aboutsummaryrefslogtreecommitdiff
diff options
context:
space:
mode:
-rw-r--r--arch/sparc/kernel/perf_event.c108
-rw-r--r--arch/x86/include/asm/perf_event_p4.h3
-rw-r--r--arch/x86/kernel/cpu/perf_event_p4.c41
-rw-r--r--include/linux/ftrace_event.h8
-rw-r--r--include/linux/perf_event.h19
-rw-r--r--include/trace/ftrace.h3
-rw-r--r--kernel/perf_event.c153
-rw-r--r--kernel/trace/trace_event_perf.c11
-rw-r--r--kernel/trace/trace_kprobe.c4
-rw-r--r--kernel/trace/trace_syscalls.c6
-rw-r--r--tools/perf/Documentation/perf-stat.txt3
-rw-r--r--tools/perf/builtin-probe.c10
-rw-r--r--tools/perf/builtin-record.c1
-rw-r--r--tools/perf/builtin-stat.c18
-rw-r--r--tools/perf/util/abspath.c81
-rw-r--r--tools/perf/util/cache.h55
-rw-r--r--tools/perf/util/config.c461
-rw-r--r--tools/perf/util/exec_cmd.c6
-rw-r--r--tools/perf/util/exec_cmd.h1
-rw-r--r--tools/perf/util/header.c84
-rw-r--r--tools/perf/util/help.c30
-rw-r--r--tools/perf/util/newt.c10
-rw-r--r--tools/perf/util/path.c204
-rw-r--r--tools/perf/util/probe-finder.c33
-rw-r--r--tools/perf/util/probe-finder.h3
-rw-r--r--tools/perf/util/quote.c433
-rw-r--r--tools/perf/util/quote.h39
-rw-r--r--tools/perf/util/run-command.c90
-rw-r--r--tools/perf/util/run-command.h30
-rw-r--r--tools/perf/util/session.c8
-rw-r--r--tools/perf/util/session.h8
-rw-r--r--tools/perf/util/sigchain.c2
-rw-r--r--tools/perf/util/sigchain.h1
-rw-r--r--tools/perf/util/strbuf.c229
-rw-r--r--tools/perf/util/strbuf.h45
-rw-r--r--tools/perf/util/symbol.c13
-rw-r--r--tools/perf/util/symbol.h1
-rw-r--r--tools/perf/util/util.h162
-rw-r--r--tools/perf/util/wrapper.c110
39 files changed, 338 insertions, 2189 deletions
diff --git a/arch/sparc/kernel/perf_event.c b/arch/sparc/kernel/perf_event.c
index e2771939341..cf4ce263ff8 100644
--- a/arch/sparc/kernel/perf_event.c
+++ b/arch/sparc/kernel/perf_event.c
@@ -91,6 +91,8 @@ struct cpu_hw_events {
/* Enabled/disable state. */
int enabled;
+
+ unsigned int group_flag;
};
DEFINE_PER_CPU(struct cpu_hw_events, cpu_hw_events) = { .enabled = 1, };
@@ -980,53 +982,6 @@ static int collect_events(struct perf_event *group, int max_count,
return n;
}
-static void event_sched_in(struct perf_event *event)
-{
- event->state = PERF_EVENT_STATE_ACTIVE;
- event->oncpu = smp_processor_id();
- event->tstamp_running += event->ctx->time - event->tstamp_stopped;
- if (is_software_event(event))
- event->pmu->enable(event);
-}
-
-int hw_perf_group_sched_in(struct perf_event *group_leader,
- struct perf_cpu_context *cpuctx,
- struct perf_event_context *ctx)
-{
- struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events);
- struct perf_event *sub;
- int n0, n;
-
- if (!sparc_pmu)
- return 0;
-
- n0 = cpuc->n_events;
- n = collect_events(group_leader, perf_max_events - n0,
- &cpuc->event[n0], &cpuc->events[n0],
- &cpuc->current_idx[n0]);
- if (n < 0)
- return -EAGAIN;
- if (check_excludes(cpuc->event, n0, n))
- return -EINVAL;
- if (sparc_check_constraints(cpuc->event, cpuc->events, n + n0))
- return -EAGAIN;
- cpuc->n_events = n0 + n;
- cpuc->n_added += n;
-
- cpuctx->active_oncpu += n;
- n = 1;
- event_sched_in(group_leader);
- list_for_each_entry(sub, &group_leader->sibling_list, group_entry) {
- if (sub->state != PERF_EVENT_STATE_OFF) {
- event_sched_in(sub);
- n++;
- }
- }
- ctx->nr_active += n;
-
- return 1;
-}
-
static int sparc_pmu_enable(struct perf_event *event)
{
struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events);
@@ -1044,11 +999,20 @@ static int sparc_pmu_enable(struct perf_event *event)
cpuc->events[n0] = event->hw.event_base;
cpuc->current_idx[n0] = PIC_NO_INDEX;
+ /*
+ * If group events scheduling transaction was started,
+ * skip the schedulability test here, it will be peformed
+ * at commit time(->commit_txn) as a whole
+ */
+ if (cpuc->group_flag & PERF_EVENT_TXN_STARTED)
+ goto nocheck;
+
if (check_excludes(cpuc->event, n0, 1))
goto out;
if (sparc_check_constraints(cpuc->event, cpuc->events, n0 + 1))
goto out;
+nocheck:
cpuc->n_events++;
cpuc->n_added++;
@@ -1128,11 +1092,61 @@ static int __hw_perf_event_init(struct perf_event *event)
return 0;
}
+/*
+ * Start group events scheduling transaction
+ * Set the flag to make pmu::enable() not perform the
+ * schedulability test, it will be performed at commit time
+ */
+static void sparc_pmu_start_txn(const struct pmu *pmu)
+{
+ struct cpu_hw_events *cpuhw = &__get_cpu_var(cpu_hw_events);
+
+ cpuhw->group_flag |= PERF_EVENT_TXN_STARTED;
+}
+
+/*
+ * Stop group events scheduling transaction
+ * Clear the flag and pmu::enable() will perform the
+ * schedulability test.
+ */
+static void sparc_pmu_cancel_txn(const struct pmu *pmu)
+{
+ struct cpu_hw_events *cpuhw = &__get_cpu_var(cpu_hw_events);
+
+ cpuhw->group_flag &= ~PERF_EVENT_TXN_STARTED;
+}
+
+/*
+ * Commit group events scheduling transaction
+ * Perform the group schedulability test as a whole
+ * Return 0 if success
+ */
+static int sparc_pmu_commit_txn(const struct pmu *pmu)
+{
+ struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events);
+ int n;
+
+ if (!sparc_pmu)
+ return -EINVAL;
+
+ cpuc = &__get_cpu_var(cpu_hw_events);
+ n = cpuc->n_events;
+ if (check_excludes(cpuc->event, 0, n))
+ return -EINVAL;
+ if (sparc_check_constraints(cpuc->event, cpuc->events, n))
+ return -EAGAIN;
+
+ return 0;
+}
+
static const struct pmu pmu = {
.enable = sparc_pmu_enable,
.disable = sparc_pmu_disable,
.read = sparc_pmu_read,
.unthrottle = sparc_pmu_unthrottle,
+ .start_txn = sparc_pmu_start_txn,
+ .cancel_txn = sparc_pmu_cancel_txn,
+ .commit_txn = sparc_pmu_commit_txn,
};
const struct pmu *hw_perf_event_init(struct perf_event *event)
diff --git a/arch/x86/include/asm/perf_event_p4.h b/arch/x86/include/asm/perf_event_p4.h
index b05400a542f..64a8ebff06f 100644
--- a/arch/x86/include/asm/perf_event_p4.h
+++ b/arch/x86/include/asm/perf_event_p4.h
@@ -89,7 +89,8 @@
P4_CCCR_ENABLE)
/* HT mask */
-#define P4_CCCR_MASK_HT (P4_CCCR_MASK | P4_CCCR_THREAD_ANY)
+#define P4_CCCR_MASK_HT \
+ (P4_CCCR_MASK | P4_CCCR_OVF_PMI_T1 | P4_CCCR_THREAD_ANY)
#define P4_GEN_ESCR_EMASK(class, name, bit) \
class##__##name = ((1 << bit) << P4_ESCR_EVENTMASK_SHIFT)
diff --git a/arch/x86/kernel/cpu/perf_event_p4.c b/arch/x86/kernel/cpu/perf_event_p4.c
index 424fc8de68e..ae85d69644d 100644
--- a/arch/x86/kernel/cpu/perf_event_p4.c
+++ b/arch/x86/kernel/cpu/perf_event_p4.c
@@ -465,15 +465,21 @@ out:
return rc;
}
-static inline void p4_pmu_clear_cccr_ovf(struct hw_perf_event *hwc)
+static inline int p4_pmu_clear_cccr_ovf(struct hw_perf_event *hwc)
{
- unsigned long dummy;
+ int overflow = 0;
+ u32 low, high;
- rdmsrl(hwc->config_base + hwc->idx, dummy);
- if (dummy & P4_CCCR_OVF) {
+ rdmsr(hwc->config_base + hwc->idx, low, high);
+
+ /* we need to check high bit for unflagged overflows */
+ if ((low & P4_CCCR_OVF) || !(high & (1 << 31))) {
+ overflow = 1;
(void)checking_wrmsrl(hwc->config_base + hwc->idx,
- ((u64)dummy) & ~P4_CCCR_OVF);
+ ((u64)low) & ~P4_CCCR_OVF);
}
+
+ return overflow;
}
static inline void p4_pmu_disable_event(struct perf_event *event)
@@ -584,21 +590,15 @@ static int p4_pmu_handle_irq(struct pt_regs *regs)
WARN_ON_ONCE(hwc->idx != idx);
- /*
- * FIXME: Redundant call, actually not needed
- * but just to check if we're screwed
- */
- p4_pmu_clear_cccr_ovf(hwc);
+ /* it might be unflagged overflow */
+ handled = p4_pmu_clear_cccr_ovf(hwc);
val = x86_perf_event_update(event);
- if (val & (1ULL << (x86_pmu.cntval_bits - 1)))
+ if (!handled && (val & (1ULL << (x86_pmu.cntval_bits - 1))))
continue;
- /*
- * event overflow
- */
- handled = 1;
- data.period = event->hw.last_period;
+ /* event overflow for sure */
+ data.period = event->hw.last_period;
if (!x86_perf_event_set_period(event))
continue;
@@ -670,7 +670,7 @@ static void p4_pmu_swap_config_ts(struct hw_perf_event *hwc, int cpu)
/*
* ESCR address hashing is tricky, ESCRs are not sequential
- * in memory but all starts from MSR_P4_BSU_ESCR0 (0x03e0) and
+ * in memory but all starts from MSR_P4_BSU_ESCR0 (0x03a0) and
* the metric between any ESCRs is laid in range [0xa0,0xe1]
*
* so we make ~70% filled hashtable
@@ -735,8 +735,9 @@ static int p4_get_escr_idx(unsigned int addr)
{
unsigned int idx = P4_ESCR_MSR_IDX(addr);
- if (unlikely(idx >= P4_ESCR_MSR_TABLE_SIZE ||
- !p4_escr_table[idx])) {
+ if (unlikely(idx >= P4_ESCR_MSR_TABLE_SIZE ||
+ !p4_escr_table[idx] ||
+ p4_escr_table[idx] != addr)) {
WARN_ONCE(1, "P4 PMU: Wrong address passed: %x\n", addr);
return -1;
}
@@ -762,7 +763,7 @@ static int p4_pmu_schedule_events(struct cpu_hw_events *cpuc, int n, int *assign
{
unsigned long used_mask[BITS_TO_LONGS(X86_PMC_IDX_MAX)];
unsigned long escr_mask[BITS_TO_LONGS(P4_ESCR_MSR_TABLE_SIZE)];
- int cpu = raw_smp_processor_id();
+ int cpu = smp_processor_id();
struct hw_perf_event *hwc;
struct p4_event_bind *bind;
unsigned int i, thread, num;
diff --git a/include/linux/ftrace_event.h b/include/linux/ftrace_event.h
index 39e71b0a3bf..a9775dd7f7f 100644
--- a/include/linux/ftrace_event.h
+++ b/include/linux/ftrace_event.h
@@ -133,6 +133,7 @@ struct ftrace_event_call {
void *data;
int perf_refcount;
+ void *perf_data;
int (*perf_event_enable)(struct ftrace_event_call *);
void (*perf_event_disable)(struct ftrace_event_call *);
};
@@ -191,7 +192,7 @@ struct perf_event;
DECLARE_PER_CPU(struct pt_regs, perf_trace_regs);
-extern int perf_trace_enable(int event_id);
+extern int perf_trace_enable(int event_id, void *data);
extern void perf_trace_disable(int event_id);
extern int ftrace_profile_set_filter(struct perf_event *event, int event_id,
char *filter_str);
@@ -202,11 +203,12 @@ perf_trace_buf_prepare(int size, unsigned short type, int *rctxp,
static inline void
perf_trace_buf_submit(void *raw_data, int size, int rctx, u64 addr,
- u64 count, unsigned long irq_flags, struct pt_regs *regs)
+ u64 count, unsigned long irq_flags, struct pt_regs *regs,
+ void *event)
{
struct trace_entry *entry = raw_data;
- perf_tp_event(entry->type, addr, count, raw_data, size, regs);
+ perf_tp_event(entry->type, addr, count, raw_data, size, regs, event);
perf_swevent_put_recursion_context(rctx);
local_irq_restore(irq_flags);
}
diff --git a/include/linux/perf_event.h b/include/linux/perf_event.h
index 3fd5c82e0e1..fe50347dc64 100644
--- a/include/linux/perf_event.h
+++ b/include/linux/perf_event.h
@@ -485,6 +485,7 @@ struct perf_guest_info_callbacks {
#include <linux/ftrace.h>
#include <linux/cpu.h>
#include <asm/atomic.h>
+#include <asm/local.h>
#define PERF_MAX_STACK_DEPTH 255
@@ -588,20 +589,18 @@ struct perf_mmap_data {
#ifdef CONFIG_PERF_USE_VMALLOC
struct work_struct work;
#endif
- int data_order;
+ int data_order; /* allocation order */
int nr_pages; /* nr of data pages */
int writable; /* are we writable */
int nr_locked; /* nr pages mlocked */
atomic_t poll; /* POLL_ for wakeups */
- atomic_t events; /* event_id limit */
- atomic_long_t head; /* write position */
- atomic_long_t done_head; /* completed head */
-
- atomic_t lock; /* concurrent writes */
- atomic_t wakeup; /* needs a wakeup */
- atomic_t lost; /* nr records lost */
+ local_t head; /* write position */
+ local_t nest; /* nested writers */
+ local_t events; /* event limit */
+ local_t wakeup; /* needs a wakeup */
+ local_t lost; /* nr records lost */
long watermark; /* wakeup watermark */
@@ -805,9 +804,9 @@ struct perf_output_handle {
struct perf_mmap_data *data;
unsigned long head;
unsigned long offset;
+ unsigned long wakeup;
int nmi;
int sample;
- int locked;
};
#ifdef CONFIG_PERF_EVENTS
@@ -994,7 +993,7 @@ static inline bool perf_paranoid_kernel(void)
extern void perf_event_init(void);
extern void perf_tp_event(int event_id, u64 addr, u64 count, void *record,
- int entry_size, struct pt_regs *regs);
+ int entry_size, struct pt_regs *regs, void *event);
extern void perf_bp_event(struct perf_event *event, void *data);
#ifndef perf_misc_flags
diff --git a/include/trace/ftrace.h b/include/trace/ftrace.h
index 16253db38d7..1016b216293 100644
--- a/include/trace/ftrace.h
+++ b/include/trace/ftrace.h
@@ -790,7 +790,8 @@ perf_trace_templ_##call(struct ftrace_event_call *event_call, \
{ assign; } \
\
perf_trace_buf_submit(entry, __entry_size, rctx, __addr, \
- __count, irq_flags, __regs); \
+ __count, irq_flags, __regs, \
+ event_call->perf_data); \
}
#undef DEFINE_EVENT
diff --git a/kernel/perf_event.c b/kernel/perf_event.c
index 511677bc1c6..2a060be3b07 100644
--- a/kernel/perf_event.c
+++ b/kernel/perf_event.c
@@ -2320,6 +2320,19 @@ perf_mmap_to_page(struct perf_mmap_data *data, unsigned long pgoff)
return virt_to_page(data->data_pages[pgoff - 1]);
}
+static void *perf_mmap_alloc_page(int cpu)
+{
+ struct page *page;
+ int node;
+
+ node = (cpu == -1) ? cpu : cpu_to_node(cpu);
+ page = alloc_pages_node(node, GFP_KERNEL | __GFP_ZERO, 0);
+ if (!page)
+ return NULL;
+
+ return page_address(page);
+}
+
static struct perf_mmap_data *
perf_mmap_data_alloc(struct perf_event *event, int nr_pages)
{
@@ -2336,12 +2349,12 @@ perf_mmap_data_alloc(struct perf_event *event, int nr_pages)
if (!data)
goto fail;
- data->user_page = (void *)get_zeroed_page(GFP_KERNEL);
+ data->user_page = perf_mmap_alloc_page(event->cpu);
if (!data->user_page)
goto fail_user_page;
for (i = 0; i < nr_pages; i++) {
- data->data_pages[i] = (void *)get_zeroed_page(GFP_KERNEL);
+ data->data_pages[i] = perf_mmap_alloc_page(event->cpu);
if (!data->data_pages[i])
goto fail_data_pages;
}
@@ -2506,8 +2519,6 @@ perf_mmap_data_init(struct perf_event *event, struct perf_mmap_data *data)
{
long max_size = perf_data_size(data);
- atomic_set(&data->lock, -1);
-
if (event->attr.watermark) {
data->watermark = min_t(long, max_size,
event->attr.wakeup_watermark);
@@ -2580,6 +2591,14 @@ static int perf_mmap(struct file *file, struct vm_area_struct *vma)
long user_extra, extra;
int ret = 0;
+ /*
+ * Don't allow mmap() of inherited per-task counters. This would
+ * create a performance issue due to all children writing to the
+ * same buffer.
+ */
+ if (event->cpu == -1 && event->attr.inherit)
+ return -EINVAL;
+
if (!(vma->vm_flags & VM_SHARED))
return -EINVAL;
@@ -2885,82 +2904,57 @@ static void perf_output_wakeup(struct perf_output_handle *handle)
}
/*
- * Curious locking construct.
- *
* We need to ensure a later event_id doesn't publish a head when a former
- * event_id isn't done writing. However since we need to deal with NMIs we
+ * event isn't done writing. However since we need to deal with NMIs we
* cannot fully serialize things.
*
- * What we do is serialize between CPUs so we only have to deal with NMI
- * nesting on a single CPU.
- *
* We only publish the head (and generate a wakeup) when the outer-most
- * event_id completes.
+ * event completes.
*/
-static void perf_output_lock(struct perf_output_handle *handle)
+static void perf_output_get_handle(struct perf_output_handle *handle)
{
struct perf_mmap_data *data = handle->data;
- int cur, cpu = get_cpu();
-
- handle->locked = 0;
-
- for (;;) {
- cur = atomic_cmpxchg(&data->lock, -1, cpu);
- if (cur == -1) {
- handle->locked = 1;
- break;
- }
- if (cur == cpu)
- break;
- cpu_relax();
- }
+ preempt_disable();
+ local_inc(&data->nest);
+ handle->wakeup = local_read(&data->wakeup);
}
-static void perf_output_unlock(struct perf_output_handle *handle)
+static void perf_output_put_handle(struct perf_output_handle *handle)
{
struct perf_mmap_data *data = handle->data;
unsigned long head;
- int cpu;
-
- data->done_head = data->head;
-
- if (!handle->locked)
- goto out;
again:
- /*
- * The xchg implies a full barrier that ensures all writes are done
- * before we publish the new head, matched by a rmb() in userspace when
- * reading this position.
- */
- while ((head = atomic_long_xchg(&data->done_head, 0)))
- data->user_page->data_head = head;
+ head = local_read(&data->head);
/*
- * NMI can happen here, which means we can miss a done_head update.
+ * IRQ/NMI can happen here, which means we can miss a head update.
*/
- cpu = atomic_xchg(&data->lock, -1);
- WARN_ON_ONCE(cpu != smp_processor_id());
+ if (!local_dec_and_test(&data->nest))
+ return;
/*
- * Therefore we have to validate we did not indeed do so.
+ * Publish the known good head. Rely on the full barrier implied
+ * by atomic_dec_and_test() order the data->head read and this
+ * write.
*/
- if (unlikely(atomic_long_read(&data->done_head))) {
- /*
- * Since we had it locked, we can lock it again.
- */
- while (atomic_cmpxchg(&data->lock, -1, cpu) != -1)
- cpu_relax();
+ data->user_page->data_head = head;
+ /*
+ * Now check if we missed an update, rely on the (compiler)
+ * barrier in atomic_dec_and_test() to re-read data->head.
+ */
+ if (unlikely(head != local_read(&data->head))) {
+ local_inc(&data->nest);
goto again;
}
- if (atomic_xchg(&data->wakeup, 0))
+ if (handle->wakeup != local_read(&data->wakeup))
perf_output_wakeup(handle);
-out:
- put_cpu();
+
+ preempt_enable();
}
void perf_output_copy(struct perf_output_handle *handle,
@@ -3036,13 +3030,13 @@ int perf_output_begin(struct perf_output_handle *handle,
handle->sample = sample;
if (!data->nr_pages)
- goto fail;
+ goto out;
- have_lost = atomic_read(&data->lost);
+ have_lost = local_read(&data->lost);
if (have_lost)
size += sizeof(lost_event);
- perf_output_lock(handle);
+ perf_output_get_handle(handle);
do {
/*
@@ -3052,24 +3046,24 @@ int perf_output_begin(struct perf_output_handle *handle,
*/
tail = ACCESS_ONCE(data->user_page->data_tail);
smp_rmb();
- offset = head = atomic_long_read(&data->head);
+ offset = head = local_read(&data->head);
head += size;
if (unlikely(!perf_output_space(data, tail, offset, head)))
goto fail;
- } while (atomic_long_cmpxchg(&data->head, offset, head) != offset);
+ } while (local_cmpxchg(&data->head, offset, head) != offset);
handle->offset = offset;
handle->head = head;
if (head - tail > data->watermark)
- atomic_set(&data->wakeup, 1);
+ local_inc(&data->wakeup);
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 = atomic_xchg(&data->lost, 0);
+ lost_event.lost = local_xchg(&data->lost, 0);
perf_output_put(handle, lost_event);
}
@@ -3077,8 +3071,8 @@ int perf_output_begin(struct perf_output_handle *handle,
return 0;
fail:
- atomic_inc(&data->lost);
- perf_output_unlock(handle);
+ local_inc(&data->lost);
+ perf_output_put_handle(handle);
out:
rcu_read_unlock();
@@ -3093,14 +3087,14 @@ void perf_output_end(struct perf_output_handle *handle)
int wakeup_events = event->attr.wakeup_events;
if (handle->sample && wakeup_events) {
- int events = atomic_inc_return(&data->events);
+ int events = local_inc_return(&data->events);
if (events >= wakeup_events) {
- atomic_sub(wakeup_events, &data->events);
- atomic_set(&data->wakeup, 1);
+ local_sub(wakeup_events, &data->events);
+ local_inc(&data->wakeup);
}
}
- perf_output_unlock(handle);
+ perf_output_put_handle(handle);
rcu_read_unlock();
}
@@ -3436,22 +3430,13 @@ static void perf_event_task_output(struct perf_event *event,
{
struct perf_output_handle handle;
struct task_struct *task = task_event->task;
- unsigned long flags;
int size, ret;
- /*
- * If this CPU attempts to acquire an rq lock held by a CPU spinning
- * in perf_output_lock() from interrupt context, it's game over.
- */
- local_irq_save(flags);
-
size = task_event->event_id.header.size;
ret = perf_output_begin(&handle, event, size, 0, 0);
- if (ret) {
- local_irq_restore(flags);
+ if (ret)
return;
- }
task_event->event_id.pid = perf_event_pid(event, task);
task_event->event_id.ppid = perf_event_pid(event, current);
@@ -3462,7 +3447,6 @@ static void perf_event_task_output(struct perf_event *event,
perf_output_put(&handle, task_event->event_id);
perf_output_end(&handle);
- local_irq_restore(flags);
}
static int perf_event_task_match(struct perf_event *event)
@@ -4502,8 +4486,9 @@ static int swevent_hlist_get(struct perf_event *event)
#ifdef CONFIG_EVENT_TRACING
void perf_tp_event(int event_id, u64 addr, u64 count, void *record,
- int entry_size, struct pt_regs *regs)
+ int entry_size, struct pt_regs *regs, void *event)
{
+ const int type = PERF_TYPE_TRACEPOINT;
struct perf_sample_data data;
struct perf_raw_record raw = {
.size = entry_size,
@@ -4513,9 +4498,13 @@ void perf_tp_event(int event_id, u64 addr, u64 count, void *record,
perf_sample_data_init(&data, addr);
data.raw = &raw;
- /* Trace events already protected against recursion */
- do_perf_sw_event(PERF_TYPE_TRACEPOINT, event_id, count, 1,
- &data, regs);
+ if (!event) {
+ do_perf_sw_event(type, event_id, count, 1, &data, regs);
+ return;
+ }
+
+ if (perf_swevent_match(event, type, event_id, &data, regs))
+ perf_swevent_add(event, count, 1, &data, regs);
}
EXPORT_SYMBOL_GPL(perf_tp_event);
@@ -4548,7 +4537,7 @@ static const struct pmu *tp_perf_event_init(struct perf_event *event)
!capable(CAP_SYS_ADMIN))
return ERR_PTR(-EPERM);
- if (perf_trace_enable(event->attr.config))
+ if (perf_trace_enable(event->attr.config, event))
return NULL;
event->destroy = tp_perf_event_destroy;
diff --git a/kernel/trace/trace_event_perf.c b/kernel/trace/trace_event_perf.c
index 0565bb42566..89b780a7c52 100644
--- a/kernel/trace/trace_event_perf.c
+++ b/kernel/trace/trace_event_perf.c
@@ -27,13 +27,15 @@ 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_enable(struct ftrace_event_call *event)
+static int perf_trace_event_enable(struct ftrace_event_call *event, void *data)
{
char *buf;
int ret = -ENOMEM;
- if (event->perf_refcount++ > 0)
+ if (event->perf_refcount++ > 0) {
+ event->perf_data = NULL;
return 0;
+ }
if (!total_ref_count) {
buf = (char *)alloc_percpu(perf_trace_t);
@@ -51,6 +53,7 @@ static int perf_trace_event_enable(struct ftrace_event_call *event)
ret = event->perf_event_enable(event);
if (!ret) {
+ event->perf_data = data;
total_ref_count++;
return 0;
}
@@ -68,7 +71,7 @@ fail_buf:
return ret;
}
-int perf_trace_enable(int event_id)
+int perf_trace_enable(int event_id, void *data)
{
struct ftrace_event_call *event;
int ret = -EINVAL;
@@ -77,7 +80,7 @@ int perf_trace_enable(int event_id)
list_for_each_entry(event, &ftrace_events, list) {
if (event->id == event_id && event->perf_event_enable &&
try_module_get(event->mod)) {
- ret = perf_trace_event_enable(event);
+ ret = perf_trace_event_enable(event, data);
break;
}
}
diff --git a/kernel/trace/trace_kprobe.c b/kernel/trace/trace_kprobe.c
index a7514326052..2d7bf4146be 100644
--- a/kernel/trace/trace_kprobe.c
+++ b/kernel/trace/trace_kprobe.c
@@ -1362,7 +1362,7 @@ static __kprobes void kprobe_perf_func(struct kprobe *kp,
for (i = 0; i < tp->nr_args; i++)
call_fetch(&tp->args[i].fetch, regs, data + tp->args[i].offset);
- perf_trace_buf_submit(entry, size, rctx, entry->ip, 1, irq_flags, regs);
+ perf_trace_buf_submit(entry, size, rctx, entry->ip, 1, irq_flags, regs, call->perf_data);
}
/* Kretprobe profile handler */
@@ -1395,7 +1395,7 @@ static __kprobes void kretprobe_perf_func(struct kretprobe_instance *ri,
call_fetch(&tp->args[i].fetch, regs, data + tp->args[i].offset);
perf_trace_buf_submit(entry, size, rctx, entry->ret_ip, 1,
- irq_flags, regs);
+ irq_flags, regs, call->perf_data);
}
static int probe_perf_enable(struct ftrace_event_call *call)
diff --git a/kernel/trace/trace_syscalls.c b/kernel/trace/trace_syscalls.c
index 4d6d711717f..9eff1a4b49b 100644
--- a/kernel/trace/trace_syscalls.c
+++ b/kernel/trace/trace_syscalls.c
@@ -468,7 +468,8 @@ static void perf_syscall_enter(struct pt_regs *regs, long id)
rec->nr = syscall_nr;
syscall_get_arguments(current, regs, 0, sys_data->nb_args,
(unsigned long *)&rec->args);
- perf_trace_buf_submit(rec, size, rctx, 0, 1, flags, regs);
+ perf_trace_buf_submit(rec, size, rctx, 0, 1, flags, regs,
+ sys_data->enter_event->perf_data);
}
int perf_sysenter_enable(struct ftrace_event_call *call)
@@ -543,7 +544,8 @@ static void perf_syscall_exit(struct pt_regs *regs, long ret)
rec->nr = sysca