libtool breaking emerges that depend on it [RESOLVED]

Message

nmp0906 · Post by **nmp0906** » Sat Feb 28, 2009 5:34 am

Emerge'ng anything results in the following line:

mkdir: cannot create directory `.libs': File exists

Which promptly calls 'die' and compiling aborts. Relevant info:

Code: Select all

mkdir .libs
mkdir: cannot create directory `.libs': File exists
mkdir .libs
mkdir: cannot create directory `.libs': File exists
make[4]: *** [proxy.lo] Error 1
make[4]: *** Waiting for unfinished jobs....
make[4]: *** [listen.lo] Error 1
mkdir .libs
mkdir: cannot create directory `.libs': File exists
make[4]: *** [dump.lo] Error 1
make[4]: Leaving directory `/var/tmp/portage/net-irc/irssi-0.8.12/work/irssi-0.8.12/src/irc/proxy'
make[3]: *** [all-recursive] Error 1
make[3]: Leaving directory `/var/tmp/portage/net-irc/irssi-0.8.12/work/irssi-0.8.12/src/irc'
make[2]: *** [all-recursive] Error 1
make[2]: Leaving directory `/var/tmp/portage/net-irc/irssi-0.8.12/work/irssi-0.8.12/src'
make[1]: *** [all-recursive] Error 1
make[1]: Leaving directory `/var/tmp/portage/net-irc/irssi-0.8.12/work/irssi-0.8.12'
make: *** [all] Error 2

Code: Select all

dev ~ # emerge --info
Portage 2.1.6.7 (hardened/linux/amd64/2008.0/no-multilib, gcc-4.3.3, glibc-2.9_p20081201-r1, 2.6.28-hardened x86_64)
=================================================================
System uname: Linux-2.6.28-hardened-x86_64-Intel-R-_Core-TM-2_Duo_CPU_T7500_@_2.20GHz-with-glibc2.2.5
Timestamp of tree: Sat, 28 Feb 2009 04:30:01 +0000
app-shells/bash:     3.2_p48-r1
dev-lang/python:     2.5.4-r2
sys-apps/baselayout: 2.0.0
sys-apps/openrc:     0.4.3-r1
sys-apps/sandbox:    1.3.8
sys-devel/autoconf:  2.63
sys-devel/automake:  1.10.2
sys-devel/binutils:  2.19.1
sys-devel/gcc-config: 1.4.1
sys-devel/libtool:   2.2.6a
virtual/os-headers:  2.6.28-r1
ACCEPT_KEYWORDS="amd64 ~amd64"
CBUILD="x86_64-pc-linux-gnu"
CFLAGS="-O2 -pipe -fforce-addr -march=nocona"
CHOST="x86_64-pc-linux-gnu"
CONFIG_PROTECT="/etc"
CONFIG_PROTECT_MASK="/etc/ca-certificates.conf /etc/env.d /etc/gconf /etc/gentoo-release /etc/php/apache2-php5/ext-active/ /etc/php/cgi-php5/ext-active/ /etc/php/cli-php5/ext-active/ /etc/revdep-rebuild /etc/sandbox.d /etc/terminfo /etc/udev/rules.d"
CXXFLAGS="-O2 -pipe -fforce-addr -march=nocona"
DISTDIR="/usr/portage/distfiles"
FEATURES="distlocks fixpackages parallel-fetch protect-owned sandbox sfperms strict unmerge-orphans userfetch"
GENTOO_MIRRORS="ftp://mirror.datapipe.net/gentoo http://mirror.datapipe.net/gentoo"
LDFLAGS="-Wl,-O1"
MAKEOPTS="-j3"
PKGDIR="/usr/portage/packages"
PORTAGE_RSYNC_OPTS="--recursive --links --safe-links --perms --times --compress --force --whole-file --delete --stats --timeout=180 --exclude=/distfiles --exclude=/local --exclude=/packages"
PORTAGE_TMPDIR="/var/tmp"
PORTDIR="/usr/portage"
SYNC="rsync://rsync.gentoo.org/gentoo-portage"
USE="acl amd64 apache2 bash-completion berkdb bzip2 cli cracklib crypt cups curl dri ftp gcj gd gdbm gpm hardened hash iconv imap isdnlog json justify midi mmx mudflap mysql mysqli ncurses nptl nptlonly openmp pam pcre perl php pic pppd python readline reflection session spl sse sse2 ssl sysfs tcpd unicode urandom xml xorg zlib" ALSA_CARDS="ali5451 als4000 atiixp atiixp-modem bt87x ca0106 cmipci emu10k1x ens1370 ens1371 es1938 es1968 fm801 hda-intel intel8x0 intel8x0m maestro3 trident usb-audio via82xx via82xx-modem ymfpci" ALSA_PCM_PLUGINS="adpcm alaw asym copy dmix dshare dsnoop empty extplug file hooks iec958 ioplug ladspa lfloat linear meter mmap_emul mulaw multi null plug rate route share shm softvol" APACHE2_MODULES="actions alias auth_basic authn_alias authn_anon authn_dbm authn_default authn_file authz_dbm authz_default authz_groupfile authz_host authz_owner authz_user autoindex cache dav dav_fs dav_lock deflate dir disk_cache env expires ext_filter file_cache filter headers include info log_config logio mem_cache mime mime_magic negotiation rewrite setenvif speling status unique_id userdir usertrack vhost_alias" ELIBC="glibc" INPUT_DEVICES="keyboard mouse evdev" KERNEL="linux" LCD_DEVICES="bayrad cfontz cfontz633 glk hd44780 lb216 lcdm001 mtxorb ncurses text" USERLAND="GNU" VIDEO_CARDS="fbdev glint i810 intel mach64 mga neomagic nv r128 radeon savage sis tdfx trident vesa vga via vmware voodoo"
Unset:  CPPFLAGS, CTARGET, EMERGE_DEFAULT_OPTS, FFLAGS, INSTALL_MASK, LANG, LC_ALL, LINGUAS, PORTAGE_COMPRESS, PORTAGE_COMPRESS_FLAGS, PORTAGE_RSYNC_EXTRA_OPTS, PORTDIR_OVERLAY

cwr · Post by **cwr** » Sat Feb 28, 2009 12:12 pm

Have you tried looking for the file ".libs"? The command:

find . -name "\.libs" -print

might produce something useful.

Will

nmp0906 · Post by **nmp0906** » Sat Feb 28, 2009 1:13 pm

That directory does not exist in /var/tmp/portage. Even looping a find during compile doesn't show one as existing at any point in time (granted, this isn't highly accurate). I tried clearing /var/tmp/portage thinking that would remove any work directories and allow the ebuild to start over from scratch, but like I said this happens on any emerge I do (which effectively means I can't emerge anything right now). All to no avail.

Post by Hu » Sat Feb 28, 2009 4:18 pm

That error message is not necessarily the problem. Any package that uses libtool may raise that message while compiling in parallel, because multiple processes all try to make the directory at once. The first one to call mkdir wins, and the others produce an error message. None of that should be fatal, though. Are you sure there are no other error messages higher up? Have you tried to emerge trivial packages that do not use libtool?

nmp0906 · Post by **nmp0906** » Sat Feb 28, 2009 5:20 pm

After further emerge tests, it would appear to be anything that uses libtool (and thus wants to create a .libs folder). So timezone-data will emerge, but xfsprogs will not. Emerging libtool itself produces the following error:

Code: Select all

chmod a+x libtoolize.tmp
chmod a-w libtoolize.tmp
mv -f libtoolize.tmp libtoolize
cd ./libltdl && \
        sed -e 's,config/mdate-sh,,' -e 's,config/texinfo.tex,,' \
            -e 's,config/mkinstalldirs,,' \
            < Makefile.in > Makefile.inT && \
        mv -f Makefile.inT Makefile.in
echo stamp > libltdl/stamp-mk
cd ./libltdl && /bin/sh /var/tmp/portage/sys-devel/libtool-2.2.6a/work/libtool-2.2.6/libltdl/config/missing --run autoheader
make[2]: *** [libltdl/lt__strl.lo] Error 1
make[2]: *** Waiting for unfinished jobs....
make[2]: *** [libltdl/loaders/dlopen.lo] Error 1
touch libltdl/config-h.in
make[2]: Leaving directory `/var/tmp/portage/sys-devel/libtool-2.2.6a/work/libtool-2.2.6'
make[1]: *** [all-recursive] Error 1
make[1]: Leaving directory `/var/tmp/portage/sys-devel/libtool-2.2.6a/work/libtool-2.2.6'
make: *** [all] Error 2
 *
 * ERROR: sys-devel/libtool-2.2.6a failed.
 * Call stack:
 *               ebuild.sh, line   49:  Called src_compile
 *             environment, line 2618:  Called _eapi0_src_compile
 *               ebuild.sh, line  609:  Called _eapi2_src_compile
 *               ebuild.sh, line  643:  Called die
 * The specific snippet of code:
 *              emake || die "emake failed"
 *  The die message:
 *   emake failed
 *

There are no other errors further up, the configure runs fine and then it starts compiling for a short bit, but then immediately dies after failing to create a .libs directory.

sera · Post by **sera** » Sun Mar 01, 2009 10:06 am

I found others with the same symptom as you. They all use hardened.

For me it seems as Hu suggested MAKEOPTS="-j1"

nmp0906 · Post by **nmp0906** » Sun Mar 01, 2009 1:11 pm

That didn't fix the issue.

Post by **zorry** » Sun Mar 01, 2009 2:05 pm

nmp0906 wrote:That didn't fix the issue.

I did get the same error when i updated it to sandbox 1.3.8.
If i use sandbox 1.3.7 it works for me.

sera · Post by **sera** » Sun Mar 01, 2009 2:10 pm

I upgraded to libtool-2.2.6a and nothing broke on my nonhardended gentoo.

Comparing the emerge --infos, then beside hardened all use libtool-2.2.6a and a kernel >=2.6.27

Do you have maybe an older kernel laying around?

nmp0906 · Post by **nmp0906** » Sun Mar 01, 2009 2:23 pm

dev ~ # uname -a
Linux dev 2.6.28-hardened #1 SMP Sun Jan 25 22:14:06 EST 2009 x86_64 Intel(R) Core(TM)2 Duo CPU T7500 @ 2.20GHz GenuineIntel GNU/Linux

I only have a 2.6.27 sitting around.

nmp0906 · Post by **nmp0906** » Sun Mar 01, 2009 2:39 pm

I can emerge a non-hardened kernel, and am in the process of compiling and installing that now.

sera · Post by **sera** » Sun Mar 01, 2009 2:47 pm

sandbox-1.3.8 isn't an issue as well on my box.

However five open bugs might be related to this. They all use sandbox-1.3.8. So would be nice if you could confirm whether downgrading sandbox fix the issue.

Post by **zorry** » Sun Mar 01, 2009 4:15 pm

It is a only hardened prob so far.
If you try to update some packages with sandbox 1.3.8 you get .lib.......
But if you update the packages with 1.3.7 and then when it is updatet you update sandbox to 1.3.8
it works well.
Some way some packages need to be update before sandbox 1.3.8.

nmp0906 · Post by **nmp0906** » Sun Mar 01, 2009 4:20 pm

So you're saying packages need to be compiled against sandbox-1.3.7 before it is updated to sandbox-1.3.8? I obviously went through 1.3.7, but there probably weren't many packages compiled against it before updating to 1.3.8.

sera · Post by **sera** » Sun Mar 01, 2009 4:22 pm

You do not compile against a sandbox. It just a tool. Downgrade to 1.3.7 and hopefully the problem is instantly gone.

So far found 9 open bugs witch seem to be the same problem. If you could confirm zorry's solution then is quite safe to assume they are the same and I will inform all of them about the workaround.

nmp0906 · Post by **nmp0906** » Sun Mar 01, 2009 4:28 pm

Yes, I was trying to understand the below quoted statement. And how this person got 1.3.8 to work. I am in the process of downgrading sandbox to 1.3.7 (and rebooting into my hardened kernel).

zorry wrote: But if you update the packages with 1.3.7 and then when it is updatet you update sandbox to 1.3.8
it works well.
Some way some packages need to be update before sandbox 1.3.8.

sera · Post by **sera** » Sun Mar 01, 2009 4:40 pm

Honestly, I do not understand it either.

Post by **zorry** » Sun Mar 01, 2009 4:48 pm

nmp0906 wrote:Yes, I was trying to understand the below quoted statement. And how this person got 1.3.8 to work. I am in the process of downgrading sandbox to 1.3.7 (and rebooting into my hardened kernel).

zorry wrote: But if you update the packages with 1.3.7 and then when it is updatet you update sandbox to 1.3.8
it works well.
Some way some packages need to be update before sandbox 1.3.8.

1 install stage3 hardened
2 mask sandbox 1.3.8
3 unmask needed packages for unstable hardened toolchain
4 update the toolchain
5 emerge -avDNu world
6 update sandbox to 1.3.8
7 emerge -e system

nmp0906 · Post by **nmp0906** » Sun Mar 01, 2009 4:53 pm

Well, downgrading to sandbox-1.3.7 seemed to fix the issue. I am in the process of updating world, and then I'll try emerge -e system, update to sandbox-1.3.8, and see if recompile still work.

nmp0906 · Post by **nmp0906** » Mon Mar 02, 2009 2:06 am

emerge -e system before re-upgrading to sandbox-1.3.8 did not fix the issue. so for now, i am sticking with 1.3.7.

sera · Post by **sera** » Mon Mar 02, 2009 9:22 am

nmp0906 wrote:emerge -e system before re-upgrading to sandbox-1.3.7 did not fix the issue. so for now, i am sticking with 1.3.8.

Well you mean the other way around, right?

Sorry for insisting but so far 8 of 11 possible bugs are linked to this and 3 of them are now closed. So zorry definitely found something here.

Post by **zorry** » Mon Mar 02, 2009 11:06 am

Tray this.
1. FEATURES=-sandbox emerge =sys-apps/sandbox-1.3.7
2. emerge coreutils
3. emerge sandbox 1.3.8
4. emerge mpfr
5 rebuild world

nmp0906 · Post by **nmp0906** » Mon Mar 02, 2009 11:49 am

sera wrote:Well you mean the other way around, right?

Yes, my apologies for that. I have corrected the post. I will try Zorry's above recommendation today as well.

nmp0906 · Post by **nmp0906** » Mon Mar 02, 2009 12:09 pm

Zorry, your suggestion did not work. Whenever sandbox-1.3.8 is installed, emerges depending on libtool fail.

Post by **zorry** » Mon Mar 02, 2009 12:27 pm

nmp0906 wrote:Zorry, your suggestion did not work. Whenever sandbox-1.3.8 is installed, emerges depending on libtool fail.

but emege world with sandbox 1.3.7 works?