From b7471c6da94d30d3deadc55986cc38d1ff57f9ca Mon Sep 17 00:00:00 2001 From: Don Zickus Date: Tue, 26 Sep 2006 10:52:26 +0200 Subject: [PATCH] i386: Add SMP support on i386 to reservation framework This patch includes the changes to make the nmi watchdog on i386 SMP aware. A bunch of code was moved around to make it simpler to read. In addition, it is now possible to determine if a particular NMI was the result of the watchdog or not. This feature allows the kernel to filter out unknown NMIs easier. Signed-off-by: Don Zickus Signed-off-by: Andi Kleen --- arch/i386/kernel/traps.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch/i386/kernel/traps.c') diff --git a/arch/i386/kernel/traps.c b/arch/i386/kernel/traps.c index 7e9edafffd8a..3a07b2677e2a 100644 --- a/arch/i386/kernel/traps.c +++ b/arch/i386/kernel/traps.c @@ -724,7 +724,7 @@ static void default_do_nmi(struct pt_regs * regs) * so it must be the NMI watchdog. */ if (nmi_watchdog) { - nmi_watchdog_tick(regs); + nmi_watchdog_tick(regs, reason); return; } #endif -- cgit From 3adbbcce9a49b900d4cc118cdccfdefa78bf1afb Mon Sep 17 00:00:00 2001 From: Don Zickus Date: Tue, 26 Sep 2006 10:52:26 +0200 Subject: [PATCH] x86: Cleanup NMI interrupt path This patch cleans up the NMI interrupt path. Instead of being gated by if the 'nmi callback' is set, the interrupt handler now calls everyone who is registered on the die_chain and additionally checks the nmi watchdog, reseting it if enabled. This allows more subsystems to hook into the NMI if they need to (without being block by set_nmi_callback). Signed-off-by: Don Zickus Signed-off-by: Andi Kleen --- arch/i386/kernel/nmi.c | 16 +++++++++++++--- arch/i386/kernel/traps.c | 24 +++++++++++------------- arch/x86_64/kernel/nmi.c | 26 +++++++++++++++++++------- arch/x86_64/kernel/traps.c | 8 ++++---- include/asm-i386/nmi.h | 2 +- include/asm-x86_64/nmi.h | 10 +++++++++- 6 files changed, 57 insertions(+), 29 deletions(-) (limited to 'arch/i386/kernel/traps.c') diff --git a/arch/i386/kernel/nmi.c b/arch/i386/kernel/nmi.c index d88004343034..bd96ea4f2942 100644 --- a/arch/i386/kernel/nmi.c +++ b/arch/i386/kernel/nmi.c @@ -781,7 +781,7 @@ EXPORT_SYMBOL(touch_nmi_watchdog); extern void die_nmi(struct pt_regs *, const char *msg); -void nmi_watchdog_tick (struct pt_regs * regs, unsigned reason) +int nmi_watchdog_tick (struct pt_regs * regs, unsigned reason) { /* @@ -794,10 +794,12 @@ void nmi_watchdog_tick (struct pt_regs * regs, unsigned reason) int cpu = smp_processor_id(); struct nmi_watchdog_ctlblk *wd = &__get_cpu_var(nmi_watchdog_ctlblk); u64 dummy; + int rc=0; /* check for other users first */ if (notify_die(DIE_NMI, "nmi", regs, reason, 2, SIGINT) == NOTIFY_STOP) { + rc = 1; touched = 1; } @@ -850,10 +852,18 @@ void nmi_watchdog_tick (struct pt_regs * regs, unsigned reason) } /* start the cycle over again */ write_watchdog_counter(wd->perfctr_msr, NULL); - } + rc = 1; + } else if (nmi_watchdog == NMI_IO_APIC) { + /* don't know how to accurately check for this. + * just assume it was a watchdog timer interrupt + * This matches the old behaviour. + */ + rc = 1; + } else + printk(KERN_WARNING "Unknown enabled NMI hardware?!\n"); } done: - return; + return rc; } #ifdef CONFIG_SYSCTL diff --git a/arch/i386/kernel/traps.c b/arch/i386/kernel/traps.c index 3a07b2677e2a..282f0bd40dfd 100644 --- a/arch/i386/kernel/traps.c +++ b/arch/i386/kernel/traps.c @@ -706,6 +706,13 @@ void die_nmi (struct pt_regs *regs, const char *msg) do_exit(SIGSEGV); } +static int dummy_nmi_callback(struct pt_regs * regs, int cpu) +{ + return 0; +} + +static nmi_callback_t nmi_callback = dummy_nmi_callback; + static void default_do_nmi(struct pt_regs * regs) { unsigned char reason = 0; @@ -723,12 +730,11 @@ static void default_do_nmi(struct pt_regs * regs) * Ok, so this is none of the documented NMI sources, * so it must be the NMI watchdog. */ - if (nmi_watchdog) { - nmi_watchdog_tick(regs, reason); + if (nmi_watchdog_tick(regs, reason)) return; - } #endif - unknown_nmi_error(reason, regs); + if (!rcu_dereference(nmi_callback)(regs, smp_processor_id())) + unknown_nmi_error(reason, regs); return; } if (notify_die(DIE_NMI, "nmi", regs, reason, 2, SIGINT) == NOTIFY_STOP) @@ -744,13 +750,6 @@ static void default_do_nmi(struct pt_regs * regs) reassert_nmi(); } -static int dummy_nmi_callback(struct pt_regs * regs, int cpu) -{ - return 0; -} - -static nmi_callback_t nmi_callback = dummy_nmi_callback; - fastcall void do_nmi(struct pt_regs * regs, long error_code) { int cpu; @@ -761,8 +760,7 @@ fastcall void do_nmi(struct pt_regs * regs, long error_code) ++nmi_count(cpu); - if (!rcu_dereference(nmi_callback)(regs, cpu)) - default_do_nmi(regs); + default_do_nmi(regs); nmi_exit(); } diff --git a/arch/x86_64/kernel/nmi.c b/arch/x86_64/kernel/nmi.c index d42374a952d7..f6b881b23a70 100644 --- a/arch/x86_64/kernel/nmi.c +++ b/arch/x86_64/kernel/nmi.c @@ -682,16 +682,18 @@ void touch_nmi_watchdog (void) touch_softlockup_watchdog(); } -void __kprobes nmi_watchdog_tick(struct pt_regs * regs, unsigned reason) +int __kprobes nmi_watchdog_tick(struct pt_regs * regs, unsigned reason) { int sum; int touched = 0; struct nmi_watchdog_ctlblk *wd = &__get_cpu_var(nmi_watchdog_ctlblk); u64 dummy; + int rc=0; /* check for other users first */ if (notify_die(DIE_NMI, "nmi", regs, reason, 2, SIGINT) == NOTIFY_STOP) { + rc = 1; touched = 1; } @@ -746,10 +748,18 @@ void __kprobes nmi_watchdog_tick(struct pt_regs * regs, unsigned reason) } /* start the cycle over again */ wrmsrl(wd->perfctr_msr, -((u64)cpu_khz * 1000 / nmi_hz)); - } + rc = 1; + } else if (nmi_watchdog == NMI_IO_APIC) { + /* don't know how to accurately check for this. + * just assume it was a watchdog timer interrupt + * This matches the old behaviour. + */ + rc = 1; + } else + printk(KERN_WARNING "Unknown enabled NMI hardware?!\n"); } done: - return; + return rc; } static __kprobes int dummy_nmi_callback(struct pt_regs * regs, int cpu) @@ -761,15 +771,17 @@ static nmi_callback_t nmi_callback = dummy_nmi_callback; asmlinkage __kprobes void do_nmi(struct pt_regs * regs, long error_code) { - int cpu = safe_smp_processor_id(); - nmi_enter(); add_pda(__nmi_count,1); - if (!rcu_dereference(nmi_callback)(regs, cpu)) - default_do_nmi(regs); + default_do_nmi(regs); nmi_exit(); } +int do_nmi_callback(struct pt_regs * regs, int cpu) +{ + return rcu_dereference(nmi_callback)(regs, cpu); +} + void set_nmi_callback(nmi_callback_t callback) { vmalloc_sync_all(); diff --git a/arch/x86_64/kernel/traps.c b/arch/x86_64/kernel/traps.c index b1249774d1e8..42bc070fdf11 100644 --- a/arch/x86_64/kernel/traps.c +++ b/arch/x86_64/kernel/traps.c @@ -781,12 +781,12 @@ asmlinkage __kprobes void default_do_nmi(struct pt_regs *regs) * Ok, so this is none of the documented NMI sources, * so it must be the NMI watchdog. */ - if (nmi_watchdog > 0) { - nmi_watchdog_tick(regs,reason); + if (nmi_watchdog_tick(regs,reason)) return; - } + if (!do_nmi_callback(regs,cpu)) #endif - unknown_nmi_error(reason, regs); + unknown_nmi_error(reason, regs); + return; } if (notify_die(DIE_NMI, "nmi", regs, reason, 2, SIGINT) == NOTIFY_STOP) diff --git a/include/asm-i386/nmi.h b/include/asm-i386/nmi.h index 4cda6801ecb8..da0e0b4e9139 100644 --- a/include/asm-i386/nmi.h +++ b/include/asm-i386/nmi.h @@ -37,7 +37,7 @@ extern int reserve_lapic_nmi(void); extern void release_lapic_nmi(void); extern void disable_timer_nmi_watchdog(void); extern void enable_timer_nmi_watchdog(void); -extern void nmi_watchdog_tick (struct pt_regs * regs, unsigned reason); +extern int nmi_watchdog_tick (struct pt_regs * regs, unsigned reason); extern atomic_t nmi_active; extern unsigned int nmi_watchdog; diff --git a/include/asm-x86_64/nmi.h b/include/asm-x86_64/nmi.h index 5918136fd853..8f02a2a416e6 100644 --- a/include/asm-x86_64/nmi.h +++ b/include/asm-x86_64/nmi.h @@ -26,6 +26,14 @@ void set_nmi_callback(nmi_callback_t callback); */ void unset_nmi_callback(void); +/** + * do_nmi_callback + * + * Check to see if a callback exists and execute it. Return 1 + * if the handler exists and was handled successfully. + */ +int do_nmi_callback(struct pt_regs *regs, int cpu); + #ifdef CONFIG_PM /** Replace the PM callback routine for NMI. */ @@ -68,7 +76,7 @@ extern int reserve_lapic_nmi(void); extern void release_lapic_nmi(void); extern void disable_timer_nmi_watchdog(void); extern void enable_timer_nmi_watchdog(void); -extern void nmi_watchdog_tick (struct pt_regs * regs, unsigned reason); +extern int nmi_watchdog_tick (struct pt_regs * regs, unsigned reason); extern void nmi_watchdog_default(void); extern int setup_nmi_watchdog(char *); -- cgit From 2fbe7b25c8edaf2d10e6c1a4cc9f8afe714c4764 Mon Sep 17 00:00:00 2001 From: Don Zickus Date: Tue, 26 Sep 2006 10:52:27 +0200 Subject: [PATCH] i386/x86-64: Remove un/set_nmi_callback and reserve/release_lapic_nmi functions Removes the un/set_nmi_callback and reserve/release_lapic_nmi functions as they are no longer needed. The various subsystems are modified to register with the die_notifier instead. Also includes compile fixes by Andrew Morton. Signed-off-by: Don Zickus Signed-off-by: Andi Kleen --- arch/i386/kernel/crash.c | 20 +++++++- arch/i386/kernel/nmi.c | 85 ++++--------------------------- arch/i386/kernel/traps.c | 23 +-------- arch/i386/oprofile/nmi_int.c | 47 ++++++++++------- arch/i386/oprofile/nmi_timer_int.c | 33 +++++++++--- arch/x86_64/kernel/crash.c | 20 +++++++- arch/x86_64/kernel/nmi.c | 102 +++---------------------------------- include/asm-i386/nmi.h | 21 ++------ include/asm-x86_64/nmi.h | 21 -------- kernel/sysctl.c | 4 +- 10 files changed, 116 insertions(+), 260 deletions(-) (limited to 'arch/i386/kernel/traps.c') diff --git a/arch/i386/kernel/crash.c b/arch/i386/kernel/crash.c index 5b96f038367f..736c76d6b31d 100644 --- a/arch/i386/kernel/crash.c +++ b/arch/i386/kernel/crash.c @@ -22,6 +22,8 @@ #include #include #include +#include + #include @@ -93,9 +95,18 @@ static void crash_save_self(struct pt_regs *regs) #if defined(CONFIG_SMP) && defined(CONFIG_X86_LOCAL_APIC) static atomic_t waiting_for_crash_ipi; -static int crash_nmi_callback(struct pt_regs *regs, int cpu) +static int crash_nmi_callback(struct notifier_block *self, + unsigned long val, void *data) { + struct pt_regs *regs; struct pt_regs fixed_regs; + int cpu; + + if (val != DIE_NMI) + return NOTIFY_OK; + + regs = ((struct die_args *)data)->regs; + cpu = raw_smp_processor_id(); /* Don't do anything if this handler is invoked on crashing cpu. * Otherwise, system will completely hang. Crashing cpu can get @@ -125,13 +136,18 @@ static void smp_send_nmi_allbutself(void) send_IPI_allbutself(NMI_VECTOR); } +static struct notifier_block crash_nmi_nb = { + .notifier_call = crash_nmi_callback, +}; + static void nmi_shootdown_cpus(void) { unsigned long msecs; atomic_set(&waiting_for_crash_ipi, num_online_cpus() - 1); /* Would it be better to replace the trap vector here? */ - set_nmi_callback(crash_nmi_callback); + if (register_die_notifier(&crash_nmi_nb)) + return; /* return what? */ /* Ensure the new callback function is set before sending * out the NMI */ diff --git a/arch/i386/kernel/nmi.c b/arch/i386/kernel/nmi.c index bd96ea4f2942..acd3fdea2a21 100644 --- a/arch/i386/kernel/nmi.c +++ b/arch/i386/kernel/nmi.c @@ -42,20 +42,6 @@ static DEFINE_PER_CPU(unsigned long, evntsel_nmi_owner[3]); */ #define NMI_MAX_COUNTER_BITS 66 -/* - * lapic_nmi_owner tracks the ownership of the lapic NMI hardware: - * - it may be reserved by some other driver, or not - * - when not reserved by some other driver, it may be used for - * the NMI watchdog, or not - * - * This is maintained separately from nmi_active because the NMI - * watchdog may also be driven from the I/O APIC timer. - */ -static DEFINE_SPINLOCK(lapic_nmi_owner_lock); -static unsigned int lapic_nmi_owner; -#define LAPIC_NMI_WATCHDOG (1<<0) -#define LAPIC_NMI_RESERVED (1<<1) - /* nmi_active: * >0: the lapic NMI watchdog is active, but can be disabled * <0: the lapic NMI watchdog has not been set up, and cannot @@ -325,33 +311,6 @@ static void enable_lapic_nmi_watchdog(void) touch_nmi_watchdog(); } -int reserve_lapic_nmi(void) -{ - unsigned int old_owner; - - spin_lock(&lapic_nmi_owner_lock); - old_owner = lapic_nmi_owner; - lapic_nmi_owner |= LAPIC_NMI_RESERVED; - spin_unlock(&lapic_nmi_owner_lock); - if (old_owner & LAPIC_NMI_RESERVED) - return -EBUSY; - if (old_owner & LAPIC_NMI_WATCHDOG) - disable_lapic_nmi_watchdog(); - return 0; -} - -void release_lapic_nmi(void) -{ - unsigned int new_owner; - - spin_lock(&lapic_nmi_owner_lock); - new_owner = lapic_nmi_owner & ~LAPIC_NMI_RESERVED; - lapic_nmi_owner = new_owner; - spin_unlock(&lapic_nmi_owner_lock); - if (new_owner & LAPIC_NMI_WATCHDOG) - enable_lapic_nmi_watchdog(); -} - void disable_timer_nmi_watchdog(void) { BUG_ON(nmi_watchdog != NMI_IO_APIC); @@ -866,6 +825,15 @@ done: return rc; } +int do_nmi_callback(struct pt_regs * regs, int cpu) +{ +#ifdef CONFIG_SYSCTL + if (unknown_nmi_panic) + return unknown_nmi_panic_callback(regs, cpu); +#endif + return 0; +} + #ifdef CONFIG_SYSCTL static int unknown_nmi_panic_callback(struct pt_regs *regs, int cpu) @@ -873,37 +841,8 @@ static int unknown_nmi_panic_callback(struct pt_regs *regs, int cpu) unsigned char reason = get_nmi_reason(); char buf[64]; - if (!(reason & 0xc0)) { - sprintf(buf, "NMI received for unknown reason %02x\n", reason); - die_nmi(regs, buf); - } - return 0; -} - -/* - * proc handler for /proc/sys/kernel/unknown_nmi_panic - */ -int proc_unknown_nmi_panic(ctl_table *table, int write, struct file *file, - void __user *buffer, size_t *length, loff_t *ppos) -{ - int old_state; - - old_state = unknown_nmi_panic; - proc_dointvec(table, write, file, buffer, length, ppos); - if (!!old_state == !!unknown_nmi_panic) - return 0; - - if (unknown_nmi_panic) { - if (reserve_lapic_nmi() < 0) { - unknown_nmi_panic = 0; - return -EBUSY; - } else { - set_nmi_callback(unknown_nmi_panic_callback); - } - } else { - release_lapic_nmi(); - unset_nmi_callback(); - } + sprintf(buf, "NMI received for unknown reason %02x\n", reason); + die_nmi(regs, buf); return 0; } @@ -917,7 +856,5 @@ EXPORT_SYMBOL(reserve_perfctr_nmi); EXPORT_SYMBOL(release_perfctr_nmi); EXPORT_SYMBOL(reserve_evntsel_nmi); EXPORT_SYMBOL(release_evntsel_nmi); -EXPORT_SYMBOL(reserve_lapic_nmi); -EXPORT_SYMBOL(release_lapic_nmi); EXPORT_SYMBOL(disable_timer_nmi_watchdog); EXPORT_SYMBOL(enable_timer_nmi_watchdog); diff --git a/arch/i386/kernel/traps.c b/arch/i386/kernel/traps.c index 282f0bd40dfd..7db664d0b25c 100644 --- a/arch/i386/kernel/traps.c +++ b/arch/i386/kernel/traps.c @@ -706,13 +706,6 @@ void die_nmi (struct pt_regs *regs, const char *msg) do_exit(SIGSEGV); } -static int dummy_nmi_callback(struct pt_regs * regs, int cpu) -{ - return 0; -} - -static nmi_callback_t nmi_callback = dummy_nmi_callback; - static void default_do_nmi(struct pt_regs * regs) { unsigned char reason = 0; @@ -732,9 +725,10 @@ static void default_do_nmi(struct pt_regs * regs) */ if (nmi_watchdog_tick(regs, reason)) return; + if (!do_nmi_callback(regs, smp_processor_id())) #endif - if (!rcu_dereference(nmi_callback)(regs, smp_processor_id())) unknown_nmi_error(reason, regs); + return; } if (notify_die(DIE_NMI, "nmi", regs, reason, 2, SIGINT) == NOTIFY_STOP) @@ -765,19 +759,6 @@ fastcall void do_nmi(struct pt_regs * regs, long error_code) nmi_exit(); } -void set_nmi_callback(nmi_callback_t callback) -{ - vmalloc_sync_all(); - rcu_assign_pointer(nmi_callback, callback); -} -EXPORT_SYMBOL_GPL(set_nmi_callback); - -void unset_nmi_callback(void) -{ - nmi_callback = dummy_nmi_callback; -} -EXPORT_SYMBOL_GPL(unset_nmi_callback); - #ifdef CONFIG_KPROBES fastcall void __kprobes do_int3(struct pt_regs *regs, long error_code) { diff --git a/arch/i386/oprofile/nmi_int.c b/arch/i386/oprofile/nmi_int.c index 8710ca081b1e..b3610188bcf0 100644 --- a/arch/i386/oprofile/nmi_int.c +++ b/arch/i386/oprofile/nmi_int.c @@ -17,14 +17,15 @@ #include #include #include +#include #include "op_counter.h" #include "op_x86_model.h" - + static struct op_x86_model_spec const * model; static struct op_msrs cpu_msrs[NR_CPUS]; static unsigned long saved_lvtpc[NR_CPUS]; - + static int nmi_start(void); static void nmi_stop(void); @@ -82,13 +83,24 @@ static void exit_driverfs(void) #define exit_driverfs() do { } while (0) #endif /* CONFIG_PM */ - -static int nmi_callback(struct pt_regs * regs, int cpu) +int profile_exceptions_notify(struct notifier_block *self, + unsigned long val, void *data) { - return model->check_ctrs(regs, &cpu_msrs[cpu]); + struct die_args *args = (struct die_args *)data; + int ret = NOTIFY_DONE; + int cpu = smp_processor_id(); + + switch(val) { + case DIE_NMI: + if (model->check_ctrs(args->regs, &cpu_msrs[cpu])) + ret = NOTIFY_STOP; + break; + default: + break; + } + return ret; } - - + static void nmi_cpu_save_registers(struct op_msrs * msrs) { unsigned int const nr_ctrs = model->num_counters; @@ -174,27 +186,29 @@ static void nmi_cpu_setup(void * dummy) apic_write(APIC_LVTPC, APIC_DM_NMI); } +static struct notifier_block profile_exceptions_nb = { + .notifier_call = profile_exceptions_notify, + .next = NULL, + .priority = 0 +}; static int nmi_setup(void) { + int err=0; + if (!allocate_msrs()) return -ENOMEM; - /* We walk a thin line between law and rape here. - * We need to be careful to install our NMI handler - * without actually triggering any NMIs as this will - * break the core code horrifically. - */ - if (reserve_lapic_nmi() < 0) { + if ((err = register_die_notifier(&profile_exceptions_nb))){ free_msrs(); - return -EBUSY; + return err; } + /* We need to serialize save and setup for HT because the subset * of msrs are distinct for save and setup operations */ on_each_cpu(nmi_save_registers, NULL, 0, 1); on_each_cpu(nmi_cpu_setup, NULL, 0, 1); - set_nmi_callback(nmi_callback); nmi_enabled = 1; return 0; } @@ -250,8 +264,7 @@ static void nmi_shutdown(void) { nmi_enabled = 0; on_each_cpu(nmi_cpu_shutdown, NULL, 0, 1); - unset_nmi_callback(); - release_lapic_nmi(); + unregister_die_notifier(&profile_exceptions_nb); free_msrs(); } diff --git a/arch/i386/oprofile/nmi_timer_int.c b/arch/i386/oprofile/nmi_timer_int.c index a33a73bb502d..93ca48f804ac 100644 --- a/arch/i386/oprofile/nmi_timer_int.c +++ b/arch/i386/oprofile/nmi_timer_int.c @@ -17,32 +17,49 @@ #include #include #include +#include -static int nmi_timer_callback(struct pt_regs * regs, int cpu) +int profile_timer_exceptions_notify(struct notifier_block *self, + unsigned long val, void *data) { - oprofile_add_sample(regs, 0); - return 1; + struct die_args *args = (struct die_args *)data; + int ret = NOTIFY_DONE; + + switch(val) { + case DIE_NMI: + oprofile_add_sample(args->regs, 0); + ret = NOTIFY_STOP; + break; + default: + break; + } + return ret; } +static struct notifier_block profile_timer_exceptions_nb = { + .notifier_call = profile_timer_exceptions_notify, + .next = NULL, + .priority = 0 +}; + static int timer_start(void) { - disable_timer_nmi_watchdog(); - set_nmi_callback(nmi_timer_callback); + if (register_die_notifier(&profile_timer_exceptions_nb)) + return 1; return 0; } static void timer_stop(void) { - enable_timer_nmi_watchdog(); - unset_nmi_callback(); + unregister_die_notifier(&profile_timer_exceptions_nb); synchronize_sched(); /* Allow already-started NMIs to complete. */ } int __init op_nmi_timer_init(struct oprofile_operations * ops) { - if (atomic_read(&nmi_active) <= 0) + if ((nmi_watchdog != NMI_IO_APIC) || (atomic_read(&nmi_active) <= 0)) return -ENODEV; ops->start = timer_start; diff --git a/arch/x86_64/kernel/crash.c b/arch/x86_64/kernel/crash.c index d8d5750d6106..44c8af65325e 100644 --- a/arch/x86_64/kernel/crash.c +++ b/arch/x86_64/kernel/crash.c @@ -23,6 +23,7 @@ #include #include #include +#include /* This keeps a track of which one is crashing cpu. */ static int crashing_cpu; @@ -95,8 +96,18 @@ static void crash_save_self(struct pt_regs *regs) #ifdef CONFIG_SMP static atomic_t waiting_for_crash_ipi; -static int crash_nmi_callback(struct pt_regs *regs, int cpu) +static int crash_nmi_callback(struct notifier_block *self, + unsigned long val, void *data) { + struct pt_regs *regs; + int cpu; + + if (val != DIE_NMI) + return NOTIFY_OK; + + regs = ((struct die_args *)data)->regs; + cpu = raw_smp_processor_id(); + /* * Don't do anything if this handler is invoked on crashing cpu. * Otherwise, system will completely hang. Crashing cpu can get @@ -127,12 +138,17 @@ static void smp_send_nmi_allbutself(void) * cpu hotplug shouldn't matter. */ +static struct notifier_block crash_nmi_nb = { + .notifier_call = crash_nmi_callback, +}; + static void nmi_shootdown_cpus(void) { unsigned long msecs; atomic_set(&waiting_for_crash_ipi, num_online_cpus() - 1); - set_nmi_callback(crash_nmi_callback); + if (register_die_notifier(&crash_nmi_nb)) + return; /* return what? */ /* * Ensure the new callback function is set before sending diff --git a/arch/x86_64/kernel/nmi.c b/arch/x86_64/kernel/nmi.c index f6b881b23a70..9d175dcf3a2d 100644 --- a/arch/x86_64/kernel/nmi.c +++ b/arch/x86_64/kernel/nmi.c @@ -41,20 +41,6 @@ static DEFINE_PER_CPU(unsigned, evntsel_nmi_owner[2]); */ #define NMI_MAX_COUNTER_BITS 66 -/* - * lapic_nmi_owner tracks the ownership of the lapic NMI hardware: - * - it may be reserved by some other driver, or not - * - when not reserved by some other driver, it may be used for - * the NMI watchdog, or not - * - * This is maintained separately from nmi_active because the NMI - * watchdog may also be driven from the I/O APIC timer. - */ -static DEFINE_SPINLOCK(lapic_nmi_owner_lock); -static unsigned int lapic_nmi_owner; -#define LAPIC_NMI_WATCHDOG (1<<0) -#define LAPIC_NMI_RESERVED (1<<1) - /* nmi_active: * >0: the lapic NMI watchdog is active, but can be disabled * <0: the lapic NMI watchdog has not been set up, and cannot @@ -321,33 +307,6 @@ static void enable_lapic_nmi_watchdog(void) touch_nmi_watchdog(); } -int reserve_lapic_nmi(void) -{ - unsigned int old_owner; - - spin_lock(&lapic_nmi_owner_lock); - old_owner = lapic_nmi_owner; - lapic_nmi_owner |= LAPIC_NMI_RESERVED; - spin_unlock(&lapic_nmi_owner_lock); - if (old_owner & LAPIC_NMI_RESERVED) - return -EBUSY; - if (old_owner & LAPIC_NMI_WATCHDOG) - disable_lapic_nmi_watchdog(); - return 0; -} - -void release_lapic_nmi(void) -{ - unsigned int new_owner; - - spin_lock(&lapic_nmi_owner_lock); - new_owner = lapic_nmi_owner & ~LAPIC_NMI_RESERVED; - lapic_nmi_owner = new_owner; - spin_unlock(&lapic_nmi_owner_lock); - if (new_owner & LAPIC_NMI_WATCHDOG) - enable_lapic_nmi_watchdog(); -} - void disable_timer_nmi_watchdog(void) { BUG_ON(nmi_watchdog != NMI_IO_APIC); @@ -762,13 +721,6 @@ done: return rc; } -static __kprobes int dummy_nmi_callback(struct pt_regs * regs, int cpu) -{ - return 0; -} - -static nmi_callback_t nmi_callback = dummy_nmi_callback; - asmlinkage __kprobes void do_nmi(struct pt_regs * regs, long error_code) { nmi_enter(); @@ -779,21 +731,12 @@ asmlinkage __kprobes void do_nmi(struct pt_regs * regs, long error_code) int do_nmi_callback(struct pt_regs * regs, int cpu) { - return rcu_dereference(nmi_callback)(regs, cpu); -} - -void set_nmi_callback(nmi_callback_t callback) -{ - vmalloc_sync_all(); - rcu_assign_pointer(nmi_callback, callback); -} -EXPORT_SYMBOL_GPL(set_nmi_callback); - -void unset_nmi_callback(void) -{ - nmi_callback = dummy_nmi_callback; +#ifdef CONFIG_SYSCTL + if (unknown_nmi_panic) + return unknown_nmi_panic_callback(regs, cpu); +#endif + return 0; } -EXPORT_SYMBOL_GPL(unset_nmi_callback); #ifdef CONFIG_SYSCTL @@ -802,37 +745,8 @@ static int unknown_nmi_panic_callback(struct pt_regs *regs, int cpu) unsigned char reason = get_nmi_reason(); char buf[64]; - if (!(reason & 0xc0)) { - sprintf(buf, "NMI received for unknown reason %02x\n", reason); - die_nmi(buf,regs); - } - return 0; -} - -/* - * proc handler for /proc/sys/kernel/unknown_nmi_panic - */ -int proc_unknown_nmi_panic(struct ctl_table *table, int write, struct file *file, - void __user *buffer, size_t *length, loff_t *ppos) -{ - int old_state; - - old_state = unknown_nmi_panic; - proc_dointvec(table, write, file, buffer, length, ppos); - if (!!old_state == !!unknown_nmi_panic) - return 0; - - if (unknown_nmi_panic) { - if (reserve_lapic_nmi() < 0) { - unknown_nmi_panic = 0; - return -EBUSY; - } else { - set_nmi_callback(unknown_nmi_panic_callback); - } - } else { - release_lapic_nmi(); - unset_nmi_callback(); - } + sprintf(buf, "NMI received for unknown reason %02x\n", reason); + die_nmi(buf,regs); return 0; } @@ -846,8 +760,6 @@ EXPORT_SYMBOL(reserve_perfctr_nmi); EXPORT_SYMBOL(release_perfctr_nmi); EXPORT_SYMBOL(reserve_evntsel_nmi); EXPORT_SYMBOL(release_evntsel_nmi); -EXPORT_SYMBOL(reserve_lapic_nmi); -EXPORT_SYMBOL(release_lapic_nmi); EXPORT_SYMBOL(disable_timer_nmi_watchdog); EXPORT_SYMBOL(enable_timer_nmi_watchdog); EXPORT_SYMBOL(touch_nmi_watchdog); diff --git a/include/asm-i386/nmi.h b/include/asm-i386/nmi.h index da0e0b4e9139..34d6bf063b6e 100644 --- a/include/asm-i386/nmi.h +++ b/include/asm-i386/nmi.h @@ -6,24 +6,13 @@ #include -struct pt_regs; - -typedef int (*nmi_callback_t)(struct pt_regs * regs, int cpu); - -/** - * set_nmi_callback - * - * Set a handler for an NMI. Only one handler may be - * set. Return 1 if the NMI was handled. - */ -void set_nmi_callback(nmi_callback_t callback); - /** - * unset_nmi_callback + * do_nmi_callback * - * Remove the handler previously set. + * Check to see if a callback exists and execute it. Return 1 + * if the handler exists and was handled successfully. */ -void unset_nmi_callback(void); +int do_nmi_callback(struct pt_regs *regs, int cpu); extern int avail_to_resrv_perfctr_nmi_bit(unsigned int); extern int avail_to_resrv_perfctr_nmi(unsigned int); @@ -33,8 +22,6 @@ extern int reserve_evntsel_nmi(unsigned int); extern void release_evntsel_nmi(unsigned int); extern void setup_apic_nmi_watchdog (void *); -extern int reserve_lapic_nmi(void); -extern void release_lapic_nmi(void); extern void disable_timer_nmi_watchdog(void); extern void enable_timer_nmi_watchdog(void); extern int nmi_watchdog_tick (struct pt_regs * regs, unsigned reason); diff --git a/include/asm-x86_64/nmi.h b/include/asm-x86_64/nmi.h index 8f02a2a416e6..8818c39d34e0 100644 --- a/include/asm-x86_64/nmi.h +++ b/include/asm-x86_64/nmi.h @@ -7,25 +7,6 @@ #include #include -struct pt_regs; - -typedef int (*nmi_callback_t)(struct pt_regs * regs, int cpu); - -/** - * set_nmi_callback - * - * Set a handler for an NMI. Only one handler may be - * set. Return 1 if the NMI was handled. - */ -void set_nmi_callback(nmi_callback_t callback); - -/** - * unset_nmi_callback - * - * Remove the handler previously set. - */ -void unset_nmi_callback(void); - /** * do_nmi_callback * @@ -72,8 +53,6 @@ extern int reserve_evntsel_nmi(unsigned int); extern void release_evntsel_nmi(unsigned int); extern void setup_apic_nmi_watchdog (void *); -extern int reserve_lapic_nmi(void); -extern void release_lapic_nmi(void); extern void disable_timer_nmi_watchdog(void); extern void enable_timer_nmi_watchdog(void); extern int nmi_watchdog_tick (struct pt_regs * regs, unsigned reason); diff --git a/kernel/sysctl.c b/kernel/sysctl.c index 362a0cc37138..83f168361624 100644 --- a/kernel/sysctl.c +++ b/kernel/sysctl.c @@ -76,8 +76,6 @@ extern int compat_log; #if defined(CONFIG_X86_LOCAL_APIC) && defined(CONFIG_X86) int unknown_nmi_panic; -extern int proc_unknown_nmi_panic(ctl_table *, int, struct file *, - void __user *, size_t *, loff_t *); #endif /* this is needed for the proc_dointvec_minmax for [fs_]overflow UID and GID */ @@ -628,7 +626,7 @@ static ctl_table kern_table[] = { .data = &unknown_nmi_panic, .maxlen = sizeof (int), .mode = 0644, - .proc_handler = &proc_unknown_nmi_panic, + .proc_handler = &proc_dointvec, }, #endif #if defined(CONFIG_X86) -- cgit From 8da5adda91df3d2fcc5300e68da491694c9af019 Mon Sep 17 00:00:00 2001 From: Don Zickus Date: Tue, 26 Sep 2006 10:52:27 +0200 Subject: [PATCH] x86: Allow users to force a panic on NMI To quote Alan Cox: The default Linux behaviour on an NMI of either memory or unknown is to continue operation. For many environments such as scientific computing it is preferable that the box is taken out and the error dealt with than an uncorrected parity/ECC error get propogated. A small number of systems do generate NMI's for bizarre random reasons such as power management so the default is unchanged. In other respects the new proc/sys entry works like the existing panic controls already in that directory. This is separate to the edac support - EDAC allows supported chipsets to handle ECC errors well, this change allows unsupported cases to at least panic rather than cause problems further down the line. Signed-off-by: Don Zickus Signed-off-by: Andi Kleen --- arch/i386/kernel/traps.c | 6 ++++++ arch/x86_64/kernel/traps.c | 6 ++++++ include/linux/kernel.h | 1 + include/linux/sysctl.h | 1 + kernel/panic.c | 1 + kernel/sysctl.c | 8 ++++++++ 6 files changed, 23 insertions(+) (limited to 'arch/i386/kernel/traps.c') diff --git a/arch/i386/kernel/traps.c b/arch/i386/kernel/traps.c index 7db664d0b25c..2f6cb8276480 100644 --- a/arch/i386/kernel/traps.c +++ b/arch/i386/kernel/traps.c @@ -635,6 +635,8 @@ static void mem_parity_error(unsigned char reason, struct pt_regs * regs) "to continue\n"); printk(KERN_EMERG "You probably have a hardware problem with your RAM " "chips\n"); + if (panic_on_unrecovered_nmi) + panic("NMI: Not continuing"); /* Clear and disable the memory parity error line. */ clear_mem_error(reason); @@ -670,6 +672,10 @@ static void unknown_nmi_error(unsigned char reason, struct pt_regs * regs) reason, smp_processor_id()); printk("Dazed and confused, but trying to continue\n"); printk("Do you have a strange power saving mode enabled?\n"); + + if (panic_on_unrecovered_nmi) + panic("NMI: Not continuing"); + } static DEFINE_SPINLOCK(nmi_print_lock); diff --git a/arch/x86_64/kernel/traps.c b/arch/x86_64/kernel/traps.c index 42bc070fdf11..b18829db2a6a 100644 --- a/arch/x86_64/kernel/traps.c +++ b/arch/x86_64/kernel/traps.c @@ -732,6 +732,8 @@ mem_parity_error(unsigned char reason, struct pt_regs * regs) { printk("Uhhuh. NMI received. Dazed and confused, but trying to continue\n"); printk("You probably have a hardware problem with your RAM chips\n"); + if (panic_on_unrecovered_nmi) + panic("NMI: Not continuing"); /* Clear and disable the memory parity error line. */ reason = (reason & 0xf) | 4; @@ -757,6 +759,10 @@ unknown_nmi_error(unsigned char reason, struct pt_regs * regs) { printk("Uhhuh. NMI received for unknown reason %02x.\n", reason); printk("Dazed and confused, but trying to continue\n"); printk("Do you have a strange power saving mode enabled?\n"); + + if (panic_on_unrecovered_nmi) + panic("NMI: Not continuing"); + } /* Runs on IST stack. This code must keep interrupts off all the time. diff --git a/include/linux/kernel.h b/include/linux/kernel.h index 2b2ae4fdce8b..1ff9609300b4 100644 --- a/include/linux/kernel.h +++ b/include/linux/kernel.h @@ -186,6 +186,7 @@ extern void bust_spinlocks(int yes); extern int oops_in_progress; /* If set, an oops, panic(), BUG() or die() is in progress */ extern int panic_timeout; extern int panic_on_oops; +extern int panic_on_unrecovered_nmi; extern int tainted; extern const char *print_tainted(void); extern void add_taint(unsigned); diff --git a/include/linux/sysctl.h b/include/linux/sysctl.h index ecb79ba52ae1..432778446ad2 100644 --- a/include/linux/sysctl.h +++ b/include/linux/sysctl.h @@ -151,6 +151,7 @@ enum KERN_COMPAT_LOG=73, /* int: print compat layer messages */ KERN_MAX_LOCK_DEPTH=74, KERN_NMI_WATCHDOG=75, /* int: enable/disable nmi watchdog */ + KERN_PANIC_ON_NMI=76, /* int: whether we will panic on an unrecovered */ }; diff --git a/kernel/panic.c b/kernel/panic.c index 8010b9b17aca..d2db3e2209e0 100644 --- a/kernel/panic.c +++ b/kernel/panic.c @@ -21,6 +21,7 @@ #include int panic_on_oops; +int panic_on_unrecovered_nmi; int tainted; static int pause_on_oops; static int pause_on_oops_flag; diff --git a/kernel/sysctl.c b/kernel/sysctl.c index 040de6bd74dd..220e20564124 100644 --- a/kernel/sysctl.c +++ b/kernel/sysctl.c @@ -641,6 +641,14 @@ static ctl_table kern_table[] = { }, #endif #if defined(CONFIG_X86) + { + .ctl_name = KERN_PANIC_ON_NMI, + .procname = "panic_on_unrecovered_nmi", + .data = &panic_on_unrecovered_nmi, + .maxlen = sizeof(int), + .mode = 0644, + .proc_handler = &proc_dointvec, + }, { .ctl_name = KERN_BOOTLOADER_TYPE, .procname = "bootloader_type", -- cgit From c41c5cd3b20a2d81c30498f13b1527847a8fdf69 Mon Sep 17 00:00:00 2001 From: Don Zickus Date: Tue, 26 Sep 2006 10:52:27 +0200 Subject: [PATCH] x86: x86 clean up nmi panic messages Clean up some of the output messages on the nmi error paths to make more sense when they are displayed. This is mainly a cosmetic fix and shouldn't impact any normal code path. Signed-off-by: Don Zickus Signed-off-by: Andi Kleen --- arch/i386/kernel/traps.c | 15 ++++++++------- arch/x86_64/kernel/traps.c | 21 ++++++++++++++------- 2 files changed, 22 insertions(+), 14 deletions(-) (limited to 'arch/i386/kernel/traps.c') diff --git a/arch/i386/kernel/traps.c b/arch/i386/kernel/traps.c index 2f6cb8276480..3c85c89f68d8 100644 --- a/arch/i386/kernel/traps.c +++ b/arch/i386/kernel/traps.c @@ -631,13 +631,15 @@ gp_in_kernel: static void mem_parity_error(unsigned char reason, struct pt_regs * regs) { - printk(KERN_EMERG "Uhhuh. NMI received. Dazed and confused, but trying " - "to continue\n"); + printk(KERN_EMERG "Uhhuh. NMI received for unknown reason %02x on " + "CPU %d.\n", reason, smp_processor_id()); printk(KERN_EMERG "You probably have a hardware problem with your RAM " "chips\n"); if (panic_on_unrecovered_nmi) panic("NMI: Not continuing"); + printk(KERN_EMERG "Dazed and confused, but trying to continue\n"); + /* Clear and disable the memory parity error line. */ clear_mem_error(reason); } @@ -668,14 +670,13 @@ static void unknown_nmi_error(unsigned char reason, struct pt_regs * regs) return; } #endif - printk("Uhhuh. NMI received for unknown reason %02x on CPU %d.\n", - reason, smp_processor_id()); - printk("Dazed and confused, but trying to continue\n"); - printk("Do you have a strange power saving mode enabled?\n"); - + printk(KERN_EMERG "Uhhuh. NMI received for unknown reason %02x on " + "CPU %d.\n", reason, smp_processor_id()); + printk(KERN_EMERG "Do you have a strange power saving mode enabled?\n"); if (panic_on_unrecovered_nmi) panic("NMI: Not continuing"); + printk(KERN_EMERG "Dazed and confused, but trying to continue\n"); } static DEFINE_SPINLOCK(nmi_print_lock); diff --git a/arch/x86_64/kernel/traps.c b/arch/x86_64/kernel/traps.c index b18829db2a6a..dae10df60926 100644 --- a/arch/x86_64/kernel/traps.c +++ b/arch/x86_64/kernel/traps.c @@ -730,10 +730,15 @@ asmlinkage void __kprobes do_general_protection(struct pt_regs * regs, static __kprobes void mem_parity_error(unsigned char reason, struct pt_regs * regs) { - printk("Uhhuh. NMI received. Dazed and confused, but trying to continue\n"); - printk("You probably have a hardware problem with your RAM chips\n"); + printk(KERN_EMERG "Uhhuh. NMI received for unknown reason %02x.\n", + reason); + printk(KERN_EMERG "You probably have a hardware problem with your " + "RAM chips\n"); + if (panic_on_unrecovered_nmi) - panic("NMI: Not continuing"); + panic("NMI: Not continuing"); + + printk(KERN_EMERG "Dazed and confused, but trying to continue\n"); /* Clear and disable the memory parity error line. */ reason = (reason & 0xf) | 4; @@ -756,13 +761,15 @@ io_check_error(unsigned char reason, struct pt_regs * regs) static __kprobes void unknown_nmi_error(unsigned char reason, struct pt_regs * regs) -{ printk("Uhhuh. NMI received for unknown reason %02x.\n", reason); - printk("Dazed and confused, but trying to continue\n"); - printk("Do you have a strange power saving mode enabled?\n"); +{ + printk(KERN_EMERG "Uhhuh. NMI received for unknown reason %02x.\n", + reason); + printk(KERN_EMERG "Do you have a strange power saving mode enabled?\n"); if (panic_on_unrecovered_nmi) - panic("NMI: Not continuing"); + panic("NMI: Not continuing"); + printk(KERN_EMERG "Dazed and confused, but trying to continue\n"); } /* Runs on IST stack. This code must keep interrupts off all the time. -- cgit From 2b14a78cd07a52001b8c3865ed615d8b9b905b78 Mon Sep 17 00:00:00 2001 From: Andi Kleen Date: Tue, 26 Sep 2006 10:52:34 +0200 Subject: [PATCH] i386: Do stacktracer conversion too Following x86-64 patches. Reuses code from them in fact. Convert the standard backtracer to do all output using callbacks. Use the x86-64 stack tracer implementation that uses these callbacks to implement the stacktrace interface. This allows to use the new dwarf2 unwinder for stacktrace and get better backtraces. Cc: mingo@elte.hu Signed-off-by: Andi Kleen --- arch/i386/kernel/Makefile | 1 + arch/i386/kernel/stacktrace.c | 93 ------------------------------------ arch/i386/kernel/traps.c | 108 +++++++++++++++++++++++++++++++----------- include/asm-i386/stacktrace.h | 1 + 4 files changed, 82 insertions(+), 121 deletions(-) delete mode 100644 arch/i386/kernel/stacktrace.c create mode 100644 include/asm-i386/stacktrace.h (limited to 'arch/i386/kernel/traps.c') diff --git a/arch/i386/kernel/Makefile b/arch/i386/kernel/Makefile index dab497472deb..1a884b6e6e5c 100644 --- a/arch/i386/kernel/Makefile +++ b/arch/i386/kernel/Makefile @@ -81,4 +81,5 @@ $(obj)/vsyscall-syms.o: $(src)/vsyscall.lds \ $(call if_changed,syscall) k8-y += ../../x86_64/kernel/k8.o +stacktrace-y += ../../x86_64/kernel/stacktrace.o diff --git a/arch/i386/kernel/stacktrace.c b/arch/i386/kernel/stacktrace.c deleted file mode 100644 index ae3c32a87add..000000000000 --- a/arch/i386/kernel/stacktrace.c +++ /dev/null @@ -1,93 +0,0 @@ -/* - * arch/i386/kernel/stacktrace.c - * - * Stack trace management functions - * - * Copyright (C) 2006 Red Hat, Inc., Ingo Molnar - */ -#include -#include - -static inline int valid_stack_ptr(struct thread_info *tinfo, void *p) -{ - return p > (void *)tinfo && - p < (void *)tinfo + THREAD_SIZE - 3; -} - -/* - * Save stack-backtrace addresses into a stack_trace buffer: - */ -static inline unsigned long -save_context_stack(struct stack_trace *trace, unsigned int skip, - struct thread_info *tinfo, unsigned long *stack, - unsigned long ebp) -{ - unsigned long addr; - -#ifdef CONFIG_FRAME_POINTER - while (valid_stack_ptr(tinfo, (void *)ebp)) { - addr = *(unsigned long *)(ebp + 4); - if (!skip) - trace->entries[trace->nr_entries++] = addr; - else - skip--; - if (trace->nr_entries >= trace->max_entries) - break; - /* - * break out of recursive entries (such as - * end_of_stack_stop_unwind_function): - */ - if (ebp == *(unsigned long *)ebp) - break; - - ebp = *(unsigned long *)ebp; - } -#else - while (valid_stack_ptr(tinfo, stack)) { - addr = *stack++; - if (__kernel_text_address(addr)) { - if (!skip) - trace->entries[trace->nr_entries++] = addr; - else - skip--; - if (trace->nr_entries >= trace->max_entries) - break; - } - } -#endif - - return ebp; -} - -/* - * Save stack-backtrace addresses into a stack_trace buffer. - */ -void save_stack_trace(struct stack_trace *trace, struct task_struct *task) -{ - unsigned long ebp; - unsigned long *stack = &ebp; - - WARN_ON(trace->nr_entries || !trace->max_entries); - - if (!task || task == current) { - /* Grab ebp right from our regs: */ - asm ("movl %%ebp, %0" : "=r" (ebp)); - } else { - /* ebp is the last reg pushed by switch_to(): */ - ebp = *(unsigned long *) task->thread.esp; - } - - while (1) { - struct thread_info *context = (struct thread_info *) - ((unsigned long)stack & (~(THREAD_SIZE - 1))); - - ebp = save_context_stack(trace, trace->skip, context, stack, ebp); - stack = (unsigned long *)context->previous_esp; - if (!stack || trace->nr_entries >= trace->max_entries) - break; - trace->entries[trace->nr_entries++] = ULONG_MAX; - if (trace->nr_entries >= trace->max_entries) - break; - } -} - diff --git a/arch/i386/kernel/traps.c b/arch/i386/kernel/traps.c index 3c85c89f68d8..4ced4285163b 100644 --- a/arch/i386/kernel/traps.c +++ b/arch/i386/kernel/traps.c @@ -51,6 +51,7 @@ #include #include #include +#include #include @@ -118,26 +119,16 @@ static inline int valid_stack_ptr(struct thread_info *tinfo, void *p) p < (void *)tinfo + THREAD_SIZE - 3; } -/* - * Print one address/symbol entries per line. - */ -static inline void print_addr_and_symbol(unsigned long addr, char *log_lvl) -{ - printk(" [<%08lx>] ", addr); - - print_symbol("%s\n", addr); -} - static inline unsigned long print_context_stack(struct thread_info *tinfo, unsigned long *stack, unsigned long ebp, - char *log_lvl) + struct stacktrace_ops *ops, void *data) { unsigned long addr; #ifdef CONFIG_FRAME_POINTER while (valid_stack_ptr(tinfo, (void *)ebp)) { addr = *(unsigned long *)(ebp + 4); - print_addr_and_symbol(addr, log_lvl); + ops->address(data, addr); /* * break out of recursive entries (such as * end_of_stack_stop_unwind_function): @@ -150,28 +141,35 @@ static inline unsigned long print_context_stack(struct thread_info *tinfo, while (valid_stack_ptr(tinfo, stack)) { addr = *stack++; if (__kernel_text_address(addr)) - print_addr_and_symbol(addr, log_lvl); + ops->address(data, addr); } #endif return ebp; } +struct ops_and_data { + struct stacktrace_ops *ops; + void *data; +}; + static asmlinkage int -show_trace_unwind(struct unwind_frame_info *info, void *log_lvl) +dump_trace_unwind(struct unwind_frame_info *info, void *data) { + struct ops_and_data *oad = (struct ops_and_data *)data; int n = 0; while (unwind(info) == 0 && UNW_PC(info)) { n++; - print_addr_and_symbol(UNW_PC(info), log_lvl); + oad->ops->address(oad->data, UNW_PC(info)); if (arch_unw_user_mode(info)) break; } return n; } -static void show_trace_log_lvl(struct task_struct *task, struct pt_regs *regs, - unsigned long *stack, char *log_lvl) +void dump_trace(struct task_struct *task, struct pt_regs *regs, + unsigned long *stack, + struct stacktrace_ops *ops, void *data) { unsigned long ebp; @@ -181,31 +179,37 @@ static void show_trace_log_lvl(struct task_struct *task, struct pt_regs *regs, if (call_trace >= 0) { int unw_ret = 0; struct unwind_frame_info info; + struct ops_and_data oad = { .ops = ops, .data = data }; if (regs) { if (unwind_init_frame_info(&info, task, regs) == 0) - unw_ret = show_trace_unwind(&info, log_lvl); + unw_ret = dump_trace_unwind(&info, &oad); } else if (task == current) - unw_ret = unwind_init_running(&info, show_trace_unwind, log_lvl); + unw_ret = unwind_init_running(&info, dump_trace_unwind, &oad); else { if (unwind_init_blocked(&info, task) == 0) - unw_ret = show_trace_unwind(&info, log_lvl); + unw_ret = dump_trace_unwind(&info, &oad); } if (unw_ret > 0) { if (call_trace == 1 && !arch_unw_user_mode(&info)) { - print_symbol("DWARF2 unwinder stuck at %s\n", + ops->warning_symbol(data, "DWARF2 unwinder stuck at %s\n", UNW_PC(&info)); if (UNW_SP(&info) >= PAGE_OFFSET) { - printk("Leftover inexact backtrace:\n"); + ops->warning(data, "Leftover inexact backtrace:\n"); stack = (void *)UNW_SP(&info); } else - printk("Full inexact backtrace again:\n"); + ops->warning(data, "Full inexact backtrace again:\n"); } else if (call_trace >= 1) return; else - printk("Full inexact backtrace again:\n"); + ops->warning(data, "Full inexact backtrace again:\n"); } else - printk("Inexact backtrace:\n"); + ops->warning(data, "Inexact backtrace:\n"); + } else if (!stack) { + unsigned long dummy; + stack = &dummy; + if (task && task != current) + stack = (unsigned long *)task->thread.esp; } if (task == current) { @@ -220,15 +224,63 @@ static void show_trace_log_lvl(struct task_struct *task, struct pt_regs *regs, struct thread_info *context; context = (struct thread_info *) ((unsigned long)stack & (~(THREAD_SIZE - 1))); - ebp = print_context_stack(context, stack, ebp, log_lvl); + ebp = print_context_stack(context, stack, ebp, ops, data); + /* Should be after the line below, but somewhere + in early boot context comes out corrupted and we + can't reference it -AK */ + if (ops->stack(data, "IRQ") < 0) + break; stack = (unsigned long*)context->previous_esp; if (!stack) break; - printk("%s =======================\n", log_lvl); } } +EXPORT_SYMBOL(dump_trace); + +static void +print_trace_warning_symbol(void *data, char *msg, unsigned long symbol) +{ + printk(data); + print_symbol(msg, symbol); + printk("\n"); +} + +static void print_trace_warning(void *data, char *msg) +{ + printk("%s%s\n", (char *)data, msg); +} + +static int print_trace_stack(void *data, char *name) +{ + return 0; +} + +/* + * Print one address/symbol entries per line. + */ +static void print_trace_address(void *data, unsigned long addr) +{ + printk("%s [<%08lx>] ", (char *)data, addr); + print_symbol("%s\n", addr); +} + +static struct stacktrace_ops print_trace_ops = { + .warning = print_trace_warning, + .warning_symbol = print_trace_warning_symbol, + .stack = print_trace_stack, + .address = print_trace_address, +}; + +static void +show_trace_log_lvl(struct task_struct *task, struct pt_regs *regs, + unsigned long * stack, char *log_lvl) +{ + dump_trace(task, regs, stack, &print_trace_ops, log_lvl); + printk("%s =======================\n", log_lvl); +} -void show_trace(struct task_struct *task, struct pt_regs *regs, unsigned long * stack) +void show_trace(struct task_struct *task, struct pt_regs *regs, + unsigned long * stack) { show_trace_log_lvl(task, regs, stack, ""); } diff --git a/include/asm-i386/stacktrace.h b/include/asm-i386/stacktrace.h new file mode 100644 index 000000000000..7d1f6a5cbfca --- /dev/null +++ b/include/asm-i386/stacktrace.h @@ -0,0 +1 @@ +#include -- cgit From 950fee84557416a3427dd404a13addc4be7b3e6c Mon Sep 17 00:00:00 2001 From: Andi Kleen Date: Tue, 26 Sep 2006 10:52:34 +0200 Subject: [PATCH] i386: Terminate backtrace fallback early if unwinder stack pointer is zero Cc: jbeulich@novell.com Signed-off-by: Andi Kleen --- arch/i386/kernel/traps.c | 2 ++ 1 file changed, 2 insertions(+) (limited to 'arch/i386/kernel/traps.c') diff --git a/arch/i386/kernel/traps.c b/arch/i386/kernel/traps.c index 4ced4285163b..86fa7e47f301 100644 --- a/arch/i386/kernel/traps.c +++ b/arch/i386/kernel/traps.c @@ -197,6 +197,8 @@ void dump_trace(struct task_struct *task, struct pt_regs *regs, if (UNW_SP(&info) >= PAGE_OFFSET) { ops->warning(data, "Leftover inexact backtrace:\n"); stack = (void *)UNW_SP(&info); + if (!stack) + return; } else ops->warning(data, "Full inexact backtrace again:\n"); } else if (call_trace >= 1) -- cgit From a32cf3975bed3b84491f8ffeb24abe8c45d86ab0 Mon Sep 17 00:00:00 2001 From: Andi Kleen Date: Tue, 26 Sep 2006 10:52:34 +0200 Subject: [PATCH] i386: Get ebp from unwinder state when continuing fallback backtrace Cc: jbeulich@novell.com Signed-off-by: Andi Kleen --- arch/i386/kernel/traps.c | 22 ++++++++++++++-------- include/asm-i386/unwind.h | 3 +++ 2 files changed, 17 insertions(+), 8 deletions(-) (limited to 'arch/i386/kernel/traps.c') diff --git a/arch/i386/kernel/traps.c b/arch/i386/kernel/traps.c index 86fa7e47f301..bdf949c30c7c 100644 --- a/arch/i386/kernel/traps.c +++ b/arch/i386/kernel/traps.c @@ -171,7 +171,7 @@ void dump_trace(struct task_struct *task, struct pt_regs *regs, unsigned long *stack, struct stacktrace_ops *ops, void *data) { - unsigned long ebp; + unsigned long ebp = 0; if (!task) task = current; @@ -199,6 +199,7 @@ void dump_trace(struct task_struct *task, struct pt_regs *regs, stack = (void *)UNW_SP(&info); if (!stack) return; + ebp = UNW_FP(&info); } else ops->warning(data, "Full inexact backtrace again:\n"); } else if (call_trace >= 1) @@ -207,20 +208,25 @@ void dump_trace(struct task_struct *task, struct pt_regs *regs, ops->warning(data, "Full inexact backtrace again:\n"); } else ops->warning(data, "Inexact backtrace:\n"); - } else if (!stack) { + } + if (!stack) { unsigned long dummy; stack = &dummy; if (task && task != current) stack = (unsigned long *)task->thread.esp; } - if (task == current) { - /* Grab ebp right from our regs */ - asm ("movl %%ebp, %0" : "=r" (ebp) : ); - } else { - /* ebp is the last reg pushed by switch_to */ - ebp = *(unsigned long *) task->thread.esp; +#ifdef CONFIG_FRAME_POINTER + if (!ebp) { + if (task == current) { + /* Grab ebp right from our regs */ + asm ("movl %%ebp, %0" : "=r" (ebp) : ); + } else { + /* ebp is the last reg pushed by switch_to */ + ebp = *(unsigned long *) task->thread.esp; + } } +#endif while (1) { struct thread_info *context; diff --git a/include/asm-i386/unwind.h b/include/asm-i386/unwind.h index 4c1a0b968569..f0ac399bae3c 100644 --- a/include/asm-i386/unwind.h +++ b/include/asm-i386/unwind.h @@ -28,6 +28,8 @@ struct unwind_frame_info #define FRAME_LINK_OFFSET 0 #define STACK_BOTTOM(tsk) STACK_LIMIT((tsk)->thread.esp0) #define STACK_TOP(tsk) ((tsk)->thread.esp0) +#else +#define UNW_FP(frame) ((void)(frame), 0) #endif #define STACK_LIMIT(ptr) (((ptr) - 1) & ~(THREAD_SIZE - 1)) @@ -88,6 +90,7 @@ static inline int arch_unw_user_mode(const struct unwind_frame_info *info) #define UNW_PC(frame) ((void)(frame), 0) #define UNW_SP(frame) ((void)(frame), 0) +#define UNW_FP(frame) ((void)(frame), 0) static inline int arch_unw_user_mode(const void *info) { -- cgit From 522e93e3fcdbf00ba85c72fde6df28cfc0486a65 Mon Sep 17 00:00:00 2001 From: Rusty Russell Date: Tue, 26 Sep 2006 10:52:35 +0200 Subject: [PATCH] i386: Descriptor and trap table cleanups. The implementation comes from Zach's [RFC, PATCH 10/24] i386 Vmi descriptor changes: Descriptor and trap table cleanups. Add cleanly written accessors for IDT and GDT gates so the subarch may override them. Note that this allows the hypervisor to transparently tweak the DPL of the descriptors as well as the RPL of segments in those descriptors, with no unnecessary kernel code modification. It also allows the hypervisor implementation of the VMI to tweak the gates, allowing for custom exception frames or extra layers of indirection above the guest fault / IRQ handlers. Signed-off-by: Zachary Amsden Signed-off-by: Rusty Russell Signed-off-by: Andi Kleen --- arch/i386/kernel/traps.c | 24 ++-------- include/asm-i386/desc.h | 121 +++++++++++++++++++++++++++++------------------ 2 files changed, 81 insertions(+), 64 deletions(-) (limited to 'arch/i386/kernel/traps.c') diff --git a/arch/i386/kernel/traps.c b/arch/i386/kernel/traps.c index bdf949c30c7c..00d643f3de41 100644 --- a/arch/i386/kernel/traps.c +++ b/arch/i386/kernel/traps.c @@ -1165,20 +1165,6 @@ void __init trap_init_f00f_bug(void) } #endif -#define _set_gate(gate_addr,type,dpl,addr,seg) \ -do { \ - int __d0, __d1; \ - __asm__ __volatile__ ("movw %%dx,%%ax\n\t" \ - "movw %4,%%dx\n\t" \ - "movl %%eax,%0\n\t" \ - "movl %%edx,%1" \ - :"=m" (*((long *) (gate_addr))), \ - "=m" (*(1+(long *) (gate_addr))), "=&a" (__d0), "=&d" (__d1) \ - :"i" ((short) (0x8000+(dpl<<13)+(type<<8))), \ - "3" ((char *) (addr)),"2" ((seg) << 16)); \ -} while (0) - - /* * This needs to use 'idt_table' rather than 'idt', and * thus use the _nonmapped_ version of the IDT, as the @@ -1187,7 +1173,7 @@ do { \ */ void set_intr_gate(unsigned int n, void *addr) { - _set_gate(idt_table+n,14,0,addr,__KERNEL_CS); + _set_gate(n, DESCTYPE_INT, addr, __KERNEL_CS); } /* @@ -1195,22 +1181,22 @@ void set_intr_gate(unsigned int n, void *addr) */ static inline void set_system_intr_gate(unsigned int n, void *addr) { - _set_gate(idt_table+n, 14, 3, addr, __KERNEL_CS); + _set_gate(n, DESCTYPE_INT | DESCTYPE_DPL3, addr, __KERNEL_CS); } static void __init set_trap_gate(unsigned int n, void *addr) { - _set_gate(idt_table+n,15,0,addr,__KERNEL_CS); + _set_gate(n, DESCTYPE_TRAP, addr, __KERNEL_CS); } static void __init set_system_gate(unsigned int n, void *addr) { - _set_gate(idt_table+n,15,3,addr,__KERNEL_CS); + _set_gate(n, DESCTYPE_TRAP | DESCTYPE_DPL3, addr, __KERNEL_CS); } static void __init set_task_gate(unsigned int n, unsigned int gdt_entry) { - _set_gate(idt_table+n,5,0,0,(gdt_entry<<3)); + _set_gate(n, DESCTYPE_TASK, (void *)0, (gdt_entry<<3)); } diff --git a/include/asm-i386/desc.h b/include/asm-i386/desc.h index 89b8b82c82b3..5db9e96e8dc1 100644 --- a/include/asm-i386/desc.h +++ b/include/asm-i386/desc.h @@ -33,50 +33,99 @@ static inline struct desc_struct *get_cpu_gdt_table(unsigned int cpu) return (struct desc_struct *)per_cpu(cpu_gdt_descr, cpu).address; } +/* + * This is the ldt that every process will get unless we need + * something other than this. + */ +extern struct desc_struct default_ldt[]; +extern struct desc_struct idt_table[]; +extern void set_intr_gate(unsigned int irq, void * addr); + +static inline void pack_descriptor(__u32 *a, __u32 *b, + unsigned long base, unsigned long limit, unsigned char type, unsigned char flags) +{ + *a = ((base & 0xffff) << 16) | (limit & 0xffff); + *b = (base & 0xff000000) | ((base & 0xff0000) >> 16) | + ((type & 0xff) << 8) | ((flags & 0xf) << 12); +} + +static inline void pack_gate(__u32 *a, __u32 *b, + unsigned long base, unsigned short seg, unsigned char type, unsigned char flags) +{ + *a = (seg << 16) | (base & 0xffff); + *b = (base & 0xffff0000) | ((type & 0xff) << 8) | (flags & 0xff); +} + +#define DESCTYPE_LDT 0x82 /* present, system, DPL-0, LDT */ +#define DESCTYPE_TSS 0x89 /* present, system, DPL-0, 32-bit TSS */ +#define DESCTYPE_TASK 0x85 /* present, system, DPL-0, task gate */ +#define DESCTYPE_INT 0x8e /* present, system, DPL-0, interrupt gate */ +#define DESCTYPE_TRAP 0x8f /* present, system, DPL-0, trap gate */ +#define DESCTYPE_DPL3 0x60 /* DPL-3 */ +#define DESCTYPE_S 0x10 /* !system */ + #define load_TR_desc() __asm__ __volatile__("ltr %w0"::"q" (GDT_ENTRY_TSS*8)) #define load_LDT_desc() __asm__ __volatile__("lldt %w0"::"q" (GDT_ENTRY_LDT*8)) #define load_gdt(dtr) __asm__ __volatile("lgdt %0"::"m" (*dtr)) #define load_idt(dtr) __asm__ __volatile("lidt %0"::"m" (*dtr)) -#define load_tr(tr) __asm__ __volatile("ltr %0"::"mr" (tr)) -#define load_ldt(ldt) __asm__ __volatile("lldt %0"::"mr" (ldt)) +#define load_tr(tr) __asm__ __volatile("ltr %0"::"m" (tr)) +#define load_ldt(ldt) __asm__ __volatile("lldt %0"::"m" (ldt)) #define store_gdt(dtr) __asm__ ("sgdt %0":"=m" (*dtr)) #define store_idt(dtr) __asm__ ("sidt %0":"=m" (*dtr)) -#define store_tr(tr) __asm__ ("str %0":"=mr" (tr)) -#define store_ldt(ldt) __asm__ ("sldt %0":"=mr" (ldt)) +#define store_tr(tr) __asm__ ("str %0":"=m" (tr)) +#define store_ldt(ldt) __asm__ ("sldt %0":"=m" (ldt)) -/* - * This is the ldt that every process will get unless we need - * something other than this. - */ -extern struct desc_struct default_ldt[]; -extern void set_intr_gate(unsigned int irq, void * addr); +#if TLS_SIZE != 24 +# error update this code. +#endif -#define _set_tssldt_desc(n,addr,limit,type) \ -__asm__ __volatile__ ("movw %w3,0(%2)\n\t" \ - "movw %w1,2(%2)\n\t" \ - "rorl $16,%1\n\t" \ - "movb %b1,4(%2)\n\t" \ - "movb %4,5(%2)\n\t" \ - "movb $0,6(%2)\n\t" \ - "movb %h1,7(%2)\n\t" \ - "rorl $16,%1" \ - : "=m"(*(n)) : "q" (addr), "r"(n), "ir"(limit), "i"(type)) - -static inline void __set_tss_desc(unsigned int cpu, unsigned int entry, void *addr) +static inline void load_TLS(struct thread_struct *t, unsigned int cpu) { - _set_tssldt_desc(&get_cpu_gdt_table(cpu)[entry], (int)addr, - offsetof(struct tss_struct, __cacheline_filler) - 1, 0x89); +#define C(i) get_cpu_gdt_table(cpu)[GDT_ENTRY_TLS_MIN + i] = t->tls_array[i] + C(0); C(1); C(2); +#undef C } -#define set_tss_desc(cpu,addr) __set_tss_desc(cpu, GDT_ENTRY_TSS, addr) +static inline void write_dt_entry(void *dt, int entry, __u32 entry_a, __u32 entry_b) +{ + __u32 *lp = (__u32 *)((char *)dt + entry*8); + *lp = entry_a; + *(lp+1) = entry_b; +} + +#define write_ldt_entry(dt, entry, a, b) write_dt_entry(dt, entry, a, b) +#define write_gdt_entry(dt, entry, a, b) write_dt_entry(dt, entry, a, b) +#define write_idt_entry(dt, entry, a, b) write_dt_entry(dt, entry, a, b) + +static inline void _set_gate(int gate, unsigned int type, void *addr, unsigned short seg) +{ + __u32 a, b; + pack_gate(&a, &b, (unsigned long)addr, seg, type, 0); + write_idt_entry(idt_table, gate, a, b); +} -static inline void set_ldt_desc(unsigned int cpu, void *addr, unsigned int size) +static inline void __set_tss_desc(unsigned int cpu, unsigned int entry, const void *addr) { - _set_tssldt_desc(&get_cpu_gdt_table(cpu)[GDT_ENTRY_LDT], (int)addr, ((size << 3)-1), 0x82); + __u32 a, b; + pack_descriptor(&a, &b, (unsigned long)addr, + offsetof(struct tss_struct, __cacheline_filler) - 1, + DESCTYPE_TSS, 0); + write_gdt_entry(get_cpu_gdt_table(cpu), entry, a, b); } +static inline void set_ldt_desc(unsigned int cpu, void *addr, unsigned int entries) +{ + __u32 a, b; + pack_descriptor(&a, &b, (unsigned long)addr, + entries * sizeof(struct desc_struct) - 1, + DESCTYPE_LDT, 0); + write_gdt_entry(get_cpu_gdt_table(cpu), GDT_ENTRY_LDT, a, b); +} + +#define set_tss_desc(cpu,addr) __set_tss_desc(cpu, GDT_ENTRY_TSS, addr) + #define LDT_entry_a(info) \ ((((info)->base_addr & 0x0000ffff) << 16) | ((info)->limit & 0x0ffff)) @@ -102,24 +151,6 @@ static inline void set_ldt_desc(unsigned int cpu, void *addr, unsigned int size) (info)->seg_not_present == 1 && \ (info)->useable == 0 ) -static inline void write_ldt_entry(void *ldt, int entry, __u32 entry_a, __u32 entry_b) -{ - __u32 *lp = (__u32 *)((char *)ldt + entry*8); - *lp = entry_a; - *(lp+1) = entry_b; -} - -#if TLS_SIZE != 24 -# error update this code. -#endif - -static inline void load_TLS(struct thread_struct *t, unsigned int cpu) -{ -#define C(i) get_cpu_gdt_table(cpu)[GDT_ENTRY_TLS_MIN + i] = t->tls_array[i] - C(0); C(1); C(2); -#undef C -} - static inline void clear_LDT(void) { int cpu = get_cpu(); -- cgit From 06039754d775d3e48e4a292e4f353321205eff53 Mon Sep 17 00:00:00 2001 From: Fernando Luis Vázquez Cao Date: Tue, 26 Sep 2006 10:52:36 +0200 Subject: [PATCH] i386: Disallow kprobes on NMI handlers A kprobe executes IRET early and that could cause NMI recursion and stack corruption. Note: This problem was originally spotted and solved by Andi Kleen in the x86_64 architecture. This patch is an adaption of his patch for i386. AK: Merged with current code which was a bit different. AK: Removed printk in nmi handler that shouldn't be there in the first time AK: Added missing include. AK: added KPROBES_END Signed-off-by: Fernando Vazquez Signed-off-by: Andi Kleen --- arch/i386/kernel/entry.S | 3 ++- arch/i386/kernel/nmi.c | 6 +++--- arch/i386/kernel/traps.c | 15 +++++++++------ 3 files changed, 14 insertions(+), 10 deletions(-) (limited to 'arch/i386/kernel/traps.c') diff --git a/arch/i386/kernel/entry.S b/arch/i386/kernel/entry.S index dede506e5bd0..0928f70639aa 100644 --- a/arch/i386/kernel/entry.S +++ b/arch/i386/kernel/entry.S @@ -729,7 +729,7 @@ KPROBE_END(debug) * check whether we got an NMI on the debug path where the debug * fault happened on the sysenter path. */ -ENTRY(nmi) +KPROBE_ENTRY(nmi) RING0_INT_FRAME pushl %eax CFI_ADJUST_CFA_OFFSET 4 @@ -805,6 +805,7 @@ nmi_16bit_stack: .align 4 .long 1b,iret_exc .previous +KPROBE_END(nmi) KPROBE_ENTRY(int3) RING0_INT_FRAME diff --git a/arch/i386/kernel/nmi.c b/arch/i386/kernel/nmi.c index 7b9a053effa3..dbda706fdd14 100644 --- a/arch/i386/kernel/nmi.c +++ b/arch/i386/kernel/nmi.c @@ -22,6 +22,7 @@ #include #include #include +#include #include #include @@ -882,7 +883,7 @@ EXPORT_SYMBOL(touch_nmi_watchdog); extern void die_nmi(struct pt_regs *, const char *msg); -int nmi_watchdog_tick (struct pt_regs * regs, unsigned reason) +__kprobes int nmi_watchdog_tick(struct pt_regs * regs, unsigned reason) { /* @@ -962,8 +963,7 @@ int nmi_watchdog_tick (struct pt_regs * regs, unsigned reason) * This matches the old behaviour. */ rc = 1; - } else - printk(KERN_WARNING "Unknown enabled NMI hardware?!\n"); + } } done: return rc; diff --git a/arch/i386/kernel/traps.c b/arch/i386/kernel/traps.c index 00d643f3de41..5c0f4960c67d 100644 --- a/arch/i386/kernel/traps.c +++ b/arch/i386/kernel/traps.c @@ -689,7 +689,8 @@ gp_in_kernel: } } -static void mem_parity_error(unsigned char reason, struct pt_regs * regs) +static __kprobes void +mem_parity_error(unsigned char reason, struct pt_regs * regs) { printk(KERN_EMERG "Uhhuh. NMI received for unknown reason %02x on " "CPU %d.\n", reason, smp_processor_id()); @@ -704,7 +705,8 @@ static void mem_parity_error(unsigned char reason, struct pt_regs * regs) clear_mem_error(reason); } -static void io_check_error(unsigned char reason, struct pt_regs * regs) +static __kprobes void +io_check_error(unsigned char reason, struct pt_regs * regs) { unsigned long i; @@ -720,7 +722,8 @@ static void io_check_error(unsigned char reason, struct pt_regs * regs) outb(reason, 0x61); } -static void unknown_nmi_error(unsigned char reason, struct pt_regs * regs) +static __kprobes void +unknown_nmi_error(unsigned char reason, struct pt_regs * regs) { #ifdef CONFIG_MCA /* Might actually be able to figure out what the guilty party @@ -741,7 +744,7 @@ static void unknown_nmi_error(unsigned char reason, struct pt_regs * regs) static DEFINE_SPINLOCK(nmi_print_lock); -void die_nmi (struct pt_regs *regs, const char *msg) +void __kprobes die_nmi(struct pt_regs *regs, const char *msg) { if (notify_die(DIE_NMIWATCHDOG, msg, regs, 0, 2, SIGINT) == NOTIFY_STOP) @@ -773,7 +776,7 @@ void die_nmi (struct pt_regs *regs, const char *msg) do_exit(SIGSEGV); } -static void default_do_nmi(struct pt_regs * regs) +static __kprobes void default_do_nmi(struct pt_regs * regs) { unsigned char reason = 0; @@ -811,7 +814,7 @@ static void default_do_nmi(struct pt_regs * regs) reassert_nmi(); } -fastcall void do_nmi(struct pt_regs * regs, long error_code) +fastcall __kprobes void do_nmi(struct pt_regs * regs, long error_code) { int cpu; -- cgit From f354b3a92af9b132b188b9c8ebbfb74de699926d Mon Sep 17 00:00:00 2001 From: Dave Jones Date: Tue, 26 Sep 2006 10:52:41 +0200 Subject: [PATCH] i386: Split multi-line printk in oops output. Sometimes, bug reports come in where we've had an oops, and the only record we have is what the reporter saw on screen shortly before the system locked up completely. Unfortunatly, syslog only prints lines beginning with KERN_EMERG to the console, so some lines get lost. An example of this can be seen at https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=203723 Some of this information isn't vital to diagnosis, but some parts are useful, such as the tainted flag. Signed-off-by: Dave Jones Signed-off-by: Andi Kleen --- arch/i386/kernel/traps.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) (limited to 'arch/i386/kernel/traps.c') diff --git a/arch/i386/kernel/traps.c b/arch/i386/kernel/traps.c index 5c0f4960c67d..c7adb076e811 100644 --- a/arch/i386/kernel/traps.c +++ b/arch/i386/kernel/traps.c @@ -351,8 +351,9 @@ void show_registers(struct pt_regs *regs) ss = regs->xss & 0xffff; } print_modules(); - printk(KERN_EMERG "CPU: %d\nEIP: %04x:[<%08lx>] %s VLI\n" - "EFLAGS: %08lx (%s %.*s) \n", + printk(KERN_EMERG "CPU: %d\n" + KERN_EMERG "EIP: %04x:[<%08lx>] %s VLI\n" + KERN_EMERG "EFLAGS: %08lx (%s %.*s)\n", smp_processor_id(), 0xffff & regs->xcs, regs->eip, print_tainted(), regs->eflags, system_utsname.release, (int)strcspn(system_utsname.version, " "), -- cgit From 99325326a57b6a56595bb097655bee9fd27d77b0 Mon Sep 17 00:00:00 2001 From: Chuck Ebbert <76306.1226@compuserve.com> Date: Mon, 25 Sep 2006 23:32:19 -0700 Subject: [PATCH] i386: show_registers(): try harder to print failing code show_registers() tries to dump failing code starting 43 bytes before the offending instruction, but this address can be bad, for example in a device driver where the failing instruction is less than 43 bytes from the start of the driver's code. When that happens, try to dump code starting at the failing instruction instead of printing no code at all. Signed-off-by: Chuck Ebbert <76306.1226@compuserve.com> Cc: Andi Kleen Cc: Keith Owens Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/i386/kernel/traps.c | 11 ++++++++--- 1 file changed, 8 insertions(+), 3 deletions(-) (limited to 'arch/i386/kernel/traps.c') diff --git a/arch/i386/kernel/traps.c b/arch/i386/kernel/traps.c index 7e9edafffd8a..4fcc6690be99 100644 --- a/arch/i386/kernel/traps.c +++ b/arch/i386/kernel/traps.c @@ -313,6 +313,8 @@ void show_registers(struct pt_regs *regs) */ if (in_kernel) { u8 __user *eip; + int code_bytes = 64; + unsigned char c; printk("\n" KERN_EMERG "Stack: "); show_stack_log_lvl(NULL, regs, (unsigned long *)esp, KERN_EMERG); @@ -320,9 +322,12 @@ void show_registers(struct pt_regs *regs) printk(KERN_EMERG "Code: "); eip = (u8 __user *)regs->eip - 43; - for (i = 0; i < 64; i++, eip++) { - unsigned char c; - + if (eip < (u8 __user *)PAGE_OFFSET || __get_user(c, eip)) { + /* try starting at EIP */ + eip = (u8 __user *)regs->eip; + code_bytes = 32; + } + for (i = 0; i < code_bytes; i++, eip++) { if (eip < (u8 __user *)PAGE_OFFSET || __get_user(c, eip)) { printk(" Bad EIP value."); break; -- cgit From 1e2af92e089d4c845248844a3e7ee06ce889af46 Mon Sep 17 00:00:00 2001 From: Andrew Morton Date: Wed, 27 Sep 2006 01:51:15 -0700 Subject: [PATCH] x86: use probe_kernel_address in handle_BUG() Avoid possible deadlock on a BUG() inside down_write(mmap_sem). The deadlock can only occur if something has gone horridly wrong, because a fault here shouldn't happen. Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/i386/kernel/traps.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) (limited to 'arch/i386/kernel/traps.c') diff --git a/arch/i386/kernel/traps.c b/arch/i386/kernel/traps.c index 21aa1cd57773..a13037fe0ee3 100644 --- a/arch/i386/kernel/traps.c +++ b/arch/i386/kernel/traps.c @@ -28,6 +28,7 @@ #include #include #include +#include #ifdef CONFIG_EISA #include @@ -40,7 +41,6 @@ #include #include -#include #include #include #include @@ -409,7 +409,7 @@ static void handle_BUG(struct pt_regs *regs) if (eip < PAGE_OFFSET) return; - if (__get_user(ud2, (unsigned short __user *)eip)) + if (probe_kernel_address((unsigned short __user *)eip, ud2)) return; if (ud2 != 0x0b0f) return; @@ -422,7 +422,8 @@ static void handle_BUG(struct pt_regs *regs) char *file; char c; - if (__get_user(line, (unsigned short __user *)(eip + 2))) + if (probe_kernel_address((unsigned short __user *)(eip + 2), + line)) break; if (__get_user(file, (char * __user *)(eip + 4)) || (unsigned long)file < PAGE_OFFSET || __get_user(c, file)) -- cgit