diff options
Diffstat (limited to 'kernel')
| -rw-r--r-- | kernel/module.c | 26 | ||||
| -rw-r--r-- | kernel/trace/ring_buffer.c | 74 | ||||
| -rw-r--r-- | kernel/trace/trace.c | 90 | ||||
| -rw-r--r-- | kernel/trace/trace.h | 1 | ||||
| -rw-r--r-- | kernel/trace/trace_power.c | 8 | ||||
| -rw-r--r-- | kernel/trace/trace_sched_switch.c | 9 | ||||
| -rw-r--r-- | kernel/trace/trace_selftest.c | 26 | ||||
| -rw-r--r-- | kernel/tracepoint.c | 7 |
8 files changed, 169 insertions, 72 deletions
diff --git a/kernel/module.c b/kernel/module.c index 8b742f2b384..7fa134e0cc2 100644 --- a/kernel/module.c +++ b/kernel/module.c @@ -2049,14 +2049,6 @@ static noinline struct module *load_module(void __user *umod, if (err < 0) goto free_mod; -#if defined(CONFIG_MODULE_UNLOAD) && defined(CONFIG_SMP) - mod->refptr = percpu_modalloc(sizeof(local_t), __alignof__(local_t), - mod->name); - if (!mod->refptr) { - err = -ENOMEM; - goto free_mod; - } -#endif if (pcpuindex) { /* We have a special allocation for this section. */ percpu = percpu_modalloc(sechdrs[pcpuindex].sh_size, @@ -2064,7 +2056,7 @@ static noinline struct module *load_module(void __user *umod, mod->name); if (!percpu) { err = -ENOMEM; - goto free_percpu; + goto free_mod; } sechdrs[pcpuindex].sh_flags &= ~(unsigned long)SHF_ALLOC; mod->percpu = percpu; @@ -2116,6 +2108,14 @@ static noinline struct module *load_module(void __user *umod, /* Module has been moved. */ mod = (void *)sechdrs[modindex].sh_addr; +#if defined(CONFIG_MODULE_UNLOAD) && defined(CONFIG_SMP) + mod->refptr = percpu_modalloc(sizeof(local_t), __alignof__(local_t), + mod->name); + if (!mod->refptr) { + err = -ENOMEM; + goto free_init; + } +#endif /* Now we've moved module, initialize linked lists, etc. */ module_unload_init(mod); @@ -2322,15 +2322,17 @@ static noinline struct module *load_module(void __user *umod, ftrace_release(mod->module_core, mod->core_size); free_unload: module_unload_free(mod); + free_init: +#if defined(CONFIG_MODULE_UNLOAD) && defined(CONFIG_SMP) + percpu_modfree(mod->refptr); +#endif module_free(mod, mod->module_init); free_core: module_free(mod, mod->module_core); + /* mod will be freed with core. Don't access it beyond this line! */ free_percpu: if (percpu) percpu_modfree(percpu); -#if defined(CONFIG_MODULE_UNLOAD) && defined(CONFIG_SMP) - percpu_modfree(mod->refptr); -#endif free_mod: kfree(args); free_hdr: diff --git a/kernel/trace/ring_buffer.c b/kernel/trace/ring_buffer.c index 58128ad2fde..384ca5d9d72 100644 --- a/kernel/trace/ring_buffer.c +++ b/kernel/trace/ring_buffer.c @@ -180,29 +180,6 @@ EXPORT_SYMBOL_GPL(tracing_is_on); #include "trace.h" -/* Up this if you want to test the TIME_EXTENTS and normalization */ -#define DEBUG_SHIFT 0 - -u64 ring_buffer_time_stamp(int cpu) -{ - u64 time; - - preempt_disable_notrace(); - /* shift to debug/test normalization and TIME_EXTENTS */ - time = trace_clock_local() << DEBUG_SHIFT; - preempt_enable_no_resched_notrace(); - - return time; -} -EXPORT_SYMBOL_GPL(ring_buffer_time_stamp); - -void ring_buffer_normalize_time_stamp(int cpu, u64 *ts) -{ - /* Just stupid testing the normalize function and deltas */ - *ts >>= DEBUG_SHIFT; -} -EXPORT_SYMBOL_GPL(ring_buffer_normalize_time_stamp); - #define RB_EVNT_HDR_SIZE (offsetof(struct ring_buffer_event, array)) #define RB_ALIGNMENT 4U #define RB_MAX_SMALL_DATA 28 @@ -374,6 +351,7 @@ struct ring_buffer { #ifdef CONFIG_HOTPLUG_CPU struct notifier_block cpu_notify; #endif + u64 (*clock)(void); }; struct ring_buffer_iter { @@ -394,6 +372,30 @@ struct ring_buffer_iter { _____ret; \ }) +/* Up this if you want to test the TIME_EXTENTS and normalization */ +#define DEBUG_SHIFT 0 + +u64 ring_buffer_time_stamp(struct ring_buffer *buffer, int cpu) +{ + u64 time; + + preempt_disable_notrace(); + /* shift to debug/test normalization and TIME_EXTENTS */ + time = buffer->clock() << DEBUG_SHIFT; + preempt_enable_no_resched_notrace(); + + return time; +} +EXPORT_SYMBOL_GPL(ring_buffer_time_stamp); + +void ring_buffer_normalize_time_stamp(struct ring_buffer *buffer, + int cpu, u64 *ts) +{ + /* Just stupid testing the normalize function and deltas */ + *ts >>= DEBUG_SHIFT; +} +EXPORT_SYMBOL_GPL(ring_buffer_normalize_time_stamp); + /** * check_pages - integrity check of buffer pages * @cpu_buffer: CPU buffer with pages to test @@ -569,13 +571,23 @@ struct ring_buffer *ring_buffer_alloc(unsigned long size, unsigned flags) buffer->pages = DIV_ROUND_UP(size, BUF_PAGE_SIZE); buffer->flags = flags; + buffer->clock = trace_clock_local; /* need at least two pages */ if (buffer->pages == 1) buffer->pages++; + /* + * In case of non-hotplug cpu, if the ring-buffer is allocated + * in early initcall, it will not be notified of secondary cpus. + * In that off case, we need to allocate for all possible cpus. + */ +#ifdef CONFIG_HOTPLUG_CPU get_online_cpus(); cpumask_copy(buffer->cpumask, cpu_online_mask); +#else + cpumask_copy(buffer->cpumask, cpu_possible_mask); +#endif buffer->cpus = nr_cpu_ids; bsize = sizeof(void *) * nr_cpu_ids; @@ -645,6 +657,12 @@ ring_buffer_free(struct ring_buffer *buffer) } EXPORT_SYMBOL_GPL(ring_buffer_free); +void ring_buffer_set_clock(struct ring_buffer *buffer, + u64 (*clock)(void)) +{ + buffer->clock = clock; +} + static void rb_reset_cpu(struct ring_buffer_per_cpu *cpu_buffer); static void @@ -1191,7 +1209,7 @@ __rb_reserve_next(struct ring_buffer_per_cpu *cpu_buffer, cpu_buffer->tail_page = next_page; /* reread the time stamp */ - *ts = ring_buffer_time_stamp(cpu_buffer->cpu); + *ts = ring_buffer_time_stamp(buffer, cpu_buffer->cpu); cpu_buffer->tail_page->page->time_stamp = *ts; } @@ -1334,7 +1352,7 @@ rb_reserve_next_event(struct ring_buffer_per_cpu *cpu_buffer, if (RB_WARN_ON(cpu_buffer, ++nr_loops > 1000)) return NULL; - ts = ring_buffer_time_stamp(cpu_buffer->cpu); + ts = ring_buffer_time_stamp(cpu_buffer->buffer, cpu_buffer->cpu); /* * Only the first commit can update the timestamp. @@ -2051,7 +2069,8 @@ rb_buffer_peek(struct ring_buffer *buffer, int cpu, u64 *ts) case RINGBUF_TYPE_DATA: if (ts) { *ts = cpu_buffer->read_stamp + event->time_delta; - ring_buffer_normalize_time_stamp(cpu_buffer->cpu, ts); + ring_buffer_normalize_time_stamp(buffer, + cpu_buffer->cpu, ts); } return event; @@ -2112,7 +2131,8 @@ rb_iter_peek(struct ring_buffer_iter *iter, u64 *ts) case RINGBUF_TYPE_DATA: if (ts) { *ts = iter->read_stamp + event->time_delta; - ring_buffer_normalize_time_stamp(cpu_buffer->cpu, ts); + ring_buffer_normalize_time_stamp(buffer, + cpu_buffer->cpu, ts); } return event; diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c index 1ce6208fd72..c95b7292be7 100644 --- a/kernel/trace/trace.c +++ b/kernel/trace/trace.c @@ -155,13 +155,6 @@ ns2usecs(cycle_t nsec) return nsec; } -cycle_t ftrace_now(int cpu) -{ - u64 ts = ring_buffer_time_stamp(cpu); - ring_buffer_normalize_time_stamp(cpu, &ts); - return ts; -} - /* * The global_trace is the descriptor that holds the tracing * buffers for the live tracing. For each CPU, it contains @@ -178,6 +171,20 @@ static struct trace_array global_trace; static DEFINE_PER_CPU(struct trace_array_cpu, global_trace_cpu); +cycle_t ftrace_now(int cpu) +{ + u64 ts; + + /* Early boot up does not have a buffer yet */ + if (!global_trace.buffer) + return trace_clock_local(); + + ts = ring_buffer_time_stamp(global_trace.buffer, cpu); + ring_buffer_normalize_time_stamp(global_trace.buffer, cpu, &ts); + + return ts; +} + /* * The max_tr is used to snapshot the global_trace when a maximum * latency is reached. Some tracers will use this to store a maximum @@ -308,6 +315,7 @@ static const char *trace_options[] = { "printk-msg-only", "context-info", "latency-format", + "global-clock", NULL }; @@ -633,6 +641,7 @@ void tracing_reset_online_cpus(struct trace_array *tr) } #define SAVED_CMDLINES 128 +#define NO_CMDLINE_MAP UINT_MAX static unsigned map_pid_to_cmdline[PID_MAX_DEFAULT+1]; static unsigned map_cmdline_to_pid[SAVED_CMDLINES]; static char saved_cmdlines[SAVED_CMDLINES][TASK_COMM_LEN]; @@ -644,8 +653,8 @@ static atomic_t trace_record_cmdline_disabled __read_mostly; static void trace_init_cmdlines(void) { - memset(&map_pid_to_cmdline, -1, sizeof(map_pid_to_cmdline)); - memset(&map_cmdline_to_pid, -1, sizeof(map_cmdline_to_pid)); + memset(&map_pid_to_cmdline, NO_CMDLINE_MAP, sizeof(map_pid_to_cmdline)); + memset(&map_cmdline_to_pid, NO_CMDLINE_MAP, sizeof(map_cmdline_to_pid)); cmdline_idx = 0; } @@ -737,8 +746,7 @@ void trace_stop_cmdline_recording(void); static void trace_save_cmdline(struct task_struct *tsk) { - unsigned map; - unsigned idx; + unsigned pid, idx; if (!tsk->pid || unlikely(tsk->pid > PID_MAX_DEFAULT)) return; @@ -753,13 +761,20 @@ static void trace_save_cmdline(struct task_struct *tsk) return; idx = map_pid_to_cmdline[tsk->pid]; - if (idx >= SAVED_CMDLINES) { + if (idx == NO_CMDLINE_MAP) { idx = (cmdline_idx + 1) % SAVED_CMDLINES; - map = map_cmdline_to_pid[idx]; - if (map <= PID_MAX_DEFAULT) - map_pid_to_cmdline[map] = (unsigned)-1; + /* + * Check whether the cmdline buffer at idx has a pid + * mapped. We are going to overwrite that entry so we + * need to clear the map_pid_to_cmdline. Otherwise we + * would read the new comm for the old pid. + */ + pid = map_cmdline_to_pid[idx]; + if (pid != NO_CMDLINE_MAP) + map_pid_to_cmdline[pid] = NO_CMDLINE_MAP; + map_cmdline_to_pid[idx] = tsk->pid; map_pid_to_cmdline[tsk->pid] = idx; cmdline_idx = idx; @@ -786,18 +801,18 @@ void trace_find_cmdline(int pid, char comm[]) __raw_spin_lock(&trace_cmdline_lock); map = map_pid_to_cmdline[pid]; - if (map >= SAVED_CMDLINES) - goto out; - - strcpy(comm, saved_cmdlines[map]); + if (map != NO_CMDLINE_MAP) + strcpy(comm, saved_cmdlines[map]); + else + strcpy(comm, "<...>"); - out: __raw_spin_unlock(&trace_cmdline_lock); } void tracing_record_cmdline(struct task_struct *tsk) { - if (atomic_read(&trace_record_cmdline_disabled) || !tracing_is_on()) + if (atomic_read(&trace_record_cmdline_disabled) || !tracer_enabled || + !tracing_is_on()) return; trace_save_cmdline(tsk); @@ -2244,6 +2259,34 @@ static int set_tracer_option(struct tracer *trace, char *cmp, int neg) return 0; } +static void set_tracer_flags(unsigned int mask, int enabled) +{ + /* do nothing if flag is already set */ + if (!!(trace_flags & mask) == !!enabled) + return; + + if (enabled) + trace_flags |= mask; + else + trace_flags &= ~mask; + + if (mask == TRACE_ITER_GLOBAL_CLK) { + u64 (*func)(void); + + if (enabled) + func = trace_clock_global; + else + func = trace_clock_local; + + mutex_lock(&trace_types_lock); + ring_buffer_set_clock(global_trace.buffer, func); + + if (max_tr.buffer) + ring_buffer_set_clock(max_tr.buffer, func); + mutex_unlock(&trace_types_lock); + } +} + static ssize_t tracing_trace_options_write(struct file *filp, const char __user *ubuf, size_t cnt, loff_t *ppos) @@ -2271,10 +2314,7 @@ tracing_trace_options_write(struct file *filp, const char __user *ubuf, int len = strlen(trace_options[i]); if (strncmp(cmp, trace_options[i], len) == 0) { - if (neg) - trace_flags &= ~(1 << i); - else - trace_flags |= (1 << i); + set_tracer_flags(1 << i, !neg); break; } } diff --git a/kernel/trace/trace.h b/kernel/trace/trace.h index 546bcbd92a0..38276d1638e 100644 --- a/kernel/trace/trace.h +++ b/kernel/trace/trace.h @@ -684,6 +684,7 @@ enum trace_iterator_flags { TRACE_ITER_PRINTK_MSGONLY = 0x10000, TRACE_ITER_CONTEXT_INFO = 0x20000, /* Print pid/cpu/time */ TRACE_ITER_LATENCY_FMT = 0x40000, + TRACE_ITER_GLOBAL_CLK = 0x80000, }; /* diff --git a/kernel/trace/trace_power.c b/kernel/trace/trace_power.c index 91ce672fb03..bae791ebcc5 100644 --- a/kernel/trace/trace_power.c +++ b/kernel/trace/trace_power.c @@ -122,12 +122,16 @@ fail_start: static void start_power_trace(struct trace_array *tr) { trace_power_enabled = 1; - tracing_power_register(); } static void stop_power_trace(struct trace_array *tr) { trace_power_enabled = 0; +} + +static void power_trace_reset(struct trace_array *tr) +{ + trace_power_enabled = 0; unregister_trace_power_start(probe_power_start); unregister_trace_power_end(probe_power_end); unregister_trace_power_mark(probe_power_mark); @@ -188,7 +192,7 @@ static struct tracer power_tracer __read_mostly = .init = power_trace_init, .start = start_power_trace, .stop = stop_power_trace, - .reset = stop_power_trace, + .reset = power_trace_reset, .print_line = power_print_line, }; diff --git a/kernel/trace/trace_sched_switch.c b/kernel/trace/trace_sched_switch.c index 77132c2cf3d..de35f200abd 100644 --- a/kernel/trace/trace_sched_switch.c +++ b/kernel/trace/trace_sched_switch.c @@ -18,6 +18,7 @@ static struct trace_array *ctx_trace; static int __read_mostly tracer_enabled; static int sched_ref; static DEFINE_MUTEX(sched_register_mutex); +static int sched_stopped; static void probe_sched_switch(struct rq *__rq, struct task_struct *prev, @@ -28,7 +29,7 @@ probe_sched_switch(struct rq *__rq, struct task_struct *prev, int cpu; int pc; - if (!sched_ref) + if (!sched_ref || sched_stopped) return; tracing_record_cmdline(prev); @@ -193,6 +194,7 @@ static void stop_sched_trace(struct trace_array *tr) static int sched_switch_trace_init(struct trace_array *tr) { ctx_trace = tr; + tracing_reset_online_cpus(tr); tracing_start_sched_switch_record(); return 0; } @@ -205,13 +207,12 @@ static void sched_switch_trace_reset(struct trace_array *tr) static void sched_switch_trace_start(struct trace_array *tr) { - tracing_reset_online_cpus(tr); - tracing_start_sched_switch(); + sched_stopped = 0; } static void sched_switch_trace_stop(struct trace_array *tr) { - tracing_stop_sched_switch(); + sched_stopped = 1; } static struct tracer sched_switch_trace __read_mostly = diff --git a/kernel/trace/trace_selftest.c b/kernel/trace/trace_selftest.c index a2ca6f0fef9..38856ba78a9 100644 --- a/kernel/trace/trace_selftest.c +++ b/kernel/trace/trace_selftest.c @@ -315,6 +315,14 @@ trace_selftest_startup_irqsoff(struct tracer *trace, struct trace_array *tr) local_irq_disable(); udelay(100); local_irq_enable(); + + /* + * Stop the tracer to avoid a warning subsequent + * to buffer flipping failure because tracing_stop() + * disables the tr and max buffers, making flipping impossible + * in case of parallels max irqs off latencies. + */ + trace->stop(tr); /* stop the tracing. */ tracing_stop(); /* check both trace buffers */ @@ -369,6 +377,14 @@ trace_selftest_startup_preemptoff(struct tracer *trace, struct trace_array *tr) preempt_disable(); udelay(100); preempt_enable(); + + /* + * Stop the tracer to avoid a warning subsequent + * to buffer flipping failure because tracing_stop() + * disables the tr and max buffers, making flipping impossible + * in case of parallels max preempt off latencies. + */ + trace->stop(tr); /* stop the tracing. */ tracing_stop(); /* check both trace buffers */ @@ -428,6 +444,13 @@ trace_selftest_startup_preemptirqsoff(struct tracer *trace, struct trace_array * /* reverse the order of preempt vs irqs */ local_irq_enable(); + /* + * Stop the tracer to avoid a warning subsequent + * to buffer flipping failure because tracing_stop() + * disables the tr and max buffers, making flipping impossible + * in case of parallels max irqs/preempt off latencies. + */ + trace->stop(tr); /* stop the tracing. */ tracing_stop(); /* check both trace buffers */ @@ -448,6 +471,8 @@ trace_selftest_startup_preemptirqsoff(struct tracer *trace, struct trace_array * /* do the test by disabling interrupts first this time */ tracing_max_latency = 0; tracing_start(); + trace->start(tr); + preempt_disable(); local_irq_disable(); udelay(100); @@ -455,6 +480,7 @@ trace_selftest_startup_preemptirqsoff(struct tracer *trace, struct trace_array * /* reverse the order of preempt vs irqs */ local_irq_enable(); + trace->stop(tr); /* stop the tracing. */ tracing_stop(); /* check both trace buffers */ diff --git a/kernel/tracepoint.c b/kernel/tracepoint.c index 79602740bbb..1ef5d3a601c 100644 --- a/kernel/tracepoint.c +++ b/kernel/tracepoint.c @@ -272,12 +272,15 @@ static void disable_tracepoint(struct tracepoint *elem) * * Updates the probe callback corresponding to a range of tracepoints. */ -void tracepoint_update_probe_range(struct tracepoint *begin, - struct tracepoint *end) +void +tracepoint_update_probe_range(struct tracepoint *begin, struct tracepoint *end) { struct tracepoint *iter; struct tracepoint_entry *mark_entry; + if (!begin) + return; + mutex_lock(&tracepoints_mutex); for (iter = begin; iter < end; iter++) { mark_entry = get_tracepoint(iter->name); |