libanl + atexit + more

[Ulrich Drepper]

Andreas Jaeger <aj@suse.de> writes:

> I don't mind if we depend on gcc 3.0 in configure as soon as it's out
> - but we shouldn't enforce a pre release.  For now we have to support
> 2.95.x.  IMO it's ok if this test fails.

It's not only the test.  It's also this hopefully-soon-standardized
behavior of atexit().  You are simply not able to build a correct
glibc with 2.95.

-- 
---------------.                          ,-.   1325 Chesapeake Terrace
Ulrich Drepper  \    ,-------------------'   \  Sunnyvale, CA 94089 USA
Red Hat          `--' drepper at redhat.com   `------------------------