On Thursday 28 of June 2007, Jeff Johnson wrote:

> > 2.6.16 actually. 2.6.17 works fine.
> >
> > So my problem is kernel related problem only.
>
> Could you supply details please? I don't understand
> why a EOWNERDEAD driven patch should fail in
> the manner described.
>
> EOWNERDEAD should not be returned by a kernel
> that does not support "robust" mutexes, and so the
> code paths in the patch should never have been
> traversed.

That's returned by glibc 2.6 nptl

http://sourceware.org/cgi-bin/cvsweb.cgi/libc/nptl/pthread_mutex_lock.c?rev=1.15&content-type=text/x-cvsweb-markup&cvsroot=glibc&only_with_tag=glibc-2_6-branch

Finding out why means digging into nptl internals.

> If there is a path that is traversed by a kernel that does
> not support "robust" mutexes, than that needs to be fixed
> for the (duh!) obvious reasons.


-- 
Arkadiusz Mi¶kiewicz        PLD/Linux Team
arekm / maven.pl            http://ftp.pld-linux.org/