From: Waiman Long <waiman.long@hpe.com>
To: Peter Zijlstra <peterz@infradead.org>
Cc: Davidlohr Bueso <dave@stgolabs.net>, <manfred@colorfullife.com>,
<mingo@kernel.org>, <torvalds@linux-foundation.org>,
<ggherdovich@suse.com>, <mgorman@techsingularity.net>,
<linux-kernel@vger.kernel.org>,
Paul McKenney <paulmck@linux.vnet.ibm.com>,
Will Deacon <will.deacon@arm.com>
Subject: Re: sem_lock() vs qspinlocks
Date: Fri, 20 May 2016 16:44:19 -0400 [thread overview]
Message-ID: <573F7723.8030201@hpe.com> (raw)
In-Reply-To: <20160520115819.GF3193@twins.programming.kicks-ass.net>
On 05/20/2016 07:58 AM, Peter Zijlstra wrote:
> On Thu, May 19, 2016 at 10:39:26PM -0700, Davidlohr Bueso wrote:
>> As such, the following restores the behavior of the ticket locks and 'fixes'
>> (or hides?) the bug in sems. Naturally incorrect approach:
>>
>> @@ -290,7 +290,8 @@ static void sem_wait_array(struct sem_array *sma)
>>
>> for (i = 0; i< sma->sem_nsems; i++) {
>> sem = sma->sem_base + i;
>> - spin_unlock_wait(&sem->lock);
>> + while (atomic_read(&sem->lock))
>> + cpu_relax();
>> }
>> ipc_smp_acquire__after_spin_is_unlocked();
>> }
> The actual bug is clear_pending_set_locked() not having acquire
> semantics. And the above 'fixes' things because it will observe the old
> pending bit or the locked bit, so it doesn't matter if the store
> flipping them is delayed.
The clear_pending_set_locked() is not the only place where the lock is
set. If there are more than one waiter, the queuing patch will be used
instead. The set_locked(), which is also an unordered store, will then
be used to set the lock.
Cheers,
Longman
next prev parent reply other threads:[~2016-05-20 20:44 UTC|newest]
Thread overview: 41+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-05-20 5:39 sem_lock() vs qspinlocks Davidlohr Bueso
2016-05-20 7:49 ` Peter Zijlstra
2016-05-20 15:00 ` Davidlohr Bueso
2016-05-20 15:05 ` Peter Zijlstra
2016-05-20 15:25 ` Davidlohr Bueso
2016-05-20 15:28 ` Peter Zijlstra
2016-05-20 20:47 ` Waiman Long
2016-05-20 20:52 ` Peter Zijlstra
2016-05-21 0:59 ` Davidlohr Bueso
2016-05-21 4:01 ` Waiman Long
2016-05-21 7:40 ` Peter Zijlstra
2016-05-20 7:53 ` Peter Zijlstra
2016-05-20 8:13 ` Peter Zijlstra
2016-05-20 8:18 ` Peter Zijlstra
2016-05-20 9:07 ` Giovanni Gherdovich
2016-05-20 9:34 ` Peter Zijlstra
2016-05-20 8:30 ` Peter Zijlstra
2016-05-20 9:00 ` Peter Zijlstra
2016-05-20 10:09 ` Ingo Molnar
2016-05-20 10:45 ` Mel Gorman
2016-05-20 11:58 ` Peter Zijlstra
2016-05-20 14:05 ` Boqun Feng
2016-05-20 15:21 ` Peter Zijlstra
2016-05-20 16:04 ` Peter Zijlstra
2016-05-20 17:00 ` Linus Torvalds
2016-05-20 21:06 ` Peter Zijlstra
2016-05-20 21:44 ` Linus Torvalds
2016-05-21 0:48 ` Davidlohr Bueso
2016-05-21 2:30 ` Linus Torvalds
2016-05-21 7:37 ` Peter Zijlstra
2016-05-21 13:49 ` Manfred Spraul
2016-05-24 10:57 ` Peter Zijlstra
2016-05-21 17:14 ` Davidlohr Bueso
2016-05-23 12:25 ` Peter Zijlstra
2016-05-23 17:52 ` Linus Torvalds
2016-05-25 6:37 ` Boqun Feng
2016-05-22 8:43 ` Manfred Spraul
2016-05-22 9:38 ` Peter Zijlstra
2016-05-20 16:20 ` Davidlohr Bueso
2016-05-20 20:44 ` Waiman Long [this message]
2016-05-20 20:53 ` Peter Zijlstra
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=573F7723.8030201@hpe.com \
--to=waiman.long@hpe.com \
--cc=dave@stgolabs.net \
--cc=ggherdovich@suse.com \
--cc=linux-kernel@vger.kernel.org \
--cc=manfred@colorfullife.com \
--cc=mgorman@techsingularity.net \
--cc=mingo@kernel.org \
--cc=paulmck@linux.vnet.ibm.com \
--cc=peterz@infradead.org \
--cc=torvalds@linux-foundation.org \
--cc=will.deacon@arm.com \
/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).