From: Haifeng Xu <haifeng.xu@shopee.com>
To: Frederic Weisbecker <frederic@kernel.org>
Cc: peterz@infradead.org, mingo@redhat.com, acme@kernel.org,
mark.rutland@arm.com, alexander.shishkin@linux.intel.com,
jolsa@kernel.org, namhyung@kernel.org, irogers@google.com,
adrian.hunter@intel.com, linux-perf-users@vger.kernel.org,
linux-kernel@vger.kernel.org
Subject: Re: [PATCH] perf/core: Fix missing wakeup when waiting for context reference
Date: Thu, 18 Apr 2024 11:26:09 +0800 [thread overview]
Message-ID: <6af2cb1b-4d93-4fd4-a5f5-0fdf2ae9d759@shopee.com> (raw)
In-Reply-To: <Zh_8FpT10qoE-x1u@localhost.localdomain>
On 2024/4/18 00:43, Frederic Weisbecker wrote:
> Le Wed, Apr 10, 2024 at 03:55:06AM +0000, Haifeng Xu a écrit :
>> In our production environment, we found many hung tasks which are
>> blocked for more than 18 hours. Their call traces are like this:
>>
>> [346278.191038] __schedule+0x2d8/0x890
>> [346278.191046] schedule+0x4e/0xb0
>> [346278.191049] perf_event_free_task+0x220/0x270
>> [346278.191056] ? init_wait_var_entry+0x50/0x50
>> [346278.191060] copy_process+0x663/0x18d0
>> [346278.191068] kernel_clone+0x9d/0x3d0
>> [346278.191072] __do_sys_clone+0x5d/0x80
>> [346278.191076] __x64_sys_clone+0x25/0x30
>> [346278.191079] do_syscall_64+0x5c/0xc0
>> [346278.191083] ? syscall_exit_to_user_mode+0x27/0x50
>> [346278.191086] ? do_syscall_64+0x69/0xc0
>> [346278.191088] ? irqentry_exit_to_user_mode+0x9/0x20
>> [346278.191092] ? irqentry_exit+0x19/0x30
>> [346278.191095] ? exc_page_fault+0x89/0x160
>> [346278.191097] ? asm_exc_page_fault+0x8/0x30
>> [346278.191102] entry_SYSCALL_64_after_hwframe+0x44/0xae
>>
>> The task was waiting for the refcount become to 1, but from the vmcore,
>> we found the refcount has already been 1. It seems that the task didn't
>> get woken up by perf_event_release_kernel() and got stuck forever. The
>> below scenario may cause the problem.
>>
>> Thread A Thread B
>> ... ...
>> perf_event_free_task perf_event_release_kernel
>> ...
>> acquire event->child_mutex
>> ...
>> get_ctx
>> ... release event->child_mutex
>> acquire ctx->mutex
>> ...
>> perf_free_event (acquire/release event->child_mutex)
>> ...
>> release ctx->mutex
>> wait_var_event
>> acquire ctx->mutex
>> acquire event->mutex
>> # move existing events to free_list
>> release event->child_mutex
>> release ctx->mutex
>> put_ctx
>> ... ...
>>
>> In this case, all events of the ctx have been freed, so we couldn't
>> find the ctx in free_list and Thread A will miss the wakeup. It's thus
>> necessary to add a wakeup after dropping the reference.
>>
>> Signed-off-by: Haifeng Xu <haifeng.xu@shopee.com>
>> ---
>> kernel/events/core.c | 16 ++++++++++++++++
>> 1 file changed, 16 insertions(+)
>>
>> diff --git a/kernel/events/core.c b/kernel/events/core.c
>> index 4f0c45ab8d7d..01dfe715f09e 100644
>> --- a/kernel/events/core.c
>> +++ b/kernel/events/core.c
>> @@ -5340,6 +5340,8 @@ int perf_event_release_kernel(struct perf_event *event)
>> again:
>> mutex_lock(&event->child_mutex);
>> list_for_each_entry(child, &event->child_list, child_list) {
>> + void *var;
>> + bool freed = false;
>>
>> /*
>> * Cannot change, child events are not migrated, see the
>> @@ -5380,11 +5382,25 @@ int perf_event_release_kernel(struct perf_event *event)
>> * this can't be the last reference.
>> */
>> put_event(event);
>> + } else {
>> + freed = true;
>> + var = &ctx->refcount;
>> }
>>
>> mutex_unlock(&event->child_mutex);
>> mutex_unlock(&ctx->mutex);
>> put_ctx(ctx);
>> +
>> + if (freed) {
>> + /*
>> + * perf_event_free_task() delete all events of the ctx and
>> + * there is no event of the ctx in free_list. It may step
>> + * into wait_var_event() before decrement the refcount. So
>> + * we should add a wakeup here.
>> + */
>> + smp_mb(); /* pairs with wait_var_event() */
>> + wake_up_var(var);
>> + }
>> goto again;
>
> Good catch!
>
> How about the following slightly simplified version?
Yes. I'll send next verion with your suggestions and add:
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
>
> diff --git a/kernel/events/core.c b/kernel/events/core.c
> index 724e6d7e128f..4082d0161b2b 100644
> --- a/kernel/events/core.c
> +++ b/kernel/events/core.c
> @@ -5365,6 +5365,7 @@ int perf_event_release_kernel(struct perf_event *event)
> again:
> mutex_lock(&event->child_mutex);
> list_for_each_entry(child, &event->child_list, child_list) {
> + void *var = NULL;
>
> /*
> * Cannot change, child events are not migrated, see the
> @@ -5405,11 +5406,23 @@ int perf_event_release_kernel(struct perf_event *event)
> * this can't be the last reference.
> */
> put_event(event);
> + } else {
> + var = &ctx->refcount;
> }
>
> mutex_unlock(&event->child_mutex);
> mutex_unlock(&ctx->mutex);
> put_ctx(ctx);
> +
> + if (var) {
> + /*
> + * If perf_event_free_task() has deleted all events from the
> + * ctx while the child_mutex got released above, make sure to
> + * notify about the preceding put_ctx().
> + */
> + smp_mb(); /* pairs with wait_var_event() */
> + wake_up_var(var);
> + }
> goto again;
> }
> mutex_unlock(&event->child_mutex);
>
>
>
>> }
>> mutex_unlock(&event->child_mutex);
>> --
>> 2.25.1
>>
>>
prev parent reply other threads:[~2024-04-18 3:26 UTC|newest]
Thread overview: 3+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-04-10 3:55 [PATCH] perf/core: Fix missing wakeup when waiting for context reference Haifeng Xu
2024-04-17 16:43 ` Frederic Weisbecker
2024-04-18 3:26 ` Haifeng Xu [this message]
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=6af2cb1b-4d93-4fd4-a5f5-0fdf2ae9d759@shopee.com \
--to=haifeng.xu@shopee.com \
--cc=acme@kernel.org \
--cc=adrian.hunter@intel.com \
--cc=alexander.shishkin@linux.intel.com \
--cc=frederic@kernel.org \
--cc=irogers@google.com \
--cc=jolsa@kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-perf-users@vger.kernel.org \
--cc=mark.rutland@arm.com \
--cc=mingo@redhat.com \
--cc=namhyung@kernel.org \
--cc=peterz@infradead.org \
/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 read-only IMAP folder(s) and NNTP newsgroup(s).