[ekato@ees.hokudai.ac.jp] libc/2271: sln core dumps: bugs of libc-start.c on powerpc?

[Etsushi Kato]

On Tue, May 22, 2001 at 04:22:24PM -0700,
Ulrich Drepper <drepper@redhat.com> wrote:

> > Looking at the generic version of libc-start, the appended patch looks
> > fine.
> 
> This patch shouldn't be necessary.  Why was it applied?  If the use of
> atexit breaks something it a problem with the tools.

So why you use __cxa_atexit on sysdeps/generic/libc-start.c?

According to CVS log...

---
CVSROOT:        /cvs/glibc
Module name:    libc
Changes by:     drepper@sources.redhat.com      2001-02-26 09:50:57

Modified files:
        sysdeps/generic: bb_init_func.c libc-start.c 
	
	Log message:
	        Use __cxa_atexit and not atexit.
---


Why only powerpc specific libc-start.c have to use atexit not
__cxa_atexit?

I'm not a developper and don't know how glibc and new gcc-3.0 are 
related.  But it just a question.

Regard,
-- 
Etsushi Kato
ekato@ees.hokudai.ac.jp