From revisions 163847 (163744 works) to 163859, I see the following ICEs with both -m32 and -m64 libgomp.graphite/force-parallel-3.c:5:6: internal compiler error: in refs_may_alias_p_1, at tree-ssa-alias.c:1031 libgomp.graphite/force-parallel-9.c:5:6: internal compiler error: in refs_may_alias_p_1, at tree-ssa-alias.c:1031 with -m64 -Os gfortran.dg/backspace_1.f:82:0: internal compiler error: in refs_may_alias_p_1, at tree-ssa-alias.c:1031 gfortran.dg/record_marker_2.f:83:0: internal compiler error: in refs_may_alias_p_1, at tree-ssa-alias.c:1031 libgomp.fortran/appendix-a/a.16.1.f90:26:0: internal compiler error: in refs_may_alias_p_1, at tree-ssa-alias.c:1031 libgomp.fortran/omp_atomic2.f90:35:0: internal compiler error: in refs_may_alias_p_1, at tree-ssa-alias.c:1031 and with -m64 -O2 -fgraphite-identity gfortran.dg/graphite/pr42393-1.f90:6:0: internal compiler error: in refs_may_alias_p_1, at tree-ssa-alias.c:1031
I can't see any of those on x86_64-linux, neither with -m32 nor with -m64.
> I can't see any of those on x86_64-linux, neither with -m32 nor with -m64. See also http://gcc.gnu.org/ml/gcc-testresults/2010-09/msg00410.html .
I also get the same ICE on powerpc-apple-darwin9 at revision 163836: [karma] f90/bug% /opt/gcc/gcc4.6w/bin/gcc -O3 /opt/gcc/_gcc_clean/gcc/testsuite/gcc.dg/vect/vect-multitypes-1.c /opt/gcc/_gcc_clean/gcc/testsuite/gcc.dg/vect/vect-multitypes-1.c: In function 'main1': /opt/gcc/_gcc_clean/gcc/testsuite/gcc.dg/vect/vect-multitypes-1.c:41:1: internal compiler error: in refs_may_alias_p_1, at tree-ssa-alias.c:1031 and gcc/testsuite/gcc.dg/vect/vect-multitypes-4.c (see also r163798 passes http://gcc.gnu.org/ml/gcc-testresults/2010-09/msg00273.html and r163836 fails http://gcc.gnu.org/ml/gcc-testresults/2010-09/msg00313.html ).
That doesn't help me to reproduce the issue. Please provide output of appending -v to the compiler command-line (this must be reproducable on x86_64 with appropriate flags).
> Please provide output of appending -v to the compiler command-line The configure options [macbook] gcc/p_build% gfcp -v Using built-in specs. COLLECT_GCC=gfcp COLLECT_LTO_WRAPPER=/opt/gcc/gcc4.6p/libexec/gcc/x86_64-apple-darwin10/4.6.0/lto-wrapper Target: x86_64-apple-darwin10 Configured with: ../p_work/configure --prefix=/opt/gcc/gcc4.6p --mandir=/opt/gcc/gcc4.6p/share/man --infodir=/opt/gcc/gcc4.6p/share/info --build=x86_64-apple-darwin10 --host=x86_64-apple-darwin10 --target=x86_64-apple-darwin10 --enable-languages=c,lto,fortran --with-gmp=/opt/sw64 --with-libiconv-prefix=/opt/sw64 --with-system-zlib --x-includes=/usr/X11R6/include --x-libraries=/usr/X11R6/lib --with-cloog=/opt/sw64 --with-ppl=/opt/sw64 --with-mpc=/opt/sw64 --enable-checking=release --with-quad=/opt/sw64 --disable-decimal-float Thread model: posix gcc version 4.6.0 20100903 (experimental) [trunk revision 163800p1] (GCC) works, building revision 163820. The other configure options can be found in the links in comment #2 and #3.
(In reply to comment #5) > > Please provide output of appending -v to the compiler command-line > > The configure options > [macbook] gcc/p_build% gfcp -v > Using built-in specs. > COLLECT_GCC=gfcp > COLLECT_LTO_WRAPPER=/opt/gcc/gcc4.6p/libexec/gcc/x86_64-apple-darwin10/4.6.0/lto-wrapper > Target: x86_64-apple-darwin10 > Configured with: ../p_work/configure --prefix=/opt/gcc/gcc4.6p > --mandir=/opt/gcc/gcc4.6p/share/man --infodir=/opt/gcc/gcc4.6p/share/info > --build=x86_64-apple-darwin10 --host=x86_64-apple-darwin10 > --target=x86_64-apple-darwin10 --enable-languages=c,lto,fortran > --with-gmp=/opt/sw64 --with-libiconv-prefix=/opt/sw64 --with-system-zlib > --x-includes=/usr/X11R6/include --x-libraries=/usr/X11R6/lib > --with-cloog=/opt/sw64 --with-ppl=/opt/sw64 --with-mpc=/opt/sw64 > --enable-checking=release --with-quad=/opt/sw64 --disable-decimal-float > Thread model: posix > gcc version 4.6.0 20100903 (experimental) [trunk revision 163800p1] (GCC) > > works, building revision 163820. > > The other configure options can be found in the links in comment #2 and #3. -v for one failing testcase - I want to see what standard -march/tune your config uses.
(In reply to comment #6) > (In reply to comment #5) > -v for one failing testcase - I want to see what standard -march/tune > your config uses. we default to generic (FWIW, the same fail occurs also for me on i686-darwin9 with -mtune=core2). $ /GCC/gcc-4-6-reghunt-build/gcc/xgcc -B/GCC/gcc-4-6-reghunt-build/gcc/ /GCC/gcc-4-6-reghunt/libgomp/testsuite/libgomp.fortran/appendix-a/a.16.1.f90 -B/GCC/gcc-4-6-reghunt-build/i686-apple-darwin9/x86_64/libgomp/ -B/GCC/gcc-4-6-reghunt-build/i686-apple-darwin9/x86_64/libgomp/.libs -I/GCC/gcc-4-6-reghunt-build/i686-apple-darwin9/x86_64/libgomp -I/GCC/gcc-4-6-reghunt/libgomp/testsuite/.. -shared-libgcc -fmessage-length=0 -fopenmp -Os -B/GCC/gcc-4-6-reghunt-build/i686-apple-darwin9/x86_64/libgomp/../libgfortran/.libs -L/GCC/gcc-4-6-reghunt-build/i686-apple-darwin9/x86_64/libgomp/.libs -lgomp -L/GCC/gcc-4-6-reghunt-build/i686-apple-darwin9/x86_64/libgomp/../libgfortran/.libs -lgfortran -lm -m64 -o ./a.16.1.exe -v Reading specs from /GCC/gcc-4-6-reghunt-build/gcc/specs COLLECT_GCC=/GCC/gcc-4-6-reghunt-build/gcc/xgcc COLLECT_LTO_WRAPPER=/GCC/gcc-4-6-reghunt-build/gcc/lto-wrapper Target: i686-apple-darwin9 Configured with: ../gcc-4-6-reghunt/configure --prefix=/GCC/tobjc --target=i686-apple-darwin9 --host=i686-apple-darwin9 --build=i686-apple-darwin9 --enable-version-specific-runtime-libs --enable-threads --enable-checking=yes --program-suffix=-4.6rh --with-libiconv-prefix=/usr --with-system-zlib --enable-languages=c,fortran CC=gcc-4.2 CXX=g++-4.2 Thread model: posix gcc version 4.6.0 20100905 (experimental) [trunk revision 163877] (GCC) COLLECT_GCC_OPTIONS='-mmacosx-version-min=10.5.8' '-B' '/GCC/gcc-4-6-reghunt-build/gcc/' '-B' '/GCC/gcc-4-6-reghunt-build/i686-apple-darwin9/x86_64/libgomp/' '-B' '/GCC/gcc-4-6-reghunt-build/i686-apple-darwin9/x86_64/libgomp/.libs' '-I' '/GCC/gcc-4-6-reghunt-build/i686-apple-darwin9/x86_64/libgomp' '-I' '/GCC/gcc-4-6-reghunt/libgomp/testsuite/..' '-shared-libgcc' '-fmessage-length=0' '-fopenmp' '-Os' '-B' '/GCC/gcc-4-6-reghunt-build/i686-apple-darwin9/x86_64/libgomp/../libgfortran/.libs' '-L/GCC/gcc-4-6-reghunt-build/i686-apple-darwin9/x86_64/libgomp/.libs' '-L/GCC/gcc-4-6-reghunt-build/i686-apple-darwin9/x86_64/libgomp/../libgfortran/.libs' '-m64' '-o' './a.16.1.exe' '-v' '-mtune=generic' /GCC/gcc-4-6-reghunt-build/gcc/f951 /GCC/gcc-4-6-reghunt/libgomp/testsuite/libgomp.fortran/appendix-a/a.16.1.f90 -I /GCC/gcc-4-6-reghunt-build/i686-apple-darwin9/x86_64/libgomp -I /GCC/gcc-4-6-reghunt/libgomp/testsuite/.. -fPIC -quiet -dumpbase a.16.1.f90 -mmacosx-version-min=10.5.8 -m64 -mtune=generic -auxbase a.16.1 -Os -version -fmessage-length=0 -fopenmp -fintrinsic-modules-path finclude -o /var/folders/OW/OW-PGOtgHbKakssxFpJpkU++-0E/-Tmp-//cc74lrzb.s GNU Fortran (GCC) version 4.6.0 20100905 (experimental) [trunk revision 163877] (i686-apple-darwin9) compiled by GNU C version 4.6.0 20100905 (experimental) [trunk revision 163877], GMP version 4.3.2, MPFR version 2.4.2, MPC version 0.8.1 GGC heuristics: --param ggc-min-expand=30 --param ggc-min-heapsize=4096 GNU Fortran (GCC) version 4.6.0 20100905 (experimental) [trunk revision 163877] (i686-apple-darwin9) compiled by GNU C version 4.6.0 20100905 (experimental) [trunk revision 163877], GMP version 4.3.2, MPFR version 2.4.2, MPC version 0.8.1 GGC heuristics: --param ggc-min-expand=30 --param ggc-min-heapsize=4096 /GCC/gcc-4-6-reghunt/libgomp/testsuite/libgomp.fortran/appendix-a/a.16.1.f90: In function ‘a16’: /GCC/gcc-4-6-reghunt/libgomp/testsuite/libgomp.fortran/appendix-a/a.16.1.f90:26:0: internal compiler error: in refs_may_alias_p_1, at tree-ssa-alias.c:1040 Please submit a full bug report, with preprocessed source if appropriate. See <http://gcc.gnu.org/bugs.html> for instructions.
The ICEs appeared between 163800 (working) and 163820 (ICE).
The ICEs are due to revision 163802.
Apparently this pr does not show up for i686-apple-darwin9 (see http://gcc.gnu.org/ml/gcc-testresults/2010-09/msg00452.html ).
The ICEs in comment #3 also show up for powerpc64-unknown-linux-gnu (see http://gcc.gnu.org/ml/gcc-testresults/2010-09/msg00417.html ).
(In reply to comment #10) > Apparently this pr does not show up for i686-apple-darwin9 (see > http://gcc.gnu.org/ml/gcc-testresults/2010-09/msg00452.html ). it does at m64 see c#7
Ok, I reproduced a failure (but none of the kind reported here) and have a fix for that issue. Maybe it fixes this bug too, which I still can't reproduce.
Subject: Bug 45534 Author: rguenth Date: Mon Sep 6 12:14:02 2010 New Revision: 163913 URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=163913 Log: 2010-09-06 Richard Guenther <rguenther@suse.de> PR tree-optimization/45534 * tree-ssa-address.c (create_mem_ref_raw): Add verify parameter. (create_mem_ref): Do verify the created TARGET_MEM_REF is valid on the target. (maybe_fold_tmr): Do not verify the created TARGET_MEM_REF is valid on the target. Modified: trunk/gcc/ChangeLog trunk/gcc/tree-ssa-address.c
(In reply to comment #14) > Subject: Bug 45534 > New Revision: 163913 > 2010-09-06 Richard Guenther <rguenther@suse.de> > > PR tree-optimization/45534 > * tree-ssa-address.c (create_mem_ref_raw): Add verify parameter. > (create_mem_ref): Do verify the created TARGET_MEM_REF is valid > on the target. > (maybe_fold_tmr): Do not verify the created TARGET_MEM_REF is > valid on the target. fixed on i686-darwin9.
> > New Revision: 163913 > fixed on i686-darwin9. also on x86_64-apple-darwin10.4 configured with --enable-checking=release.
Also fixed on x86_64-apple-darwin10.4, trunk configured with the default value for --enable-checking, see http://gcc.gnu.org/ml/gcc-testresults/2010-09/msg00616.html The failures on ppc reported in comment #3 have also been fixed, see http://gcc.gnu.org/ml/gcc-testresults/2010-09/msg00613.html . Closing as fixed.