http://gcc.gnu.org/bugzilla/show_bug.cgi?id=58528
--- Comment #2 from Charles <charles.frasch at gmail dot com> --- Hi, I think I got this right. Found a file in /tmp that had stuff like this: CMakeFiles/Fix_Engine_Detail_ut.dir/Message_Index_ut.cpp.o CMakeFiles/Fix_Engine_Detail_ut.dir/Message_Log_ut.cpp.o CMakeFiles/Fix_Engine_Detail_ut.dir/Sequential_Message_Log_ut.cpp.o ../../libfix_engine.a@0x3c558ba ../../libfix_engine.a@0x4037b36 ../../libfix_engine.a@0x42ded96 ../../libfix_engine.a@0x43f27d2 ../../libfix_engine.a@0x44a830e ../../libfix_engine.a@0x457237a ../../libfix_engine.a@0x14fd36a ../../libfix_engine.a@0x3ad7d36 ../../../middleware/libmiddleware.a@0x9f45c0 ../../../middleware/libmiddleware.a@0xff39dc ../../../middleware/libmiddleware.a@0x100c350 ../../../middleware/libmiddleware.a@0x102909c ../../../middleware/libmiddleware.a@0x98fd04 ../../../test/gtest/libtheme-gtest.a@0xd2d78 ../../../test/gtest/libtheme-gtest.a@0x142d8c ../../../test/libtheme-test.a@0x632 ../../../json/libjsoncpp.a@0x3816 ../../../libtheme.a@0x5ef90 ../../../libtheme.a@0xef438 ../../../libtheme.a@0x10637c ../../../libtheme.a@0x4743b0 ../../../libtheme.a@0x4a2000 ../../../libtheme.a@0x64806c ../../../libtheme.a@0x6ae8c8 ../../../libtheme.a@0x7d3a90 ../../../libtheme.a@0x92f87c ../../../libtheme.a@0x2c1dc Used that file as @ input to lto1 with all of the options from the failure invocation, thusly: /usr/lib/gcc/x86_64-linux-gnu/4.8.1/lto1 -o /dev/null @$1 -O2 -g -DNDEBUG -static-libstdc++ -flto -O2 This runs to completion without failure and no ICE. and produces what looks to me to be an assembly file (after changing /dev/null). If this approach is correct then I can't reproduce the error following the instructions at "Reducing LTO Bugs" On a whim I tried the command line with -save-temps and -v. It seems to be failing on the ltrans16 file: COLLECT_GCC_OPTIONS='-c' '-fexceptions' '-mtune=generic' '-march=x86-64' '-O2' '-g' '-D' 'NDEBUG' '-O2' '-rdynamic' '-v' '-save-temps' '-shared-libgcc' '-mtune=generic' '-march=x86-64' '-dumpdir' './' '-dumpbase' 'Fix_Engine_Detail_ut.ltrans16' '-fltrans' '-o' 'Fix_Engine_Detail_ut.ltrans16.ltrans.o' '-shared-libgcc' /usr/lib/gcc/x86_64-linux-gnu/4.8/lto1 -quiet -dumpdir ./ -dumpbase Fix_Engine_Detail_ut.ltrans16 -mtune=generic -march=x86-64 -mtune=generic -march=x86-64 -auxbase-strip Fix_Engine_Detail_ut.ltrans16.ltrans.o -g -O2 -O2 -version -fexceptions -fltrans @/tmp/ccibKIU0 -o Fix_Engine_Detail_ut.ltrans16.s causes the ICE. Would the Fix_Engine_Detail_ut.ltrans16.s or .o help. The .s file appears to truncated stopping right after the .Letext0: label. If not which other reduction approach to you think would be the most fruitful. Thanks, Charlie Frasch On Wed, Sep 25, 2013 at 8:35 AM, rguenth at gcc dot gnu.org < gcc-bugzi...@gcc.gnu.org> wrote: > http://gcc.gnu.org/bugzilla/show_bug.cgi?id=58528 > > --- Comment #1 from Richard Biener <rguenth at gcc dot gnu.org> --- > Please try to reduce the testcase, see > http://gcc.gnu.org/wiki/A_guide_to_testcase_reduction > > -- > You are receiving this mail because: > You reported the bug. >
