This is the mail archive of the gcc@gcc.gnu.org mailing list for the GCC project.

Index Nav:	[Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav:	[Date Prev] [Date Next]	[Thread Prev] [Thread Next]
Other format:	[Raw text]

Re: [tree-ssa] Miscompilation (SIGSEGV) with -ffast-math (!?), hugecompile time regression

From: Richard Guenther <rguenth at tat dot physik dot uni-tuebingen dot de>
To: Richard Guenther <rguenth at tat dot physik dot uni-tuebingen dot de>
Cc: "gcc at gcc dot gnu dot org" <gcc at gcc dot gnu dot org>
Date: Mon, 23 Feb 2004 15:01:47 +0100
Subject: Re: [tree-ssa] Miscompilation (SIGSEGV) with -ffast-math (!?), hugecompile time regression
References: <403A0162.2070207@tat.physik.uni-tuebingen.de>

Richard Guenther wrote:

My favorite testcase is miscompiled by g++-ssa (GCC) 3.5-tree-ssa 20040220 (merged 20040211) causing it to segfault when using -O2 -ffast-math as opposed to -O2 which is fine (cmdline params "-n 10").
The testcase can be fetched from
http://www.tat.physik.uni-tuebingen.de/~rguenth/gcc/tramp3d-v2.cpp.gz
Trying to hunt it down with available compilers now, and check todays CVS.

g++ (GCC) 3.5-tree-ssa 20040217 (merged 20040211) endlessly loops, g++ (GCC) 3.5-tree-ssa 20040209 (merged 20040126) is fine. Note that between these two a huge compile time regression occoured, 20040209 takes 6m36s, 20040217 over 12m!

So maybe its the merge from mainline that brought all these problems? Though enough, I never checked mainline, but stayed with the 3.4 branch which is fine in both regards.

Trying to build mainline now.

Richard.

Follow-Ups:
- Re: [tree-ssa] Miscompilation (SIGSEGV) with -ffast-math (!?),huge compile time regression
  - From: Diego Novillo

References:
- [tree-ssa] Miscompilation (SIGSEGV) with -ffast-math (!?)
  - From: Richard Guenther

Index Nav:	[Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav:	[Date Prev] [Date Next]	[Thread Prev] [Thread Next]