During the emerge -e world that I performed after switching from amd64 to ~amd64 I discovered that glib-1.2.10-r5 fails to build with the following errors. /libtool: line 297: s,^.*/,,g: No such file or directory ./libtool: line 297: s,^.*/,,g: No such file or directory ./libtool: line 711: -e: command not found See attached build.log for more info. I have also noticed that this has been posted to the forums http://forums.gentoo.org/viewtopic-t-541928.html Of note is within the forum thread is this link http://www.arcknowledge.com/gmane.comp.gnu.libtool.general/2002-08/msg00104.html Which posts that the solution to this issue is... export SED=sed Im presently still emerge -e world (ing) so I have not been able to test this solution, but one person has reported success with this fix. I will test this as soon as possible. Reproducible: Always Steps to Reproduce: 1. Ensure dev-libs/glib-1.2.10-r5 is ~arch'd 2. emerge -auv glib 3.
Created attachment 111100 [details] build.log build.log showing build errors
Sorry for missing my emerge --info Portage 2.1.2-r10 (default-linux/amd64/2006.1, gcc-4.1.2, glibc-2.5-r0, 2.6.19-gentoo-r5 x86_64) ================================================================= System uname: 2.6.19-gentoo-r5 x86_64 AMD Athlon(tm) 64 Processor 3500+ Gentoo Base System release 1.12.9 Timestamp of tree: Thu, 22 Feb 2007 07:00:01 +0000 dev-java/java-config: 1.3.7, 2.0.31-r3 dev-lang/python: 2.4.4 dev-python/pycrypto: 2.0.1-r5 sys-apps/sandbox: 1.2.18.1 sys-devel/autoconf: 2.13, 2.61 sys-devel/automake: 1.4_p6, 1.5, 1.6.3, 1.7.9-r1, 1.8.5-r3, 1.9.6-r2, 1.10 sys-devel/binutils: 2.17 sys-devel/gcc-config: 1.3.14 sys-devel/libtool: 1.5.23b virtual/os-headers: 2.6.20 ACCEPT_KEYWORDS="amd64 ~amd64" AUTOCLEAN="yes" CBUILD="x86_64-pc-linux-gnu" CFLAGS="-march=k8 -pipe -O2" CHOST="x86_64-pc-linux-gnu" CONFIG_PROTECT="/etc /usr/kde/3.5/env /usr/kde/3.5/share/config /usr/kde/3.5/shutdown /usr/share/X11/xkb /usr/share/config /var/lib/hsqldb" CONFIG_PROTECT_MASK="/etc/env.d /etc/env.d/java/ /etc/gconf /etc/java-config/vms/ /etc/revdep-rebuild /etc/splash /etc/terminfo /etc/texmf/web2c" CXXFLAGS="-march=k8 -pipe -O2" DISTDIR="/usr/portage/distfiles" FEATURES="autoconfig debug distlocks metadata-transfer parallel-fetchi sandbox sfperms strict" GENTOO_MIRRORS="http://distfiles.gentoo.org http://distro.ibiblio.org/pub/linux/distributions/gentoo" LINGUAS="en en_GB" MAKEOPTS="-j2" PKGDIR="/usr/portage/packages" PORTAGE_RSYNC_OPTS="--recursive --links --safe-links --perms --times --compress --force --whole-file --delete --delete-after --stats --timeout=180 --exclude=/distfiles --exclude=/local --exclude=/packages" PORTAGE_TMPDIR="/var/tmp" PORTDIR="/usr/portage" PORTDIR_OVERLAY="/home/alistair/overlays/java-overlays/migrated-java-experimental-overlay /home/alistair/overlays/java-overlays/maven-overlay /usr/portage/local/layman/sunrise" SYNC="rsync://rsync.gentoo.org/gentoo-portage" USE="X aac admin alsa amd64 apache2 avi berkdb bitmap-fonts bzip2 c3p0 cdr cli cracklib crypt cups dbus dlloader doc dri dvd dvdr dvdread fortran gd gdbm glut gpm hal iconv ipv6 isdnlog jpeg jpeg2k kde libg++ live log4j midi mp3 mpeg mplayer mysql ncurses nls nptl nptlonly nsplugin nvidia offensive ogg openal opengl pam pcre pdf perl php png ppds pppd python qt readline reflection ruby session spell spl ssl tcpd threads tiff truetype truetype-fonts type1-fonts unicode userlocales vorbis xml xorg zlib" ALSA_CARDS="intel8x0" ALSA_PCM_PLUGINS="adpcm alaw asym copy dmix dshare dsnoop empty extplug file hooks iec958 ioplug ladspa lfloat linear meter mulaw multi null plug rate route share shm softvol" ELIBC="glibc" INPUT_DEVICES="keyboard mouse" KERNEL="linux" LCD_DEVICES="bayrad cfontz cfontz633 glk hd44780 lb216 lcdm001 mtxorb ncurses text" LINGUAS="en en_GB" USERLAND="GNU" VIDEO_CARDS="nvidia vesa" Unset: CTARGET, EMERGE_DEFAULT_OPTS, INSTALL_MASK, LANG, LC_ALL, LDFLAGS, PORTAGE_RSYNC_EXTRA_OPTS
Thanks. SED=sed emerge "<glib-2" worked.
I am now able to confirm that this fix works SED=sed emerge "<glib-2"
Works for me as well. Thank you all.
confirmed! worked for me, too, thanks =)
*** Bug 168266 has been marked as a duplicate of this bug. ***
exporting SED is a workaround, not a solution glib-1.x sucks and bundles an old libtool m4 file ... solution here would be to: rm -f acinclude.m4 before running eautoreconf
*** Bug 169566 has been marked as a duplicate of this bug. ***
*** Bug 169568 has been marked as a duplicate of this bug. ***
SED=sed emerge -v unixODBC Found this on my system too. I'll generate a list of current package versions shortly...
Created attachment 112242 [details] List of all currently installed packages Here are all of the installed packages at the time of the bug. I can try to reproduce it again after a full world upgrade completes.
*** Bug 169573 has been marked as a duplicate of this bug. ***
*** Bug 169663 has been marked as a duplicate of this bug. ***
This happens on more packages than just glib>2. It also happens on older versions of gtk as well. It looks like the same problem. I've noticed that if you change to the source directories and just remove the sh from the start of the ./libtool command, that everything compiles fine. So is this a bash bug since sh links to /bin/bash? I've noticed that: /bin/bash echo 1 does not do what I think it should do which is print out 1. It gives me an error: /usr/bin/echo: /usr/bin/echo: cannot execute binary file I've tried a few different commands and get the same result. Also, /bin/bash -c "echo 1" works fine and prints the expected "1". I have no idea why the SED environment variable fixes it. It still gives me the above error even if I export SED=sed Maybe the bugs are related, maybe not. (In reply to comment #0) > During the emerge -e world that I performed after switching from amd64 to > ~amd64 I discovered that glib-1.2.10-r5 fails to build with the following > errors. > > /libtool: line 297: s,^.*/,,g: No such file or directory > ./libtool: line 297: s,^.*/,,g: No such file or directory > ./libtool: line 711: -e: command not found >
Error is still reproducible with the test I outlined above. I believe that this has a root cause in one of the automatic build tools packages, though since most other things compile, probably the more outdated version support. Can anyone think of any tests I could run at this point?
Even when I cd to the build directory and type "make" I get the same error. angeliqe glib-1.2.10 # /bin/sh ./libtool --mode=compile i686-pc-linux-gnu-gcc -DHAVE_CONFIG_H -I. -I. -I. -DG_LOG_DOMAIN=g_log_domain_glib -DG_ENABLE_DEBUG -O3 -march=athlon-xp -pipe -fPIC -Wall -D_REENTRANT -c garray.c ./libtool: line 297: s,^.*/,,g: No such file or directory ./libtool: line 711: -e: command not found *** Warning: inferring the mode of operation is deprecated. *** Future versions of Libtool will require --mode=MODE be specified. ./libtool: line 1022: -e: command not found ./libtool: line 1022: -e: command not found ./libtool: line 1022: -e: command not found ./libtool: line 1022: -e: command not found ./libtool: line 1022: -e: command not found ./libtool: line 1022: -e: command not found ./libtool: line 1022: -e: command not found ./libtool: line 1022: -e: command not found ./libtool: line 1022: -e: command not found ./libtool: line 1022: -e: command not found ./libtool: line 1022: -e: command not found ./libtool: line 1022: -e: command not found ./libtool: line 1022: -e: command not found ./libtool: line 1022: -e: command not found ./libtool: line 1050: -e: command not found ./libtool: line 1074: -e: command not found : compile: cannot determine name of library object from `' You could update this package to use a current version of libtool as well as gtk+-1.2.10-r12 which has the same issue. Something like: ---------------------------------------------------------------------------- --- glib-1.2.10/aclocal.m4.orig 2007-03-07 14:52:20.000000000 -0500 +++ glib-1.2.10/aclocal.m4 2007-03-07 14:52:36.000000000 -0500 @@ -34,7 +34,7 @@ LIBTOOL_DEPS="$ac_aux_dir/ltconfig $ac_aux_dir/ltmain.sh" # Always use our own libtool. -LIBTOOL='$(SHELL) $(top_builddir)/libtool' +LIBTOOL='$(SHELL) /usr/bin/libtool' AC_SUBST(LIBTOOL)dnl # Redirect the config.log output again, so that the ltconfig log is not ------------------------------------------------------------------------------ (In reply to comment #16) > Error is still reproducible with the test I outlined above. > > I believe that this has a root cause in one of the automatic build tools > packages, though since most other things compile, probably the more outdated > version support. > > Can anyone think of any tests I could run at this point? >
*** Bug 170474 has been marked as a duplicate of this bug. ***
*** Bug 171387 has been marked as a duplicate of this bug. ***
mplayer is using a really old glib version
*** Bug 171893 has been marked as a duplicate of this bug. ***
Um, solution is in comment #8. We don't need further diagnosis. ;P
(In reply to comment #22) > Um, solution is in comment #8. We don't need further diagnosis. ;P > Half right about that. "rm -f acinclude.m4" works fine with glib, but w/ gtk+ it gets unhappy: * Running eautoreconf in '/var/tmp/portage/x11-libs/gtk+-1.2.10-r12/work/gtk+-1.2.10' ... * Running aclocal ... [ !! ] * Failed Running aclocal ! * * Include in your bugreport the contents of: * * /var/tmp/portage/x11-libs/gtk+-1.2.10-r12/temp/aclocal-867529.out !!! ERROR: x11-libs/gtk+-1.2.10-r12 failed. Call stack: ebuild.sh, line 1614: Called dyn_unpack ebuild.sh, line 751: Called qa_call 'src_unpack' environment, line 3595: Called src_unpack gtk+-1.2.10-r12.ebuild, line 43: Called eautoreconf autotools.eclass, line 93: Called eaclocal autotools.eclass, line 130: Called autotools_run_tool 'aclocal' autotools.eclass, line 218: Called die Here's the aclocal error log: aclocal: configure.in: 178: macro `AM_GTK_GNU_GETTEXT' not found in library I'm not sure what's up with this. The macro is in the aclocal.m4 file. However, if you strip all the archaic libtool stuff from acinclude.m4, then everything's happy. Probably easiest to just copy the shortened acinclude, since a patch would actually be longer than the new file. Or have sed delete the first 400 or so lines.
force use system libtool on aclocal.m4 and acinclude.m4. this is usable for glib and gtk+. --- glib-1.2.10/aclocal.m4.orig 2007-03-07 14:52:20.000000000 -0500 +++ glib-1.2.10/aclocal.m4 2007-03-07 14:52:36.000000000 -0500 @@ -34,5 +34,5 @@ LIBTOOL_DEPS="$ac_aux_dir/ltconfig $ac_aux_dir/ltmain.sh" # Always use our own libtool. -LIBTOOL='$(SHELL) $(top_builddir)/libtool' +LIBTOOL='$(SHELL) /usr/bin/libtool' AC_SUBST(LIBTOOL)dnl --- glib-1.2.10/acinclude.m4.orig 2007-03-07 14:52:20.000000000 -0500 +++ glib-1.2.10/acinclude.m4 2007-03-07 14:52:36.000000000 -0500 @@ -34,5 +34,5 @@ LIBTOOL_DEPS="$ac_aux_dir/ltconfig $ac_aux_dir/ltmain.sh" # Always use our own libtool. -LIBTOOL='$(SHELL) $(top_builddir)/libtool' +LIBTOOL='$(SHELL) /usr/bin/libtool' AC_SUBST(LIBTOOL)dnl
more better is maybe `which libtool`.
(In reply to comment #25) > more better is maybe `which libtool`. > It is probably better to be using `type -p libtool` I will refer you to the dev mailing list message http://archives.gentoo.org/gentoo-dev/msg_143424.xml as to why.
I currently have to rebuild gtk+ 1.2.10 and got bitten by this bug for the 2nd time. Is there a chance, that we'll see a fixed gtk+ 1.2.10 sometime?
I tried applying the patch from Comment #24, but it did not fix the problem. I tried export SED=sed then emerge "<glib-2", and it worked perfectly. I know the SED=sed is only a workaround, but it seems that the patch did not seem to help any. I am using libtool-1.5.23b. Haven't tried gtk+ yet, but performing an emerge -e world currently, so we'll see. BTW, I'm running ~x86 arch, with CFLAGS="-O2 -march=athlon-xp -pipe -fomit-frame-pointer".
(In reply to comment #28) > I tried applying the patch from Comment #24, but it did not fix the problem. I > tried export SED=sed then emerge "<glib-2", and it worked perfectly. I know > the SED=sed is only a workaround, but it seems that the patch did not seem to > help any. I am using libtool-1.5.23b. > > Haven't tried gtk+ yet, but performing an emerge -e world currently, so we'll > see. > > BTW, I'm running ~x86 arch, with CFLAGS="-O2 -march=athlon-xp -pipe > -fomit-frame-pointer". > same config, but the solution doesn't work for me :( Saurabh
*** Bug 176974 has been marked as a duplicate of this bug. ***
>>> Regenerating /etc/ld.so.cache... >>> Original instance of package unmerged safely. >>> dev-libs/glib-1.2.10-r5 merged. >>> No packages selected for removal by clean >>> Auto-cleaning packages... >>> No outdated packages were found on your system. * Regenerating GNU info directory index... * Processed 210 info files. * IMPORTANT: 2 config files in '/etc' need updating. * Type emerge --help config to learn how to update config files. # diff -u /usr/portage/dev-libs/glib/glib-1.2.10-r5.ebuild /usr/local/portage/dev-libs/glib/glib-1.2.10-r5.ebuild --- /usr/portage/dev-libs/glib/glib-1.2.10-r5.ebuild 2007-03-03 11:05:49.000000000 -0500 +++ /usr/local/portage/dev-libs/glib/glib-1.2.10-r5.ebuild 2007-05-03 20:46:01.000000000 -0400 @@ -35,6 +35,8 @@ use ppc64 && use hardened && replace-flags -O[2-3] -O1 append-ldflags $(dlopen_lib) + + rm -f acinclude.m4 eautoreconf elibtoolize }
Created attachment 118749 [details] gtk+-1 and glib-1 tar-bz2'd for easy overlaying. (In reply to comment #24 and #26) I've merged that patch to gtk+-1 and glib-1 and tar-bz2'd them together for easy overlaying. see attached file "gtk+_glib-force_system_libtool-ebuilds.tar.bz2"
Created attachment 118902 [details] octave emerge log
Comment on attachment 118902 [details] octave emerge log The same error seems to prevent octave to emerge. export SED=sed does not help to work around.
(In reply to comment #28 and #29) hmm... if you still has error with #32 overlay, how error was output?
*** Bug 180124 has been marked as a duplicate of this bug. ***
(In reply to comment #17) On solution for this are fix_libtool_files.sh 3.3.4 as root works for glib and gtk+ problems for libtool because if change you gcc, you get this problem.
(In reply to comment #38) > fix_libtool_files.sh 3.3.4 as root > > works for glib and gtk+ problems for libtool Not for me. glib-1.2.10-r5 still fails to compile. I don't know which version of glib was used to compile the version on the 2006.1 installation CD, which I assume is the version I have installed here since I've never been able to recompile it, but I've tried passing every version reported by eix to fix_libtool_files.sh, and still no joy.
- but the SED fix did work for me. Sorry, I meant to say that but forgot until just after clicking Submit.
*** Bug 182545 has been marked as a duplicate of this bug. ***
glib-1.x is fixed now
and now gtk+-1.x is fixed
