Date: Fri, 26 Mar 2004 14:23:56 -0800 From: Sean McNeil <sean@mcneil.com> To: "Jacques A. Vidrine" <nectar@freebsd.org> Cc: freebsd-current@freebsd.org Subject: libc using pthread routines (was nss_ldap broken) Message-ID: <1080339836.31925.17.camel@server.mcneil.com> In-Reply-To: <1080335779.11426.15.camel@server.mcneil.com> References: <1080273717.18231.10.camel@server.mcneil.com> <20040326125934.GA68357@madman.celabo.org> <1080334840.11426.12.camel@server.mcneil.com> <1080335779.11426.15.camel@server.mcneil.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On Fri, 2004-03-26 at 13:16, Sean McNeil wrote:
> On Fri, 2004-03-26 at 13:00, Sean McNeil wrote:
> > OK, I think I understand this problem...
> >
> > When I have my nsswitch.conf setup as follows, I get seg11s:
> >
> > passwd: files ldap
> > group: files ldap
> >
> > This appears to be an issue with any external nss_*.so.1 module that
> > uses pthread. It looks to me it is about the following:
> >
> > /*
> > * Cleanup
> > */
> > static void
> > nss_atexit(void)
> > {
> > (void)_pthread_rwlock_wrlock(&nss_lock);
> > VECTOR_FREE(_nsmap, &_nsmapsize, sizeof(*_nsmap),
> > (vector_free_elem)ns_dbt_free);
> > VECTOR_FREE(_nsmod, &_nsmodsize, sizeof(*_nsmod),
> > (vector_free_elem)ns_mod_free);
> > (void)_pthread_rwlock_unlock(&nss_lock);
> > }
> >
> > In my case, the nss_ldap.so.1 module was loaded which pulls in
> > libpthread. I'm not sure how this works without a libpthred, but it
> > would appear that unless libpthread.so is loaded everything is OK. But
> > now, it has been loaded and the rwlock_wrlock() works, but then it has
> > been unloaded before rwlock_unlock() gets called.
> >
> > Would using
> >
> > #include <reentrant.h>
> > rwlock_wrlock()
> > rwlock_unlock()
> >
> > macros fix this?
> >
> > Cheers,
> > Sean
> >
>
> Final thought: exit() sets __isthreaded to 0, so I am hazarding a guess
> that the lock isn't necessary. Removing the lock/unlock from
> nss_atexit() did the trick.
>
> Sean
>
I was looking at this wrong. __isthreaded is set to 0 as an initializer,
so I guess there is a need to lock and unlock in the atexit routine.
I am very puzzled about these routines. net/nsdispatch.c is the only
file in libc that uses the _pthread_ routines directly. Others use the
macros that test __isthreaded. The macros, however, will not provide
the status of the call back. The most peculiar thing about it all is I
cannot see how these are resolved if no thread library is loaded.
How does this mechanism work? I see that the routines in question are
marked as "W" by nm. So if a thread library doesn't get loaded in this
routine resolves to 0? And a call to 0 returns ok? Can someone help
out this confused soul?
Sean
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?1080339836.31925.17.camel>