diff options
Diffstat (limited to 'arch/x86/kernel/nmi.c')
| -rw-r--r-- | arch/x86/kernel/nmi.c | 119 |
1 files changed, 48 insertions, 71 deletions
diff --git a/arch/x86/kernel/nmi.c b/arch/x86/kernel/nmi.c index 54c21d6abd5a..cec0bfa3bc04 100644 --- a/arch/x86/kernel/nmi.c +++ b/arch/x86/kernel/nmi.c @@ -25,10 +25,6 @@ #include <linux/atomic.h> #include <linux/sched/clock.h> -#if defined(CONFIG_EDAC) -#include <linux/edac.h> -#endif - #include <asm/cpu_entry_area.h> #include <asm/traps.h> #include <asm/mach_traps.h> @@ -37,6 +33,7 @@ #include <asm/reboot.h> #include <asm/cache.h> #include <asm/nospec-branch.h> +#include <asm/sev.h> #define CREATE_TRACE_POINTS #include <trace/events/nmi.h> @@ -106,7 +103,6 @@ fs_initcall(nmi_warning_debugfs); static void nmi_check_duration(struct nmiaction *action, u64 duration) { - u64 whole_msecs = READ_ONCE(action->max_duration); int remainder_ns, decimal_msecs; if (duration < nmi_longest_ns || duration < action->max_duration) @@ -114,12 +110,12 @@ static void nmi_check_duration(struct nmiaction *action, u64 duration) action->max_duration = duration; - remainder_ns = do_div(whole_msecs, (1000 * 1000)); + remainder_ns = do_div(duration, (1000 * 1000)); decimal_msecs = remainder_ns / 1000; printk_ratelimited(KERN_INFO "INFO: NMI handler (%ps) took too long to run: %lld.%03d msecs\n", - action->handler, whole_msecs, decimal_msecs); + action->handler, duration, decimal_msecs); } static int nmi_handle(unsigned int type, struct pt_regs *regs) @@ -161,7 +157,7 @@ int __register_nmi_handler(unsigned int type, struct nmiaction *action) struct nmi_desc *desc = nmi_to_desc(type); unsigned long flags; - if (!action->handler) + if (WARN_ON_ONCE(!action->handler || !list_empty(&action->list))) return -EINVAL; raw_spin_lock_irqsave(&desc->lock, flags); @@ -181,7 +177,7 @@ int __register_nmi_handler(unsigned int type, struct nmiaction *action) list_add_rcu(&action->list, &desc->head); else list_add_tail_rcu(&action->list, &desc->head); - + raw_spin_unlock_irqrestore(&desc->lock, flags); return 0; } @@ -190,7 +186,7 @@ EXPORT_SYMBOL(__register_nmi_handler); void unregister_nmi_handler(unsigned int type, const char *name) { struct nmi_desc *desc = nmi_to_desc(type); - struct nmiaction *n; + struct nmiaction *n, *found = NULL; unsigned long flags; raw_spin_lock_irqsave(&desc->lock, flags); @@ -204,12 +200,16 @@ void unregister_nmi_handler(unsigned int type, const char *name) WARN(in_nmi(), "Trying to free NMI (%s) from NMI context!\n", n->name); list_del_rcu(&n->list); + found = n; break; } } raw_spin_unlock_irqrestore(&desc->lock, flags); - synchronize_rcu(); + if (found) { + synchronize_rcu(); + INIT_LIST_HEAD(&found->list); + } } EXPORT_SYMBOL_GPL(unregister_nmi_handler); @@ -296,7 +296,6 @@ unknown_nmi_error(unsigned char reason, struct pt_regs *regs) pr_emerg("Uhhuh. NMI received for unknown reason %02x on CPU %d.\n", reason, smp_processor_id()); - pr_emerg("Do you have a strange power saving mode enabled?\n"); if (unknown_nmi_panic || panic_on_unrecovered_nmi) nmi_panic(regs, "NMI: Not continuing"); @@ -307,7 +306,7 @@ NOKPROBE_SYMBOL(unknown_nmi_error); static DEFINE_PER_CPU(bool, swallow_nmi); static DEFINE_PER_CPU(unsigned long, last_nmi_rip); -static void default_do_nmi(struct pt_regs *regs) +static noinstr void default_do_nmi(struct pt_regs *regs) { unsigned char reason = 0; int handled; @@ -333,6 +332,8 @@ static void default_do_nmi(struct pt_regs *regs) __this_cpu_write(last_nmi_rip, regs->ip); + instrumentation_begin(); + handled = nmi_handle(NMI_LOCAL, regs); __this_cpu_add(nmi_stats.normal, handled); if (handled) { @@ -346,7 +347,7 @@ static void default_do_nmi(struct pt_regs *regs) */ if (handled > 1) __this_cpu_write(swallow_nmi, true); - return; + goto out; } /* @@ -378,7 +379,7 @@ static void default_do_nmi(struct pt_regs *regs) #endif __this_cpu_add(nmi_stats.external, 1); raw_spin_unlock(&nmi_reason_lock); - return; + goto out; } raw_spin_unlock(&nmi_reason_lock); @@ -403,9 +404,9 @@ static void default_do_nmi(struct pt_regs *regs) * a 'real' unknown NMI. For example, while processing * a perf NMI another perf NMI comes in along with a * 'real' unknown NMI. These two NMIs get combined into - * one (as descibed above). When the next NMI gets + * one (as described above). When the next NMI gets * processed, it will be flagged by perf as handled, but - * noone will know that there was a 'real' unknown NMI sent + * no one will know that there was a 'real' unknown NMI sent * also. As a result it gets swallowed. Or if the first * perf NMI returns two events handled then the second * NMI will get eaten by the logic below, again losing a @@ -416,8 +417,10 @@ static void default_do_nmi(struct pt_regs *regs) __this_cpu_add(nmi_stats.swallow, 1); else unknown_nmi_error(reason, regs); + +out: + instrumentation_end(); } -NOKPROBE_SYMBOL(default_do_nmi); /* * NMIs can page fault or hit breakpoints which will cause it to lose @@ -471,46 +474,19 @@ enum nmi_states { }; static DEFINE_PER_CPU(enum nmi_states, nmi_state); static DEFINE_PER_CPU(unsigned long, nmi_cr2); +static DEFINE_PER_CPU(unsigned long, nmi_dr7); -#ifdef CONFIG_X86_64 -/* - * In x86_64, we need to handle breakpoint -> NMI -> breakpoint. Without - * some care, the inner breakpoint will clobber the outer breakpoint's - * stack. - * - * If a breakpoint is being processed, and the debug stack is being - * used, if an NMI comes in and also hits a breakpoint, the stack - * pointer will be set to the same fixed address as the breakpoint that - * was interrupted, causing that stack to be corrupted. To handle this - * case, check if the stack that was interrupted is the debug stack, and - * if so, change the IDT so that new breakpoints will use the current - * stack and not switch to the fixed address. On return of the NMI, - * switch back to the original IDT. - */ -static DEFINE_PER_CPU(int, update_debug_stack); - -static bool notrace is_debug_stack(unsigned long addr) +DEFINE_IDTENTRY_RAW(exc_nmi) { - struct cea_exception_stacks *cs = __this_cpu_read(cea_exception_stacks); - unsigned long top = CEA_ESTACK_TOP(cs, DB); - unsigned long bot = CEA_ESTACK_BOT(cs, DB1); + irqentry_state_t irq_state; - if (__this_cpu_read(debug_stack_usage)) - return true; /* - * Note, this covers the guard page between DB and DB1 as well to - * avoid two checks. But by all means @addr can never point into - * the guard page. + * Re-enable NMIs right here when running as an SEV-ES guest. This might + * cause nested NMIs, but those can be handled safely. */ - return addr >= bot && addr < top; -} -NOKPROBE_SYMBOL(is_debug_stack); -#endif + sev_es_nmi_complete(); -dotraplinkage notrace void -do_nmi(struct pt_regs *regs, long error_code) -{ - if (IS_ENABLED(CONFIG_SMP) && cpu_is_offline(smp_processor_id())) + if (IS_ENABLED(CONFIG_SMP) && arch_cpu_is_offline(smp_processor_id())) return; if (this_cpu_read(nmi_state) != NMI_NOT_RUNNING) { @@ -521,34 +497,26 @@ do_nmi(struct pt_regs *regs, long error_code) this_cpu_write(nmi_cr2, read_cr2()); nmi_restart: -#ifdef CONFIG_X86_64 /* - * If we interrupted a breakpoint, it is possible that - * the nmi handler will have breakpoints too. We need to - * change the IDT such that breakpoints that happen here - * continue to use the NMI stack. + * Needs to happen before DR7 is accessed, because the hypervisor can + * intercept DR7 reads/writes, turning those into #VC exceptions. */ - if (unlikely(is_debug_stack(regs->sp))) { - debug_stack_set_zero(); - this_cpu_write(update_debug_stack, 1); - } -#endif + sev_es_ist_enter(regs); - nmi_enter(); + this_cpu_write(nmi_dr7, local_db_save()); + + irq_state = irqentry_nmi_enter(regs); inc_irq_stat(__nmi_count); if (!ignore_nmis) default_do_nmi(regs); - nmi_exit(); + irqentry_nmi_exit(regs, irq_state); -#ifdef CONFIG_X86_64 - if (unlikely(this_cpu_read(update_debug_stack))) { - debug_stack_reset(); - this_cpu_write(update_debug_stack, 0); - } -#endif + local_db_restore(this_cpu_read(nmi_dr7)); + + sev_es_ist_exit(); if (unlikely(this_cpu_read(nmi_cr2) != read_cr2())) write_cr2(this_cpu_read(nmi_cr2)); @@ -558,7 +526,16 @@ nmi_restart: if (user_mode(regs)) mds_user_clear_cpu_buffers(); } -NOKPROBE_SYMBOL(do_nmi); + +#if defined(CONFIG_X86_64) && IS_ENABLED(CONFIG_KVM_INTEL) +DEFINE_IDTENTRY_RAW(exc_nmi_noist) +{ + exc_nmi(regs); +} +#endif +#if IS_MODULE(CONFIG_KVM_INTEL) +EXPORT_SYMBOL_GPL(asm_exc_nmi_noist); +#endif void stop_nmi(void) { |