[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]

Re: pingpong, etc. programs



-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Saurabh Desai wrote:

> The pthread_cond_broadcast wakes up 200+ threads and they all contend
> for 
> an internal condvar lock. And again they wait in kernel using futex. 
> To verify this I tried adding this patch in the library and that made a
> hugh difference.

This is just what I described earlier.

But the solution is not to aritificially serialize the wakeup.  This
won't help either; the wakeup happens in the condvar function while the
contended lock is freed by a pthread_mutex_unlock call (or inside the
next pthread_cond_wait call).  The fact that your patch has good effects
is noteworthy but still a hack.

The correct solution is to have the kernel handle the wakeup correctly.
 Our current thinking is to wake up only one thread.  All the other
threads are moved from the wait-queue of the condvar to the wait-queue
of the mutex.  Since there is exactly one mutex involved with every
condvar at any one time the pthread_cond_broadcast function knows where
the mutex wait list is.

I just have to wait for Ingo to have some time to implement this.

- -- 
- --------------.                        ,-.            444 Castro Street
Ulrich Drepper \    ,-----------------'   \ Mountain View, CA 94041 USA
Red Hat         `--' drepper at redhat.com `---------------------------
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.1 (GNU/Linux)

iD8DBQE+sG/m2ijCOnn/RHQRAlxDAJ98J6DHF1mgHTj0m83ZZovhUr++OgCguai2
2f+Dt6rH/rg5yxtyLMbgLjs=
=818N
-----END PGP SIGNATURE-----





[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]