On Fri, 22 Feb 2013, Dimitry Andric wrote: > On 2013-02-22 22:30, Daniel Eischen wrote: >> In trying to debug an unrelated problem, I switched CC from Clang >> back to GCC. I had a -current kernel and world r247050 built and >> installed with Clang as the system compiler I have nothing special >> in /etc/make.conf: >> >> BATCH=yes >> WITH_NEW_XORG=true >> WITH_KMS=true >> WITH_PKGNG=yes >> PERL_VERSION=5.14.2 >> >> I added WITHOUT_CLANG_IS_CC=yes to this, then rebuilt kernel and >> world. I installed the kernel rebooted, everthing worked, so I >> then installed world. Installword stopped here: >> >> ===> libexec/rtld-elf (install) >> chflags -h noschg /usr/libexec/ld-elf.so.1 >> install -s -o root -g wheel -m 555 -C -b -fschg -S ld-elf.so.1 >> /libexec >> install -o root -g wheel -m 444 rtld.1.gz /usr/share/man/man1 >> *** [_maninstall] Signal 11 >> >> Stop in /opt/FreeBSD/current/src/libexec/rtld-elf. >> *** [realinstall] Error code 1 >> >> At that point my system was completely hosed. Every binary (/bin, >> /sbin, etc) would sig 11. I had to build a world on another >> system, then use /rescue to NFS mount the other system and >> copy over /libexec, /lib, and /usr/lib. This let me recover >> enough to svn up to r247164. remove WITHOUT_CLANG_IS_CC from >> /etc/make.conf, and build/install a working world. >> >> Is switching from Clang to GCC suppose to work? > > This might have had nothing to do with either clang or gcc. Between > r247012 and r247117, binutils was broken, and this apparently resulted > in various nasty problems. > > Maybe you can try it again, since you are now at r247164? Ahh, okay, thanks to you (& to Steve). I guess my timing was just off. I'll try it again. -- DEReceived on Fri Feb 22 2013 - 22:55:36 UTC
This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:40:35 UTC