ICE (segfault) while building Tremolo with 11.12 with -O -fmodulo-sched

[Bernhard Rosenkraenzer]

[bero@matterhorn imx-4.0.3]$ /opt/android-toolchain-eabi/bin/arm-linux-androideabi-gcc -O -fmodulo-sched -o res012.o res012.i
external/tremolo/Tremolo/res012.c: In function 'res_inverse':
external/tremolo/Tremolo/res012.c:243:1: internal compiler error: Segmentation fault
Please submit a full bug report,
with preprocessed source if appropriate.
See <http://gcc.gnu.org/bugs.html> for instructions.

[Bernhard Rosenkraenzer (berolinux)]

Attaching reduced test case

[Michael Hope (michaelh1)]

Thank you for the bug report. I've confirmed this with gcc-linaro-4.6-2011.12 on ARM:

michaelh@ursa2:~/linaro/bugs$ /tools/toolchains/arch/armv7l/gcc-linaro-4.6-2011.12-armv7l-natty-cbuild212-tcpanda03-cortexa9r1/bin/gcc -O -fmodulo-sched -c res012.i
external/tremolo/Tremolo/res012.c: In function 'res_inverse':
external/tremolo/Tremolo/res012.c:243:1: internal compiler error: Segmentation fault

The (uninteresting) work-around is to not use -fmodulo-sched. The fault was introduced in the 4.6-2011.10 release. The fault also exists in gcc-4.7~svn182780 and was introduced somewhere between r178624 and r182152. Could you please also report this in GCC bugzilla and attach it to this ticket?

I've set it to low priority as it is a ftbfs, occurs at with module sched on, and the fault exists upstream.

[Michael Hope (michaelh1)]

Looking through the Linaro bzr log, r179381, r179380, and r178804 are candidates.

[Bernhard Rosenkraenzer (berolinux)]

Created attachment 26275
Reduced test case

Current gcc 4.7 and 4.6 branches result in a segfault while building the attached (reduced) test case for arm:

[bero@matterhorn bero]$ arm-linux-eabi-gcc -O -fmodulo-sched -o test.o -c test.c
test.c: In function 'res_inverse':
test.c:7:1: internal compiler error: Segmentation fault
Please submit a full bug report,
with preprocessed source if appropriate.
See <http://gcc.gnu.org/bugs.html> for instructions.

[Bernhard Rosenkraenzer (berolinux)]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51794

[Ibolton (ibolton)]

I could not reproduce with trunk (4.7.0) r183218 or r183840.

Please can you provide details of your cc1.

This is mine:

/work/ian/cross-build/trunk-r183218-thumb/arm-none-linux-gnueabi/tools/libexec/gcc/arm-none-linux-gnueabi/4.7.0/cc1 -quiet -v /home/ian/bugzilla/pr51794.c -quiet -dumpbase pr51794.c -mcpu=cortex-a9 -mfloat-abi=softfp -mfpu=vfpv3-d16 -marm -mtls-dialect=gnu -auxbase-strip test.o -O -version -fmodulo-sched -o /tmp/cc76gYW3.s

[Bernhard Rosenkraenzer (berolinux)]

I've updated my 4.7 build since and can't reproduce it there anymore.

Problem still occurs on 4.6 with

GNU C (Linaro GCC 4.6-2012.02-1~dev) version 4.6.3 20120201 (prerelease) [gcc-linaro/4.6 revision 106874] (arm-linux-androideabi)
        compiled by GNU C version 4.5.2, GMP version 5.0.4, MPFR version 3.1.0-p2, MPC version 0.9
GGC heuristics: --param ggc-min-expand=100 --param ggc-min-heapsize=131072

[vishal (vishalbhoj)]

Marking this bug as fixed since it is not reproducable on gcc 4.7

[Viktor (vchong)]

Not reproducable on gcc 4.7.