type mismatch warnings from access_ok()

2022-04-08T12:24:01+00:00

[ Upstream commit 23fc539e81295b14b50c6ccc5baeb4f3d59d822d ] On some architectures, access_ok() does not do any argument type checking, so replacing the definition with a generic one causes a few warnings for harmless issues that were never caught before. Fix the ones that I found either through my own test builds or that were reported by the 0-day bot. Reported-by: kernel test robot Reviewed-by: Christoph Hellwig Acked-by: Dinh Nguyen Signed-off-by: Arnd Bergmann Signed-off-by: Sasha Levin

riscv: Fix fill_callchain return value

2022-04-08T12:22:55+00:00

commit 2b2b574ac587ec5bd7716a356492a85ab8b0ce9f upstream. perf_callchain_store return 0 on success, -1 otherwise, fix fill_callchain to return correct bool value. Fixes: dbeb90b0c1eb ("riscv: Add perf callchain support") Signed-off-by: Nikita Shubin Cc: stable@vger.kernel.org Signed-off-by: Palmer Dabbelt Signed-off-by: Greg Kroah-Hartman

riscv: Fix auipc+jalr relocation range checks

2022-03-16T13:23:43+00:00

commit 0966d385830de3470b7131db8e86c0c5bc9c52dc upstream. RISC-V can do PC-relative jumps with a 32bit range using the following two instructions: auipc t0, imm20 ; t0 = PC + imm20 * 2^12 jalr ra, t0, imm12 ; ra = PC + 4, PC = t0 + imm12 Crucially both the 20bit immediate imm20 and the 12bit immediate imm12 are treated as two's-complement signed values. For this reason the immediates are usually calculated like this: imm20 = (offset + 0x800) >> 12 imm12 = offset & 0xfff ..where offset is the signed offset from the auipc instruction. When the 11th bit of offset is 0 the addition of 0x800 doesn't change the top 20 bits and imm12 considered positive. When the 11th bit is 1 the carry of the addition by 0x800 means imm20 is one higher, but since imm12 is then considered negative the two's complement representation means it all cancels out nicely. However, this addition by 0x800 (2^11) means an offset greater than or equal to 2^31 - 2^11 would overflow so imm20 is considered negative and result in a backwards jump. Similarly the lower range of offset is also moved down by 2^11 and hence the true 32bit range is [-2^31 - 2^11, 2^31 - 2^11) Signed-off-by: Emil Renner Berthing Fixes: e2c0cdfba7f6 ("RISC-V: User-facing API") Cc: stable@vger.kernel.org Signed-off-by: Palmer Dabbelt Signed-off-by: Greg Kroah-Hartman

riscv: fix oops caused by irqsoff latency tracer

2022-03-02T10:48:08+00:00

commit 22e2100b1b07d6f5acc71cc1acb53f680c677d77 upstream. The trace_hardirqs_{on,off}() require the caller to setup frame pointer properly. This because these two functions use macro 'CALLER_ADDR1' (aka. __builtin_return_address(1)) to acquire caller info. If the $fp is used for other purpose, the code generated this macro (as below) could trigger memory access fault. 0xffffffff8011510e <+80>: ld a1,-16(s0) 0xffffffff80115112 <+84>: ld s2,-8(a1) # <-- paging fault here The oops message during booting if compiled with 'irqoff' tracer enabled: [ 0.039615][ T0] Unable to handle kernel NULL pointer dereference at virtual address 00000000000000f8 [ 0.041925][ T0] Oops [#1] [ 0.042063][ T0] Modules linked in: [ 0.042864][ T0] CPU: 0 PID: 0 Comm: swapper/0 Not tainted 5.17.0-rc1-00233-g9a20c48d1ed2 #29 [ 0.043568][ T0] Hardware name: riscv-virtio,qemu (DT) [ 0.044343][ T0] epc : trace_hardirqs_on+0x56/0xe2 [ 0.044601][ T0] ra : restore_all+0x12/0x6e [ 0.044721][ T0] epc : ffffffff80126a5c ra : ffffffff80003b94 sp : ffffffff81403db0 [ 0.044801][ T0] gp : ffffffff8163acd8 tp : ffffffff81414880 t0 : 0000000000000020 [ 0.044882][ T0] t1 : 0098968000000000 t2 : 0000000000000000 s0 : ffffffff81403de0 [ 0.044967][ T0] s1 : 0000000000000000 a0 : 0000000000000001 a1 : 0000000000000100 [ 0.045046][ T0] a2 : 0000000000000000 a3 : 0000000000000000 a4 : 0000000000000000 [ 0.045124][ T0] a5 : 0000000000000000 a6 : 0000000000000000 a7 : 0000000054494d45 [ 0.045210][ T0] s2 : ffffffff80003b94 s3 : ffffffff81a8f1b0 s4 : ffffffff80e27b50 [ 0.045289][ T0] s5 : ffffffff81414880 s6 : ffffffff8160fa00 s7 : 00000000800120e8 [ 0.045389][ T0] s8 : 0000000080013100 s9 : 000000000000007f s10: 0000000000000000 [ 0.045474][ T0] s11: 0000000000000000 t3 : 7fffffffffffffff t4 : 0000000000000000 [ 0.045548][ T0] t5 : 0000000000000000 t6 : ffffffff814aa368 [ 0.045620][ T0] status: 0000000200000100 badaddr: 00000000000000f8 cause: 000000000000000d [ 0.046402][ T0] [] restore_all+0x12/0x6e This because the $fp(aka. $s0) register is not used as frame pointer in the assembly entry code. resume_kernel: REG_L s0, TASK_TI_PREEMPT_COUNT(tp) bnez s0, restore_all REG_L s0, TASK_TI_FLAGS(tp) andi s0, s0, _TIF_NEED_RESCHED beqz s0, restore_all call preempt_schedule_irq j restore_all To fix above issue, here we add one extra level wrapper for function trace_hardirqs_{on,off}() so they can be safely called by low level entry code. Signed-off-by: Changbin Du Fixes: 3c4697982982 ("riscv: Enable LOCKDEP_SUPPORT & fixup TRACE_IRQFLAGS_SUPPORT") Cc: stable@vger.kernel.org Signed-off-by: Palmer Dabbelt Signed-off-by: Greg Kroah-Hartman

riscv: eliminate unreliable __builtin_frame_address(1)

2022-02-16T11:56:18+00:00

commit 6a00ef4493706a23120057fafbc62379bcde11ec upstream. I tried different pieces of code which uses __builtin_frame_address(1) (with both gcc version 7.5.0 and 10.3.0) to verify whether it works as expected on riscv64. The result is negative. What the compiler had generated is as below: 31 fp = (unsigned long)__builtin_frame_address(1); 0xffffffff80006024 <+200>: ld s1,0(s0) It takes '0(s0)' as the address of frame 1 (caller), but the actual address should be '-16(s0)'. | ... | <-+ +-----------------+ | | return address | | | previous fp | | | saved registers | | | local variables | | $fp --> | ... | | +-----------------+ | | return address | | | previous fp --------+ | saved registers | $sp --> | local variables | +-----------------+ This leads the kernel can not dump the full stack trace on riscv. [ 7.222126][ T1] Call Trace: [ 7.222804][ T1] [] dump_backtrace+0x2c/0x3a This problem is not exposed on most riscv builds just because the '0(s0)' occasionally is the address frame 2 (caller's caller), if only ra and fp are stored in frame 1 (caller). | ... | <-+ +-----------------+ | | return address | | $fp --> | previous fp | | +-----------------+ | | return address | | | previous fp --------+ | saved registers | $sp --> | local variables | +-----------------+ This could be a *bug* of gcc that should be fixed. But as noted in gcc manual "Calling this function with a nonzero argument can have unpredictable effects, including crashing the calling program.", let's remove the '__builtin_frame_address(1)' in backtrace code. With this fix now it can show full stack trace: [ 10.444838][ T1] Call Trace: [ 10.446199][ T1] [] dump_backtrace+0x2c/0x3a [ 10.447711][ T1] [] show_stack+0x32/0x3e [ 10.448710][ T1] [] dump_stack_lvl+0x58/0x7a [ 10.449941][ T1] [] dump_stack+0x14/0x1c [ 10.450929][ T1] [] ubsan_epilogue+0x10/0x5a [ 10.451869][ T1] [] __ubsan_handle_load_invalid_value+0x6c/0x78 [ 10.453049][ T1] [] __pagevec_release+0x62/0x64 [ 10.455476][ T1] [] truncate_inode_pages_range+0x132/0x5be [ 10.456798][ T1] [] truncate_inode_pages+0x24/0x30 [ 10.457853][ T1] [] kill_bdev+0x32/0x3c ... Signed-off-by: Changbin Du Fixes: eac2f3059e02 ("riscv: stacktrace: fix the riscv stacktrace when CONFIG_FRAME_POINTER enabled") Cc: stable@vger.kernel.org Signed-off-by: Palmer Dabbelt Signed-off-by: Greg Kroah-Hartman

riscv: cpu-hotplug: clear cpu from numa map when teardown

2022-02-16T11:56:18+00:00

commit f40fe31c01445f31253b15bef2412b33ae31093b upstream. There is numa_add_cpu() when cpus online, accordingly, there should be numa_remove_cpu() when cpus offline. Signed-off-by: Pingfan Liu Fixes: 4f0e8eef772e ("riscv: Add numa support for riscv64 platform") Cc: stable@vger.kernel.org [Palmer: Add missing NUMA include] Signed-off-by: Palmer Dabbelt Signed-off-by: Greg Kroah-Hartman

riscv: use hart id instead of cpu id on machine_kexec

2022-01-27T10:02:50+00:00

commit 0e105f1d0037d677dff3c697d22f9551e6c39af8 upstream. raw_smp_processor_id() doesn't return the hart id as stated in arch/riscv/include/asm/smp.h, use smp_processor_id() instead to get the cpu id, and cpuid_to_hartid_map() to pass the hart id to the next kernel. This fixes kexec on HiFive Unleashed/Unmatched where cpu ids and hart ids don't match (on qemu-virt they match). Fixes: fba8a8674f68 ("RISC-V: Add kexec support") Signed-off-by: Nick Kossifidis Cc: stable@vger.kernel.org Signed-off-by: Palmer Dabbelt Signed-off-by: Greg Kroah-Hartman

riscv: Don't use va_pa_offset on kdump

2022-01-27T10:02:50+00:00

commit a11c07f032a0e9a562a32ece73af96b0e754c4b3 upstream. On kdump instead of using an intermediate step to relocate the kernel, that lives in a "control buffer" outside the current kernel's mapping, we jump to the crash kernel directly by calling riscv_kexec_norelocate(). The current implementation uses va_pa_offset while switching to physical addressing, however since we moved the kernel outside the linear mapping this won't work anymore since riscv_kexec_norelocate() is part of the kernel mapping and we should use kernel_map.va_kernel_pa_offset, and also take XIP kernel into account. We don't really need to use va_pa_offset on riscv_kexec_norelocate, we can just set STVEC to the physical address of the new kernel instead and let the hart jump to the new kernel on the next instruction after setting SATP to zero. This fixes kdump and is also simpler/cleaner. I tested this on the latest qemu and HiFive Unmatched and works as expected. Fixes: 2bfc6cd81bd1 ("riscv: Move kernel mapping outside of linear mapping") Signed-off-by: Nick Kossifidis Reviewed-by: Alexandre Ghiti Cc: stable@vger.kernel.org Signed-off-by: Palmer Dabbelt Signed-off-by: Greg Kroah-Hartman

RISC-V: Use common riscv_cpuid_to_hartid_mask() for both SMP=y and SMP=n

2022-01-27T10:02:50+00:00

commit 869c70609248102f3a2e95a39b6233ff6ea2c932 upstream. Use what is currently the SMP=y version of riscv_cpuid_to_hartid_mask() for both SMP=y and SMP=n to fix a build failure with KVM=m and SMP=n due to boot_cpu_hartid not being exported. This also fixes a second bug where the SMP=n version assumes the sole CPU in the system is in the incoming mask, which may not hold true in kvm_riscv_vcpu_sbi_ecall() if the KVM guest VM has multiple vCPUs (on a SMP=n system). Fixes: 1ef46c231df4 ("RISC-V: Implement new SBI v0.2 extensions") Reported-by: Adam Borowski Reviewed-by: Anup Patel Signed-off-by: Sean Christopherson Cc: stable@vger.kernel.org Signed-off-by: Palmer Dabbelt Signed-off-by: Greg Kroah-Hartman

perf: Protect perf_guest_cbs with RCU

2022-01-20T08:13:14+00:00

commit ff083a2d972f56bebfd82409ca62e5dfce950961 upstream. Protect perf_guest_cbs with RCU to fix multiple possible errors. Luckily, all paths that read perf_guest_cbs already require RCU protection, e.g. to protect the callback chains, so only the direct perf_guest_cbs touchpoints need to be modified. Bug #1 is a simple lack of WRITE_ONCE/READ_ONCE behavior to ensure perf_guest_cbs isn't reloaded between a !NULL check and a dereference. Fixed via the READ_ONCE() in rcu_dereference(). Bug #2 is that on weakly-ordered architectures, updates to the callbacks themselves are not guaranteed to be visible before the pointer is made visible to readers. Fixed by the smp_store_release() in rcu_assign_pointer() when the new pointer is non-NULL. Bug #3 is that, because the callbacks are global, it's possible for readers to run in parallel with an unregisters, and thus a module implementing the callbacks can be unloaded while readers are in flight, resulting in a use-after-free. Fixed by a synchronize_rcu() call when unregistering callbacks. Bug #1 escaped notice because it's extremely unlikely a compiler will reload perf_guest_cbs in this sequence. perf_guest_cbs does get reloaded for future derefs, e.g. for ->is_user_mode(), but the ->is_in_guest() guard all but guarantees the consumer will win the race, e.g. to nullify perf_guest_cbs, KVM has to completely exit the guest and teardown down all VMs before KVM start its module unload / unregister sequence. This also makes it all but impossible to encounter bug #3. Bug #2 has not been a problem because all architectures that register callbacks are strongly ordered and/or have a static set of callbacks. But with help, unloading kvm_intel can trigger bug #1 e.g. wrapping perf_guest_cbs with READ_ONCE in perf_misc_flags() while spamming kvm_intel module load/unload leads to: BUG: kernel NULL pointer dereference, address: 0000000000000000 #PF: supervisor read access in kernel mode #PF: error_code(0x0000) - not-present page PGD 0 P4D 0 Oops: 0000 [#1] PREEMPT SMP CPU: 6 PID: 1825 Comm: stress Not tainted 5.14.0-rc2+ #459 Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 0.0.0 02/06/2015 RIP: 0010:perf_misc_flags+0x1c/0x70 Call Trace: perf_prepare_sample+0x53/0x6b0 perf_event_output_forward+0x67/0x160 __perf_event_overflow+0x52/0xf0 handle_pmi_common+0x207/0x300 intel_pmu_handle_irq+0xcf/0x410 perf_event_nmi_handler+0x28/0x50 nmi_handle+0xc7/0x260 default_do_nmi+0x6b/0x170 exc_nmi+0x103/0x130 asm_exc_nmi+0x76/0xbf Fixes: 39447b386c84 ("perf: Enhance perf to allow for guest statistic collection from host") Signed-off-by: Sean Christopherson Signed-off-by: Peter Zijlstra (Intel) Reviewed-by: Paolo Bonzini Cc: stable@vger.kernel.org Link: https://lore.kernel.org/r/20211111020738.2512932-2-seanjc@google.com Signed-off-by: Greg Kroah-Hartman