From: Peter Zijlstra <peterz@infradead.org>
To: Vince Weaver <vincent.weaver@maine.edu>
Cc: Thomas Gleixner <tglx@linutronix.de>,
Ingo Molnar <mingo@kernel.org>,
linux-kernel@vger.kernel.org,
Steven Rostedt <rostedt@goodmis.org>
Subject: Re: [perf] more perf_fuzzer memory corruption
Date: Fri, 2 May 2014 17:42:17 +0200 [thread overview]
Message-ID: <20140502154217.GW11096@twins.programming.kicks-ass.net> (raw)
In-Reply-To: <alpine.DEB.2.10.1405011443410.19874@vincent-weaver-1.umelst.maine.edu>
[-- Attachment #1: Type: text/plain, Size: 6331 bytes --]
On Thu, May 01, 2014 at 02:49:01PM -0400, Vince Weaver wrote:
> It is a rance condition of sorts, because it's just a 10us or so
> interleaving of calls that causes the bug to happen or not.
>
> In the good trace:
>
> [parent] __perf_event_task_sched_out (and hence perf_swevent_del)
> [child] perf_release
>
> In the buggy trace:
>
> [child] perf_release
> [parent] __perf_event_task_sched_out (perf_swevent_del never happens)
>
Can you give this a spin?
---
Subject: perf: Fix race in removing an event
From: Peter Zijlstra <peterz@infradead.org>
Date: Fri May 2 16:56:01 CEST 2014
When removing a (sibling) event we do:
raw_spin_lock_irq(&ctx->lock);
perf_group_detach(event);
raw_spin_unlock_irq(&ctx->lock);
<hole>
perf_remove_from_context(event);
raw_spin_lock_irq(&ctx->lock);
...
raw_spin_unlock_irq(&ctx->lock);
Now, assuming the event is a sibling, it will be 'unreachable' for
things like ctx_sched_out() because that iterates the
groups->siblings, and we just unhooked the sibling.
So, if during <hole> we get ctx_sched_out(), it will miss the event
and not call event_sched_out() on it, leaving it programmed on the
PMU.
The subsequent perf_remove_from_context() call will find the ctx is
inactive and only call list_del_event() to remove the event from all
other lists.
Hereafter we can proceed to free the event; while still programmed!
Close this hole by moving perf_group_detach() inside the same
ctx->lock region(s) perf_remove_from_context() has.
The condition on inherited events only in __perf_event_exit_task() is
likely complete crap because non-inherited events are part of groups
too and we're tearing down just the same. But leave that for another
patch.
Reported-by: Vince Weaver <vincent.weaver@maine.edu>
Much-staring-at-traces-by: Vince Weaver <vincent.weaver@maine.edu>
Much-staring-at-traces-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Peter Zijlstra <peterz@infradead.org>
---
kernel/events/core.c | 41 +++++++++++++++++++++++------------------
1 file changed, 23 insertions(+), 18 deletions(-)
--- a/kernel/events/core.c
+++ b/kernel/events/core.c
@@ -1444,6 +1444,11 @@ group_sched_out(struct perf_event *group
cpuctx->exclusive = 0;
}
+struct remove_event {
+ struct perf_event *event;
+ bool detach_group;
+};
+
/*
* Cross CPU call to remove a performance event
*
@@ -1452,12 +1457,15 @@ group_sched_out(struct perf_event *group
*/
static int __perf_remove_from_context(void *info)
{
- struct perf_event *event = info;
+ struct remove_event *re = info;
+ struct perf_event *event = re->event;
struct perf_event_context *ctx = event->ctx;
struct perf_cpu_context *cpuctx = __get_cpu_context(ctx);
raw_spin_lock(&ctx->lock);
event_sched_out(event, cpuctx, ctx);
+ if (re->detach_group)
+ perf_group_detach(event);
list_del_event(event, ctx);
if (!ctx->nr_events && cpuctx->task_ctx == ctx) {
ctx->is_active = 0;
@@ -1482,10 +1490,14 @@ static int __perf_remove_from_context(vo
* When called from perf_event_exit_task, it's OK because the
* context has been detached from its task.
*/
-static void perf_remove_from_context(struct perf_event *event)
+static void perf_remove_from_context(struct perf_event *event, bool detach_group)
{
struct perf_event_context *ctx = event->ctx;
struct task_struct *task = ctx->task;
+ struct remove_event re = {
+ .event = event,
+ .detach_group = detach_group,
+ };
lockdep_assert_held(&ctx->mutex);
@@ -1494,12 +1506,12 @@ static void perf_remove_from_context(str
* Per cpu events are removed via an smp call and
* the removal is always successful.
*/
- cpu_function_call(event->cpu, __perf_remove_from_context, event);
+ cpu_function_call(event->cpu, __perf_remove_from_context, &re);
return;
}
retry:
- if (!task_function_call(task, __perf_remove_from_context, event))
+ if (!task_function_call(task, __perf_remove_from_context, &re))
return;
raw_spin_lock_irq(&ctx->lock);
@@ -1516,6 +1528,8 @@ static void perf_remove_from_context(str
* Since the task isn't running, its safe to remove the event, us
* holding the ctx->lock ensures the task won't get scheduled in.
*/
+ if (detach_group)
+ perf_group_detach(event);
list_del_event(event, ctx);
raw_spin_unlock_irq(&ctx->lock);
}
@@ -3285,10 +3299,7 @@ int perf_event_release_kernel(struct per
* to trigger the AB-BA case.
*/
mutex_lock_nested(&ctx->mutex, SINGLE_DEPTH_NESTING);
- raw_spin_lock_irq(&ctx->lock);
- perf_group_detach(event);
- raw_spin_unlock_irq(&ctx->lock);
- perf_remove_from_context(event);
+ perf_remove_from_context(event, true);
mutex_unlock(&ctx->mutex);
free_event(event);
@@ -7180,7 +7191,7 @@ SYSCALL_DEFINE5(perf_event_open,
struct perf_event_context *gctx = group_leader->ctx;
mutex_lock(&gctx->mutex);
- perf_remove_from_context(group_leader);
+ perf_remove_from_context(group_leader, false);
/*
* Removing from the context ends up with disabled
@@ -7190,7 +7201,7 @@ SYSCALL_DEFINE5(perf_event_open,
perf_event__state_init(group_leader);
list_for_each_entry(sibling, &group_leader->sibling_list,
group_entry) {
- perf_remove_from_context(sibling);
+ perf_remove_from_context(sibling, false);
perf_event__state_init(sibling);
put_ctx(gctx);
}
@@ -7320,7 +7331,7 @@ void perf_pmu_migrate_context(struct pmu
mutex_lock(&src_ctx->mutex);
list_for_each_entry_safe(event, tmp, &src_ctx->event_list,
event_entry) {
- perf_remove_from_context(event);
+ perf_remove_from_context(event, false);
unaccount_event_cpu(event, src_cpu);
put_ctx(src_ctx);
list_add(&event->migrate_entry, &events);
@@ -7382,13 +7393,7 @@ __perf_event_exit_task(struct perf_event
struct perf_event_context *child_ctx,
struct task_struct *child)
{
- if (child_event->parent) {
- raw_spin_lock_irq(&child_ctx->lock);
- perf_group_detach(child_event);
- raw_spin_unlock_irq(&child_ctx->lock);
- }
-
- perf_remove_from_context(child_event);
+ perf_remove_from_context(child_event, !!child_event->parent);
/*
* It can happen that the parent exits first, and has events
[-- Attachment #2: Type: application/pgp-signature, Size: 836 bytes --]
next prev parent reply other threads:[~2014-05-02 15:42 UTC|newest]
Thread overview: 81+ messages / expand[flat|nested] mbox.gz Atom feed top
2014-04-15 21:37 [perf] more perf_fuzzer memory corruption Vince Weaver
2014-04-15 21:49 ` Thomas Gleixner
2014-04-16 3:21 ` Vince Weaver
2014-04-16 4:18 ` Vince Weaver
2014-04-16 14:15 ` Peter Zijlstra
2014-04-16 17:30 ` Vince Weaver
2014-04-16 17:43 ` Vince Weaver
2014-04-16 17:47 ` Peter Zijlstra
2014-04-17 9:48 ` Ingo Molnar
2014-04-17 11:45 ` Peter Zijlstra
2014-04-17 14:22 ` Ingo Molnar
2014-04-17 14:42 ` Vince Weaver
2014-04-17 14:54 ` Peter Zijlstra
2014-04-17 15:35 ` Vince Weaver
2014-04-18 14:45 ` Vince Weaver
2014-04-18 14:51 ` Vince Weaver
2014-04-18 15:23 ` Peter Zijlstra
2014-04-18 16:59 ` Peter Zijlstra
2014-04-18 17:15 ` Peter Zijlstra
2014-04-23 20:58 ` Vince Weaver
2014-04-25 2:51 ` Vince Weaver
2014-04-28 14:21 ` Vince Weaver
2014-04-28 19:38 ` Vince Weaver
2014-04-29 9:46 ` Peter Zijlstra
2014-04-29 18:21 ` Vince Weaver
2014-04-29 19:01 ` Peter Zijlstra
2014-04-29 20:59 ` Vince Weaver
2014-04-30 18:44 ` Peter Zijlstra
2014-04-30 21:08 ` Vince Weaver
2014-04-30 22:51 ` Thomas Gleixner
2014-05-01 10:26 ` Peter Zijlstra
2014-05-01 11:50 ` Peter Zijlstra
2014-05-01 12:35 ` Thomas Gleixner
2014-05-01 13:12 ` Peter Zijlstra
2014-05-01 13:29 ` Thomas Gleixner
2014-05-01 13:22 ` Vince Weaver
2014-05-01 14:07 ` Vince Weaver
2014-05-01 14:27 ` Vince Weaver
2014-05-01 15:09 ` Peter Zijlstra
2014-05-01 15:50 ` Vince Weaver
2014-05-01 16:31 ` Thomas Gleixner
2014-05-01 17:18 ` Vince Weaver
2014-05-01 18:49 ` Vince Weaver
2014-05-01 21:32 ` Vince Weaver
2014-05-02 11:15 ` Peter Zijlstra
2014-05-02 15:42 ` Peter Zijlstra [this message]
2014-05-02 16:22 ` Vince Weaver
2014-05-02 16:22 ` Peter Zijlstra
2014-05-02 16:43 ` Vince Weaver
2014-05-02 17:27 ` Peter Zijlstra
2014-05-02 17:46 ` Vince Weaver
2014-05-02 19:12 ` Thomas Gleixner
2014-05-02 20:15 ` Vince Weaver
2014-05-02 20:45 ` Thomas Gleixner
2014-05-03 2:32 ` Vince Weaver
2014-05-03 3:02 ` Vince Weaver
2014-05-03 7:33 ` Peter Zijlstra
2014-05-05 9:31 ` Peter Zijlstra
2014-05-05 16:00 ` Vince Weaver
2014-05-05 17:10 ` Vince Weaver
2014-05-05 17:14 ` Peter Zijlstra
2014-05-05 18:47 ` Vince Weaver
2014-05-05 19:36 ` Peter Zijlstra
2014-05-05 19:51 ` Vince Weaver
2014-05-06 1:06 ` Vince Weaver
2014-05-06 16:57 ` Vince Weaver
2014-05-07 16:45 ` Peter Zijlstra
2014-05-08 10:40 ` [tip:perf/core] perf: Fix perf_event_init_context() tip-bot for Peter Zijlstra
2014-05-05 17:29 ` [perf] more perf_fuzzer memory corruption Ingo Molnar
2014-05-06 4:51 ` Vince Weaver
2014-05-06 17:06 ` Vince Weaver
2014-05-07 19:12 ` Ingo Molnar
2014-05-07 19:11 ` Ingo Molnar
2014-05-08 10:40 ` [tip:perf/core] perf: Fix race in removing an event tip-bot for Peter Zijlstra
2014-05-02 17:06 ` [perf] more perf_fuzzer memory corruption Vince Weaver
2014-05-02 17:04 ` Peter Zijlstra
2014-04-29 19:26 ` Steven Rostedt
2014-04-29 8:52 ` Peter Zijlstra
2014-04-29 18:11 ` Vince Weaver
2014-04-29 19:21 ` Steven Rostedt
2014-04-28 17:48 ` Thomas Gleixner
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20140502154217.GW11096@twins.programming.kicks-ass.net \
--to=peterz@infradead.org \
--cc=linux-kernel@vger.kernel.org \
--cc=mingo@kernel.org \
--cc=rostedt@goodmis.org \
--cc=tglx@linutronix.de \
--cc=vincent.weaver@maine.edu \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).