This is the mail archive of the
gcc-bugs@gcc.gnu.org
mailing list for the GCC project.
Re: stack alignment bug in gcc 2.95.2
- To: gcc-bugs at gcc dot gnu dot org
- Subject: Re: stack alignment bug in gcc 2.95.2
- From: David Ronis <ronis at ronispc dot chem dot mcgill dot ca>
- Date: Thu, 04 Nov 1999 17:42:03 -0500
- Organization: McGill University
- Reply-To: ronis at onsager dot chem dot mcgill dot ca
Jeffrey Law wrote:
>Alignment can only be preserved if the OS & dld arrange to
provide main
>with an aligned stack.
>
>If it's mis-aligned when main is called, there's little
we're going
>to be able to do about the problem.
Steven G. Johnson answered:
I've modified my test program slightly to check whether the
stack is
aligned in main() (see below). It is. Then, in the
subroutine "blah", it
is misaligned. So, it really seems like a bug in gcc's stack
alignment.
Can you suggest anything else to try?
Could someone post a few words on just where this misalignment is
creeping in. It clearly shouldn't be gcc's
startup files' job to fix a sloppy loader or worse OS (but it might be
the quickest way to fix the problem).
I'm running linux-2.2.13; should I be talking to the linux developers?
The ld-so folks? The glibc2 folks?
It may be my imagination, but lots of things seem to be running slower
since upgrading from libc5 to glibc-2.1.2, but I don't
want to draw any conclusions...
David