Bruce Dubbs wrote: > I think we have a problem with gcc-4.5. Every kernel I build with it > segfaults. > > My last try was linux-2.6.32.8 which is the stable lfs-6.6 kernel. I > used the identical configuration. > > Either we are not building gcc-4.5 correctly or there is a bug in it > that causes the segfault. If I take a kernel built with gcc-4.4.3 and > everything else built with gcc-4.5.0, everything is fine. I'm on a > x86_64 system. > > Has anyone else tried building a kernel with gcc-4.5, e.g. lfs-svn?
I've narrowed it down to what appears to be a gcc-4.5 problem. I booted into an lfs-6.5 (gcc-4.4.1) system and built the 2.6.33.4 kernel and put that on /boot. I used the same .config file that I used earlier when using gcc-4.5.0. I rebooted into a partition with all the other packages built with gcc-4.5.0 and there were no problems. The only difference then between a kernel panic and a successful boot is building the kernel with gcc-4.5.0. A google search doesn't really show anything, so I suspect that we have something wrong with the gcc build. Rechecking my logs that were generated with gcc-4.5.0: MPC passes all 57 tests. GMP passes al tests. MPFR passes all 148 tests. GCC-4.5.0 has 10 unexpected failures in libmudflap, but that's pretty normal. FAIL: libmudflap.c/fail31-frag.c (-O3) output pattern test FAIL: libmudflap.c/pass45-frag.c (-O3) execution test FAIL: libmudflap.c/pass45-frag.c (-O3) output pattern test FAIL: libmudflap.c/pass45-frag.c (-O3) execution test FAIL: libmudflap.c/pass45-frag.c (-O3) output pattern test FAIL: libmudflap.c++/pass41-frag.cxx execution test FAIL: libmudflap.c++/pass41-frag.cxx (-static) execution test FAIL: libmudflap.c++/pass41-frag.cxx ( -O) execution test FAIL: libmudflap.c++/pass41-frag.cxx (-O2) execution test FAIL: libmudflap.c++/pass41-frag.cxx (-O3) execution test -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-dev FAQ: http://www.linuxfromscratch.org/faq/ Unsubscribe: See the above information page
