--=-FX3hdJRzauHV7D5x423w
Content-Type: text/plain
Content-Transfer-Encoding: quoted-printable

On Wed, 2005-09-21 at 18:50 +0200, Martin Husemann wrote:
> On Wed, Sep 21, 2005 at 12:10:09PM -0400, Jean-Luc Wasmer wrote:
> > Program received signal SIGABRT, Aborted.
> > [Switching to LWP 1]
> > 0xbd86071b in kill () from /usr/lib/libc.so.12
> > (gdb) backtrace
> > #0  0xbd86071b in kill () from /usr/lib/libc.so.12
> > #1  0xbd91c009 in pthread__errorfunc () from /usr/lib/libpthread.so.0
> > #2  0xbd919047 in pthread_mutex_unlock () from /usr/lib/libpthread.so.0
> > #3  0xbd965e6f in apr_thread_mutex_unlock () from /usr/pkg/lib/libapr-0=
.so.0
> > #4  0xbd964eb5 in apr_global_mutex_unlock () from /usr/pkg/lib/libapr-0=
.so.0
> > #5  0x0806c656 in util_ldap_cache_checkuserid ()
> > #6  0x0806f27a in mod_auth_ldap_check_user_id ()
> > #7  0x080e1e8c in ap_run_check_user_id ()
>=20
> Maybe you should forward this to the developers of this ldap module and t=
ell
> them that unlocking mutexes that have not been locked before (or have bee=
n
> unlocked already) is invoking undefined behaviour, according to POSIX.

Has this been reported upstream to the Apache developers? I've just
found this thread after getting bitten by this problem and I was
wondering if there was an upstream bug I could contribute to and/or
track. I couldn't see anything obvious in the Apache bugzilla.

Matt

--=-FX3hdJRzauHV7D5x423w
Content-Type: application/pgp-signature; name=signature.asc
Content-Description: This is a digitally signed message part

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.1 (GNU/Linux)

iD8DBQBDbRNlKP58eR+X2TMRAqMWAKC10wccDD2vuvBA6z2k2wHoW6ZEbgCfeedc
bvtem+o4FdORBvhfYPo6QUo=
=Cs8x
-----END PGP SIGNATURE-----

--=-FX3hdJRzauHV7D5x423w--