Gentoo Websites Logo
Go to: Gentoo Home Documentation Forums Lists Bugs Planet Store Wiki Get Gentoo!
Bug 223787 - sys-fs/e2fsprogs-1.40.8/9 compile/link failure, sys-libs/com_err-1.40.8/9 error, e.g. mc/e2fsck failing, system bootup failure, TLS data error
Summary: sys-fs/e2fsprogs-1.40.8/9 compile/link failure, sys-libs/com_err-1.40.8/9 err...
Status: RESOLVED TEST-REQUEST
Alias: None
Product: Gentoo Linux
Classification: Unclassified
Component: [OLD] Core system (show other bugs)
Hardware: x86 Linux
: High normal (vote)
Assignee: Gentoo Linux bug wranglers
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2008-05-27 05:30 UTC by Werner M. Mayer
Modified: 2008-05-31 04:50 UTC (History)
0 users

See Also:
Package list:
Runtime testing required: ---


Attachments

Note You need to log in before you can comment on or make changes to this bug.
Description Werner M. Mayer 2008-05-27 05:30:29 UTC
This bug report may be a duplicate of bug #222949. Please check.

A while ago I already did a emerge --sync;emerge portage;emerge -vun system
which ended in error at sys-fs/e2fsprogs-1.40.8 compile/link failure.

I ran into the problem, that my mc suddenly did no longer start but abort with
mc: error while loading shared libraries: libcom_err.so.2: cannot handle TLS data
and my system only booting to maintenance prompt.

After googling around and finding nothing suitable, not even here,
I fixed that by copying libcom_err.so.2.1 file from CD 2008beta2 to current
/lib. Not a fix, rather a workaround.

After waiting a few weeks, I repeated the whole procedure beginning with 
syncing, but again I had to use my workaround to able to keep using my mc.
Other programs failing on this annoying bug:
- System bootup with ext3fs partition failing to be checked even though
  partition is clean (cleaned by CD-2008beta) with TLS data error
- manual e2fsck failing with TLS data error

After application of my workaround, booting was fine but emerge process
ends here with failing package sys-fs/e2fsprogs-1.40.8

Most other packages had to be emerged manually to complete system upgrade,
with emerge of gcc and glibc still going during this reporting effort.

As a last chance, I applied ~x86 in package.keywords for emerging experimental
sys-libs/com_err-1.40.9 and its' dependencies. It didn't solve my problems.

Just after successful install of sys-libs/com_err-1.40.9 my mc and most 
probably my other programs began to fail again with the above mentioned 
TLS data error. I therefore reapplied my workaround again to get mc going
and most probably have a bootable system on next reboot.

My system is a Pentium I 133Mhz with 64mb RAM, Toshiba Portege 300ct.

This bug report was made via links with text mode editing on the 300ct.

Reproducible: Always

Steps to Reproduce:
1.emerge --sync
2.emerge -va portage
3.emerge -vun system
emerge #3 stops after compile of sys-fs/e2fsprogs-1.40.8 and 
sys-fs/e2fsprogs-1.40.9 leaving a broken libcom_err.so.2.1 installed behind, 
even when using ~x86 keyword
Actual Results:  
 See above report, try emerging sys-fs/e2fsprogs-1.40.8 
 or sys-fs/e2fsprogs-1.40.9 on x86 system

Expected Results:  
 See above report, after installation of dependancy sys-libs/com_err-1.40.8
 or sys-libs/com_err-1.40.9 critical programs fail and system reboot aborts
 to maintenance mode.

  compile link and install e2fsprogs.

 Portage 2.1.4.4 (default-linux/x86/2006.1, gcc-3.3.2, glibc-2.3.5-r0, 2.6.25 i5 86)
 =================================================================
 System uname: 2.6.25 i586 Pentium MMX
 Timestamp of tree: Sun, 25 May 2008 12:46:01 +0000
 ccache version 2.3 [enabled]
 app-shells/bash:     3.2_p33
 dev-lang/python:     2.4.4-r9
 dev-python/pycrypto: 2.0.1-r6
 dev-util/ccache:     2.4-r7
 sys-apps/baselayout: 1.12.11.1
 sys-apps/sandbox:    1.2.18.1-r2
 sys-devel/autoconf:  2.13, 2.61-r1
 sys-devel/automake:  1.4_p6, 1.5, 1.6.3, 1.7.9-r1, 1.8.5-r3, 1.9.6-r2, 1.10.1
 sys-devel/binutils:  2.18-r1
 sys-devel/gcc-config: 1.4.0-r4
 sys-devel/libtool:   1.5.26
 virtual/os-headers:  2.6.23-r3
 ACCEPT_KEYWORDS="x86"
 CBUILD="i586-pc-linux-gnu"
 CFLAGS="-Os -fomit-frame-pointer"
 CHOST="i586-pc-linux-gnu"
 CONFIG_PROTECT="/etc /etc/conf.d/gpm /etc/fstab /etc/init.d/gpm /etc/inittab /e tc/profile /etc/profile.local /usr/X11R6/lib/X11/xkb /usr/kde/3.4/env /usr/kde/ 3.4/share/config /usr/kde/3.4/shutdown /usr/share/config"
 CONFIG_PROTECT_MASK="/etc/env.d /etc/gconf /etc/rc.d /etc/revdep-rebuild /etc/t erminfo /etc/udev/rules.d /etc/wget"
 CXXFLAGS="-Os -fomit-frame-pointer"
 DISTDIR="/usr/portage/distfiles"
 FEATURES="autoaddcvs ccache cvs distlocks metadata-transfer sandbox sfperms str ict unmerge-orphans userfetch userpriv usersandbox"
 GENTOO_MIRRORS="http://ftp-stud.fht-esslingen.de/pub/Mirrors/gentoo/ ..."
 LANG="de_DE@euro"
 LINGUAS="de"
 MAKEOPTS="-j2"
 PKGDIR="/usr/portage/packages"
 PORTAGE_RSYNC_EXTRA_OPTS="--timeout=600"
 PORTAGE_RSYNC_OPTS="--recursive --links --safe-links --perms --times --compress  --force --whole-file --delete --stats --timeout=180 --exclude=/distfiles --exc lude=/local --exclude=/packages"
 PORTAGE_TMPDIR="/var/tmp"
 PORTDIR="/usr/portage"
 PORTDIR_OVERLAY="/usr/local/portage"
 SYNC="rsync://rsync.europe.gentoo.org/gentoo-portage"
 USE="7zip X X509 aac aalib ace acpi alsa apm audiofile bluetooth bzip2 cardbus
 ccache cdaudio cdb cdda cddb cdparanoia cdr cdrom cdsound chipcard clamav clamd  cli colordiff compress config-log config_wizard cpudetection cpufreq cpulimit
 cracklib crypt cups dbase dhcp dillo directfb dri dvd dvdr dvdread encode epson  exif fbcon fbdev fbsplash firefox fping ftp gdbm geldkarte gif gimp gimpprint
 glsa gphoto2 gpm gs gtk gzip h323 hacking hal hbci hddtemp highlight id3 ieee13 94 ipv6 irda isdn isdnlog jabber joystick jpeg jpeg2k kaspersky lame largefile
 lesstif lm_sensors logrotate logwatch lowmem lzo lzw mad madwifi mailbox mbox m idi mikmod mmx mng mod mozilla mp2 mp3 mp3rtp mp4 mp4live mpeg mpeg2 mplayer mp u401 mudflap nas ncurses network network-cron nfconntrack nfs nids nls nptl npt lonly ntfs numeric office ogg oggvorbis openssh openssl opensslcrypt oss pam pc mcia pcsc-lite pda pdf pentium perl php pipe png pnm postfix postscript ppds pp pd ps python rar raw readline recode reflection reiser4 reiserfs resolvconf rfk ill rplay rpm rtc scanner seamonkey sendmail sensord session sftp shout smartca rd sndfile snortsam sortrecords sox spl ssl svga sysfs syslog tcp tcpd tcpwrapp er tftp tga tgif thumbnail thumbnailing thunderbird tidy tiff tk truetype twola me type1 unicode unzip ups usb userlocales uudeview uuencode v4l v4l2 vcd verbo se vga vorbis vorbis-psy wav wavpack wifi win32codecs wireshark wma wmf wmp x26 4 x86 xanim xine xinerama xorg xorg-x11 xpm xv xvid yaz yv12 zip zip-external z lib" ALSA_CARDS="ali5451 als4000 atiixp atiixp-modem bt87x ca0106 cmipci emu10k 1 emu10k1x ens1370 ens1371 es1938 es1968 fm801 hda-intel intel8x0 intel8x0m mae stro3 trident usb-audio via82xx via82xx-modem ymfpci" ALSA_PCM_PLUGINS="adpcm a law asym copy dmix dshare dsnoop empty extplug file hooks iec958 ioplug ladspa
 lfloat linear meter mulaw multi null plug rate route share shm softvol" APACHE2
 _MODULES="actions alias auth_basic authn_alias authn_anon authn_dbm authn_defau lt authn_file authz_dbm authz_default authz_groupfile authz_host authz_owner au thz_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 m ime mime_magic negotiation rewrite setenvif speling status unique_id userdir us ertrack vhost_alias" ELIBC="glibc" INPUT_DEVICES="keyboard mouse evdev" KERNEL= "linux" LCD_DEVICES="bayrad cfontz cfontz633 glk hd44780 lb216 lcdm001 mtxorb n curses text" LINGUAS="de" USERLAND="GNU" VIDEO_CARDS="apm ark chips cirrus cyri x dummy fbdev glint i128 i740 i810 imstt mach64 mga neomagic nsc nv r128 radeon  rendition s3 s3virge savage siliconmotion sis sisusb tdfx tga trident tseng v4 l vesa vga via vmware voodoo"
 Unset:  CPPFLAGS, CTARGET, EMERGE_DEFAULT_OPTS, INSTALL_MASK, LC_ALL, LDFLAGS,
 PORTAGE_COMPRESS, PORTAGE_COMPRESS_FLAGS

It is hard for me to say how serious this bug is:
- if I were a new gentoo user, I'd say it's major to critical
- at my current gentoo knowledge level I can work around it, so it's a bug
  that should be fixed.
Comment 1 Jeroen Roovers (RETIRED) gentoo-dev 2008-05-27 05:42:24 UTC
Did you know that sys-fs/e2fsprogs, sys-libs/com_err and sys-libs/ss are basically different parts of one and the same package? You could try getting all three of the same version emerged. The point is to update com_err first, then ss and then e2fsprogs.

And surely enough, the com_err ebuild even reminds you what you should do after you have emerged com_err:


 * PLEASE PLEASE take note of this
 * Please make *sure* to run revdep-rebuild now.
 * Certain things on your system may have linked against a different
 * version of com_err -- those things need to be recompiled.
 * Sorry for the inconvenience

Please try that and reopen this bug only if you find the problem persists after you follow the instructions that should fix your system.
Comment 2 Jeroen Roovers (RETIRED) gentoo-dev 2008-05-27 05:43:59 UTC
Btw, looking at your emerge --info, you might also _really_ update your system by running emerge -vuaDN world. It will definitely be a tough ride, but in the end you may end up with a _current_ system. :)
Comment 3 Werner M. Mayer 2008-05-27 06:38:26 UTC
  WOW, what a fast response. I was just about to include the error outputs 
  of the log files when a collision was reported.

  Let me explain:
  When I run updates on this slow system which can take up to two weeks of 
  compilation, I do this mainly at night without attending my system, so
  I only see errors the next day.

  Also, when updating, I do this in two steps:
  First, I do an emerge -vun system and when that compiles cleanly I do an
  emerge -vuanDN world in mostly convenient small parts, as I also 
  sometimes need this notebook inbetween to do productional tasks.

My questions:  
  Is there a difference between 
     emerge -vuanDN world
   and emerge -vuaDN world
  other than the extra 'n' should keep emerge from reinstalling already
  installed packages?

  After an unattended emerge has been done, does there exist a way of
  quickly sifting through the log files to find such important notices
  of the kind you mentioned about com_err telling me to run revdep-rebuild?

  Or does there exist a file I don't know about that should notify me about 
  critical package installation messages?
  If not, is there a chance of that such a file could be generated
  in /var/log/portage during emerge process, i.e. like the file
  emerge.log?
  Do you have a script that does the job?

  At what point exactly in my 'emerge --info' do you see my system 
  lacks updating?

  It's true, I've tried to update my system many times in the past, 
  but errors in compiling and bugs have kept me from getting through
  the update phase completely.

  In the meantime, running revdep-rebuild...
  I will come back as soon as revdep is through or 
  if I run into new problems.
Comment 4 Werner M. Mayer 2008-05-31 03:29:03 UTC
WOW, fast response.   While writing on an additional error log I wanted to post I got a collision   notice. I will postbone the error logs until revdep-rebuild is complete and  a reemerge of system update has been launched.   Sorry this reply took longer to post, because compile times on P1 take longer  and having multiple bug report postings open on this "bug" I noticed I had  produced a lot of collision notices myself on my open and not yet sent postings.   I am proceeding as requested and will report on what I find.   To comment #1:  When I run updates on this slow system which can take up to two weeks or  longer of compilation, I do this mainly at night without attending my system,   so I only see errors the next day, if at all.   Also, when updating, I do this in two steps:  First, I do an emerge -vun system and when that compiles cleanly I do an  emerge -vuanDN world in mostly convenient small parts, as I also  sometimes need this notebook inbetween to do productional tasks.   My questions:  Is there a difference between     emerge -vuanDN world   and emerge -vuaDN world     other than the extra 'n',which should keep emerge from reinstalling      already installed packages?   After an unattended emerge has been done, does there exist a way of  quickly sifting through the log files to find such important notices  of the kind you mentioned about com_err telling me to run revdep-rebuild?   Or does there exist a file I don't know about that should notify me about  critical package installation messages?   If not, is there a chance of that such a file could be generated in  /var/log/portage during emerge process, i.e. like the file emerge.log?   Do you have a script that does the job?  To Comment #2: I would like to know where you can see my system has not successfully emerged  in the last few years due to revdep-rebuild errors I could not manage to  resolve (versions of installed packages?).  It's true, I've tried to update my system many times in the past, but errors in compiling and bugs have kept me from getting through the update phase completely.  Also I would like to know how and if unmerging one of two installed almost  identical packages the older versioned number can be removed, e.g. two  installed gcc's, two or more installed automake's, etc.?  I am running into packages with following error I can't resolve at this time, so the question is howto emerge packages or what to do with them: =media-libs/gdk-pixbuf-0.22.0-r3 (masked by: ) media-libs/gdk-pixbuf (masked by: )  This just an example, there are many more of them...  What I can't handle are packages masked by "missing keyword" and  "(missing by: )".  On one of my other systems (P III/500) I also have a revdep-rebuild running and emerge there fails with invalid atom... while trying to emerge packages with  of this sort for all packages: =app-arch/rpm:0 =app-pda/jpilot:0 =dev-perl/DBD-mysql:0 ...  Almost all versions of all packages have the same queer version number 0, some are :2, another :5. Could it be that I have a version of revdep-rebuild that has a bug, one or two are like net-analyzer/midas-nms:2.2f and kde-base/kdelibs:3.5 .  How can that be? Or should I file a new bug?  Meanwhile, I am trying to emerge the rest on my P III/500 and on my P I/133.  I'll be back, soon.
Comment 5 Werner M. Mayer 2008-05-31 04:50:40 UTC
Please ignore Comment #3.

It was never actually uploaded to bugzilla by me, 
but due to posting collision while actively editing on 
comment #3 on my LOCAL system it seems to have been
uploaded due to collision-side merging and messing up 
my well structured Comment #4.

Sorry for the mess in Comment #4.

Viewing my postings before submitting Comment #4 as 
#3 showed that after Comment #2 no other comments had 
been posted, not even by me.

After committing the well structured contents of 
Comment #4 as my originally planned Comment #3
and having quit links on the prior not committed 
Comments #3 I am somewhat astonished that even not 
having posted Comment #3 that the contents do show up
in this bug report. Seems to be a bug in bugzilla.

Is there any way of editing Comment #4?