This is the mail archive of the
mailing list for the GCC project.
Re: Potential builtin memcpy bug in 4.9
- From: Andrew Haley <aph at redhat dot com>
- To: Zan Lynx <zlynx at acm dot org>, gcc-help at gcc dot gnu dot org
- Date: Wed, 11 Mar 2015 08:54:24 +0000
- Subject: Re: Potential builtin memcpy bug in 4.9
- Authentication-results: sourceware.org; auth=none
- References: <54FF5BB6 dot 70409 at acm dot org>
[redirect to gcc-help]
On 10/03/15 21:01, Zan Lynx wrote:
> Looking for some ideas. Maybe someone could audit the ASM code for the
> memcpy builtin, see if anything jumps out at you. I haven't tried that
> yet. Is it all one piece, or is it multiple chunks? Could it have bad
> ASM specifications which are allowing the optimizer to write into a
> register that should be preserve?
We can't really do anything without a way to reproduce the problem.
We don't even know the source which generated this. The naked
assembly code makes no sense if you can't see the source.
The fact that you I can't seem to build a stand-alone example is a
huge clue. It tells me that the real problem may be elsewhere. Did
you compile with -Wall? Any warnings?