diff options
| author | Peter Zijlstra <a.p.zijlstra@chello.nl> | 2009-06-19 13:22:51 +0200 |
|---|---|---|
| committer | Ingo Molnar <mingo@elte.hu> | 2009-06-19 13:43:12 +0200 |
| commit | e5289d4a181fb6c0b7a7607649af2ffdc491335c (patch) | |
| tree | 6c9e9ab46a29c538e4c505d3d3cd135265db08a2 /kernel | |
| parent | f5970550d5ccf90453cbd7d260370ea99d1f6513 (diff) | |
| download | linux-e5289d4a181fb6c0b7a7607649af2ffdc491335c.tar.gz linux-e5289d4a181fb6c0b7a7607649af2ffdc491335c.tar.bz2 linux-e5289d4a181fb6c0b7a7607649af2ffdc491335c.zip | |
perf_counter: Simplify and fix task migration counting
The task migrations counter was causing rare and hard to decypher
memory corruptions under load. After a day of debugging and bisection
we found that the problem was introduced with:
3f731ca: perf_counter: Fix cpu migration counter
Turning them off fixes the crashes. Incidentally, the whole
perf_counter_task_migration() logic can be done simpler as well,
by injecting a proper sw-counter event.
This cleanup also fixed the crashes. The precise failure mode is
not completely clear yet, but we are clearly not unhappy about
having a fix ;-)
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Corey Ashford <cjashfor@linux.vnet.ibm.com>
Cc: Marcelo Tosatti <mtosatti@redhat.com>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
LKML-Reference: <new-submission>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Diffstat (limited to 'kernel')
| -rw-r--r-- | kernel/perf_counter.c | 23 | ||||
| -rw-r--r-- | kernel/sched.c | 3 |
2 files changed, 3 insertions, 23 deletions
diff --git a/kernel/perf_counter.c b/kernel/perf_counter.c index 7e9108efd305..8d4f0dd41c22 100644 --- a/kernel/perf_counter.c +++ b/kernel/perf_counter.c @@ -124,7 +124,7 @@ void perf_enable(void) static void get_ctx(struct perf_counter_context *ctx) { - atomic_inc(&ctx->refcount); + WARN_ON(!atomic_inc_not_zero(&ctx->refcount)); } static void free_ctx(struct rcu_head *head) @@ -3467,27 +3467,6 @@ static const struct pmu perf_ops_task_clock = { .read = task_clock_perf_counter_read, }; -/* - * Software counter: cpu migrations - */ -void perf_counter_task_migration(struct task_struct *task, int cpu) -{ - struct perf_cpu_context *cpuctx = &per_cpu(perf_cpu_context, cpu); - struct perf_counter_context *ctx; - - perf_swcounter_ctx_event(&cpuctx->ctx, PERF_TYPE_SOFTWARE, - PERF_COUNT_SW_CPU_MIGRATIONS, - 1, 1, NULL, 0); - - ctx = perf_pin_task_context(task); - if (ctx) { - perf_swcounter_ctx_event(ctx, PERF_TYPE_SOFTWARE, - PERF_COUNT_SW_CPU_MIGRATIONS, - 1, 1, NULL, 0); - perf_unpin_context(ctx); - } -} - #ifdef CONFIG_EVENT_PROFILE void perf_tpcounter_event(int event_id) { diff --git a/kernel/sched.c b/kernel/sched.c index 8fb88a906aaa..f46540b359c0 100644 --- a/kernel/sched.c +++ b/kernel/sched.c @@ -1978,7 +1978,8 @@ void set_task_cpu(struct task_struct *p, unsigned int new_cpu) if (task_hot(p, old_rq->clock, NULL)) schedstat_inc(p, se.nr_forced2_migrations); #endif - perf_counter_task_migration(p, new_cpu); + perf_swcounter_event(PERF_COUNT_SW_CPU_MIGRATIONS, + 1, 1, NULL, 0); } p->se.vruntime -= old_cfsrq->min_vruntime - new_cfsrq->min_vruntime; |