Bug 14433

Summary:	[tree-ssa] ICE, seg fault
Product:	gcc	Reporter:	Erik Schnetter <schnetter>
Component:	target	Assignee:	Not yet assigned to anyone <unassigned>
Status:	RESOLVED FIXED
Severity:	normal	CC:	gcc-bugs, reichelt
Priority:	P2	Keywords:	ice-on-valid-code
Version:	tree-ssa
Target Milestone:	tree-ssa
Host:	i686-pc-linux-gnu	Target:	i686-pc-linux-gnu
Build:	i686-pc-linux-gnu	Known to work:	3.4.0 4.0.0
Known to fail:		Last reconfirmed:	2004-03-05 10:36:07
Attachments:	failing source code

Description Erik Schnetter 2004-03-04 16:30:36 UTC

When compiling the attached C code with optimisation, I get an ICE: 
 
$ /home/eschnett/gcc-3.5-tree-ssa/bin/gcc -v 
Reading specs 
from /home/eschnett/gcc-3.5-tree-ssa/lib/gcc/i686-pc-linux-gnu/3.5-tree-ssa/specs 
Configured with: ../gcc/configure --prefix=/home/eschnett/gcc-3.5-tree-ssa 
--enable-languages=c,c++,f95 
Thread model: posix 
gcc version 3.5-tree-ssa 20040304 (merged 20040227) 
 
$ /home/eschnett/gcc-3.5-tree-ssa/bin/gcc -DCARPET_INT -DCARPET_REAL 
-DCARPET_COMPLEX -march=pentium3 -malign-double -m128bit-long-double -g3 
-gdwarf-2 -O3 -funroll-loops -Wall -Wpointer-arith -Wbad-function-cast 
-Wcast-qual -Wcast-align -Wstrict-prototypes -Wmissing-declarations 
simpleexcision.i -c 
In file included 
from /home/eschnett/Calpha/Cactus/configs/einstein-gfortran/build/BAM_Elliptic/simpleexcision.c:8: 
/home/eschnett/Calpha/Cactus/arrangements/AEIThorns/BAM_Elliptic/src/bbmg.h:44: 
warning: function declaration isn't a prototype 
/home/eschnett/Calpha/Cactus/configs/einstein-gfortran/build/BAM_Elliptic/simpleexcision.c: 
In function `iniboundmask': 
/home/eschnett/Calpha/Cactus/configs/einstein-gfortran/build/BAM_Elliptic/simpleexcision.c:63: 
internal compiler error: Segmentation fault

Comment 1 Erik Schnetter 2004-03-04 16:56:55 UTC

Created attachment 5859 [details]
failing source code

Comment 2 Volker Reichelt 2004-03-05 10:36:06 UTC

Reduced testcase that segfaults when compiled with "-O2 -march=i686"

=========================
int i;

void foo(int *p)
{
  i=0;
  if (*p) i=1;
}
=========================

Comment 3 Andrew Pinski 2004-03-05 15:19:54 UTC

This is almost definitely a RTL problem.

Comment 4 Volker Reichelt 2004-03-12 09:36:28 UTC

The problem disappeared today.