[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20140806231539.GE1674@brightrain.aerifal.cx>
Date: Wed, 6 Aug 2014 19:15:39 -0400
From: Rich Felker <dalias@...c.org>
To: musl@...ts.openwall.com
Subject: Re: Explaining cond var destroy [Re: C threads, v3.0]
On Thu, Aug 07, 2014 at 12:43:35AM +0200, Jens Gustedt wrote:
> Am Mittwoch, den 06.08.2014, 18:04 -0400 schrieb Rich Felker:
> > No. The problem is that the memory in which the cond var resides could
> > previously have been used for a mutex, semaphore, or other object for
> > which a futex wait arrives "very late" due to scheduling. It has
> > nothing to do with the destruction of cond vars.
>
> Hm, but then we have a problem on the side that has this late wait,
> that is an access to a dead object. So the UB is there, in that thread
> that is doing the wait. (Suppose it is just threads and not processes,
> that would even be more complicated.) That spurious wakeup that you
> are talking about is a consequence of UB.
No, it's not. The wait happens prior to the deallocation, in the same
thread that performs the deallocation. The interleaving looks like
this:
Thread A Thread B
-------- --------
waiters++
save waiters count
atomic unlock
futex wait fails with EAGAIN
cas succeeds & gets lock
waiters--
[unlock operation]
[free operation]
futex wake to freed address
The free operation in thread A is valid since A knows it is the last
user of the mutex and thread B's use/ownership of the mutex formally
ends with the atomic unlock.
Rich
Powered by blists - more mailing lists
Confused about mailing lists and their use?
Read about mailing lists on Wikipedia
and check out these
guidelines on proper formatting of your messages.
