diff options
author | 2011-01-06 10:17:26 -0800 | |
---|---|---|
committer | 2011-01-06 10:17:26 -0800 | |
commit | 28d9bfc37c861aa9c8386dff1ac7e9a10e5c5162 (patch) | |
tree | 85bcc2db18ff20e380a40aba375e70d14c2671b4 /kernel | |
parent | Merge branch 'core-futexes-for-linus' of git://git.kernel.org/pub/scm/linux/k... (diff) | |
parent | tools, perf: Documentation for the power events API (diff) | |
download | linux-28d9bfc37c861aa9c8386dff1ac7e9a10e5c5162.tar.gz linux-28d9bfc37c861aa9c8386dff1ac7e9a10e5c5162.tar.bz2 linux-28d9bfc37c861aa9c8386dff1ac7e9a10e5c5162.zip |
Merge branch 'perf-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip
* 'perf-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip: (146 commits)
tools, perf: Documentation for the power events API
perf: Add calls to suspend trace point
perf script: Make some lists static
perf script: Use the default lost event handler
perf session: Warn about errors when processing pipe events too
perf tools: Fix perf_event.h header usage
perf test: Clarify some error reports in the open syscall test
x86, NMI: Add touch_nmi_watchdog to io_check_error delay
x86: Avoid calling arch_trigger_all_cpu_backtrace() at the same time
x86: Only call smp_processor_id in non-preempt cases
perf timechart: Adjust perf timechart to the new power events
perf: Clean up power events by introducing new, more generic ones
perf: Do not export power_frequency, but power_start event
perf test: Add test for counting open syscalls
perf evsel: Auto allocate resources needed for some methods
perf evsel: Use {cpu,thread}_map to shorten list of parameters
perf tools: Refactor all_tids to hold nr and the map
perf tools: Refactor cpumap to hold nr and the map
perf evsel: Introduce per cpu and per thread open helpers
perf evsel: Steal the counter reading routines from stat
...
Diffstat (limited to 'kernel')
-rw-r--r-- | kernel/hw_breakpoint.c | 2 | ||||
-rw-r--r-- | kernel/kprobes.c | 565 | ||||
-rw-r--r-- | kernel/perf_event.c | 573 | ||||
-rw-r--r-- | kernel/power/suspend.c | 3 | ||||
-rw-r--r-- | kernel/sched.c | 2 | ||||
-rw-r--r-- | kernel/sysctl.c | 16 | ||||
-rw-r--r-- | kernel/sysctl_binary.c | 1 | ||||
-rw-r--r-- | kernel/trace/Kconfig | 15 | ||||
-rw-r--r-- | kernel/trace/power-traces.c | 5 | ||||
-rw-r--r-- | kernel/trace/trace_event_perf.c | 31 | ||||
-rw-r--r-- | kernel/trace/trace_events.c | 6 | ||||
-rw-r--r-- | kernel/trace/trace_export.c | 14 | ||||
-rw-r--r-- | kernel/watchdog.c | 9 |
13 files changed, 894 insertions, 348 deletions
diff --git a/kernel/hw_breakpoint.c b/kernel/hw_breakpoint.c index e5325825aeb6..086adf25a55e 100644 --- a/kernel/hw_breakpoint.c +++ b/kernel/hw_breakpoint.c @@ -641,7 +641,7 @@ int __init init_hw_breakpoint(void) constraints_initialized = 1; - perf_pmu_register(&perf_breakpoint); + perf_pmu_register(&perf_breakpoint, "breakpoint", PERF_TYPE_BREAKPOINT); return register_die_notifier(&hw_breakpoint_exceptions_nb); diff --git a/kernel/kprobes.c b/kernel/kprobes.c index 9737a76e106f..7663e5df0e6f 100644 --- a/kernel/kprobes.c +++ b/kernel/kprobes.c @@ -354,13 +354,20 @@ static inline int kprobe_aggrprobe(struct kprobe *p) return p->pre_handler == aggr_pre_handler; } +/* Return true(!0) if the kprobe is unused */ +static inline int kprobe_unused(struct kprobe *p) +{ + return kprobe_aggrprobe(p) && kprobe_disabled(p) && + list_empty(&p->list); +} + /* * Keep all fields in the kprobe consistent */ -static inline void copy_kprobe(struct kprobe *old_p, struct kprobe *p) +static inline void copy_kprobe(struct kprobe *ap, struct kprobe *p) { - memcpy(&p->opcode, &old_p->opcode, sizeof(kprobe_opcode_t)); - memcpy(&p->ainsn, &old_p->ainsn, sizeof(struct arch_specific_insn)); + memcpy(&p->opcode, &ap->opcode, sizeof(kprobe_opcode_t)); + memcpy(&p->ainsn, &ap->ainsn, sizeof(struct arch_specific_insn)); } #ifdef CONFIG_OPTPROBES @@ -384,6 +391,17 @@ void __kprobes opt_pre_handler(struct kprobe *p, struct pt_regs *regs) } } +/* Free optimized instructions and optimized_kprobe */ +static __kprobes void free_aggr_kprobe(struct kprobe *p) +{ + struct optimized_kprobe *op; + + op = container_of(p, struct optimized_kprobe, kp); + arch_remove_optimized_kprobe(op); + arch_remove_kprobe(p); + kfree(op); +} + /* Return true(!0) if the kprobe is ready for optimization. */ static inline int kprobe_optready(struct kprobe *p) { @@ -397,6 +415,33 @@ static inline int kprobe_optready(struct kprobe *p) return 0; } +/* Return true(!0) if the kprobe is disarmed. Note: p must be on hash list */ +static inline int kprobe_disarmed(struct kprobe *p) +{ + struct optimized_kprobe *op; + + /* If kprobe is not aggr/opt probe, just return kprobe is disabled */ + if (!kprobe_aggrprobe(p)) + return kprobe_disabled(p); + + op = container_of(p, struct optimized_kprobe, kp); + + return kprobe_disabled(p) && list_empty(&op->list); +} + +/* Return true(!0) if the probe is queued on (un)optimizing lists */ +static int __kprobes kprobe_queued(struct kprobe *p) +{ + struct optimized_kprobe *op; + + if (kprobe_aggrprobe(p)) { + op = container_of(p, struct optimized_kprobe, kp); + if (!list_empty(&op->list)) + return 1; + } + return 0; +} + /* * Return an optimized kprobe whose optimizing code replaces * instructions including addr (exclude breakpoint). @@ -422,30 +467,23 @@ static struct kprobe *__kprobes get_optimized_kprobe(unsigned long addr) /* Optimization staging list, protected by kprobe_mutex */ static LIST_HEAD(optimizing_list); +static LIST_HEAD(unoptimizing_list); static void kprobe_optimizer(struct work_struct *work); static DECLARE_DELAYED_WORK(optimizing_work, kprobe_optimizer); +static DECLARE_COMPLETION(optimizer_comp); #define OPTIMIZE_DELAY 5 -/* Kprobe jump optimizer */ -static __kprobes void kprobe_optimizer(struct work_struct *work) +/* + * Optimize (replace a breakpoint with a jump) kprobes listed on + * optimizing_list. + */ +static __kprobes void do_optimize_kprobes(void) { - struct optimized_kprobe *op, *tmp; - - /* Lock modules while optimizing kprobes */ - mutex_lock(&module_mutex); - mutex_lock(&kprobe_mutex); - if (kprobes_all_disarmed || !kprobes_allow_optimization) - goto end; - - /* - * Wait for quiesence period to ensure all running interrupts - * are done. Because optprobe may modify multiple instructions - * there is a chance that Nth instruction is interrupted. In that - * case, running interrupt can return to 2nd-Nth byte of jump - * instruction. This wait is for avoiding it. - */ - synchronize_sched(); + /* Optimization never be done when disarmed */ + if (kprobes_all_disarmed || !kprobes_allow_optimization || + list_empty(&optimizing_list)) + return; /* * The optimization/unoptimization refers online_cpus via @@ -459,17 +497,111 @@ static __kprobes void kprobe_optimizer(struct work_struct *work) */ get_online_cpus(); mutex_lock(&text_mutex); - list_for_each_entry_safe(op, tmp, &optimizing_list, list) { - WARN_ON(kprobe_disabled(&op->kp)); - if (arch_optimize_kprobe(op) < 0) - op->kp.flags &= ~KPROBE_FLAG_OPTIMIZED; - list_del_init(&op->list); + arch_optimize_kprobes(&optimizing_list); + mutex_unlock(&text_mutex); + put_online_cpus(); +} + +/* + * Unoptimize (replace a jump with a breakpoint and remove the breakpoint + * if need) kprobes listed on unoptimizing_list. + */ +static __kprobes void do_unoptimize_kprobes(struct list_head *free_list) +{ + struct optimized_kprobe *op, *tmp; + + /* Unoptimization must be done anytime */ + if (list_empty(&unoptimizing_list)) + return; + + /* Ditto to do_optimize_kprobes */ + get_online_cpus(); + mutex_lock(&text_mutex); + arch_unoptimize_kprobes(&unoptimizing_list, free_list); + /* Loop free_list for disarming */ + list_for_each_entry_safe(op, tmp, free_list, list) { + /* Disarm probes if marked disabled */ + if (kprobe_disabled(&op->kp)) + arch_disarm_kprobe(&op->kp); + if (kprobe_unused(&op->kp)) { + /* + * Remove unused probes from hash list. After waiting + * for synchronization, these probes are reclaimed. + * (reclaiming is done by do_free_cleaned_kprobes.) + */ + hlist_del_rcu(&op->kp.hlist); + } else + list_del_init(&op->list); } mutex_unlock(&text_mutex); put_online_cpus(); -end: +} + +/* Reclaim all kprobes on the free_list */ +static __kprobes void do_free_cleaned_kprobes(struct list_head *free_list) +{ + struct optimized_kprobe *op, *tmp; + + list_for_each_entry_safe(op, tmp, free_list, list) { + BUG_ON(!kprobe_unused(&op->kp)); + list_del_init(&op->list); + free_aggr_kprobe(&op->kp); + } +} + +/* Start optimizer after OPTIMIZE_DELAY passed */ +static __kprobes void kick_kprobe_optimizer(void) +{ + if (!delayed_work_pending(&optimizing_work)) + schedule_delayed_work(&optimizing_work, OPTIMIZE_DELAY); +} + +/* Kprobe jump optimizer */ +static __kprobes void kprobe_optimizer(struct work_struct *work) +{ + LIST_HEAD(free_list); + + /* Lock modules while optimizing kprobes */ + mutex_lock(&module_mutex); + mutex_lock(&kprobe_mutex); + + /* + * Step 1: Unoptimize kprobes and collect cleaned (unused and disarmed) + * kprobes before waiting for quiesence period. + */ + do_unoptimize_kprobes(&free_list); + + /* + * Step 2: Wait for quiesence period to ensure all running interrupts + * are done. Because optprobe may modify multiple instructions + * there is a chance that Nth instruction is interrupted. In that + * case, running interrupt can return to 2nd-Nth byte of jump + * instruction. This wait is for avoiding it. + */ + synchronize_sched(); + + /* Step 3: Optimize kprobes after quiesence period */ + do_optimize_kprobes(); + + /* Step 4: Free cleaned kprobes after quiesence period */ + do_free_cleaned_kprobes(&free_list); + mutex_unlock(&kprobe_mutex); mutex_unlock(&module_mutex); + + /* Step 5: Kick optimizer again if needed */ + if (!list_empty(&optimizing_list) || !list_empty(&unoptimizing_list)) + kick_kprobe_optimizer(); + else + /* Wake up all waiters */ + complete_all(&optimizer_comp); +} + +/* Wait for completing optimization and unoptimization */ +static __kprobes void wait_for_kprobe_optimizer(void) +{ + if (delayed_work_pending(&optimizing_work)) + wait_for_completion(&optimizer_comp); } /* Optimize kprobe if p is ready to be optimized */ @@ -495,42 +627,99 @@ static __kprobes void optimize_kprobe(struct kprobe *p) /* Check if it is already optimized. */ if (op->kp.flags & KPROBE_FLAG_OPTIMIZED) return; - op->kp.flags |= KPROBE_FLAG_OPTIMIZED; - list_add(&op->list, &optimizing_list); - if (!delayed_work_pending(&optimizing_work)) - schedule_delayed_work(&optimizing_work, OPTIMIZE_DELAY); + + if (!list_empty(&op->list)) + /* This is under unoptimizing. Just dequeue the probe */ + list_del_init(&op->list); + else { + list_add(&op->list, &optimizing_list); + kick_kprobe_optimizer(); + } +} + +/* Short cut to direct unoptimizing */ +static __kprobes void force_unoptimize_kprobe(struct optimized_kprobe *op) +{ + get_online_cpus(); + arch_unoptimize_kprobe(op); + put_online_cpus(); + if (kprobe_disabled(&op->kp)) + arch_disarm_kprobe(&op->kp); } /* Unoptimize a kprobe if p is optimized */ -static __kprobes void unoptimize_kprobe(struct kprobe *p) +static __kprobes void unoptimize_kprobe(struct kprobe *p, bool force) { struct optimized_kprobe *op; - if ((p->flags & KPROBE_FLAG_OPTIMIZED) && kprobe_aggrprobe(p)) { - op = container_of(p, struct optimized_kprobe, kp); - if (!list_empty(&op->list)) - /* Dequeue from the optimization queue */ + if (!kprobe_aggrprobe(p) || kprobe_disarmed(p)) + return; /* This is not an optprobe nor optimized */ + + op = container_of(p, struct optimized_kprobe, kp); + if (!kprobe_optimized(p)) { + /* Unoptimized or unoptimizing case */ + if (force && !list_empty(&op->list)) { + /* + * Only if this is unoptimizing kprobe and forced, + * forcibly unoptimize it. (No need to unoptimize + * unoptimized kprobe again :) + */ list_del_init(&op->list); - else - /* Replace jump with break */ - arch_unoptimize_kprobe(op); - op->kp.flags &= ~KPROBE_FLAG_OPTIMIZED; + force_unoptimize_kprobe(op); + } + return; + } + + op->kp.flags &= ~KPROBE_FLAG_OPTIMIZED; + if (!list_empty(&op->list)) { + /* Dequeue from the optimization queue */ + list_del_init(&op->list); + return; + } + /* Optimized kprobe case */ + if (force) + /* Forcibly update the code: this is a special case */ + force_unoptimize_kprobe(op); + else { + list_add(&op->list, &unoptimizing_list); + kick_kprobe_optimizer(); } } +/* Cancel unoptimizing for reusing */ +static void reuse_unused_kprobe(struct kprobe *ap) +{ + struct optimized_kprobe *op; + + BUG_ON(!kprobe_unused(ap)); + /* + * Unused kprobe MUST be on the way of delayed unoptimizing (means + * there is still a relative jump) and disabled. + */ + op = container_of(ap, struct optimized_kprobe, kp); + if (unlikely(list_empty(&op->list))) + printk(KERN_WARNING "Warning: found a stray unused " + "aggrprobe@%p\n", ap->addr); + /* Enable the probe again */ + ap->flags &= ~KPROBE_FLAG_DISABLED; + /* Optimize it again (remove from op->list) */ + BUG_ON(!kprobe_optready(ap)); + optimize_kprobe(ap); +} + /* Remove optimized instructions */ static void __kprobes kill_optimized_kprobe(struct kprobe *p) { struct optimized_kprobe *op; op = container_of(p, struct optimized_kprobe, kp); - if (!list_empty(&op->list)) { - /* Dequeue from the optimization queue */ + if (!list_empty(&op->list)) + /* Dequeue from the (un)optimization queue */ list_del_init(&op->list); - op->kp.flags &= ~KPROBE_FLAG_OPTIMIZED; - } - /* Don't unoptimize, because the target code will be freed. */ + + op->kp.flags &= ~KPROBE_FLAG_OPTIMIZED; + /* Don't touch the code, because it is already freed. */ arch_remove_optimized_kprobe(op); } @@ -543,16 +732,6 @@ static __kprobes void prepare_optimized_kprobe(struct kprobe *p) arch_prepare_optimized_kprobe(op); } -/* Free optimized instructions and optimized_kprobe */ -static __kprobes void free_aggr_kprobe(struct kprobe *p) -{ - struct optimized_kprobe *op; - - op = container_of(p, struct optimized_kprobe, kp); - arch_remove_optimized_kprobe(op); - kfree(op); -} - /* Allocate new optimized_kprobe and try to prepare optimized instructions */ static __kprobes struct kprobe *alloc_aggr_kprobe(struct kprobe *p) { @@ -587,7 +766,8 @@ static __kprobes void try_to_optimize_kprobe(struct kprobe *p) op = container_of(ap, struct optimized_kprobe, kp); if (!arch_prepared_optinsn(&op->optinsn)) { /* If failed to setup optimizing, fallback to kprobe */ - free_aggr_kprobe(ap); + arch_remove_optimized_kprobe(op); + kfree(op); return; } @@ -631,21 +811,16 @@ static void __kprobes unoptimize_all_kprobes(void) return; kprobes_allow_optimization = false; - printk(KERN_INFO "Kprobes globally unoptimized\n"); - get_online_cpus(); /* For avoiding text_mutex deadlock */ - mutex_lock(&text_mutex); for (i = 0; i < KPROBE_TABLE_SIZE; i++) { head = &kprobe_table[i]; hlist_for_each_entry_rcu(p, node, head, hlist) { if (!kprobe_disabled(p)) - unoptimize_kprobe(p); + unoptimize_kprobe(p, false); } } - - mutex_unlock(&text_mutex); - put_online_cpus(); - /* Allow all currently running kprobes to complete */ - synchronize_sched(); + /* Wait for unoptimizing completion */ + wait_for_kprobe_optimizer(); + printk(KERN_INFO "Kprobes globally unoptimized\n"); } int sysctl_kprobes_optimization; @@ -669,44 +844,60 @@ int proc_kprobes_optimization_handler(struct ctl_table *table, int write, } #endif /* CONFIG_SYSCTL */ +/* Put a breakpoint for a probe. Must be called with text_mutex locked */ static void __kprobes __arm_kprobe(struct kprobe *p) { - struct kprobe *old_p; + struct kprobe *_p; /* Check collision with other optimized kprobes */ - old_p = get_optimized_kprobe((unsigned long)p->addr); - if (unlikely(old_p)) - unoptimize_kprobe(old_p); /* Fallback to unoptimized kprobe */ + _p = get_optimized_kprobe((unsigned long)p->addr); + if (unlikely(_p)) + /* Fallback to unoptimized kprobe */ + unoptimize_kprobe(_p, true); arch_arm_kprobe(p); optimize_kprobe(p); /* Try to optimize (add kprobe to a list) */ } -static void __kprobes __disarm_kprobe(struct kprobe *p) +/* Remove the breakpoint of a probe. Must be called with text_mutex locked */ +static void __kprobes __disarm_kprobe(struct kprobe *p, bool reopt) { - struct kprobe *old_p; + struct kprobe *_p; - unoptimize_kprobe(p); /* Try to unoptimize */ - arch_disarm_kprobe(p); + unoptimize_kprobe(p, false); /* Try to unoptimize */ - /* If another kprobe was blocked, optimize it. */ - old_p = get_optimized_kprobe((unsigned long)p->addr); - if (unlikely(old_p)) - optimize_kprobe(old_p); + if (!kprobe_queued(p)) { + arch_disarm_kprobe(p); + /* If another kprobe was blocked, optimize it. */ + _p = get_optimized_kprobe((unsigned long)p->addr); + if (unlikely(_p) && reopt) + optimize_kprobe(_p); + } + /* TODO: reoptimize others after unoptimized this probe */ } #else /* !CONFIG_OPTPROBES */ #define optimize_kprobe(p) do {} while (0) -#define unoptimize_kprobe(p) do {} while (0) +#define unoptimize_kprobe(p, f) do {} while (0) #define kill_optimized_kprobe(p) do {} while (0) #define prepare_optimized_kprobe(p) do {} while (0) #define try_to_optimize_kprobe(p) do {} while (0) #define __arm_kprobe(p) arch_arm_kprobe(p) -#define __disarm_kprobe(p) arch_disarm_kprobe(p) +#define __disarm_kprobe(p, o) arch_disarm_kprobe(p) +#define kprobe_disarmed(p) kprobe_disabled(p) +#define wait_for_kprobe_optimizer() do {} while (0) + +/* There should be no unused kprobes can be reused without optimization */ +static void reuse_unused_kprobe(struct kprobe *ap) +{ + printk(KERN_ERR "Error: There should be no unused kprobe here.\n"); + BUG_ON(kprobe_unused(ap)); +} static __kprobes void free_aggr_kprobe(struct kprobe *p) { + arch_remove_kprobe(p); kfree(p); } @@ -732,11 +923,10 @@ static void __kprobes arm_kprobe(struct kprobe *kp) /* Disarm a kprobe with text_mutex */ static void __kprobes disarm_kprobe(struct kprobe *kp) { - get_online_cpus(); /* For avoiding text_mutex deadlock */ + /* Ditto */ mutex_lock(&text_mutex); - __disarm_kprobe(kp); + __disarm_kprobe(kp, true); mutex_unlock(&text_mutex); - put_online_cpus(); } /* @@ -942,7 +1132,7 @@ static int __kprobes add_new_kprobe(struct kprobe *ap, struct kprobe *p) BUG_ON(kprobe_gone(ap) || kprobe_gone(p)); if (p->break_handler || p->post_handler) - unoptimize_kprobe(ap); /* Fall back to normal kprobe */ + unoptimize_kprobe(ap, true); /* Fall back to normal kprobe */ if (p->break_handler) { if (ap->break_handler) @@ -993,19 +1183,21 @@ static void __kprobes init_aggr_kprobe(struct kprobe *ap, struct kprobe *p) * This is the second or subsequent kprobe at the address - handle * the intricacies */ -static int __kprobes register_aggr_kprobe(struct kprobe *old_p, +static int __kprobes register_aggr_kprobe(struct kprobe *orig_p, struct kprobe *p) { int ret = 0; - struct kprobe *ap = old_p; + struct kprobe *ap = orig_p; - if (!kprobe_aggrprobe(old_p)) { - /* If old_p is not an aggr_kprobe, create new aggr_kprobe. */ - ap = alloc_aggr_kprobe(old_p); + if (!kprobe_aggrprobe(orig_p)) { + /* If orig_p is not an aggr_kprobe, create new aggr_kprobe. */ + ap = alloc_aggr_kprobe(orig_p); if (!ap) return -ENOMEM; - init_aggr_kprobe(ap, old_p); - } + init_aggr_kprobe(ap, orig_p); + } else if (kprobe_unused(ap)) + /* This probe is going to die. Rescue it */ + reuse_unused_kprobe(ap); if (kprobe_gone(ap)) { /* @@ -1039,23 +1231,6 @@ static int __kprobes register_aggr_kprobe(struct kprobe *old_p, return add_new_kprobe(ap, p); } -/* Try to disable aggr_kprobe, and return 1 if succeeded.*/ -static int __kprobes try_to_disable_aggr_kprobe(struct kprobe *p) -{ - struct kprobe *kp; - - list_for_each_entry_rcu(kp, &p->list, list) { - if (!kprobe_disabled(kp)) - /* - * There is an active probe on the list. - * We can't disable aggr_kprobe. - */ - return 0; - } - p->flags |= KPROBE_FLAG_DISABLED; - return 1; -} - static int __kprobes in_kprobes_functions(unsigned long addr) { struct kprobe_blackpoint *kb; @@ -1098,34 +1273,33 @@ static kprobe_opcode_t __kprobes *kprobe_addr(struct kprobe *p) /* Check passed kprobe is valid and return kprobe in kprobe_table. */ static struct kprobe * __kprobes __get_valid_kprobe(struct kprobe *p) { - struct kprobe *old_p, *list_p; + struct kprobe *ap, *list_p; - old_p = get_kprobe(p->addr); - if (unlikely(!old_p)) + ap = get_kprobe(p->addr); + if (unlikely(!ap)) return NULL; - if (p != old_p) { - list_for_each_entry_rcu(list_p, &old_p->list, list) + if (p != ap) { + list_for_each_entry_rcu(list_p, &ap->list, list) if (list_p == p) /* kprobe p is a valid probe */ goto valid; return NULL; } valid: - return old_p; + return ap; } /* Return error if the kprobe is being re-registered */ static inline int check_kprobe_rereg(struct kprobe *p) { int ret = 0; - struct kprobe *old_p; mutex_lock(&kprobe_mutex); - old_p = __get_valid_kprobe(p); - if (old_p) + if (__get_valid_kprobe(p)) ret = -EINVAL; mutex_unlock(&kprobe_mutex); + return ret; } @@ -1229,67 +1403,121 @@ fail_with_jump_label: } EXPORT_SYMBOL_GPL(register_kprobe); +/* Check if all probes on the aggrprobe are disabled */ +static int __kprobes aggr_kprobe_disabled(struct kprobe *ap) +{ + struct kprobe *kp; + + list_for_each_entry_rcu(kp, &ap->list, list) + if (!kprobe_disabled(kp)) + /* + * There is an active probe on the list. + * We can't disable this ap. + */ + return 0; + + return 1; +} + +/* Disable one kprobe: Make sure called under kprobe_mutex is locked */ +static struct kprobe *__kprobes __disable_kprobe(struct kprobe *p) +{ + struct kprobe *orig_p; + + /* Get an original kprobe for return */ + orig_p = __get_valid_kprobe(p); + if (unlikely(orig_p == NULL)) + return NULL; + + if (!kprobe_disabled(p)) { + /* Disable probe if it is a child probe */ + if (p != orig_p) + p->flags |= KPROBE_FLAG_DISABLED; + + /* Try to disarm and disable this/parent probe */ + if (p == orig_p || aggr_kprobe_disabled(orig_p)) { + disarm_kprobe(orig_p); + orig_p->flags |= KPROBE_FLAG_DISABLED; + } + } + + return orig_p; +} + /* * Unregister a kprobe without a scheduler synchronization. */ static int __kprobes __unregister_kprobe_top(struct kprobe *p) { - struct kprobe *old_p, *list_p; + struct kprobe *ap, *list_p; - old_p = __get_valid_kprobe(p); - if (old_p == NULL) + /* Disable kprobe. This will disarm it if needed. */ + ap = __disable_kprobe(p); + if (ap == NULL) return -EINVAL; - if (old_p == p || - (kprobe_aggrprobe(old_p) && - list_is_singular(&old_p->list))) { + if (ap == p) /* - * Only probe on the hash list. Disarm only if kprobes are - * enabled and not gone - otherwise, the breakpoint would - * already have been removed. We save on flushing icache. + * This probe is an independent(and non-optimized) kprobe + * (not an aggrprobe). Remove from the hash list. */ - if (!kprobes_all_disarmed && !kprobe_disabled(old_p)) - disarm_kprobe(old_p); - hlist_del_rcu(&old_p->hlist); - } else { + goto disarmed; + + /* Following process expects this probe is an aggrprobe */ + WARN_ON(!kprobe_aggrprobe(ap)); + + if (list_is_singular(&ap->list) && kprobe_disarmed(ap)) + /* + * !disarmed could be happen if the probe is under delayed + * unoptimizing. + */ + goto disarmed; + else { + /* If disabling probe has special handlers, update aggrprobe */ if (p->break_handler && !kprobe_gone(p)) - old_p->break_handler = NULL; + ap->break_handler = NULL; if (p->post_handler && !kprobe_gone(p)) { - list_for_each_entry_rcu(list_p, &old_p->list, list) { + list_for_each_entry_rcu(list_p, &ap->list, list) { if ((list_p != p) && (list_p->post_handler)) goto noclean; } - old_p->post_handler = NULL; + ap->post_handler = NULL; } noclean: + /* + * Remove from the aggrprobe: this path will do nothing in + * __unregister_kprobe_bottom(). + */ list_del_rcu(&p->list); - if (!kprobe_disabled(old_p)) { - try_to_disable_aggr_kprobe(old_p); - if (!kprobes_all_disarmed) { - if (kprobe_disabled(old_p)) - disarm_kprobe(old_p); - else - /* Try to optimize this probe again */ - optimize_kprobe(old_p); - } - } + if (!kprobe_disabled(ap) && !kprobes_all_disarmed) + /* + * Try to optimize this probe again, because post + * handler may have been changed. + */ + optimize_kprobe(ap); } return 0; + +disarmed: + BUG_ON(!kprobe_disarmed(ap)); + hlist_del_rcu(&ap->hlist); + return 0; } static void __kprobes __unregister_kprobe_bottom(struct kprobe *p) { - struct kprobe *old_p; + struct kprobe *ap; if (list_empty(&p->list)) + /* This is an independent kprobe */ arch_remove_kprobe(p); else if (list_is_singular(&p->list)) { - /* "p" is the last child of an aggr_kprobe */ - old_p = list_entry(p->list.next, struct kprobe, list); + /* This is the last child of an aggrprobe */ + ap = list_entry(p->list.next, struct kprobe, list); list_del(&p->list); - arch_remove_kprobe(old_p); - free_aggr_kprobe(old_p); + free_aggr_kprobe(ap); } + /* Otherwise, do nothing. */ } int __kprobes register_kprobes(struct kprobe **kps, int num) @@ -1607,29 +1835,13 @@ static void __kprobes kill_kprobe(struct kprobe *p) int __kprobes disable_kprobe(struct kprobe *kp) { int ret = 0; - struct kprobe *p; mutex_lock(&kprobe_mutex); - /* Check whether specified probe is valid. */ - p = __get_valid_kprobe(kp); - if (unlikely(p == NULL)) { + /* Disable this kprobe */ + if (__disable_kprobe(kp) == NULL) ret = -EINVAL; - goto out; - } - /* If the probe is already disabled (or gone), just return */ - if (kprobe_disabled(kp)) - goto out; - - kp->flags |= KPROBE_FLAG_DISABLED; - if (p != kp) - /* When kp != p, p is always enabled. */ - try_to_disable_aggr_kprobe(p); - - if (!kprobes_all_disarmed && kprobe_disabled(p)) - disarm_kprobe(p); -out: mutex_unlock(&kprobe_mutex); return ret; } @@ -1927,36 +2139,27 @@ static void __kprobes disarm_all_kprobes(void) mutex_lock(&kprobe_mutex); /* If kprobes are already disarmed, just return */ - if (kprobes_all_disarmed) - goto already_disabled; + if (kprobes_all_disarmed) { + mutex_unlock(&kprobe_mutex); + return; + } kprobes_all_disarmed = true; printk(KERN_INFO "Kprobes globally disabled\n"); - /* - * Here we call get_online_cpus() for avoiding text_mutex deadlock, - * because disarming may also unoptimize kprobes. - */ - get_online_cpus(); mutex_lock(&text_mutex); for (i = 0; i < KPROBE_TABLE_SIZE; i++) { head = &kprobe_table[i]; hlist_for_each_entry_rcu(p, node, head, hlist) { if (!arch_trampoline_kprobe(p) && !kprobe_disabled(p)) - __disarm_kprobe(p); + __disarm_kprobe(p, false); } } - mutex_unlock(&text_mutex); - put_online_cpus(); mutex_unlock(&kprobe_mutex); - /* Allow all currently running kprobes to complete */ - synchronize_sched(); - return; -already_disabled: - mutex_unlock(&kprobe_mutex); - return; + /* Wait for disarming all kprobes by optimizer */ + wait_for_kprobe_optimizer(); } /* diff --git a/kernel/perf_event.c b/kernel/perf_event.c index 2870feee81dd..11847bf1e8cc 100644 --- a/kernel/perf_event.c +++ b/kernel/perf_event.c @@ -13,6 +13,7 @@ #include <linux/mm.h> #include <linux/cpu.h> #include <linux/smp.h> +#include <linux/idr.h> #include <linux/file.h> #include <linux/poll.h> #include <linux/slab.h> @@ -21,7 +22,9 @@ #include <linux/dcache.h> #include <linux/percpu.h> #include <linux/ptrace.h> +#include <linux/reboot.h> #include <linux/vmstat.h> +#include <linux/device.h> #include <linux/vmalloc.h> #include <linux/hardirq.h> #include <linux/rculist.h> @@ -133,6 +136,28 @@ static void unclone_ctx(struct perf_event_context *ctx) } } +static u32 perf_event_pid(struct perf_event *event, struct task_struct *p) +{ + /* + * only top level events have the pid namespace they were created in + */ + if (event->parent) + event = event->parent; + + return task_tgid_nr_ns(p, event->ns); +} + +static u32 perf_event_tid(struct perf_event *event, struct task_struct *p) +{ + /* + * only top level events have the pid namespace they were created in + */ + if (event->parent) + event = event->parent; + + return task_pid_nr_ns(p, event->ns); +} + /* * If we inherit events we want to return the parent event id * to userspace. @@ -312,9 +337,84 @@ list_add_event(struct perf_event *event, struct perf_event_context *ctx) ctx->nr_stat++; } +/* + * Called at perf_event creation and when events are attached/detached from a + * group. + */ +static void perf_event__read_size(struct perf_event *event) +{ + int entry = sizeof(u64); /* value */ + int size = 0; + int nr = 1; + + if (event->attr.read_format & PERF_FORMAT_TOTAL_TIME_ENABLED) + size += sizeof(u64); + + if (event->attr.read_format & PERF_FORMAT_TOTAL_TIME_RUNNING) + size += sizeof(u64); + + if (event->attr.read_format & PERF_FORMAT_ID) + entry += sizeof(u64); + + if (event->attr.read_format & PERF_FORMAT_GROUP) { + nr += event->group_leader->nr_siblings; + size += sizeof(u64); + } + + size += entry * nr; + event->read_size = size; +} + +static void perf_event__header_size(struct perf_event *event) +{ + struct perf_sample_data *data; + u64 sample_type = event->attr.sample_type; + u16 size = 0; + + perf_event__read_size(event); + + if (sample_type & PERF_SAMPLE_IP) + size += sizeof(data->ip); + + if (sample_type & PERF_SAMPLE_ADDR) + size += sizeof(data->addr); + + if (sample_type & PERF_SAMPLE_PERIOD) + size += sizeof(data->period); + + if (sample_type & PERF_SAMPLE_READ) + size += event->read_size; + + event->header_size = size; +} + +static void perf_event__id_header_size(struct perf_event *event) +{ + struct perf_sample_data *data; + u64 sample_type = event->attr.sample_type; + u16 size = 0; + + if (sample_type & PERF_SAMPLE_TID) + size += sizeof(data->tid_entry); + + if (sample_type & PERF_SAMPLE_TIME) + size += sizeof(data->time); + + if (sample_type & PERF_SAMPLE_ID) + size += sizeof(data->id); + + if (sample_type & PERF_SAMPLE_STREAM_ID) + size += sizeof(data->stream_id); + + if (sample_type & PERF_SAMPLE_CPU) + size += sizeof(data->cpu_entry); + + event->id_header_size = size; +} + static void perf_group_attach(struct perf_event *event) { - struct perf_event *group_leader = event->group_leader; + struct perf_event *group_leader = event->group_leader, *pos; /* * We can have double attach due to group movement in perf_event_open. @@ -333,6 +433,11 @@ static void perf_group_attach(struct perf_event *event) list_add_tail(&event->group_entry, &group_leader->sibling_list); group_leader->nr_siblings++; + + perf_event__header_size(group_leader); + + list_for_each_entry(pos, &group_leader->sibling_list, group_entry) + perf_event__header_size(pos); } /* @@ -391,7 +496,7 @@ static void perf_group_detach(struct perf_event *event) if (event->group_leader != event) { list_del_init(&event->group_entry); event->group_leader->nr_siblings--; - return; + goto out; } if (!list_empty(&event->group_entry)) @@ -410,6 +515,12 @@ static void perf_group_detach(struct perf_event *event) /* Inherit group flags from the previous leader */ sibling->group_flags = event->group_flags; } + +out: + perf_event__header_size(event->group_leader); + + list_for_each_entry(tmp, &event->group_leader->sibling_list, group_entry) + perf_event__header_size(tmp); } static inline int @@ -1073,7 +1184,7 @@ static int perf_event_refresh(struct perf_event *event, int refresh) /* * not supported on inherited events */ - if (event->attr.inherit) + if (event->attr.inherit || !is_sampling_event(event)) return -EINVAL; atomic_add(refresh, &event->event_limit); @@ -2289,31 +2400,6 @@ static int perf_release(struct inode *inode, struct file *file) return perf_event_release_kernel(event); } -static int perf_event_read_size(struct perf_event *event) -{ - int entry = sizeof(u64); /* value */ - int size = 0; - int nr = 1; - - if (event->attr.read_format & PERF_FORMAT_TOTAL_TIME_ENABLED) - size += sizeof(u64); - - if (event->attr.read_format & PERF_FORMAT_TOTAL_TIME_RUNNING) - size += sizeof(u64); - - if (event->attr.read_format & PERF_FORMAT_ID) - entry += sizeof(u64); - - if (event->attr.read_format & PERF_FORMAT_GROUP) { - nr += event->group_leader->nr_siblings; - size += sizeof(u64); - } - - size += entry * nr; - - return size; -} - u64 perf_event_read_value(struct perf_event *event, u64 *enabled, u64 *running) { struct perf_event *child; @@ -2428,7 +2514,7 @@ perf_read_hw(struct perf_event *event, char __user *buf, size_t count) if (event->state == PERF_EVENT_STATE_ERROR) return 0; - if (count < perf_event_read_size(event)) + if (count < event->read_size) return -ENOSPC; WARN_ON_ONCE(event->ctx->parent_ctx); @@ -2514,7 +2600,7 @@ static int perf_event_period(struct perf_event *event, u64 __user *arg) int ret = 0; u64 value; - if (!event->attr.sample_period) + if (!is_sampling_event(event)) return -EINVAL; if (copy_from_user(&value, arg, sizeof(value))) @@ -3305,6 +3391,73 @@ __always_inline void perf_output_copy(struct perf_output_handle *handle, } while (len); } +static void __perf_event_header__init_id(struct perf_event_header *header, + struct perf_sample_data *data, + struct perf_event *event) +{ + u64 sample_type = event->attr.sample_type; + + data->type = sample_type; + header->size += event->id_header_size; + + if (sample_type & PERF_SAMPLE_TID) { + /* namespace issues */ + data->tid_entry.pid = perf_event_pid(event, current); + data->tid_entry.tid = perf_event_tid(event, current); + } + + if (sample_type & PERF_SAMPLE_TIME) + data->time = perf_clock(); + + if (sample_type & PERF_SAMPLE_ID) + data->id = primary_event_id(event); + + if (sample_type & PERF_SAMPLE_STREAM_ID) + data->stream_id = event->id; + + if (sample_type & PERF_SAMPLE_CPU) { + data->cpu_entry.cpu = raw_smp_processor_id(); + data->cpu_entry.reserved = 0; + } +} + +static void perf_event_header__init_id(struct perf_event_header *header, + struct perf_sample_data *data, + struct perf_event *event) +{ + if (event->attr.sample_id_all) + __perf_event_header__init_id(header, data, event); +} + +static void __perf_event__output_id_sample(struct perf_output_handle *handle, + struct perf_sample_data *data) +{ + u64 sample_type = data->type; + + if (sample_type & PERF_SAMPLE_TID) + perf_output_put(handle, data->tid_entry); + + if (sample_type & PERF_SAMPLE_TIME) + perf_output_put(handle, data->time); + + if (sample_type & PERF_SAMPLE_ID) + perf_output_put(handle, data->id); + + if (sample_type & PERF_SAMPLE_STREAM_ID) + perf_output_put(handle, data->stream_id); + + if (sample_type & PERF_SAMPLE_CPU) + perf_output_put(handle, data->cpu_entry); +} + +static void perf_event__output_id_sample(struct perf_event *event, + struct perf_output_handle *handle, + struct perf_sample_data *sample) +{ + if (event->attr.sample_id_all) + __perf_event__output_id_sample(handle, sample); +} + int perf_output_begin(struct perf_output_handle *handle, struct perf_event *event, unsigned int size, int nmi, int sample) @@ -3312,6 +3465,7 @@ int perf_output_begin(struct perf_output_handle *handle, struct perf_buffer *buffer; unsigned long tail, offset, head; int have_lost; + struct perf_sample_data sample_data; struct { struct perf_event_header header; u64 id; @@ -3338,8 +3492,12 @@ int perf_output_begin(struct perf_output_handle *handle, goto out; have_lost = local_read(&buffer->lost); - if (have_lost) - size += sizeof(lost_event); + if (have_lost) { + lost_event.header.size = sizeof(lost_event); + perf_event_header__init_id(&lost_event.header, &sample_data, + event); + size += lost_event.header.size; + } perf_output_get_handle(handle); @@ -3370,11 +3528,11 @@ int perf_output_begin(struct perf_output_handle *handle, if (have_lost) { lost_event.header.type = PERF_RECORD_LOST; lost_event.header.misc = 0; - lost_event.header.size = sizeof(lost_event); lost_event.id = event->id; lost_event.lost = local_xchg(&buffer->lost, 0); perf_output_put(handle, lost_event); + perf_event__output_id_sample(event, handle, &sample_data); } return 0; @@ -3407,28 +3565,6 @@ void perf_output_end(struct perf_output_handle *handle) rcu_read_unlock(); } -static u32 perf_event_pid(struct perf_event *event, struct task_struct *p) -{ - /* - * only top level events have the pid namespace they were created in - */ - if (event->parent) - event = event->parent; - - return task_tgid_nr_ns(p, event->ns); -} - -static u32 perf_event_tid(struct perf_event *event, struct task_struct *p) -{ - /* - * only top level events have the pid namespace they were created in - */ - if (event->parent) - event = event->parent; - - return task_pid_nr_ns(p, event->ns); -} - static void perf_output_read_one(struct perf_output_handle *handle, struct perf_event *event, u64 enabled, u64 running) @@ -3603,61 +3739,16 @@ void perf_prepare_sample(struct perf_event_header *header, { u64 sample_type = event->attr.sample_type; - data->type = sample_type; - header->type = PERF_RECORD_SAMPLE; - header->size = sizeof(*header); + header->size = sizeof(*header) + event->header_size; header->misc = 0; header->misc |= perf_misc_flags(regs); - if (sample_type & PERF_SAMPLE_IP) { - data->ip = perf_instruction_pointer(regs); - - header->size += sizeof(data->ip); - } - - if (sample_type & PERF_SAMPLE_TID) { - /* namespace issues */ - data->tid_entry.pid = perf_event_pid(event, current); - data->tid_entry.tid = perf_event_tid(event, current); - - header->size += sizeof(data->tid_entry); - } - - if (sample_type & PERF_SAMPLE_TIME) { - data->time = perf_clock(); - - header->size += sizeof(data->time); - } - - if (sample_type & PERF_SAMPLE_ADDR) - header->size += sizeof(data->addr); - - if (sample_type & PERF_SAMPLE_ID) { - data->id = primary_event_id(event); - - header->size += sizeof(data->id); - } - - if (sample_type & PERF_SAMPLE_STREAM_ID) { - data->stream_id = event->id; - - header->size += sizeof(data->stream_id); - } - - if (sample_type & PERF_SAMPLE_CPU) { - data->cpu_entry.cpu = raw_smp_processor_id(); - data->cpu_entry.reserved = 0; - - header->size += sizeof(data->cpu_entry); - } - - if (sample_type & PERF_SAMPLE_PERIOD) - header->size += sizeof(data->period); + __perf_event_header__init_id(header, data, event); - if (sample_type & PERF_SAMPLE_READ) - header->size += perf_event_read_size(event); + if (sample_type & PERF_SAMPLE_IP) + data->ip = perf_instruction_pointer(regs); if (sample_type & PERF_SAMPLE_CALLCHAIN) { int size = 1; @@ -3722,23 +3813,26 @@ perf_event_read_event(struct perf_event *event, struct task_struct *task) { struct perf_output_handle handle; + struct perf_sample_data sample; struct perf_read_event read_event = { .header = { .type = PERF_RECORD_READ, .misc = 0, - .size = sizeof(read_event) + perf_event_read_size(event), + .size = sizeof(read_event) + event->read_size, }, .pid = perf_event_pid(event, task), .tid = perf_event_tid(event, task), }; int ret; + perf_event_header__init_id(&read_event.header, &sample, event); ret = perf_output_begin(&handle, event, read_event.header.size, 0, 0); if (ret) return; perf_output_put(&handle, read_event); perf_output_read(&handle, event); + perf_event__output_id_sample(event, &handle, &sample); perf_output_end(&handle); } @@ -3768,14 +3862,16 @@ static void perf_event_task_output(struct perf_event *event, struct perf_task_event *task_event) { struct perf_output_handle handle; + struct perf_sample_data sample; struct task_struct *task = task_event->task; - int size, ret; + int ret, size = task_event->event_id.header.size; - size = task_event->event_id.header.size; - ret = perf_output_begin(&handle, event, size, 0, 0); + perf_event_header__init_id(&task_event->event_id.header, &sample, event); + ret = perf_output_begin(&handle, event, + task_event->event_id.header.size, 0, 0); if (ret) - return; + goto out; task_event->event_id.pid = perf_event_pid(event, task); task_event->event_id.ppid = perf_event_pid(event, current); @@ -3785,7 +3881,11 @@ static void perf_event_task_output(struct perf_event *event, perf_output_put(&handle, task_event->event_id); + perf_event__output_id_sample(event, &handle, &sample); + perf_output_end(&handle); +out: + task_event->event_id.header.size = size; } static int perf_event_task_match(struct perf_event *event) @@ -3900,11 +4000,16 @@ static void perf_event_comm_output(struct perf_event *event, struct perf_comm_event *comm_event) { struct perf_output_handle handle; + struct perf_sample_data sample; int size = comm_event->event_id.header.size; - int ret = perf_output_begin(&handle, event, size, 0, 0); + int ret; + + perf_event_header__init_id(&comm_event->event_id.header, &sample, event); + ret = perf_output_begin(&handle, event, + comm_event->event_id.header.size, 0, 0); if (ret) - return; + goto out; comm_event->event_id.pid = perf_event_pid(event, comm_event->task); comm_event->event_id.tid = perf_event_tid(event, comm_event->task); @@ -3912,7 +4017,12 @@ static void perf_event_comm_output(struct perf_event *event, perf_output_put(&handle, comm_event->event_id); perf_output_copy(&handle, comm_event->comm, comm_event->comm_size); + + perf_event__output_id_sample(event, &handle, &sample); + perf_output_end(&handle); +out: + comm_event->event_id.header.size = size; } static int perf_event_comm_match(struct perf_event *event) @@ -3957,7 +4067,6 @@ static void perf_event_comm_event(struct perf_comm_event *comm_event) comm_event->comm_size = size; comm_event->event_id.header.size = sizeof(comm_event->event_id) + size; - rcu_read_lock(); list_for_each_entry_rcu(pmu, &pmus, entry) { cpuctx = get_cpu_ptr(pmu->pmu_cpu_context); @@ -4038,11 +4147,15 @@ static void perf_event_mmap_output(struct perf_event *event, struct perf_mmap_event *mmap_event) { struct perf_output_handle handle; + struct perf_sample_data sample; int size = mmap_event->event_id.header.size; - int ret = perf_output_begin(&handle, event, size, 0, 0); + int ret; + perf_event_header__init_id(&mmap_event->event_id.header, &sample, event); + ret = perf_output_begin(&handle, event, + mmap_event->event_id.header.size, 0, 0); if (ret) - return; + goto out; mmap_event->event_id.pid = perf_event_pid(event, current); mmap_event->event_id.tid = perf_event_tid(event, current); @@ -4050,7 +4163,12 @@ static void perf_event_mmap_output(struct perf_event *event, perf_output_put(&handle, mmap_event->event_id); perf_output_copy(&handle, mmap_event->file_name, mmap_event->file_size); + + perf_event__output_id_sample(event, &handle, &sample); + perf_output_end(&handle); +out: + mmap_event->event_id.header.size = size; } static int perf_event_mmap_match(struct perf_event *event, @@ -4205,6 +4323,7 @@ void perf_event_mmap(struct vm_area_struct *vma) static void perf_log_throttle(struct perf_event *event, int enable) { struct perf_output_handle handle; + struct perf_sample_data sample; int ret; struct { @@ -4226,11 +4345,15 @@ static void perf_log_throttle(struct perf_event *event, int enable) if (enable) throttle_event.header.type = PERF_RECORD_UNTHROTTLE; - ret = perf_output_begin(&handle, event, sizeof(throttle_event), 1, 0); + perf_event_header__init_id(&throttle_event.header, &sample, event); + + ret = perf_output_begin(&handle, event, + throttle_event.header.size, 1, 0); if (ret) return; perf_output_put(&handle, throttle_event); + perf_event__output_id_sample(event, &handle, &sample); perf_output_end(&handle); } @@ -4246,6 +4369,13 @@ static int __perf_event_overflow(struct perf_event *event, int nmi, struct hw_perf_event *hwc = &event->hw; int ret = 0; + /* + * Non-sampling counters might still use the PMI to fold short + * hardware counters, ignore those. + */ + if (unlikely(!is_sampling_event(event))) + return 0; + if (!throttle) { hwc->interrupts++; } else { @@ -4391,7 +4521,7 @@ static void perf_swevent_event(struct perf_event *event, u64 nr, if (!regs) return; - if (!hwc->sample_period) + if (!is_sampling_event(event)) return; if (nr == 1 && hwc->sample_period == 1 && !event->attr.freq) @@ -4554,7 +4684,7 @@ static int perf_swevent_add(struct perf_event *event, int flags) struct hw_perf_event *hwc = &event->hw; struct hlist_head *head; - if (hwc->sample_period) { + if (is_sampling_event(event)) { hwc->last_period = hwc->sample_period; perf_swevent_set_period(event); } @@ -4811,15 +4941,6 @@ static int perf_tp_event_init(struct perf_event *event) if (event->attr.type != PERF_TYPE_TRACEPOINT) return -ENOENT; - /* - * Raw tracepoint data is a severe data leak, only allow root to - * have these. - */ - if ((event->attr.sample_type & PERF_SAMPLE_RAW) && - perf_paranoid_tracepoint_raw() && - !capable(CAP_SYS_ADMIN)) - return -EPERM; - err = perf_trace_init(event); if (err) return err; @@ -4842,7 +4963,7 @@ static struct pmu perf_tracepoint = { static inline void perf_tp_register(void) { - perf_pmu_register(&perf_tracepoint); + perf_pmu_register(&perf_tracepoint, "tracepoint", PERF_TYPE_TRACEPOINT); } static int perf_event_set_filter(struct perf_event *event, void __user *arg) @@ -4932,31 +5053,33 @@ static enum hrtimer_restart perf_swevent_hrtimer(struct hrtimer *hrtimer) static void perf_swevent_start_hrtimer(struct perf_event *event) { struct hw_perf_event *hwc = &event->hw; + s64 period; + + if (!is_sampling_event(event)) + return; hrtimer_init(&hwc->hrtimer, CLOCK_MONOTONIC, HRTIMER_MODE_REL); hwc->hrtimer.function = perf_swevent_hrtimer; - if (hwc->sample_period) { - s64 period = local64_read(&hwc->period_left); - if (period) { - if (period < 0) - period = 10000; + period = local64_read(&hwc->period_left); + if (period) { + if (period < 0) + period = 10000; - local64_set(&hwc->period_left, 0); - } else { - period = max_t(u64, 10000, hwc->sample_period); - } - __hrtimer_start_range_ns(&hwc->hrtimer, + local64_set(&hwc->period_left, 0); + } else { + period = max_t(u64, 10000, hwc->sample_period); + } + __hrtimer_start_range_ns(&hwc->hrtimer, ns_to_ktime(period), 0, HRTIMER_MODE_REL_PINNED, 0); - } } static void perf_swevent_cancel_hrtimer(struct perf_event *event) { struct hw_perf_event *hwc = &event->hw; - if (hwc->sample_period) { + if (is_sampling_event(event)) { ktime_t remaining = hrtimer_get_remaining(&hwc->hrtimer); local64_set(&hwc->period_left, ktime_to_ns(remaining)); @@ -5184,8 +5307,61 @@ static void free_pmu_context(struct pmu *pmu) out: mutex_unlock(&pmus_lock); } +static struct idr pmu_idr; + +static ssize_t +type_show(struct device *dev, struct device_attribute *attr, char *page) +{ + struct pmu *pmu = dev_get_drvdata(dev); + + return snprintf(page, PAGE_SIZE-1, "%d\n", pmu->type); +} + +static struct device_attribute pmu_dev_attrs[] = { + __ATTR_RO(type), + __ATTR_NULL, +}; + +static int pmu_bus_running; +static struct bus_type pmu_bus = { + .name = "event_source", + .dev_attrs = pmu_dev_attrs, +}; + +static void pmu_dev_release(struct device *dev) +{ + kfree(dev); +} + +static int pmu_dev_alloc(struct pmu *pmu) +{ + int ret = -ENOMEM; + + pmu->dev = kzalloc(sizeof(struct device), GFP_KERNEL); + if (!pmu->dev) + goto out; + + device_initialize(pmu->dev); + ret = dev_set_name(pmu->dev, "%s", pmu->name); + if (ret) + goto free_dev; + + dev_set_drvdata(pmu->dev, pmu); + pmu->dev->bus = &pmu_bus; + pmu->dev->release = pmu_dev_release; + ret = device_add(pmu->dev); + if (ret) + goto free_dev; + +out: + return ret; + +free_dev: + put_device(pmu->dev); + goto out; +} -int perf_pmu_register(struct pmu *pmu) +int perf_pmu_register(struct pmu *pmu, char *name, int type) { int cpu, ret; @@ -5195,13 +5371,38 @@ int perf_pmu_register(struct pmu *pmu) if (!pmu->pmu_disable_count) goto unlock; + pmu->type = -1; + if (!name) + goto skip_type; + pmu->name = name; + + if (type < 0) { + int err = idr_pre_get(&pmu_idr, GFP_KERNEL); + if (!err) + goto free_pdc; + + err = idr_get_new_above(&pmu_idr, pmu, PERF_TYPE_MAX, &type); + if (err) { + ret = err; + goto free_pdc; + } + } + pmu->type = type; + + if (pmu_bus_running) { + ret = pmu_dev_alloc(pmu); + if (ret) + goto free_idr; + } + +skip_type: pmu->pmu_cpu_context = find_pmu_context(pmu->task_ctx_nr); if (pmu->pmu_cpu_context) goto got_cpu_context; pmu->pmu_cpu_context = alloc_percpu(struct perf_cpu_context); if (!pmu->pmu_cpu_context) - goto free_pdc; + goto free_dev; for_each_possible_cpu(cpu) { struct perf_cpu_context *cpuctx; @@ -5245,6 +5446,14 @@ unlock: return ret; +free_dev: + device_del(pmu->dev); + put_device(pmu->dev); + +free_idr: + if (pmu->type >= PERF_TYPE_MAX) + idr_remove(&pmu_idr, pmu->type); + free_pdc: free_percpu(pmu->pmu_disable_count); goto unlock; @@ -5264,6 +5473,10 @@ void perf_pmu_unregister(struct pmu *pmu) synchronize_rcu(); free_percpu(pmu->pmu_disable_count); + if (pmu->type >= PERF_TYPE_MAX) + idr_remove(&pmu_idr, pmu->type); + device_del(pmu->dev); + put_device(pmu->dev); free_pmu_context(pmu); } @@ -5273,6 +5486,13 @@ struct pmu *perf_init_event(struct perf_event *event) int idx; idx = srcu_read_lock(&pmus_srcu); + + rcu_read_lock(); + pmu = idr_find(&pmu_idr, event->attr.type); + rcu_read_unlock(); + if (pmu) + goto unlock; + list_for_each_entry_rcu(pmu, &pmus, entry) { int ret = pmu->event_init(event); if (!ret) @@ -5738,6 +5958,12 @@ SYSCALL_DEFINE5(perf_event_open, mutex_unlock(¤t->perf_event_mutex); /* + * Precalculate sample_data sizes + */ + perf_event__header_size(event); + perf_event__id_header_size(event); + + /* * Drop the reference on the group_event after placing the * new event on the sibling_list. This ensures destruction * of the group leader will find the pointer to itself in @@ -6090,6 +6316,12 @@ inherit_event(struct perf_event *parent_event, child_event->overflow_handler = parent_event->overflow_handler; /* + * Precalculate sample_data sizes + */ + perf_event__header_size(child_event); + perf_event__id_header_size(child_event); + + /* * Link it up in the child's context: */ raw_spin_lock_irqsave(&child_ctx->lock, flags); @@ -6320,7 +6552,7 @@ static void __cpuinit perf_event_init_cpu(int cpu) mutex_unlock(&swhash->hlist_mutex); } -#ifdef CONFIG_HOTPLUG_CPU +#if defined CONFIG_HOTPLUG_CPU || defined CONFIG_KEXEC static void perf_pmu_rotate_stop(struct pmu *pmu) { struct perf_cpu_context *cpuctx = this_cpu_ptr(pmu->pmu_cpu_context); @@ -6374,6 +6606,26 @@ static void perf_event_exit_cpu(int cpu) static inline void perf_event_exit_cpu(int cpu) { } #endif +static int +perf_reboot(struct notifier_block *notifier, unsigned long val, void *v) +{ + int cpu; + + for_each_online_cpu(cpu) + perf_event_exit_cpu(cpu); + + return NOTIFY_OK; +} + +/* + * Run the perf reboot notifier at the very last possible moment so that + * the generic watchdog code runs as long as possible. + */ +static struct notifier_block perf_reboot_notifier = { + .notifier_call = perf_reboot, + .priority = INT_MIN, +}; + static int __cpuinit perf_cpu_notify(struct notifier_block *self, unsigned long action, void *hcpu) { @@ -6402,14 +6654,45 @@ void __init perf_event_init(void) { int ret; + idr_init(&pmu_idr); + perf_event_init_all_cpus(); init_srcu_struct(&pmus_srcu); - perf_pmu_register(&perf_swevent); - perf_pmu_register(&perf_cpu_clock); - perf_pmu_register(&perf_task_clock); + perf_pmu_register(&perf_swevent, "software", PERF_TYPE_SOFTWARE); + perf_pmu_register(&perf_cpu_clock, NULL, -1); + perf_pmu_register(&perf_task_clock, NULL, -1); perf_tp_register(); perf_cpu_notifier(perf_cpu_notify); + register_reboot_notifier(&perf_reboot_notifier); ret = init_hw_breakpoint(); WARN(ret, "hw_breakpoint initialization failed with: %d", ret); } + +static int __init perf_event_sysfs_init(void) +{ + struct pmu *pmu; + int ret; + + mutex_lock(&pmus_lock); + + ret = bus_register(&pmu_bus); + if (ret) + goto unlock; + + list_for_each_entry(pmu, &pmus, entry) { + if (!pmu->name || pmu->type < 0) + continue; + + ret = pmu_dev_alloc(pmu); + WARN(ret, "Failed to register pmu: %s, reason %d\n", pmu->name, ret); + } + pmu_bus_running = 1; + ret = 0; + +unlock: + mutex_unlock(&pmus_lock); + + return ret; +} +device_initcall(perf_event_sysfs_init); diff --git a/kernel/power/suspend.c b/kernel/power/suspend.c index ecf770509d0d..031d5e3a6197 100644 --- a/kernel/power/suspend.c +++ b/kernel/power/suspend.c @@ -22,6 +22,7 @@ #include <linux/mm.h> #include <linux/slab.h> #include <linux/suspend.h> +#include <trace/events/power.h> #include "power.h" @@ -201,6 +202,7 @@ int suspend_devices_and_enter(suspend_state_t state) if (!suspend_ops) return -ENOSYS; + trace_machine_suspend(state); if (suspend_ops->begin) { error = suspend_ops->begin(state); if (error) @@ -229,6 +231,7 @@ int suspend_devices_and_enter(suspend_state_t state) Close: if (suspend_ops->end) suspend_ops->end(); + trace_machine_suspend(PWR_EVENT_EXIT); return error; Recover_platform: diff --git a/kernel/sched.c b/kernel/sched.c index e6f8f1254319..260132961a99 100644 --- a/kernel/sched.c +++ b/kernel/sched.c @@ -8293,8 +8293,6 @@ void __init sched_init(void) zalloc_cpumask_var(&cpu_isolated_map, GFP_NOWAIT); #endif /* SMP */ - perf_event_init(); - scheduler_running = 1; } diff --git a/kernel/sysctl.c b/kernel/sysctl.c index 5abfa1518554..46404414d8a7 100644 --- a/kernel/sysctl.c +++ b/kernel/sysctl.c @@ -745,21 +745,21 @@ static struct ctl_table kern_table[] = { .extra1 = &zero, .extra2 = &one, }, -#endif -#if defined(CONFIG_X86_LOCAL_APIC) && defined(CONFIG_X86) && !defined(CONFIG_LOCKUP_DETECTOR) { - .procname = "unknown_nmi_panic", - .data = &unknown_nmi_panic, + .procname = "nmi_watchdog", + .data = &watchdog_enabled, .maxlen = sizeof (int), .mode = 0644, - .proc_handler = proc_dointvec, + .proc_handler = proc_dowatchdog_enabled, }, +#endif +#if defined(CONFIG_X86_LOCAL_APIC) && defined(CONFIG_X86) { - .procname = "nmi_watchdog", - .data = &nmi_watchdog_enabled, + .procname = "unknown_nmi_panic", + .data = &unknown_nmi_panic, .maxlen = sizeof (int), .mode = 0644, - .proc_handler = proc_nmi_enabled, + .proc_handler = proc_dointvec, }, #endif #if defined(CONFIG_X86) diff --git a/kernel/sysctl_binary.c b/kernel/sysctl_binary.c index 1357c5786064..4b2545a136ff 100644 --- a/kernel/sysctl_binary.c +++ b/kernel/sysctl_binary.c @@ -136,7 +136,6 @@ static const struct bin_table bin_kern_table[] = { { CTL_INT, KERN_IA64_UNALIGNED, "ignore-unaligned-usertrap" }, { CTL_INT, KERN_COMPAT_LOG, "compat-log" }, { CTL_INT, KERN_MAX_LOCK_DEPTH, "max_lock_depth" }, - { CTL_INT, KERN_NMI_WATCHDOG, "nmi_watchdog" }, { CTL_INT, KERN_PANIC_ON_NMI, "panic_on_unrecovered_nmi" }, {} }; diff --git a/kernel/trace/Kconfig b/kernel/trace/Kconfig index ea37e2ff4164..14674dce77a6 100644 --- a/kernel/trace/Kconfig +++ b/kernel/trace/Kconfig @@ -69,6 +69,21 @@ config EVENT_TRACING select CONTEXT_SWITCH_TRACER bool +config EVENT_POWER_TRACING_DEPRECATED + depends on EVENT_TRACING + bool "Deprecated power event trace API, to be removed" + default y + help + Provides old power event types: + C-state/idle accounting events: + power:power_start + power:power_end + and old cpufreq accounting event: + power:power_frequency + This is for userspace compatibility + and will vanish after 5 kernel iterations, + namely 2.6.41. + config CONTEXT_SWITCH_TRACER bool diff --git a/kernel/trace/power-traces.c b/kernel/trace/power-traces.c index a22582a06161..f55fcf61b223 100644 --- a/kernel/trace/power-traces.c +++ b/kernel/trace/power-traces.c @@ -13,5 +13,8 @@ #define CREATE_TRACE_POINTS #include <trace/events/power.h> -EXPORT_TRACEPOINT_SYMBOL_GPL(power_frequency); +#ifdef EVENT_POWER_TRACING_DEPRECATED +EXPORT_TRACEPOINT_SYMBOL_GPL(power_start); +#endif +EXPORT_TRACEPOINT_SYMBOL_GPL(cpu_idle); diff --git a/kernel/trace/trace_event_perf.c b/kernel/trace/trace_event_perf.c index 39c059ca670e..19a359d5e6d5 100644 --- a/kernel/trace/trace_event_perf.c +++ b/kernel/trace/trace_event_perf.c @@ -21,17 +21,46 @@ typedef typeof(unsigned long [PERF_MAX_TRACE_SIZE / sizeof(unsigned long)]) /* Count the events in use (per event id, not per instance) */ static int total_ref_count; +static int perf_trace_event_perm(struct ftrace_event_call *tp_event, + struct perf_event *p_event) +{ + /* No tracing, just counting, so no obvious leak */ + if (!(p_event->attr.sample_type & PERF_SAMPLE_RAW)) + return 0; + + /* Some events are ok to be traced by non-root users... */ + if (p_event->attach_state == PERF_ATTACH_TASK) { + if (tp_event->flags & TRACE_EVENT_FL_CAP_ANY) + return 0; + } + + /* + * ...otherwise raw tracepoint data can be a severe data leak, + * only allow root to have these. + */ + if (perf_paranoid_tracepoint_raw() && !capable(CAP_SYS_ADMIN)) + return -EPERM; + + return 0; +} + static int perf_trace_event_init(struct ftrace_event_call *tp_event, struct perf_event *p_event) { struct hlist_head __percpu *list; - int ret = -ENOMEM; + int ret; int cpu; + ret = perf_trace_event_perm(tp_event, p_event); + if (ret) + return ret; + p_event->tp_event = tp_event; if (tp_event->perf_refcount++ > 0) return 0; + ret = -ENOMEM; + list = alloc_percpu(struct hlist_head); if (!list) goto fail; diff --git a/kernel/trace/trace_events.c b/kernel/trace/trace_events.c index 0725eeab1937..35fde09b81de 100644 --- a/kernel/trace/trace_events.c +++ b/kernel/trace/trace_events.c @@ -27,6 +27,12 @@ DEFINE_MUTEX(event_mutex); +DEFINE_MUTEX(event_storage_mutex); +EXPORT_SYMBOL_GPL(event_storage_mutex); + +char event_storage[EVENT_STORAGE_SIZE]; +EXPORT_SYMBOL_GPL(event_storage); + LIST_HEAD(ftrace_events); LIST_HEAD(ftrace_common_fields); diff --git a/kernel/trace/trace_export.c b/kernel/trace/trace_export.c index 4ba44deaac25..4b74d71705c0 100644 --- a/kernel/trace/trace_export.c +++ b/kernel/trace/trace_export.c @@ -83,13 +83,19 @@ static void __always_unused ____ftrace_check_##name(void) \ #undef __array #define __array(type, item, len) \ - BUILD_BUG_ON(len > MAX_FILTER_STR_VAL); \ - ret = trace_define_field(event_call, #type "[" #len "]", #item, \ + do { \ + BUILD_BUG_ON(len > MAX_FILTER_STR_VAL); \ + mutex_lock(&event_storage_mutex); \ + snprintf(event_storage, sizeof(event_storage), \ + "%s[%d]", #type, len); \ + ret = trace_define_field(event_call, event_storage, #item, \ offsetof(typeof(field), item), \ sizeof(field.item), \ is_signed_type(type), FILTER_OTHER); \ - if (ret) \ - return ret; + mutex_unlock(&event_storage_mutex); \ + if (ret) \ + return ret; \ + } while (0); #undef __array_desc #define __array_desc(type, container, item, len) \ diff --git a/kernel/watchdog.c b/kernel/watchdog.c index 5b082156cd21..aaa8dae08236 100644 --- a/kernel/watchdog.c +++ b/kernel/watchdog.c @@ -57,6 +57,8 @@ static int __init hardlockup_panic_setup(char *str) { if (!strncmp(str, "panic", 5)) hardlockup_panic = 1; + else if (!strncmp(str, "0", 1)) + no_watchdog = 1; return 1; } __setup("nmi_watchdog=", hardlockup_panic_setup); @@ -548,13 +550,13 @@ static struct notifier_block __cpuinitdata cpu_nfb = { .notifier_call = cpu_callback }; -static int __init spawn_watchdog_task(void) +void __init lockup_detector_init(void) { void *cpu = (void *)(long)smp_processor_id(); int err; if (no_watchdog) - return 0; + return; err = cpu_callback(&cpu_nfb, CPU_UP_PREPARE, cpu); WARN_ON(notifier_to_errno(err)); @@ -562,6 +564,5 @@ static int __init spawn_watchdog_task(void) cpu_callback(&cpu_nfb, CPU_ONLINE, cpu); register_cpu_notifier(&cpu_nfb); - return 0; + return; } -early_initcall(spawn_watchdog_task); |