From: Darren Hart <dvhltc@us.ibm.com>
To: linux-rt-users@vger.kernel.org, linux-kernel@vger.kernel.org
Cc: tglx@linutronix.de, peterz@infradead.org, rostedt@goodmis.org,
mingo@elte.hu, dino@in.ibm.com, johnstul@us.ibm.com,
John Kacur <jkacur@redhat.com>
Subject: Re: [PATCH 1/2] Update woken requeued futex_q lock_ptr
Date: Thu, 06 Aug 2009 17:24:41 -0700 [thread overview]
Message-ID: <4A7B7449.7070008@us.ibm.com> (raw)
In-Reply-To: <4A7A66D7.3090203@us.ibm.com>
Darren Hart wrote:
> Darren Hart wrote:
>> futex_requeue() can acquire the lock on behalf of a waiter during the
>> requeue
>> loop in the event of a lock steal or owner died.
>> futex_wait_requeue_pi() cleans
>> up the pi_state owner, using the lock_ptr to protect against
>> concurrent access
>> to the pi_state. The pi_state is found on the requeue target futex
>> hash bucket
>> so the lock_ptr needs to be updated accordingly. The problem
>> manifested by
>> triggering the WARN_ON in lookup_pi_state() about the pid !=
>> pi_state->owner
>> pid.
>> The astute reviewer will note that still exists a race between the time
>> futex_requeue() releases hb2->lock() and the time when
>> futex_wait_requeue_pi()
>> acquires it. During this time the pi_state and the futex uaddr are
>> not in sync
>> with the rt_mutex ownership. This patch closes the window to the
>> point where
>> my tests now pass, but we still need to address it.
>>
>> Note: Please apply to mainline and rt
>>
>
>
>> static inline
>> -void requeue_pi_wake_futex(struct futex_q *q, union futex_key *key)
>> +void requeue_pi_wake_futex(struct futex_q *q, union futex_key *key,
>> + struct futex_hash_bucket *hb)
>> {
>> drop_futex_key_refs(&q->key);
>> get_futex_key_refs(key);
>> q->key = *key;
>> + q->lock_ptr = &hb->lock;
>
> Hrm... turns out changing this breaks the
> handle_early_requeue_pi_wakeup() logic. I'll have to respin this patch
> to account for that as well. Please hold off on this patch.
In fact, this doesn't affect the handle_early_requeue_pi_wakeup() code
in the slightest. It only needs to hold a hb->lock (either one is
adequate) to ensure the requeue routine has completed. By changing the
q->lock_ptr of the waiter to hb2->lock we ensure the pi_state is
protected from concurrent access by futex_wait_requeue_pi() and new
contending threads.
Ingo, please apply to tip/urgent.
Thanks,
--
Darren Hart
IBM Linux Technology Center
Real-Time Linux Team
next prev parent reply other threads:[~2009-08-07 0:24 UTC|newest]
Thread overview: 10+ messages / expand[flat|nested] mbox.gz Atom feed top
2009-08-05 21:58 [PATCH 0/2] futex: requeue_pi lock steal deadlock fixes Darren Hart
2009-08-05 22:02 ` [PATCH 1/2] Update woken requeued futex_q lock_ptr Darren Hart
2009-08-06 5:15 ` Darren Hart
2009-08-07 0:24 ` Darren Hart [this message]
2009-08-08 15:27 ` [tip:core/urgent] futex: " tip-bot for Darren Hart
2009-08-09 20:24 ` tip-bot for Darren Hart
2009-08-09 20:56 ` Ingo Molnar
2009-08-09 22:03 ` Darren Hart
2009-08-09 22:18 ` Darren Hart
2009-08-05 22:04 ` [PATCH 2/2][RT] Avoid deadlock in rt_mutex_start_proxy_lock() Darren Hart
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=4A7B7449.7070008@us.ibm.com \
--to=dvhltc@us.ibm.com \
--cc=dino@in.ibm.com \
--cc=jkacur@redhat.com \
--cc=johnstul@us.ibm.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-rt-users@vger.kernel.org \
--cc=mingo@elte.hu \
--cc=peterz@infradead.org \
--cc=rostedt@goodmis.org \
--cc=tglx@linutronix.de \
/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