From 4bbc84ffd137fe43d68aa633d317b0a96de8a828 Mon Sep 17 00:00:00 2001 From: Mike Kravetz Date: Mon, 19 Dec 2016 19:17:08 -0800 Subject: sparc: use symbolic names for tsb indexing Use symbolic names MM_TSB_BASE and MM_TSB_HUGE instead of numeric values 0 and 1 in __tsb_context_switch. Code cleanup only, no functional change. Suggested-by: Sam Ravnborg Signed-off-by: Mike Kravetz Acked-by: Sam Ravnborg Signed-off-by: David S. Miller --- arch/sparc/include/asm/mmu_context_64.h | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) (limited to 'arch') diff --git a/arch/sparc/include/asm/mmu_context_64.h b/arch/sparc/include/asm/mmu_context_64.h index b84be675e507..d0317993e947 100644 --- a/arch/sparc/include/asm/mmu_context_64.h +++ b/arch/sparc/include/asm/mmu_context_64.h @@ -35,15 +35,15 @@ void __tsb_context_switch(unsigned long pgd_pa, static inline void tsb_context_switch(struct mm_struct *mm) { __tsb_context_switch(__pa(mm->pgd), - &mm->context.tsb_block[0], + &mm->context.tsb_block[MM_TSB_BASE], #if defined(CONFIG_HUGETLB_PAGE) || defined(CONFIG_TRANSPARENT_HUGEPAGE) - (mm->context.tsb_block[1].tsb ? - &mm->context.tsb_block[1] : + (mm->context.tsb_block[MM_TSB_HUGE].tsb ? + &mm->context.tsb_block[MM_TSB_HUGE] : NULL) #else NULL #endif - , __pa(&mm->context.tsb_descr[0])); + , __pa(&mm->context.tsb_descr[MM_TSB_BASE])); } void tsb_grow(struct mm_struct *mm, -- cgit v1.2.3 From 5d0e7705774dd412a465896d08d59a81a345c1e4 Mon Sep 17 00:00:00 2001 From: Tom Hromatka Date: Tue, 10 Jan 2017 10:57:56 -0700 Subject: sparc: Fixed typo in sstate.c. Replaced panicing with panicking Signed-off-by: Tom Hromatka Signed-off-by: David S. Miller --- arch/sparc/kernel/sstate.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) (limited to 'arch') diff --git a/arch/sparc/kernel/sstate.c b/arch/sparc/kernel/sstate.c index c59af546f522..3caed4023589 100644 --- a/arch/sparc/kernel/sstate.c +++ b/arch/sparc/kernel/sstate.c @@ -43,8 +43,8 @@ static const char poweroff_msg[32] __attribute__((aligned(32))) = "Linux powering off"; static const char rebooting_msg[32] __attribute__((aligned(32))) = "Linux rebooting"; -static const char panicing_msg[32] __attribute__((aligned(32))) = - "Linux panicing"; +static const char panicking_msg[32] __attribute__((aligned(32))) = + "Linux panicking"; static int sstate_reboot_call(struct notifier_block *np, unsigned long type, void *_unused) { @@ -76,7 +76,7 @@ static struct notifier_block sstate_reboot_notifier = { static int sstate_panic_event(struct notifier_block *n, unsigned long event, void *ptr) { - do_set_sstate(HV_SOFT_STATE_TRANSITION, panicing_msg); + do_set_sstate(HV_SOFT_STATE_TRANSITION, panicking_msg); return NOTIFY_DONE; } -- cgit v1.2.3 From c26665ab5c49ad3e142e0f054ca3204f259ba09c Mon Sep 17 00:00:00 2001 From: Borislav Petkov Date: Fri, 20 Jan 2017 21:29:40 +0100 Subject: x86/microcode/intel: Drop stashed AP patch pointer optimization This was meant to save us the scanning of the microcode containter in the initrd since the first AP had already done that but it can also hurt us: Imagine a single hyperthreaded CPU (Intel(R) Atom(TM) CPU N270, for example) which updates the microcode on the BSP but since the microcode engine is shared between the two threads, the update on CPU1 doesn't happen because it has already happened on CPU0 and we don't find a newer microcode revision on CPU1. Which doesn't set the intel_ucode_patch pointer and at initrd jettisoning time we don't save the microcode patch for later application. Now, when we suspend to RAM, the loaded microcode gets cleared so we need to reload but there's no patch saved in the cache. Removing the optimization fixes this issue and all is fine and dandy. Fixes: 06b8534cb728 ("x86/microcode: Rework microcode loading") Signed-off-by: Borislav Petkov Reviewed-by: Thomas Gleixner Link: http://lkml.kernel.org/r/20170120202955.4091-2-bp@alien8.de Signed-off-by: Thomas Gleixner --- arch/x86/kernel/cpu/microcode/intel.c | 9 +-------- 1 file changed, 1 insertion(+), 8 deletions(-) (limited to 'arch') diff --git a/arch/x86/kernel/cpu/microcode/intel.c b/arch/x86/kernel/cpu/microcode/intel.c index 3f329b74e040..8325d8a09ab0 100644 --- a/arch/x86/kernel/cpu/microcode/intel.c +++ b/arch/x86/kernel/cpu/microcode/intel.c @@ -41,7 +41,7 @@ static const char ucode_path[] = "kernel/x86/microcode/GenuineIntel.bin"; -/* Current microcode patch used in early patching */ +/* Current microcode patch used in early patching on the APs. */ struct microcode_intel *intel_ucode_patch; static inline bool cpu_signatures_match(unsigned int s1, unsigned int p1, @@ -607,12 +607,6 @@ int __init save_microcode_in_initrd_intel(void) struct ucode_cpu_info uci; struct cpio_data cp; - /* - * AP loading didn't find any microcode patch, no need to save anything. - */ - if (!intel_ucode_patch || IS_ERR(intel_ucode_patch)) - return 0; - if (!load_builtin_intel_microcode(&cp)) cp = find_microcode_in_initrd(ucode_path, false); @@ -628,7 +622,6 @@ int __init save_microcode_in_initrd_intel(void) return 0; } - /* * @res_patch, output: a pointer to the patch we found. */ -- cgit v1.2.3 From 4c833368f0bf748d4147bf301b1f95bc8eccb3c0 Mon Sep 17 00:00:00 2001 From: Kevin Hao Date: Sun, 22 Jan 2017 16:50:23 +0800 Subject: x86/fpu: Set the xcomp_bv when we fake up a XSAVES area I got the following calltrace on a Apollo Lake SoC with 32-bit kernel: WARNING: CPU: 2 PID: 261 at arch/x86/include/asm/fpu/internal.h:363 fpu__restore+0x1f5/0x260 [...] Hardware name: Intel Corp. Broxton P/NOTEBOOK, BIOS APLIRVPA.X64.0138.B35.1608091058 08/09/2016 Call Trace: dump_stack() __warn() ? fpu__restore() warn_slowpath_null() fpu__restore() __fpu__restore_sig() fpu__restore_sig() restore_sigcontext.isra.9() sys_sigreturn() do_int80_syscall_32() entry_INT80_32() The reason is that a #GP occurs when executing XRSTORS. The root cause is that we forget to set the xcomp_bv when we fake up the XSAVES area in the copyin_to_xsaves() function. Signed-off-by: Kevin Hao Cc: Andy Lutomirski Cc: Borislav Petkov Cc: Brian Gerst Cc: Dave Hansen Cc: Denys Vlasenko Cc: Fenghua Yu Cc: H. Peter Anvin Cc: Josh Poimboeuf Cc: Linus Torvalds Cc: Oleg Nesterov Cc: Peter Zijlstra Cc: Quentin Casasnovas Cc: Rik van Riel Cc: Thomas Gleixner Cc: Yu-cheng Yu Link: http://lkml.kernel.org/r/1485075023-30161-1-git-send-email-haokexin@gmail.com Signed-off-by: Ingo Molnar Signed-off-by: Thomas Gleixner --- arch/x86/kernel/fpu/xstate.c | 1 + 1 file changed, 1 insertion(+) (limited to 'arch') diff --git a/arch/x86/kernel/fpu/xstate.c b/arch/x86/kernel/fpu/xstate.c index 1d7770447b3e..e287b9075527 100644 --- a/arch/x86/kernel/fpu/xstate.c +++ b/arch/x86/kernel/fpu/xstate.c @@ -1070,6 +1070,7 @@ int copyin_to_xsaves(const void *kbuf, const void __user *ubuf, * Add back in the features that came in from userspace: */ xsave->header.xfeatures |= xfeatures; + xsave->header.xcomp_bv = XCOMP_BV_COMPACTED_FORMAT | xsave->header.xfeatures; return 0; } -- cgit v1.2.3 From 11e3b725cfc282efe9d4a354153e99d86a16af08 Mon Sep 17 00:00:00 2001 From: Ard Biesheuvel Date: Tue, 17 Jan 2017 13:46:29 +0000 Subject: crypto: arm64/aes-blk - honour iv_out requirement in CBC and CTR modes Update the ARMv8 Crypto Extensions and the plain NEON AES implementations in CBC and CTR modes to return the next IV back to the skcipher API client. This is necessary for chaining to work correctly. Note that for CTR, this is only done if the request is a round multiple of the block size, since otherwise, chaining is impossible anyway. Cc: # v3.16+ Signed-off-by: Ard Biesheuvel Signed-off-by: Herbert Xu --- arch/arm64/crypto/aes-modes.S | 88 +++++++++++++++++++++---------------------- 1 file changed, 42 insertions(+), 46 deletions(-) (limited to 'arch') diff --git a/arch/arm64/crypto/aes-modes.S b/arch/arm64/crypto/aes-modes.S index c53dbeae79f2..838dad5c209f 100644 --- a/arch/arm64/crypto/aes-modes.S +++ b/arch/arm64/crypto/aes-modes.S @@ -193,15 +193,16 @@ AES_ENTRY(aes_cbc_encrypt) cbz w6, .Lcbcencloop ld1 {v0.16b}, [x5] /* get iv */ - enc_prepare w3, x2, x5 + enc_prepare w3, x2, x6 .Lcbcencloop: ld1 {v1.16b}, [x1], #16 /* get next pt block */ eor v0.16b, v0.16b, v1.16b /* ..and xor with iv */ - encrypt_block v0, w3, x2, x5, w6 + encrypt_block v0, w3, x2, x6, w7 st1 {v0.16b}, [x0], #16 subs w4, w4, #1 bne .Lcbcencloop + st1 {v0.16b}, [x5] /* return iv */ ret AES_ENDPROC(aes_cbc_encrypt) @@ -211,7 +212,7 @@ AES_ENTRY(aes_cbc_decrypt) cbz w6, .LcbcdecloopNx ld1 {v7.16b}, [x5] /* get iv */ - dec_prepare w3, x2, x5 + dec_prepare w3, x2, x6 .LcbcdecloopNx: #if INTERLEAVE >= 2 @@ -248,7 +249,7 @@ AES_ENTRY(aes_cbc_decrypt) .Lcbcdecloop: ld1 {v1.16b}, [x1], #16 /* get next ct block */ mov v0.16b, v1.16b /* ...and copy to v0 */ - decrypt_block v0, w3, x2, x5, w6 + decrypt_block v0, w3, x2, x6, w7 eor v0.16b, v0.16b, v7.16b /* xor with iv => pt */ mov v7.16b, v1.16b /* ct is next iv */ st1 {v0.16b}, [x0], #16 @@ -256,6 +257,7 @@ AES_ENTRY(aes_cbc_decrypt) bne .Lcbcdecloop .Lcbcdecout: FRAME_POP + st1 {v7.16b}, [x5] /* return iv */ ret AES_ENDPROC(aes_cbc_decrypt) @@ -267,24 +269,15 @@ AES_ENDPROC(aes_cbc_decrypt) AES_ENTRY(aes_ctr_encrypt) FRAME_PUSH - cbnz w6, .Lctrfirst /* 1st time around? */ - umov x5, v4.d[1] /* keep swabbed ctr in reg */ - rev x5, x5 -#if INTERLEAVE >= 2 - cmn w5, w4 /* 32 bit overflow? */ - bcs .Lctrinc - add x5, x5, #1 /* increment BE ctr */ - b .LctrincNx -#else - b .Lctrinc -#endif -.Lctrfirst: + cbz w6, .Lctrnotfirst /* 1st time around? */ enc_prepare w3, x2, x6 ld1 {v4.16b}, [x5] - umov x5, v4.d[1] /* keep swabbed ctr in reg */ - rev x5, x5 + +.Lctrnotfirst: + umov x8, v4.d[1] /* keep swabbed ctr in reg */ + rev x8, x8 #if INTERLEAVE >= 2 - cmn w5, w4 /* 32 bit overflow? */ + cmn w8, w4 /* 32 bit overflow? */ bcs .Lctrloop .LctrloopNx: subs w4, w4, #INTERLEAVE @@ -292,11 +285,11 @@ AES_ENTRY(aes_ctr_encrypt) #if INTERLEAVE == 2 mov v0.8b, v4.8b mov v1.8b, v4.8b - rev x7, x5 - add x5, x5, #1 + rev x7, x8 + add x8, x8, #1 ins v0.d[1], x7 - rev x7, x5 - add x5, x5, #1 + rev x7, x8 + add x8, x8, #1 ins v1.d[1], x7 ld1 {v2.16b-v3.16b}, [x1], #32 /* get 2 input blocks */ do_encrypt_block2x @@ -305,7 +298,7 @@ AES_ENTRY(aes_ctr_encrypt) st1 {v0.16b-v1.16b}, [x0], #32 #else ldr q8, =0x30000000200000001 /* addends 1,2,3[,0] */ - dup v7.4s, w5 + dup v7.4s, w8 mov v0.16b, v4.16b add v7.4s, v7.4s, v8.4s mov v1.16b, v4.16b @@ -323,18 +316,12 @@ AES_ENTRY(aes_ctr_encrypt) eor v2.16b, v7.16b, v2.16b eor v3.16b, v5.16b, v3.16b st1 {v0.16b-v3.16b}, [x0], #64 - add x5, x5, #INTERLEAVE + add x8, x8, #INTERLEAVE #endif - cbz w4, .LctroutNx -.LctrincNx: - rev x7, x5 + rev x7, x8 ins v4.d[1], x7 + cbz w4, .Lctrout b .LctrloopNx -.LctroutNx: - sub x5, x5, #1 - rev x7, x5 - ins v4.d[1], x7 - b .Lctrout .Lctr1x: adds w4, w4, #INTERLEAVE beq .Lctrout @@ -342,30 +329,39 @@ AES_ENTRY(aes_ctr_encrypt) .Lctrloop: mov v0.16b, v4.16b encrypt_block v0, w3, x2, x6, w7 + + adds x8, x8, #1 /* increment BE ctr */ + rev x7, x8 + ins v4.d[1], x7 + bcs .Lctrcarry /* overflow? */ + +.Lctrcarrydone: subs w4, w4, #1 bmi .Lctrhalfblock /* blocks < 0 means 1/2 block */ ld1 {v3.16b}, [x1], #16 eor v3.16b, v0.16b, v3.16b st1 {v3.16b}, [x0], #16 - beq .Lctrout -.Lctrinc: - adds x5, x5, #1 /* increment BE ctr */ - rev x7, x5 - ins v4.d[1], x7 - bcc .Lctrloop /* no overflow? */ - umov x7, v4.d[0] /* load upper word of ctr */ - rev x7, x7 /* ... to handle the carry */ - add x7, x7, #1 - rev x7, x7 - ins v4.d[0], x7 - b .Lctrloop + bne .Lctrloop + +.Lctrout: + st1 {v4.16b}, [x5] /* return next CTR value */ + FRAME_POP + ret + .Lctrhalfblock: ld1 {v3.8b}, [x1] eor v3.8b, v0.8b, v3.8b st1 {v3.8b}, [x0] -.Lctrout: FRAME_POP ret + +.Lctrcarry: + umov x7, v4.d[0] /* load upper word of ctr */ + rev x7, x7 /* ... to handle the carry */ + add x7, x7, #1 + rev x7, x7 + ins v4.d[0], x7 + b .Lctrcarrydone AES_ENDPROC(aes_ctr_encrypt) .ltorg -- cgit v1.2.3 From dffba9a31c7769be3231c420d4b364c92ba3f1ac Mon Sep 17 00:00:00 2001 From: Yu-cheng Yu Date: Mon, 23 Jan 2017 14:54:44 -0800 Subject: x86/fpu/xstate: Fix xcomp_bv in XSAVES header The compacted-format XSAVES area is determined at boot time and never changed after. The field xsave.header.xcomp_bv indicates which components are in the fixed XSAVES format. In fpstate_init() we did not set xcomp_bv to reflect the XSAVES format since at the time there is no valid data. However, after we do copy_init_fpstate_to_fpregs() in fpu__clear(), as in commit: b22cbe404a9c x86/fpu: Fix invalid FPU ptrace state after execve() and when __fpu_restore_sig() does fpu__restore() for a COMPAT-mode app, a #GP occurs. This can be easily triggered by doing valgrind on a COMPAT-mode "Hello World," as reported by Joakim Tjernlund and others: https://bugzilla.kernel.org/show_bug.cgi?id=190061 Fix it by setting xcomp_bv correctly. This patch also moves the xcomp_bv initialization to the proper place, which was in copyin_to_xsaves() as of: 4c833368f0bf x86/fpu: Set the xcomp_bv when we fake up a XSAVES area which fixed the bug too, but it's more efficient and cleaner to initialize things once per boot, not for every signal handling operation. Reported-by: Kevin Hao Reported-by: Joakim Tjernlund Signed-off-by: Yu-cheng Yu Cc: Andy Lutomirski Cc: Borislav Petkov Cc: Dave Hansen Cc: Fenghua Yu Cc: Linus Torvalds Cc: Peter Zijlstra Cc: Ravi V. Shankar Cc: Thomas Gleixner Cc: haokexin@gmail.com Link: http://lkml.kernel.org/r/1485212084-4418-1-git-send-email-yu-cheng.yu@intel.com [ Combined it with 4c833368f0bf. ] Signed-off-by: Ingo Molnar --- arch/x86/kernel/fpu/core.c | 4 +++- arch/x86/kernel/fpu/xstate.c | 1 - 2 files changed, 3 insertions(+), 2 deletions(-) (limited to 'arch') diff --git a/arch/x86/kernel/fpu/core.c b/arch/x86/kernel/fpu/core.c index e4e97a5355ce..de7234401275 100644 --- a/arch/x86/kernel/fpu/core.c +++ b/arch/x86/kernel/fpu/core.c @@ -9,6 +9,7 @@ #include #include #include +#include #include #include @@ -183,7 +184,8 @@ void fpstate_init(union fpregs_state *state) * it will #GP. Make sure it is replaced after the memset(). */ if (static_cpu_has(X86_FEATURE_XSAVES)) - state->xsave.header.xcomp_bv = XCOMP_BV_COMPACTED_FORMAT; + state->xsave.header.xcomp_bv = XCOMP_BV_COMPACTED_FORMAT | + xfeatures_mask; if (static_cpu_has(X86_FEATURE_FXSR)) fpstate_init_fxstate(&state->fxsave); diff --git a/arch/x86/kernel/fpu/xstate.c b/arch/x86/kernel/fpu/xstate.c index e287b9075527..1d7770447b3e 100644 --- a/arch/x86/kernel/fpu/xstate.c +++ b/arch/x86/kernel/fpu/xstate.c @@ -1070,7 +1070,6 @@ int copyin_to_xsaves(const void *kbuf, const void __user *ubuf, * Add back in the features that came in from userspace: */ xsave->header.xfeatures |= xfeatures; - xsave->header.xcomp_bv = XCOMP_BV_COMPACTED_FORMAT | xsave->header.xfeatures; return 0; } -- cgit v1.2.3 From bf29bddf0417a4783da3b24e8c9e017ac649326f Mon Sep 17 00:00:00 2001 From: Jiri Kosina Date: Fri, 27 Jan 2017 22:25:52 +0000 Subject: x86/efi: Always map the first physical page into the EFI pagetables Commit: 129766708 ("x86/efi: Only map RAM into EFI page tables if in mixed-mode") stopped creating 1:1 mappings for all RAM, when running in native 64-bit mode. It turns out though that there are 64-bit EFI implementations in the wild (this particular problem has been reported on a Lenovo Yoga 710-11IKB), which still make use of the first physical page for their own private use, even though they explicitly mark it EFI_CONVENTIONAL_MEMORY in the memory map. In case there is no mapping for this particular frame in the EFI pagetables, as soon as firmware tries to make use of it, a triple fault occurs and the system reboots (in case of the Yoga 710-11IKB this is very early during bootup). Fix that by always mapping the first page of physical memory into the EFI pagetables. We're free to hand this page to the BIOS, as trim_bios_range() will reserve the first page and isolate it away from memory allocators anyway. Note that just reverting 129766708 alone is not enough on v4.9-rc1+ to fix the regression on affected hardware, as this commit: ab72a27da ("x86/efi: Consolidate region mapping logic") later made the first physical frame not to be mapped anyway. Reported-by: Hanka Pavlikova Signed-off-by: Jiri Kosina Signed-off-by: Matt Fleming Cc: Ard Biesheuvel Cc: Borislav Petkov Cc: Borislav Petkov Cc: Laura Abbott Cc: Linus Torvalds Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: Vojtech Pavlik Cc: Waiman Long Cc: linux-efi@vger.kernel.org Cc: stable@kernel.org # v4.8+ Fixes: 129766708 ("x86/efi: Only map RAM into EFI page tables if in mixed-mode") Link: http://lkml.kernel.org/r/20170127222552.22336-1-matt@codeblueprint.co.uk [ Tidied up the changelog and the comment. ] Signed-off-by: Ingo Molnar --- arch/x86/platform/efi/efi_64.c | 16 ++++++++++++++++ 1 file changed, 16 insertions(+) (limited to 'arch') diff --git a/arch/x86/platform/efi/efi_64.c b/arch/x86/platform/efi/efi_64.c index 319148bd4b05..2f25a363068c 100644 --- a/arch/x86/platform/efi/efi_64.c +++ b/arch/x86/platform/efi/efi_64.c @@ -268,6 +268,22 @@ int __init efi_setup_page_tables(unsigned long pa_memmap, unsigned num_pages) efi_scratch.use_pgd = true; + /* + * Certain firmware versions are way too sentimential and still believe + * they are exclusive and unquestionable owners of the first physical page, + * even though they explicitly mark it as EFI_CONVENTIONAL_MEMORY + * (but then write-access it later during SetVirtualAddressMap()). + * + * Create a 1:1 mapping for this page, to avoid triple faults during early + * boot with such firmware. We are free to hand this page to the BIOS, + * as trim_bios_range() will reserve the first page and isolate it away + * from memory allocators anyway. + */ + if (kernel_map_pages_in_pgd(pgd, 0x0, 0x0, 1, _PAGE_RW)) { + pr_err("Failed to create 1:1 mapping for the first page!\n"); + return 1; + } + /* * When making calls to the firmware everything needs to be 1:1 * mapped and addressable with 32-bit pointers. Map the kernel -- cgit v1.2.3 From 24c2503255d35c269b67162c397a1a1c1e02f6ce Mon Sep 17 00:00:00 2001 From: Borislav Petkov Date: Wed, 25 Jan 2017 21:00:48 +0100 Subject: x86/microcode: Do not access the initrd after it has been freed When we look for microcode blobs, we first try builtin and if that doesn't succeed, we fallback to the initrd supplied to the kernel. However, at some point doing boot, that initrd gets jettisoned and we shouldn't access it anymore. But we do, as the below KASAN report shows. That's because find_microcode_in_initrd() doesn't check whether the initrd is still valid or not. So do that. ================================================================== BUG: KASAN: use-after-free in find_cpio_data Read of size 1 by task swapper/1/0 page:ffffea0000db9d40 count:0 mapcount:0 mapping: (null) index:0x1 flags: 0x100000000000000() raw: 0100000000000000 0000000000000000 0000000000000001 00000000ffffffff raw: dead000000000100 dead000000000200 0000000000000000 0000000000000000 page dumped because: kasan: bad access detected CPU: 1 PID: 0 Comm: swapper/1 Tainted: G W 4.10.0-rc5-debug-00075-g2dbde22 #3 Hardware name: Dell Inc. XPS 13 9360/0839Y6, BIOS 1.2.3 12/01/2016 Call Trace: dump_stack ? _atomic_dec_and_lock ? __dump_page kasan_report_error ? pointer ? find_cpio_data __asan_report_load1_noabort ? find_cpio_data find_cpio_data ? vsprintf ? dump_stack ? get_ucode_user ? print_usage_bug find_microcode_in_initrd __load_ucode_intel ? collect_cpu_info_early ? debug_check_no_locks_freed load_ucode_intel_ap ? collect_cpu_info ? trace_hardirqs_on ? flat_send_IPI_mask_allbutself load_ucode_ap ? get_builtin_firmware ? flush_tlb_func ? do_raw_spin_trylock ? cpumask_weight cpu_init ? trace_hardirqs_off ? play_dead_common ? native_play_dead ? hlt_play_dead ? syscall_init ? arch_cpu_idle_dead ? do_idle start_secondary start_cpu Memory state around the buggy address: ffff880036e74f00: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ffff880036e74f80: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff >ffff880036e75000: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ^ ffff880036e75080: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ffff880036e75100: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ================================================================== Reported-by: Andrey Ryabinin Tested-by: Andrey Ryabinin Signed-off-by: Borislav Petkov Cc: Linus Torvalds Cc: Peter Zijlstra Cc: Thomas Gleixner Link: http://lkml.kernel.org/r/20170126165833.evjemhbqzaepirxo@pd.tnic Signed-off-by: Ingo Molnar --- arch/x86/include/asm/microcode.h | 1 + arch/x86/kernel/cpu/microcode/amd.c | 5 +++-- arch/x86/kernel/cpu/microcode/core.c | 22 +++++++++++++++++----- 3 files changed, 21 insertions(+), 7 deletions(-) (limited to 'arch') diff --git a/arch/x86/include/asm/microcode.h b/arch/x86/include/asm/microcode.h index 38711df3bcb5..2266f864b747 100644 --- a/arch/x86/include/asm/microcode.h +++ b/arch/x86/include/asm/microcode.h @@ -140,6 +140,7 @@ extern void __init load_ucode_bsp(void); extern void load_ucode_ap(void); void reload_early_microcode(void); extern bool get_builtin_firmware(struct cpio_data *cd, const char *name); +extern bool initrd_gone; #else static inline int __init microcode_init(void) { return 0; }; static inline void __init load_ucode_bsp(void) { } diff --git a/arch/x86/kernel/cpu/microcode/amd.c b/arch/x86/kernel/cpu/microcode/amd.c index 6a31e2691f3a..079e81733a58 100644 --- a/arch/x86/kernel/cpu/microcode/amd.c +++ b/arch/x86/kernel/cpu/microcode/amd.c @@ -384,8 +384,9 @@ void load_ucode_amd_ap(unsigned int family) reget: if (!get_builtin_microcode(&cp, family)) { #ifdef CONFIG_BLK_DEV_INITRD - cp = find_cpio_data(ucode_path, (void *)initrd_start, - initrd_end - initrd_start, NULL); + if (!initrd_gone) + cp = find_cpio_data(ucode_path, (void *)initrd_start, + initrd_end - initrd_start, NULL); #endif if (!(cp.data && cp.size)) { /* diff --git a/arch/x86/kernel/cpu/microcode/core.c b/arch/x86/kernel/cpu/microcode/core.c index 2af69d27da62..73102d932760 100644 --- a/arch/x86/kernel/cpu/microcode/core.c +++ b/arch/x86/kernel/cpu/microcode/core.c @@ -46,6 +46,8 @@ static struct microcode_ops *microcode_ops; static bool dis_ucode_ldr = true; +bool initrd_gone; + LIST_HEAD(microcode_cache); /* @@ -190,21 +192,24 @@ void load_ucode_ap(void) static int __init save_microcode_in_initrd(void) { struct cpuinfo_x86 *c = &boot_cpu_data; + int ret = -EINVAL; switch (c->x86_vendor) { case X86_VENDOR_INTEL: if (c->x86 >= 6) - return save_microcode_in_initrd_intel(); + ret = save_microcode_in_initrd_intel(); break; case X86_VENDOR_AMD: if (c->x86 >= 0x10) - return save_microcode_in_initrd_amd(c->x86); + ret = save_microcode_in_initrd_amd(c->x86); break; default: break; } - return -EINVAL; + initrd_gone = true; + + return ret; } struct cpio_data find_microcode_in_initrd(const char *path, bool use_pa) @@ -247,9 +252,16 @@ struct cpio_data find_microcode_in_initrd(const char *path, bool use_pa) * has the virtual address of the beginning of the initrd. It also * possibly relocates the ramdisk. In either case, initrd_start contains * the updated address so use that instead. + * + * initrd_gone is for the hotplug case where we've thrown out initrd + * already. */ - if (!use_pa && initrd_start) - start = initrd_start; + if (!use_pa) { + if (initrd_gone) + return (struct cpio_data){ NULL, 0, "" }; + if (initrd_start) + start = initrd_start; + } return find_cpio_data(path, (void *)start, size, NULL); #else /* !CONFIG_BLK_DEV_INITRD */ -- cgit v1.2.3 From 7a7dc961a28b965a0d0303c2e989df17b411708b Mon Sep 17 00:00:00 2001 From: "Liam R. Howlett" Date: Tue, 17 Jan 2017 10:59:02 -0500 Subject: sparc64: Zero pages on allocation for mondo and error queues. Error queues use a non-zero first word to detect if the queues are full. Using pages that have not been zeroed may result in false positive overflow events. These queues are set up once during boot so zeroing all mondo and error queue pages is safe. Note that the false positive overflow does not always occur because the page allocation for these queues is so early in the boot cycle that higher number CPUs get fresh pages. It is only when traps are serviced with lower number CPUs who were given already used pages that this issue is exposed. Signed-off-by: Liam R. Howlett Signed-off-by: David S. Miller --- arch/sparc/kernel/irq_64.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch') diff --git a/arch/sparc/kernel/irq_64.c b/arch/sparc/kernel/irq_64.c index 34a7930b76ef..baed4cdeda75 100644 --- a/arch/sparc/kernel/irq_64.c +++ b/arch/sparc/kernel/irq_64.c @@ -1021,7 +1021,7 @@ static void __init alloc_one_queue(unsigned long *pa_ptr, unsigned long qmask) unsigned long order = get_order(size); unsigned long p; - p = __get_free_pages(GFP_KERNEL, order); + p = __get_free_pages(GFP_KERNEL | __GFP_ZERO, order); if (!p) { prom_printf("SUN4V: Error, cannot allocate queue.\n"); prom_halt(); -- cgit v1.2.3 From 047487241ff59374fded8c477f21453681f5995c Mon Sep 17 00:00:00 2001 From: "Liam R. Howlett" Date: Tue, 17 Jan 2017 10:59:03 -0500 Subject: sparc64: Handle PIO & MEM non-resumable errors. User processes trying to access an invalid memory address via PIO will receive a SIGBUS signal instead of causing a panic. Memory errors will receive a SIGKILL since a SIGBUS may result in a coredump which may attempt to repeat the faulting access. Signed-off-by: Liam R. Howlett Signed-off-by: David S. Miller --- arch/sparc/kernel/traps_64.c | 73 ++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 73 insertions(+) (limited to 'arch') diff --git a/arch/sparc/kernel/traps_64.c b/arch/sparc/kernel/traps_64.c index 496fa926e1e0..d44fb806bbd7 100644 --- a/arch/sparc/kernel/traps_64.c +++ b/arch/sparc/kernel/traps_64.c @@ -2051,6 +2051,73 @@ void sun4v_resum_overflow(struct pt_regs *regs) atomic_inc(&sun4v_resum_oflow_cnt); } +/* Given a set of registers, get the virtual addressi that was being accessed + * by the faulting instructions at tpc. + */ +static unsigned long sun4v_get_vaddr(struct pt_regs *regs) +{ + unsigned int insn; + + if (!copy_from_user(&insn, (void __user *)regs->tpc, 4)) { + return compute_effective_address(regs, insn, + (insn >> 25) & 0x1f); + } + return 0; +} + +/* Attempt to handle non-resumable errors generated from userspace. + * Returns true if the signal was handled, false otherwise. + */ +bool sun4v_nonresum_error_user_handled(struct pt_regs *regs, + struct sun4v_error_entry *ent) { + + unsigned int attrs = ent->err_attrs; + + if (attrs & SUN4V_ERR_ATTRS_MEMORY) { + unsigned long addr = ent->err_raddr; + siginfo_t info; + + if (addr == ~(u64)0) { + /* This seems highly unlikely to ever occur */ + pr_emerg("SUN4V NON-RECOVERABLE ERROR: Memory error detected in unknown location!\n"); + } else { + unsigned long page_cnt = DIV_ROUND_UP(ent->err_size, + PAGE_SIZE); + + /* Break the unfortunate news. */ + pr_emerg("SUN4V NON-RECOVERABLE ERROR: Memory failed at %016lX\n", + addr); + pr_emerg("SUN4V NON-RECOVERABLE ERROR: Claiming %lu ages.\n", + page_cnt); + + while (page_cnt-- > 0) { + if (pfn_valid(addr >> PAGE_SHIFT)) + get_page(pfn_to_page(addr >> PAGE_SHIFT)); + addr += PAGE_SIZE; + } + } + info.si_signo = SIGKILL; + info.si_errno = 0; + info.si_trapno = 0; + force_sig_info(info.si_signo, &info, current); + + return true; + } + if (attrs & SUN4V_ERR_ATTRS_PIO) { + siginfo_t info; + + info.si_signo = SIGBUS; + info.si_code = BUS_ADRERR; + info.si_addr = (void __user *)sun4v_get_vaddr(regs); + force_sig_info(info.si_signo, &info, current); + + return true; + } + + /* Default to doing nothing */ + return false; +} + /* We run with %pil set to PIL_NORMAL_MAX and PSTATE_IE enabled in %pstate. * Log the event, clear the first word of the entry, and die. */ @@ -2075,6 +2142,12 @@ void sun4v_nonresum_error(struct pt_regs *regs, unsigned long offset) put_cpu(); + if (!(regs->tstate & TSTATE_PRIV) && + sun4v_nonresum_error_user_handled(regs, &local_copy)) { + /* DON'T PANIC: This userspace error was handled. */ + return; + } + #ifdef CONFIG_PCI /* Check for the special PCI poke sequence. */ if (pci_poke_in_progress && pci_poke_cpu == cpu) { -- cgit v1.2.3 From 0becc0ae5b42828785b589f686725ff5bc3b9b25 Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Tue, 31 Jan 2017 09:37:34 +0100 Subject: x86/mce: Make timer handling more robust Erik reported that on a preproduction hardware a CMCI storm triggers the BUG_ON in add_timer_on(). The reason is that the per CPU MCE timer is started by the CMCI logic before the MCE CPU hotplug callback starts the timer with add_timer_on(). So the timer is already queued which triggers the BUG. Using add_timer_on() is pretty pointless in this code because the timer is strictlty per CPU, initialized as pinned and all operations which arm the timer happen on the CPU to which the timer belongs. Simplify the whole machinery by using mod_timer() instead of add_timer_on() which avoids the problem because mod_timer() can handle already queued timers. Use __start_timer() everywhere so the earliest armed expiry time is preserved. Reported-by: Erik Veijola Tested-by: Borislav Petkov Signed-off-by: Thomas Gleixner Reviewed-by: Borislav Petkov Cc: Tony Luck Link: http://lkml.kernel.org/r/alpine.DEB.2.20.1701310936080.3457@nanos Signed-off-by: Thomas Gleixner --- arch/x86/kernel/cpu/mcheck/mce.c | 31 ++++++++++++------------------- 1 file changed, 12 insertions(+), 19 deletions(-) (limited to 'arch') diff --git a/arch/x86/kernel/cpu/mcheck/mce.c b/arch/x86/kernel/cpu/mcheck/mce.c index 00ef43233e03..537c6647d84c 100644 --- a/arch/x86/kernel/cpu/mcheck/mce.c +++ b/arch/x86/kernel/cpu/mcheck/mce.c @@ -1373,20 +1373,15 @@ static unsigned long mce_adjust_timer_default(unsigned long interval) static unsigned long (*mce_adjust_timer)(unsigned long interval) = mce_adjust_timer_default; -static void __restart_timer(struct timer_list *t, unsigned long interval) +static void __start_timer(struct timer_list *t, unsigned long interval) { unsigned long when = jiffies + interval; unsigned long flags; local_irq_save(flags); - if (timer_pending(t)) { - if (time_before(when, t->expires)) - mod_timer(t, when); - } else { - t->expires = round_jiffies(when); - add_timer_on(t, smp_processor_id()); - } + if (!timer_pending(t) || time_before(when, t->expires)) + mod_timer(t, round_jiffies(when)); local_irq_restore(flags); } @@ -1421,7 +1416,7 @@ static void mce_timer_fn(unsigned long data) done: __this_cpu_write(mce_next_interval, iv); - __restart_timer(t, iv); + __start_timer(t, iv); } /* @@ -1432,7 +1427,7 @@ void mce_timer_kick(unsigned long interval) struct timer_list *t = this_cpu_ptr(&mce_timer); unsigned long iv = __this_cpu_read(mce_next_interval); - __restart_timer(t, interval); + __start_timer(t, interval); if (interval < iv) __this_cpu_write(mce_next_interval, interval); @@ -1779,17 +1774,15 @@ static void __mcheck_cpu_clear_vendor(struct cpuinfo_x86 *c) } } -static void mce_start_timer(unsigned int cpu, struct timer_list *t) +static void mce_start_timer(struct timer_list *t) { unsigned long iv = check_interval * HZ; if (mca_cfg.ignore_ce || !iv) return; - per_cpu(mce_next_interval, cpu) = iv; - - t->expires = round_jiffies(jiffies + iv); - add_timer_on(t, cpu); + this_cpu_write(mce_next_interval, iv); + __start_timer(t, iv); } static void __mcheck_cpu_setup_timer(void) @@ -1806,7 +1799,7 @@ static void __mcheck_cpu_init_timer(void) unsigned int cpu = smp_processor_id(); setup_pinned_timer(t, mce_timer_fn, cpu); - mce_start_timer(cpu, t); + mce_start_timer(t); } /* Handle unconfigured int18 (should never happen) */ @@ -2566,7 +2559,7 @@ static int mce_cpu_dead(unsigned int cpu) static int mce_cpu_online(unsigned int cpu) { - struct timer_list *t = &per_cpu(mce_timer, cpu); + struct timer_list *t = this_cpu_ptr(&mce_timer); int ret; mce_device_create(cpu); @@ -2577,13 +2570,13 @@ static int mce_cpu_online(unsigned int cpu) return ret; } mce_reenable_cpu(); - mce_start_timer(cpu, t); + mce_start_timer(t); return 0; } static int mce_cpu_pre_down(unsigned int cpu) { - struct timer_list *t = &per_cpu(mce_timer, cpu); + struct timer_list *t = this_cpu_ptr(&mce_timer); mce_disable_cpu(); del_timer_sync(t); -- cgit v1.2.3 From 4b3e6f2ef3722f1a6a97b6034ed492c1a21fd4ae Mon Sep 17 00:00:00 2001 From: Max Filippov Date: Tue, 31 Jan 2017 18:35:37 -0800 Subject: xtensa: fix noMMU build on cores with MMU Commit bf15f86b343ed8 ("xtensa: initialize MMU before jumping to reset vector") calls MMU management functions even when CONFIG_MMU is not selected. That breaks noMMU build on cores with MMU. Don't manage MMU when CONFIG_MMU is not selected. Cc: stable@vger.kernel.org Signed-off-by: Max Filippov --- arch/xtensa/kernel/setup.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch') diff --git a/arch/xtensa/kernel/setup.c b/arch/xtensa/kernel/setup.c index 88a044af7504..32cdc2c52e98 100644 --- a/arch/xtensa/kernel/setup.c +++ b/arch/xtensa/kernel/setup.c @@ -540,7 +540,7 @@ subsys_initcall(topology_init); void cpu_reset(void) { -#if XCHAL_HAVE_PTP_MMU +#if XCHAL_HAVE_PTP_MMU && IS_ENABLED(CONFIG_MMU) local_irq_disable(); /* * We have full MMU: all autoload ways, ways 7, 8 and 9 of DTLB must -- cgit v1.2.3 From dd86e373e09fb16b83e8adf5c48c421a4ca76468 Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Tue, 31 Jan 2017 23:58:38 +0100 Subject: perf/x86/intel/rapl: Make package handling more robust The package management code in RAPL relies on package mapping being available before a CPU is started. This changed with: 9d85eb9119f4 ("x86/smpboot: Make logical package management more robust") because the ACPI/BIOS information turned out to be unreliable, but that left RAPL in broken state. This was not noticed because on a regular boot all CPUs are online before RAPL is initialized. A possible fix would be to reintroduce the mess which allocates a package data structure in CPU prepare and when it turns out to already exist in starting throw it away later in the CPU online callback. But that's a horrible hack and not required at all because RAPL becomes functional for perf only in the CPU online callback. That's correct because user space is not yet informed about the CPU being onlined, so nothing caan rely on RAPL being available on that particular CPU. Move the allocation to the CPU online callback and simplify the hotplug handling. At this point the package mapping is established and correct. This also adds a missing check for available package data in the event_init() function. Reported-by: Yasuaki Ishimatsu Signed-off-by: Thomas Gleixner Cc: Alexander Shishkin Cc: Arnaldo Carvalho de Melo Cc: Jiri Olsa Cc: Linus Torvalds Cc: Peter Zijlstra Cc: Sebastian Siewior Cc: Stephane Eranian Cc: Vince Weaver Fixes: 9d85eb9119f4 ("x86/smpboot: Make logical package management more robust") Link: http://lkml.kernel.org/r/20170131230141.212593966@linutronix.de Signed-off-by: Ingo Molnar --- arch/x86/events/intel/rapl.c | 60 +++++++++++++++++++------------------------- 1 file changed, 26 insertions(+), 34 deletions(-) (limited to 'arch') diff --git a/arch/x86/events/intel/rapl.c b/arch/x86/events/intel/rapl.c index 17c3564d087a..22ef4f72cf32 100644 --- a/arch/x86/events/intel/rapl.c +++ b/arch/x86/events/intel/rapl.c @@ -161,7 +161,13 @@ static u64 rapl_timer_ms; static inline struct rapl_pmu *cpu_to_rapl_pmu(unsigned int cpu) { - return rapl_pmus->pmus[topology_logical_package_id(cpu)]; + unsigned int pkgid = topology_logical_package_id(cpu); + + /* + * The unsigned check also catches the '-1' return value for non + * existent mappings in the topology map. + */ + return pkgid < rapl_pmus->maxpkg ? rapl_pmus->pmus[pkgid] : NULL; } static inline u64 rapl_read_counter(struct perf_event *event) @@ -402,6 +408,8 @@ static int rapl_pmu_event_init(struct perf_event *event) /* must be done before validate_group */ pmu = cpu_to_rapl_pmu(event->cpu); + if (!pmu) + return -EINVAL; event->cpu = pmu->cpu; event->pmu_private = pmu; event->hw.event_base = msr; @@ -585,6 +593,20 @@ static int rapl_cpu_online(unsigned int cpu) struct rapl_pmu *pmu = cpu_to_rapl_pmu(cpu); int target; + if (!pmu) { + pmu = kzalloc_node(sizeof(*pmu), GFP_KERNEL, cpu_to_node(cpu)); + if (!pmu) + return -ENOMEM; + + raw_spin_lock_init(&pmu->lock); + INIT_LIST_HEAD(&pmu->active_list); + pmu->pmu = &rapl_pmus->pmu; + pmu->timer_interval = ms_to_ktime(rapl_timer_ms); + rapl_hrtimer_init(pmu); + + rapl_pmus->pmus[topology_logical_package_id(cpu)] = pmu; + } + /* * Check if there is an online cpu in the package which collects rapl * events already. @@ -598,27 +620,6 @@ static int rapl_cpu_online(unsigned int cpu) return 0; } -static int rapl_cpu_prepare(unsigned int cpu) -{ - struct rapl_pmu *pmu = cpu_to_rapl_pmu(cpu); - - if (pmu) - return 0; - - pmu = kzalloc_node(sizeof(*pmu), GFP_KERNEL, cpu_to_node(cpu)); - if (!pmu) - return -ENOMEM; - - raw_spin_lock_init(&pmu->lock); - INIT_LIST_HEAD(&pmu->active_list); - pmu->pmu = &rapl_pmus->pmu; - pmu->timer_interval = ms_to_ktime(rapl_timer_ms); - pmu->cpu = -1; - rapl_hrtimer_init(pmu); - rapl_pmus->pmus[topology_logical_package_id(cpu)] = pmu; - return 0; -} - static int rapl_check_hw_unit(bool apply_quirk) { u64 msr_rapl_power_unit_bits; @@ -803,29 +804,21 @@ static int __init rapl_pmu_init(void) /* * Install callbacks. Core will call them for each online cpu. */ - - ret = cpuhp_setup_state(CPUHP_PERF_X86_RAPL_PREP, "perf/x86/rapl:prepare", - rapl_cpu_prepare, NULL); - if (ret) - goto out; - ret = cpuhp_setup_state(CPUHP_AP_PERF_X86_RAPL_ONLINE, "perf/x86/rapl:online", rapl_cpu_online, rapl_cpu_offline); if (ret) - goto out1; + goto out; ret = perf_pmu_register(&rapl_pmus->pmu, "power", -1); if (ret) - goto out2; + goto out1; rapl_advertise(); return 0; -out2: - cpuhp_remove_state(CPUHP_AP_PERF_X86_RAPL_ONLINE); out1: - cpuhp_remove_state(CPUHP_PERF_X86_RAPL_PREP); + cpuhp_remove_state(CPUHP_AP_PERF_X86_RAPL_ONLINE); out: pr_warn("Initialization failed (%d), disabled\n", ret); cleanup_rapl_pmus(); @@ -836,7 +829,6 @@ module_init(rapl_pmu_init); static void __exit intel_rapl_exit(void) { cpuhp_remove_state_nocalls(CPUHP_AP_PERF_X86_RAPL_ONLINE); - cpuhp_remove_state_nocalls(CPUHP_PERF_X86_RAPL_PREP); perf_pmu_unregister(&rapl_pmus->pmu); cleanup_rapl_pmus(); } -- cgit v1.2.3 From 1aa6cfd33df492939b0be15ebdbcff1f8ae5ddb6 Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Tue, 31 Jan 2017 23:58:39 +0100 Subject: perf/x86/intel/uncore: Clean up hotplug conversion fallout The recent conversion to the hotplug state machine kept two mechanisms from the original code: 1) The first_init logic which adds the number of online CPUs in a package to the refcount. That's wrong because the callbacks are executed for all online CPUs. Remove it so the refcounting is correct. 2) The on_each_cpu() call to undo box->init() in the error handling path. That's bogus because when the prepare callback fails no box has been initialized yet. Remove it. Signed-off-by: Thomas Gleixner Cc: Alexander Shishkin Cc: Arnaldo Carvalho de Melo Cc: Jiri Olsa Cc: Linus Torvalds Cc: Peter Zijlstra Cc: Sebastian Siewior Cc: Stephane Eranian Cc: Vince Weaver Cc: Yasuaki Ishimatsu Fixes: 1a246b9f58c6 ("perf/x86/intel/uncore: Convert to hotplug state machine") Link: http://lkml.kernel.org/r/20170131230141.298032324@linutronix.de Signed-off-by: Ingo Molnar --- arch/x86/events/intel/uncore.c | 44 ++++-------------------------------------- 1 file changed, 4 insertions(+), 40 deletions(-) (limited to 'arch') diff --git a/arch/x86/events/intel/uncore.c b/arch/x86/events/intel/uncore.c index 8c4ccdc3a3f3..56c5235dcc29 100644 --- a/arch/x86/events/intel/uncore.c +++ b/arch/x86/events/intel/uncore.c @@ -764,30 +764,6 @@ static void uncore_pmu_unregister(struct intel_uncore_pmu *pmu) pmu->registered = false; } -static void __uncore_exit_boxes(struct intel_uncore_type *type, int cpu) -{ - struct intel_uncore_pmu *pmu = type->pmus; - struct intel_uncore_box *box; - int i, pkg; - - if (pmu) { - pkg = topology_physical_package_id(cpu); - for (i = 0; i < type->num_boxes; i++, pmu++) { - box = pmu->boxes[pkg]; - if (box) - uncore_box_exit(box); - } - } -} - -static void uncore_exit_boxes(void *dummy) -{ - struct intel_uncore_type **types; - - for (types = uncore_msr_uncores; *types; types++) - __uncore_exit_boxes(*types++, smp_processor_id()); -} - static void uncore_free_boxes(struct intel_uncore_pmu *pmu) { int pkg; @@ -1078,22 +1054,12 @@ static int uncore_cpu_dying(unsigned int cpu) return 0; } -static int first_init; - static int uncore_cpu_starting(unsigned int cpu) { struct intel_uncore_type *type, **types = uncore_msr_uncores; struct intel_uncore_pmu *pmu; struct intel_uncore_box *box; - int i, pkg, ncpus = 1; - - if (first_init) { - /* - * On init we get the number of online cpus in the package - * and set refcount for all of them. - */ - ncpus = cpumask_weight(topology_core_cpumask(cpu)); - } + int i, pkg; pkg = topology_logical_package_id(cpu); for (; *types; types++) { @@ -1104,7 +1070,7 @@ static int uncore_cpu_starting(unsigned int cpu) if (!box) continue; /* The first cpu on a package activates the box */ - if (atomic_add_return(ncpus, &box->refcnt) == ncpus) + if (atomic_inc_return(&box->refcnt) == 1) uncore_box_init(box); } } @@ -1408,19 +1374,17 @@ static int __init intel_uncore_init(void) "perf/x86/intel/uncore:prepare", uncore_cpu_prepare, NULL); } - first_init = 1; + cpuhp_setup_state(CPUHP_AP_PERF_X86_UNCORE_STARTING, "perf/x86/uncore:starting", uncore_cpu_starting, uncore_cpu_dying); - first_init = 0; + cpuhp_setup_state(CPUHP_AP_PERF_X86_UNCORE_ONLINE, "perf/x86/uncore:online", uncore_event_cpu_online, uncore_event_cpu_offline); return 0; err: - /* Undo box->init_box() */ - on_each_cpu_mask(&uncore_cpu_mask, uncore_exit_boxes, NULL, 1); uncore_types_exit(uncore_msr_uncores); uncore_pci_exit(); return ret; -- cgit v1.2.3 From fff4b87e594ad3d2e4f51e8d3d86a6f9d3d8b654 Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Tue, 31 Jan 2017 23:58:40 +0100 Subject: perf/x86/intel/uncore: Make package handling more robust The package management code in uncore relies on package mapping being available before a CPU is started. This changed with: 9d85eb9119f4 ("x86/smpboot: Make logical package management more robust") because the ACPI/BIOS information turned out to be unreliable, but that left uncore in broken state. This was not noticed because on a regular boot all CPUs are online before uncore is initialized. Move the allocation to the CPU online callback and simplify the hotplug handling. At this point the package mapping is established and correct. Signed-off-by: Thomas Gleixner Cc: Alexander Shishkin Cc: Arnaldo Carvalho de Melo Cc: Jiri Olsa Cc: Linus Torvalds Cc: Peter Zijlstra Cc: Sebastian Siewior Cc: Stephane Eranian Cc: Vince Weaver Cc: Yasuaki Ishimatsu Fixes: 9d85eb9119f4 ("x86/smpboot: Make logical package management more robust") Link: http://lkml.kernel.org/r/20170131230141.377156255@linutronix.de Signed-off-by: Ingo Molnar --- arch/x86/events/intel/uncore.c | 196 +++++++++++++++++++---------------------- 1 file changed, 91 insertions(+), 105 deletions(-) (limited to 'arch') diff --git a/arch/x86/events/intel/uncore.c b/arch/x86/events/intel/uncore.c index 56c5235dcc29..1ab45976474d 100644 --- a/arch/x86/events/intel/uncore.c +++ b/arch/x86/events/intel/uncore.c @@ -100,7 +100,13 @@ ssize_t uncore_event_show(struct kobject *kobj, struct intel_uncore_box *uncore_pmu_to_box(struct intel_uncore_pmu *pmu, int cpu) { - return pmu->boxes[topology_logical_package_id(cpu)]; + unsigned int pkgid = topology_logical_package_id(cpu); + + /* + * The unsigned check also catches the '-1' return value for non + * existent mappings in the topology map. + */ + return pkgid < max_packages ? pmu->boxes[pkgid] : NULL; } u64 uncore_msr_read_counter(struct intel_uncore_box *box, struct perf_event *event) @@ -1034,76 +1040,6 @@ static void uncore_pci_exit(void) } } -static int uncore_cpu_dying(unsigned int cpu) -{ - struct intel_uncore_type *type, **types = uncore_msr_uncores; - struct intel_uncore_pmu *pmu; - struct intel_uncore_box *box; - int i, pkg; - - pkg = topology_logical_package_id(cpu); - for (; *types; types++) { - type = *types; - pmu = type->pmus; - for (i = 0; i < type->num_boxes; i++, pmu++) { - box = pmu->boxes[pkg]; - if (box && atomic_dec_return(&box->refcnt) == 0) - uncore_box_exit(box); - } - } - return 0; -} - -static int uncore_cpu_starting(unsigned int cpu) -{ - struct intel_uncore_type *type, **types = uncore_msr_uncores; - struct intel_uncore_pmu *pmu; - struct intel_uncore_box *box; - int i, pkg; - - pkg = topology_logical_package_id(cpu); - for (; *types; types++) { - type = *types; - pmu = type->pmus; - for (i = 0; i < type->num_boxes; i++, pmu++) { - box = pmu->boxes[pkg]; - if (!box) - continue; - /* The first cpu on a package activates the box */ - if (atomic_inc_return(&box->refcnt) == 1) - uncore_box_init(box); - } - } - - return 0; -} - -static int uncore_cpu_prepare(unsigned int cpu) -{ - struct intel_uncore_type *type, **types = uncore_msr_uncores; - struct intel_uncore_pmu *pmu; - struct intel_uncore_box *box; - int i, pkg; - - pkg = topology_logical_package_id(cpu); - for (; *types; types++) { - type = *types; - pmu = type->pmus; - for (i = 0; i < type->num_boxes; i++, pmu++) { - if (pmu->boxes[pkg]) - continue; - /* First cpu of a package allocates the box */ - box = uncore_alloc_box(type, cpu_to_node(cpu)); - if (!box) - return -ENOMEM; - box->pmu = pmu; - box->pkgid = pkg; - pmu->boxes[pkg] = box; - } - } - return 0; -} - static void uncore_change_type_ctx(struct intel_uncore_type *type, int old_cpu, int new_cpu) { @@ -1143,12 +1079,14 @@ static void uncore_change_context(struct intel_uncore_type **uncores, static int uncore_event_cpu_offline(unsigned int cpu) { - int target; + struct intel_uncore_type *type, **types = uncore_msr_uncores; + struct intel_uncore_pmu *pmu; + struct intel_uncore_box *box; + int i, pkg, target; /* Check if exiting cpu is used for collecting uncore events */ if (!cpumask_test_and_clear_cpu(cpu, &uncore_cpu_mask)) - return 0; - + goto unref; /* Find a new cpu to collect uncore events */ target = cpumask_any_but(topology_core_cpumask(cpu), cpu); @@ -1160,12 +1098,82 @@ static int uncore_event_cpu_offline(unsigned int cpu) uncore_change_context(uncore_msr_uncores, cpu, target); uncore_change_context(uncore_pci_uncores, cpu, target); + +unref: + /* Clear the references */ + pkg = topology_logical_package_id(cpu); + for (; *types; types++) { + type = *types; + pmu = type->pmus; + for (i = 0; i < type->num_boxes; i++, pmu++) { + box = pmu->boxes[pkg]; + if (box && atomic_dec_return(&box->refcnt) == 0) + uncore_box_exit(box); + } + } return 0; } +static int allocate_boxes(struct intel_uncore_type **types, + unsigned int pkg, unsigned int cpu) +{ + struct intel_uncore_box *box, *tmp; + struct intel_uncore_type *type; + struct intel_uncore_pmu *pmu; + LIST_HEAD(allocated); + int i; + + /* Try to allocate all required boxes */ + for (; *types; types++) { + type = *types; + pmu = type->pmus; + for (i = 0; i < type->num_boxes; i++, pmu++) { + if (pmu->boxes[pkg]) + continue; + box = uncore_alloc_box(type, cpu_to_node(cpu)); + if (!box) + goto cleanup; + box->pmu = pmu; + box->pkgid = pkg; + list_add(&box->active_list, &allocated); + } + } + /* Install them in the pmus */ + list_for_each_entry_safe(box, tmp, &allocated, active_list) { + list_del_init(&box->active_list); + box->pmu->boxes[pkg] = box; + } + return 0; + +cleanup: + list_for_each_entry_safe(box, tmp, &allocated, active_list) { + list_del_init(&box->active_list); + kfree(box); + } + return -ENOMEM; +} + static int uncore_event_cpu_online(unsigned int cpu) { - int target; + struct intel_uncore_type *type, **types = uncore_msr_uncores; + struct intel_uncore_pmu *pmu; + struct intel_uncore_box *box; + int i, ret, pkg, target; + + pkg = topology_logical_package_id(cpu); + ret = allocate_boxes(types, pkg, cpu); + if (ret) + return ret; + + for (; *types; types++) { + type = *types; + pmu = type->pmus; + for (i = 0; i < type->num_boxes; i++, pmu++) { + box = pmu->boxes[pkg]; + if (!box && atomic_inc_return(&box->refcnt) == 1) + uncore_box_init(box); + } + } /* * Check if there is an online cpu in the package @@ -1355,33 +1363,13 @@ static int __init intel_uncore_init(void) if (cret && pret) return -ENODEV; - /* - * Install callbacks. Core will call them for each online cpu. - * - * The first online cpu of each package allocates and takes - * the refcounts for all other online cpus in that package. - * If msrs are not enabled no allocation is required and - * uncore_cpu_prepare() is not called for each online cpu. - */ - if (!cret) { - ret = cpuhp_setup_state(CPUHP_PERF_X86_UNCORE_PREP, - "perf/x86/intel/uncore:prepare", - uncore_cpu_prepare, NULL); - if (ret) - goto err; - } else { - cpuhp_setup_state_nocalls(CPUHP_PERF_X86_UNCORE_PREP, - "perf/x86/intel/uncore:prepare", - uncore_cpu_prepare, NULL); - } - - cpuhp_setup_state(CPUHP_AP_PERF_X86_UNCORE_STARTING, - "perf/x86/uncore:starting", - uncore_cpu_starting, uncore_cpu_dying); - - cpuhp_setup_state(CPUHP_AP_PERF_X86_UNCORE_ONLINE, - "perf/x86/uncore:online", - uncore_event_cpu_online, uncore_event_cpu_offline); + /* Install hotplug callbacks to setup the targets for each package */ + ret = cpuhp_setup_state(CPUHP_AP_PERF_X86_UNCORE_ONLINE, + "perf/x86/intel/uncore:online", + uncore_event_cpu_online, + uncore_event_cpu_offline); + if (ret) + goto err; return 0; err: @@ -1393,9 +1381,7 @@ module_init(intel_uncore_init); static void __exit intel_uncore_exit(void) { - cpuhp_remove_state_nocalls(CPUHP_AP_PERF_X86_UNCORE_ONLINE); - cpuhp_remove_state_nocalls(CPUHP_AP_PERF_X86_UNCORE_STARTING); - cpuhp_remove_state_nocalls(CPUHP_PERF_X86_UNCORE_PREP); + cpuhp_remove_state(CPUHP_AP_PERF_X86_UNCORE_ONLINE); uncore_types_exit(uncore_msr_uncores); uncore_pci_exit(); } -- cgit v1.2.3