Hello, glibc fails to compile if the /bin/sh symlink is not set to to /bin/bash. Mine is set to /bin/ash. A posting to the forums suggested I file a bug on this so here we go. I ran the commands inside of script so I will attach a full log to this in a minute. However, here are the things that stand out to me: First we have a portage error with traceback: >>> md5 src_uri ;-) glibc-2.3.2.tar.bz2 >>> md5 src_uri ;-) glibc-linuxthreads-2.3.2.tar.bz2 >>> md5 src_uri ;-) glibc-2.3.3-branch-update-20040420.patch.bz2 Traceback (most recent call last): File "<string>", line 1, in ? File "/usr/lib/portage/pym/portage.py", line 2817, in vercmp cmp2=relparse(val2[x]) File "/usr/lib/portage/pym/portage.py", line 2534, in relparse number=string.atof(myver) File "/usr/lib/python2.3/string.py", line 205, in atof return _float(s) ValueError: invalid literal for float(): r1 /usr/sbin/ebuild.sh: line 144: [: -lt: unary operator expected Then it seemingly keeps going fine. It configures and compiles for a little bit and then gcc crashes with a missing terminating " character. Reproducible: Always Steps to Reproduce: 1.Set the /bin/sh symlink to /bin/ash 2.emerge glibc 3.ebuild fails as per attached log Actual Results: Please see attached glibc.txt.gz which is a full log. Expected Results: glibc should have emerged. If I go back and change the /bin/sh symlink back to /bin/bash it compiles fine. So I guess the ebuild just needs to know to use /bin/bash as the shell. Gentoo Base System version 1.4.16 Portage 2.0.50-r9 (default-x86-2004.2, gcc-3.3.4, glibc-2.3.3.20040420-r1, 2.6.8.1-ck1) ================================================================= System uname: 2.6.8.1-ck1 i686 Mobile Pentium II distcc 2.13 i686-pc-linux-gnu (protocols 1 and 2) (default port 3632) [enabled] Autoconf: sys-devel/autoconf-2.59-r4 Automake: sys-devel/automake-1.8.3 ACCEPT_KEYWORDS="x86" AUTOCLEAN="yes" CFLAGS="-march=pentium2 -O2 -pipe" CHOST="i686-pc-linux-gnu" COMPILER="gcc3" CONFIG_PROTECT="/etc /usr/X11R6/lib/X11/xkb /usr/kde/2/share/config /usr/kde/3/share/config /usr/share/config /var/qmail/control" CONFIG_PROTECT_MASK="/etc/gconf /etc/terminfo /etc/env.d" CXXFLAGS="-march=pentium2 -O2 -pipe" DISTDIR="/usr/portage/distfiles" FEATURES="autoaddcvs ccache distcc sandbox userpriv" GENTOO_MIRRORS="http://gentoo.mirrors.pair.com/" MAKEOPTS="-j2" PKGDIR="/usr/portage/packages" PORTAGE_TMPDIR="/var/tmp" PORTDIR="/usr/portage" PORTDIR_OVERLAY="/usr/local/portage" SYNC="rsync://192.168.42.25/gentoo-portage" USE="X acl adns aim alsa apm audiofile avi berkdb bonobo bzlib cdr crypt cups curl exif flac flash gdbm geoip gimpprint gtk gtk2 gtkhtml iconv imagemagick imap ipv6 jabber jpeg lcms libwww lzw-tiff mad mbox mime ming mmap mmx mng mpeg ncurses nptl oggvorbis oscar oss pam pcmcia pcre pdflib perl png pnp ppds python readline sdl snmp speex spell ssl svg tcpd theora tidy tiff truetype usb wmf x86 xine xml2 xv xvid zlib"
Created attachment 37557 [details] Gziped txt file containing log of glibc failure This is a gziped txt file of the log of glibc failing with /bin/sh set to /bin/ash. The portage traceback is in there and as you see, it doesn't get very far into the compiling.
Alot of things wont work if /bin/sh != a sunlink to bash. The advice you found on the forums bad suggestion. baselayout will also fail in misc ways. Solution is to leave the orig symlink alone.
that traceback you see has nothing to do with your shell, that's Bug 59704 but yes, if glibc itself fails because of /bin/sh not being bash, well that's a problem to be solved with the glibc peeps
