I installed a new system on a amd64 3200 departing from a stage1 tarball. When I did an "emerge system", the emerge process stopped during the compilation of e2fsprogs-1.34 with the following lines: "ln -s -f /var/tmp/portage/e2fsprogs-1.34/image//usr/lib/libcom_err.so.2 \ zapme/libcom_err.so /sbin/ldconfig make[1]: Leaving directory `/var/tmp/portage/e2fsprogs-1.34/work/e2fsprogs-1.34/lib/et' making install-shlibs in lib/ss make: *** [install-shlibs-libs-recursive] Error 1 !!! ERROR: sys-fs/e2fsprogs-1.34 failed. !!! Function einstall, Line 388, Exitcode 2 !!! einstall failed" After this, every binary gives a segmentation fault. Reproducible: Always Steps to Reproduce: 1. Do a basic gentoo install with 2004.1 livecd, using a stage1 tarball 2. emerge system 3. Actual Results: emerge system fails and everything segfaults Expected Results: completed the emerge system
Apperently the stage 1 tarball contains glibc-2.3.3, while the bootstrapping/emerge system process downgrades to glibc-2.3.2. Unmasking glibc-2.3.3-20040404 should solve the problem. Ckruse already confirmed this as a working solution (http://forums.gentoo.org/viewtopic.php?t=167086), I'm trying it out right now. I will post the results here later this evening or tomorrow morning. Sam
e2fsprogs compiles fine with glibc-2.3.3, but the next problem is then hdparm. See bug #40908 for this one.
Workaround for me, which has let "emerge system" run through smooth: Edit /usr/portage/sys-libs/glibc/glibc-2.3.3_pre20040420.ebuild => unmask ~amd64 Edit /usr/portage/sys-apps/hdparm/hdparm-5.4.ebuild => unmask ~amd64 (which is the only architecture masked). Greetings Chris
closing then ..
