WSL2-Linux-Kernel/arch/x86/entry/entry_64.S

1397 строки
38 KiB
ArmAsm
Исходник Обычный вид История

/*
* linux/arch/x86_64/entry.S
*
* Copyright (C) 1991, 1992 Linus Torvalds
* Copyright (C) 2000, 2001, 2002 Andi Kleen SuSE Labs
* Copyright (C) 2000 Pavel Machek <pavel@suse.cz>
*
* entry.S contains the system-call and fault low-level handling routines.
*
* Some of this is documented in Documentation/x86/entry_64.txt
*
* A note on terminology:
* - iret frame: Architecture defined interrupt frame from SS to RIP
* at the top of the kernel process stack.
*
* Some macro usage:
* - ENTRY/END: Define functions in the symbol table.
* - TRACE_IRQ_*: Trace hardirq state for lock debugging.
* - idtentry: Define exception entry points.
*/
#include <linux/linkage.h>
#include <asm/segment.h>
#include <asm/cache.h>
#include <asm/errno.h>
#include "calling.h"
#include <asm/asm-offsets.h>
#include <asm/msr.h>
#include <asm/unistd.h>
#include <asm/thread_info.h>
#include <asm/hw_irq.h>
#include <asm/page_types.h>
#include <asm/irqflags.h>
#include <asm/paravirt.h>
#include <asm/percpu.h>
#include <asm/asm.h>
#include <asm/context_tracking.h>
#include <asm/smap.h>
x86-64, espfix: Don't leak bits 31:16 of %esp returning to 16-bit stack The IRET instruction, when returning to a 16-bit segment, only restores the bottom 16 bits of the user space stack pointer. This causes some 16-bit software to break, but it also leaks kernel state to user space. We have a software workaround for that ("espfix") for the 32-bit kernel, but it relies on a nonzero stack segment base which is not available in 64-bit mode. In checkin: b3b42ac2cbae x86-64, modify_ldt: Ban 16-bit segments on 64-bit kernels we "solved" this by forbidding 16-bit segments on 64-bit kernels, with the logic that 16-bit support is crippled on 64-bit kernels anyway (no V86 support), but it turns out that people are doing stuff like running old Win16 binaries under Wine and expect it to work. This works around this by creating percpu "ministacks", each of which is mapped 2^16 times 64K apart. When we detect that the return SS is on the LDT, we copy the IRET frame to the ministack and use the relevant alias to return to userspace. The ministacks are mapped readonly, so if IRET faults we promote #GP to #DF which is an IST vector and thus has its own stack; we then do the fixup in the #DF handler. (Making #GP an IST exception would make the msr_safe functions unsafe in NMI/MC context, and quite possibly have other effects.) Special thanks to: - Andy Lutomirski, for the suggestion of using very small stack slots and copy (as opposed to map) the IRET frame there, and for the suggestion to mark them readonly and let the fault promote to #DF. - Konrad Wilk for paravirt fixup and testing. - Borislav Petkov for testing help and useful comments. Reported-by: Brian Gerst <brgerst@gmail.com> Signed-off-by: H. Peter Anvin <hpa@linux.intel.com> Link: http://lkml.kernel.org/r/1398816946-3351-1-git-send-email-hpa@linux.intel.com Cc: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com> Cc: Borislav Petkov <bp@alien8.de> Cc: Andrew Lutomriski <amluto@gmail.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Dirk Hohndel <dirk@hohndel.org> Cc: Arjan van de Ven <arjan.van.de.ven@intel.com> Cc: comex <comexk@gmail.com> Cc: Alexander van Heukelum <heukelum@fastmail.fm> Cc: Boris Ostrovsky <boris.ostrovsky@oracle.com> Cc: <stable@vger.kernel.org> # consider after upstream merge
2014-04-30 03:46:09 +04:00
#include <asm/pgtable_types.h>
Audit: push audit success and retcode into arch ptrace.h The audit system previously expected arches calling to audit_syscall_exit to supply as arguments if the syscall was a success and what the return code was. Audit also provides a helper AUDITSC_RESULT which was supposed to simplify things by converting from negative retcodes to an audit internal magic value stating success or failure. This helper was wrong and could indicate that a valid pointer returned to userspace was a failed syscall. The fix is to fix the layering foolishness. We now pass audit_syscall_exit a struct pt_reg and it in turns calls back into arch code to collect the return value and to determine if the syscall was a success or failure. We also define a generic is_syscall_success() macro which determines success/failure based on if the value is < -MAX_ERRNO. This works for arches like x86 which do not use a separate mechanism to indicate syscall failure. We make both the is_syscall_success() and regs_return_value() static inlines instead of macros. The reason is because the audit function must take a void* for the regs. (uml calls theirs struct uml_pt_regs instead of just struct pt_regs so audit_syscall_exit can't take a struct pt_regs). Since the audit function takes a void* we need to use static inlines to cast it back to the arch correct structure to dereference it. The other major change is that on some arches, like ia64, MIPS and ppc, we change regs_return_value() to give us the negative value on syscall failure. THE only other user of this macro, kretprobe_example.c, won't notice and it makes the value signed consistently for the audit functions across all archs. In arch/sh/kernel/ptrace_64.c I see that we were using regs[9] in the old audit code as the return value. But the ptrace_64.h code defined the macro regs_return_value() as regs[3]. I have no idea which one is correct, but this patch now uses the regs_return_value() function, so it now uses regs[3]. For powerpc we previously used regs->result but now use the regs_return_value() function which uses regs->gprs[3]. regs->gprs[3] is always positive so the regs_return_value(), much like ia64 makes it negative before calling the audit code when appropriate. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: H. Peter Anvin <hpa@zytor.com> [for x86 portion] Acked-by: Tony Luck <tony.luck@intel.com> [for ia64] Acked-by: Richard Weinberger <richard@nod.at> [for uml] Acked-by: David S. Miller <davem@davemloft.net> [for sparc] Acked-by: Ralf Baechle <ralf@linux-mips.org> [for mips] Acked-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> [for ppc]
2012-01-03 23:23:06 +04:00
#include <linux/err.h>
/* Avoid __ASSEMBLER__'ifying <linux/audit.h> just for this. */
#include <linux/elf-em.h>
#define AUDIT_ARCH_X86_64 (EM_X86_64|__AUDIT_ARCH_64BIT|__AUDIT_ARCH_LE)
#define __AUDIT_ARCH_64BIT 0x80000000
#define __AUDIT_ARCH_LE 0x40000000
x86: Separate out entry text section Put x86 entry code into a separate link section: .entry.text. Separating the entry text section seems to have performance benefits - caused by more efficient instruction cache usage. Running hackbench with perf stat --repeat showed that the change compresses the icache footprint. The icache load miss rate went down by about 15%: before patch: 19417627 L1-icache-load-misses ( +- 0.147% ) after patch: 16490788 L1-icache-load-misses ( +- 0.180% ) The motivation of the patch was to fix a particular kprobes bug that relates to the entry text section, the performance advantage was discovered accidentally. Whole perf output follows: - results for current tip tree: Performance counter stats for './hackbench/hackbench 10' (500 runs): 19417627 L1-icache-load-misses ( +- 0.147% ) 2676914223 instructions # 0.497 IPC ( +- 0.079% ) 5389516026 cycles ( +- 0.144% ) 0.206267711 seconds time elapsed ( +- 0.138% ) - results for current tip tree with the patch applied: Performance counter stats for './hackbench/hackbench 10' (500 runs): 16490788 L1-icache-load-misses ( +- 0.180% ) 2717734941 instructions # 0.502 IPC ( +- 0.079% ) 5414756975 cycles ( +- 0.148% ) 0.206747566 seconds time elapsed ( +- 0.137% ) Signed-off-by: Jiri Olsa <jolsa@redhat.com> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Andrew Morton <akpm@linux-foundation.org> Cc: Nick Piggin <npiggin@kernel.dk> Cc: Eric Dumazet <eric.dumazet@gmail.com> Cc: masami.hiramatsu.pt@hitachi.com Cc: ananth@in.ibm.com Cc: davem@davemloft.net Cc: 2nddept-manager@sdl.hitachi.co.jp LKML-Reference: <20110307181039.GB15197@jolsa.redhat.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2011-03-07 21:10:39 +03:00
.code64
.section .entry.text, "ax"
#ifdef CONFIG_PARAVIRT
ENTRY(native_usergs_sysret64)
swapgs
sysretq
ENDPROC(native_usergs_sysret64)
#endif /* CONFIG_PARAVIRT */
.macro TRACE_IRQS_IRETQ
#ifdef CONFIG_TRACE_IRQFLAGS
bt $9, EFLAGS(%rsp) /* interrupts off? */
jnc 1f
TRACE_IRQS_ON
1:
#endif
.endm
ftrace/x86: Do not change stacks in DEBUG when calling lockdep When both DYNAMIC_FTRACE and LOCKDEP are set, the TRACE_IRQS_ON/OFF will call into the lockdep code. The lockdep code can call lots of functions that may be traced by ftrace. When ftrace is updating its code and hits a breakpoint, the breakpoint handler will call into lockdep. If lockdep happens to call a function that also has a breakpoint attached, it will jump back into the breakpoint handler resetting the stack to the debug stack and corrupt the contents currently on that stack. The 'do_sym' call that calls do_int3() is protected by modifying the IST table to point to a different location if another breakpoint is hit. But the TRACE_IRQS_OFF/ON are outside that protection, and if a breakpoint is hit from those, the stack will get corrupted, and the kernel will crash: [ 1013.243754] BUG: unable to handle kernel NULL pointer dereference at 0000000000000002 [ 1013.272665] IP: [<ffff880145cc0000>] 0xffff880145cbffff [ 1013.285186] PGD 1401b2067 PUD 14324c067 PMD 0 [ 1013.298832] Oops: 0010 [#1] PREEMPT SMP [ 1013.310600] CPU 2 [ 1013.317904] Modules linked in: ip6t_REJECT nf_conntrack_ipv6 nf_defrag_ipv6 xt_state nf_conntrack ip6table_filter ip6_tables crc32c_intel ghash_clmulni_intel microcode usb_debug serio_raw pcspkr iTCO_wdt i2c_i801 iTCO_vendor_support e1000e nfsd nfs_acl auth_rpcgss lockd sunrpc i915 video i2c_algo_bit drm_kms_helper drm i2c_core [last unloaded: scsi_wait_scan] [ 1013.401848] [ 1013.407399] Pid: 112, comm: kworker/2:1 Not tainted 3.4.0+ #30 [ 1013.437943] RIP: 8eb8:[<ffff88014630a000>] [<ffff88014630a000>] 0xffff880146309fff [ 1013.459871] RSP: ffffffff8165e919:ffff88014780f408 EFLAGS: 00010046 [ 1013.477909] RAX: 0000000000000001 RBX: ffffffff81104020 RCX: 0000000000000000 [ 1013.499458] RDX: ffff880148008ea8 RSI: ffffffff8131ef40 RDI: ffffffff82203b20 [ 1013.521612] RBP: ffffffff81005751 R08: 0000000000000000 R09: 0000000000000000 [ 1013.543121] R10: ffffffff82cdc318 R11: 0000000000000000 R12: ffff880145cc0000 [ 1013.564614] R13: ffff880148008eb8 R14: 0000000000000002 R15: ffff88014780cb40 [ 1013.586108] FS: 0000000000000000(0000) GS:ffff880148000000(0000) knlGS:0000000000000000 [ 1013.609458] CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b [ 1013.627420] CR2: 0000000000000002 CR3: 0000000141f10000 CR4: 00000000001407e0 [ 1013.649051] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 [ 1013.670724] DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400 [ 1013.692376] Process kworker/2:1 (pid: 112, threadinfo ffff88013fe0e000, task ffff88014020a6a0) [ 1013.717028] Stack: [ 1013.724131] ffff88014780f570 ffff880145cc0000 0000400000004000 0000000000000000 [ 1013.745918] cccccccccccccccc ffff88014780cca8 ffffffff811072bb ffffffff81651627 [ 1013.767870] ffffffff8118f8a7 ffffffff811072bb ffffffff81f2b6c5 ffffffff81f11bdb [ 1013.790021] Call Trace: [ 1013.800701] Code: 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a <e7> d7 64 81 ff ff ff ff 01 00 00 00 00 00 00 00 65 d9 64 81 ff [ 1013.861443] RIP [<ffff88014630a000>] 0xffff880146309fff [ 1013.884466] RSP <ffff88014780f408> [ 1013.901507] CR2: 0000000000000002 The solution was to reuse the NMI functions that change the IDT table to make the debug stack keep its current stack (in kernel mode) when hitting a breakpoint: call debug_stack_set_zero TRACE_IRQS_ON call debug_stack_reset If the TRACE_IRQS_ON happens to hit a breakpoint then it will keep the current stack and not crash the box. Reported-by: Dave Jones <davej@redhat.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2012-05-30 19:54:53 +04:00
/*
* When dynamic function tracer is enabled it will add a breakpoint
* to all locations that it is about to modify, sync CPUs, update
* all the code, sync CPUs, then remove the breakpoints. In this time
* if lockdep is enabled, it might jump back into the debug handler
* outside the updating of the IST protection. (TRACE_IRQS_ON/OFF).
*
* We need to change the IDT table before calling TRACE_IRQS_ON/OFF to
* make sure the stack pointer does not get reset back to the top
* of the debug stack, and instead just reuses the current stack.
*/
#if defined(CONFIG_DYNAMIC_FTRACE) && defined(CONFIG_TRACE_IRQFLAGS)
.macro TRACE_IRQS_OFF_DEBUG
call debug_stack_set_zero
ftrace/x86: Do not change stacks in DEBUG when calling lockdep When both DYNAMIC_FTRACE and LOCKDEP are set, the TRACE_IRQS_ON/OFF will call into the lockdep code. The lockdep code can call lots of functions that may be traced by ftrace. When ftrace is updating its code and hits a breakpoint, the breakpoint handler will call into lockdep. If lockdep happens to call a function that also has a breakpoint attached, it will jump back into the breakpoint handler resetting the stack to the debug stack and corrupt the contents currently on that stack. The 'do_sym' call that calls do_int3() is protected by modifying the IST table to point to a different location if another breakpoint is hit. But the TRACE_IRQS_OFF/ON are outside that protection, and if a breakpoint is hit from those, the stack will get corrupted, and the kernel will crash: [ 1013.243754] BUG: unable to handle kernel NULL pointer dereference at 0000000000000002 [ 1013.272665] IP: [<ffff880145cc0000>] 0xffff880145cbffff [ 1013.285186] PGD 1401b2067 PUD 14324c067 PMD 0 [ 1013.298832] Oops: 0010 [#1] PREEMPT SMP [ 1013.310600] CPU 2 [ 1013.317904] Modules linked in: ip6t_REJECT nf_conntrack_ipv6 nf_defrag_ipv6 xt_state nf_conntrack ip6table_filter ip6_tables crc32c_intel ghash_clmulni_intel microcode usb_debug serio_raw pcspkr iTCO_wdt i2c_i801 iTCO_vendor_support e1000e nfsd nfs_acl auth_rpcgss lockd sunrpc i915 video i2c_algo_bit drm_kms_helper drm i2c_core [last unloaded: scsi_wait_scan] [ 1013.401848] [ 1013.407399] Pid: 112, comm: kworker/2:1 Not tainted 3.4.0+ #30 [ 1013.437943] RIP: 8eb8:[<ffff88014630a000>] [<ffff88014630a000>] 0xffff880146309fff [ 1013.459871] RSP: ffffffff8165e919:ffff88014780f408 EFLAGS: 00010046 [ 1013.477909] RAX: 0000000000000001 RBX: ffffffff81104020 RCX: 0000000000000000 [ 1013.499458] RDX: ffff880148008ea8 RSI: ffffffff8131ef40 RDI: ffffffff82203b20 [ 1013.521612] RBP: ffffffff81005751 R08: 0000000000000000 R09: 0000000000000000 [ 1013.543121] R10: ffffffff82cdc318 R11: 0000000000000000 R12: ffff880145cc0000 [ 1013.564614] R13: ffff880148008eb8 R14: 0000000000000002 R15: ffff88014780cb40 [ 1013.586108] FS: 0000000000000000(0000) GS:ffff880148000000(0000) knlGS:0000000000000000 [ 1013.609458] CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b [ 1013.627420] CR2: 0000000000000002 CR3: 0000000141f10000 CR4: 00000000001407e0 [ 1013.649051] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 [ 1013.670724] DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400 [ 1013.692376] Process kworker/2:1 (pid: 112, threadinfo ffff88013fe0e000, task ffff88014020a6a0) [ 1013.717028] Stack: [ 1013.724131] ffff88014780f570 ffff880145cc0000 0000400000004000 0000000000000000 [ 1013.745918] cccccccccccccccc ffff88014780cca8 ffffffff811072bb ffffffff81651627 [ 1013.767870] ffffffff8118f8a7 ffffffff811072bb ffffffff81f2b6c5 ffffffff81f11bdb [ 1013.790021] Call Trace: [ 1013.800701] Code: 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a <e7> d7 64 81 ff ff ff ff 01 00 00 00 00 00 00 00 65 d9 64 81 ff [ 1013.861443] RIP [<ffff88014630a000>] 0xffff880146309fff [ 1013.884466] RSP <ffff88014780f408> [ 1013.901507] CR2: 0000000000000002 The solution was to reuse the NMI functions that change the IDT table to make the debug stack keep its current stack (in kernel mode) when hitting a breakpoint: call debug_stack_set_zero TRACE_IRQS_ON call debug_stack_reset If the TRACE_IRQS_ON happens to hit a breakpoint then it will keep the current stack and not crash the box. Reported-by: Dave Jones <davej@redhat.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2012-05-30 19:54:53 +04:00
TRACE_IRQS_OFF
call debug_stack_reset
ftrace/x86: Do not change stacks in DEBUG when calling lockdep When both DYNAMIC_FTRACE and LOCKDEP are set, the TRACE_IRQS_ON/OFF will call into the lockdep code. The lockdep code can call lots of functions that may be traced by ftrace. When ftrace is updating its code and hits a breakpoint, the breakpoint handler will call into lockdep. If lockdep happens to call a function that also has a breakpoint attached, it will jump back into the breakpoint handler resetting the stack to the debug stack and corrupt the contents currently on that stack. The 'do_sym' call that calls do_int3() is protected by modifying the IST table to point to a different location if another breakpoint is hit. But the TRACE_IRQS_OFF/ON are outside that protection, and if a breakpoint is hit from those, the stack will get corrupted, and the kernel will crash: [ 1013.243754] BUG: unable to handle kernel NULL pointer dereference at 0000000000000002 [ 1013.272665] IP: [<ffff880145cc0000>] 0xffff880145cbffff [ 1013.285186] PGD 1401b2067 PUD 14324c067 PMD 0 [ 1013.298832] Oops: 0010 [#1] PREEMPT SMP [ 1013.310600] CPU 2 [ 1013.317904] Modules linked in: ip6t_REJECT nf_conntrack_ipv6 nf_defrag_ipv6 xt_state nf_conntrack ip6table_filter ip6_tables crc32c_intel ghash_clmulni_intel microcode usb_debug serio_raw pcspkr iTCO_wdt i2c_i801 iTCO_vendor_support e1000e nfsd nfs_acl auth_rpcgss lockd sunrpc i915 video i2c_algo_bit drm_kms_helper drm i2c_core [last unloaded: scsi_wait_scan] [ 1013.401848] [ 1013.407399] Pid: 112, comm: kworker/2:1 Not tainted 3.4.0+ #30 [ 1013.437943] RIP: 8eb8:[<ffff88014630a000>] [<ffff88014630a000>] 0xffff880146309fff [ 1013.459871] RSP: ffffffff8165e919:ffff88014780f408 EFLAGS: 00010046 [ 1013.477909] RAX: 0000000000000001 RBX: ffffffff81104020 RCX: 0000000000000000 [ 1013.499458] RDX: ffff880148008ea8 RSI: ffffffff8131ef40 RDI: ffffffff82203b20 [ 1013.521612] RBP: ffffffff81005751 R08: 0000000000000000 R09: 0000000000000000 [ 1013.543121] R10: ffffffff82cdc318 R11: 0000000000000000 R12: ffff880145cc0000 [ 1013.564614] R13: ffff880148008eb8 R14: 0000000000000002 R15: ffff88014780cb40 [ 1013.586108] FS: 0000000000000000(0000) GS:ffff880148000000(0000) knlGS:0000000000000000 [ 1013.609458] CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b [ 1013.627420] CR2: 0000000000000002 CR3: 0000000141f10000 CR4: 00000000001407e0 [ 1013.649051] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 [ 1013.670724] DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400 [ 1013.692376] Process kworker/2:1 (pid: 112, threadinfo ffff88013fe0e000, task ffff88014020a6a0) [ 1013.717028] Stack: [ 1013.724131] ffff88014780f570 ffff880145cc0000 0000400000004000 0000000000000000 [ 1013.745918] cccccccccccccccc ffff88014780cca8 ffffffff811072bb ffffffff81651627 [ 1013.767870] ffffffff8118f8a7 ffffffff811072bb ffffffff81f2b6c5 ffffffff81f11bdb [ 1013.790021] Call Trace: [ 1013.800701] Code: 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a <e7> d7 64 81 ff ff ff ff 01 00 00 00 00 00 00 00 65 d9 64 81 ff [ 1013.861443] RIP [<ffff88014630a000>] 0xffff880146309fff [ 1013.884466] RSP <ffff88014780f408> [ 1013.901507] CR2: 0000000000000002 The solution was to reuse the NMI functions that change the IDT table to make the debug stack keep its current stack (in kernel mode) when hitting a breakpoint: call debug_stack_set_zero TRACE_IRQS_ON call debug_stack_reset If the TRACE_IRQS_ON happens to hit a breakpoint then it will keep the current stack and not crash the box. Reported-by: Dave Jones <davej@redhat.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2012-05-30 19:54:53 +04:00
.endm
.macro TRACE_IRQS_ON_DEBUG
call debug_stack_set_zero
ftrace/x86: Do not change stacks in DEBUG when calling lockdep When both DYNAMIC_FTRACE and LOCKDEP are set, the TRACE_IRQS_ON/OFF will call into the lockdep code. The lockdep code can call lots of functions that may be traced by ftrace. When ftrace is updating its code and hits a breakpoint, the breakpoint handler will call into lockdep. If lockdep happens to call a function that also has a breakpoint attached, it will jump back into the breakpoint handler resetting the stack to the debug stack and corrupt the contents currently on that stack. The 'do_sym' call that calls do_int3() is protected by modifying the IST table to point to a different location if another breakpoint is hit. But the TRACE_IRQS_OFF/ON are outside that protection, and if a breakpoint is hit from those, the stack will get corrupted, and the kernel will crash: [ 1013.243754] BUG: unable to handle kernel NULL pointer dereference at 0000000000000002 [ 1013.272665] IP: [<ffff880145cc0000>] 0xffff880145cbffff [ 1013.285186] PGD 1401b2067 PUD 14324c067 PMD 0 [ 1013.298832] Oops: 0010 [#1] PREEMPT SMP [ 1013.310600] CPU 2 [ 1013.317904] Modules linked in: ip6t_REJECT nf_conntrack_ipv6 nf_defrag_ipv6 xt_state nf_conntrack ip6table_filter ip6_tables crc32c_intel ghash_clmulni_intel microcode usb_debug serio_raw pcspkr iTCO_wdt i2c_i801 iTCO_vendor_support e1000e nfsd nfs_acl auth_rpcgss lockd sunrpc i915 video i2c_algo_bit drm_kms_helper drm i2c_core [last unloaded: scsi_wait_scan] [ 1013.401848] [ 1013.407399] Pid: 112, comm: kworker/2:1 Not tainted 3.4.0+ #30 [ 1013.437943] RIP: 8eb8:[<ffff88014630a000>] [<ffff88014630a000>] 0xffff880146309fff [ 1013.459871] RSP: ffffffff8165e919:ffff88014780f408 EFLAGS: 00010046 [ 1013.477909] RAX: 0000000000000001 RBX: ffffffff81104020 RCX: 0000000000000000 [ 1013.499458] RDX: ffff880148008ea8 RSI: ffffffff8131ef40 RDI: ffffffff82203b20 [ 1013.521612] RBP: ffffffff81005751 R08: 0000000000000000 R09: 0000000000000000 [ 1013.543121] R10: ffffffff82cdc318 R11: 0000000000000000 R12: ffff880145cc0000 [ 1013.564614] R13: ffff880148008eb8 R14: 0000000000000002 R15: ffff88014780cb40 [ 1013.586108] FS: 0000000000000000(0000) GS:ffff880148000000(0000) knlGS:0000000000000000 [ 1013.609458] CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b [ 1013.627420] CR2: 0000000000000002 CR3: 0000000141f10000 CR4: 00000000001407e0 [ 1013.649051] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 [ 1013.670724] DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400 [ 1013.692376] Process kworker/2:1 (pid: 112, threadinfo ffff88013fe0e000, task ffff88014020a6a0) [ 1013.717028] Stack: [ 1013.724131] ffff88014780f570 ffff880145cc0000 0000400000004000 0000000000000000 [ 1013.745918] cccccccccccccccc ffff88014780cca8 ffffffff811072bb ffffffff81651627 [ 1013.767870] ffffffff8118f8a7 ffffffff811072bb ffffffff81f2b6c5 ffffffff81f11bdb [ 1013.790021] Call Trace: [ 1013.800701] Code: 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a <e7> d7 64 81 ff ff ff ff 01 00 00 00 00 00 00 00 65 d9 64 81 ff [ 1013.861443] RIP [<ffff88014630a000>] 0xffff880146309fff [ 1013.884466] RSP <ffff88014780f408> [ 1013.901507] CR2: 0000000000000002 The solution was to reuse the NMI functions that change the IDT table to make the debug stack keep its current stack (in kernel mode) when hitting a breakpoint: call debug_stack_set_zero TRACE_IRQS_ON call debug_stack_reset If the TRACE_IRQS_ON happens to hit a breakpoint then it will keep the current stack and not crash the box. Reported-by: Dave Jones <davej@redhat.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2012-05-30 19:54:53 +04:00
TRACE_IRQS_ON
call debug_stack_reset
ftrace/x86: Do not change stacks in DEBUG when calling lockdep When both DYNAMIC_FTRACE and LOCKDEP are set, the TRACE_IRQS_ON/OFF will call into the lockdep code. The lockdep code can call lots of functions that may be traced by ftrace. When ftrace is updating its code and hits a breakpoint, the breakpoint handler will call into lockdep. If lockdep happens to call a function that also has a breakpoint attached, it will jump back into the breakpoint handler resetting the stack to the debug stack and corrupt the contents currently on that stack. The 'do_sym' call that calls do_int3() is protected by modifying the IST table to point to a different location if another breakpoint is hit. But the TRACE_IRQS_OFF/ON are outside that protection, and if a breakpoint is hit from those, the stack will get corrupted, and the kernel will crash: [ 1013.243754] BUG: unable to handle kernel NULL pointer dereference at 0000000000000002 [ 1013.272665] IP: [<ffff880145cc0000>] 0xffff880145cbffff [ 1013.285186] PGD 1401b2067 PUD 14324c067 PMD 0 [ 1013.298832] Oops: 0010 [#1] PREEMPT SMP [ 1013.310600] CPU 2 [ 1013.317904] Modules linked in: ip6t_REJECT nf_conntrack_ipv6 nf_defrag_ipv6 xt_state nf_conntrack ip6table_filter ip6_tables crc32c_intel ghash_clmulni_intel microcode usb_debug serio_raw pcspkr iTCO_wdt i2c_i801 iTCO_vendor_support e1000e nfsd nfs_acl auth_rpcgss lockd sunrpc i915 video i2c_algo_bit drm_kms_helper drm i2c_core [last unloaded: scsi_wait_scan] [ 1013.401848] [ 1013.407399] Pid: 112, comm: kworker/2:1 Not tainted 3.4.0+ #30 [ 1013.437943] RIP: 8eb8:[<ffff88014630a000>] [<ffff88014630a000>] 0xffff880146309fff [ 1013.459871] RSP: ffffffff8165e919:ffff88014780f408 EFLAGS: 00010046 [ 1013.477909] RAX: 0000000000000001 RBX: ffffffff81104020 RCX: 0000000000000000 [ 1013.499458] RDX: ffff880148008ea8 RSI: ffffffff8131ef40 RDI: ffffffff82203b20 [ 1013.521612] RBP: ffffffff81005751 R08: 0000000000000000 R09: 0000000000000000 [ 1013.543121] R10: ffffffff82cdc318 R11: 0000000000000000 R12: ffff880145cc0000 [ 1013.564614] R13: ffff880148008eb8 R14: 0000000000000002 R15: ffff88014780cb40 [ 1013.586108] FS: 0000000000000000(0000) GS:ffff880148000000(0000) knlGS:0000000000000000 [ 1013.609458] CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b [ 1013.627420] CR2: 0000000000000002 CR3: 0000000141f10000 CR4: 00000000001407e0 [ 1013.649051] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 [ 1013.670724] DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400 [ 1013.692376] Process kworker/2:1 (pid: 112, threadinfo ffff88013fe0e000, task ffff88014020a6a0) [ 1013.717028] Stack: [ 1013.724131] ffff88014780f570 ffff880145cc0000 0000400000004000 0000000000000000 [ 1013.745918] cccccccccccccccc ffff88014780cca8 ffffffff811072bb ffffffff81651627 [ 1013.767870] ffffffff8118f8a7 ffffffff811072bb ffffffff81f2b6c5 ffffffff81f11bdb [ 1013.790021] Call Trace: [ 1013.800701] Code: 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a <e7> d7 64 81 ff ff ff ff 01 00 00 00 00 00 00 00 65 d9 64 81 ff [ 1013.861443] RIP [<ffff88014630a000>] 0xffff880146309fff [ 1013.884466] RSP <ffff88014780f408> [ 1013.901507] CR2: 0000000000000002 The solution was to reuse the NMI functions that change the IDT table to make the debug stack keep its current stack (in kernel mode) when hitting a breakpoint: call debug_stack_set_zero TRACE_IRQS_ON call debug_stack_reset If the TRACE_IRQS_ON happens to hit a breakpoint then it will keep the current stack and not crash the box. Reported-by: Dave Jones <davej@redhat.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2012-05-30 19:54:53 +04:00
.endm
.macro TRACE_IRQS_IRETQ_DEBUG
bt $9, EFLAGS(%rsp) /* interrupts off? */
jnc 1f
ftrace/x86: Do not change stacks in DEBUG when calling lockdep When both DYNAMIC_FTRACE and LOCKDEP are set, the TRACE_IRQS_ON/OFF will call into the lockdep code. The lockdep code can call lots of functions that may be traced by ftrace. When ftrace is updating its code and hits a breakpoint, the breakpoint handler will call into lockdep. If lockdep happens to call a function that also has a breakpoint attached, it will jump back into the breakpoint handler resetting the stack to the debug stack and corrupt the contents currently on that stack. The 'do_sym' call that calls do_int3() is protected by modifying the IST table to point to a different location if another breakpoint is hit. But the TRACE_IRQS_OFF/ON are outside that protection, and if a breakpoint is hit from those, the stack will get corrupted, and the kernel will crash: [ 1013.243754] BUG: unable to handle kernel NULL pointer dereference at 0000000000000002 [ 1013.272665] IP: [<ffff880145cc0000>] 0xffff880145cbffff [ 1013.285186] PGD 1401b2067 PUD 14324c067 PMD 0 [ 1013.298832] Oops: 0010 [#1] PREEMPT SMP [ 1013.310600] CPU 2 [ 1013.317904] Modules linked in: ip6t_REJECT nf_conntrack_ipv6 nf_defrag_ipv6 xt_state nf_conntrack ip6table_filter ip6_tables crc32c_intel ghash_clmulni_intel microcode usb_debug serio_raw pcspkr iTCO_wdt i2c_i801 iTCO_vendor_support e1000e nfsd nfs_acl auth_rpcgss lockd sunrpc i915 video i2c_algo_bit drm_kms_helper drm i2c_core [last unloaded: scsi_wait_scan] [ 1013.401848] [ 1013.407399] Pid: 112, comm: kworker/2:1 Not tainted 3.4.0+ #30 [ 1013.437943] RIP: 8eb8:[<ffff88014630a000>] [<ffff88014630a000>] 0xffff880146309fff [ 1013.459871] RSP: ffffffff8165e919:ffff88014780f408 EFLAGS: 00010046 [ 1013.477909] RAX: 0000000000000001 RBX: ffffffff81104020 RCX: 0000000000000000 [ 1013.499458] RDX: ffff880148008ea8 RSI: ffffffff8131ef40 RDI: ffffffff82203b20 [ 1013.521612] RBP: ffffffff81005751 R08: 0000000000000000 R09: 0000000000000000 [ 1013.543121] R10: ffffffff82cdc318 R11: 0000000000000000 R12: ffff880145cc0000 [ 1013.564614] R13: ffff880148008eb8 R14: 0000000000000002 R15: ffff88014780cb40 [ 1013.586108] FS: 0000000000000000(0000) GS:ffff880148000000(0000) knlGS:0000000000000000 [ 1013.609458] CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b [ 1013.627420] CR2: 0000000000000002 CR3: 0000000141f10000 CR4: 00000000001407e0 [ 1013.649051] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 [ 1013.670724] DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400 [ 1013.692376] Process kworker/2:1 (pid: 112, threadinfo ffff88013fe0e000, task ffff88014020a6a0) [ 1013.717028] Stack: [ 1013.724131] ffff88014780f570 ffff880145cc0000 0000400000004000 0000000000000000 [ 1013.745918] cccccccccccccccc ffff88014780cca8 ffffffff811072bb ffffffff81651627 [ 1013.767870] ffffffff8118f8a7 ffffffff811072bb ffffffff81f2b6c5 ffffffff81f11bdb [ 1013.790021] Call Trace: [ 1013.800701] Code: 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a <e7> d7 64 81 ff ff ff ff 01 00 00 00 00 00 00 00 65 d9 64 81 ff [ 1013.861443] RIP [<ffff88014630a000>] 0xffff880146309fff [ 1013.884466] RSP <ffff88014780f408> [ 1013.901507] CR2: 0000000000000002 The solution was to reuse the NMI functions that change the IDT table to make the debug stack keep its current stack (in kernel mode) when hitting a breakpoint: call debug_stack_set_zero TRACE_IRQS_ON call debug_stack_reset If the TRACE_IRQS_ON happens to hit a breakpoint then it will keep the current stack and not crash the box. Reported-by: Dave Jones <davej@redhat.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2012-05-30 19:54:53 +04:00
TRACE_IRQS_ON_DEBUG
1:
.endm
#else
# define TRACE_IRQS_OFF_DEBUG TRACE_IRQS_OFF
# define TRACE_IRQS_ON_DEBUG TRACE_IRQS_ON
# define TRACE_IRQS_IRETQ_DEBUG TRACE_IRQS_IRETQ
ftrace/x86: Do not change stacks in DEBUG when calling lockdep When both DYNAMIC_FTRACE and LOCKDEP are set, the TRACE_IRQS_ON/OFF will call into the lockdep code. The lockdep code can call lots of functions that may be traced by ftrace. When ftrace is updating its code and hits a breakpoint, the breakpoint handler will call into lockdep. If lockdep happens to call a function that also has a breakpoint attached, it will jump back into the breakpoint handler resetting the stack to the debug stack and corrupt the contents currently on that stack. The 'do_sym' call that calls do_int3() is protected by modifying the IST table to point to a different location if another breakpoint is hit. But the TRACE_IRQS_OFF/ON are outside that protection, and if a breakpoint is hit from those, the stack will get corrupted, and the kernel will crash: [ 1013.243754] BUG: unable to handle kernel NULL pointer dereference at 0000000000000002 [ 1013.272665] IP: [<ffff880145cc0000>] 0xffff880145cbffff [ 1013.285186] PGD 1401b2067 PUD 14324c067 PMD 0 [ 1013.298832] Oops: 0010 [#1] PREEMPT SMP [ 1013.310600] CPU 2 [ 1013.317904] Modules linked in: ip6t_REJECT nf_conntrack_ipv6 nf_defrag_ipv6 xt_state nf_conntrack ip6table_filter ip6_tables crc32c_intel ghash_clmulni_intel microcode usb_debug serio_raw pcspkr iTCO_wdt i2c_i801 iTCO_vendor_support e1000e nfsd nfs_acl auth_rpcgss lockd sunrpc i915 video i2c_algo_bit drm_kms_helper drm i2c_core [last unloaded: scsi_wait_scan] [ 1013.401848] [ 1013.407399] Pid: 112, comm: kworker/2:1 Not tainted 3.4.0+ #30 [ 1013.437943] RIP: 8eb8:[<ffff88014630a000>] [<ffff88014630a000>] 0xffff880146309fff [ 1013.459871] RSP: ffffffff8165e919:ffff88014780f408 EFLAGS: 00010046 [ 1013.477909] RAX: 0000000000000001 RBX: ffffffff81104020 RCX: 0000000000000000 [ 1013.499458] RDX: ffff880148008ea8 RSI: ffffffff8131ef40 RDI: ffffffff82203b20 [ 1013.521612] RBP: ffffffff81005751 R08: 0000000000000000 R09: 0000000000000000 [ 1013.543121] R10: ffffffff82cdc318 R11: 0000000000000000 R12: ffff880145cc0000 [ 1013.564614] R13: ffff880148008eb8 R14: 0000000000000002 R15: ffff88014780cb40 [ 1013.586108] FS: 0000000000000000(0000) GS:ffff880148000000(0000) knlGS:0000000000000000 [ 1013.609458] CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b [ 1013.627420] CR2: 0000000000000002 CR3: 0000000141f10000 CR4: 00000000001407e0 [ 1013.649051] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 [ 1013.670724] DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400 [ 1013.692376] Process kworker/2:1 (pid: 112, threadinfo ffff88013fe0e000, task ffff88014020a6a0) [ 1013.717028] Stack: [ 1013.724131] ffff88014780f570 ffff880145cc0000 0000400000004000 0000000000000000 [ 1013.745918] cccccccccccccccc ffff88014780cca8 ffffffff811072bb ffffffff81651627 [ 1013.767870] ffffffff8118f8a7 ffffffff811072bb ffffffff81f2b6c5 ffffffff81f11bdb [ 1013.790021] Call Trace: [ 1013.800701] Code: 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a <e7> d7 64 81 ff ff ff ff 01 00 00 00 00 00 00 00 65 d9 64 81 ff [ 1013.861443] RIP [<ffff88014630a000>] 0xffff880146309fff [ 1013.884466] RSP <ffff88014780f408> [ 1013.901507] CR2: 0000000000000002 The solution was to reuse the NMI functions that change the IDT table to make the debug stack keep its current stack (in kernel mode) when hitting a breakpoint: call debug_stack_set_zero TRACE_IRQS_ON call debug_stack_reset If the TRACE_IRQS_ON happens to hit a breakpoint then it will keep the current stack and not crash the box. Reported-by: Dave Jones <davej@redhat.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2012-05-30 19:54:53 +04:00
#endif
/*
* 64-bit SYSCALL instruction entry. Up to 6 arguments in registers.
*
* 64-bit SYSCALL saves rip to rcx, clears rflags.RF, then saves rflags to r11,
* then loads new ss, cs, and rip from previously programmed MSRs.
* rflags gets masked by a value from another MSR (so CLD and CLAC
* are not needed). SYSCALL does not save anything on the stack
* and does not change rsp.
*
* Registers on entry:
* rax system call number
* rcx return address
* r11 saved rflags (note: r11 is callee-clobbered register in C ABI)
* rdi arg0
* rsi arg1
* rdx arg2
* r10 arg3 (needs to be moved to rcx to conform to C ABI)
* r8 arg4
* r9 arg5
* (note: r12-r15, rbp, rbx are callee-preserved in C ABI)
*
* Only called from user space.
*
* When user can change pt_regs->foo always force IRET. That is because
* it deals with uncanonical addresses better. SYSRET has trouble
* with them due to bugs in both AMD and Intel CPUs.
*/
ENTRY(entry_SYSCALL_64)
x86/asm/entry/64: Use PUSH instructions to build pt_regs on stack With this change, on SYSCALL64 code path we are now populating pt_regs->cs, pt_regs->ss and pt_regs->rcx unconditionally and therefore don't need to do that in FIXUP_TOP_OF_STACK. We lose a number of large instructions there: text data bss dec hex filename 13298 0 0 13298 33f2 entry_64_before.o 12978 0 0 12978 32b2 entry_64.o What's more important, we convert two "MOVQ $imm,off(%rsp)" to "PUSH $imm" (the ones which fill pt_regs->cs,ss). Before this patch, placing them on fast path was slowing it down by two cycles: this form of MOV is very large, 12 bytes, and this probably reduces decode bandwidth to one instruction per cycle when CPU sees them. Therefore they were living in FIXUP_TOP_OF_STACK instead (away from fast path). "PUSH $imm" is a small 2-byte instruction. Moving it to fast path does not slow it down in my measurements. Signed-off-by: Denys Vlasenko <dvlasenk@redhat.com> Acked-by: Borislav Petkov <bp@suse.de> Acked-by: Andy Lutomirski <luto@kernel.org> Cc: Alexei Starovoitov <ast@plumgrid.com> Cc: Andy Lutomirski <luto@amacapital.net> Cc: Borislav Petkov <bp@alien8.de> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: H. Peter Anvin <hpa@zytor.com> Cc: Kees Cook <keescook@chromium.org> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Will Drewry <wad@chromium.org> Link: http://lkml.kernel.org/r/1426785469-15125-3-git-send-email-dvlasenk@redhat.com Signed-off-by: Ingo Molnar <mingo@kernel.org>
2015-03-19 20:17:47 +03:00
/*
* Interrupts are off on entry.
* We do not frame this tiny irq-off block with TRACE_IRQS_OFF/ON,
* it is too small to ever cause noticeable irq latency.
*/
SWAPGS_UNSAFE_STACK
/*
* A hypervisor implementation might want to use a label
* after the swapgs, so that it can do the swapgs
* for the guest and jump here on syscall.
*/
GLOBAL(entry_SYSCALL_64_after_swapgs)
movq %rsp, PER_CPU_VAR(rsp_scratch)
movq PER_CPU_VAR(cpu_current_top_of_stack), %rsp
x86/asm/entry/64: Use PUSH instructions to build pt_regs on stack With this change, on SYSCALL64 code path we are now populating pt_regs->cs, pt_regs->ss and pt_regs->rcx unconditionally and therefore don't need to do that in FIXUP_TOP_OF_STACK. We lose a number of large instructions there: text data bss dec hex filename 13298 0 0 13298 33f2 entry_64_before.o 12978 0 0 12978 32b2 entry_64.o What's more important, we convert two "MOVQ $imm,off(%rsp)" to "PUSH $imm" (the ones which fill pt_regs->cs,ss). Before this patch, placing them on fast path was slowing it down by two cycles: this form of MOV is very large, 12 bytes, and this probably reduces decode bandwidth to one instruction per cycle when CPU sees them. Therefore they were living in FIXUP_TOP_OF_STACK instead (away from fast path). "PUSH $imm" is a small 2-byte instruction. Moving it to fast path does not slow it down in my measurements. Signed-off-by: Denys Vlasenko <dvlasenk@redhat.com> Acked-by: Borislav Petkov <bp@suse.de> Acked-by: Andy Lutomirski <luto@kernel.org> Cc: Alexei Starovoitov <ast@plumgrid.com> Cc: Andy Lutomirski <luto@amacapital.net> Cc: Borislav Petkov <bp@alien8.de> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: H. Peter Anvin <hpa@zytor.com> Cc: Kees Cook <keescook@chromium.org> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Will Drewry <wad@chromium.org> Link: http://lkml.kernel.org/r/1426785469-15125-3-git-send-email-dvlasenk@redhat.com Signed-off-by: Ingo Molnar <mingo@kernel.org>
2015-03-19 20:17:47 +03:00
/* Construct struct pt_regs on stack */
pushq $__USER_DS /* pt_regs->ss */
pushq PER_CPU_VAR(rsp_scratch) /* pt_regs->sp */
/*
x86/asm/entry/64: Use PUSH instructions to build pt_regs on stack With this change, on SYSCALL64 code path we are now populating pt_regs->cs, pt_regs->ss and pt_regs->rcx unconditionally and therefore don't need to do that in FIXUP_TOP_OF_STACK. We lose a number of large instructions there: text data bss dec hex filename 13298 0 0 13298 33f2 entry_64_before.o 12978 0 0 12978 32b2 entry_64.o What's more important, we convert two "MOVQ $imm,off(%rsp)" to "PUSH $imm" (the ones which fill pt_regs->cs,ss). Before this patch, placing them on fast path was slowing it down by two cycles: this form of MOV is very large, 12 bytes, and this probably reduces decode bandwidth to one instruction per cycle when CPU sees them. Therefore they were living in FIXUP_TOP_OF_STACK instead (away from fast path). "PUSH $imm" is a small 2-byte instruction. Moving it to fast path does not slow it down in my measurements. Signed-off-by: Denys Vlasenko <dvlasenk@redhat.com> Acked-by: Borislav Petkov <bp@suse.de> Acked-by: Andy Lutomirski <luto@kernel.org> Cc: Alexei Starovoitov <ast@plumgrid.com> Cc: Andy Lutomirski <luto@amacapital.net> Cc: Borislav Petkov <bp@alien8.de> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: H. Peter Anvin <hpa@zytor.com> Cc: Kees Cook <keescook@chromium.org> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Will Drewry <wad@chromium.org> Link: http://lkml.kernel.org/r/1426785469-15125-3-git-send-email-dvlasenk@redhat.com Signed-off-by: Ingo Molnar <mingo@kernel.org>
2015-03-19 20:17:47 +03:00
* Re-enable interrupts.
* We use 'rsp_scratch' as a scratch space, hence irq-off block above
* must execute atomically in the face of possible interrupt-driven
* task preemption. We must enable interrupts only after we're done
* with using rsp_scratch:
*/
ENABLE_INTERRUPTS(CLBR_NONE)
pushq %r11 /* pt_regs->flags */
pushq $__USER_CS /* pt_regs->cs */
pushq %rcx /* pt_regs->ip */
pushq %rax /* pt_regs->orig_ax */
pushq %rdi /* pt_regs->di */
pushq %rsi /* pt_regs->si */
pushq %rdx /* pt_regs->dx */
pushq %rcx /* pt_regs->cx */
pushq $-ENOSYS /* pt_regs->ax */
pushq %r8 /* pt_regs->r8 */
pushq %r9 /* pt_regs->r9 */
pushq %r10 /* pt_regs->r10 */
pushq %r11 /* pt_regs->r11 */
sub $(6*8), %rsp /* pt_regs->bp, bx, r12-15 not saved */
testl $_TIF_WORK_SYSCALL_ENTRY, ASM_THREAD_INFO(TI_flags, %rsp, SIZEOF_PTREGS)
jnz tracesys
entry_SYSCALL_64_fastpath:
#if __SYSCALL_MASK == ~0
cmpq $__NR_syscall_max, %rax
#else
andl $__SYSCALL_MASK, %eax
cmpl $__NR_syscall_max, %eax
#endif
ja 1f /* return -ENOSYS (already in pt_regs->ax) */
movq %r10, %rcx
call *sys_call_table(, %rax, 8)
movq %rax, RAX(%rsp)
1:
/*
* Syscall return path ending with SYSRET (fast path).
* Has incompletely filled pt_regs.
*/
LOCKDEP_SYS_EXIT
/*
* We do not frame this tiny irq-off block with TRACE_IRQS_OFF/ON,
* it is too small to ever cause noticeable irq latency.
*/
DISABLE_INTERRUPTS(CLBR_NONE)
/*
* We must check ti flags with interrupts (or at least preemption)
* off because we must *never* return to userspace without
* processing exit work that is enqueued if we're preempted here.
* In particular, returning to userspace with any of the one-shot
* flags (TIF_NOTIFY_RESUME, TIF_USER_RETURN_NOTIFY, etc) set is
* very bad.
*/
testl $_TIF_ALLWORK_MASK, ASM_THREAD_INFO(TI_flags, %rsp, SIZEOF_PTREGS)
jnz int_ret_from_sys_call_irqs_off /* Go to the slow path */
RESTORE_C_REGS_EXCEPT_RCX_R11
movq RIP(%rsp), %rcx
movq EFLAGS(%rsp), %r11
movq RSP(%rsp), %rsp
/*
* 64-bit SYSRET restores rip from rcx,
* rflags from r11 (but RF and VM bits are forced to 0),
* cs and ss are loaded from MSRs.
* Restoration of rflags re-enables interrupts.
*
* NB: On AMD CPUs with the X86_BUG_SYSRET_SS_ATTRS bug, the ss
* descriptor is not reinitialized. This means that we should
* avoid SYSRET with SS == NULL, which could happen if we schedule,
* exit the kernel, and re-enter using an interrupt vector. (All
* interrupt entries on x86_64 set SS to NULL.) We prevent that
* from happening by reloading SS in __switch_to. (Actually
* detecting the failure in 64-bit userspace is tricky but can be
* done.)
*/
USERGS_SYSRET64
GLOBAL(int_ret_from_sys_call_irqs_off)
TRACE_IRQS_ON
ENABLE_INTERRUPTS(CLBR_NONE)
jmp int_ret_from_sys_call
/* Do syscall entry tracing */
tracesys:
movq %rsp, %rdi
movl $AUDIT_ARCH_X86_64, %esi
call syscall_trace_enter_phase1
test %rax, %rax
jnz tracesys_phase2 /* if needed, run the slow path */
RESTORE_C_REGS_EXCEPT_RAX /* else restore clobbered regs */
movq ORIG_RAX(%rsp), %rax
jmp entry_SYSCALL_64_fastpath /* and return to the fast path */
tracesys_phase2:
x86/asm/entry/64: Always allocate a complete "struct pt_regs" on the kernel stack The 64-bit entry code was using six stack slots less by not saving/restoring registers which are callee-preserved according to the C ABI, and was not allocating space for them. Only when syscalls needed a complete "struct pt_regs" was the complete area allocated and filled in. As an additional twist, on interrupt entry a "slightly less truncated pt_regs" trick is used, to make nested interrupt stacks easier to unwind. This proved to be a source of significant obfuscation and subtle bugs. For example, 'stub_fork' had to pop the return address, extend the struct, save registers, and push return address back. Ugly. 'ia32_ptregs_common' pops return address and "returns" via jmp insn, throwing a wrench into CPU return stack cache. This patch changes the code to always allocate a complete "struct pt_regs" on the kernel stack. The saving of registers is still done lazily. "Partial pt_regs" trick on interrupt stack is retained. Macros which manipulate "struct pt_regs" on stack are reworked: - ALLOC_PT_GPREGS_ON_STACK allocates the structure. - SAVE_C_REGS saves to it those registers which are clobbered by C code. - SAVE_EXTRA_REGS saves to it all other registers. - Corresponding RESTORE_* and REMOVE_PT_GPREGS_FROM_STACK macros reverse it. 'ia32_ptregs_common', 'stub_fork' and friends lost their ugly dance with the return pointer. LOAD_ARGS32 in ia32entry.S now uses symbolic stack offsets instead of magic numbers. 'error_entry' and 'save_paranoid' now use SAVE_C_REGS + SAVE_EXTRA_REGS instead of having it open-coded yet again. Patch was run-tested: 64-bit executables, 32-bit executables, strace works. Timing tests did not show measurable difference in 32-bit and 64-bit syscalls. Signed-off-by: Denys Vlasenko <dvlasenk@redhat.com> Signed-off-by: Andy Lutomirski <luto@amacapital.net> Cc: Alexei Starovoitov <ast@plumgrid.com> Cc: Borislav Petkov <bp@alien8.de> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: H. Peter Anvin <hpa@zytor.com> Cc: Kees Cook <keescook@chromium.org> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Will Drewry <wad@chromium.org> Link: http://lkml.kernel.org/r/1423778052-21038-2-git-send-email-dvlasenk@redhat.com Link: http://lkml.kernel.org/r/b89763d354aa23e670b9bdf3a40ae320320a7c2e.1424989793.git.luto@amacapital.net Signed-off-by: Ingo Molnar <mingo@kernel.org>
2015-02-27 01:40:27 +03:00
SAVE_EXTRA_REGS
movq %rsp, %rdi
movl $AUDIT_ARCH_X86_64, %esi
movq %rax, %rdx
call syscall_trace_enter_phase2
/*
* Reload registers from stack in case ptrace changed them.
* We don't reload %rax because syscall_trace_entry_phase2() returned
* the value it wants us to use in the table lookup.
*/
x86/asm/entry/64: Always allocate a complete "struct pt_regs" on the kernel stack The 64-bit entry code was using six stack slots less by not saving/restoring registers which are callee-preserved according to the C ABI, and was not allocating space for them. Only when syscalls needed a complete "struct pt_regs" was the complete area allocated and filled in. As an additional twist, on interrupt entry a "slightly less truncated pt_regs" trick is used, to make nested interrupt stacks easier to unwind. This proved to be a source of significant obfuscation and subtle bugs. For example, 'stub_fork' had to pop the return address, extend the struct, save registers, and push return address back. Ugly. 'ia32_ptregs_common' pops return address and "returns" via jmp insn, throwing a wrench into CPU return stack cache. This patch changes the code to always allocate a complete "struct pt_regs" on the kernel stack. The saving of registers is still done lazily. "Partial pt_regs" trick on interrupt stack is retained. Macros which manipulate "struct pt_regs" on stack are reworked: - ALLOC_PT_GPREGS_ON_STACK allocates the structure. - SAVE_C_REGS saves to it those registers which are clobbered by C code. - SAVE_EXTRA_REGS saves to it all other registers. - Corresponding RESTORE_* and REMOVE_PT_GPREGS_FROM_STACK macros reverse it. 'ia32_ptregs_common', 'stub_fork' and friends lost their ugly dance with the return pointer. LOAD_ARGS32 in ia32entry.S now uses symbolic stack offsets instead of magic numbers. 'error_entry' and 'save_paranoid' now use SAVE_C_REGS + SAVE_EXTRA_REGS instead of having it open-coded yet again. Patch was run-tested: 64-bit executables, 32-bit executables, strace works. Timing tests did not show measurable difference in 32-bit and 64-bit syscalls. Signed-off-by: Denys Vlasenko <dvlasenk@redhat.com> Signed-off-by: Andy Lutomirski <luto@amacapital.net> Cc: Alexei Starovoitov <ast@plumgrid.com> Cc: Borislav Petkov <bp@alien8.de> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: H. Peter Anvin <hpa@zytor.com> Cc: Kees Cook <keescook@chromium.org> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Will Drewry <wad@chromium.org> Link: http://lkml.kernel.org/r/1423778052-21038-2-git-send-email-dvlasenk@redhat.com Link: http://lkml.kernel.org/r/b89763d354aa23e670b9bdf3a40ae320320a7c2e.1424989793.git.luto@amacapital.net Signed-off-by: Ingo Molnar <mingo@kernel.org>
2015-02-27 01:40:27 +03:00
RESTORE_C_REGS_EXCEPT_RAX
RESTORE_EXTRA_REGS
#if __SYSCALL_MASK == ~0
cmpq $__NR_syscall_max, %rax
#else
andl $__SYSCALL_MASK, %eax
cmpl $__NR_syscall_max, %eax
#endif
ja 1f /* return -ENOSYS (already in pt_regs->ax) */
movq %r10, %rcx /* fixup for C */
call *sys_call_table(, %rax, 8)
movq %rax, RAX(%rsp)
1:
/* Use IRET because user could have changed pt_regs->foo */
/*
* Syscall return path ending with IRET.
* Has correct iret frame.
*/
GLOBAL(int_ret_from_sys_call)
x86/asm/entry/64: Always allocate a complete "struct pt_regs" on the kernel stack The 64-bit entry code was using six stack slots less by not saving/restoring registers which are callee-preserved according to the C ABI, and was not allocating space for them. Only when syscalls needed a complete "struct pt_regs" was the complete area allocated and filled in. As an additional twist, on interrupt entry a "slightly less truncated pt_regs" trick is used, to make nested interrupt stacks easier to unwind. This proved to be a source of significant obfuscation and subtle bugs. For example, 'stub_fork' had to pop the return address, extend the struct, save registers, and push return address back. Ugly. 'ia32_ptregs_common' pops return address and "returns" via jmp insn, throwing a wrench into CPU return stack cache. This patch changes the code to always allocate a complete "struct pt_regs" on the kernel stack. The saving of registers is still done lazily. "Partial pt_regs" trick on interrupt stack is retained. Macros which manipulate "struct pt_regs" on stack are reworked: - ALLOC_PT_GPREGS_ON_STACK allocates the structure. - SAVE_C_REGS saves to it those registers which are clobbered by C code. - SAVE_EXTRA_REGS saves to it all other registers. - Corresponding RESTORE_* and REMOVE_PT_GPREGS_FROM_STACK macros reverse it. 'ia32_ptregs_common', 'stub_fork' and friends lost their ugly dance with the return pointer. LOAD_ARGS32 in ia32entry.S now uses symbolic stack offsets instead of magic numbers. 'error_entry' and 'save_paranoid' now use SAVE_C_REGS + SAVE_EXTRA_REGS instead of having it open-coded yet again. Patch was run-tested: 64-bit executables, 32-bit executables, strace works. Timing tests did not show measurable difference in 32-bit and 64-bit syscalls. Signed-off-by: Denys Vlasenko <dvlasenk@redhat.com> Signed-off-by: Andy Lutomirski <luto@amacapital.net> Cc: Alexei Starovoitov <ast@plumgrid.com> Cc: Borislav Petkov <bp@alien8.de> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: H. Peter Anvin <hpa@zytor.com> Cc: Kees Cook <keescook@chromium.org> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Will Drewry <wad@chromium.org> Link: http://lkml.kernel.org/r/1423778052-21038-2-git-send-email-dvlasenk@redhat.com Link: http://lkml.kernel.org/r/b89763d354aa23e670b9bdf3a40ae320320a7c2e.1424989793.git.luto@amacapital.net Signed-off-by: Ingo Molnar <mingo@kernel.org>
2015-02-27 01:40:27 +03:00
SAVE_EXTRA_REGS
movq %rsp, %rdi
call syscall_return_slowpath /* returns with IRQs disabled */
x86/asm/entry/64: Always allocate a complete "struct pt_regs" on the kernel stack The 64-bit entry code was using six stack slots less by not saving/restoring registers which are callee-preserved according to the C ABI, and was not allocating space for them. Only when syscalls needed a complete "struct pt_regs" was the complete area allocated and filled in. As an additional twist, on interrupt entry a "slightly less truncated pt_regs" trick is used, to make nested interrupt stacks easier to unwind. This proved to be a source of significant obfuscation and subtle bugs. For example, 'stub_fork' had to pop the return address, extend the struct, save registers, and push return address back. Ugly. 'ia32_ptregs_common' pops return address and "returns" via jmp insn, throwing a wrench into CPU return stack cache. This patch changes the code to always allocate a complete "struct pt_regs" on the kernel stack. The saving of registers is still done lazily. "Partial pt_regs" trick on interrupt stack is retained. Macros which manipulate "struct pt_regs" on stack are reworked: - ALLOC_PT_GPREGS_ON_STACK allocates the structure. - SAVE_C_REGS saves to it those registers which are clobbered by C code. - SAVE_EXTRA_REGS saves to it all other registers. - Corresponding RESTORE_* and REMOVE_PT_GPREGS_FROM_STACK macros reverse it. 'ia32_ptregs_common', 'stub_fork' and friends lost their ugly dance with the return pointer. LOAD_ARGS32 in ia32entry.S now uses symbolic stack offsets instead of magic numbers. 'error_entry' and 'save_paranoid' now use SAVE_C_REGS + SAVE_EXTRA_REGS instead of having it open-coded yet again. Patch was run-tested: 64-bit executables, 32-bit executables, strace works. Timing tests did not show measurable difference in 32-bit and 64-bit syscalls. Signed-off-by: Denys Vlasenko <dvlasenk@redhat.com> Signed-off-by: Andy Lutomirski <luto@amacapital.net> Cc: Alexei Starovoitov <ast@plumgrid.com> Cc: Borislav Petkov <bp@alien8.de> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: H. Peter Anvin <hpa@zytor.com> Cc: Kees Cook <keescook@chromium.org> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Will Drewry <wad@chromium.org> Link: http://lkml.kernel.org/r/1423778052-21038-2-git-send-email-dvlasenk@redhat.com Link: http://lkml.kernel.org/r/b89763d354aa23e670b9bdf3a40ae320320a7c2e.1424989793.git.luto@amacapital.net Signed-off-by: Ingo Molnar <mingo@kernel.org>
2015-02-27 01:40:27 +03:00
RESTORE_EXTRA_REGS
TRACE_IRQS_IRETQ /* we're about to change IF */
/*
* Try to use SYSRET instead of IRET if we're returning to
* a completely clean 64-bit userspace context.
*/
movq RCX(%rsp), %rcx
movq RIP(%rsp), %r11
cmpq %rcx, %r11 /* RCX == RIP */
jne opportunistic_sysret_failed
/*
* On Intel CPUs, SYSRET with non-canonical RCX/RIP will #GP
* in kernel space. This essentially lets the user take over
x86/asm/entry/64: Implement better check for canonical addresses This change makes the check exact (no more false positives on "negative" addresses). Andy explains: "Canonical addresses either start with 17 zeros or 17 ones. In the old code, we checked that the top (64-47) = 17 bits were all zero. We did this by shifting right by 47 bits and making sure that nothing was left. In the new code, we're shifting left by (64 - 48) = 16 bits and then signed shifting right by the same amount, this propagating the 17th highest bit to all positions to its left. If we get the same value we started with, then we're good to go." While it isn't really important to be fully correct here - almost all addresses we'll ever see will be userspace ones, but OTOH it looks to be cheap enough: the new code uses two more ALU ops but preserves %rcx, allowing to not reload it from pt_regs->cx again. On disassembly level, the changes are: cmp %rcx,0x80(%rsp) -> mov 0x80(%rsp),%r11; cmp %rcx,%r11 shr $0x2f,%rcx -> shl $0x10,%rcx; sar $0x10,%rcx; cmp %rcx,%r11 mov 0x58(%rsp),%rcx -> (eliminated) Signed-off-by: Denys Vlasenko <dvlasenk@redhat.com> Acked-by: Andy Lutomirski <luto@kernel.org> Cc: Alexei Starovoitov <ast@plumgrid.com> Cc: Andy Lutomirski <luto@amacapital.net> Cc: Borislav Petkov <bp@alien8.de> Cc: Brian Gerst <brgerst@gmail.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: H. Peter Anvin <hpa@zytor.com> Cc: Kees Cook <keescook@chromium.org> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Will Drewry <wad@chromium.org> Link: http://lkml.kernel.org/r/1429633649-20169-1-git-send-email-dvlasenk@redhat.com [ Changelog massage. ] Signed-off-by: Ingo Molnar <mingo@kernel.org>
2015-04-21 19:27:29 +03:00
* the kernel, since userspace controls RSP.
*
x86/asm/entry/64: Implement better check for canonical addresses This change makes the check exact (no more false positives on "negative" addresses). Andy explains: "Canonical addresses either start with 17 zeros or 17 ones. In the old code, we checked that the top (64-47) = 17 bits were all zero. We did this by shifting right by 47 bits and making sure that nothing was left. In the new code, we're shifting left by (64 - 48) = 16 bits and then signed shifting right by the same amount, this propagating the 17th highest bit to all positions to its left. If we get the same value we started with, then we're good to go." While it isn't really important to be fully correct here - almost all addresses we'll ever see will be userspace ones, but OTOH it looks to be cheap enough: the new code uses two more ALU ops but preserves %rcx, allowing to not reload it from pt_regs->cx again. On disassembly level, the changes are: cmp %rcx,0x80(%rsp) -> mov 0x80(%rsp),%r11; cmp %rcx,%r11 shr $0x2f,%rcx -> shl $0x10,%rcx; sar $0x10,%rcx; cmp %rcx,%r11 mov 0x58(%rsp),%rcx -> (eliminated) Signed-off-by: Denys Vlasenko <dvlasenk@redhat.com> Acked-by: Andy Lutomirski <luto@kernel.org> Cc: Alexei Starovoitov <ast@plumgrid.com> Cc: Andy Lutomirski <luto@amacapital.net> Cc: Borislav Petkov <bp@alien8.de> Cc: Brian Gerst <brgerst@gmail.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: H. Peter Anvin <hpa@zytor.com> Cc: Kees Cook <keescook@chromium.org> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Will Drewry <wad@chromium.org> Link: http://lkml.kernel.org/r/1429633649-20169-1-git-send-email-dvlasenk@redhat.com [ Changelog massage. ] Signed-off-by: Ingo Molnar <mingo@kernel.org>
2015-04-21 19:27:29 +03:00
* If width of "canonical tail" ever becomes variable, this will need
* to be updated to remain correct on both old and new CPUs.
*/
.ifne __VIRTUAL_MASK_SHIFT - 47
.error "virtual address width changed -- SYSRET checks need update"
.endif
x86/asm/entry/64: Implement better check for canonical addresses This change makes the check exact (no more false positives on "negative" addresses). Andy explains: "Canonical addresses either start with 17 zeros or 17 ones. In the old code, we checked that the top (64-47) = 17 bits were all zero. We did this by shifting right by 47 bits and making sure that nothing was left. In the new code, we're shifting left by (64 - 48) = 16 bits and then signed shifting right by the same amount, this propagating the 17th highest bit to all positions to its left. If we get the same value we started with, then we're good to go." While it isn't really important to be fully correct here - almost all addresses we'll ever see will be userspace ones, but OTOH it looks to be cheap enough: the new code uses two more ALU ops but preserves %rcx, allowing to not reload it from pt_regs->cx again. On disassembly level, the changes are: cmp %rcx,0x80(%rsp) -> mov 0x80(%rsp),%r11; cmp %rcx,%r11 shr $0x2f,%rcx -> shl $0x10,%rcx; sar $0x10,%rcx; cmp %rcx,%r11 mov 0x58(%rsp),%rcx -> (eliminated) Signed-off-by: Denys Vlasenko <dvlasenk@redhat.com> Acked-by: Andy Lutomirski <luto@kernel.org> Cc: Alexei Starovoitov <ast@plumgrid.com> Cc: Andy Lutomirski <luto@amacapital.net> Cc: Borislav Petkov <bp@alien8.de> Cc: Brian Gerst <brgerst@gmail.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: H. Peter Anvin <hpa@zytor.com> Cc: Kees Cook <keescook@chromium.org> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Will Drewry <wad@chromium.org> Link: http://lkml.kernel.org/r/1429633649-20169-1-git-send-email-dvlasenk@redhat.com [ Changelog massage. ] Signed-off-by: Ingo Molnar <mingo@kernel.org>
2015-04-21 19:27:29 +03:00
/* Change top 16 bits to be the sign-extension of 47th bit */
shl $(64 - (__VIRTUAL_MASK_SHIFT+1)), %rcx
sar $(64 - (__VIRTUAL_MASK_SHIFT+1)), %rcx
x86/asm/entry/64: Implement better check for canonical addresses This change makes the check exact (no more false positives on "negative" addresses). Andy explains: "Canonical addresses either start with 17 zeros or 17 ones. In the old code, we checked that the top (64-47) = 17 bits were all zero. We did this by shifting right by 47 bits and making sure that nothing was left. In the new code, we're shifting left by (64 - 48) = 16 bits and then signed shifting right by the same amount, this propagating the 17th highest bit to all positions to its left. If we get the same value we started with, then we're good to go." While it isn't really important to be fully correct here - almost all addresses we'll ever see will be userspace ones, but OTOH it looks to be cheap enough: the new code uses two more ALU ops but preserves %rcx, allowing to not reload it from pt_regs->cx again. On disassembly level, the changes are: cmp %rcx,0x80(%rsp) -> mov 0x80(%rsp),%r11; cmp %rcx,%r11 shr $0x2f,%rcx -> shl $0x10,%rcx; sar $0x10,%rcx; cmp %rcx,%r11 mov 0x58(%rsp),%rcx -> (eliminated) Signed-off-by: Denys Vlasenko <dvlasenk@redhat.com> Acked-by: Andy Lutomirski <luto@kernel.org> Cc: Alexei Starovoitov <ast@plumgrid.com> Cc: Andy Lutomirski <luto@amacapital.net> Cc: Borislav Petkov <bp@alien8.de> Cc: Brian Gerst <brgerst@gmail.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: H. Peter Anvin <hpa@zytor.com> Cc: Kees Cook <keescook@chromium.org> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Will Drewry <wad@chromium.org> Link: http://lkml.kernel.org/r/1429633649-20169-1-git-send-email-dvlasenk@redhat.com [ Changelog massage. ] Signed-off-by: Ingo Molnar <mingo@kernel.org>
2015-04-21 19:27:29 +03:00
/* If this changed %rcx, it was not canonical */
cmpq %rcx, %r11
jne opportunistic_sysret_failed
cmpq $__USER_CS, CS(%rsp) /* CS must match SYSRET */
jne opportunistic_sysret_failed
movq R11(%rsp), %r11
cmpq %r11, EFLAGS(%rsp) /* R11 == RFLAGS */
jne opportunistic_sysret_failed
/*
* SYSRET can't restore RF. SYSRET can restore TF, but unlike IRET,
* restoring TF results in a trap from userspace immediately after
* SYSRET. This would cause an infinite loop whenever #DB happens
* with register state that satisfies the opportunistic SYSRET
* conditions. For example, single-stepping this user code:
*
* movq $stuck_here, %rcx
* pushfq
* popq %r11
* stuck_here:
*
* would never get past 'stuck_here'.
*/
testq $(X86_EFLAGS_RF|X86_EFLAGS_TF), %r11
jnz opportunistic_sysret_failed
/* nothing to check for RSP */
cmpq $__USER_DS, SS(%rsp) /* SS must match SYSRET */
jne opportunistic_sysret_failed
/*
* We win! This label is here just for ease of understanding
* perf profiles. Nothing jumps here.
*/
syscall_return_via_sysret:
x86/asm/entry/64: Implement better check for canonical addresses This change makes the check exact (no more false positives on "negative" addresses). Andy explains: "Canonical addresses either start with 17 zeros or 17 ones. In the old code, we checked that the top (64-47) = 17 bits were all zero. We did this by shifting right by 47 bits and making sure that nothing was left. In the new code, we're shifting left by (64 - 48) = 16 bits and then signed shifting right by the same amount, this propagating the 17th highest bit to all positions to its left. If we get the same value we started with, then we're good to go." While it isn't really important to be fully correct here - almost all addresses we'll ever see will be userspace ones, but OTOH it looks to be cheap enough: the new code uses two more ALU ops but preserves %rcx, allowing to not reload it from pt_regs->cx again. On disassembly level, the changes are: cmp %rcx,0x80(%rsp) -> mov 0x80(%rsp),%r11; cmp %rcx,%r11 shr $0x2f,%rcx -> shl $0x10,%rcx; sar $0x10,%rcx; cmp %rcx,%r11 mov 0x58(%rsp),%rcx -> (eliminated) Signed-off-by: Denys Vlasenko <dvlasenk@redhat.com> Acked-by: Andy Lutomirski <luto@kernel.org> Cc: Alexei Starovoitov <ast@plumgrid.com> Cc: Andy Lutomirski <luto@amacapital.net> Cc: Borislav Petkov <bp@alien8.de> Cc: Brian Gerst <brgerst@gmail.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: H. Peter Anvin <hpa@zytor.com> Cc: Kees Cook <keescook@chromium.org> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Will Drewry <wad@chromium.org> Link: http://lkml.kernel.org/r/1429633649-20169-1-git-send-email-dvlasenk@redhat.com [ Changelog massage. ] Signed-off-by: Ingo Molnar <mingo@kernel.org>
2015-04-21 19:27:29 +03:00
/* rcx and r11 are already restored (see code above) */
RESTORE_C_REGS_EXCEPT_RCX_R11
movq RSP(%rsp), %rsp
USERGS_SYSRET64
opportunistic_sysret_failed:
SWAPGS
jmp restore_c_regs_and_iret
END(entry_SYSCALL_64)
.macro FORK_LIKE func
ENTRY(stub_\func)
x86/asm/entry/64: Always allocate a complete "struct pt_regs" on the kernel stack The 64-bit entry code was using six stack slots less by not saving/restoring registers which are callee-preserved according to the C ABI, and was not allocating space for them. Only when syscalls needed a complete "struct pt_regs" was the complete area allocated and filled in. As an additional twist, on interrupt entry a "slightly less truncated pt_regs" trick is used, to make nested interrupt stacks easier to unwind. This proved to be a source of significant obfuscation and subtle bugs. For example, 'stub_fork' had to pop the return address, extend the struct, save registers, and push return address back. Ugly. 'ia32_ptregs_common' pops return address and "returns" via jmp insn, throwing a wrench into CPU return stack cache. This patch changes the code to always allocate a complete "struct pt_regs" on the kernel stack. The saving of registers is still done lazily. "Partial pt_regs" trick on interrupt stack is retained. Macros which manipulate "struct pt_regs" on stack are reworked: - ALLOC_PT_GPREGS_ON_STACK allocates the structure. - SAVE_C_REGS saves to it those registers which are clobbered by C code. - SAVE_EXTRA_REGS saves to it all other registers. - Corresponding RESTORE_* and REMOVE_PT_GPREGS_FROM_STACK macros reverse it. 'ia32_ptregs_common', 'stub_fork' and friends lost their ugly dance with the return pointer. LOAD_ARGS32 in ia32entry.S now uses symbolic stack offsets instead of magic numbers. 'error_entry' and 'save_paranoid' now use SAVE_C_REGS + SAVE_EXTRA_REGS instead of having it open-coded yet again. Patch was run-tested: 64-bit executables, 32-bit executables, strace works. Timing tests did not show measurable difference in 32-bit and 64-bit syscalls. Signed-off-by: Denys Vlasenko <dvlasenk@redhat.com> Signed-off-by: Andy Lutomirski <luto@amacapital.net> Cc: Alexei Starovoitov <ast@plumgrid.com> Cc: Borislav Petkov <bp@alien8.de> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: H. Peter Anvin <hpa@zytor.com> Cc: Kees Cook <keescook@chromium.org> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Will Drewry <wad@chromium.org> Link: http://lkml.kernel.org/r/1423778052-21038-2-git-send-email-dvlasenk@redhat.com Link: http://lkml.kernel.org/r/b89763d354aa23e670b9bdf3a40ae320320a7c2e.1424989793.git.luto@amacapital.net Signed-off-by: Ingo Molnar <mingo@kernel.org>
2015-02-27 01:40:27 +03:00
SAVE_EXTRA_REGS 8
jmp sys_\func
END(stub_\func)
.endm
FORK_LIKE clone
FORK_LIKE fork
FORK_LIKE vfork
ENTRY(stub_execve)
call sys_execve
return_from_execve:
testl %eax, %eax
jz 1f
/* exec failed, can use fast SYSRET code path in this case */
ret
1:
/* must use IRET code path (pt_regs->cs may have changed) */
addq $8, %rsp
ZERO_EXTRA_REGS
movq %rax, RAX(%rsp)
jmp int_ret_from_sys_call
END(stub_execve)
/*
* Remaining execve stubs are only 7 bytes long.
* ENTRY() often aligns to 16 bytes, which in this case has no benefits.
*/
.align 8
GLOBAL(stub_execveat)
call sys_execveat
jmp return_from_execve
END(stub_execveat)
#if defined(CONFIG_X86_X32_ABI) || defined(CONFIG_IA32_EMULATION)
.align 8
GLOBAL(stub_x32_execve)
GLOBAL(stub32_execve)
call compat_sys_execve
jmp return_from_execve
END(stub32_execve)
END(stub_x32_execve)
.align 8
GLOBAL(stub_x32_execveat)
GLOBAL(stub32_execveat)
call compat_sys_execveat
jmp return_from_execve
END(stub32_execveat)
END(stub_x32_execveat)
#endif
/*
* sigreturn is special because it needs to restore all registers on return.
* This cannot be done with SYSRET, so use the IRET return path instead.
*/
ENTRY(stub_rt_sigreturn)
/*
* SAVE_EXTRA_REGS result is not normally needed:
* sigreturn overwrites all pt_regs->GPREGS.
* But sigreturn can fail (!), and there is no easy way to detect that.
* To make sure RESTORE_EXTRA_REGS doesn't restore garbage on error,
* we SAVE_EXTRA_REGS here.
*/
SAVE_EXTRA_REGS 8
call sys_rt_sigreturn
return_from_stub:
addq $8, %rsp
x86/asm/entry/64: Always allocate a complete "struct pt_regs" on the kernel stack The 64-bit entry code was using six stack slots less by not saving/restoring registers which are callee-preserved according to the C ABI, and was not allocating space for them. Only when syscalls needed a complete "struct pt_regs" was the complete area allocated and filled in. As an additional twist, on interrupt entry a "slightly less truncated pt_regs" trick is used, to make nested interrupt stacks easier to unwind. This proved to be a source of significant obfuscation and subtle bugs. For example, 'stub_fork' had to pop the return address, extend the struct, save registers, and push return address back. Ugly. 'ia32_ptregs_common' pops return address and "returns" via jmp insn, throwing a wrench into CPU return stack cache. This patch changes the code to always allocate a complete "struct pt_regs" on the kernel stack. The saving of registers is still done lazily. "Partial pt_regs" trick on interrupt stack is retained. Macros which manipulate "struct pt_regs" on stack are reworked: - ALLOC_PT_GPREGS_ON_STACK allocates the structure. - SAVE_C_REGS saves to it those registers which are clobbered by C code. - SAVE_EXTRA_REGS saves to it all other registers. - Corresponding RESTORE_* and REMOVE_PT_GPREGS_FROM_STACK macros reverse it. 'ia32_ptregs_common', 'stub_fork' and friends lost their ugly dance with the return pointer. LOAD_ARGS32 in ia32entry.S now uses symbolic stack offsets instead of magic numbers. 'error_entry' and 'save_paranoid' now use SAVE_C_REGS + SAVE_EXTRA_REGS instead of having it open-coded yet again. Patch was run-tested: 64-bit executables, 32-bit executables, strace works. Timing tests did not show measurable difference in 32-bit and 64-bit syscalls. Signed-off-by: Denys Vlasenko <dvlasenk@redhat.com> Signed-off-by: Andy Lutomirski <luto@amacapital.net> Cc: Alexei Starovoitov <ast@plumgrid.com> Cc: Borislav Petkov <bp@alien8.de> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: H. Peter Anvin <hpa@zytor.com> Cc: Kees Cook <keescook@chromium.org> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Will Drewry <wad@chromium.org> Link: http://lkml.kernel.org/r/1423778052-21038-2-git-send-email-dvlasenk@redhat.com Link: http://lkml.kernel.org/r/b89763d354aa23e670b9bdf3a40ae320320a7c2e.1424989793.git.luto@amacapital.net Signed-off-by: Ingo Molnar <mingo@kernel.org>
2015-02-27 01:40:27 +03:00
RESTORE_EXTRA_REGS
movq %rax, RAX(%rsp)
jmp int_ret_from_sys_call
END(stub_rt_sigreturn)
#ifdef CONFIG_X86_X32_ABI
ENTRY(stub_x32_rt_sigreturn)
SAVE_EXTRA_REGS 8
call sys32_x32_rt_sigreturn
jmp return_from_stub
END(stub_x32_rt_sigreturn)
#endif
/*
* A newly forked process directly context switches into this address.
*
* rdi: prev task we switched from
*/
ENTRY(ret_from_fork)
LOCK ; btr $TIF_FORK, TI_flags(%r8)
pushq $0x0002
popfq /* reset kernel eflags */
call schedule_tail /* rdi: 'prev' task parameter */
RESTORE_EXTRA_REGS
testb $3, CS(%rsp) /* from kernel_thread? */
/*
* By the time we get here, we have no idea whether our pt_regs,
* ti flags, and ti status came from the 64-bit SYSCALL fast path,
* the slow path, or one of the 32-bit compat paths.
* Use IRET code path to return, since it can safely handle
* all of the above.
*/
jnz int_ret_from_sys_call
/*
* We came from kernel_thread
* nb: we depend on RESTORE_EXTRA_REGS above
*/
movq %rbp, %rdi
call *%rbx
movl $0, RAX(%rsp)
RESTORE_EXTRA_REGS
jmp int_ret_from_sys_call
END(ret_from_fork)
/*
x86/asm/entry/irq: Simplify interrupt dispatch table (IDT) layout Interrupt entry points are handled with the following code, each 32-byte code block contains seven entry points: ... [push][jump 22] // 4 bytes [push][jump 18] // 4 bytes [push][jump 14] // 4 bytes [push][jump 10] // 4 bytes [push][jump 6] // 4 bytes [push][jump 2] // 4 bytes [push][jump common_interrupt][padding] // 8 bytes [push][jump] [push][jump] [push][jump] [push][jump] [push][jump] [push][jump] [push][jump common_interrupt][padding] [padding_2] common_interrupt: And there is a table which holds pointers to every entry point, IOW: to every push. In cold cache, two jumps are still costlier than one, even though we get the benefit of them residing in the same cacheline. This change replaces short jumps with near ones to 'common_interrupt', and pads every push+jump pair to 8 bytes. This way, each interrupt takes only one jump. This change replaces ".p2align CONFIG_X86_L1_CACHE_SHIFT" before dispatch table with ".align 8" - we do not need anything stronger than that. The table of entry addresses (the interrupt[] array) is no longer necessary, the address of entries can be easily calculated as (irq_entries_start + i*8). text data bss dec hex filename 12546 0 0 12546 3102 entry_64.o.before 11626 0 0 11626 2d6a entry_64.o The size decrease is because 1656 bytes of .init.rodata are gone. That's initdata, though. The resident size does go up a bit. Run-tested (32 and 64 bits). Acked-and-Tested-by: Borislav Petkov <bp@suse.de> Signed-off-by: Denys Vlasenko <dvlasenk@redhat.com> Cc: Alexei Starovoitov <ast@plumgrid.com> Cc: Andy Lutomirski <luto@amacapital.net> Cc: Borislav Petkov <bp@alien8.de> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: H. Peter Anvin <hpa@zytor.com> Cc: Kees Cook <keescook@chromium.org> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Will Drewry <wad@chromium.org> Link: http://lkml.kernel.org/r/1428090553-7283-1-git-send-email-dvlasenk@redhat.com Signed-off-by: Ingo Molnar <mingo@kernel.org>
2015-04-03 22:49:13 +03:00
* Build the entry stubs with some assembler magic.
* We pack 1 stub into every 8-byte block.
*/
x86/asm/entry/irq: Simplify interrupt dispatch table (IDT) layout Interrupt entry points are handled with the following code, each 32-byte code block contains seven entry points: ... [push][jump 22] // 4 bytes [push][jump 18] // 4 bytes [push][jump 14] // 4 bytes [push][jump 10] // 4 bytes [push][jump 6] // 4 bytes [push][jump 2] // 4 bytes [push][jump common_interrupt][padding] // 8 bytes [push][jump] [push][jump] [push][jump] [push][jump] [push][jump] [push][jump] [push][jump common_interrupt][padding] [padding_2] common_interrupt: And there is a table which holds pointers to every entry point, IOW: to every push. In cold cache, two jumps are still costlier than one, even though we get the benefit of them residing in the same cacheline. This change replaces short jumps with near ones to 'common_interrupt', and pads every push+jump pair to 8 bytes. This way, each interrupt takes only one jump. This change replaces ".p2align CONFIG_X86_L1_CACHE_SHIFT" before dispatch table with ".align 8" - we do not need anything stronger than that. The table of entry addresses (the interrupt[] array) is no longer necessary, the address of entries can be easily calculated as (irq_entries_start + i*8). text data bss dec hex filename 12546 0 0 12546 3102 entry_64.o.before 11626 0 0 11626 2d6a entry_64.o The size decrease is because 1656 bytes of .init.rodata are gone. That's initdata, though. The resident size does go up a bit. Run-tested (32 and 64 bits). Acked-and-Tested-by: Borislav Petkov <bp@suse.de> Signed-off-by: Denys Vlasenko <dvlasenk@redhat.com> Cc: Alexei Starovoitov <ast@plumgrid.com> Cc: Andy Lutomirski <luto@amacapital.net> Cc: Borislav Petkov <bp@alien8.de> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: H. Peter Anvin <hpa@zytor.com> Cc: Kees Cook <keescook@chromium.org> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Will Drewry <wad@chromium.org> Link: http://lkml.kernel.org/r/1428090553-7283-1-git-send-email-dvlasenk@redhat.com Signed-off-by: Ingo Molnar <mingo@kernel.org>
2015-04-03 22:49:13 +03:00
.align 8
ENTRY(irq_entries_start)
x86/asm/entry/irq: Simplify interrupt dispatch table (IDT) layout Interrupt entry points are handled with the following code, each 32-byte code block contains seven entry points: ... [push][jump 22] // 4 bytes [push][jump 18] // 4 bytes [push][jump 14] // 4 bytes [push][jump 10] // 4 bytes [push][jump 6] // 4 bytes [push][jump 2] // 4 bytes [push][jump common_interrupt][padding] // 8 bytes [push][jump] [push][jump] [push][jump] [push][jump] [push][jump] [push][jump] [push][jump common_interrupt][padding] [padding_2] common_interrupt: And there is a table which holds pointers to every entry point, IOW: to every push. In cold cache, two jumps are still costlier than one, even though we get the benefit of them residing in the same cacheline. This change replaces short jumps with near ones to 'common_interrupt', and pads every push+jump pair to 8 bytes. This way, each interrupt takes only one jump. This change replaces ".p2align CONFIG_X86_L1_CACHE_SHIFT" before dispatch table with ".align 8" - we do not need anything stronger than that. The table of entry addresses (the interrupt[] array) is no longer necessary, the address of entries can be easily calculated as (irq_entries_start + i*8). text data bss dec hex filename 12546 0 0 12546 3102 entry_64.o.before 11626 0 0 11626 2d6a entry_64.o The size decrease is because 1656 bytes of .init.rodata are gone. That's initdata, though. The resident size does go up a bit. Run-tested (32 and 64 bits). Acked-and-Tested-by: Borislav Petkov <bp@suse.de> Signed-off-by: Denys Vlasenko <dvlasenk@redhat.com> Cc: Alexei Starovoitov <ast@plumgrid.com> Cc: Andy Lutomirski <luto@amacapital.net> Cc: Borislav Petkov <bp@alien8.de> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: H. Peter Anvin <hpa@zytor.com> Cc: Kees Cook <keescook@chromium.org> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Will Drewry <wad@chromium.org> Link: http://lkml.kernel.org/r/1428090553-7283-1-git-send-email-dvlasenk@redhat.com Signed-off-by: Ingo Molnar <mingo@kernel.org>
2015-04-03 22:49:13 +03:00
vector=FIRST_EXTERNAL_VECTOR
.rept (FIRST_SYSTEM_VECTOR - FIRST_EXTERNAL_VECTOR)
pushq $(~vector+0x80) /* Note: always in signed byte range */
x86/asm/entry/irq: Simplify interrupt dispatch table (IDT) layout Interrupt entry points are handled with the following code, each 32-byte code block contains seven entry points: ... [push][jump 22] // 4 bytes [push][jump 18] // 4 bytes [push][jump 14] // 4 bytes [push][jump 10] // 4 bytes [push][jump 6] // 4 bytes [push][jump 2] // 4 bytes [push][jump common_interrupt][padding] // 8 bytes [push][jump] [push][jump] [push][jump] [push][jump] [push][jump] [push][jump] [push][jump common_interrupt][padding] [padding_2] common_interrupt: And there is a table which holds pointers to every entry point, IOW: to every push. In cold cache, two jumps are still costlier than one, even though we get the benefit of them residing in the same cacheline. This change replaces short jumps with near ones to 'common_interrupt', and pads every push+jump pair to 8 bytes. This way, each interrupt takes only one jump. This change replaces ".p2align CONFIG_X86_L1_CACHE_SHIFT" before dispatch table with ".align 8" - we do not need anything stronger than that. The table of entry addresses (the interrupt[] array) is no longer necessary, the address of entries can be easily calculated as (irq_entries_start + i*8). text data bss dec hex filename 12546 0 0 12546 3102 entry_64.o.before 11626 0 0 11626 2d6a entry_64.o The size decrease is because 1656 bytes of .init.rodata are gone. That's initdata, though. The resident size does go up a bit. Run-tested (32 and 64 bits). Acked-and-Tested-by: Borislav Petkov <bp@suse.de> Signed-off-by: Denys Vlasenko <dvlasenk@redhat.com> Cc: Alexei Starovoitov <ast@plumgrid.com> Cc: Andy Lutomirski <luto@amacapital.net> Cc: Borislav Petkov <bp@alien8.de> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: H. Peter Anvin <hpa@zytor.com> Cc: Kees Cook <keescook@chromium.org> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Will Drewry <wad@chromium.org> Link: http://lkml.kernel.org/r/1428090553-7283-1-git-send-email-dvlasenk@redhat.com Signed-off-by: Ingo Molnar <mingo@kernel.org>
2015-04-03 22:49:13 +03:00
vector=vector+1
jmp common_interrupt
.align 8
.endr
END(irq_entries_start)
x86: move entry_64.S register saving out of the macros Here is a combined patch that moves "save_args" out-of-line for the interrupt macro and moves "error_entry" mostly out-of-line for the zeroentry and errorentry macros. The save_args function becomes really straightforward and easy to understand, with the possible exception of the stack switch code, which now needs to copy the return address of to the calling function. Normal interrupts arrive with ((~vector)-0x80) on the stack, which gets adjusted in common_interrupt: <common_interrupt>: (5) addq $0xffffffffffffff80,(%rsp) /* -> ~(vector) */ (4) sub $0x50,%rsp /* space for registers */ (5) callq ffffffff80211290 <save_args> (5) callq ffffffff80214290 <do_IRQ> <ret_from_intr>: ... An apic interrupt stub now look like this: <thermal_interrupt>: (5) pushq $0xffffffffffffff05 /* ~(vector) */ (4) sub $0x50,%rsp /* space for registers */ (5) callq ffffffff80211290 <save_args> (5) callq ffffffff80212b8f <smp_thermal_interrupt> (5) jmpq ffffffff80211f93 <ret_from_intr> Similarly the exception handler register saving function becomes simpler, without the need of any parameter shuffling. The stub for an exception without errorcode looks like this: <overflow>: (6) callq *0x1cad12(%rip) # ffffffff803dd448 <pv_irq_ops+0x38> (2) pushq $0xffffffffffffffff /* no syscall */ (4) sub $0x78,%rsp /* space for registers */ (5) callq ffffffff8030e3b0 <error_entry> (3) mov %rsp,%rdi /* pt_regs pointer */ (2) xor %esi,%esi /* no error code */ (5) callq ffffffff80213446 <do_overflow> (5) jmpq ffffffff8030e460 <error_exit> And one for an exception with errorcode like this: <segment_not_present>: (6) callq *0x1cab92(%rip) # ffffffff803dd448 <pv_irq_ops+0x38> (4) sub $0x78,%rsp /* space for registers */ (5) callq ffffffff8030e3b0 <error_entry> (3) mov %rsp,%rdi /* pt_regs pointer */ (5) mov 0x78(%rsp),%rsi /* load error code */ (9) movq $0xffffffffffffffff,0x78(%rsp) /* no syscall */ (5) callq ffffffff80213209 <do_segment_not_present> (5) jmpq ffffffff8030e460 <error_exit> Unfortunately, this last type is more than 32 bytes. But the total space savings due to this patch is about 2500 bytes on an smp-configuration, and I think the code is clearer than it was before. The tested kernels were non-paravirt ones (i.e., without the indirect call at the top of the exception handlers). Anyhow, I tested this patch on top of a recent -tip. The machine was an 2x4-core Xeon at 2333MHz. Measured where the delays between (almost-)adjacent rdtsc instructions. The graphs show how much time is spent outside of the program as a function of the measured delay. The area under the graph represents the total time spent outside the program. Eight instances of the rdtsctest were started, each pinned to a single cpu. The histogams are added. For each kernel two measurements were done: one in mostly idle condition, the other while running "bonnie++ -f", bound to cpu 0. Each measurement took 40 minutes runtime. See the attached graphs for the results. The graphs overlap almost everywhere, but there are small differences. Signed-off-by: Alexander van Heukelum <heukelum@fastmail.fm> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2008-11-19 03:18:11 +03:00
/*
* Interrupt entry/exit.
*
* Interrupt entry points save only callee clobbered registers in fast path.
x86: move entry_64.S register saving out of the macros Here is a combined patch that moves "save_args" out-of-line for the interrupt macro and moves "error_entry" mostly out-of-line for the zeroentry and errorentry macros. The save_args function becomes really straightforward and easy to understand, with the possible exception of the stack switch code, which now needs to copy the return address of to the calling function. Normal interrupts arrive with ((~vector)-0x80) on the stack, which gets adjusted in common_interrupt: <common_interrupt>: (5) addq $0xffffffffffffff80,(%rsp) /* -> ~(vector) */ (4) sub $0x50,%rsp /* space for registers */ (5) callq ffffffff80211290 <save_args> (5) callq ffffffff80214290 <do_IRQ> <ret_from_intr>: ... An apic interrupt stub now look like this: <thermal_interrupt>: (5) pushq $0xffffffffffffff05 /* ~(vector) */ (4) sub $0x50,%rsp /* space for registers */ (5) callq ffffffff80211290 <save_args> (5) callq ffffffff80212b8f <smp_thermal_interrupt> (5) jmpq ffffffff80211f93 <ret_from_intr> Similarly the exception handler register saving function becomes simpler, without the need of any parameter shuffling. The stub for an exception without errorcode looks like this: <overflow>: (6) callq *0x1cad12(%rip) # ffffffff803dd448 <pv_irq_ops+0x38> (2) pushq $0xffffffffffffffff /* no syscall */ (4) sub $0x78,%rsp /* space for registers */ (5) callq ffffffff8030e3b0 <error_entry> (3) mov %rsp,%rdi /* pt_regs pointer */ (2) xor %esi,%esi /* no error code */ (5) callq ffffffff80213446 <do_overflow> (5) jmpq ffffffff8030e460 <error_exit> And one for an exception with errorcode like this: <segment_not_present>: (6) callq *0x1cab92(%rip) # ffffffff803dd448 <pv_irq_ops+0x38> (4) sub $0x78,%rsp /* space for registers */ (5) callq ffffffff8030e3b0 <error_entry> (3) mov %rsp,%rdi /* pt_regs pointer */ (5) mov 0x78(%rsp),%rsi /* load error code */ (9) movq $0xffffffffffffffff,0x78(%rsp) /* no syscall */ (5) callq ffffffff80213209 <do_segment_not_present> (5) jmpq ffffffff8030e460 <error_exit> Unfortunately, this last type is more than 32 bytes. But the total space savings due to this patch is about 2500 bytes on an smp-configuration, and I think the code is clearer than it was before. The tested kernels were non-paravirt ones (i.e., without the indirect call at the top of the exception handlers). Anyhow, I tested this patch on top of a recent -tip. The machine was an 2x4-core Xeon at 2333MHz. Measured where the delays between (almost-)adjacent rdtsc instructions. The graphs show how much time is spent outside of the program as a function of the measured delay. The area under the graph represents the total time spent outside the program. Eight instances of the rdtsctest were started, each pinned to a single cpu. The histogams are added. For each kernel two measurements were done: one in mostly idle condition, the other while running "bonnie++ -f", bound to cpu 0. Each measurement took 40 minutes runtime. See the attached graphs for the results. The graphs overlap almost everywhere, but there are small differences. Signed-off-by: Alexander van Heukelum <heukelum@fastmail.fm> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2008-11-19 03:18:11 +03:00
*
* Entry runs with interrupts off.
*/
/* 0(%rsp): ~(interrupt number) */
.macro interrupt func
cld
ALLOC_PT_GPREGS_ON_STACK
SAVE_C_REGS
SAVE_EXTRA_REGS
x86/asm/entry/64: Always allocate a complete "struct pt_regs" on the kernel stack The 64-bit entry code was using six stack slots less by not saving/restoring registers which are callee-preserved according to the C ABI, and was not allocating space for them. Only when syscalls needed a complete "struct pt_regs" was the complete area allocated and filled in. As an additional twist, on interrupt entry a "slightly less truncated pt_regs" trick is used, to make nested interrupt stacks easier to unwind. This proved to be a source of significant obfuscation and subtle bugs. For example, 'stub_fork' had to pop the return address, extend the struct, save registers, and push return address back. Ugly. 'ia32_ptregs_common' pops return address and "returns" via jmp insn, throwing a wrench into CPU return stack cache. This patch changes the code to always allocate a complete "struct pt_regs" on the kernel stack. The saving of registers is still done lazily. "Partial pt_regs" trick on interrupt stack is retained. Macros which manipulate "struct pt_regs" on stack are reworked: - ALLOC_PT_GPREGS_ON_STACK allocates the structure. - SAVE_C_REGS saves to it those registers which are clobbered by C code. - SAVE_EXTRA_REGS saves to it all other registers. - Corresponding RESTORE_* and REMOVE_PT_GPREGS_FROM_STACK macros reverse it. 'ia32_ptregs_common', 'stub_fork' and friends lost their ugly dance with the return pointer. LOAD_ARGS32 in ia32entry.S now uses symbolic stack offsets instead of magic numbers. 'error_entry' and 'save_paranoid' now use SAVE_C_REGS + SAVE_EXTRA_REGS instead of having it open-coded yet again. Patch was run-tested: 64-bit executables, 32-bit executables, strace works. Timing tests did not show measurable difference in 32-bit and 64-bit syscalls. Signed-off-by: Denys Vlasenko <dvlasenk@redhat.com> Signed-off-by: Andy Lutomirski <luto@amacapital.net> Cc: Alexei Starovoitov <ast@plumgrid.com> Cc: Borislav Petkov <bp@alien8.de> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: H. Peter Anvin <hpa@zytor.com> Cc: Kees Cook <keescook@chromium.org> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Will Drewry <wad@chromium.org> Link: http://lkml.kernel.org/r/1423778052-21038-2-git-send-email-dvlasenk@redhat.com Link: http://lkml.kernel.org/r/b89763d354aa23e670b9bdf3a40ae320320a7c2e.1424989793.git.luto@amacapital.net Signed-off-by: Ingo Molnar <mingo@kernel.org>
2015-02-27 01:40:27 +03:00
testb $3, CS(%rsp)
jz 1f
SWAPGS
x86/asm/entry/64: Always allocate a complete "struct pt_regs" on the kernel stack The 64-bit entry code was using six stack slots less by not saving/restoring registers which are callee-preserved according to the C ABI, and was not allocating space for them. Only when syscalls needed a complete "struct pt_regs" was the complete area allocated and filled in. As an additional twist, on interrupt entry a "slightly less truncated pt_regs" trick is used, to make nested interrupt stacks easier to unwind. This proved to be a source of significant obfuscation and subtle bugs. For example, 'stub_fork' had to pop the return address, extend the struct, save registers, and push return address back. Ugly. 'ia32_ptregs_common' pops return address and "returns" via jmp insn, throwing a wrench into CPU return stack cache. This patch changes the code to always allocate a complete "struct pt_regs" on the kernel stack. The saving of registers is still done lazily. "Partial pt_regs" trick on interrupt stack is retained. Macros which manipulate "struct pt_regs" on stack are reworked: - ALLOC_PT_GPREGS_ON_STACK allocates the structure. - SAVE_C_REGS saves to it those registers which are clobbered by C code. - SAVE_EXTRA_REGS saves to it all other registers. - Corresponding RESTORE_* and REMOVE_PT_GPREGS_FROM_STACK macros reverse it. 'ia32_ptregs_common', 'stub_fork' and friends lost their ugly dance with the return pointer. LOAD_ARGS32 in ia32entry.S now uses symbolic stack offsets instead of magic numbers. 'error_entry' and 'save_paranoid' now use SAVE_C_REGS + SAVE_EXTRA_REGS instead of having it open-coded yet again. Patch was run-tested: 64-bit executables, 32-bit executables, strace works. Timing tests did not show measurable difference in 32-bit and 64-bit syscalls. Signed-off-by: Denys Vlasenko <dvlasenk@redhat.com> Signed-off-by: Andy Lutomirski <luto@amacapital.net> Cc: Alexei Starovoitov <ast@plumgrid.com> Cc: Borislav Petkov <bp@alien8.de> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: H. Peter Anvin <hpa@zytor.com> Cc: Kees Cook <keescook@chromium.org> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Will Drewry <wad@chromium.org> Link: http://lkml.kernel.org/r/1423778052-21038-2-git-send-email-dvlasenk@redhat.com Link: http://lkml.kernel.org/r/b89763d354aa23e670b9bdf3a40ae320320a7c2e.1424989793.git.luto@amacapital.net Signed-off-by: Ingo Molnar <mingo@kernel.org>
2015-02-27 01:40:27 +03:00
1:
/*
* Save previous stack pointer, optionally switch to interrupt stack.
* irq_count is used to check if a CPU is already on an interrupt stack
* or not. While this is essentially redundant with preempt_count it is
* a little cheaper to use a separate counter in the PDA (short of
* moving irq_enter into assembly, which would be too much work)
*/
movq %rsp, %rdi
incl PER_CPU_VAR(irq_count)
cmovzq PER_CPU_VAR(irq_stack_ptr), %rsp
pushq %rdi
/* We entered an interrupt context - irqs are off: */
TRACE_IRQS_OFF
call \func /* rdi points to pt_regs */
.endm
/*
* The interrupt stubs push (~vector+0x80) onto the stack and
* then jump to common_interrupt.
*/
.p2align CONFIG_X86_L1_CACHE_SHIFT
common_interrupt:
ASM_CLAC
addq $-0x80, (%rsp) /* Adjust vector to [-256, -1] range */
interrupt do_IRQ
/* 0(%rsp): old RSP */
ret_from_intr:
DISABLE_INTERRUPTS(CLBR_NONE)
TRACE_IRQS_OFF
decl PER_CPU_VAR(irq_count)
x86: Save rbp in pt_regs on irq entry From the x86_64 low level interrupt handlers, the frame pointer is saved right after the partial pt_regs frame. rbp is not supposed to be part of the irq partial saved registers, but it only requires to extend the pt_regs frame by 8 bytes to do so, plus a tiny stack offset fixup on irq exit. This changes a bit the semantics or get_irq_entry() that is supposed to provide only the value of caller saved registers and the cpu saved frame. However it's a win for unwinders that can walk through stack frames on top of get_irq_regs() snapshots. A noticeable impact is that it makes perf events cpu-clock and task-clock events based callchains working on x86_64. Let's then save rbp into the irq pt_regs. As a result with: perf record -e cpu-clock perf bench sched messaging perf report --stdio Before: 20.94% perf [kernel.kallsyms] [k] lock_acquire | --- lock_acquire | |--44.01%-- __write_nocancel | |--43.18%-- __read | |--6.08%-- fork | create_worker | |--0.88%-- _dl_fixup | |--0.65%-- do_lookup_x | |--0.53%-- __GI___libc_read --4.67%-- [...] After: 19.23% perf [kernel.kallsyms] [k] __lock_acquire | --- __lock_acquire | |--97.74%-- lock_acquire | | | |--21.82%-- _raw_spin_lock | | | | | |--37.26%-- unix_stream_recvmsg | | | sock_aio_read | | | do_sync_read | | | vfs_read | | | sys_read | | | system_call | | | __read | | | | | |--24.09%-- unix_stream_sendmsg | | | sock_aio_write | | | do_sync_write | | | vfs_write | | | sys_write | | | system_call | | | __write_nocancel v2: Fix cfi annotations. Reported-by: Soeren Sandmann Pedersen <sandmann@redhat.com> Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com> Cc: Ingo Molnar <mingo@elte.hu> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: H. Peter Anvin <hpa@zytor.com Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Stephane Eranian <eranian@google.com> Cc: Jan Beulich <JBeulich@novell.com>
2011-01-06 17:22:47 +03:00
x86: Don't use frame pointer to save old stack on irq entry rbp is used in SAVE_ARGS_IRQ to save the old stack pointer in order to restore it later in ret_from_intr. It is convenient because we save its value in the irq regs and it's easily restored using the leave instruction. However this is a kind of abuse of the frame pointer which role is to help unwinding the kernel by chaining frames together, each node following the return address to the previous frame. But although we are breaking the frame by changing the stack pointer, there is no preceding return address before the new frame. Hence using the frame pointer to link the two stacks breaks the stack unwinders that find a random value instead of a return address here. There is no workaround that can work in every case. We are using the fixup_bp_irq_link() function to dereference that abused frame pointer in the case of non nesting interrupt (which means stack changed). But that doesn't fix the case of interrupts that don't change the stack (but we still have the unconditional frame link), which is the case of hardirq interrupting softirq. We have no way to detect this transition so the frame irq link is considered as a real frame pointer and the return address is dereferenced but it is still a spurious one. There are two possible results of this: either the spurious return address, a random stack value, luckily belongs to the kernel text and then the unwinding can continue and we just have a weird entry in the stack trace. Or it doesn't belong to the kernel text and unwinding stops there. This is the reason why stacktraces (including perf callchains) on irqs that interrupted softirqs don't work very well. To solve this, we don't save the old stack pointer on rbp anymore but we save it to a scratch register that we push on the new stack and that we pop back later on irq return. This preserves the whole frame chain without spurious return addresses in the middle and drops the need for the horrid fixup_bp_irq_link() workaround. And finally irqs that interrupt softirq are sanely unwinded. Before: 99.81% perf [kernel.kallsyms] [k] perf_pending_event | --- perf_pending_event irq_work_run smp_irq_work_interrupt irq_work_interrupt | |--41.60%-- __read | | | |--99.90%-- create_worker | | bench_sched_messaging | | cmd_bench | | run_builtin | | main | | __libc_start_main | --0.10%-- [...] After: 1.64% swapper [kernel.kallsyms] [k] perf_pending_event | --- perf_pending_event irq_work_run smp_irq_work_interrupt irq_work_interrupt | |--95.00%-- arch_irq_work_raise | irq_work_queue | __perf_event_overflow | perf_swevent_overflow | perf_swevent_event | perf_tp_event | perf_trace_softirq | __do_softirq | call_softirq | do_softirq | irq_exit | | | |--73.68%-- smp_apic_timer_interrupt | | apic_timer_interrupt | | | | | |--96.43%-- amd_e400_idle | | | cpu_idle | | | start_secondary Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com> Cc: Ingo Molnar <mingo@elte.hu> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: H. Peter Anvin <hpa@zytor.com> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Jan Beulich <JBeulich@novell.com>
2011-07-02 18:52:45 +04:00
/* Restore saved previous stack */
popq %rsp
x86: Save rbp in pt_regs on irq entry From the x86_64 low level interrupt handlers, the frame pointer is saved right after the partial pt_regs frame. rbp is not supposed to be part of the irq partial saved registers, but it only requires to extend the pt_regs frame by 8 bytes to do so, plus a tiny stack offset fixup on irq exit. This changes a bit the semantics or get_irq_entry() that is supposed to provide only the value of caller saved registers and the cpu saved frame. However it's a win for unwinders that can walk through stack frames on top of get_irq_regs() snapshots. A noticeable impact is that it makes perf events cpu-clock and task-clock events based callchains working on x86_64. Let's then save rbp into the irq pt_regs. As a result with: perf record -e cpu-clock perf bench sched messaging perf report --stdio Before: 20.94% perf [kernel.kallsyms] [k] lock_acquire | --- lock_acquire | |--44.01%-- __write_nocancel | |--43.18%-- __read | |--6.08%-- fork | create_worker | |--0.88%-- _dl_fixup | |--0.65%-- do_lookup_x | |--0.53%-- __GI___libc_read --4.67%-- [...] After: 19.23% perf [kernel.kallsyms] [k] __lock_acquire | --- __lock_acquire | |--97.74%-- lock_acquire | | | |--21.82%-- _raw_spin_lock | | | | | |--37.26%-- unix_stream_recvmsg | | | sock_aio_read | | | do_sync_read | | | vfs_read | | | sys_read | | | system_call | | | __read | | | | | |--24.09%-- unix_stream_sendmsg | | | sock_aio_write | | | do_sync_write | | | vfs_write | | | sys_write | | | system_call | | | __write_nocancel v2: Fix cfi annotations. Reported-by: Soeren Sandmann Pedersen <sandmann@redhat.com> Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com> Cc: Ingo Molnar <mingo@elte.hu> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: H. Peter Anvin <hpa@zytor.com Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Stephane Eranian <eranian@google.com> Cc: Jan Beulich <JBeulich@novell.com>
2011-01-06 17:22:47 +03:00
testb $3, CS(%rsp)
jz retint_kernel
/* Interrupt came from user space */
GLOBAL(retint_user)
GET_THREAD_INFO(%rcx)
/* %rcx: thread info. Interrupts are off. */
retint_with_reschedule:
movl $_TIF_WORK_MASK, %edi
retint_check:
LOCKDEP_SYS_EXIT_IRQ
movl TI_flags(%rcx), %edx
andl %edi, %edx
jnz retint_careful
retint_swapgs: /* return to user-space */
/*
* The iretq could re-enable interrupts:
*/
DISABLE_INTERRUPTS(CLBR_ANY)
TRACE_IRQS_IRETQ
x86_64, entry: Use sysret to return to userspace when possible The x86_64 entry code currently jumps through complex and inconsistent hoops to try to minimize the impact of syscall exit work. For a true fast-path syscall, almost nothing needs to be done, so returning is just a check for exit work and sysret. For a full slow-path return from a syscall, the C exit hook is invoked if needed and we join the iret path. Using iret to return to userspace is very slow, so the entry code has accumulated various special cases to try to do certain forms of exit work without invoking iret. This is error-prone, since it duplicates assembly code paths, and it's dangerous, since sysret can malfunction in interesting ways if used carelessly. It's also inefficient, since a lot of useful cases aren't optimized and therefore force an iret out of a combination of paranoia and the fact that no one has bothered to write even more asm code to avoid it. I would argue that this approach is backwards. Rather than trying to avoid the iret path, we should instead try to make the iret path fast. Under a specific set of conditions, iret is unnecessary. In particular, if RIP==RCX, RFLAGS==R11, RIP is canonical, RF is not set, and both SS and CS are as expected, then movq 32(%rsp),%rsp;sysret does the same thing as iret. This set of conditions is nearly always satisfied on return from syscalls, and it can even occasionally be satisfied on return from an irq. Even with the careful checks for sysret applicability, this cuts nearly 80ns off of the overhead from syscalls with unoptimized exit work. This includes tracing and context tracking, and any return that invokes KVM's user return notifier. For example, the cost of getpid with CONFIG_CONTEXT_TRACKING_FORCE=y drops from ~360ns to ~280ns on my computer. This may allow the removal and even eventual conversion to C of a respectable amount of exit asm. This may require further tweaking to give the full benefit on Xen. It may be worthwhile to adjust signal delivery and exec to try hit the sysret path. This does not optimize returns to 32-bit userspace. Making the same optimization for CS == __USER32_CS is conceptually straightforward, but it will require some tedious code to handle the differences between sysretl and sysexitl. Link: http://lkml.kernel.org/r/71428f63e681e1b4aa1a781e3ef7c27f027d1103.1421453410.git.luto@amacapital.net Signed-off-by: Andy Lutomirski <luto@amacapital.net>
2014-07-22 23:46:50 +04:00
SWAPGS
jmp restore_regs_and_iret
/* Returning to kernel space */
retint_kernel:
#ifdef CONFIG_PREEMPT
/* Interrupts are off */
/* Check if we need preemption */
bt $9, EFLAGS(%rsp) /* were interrupts off? */
jnc 1f
0: cmpl $0, PER_CPU_VAR(__preempt_count)
jnz 1f
call preempt_schedule_irq
jmp 0b
1:
#endif
/*
* The iretq could re-enable interrupts:
*/
TRACE_IRQS_IRETQ
/*
* At this label, code paths which return to kernel and to user,
* which come from interrupts/exception and from syscalls, merge.
*/
restore_regs_and_iret:
RESTORE_EXTRA_REGS
restore_c_regs_and_iret:
x86/asm/entry/64: Always allocate a complete "struct pt_regs" on the kernel stack The 64-bit entry code was using six stack slots less by not saving/restoring registers which are callee-preserved according to the C ABI, and was not allocating space for them. Only when syscalls needed a complete "struct pt_regs" was the complete area allocated and filled in. As an additional twist, on interrupt entry a "slightly less truncated pt_regs" trick is used, to make nested interrupt stacks easier to unwind. This proved to be a source of significant obfuscation and subtle bugs. For example, 'stub_fork' had to pop the return address, extend the struct, save registers, and push return address back. Ugly. 'ia32_ptregs_common' pops return address and "returns" via jmp insn, throwing a wrench into CPU return stack cache. This patch changes the code to always allocate a complete "struct pt_regs" on the kernel stack. The saving of registers is still done lazily. "Partial pt_regs" trick on interrupt stack is retained. Macros which manipulate "struct pt_regs" on stack are reworked: - ALLOC_PT_GPREGS_ON_STACK allocates the structure. - SAVE_C_REGS saves to it those registers which are clobbered by C code. - SAVE_EXTRA_REGS saves to it all other registers. - Corresponding RESTORE_* and REMOVE_PT_GPREGS_FROM_STACK macros reverse it. 'ia32_ptregs_common', 'stub_fork' and friends lost their ugly dance with the return pointer. LOAD_ARGS32 in ia32entry.S now uses symbolic stack offsets instead of magic numbers. 'error_entry' and 'save_paranoid' now use SAVE_C_REGS + SAVE_EXTRA_REGS instead of having it open-coded yet again. Patch was run-tested: 64-bit executables, 32-bit executables, strace works. Timing tests did not show measurable difference in 32-bit and 64-bit syscalls. Signed-off-by: Denys Vlasenko <dvlasenk@redhat.com> Signed-off-by: Andy Lutomirski <luto@amacapital.net> Cc: Alexei Starovoitov <ast@plumgrid.com> Cc: Borislav Petkov <bp@alien8.de> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: H. Peter Anvin <hpa@zytor.com> Cc: Kees Cook <keescook@chromium.org> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Will Drewry <wad@chromium.org> Link: http://lkml.kernel.org/r/1423778052-21038-2-git-send-email-dvlasenk@redhat.com Link: http://lkml.kernel.org/r/b89763d354aa23e670b9bdf3a40ae320320a7c2e.1424989793.git.luto@amacapital.net Signed-off-by: Ingo Molnar <mingo@kernel.org>
2015-02-27 01:40:27 +03:00
RESTORE_C_REGS
REMOVE_PT_GPREGS_FROM_STACK 8
INTERRUPT_RETURN
ENTRY(native_iret)
x86-64, espfix: Don't leak bits 31:16 of %esp returning to 16-bit stack The IRET instruction, when returning to a 16-bit segment, only restores the bottom 16 bits of the user space stack pointer. This causes some 16-bit software to break, but it also leaks kernel state to user space. We have a software workaround for that ("espfix") for the 32-bit kernel, but it relies on a nonzero stack segment base which is not available in 64-bit mode. In checkin: b3b42ac2cbae x86-64, modify_ldt: Ban 16-bit segments on 64-bit kernels we "solved" this by forbidding 16-bit segments on 64-bit kernels, with the logic that 16-bit support is crippled on 64-bit kernels anyway (no V86 support), but it turns out that people are doing stuff like running old Win16 binaries under Wine and expect it to work. This works around this by creating percpu "ministacks", each of which is mapped 2^16 times 64K apart. When we detect that the return SS is on the LDT, we copy the IRET frame to the ministack and use the relevant alias to return to userspace. The ministacks are mapped readonly, so if IRET faults we promote #GP to #DF which is an IST vector and thus has its own stack; we then do the fixup in the #DF handler. (Making #GP an IST exception would make the msr_safe functions unsafe in NMI/MC context, and quite possibly have other effects.) Special thanks to: - Andy Lutomirski, for the suggestion of using very small stack slots and copy (as opposed to map) the IRET frame there, and for the suggestion to mark them readonly and let the fault promote to #DF. - Konrad Wilk for paravirt fixup and testing. - Borislav Petkov for testing help and useful comments. Reported-by: Brian Gerst <brgerst@gmail.com> Signed-off-by: H. Peter Anvin <hpa@linux.intel.com> Link: http://lkml.kernel.org/r/1398816946-3351-1-git-send-email-hpa@linux.intel.com Cc: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com> Cc: Borislav Petkov <bp@alien8.de> Cc: Andrew Lutomriski <amluto@gmail.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Dirk Hohndel <dirk@hohndel.org> Cc: Arjan van de Ven <arjan.van.de.ven@intel.com> Cc: comex <comexk@gmail.com> Cc: Alexander van Heukelum <heukelum@fastmail.fm> Cc: Boris Ostrovsky <boris.ostrovsky@oracle.com> Cc: <stable@vger.kernel.org> # consider after upstream merge
2014-04-30 03:46:09 +04:00
/*
* Are we returning to a stack segment from the LDT? Note: in
* 64-bit mode SS:RSP on the exception stack is always valid.
*/
#ifdef CONFIG_X86_ESPFIX64
testb $4, (SS-RIP)(%rsp)
jnz native_irq_return_ldt
#endif
x86-64, espfix: Don't leak bits 31:16 of %esp returning to 16-bit stack The IRET instruction, when returning to a 16-bit segment, only restores the bottom 16 bits of the user space stack pointer. This causes some 16-bit software to break, but it also leaks kernel state to user space. We have a software workaround for that ("espfix") for the 32-bit kernel, but it relies on a nonzero stack segment base which is not available in 64-bit mode. In checkin: b3b42ac2cbae x86-64, modify_ldt: Ban 16-bit segments on 64-bit kernels we "solved" this by forbidding 16-bit segments on 64-bit kernels, with the logic that 16-bit support is crippled on 64-bit kernels anyway (no V86 support), but it turns out that people are doing stuff like running old Win16 binaries under Wine and expect it to work. This works around this by creating percpu "ministacks", each of which is mapped 2^16 times 64K apart. When we detect that the return SS is on the LDT, we copy the IRET frame to the ministack and use the relevant alias to return to userspace. The ministacks are mapped readonly, so if IRET faults we promote #GP to #DF which is an IST vector and thus has its own stack; we then do the fixup in the #DF handler. (Making #GP an IST exception would make the msr_safe functions unsafe in NMI/MC context, and quite possibly have other effects.) Special thanks to: - Andy Lutomirski, for the suggestion of using very small stack slots and copy (as opposed to map) the IRET frame there, and for the suggestion to mark them readonly and let the fault promote to #DF. - Konrad Wilk for paravirt fixup and testing. - Borislav Petkov for testing help and useful comments. Reported-by: Brian Gerst <brgerst@gmail.com> Signed-off-by: H. Peter Anvin <hpa@linux.intel.com> Link: http://lkml.kernel.org/r/1398816946-3351-1-git-send-email-hpa@linux.intel.com Cc: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com> Cc: Borislav Petkov <bp@alien8.de> Cc: Andrew Lutomriski <amluto@gmail.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Dirk Hohndel <dirk@hohndel.org> Cc: Arjan van de Ven <arjan.van.de.ven@intel.com> Cc: comex <comexk@gmail.com> Cc: Alexander van Heukelum <heukelum@fastmail.fm> Cc: Boris Ostrovsky <boris.ostrovsky@oracle.com> Cc: <stable@vger.kernel.org> # consider after upstream merge
2014-04-30 03:46:09 +04:00
.global native_irq_return_iret
native_irq_return_iret:
/*
* This may fault. Non-paranoid faults on return to userspace are
* handled by fixup_bad_iret. These include #SS, #GP, and #NP.
* Double-faults due to espfix64 are handled in do_double_fault.
* Other faults here are fatal.
*/
iretq
#ifdef CONFIG_X86_ESPFIX64
native_irq_return_ldt:
pushq %rax
pushq %rdi
x86-64, espfix: Don't leak bits 31:16 of %esp returning to 16-bit stack The IRET instruction, when returning to a 16-bit segment, only restores the bottom 16 bits of the user space stack pointer. This causes some 16-bit software to break, but it also leaks kernel state to user space. We have a software workaround for that ("espfix") for the 32-bit kernel, but it relies on a nonzero stack segment base which is not available in 64-bit mode. In checkin: b3b42ac2cbae x86-64, modify_ldt: Ban 16-bit segments on 64-bit kernels we "solved" this by forbidding 16-bit segments on 64-bit kernels, with the logic that 16-bit support is crippled on 64-bit kernels anyway (no V86 support), but it turns out that people are doing stuff like running old Win16 binaries under Wine and expect it to work. This works around this by creating percpu "ministacks", each of which is mapped 2^16 times 64K apart. When we detect that the return SS is on the LDT, we copy the IRET frame to the ministack and use the relevant alias to return to userspace. The ministacks are mapped readonly, so if IRET faults we promote #GP to #DF which is an IST vector and thus has its own stack; we then do the fixup in the #DF handler. (Making #GP an IST exception would make the msr_safe functions unsafe in NMI/MC context, and quite possibly have other effects.) Special thanks to: - Andy Lutomirski, for the suggestion of using very small stack slots and copy (as opposed to map) the IRET frame there, and for the suggestion to mark them readonly and let the fault promote to #DF. - Konrad Wilk for paravirt fixup and testing. - Borislav Petkov for testing help and useful comments. Reported-by: Brian Gerst <brgerst@gmail.com> Signed-off-by: H. Peter Anvin <hpa@linux.intel.com> Link: http://lkml.kernel.org/r/1398816946-3351-1-git-send-email-hpa@linux.intel.com Cc: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com> Cc: Borislav Petkov <bp@alien8.de> Cc: Andrew Lutomriski <amluto@gmail.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Dirk Hohndel <dirk@hohndel.org> Cc: Arjan van de Ven <arjan.van.de.ven@intel.com> Cc: comex <comexk@gmail.com> Cc: Alexander van Heukelum <heukelum@fastmail.fm> Cc: Boris Ostrovsky <boris.ostrovsky@oracle.com> Cc: <stable@vger.kernel.org> # consider after upstream merge
2014-04-30 03:46:09 +04:00
SWAPGS
movq PER_CPU_VAR(espfix_waddr), %rdi
movq %rax, (0*8)(%rdi) /* RAX */
movq (2*8)(%rsp), %rax /* RIP */
movq %rax, (1*8)(%rdi)
movq (3*8)(%rsp), %rax /* CS */
movq %rax, (2*8)(%rdi)
movq (4*8)(%rsp), %rax /* RFLAGS */
movq %rax, (3*8)(%rdi)
movq (6*8)(%rsp), %rax /* SS */
movq %rax, (5*8)(%rdi)
movq (5*8)(%rsp), %rax /* RSP */
movq %rax, (4*8)(%rdi)
andl $0xffff0000, %eax
popq %rdi
orq PER_CPU_VAR(espfix_stack), %rax
x86-64, espfix: Don't leak bits 31:16 of %esp returning to 16-bit stack The IRET instruction, when returning to a 16-bit segment, only restores the bottom 16 bits of the user space stack pointer. This causes some 16-bit software to break, but it also leaks kernel state to user space. We have a software workaround for that ("espfix") for the 32-bit kernel, but it relies on a nonzero stack segment base which is not available in 64-bit mode. In checkin: b3b42ac2cbae x86-64, modify_ldt: Ban 16-bit segments on 64-bit kernels we "solved" this by forbidding 16-bit segments on 64-bit kernels, with the logic that 16-bit support is crippled on 64-bit kernels anyway (no V86 support), but it turns out that people are doing stuff like running old Win16 binaries under Wine and expect it to work. This works around this by creating percpu "ministacks", each of which is mapped 2^16 times 64K apart. When we detect that the return SS is on the LDT, we copy the IRET frame to the ministack and use the relevant alias to return to userspace. The ministacks are mapped readonly, so if IRET faults we promote #GP to #DF which is an IST vector and thus has its own stack; we then do the fixup in the #DF handler. (Making #GP an IST exception would make the msr_safe functions unsafe in NMI/MC context, and quite possibly have other effects.) Special thanks to: - Andy Lutomirski, for the suggestion of using very small stack slots and copy (as opposed to map) the IRET frame there, and for the suggestion to mark them readonly and let the fault promote to #DF. - Konrad Wilk for paravirt fixup and testing. - Borislav Petkov for testing help and useful comments. Reported-by: Brian Gerst <brgerst@gmail.com> Signed-off-by: H. Peter Anvin <hpa@linux.intel.com> Link: http://lkml.kernel.org/r/1398816946-3351-1-git-send-email-hpa@linux.intel.com Cc: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com> Cc: Borislav Petkov <bp@alien8.de> Cc: Andrew Lutomriski <amluto@gmail.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Dirk Hohndel <dirk@hohndel.org> Cc: Arjan van de Ven <arjan.van.de.ven@intel.com> Cc: comex <comexk@gmail.com> Cc: Alexander van Heukelum <heukelum@fastmail.fm> Cc: Boris Ostrovsky <boris.ostrovsky@oracle.com> Cc: <stable@vger.kernel.org> # consider after upstream merge
2014-04-30 03:46:09 +04:00
SWAPGS
movq %rax, %rsp
popq %rax
jmp native_irq_return_iret
#endif
x86-64, espfix: Don't leak bits 31:16 of %esp returning to 16-bit stack The IRET instruction, when returning to a 16-bit segment, only restores the bottom 16 bits of the user space stack pointer. This causes some 16-bit software to break, but it also leaks kernel state to user space. We have a software workaround for that ("espfix") for the 32-bit kernel, but it relies on a nonzero stack segment base which is not available in 64-bit mode. In checkin: b3b42ac2cbae x86-64, modify_ldt: Ban 16-bit segments on 64-bit kernels we "solved" this by forbidding 16-bit segments on 64-bit kernels, with the logic that 16-bit support is crippled on 64-bit kernels anyway (no V86 support), but it turns out that people are doing stuff like running old Win16 binaries under Wine and expect it to work. This works around this by creating percpu "ministacks", each of which is mapped 2^16 times 64K apart. When we detect that the return SS is on the LDT, we copy the IRET frame to the ministack and use the relevant alias to return to userspace. The ministacks are mapped readonly, so if IRET faults we promote #GP to #DF which is an IST vector and thus has its own stack; we then do the fixup in the #DF handler. (Making #GP an IST exception would make the msr_safe functions unsafe in NMI/MC context, and quite possibly have other effects.) Special thanks to: - Andy Lutomirski, for the suggestion of using very small stack slots and copy (as opposed to map) the IRET frame there, and for the suggestion to mark them readonly and let the fault promote to #DF. - Konrad Wilk for paravirt fixup and testing. - Borislav Petkov for testing help and useful comments. Reported-by: Brian Gerst <brgerst@gmail.com> Signed-off-by: H. Peter Anvin <hpa@linux.intel.com> Link: http://lkml.kernel.org/r/1398816946-3351-1-git-send-email-hpa@linux.intel.com Cc: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com> Cc: Borislav Petkov <bp@alien8.de> Cc: Andrew Lutomriski <amluto@gmail.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Dirk Hohndel <dirk@hohndel.org> Cc: Arjan van de Ven <arjan.van.de.ven@intel.com> Cc: comex <comexk@gmail.com> Cc: Alexander van Heukelum <heukelum@fastmail.fm> Cc: Boris Ostrovsky <boris.ostrovsky@oracle.com> Cc: <stable@vger.kernel.org> # consider after upstream merge
2014-04-30 03:46:09 +04:00
/* edi: workmask, edx: work */
retint_careful:
bt $TIF_NEED_RESCHED, %edx
jnc retint_signal
TRACE_IRQS_ON
ENABLE_INTERRUPTS(CLBR_NONE)
pushq %rdi
SCHEDULE_USER
popq %rdi
GET_THREAD_INFO(%rcx)
DISABLE_INTERRUPTS(CLBR_NONE)
TRACE_IRQS_OFF
jmp retint_check
retint_signal:
testl $_TIF_DO_NOTIFY_MASK, %edx
jz retint_swapgs
TRACE_IRQS_ON
ENABLE_INTERRUPTS(CLBR_NONE)
movq $-1, ORIG_RAX(%rsp)
xorl %esi, %esi /* oldset */
movq %rsp, %rdi /* &pt_regs */
call do_notify_resume
DISABLE_INTERRUPTS(CLBR_NONE)
TRACE_IRQS_OFF
GET_THREAD_INFO(%rcx)
jmp retint_with_reschedule
END(common_interrupt)
x86-64, espfix: Don't leak bits 31:16 of %esp returning to 16-bit stack The IRET instruction, when returning to a 16-bit segment, only restores the bottom 16 bits of the user space stack pointer. This causes some 16-bit software to break, but it also leaks kernel state to user space. We have a software workaround for that ("espfix") for the 32-bit kernel, but it relies on a nonzero stack segment base which is not available in 64-bit mode. In checkin: b3b42ac2cbae x86-64, modify_ldt: Ban 16-bit segments on 64-bit kernels we "solved" this by forbidding 16-bit segments on 64-bit kernels, with the logic that 16-bit support is crippled on 64-bit kernels anyway (no V86 support), but it turns out that people are doing stuff like running old Win16 binaries under Wine and expect it to work. This works around this by creating percpu "ministacks", each of which is mapped 2^16 times 64K apart. When we detect that the return SS is on the LDT, we copy the IRET frame to the ministack and use the relevant alias to return to userspace. The ministacks are mapped readonly, so if IRET faults we promote #GP to #DF which is an IST vector and thus has its own stack; we then do the fixup in the #DF handler. (Making #GP an IST exception would make the msr_safe functions unsafe in NMI/MC context, and quite possibly have other effects.) Special thanks to: - Andy Lutomirski, for the suggestion of using very small stack slots and copy (as opposed to map) the IRET frame there, and for the suggestion to mark them readonly and let the fault promote to #DF. - Konrad Wilk for paravirt fixup and testing. - Borislav Petkov for testing help and useful comments. Reported-by: Brian Gerst <brgerst@gmail.com> Signed-off-by: H. Peter Anvin <hpa@linux.intel.com> Link: http://lkml.kernel.org/r/1398816946-3351-1-git-send-email-hpa@linux.intel.com Cc: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com> Cc: Borislav Petkov <bp@alien8.de> Cc: Andrew Lutomriski <amluto@gmail.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Dirk Hohndel <dirk@hohndel.org> Cc: Arjan van de Ven <arjan.van.de.ven@intel.com> Cc: comex <comexk@gmail.com> Cc: Alexander van Heukelum <heukelum@fastmail.fm> Cc: Boris Ostrovsky <boris.ostrovsky@oracle.com> Cc: <stable@vger.kernel.org> # consider after upstream merge
2014-04-30 03:46:09 +04:00
/*
* APIC interrupts.
*/
x86, trace: Add irq vector tracepoints [Purpose of this patch] As Vaibhav explained in the thread below, tracepoints for irq vectors are useful. http://www.spinics.net/lists/mm-commits/msg85707.html <snip> The current interrupt traces from irq_handler_entry and irq_handler_exit provide when an interrupt is handled. They provide good data about when the system has switched to kernel space and how it affects the currently running processes. There are some IRQ vectors which trigger the system into kernel space, which are not handled in generic IRQ handlers. Tracing such events gives us the information about IRQ interaction with other system events. The trace also tells where the system is spending its time. We want to know which cores are handling interrupts and how they are affecting other processes in the system. Also, the trace provides information about when the cores are idle and which interrupts are changing that state. <snip> On the other hand, my usecase is tracing just local timer event and getting a value of instruction pointer. I suggested to add an argument local timer event to get instruction pointer before. But there is another way to get it with external module like systemtap. So, I don't need to add any argument to irq vector tracepoints now. [Patch Description] Vaibhav's patch shared a trace point ,irq_vector_entry/irq_vector_exit, in all events. But there is an above use case to trace specific irq_vector rather than tracing all events. In this case, we are concerned about overhead due to unwanted events. So, add following tracepoints instead of introducing irq_vector_entry/exit. so that we can enable them independently. - local_timer_vector - reschedule_vector - call_function_vector - call_function_single_vector - irq_work_entry_vector - error_apic_vector - thermal_apic_vector - threshold_apic_vector - spurious_apic_vector - x86_platform_ipi_vector Also, introduce a logic switching IDT at enabling/disabling time so that a time penalty makes a zero when tracepoints are disabled. Detailed explanations are as follows. - Create trace irq handlers with entering_irq()/exiting_irq(). - Create a new IDT, trace_idt_table, at boot time by adding a logic to _set_gate(). It is just a copy of original idt table. - Register the new handlers for tracpoints to the new IDT by introducing macros to alloc_intr_gate() called at registering time of irq_vector handlers. - Add checking, whether irq vector tracing is on/off, into load_current_idt(). This has to be done below debug checking for these reasons. - Switching to debug IDT may be kicked while tracing is enabled. - On the other hands, switching to trace IDT is kicked only when debugging is disabled. In addition, the new IDT is created only when CONFIG_TRACING is enabled to avoid being used for other purposes. Signed-off-by: Seiji Aguchi <seiji.aguchi@hds.com> Link: http://lkml.kernel.org/r/51C323ED.5050708@hds.com Signed-off-by: H. Peter Anvin <hpa@linux.intel.com> Cc: Steven Rostedt <rostedt@goodmis.org>
2013-06-20 19:46:53 +04:00
.macro apicinterrupt3 num sym do_sym
ENTRY(\sym)
ASM_CLAC
pushq $~(\num)
.Lcommon_\sym:
interrupt \do_sym
jmp ret_from_intr
END(\sym)
.endm
x86, trace: Add irq vector tracepoints [Purpose of this patch] As Vaibhav explained in the thread below, tracepoints for irq vectors are useful. http://www.spinics.net/lists/mm-commits/msg85707.html <snip> The current interrupt traces from irq_handler_entry and irq_handler_exit provide when an interrupt is handled. They provide good data about when the system has switched to kernel space and how it affects the currently running processes. There are some IRQ vectors which trigger the system into kernel space, which are not handled in generic IRQ handlers. Tracing such events gives us the information about IRQ interaction with other system events. The trace also tells where the system is spending its time. We want to know which cores are handling interrupts and how they are affecting other processes in the system. Also, the trace provides information about when the cores are idle and which interrupts are changing that state. <snip> On the other hand, my usecase is tracing just local timer event and getting a value of instruction pointer. I suggested to add an argument local timer event to get instruction pointer before. But there is another way to get it with external module like systemtap. So, I don't need to add any argument to irq vector tracepoints now. [Patch Description] Vaibhav's patch shared a trace point ,irq_vector_entry/irq_vector_exit, in all events. But there is an above use case to trace specific irq_vector rather than tracing all events. In this case, we are concerned about overhead due to unwanted events. So, add following tracepoints instead of introducing irq_vector_entry/exit. so that we can enable them independently. - local_timer_vector - reschedule_vector - call_function_vector - call_function_single_vector - irq_work_entry_vector - error_apic_vector - thermal_apic_vector - threshold_apic_vector - spurious_apic_vector - x86_platform_ipi_vector Also, introduce a logic switching IDT at enabling/disabling time so that a time penalty makes a zero when tracepoints are disabled. Detailed explanations are as follows. - Create trace irq handlers with entering_irq()/exiting_irq(). - Create a new IDT, trace_idt_table, at boot time by adding a logic to _set_gate(). It is just a copy of original idt table. - Register the new handlers for tracpoints to the new IDT by introducing macros to alloc_intr_gate() called at registering time of irq_vector handlers. - Add checking, whether irq vector tracing is on/off, into load_current_idt(). This has to be done below debug checking for these reasons. - Switching to debug IDT may be kicked while tracing is enabled. - On the other hands, switching to trace IDT is kicked only when debugging is disabled. In addition, the new IDT is created only when CONFIG_TRACING is enabled to avoid being used for other purposes. Signed-off-by: Seiji Aguchi <seiji.aguchi@hds.com> Link: http://lkml.kernel.org/r/51C323ED.5050708@hds.com Signed-off-by: H. Peter Anvin <hpa@linux.intel.com> Cc: Steven Rostedt <rostedt@goodmis.org>
2013-06-20 19:46:53 +04:00
#ifdef CONFIG_TRACING
#define trace(sym) trace_##sym
#define smp_trace(sym) smp_trace_##sym
.macro trace_apicinterrupt num sym
apicinterrupt3 \num trace(\sym) smp_trace(\sym)
.endm
#else
.macro trace_apicinterrupt num sym do_sym
.endm
#endif
.macro apicinterrupt num sym do_sym
apicinterrupt3 \num \sym \do_sym
trace_apicinterrupt \num \sym
.endm
#ifdef CONFIG_SMP
apicinterrupt3 IRQ_MOVE_CLEANUP_VECTOR irq_move_cleanup_interrupt smp_irq_move_cleanup_interrupt
apicinterrupt3 REBOOT_VECTOR reboot_interrupt smp_reboot_interrupt
#endif
#ifdef CONFIG_X86_UV
apicinterrupt3 UV_BAU_MESSAGE uv_bau_message_intr1 uv_bau_message_interrupt
#endif
apicinterrupt LOCAL_TIMER_VECTOR apic_timer_interrupt smp_apic_timer_interrupt
apicinterrupt X86_PLATFORM_IPI_VECTOR x86_platform_ipi smp_x86_platform_ipi
#ifdef CONFIG_HAVE_KVM
apicinterrupt3 POSTED_INTR_VECTOR kvm_posted_intr_ipi smp_kvm_posted_intr_ipi
apicinterrupt3 POSTED_INTR_WAKEUP_VECTOR kvm_posted_intr_wakeup_ipi smp_kvm_posted_intr_wakeup_ipi
#endif
#ifdef CONFIG_X86_MCE_THRESHOLD
apicinterrupt THRESHOLD_APIC_VECTOR threshold_interrupt smp_threshold_interrupt
#endif
#ifdef CONFIG_X86_MCE_AMD
apicinterrupt DEFERRED_ERROR_VECTOR deferred_error_interrupt smp_deferred_error_interrupt
#endif
#ifdef CONFIG_X86_THERMAL_VECTOR
apicinterrupt THERMAL_APIC_VECTOR thermal_interrupt smp_thermal_interrupt
#endif
#ifdef CONFIG_SMP
apicinterrupt CALL_FUNCTION_SINGLE_VECTOR call_function_single_interrupt smp_call_function_single_interrupt
apicinterrupt CALL_FUNCTION_VECTOR call_function_interrupt smp_call_function_interrupt
apicinterrupt RESCHEDULE_VECTOR reschedule_interrupt smp_reschedule_interrupt
#endif
apicinterrupt ERROR_APIC_VECTOR error_interrupt smp_error_interrupt
apicinterrupt SPURIOUS_APIC_VECTOR spurious_interrupt smp_spurious_interrupt
#ifdef CONFIG_IRQ_WORK
apicinterrupt IRQ_WORK_VECTOR irq_work_interrupt smp_irq_work_interrupt
#endif
/*
* Exception entry points.
*/
#define CPU_TSS_IST(x) PER_CPU_VAR(cpu_tss) + (TSS_ist + ((x) - 1) * 8)
.macro idtentry sym do_sym has_error_code:req paranoid=0 shift_ist=-1
ENTRY(\sym)
/* Sanity check */
.if \shift_ist != -1 && \paranoid == 0
.error "using shift_ist requires paranoid=1"
.endif
ASM_CLAC
PARAVIRT_ADJUST_EXCEPTION_FRAME
.ifeq \has_error_code
pushq $-1 /* ORIG_RAX: no syscall to restart */
.endif
x86/asm/entry/64: Always allocate a complete "struct pt_regs" on the kernel stack The 64-bit entry code was using six stack slots less by not saving/restoring registers which are callee-preserved according to the C ABI, and was not allocating space for them. Only when syscalls needed a complete "struct pt_regs" was the complete area allocated and filled in. As an additional twist, on interrupt entry a "slightly less truncated pt_regs" trick is used, to make nested interrupt stacks easier to unwind. This proved to be a source of significant obfuscation and subtle bugs. For example, 'stub_fork' had to pop the return address, extend the struct, save registers, and push return address back. Ugly. 'ia32_ptregs_common' pops return address and "returns" via jmp insn, throwing a wrench into CPU return stack cache. This patch changes the code to always allocate a complete "struct pt_regs" on the kernel stack. The saving of registers is still done lazily. "Partial pt_regs" trick on interrupt stack is retained. Macros which manipulate "struct pt_regs" on stack are reworked: - ALLOC_PT_GPREGS_ON_STACK allocates the structure. - SAVE_C_REGS saves to it those registers which are clobbered by C code. - SAVE_EXTRA_REGS saves to it all other registers. - Corresponding RESTORE_* and REMOVE_PT_GPREGS_FROM_STACK macros reverse it. 'ia32_ptregs_common', 'stub_fork' and friends lost their ugly dance with the return pointer. LOAD_ARGS32 in ia32entry.S now uses symbolic stack offsets instead of magic numbers. 'error_entry' and 'save_paranoid' now use SAVE_C_REGS + SAVE_EXTRA_REGS instead of having it open-coded yet again. Patch was run-tested: 64-bit executables, 32-bit executables, strace works. Timing tests did not show measurable difference in 32-bit and 64-bit syscalls. Signed-off-by: Denys Vlasenko <dvlasenk@redhat.com> Signed-off-by: Andy Lutomirski <luto@amacapital.net> Cc: Alexei Starovoitov <ast@plumgrid.com> Cc: Borislav Petkov <bp@alien8.de> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: H. Peter Anvin <hpa@zytor.com> Cc: Kees Cook <keescook@chromium.org> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Will Drewry <wad@chromium.org> Link: http://lkml.kernel.org/r/1423778052-21038-2-git-send-email-dvlasenk@redhat.com Link: http://lkml.kernel.org/r/b89763d354aa23e670b9bdf3a40ae320320a7c2e.1424989793.git.luto@amacapital.net Signed-off-by: Ingo Molnar <mingo@kernel.org>
2015-02-27 01:40:27 +03:00
ALLOC_PT_GPREGS_ON_STACK
.if \paranoid
.if \paranoid == 1
testb $3, CS(%rsp) /* If coming from userspace, switch stacks */
jnz 1f
.endif
call paranoid_entry
.else
call error_entry
.endif
/* returned flag: ebx=0: need swapgs on exit, ebx=1: don't need it */
.if \paranoid
.if \shift_ist != -1
TRACE_IRQS_OFF_DEBUG /* reload IDT in case of recursion */
.else
TRACE_IRQS_OFF
.endif
.endif
movq %rsp, %rdi /* pt_regs pointer */
.if \has_error_code
movq ORIG_RAX(%rsp), %rsi /* get error code */
movq $-1, ORIG_RAX(%rsp) /* no syscall to restart */
.else
xorl %esi, %esi /* no error code */
.endif
.if \shift_ist != -1
subq $EXCEPTION_STKSZ, CPU_TSS_IST(\shift_ist)
.endif
call \do_sym
.if \shift_ist != -1
addq $EXCEPTION_STKSZ, CPU_TSS_IST(\shift_ist)
.endif
/* these procedures expect "no swapgs" flag in ebx */
.if \paranoid
jmp paranoid_exit
.else
jmp error_exit
.endif
.if \paranoid == 1
/*
* Paranoid entry from userspace. Switch stacks and treat it
* as a normal entry. This means that paranoid handlers
* run in real process context if user_mode(regs).
*/
1:
call error_entry
movq %rsp, %rdi /* pt_regs pointer */
call sync_regs
movq %rax, %rsp /* switch stack */
movq %rsp, %rdi /* pt_regs pointer */
.if \has_error_code
movq ORIG_RAX(%rsp), %rsi /* get error code */
movq $-1, ORIG_RAX(%rsp) /* no syscall to restart */
.else
xorl %esi, %esi /* no error code */
.endif
call \do_sym
jmp error_exit /* %ebx: no swapgs flag */
.endif
END(\sym)
.endm
#ifdef CONFIG_TRACING
.macro trace_idtentry sym do_sym has_error_code:req
idtentry trace(\sym) trace(\do_sym) has_error_code=\has_error_code
idtentry \sym \do_sym has_error_code=\has_error_code
.endm
#else
.macro trace_idtentry sym do_sym has_error_code:req
idtentry \sym \do_sym has_error_code=\has_error_code
.endm
#endif
idtentry divide_error do_divide_error has_error_code=0
idtentry overflow do_overflow has_error_code=0
idtentry bounds do_bounds has_error_code=0
idtentry invalid_op do_invalid_op has_error_code=0
idtentry device_not_available do_device_not_available has_error_code=0
idtentry double_fault do_double_fault has_error_code=1 paranoid=2
idtentry coprocessor_segment_overrun do_coprocessor_segment_overrun has_error_code=0
idtentry invalid_TSS do_invalid_TSS has_error_code=1
idtentry segment_not_present do_segment_not_present has_error_code=1
idtentry spurious_interrupt_bug do_spurious_interrupt_bug has_error_code=0
idtentry coprocessor_error do_coprocessor_error has_error_code=0
idtentry alignment_check do_alignment_check has_error_code=1
idtentry simd_coprocessor_error do_simd_coprocessor_error has_error_code=0
/*
* Reload gs selector with exception handling
* edi: new selector
*/
ENTRY(native_load_gs_index)
x86/debug: Remove perpetually broken, unmaintainable dwarf annotations So the dwarf2 annotations in low level assembly code have become an increasing hindrance: unreadable, messy macros mixed into some of the most security sensitive code paths of the Linux kernel. These debug info annotations don't even buy the upstream kernel anything: dwarf driven stack unwinding has caused problems in the past so it's out of tree, and the upstream kernel only uses the much more robust framepointers based stack unwinding method. In addition to that there's a steady, slow bitrot going on with these annotations, requiring frequent fixups. There's no tooling and no functionality upstream that keeps it correct. So burn down the sick forest, allowing new, healthier growth: 27 files changed, 350 insertions(+), 1101 deletions(-) Someone who has the willingness and time to do this properly can attempt to reintroduce dwarf debuginfo in x86 assembly code plus dwarf unwinding from first principles, with the following conditions: - it should be maximally readable, and maximally low-key to 'ordinary' code reading and maintenance. - find a build time method to insert dwarf annotations automatically in the most common cases, for pop/push instructions that manipulate the stack pointer. This could be done for example via a preprocessing step that just looks for common patterns - plus special annotations for the few cases where we want to depart from the default. We have hundreds of CFI annotations, so automating most of that makes sense. - it should come with build tooling checks that ensure that CFI annotations are sensible. We've seen such efforts from the framepointer side, and there's no reason it couldn't be done on the dwarf side. Cc: Andy Lutomirski <luto@amacapital.net> Cc: Borislav Petkov <bp@alien8.de> Cc: Brian Gerst <brgerst@gmail.com> Cc: Denys Vlasenko <dvlasenk@redhat.com> Cc: Frédéric Weisbecker <fweisbec@gmail.com Cc: H. Peter Anvin <hpa@zytor.com> Cc: Jan Beulich <JBeulich@suse.com> Cc: Josh Poimboeuf <jpoimboe@redhat.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: linux-kernel@vger.kernel.org Signed-off-by: Ingo Molnar <mingo@kernel.org>
2015-05-28 13:21:47 +03:00
pushfq
DISABLE_INTERRUPTS(CLBR_ANY & ~CLBR_RDI)
SWAPGS
gs_change:
movl %edi, %gs
2: mfence /* workaround */
SWAPGS
x86/debug: Remove perpetually broken, unmaintainable dwarf annotations So the dwarf2 annotations in low level assembly code have become an increasing hindrance: unreadable, messy macros mixed into some of the most security sensitive code paths of the Linux kernel. These debug info annotations don't even buy the upstream kernel anything: dwarf driven stack unwinding has caused problems in the past so it's out of tree, and the upstream kernel only uses the much more robust framepointers based stack unwinding method. In addition to that there's a steady, slow bitrot going on with these annotations, requiring frequent fixups. There's no tooling and no functionality upstream that keeps it correct. So burn down the sick forest, allowing new, healthier growth: 27 files changed, 350 insertions(+), 1101 deletions(-) Someone who has the willingness and time to do this properly can attempt to reintroduce dwarf debuginfo in x86 assembly code plus dwarf unwinding from first principles, with the following conditions: - it should be maximally readable, and maximally low-key to 'ordinary' code reading and maintenance. - find a build time method to insert dwarf annotations automatically in the most common cases, for pop/push instructions that manipulate the stack pointer. This could be done for example via a preprocessing step that just looks for common patterns - plus special annotations for the few cases where we want to depart from the default. We have hundreds of CFI annotations, so automating most of that makes sense. - it should come with build tooling checks that ensure that CFI annotations are sensible. We've seen such efforts from the framepointer side, and there's no reason it couldn't be done on the dwarf side. Cc: Andy Lutomirski <luto@amacapital.net> Cc: Borislav Petkov <bp@alien8.de> Cc: Brian Gerst <brgerst@gmail.com> Cc: Denys Vlasenko <dvlasenk@redhat.com> Cc: Frédéric Weisbecker <fweisbec@gmail.com Cc: H. Peter Anvin <hpa@zytor.com> Cc: Jan Beulich <JBeulich@suse.com> Cc: Josh Poimboeuf <jpoimboe@redhat.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: linux-kernel@vger.kernel.org Signed-off-by: Ingo Molnar <mingo@kernel.org>
2015-05-28 13:21:47 +03:00
popfq
ret
END(native_load_gs_index)
_ASM_EXTABLE(gs_change, bad_gs)
.section .fixup, "ax"
/* running with kernelgs */
bad_gs:
SWAPGS /* switch back to user gs */
xorl %eax, %eax
movl %eax, %gs
jmp 2b
.previous
/* Call softirq on interrupt stack. Interrupts are off. */
ENTRY(do_softirq_own_stack)
pushq %rbp
mov %rsp, %rbp
incl PER_CPU_VAR(irq_count)
cmove PER_CPU_VAR(irq_stack_ptr), %rsp
push %rbp /* frame pointer backlink */
call __do_softirq
leaveq
decl PER_CPU_VAR(irq_count)
ret
END(do_softirq_own_stack)
#ifdef CONFIG_XEN
idtentry xen_hypervisor_callback xen_do_hypervisor_callback has_error_code=0
/*
* A note on the "critical region" in our callback handler.
* We want to avoid stacking callback handlers due to events occurring
* during handling of the last event. To do this, we keep events disabled
* until we've done all processing. HOWEVER, we must enable events before
* popping the stack frame (can't be done atomically) and so it would still
* be possible to get enough handler activations to overflow the stack.
* Although unlikely, bugs of that kind are hard to track down, so we'd
* like to avoid the possibility.
* So, on entry to the handler we detect whether we interrupted an
* existing activation in its critical region -- if so, we pop the current
* activation and restart the handler using the previous one.
*/
ENTRY(xen_do_hypervisor_callback) /* do_hypervisor_callback(struct *pt_regs) */
/*
* Since we don't modify %rdi, evtchn_do_upall(struct *pt_regs) will
* see the correct pointer to the pt_regs
*/
movq %rdi, %rsp /* we don't return, adjust the stack frame */
11: incl PER_CPU_VAR(irq_count)
movq %rsp, %rbp
cmovzq PER_CPU_VAR(irq_stack_ptr), %rsp
pushq %rbp /* frame pointer backlink */
call xen_evtchn_do_upcall
popq %rsp
decl PER_CPU_VAR(irq_count)
#ifndef CONFIG_PREEMPT
call xen_maybe_preempt_hcall
#endif
jmp error_exit
x86, binutils, xen: Fix another wrong size directive The latest binutils (2.21.0.20110302/Ubuntu) breaks the build yet another time, under CONFIG_XEN=y due to a .size directive that refers to a slightly differently named (hence, to the now very strict and unforgiving assembler, non-existent) symbol. [ mingo: This unnecessary build breakage caused by new binutils version 2.21 gets escallated back several kernel releases spanning several years of Linux history, affecting over 130,000 upstream kernel commits (!), on CONFIG_XEN=y 64-bit kernels (i.e. essentially affecting all major Linux distro kernel configs). Git annotate tells us that this slight debug symbol code mismatch bug has been introduced in 2008 in commit 3d75e1b8: 3d75e1b8 (Jeremy Fitzhardinge 2008-07-08 15:06:49 -0700 1231) ENTRY(xen_do_hypervisor_callback) # do_hypervisor_callback(struct *pt_regs) The 'bug' is just a slight assymetry in ENTRY()/END() debug-symbols sequences, with lots of assembly code between the ENTRY() and the END(): ENTRY(xen_do_hypervisor_callback) # do_hypervisor_callback(struct *pt_regs) ... END(do_hypervisor_callback) Human reviewers almost never catch such small mismatches, and binutils never even warned about it either. This new binutils version thus breaks the Xen build on all upstream kernels since v2.6.27, out of the blue. This makes a straightforward Git bisection of all 64-bit Xen-enabled kernels impossible on such binutils, for a bisection window of over hundred thousand historic commits. (!) This is a major fail on the side of binutils and binutils needs to turn this show-stopper build failure into a warning ASAP. ] Signed-off-by: Alexander van Heukelum <heukelum@fastmail.fm> Cc: Jeremy Fitzhardinge <jeremy@goop.org> Cc: Jan Beulich <jbeulich@novell.com> Cc: H.J. Lu <hjl.tools@gmail.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Andrew Morton <akpm@linux-foundation.org> Cc: "H. Peter Anvin" <hpa@zytor.com> Cc: Kees Cook <kees.cook@canonical.com> LKML-Reference: <1299877178-26063-1-git-send-email-heukelum@fastmail.fm> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2011-03-11 23:59:38 +03:00
END(xen_do_hypervisor_callback)
/*
* Hypervisor uses this for application faults while it executes.
* We get here for two reasons:
* 1. Fault while reloading DS, ES, FS or GS
* 2. Fault while executing IRET
* Category 1 we do not need to fix up as Xen has already reloaded all segment
* registers that could be reloaded and zeroed the others.
* Category 2 we fix up by killing the current process. We cannot use the
* normal Linux return path in this case because if we use the IRET hypercall
* to pop the stack frame we end up in an infinite loop of failsafe callbacks.
* We distinguish between categories by comparing each saved segment register
* with its current contents: any discrepancy means we in category 1.
*/
ENTRY(xen_failsafe_callback)
movl %ds, %ecx
cmpw %cx, 0x10(%rsp)
jne 1f
movl %es, %ecx
cmpw %cx, 0x18(%rsp)
jne 1f
movl %fs, %ecx
cmpw %cx, 0x20(%rsp)
jne 1f
movl %gs, %ecx
cmpw %cx, 0x28(%rsp)
jne 1f
/* All segments match their saved values => Category 2 (Bad IRET). */
movq (%rsp), %rcx
movq 8(%rsp), %r11
addq $0x30, %rsp
pushq $0 /* RIP */
pushq %r11
pushq %rcx
jmp general_protection
1: /* Segment mismatch => Category 1 (Bad segment). Retry the IRET. */
movq (%rsp), %rcx
movq 8(%rsp), %r11
addq $0x30, %rsp
pushq $-1 /* orig_ax = -1 => not a system call */
x86/asm/entry/64: Always allocate a complete "struct pt_regs" on the kernel stack The 64-bit entry code was using six stack slots less by not saving/restoring registers which are callee-preserved according to the C ABI, and was not allocating space for them. Only when syscalls needed a complete "struct pt_regs" was the complete area allocated and filled in. As an additional twist, on interrupt entry a "slightly less truncated pt_regs" trick is used, to make nested interrupt stacks easier to unwind. This proved to be a source of significant obfuscation and subtle bugs. For example, 'stub_fork' had to pop the return address, extend the struct, save registers, and push return address back. Ugly. 'ia32_ptregs_common' pops return address and "returns" via jmp insn, throwing a wrench into CPU return stack cache. This patch changes the code to always allocate a complete "struct pt_regs" on the kernel stack. The saving of registers is still done lazily. "Partial pt_regs" trick on interrupt stack is retained. Macros which manipulate "struct pt_regs" on stack are reworked: - ALLOC_PT_GPREGS_ON_STACK allocates the structure. - SAVE_C_REGS saves to it those registers which are clobbered by C code. - SAVE_EXTRA_REGS saves to it all other registers. - Corresponding RESTORE_* and REMOVE_PT_GPREGS_FROM_STACK macros reverse it. 'ia32_ptregs_common', 'stub_fork' and friends lost their ugly dance with the return pointer. LOAD_ARGS32 in ia32entry.S now uses symbolic stack offsets instead of magic numbers. 'error_entry' and 'save_paranoid' now use SAVE_C_REGS + SAVE_EXTRA_REGS instead of having it open-coded yet again. Patch was run-tested: 64-bit executables, 32-bit executables, strace works. Timing tests did not show measurable difference in 32-bit and 64-bit syscalls. Signed-off-by: Denys Vlasenko <dvlasenk@redhat.com> Signed-off-by: Andy Lutomirski <luto@amacapital.net> Cc: Alexei Starovoitov <ast@plumgrid.com> Cc: Borislav Petkov <bp@alien8.de> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: H. Peter Anvin <hpa@zytor.com> Cc: Kees Cook <keescook@chromium.org> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Will Drewry <wad@chromium.org> Link: http://lkml.kernel.org/r/1423778052-21038-2-git-send-email-dvlasenk@redhat.com Link: http://lkml.kernel.org/r/b89763d354aa23e670b9bdf3a40ae320320a7c2e.1424989793.git.luto@amacapital.net Signed-off-by: Ingo Molnar <mingo@kernel.org>
2015-02-27 01:40:27 +03:00
ALLOC_PT_GPREGS_ON_STACK
SAVE_C_REGS
SAVE_EXTRA_REGS
jmp error_exit
END(xen_failsafe_callback)
x86, trace: Add irq vector tracepoints [Purpose of this patch] As Vaibhav explained in the thread below, tracepoints for irq vectors are useful. http://www.spinics.net/lists/mm-commits/msg85707.html <snip> The current interrupt traces from irq_handler_entry and irq_handler_exit provide when an interrupt is handled. They provide good data about when the system has switched to kernel space and how it affects the currently running processes. There are some IRQ vectors which trigger the system into kernel space, which are not handled in generic IRQ handlers. Tracing such events gives us the information about IRQ interaction with other system events. The trace also tells where the system is spending its time. We want to know which cores are handling interrupts and how they are affecting other processes in the system. Also, the trace provides information about when the cores are idle and which interrupts are changing that state. <snip> On the other hand, my usecase is tracing just local timer event and getting a value of instruction pointer. I suggested to add an argument local timer event to get instruction pointer before. But there is another way to get it with external module like systemtap. So, I don't need to add any argument to irq vector tracepoints now. [Patch Description] Vaibhav's patch shared a trace point ,irq_vector_entry/irq_vector_exit, in all events. But there is an above use case to trace specific irq_vector rather than tracing all events. In this case, we are concerned about overhead due to unwanted events. So, add following tracepoints instead of introducing irq_vector_entry/exit. so that we can enable them independently. - local_timer_vector - reschedule_vector - call_function_vector - call_function_single_vector - irq_work_entry_vector - error_apic_vector - thermal_apic_vector - threshold_apic_vector - spurious_apic_vector - x86_platform_ipi_vector Also, introduce a logic switching IDT at enabling/disabling time so that a time penalty makes a zero when tracepoints are disabled. Detailed explanations are as follows. - Create trace irq handlers with entering_irq()/exiting_irq(). - Create a new IDT, trace_idt_table, at boot time by adding a logic to _set_gate(). It is just a copy of original idt table. - Register the new handlers for tracpoints to the new IDT by introducing macros to alloc_intr_gate() called at registering time of irq_vector handlers. - Add checking, whether irq vector tracing is on/off, into load_current_idt(). This has to be done below debug checking for these reasons. - Switching to debug IDT may be kicked while tracing is enabled. - On the other hands, switching to trace IDT is kicked only when debugging is disabled. In addition, the new IDT is created only when CONFIG_TRACING is enabled to avoid being used for other purposes. Signed-off-by: Seiji Aguchi <seiji.aguchi@hds.com> Link: http://lkml.kernel.org/r/51C323ED.5050708@hds.com Signed-off-by: H. Peter Anvin <hpa@linux.intel.com> Cc: Steven Rostedt <rostedt@goodmis.org>
2013-06-20 19:46:53 +04:00
apicinterrupt3 HYPERVISOR_CALLBACK_VECTOR \
xen_hvm_callback_vector xen_evtchn_do_upcall
#endif /* CONFIG_XEN */
#if IS_ENABLED(CONFIG_HYPERV)
x86, trace: Add irq vector tracepoints [Purpose of this patch] As Vaibhav explained in the thread below, tracepoints for irq vectors are useful. http://www.spinics.net/lists/mm-commits/msg85707.html <snip> The current interrupt traces from irq_handler_entry and irq_handler_exit provide when an interrupt is handled. They provide good data about when the system has switched to kernel space and how it affects the currently running processes. There are some IRQ vectors which trigger the system into kernel space, which are not handled in generic IRQ handlers. Tracing such events gives us the information about IRQ interaction with other system events. The trace also tells where the system is spending its time. We want to know which cores are handling interrupts and how they are affecting other processes in the system. Also, the trace provides information about when the cores are idle and which interrupts are changing that state. <snip> On the other hand, my usecase is tracing just local timer event and getting a value of instruction pointer. I suggested to add an argument local timer event to get instruction pointer before. But there is another way to get it with external module like systemtap. So, I don't need to add any argument to irq vector tracepoints now. [Patch Description] Vaibhav's patch shared a trace point ,irq_vector_entry/irq_vector_exit, in all events. But there is an above use case to trace specific irq_vector rather than tracing all events. In this case, we are concerned about overhead due to unwanted events. So, add following tracepoints instead of introducing irq_vector_entry/exit. so that we can enable them independently. - local_timer_vector - reschedule_vector - call_function_vector - call_function_single_vector - irq_work_entry_vector - error_apic_vector - thermal_apic_vector - threshold_apic_vector - spurious_apic_vector - x86_platform_ipi_vector Also, introduce a logic switching IDT at enabling/disabling time so that a time penalty makes a zero when tracepoints are disabled. Detailed explanations are as follows. - Create trace irq handlers with entering_irq()/exiting_irq(). - Create a new IDT, trace_idt_table, at boot time by adding a logic to _set_gate(). It is just a copy of original idt table. - Register the new handlers for tracpoints to the new IDT by introducing macros to alloc_intr_gate() called at registering time of irq_vector handlers. - Add checking, whether irq vector tracing is on/off, into load_current_idt(). This has to be done below debug checking for these reasons. - Switching to debug IDT may be kicked while tracing is enabled. - On the other hands, switching to trace IDT is kicked only when debugging is disabled. In addition, the new IDT is created only when CONFIG_TRACING is enabled to avoid being used for other purposes. Signed-off-by: Seiji Aguchi <seiji.aguchi@hds.com> Link: http://lkml.kernel.org/r/51C323ED.5050708@hds.com Signed-off-by: H. Peter Anvin <hpa@linux.intel.com> Cc: Steven Rostedt <rostedt@goodmis.org>
2013-06-20 19:46:53 +04:00
apicinterrupt3 HYPERVISOR_CALLBACK_VECTOR \
hyperv_callback_vector hyperv_vector_handler
#endif /* CONFIG_HYPERV */
idtentry debug do_debug has_error_code=0 paranoid=1 shift_ist=DEBUG_STACK
idtentry int3 do_int3 has_error_code=0 paranoid=1 shift_ist=DEBUG_STACK
idtentry stack_segment do_stack_segment has_error_code=1
#ifdef CONFIG_XEN
idtentry xen_debug do_debug has_error_code=0
idtentry xen_int3 do_int3 has_error_code=0
idtentry xen_stack_segment do_stack_segment has_error_code=1
#endif
idtentry general_protection do_general_protection has_error_code=1
trace_idtentry page_fault do_page_fault has_error_code=1
#ifdef CONFIG_KVM_GUEST
idtentry async_page_fault do_async_page_fault has_error_code=1
#endif
#ifdef CONFIG_X86_MCE
idtentry machine_check has_error_code=0 paranoid=1 do_sym=*machine_check_vector(%rip)
#endif
/*
* Save all registers in pt_regs, and switch gs if needed.
* Use slow, but surefire "are we in kernel?" check.
* Return: ebx=0: need swapgs on exit, ebx=1: otherwise
*/
ENTRY(paranoid_entry)
cld
SAVE_C_REGS 8
SAVE_EXTRA_REGS 8
movl $1, %ebx
movl $MSR_GS_BASE, %ecx
rdmsr
testl %edx, %edx
js 1f /* negative -> in kernel */
SWAPGS
xorl %ebx, %ebx
1: ret
END(paranoid_entry)
/*
* "Paranoid" exit path from exception stack. This is invoked
* only on return from non-NMI IST interrupts that came
* from kernel space.
*
* We may be returning to very strange contexts (e.g. very early
* in syscall entry), so checking for preemption here would
* be complicated. Fortunately, we there's no good reason
* to try to handle preemption here.
*
* On entry, ebx is "no swapgs" flag (1: don't need swapgs, 0: need it)
*/
ENTRY(paranoid_exit)
DISABLE_INTERRUPTS(CLBR_NONE)
ftrace/x86: Do not change stacks in DEBUG when calling lockdep When both DYNAMIC_FTRACE and LOCKDEP are set, the TRACE_IRQS_ON/OFF will call into the lockdep code. The lockdep code can call lots of functions that may be traced by ftrace. When ftrace is updating its code and hits a breakpoint, the breakpoint handler will call into lockdep. If lockdep happens to call a function that also has a breakpoint attached, it will jump back into the breakpoint handler resetting the stack to the debug stack and corrupt the contents currently on that stack. The 'do_sym' call that calls do_int3() is protected by modifying the IST table to point to a different location if another breakpoint is hit. But the TRACE_IRQS_OFF/ON are outside that protection, and if a breakpoint is hit from those, the stack will get corrupted, and the kernel will crash: [ 1013.243754] BUG: unable to handle kernel NULL pointer dereference at 0000000000000002 [ 1013.272665] IP: [<ffff880145cc0000>] 0xffff880145cbffff [ 1013.285186] PGD 1401b2067 PUD 14324c067 PMD 0 [ 1013.298832] Oops: 0010 [#1] PREEMPT SMP [ 1013.310600] CPU 2 [ 1013.317904] Modules linked in: ip6t_REJECT nf_conntrack_ipv6 nf_defrag_ipv6 xt_state nf_conntrack ip6table_filter ip6_tables crc32c_intel ghash_clmulni_intel microcode usb_debug serio_raw pcspkr iTCO_wdt i2c_i801 iTCO_vendor_support e1000e nfsd nfs_acl auth_rpcgss lockd sunrpc i915 video i2c_algo_bit drm_kms_helper drm i2c_core [last unloaded: scsi_wait_scan] [ 1013.401848] [ 1013.407399] Pid: 112, comm: kworker/2:1 Not tainted 3.4.0+ #30 [ 1013.437943] RIP: 8eb8:[<ffff88014630a000>] [<ffff88014630a000>] 0xffff880146309fff [ 1013.459871] RSP: ffffffff8165e919:ffff88014780f408 EFLAGS: 00010046 [ 1013.477909] RAX: 0000000000000001 RBX: ffffffff81104020 RCX: 0000000000000000 [ 1013.499458] RDX: ffff880148008ea8 RSI: ffffffff8131ef40 RDI: ffffffff82203b20 [ 1013.521612] RBP: ffffffff81005751 R08: 0000000000000000 R09: 0000000000000000 [ 1013.543121] R10: ffffffff82cdc318 R11: 0000000000000000 R12: ffff880145cc0000 [ 1013.564614] R13: ffff880148008eb8 R14: 0000000000000002 R15: ffff88014780cb40 [ 1013.586108] FS: 0000000000000000(0000) GS:ffff880148000000(0000) knlGS:0000000000000000 [ 1013.609458] CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b [ 1013.627420] CR2: 0000000000000002 CR3: 0000000141f10000 CR4: 00000000001407e0 [ 1013.649051] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 [ 1013.670724] DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400 [ 1013.692376] Process kworker/2:1 (pid: 112, threadinfo ffff88013fe0e000, task ffff88014020a6a0) [ 1013.717028] Stack: [ 1013.724131] ffff88014780f570 ffff880145cc0000 0000400000004000 0000000000000000 [ 1013.745918] cccccccccccccccc ffff88014780cca8 ffffffff811072bb ffffffff81651627 [ 1013.767870] ffffffff8118f8a7 ffffffff811072bb ffffffff81f2b6c5 ffffffff81f11bdb [ 1013.790021] Call Trace: [ 1013.800701] Code: 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a <e7> d7 64 81 ff ff ff ff 01 00 00 00 00 00 00 00 65 d9 64 81 ff [ 1013.861443] RIP [<ffff88014630a000>] 0xffff880146309fff [ 1013.884466] RSP <ffff88014780f408> [ 1013.901507] CR2: 0000000000000002 The solution was to reuse the NMI functions that change the IDT table to make the debug stack keep its current stack (in kernel mode) when hitting a breakpoint: call debug_stack_set_zero TRACE_IRQS_ON call debug_stack_reset If the TRACE_IRQS_ON happens to hit a breakpoint then it will keep the current stack and not crash the box. Reported-by: Dave Jones <davej@redhat.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2012-05-30 19:54:53 +04:00
TRACE_IRQS_OFF_DEBUG
testl %ebx, %ebx /* swapgs needed? */
jnz paranoid_exit_no_swapgs
TRACE_IRQS_IRETQ
SWAPGS_UNSAFE_STACK
jmp paranoid_exit_restore
paranoid_exit_no_swapgs:
TRACE_IRQS_IRETQ_DEBUG
paranoid_exit_restore:
x86/asm/entry/64: Always allocate a complete "struct pt_regs" on the kernel stack The 64-bit entry code was using six stack slots less by not saving/restoring registers which are callee-preserved according to the C ABI, and was not allocating space for them. Only when syscalls needed a complete "struct pt_regs" was the complete area allocated and filled in. As an additional twist, on interrupt entry a "slightly less truncated pt_regs" trick is used, to make nested interrupt stacks easier to unwind. This proved to be a source of significant obfuscation and subtle bugs. For example, 'stub_fork' had to pop the return address, extend the struct, save registers, and push return address back. Ugly. 'ia32_ptregs_common' pops return address and "returns" via jmp insn, throwing a wrench into CPU return stack cache. This patch changes the code to always allocate a complete "struct pt_regs" on the kernel stack. The saving of registers is still done lazily. "Partial pt_regs" trick on interrupt stack is retained. Macros which manipulate "struct pt_regs" on stack are reworked: - ALLOC_PT_GPREGS_ON_STACK allocates the structure. - SAVE_C_REGS saves to it those registers which are clobbered by C code. - SAVE_EXTRA_REGS saves to it all other registers. - Corresponding RESTORE_* and REMOVE_PT_GPREGS_FROM_STACK macros reverse it. 'ia32_ptregs_common', 'stub_fork' and friends lost their ugly dance with the return pointer. LOAD_ARGS32 in ia32entry.S now uses symbolic stack offsets instead of magic numbers. 'error_entry' and 'save_paranoid' now use SAVE_C_REGS + SAVE_EXTRA_REGS instead of having it open-coded yet again. Patch was run-tested: 64-bit executables, 32-bit executables, strace works. Timing tests did not show measurable difference in 32-bit and 64-bit syscalls. Signed-off-by: Denys Vlasenko <dvlasenk@redhat.com> Signed-off-by: Andy Lutomirski <luto@amacapital.net> Cc: Alexei Starovoitov <ast@plumgrid.com> Cc: Borislav Petkov <bp@alien8.de> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: H. Peter Anvin <hpa@zytor.com> Cc: Kees Cook <keescook@chromium.org> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Will Drewry <wad@chromium.org> Link: http://lkml.kernel.org/r/1423778052-21038-2-git-send-email-dvlasenk@redhat.com Link: http://lkml.kernel.org/r/b89763d354aa23e670b9bdf3a40ae320320a7c2e.1424989793.git.luto@amacapital.net Signed-off-by: Ingo Molnar <mingo@kernel.org>
2015-02-27 01:40:27 +03:00
RESTORE_EXTRA_REGS
RESTORE_C_REGS
REMOVE_PT_GPREGS_FROM_STACK 8
INTERRUPT_RETURN
END(paranoid_exit)
/*
* Save all registers in pt_regs, and switch gs if needed.
* Return: EBX=0: came from user mode; EBX=1: otherwise
*/
ENTRY(error_entry)
cld
x86/asm/entry/64: Always allocate a complete "struct pt_regs" on the kernel stack The 64-bit entry code was using six stack slots less by not saving/restoring registers which are callee-preserved according to the C ABI, and was not allocating space for them. Only when syscalls needed a complete "struct pt_regs" was the complete area allocated and filled in. As an additional twist, on interrupt entry a "slightly less truncated pt_regs" trick is used, to make nested interrupt stacks easier to unwind. This proved to be a source of significant obfuscation and subtle bugs. For example, 'stub_fork' had to pop the return address, extend the struct, save registers, and push return address back. Ugly. 'ia32_ptregs_common' pops return address and "returns" via jmp insn, throwing a wrench into CPU return stack cache. This patch changes the code to always allocate a complete "struct pt_regs" on the kernel stack. The saving of registers is still done lazily. "Partial pt_regs" trick on interrupt stack is retained. Macros which manipulate "struct pt_regs" on stack are reworked: - ALLOC_PT_GPREGS_ON_STACK allocates the structure. - SAVE_C_REGS saves to it those registers which are clobbered by C code. - SAVE_EXTRA_REGS saves to it all other registers. - Corresponding RESTORE_* and REMOVE_PT_GPREGS_FROM_STACK macros reverse it. 'ia32_ptregs_common', 'stub_fork' and friends lost their ugly dance with the return pointer. LOAD_ARGS32 in ia32entry.S now uses symbolic stack offsets instead of magic numbers. 'error_entry' and 'save_paranoid' now use SAVE_C_REGS + SAVE_EXTRA_REGS instead of having it open-coded yet again. Patch was run-tested: 64-bit executables, 32-bit executables, strace works. Timing tests did not show measurable difference in 32-bit and 64-bit syscalls. Signed-off-by: Denys Vlasenko <dvlasenk@redhat.com> Signed-off-by: Andy Lutomirski <luto@amacapital.net> Cc: Alexei Starovoitov <ast@plumgrid.com> Cc: Borislav Petkov <bp@alien8.de> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: H. Peter Anvin <hpa@zytor.com> Cc: Kees Cook <keescook@chromium.org> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Will Drewry <wad@chromium.org> Link: http://lkml.kernel.org/r/1423778052-21038-2-git-send-email-dvlasenk@redhat.com Link: http://lkml.kernel.org/r/b89763d354aa23e670b9bdf3a40ae320320a7c2e.1424989793.git.luto@amacapital.net Signed-off-by: Ingo Molnar <mingo@kernel.org>
2015-02-27 01:40:27 +03:00
SAVE_C_REGS 8
SAVE_EXTRA_REGS 8
xorl %ebx, %ebx
testb $3, CS+8(%rsp)
jz .Lerror_kernelspace
.Lerror_entry_from_usermode_swapgs:
/*
* We entered from user mode or we're pretending to have entered
* from user mode due to an IRET fault.
*/
SWAPGS
.Lerror_entry_from_usermode_after_swapgs:
.Lerror_entry_done:
TRACE_IRQS_OFF
ret
/*
* There are two places in the kernel that can potentially fault with
* usergs. Handle them here. B stepping K8s sometimes report a
* truncated RIP for IRET exceptions returning to compat mode. Check
* for these here too.
*/
.Lerror_kernelspace:
incl %ebx
leaq native_irq_return_iret(%rip), %rcx
cmpq %rcx, RIP+8(%rsp)
je .Lerror_bad_iret
movl %ecx, %eax /* zero extend */
cmpq %rax, RIP+8(%rsp)
je .Lbstep_iret
cmpq $gs_change, RIP+8(%rsp)
jne .Lerror_entry_done
/*
* hack: gs_change can fail with user gsbase. If this happens, fix up
* gsbase and proceed. We'll fix up the exception and land in
* gs_change's error handler with kernel gsbase.
*/
jmp .Lerror_entry_from_usermode_swapgs
.Lbstep_iret:
/* Fix truncated RIP */
movq %rcx, RIP+8(%rsp)
/* fall through */
.Lerror_bad_iret:
/*
* We came from an IRET to user mode, so we have user gsbase.
* Switch to kernel gsbase:
*/
SWAPGS
/*
* Pretend that the exception came from user mode: set up pt_regs
* as if we faulted immediately after IRET and clear EBX so that
* error_exit knows that we will be returning to user mode.
*/
mov %rsp, %rdi
call fixup_bad_iret
mov %rax, %rsp
decl %ebx
jmp .Lerror_entry_from_usermode_after_swapgs
END(error_entry)
/*
* On entry, EBS is a "return to kernel mode" flag:
* 1: already in kernel mode, don't need SWAPGS
* 0: user gsbase is loaded, we need SWAPGS and standard preparation for return to usermode
*/
ENTRY(error_exit)
movl %ebx, %eax
DISABLE_INTERRUPTS(CLBR_NONE)
TRACE_IRQS_OFF
testl %eax, %eax
jnz retint_kernel
jmp retint_user
END(error_exit)
/* Runs on exception stack */
ENTRY(nmi)
PARAVIRT_ADJUST_EXCEPTION_FRAME
x86: Add workaround to NMI iret woes In x86, when an NMI goes off, the CPU goes into an NMI context that prevents other NMIs to trigger on that CPU. If an NMI is suppose to trigger, it has to wait till the previous NMI leaves NMI context. At that time, the next NMI can trigger (note, only one more NMI will trigger, as only one can be latched at a time). The way x86 gets out of NMI context is by calling iret. The problem with this is that this causes problems if the NMI handle either triggers an exception, or a breakpoint. Both the exception and the breakpoint handlers will finish with an iret. If this happens while in NMI context, the CPU will leave NMI context and a new NMI may come in. As NMI handlers are not made to be re-entrant, this can cause havoc with the system, not to mention, the nested NMI will write all over the previous NMI's stack. Linus Torvalds proposed the following workaround to this problem: https://lkml.org/lkml/2010/7/14/264 "In fact, I wonder if we couldn't just do a software NMI disable instead? Hav ea per-cpu variable (in the _core_ percpu areas that get allocated statically) that points to the NMI stack frame, and just make the NMI code itself do something like NMI entry: - load percpu NMI stack frame pointer - if non-zero we know we're nested, and should ignore this NMI: - we're returning to kernel mode, so return immediately by using "popf/ret", which also keeps NMI's disabled in the hardware until the "real" NMI iret happens. - before the popf/iret, use the NMI stack pointer to make the NMI return stack be invalid and cause a fault - set the NMI stack pointer to the current stack pointer NMI exit (not the above "immediate exit because we nested"): clear the percpu NMI stack pointer Just do the iret. Now, the thing is, now the "iret" is atomic. If we had a nested NMI, we'll take a fault, and that re-does our "delayed" NMI - and NMI's will stay masked. And if we didn't have a nested NMI, that iret will now unmask NMI's, and everything is happy." I first tried to follow this advice but as I started implementing this code, a few gotchas showed up. One, is accessing per-cpu variables in the NMI handler. The problem is that per-cpu variables use the %gs register to get the variable for the given CPU. But as the NMI may happen in userspace, we must first perform a SWAPGS to get to it. The NMI handler already does this later in the code, but its too late as we have saved off all the registers and we don't want to do that for a disabled NMI. Peter Zijlstra suggested to keep all variables on the stack. This simplifies things greatly and it has the added benefit of cache locality. Two, faulting on the iret. I really wanted to make this work, but it was becoming very hacky, and I never got it to be stable. The iret already had a fault handler for userspace faulting with bad segment registers, and getting NMI to trigger a fault and detect it was very tricky. But for strange reasons, the system would usually take a double fault and crash. I never figured out why and decided to go with a simple "jmp" approach. The new approach I took also simplified things. Finally, the last problem with Linus's approach was to have the nested NMI handler do a ret instead of an iret to give the first NMI NMI-context again. The problem is that ret is much more limited than an iret. I couldn't figure out how to get the stack back where it belonged. I could have copied the current stack, pushed the return onto it, but my fear here is that there may be some place that writes data below the stack pointer. I know that is not something code should depend on, but I don't want to chance it. I may add this feature later, but for now, an NMI handler that loses NMI context will not get it back. Here's what is done: When an NMI comes in, the HW pushes the interrupt stack frame onto the per cpu NMI stack that is selected by the IST. A special location on the NMI stack holds a variable that is set when the first NMI handler runs. If this variable is set then we know that this is a nested NMI and we process the nested NMI code. There is still a race when this variable is cleared and an NMI comes in just before the first NMI does the return. For this case, if the variable is cleared, we also check if the interrupted stack is the NMI stack. If it is, then we process the nested NMI code. Why the two tests and not just test the interrupted stack? If the first NMI hits a breakpoint and loses NMI context, and then it hits another breakpoint and while processing that breakpoint we get a nested NMI. When processing a breakpoint, the stack changes to the breakpoint stack. If another NMI comes in here we can't rely on the interrupted stack to be the NMI stack. If the variable is not set and the interrupted task's stack is not the NMI stack, then we know this is the first NMI and we can process things normally. But in order to do so, we need to do a few things first. 1) Set the stack variable that tells us that we are in an NMI handler 2) Make two copies of the interrupt stack frame. One copy is used to return on iret The other is used to restore the first one if we have a nested NMI. This is what the stack will look like: +-------------------------+ | original SS | | original Return RSP | | original RFLAGS | | original CS | | original RIP | +-------------------------+ | temp storage for rdx | +-------------------------+ | NMI executing variable | +-------------------------+ | Saved SS | | Saved Return RSP | | Saved RFLAGS | | Saved CS | | Saved RIP | +-------------------------+ | copied SS | | copied Return RSP | | copied RFLAGS | | copied CS | | copied RIP | +-------------------------+ | pt_regs | +-------------------------+ The original stack frame contains what the HW put in when we entered the NMI. We store %rdx as a temp variable to use. Both the original HW stack frame and this %rdx storage will be clobbered by nested NMIs so we can not rely on them later in the first NMI handler. The next item is the special stack variable that is set when we execute the rest of the NMI handler. Then we have two copies of the interrupt stack. The second copy is modified by any nested NMIs to let the first NMI know that we triggered a second NMI (latched) and that we should repeat the NMI handler. If the first NMI hits an exception or breakpoint that takes it out of NMI context, if a second NMI comes in before the first one finishes, it will update the copied interrupt stack to point to a fix up location to trigger another NMI. When the first NMI calls iret, it will instead jump to the fix up location. This fix up location will copy the saved interrupt stack back to the copy and execute the nmi handler again. Note, the nested NMI knows enough to check if it preempted a previous NMI handler while it is in the fixup location. If it has, it will not modify the copied interrupt stack and will just leave as if nothing happened. As the NMI handle is about to execute again, there's no reason to latch now. To test all this, I forced the NMI handler to call iret and take itself out of NMI context. I also added assemble code to write to the serial to make sure that it hits the nested path as well as the fix up path. Everything seems to be working fine. Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: H. Peter Anvin <hpa@linux.intel.com> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Paul Turner <pjt@google.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Mathieu Desnoyers <mathieu.desnoyers@efficios.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2011-12-08 21:36:23 +04:00
/*
* We allow breakpoints in NMIs. If a breakpoint occurs, then
* the iretq it performs will take us out of NMI context.
* This means that we can have nested NMIs where the next
* NMI is using the top of the stack of the previous NMI. We
* can't let it execute because the nested NMI will corrupt the
* stack of the previous NMI. NMI handlers are not re-entrant
* anyway.
*
* To handle this case we do the following:
* Check the a special location on the stack that contains
* a variable that is set when NMIs are executing.
* The interrupted task's stack is also checked to see if it
* is an NMI stack.
* If the variable is not set and the stack is not the NMI
* stack then:
* o Set the special variable on the stack
* o Copy the interrupt frame into a "saved" location on the stack
* o Copy the interrupt frame into a "copy" location on the stack
* o Continue processing the NMI
* If the variable is set or the previous stack is the NMI stack:
* o Modify the "copy" location to jump to the repeate_nmi
* o return back to the first NMI
*
* Now on exit of the first NMI, we first clear the stack variable
* The NMI stack will tell any nested NMIs at that point that it is
* nested. Then we pop the stack normally with iret, and if there was
* a nested NMI that updated the copy interrupt stack frame, a
* jump will be made to the repeat_nmi code that will handle the second
* NMI.
*/
/* Use %rdx as our temp variable throughout */
pushq %rdx
x86: Add workaround to NMI iret woes In x86, when an NMI goes off, the CPU goes into an NMI context that prevents other NMIs to trigger on that CPU. If an NMI is suppose to trigger, it has to wait till the previous NMI leaves NMI context. At that time, the next NMI can trigger (note, only one more NMI will trigger, as only one can be latched at a time). The way x86 gets out of NMI context is by calling iret. The problem with this is that this causes problems if the NMI handle either triggers an exception, or a breakpoint. Both the exception and the breakpoint handlers will finish with an iret. If this happens while in NMI context, the CPU will leave NMI context and a new NMI may come in. As NMI handlers are not made to be re-entrant, this can cause havoc with the system, not to mention, the nested NMI will write all over the previous NMI's stack. Linus Torvalds proposed the following workaround to this problem: https://lkml.org/lkml/2010/7/14/264 "In fact, I wonder if we couldn't just do a software NMI disable instead? Hav ea per-cpu variable (in the _core_ percpu areas that get allocated statically) that points to the NMI stack frame, and just make the NMI code itself do something like NMI entry: - load percpu NMI stack frame pointer - if non-zero we know we're nested, and should ignore this NMI: - we're returning to kernel mode, so return immediately by using "popf/ret", which also keeps NMI's disabled in the hardware until the "real" NMI iret happens. - before the popf/iret, use the NMI stack pointer to make the NMI return stack be invalid and cause a fault - set the NMI stack pointer to the current stack pointer NMI exit (not the above "immediate exit because we nested"): clear the percpu NMI stack pointer Just do the iret. Now, the thing is, now the "iret" is atomic. If we had a nested NMI, we'll take a fault, and that re-does our "delayed" NMI - and NMI's will stay masked. And if we didn't have a nested NMI, that iret will now unmask NMI's, and everything is happy." I first tried to follow this advice but as I started implementing this code, a few gotchas showed up. One, is accessing per-cpu variables in the NMI handler. The problem is that per-cpu variables use the %gs register to get the variable for the given CPU. But as the NMI may happen in userspace, we must first perform a SWAPGS to get to it. The NMI handler already does this later in the code, but its too late as we have saved off all the registers and we don't want to do that for a disabled NMI. Peter Zijlstra suggested to keep all variables on the stack. This simplifies things greatly and it has the added benefit of cache locality. Two, faulting on the iret. I really wanted to make this work, but it was becoming very hacky, and I never got it to be stable. The iret already had a fault handler for userspace faulting with bad segment registers, and getting NMI to trigger a fault and detect it was very tricky. But for strange reasons, the system would usually take a double fault and crash. I never figured out why and decided to go with a simple "jmp" approach. The new approach I took also simplified things. Finally, the last problem with Linus's approach was to have the nested NMI handler do a ret instead of an iret to give the first NMI NMI-context again. The problem is that ret is much more limited than an iret. I couldn't figure out how to get the stack back where it belonged. I could have copied the current stack, pushed the return onto it, but my fear here is that there may be some place that writes data below the stack pointer. I know that is not something code should depend on, but I don't want to chance it. I may add this feature later, but for now, an NMI handler that loses NMI context will not get it back. Here's what is done: When an NMI comes in, the HW pushes the interrupt stack frame onto the per cpu NMI stack that is selected by the IST. A special location on the NMI stack holds a variable that is set when the first NMI handler runs. If this variable is set then we know that this is a nested NMI and we process the nested NMI code. There is still a race when this variable is cleared and an NMI comes in just before the first NMI does the return. For this case, if the variable is cleared, we also check if the interrupted stack is the NMI stack. If it is, then we process the nested NMI code. Why the two tests and not just test the interrupted stack? If the first NMI hits a breakpoint and loses NMI context, and then it hits another breakpoint and while processing that breakpoint we get a nested NMI. When processing a breakpoint, the stack changes to the breakpoint stack. If another NMI comes in here we can't rely on the interrupted stack to be the NMI stack. If the variable is not set and the interrupted task's stack is not the NMI stack, then we know this is the first NMI and we can process things normally. But in order to do so, we need to do a few things first. 1) Set the stack variable that tells us that we are in an NMI handler 2) Make two copies of the interrupt stack frame. One copy is used to return on iret The other is used to restore the first one if we have a nested NMI. This is what the stack will look like: +-------------------------+ | original SS | | original Return RSP | | original RFLAGS | | original CS | | original RIP | +-------------------------+ | temp storage for rdx | +-------------------------+ | NMI executing variable | +-------------------------+ | Saved SS | | Saved Return RSP | | Saved RFLAGS | | Saved CS | | Saved RIP | +-------------------------+ | copied SS | | copied Return RSP | | copied RFLAGS | | copied CS | | copied RIP | +-------------------------+ | pt_regs | +-------------------------+ The original stack frame contains what the HW put in when we entered the NMI. We store %rdx as a temp variable to use. Both the original HW stack frame and this %rdx storage will be clobbered by nested NMIs so we can not rely on them later in the first NMI handler. The next item is the special stack variable that is set when we execute the rest of the NMI handler. Then we have two copies of the interrupt stack. The second copy is modified by any nested NMIs to let the first NMI know that we triggered a second NMI (latched) and that we should repeat the NMI handler. If the first NMI hits an exception or breakpoint that takes it out of NMI context, if a second NMI comes in before the first one finishes, it will update the copied interrupt stack to point to a fix up location to trigger another NMI. When the first NMI calls iret, it will instead jump to the fix up location. This fix up location will copy the saved interrupt stack back to the copy and execute the nmi handler again. Note, the nested NMI knows enough to check if it preempted a previous NMI handler while it is in the fixup location. If it has, it will not modify the copied interrupt stack and will just leave as if nothing happened. As the NMI handle is about to execute again, there's no reason to latch now. To test all this, I forced the NMI handler to call iret and take itself out of NMI context. I also added assemble code to write to the serial to make sure that it hits the nested path as well as the fix up path. Everything seems to be working fine. Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: H. Peter Anvin <hpa@linux.intel.com> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Paul Turner <pjt@google.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Mathieu Desnoyers <mathieu.desnoyers@efficios.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2011-12-08 21:36:23 +04:00
/*
* If %cs was not the kernel segment, then the NMI triggered in user
* space, which means it is definitely not nested.
*/
cmpl $__KERNEL_CS, 16(%rsp)
jne first_nmi
x86: Add workaround to NMI iret woes In x86, when an NMI goes off, the CPU goes into an NMI context that prevents other NMIs to trigger on that CPU. If an NMI is suppose to trigger, it has to wait till the previous NMI leaves NMI context. At that time, the next NMI can trigger (note, only one more NMI will trigger, as only one can be latched at a time). The way x86 gets out of NMI context is by calling iret. The problem with this is that this causes problems if the NMI handle either triggers an exception, or a breakpoint. Both the exception and the breakpoint handlers will finish with an iret. If this happens while in NMI context, the CPU will leave NMI context and a new NMI may come in. As NMI handlers are not made to be re-entrant, this can cause havoc with the system, not to mention, the nested NMI will write all over the previous NMI's stack. Linus Torvalds proposed the following workaround to this problem: https://lkml.org/lkml/2010/7/14/264 "In fact, I wonder if we couldn't just do a software NMI disable instead? Hav ea per-cpu variable (in the _core_ percpu areas that get allocated statically) that points to the NMI stack frame, and just make the NMI code itself do something like NMI entry: - load percpu NMI stack frame pointer - if non-zero we know we're nested, and should ignore this NMI: - we're returning to kernel mode, so return immediately by using "popf/ret", which also keeps NMI's disabled in the hardware until the "real" NMI iret happens. - before the popf/iret, use the NMI stack pointer to make the NMI return stack be invalid and cause a fault - set the NMI stack pointer to the current stack pointer NMI exit (not the above "immediate exit because we nested"): clear the percpu NMI stack pointer Just do the iret. Now, the thing is, now the "iret" is atomic. If we had a nested NMI, we'll take a fault, and that re-does our "delayed" NMI - and NMI's will stay masked. And if we didn't have a nested NMI, that iret will now unmask NMI's, and everything is happy." I first tried to follow this advice but as I started implementing this code, a few gotchas showed up. One, is accessing per-cpu variables in the NMI handler. The problem is that per-cpu variables use the %gs register to get the variable for the given CPU. But as the NMI may happen in userspace, we must first perform a SWAPGS to get to it. The NMI handler already does this later in the code, but its too late as we have saved off all the registers and we don't want to do that for a disabled NMI. Peter Zijlstra suggested to keep all variables on the stack. This simplifies things greatly and it has the added benefit of cache locality. Two, faulting on the iret. I really wanted to make this work, but it was becoming very hacky, and I never got it to be stable. The iret already had a fault handler for userspace faulting with bad segment registers, and getting NMI to trigger a fault and detect it was very tricky. But for strange reasons, the system would usually take a double fault and crash. I never figured out why and decided to go with a simple "jmp" approach. The new approach I took also simplified things. Finally, the last problem with Linus's approach was to have the nested NMI handler do a ret instead of an iret to give the first NMI NMI-context again. The problem is that ret is much more limited than an iret. I couldn't figure out how to get the stack back where it belonged. I could have copied the current stack, pushed the return onto it, but my fear here is that there may be some place that writes data below the stack pointer. I know that is not something code should depend on, but I don't want to chance it. I may add this feature later, but for now, an NMI handler that loses NMI context will not get it back. Here's what is done: When an NMI comes in, the HW pushes the interrupt stack frame onto the per cpu NMI stack that is selected by the IST. A special location on the NMI stack holds a variable that is set when the first NMI handler runs. If this variable is set then we know that this is a nested NMI and we process the nested NMI code. There is still a race when this variable is cleared and an NMI comes in just before the first NMI does the return. For this case, if the variable is cleared, we also check if the interrupted stack is the NMI stack. If it is, then we process the nested NMI code. Why the two tests and not just test the interrupted stack? If the first NMI hits a breakpoint and loses NMI context, and then it hits another breakpoint and while processing that breakpoint we get a nested NMI. When processing a breakpoint, the stack changes to the breakpoint stack. If another NMI comes in here we can't rely on the interrupted stack to be the NMI stack. If the variable is not set and the interrupted task's stack is not the NMI stack, then we know this is the first NMI and we can process things normally. But in order to do so, we need to do a few things first. 1) Set the stack variable that tells us that we are in an NMI handler 2) Make two copies of the interrupt stack frame. One copy is used to return on iret The other is used to restore the first one if we have a nested NMI. This is what the stack will look like: +-------------------------+ | original SS | | original Return RSP | | original RFLAGS | | original CS | | original RIP | +-------------------------+ | temp storage for rdx | +-------------------------+ | NMI executing variable | +-------------------------+ | Saved SS | | Saved Return RSP | | Saved RFLAGS | | Saved CS | | Saved RIP | +-------------------------+ | copied SS | | copied Return RSP | | copied RFLAGS | | copied CS | | copied RIP | +-------------------------+ | pt_regs | +-------------------------+ The original stack frame contains what the HW put in when we entered the NMI. We store %rdx as a temp variable to use. Both the original HW stack frame and this %rdx storage will be clobbered by nested NMIs so we can not rely on them later in the first NMI handler. The next item is the special stack variable that is set when we execute the rest of the NMI handler. Then we have two copies of the interrupt stack. The second copy is modified by any nested NMIs to let the first NMI know that we triggered a second NMI (latched) and that we should repeat the NMI handler. If the first NMI hits an exception or breakpoint that takes it out of NMI context, if a second NMI comes in before the first one finishes, it will update the copied interrupt stack to point to a fix up location to trigger another NMI. When the first NMI calls iret, it will instead jump to the fix up location. This fix up location will copy the saved interrupt stack back to the copy and execute the nmi handler again. Note, the nested NMI knows enough to check if it preempted a previous NMI handler while it is in the fixup location. If it has, it will not modify the copied interrupt stack and will just leave as if nothing happened. As the NMI handle is about to execute again, there's no reason to latch now. To test all this, I forced the NMI handler to call iret and take itself out of NMI context. I also added assemble code to write to the serial to make sure that it hits the nested path as well as the fix up path. Everything seems to be working fine. Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: H. Peter Anvin <hpa@linux.intel.com> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Paul Turner <pjt@google.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Mathieu Desnoyers <mathieu.desnoyers@efficios.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2011-12-08 21:36:23 +04:00
/*
* Check the special variable on the stack to see if NMIs are
* executing.
*/
cmpl $1, -8(%rsp)
je nested_nmi
x86: Add workaround to NMI iret woes In x86, when an NMI goes off, the CPU goes into an NMI context that prevents other NMIs to trigger on that CPU. If an NMI is suppose to trigger, it has to wait till the previous NMI leaves NMI context. At that time, the next NMI can trigger (note, only one more NMI will trigger, as only one can be latched at a time). The way x86 gets out of NMI context is by calling iret. The problem with this is that this causes problems if the NMI handle either triggers an exception, or a breakpoint. Both the exception and the breakpoint handlers will finish with an iret. If this happens while in NMI context, the CPU will leave NMI context and a new NMI may come in. As NMI handlers are not made to be re-entrant, this can cause havoc with the system, not to mention, the nested NMI will write all over the previous NMI's stack. Linus Torvalds proposed the following workaround to this problem: https://lkml.org/lkml/2010/7/14/264 "In fact, I wonder if we couldn't just do a software NMI disable instead? Hav ea per-cpu variable (in the _core_ percpu areas that get allocated statically) that points to the NMI stack frame, and just make the NMI code itself do something like NMI entry: - load percpu NMI stack frame pointer - if non-zero we know we're nested, and should ignore this NMI: - we're returning to kernel mode, so return immediately by using "popf/ret", which also keeps NMI's disabled in the hardware until the "real" NMI iret happens. - before the popf/iret, use the NMI stack pointer to make the NMI return stack be invalid and cause a fault - set the NMI stack pointer to the current stack pointer NMI exit (not the above "immediate exit because we nested"): clear the percpu NMI stack pointer Just do the iret. Now, the thing is, now the "iret" is atomic. If we had a nested NMI, we'll take a fault, and that re-does our "delayed" NMI - and NMI's will stay masked. And if we didn't have a nested NMI, that iret will now unmask NMI's, and everything is happy." I first tried to follow this advice but as I started implementing this code, a few gotchas showed up. One, is accessing per-cpu variables in the NMI handler. The problem is that per-cpu variables use the %gs register to get the variable for the given CPU. But as the NMI may happen in userspace, we must first perform a SWAPGS to get to it. The NMI handler already does this later in the code, but its too late as we have saved off all the registers and we don't want to do that for a disabled NMI. Peter Zijlstra suggested to keep all variables on the stack. This simplifies things greatly and it has the added benefit of cache locality. Two, faulting on the iret. I really wanted to make this work, but it was becoming very hacky, and I never got it to be stable. The iret already had a fault handler for userspace faulting with bad segment registers, and getting NMI to trigger a fault and detect it was very tricky. But for strange reasons, the system would usually take a double fault and crash. I never figured out why and decided to go with a simple "jmp" approach. The new approach I took also simplified things. Finally, the last problem with Linus's approach was to have the nested NMI handler do a ret instead of an iret to give the first NMI NMI-context again. The problem is that ret is much more limited than an iret. I couldn't figure out how to get the stack back where it belonged. I could have copied the current stack, pushed the return onto it, but my fear here is that there may be some place that writes data below the stack pointer. I know that is not something code should depend on, but I don't want to chance it. I may add this feature later, but for now, an NMI handler that loses NMI context will not get it back. Here's what is done: When an NMI comes in, the HW pushes the interrupt stack frame onto the per cpu NMI stack that is selected by the IST. A special location on the NMI stack holds a variable that is set when the first NMI handler runs. If this variable is set then we know that this is a nested NMI and we process the nested NMI code. There is still a race when this variable is cleared and an NMI comes in just before the first NMI does the return. For this case, if the variable is cleared, we also check if the interrupted stack is the NMI stack. If it is, then we process the nested NMI code. Why the two tests and not just test the interrupted stack? If the first NMI hits a breakpoint and loses NMI context, and then it hits another breakpoint and while processing that breakpoint we get a nested NMI. When processing a breakpoint, the stack changes to the breakpoint stack. If another NMI comes in here we can't rely on the interrupted stack to be the NMI stack. If the variable is not set and the interrupted task's stack is not the NMI stack, then we know this is the first NMI and we can process things normally. But in order to do so, we need to do a few things first. 1) Set the stack variable that tells us that we are in an NMI handler 2) Make two copies of the interrupt stack frame. One copy is used to return on iret The other is used to restore the first one if we have a nested NMI. This is what the stack will look like: +-------------------------+ | original SS | | original Return RSP | | original RFLAGS | | original CS | | original RIP | +-------------------------+ | temp storage for rdx | +-------------------------+ | NMI executing variable | +-------------------------+ | Saved SS | | Saved Return RSP | | Saved RFLAGS | | Saved CS | | Saved RIP | +-------------------------+ | copied SS | | copied Return RSP | | copied RFLAGS | | copied CS | | copied RIP | +-------------------------+ | pt_regs | +-------------------------+ The original stack frame contains what the HW put in when we entered the NMI. We store %rdx as a temp variable to use. Both the original HW stack frame and this %rdx storage will be clobbered by nested NMIs so we can not rely on them later in the first NMI handler. The next item is the special stack variable that is set when we execute the rest of the NMI handler. Then we have two copies of the interrupt stack. The second copy is modified by any nested NMIs to let the first NMI know that we triggered a second NMI (latched) and that we should repeat the NMI handler. If the first NMI hits an exception or breakpoint that takes it out of NMI context, if a second NMI comes in before the first one finishes, it will update the copied interrupt stack to point to a fix up location to trigger another NMI. When the first NMI calls iret, it will instead jump to the fix up location. This fix up location will copy the saved interrupt stack back to the copy and execute the nmi handler again. Note, the nested NMI knows enough to check if it preempted a previous NMI handler while it is in the fixup location. If it has, it will not modify the copied interrupt stack and will just leave as if nothing happened. As the NMI handle is about to execute again, there's no reason to latch now. To test all this, I forced the NMI handler to call iret and take itself out of NMI context. I also added assemble code to write to the serial to make sure that it hits the nested path as well as the fix up path. Everything seems to be working fine. Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: H. Peter Anvin <hpa@linux.intel.com> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Paul Turner <pjt@google.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Mathieu Desnoyers <mathieu.desnoyers@efficios.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2011-12-08 21:36:23 +04:00
/*
* Now test if the previous stack was an NMI stack.
* We need the double check. We check the NMI stack to satisfy the
* race when the first NMI clears the variable before returning.
* We check the variable because the first NMI could be in a
* breakpoint routine using a breakpoint stack.
*/
lea 6*8(%rsp), %rdx
/* Compare the NMI stack (rdx) with the stack we came from (4*8(%rsp)) */
cmpq %rdx, 4*8(%rsp)
/* If the stack pointer is above the NMI stack, this is a normal NMI */
ja first_nmi
subq $EXCEPTION_STKSZ, %rdx
cmpq %rdx, 4*8(%rsp)
/* If it is below the NMI stack, it is a normal NMI */
jb first_nmi
/* Ah, it is within the NMI stack, treat it as nested */
x86: Add workaround to NMI iret woes In x86, when an NMI goes off, the CPU goes into an NMI context that prevents other NMIs to trigger on that CPU. If an NMI is suppose to trigger, it has to wait till the previous NMI leaves NMI context. At that time, the next NMI can trigger (note, only one more NMI will trigger, as only one can be latched at a time). The way x86 gets out of NMI context is by calling iret. The problem with this is that this causes problems if the NMI handle either triggers an exception, or a breakpoint. Both the exception and the breakpoint handlers will finish with an iret. If this happens while in NMI context, the CPU will leave NMI context and a new NMI may come in. As NMI handlers are not made to be re-entrant, this can cause havoc with the system, not to mention, the nested NMI will write all over the previous NMI's stack. Linus Torvalds proposed the following workaround to this problem: https://lkml.org/lkml/2010/7/14/264 "In fact, I wonder if we couldn't just do a software NMI disable instead? Hav ea per-cpu variable (in the _core_ percpu areas that get allocated statically) that points to the NMI stack frame, and just make the NMI code itself do something like NMI entry: - load percpu NMI stack frame pointer - if non-zero we know we're nested, and should ignore this NMI: - we're returning to kernel mode, so return immediately by using "popf/ret", which also keeps NMI's disabled in the hardware until the "real" NMI iret happens. - before the popf/iret, use the NMI stack pointer to make the NMI return stack be invalid and cause a fault - set the NMI stack pointer to the current stack pointer NMI exit (not the above "immediate exit because we nested"): clear the percpu NMI stack pointer Just do the iret. Now, the thing is, now the "iret" is atomic. If we had a nested NMI, we'll take a fault, and that re-does our "delayed" NMI - and NMI's will stay masked. And if we didn't have a nested NMI, that iret will now unmask NMI's, and everything is happy." I first tried to follow this advice but as I started implementing this code, a few gotchas showed up. One, is accessing per-cpu variables in the NMI handler. The problem is that per-cpu variables use the %gs register to get the variable for the given CPU. But as the NMI may happen in userspace, we must first perform a SWAPGS to get to it. The NMI handler already does this later in the code, but its too late as we have saved off all the registers and we don't want to do that for a disabled NMI. Peter Zijlstra suggested to keep all variables on the stack. This simplifies things greatly and it has the added benefit of cache locality. Two, faulting on the iret. I really wanted to make this work, but it was becoming very hacky, and I never got it to be stable. The iret already had a fault handler for userspace faulting with bad segment registers, and getting NMI to trigger a fault and detect it was very tricky. But for strange reasons, the system would usually take a double fault and crash. I never figured out why and decided to go with a simple "jmp" approach. The new approach I took also simplified things. Finally, the last problem with Linus's approach was to have the nested NMI handler do a ret instead of an iret to give the first NMI NMI-context again. The problem is that ret is much more limited than an iret. I couldn't figure out how to get the stack back where it belonged. I could have copied the current stack, pushed the return onto it, but my fear here is that there may be some place that writes data below the stack pointer. I know that is not something code should depend on, but I don't want to chance it. I may add this feature later, but for now, an NMI handler that loses NMI context will not get it back. Here's what is done: When an NMI comes in, the HW pushes the interrupt stack frame onto the per cpu NMI stack that is selected by the IST. A special location on the NMI stack holds a variable that is set when the first NMI handler runs. If this variable is set then we know that this is a nested NMI and we process the nested NMI code. There is still a race when this variable is cleared and an NMI comes in just before the first NMI does the return. For this case, if the variable is cleared, we also check if the interrupted stack is the NMI stack. If it is, then we process the nested NMI code. Why the two tests and not just test the interrupted stack? If the first NMI hits a breakpoint and loses NMI context, and then it hits another breakpoint and while processing that breakpoint we get a nested NMI. When processing a breakpoint, the stack changes to the breakpoint stack. If another NMI comes in here we can't rely on the interrupted stack to be the NMI stack. If the variable is not set and the interrupted task's stack is not the NMI stack, then we know this is the first NMI and we can process things normally. But in order to do so, we need to do a few things first. 1) Set the stack variable that tells us that we are in an NMI handler 2) Make two copies of the interrupt stack frame. One copy is used to return on iret The other is used to restore the first one if we have a nested NMI. This is what the stack will look like: +-------------------------+ | original SS | | original Return RSP | | original RFLAGS | | original CS | | original RIP | +-------------------------+ | temp storage for rdx | +-------------------------+ | NMI executing variable | +-------------------------+ | Saved SS | | Saved Return RSP | | Saved RFLAGS | | Saved CS | | Saved RIP | +-------------------------+ | copied SS | | copied Return RSP | | copied RFLAGS | | copied CS | | copied RIP | +-------------------------+ | pt_regs | +-------------------------+ The original stack frame contains what the HW put in when we entered the NMI. We store %rdx as a temp variable to use. Both the original HW stack frame and this %rdx storage will be clobbered by nested NMIs so we can not rely on them later in the first NMI handler. The next item is the special stack variable that is set when we execute the rest of the NMI handler. Then we have two copies of the interrupt stack. The second copy is modified by any nested NMIs to let the first NMI know that we triggered a second NMI (latched) and that we should repeat the NMI handler. If the first NMI hits an exception or breakpoint that takes it out of NMI context, if a second NMI comes in before the first one finishes, it will update the copied interrupt stack to point to a fix up location to trigger another NMI. When the first NMI calls iret, it will instead jump to the fix up location. This fix up location will copy the saved interrupt stack back to the copy and execute the nmi handler again. Note, the nested NMI knows enough to check if it preempted a previous NMI handler while it is in the fixup location. If it has, it will not modify the copied interrupt stack and will just leave as if nothing happened. As the NMI handle is about to execute again, there's no reason to latch now. To test all this, I forced the NMI handler to call iret and take itself out of NMI context. I also added assemble code to write to the serial to make sure that it hits the nested path as well as the fix up path. Everything seems to be working fine. Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: H. Peter Anvin <hpa@linux.intel.com> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Paul Turner <pjt@google.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Mathieu Desnoyers <mathieu.desnoyers@efficios.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2011-12-08 21:36:23 +04:00
nested_nmi:
/*
* Do nothing if we interrupted the fixup in repeat_nmi.
* It's about to repeat the NMI handler, so we are fine
* with ignoring this one.
*/
movq $repeat_nmi, %rdx
cmpq 8(%rsp), %rdx
ja 1f
movq $end_repeat_nmi, %rdx
cmpq 8(%rsp), %rdx
ja nested_nmi_out
x86: Add workaround to NMI iret woes In x86, when an NMI goes off, the CPU goes into an NMI context that prevents other NMIs to trigger on that CPU. If an NMI is suppose to trigger, it has to wait till the previous NMI leaves NMI context. At that time, the next NMI can trigger (note, only one more NMI will trigger, as only one can be latched at a time). The way x86 gets out of NMI context is by calling iret. The problem with this is that this causes problems if the NMI handle either triggers an exception, or a breakpoint. Both the exception and the breakpoint handlers will finish with an iret. If this happens while in NMI context, the CPU will leave NMI context and a new NMI may come in. As NMI handlers are not made to be re-entrant, this can cause havoc with the system, not to mention, the nested NMI will write all over the previous NMI's stack. Linus Torvalds proposed the following workaround to this problem: https://lkml.org/lkml/2010/7/14/264 "In fact, I wonder if we couldn't just do a software NMI disable instead? Hav ea per-cpu variable (in the _core_ percpu areas that get allocated statically) that points to the NMI stack frame, and just make the NMI code itself do something like NMI entry: - load percpu NMI stack frame pointer - if non-zero we know we're nested, and should ignore this NMI: - we're returning to kernel mode, so return immediately by using "popf/ret", which also keeps NMI's disabled in the hardware until the "real" NMI iret happens. - before the popf/iret, use the NMI stack pointer to make the NMI return stack be invalid and cause a fault - set the NMI stack pointer to the current stack pointer NMI exit (not the above "immediate exit because we nested"): clear the percpu NMI stack pointer Just do the iret. Now, the thing is, now the "iret" is atomic. If we had a nested NMI, we'll take a fault, and that re-does our "delayed" NMI - and NMI's will stay masked. And if we didn't have a nested NMI, that iret will now unmask NMI's, and everything is happy." I first tried to follow this advice but as I started implementing this code, a few gotchas showed up. One, is accessing per-cpu variables in the NMI handler. The problem is that per-cpu variables use the %gs register to get the variable for the given CPU. But as the NMI may happen in userspace, we must first perform a SWAPGS to get to it. The NMI handler already does this later in the code, but its too late as we have saved off all the registers and we don't want to do that for a disabled NMI. Peter Zijlstra suggested to keep all variables on the stack. This simplifies things greatly and it has the added benefit of cache locality. Two, faulting on the iret. I really wanted to make this work, but it was becoming very hacky, and I never got it to be stable. The iret already had a fault handler for userspace faulting with bad segment registers, and getting NMI to trigger a fault and detect it was very tricky. But for strange reasons, the system would usually take a double fault and crash. I never figured out why and decided to go with a simple "jmp" approach. The new approach I took also simplified things. Finally, the last problem with Linus's approach was to have the nested NMI handler do a ret instead of an iret to give the first NMI NMI-context again. The problem is that ret is much more limited than an iret. I couldn't figure out how to get the stack back where it belonged. I could have copied the current stack, pushed the return onto it, but my fear here is that there may be some place that writes data below the stack pointer. I know that is not something code should depend on, but I don't want to chance it. I may add this feature later, but for now, an NMI handler that loses NMI context will not get it back. Here's what is done: When an NMI comes in, the HW pushes the interrupt stack frame onto the per cpu NMI stack that is selected by the IST. A special location on the NMI stack holds a variable that is set when the first NMI handler runs. If this variable is set then we know that this is a nested NMI and we process the nested NMI code. There is still a race when this variable is cleared and an NMI comes in just before the first NMI does the return. For this case, if the variable is cleared, we also check if the interrupted stack is the NMI stack. If it is, then we process the nested NMI code. Why the two tests and not just test the interrupted stack? If the first NMI hits a breakpoint and loses NMI context, and then it hits another breakpoint and while processing that breakpoint we get a nested NMI. When processing a breakpoint, the stack changes to the breakpoint stack. If another NMI comes in here we can't rely on the interrupted stack to be the NMI stack. If the variable is not set and the interrupted task's stack is not the NMI stack, then we know this is the first NMI and we can process things normally. But in order to do so, we need to do a few things first. 1) Set the stack variable that tells us that we are in an NMI handler 2) Make two copies of the interrupt stack frame. One copy is used to return on iret The other is used to restore the first one if we have a nested NMI. This is what the stack will look like: +-------------------------+ | original SS | | original Return RSP | | original RFLAGS | | original CS | | original RIP | +-------------------------+ | temp storage for rdx | +-------------------------+ | NMI executing variable | +-------------------------+ | Saved SS | | Saved Return RSP | | Saved RFLAGS | | Saved CS | | Saved RIP | +-------------------------+ | copied SS | | copied Return RSP | | copied RFLAGS | | copied CS | | copied RIP | +-------------------------+ | pt_regs | +-------------------------+ The original stack frame contains what the HW put in when we entered the NMI. We store %rdx as a temp variable to use. Both the original HW stack frame and this %rdx storage will be clobbered by nested NMIs so we can not rely on them later in the first NMI handler. The next item is the special stack variable that is set when we execute the rest of the NMI handler. Then we have two copies of the interrupt stack. The second copy is modified by any nested NMIs to let the first NMI know that we triggered a second NMI (latched) and that we should repeat the NMI handler. If the first NMI hits an exception or breakpoint that takes it out of NMI context, if a second NMI comes in before the first one finishes, it will update the copied interrupt stack to point to a fix up location to trigger another NMI. When the first NMI calls iret, it will instead jump to the fix up location. This fix up location will copy the saved interrupt stack back to the copy and execute the nmi handler again. Note, the nested NMI knows enough to check if it preempted a previous NMI handler while it is in the fixup location. If it has, it will not modify the copied interrupt stack and will just leave as if nothing happened. As the NMI handle is about to execute again, there's no reason to latch now. To test all this, I forced the NMI handler to call iret and take itself out of NMI context. I also added assemble code to write to the serial to make sure that it hits the nested path as well as the fix up path. Everything seems to be working fine. Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: H. Peter Anvin <hpa@linux.intel.com> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Paul Turner <pjt@google.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Mathieu Desnoyers <mathieu.desnoyers@efficios.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2011-12-08 21:36:23 +04:00
1:
/* Set up the interrupted NMIs stack to jump to repeat_nmi */
leaq -1*8(%rsp), %rdx
movq %rdx, %rsp
leaq -10*8(%rsp), %rdx
pushq $__KERNEL_DS
pushq %rdx
x86/debug: Remove perpetually broken, unmaintainable dwarf annotations So the dwarf2 annotations in low level assembly code have become an increasing hindrance: unreadable, messy macros mixed into some of the most security sensitive code paths of the Linux kernel. These debug info annotations don't even buy the upstream kernel anything: dwarf driven stack unwinding has caused problems in the past so it's out of tree, and the upstream kernel only uses the much more robust framepointers based stack unwinding method. In addition to that there's a steady, slow bitrot going on with these annotations, requiring frequent fixups. There's no tooling and no functionality upstream that keeps it correct. So burn down the sick forest, allowing new, healthier growth: 27 files changed, 350 insertions(+), 1101 deletions(-) Someone who has the willingness and time to do this properly can attempt to reintroduce dwarf debuginfo in x86 assembly code plus dwarf unwinding from first principles, with the following conditions: - it should be maximally readable, and maximally low-key to 'ordinary' code reading and maintenance. - find a build time method to insert dwarf annotations automatically in the most common cases, for pop/push instructions that manipulate the stack pointer. This could be done for example via a preprocessing step that just looks for common patterns - plus special annotations for the few cases where we want to depart from the default. We have hundreds of CFI annotations, so automating most of that makes sense. - it should come with build tooling checks that ensure that CFI annotations are sensible. We've seen such efforts from the framepointer side, and there's no reason it couldn't be done on the dwarf side. Cc: Andy Lutomirski <luto@amacapital.net> Cc: Borislav Petkov <bp@alien8.de> Cc: Brian Gerst <brgerst@gmail.com> Cc: Denys Vlasenko <dvlasenk@redhat.com> Cc: Frédéric Weisbecker <fweisbec@gmail.com Cc: H. Peter Anvin <hpa@zytor.com> Cc: Jan Beulich <JBeulich@suse.com> Cc: Josh Poimboeuf <jpoimboe@redhat.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: linux-kernel@vger.kernel.org Signed-off-by: Ingo Molnar <mingo@kernel.org>
2015-05-28 13:21:47 +03:00
pushfq
pushq $__KERNEL_CS
pushq $repeat_nmi
x86: Add workaround to NMI iret woes In x86, when an NMI goes off, the CPU goes into an NMI context that prevents other NMIs to trigger on that CPU. If an NMI is suppose to trigger, it has to wait till the previous NMI leaves NMI context. At that time, the next NMI can trigger (note, only one more NMI will trigger, as only one can be latched at a time). The way x86 gets out of NMI context is by calling iret. The problem with this is that this causes problems if the NMI handle either triggers an exception, or a breakpoint. Both the exception and the breakpoint handlers will finish with an iret. If this happens while in NMI context, the CPU will leave NMI context and a new NMI may come in. As NMI handlers are not made to be re-entrant, this can cause havoc with the system, not to mention, the nested NMI will write all over the previous NMI's stack. Linus Torvalds proposed the following workaround to this problem: https://lkml.org/lkml/2010/7/14/264 "In fact, I wonder if we couldn't just do a software NMI disable instead? Hav ea per-cpu variable (in the _core_ percpu areas that get allocated statically) that points to the NMI stack frame, and just make the NMI code itself do something like NMI entry: - load percpu NMI stack frame pointer - if non-zero we know we're nested, and should ignore this NMI: - we're returning to kernel mode, so return immediately by using "popf/ret", which also keeps NMI's disabled in the hardware until the "real" NMI iret happens. - before the popf/iret, use the NMI stack pointer to make the NMI return stack be invalid and cause a fault - set the NMI stack pointer to the current stack pointer NMI exit (not the above "immediate exit because we nested"): clear the percpu NMI stack pointer Just do the iret. Now, the thing is, now the "iret" is atomic. If we had a nested NMI, we'll take a fault, and that re-does our "delayed" NMI - and NMI's will stay masked. And if we didn't have a nested NMI, that iret will now unmask NMI's, and everything is happy." I first tried to follow this advice but as I started implementing this code, a few gotchas showed up. One, is accessing per-cpu variables in the NMI handler. The problem is that per-cpu variables use the %gs register to get the variable for the given CPU. But as the NMI may happen in userspace, we must first perform a SWAPGS to get to it. The NMI handler already does this later in the code, but its too late as we have saved off all the registers and we don't want to do that for a disabled NMI. Peter Zijlstra suggested to keep all variables on the stack. This simplifies things greatly and it has the added benefit of cache locality. Two, faulting on the iret. I really wanted to make this work, but it was becoming very hacky, and I never got it to be stable. The iret already had a fault handler for userspace faulting with bad segment registers, and getting NMI to trigger a fault and detect it was very tricky. But for strange reasons, the system would usually take a double fault and crash. I never figured out why and decided to go with a simple "jmp" approach. The new approach I took also simplified things. Finally, the last problem with Linus's approach was to have the nested NMI handler do a ret instead of an iret to give the first NMI NMI-context again. The problem is that ret is much more limited than an iret. I couldn't figure out how to get the stack back where it belonged. I could have copied the current stack, pushed the return onto it, but my fear here is that there may be some place that writes data below the stack pointer. I know that is not something code should depend on, but I don't want to chance it. I may add this feature later, but for now, an NMI handler that loses NMI context will not get it back. Here's what is done: When an NMI comes in, the HW pushes the interrupt stack frame onto the per cpu NMI stack that is selected by the IST. A special location on the NMI stack holds a variable that is set when the first NMI handler runs. If this variable is set then we know that this is a nested NMI and we process the nested NMI code. There is still a race when this variable is cleared and an NMI comes in just before the first NMI does the return. For this case, if the variable is cleared, we also check if the interrupted stack is the NMI stack. If it is, then we process the nested NMI code. Why the two tests and not just test the interrupted stack? If the first NMI hits a breakpoint and loses NMI context, and then it hits another breakpoint and while processing that breakpoint we get a nested NMI. When processing a breakpoint, the stack changes to the breakpoint stack. If another NMI comes in here we can't rely on the interrupted stack to be the NMI stack. If the variable is not set and the interrupted task's stack is not the NMI stack, then we know this is the first NMI and we can process things normally. But in order to do so, we need to do a few things first. 1) Set the stack variable that tells us that we are in an NMI handler 2) Make two copies of the interrupt stack frame. One copy is used to return on iret The other is used to restore the first one if we have a nested NMI. This is what the stack will look like: +-------------------------+ | original SS | | original Return RSP | | original RFLAGS | | original CS | | original RIP | +-------------------------+ | temp storage for rdx | +-------------------------+ | NMI executing variable | +-------------------------+ | Saved SS | | Saved Return RSP | | Saved RFLAGS | | Saved CS | | Saved RIP | +-------------------------+ | copied SS | | copied Return RSP | | copied RFLAGS | | copied CS | | copied RIP | +-------------------------+ | pt_regs | +-------------------------+ The original stack frame contains what the HW put in when we entered the NMI. We store %rdx as a temp variable to use. Both the original HW stack frame and this %rdx storage will be clobbered by nested NMIs so we can not rely on them later in the first NMI handler. The next item is the special stack variable that is set when we execute the rest of the NMI handler. Then we have two copies of the interrupt stack. The second copy is modified by any nested NMIs to let the first NMI know that we triggered a second NMI (latched) and that we should repeat the NMI handler. If the first NMI hits an exception or breakpoint that takes it out of NMI context, if a second NMI comes in before the first one finishes, it will update the copied interrupt stack to point to a fix up location to trigger another NMI. When the first NMI calls iret, it will instead jump to the fix up location. This fix up location will copy the saved interrupt stack back to the copy and execute the nmi handler again. Note, the nested NMI knows enough to check if it preempted a previous NMI handler while it is in the fixup location. If it has, it will not modify the copied interrupt stack and will just leave as if nothing happened. As the NMI handle is about to execute again, there's no reason to latch now. To test all this, I forced the NMI handler to call iret and take itself out of NMI context. I also added assemble code to write to the serial to make sure that it hits the nested path as well as the fix up path. Everything seems to be working fine. Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: H. Peter Anvin <hpa@linux.intel.com> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Paul Turner <pjt@google.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Mathieu Desnoyers <mathieu.desnoyers@efficios.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2011-12-08 21:36:23 +04:00
/* Put stack back */
addq $(6*8), %rsp
x86: Add workaround to NMI iret woes In x86, when an NMI goes off, the CPU goes into an NMI context that prevents other NMIs to trigger on that CPU. If an NMI is suppose to trigger, it has to wait till the previous NMI leaves NMI context. At that time, the next NMI can trigger (note, only one more NMI will trigger, as only one can be latched at a time). The way x86 gets out of NMI context is by calling iret. The problem with this is that this causes problems if the NMI handle either triggers an exception, or a breakpoint. Both the exception and the breakpoint handlers will finish with an iret. If this happens while in NMI context, the CPU will leave NMI context and a new NMI may come in. As NMI handlers are not made to be re-entrant, this can cause havoc with the system, not to mention, the nested NMI will write all over the previous NMI's stack. Linus Torvalds proposed the following workaround to this problem: https://lkml.org/lkml/2010/7/14/264 "In fact, I wonder if we couldn't just do a software NMI disable instead? Hav ea per-cpu variable (in the _core_ percpu areas that get allocated statically) that points to the NMI stack frame, and just make the NMI code itself do something like NMI entry: - load percpu NMI stack frame pointer - if non-zero we know we're nested, and should ignore this NMI: - we're returning to kernel mode, so return immediately by using "popf/ret", which also keeps NMI's disabled in the hardware until the "real" NMI iret happens. - before the popf/iret, use the NMI stack pointer to make the NMI return stack be invalid and cause a fault - set the NMI stack pointer to the current stack pointer NMI exit (not the above "immediate exit because we nested"): clear the percpu NMI stack pointer Just do the iret. Now, the thing is, now the "iret" is atomic. If we had a nested NMI, we'll take a fault, and that re-does our "delayed" NMI - and NMI's will stay masked. And if we didn't have a nested NMI, that iret will now unmask NMI's, and everything is happy." I first tried to follow this advice but as I started implementing this code, a few gotchas showed up. One, is accessing per-cpu variables in the NMI handler. The problem is that per-cpu variables use the %gs register to get the variable for the given CPU. But as the NMI may happen in userspace, we must first perform a SWAPGS to get to it. The NMI handler already does this later in the code, but its too late as we have saved off all the registers and we don't want to do that for a disabled NMI. Peter Zijlstra suggested to keep all variables on the stack. This simplifies things greatly and it has the added benefit of cache locality. Two, faulting on the iret. I really wanted to make this work, but it was becoming very hacky, and I never got it to be stable. The iret already had a fault handler for userspace faulting with bad segment registers, and getting NMI to trigger a fault and detect it was very tricky. But for strange reasons, the system would usually take a double fault and crash. I never figured out why and decided to go with a simple "jmp" approach. The new approach I took also simplified things. Finally, the last problem with Linus's approach was to have the nested NMI handler do a ret instead of an iret to give the first NMI NMI-context again. The problem is that ret is much more limited than an iret. I couldn't figure out how to get the stack back where it belonged. I could have copied the current stack, pushed the return onto it, but my fear here is that there may be some place that writes data below the stack pointer. I know that is not something code should depend on, but I don't want to chance it. I may add this feature later, but for now, an NMI handler that loses NMI context will not get it back. Here's what is done: When an NMI comes in, the HW pushes the interrupt stack frame onto the per cpu NMI stack that is selected by the IST. A special location on the NMI stack holds a variable that is set when the first NMI handler runs. If this variable is set then we know that this is a nested NMI and we process the nested NMI code. There is still a race when this variable is cleared and an NMI comes in just before the first NMI does the return. For this case, if the variable is cleared, we also check if the interrupted stack is the NMI stack. If it is, then we process the nested NMI code. Why the two tests and not just test the interrupted stack? If the first NMI hits a breakpoint and loses NMI context, and then it hits another breakpoint and while processing that breakpoint we get a nested NMI. When processing a breakpoint, the stack changes to the breakpoint stack. If another NMI comes in here we can't rely on the interrupted stack to be the NMI stack. If the variable is not set and the interrupted task's stack is not the NMI stack, then we know this is the first NMI and we can process things normally. But in order to do so, we need to do a few things first. 1) Set the stack variable that tells us that we are in an NMI handler 2) Make two copies of the interrupt stack frame. One copy is used to return on iret The other is used to restore the first one if we have a nested NMI. This is what the stack will look like: +-------------------------+ | original SS | | original Return RSP | | original RFLAGS | | original CS | | original RIP | +-------------------------+ | temp storage for rdx | +-------------------------+ | NMI executing variable | +-------------------------+ | Saved SS | | Saved Return RSP | | Saved RFLAGS | | Saved CS | | Saved RIP | +-------------------------+ | copied SS | | copied Return RSP | | copied RFLAGS | | copied CS | | copied RIP | +-------------------------+ | pt_regs | +-------------------------+ The original stack frame contains what the HW put in when we entered the NMI. We store %rdx as a temp variable to use. Both the original HW stack frame and this %rdx storage will be clobbered by nested NMIs so we can not rely on them later in the first NMI handler. The next item is the special stack variable that is set when we execute the rest of the NMI handler. Then we have two copies of the interrupt stack. The second copy is modified by any nested NMIs to let the first NMI know that we triggered a second NMI (latched) and that we should repeat the NMI handler. If the first NMI hits an exception or breakpoint that takes it out of NMI context, if a second NMI comes in before the first one finishes, it will update the copied interrupt stack to point to a fix up location to trigger another NMI. When the first NMI calls iret, it will instead jump to the fix up location. This fix up location will copy the saved interrupt stack back to the copy and execute the nmi handler again. Note, the nested NMI knows enough to check if it preempted a previous NMI handler while it is in the fixup location. If it has, it will not modify the copied interrupt stack and will just leave as if nothing happened. As the NMI handle is about to execute again, there's no reason to latch now. To test all this, I forced the NMI handler to call iret and take itself out of NMI context. I also added assemble code to write to the serial to make sure that it hits the nested path as well as the fix up path. Everything seems to be working fine. Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: H. Peter Anvin <hpa@linux.intel.com> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Paul Turner <pjt@google.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Mathieu Desnoyers <mathieu.desnoyers@efficios.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2011-12-08 21:36:23 +04:00
nested_nmi_out:
popq %rdx
x86: Add workaround to NMI iret woes In x86, when an NMI goes off, the CPU goes into an NMI context that prevents other NMIs to trigger on that CPU. If an NMI is suppose to trigger, it has to wait till the previous NMI leaves NMI context. At that time, the next NMI can trigger (note, only one more NMI will trigger, as only one can be latched at a time). The way x86 gets out of NMI context is by calling iret. The problem with this is that this causes problems if the NMI handle either triggers an exception, or a breakpoint. Both the exception and the breakpoint handlers will finish with an iret. If this happens while in NMI context, the CPU will leave NMI context and a new NMI may come in. As NMI handlers are not made to be re-entrant, this can cause havoc with the system, not to mention, the nested NMI will write all over the previous NMI's stack. Linus Torvalds proposed the following workaround to this problem: https://lkml.org/lkml/2010/7/14/264 "In fact, I wonder if we couldn't just do a software NMI disable instead? Hav ea per-cpu variable (in the _core_ percpu areas that get allocated statically) that points to the NMI stack frame, and just make the NMI code itself do something like NMI entry: - load percpu NMI stack frame pointer - if non-zero we know we're nested, and should ignore this NMI: - we're returning to kernel mode, so return immediately by using "popf/ret", which also keeps NMI's disabled in the hardware until the "real" NMI iret happens. - before the popf/iret, use the NMI stack pointer to make the NMI return stack be invalid and cause a fault - set the NMI stack pointer to the current stack pointer NMI exit (not the above "immediate exit because we nested"): clear the percpu NMI stack pointer Just do the iret. Now, the thing is, now the "iret" is atomic. If we had a nested NMI, we'll take a fault, and that re-does our "delayed" NMI - and NMI's will stay masked. And if we didn't have a nested NMI, that iret will now unmask NMI's, and everything is happy." I first tried to follow this advice but as I started implementing this code, a few gotchas showed up. One, is accessing per-cpu variables in the NMI handler. The problem is that per-cpu variables use the %gs register to get the variable for the given CPU. But as the NMI may happen in userspace, we must first perform a SWAPGS to get to it. The NMI handler already does this later in the code, but its too late as we have saved off all the registers and we don't want to do that for a disabled NMI. Peter Zijlstra suggested to keep all variables on the stack. This simplifies things greatly and it has the added benefit of cache locality. Two, faulting on the iret. I really wanted to make this work, but it was becoming very hacky, and I never got it to be stable. The iret already had a fault handler for userspace faulting with bad segment registers, and getting NMI to trigger a fault and detect it was very tricky. But for strange reasons, the system would usually take a double fault and crash. I never figured out why and decided to go with a simple "jmp" approach. The new approach I took also simplified things. Finally, the last problem with Linus's approach was to have the nested NMI handler do a ret instead of an iret to give the first NMI NMI-context again. The problem is that ret is much more limited than an iret. I couldn't figure out how to get the stack back where it belonged. I could have copied the current stack, pushed the return onto it, but my fear here is that there may be some place that writes data below the stack pointer. I know that is not something code should depend on, but I don't want to chance it. I may add this feature later, but for now, an NMI handler that loses NMI context will not get it back. Here's what is done: When an NMI comes in, the HW pushes the interrupt stack frame onto the per cpu NMI stack that is selected by the IST. A special location on the NMI stack holds a variable that is set when the first NMI handler runs. If this variable is set then we know that this is a nested NMI and we process the nested NMI code. There is still a race when this variable is cleared and an NMI comes in just before the first NMI does the return. For this case, if the variable is cleared, we also check if the interrupted stack is the NMI stack. If it is, then we process the nested NMI code. Why the two tests and not just test the interrupted stack? If the first NMI hits a breakpoint and loses NMI context, and then it hits another breakpoint and while processing that breakpoint we get a nested NMI. When processing a breakpoint, the stack changes to the breakpoint stack. If another NMI comes in here we can't rely on the interrupted stack to be the NMI stack. If the variable is not set and the interrupted task's stack is not the NMI stack, then we know this is the first NMI and we can process things normally. But in order to do so, we need to do a few things first. 1) Set the stack variable that tells us that we are in an NMI handler 2) Make two copies of the interrupt stack frame. One copy is used to return on iret The other is used to restore the first one if we have a nested NMI. This is what the stack will look like: +-------------------------+ | original SS | | original Return RSP | | original RFLAGS | | original CS | | original RIP | +-------------------------+ | temp storage for rdx | +-------------------------+ | NMI executing variable | +-------------------------+ | Saved SS | | Saved Return RSP | | Saved RFLAGS | | Saved CS | | Saved RIP | +-------------------------+ | copied SS | | copied Return RSP | | copied RFLAGS | | copied CS | | copied RIP | +-------------------------+ | pt_regs | +-------------------------+ The original stack frame contains what the HW put in when we entered the NMI. We store %rdx as a temp variable to use. Both the original HW stack frame and this %rdx storage will be clobbered by nested NMIs so we can not rely on them later in the first NMI handler. The next item is the special stack variable that is set when we execute the rest of the NMI handler. Then we have two copies of the interrupt stack. The second copy is modified by any nested NMIs to let the first NMI know that we triggered a second NMI (latched) and that we should repeat the NMI handler. If the first NMI hits an exception or breakpoint that takes it out of NMI context, if a second NMI comes in before the first one finishes, it will update the copied interrupt stack to point to a fix up location to trigger another NMI. When the first NMI calls iret, it will instead jump to the fix up location. This fix up location will copy the saved interrupt stack back to the copy and execute the nmi handler again. Note, the nested NMI knows enough to check if it preempted a previous NMI handler while it is in the fixup location. If it has, it will not modify the copied interrupt stack and will just leave as if nothing happened. As the NMI handle is about to execute again, there's no reason to latch now. To test all this, I forced the NMI handler to call iret and take itself out of NMI context. I also added assemble code to write to the serial to make sure that it hits the nested path as well as the fix up path. Everything seems to be working fine. Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: H. Peter Anvin <hpa@linux.intel.com> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Paul Turner <pjt@google.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Mathieu Desnoyers <mathieu.desnoyers@efficios.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2011-12-08 21:36:23 +04:00
/* No need to check faults here */
INTERRUPT_RETURN
first_nmi:
/*
* Because nested NMIs will use the pushed location that we
* stored in rdx, we must keep that space available.
* Here's what our stack frame will look like:
* +-------------------------+
* | original SS |
* | original Return RSP |
* | original RFLAGS |
* | original CS |
* | original RIP |
* +-------------------------+
* | temp storage for rdx |
* +-------------------------+
* | NMI executing variable |
* +-------------------------+
* | copied SS |
* | copied Return RSP |
* | copied RFLAGS |
* | copied CS |
* | copied RIP |
* +-------------------------+
* | Saved SS |
* | Saved Return RSP |
* | Saved RFLAGS |
* | Saved CS |
* | Saved RIP |
* +-------------------------+
x86: Add workaround to NMI iret woes In x86, when an NMI goes off, the CPU goes into an NMI context that prevents other NMIs to trigger on that CPU. If an NMI is suppose to trigger, it has to wait till the previous NMI leaves NMI context. At that time, the next NMI can trigger (note, only one more NMI will trigger, as only one can be latched at a time). The way x86 gets out of NMI context is by calling iret. The problem with this is that this causes problems if the NMI handle either triggers an exception, or a breakpoint. Both the exception and the breakpoint handlers will finish with an iret. If this happens while in NMI context, the CPU will leave NMI context and a new NMI may come in. As NMI handlers are not made to be re-entrant, this can cause havoc with the system, not to mention, the nested NMI will write all over the previous NMI's stack. Linus Torvalds proposed the following workaround to this problem: https://lkml.org/lkml/2010/7/14/264 "In fact, I wonder if we couldn't just do a software NMI disable instead? Hav ea per-cpu variable (in the _core_ percpu areas that get allocated statically) that points to the NMI stack frame, and just make the NMI code itself do something like NMI entry: - load percpu NMI stack frame pointer - if non-zero we know we're nested, and should ignore this NMI: - we're returning to kernel mode, so return immediately by using "popf/ret", which also keeps NMI's disabled in the hardware until the "real" NMI iret happens. - before the popf/iret, use the NMI stack pointer to make the NMI return stack be invalid and cause a fault - set the NMI stack pointer to the current stack pointer NMI exit (not the above "immediate exit because we nested"): clear the percpu NMI stack pointer Just do the iret. Now, the thing is, now the "iret" is atomic. If we had a nested NMI, we'll take a fault, and that re-does our "delayed" NMI - and NMI's will stay masked. And if we didn't have a nested NMI, that iret will now unmask NMI's, and everything is happy." I first tried to follow this advice but as I started implementing this code, a few gotchas showed up. One, is accessing per-cpu variables in the NMI handler. The problem is that per-cpu variables use the %gs register to get the variable for the given CPU. But as the NMI may happen in userspace, we must first perform a SWAPGS to get to it. The NMI handler already does this later in the code, but its too late as we have saved off all the registers and we don't want to do that for a disabled NMI. Peter Zijlstra suggested to keep all variables on the stack. This simplifies things greatly and it has the added benefit of cache locality. Two, faulting on the iret. I really wanted to make this work, but it was becoming very hacky, and I never got it to be stable. The iret already had a fault handler for userspace faulting with bad segment registers, and getting NMI to trigger a fault and detect it was very tricky. But for strange reasons, the system would usually take a double fault and crash. I never figured out why and decided to go with a simple "jmp" approach. The new approach I took also simplified things. Finally, the last problem with Linus's approach was to have the nested NMI handler do a ret instead of an iret to give the first NMI NMI-context again. The problem is that ret is much more limited than an iret. I couldn't figure out how to get the stack back where it belonged. I could have copied the current stack, pushed the return onto it, but my fear here is that there may be some place that writes data below the stack pointer. I know that is not something code should depend on, but I don't want to chance it. I may add this feature later, but for now, an NMI handler that loses NMI context will not get it back. Here's what is done: When an NMI comes in, the HW pushes the interrupt stack frame onto the per cpu NMI stack that is selected by the IST. A special location on the NMI stack holds a variable that is set when the first NMI handler runs. If this variable is set then we know that this is a nested NMI and we process the nested NMI code. There is still a race when this variable is cleared and an NMI comes in just before the first NMI does the return. For this case, if the variable is cleared, we also check if the interrupted stack is the NMI stack. If it is, then we process the nested NMI code. Why the two tests and not just test the interrupted stack? If the first NMI hits a breakpoint and loses NMI context, and then it hits another breakpoint and while processing that breakpoint we get a nested NMI. When processing a breakpoint, the stack changes to the breakpoint stack. If another NMI comes in here we can't rely on the interrupted stack to be the NMI stack. If the variable is not set and the interrupted task's stack is not the NMI stack, then we know this is the first NMI and we can process things normally. But in order to do so, we need to do a few things first. 1) Set the stack variable that tells us that we are in an NMI handler 2) Make two copies of the interrupt stack frame. One copy is used to return on iret The other is used to restore the first one if we have a nested NMI. This is what the stack will look like: +-------------------------+ | original SS | | original Return RSP | | original RFLAGS | | original CS | | original RIP | +-------------------------+ | temp storage for rdx | +-------------------------+ | NMI executing variable | +-------------------------+ | Saved SS | | Saved Return RSP | | Saved RFLAGS | | Saved CS | | Saved RIP | +-------------------------+ | copied SS | | copied Return RSP | | copied RFLAGS | | copied CS | | copied RIP | +-------------------------+ | pt_regs | +-------------------------+ The original stack frame contains what the HW put in when we entered the NMI. We store %rdx as a temp variable to use. Both the original HW stack frame and this %rdx storage will be clobbered by nested NMIs so we can not rely on them later in the first NMI handler. The next item is the special stack variable that is set when we execute the rest of the NMI handler. Then we have two copies of the interrupt stack. The second copy is modified by any nested NMIs to let the first NMI know that we triggered a second NMI (latched) and that we should repeat the NMI handler. If the first NMI hits an exception or breakpoint that takes it out of NMI context, if a second NMI comes in before the first one finishes, it will update the copied interrupt stack to point to a fix up location to trigger another NMI. When the first NMI calls iret, it will instead jump to the fix up location. This fix up location will copy the saved interrupt stack back to the copy and execute the nmi handler again. Note, the nested NMI knows enough to check if it preempted a previous NMI handler while it is in the fixup location. If it has, it will not modify the copied interrupt stack and will just leave as if nothing happened. As the NMI handle is about to execute again, there's no reason to latch now. To test all this, I forced the NMI handler to call iret and take itself out of NMI context. I also added assemble code to write to the serial to make sure that it hits the nested path as well as the fix up path. Everything seems to be working fine. Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: H. Peter Anvin <hpa@linux.intel.com> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Paul Turner <pjt@google.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Mathieu Desnoyers <mathieu.desnoyers@efficios.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2011-12-08 21:36:23 +04:00
* | pt_regs |
* +-------------------------+
*
* The saved stack frame is used to fix up the copied stack frame
* that a nested NMI may change to make the interrupted NMI iret jump
* to the repeat_nmi. The original stack frame and the temp storage
x86: Add workaround to NMI iret woes In x86, when an NMI goes off, the CPU goes into an NMI context that prevents other NMIs to trigger on that CPU. If an NMI is suppose to trigger, it has to wait till the previous NMI leaves NMI context. At that time, the next NMI can trigger (note, only one more NMI will trigger, as only one can be latched at a time). The way x86 gets out of NMI context is by calling iret. The problem with this is that this causes problems if the NMI handle either triggers an exception, or a breakpoint. Both the exception and the breakpoint handlers will finish with an iret. If this happens while in NMI context, the CPU will leave NMI context and a new NMI may come in. As NMI handlers are not made to be re-entrant, this can cause havoc with the system, not to mention, the nested NMI will write all over the previous NMI's stack. Linus Torvalds proposed the following workaround to this problem: https://lkml.org/lkml/2010/7/14/264 "In fact, I wonder if we couldn't just do a software NMI disable instead? Hav ea per-cpu variable (in the _core_ percpu areas that get allocated statically) that points to the NMI stack frame, and just make the NMI code itself do something like NMI entry: - load percpu NMI stack frame pointer - if non-zero we know we're nested, and should ignore this NMI: - we're returning to kernel mode, so return immediately by using "popf/ret", which also keeps NMI's disabled in the hardware until the "real" NMI iret happens. - before the popf/iret, use the NMI stack pointer to make the NMI return stack be invalid and cause a fault - set the NMI stack pointer to the current stack pointer NMI exit (not the above "immediate exit because we nested"): clear the percpu NMI stack pointer Just do the iret. Now, the thing is, now the "iret" is atomic. If we had a nested NMI, we'll take a fault, and that re-does our "delayed" NMI - and NMI's will stay masked. And if we didn't have a nested NMI, that iret will now unmask NMI's, and everything is happy." I first tried to follow this advice but as I started implementing this code, a few gotchas showed up. One, is accessing per-cpu variables in the NMI handler. The problem is that per-cpu variables use the %gs register to get the variable for the given CPU. But as the NMI may happen in userspace, we must first perform a SWAPGS to get to it. The NMI handler already does this later in the code, but its too late as we have saved off all the registers and we don't want to do that for a disabled NMI. Peter Zijlstra suggested to keep all variables on the stack. This simplifies things greatly and it has the added benefit of cache locality. Two, faulting on the iret. I really wanted to make this work, but it was becoming very hacky, and I never got it to be stable. The iret already had a fault handler for userspace faulting with bad segment registers, and getting NMI to trigger a fault and detect it was very tricky. But for strange reasons, the system would usually take a double fault and crash. I never figured out why and decided to go with a simple "jmp" approach. The new approach I took also simplified things. Finally, the last problem with Linus's approach was to have the nested NMI handler do a ret instead of an iret to give the first NMI NMI-context again. The problem is that ret is much more limited than an iret. I couldn't figure out how to get the stack back where it belonged. I could have copied the current stack, pushed the return onto it, but my fear here is that there may be some place that writes data below the stack pointer. I know that is not something code should depend on, but I don't want to chance it. I may add this feature later, but for now, an NMI handler that loses NMI context will not get it back. Here's what is done: When an NMI comes in, the HW pushes the interrupt stack frame onto the per cpu NMI stack that is selected by the IST. A special location on the NMI stack holds a variable that is set when the first NMI handler runs. If this variable is set then we know that this is a nested NMI and we process the nested NMI code. There is still a race when this variable is cleared and an NMI comes in just before the first NMI does the return. For this case, if the variable is cleared, we also check if the interrupted stack is the NMI stack. If it is, then we process the nested NMI code. Why the two tests and not just test the interrupted stack? If the first NMI hits a breakpoint and loses NMI context, and then it hits another breakpoint and while processing that breakpoint we get a nested NMI. When processing a breakpoint, the stack changes to the breakpoint stack. If another NMI comes in here we can't rely on the interrupted stack to be the NMI stack. If the variable is not set and the interrupted task's stack is not the NMI stack, then we know this is the first NMI and we can process things normally. But in order to do so, we need to do a few things first. 1) Set the stack variable that tells us that we are in an NMI handler 2) Make two copies of the interrupt stack frame. One copy is used to return on iret The other is used to restore the first one if we have a nested NMI. This is what the stack will look like: +-------------------------+ | original SS | | original Return RSP | | original RFLAGS | | original CS | | original RIP | +-------------------------+ | temp storage for rdx | +-------------------------+ | NMI executing variable | +-------------------------+ | Saved SS | | Saved Return RSP | | Saved RFLAGS | | Saved CS | | Saved RIP | +-------------------------+ | copied SS | | copied Return RSP | | copied RFLAGS | | copied CS | | copied RIP | +-------------------------+ | pt_regs | +-------------------------+ The original stack frame contains what the HW put in when we entered the NMI. We store %rdx as a temp variable to use. Both the original HW stack frame and this %rdx storage will be clobbered by nested NMIs so we can not rely on them later in the first NMI handler. The next item is the special stack variable that is set when we execute the rest of the NMI handler. Then we have two copies of the interrupt stack. The second copy is modified by any nested NMIs to let the first NMI know that we triggered a second NMI (latched) and that we should repeat the NMI handler. If the first NMI hits an exception or breakpoint that takes it out of NMI context, if a second NMI comes in before the first one finishes, it will update the copied interrupt stack to point to a fix up location to trigger another NMI. When the first NMI calls iret, it will instead jump to the fix up location. This fix up location will copy the saved interrupt stack back to the copy and execute the nmi handler again. Note, the nested NMI knows enough to check if it preempted a previous NMI handler while it is in the fixup location. If it has, it will not modify the copied interrupt stack and will just leave as if nothing happened. As the NMI handle is about to execute again, there's no reason to latch now. To test all this, I forced the NMI handler to call iret and take itself out of NMI context. I also added assemble code to write to the serial to make sure that it hits the nested path as well as the fix up path. Everything seems to be working fine. Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: H. Peter Anvin <hpa@linux.intel.com> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Paul Turner <pjt@google.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Mathieu Desnoyers <mathieu.desnoyers@efficios.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2011-12-08 21:36:23 +04:00
* is also used by nested NMIs and can not be trusted on exit.
*/
/* Do not pop rdx, nested NMIs will corrupt that part of the stack */
movq (%rsp), %rdx
x86: Add workaround to NMI iret woes In x86, when an NMI goes off, the CPU goes into an NMI context that prevents other NMIs to trigger on that CPU. If an NMI is suppose to trigger, it has to wait till the previous NMI leaves NMI context. At that time, the next NMI can trigger (note, only one more NMI will trigger, as only one can be latched at a time). The way x86 gets out of NMI context is by calling iret. The problem with this is that this causes problems if the NMI handle either triggers an exception, or a breakpoint. Both the exception and the breakpoint handlers will finish with an iret. If this happens while in NMI context, the CPU will leave NMI context and a new NMI may come in. As NMI handlers are not made to be re-entrant, this can cause havoc with the system, not to mention, the nested NMI will write all over the previous NMI's stack. Linus Torvalds proposed the following workaround to this problem: https://lkml.org/lkml/2010/7/14/264 "In fact, I wonder if we couldn't just do a software NMI disable instead? Hav ea per-cpu variable (in the _core_ percpu areas that get allocated statically) that points to the NMI stack frame, and just make the NMI code itself do something like NMI entry: - load percpu NMI stack frame pointer - if non-zero we know we're nested, and should ignore this NMI: - we're returning to kernel mode, so return immediately by using "popf/ret", which also keeps NMI's disabled in the hardware until the "real" NMI iret happens. - before the popf/iret, use the NMI stack pointer to make the NMI return stack be invalid and cause a fault - set the NMI stack pointer to the current stack pointer NMI exit (not the above "immediate exit because we nested"): clear the percpu NMI stack pointer Just do the iret. Now, the thing is, now the "iret" is atomic. If we had a nested NMI, we'll take a fault, and that re-does our "delayed" NMI - and NMI's will stay masked. And if we didn't have a nested NMI, that iret will now unmask NMI's, and everything is happy." I first tried to follow this advice but as I started implementing this code, a few gotchas showed up. One, is accessing per-cpu variables in the NMI handler. The problem is that per-cpu variables use the %gs register to get the variable for the given CPU. But as the NMI may happen in userspace, we must first perform a SWAPGS to get to it. The NMI handler already does this later in the code, but its too late as we have saved off all the registers and we don't want to do that for a disabled NMI. Peter Zijlstra suggested to keep all variables on the stack. This simplifies things greatly and it has the added benefit of cache locality. Two, faulting on the iret. I really wanted to make this work, but it was becoming very hacky, and I never got it to be stable. The iret already had a fault handler for userspace faulting with bad segment registers, and getting NMI to trigger a fault and detect it was very tricky. But for strange reasons, the system would usually take a double fault and crash. I never figured out why and decided to go with a simple "jmp" approach. The new approach I took also simplified things. Finally, the last problem with Linus's approach was to have the nested NMI handler do a ret instead of an iret to give the first NMI NMI-context again. The problem is that ret is much more limited than an iret. I couldn't figure out how to get the stack back where it belonged. I could have copied the current stack, pushed the return onto it, but my fear here is that there may be some place that writes data below the stack pointer. I know that is not something code should depend on, but I don't want to chance it. I may add this feature later, but for now, an NMI handler that loses NMI context will not get it back. Here's what is done: When an NMI comes in, the HW pushes the interrupt stack frame onto the per cpu NMI stack that is selected by the IST. A special location on the NMI stack holds a variable that is set when the first NMI handler runs. If this variable is set then we know that this is a nested NMI and we process the nested NMI code. There is still a race when this variable is cleared and an NMI comes in just before the first NMI does the return. For this case, if the variable is cleared, we also check if the interrupted stack is the NMI stack. If it is, then we process the nested NMI code. Why the two tests and not just test the interrupted stack? If the first NMI hits a breakpoint and loses NMI context, and then it hits another breakpoint and while processing that breakpoint we get a nested NMI. When processing a breakpoint, the stack changes to the breakpoint stack. If another NMI comes in here we can't rely on the interrupted stack to be the NMI stack. If the variable is not set and the interrupted task's stack is not the NMI stack, then we know this is the first NMI and we can process things normally. But in order to do so, we need to do a few things first. 1) Set the stack variable that tells us that we are in an NMI handler 2) Make two copies of the interrupt stack frame. One copy is used to return on iret The other is used to restore the first one if we have a nested NMI. This is what the stack will look like: +-------------------------+ | original SS | | original Return RSP | | original RFLAGS | | original CS | | original RIP | +-------------------------+ | temp storage for rdx | +-------------------------+ | NMI executing variable | +-------------------------+ | Saved SS | | Saved Return RSP | | Saved RFLAGS | | Saved CS | | Saved RIP | +-------------------------+ | copied SS | | copied Return RSP | | copied RFLAGS | | copied CS | | copied RIP | +-------------------------+ | pt_regs | +-------------------------+ The original stack frame contains what the HW put in when we entered the NMI. We store %rdx as a temp variable to use. Both the original HW stack frame and this %rdx storage will be clobbered by nested NMIs so we can not rely on them later in the first NMI handler. The next item is the special stack variable that is set when we execute the rest of the NMI handler. Then we have two copies of the interrupt stack. The second copy is modified by any nested NMIs to let the first NMI know that we triggered a second NMI (latched) and that we should repeat the NMI handler. If the first NMI hits an exception or breakpoint that takes it out of NMI context, if a second NMI comes in before the first one finishes, it will update the copied interrupt stack to point to a fix up location to trigger another NMI. When the first NMI calls iret, it will instead jump to the fix up location. This fix up location will copy the saved interrupt stack back to the copy and execute the nmi handler again. Note, the nested NMI knows enough to check if it preempted a previous NMI handler while it is in the fixup location. If it has, it will not modify the copied interrupt stack and will just leave as if nothing happened. As the NMI handle is about to execute again, there's no reason to latch now. To test all this, I forced the NMI handler to call iret and take itself out of NMI context. I also added assemble code to write to the serial to make sure that it hits the nested path as well as the fix up path. Everything seems to be working fine. Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: H. Peter Anvin <hpa@linux.intel.com> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Paul Turner <pjt@google.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Mathieu Desnoyers <mathieu.desnoyers@efficios.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2011-12-08 21:36:23 +04:00
/* Set the NMI executing variable on the stack. */
pushq $1
x86: Add workaround to NMI iret woes In x86, when an NMI goes off, the CPU goes into an NMI context that prevents other NMIs to trigger on that CPU. If an NMI is suppose to trigger, it has to wait till the previous NMI leaves NMI context. At that time, the next NMI can trigger (note, only one more NMI will trigger, as only one can be latched at a time). The way x86 gets out of NMI context is by calling iret. The problem with this is that this causes problems if the NMI handle either triggers an exception, or a breakpoint. Both the exception and the breakpoint handlers will finish with an iret. If this happens while in NMI context, the CPU will leave NMI context and a new NMI may come in. As NMI handlers are not made to be re-entrant, this can cause havoc with the system, not to mention, the nested NMI will write all over the previous NMI's stack. Linus Torvalds proposed the following workaround to this problem: https://lkml.org/lkml/2010/7/14/264 "In fact, I wonder if we couldn't just do a software NMI disable instead? Hav ea per-cpu variable (in the _core_ percpu areas that get allocated statically) that points to the NMI stack frame, and just make the NMI code itself do something like NMI entry: - load percpu NMI stack frame pointer - if non-zero we know we're nested, and should ignore this NMI: - we're returning to kernel mode, so return immediately by using "popf/ret", which also keeps NMI's disabled in the hardware until the "real" NMI iret happens. - before the popf/iret, use the NMI stack pointer to make the NMI return stack be invalid and cause a fault - set the NMI stack pointer to the current stack pointer NMI exit (not the above "immediate exit because we nested"): clear the percpu NMI stack pointer Just do the iret. Now, the thing is, now the "iret" is atomic. If we had a nested NMI, we'll take a fault, and that re-does our "delayed" NMI - and NMI's will stay masked. And if we didn't have a nested NMI, that iret will now unmask NMI's, and everything is happy." I first tried to follow this advice but as I started implementing this code, a few gotchas showed up. One, is accessing per-cpu variables in the NMI handler. The problem is that per-cpu variables use the %gs register to get the variable for the given CPU. But as the NMI may happen in userspace, we must first perform a SWAPGS to get to it. The NMI handler already does this later in the code, but its too late as we have saved off all the registers and we don't want to do that for a disabled NMI. Peter Zijlstra suggested to keep all variables on the stack. This simplifies things greatly and it has the added benefit of cache locality. Two, faulting on the iret. I really wanted to make this work, but it was becoming very hacky, and I never got it to be stable. The iret already had a fault handler for userspace faulting with bad segment registers, and getting NMI to trigger a fault and detect it was very tricky. But for strange reasons, the system would usually take a double fault and crash. I never figured out why and decided to go with a simple "jmp" approach. The new approach I took also simplified things. Finally, the last problem with Linus's approach was to have the nested NMI handler do a ret instead of an iret to give the first NMI NMI-context again. The problem is that ret is much more limited than an iret. I couldn't figure out how to get the stack back where it belonged. I could have copied the current stack, pushed the return onto it, but my fear here is that there may be some place that writes data below the stack pointer. I know that is not something code should depend on, but I don't want to chance it. I may add this feature later, but for now, an NMI handler that loses NMI context will not get it back. Here's what is done: When an NMI comes in, the HW pushes the interrupt stack frame onto the per cpu NMI stack that is selected by the IST. A special location on the NMI stack holds a variable that is set when the first NMI handler runs. If this variable is set then we know that this is a nested NMI and we process the nested NMI code. There is still a race when this variable is cleared and an NMI comes in just before the first NMI does the return. For this case, if the variable is cleared, we also check if the interrupted stack is the NMI stack. If it is, then we process the nested NMI code. Why the two tests and not just test the interrupted stack? If the first NMI hits a breakpoint and loses NMI context, and then it hits another breakpoint and while processing that breakpoint we get a nested NMI. When processing a breakpoint, the stack changes to the breakpoint stack. If another NMI comes in here we can't rely on the interrupted stack to be the NMI stack. If the variable is not set and the interrupted task's stack is not the NMI stack, then we know this is the first NMI and we can process things normally. But in order to do so, we need to do a few things first. 1) Set the stack variable that tells us that we are in an NMI handler 2) Make two copies of the interrupt stack frame. One copy is used to return on iret The other is used to restore the first one if we have a nested NMI. This is what the stack will look like: +-------------------------+ | original SS | | original Return RSP | | original RFLAGS | | original CS | | original RIP | +-------------------------+ | temp storage for rdx | +-------------------------+ | NMI executing variable | +-------------------------+ | Saved SS | | Saved Return RSP | | Saved RFLAGS | | Saved CS | | Saved RIP | +-------------------------+ | copied SS | | copied Return RSP | | copied RFLAGS | | copied CS | | copied RIP | +-------------------------+ | pt_regs | +-------------------------+ The original stack frame contains what the HW put in when we entered the NMI. We store %rdx as a temp variable to use. Both the original HW stack frame and this %rdx storage will be clobbered by nested NMIs so we can not rely on them later in the first NMI handler. The next item is the special stack variable that is set when we execute the rest of the NMI handler. Then we have two copies of the interrupt stack. The second copy is modified by any nested NMIs to let the first NMI know that we triggered a second NMI (latched) and that we should repeat the NMI handler. If the first NMI hits an exception or breakpoint that takes it out of NMI context, if a second NMI comes in before the first one finishes, it will update the copied interrupt stack to point to a fix up location to trigger another NMI. When the first NMI calls iret, it will instead jump to the fix up location. This fix up location will copy the saved interrupt stack back to the copy and execute the nmi handler again. Note, the nested NMI knows enough to check if it preempted a previous NMI handler while it is in the fixup location. If it has, it will not modify the copied interrupt stack and will just leave as if nothing happened. As the NMI handle is about to execute again, there's no reason to latch now. To test all this, I forced the NMI handler to call iret and take itself out of NMI context. I also added assemble code to write to the serial to make sure that it hits the nested path as well as the fix up path. Everything seems to be working fine. Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: H. Peter Anvin <hpa@linux.intel.com> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Paul Turner <pjt@google.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Mathieu Desnoyers <mathieu.desnoyers@efficios.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2011-12-08 21:36:23 +04:00
/* Leave room for the "copied" frame */
subq $(5*8), %rsp
x86: Add workaround to NMI iret woes In x86, when an NMI goes off, the CPU goes into an NMI context that prevents other NMIs to trigger on that CPU. If an NMI is suppose to trigger, it has to wait till the previous NMI leaves NMI context. At that time, the next NMI can trigger (note, only one more NMI will trigger, as only one can be latched at a time). The way x86 gets out of NMI context is by calling iret. The problem with this is that this causes problems if the NMI handle either triggers an exception, or a breakpoint. Both the exception and the breakpoint handlers will finish with an iret. If this happens while in NMI context, the CPU will leave NMI context and a new NMI may come in. As NMI handlers are not made to be re-entrant, this can cause havoc with the system, not to mention, the nested NMI will write all over the previous NMI's stack. Linus Torvalds proposed the following workaround to this problem: https://lkml.org/lkml/2010/7/14/264 "In fact, I wonder if we couldn't just do a software NMI disable instead? Hav ea per-cpu variable (in the _core_ percpu areas that get allocated statically) that points to the NMI stack frame, and just make the NMI code itself do something like NMI entry: - load percpu NMI stack frame pointer - if non-zero we know we're nested, and should ignore this NMI: - we're returning to kernel mode, so return immediately by using "popf/ret", which also keeps NMI's disabled in the hardware until the "real" NMI iret happens. - before the popf/iret, use the NMI stack pointer to make the NMI return stack be invalid and cause a fault - set the NMI stack pointer to the current stack pointer NMI exit (not the above "immediate exit because we nested"): clear the percpu NMI stack pointer Just do the iret. Now, the thing is, now the "iret" is atomic. If we had a nested NMI, we'll take a fault, and that re-does our "delayed" NMI - and NMI's will stay masked. And if we didn't have a nested NMI, that iret will now unmask NMI's, and everything is happy." I first tried to follow this advice but as I started implementing this code, a few gotchas showed up. One, is accessing per-cpu variables in the NMI handler. The problem is that per-cpu variables use the %gs register to get the variable for the given CPU. But as the NMI may happen in userspace, we must first perform a SWAPGS to get to it. The NMI handler already does this later in the code, but its too late as we have saved off all the registers and we don't want to do that for a disabled NMI. Peter Zijlstra suggested to keep all variables on the stack. This simplifies things greatly and it has the added benefit of cache locality. Two, faulting on the iret. I really wanted to make this work, but it was becoming very hacky, and I never got it to be stable. The iret already had a fault handler for userspace faulting with bad segment registers, and getting NMI to trigger a fault and detect it was very tricky. But for strange reasons, the system would usually take a double fault and crash. I never figured out why and decided to go with a simple "jmp" approach. The new approach I took also simplified things. Finally, the last problem with Linus's approach was to have the nested NMI handler do a ret instead of an iret to give the first NMI NMI-context again. The problem is that ret is much more limited than an iret. I couldn't figure out how to get the stack back where it belonged. I could have copied the current stack, pushed the return onto it, but my fear here is that there may be some place that writes data below the stack pointer. I know that is not something code should depend on, but I don't want to chance it. I may add this feature later, but for now, an NMI handler that loses NMI context will not get it back. Here's what is done: When an NMI comes in, the HW pushes the interrupt stack frame onto the per cpu NMI stack that is selected by the IST. A special location on the NMI stack holds a variable that is set when the first NMI handler runs. If this variable is set then we know that this is a nested NMI and we process the nested NMI code. There is still a race when this variable is cleared and an NMI comes in just before the first NMI does the return. For this case, if the variable is cleared, we also check if the interrupted stack is the NMI stack. If it is, then we process the nested NMI code. Why the two tests and not just test the interrupted stack? If the first NMI hits a breakpoint and loses NMI context, and then it hits another breakpoint and while processing that breakpoint we get a nested NMI. When processing a breakpoint, the stack changes to the breakpoint stack. If another NMI comes in here we can't rely on the interrupted stack to be the NMI stack. If the variable is not set and the interrupted task's stack is not the NMI stack, then we know this is the first NMI and we can process things normally. But in order to do so, we need to do a few things first. 1) Set the stack variable that tells us that we are in an NMI handler 2) Make two copies of the interrupt stack frame. One copy is used to return on iret The other is used to restore the first one if we have a nested NMI. This is what the stack will look like: +-------------------------+ | original SS | | original Return RSP | | original RFLAGS | | original CS | | original RIP | +-------------------------+ | temp storage for rdx | +-------------------------+ | NMI executing variable | +-------------------------+ | Saved SS | | Saved Return RSP | | Saved RFLAGS | | Saved CS | | Saved RIP | +-------------------------+ | copied SS | | copied Return RSP | | copied RFLAGS | | copied CS | | copied RIP | +-------------------------+ | pt_regs | +-------------------------+ The original stack frame contains what the HW put in when we entered the NMI. We store %rdx as a temp variable to use. Both the original HW stack frame and this %rdx storage will be clobbered by nested NMIs so we can not rely on them later in the first NMI handler. The next item is the special stack variable that is set when we execute the rest of the NMI handler. Then we have two copies of the interrupt stack. The second copy is modified by any nested NMIs to let the first NMI know that we triggered a second NMI (latched) and that we should repeat the NMI handler. If the first NMI hits an exception or breakpoint that takes it out of NMI context, if a second NMI comes in before the first one finishes, it will update the copied interrupt stack to point to a fix up location to trigger another NMI. When the first NMI calls iret, it will instead jump to the fix up location. This fix up location will copy the saved interrupt stack back to the copy and execute the nmi handler again. Note, the nested NMI knows enough to check if it preempted a previous NMI handler while it is in the fixup location. If it has, it will not modify the copied interrupt stack and will just leave as if nothing happened. As the NMI handle is about to execute again, there's no reason to latch now. To test all this, I forced the NMI handler to call iret and take itself out of NMI context. I also added assemble code to write to the serial to make sure that it hits the nested path as well as the fix up path. Everything seems to be working fine. Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: H. Peter Anvin <hpa@linux.intel.com> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Paul Turner <pjt@google.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Mathieu Desnoyers <mathieu.desnoyers@efficios.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2011-12-08 21:36:23 +04:00
/* Copy the stack frame to the Saved frame */
.rept 5
pushq 11*8(%rsp)
x86: Add workaround to NMI iret woes In x86, when an NMI goes off, the CPU goes into an NMI context that prevents other NMIs to trigger on that CPU. If an NMI is suppose to trigger, it has to wait till the previous NMI leaves NMI context. At that time, the next NMI can trigger (note, only one more NMI will trigger, as only one can be latched at a time). The way x86 gets out of NMI context is by calling iret. The problem with this is that this causes problems if the NMI handle either triggers an exception, or a breakpoint. Both the exception and the breakpoint handlers will finish with an iret. If this happens while in NMI context, the CPU will leave NMI context and a new NMI may come in. As NMI handlers are not made to be re-entrant, this can cause havoc with the system, not to mention, the nested NMI will write all over the previous NMI's stack. Linus Torvalds proposed the following workaround to this problem: https://lkml.org/lkml/2010/7/14/264 "In fact, I wonder if we couldn't just do a software NMI disable instead? Hav ea per-cpu variable (in the _core_ percpu areas that get allocated statically) that points to the NMI stack frame, and just make the NMI code itself do something like NMI entry: - load percpu NMI stack frame pointer - if non-zero we know we're nested, and should ignore this NMI: - we're returning to kernel mode, so return immediately by using "popf/ret", which also keeps NMI's disabled in the hardware until the "real" NMI iret happens. - before the popf/iret, use the NMI stack pointer to make the NMI return stack be invalid and cause a fault - set the NMI stack pointer to the current stack pointer NMI exit (not the above "immediate exit because we nested"): clear the percpu NMI stack pointer Just do the iret. Now, the thing is, now the "iret" is atomic. If we had a nested NMI, we'll take a fault, and that re-does our "delayed" NMI - and NMI's will stay masked. And if we didn't have a nested NMI, that iret will now unmask NMI's, and everything is happy." I first tried to follow this advice but as I started implementing this code, a few gotchas showed up. One, is accessing per-cpu variables in the NMI handler. The problem is that per-cpu variables use the %gs register to get the variable for the given CPU. But as the NMI may happen in userspace, we must first perform a SWAPGS to get to it. The NMI handler already does this later in the code, but its too late as we have saved off all the registers and we don't want to do that for a disabled NMI. Peter Zijlstra suggested to keep all variables on the stack. This simplifies things greatly and it has the added benefit of cache locality. Two, faulting on the iret. I really wanted to make this work, but it was becoming very hacky, and I never got it to be stable. The iret already had a fault handler for userspace faulting with bad segment registers, and getting NMI to trigger a fault and detect it was very tricky. But for strange reasons, the system would usually take a double fault and crash. I never figured out why and decided to go with a simple "jmp" approach. The new approach I took also simplified things. Finally, the last problem with Linus's approach was to have the nested NMI handler do a ret instead of an iret to give the first NMI NMI-context again. The problem is that ret is much more limited than an iret. I couldn't figure out how to get the stack back where it belonged. I could have copied the current stack, pushed the return onto it, but my fear here is that there may be some place that writes data below the stack pointer. I know that is not something code should depend on, but I don't want to chance it. I may add this feature later, but for now, an NMI handler that loses NMI context will not get it back. Here's what is done: When an NMI comes in, the HW pushes the interrupt stack frame onto the per cpu NMI stack that is selected by the IST. A special location on the NMI stack holds a variable that is set when the first NMI handler runs. If this variable is set then we know that this is a nested NMI and we process the nested NMI code. There is still a race when this variable is cleared and an NMI comes in just before the first NMI does the return. For this case, if the variable is cleared, we also check if the interrupted stack is the NMI stack. If it is, then we process the nested NMI code. Why the two tests and not just test the interrupted stack? If the first NMI hits a breakpoint and loses NMI context, and then it hits another breakpoint and while processing that breakpoint we get a nested NMI. When processing a breakpoint, the stack changes to the breakpoint stack. If another NMI comes in here we can't rely on the interrupted stack to be the NMI stack. If the variable is not set and the interrupted task's stack is not the NMI stack, then we know this is the first NMI and we can process things normally. But in order to do so, we need to do a few things first. 1) Set the stack variable that tells us that we are in an NMI handler 2) Make two copies of the interrupt stack frame. One copy is used to return on iret The other is used to restore the first one if we have a nested NMI. This is what the stack will look like: +-------------------------+ | original SS | | original Return RSP | | original RFLAGS | | original CS | | original RIP | +-------------------------+ | temp storage for rdx | +-------------------------+ | NMI executing variable | +-------------------------+ | Saved SS | | Saved Return RSP | | Saved RFLAGS | | Saved CS | | Saved RIP | +-------------------------+ | copied SS | | copied Return RSP | | copied RFLAGS | | copied CS | | copied RIP | +-------------------------+ | pt_regs | +-------------------------+ The original stack frame contains what the HW put in when we entered the NMI. We store %rdx as a temp variable to use. Both the original HW stack frame and this %rdx storage will be clobbered by nested NMIs so we can not rely on them later in the first NMI handler. The next item is the special stack variable that is set when we execute the rest of the NMI handler. Then we have two copies of the interrupt stack. The second copy is modified by any nested NMIs to let the first NMI know that we triggered a second NMI (latched) and that we should repeat the NMI handler. If the first NMI hits an exception or breakpoint that takes it out of NMI context, if a second NMI comes in before the first one finishes, it will update the copied interrupt stack to point to a fix up location to trigger another NMI. When the first NMI calls iret, it will instead jump to the fix up location. This fix up location will copy the saved interrupt stack back to the copy and execute the nmi handler again. Note, the nested NMI knows enough to check if it preempted a previous NMI handler while it is in the fixup location. If it has, it will not modify the copied interrupt stack and will just leave as if nothing happened. As the NMI handle is about to execute again, there's no reason to latch now. To test all this, I forced the NMI handler to call iret and take itself out of NMI context. I also added assemble code to write to the serial to make sure that it hits the nested path as well as the fix up path. Everything seems to be working fine. Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: H. Peter Anvin <hpa@linux.intel.com> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Paul Turner <pjt@google.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Mathieu Desnoyers <mathieu.desnoyers@efficios.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2011-12-08 21:36:23 +04:00
.endr
/* Everything up to here is safe from nested NMIs */
/*
* If there was a nested NMI, the first NMI's iret will return
* here. But NMIs are still enabled and we can take another
* nested NMI. The nested NMI checks the interrupted RIP to see
* if it is between repeat_nmi and end_repeat_nmi, and if so
* it will just return, as we are about to repeat an NMI anyway.
* This makes it safe to copy to the stack frame that a nested
* NMI will update.
*/
repeat_nmi:
/*
* Update the stack variable to say we are still in NMI (the update
* is benign for the non-repeat case, where 1 was pushed just above
* to this very stack slot).
*/
movq $1, 10*8(%rsp)
x86: Add workaround to NMI iret woes In x86, when an NMI goes off, the CPU goes into an NMI context that prevents other NMIs to trigger on that CPU. If an NMI is suppose to trigger, it has to wait till the previous NMI leaves NMI context. At that time, the next NMI can trigger (note, only one more NMI will trigger, as only one can be latched at a time). The way x86 gets out of NMI context is by calling iret. The problem with this is that this causes problems if the NMI handle either triggers an exception, or a breakpoint. Both the exception and the breakpoint handlers will finish with an iret. If this happens while in NMI context, the CPU will leave NMI context and a new NMI may come in. As NMI handlers are not made to be re-entrant, this can cause havoc with the system, not to mention, the nested NMI will write all over the previous NMI's stack. Linus Torvalds proposed the following workaround to this problem: https://lkml.org/lkml/2010/7/14/264 "In fact, I wonder if we couldn't just do a software NMI disable instead? Hav ea per-cpu variable (in the _core_ percpu areas that get allocated statically) that points to the NMI stack frame, and just make the NMI code itself do something like NMI entry: - load percpu NMI stack frame pointer - if non-zero we know we're nested, and should ignore this NMI: - we're returning to kernel mode, so return immediately by using "popf/ret", which also keeps NMI's disabled in the hardware until the "real" NMI iret happens. - before the popf/iret, use the NMI stack pointer to make the NMI return stack be invalid and cause a fault - set the NMI stack pointer to the current stack pointer NMI exit (not the above "immediate exit because we nested"): clear the percpu NMI stack pointer Just do the iret. Now, the thing is, now the "iret" is atomic. If we had a nested NMI, we'll take a fault, and that re-does our "delayed" NMI - and NMI's will stay masked. And if we didn't have a nested NMI, that iret will now unmask NMI's, and everything is happy." I first tried to follow this advice but as I started implementing this code, a few gotchas showed up. One, is accessing per-cpu variables in the NMI handler. The problem is that per-cpu variables use the %gs register to get the variable for the given CPU. But as the NMI may happen in userspace, we must first perform a SWAPGS to get to it. The NMI handler already does this later in the code, but its too late as we have saved off all the registers and we don't want to do that for a disabled NMI. Peter Zijlstra suggested to keep all variables on the stack. This simplifies things greatly and it has the added benefit of cache locality. Two, faulting on the iret. I really wanted to make this work, but it was becoming very hacky, and I never got it to be stable. The iret already had a fault handler for userspace faulting with bad segment registers, and getting NMI to trigger a fault and detect it was very tricky. But for strange reasons, the system would usually take a double fault and crash. I never figured out why and decided to go with a simple "jmp" approach. The new approach I took also simplified things. Finally, the last problem with Linus's approach was to have the nested NMI handler do a ret instead of an iret to give the first NMI NMI-context again. The problem is that ret is much more limited than an iret. I couldn't figure out how to get the stack back where it belonged. I could have copied the current stack, pushed the return onto it, but my fear here is that there may be some place that writes data below the stack pointer. I know that is not something code should depend on, but I don't want to chance it. I may add this feature later, but for now, an NMI handler that loses NMI context will not get it back. Here's what is done: When an NMI comes in, the HW pushes the interrupt stack frame onto the per cpu NMI stack that is selected by the IST. A special location on the NMI stack holds a variable that is set when the first NMI handler runs. If this variable is set then we know that this is a nested NMI and we process the nested NMI code. There is still a race when this variable is cleared and an NMI comes in just before the first NMI does the return. For this case, if the variable is cleared, we also check if the interrupted stack is the NMI stack. If it is, then we process the nested NMI code. Why the two tests and not just test the interrupted stack? If the first NMI hits a breakpoint and loses NMI context, and then it hits another breakpoint and while processing that breakpoint we get a nested NMI. When processing a breakpoint, the stack changes to the breakpoint stack. If another NMI comes in here we can't rely on the interrupted stack to be the NMI stack. If the variable is not set and the interrupted task's stack is not the NMI stack, then we know this is the first NMI and we can process things normally. But in order to do so, we need to do a few things first. 1) Set the stack variable that tells us that we are in an NMI handler 2) Make two copies of the interrupt stack frame. One copy is used to return on iret The other is used to restore the first one if we have a nested NMI. This is what the stack will look like: +-------------------------+ | original SS | | original Return RSP | | original RFLAGS | | original CS | | original RIP | +-------------------------+ | temp storage for rdx | +-------------------------+ | NMI executing variable | +-------------------------+ | Saved SS | | Saved Return RSP | | Saved RFLAGS | | Saved CS | | Saved RIP | +-------------------------+ | copied SS | | copied Return RSP | | copied RFLAGS | | copied CS | | copied RIP | +-------------------------+ | pt_regs | +-------------------------+ The original stack frame contains what the HW put in when we entered the NMI. We store %rdx as a temp variable to use. Both the original HW stack frame and this %rdx storage will be clobbered by nested NMIs so we can not rely on them later in the first NMI handler. The next item is the special stack variable that is set when we execute the rest of the NMI handler. Then we have two copies of the interrupt stack. The second copy is modified by any nested NMIs to let the first NMI know that we triggered a second NMI (latched) and that we should repeat the NMI handler. If the first NMI hits an exception or breakpoint that takes it out of NMI context, if a second NMI comes in before the first one finishes, it will update the copied interrupt stack to point to a fix up location to trigger another NMI. When the first NMI calls iret, it will instead jump to the fix up location. This fix up location will copy the saved interrupt stack back to the copy and execute the nmi handler again. Note, the nested NMI knows enough to check if it preempted a previous NMI handler while it is in the fixup location. If it has, it will not modify the copied interrupt stack and will just leave as if nothing happened. As the NMI handle is about to execute again, there's no reason to latch now. To test all this, I forced the NMI handler to call iret and take itself out of NMI context. I also added assemble code to write to the serial to make sure that it hits the nested path as well as the fix up path. Everything seems to be working fine. Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: H. Peter Anvin <hpa@linux.intel.com> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Paul Turner <pjt@google.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Mathieu Desnoyers <mathieu.desnoyers@efficios.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2011-12-08 21:36:23 +04:00
/* Make another copy, this one may be modified by nested NMIs */
addq $(10*8), %rsp
x86: Add workaround to NMI iret woes In x86, when an NMI goes off, the CPU goes into an NMI context that prevents other NMIs to trigger on that CPU. If an NMI is suppose to trigger, it has to wait till the previous NMI leaves NMI context. At that time, the next NMI can trigger (note, only one more NMI will trigger, as only one can be latched at a time). The way x86 gets out of NMI context is by calling iret. The problem with this is that this causes problems if the NMI handle either triggers an exception, or a breakpoint. Both the exception and the breakpoint handlers will finish with an iret. If this happens while in NMI context, the CPU will leave NMI context and a new NMI may come in. As NMI handlers are not made to be re-entrant, this can cause havoc with the system, not to mention, the nested NMI will write all over the previous NMI's stack. Linus Torvalds proposed the following workaround to this problem: https://lkml.org/lkml/2010/7/14/264 "In fact, I wonder if we couldn't just do a software NMI disable instead? Hav ea per-cpu variable (in the _core_ percpu areas that get allocated statically) that points to the NMI stack frame, and just make the NMI code itself do something like NMI entry: - load percpu NMI stack frame pointer - if non-zero we know we're nested, and should ignore this NMI: - we're returning to kernel mode, so return immediately by using "popf/ret", which also keeps NMI's disabled in the hardware until the "real" NMI iret happens. - before the popf/iret, use the NMI stack pointer to make the NMI return stack be invalid and cause a fault - set the NMI stack pointer to the current stack pointer NMI exit (not the above "immediate exit because we nested"): clear the percpu NMI stack pointer Just do the iret. Now, the thing is, now the "iret" is atomic. If we had a nested NMI, we'll take a fault, and that re-does our "delayed" NMI - and NMI's will stay masked. And if we didn't have a nested NMI, that iret will now unmask NMI's, and everything is happy." I first tried to follow this advice but as I started implementing this code, a few gotchas showed up. One, is accessing per-cpu variables in the NMI handler. The problem is that per-cpu variables use the %gs register to get the variable for the given CPU. But as the NMI may happen in userspace, we must first perform a SWAPGS to get to it. The NMI handler already does this later in the code, but its too late as we have saved off all the registers and we don't want to do that for a disabled NMI. Peter Zijlstra suggested to keep all variables on the stack. This simplifies things greatly and it has the added benefit of cache locality. Two, faulting on the iret. I really wanted to make this work, but it was becoming very hacky, and I never got it to be stable. The iret already had a fault handler for userspace faulting with bad segment registers, and getting NMI to trigger a fault and detect it was very tricky. But for strange reasons, the system would usually take a double fault and crash. I never figured out why and decided to go with a simple "jmp" approach. The new approach I took also simplified things. Finally, the last problem with Linus's approach was to have the nested NMI handler do a ret instead of an iret to give the first NMI NMI-context again. The problem is that ret is much more limited than an iret. I couldn't figure out how to get the stack back where it belonged. I could have copied the current stack, pushed the return onto it, but my fear here is that there may be some place that writes data below the stack pointer. I know that is not something code should depend on, but I don't want to chance it. I may add this feature later, but for now, an NMI handler that loses NMI context will not get it back. Here's what is done: When an NMI comes in, the HW pushes the interrupt stack frame onto the per cpu NMI stack that is selected by the IST. A special location on the NMI stack holds a variable that is set when the first NMI handler runs. If this variable is set then we know that this is a nested NMI and we process the nested NMI code. There is still a race when this variable is cleared and an NMI comes in just before the first NMI does the return. For this case, if the variable is cleared, we also check if the interrupted stack is the NMI stack. If it is, then we process the nested NMI code. Why the two tests and not just test the interrupted stack? If the first NMI hits a breakpoint and loses NMI context, and then it hits another breakpoint and while processing that breakpoint we get a nested NMI. When processing a breakpoint, the stack changes to the breakpoint stack. If another NMI comes in here we can't rely on the interrupted stack to be the NMI stack. If the variable is not set and the interrupted task's stack is not the NMI stack, then we know this is the first NMI and we can process things normally. But in order to do so, we need to do a few things first. 1) Set the stack variable that tells us that we are in an NMI handler 2) Make two copies of the interrupt stack frame. One copy is used to return on iret The other is used to restore the first one if we have a nested NMI. This is what the stack will look like: +-------------------------+ | original SS | | original Return RSP | | original RFLAGS | | original CS | | original RIP | +-------------------------+ | temp storage for rdx | +-------------------------+ | NMI executing variable | +-------------------------+ | Saved SS | | Saved Return RSP | | Saved RFLAGS | | Saved CS | | Saved RIP | +-------------------------+ | copied SS | | copied Return RSP | | copied RFLAGS | | copied CS | | copied RIP | +-------------------------+ | pt_regs | +-------------------------+ The original stack frame contains what the HW put in when we entered the NMI. We store %rdx as a temp variable to use. Both the original HW stack frame and this %rdx storage will be clobbered by nested NMIs so we can not rely on them later in the first NMI handler. The next item is the special stack variable that is set when we execute the rest of the NMI handler. Then we have two copies of the interrupt stack. The second copy is modified by any nested NMIs to let the first NMI know that we triggered a second NMI (latched) and that we should repeat the NMI handler. If the first NMI hits an exception or breakpoint that takes it out of NMI context, if a second NMI comes in before the first one finishes, it will update the copied interrupt stack to point to a fix up location to trigger another NMI. When the first NMI calls iret, it will instead jump to the fix up location. This fix up location will copy the saved interrupt stack back to the copy and execute the nmi handler again. Note, the nested NMI knows enough to check if it preempted a previous NMI handler while it is in the fixup location. If it has, it will not modify the copied interrupt stack and will just leave as if nothing happened. As the NMI handle is about to execute again, there's no reason to latch now. To test all this, I forced the NMI handler to call iret and take itself out of NMI context. I also added assemble code to write to the serial to make sure that it hits the nested path as well as the fix up path. Everything seems to be working fine. Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: H. Peter Anvin <hpa@linux.intel.com> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Paul Turner <pjt@google.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Mathieu Desnoyers <mathieu.desnoyers@efficios.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2011-12-08 21:36:23 +04:00
.rept 5
pushq -6*8(%rsp)
x86: Add workaround to NMI iret woes In x86, when an NMI goes off, the CPU goes into an NMI context that prevents other NMIs to trigger on that CPU. If an NMI is suppose to trigger, it has to wait till the previous NMI leaves NMI context. At that time, the next NMI can trigger (note, only one more NMI will trigger, as only one can be latched at a time). The way x86 gets out of NMI context is by calling iret. The problem with this is that this causes problems if the NMI handle either triggers an exception, or a breakpoint. Both the exception and the breakpoint handlers will finish with an iret. If this happens while in NMI context, the CPU will leave NMI context and a new NMI may come in. As NMI handlers are not made to be re-entrant, this can cause havoc with the system, not to mention, the nested NMI will write all over the previous NMI's stack. Linus Torvalds proposed the following workaround to this problem: https://lkml.org/lkml/2010/7/14/264 "In fact, I wonder if we couldn't just do a software NMI disable instead? Hav ea per-cpu variable (in the _core_ percpu areas that get allocated statically) that points to the NMI stack frame, and just make the NMI code itself do something like NMI entry: - load percpu NMI stack frame pointer - if non-zero we know we're nested, and should ignore this NMI: - we're returning to kernel mode, so return immediately by using "popf/ret", which also keeps NMI's disabled in the hardware until the "real" NMI iret happens. - before the popf/iret, use the NMI stack pointer to make the NMI return stack be invalid and cause a fault - set the NMI stack pointer to the current stack pointer NMI exit (not the above "immediate exit because we nested"): clear the percpu NMI stack pointer Just do the iret. Now, the thing is, now the "iret" is atomic. If we had a nested NMI, we'll take a fault, and that re-does our "delayed" NMI - and NMI's will stay masked. And if we didn't have a nested NMI, that iret will now unmask NMI's, and everything is happy." I first tried to follow this advice but as I started implementing this code, a few gotchas showed up. One, is accessing per-cpu variables in the NMI handler. The problem is that per-cpu variables use the %gs register to get the variable for the given CPU. But as the NMI may happen in userspace, we must first perform a SWAPGS to get to it. The NMI handler already does this later in the code, but its too late as we have saved off all the registers and we don't want to do that for a disabled NMI. Peter Zijlstra suggested to keep all variables on the stack. This simplifies things greatly and it has the added benefit of cache locality. Two, faulting on the iret. I really wanted to make this work, but it was becoming very hacky, and I never got it to be stable. The iret already had a fault handler for userspace faulting with bad segment registers, and getting NMI to trigger a fault and detect it was very tricky. But for strange reasons, the system would usually take a double fault and crash. I never figured out why and decided to go with a simple "jmp" approach. The new approach I took also simplified things. Finally, the last problem with Linus's approach was to have the nested NMI handler do a ret instead of an iret to give the first NMI NMI-context again. The problem is that ret is much more limited than an iret. I couldn't figure out how to get the stack back where it belonged. I could have copied the current stack, pushed the return onto it, but my fear here is that there may be some place that writes data below the stack pointer. I know that is not something code should depend on, but I don't want to chance it. I may add this feature later, but for now, an NMI handler that loses NMI context will not get it back. Here's what is done: When an NMI comes in, the HW pushes the interrupt stack frame onto the per cpu NMI stack that is selected by the IST. A special location on the NMI stack holds a variable that is set when the first NMI handler runs. If this variable is set then we know that this is a nested NMI and we process the nested NMI code. There is still a race when this variable is cleared and an NMI comes in just before the first NMI does the return. For this case, if the variable is cleared, we also check if the interrupted stack is the NMI stack. If it is, then we process the nested NMI code. Why the two tests and not just test the interrupted stack? If the first NMI hits a breakpoint and loses NMI context, and then it hits another breakpoint and while processing that breakpoint we get a nested NMI. When processing a breakpoint, the stack changes to the breakpoint stack. If another NMI comes in here we can't rely on the interrupted stack to be the NMI stack. If the variable is not set and the interrupted task's stack is not the NMI stack, then we know this is the first NMI and we can process things normally. But in order to do so, we need to do a few things first. 1) Set the stack variable that tells us that we are in an NMI handler 2) Make two copies of the interrupt stack frame. One copy is used to return on iret The other is used to restore the first one if we have a nested NMI. This is what the stack will look like: +-------------------------+ | original SS | | original Return RSP | | original RFLAGS | | original CS | | original RIP | +-------------------------+ | temp storage for rdx | +-------------------------+ | NMI executing variable | +-------------------------+ | Saved SS | | Saved Return RSP | | Saved RFLAGS | | Saved CS | | Saved RIP | +-------------------------+ | copied SS | | copied Return RSP | | copied RFLAGS | | copied CS | | copied RIP | +-------------------------+ | pt_regs | +-------------------------+ The original stack frame contains what the HW put in when we entered the NMI. We store %rdx as a temp variable to use. Both the original HW stack frame and this %rdx storage will be clobbered by nested NMIs so we can not rely on them later in the first NMI handler. The next item is the special stack variable that is set when we execute the rest of the NMI handler. Then we have two copies of the interrupt stack. The second copy is modified by any nested NMIs to let the first NMI know that we triggered a second NMI (latched) and that we should repeat the NMI handler. If the first NMI hits an exception or breakpoint that takes it out of NMI context, if a second NMI comes in before the first one finishes, it will update the copied interrupt stack to point to a fix up location to trigger another NMI. When the first NMI calls iret, it will instead jump to the fix up location. This fix up location will copy the saved interrupt stack back to the copy and execute the nmi handler again. Note, the nested NMI knows enough to check if it preempted a previous NMI handler while it is in the fixup location. If it has, it will not modify the copied interrupt stack and will just leave as if nothing happened. As the NMI handle is about to execute again, there's no reason to latch now. To test all this, I forced the NMI handler to call iret and take itself out of NMI context. I also added assemble code to write to the serial to make sure that it hits the nested path as well as the fix up path. Everything seems to be working fine. Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: H. Peter Anvin <hpa@linux.intel.com> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Paul Turner <pjt@google.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Mathieu Desnoyers <mathieu.desnoyers@efficios.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2011-12-08 21:36:23 +04:00
.endr
subq $(5*8), %rsp
end_repeat_nmi:
x86: Add workaround to NMI iret woes In x86, when an NMI goes off, the CPU goes into an NMI context that prevents other NMIs to trigger on that CPU. If an NMI is suppose to trigger, it has to wait till the previous NMI leaves NMI context. At that time, the next NMI can trigger (note, only one more NMI will trigger, as only one can be latched at a time). The way x86 gets out of NMI context is by calling iret. The problem with this is that this causes problems if the NMI handle either triggers an exception, or a breakpoint. Both the exception and the breakpoint handlers will finish with an iret. If this happens while in NMI context, the CPU will leave NMI context and a new NMI may come in. As NMI handlers are not made to be re-entrant, this can cause havoc with the system, not to mention, the nested NMI will write all over the previous NMI's stack. Linus Torvalds proposed the following workaround to this problem: https://lkml.org/lkml/2010/7/14/264 "In fact, I wonder if we couldn't just do a software NMI disable instead? Hav ea per-cpu variable (in the _core_ percpu areas that get allocated statically) that points to the NMI stack frame, and just make the NMI code itself do something like NMI entry: - load percpu NMI stack frame pointer - if non-zero we know we're nested, and should ignore this NMI: - we're returning to kernel mode, so return immediately by using "popf/ret", which also keeps NMI's disabled in the hardware until the "real" NMI iret happens. - before the popf/iret, use the NMI stack pointer to make the NMI return stack be invalid and cause a fault - set the NMI stack pointer to the current stack pointer NMI exit (not the above "immediate exit because we nested"): clear the percpu NMI stack pointer Just do the iret. Now, the thing is, now the "iret" is atomic. If we had a nested NMI, we'll take a fault, and that re-does our "delayed" NMI - and NMI's will stay masked. And if we didn't have a nested NMI, that iret will now unmask NMI's, and everything is happy." I first tried to follow this advice but as I started implementing this code, a few gotchas showed up. One, is accessing per-cpu variables in the NMI handler. The problem is that per-cpu variables use the %gs register to get the variable for the given CPU. But as the NMI may happen in userspace, we must first perform a SWAPGS to get to it. The NMI handler already does this later in the code, but its too late as we have saved off all the registers and we don't want to do that for a disabled NMI. Peter Zijlstra suggested to keep all variables on the stack. This simplifies things greatly and it has the added benefit of cache locality. Two, faulting on the iret. I really wanted to make this work, but it was becoming very hacky, and I never got it to be stable. The iret already had a fault handler for userspace faulting with bad segment registers, and getting NMI to trigger a fault and detect it was very tricky. But for strange reasons, the system would usually take a double fault and crash. I never figured out why and decided to go with a simple "jmp" approach. The new approach I took also simplified things. Finally, the last problem with Linus's approach was to have the nested NMI handler do a ret instead of an iret to give the first NMI NMI-context again. The problem is that ret is much more limited than an iret. I couldn't figure out how to get the stack back where it belonged. I could have copied the current stack, pushed the return onto it, but my fear here is that there may be some place that writes data below the stack pointer. I know that is not something code should depend on, but I don't want to chance it. I may add this feature later, but for now, an NMI handler that loses NMI context will not get it back. Here's what is done: When an NMI comes in, the HW pushes the interrupt stack frame onto the per cpu NMI stack that is selected by the IST. A special location on the NMI stack holds a variable that is set when the first NMI handler runs. If this variable is set then we know that this is a nested NMI and we process the nested NMI code. There is still a race when this variable is cleared and an NMI comes in just before the first NMI does the return. For this case, if the variable is cleared, we also check if the interrupted stack is the NMI stack. If it is, then we process the nested NMI code. Why the two tests and not just test the interrupted stack? If the first NMI hits a breakpoint and loses NMI context, and then it hits another breakpoint and while processing that breakpoint we get a nested NMI. When processing a breakpoint, the stack changes to the breakpoint stack. If another NMI comes in here we can't rely on the interrupted stack to be the NMI stack. If the variable is not set and the interrupted task's stack is not the NMI stack, then we know this is the first NMI and we can process things normally. But in order to do so, we need to do a few things first. 1) Set the stack variable that tells us that we are in an NMI handler 2) Make two copies of the interrupt stack frame. One copy is used to return on iret The other is used to restore the first one if we have a nested NMI. This is what the stack will look like: +-------------------------+ | original SS | | original Return RSP | | original RFLAGS | | original CS | | original RIP | +-------------------------+ | temp storage for rdx | +-------------------------+ | NMI executing variable | +-------------------------+ | Saved SS | | Saved Return RSP | | Saved RFLAGS | | Saved CS | | Saved RIP | +-------------------------+ | copied SS | | copied Return RSP | | copied RFLAGS | | copied CS | | copied RIP | +-------------------------+ | pt_regs | +-------------------------+ The original stack frame contains what the HW put in when we entered the NMI. We store %rdx as a temp variable to use. Both the original HW stack frame and this %rdx storage will be clobbered by nested NMIs so we can not rely on them later in the first NMI handler. The next item is the special stack variable that is set when we execute the rest of the NMI handler. Then we have two copies of the interrupt stack. The second copy is modified by any nested NMIs to let the first NMI know that we triggered a second NMI (latched) and that we should repeat the NMI handler. If the first NMI hits an exception or breakpoint that takes it out of NMI context, if a second NMI comes in before the first one finishes, it will update the copied interrupt stack to point to a fix up location to trigger another NMI. When the first NMI calls iret, it will instead jump to the fix up location. This fix up location will copy the saved interrupt stack back to the copy and execute the nmi handler again. Note, the nested NMI knows enough to check if it preempted a previous NMI handler while it is in the fixup location. If it has, it will not modify the copied interrupt stack and will just leave as if nothing happened. As the NMI handle is about to execute again, there's no reason to latch now. To test all this, I forced the NMI handler to call iret and take itself out of NMI context. I also added assemble code to write to the serial to make sure that it hits the nested path as well as the fix up path. Everything seems to be working fine. Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: H. Peter Anvin <hpa@linux.intel.com> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Paul Turner <pjt@google.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Mathieu Desnoyers <mathieu.desnoyers@efficios.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2011-12-08 21:36:23 +04:00
/*
* Everything below this point can be preempted by a nested
* NMI if the first NMI took an exception and reset our iret stack
* so that we repeat another NMI.
x86: Add workaround to NMI iret woes In x86, when an NMI goes off, the CPU goes into an NMI context that prevents other NMIs to trigger on that CPU. If an NMI is suppose to trigger, it has to wait till the previous NMI leaves NMI context. At that time, the next NMI can trigger (note, only one more NMI will trigger, as only one can be latched at a time). The way x86 gets out of NMI context is by calling iret. The problem with this is that this causes problems if the NMI handle either triggers an exception, or a breakpoint. Both the exception and the breakpoint handlers will finish with an iret. If this happens while in NMI context, the CPU will leave NMI context and a new NMI may come in. As NMI handlers are not made to be re-entrant, this can cause havoc with the system, not to mention, the nested NMI will write all over the previous NMI's stack. Linus Torvalds proposed the following workaround to this problem: https://lkml.org/lkml/2010/7/14/264 "In fact, I wonder if we couldn't just do a software NMI disable instead? Hav ea per-cpu variable (in the _core_ percpu areas that get allocated statically) that points to the NMI stack frame, and just make the NMI code itself do something like NMI entry: - load percpu NMI stack frame pointer - if non-zero we know we're nested, and should ignore this NMI: - we're returning to kernel mode, so return immediately by using "popf/ret", which also keeps NMI's disabled in the hardware until the "real" NMI iret happens. - before the popf/iret, use the NMI stack pointer to make the NMI return stack be invalid and cause a fault - set the NMI stack pointer to the current stack pointer NMI exit (not the above "immediate exit because we nested"): clear the percpu NMI stack pointer Just do the iret. Now, the thing is, now the "iret" is atomic. If we had a nested NMI, we'll take a fault, and that re-does our "delayed" NMI - and NMI's will stay masked. And if we didn't have a nested NMI, that iret will now unmask NMI's, and everything is happy." I first tried to follow this advice but as I started implementing this code, a few gotchas showed up. One, is accessing per-cpu variables in the NMI handler. The problem is that per-cpu variables use the %gs register to get the variable for the given CPU. But as the NMI may happen in userspace, we must first perform a SWAPGS to get to it. The NMI handler already does this later in the code, but its too late as we have saved off all the registers and we don't want to do that for a disabled NMI. Peter Zijlstra suggested to keep all variables on the stack. This simplifies things greatly and it has the added benefit of cache locality. Two, faulting on the iret. I really wanted to make this work, but it was becoming very hacky, and I never got it to be stable. The iret already had a fault handler for userspace faulting with bad segment registers, and getting NMI to trigger a fault and detect it was very tricky. But for strange reasons, the system would usually take a double fault and crash. I never figured out why and decided to go with a simple "jmp" approach. The new approach I took also simplified things. Finally, the last problem with Linus's approach was to have the nested NMI handler do a ret instead of an iret to give the first NMI NMI-context again. The problem is that ret is much more limited than an iret. I couldn't figure out how to get the stack back where it belonged. I could have copied the current stack, pushed the return onto it, but my fear here is that there may be some place that writes data below the stack pointer. I know that is not something code should depend on, but I don't want to chance it. I may add this feature later, but for now, an NMI handler that loses NMI context will not get it back. Here's what is done: When an NMI comes in, the HW pushes the interrupt stack frame onto the per cpu NMI stack that is selected by the IST. A special location on the NMI stack holds a variable that is set when the first NMI handler runs. If this variable is set then we know that this is a nested NMI and we process the nested NMI code. There is still a race when this variable is cleared and an NMI comes in just before the first NMI does the return. For this case, if the variable is cleared, we also check if the interrupted stack is the NMI stack. If it is, then we process the nested NMI code. Why the two tests and not just test the interrupted stack? If the first NMI hits a breakpoint and loses NMI context, and then it hits another breakpoint and while processing that breakpoint we get a nested NMI. When processing a breakpoint, the stack changes to the breakpoint stack. If another NMI comes in here we can't rely on the interrupted stack to be the NMI stack. If the variable is not set and the interrupted task's stack is not the NMI stack, then we know this is the first NMI and we can process things normally. But in order to do so, we need to do a few things first. 1) Set the stack variable that tells us that we are in an NMI handler 2) Make two copies of the interrupt stack frame. One copy is used to return on iret The other is used to restore the first one if we have a nested NMI. This is what the stack will look like: +-------------------------+ | original SS | | original Return RSP | | original RFLAGS | | original CS | | original RIP | +-------------------------+ | temp storage for rdx | +-------------------------+ | NMI executing variable | +-------------------------+ | Saved SS | | Saved Return RSP | | Saved RFLAGS | | Saved CS | | Saved RIP | +-------------------------+ | copied SS | | copied Return RSP | | copied RFLAGS | | copied CS | | copied RIP | +-------------------------+ | pt_regs | +-------------------------+ The original stack frame contains what the HW put in when we entered the NMI. We store %rdx as a temp variable to use. Both the original HW stack frame and this %rdx storage will be clobbered by nested NMIs so we can not rely on them later in the first NMI handler. The next item is the special stack variable that is set when we execute the rest of the NMI handler. Then we have two copies of the interrupt stack. The second copy is modified by any nested NMIs to let the first NMI know that we triggered a second NMI (latched) and that we should repeat the NMI handler. If the first NMI hits an exception or breakpoint that takes it out of NMI context, if a second NMI comes in before the first one finishes, it will update the copied interrupt stack to point to a fix up location to trigger another NMI. When the first NMI calls iret, it will instead jump to the fix up location. This fix up location will copy the saved interrupt stack back to the copy and execute the nmi handler again. Note, the nested NMI knows enough to check if it preempted a previous NMI handler while it is in the fixup location. If it has, it will not modify the copied interrupt stack and will just leave as if nothing happened. As the NMI handle is about to execute again, there's no reason to latch now. To test all this, I forced the NMI handler to call iret and take itself out of NMI context. I also added assemble code to write to the serial to make sure that it hits the nested path as well as the fix up path. Everything seems to be working fine. Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: H. Peter Anvin <hpa@linux.intel.com> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Paul Turner <pjt@google.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Mathieu Desnoyers <mathieu.desnoyers@efficios.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2011-12-08 21:36:23 +04:00
*/
pushq $-1 /* ORIG_RAX: no syscall to restart */
x86/asm/entry/64: Always allocate a complete "struct pt_regs" on the kernel stack The 64-bit entry code was using six stack slots less by not saving/restoring registers which are callee-preserved according to the C ABI, and was not allocating space for them. Only when syscalls needed a complete "struct pt_regs" was the complete area allocated and filled in. As an additional twist, on interrupt entry a "slightly less truncated pt_regs" trick is used, to make nested interrupt stacks easier to unwind. This proved to be a source of significant obfuscation and subtle bugs. For example, 'stub_fork' had to pop the return address, extend the struct, save registers, and push return address back. Ugly. 'ia32_ptregs_common' pops return address and "returns" via jmp insn, throwing a wrench into CPU return stack cache. This patch changes the code to always allocate a complete "struct pt_regs" on the kernel stack. The saving of registers is still done lazily. "Partial pt_regs" trick on interrupt stack is retained. Macros which manipulate "struct pt_regs" on stack are reworked: - ALLOC_PT_GPREGS_ON_STACK allocates the structure. - SAVE_C_REGS saves to it those registers which are clobbered by C code. - SAVE_EXTRA_REGS saves to it all other registers. - Corresponding RESTORE_* and REMOVE_PT_GPREGS_FROM_STACK macros reverse it. 'ia32_ptregs_common', 'stub_fork' and friends lost their ugly dance with the return pointer. LOAD_ARGS32 in ia32entry.S now uses symbolic stack offsets instead of magic numbers. 'error_entry' and 'save_paranoid' now use SAVE_C_REGS + SAVE_EXTRA_REGS instead of having it open-coded yet again. Patch was run-tested: 64-bit executables, 32-bit executables, strace works. Timing tests did not show measurable difference in 32-bit and 64-bit syscalls. Signed-off-by: Denys Vlasenko <dvlasenk@redhat.com> Signed-off-by: Andy Lutomirski <luto@amacapital.net> Cc: Alexei Starovoitov <ast@plumgrid.com> Cc: Borislav Petkov <bp@alien8.de> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: H. Peter Anvin <hpa@zytor.com> Cc: Kees Cook <keescook@chromium.org> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Will Drewry <wad@chromium.org> Link: http://lkml.kernel.org/r/1423778052-21038-2-git-send-email-dvlasenk@redhat.com Link: http://lkml.kernel.org/r/b89763d354aa23e670b9bdf3a40ae320320a7c2e.1424989793.git.luto@amacapital.net Signed-off-by: Ingo Molnar <mingo@kernel.org>
2015-02-27 01:40:27 +03:00
ALLOC_PT_GPREGS_ON_STACK
/*
* Use paranoid_entry to handle SWAPGS, but no need to use paranoid_exit
* as we should not be calling schedule in NMI context.
* Even with normal interrupts enabled. An NMI should not be
* setting NEED_RESCHED or anything that normal interrupts and
* exceptions might do.
*/
call paranoid_entry
/*
* Save off the CR2 register. If we take a page fault in the NMI then
* it could corrupt the CR2 value. If the NMI preempts a page fault
* handler before it was able to read the CR2 register, and then the
* NMI itself takes a page fault, the page fault that was preempted
* will read the information from the NMI page fault and not the
* origin fault. Save it off and restore it if it changes.
* Use the r12 callee-saved register.
*/
movq %cr2, %r12
/* paranoidentry do_nmi, 0; without TRACE_IRQS_OFF */
movq %rsp, %rdi
movq $-1, %rsi
call do_nmi
/* Did the NMI take a page fault? Restore cr2 if it did */
movq %cr2, %rcx
cmpq %rcx, %r12
je 1f
movq %r12, %cr2
1:
testl %ebx, %ebx /* swapgs needed? */
jnz nmi_restore
nmi_swapgs:
SWAPGS_UNSAFE_STACK
nmi_restore:
x86/asm/entry/64: Always allocate a complete "struct pt_regs" on the kernel stack The 64-bit entry code was using six stack slots less by not saving/restoring registers which are callee-preserved according to the C ABI, and was not allocating space for them. Only when syscalls needed a complete "struct pt_regs" was the complete area allocated and filled in. As an additional twist, on interrupt entry a "slightly less truncated pt_regs" trick is used, to make nested interrupt stacks easier to unwind. This proved to be a source of significant obfuscation and subtle bugs. For example, 'stub_fork' had to pop the return address, extend the struct, save registers, and push return address back. Ugly. 'ia32_ptregs_common' pops return address and "returns" via jmp insn, throwing a wrench into CPU return stack cache. This patch changes the code to always allocate a complete "struct pt_regs" on the kernel stack. The saving of registers is still done lazily. "Partial pt_regs" trick on interrupt stack is retained. Macros which manipulate "struct pt_regs" on stack are reworked: - ALLOC_PT_GPREGS_ON_STACK allocates the structure. - SAVE_C_REGS saves to it those registers which are clobbered by C code. - SAVE_EXTRA_REGS saves to it all other registers. - Corresponding RESTORE_* and REMOVE_PT_GPREGS_FROM_STACK macros reverse it. 'ia32_ptregs_common', 'stub_fork' and friends lost their ugly dance with the return pointer. LOAD_ARGS32 in ia32entry.S now uses symbolic stack offsets instead of magic numbers. 'error_entry' and 'save_paranoid' now use SAVE_C_REGS + SAVE_EXTRA_REGS instead of having it open-coded yet again. Patch was run-tested: 64-bit executables, 32-bit executables, strace works. Timing tests did not show measurable difference in 32-bit and 64-bit syscalls. Signed-off-by: Denys Vlasenko <dvlasenk@redhat.com> Signed-off-by: Andy Lutomirski <luto@amacapital.net> Cc: Alexei Starovoitov <ast@plumgrid.com> Cc: Borislav Petkov <bp@alien8.de> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: H. Peter Anvin <hpa@zytor.com> Cc: Kees Cook <keescook@chromium.org> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Will Drewry <wad@chromium.org> Link: http://lkml.kernel.org/r/1423778052-21038-2-git-send-email-dvlasenk@redhat.com Link: http://lkml.kernel.org/r/b89763d354aa23e670b9bdf3a40ae320320a7c2e.1424989793.git.luto@amacapital.net Signed-off-by: Ingo Molnar <mingo@kernel.org>
2015-02-27 01:40:27 +03:00
RESTORE_EXTRA_REGS
RESTORE_C_REGS
/* Pop the extra iret frame at once */
x86/asm/entry/64: Always allocate a complete "struct pt_regs" on the kernel stack The 64-bit entry code was using six stack slots less by not saving/restoring registers which are callee-preserved according to the C ABI, and was not allocating space for them. Only when syscalls needed a complete "struct pt_regs" was the complete area allocated and filled in. As an additional twist, on interrupt entry a "slightly less truncated pt_regs" trick is used, to make nested interrupt stacks easier to unwind. This proved to be a source of significant obfuscation and subtle bugs. For example, 'stub_fork' had to pop the return address, extend the struct, save registers, and push return address back. Ugly. 'ia32_ptregs_common' pops return address and "returns" via jmp insn, throwing a wrench into CPU return stack cache. This patch changes the code to always allocate a complete "struct pt_regs" on the kernel stack. The saving of registers is still done lazily. "Partial pt_regs" trick on interrupt stack is retained. Macros which manipulate "struct pt_regs" on stack are reworked: - ALLOC_PT_GPREGS_ON_STACK allocates the structure. - SAVE_C_REGS saves to it those registers which are clobbered by C code. - SAVE_EXTRA_REGS saves to it all other registers. - Corresponding RESTORE_* and REMOVE_PT_GPREGS_FROM_STACK macros reverse it. 'ia32_ptregs_common', 'stub_fork' and friends lost their ugly dance with the return pointer. LOAD_ARGS32 in ia32entry.S now uses symbolic stack offsets instead of magic numbers. 'error_entry' and 'save_paranoid' now use SAVE_C_REGS + SAVE_EXTRA_REGS instead of having it open-coded yet again. Patch was run-tested: 64-bit executables, 32-bit executables, strace works. Timing tests did not show measurable difference in 32-bit and 64-bit syscalls. Signed-off-by: Denys Vlasenko <dvlasenk@redhat.com> Signed-off-by: Andy Lutomirski <luto@amacapital.net> Cc: Alexei Starovoitov <ast@plumgrid.com> Cc: Borislav Petkov <bp@alien8.de> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: H. Peter Anvin <hpa@zytor.com> Cc: Kees Cook <keescook@chromium.org> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Will Drewry <wad@chromium.org> Link: http://lkml.kernel.org/r/1423778052-21038-2-git-send-email-dvlasenk@redhat.com Link: http://lkml.kernel.org/r/b89763d354aa23e670b9bdf3a40ae320320a7c2e.1424989793.git.luto@amacapital.net Signed-off-by: Ingo Molnar <mingo@kernel.org>
2015-02-27 01:40:27 +03:00
REMOVE_PT_GPREGS_FROM_STACK 6*8
x86: Add workaround to NMI iret woes In x86, when an NMI goes off, the CPU goes into an NMI context that prevents other NMIs to trigger on that CPU. If an NMI is suppose to trigger, it has to wait till the previous NMI leaves NMI context. At that time, the next NMI can trigger (note, only one more NMI will trigger, as only one can be latched at a time). The way x86 gets out of NMI context is by calling iret. The problem with this is that this causes problems if the NMI handle either triggers an exception, or a breakpoint. Both the exception and the breakpoint handlers will finish with an iret. If this happens while in NMI context, the CPU will leave NMI context and a new NMI may come in. As NMI handlers are not made to be re-entrant, this can cause havoc with the system, not to mention, the nested NMI will write all over the previous NMI's stack. Linus Torvalds proposed the following workaround to this problem: https://lkml.org/lkml/2010/7/14/264 "In fact, I wonder if we couldn't just do a software NMI disable instead? Hav ea per-cpu variable (in the _core_ percpu areas that get allocated statically) that points to the NMI stack frame, and just make the NMI code itself do something like NMI entry: - load percpu NMI stack frame pointer - if non-zero we know we're nested, and should ignore this NMI: - we're returning to kernel mode, so return immediately by using "popf/ret", which also keeps NMI's disabled in the hardware until the "real" NMI iret happens. - before the popf/iret, use the NMI stack pointer to make the NMI return stack be invalid and cause a fault - set the NMI stack pointer to the current stack pointer NMI exit (not the above "immediate exit because we nested"): clear the percpu NMI stack pointer Just do the iret. Now, the thing is, now the "iret" is atomic. If we had a nested NMI, we'll take a fault, and that re-does our "delayed" NMI - and NMI's will stay masked. And if we didn't have a nested NMI, that iret will now unmask NMI's, and everything is happy." I first tried to follow this advice but as I started implementing this code, a few gotchas showed up. One, is accessing per-cpu variables in the NMI handler. The problem is that per-cpu variables use the %gs register to get the variable for the given CPU. But as the NMI may happen in userspace, we must first perform a SWAPGS to get to it. The NMI handler already does this later in the code, but its too late as we have saved off all the registers and we don't want to do that for a disabled NMI. Peter Zijlstra suggested to keep all variables on the stack. This simplifies things greatly and it has the added benefit of cache locality. Two, faulting on the iret. I really wanted to make this work, but it was becoming very hacky, and I never got it to be stable. The iret already had a fault handler for userspace faulting with bad segment registers, and getting NMI to trigger a fault and detect it was very tricky. But for strange reasons, the system would usually take a double fault and crash. I never figured out why and decided to go with a simple "jmp" approach. The new approach I took also simplified things. Finally, the last problem with Linus's approach was to have the nested NMI handler do a ret instead of an iret to give the first NMI NMI-context again. The problem is that ret is much more limited than an iret. I couldn't figure out how to get the stack back where it belonged. I could have copied the current stack, pushed the return onto it, but my fear here is that there may be some place that writes data below the stack pointer. I know that is not something code should depend on, but I don't want to chance it. I may add this feature later, but for now, an NMI handler that loses NMI context will not get it back. Here's what is done: When an NMI comes in, the HW pushes the interrupt stack frame onto the per cpu NMI stack that is selected by the IST. A special location on the NMI stack holds a variable that is set when the first NMI handler runs. If this variable is set then we know that this is a nested NMI and we process the nested NMI code. There is still a race when this variable is cleared and an NMI comes in just before the first NMI does the return. For this case, if the variable is cleared, we also check if the interrupted stack is the NMI stack. If it is, then we process the nested NMI code. Why the two tests and not just test the interrupted stack? If the first NMI hits a breakpoint and loses NMI context, and then it hits another breakpoint and while processing that breakpoint we get a nested NMI. When processing a breakpoint, the stack changes to the breakpoint stack. If another NMI comes in here we can't rely on the interrupted stack to be the NMI stack. If the variable is not set and the interrupted task's stack is not the NMI stack, then we know this is the first NMI and we can process things normally. But in order to do so, we need to do a few things first. 1) Set the stack variable that tells us that we are in an NMI handler 2) Make two copies of the interrupt stack frame. One copy is used to return on iret The other is used to restore the first one if we have a nested NMI. This is what the stack will look like: +-------------------------+ | original SS | | original Return RSP | | original RFLAGS | | original CS | | original RIP | +-------------------------+ | temp storage for rdx | +-------------------------+ | NMI executing variable | +-------------------------+ | Saved SS | | Saved Return RSP | | Saved RFLAGS | | Saved CS | | Saved RIP | +-------------------------+ | copied SS | | copied Return RSP | | copied RFLAGS | | copied CS | | copied RIP | +-------------------------+ | pt_regs | +-------------------------+ The original stack frame contains what the HW put in when we entered the NMI. We store %rdx as a temp variable to use. Both the original HW stack frame and this %rdx storage will be clobbered by nested NMIs so we can not rely on them later in the first NMI handler. The next item is the special stack variable that is set when we execute the rest of the NMI handler. Then we have two copies of the interrupt stack. The second copy is modified by any nested NMIs to let the first NMI know that we triggered a second NMI (latched) and that we should repeat the NMI handler. If the first NMI hits an exception or breakpoint that takes it out of NMI context, if a second NMI comes in before the first one finishes, it will update the copied interrupt stack to point to a fix up location to trigger another NMI. When the first NMI calls iret, it will instead jump to the fix up location. This fix up location will copy the saved interrupt stack back to the copy and execute the nmi handler again. Note, the nested NMI knows enough to check if it preempted a previous NMI handler while it is in the fixup location. If it has, it will not modify the copied interrupt stack and will just leave as if nothing happened. As the NMI handle is about to execute again, there's no reason to latch now. To test all this, I forced the NMI handler to call iret and take itself out of NMI context. I also added assemble code to write to the serial to make sure that it hits the nested path as well as the fix up path. Everything seems to be working fine. Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: H. Peter Anvin <hpa@linux.intel.com> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Paul Turner <pjt@google.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Mathieu Desnoyers <mathieu.desnoyers@efficios.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2011-12-08 21:36:23 +04:00
/* Clear the NMI executing stack variable */
movq $0, 5*8(%rsp)
INTERRUPT_RETURN
END(nmi)
ENTRY(ignore_sysret)
mov $-ENOSYS, %eax
sysret
END(ignore_sysret)