Suddenly, after updating some system packages, firefox and thunderbird stopped working (got a SEGFAULT right on startup). I really tried hard to find find the issue and I came to this: when libgnome is installed, they fail, if I unmerge it (and remove "/usr/lib64/libgnome-2.so.0*"), they both work. Maybe the root cause is somewhere else but this is as far as I got. Another user seemed to have the same problem: http://forums.gentoo.org/viewtopic-p-6983838.html backtrace and emerge --info will be appended Reproducible: Always
Created attachment 305871 [details] emerge --info
Created attachment 305873 [details] gdb backtrace
You need to recompile packages providing listed libs in: #0 0x00007ffff0897600 in g_utf8_validate () from /usr/lib64/libglib-2.0.so.0 #1 0x00007fffe6c07162 in gconf_value_from_corba_value () from /usr/lib64/libgconf-2.so.4 #2 0x00007fffe6c1468e in gconf_engine_all_entries () from /usr/lib64/libgconf-2.so.4 #3 0x00007fffe6c17de4 in ?? () from /usr/lib64/libgconf-2.so.4 #4 0x00007fffe6c1aebb in gconf_client_add_dir () from /usr/lib64/libgconf-2.so.4 #5 0x00007fffe79d66fc in ?? () from /usr/lib64/libgnomeui-2.so.0 #6 0x00007fffe72c74ce in gnome_program_postinit () from /usr/lib64/libgnome-2.so.0 #7 0x00007fffe72c78f9 in ?? () from /usr/lib64/libgnome-2.so.0 #8 0x00007fffe72c7b6d in gnome_program_initv () from /usr/lib64/libgnome-2.so.0 #9 0x00007fffe72c7c6a in gnome_program_init () from /usr/lib64/libgnome-2.so.0 #10 0x00007ffff51cbccb in ?? () from /usr/lib64/firefox/libxul.so #11 0x00007ffff51c4f98 in XRE_main () from /usr/lib64/firefox/libxul.so to get an useful backtrace: http://www.gentoo.org/proj/en/qa/backtraces.xml
Created attachment 305953 [details] gdb backtrace with debugging symbols Firefox not built with debugging symbols due to time constraints. I saw gnucash trigger this, too but now it won't start at all since it won't start w/o libgnome (in contrast to firefox)
Looks something related with /desktop/gnome/sound, maybe because of esd support removed? :-/
Are you able to reproduce on a new created user account with a new home?
Created attachment 306397 [details] /etc/gconf/schemas/desktop_gnome_sound.schemas Yes, this is reproducable: test@miramis ~ $ ls -lha; firefox; ls -lha total 32K drwxr-xr-x 4 test test 4.0K Mar 23 10:57 . drwxr-xr-x 6 root root 4.0K Mar 23 10:46 .. -rw------- 1 test test 16 Mar 23 10:47 .bash_history -rw-r--r-- 1 test test 127 Mar 19 01:59 .bash_logout -rw-r--r-- 1 test test 193 Mar 19 01:59 .bash_profile -rw-r--r-- 1 test test 551 Mar 19 01:59 .bashrc drwx------ 2 test test 4.0K Mar 23 10:57 .gconfd drwx------ 2 test test 4.0K Jul 3 2007 .ssh Segmentation fault total 44K drwxr-xr-x 7 test test 4.0K Mar 23 10:57 . drwxr-xr-x 6 root root 4.0K Mar 23 10:46 .. -rw------- 1 test test 16 Mar 23 10:47 .bash_history -rw-r--r-- 1 test test 127 Mar 19 01:59 .bash_logout -rw-r--r-- 1 test test 193 Mar 19 01:59 .bash_profile -rw-r--r-- 1 test test 551 Mar 19 01:59 .bashrc drwx------ 2 test test 4.0K Mar 23 10:57 .gconf drwx------ 2 test test 4.0K Mar 23 10:57 .gconfd drwx------ 3 test test 4.0K Mar 23 10:57 .gnome2 drwx------ 2 test test 4.0K Mar 23 10:57 .gnome2_private drwx------ 2 test test 4.0K Jul 3 2007 .ssh Since you mentioned something about desktop/gnome/sound, I attache the file "/etc/gconf/schemas/desktop_gnome_sound.schemas" here. Contains strings with german umlauts but they look like valid UTF-8 to me
(In reply to comment #7) > ... > -rw-r--r-- 1 test test 551 Mar 19 01:59 .bashrc > drwx------ 2 test test 4.0K Mar 23 10:57 .gconfd > drwx------ 2 test test 4.0K Jul 3 2007 .ssh > ... That .gconfd was a leftover from a previous try but it fails the same way when deleting that folder: test@miramis ~ $ ls -lha; firefox; ls -lha total 28K drwxr-xr-x 3 test test 4.0K Mar 23 11:01 . drwxr-xr-x 6 root root 4.0K Mar 23 10:46 .. -rw------- 1 test test 16 Mar 23 10:47 .bash_history -rw-r--r-- 1 test test 127 Mar 19 01:59 .bash_logout -rw-r--r-- 1 test test 193 Mar 19 01:59 .bash_profile -rw-r--r-- 1 test test 551 Mar 19 01:59 .bashrc drwx------ 2 test test 4.0K Jul 3 2007 .ssh Segmentation fault total 44K drwxr-xr-x 7 test test 4.0K Mar 23 11:02 . drwxr-xr-x 6 root root 4.0K Mar 23 10:46 .. -rw------- 1 test test 16 Mar 23 10:47 .bash_history -rw-r--r-- 1 test test 127 Mar 19 01:59 .bash_logout -rw-r--r-- 1 test test 193 Mar 19 01:59 .bash_profile -rw-r--r-- 1 test test 551 Mar 19 01:59 .bashrc drwx------ 2 test test 4.0K Mar 23 11:02 .gconf drwx------ 2 test test 4.0K Mar 23 11:02 .gconfd drwx------ 3 test test 4.0K Mar 23 11:02 .gnome2 drwx------ 2 test test 4.0K Mar 23 11:02 .gnome2_private drwx------ 2 test test 4.0K Jul 3 2007 .ssh test@miramis ~ $ Now the same after "rm /usr/lib64/libgnome-2.so*" as root: test@miramis ~ $ ls -lha; firefox; ls -lha total 28K drwxr-xr-x 3 test test 4.0K Mar 23 11:03 . drwxr-xr-x 6 root root 4.0K Mar 23 10:46 .. -rw------- 1 test test 16 Mar 23 10:47 .bash_history -rw-r--r-- 1 test test 127 Mar 19 01:59 .bash_logout -rw-r--r-- 1 test test 193 Mar 19 01:59 .bash_profile -rw-r--r-- 1 test test 551 Mar 19 01:59 .bashrc drwx------ 2 test test 4.0K Jul 3 2007 .ssh Terminated total 36K drwxr-xr-x 5 test test 4.0K Mar 23 11:03 . drwxr-xr-x 6 root root 4.0K Mar 23 10:46 .. -rw------- 1 test test 16 Mar 23 10:47 .bash_history -rw-r--r-- 1 test test 127 Mar 19 01:59 .bash_logout -rw-r--r-- 1 test test 193 Mar 19 01:59 .bash_profile -rw-r--r-- 1 test test 551 Mar 19 01:59 .bashrc drwx------ 3 test test 4.0K Mar 23 11:03 .cache drwx------ 4 test test 4.0K Mar 23 11:03 .mozilla drwx------ 2 test test 4.0K Jul 3 2007 .ssh test@miramis ~ $
Created attachment 306399 [details] gdb backtrace of gnucash I tried gnucash again (to check that this is not firefox/thunderbird specific): test@miramis ~ $ ls -lha; gnucash total 28K drwxr-xr-x 3 test test 4.0K Mar 23 11:05 . drwxr-xr-x 6 root root 4.0K Mar 23 10:46 .. -rw------- 1 test test 16 Mar 23 10:47 .bash_history -rw-r--r-- 1 test test 127 Mar 19 01:59 .bash_logout -rw-r--r-- 1 test test 193 Mar 19 01:59 .bash_profile -rw-r--r-- 1 test test 551 Mar 19 01:59 .bashrc drwx------ 2 test test 4.0K Jul 3 2007 .ssh gnc.bin-Message: main: binreloc relocation support was disabled at configure time. Segmentation fault
Created attachment 306403 [details] /etc/gconf/schemas/desktop_gnome_sound.schemas try with my /etc/gconf/schemas/desktop_gnome_sound.schemas instead (better on a new created account with a new home to get it used instead of any other could be inside your home)
Behavior is still the same. I tried your "desktop_gnome_sound.schemas"-file and after that tried one without any translations (only strings for locale="C") and firefox still crashes the same way
Have you tried to compile firefox without "gconf" USE?
(In reply to comment #12) > Have you tried to compile firefox without "gconf" USE? That flag doesn't exist for firefox (but for thunderbird, where it is disabled): [ebuild R ] www-client/firefox-11.0-r1 USE="alsa crashreporter dbus ipc minimal system-sqlite webm wifi -bindist -custom-cflags -custom-optimization -debug -libnotify -pgo (-selinux) -startup-notification" LINGUAS="de -af -ak -ar -as -ast -be -bg -bn_BD -bn_IN -br -bs -ca -cs -csb -cy -da -el -en_GB -en_ZA -eo -es_AR -es_CL -es_ES -es_MX -et -eu -fa -fi -fr -fy_NL -ga_IE -gd -gl -gu_IN -he -hi_IN -hr -hu -hy_AM -id -is -it -ja -kk -kn -ko -ku -lg -lt -lv -mai -mk -ml -mr -nb_NO -nl -nn_NO -nso -or -pa_IN -pl -pt_BR -pt_PT -rm -ro -ru -si -sk -sl -son -sq -sr -sv_SE -ta -ta_LK -te -th -tr -uk -vi -zh_CN -zh_TW -zu" 76,957 kB [ebuild R ] mail-client/thunderbird-11.0.1 USE="alsa crashreporter crypt dbus ipc lightning minimal system-sqlite webm wifi -bindist -custom-cflags -custom-optimization -debug -gconf -libnotify -mozdom -startup-notification" LINGUAS="de -ar -ast -be -bg -bn_BD -br -ca -cs -da -el -en_GB -es_AR -es_ES -et -eu -fi -fr -fy_NL -ga_IE -gd -gl -he -hu -id -is -it -ja -ko -lt -nb_NO -nl -nn_NO -pa_IN -pl -pt_BR -pt_PT -rm -ro -ru -si -sk -sl -sq -sr -sv_SE -ta_LK -tr -uk -vi -zh_CN -zh_TW" 0 kB I'm using KDE as my DE and only have gnome-packages installed only when needed as dependencies for some other applications.
Please attach the file(s) in your ~/.gconf/desktop/gnome/sound
(In reply to comment #14) > Please attach the file(s) in your ~/.gconf/desktop/gnome/sound Not existant/empty: kripton@miramis ~ $ find ~/.gconf /home/kripton/.gconf kripton@miramis ~ $ file ~/.gconf /home/kripton/.gconf: directory kripton@miramis ~ $ Also see comment #8 where I tested stuff under a new user account just created for this purpose.
(In reply to comment #14) > Please attach the file(s) in your ~/.gconf/desktop/gnome/sound This bnug has nothing to do with sound issue. Esd is not required and has been patched out of both thunderbird and firefox. With the use of so many overlays, we could have a bug that is invalid due to a lib being included that is not part of official tree that is breaking libgnome support.
Created attachment 307481 [details] output of "eix -I -J" Sorry that this is quite verbose. It is a list of packages that have been installed from any overlay I'm using (using "eix -I -J"). If you think any package in that list might interfer with firefox/thunderbird/gnucash/libgnome-stuff, please tell me and I'll uninstall and retest.
I am having the same issue and also using gcc 3.5.4 on amd64 (I have an older kernel, though: 3.2.1-r2.). In addtion to firefox, balsa is also affected. $ gdb balsa GNU gdb (Gentoo 7.3.1 p2) 7.3.1 Copyright (C) 2011 Free Software Foundation, Inc. License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html> This is free software: you are free to change and redistribute it. There is NO WARRANTY, to the extent permitted by law. Type "show copying" and "show warranty" for details. This GDB was configured as "x86_64-pc-linux-gnu". For bug reporting instructions, please see: <http://bugs.gentoo.org/>... Reading symbols from /usr/bin/balsa...(no debugging symbols found)...done. (gdb) r Starting program: /usr/bin/balsa [Thread debugging using libthread_db enabled] [New Thread 0x7fffe6707700 (LWP 28957)] Program received signal SIGSEGV, Segmentation fault. 0x00007ffff2e0d028 in g_utf8_validate () from /usr/lib64/libglib-2.0.so.0 (gdb) I just rebuild every package on my system and found that firefox and mail-client/balsa are broken, but gimp, inkscape, and blender all work fine. I had portage-2.2.0_alpha95 installed at the time. I have since upgraded to portage-2.2.0_alpha99. I have tried glib-2.30.2 and glib-2.30.3. No difference: same segfault either way. I have found an interesting clue: All broken packages work if I: emerge --unmerge gconf firefox and balsa will present and error "An error occurred while loading or saving configuration information for balsa. Some of your configuration settings may not work properly". If you click "OK" on this error both applications run fine. If after that, I run emerge -1 gconf everything goes back to broken (segfault in /usr/lib64/libglib-2.0.so.0). (Using gconf-2.32.4 and I just got the same results with gconf-2.32.3, but I did not rebuild other packages after installing 2.32.3). I have repeated these steps several times with the same result every time. Not having gconf installed seems to skip the initialization steps that cause the issue. I will be running without gconf installed in the short term. I ran the above tests with balsa (because I am using firefox to compile and post the results). Firefox exhits the issues the same way. I have confirmed that it always works with gconf uninstalled and always segfaults with gconf installed. jannis: Can you try emerge --unmerge gconf and see if that "fixes" the issue for you? $ emerge --info Portage 2.2.0_alpha99 (default/linux/amd64/10.0/desktop, gcc-4.5.3, glibc-2.13-r4, 3.2.1-gentoo-r2 x86_64) ================================================================= System uname: Linux-3.2.1-gentoo-r2-x86_64-AMD_Athlon-tm-_64_X2_Dual_Core_Processor_3800+-with-gentoo-2.1 Timestamp of tree: Thu, 05 Apr 2012 03:45:01 +0000 app-shells/bash: 4.2_p20 dev-java/java-config: 2.1.11-r3 dev-lang/python: 2.7.2-r3, 3.2.2 dev-util/cmake: 2.8.6-r4 dev-util/pkgconfig: 0.26 sys-apps/baselayout: 2.1 sys-apps/openrc: 0.9.9.3 sys-apps/sandbox: 2.5 sys-devel/autoconf: 2.13, 2.68 sys-devel/automake: 1.11.1 sys-devel/binutils: 2.21.1-r1 sys-devel/gcc: 4.5.3-r2 sys-devel/gcc-config: 1.5-r2 sys-devel/libtool: 2.4-r1 sys-devel/make: 3.82-r1 sys-kernel/linux-headers: 3.1 (virtual/os-headers) sys-libs/glibc: 2.13-r4 Repositories: gentoo desktop-effects arduino portage marathon_reprap compiz Installed sets: ACCEPT_KEYWORDS="amd64" ACCEPT_LICENSE="* -@EULA AdobeFlash-10.3" CBUILD="x86_64-pc-linux-gnu" CFLAGS="-march=athlon64 -mtune=native -msse3 -O2 -pipe" CHOST="x86_64-pc-linux-gnu" CONFIG_PROTECT="/etc /usr/share/gnupg/qualified.txt /var/lib/hsqldb" CONFIG_PROTECT_MASK="/etc/ca-certificates.conf /etc/dconf /etc/env.d /etc/env.d/java/ /etc/fonts/fonts.conf /etc/gconf /etc/gentoo-release /etc/revdep-rebuild /etc/sandbox.d /etc/terminfo /etc/texmf/language.dat.d /etc/texmf/language.def.d /etc/texmf/updmap.d /etc/texmf/web2c" CXXFLAGS="-march=athlon64 -mtune=native -msse3 -O2 -pipe" DISTDIR="/usr/portage/distfiles" FEATURES="assume-digests binpkg-logs distlocks ebuild-locks fixlafiles news parallel-fetch preserve-libs protect-owned sandbox sfperms strict unknown-features-warn unmerge-logs unmerge-orphans userfetch" FFLAGS="-march=athlon64 -mtune=native -msse3 -O2 -pipe" GENTOO_MIRRORS="ftp://gentoo.cites.uiuc.edu/pub/gentoo/ http://gentoo.cs.uni.edu/ ftp://gentoo.in.th/ ftp://gentoo.mirrors.tds.net/gentoo" LDFLAGS="-Wl,-O1 -Wl,--as-needed" LINGUAS="en_US en" MAKEOPTS="-j3" PKGDIR="/usr/portage/packages" PORTAGE_CONFIGROOT="/" PORTAGE_RSYNC_OPTS="--recursive --links --safe-links --perms --times --compress --force --whole-file --delete --stats --human-readable --timeout=180 --exclude=/distfiles --exclude=/local --exclude=/packages" PORTAGE_TMPDIR="/var/tmp" PORTDIR="/usr/portage" PORTDIR_OVERLAY="/var/lib/layman/desktop-effects /var/lib/layman/luksans-arduino /usr/local/overlay/maurerpe /usr/local/overlay/reprap /usr/local/overlay/compiz" SYNC="rsync://rsync.namerica.gentoo.org/gentoo-portage" USE="3dnow 3dnowext X a52 aac acpi alsa amd64 apm apng bash-completion berkdb bidi branding bzip2 cairo cdda cdio cdr chroot cleartype cli colordiff consolekit cpudetection cracklib crypt cups cxx dbus dts dv dvd dvdr dvi emacs emboss emerald encode exif fam ffmpeg fftw firefox flac foomaticdb fortran gd gdbm gdu gif gimp glitz gmp gnome gnuplot gpm gs gstreamer gtk gtkhtml gtkspell guile gzip iconv icu id3tag imagemagick ipv6 jpeg jpeg2k lame lapack lcms libnotify lm_sensors lucene mad maildir matroska mbox mdb mime mmx mmxext mng modules mp3 mp4 mpeg mplayer mudflap multilib musicbrainz nautilus ncurses networking nls nodrm nptl nptlonly nsplugin nvidia ogg openexr opengl openssl pam pango pcre pdf plotutils png policykit posix ppds pppd pulseaudio qt3support quicktime raw readline regex rtc sbcl scanner sdl secure-delete session sftp sharedmem slideshow smp sockets sox sparse spell sqlite sqlite3 sse sse2 ssh ssl startup-notification svg sysfs syslog system-sqlite t1lib threads threadsafe tiff timidity toolkit-scroll-bars truetype udev unicode usb vorbis vpx wav wma wmf wxwidgets wxwindows x264 xcb xcomposite xine xml xmp xorg xpm xulrunner xv xvid xvmc 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 cgi cgid 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" CALLIGRA_FEATURES="kexi words flow plan sheets stage tables krita karbon braindump" CAMERAS="ptp2 canon" COLLECTD_PLUGINS="df interface irq load memory rrdtool swap syslog" ELIBC="glibc" GPSD_PROTOCOLS="ashtech aivdm earthmate evermore fv18 garmin garmintxt gpsclock itrax mtk3301 nmea ntrip navcom oceanserver oldstyle oncore rtcm104v2 rtcm104v3 sirf superstar2 timing tsip tripmate tnt ubx" INPUT_DEVICES="keyboard mouse evdev" KERNEL="linux" LCD_DEVICES="bayrad cfontz cfontz633 glk hd44780 lb216 lcdm001 mtxorb ncurses text" LINGUAS="en_US en" PHP_TARGETS="php5-3" RUBY_TARGETS="ruby18" SANE_BACKENDS="hp" USERLAND="GNU" VIDEO_CARDS="nv vesa vga fbdev apm nvidia" XTABLES_ADDONS="quota2 psd pknock lscan length2 ipv4options ipset ipp2p iface geoip fuzzy condition tee tarpit sysrq steal rawnat logmark ipmark dhcpmac delude chaos account" Unset: CPPFLAGS, CTARGET, EMERGE_DEFAULT_OPTS, INSTALL_MASK, LANG, LC_ALL, PORTAGE_BUNZIP2_COMMAND, PORTAGE_COMPRESS, PORTAGE_COMPRESS_FLAGS, PORTAGE_RSYNC_EXTRA_OPTS, USE_PYTHON
Have you tried to set your locale to "C" or english?
Just a short test with gnucash (used to crash, too) showed: After unmerging gconf, it doesn't crash anymore. Still don't know where exactly the cause is... libgnome or gconf :/
(In reply to comment #19) > Have you tried to set your locale to "C" or english? I am not exactly sure how to do this. I found the page: http://www.gentoo.org/doc/en/guide-localization.xml Somehow, I didn't even have an /etc/env.d/02locale, so I created one and set it to: LANG="en_US.UTF-8" LC_COLLATE="C" Then I ran as root: # env-update && source /etc/profile # emerge -1 =gnome-base/gconf-2.32.4 Then, as my normal user, I killed my x session, logged completely out, logged back in, restarted x and ran $ locale LANG=en_US.UTF-8 LC_CTYPE="en_US.UTF-8" LC_NUMERIC="en_US.UTF-8" LC_TIME="en_US.UTF-8" LC_COLLATE=C LC_MONETARY="en_US.UTF-8" LC_MESSAGES="en_US.UTF-8" LC_PAPER="en_US.UTF-8" LC_NAME="en_US.UTF-8" LC_ADDRESS="en_US.UTF-8" LC_TELEPHONE="en_US.UTF-8" LC_MEASUREMENT="en_US.UTF-8" LC_IDENTIFICATION="en_US.UTF-8" LC_ALL= $ balsa Segmentation fault $ LC_AL="C" balsa Segmentation fault It doesn't seem like this helped. Did I do it right? Is there something that I should have done differently?
(In reply to comment #21) [...] > $ LC_AL="C" balsa > Segmentation fault > > It doesn't seem like this helped. Did I do it right? Is there something > that I should have done differently? are YOU also able to reproduce that balsa crash on a new created user account with a NEW home?
Also confirm downgrading to gnome-base/libgnome-2.32.1 "fixes" this problem
(In reply to comment #23) > Also confirm downgrading to gnome-base/libgnome-2.32.1 "fixes" this problem For me, the crash also happens with gnome-base/libgnome-2.32.1
(In reply to comment #0) > Suddenly, after updating some system packages [...] Since we are lost here, I guess you will need to investigate what packages where updated when they broke. You need to check /var/log/emerge.log and try to extract that information, I know it's difficult, but I don't know what else try :(
(In reply to comment #22) > (In reply to comment #21) > [...] > > $ LC_AL="C" balsa > > Segmentation fault > > > > It doesn't seem like this helped. Did I do it right? Is there something > > that I should have done differently? > > are YOU also able to reproduce that balsa crash on a new created user > account with a NEW home? I created a new user and did a minimal configuration to start X, get a termal and execute: testuser$ LC_ALL="C" balsa Segmentation fault So no change. I will look at my emerge.log and generate a timeline.
(In reply to comment #25) > (In reply to comment #0) > > Suddenly, after updating some system packages > [...] > > Since we are lost here, I guess you will need to investigate what packages > where updated when they broke. You need to check /var/log/emerge.log and try > to extract that information, I know it's difficult, but I don't know what > else try :( Is there some kind of tool like "git bisect" but for package versions in the portage-tree? Maybe some old ebuilds are no longer available so we can't downgrade specific packages?
Ah wait, I have a full system backup from 20120316 so I'll try starting from that. Maybe in a virtual machine :)
(In reply to comment #28) > Ah wait, I have a full system backup from 20120316 so I'll try starting from > that. Maybe in a virtual machine :) Sounds promising :D
(In reply to comment #25) > (In reply to comment #0) > > Suddenly, after updating some system packages > [...] > > Since we are lost here, I guess you will need to investigate what packages > where updated when they broke. You need to check /var/log/emerge.log and try > to extract that information, I know it's difficult, but I don't know what > else try :( Here is my timeline: On 2012-03-25, I updated 45 packages, including firefox & dbus. NOTE THAT LIBGNOME WAS A REBUILD NOT AN UPDATE, both old and new versions were 2.32.1. I don't know what triggered the rebuild. emerge --newuse --deep --update --with-bdeps=y @world 2012-03-25 17:07: >>> emerge (1 of 45) dev-libs/expat-2.1.0_beta3 to / 2012-03-25 17:07: >>> emerge (2 of 45) sys-apps/pciutils-3.1.9-r1 to / 2012-03-25 17:08: >>> emerge (3 of 45) media-plugins/live-2012.01.07 to / 2012-03-25 17:09: >>> emerge (4 of 45) app-emulation/emul-linux-x86-baselibs-20120127 to / 2012-03-25 17:10: >>> emerge (5 of 45) app-emulation/emul-linux-x86-db-20120127 to / 2012-03-25 17:10: >>> emerge (6 of 45) app-shells/bash-4.2_p20 to / 2012-03-25 17:11: >>> emerge (7 of 45) dev-libs/libtasn1-2.12 to / 2012-03-25 17:12: >>> emerge (8 of 45) dev-libs/openssl-1.0.0h to / 2012-03-25 17:15: >>> emerge (9 of 45) net-libs/libssh2-1.4.0 to / 2012-03-25 17:16: >>> emerge (10 of 45) app-portage/conf-update-1.0.2 to / 2012-03-25 17:16: >>> emerge (11 of 45) media-libs/freetype-2.4.9 to / 2012-03-25 17:17: >>> emerge (12 of 45) sys-apps/dbus-1.4.18 to / 2012-03-25 17:18: >>> emerge (13 of 45) dev-db/sqlite-3.7.11 to / 2012-03-25 17:19: >>> emerge (14 of 45) media-libs/libmad-0.15.1b-r7 to / 2012-03-25 17:20: >>> emerge (15 of 45) dev-libs/nss-3.13.3 to / 2012-03-25 17:23: >>> emerge (16 of 45) media-fonts/liberation-fonts-1.07.2 to / 2012-03-25 17:23: >>> emerge (17 of 45) sys-apps/sg3_utils-1.33 to / 2012-03-25 17:24: >>> emerge (18 of 45) sys-kernel/gentoo-sources-3.2.12 to / 2012-03-25 17:25: >>> emerge (19 of 45) x11-libs/gtk+-3.2.4-r1 to / 2012-03-25 17:30: >>> emerge (20 of 45) app-admin/eselect-opencl-1.1.0-r1 to / 2012-03-25 17:31: >>> emerge (21 of 45) dev-python/dbus-python-1.0.0 to / 2012-03-25 17:31: >>> emerge (22 of 45) media-libs/ftgl-2.1.3_rc5 to / 2012-03-25 17:32: >>> emerge (23 of 45) media-libs/openjpeg-1.5.0 to / 2012-03-25 17:33: >>> emerge (24 of 45) x11-themes/gnome-themes-standard-3.2.1 to / 2012-03-25 17:33: >>> emerge (25 of 45) media-libs/taglib-1.7-r1 to / 2012-03-25 17:34: >>> emerge (26 of 45) x11-themes/gnome-icon-theme-symbolic-3.2.2 to / 2012-03-25 17:34: >>> emerge (27 of 45) media-sound/pulseaudio-1.1-r1 to / 2012-03-25 17:38: >>> emerge (28 of 45) sys-apps/openrc-0.9.9.3 to / 2012-03-25 17:38: >>> emerge (29 of 45) virtual/shadow-0 to / 2012-03-25 17:38: >>> emerge (30 of 45) media-video/ffmpeg-0.10.2 to / 2012-03-25 17:44: >>> emerge (31 of 45) media-video/mplayer-1.0_rc4_p20120213 to / 2012-03-25 17:45: >>> emerge (32 of 45) app-text/texlive-core-2011-r6 to / 2012-03-25 17:53: >>> emerge (33 of 45) media-plugins/gst-plugins-meta-0.10-r6 to / 2012-03-25 17:53: >>> emerge (34 of 45) dev-java/icedtea-bin-6.1.11.1-r1 to / 2012-03-25 17:54: >>> emerge (35 of 45) www-client/firefox-11.0-r1 to / 2012-03-25 18:44: >>> emerge (36 of 45) media-gfx/darktable-1.0 to / 2012-03-25 18:47: >>> emerge (37 of 45) gnome-base/libgnome-2.32.1 to / 2012-03-25 18:47: >>> emerge (38 of 45) app-emulation/emul-linux-x86-xlibs-20120127 to / 2012-03-25 18:48: >>> emerge (39 of 45) app-emulation/emul-linux-x86-opengl-20120127 to / 2012-03-25 18:48: >>> emerge (40 of 45) app-emulation/emul-linux-x86-gtklibs-20120127 to / 2012-03-25 18:48: >>> emerge (41 of 45) app-emulation/emul-linux-x86-medialibs-20120127 to / 2012-03-25 18:49: >>> emerge (42 of 45) app-emulation/emul-linux-x86-soundlibs-20120127 to / 2012-03-25 18:49: >>> emerge (43 of 45) app-emulation/emul-linux-x86-sdl-20120127 to / 2012-03-25 18:49: >>> emerge (44 of 45) app-emulation/wine-1.5.0 to / 2012-03-25 19:25: >>> emerge (45 of 45) x11-drivers/nvidia-drivers-295.20-r1 to / At this point EVERYTHING STILL WORKED: NO SEGFAULTS Over the next few days, I did some small emerges 2012-03-25 21:34: >>> emerge (1 of 1) app-text/poppler-0.16.7 to / 2012-03-28 19:05: >>> emerge (1 of 1) x11-drivers/nvidia-drivers-295.20-r1 to / 2012-03-28 19:06: >>> emerge (1 of 1) x11-drivers/nvidia-drivers-295.20-r1 to / 2012-03-28 19:13: >>> emerge (1 of 1) x11-drivers/nvidia-drivers-290.10 to / 2012-03-28 19:52: >>> emerge (1 of 1) sys-apps/pciutils-3.1.7 to / 2012-03-28 19:53: >>> emerge (1 of 1) sys-apps/openrc-0.9.9.2 to / 2012-03-28 20:13: >>> emerge (1 of 3) sys-apps/pciutils-3.1.9-r1 to / 2012-03-28 20:14: >>> emerge (2 of 3) sys-apps/openrc-0.9.9.3 to / 2012-03-28 20:14: >>> emerge (3 of 3) x11-drivers/nvidia-drivers-295.20-r1 to / 2012-03-28 20:23: >>> emerge (1 of 1) x11-drivers/nvidia-drivers-295.20-r1 to / 2012-03-28 20:24: >>> emerge (1 of 1) x11-drivers/nvidia-drivers-295.20-r1 to / 2012-04-02 20:28: >>> emerge (1 of 2) dev-tex/natbib-8.31b to / 2012-04-02 20:28: >>> emerge (2 of 2) dev-tex/revtex-4.1_p2-r1 to / Again, Everything still worked, NO SEGFAULTS Then on starting 2012-04-03, I rebuilt my entire system with $ emerge --deep --keep-going --emptytree @world This merged all 991 packages on my system and took more than 24 hours. The breakage happened before this build completed. From what I could tell using the system as things compiled, firefox broke at the moment the it was emerged. I noticed the breakage the next time I started firefox, which was within minutes. I had also started firefox a few minutes before the emerge (while it was still compiling) and it worked fine. I was not using the system when balsa was emerged, I cannot pinout its breakage as accurately. Let me know if you need any more details.
Created attachment 308205 [details] 1.png Looking to: https://bugs.launchpad.net/ubuntu/+source/gnome-utils/+bug/284923 looks like crashes like this would be caused by wrong gconf options, offending one looks to be: /desktop/gnome/sound I attach you what I see from gconf-editor in my systems, compare it with yours
Created attachment 308217 [details] gdb backtrace from gconf-editor I emerged gconf-editor and tried to navigate to /desktop/gnome/sound. As soon as I click the sound-folder, gconf-editor segfaults with attached backtrace
Created attachment 308221 [details] gdb backtrace of gconf-editor Forgot to tell portage not to strip debugging symbols via FEATURES
(In reply to comment #33) > Created attachment 308221 [details] > gdb backtrace of gconf-editor > > Forgot to tell portage not to strip debugging symbols via FEATURES I have the exact same result. Here are my exact steps: 1) launch gconf-editor -> Window appears 2) On the left panel, double click on desktop -> gnome appears under desktop 3) double click on gnome. 4) Single or double click on any item *except* sound -> settings appear on right pannel 5) Single click on sound -> program immedately segfaults. Settings on right do not change, highlight does not move to indicate sound active. Segfault is immediate. Backtrace is very similar to jannis: #0 0x00007ffff6943028 in g_utf8_validate () from /usr/lib64/libglib-2.0.so.0 #1 0x00007ffff7bb20c2 in gconf_value_from_corba_value () from /usr/lib64/libgconf-2.so.4 #2 0x00007ffff7bbf216 in gconf_engine_all_entries () from /usr/lib64/libgconf-2.so.4 #3 0x00007ffff7bc27ec in ?? () from /usr/lib64/libgconf-2.so.4 #4 0x00007ffff7bc57fb in gconf_client_add_dir () from /usr/lib64/libgconf-2.so.4 #5 0x000000000041476a in ?? () #6 0x000000000040f8c0 in ?? () #7 0x00007ffff6bfd5b4 in g_closure_invoke () from /usr/lib64/libgobject-2.0.so.0 #8 0x00007ffff6c0f682 in ?? () from /usr/lib64/libgobject-2.0.so.0 #9 0x00007ffff6c1886c in g_signal_emit_valist () from /usr/lib64/libgobject-2.0.so.0 <snip>
I repeated the gconf-editor test with a new user starting with a (nearly) empty home directory (no .gconf or .gconfd). 1) Start gconf-editor Window opens and the following error is printed: (gconf-editor:10741): GConf-CRITICAL **: Received list from gconfd with a bad list type GConf Error: Type mismatch: Expected list of String, got list of UNKNOWN, 0 2) Double click on desktop gnome appears under desktop 3) Double click on gnome Gnome expands to many sub-items 4) What happens next depends on which sub-item is selected. The following sub-items display settings on the right panel: thumbnail_cache typing_break The following sub-items print an error to the terminal and display no settings on the right: background font_rendering interface The following sub-items immediately segfault: fileviews sound Perhaps something is wrong with my system gconf settings? Where are these stored and what package sets them? Or does each package install its own system defaults?
Rebuild gnome-base/gconf-3.2.3 with USE="-orbit" and reboot?
(In reply to comment #36) > Rebuild gnome-base/gconf-3.2.3 with USE="-orbit" and reboot? Neither gnome-base/gconf-3.2.3 or gnome-base/gconf-3.2.4 have an "orbit" use flag. gconf-3.2.3: IUSE="debug doc +introspection ldap policykit" gconf-3.2.4: IUSE="debug doc +introspection ldap policykit" $ emerge -pv gconf These are the packages that would be merged, in order: Calculating dependencies... done! [ebuild N ] gnome-base/gconf-2.32.4 USE="introspection policykit -debug -doc -ldap" 0 kB Total: 1 package (1 new), Size of downloads: 0 kB
(gconf-editor:5541): GConf-CRITICAL **: Received list from gconfd with a bad list type this suggests that either the wrong kind of data is stored in gconf, or this file was corrupted somehow. This seems strange if the issue happens on a freshly created user as well. Anyway, here is how to get the values with gconftool-2: $ gconftool-2 --recursive-list /desktop/gnome/sound default_mixer_tracks = [alsamixer:hw:0] default_mixer_device = alsamixer:hw:0 theme_name = freedesktop theme-name = freedesktop input-feedback-sounds = false input_feedback_sounds = true event-sounds = true enable_esd = true event_sounds = true Could you paste yours ?
FTR, I'm running gconf-3.2.3 but this should not matter anyway. $ egrep "desktop/gnome/sound" /etc/gconf/schemas/* desktop_gnome_sound.schemas $ qfile desktop_gnome_sound.schemas gnome-base/libgnome (/etc/gconf/schemas/desktop_gnome_sound.schemas) now to figure out who/what is wrong on your system :)
(In reply to comment #38) > (gconf-editor:5541): GConf-CRITICAL **: Received list from gconfd with a bad > list type > > this suggests that either the wrong kind of data is stored in gconf, or this > file was corrupted somehow. This seems strange if the issue happens on a > freshly created user as well. > > Anyway, here is how to get the values with gconftool-2: > > $ gconftool-2 --recursive-list /desktop/gnome/sound > default_mixer_tracks = [alsamixer:hw:0] > default_mixer_device = alsamixer:hw:0 > theme_name = freedesktop > theme-name = freedesktop > input-feedback-sounds = false > input_feedback_sounds = true > event-sounds = true > enable_esd = true > event_sounds = true > > Could you paste yours ? $ gconftool-2 --recursive-list /desktop/gnome/sound input_feedback_sounds = false theme_name = freedesktop event_sounds = false enable_esd = false default_mixer_device = default_mixer_tracks = [] $ egrep "desktop/gnome/sound" /etc/gconf/schemas/* /etc/gconf/schemas/desktop_gnome_sound.schemas: <key>/schemas/desktop/gnome/sound/default_mixer_device</key> /etc/gconf/schemas/desktop_gnome_sound.schemas: <applyto>/desktop/gnome/sound/default_mixer_device</applyto> /etc/gconf/schemas/desktop_gnome_sound.schemas: <key>/schemas/desktop/gnome/sound/default_mixer_tracks</key> /etc/gconf/schemas/desktop_gnome_sound.schemas: <applyto>/desktop/gnome/sound/default_mixer_tracks</applyto> /etc/gconf/schemas/desktop_gnome_sound.schemas: <key>/schemas/desktop/gnome/sound/enable_esd</key> /etc/gconf/schemas/desktop_gnome_sound.schemas: <applyto>/desktop/gnome/sound/enable_esd</applyto> /etc/gconf/schemas/desktop_gnome_sound.schemas: <key>/schemas/desktop/gnome/sound/event_sounds</key> /etc/gconf/schemas/desktop_gnome_sound.schemas: <applyto>/desktop/gnome/sound/event_sounds</applyto> /etc/gconf/schemas/desktop_gnome_sound.schemas: <key>/schemas/desktop/gnome/sound/theme_name</key> /etc/gconf/schemas/desktop_gnome_sound.schemas: <applyto>/desktop/gnome/sound/theme_name</applyto> /etc/gconf/schemas/desktop_gnome_sound.schemas: <key>/schemas/desktop/gnome/sound/input_feedback_sounds</key> /etc/gconf/schemas/desktop_gnome_sound.schemas: <applyto>/desktop/gnome/sound/input_feedback_sounds</applyto> $ qfile desktop_gnome_sound.schemas gnome-base/libgnome (/etc/gconf/schemas/desktop_gnome_sound.schemas) $ equery list -p libgnome * Searching for libgnome ... [IP-] [ ] gnome-base/libgnome-2.32.1:0 [-P-] [ ~] gnome-base/libgnome-2.32.1-r1:0
Created attachment 308467 [details] Marathon's desktop_gnome_sound.schemas Here is my /etc/gconf/schemas/desktop_gnome_sound.schemas file.
The following was ran as my test user: $ gconftool-2 --recursive-list /desktop/gnome/sound input_feedback_sounds = false theme_name = freedesktop event_sounds = false enable_esd = false default_mixer_device = default_mixer_tracks = []
Created attachment 308501 [details] gdb backtrace (In reply to comment #38) Here's mine (segfault) kripton@miramis ~ $ egrep "desktop/gnome/sound" /etc/gconf/schemas/* /etc/gconf/schemas/desktop_gnome_sound.schemas: <key>/schemas/desktop/gnome/sound/default_mixer_device</key> /etc/gconf/schemas/desktop_gnome_sound.schemas: <applyto>/desktop/gnome/sound/default_mixer_device</applyto> /etc/gconf/schemas/desktop_gnome_sound.schemas: <key>/schemas/desktop/gnome/sound/default_mixer_tracks</key> /etc/gconf/schemas/desktop_gnome_sound.schemas: <applyto>/desktop/gnome/sound/default_mixer_tracks</applyto> /etc/gconf/schemas/desktop_gnome_sound.schemas: <key>/schemas/desktop/gnome/sound/enable_esd</key> /etc/gconf/schemas/desktop_gnome_sound.schemas: <applyto>/desktop/gnome/sound/enable_esd</applyto> /etc/gconf/schemas/desktop_gnome_sound.schemas: <key>/schemas/desktop/gnome/sound/event_sounds</key> /etc/gconf/schemas/desktop_gnome_sound.schemas: <applyto>/desktop/gnome/sound/event_sounds</applyto> /etc/gconf/schemas/desktop_gnome_sound.schemas: <key>/schemas/desktop/gnome/sound/theme_name</key> /etc/gconf/schemas/desktop_gnome_sound.schemas: <applyto>/desktop/gnome/sound/theme_name</applyto> /etc/gconf/schemas/desktop_gnome_sound.schemas: <key>/schemas/desktop/gnome/sound/input_feedback_sounds</key> /etc/gconf/schemas/desktop_gnome_sound.schemas: <applyto>/desktop/gnome/sound/input_feedback_sounds</applyto> /etc/gconf/schemas/desktop_gnome_sound.schemas.old: <key>/schemas/desktop/gnome/sound/default_mixer_device</key> /etc/gconf/schemas/desktop_gnome_sound.schemas.old: <applyto>/desktop/gnome/sound/default_mixer_device</applyto> /etc/gconf/schemas/desktop_gnome_sound.schemas.old: <key>/schemas/desktop/gnome/sound/default_mixer_tracks</key> /etc/gconf/schemas/desktop_gnome_sound.schemas.old: <applyto>/desktop/gnome/sound/default_mixer_tracks</applyto> /etc/gconf/schemas/desktop_gnome_sound.schemas.old: <key>/schemas/desktop/gnome/sound/enable_esd</key> /etc/gconf/schemas/desktop_gnome_sound.schemas.old: <applyto>/desktop/gnome/sound/enable_esd</applyto> /etc/gconf/schemas/desktop_gnome_sound.schemas.old: <key>/schemas/desktop/gnome/sound/event_sounds</key> /etc/gconf/schemas/desktop_gnome_sound.schemas.old: <applyto>/desktop/gnome/sound/event_sounds</applyto> /etc/gconf/schemas/desktop_gnome_sound.schemas.old: <key>/schemas/desktop/gnome/sound/theme_name</key> /etc/gconf/schemas/desktop_gnome_sound.schemas.old: <applyto>/desktop/gnome/sound/theme_name</applyto> /etc/gconf/schemas/desktop_gnome_sound.schemas.old: <key>/schemas/desktop/gnome/sound/input_feedback_sounds</key> /etc/gconf/schemas/desktop_gnome_sound.schemas.old: <applyto>/desktop/gnome/sound/input_feedback_sounds</applyto> kripton@miramis ~ $ qfile desktop_gnome_sound.schemas gnome-base/libgnome (/etc/gconf/schemas/desktop_gnome_sound.schemas) kripton@miramis ~ $ equery list -p libgnome * Searching for libgnome ... [-P-] [ ] gnome-base/libgnome-2.32.1:0 [IP-] [ ] gnome-base/libgnome-2.32.1-r1:0 kripton@miramis ~ $
Building gconf with USE="-orbit" (which is default by now) fixes the issue for me. See also https://bugs.gentoo.org/show_bug.cgi?id=411237 comment 5
Are you both using gconf-3?
(In reply to comment #45) > Are you both using gconf-3? I do: ekripton@miramis ~ $ eix -e gconf [I] gnome-base/gconf Available versions: (2) 2.32.3 2.32.4 (~)3.2.3 (~)3.2.3-r1 {{debug doc +introspection ldap (+)orbit policykit}} Installed versions: 3.2.3-r1(2)(10:16:11 11.04.2012)(doc introspection policykit -debug -ldap -orbit) Now having gconf with -orbit I tried running media-sound/grip (that hard-depends on orbit). It works but according to strace doesn't use orbit at all (but it does IPC using dbus).
(In reply to comment #44) > Building gconf with USE="-orbit" (which is default by now) fixes the issue > for me. See also https://bugs.gentoo.org/show_bug.cgi?id=411237 comment 5 I had been using gconf-2.32.4 which does not have an orbit use flag. I noticed that there is an unstable gconf-3.2.3-r1 in the tree. I emerged that with USE=-orbit (default) and the problem is fixed.
(In reply to comment #45) > Are you both using gconf-3? I was not. I have upgraded to gconf-3 with USE=-orbit and the problem is fixed. Thank you for sticking with us to the end, Pacho Ramos.
Does anything speak against closing this as RESOLVED FIXED?
But, does it still fail with USE "orbit"?
(In reply to comment #50) > But, does it still fail with USE "orbit"? Oh, you're right - it still fails (just tried again) so it's not FIXED. The workaround USE="-orbit" is just the default now so fewer people might hit this bug. What still might cause trouble is that gconf:2 uses orbit and can't be stopped doing so via USE-flag (according to the reports by Marathon
I have been fighting with this issue for a day now. I am wondering why the status of the bug is UNCONFIRMED? Anyway, I'm glad I'm not alone - I was beginning to suspect hardware failures. Here is a backtrace of gconfd-2 while trying to start evolution, which crashes as well with segfault (from a strace it seems none of the gconf-relying apps check that the socket they open to contact gconf is closed externally - but I'm not familiar with the code at all). cipi@pink ~ $ gdb /usr/libexec/gconfd-2 GNU gdb (Gentoo 7.3.1 p2) 7.3.1 Copyright (C) 2011 Free Software Foundation, Inc. License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html> This is free software: you are free to change and redistribute it. There is NO WARRANTY, to the extent permitted by law. Type "show copying" and "show warranty" for details. This GDB was configured as "x86_64-pc-linux-gnu". For bug reporting instructions, please see: <http://bugs.gentoo.org/>... Reading symbols from /usr/libexec/gconfd-2...(no debugging symbols found)...done. (gdb) run Starting program: /usr/libexec/gconfd-2 [Thread debugging using libthread_db enabled] Program received signal SIGSEGV, Segmentation fault. 0x00007ffff6d0e3bf in ?? () from /lib64/libc.so.6 (gdb) bt #0 0x00007ffff6d0e3bf in ?? () from /lib64/libc.so.6 #1 0x00007ffff7955af1 in giop_send_buffer_append_string (buf=0x625020, str=0x63d32100000000 <Address 0x63d32100000000 out of bounds>) at giop-send-buffer.c:542 #2 0x00007ffff7961e3c in ORBit_marshal_value (buf=0x625020, val=0x7fffffffc9c0, tc=0x7ffff7b94a20) at corba-any.c:229 #3 0x00007ffff7961f1f in ORBit_marshal_value (buf=0x625020, val=0x7fffffffc9c0, tc=0x7ffff7dd7d00) at corba-any.c:187 #4 0x00007ffff7961f8c in ORBit_marshal_value (buf=0x625020, val=0x7fffffffca28, tc=0x7ffff7dd7f80) at corba-any.c:210 #5 0x00007ffff7961fc3 in ORBit_marshal_arg (buf=<optimized out>, val=0x976f34, tc=<optimized out>) at corba-any.c:385 #6 0x00007ffff795ae60 in ORBit_small_invoke_adaptor (adaptor_obj=0x62d830, recv_buffer=0x631ce0, m_data=<optimized out>, data=0x7fffffffccf0, ev=0x7fffffffcde0) at orbit-small.c:918 #7 0x00007ffff796892e in ORBit_POAObject_handle_request (pobj=0x62d830, opname=0x97ceec "lookup_with_schema_name", ret=0x0, args=0x0, ctx=0x0, recv_buffer=0x631ce0, ev=0x7fffffffcde0) at poa.c:1357 #8 0x00007ffff7968f4a in ORBit_POAObject_invoke_incoming_request (pobj=0x62d830, recv_buffer=0x631ce0, opt_ev=0x7fffffffcde0) at poa.c:1427 #9 0x00007ffff79690d7 in ORBit_POA_handle_request (poa=0x6239e0, recv_buffer=0x631ce0, objkey=<optimized out>) at poa.c:1649 #10 0x00007ffff796cf1c in ORBit_handle_request (orb=0x6238c0, recv_buffer=0x631ce0) at orbit-adaptor.c:300 #11 0x00007ffff795768f in giop_connection_handle_input (lcnx=0x632020) at giop-recv-buffer.c:1312 #12 0x00007ffff7972f61 in link_connection_io_handler (gioc=<optimized out>, condition=G_IO_IN, data=0x632020) at linc-connection.c:1475 #13 0x00007ffff6fc888f in g_main_context_dispatch () from /usr/lib64/libglib-2.0.so.0 #14 0x00007ffff6fc9098 in ?? () from /usr/lib64/libglib-2.0.so.0 #15 0x00007ffff6fc95ea in g_main_loop_run () from /usr/lib64/libglib-2.0.so.0 #16 0x000000000040acaf in main () (gdb) Evolution's log: cipi@pink ~ $ evolution (evolution:14869): evolution-shell-WARNING **: Configuration server couldn't be contacted: CORBA error: IDL:omg.org/CORBA/COMM_FAILURE:1.0 (evolution:14869): evolution-shell-WARNING **: Configuration server couldn't be contacted: CORBA error: IDL:omg.org/CORBA/COMM_FAILURE:1.0 (evolution:14869): evolution-shell-WARNING **: Configuration server couldn't be contacted: CORBA error: IDL:omg.org/CORBA/COMM_FAILURE:1.0 (evolution:14869): evolution-shell-WARNING **: Configuration server couldn't be contacted: CORBA error: IDL:omg.org/CORBA/COMM_FAILURE:1.0 (evolution:14869): evolution-shell-WARNING **: Configuration server couldn't be contacted: CORBA error: IDL:omg.org/CORBA/COMM_FAILURE:1.0 (evolution:14869): evolution-shell-WARNING **: Configuration server couldn't be contacted: CORBA error: IDL:omg.org/CORBA/COMM_FAILURE:1.0 (evolution:14869): evolution-shell-WARNING **: Configuration server couldn't be contacted: CORBA error: IDL:omg.org/CORBA/COMM_FAILURE:1.0 Segmentation fault
I have migrated to gconf-3.2.5 with USE=-orbit (the default) and evolution started working again. Too bad all my settings got lost... but at least I have a mail client and firefox working again :)
Created attachment 316301 [details, diff] add_null_pointer_check.patch gboolean g_utf8_validate (const gchar *str, gssize max_len, const gchar **end); this function segfault if str == 0x0, instead considering it a invalid utf-8 string seem more reasonable, the patch attached add such test and fix all broken gnome application for me. (gdb) bt #0 gconf_value_from_corba_value (value=0x9002d0) at gconf-internals.c:136 #1 0x00007ffff7bbf8d9 in gconf_engine_all_entries (conf=<optimized out>, dir=0x8d70a0 "/desktop/gnome/sound", err=<optimized out>) at gconf.c:1781 #2 0x00007ffff7bc2999 in cache_pairs_in_dir (client=0x757800, dir=0x8d70a0 "/desktop/gnome/sound") at gconf-client.c:966 #3 0x00007ffff7bc58cb in gconf_client_add_dir (client=0x757800, dirname=0x8d70a0 "/desktop/gnome/sound", preload=GCONF_CLIENT_PRELOAD_ONELEVEL, err=0x0) at gconf-client.c:626 #4 0x00000000004133eb in gconf_list_model_set_root_path (model=0x811c50, root_path=0x8d70a0 "/desktop/gnome/sound") at gconf-list-model.c:156 #5 0x000000000040e0bf in gconf_editor_window_selection_changed (selection=<optimized out>, window=0x72c000) at gconf-editor-window.c:906 #6 0x00007ffff6c13e8d in _g_closure_invoke_va () from /lib64/libgobject-2.0.so.0 #7 0x00007ffff6c2c5c1 in g_signal_emit_valist () from /lib64/libgobject-2.0.so.0 #8 0x00007ffff6c2d0d2 in g_signal_emit () from /lib64/libgobject-2.0.so.0 #9 0x00007ffff77a47bc in gtk_tree_view_real_set_cursor (tree_view=0x7fe380, path=0x8e3450, clear_and_select=1, clamp_node=1) at gtktreeview.c:12594 #10 0x00007ffff77a82db in gtk_tree_view_button_press (widget=0x7fe380, event=0x7ca990) at gtktreeview.c:2797 #11 0x00007ffff76aafb8 in _gtk_marshal_BOOLEAN__BOXED (closure=0x716500, return_value=0x7fffffffd2e0, n_param_values=<optimized out>, param_values=0x7fffffffd350, invocation_hint=<optimized out>, marshal_data=<optimized out>) at gtkmarshalers.c:86 #12 0x00007ffff6c13c62 in g_closure_invoke () from /lib64/libgobject-2.0.so.0 #13 0x00007ffff6c25289 in signal_emit_unlocked_R () from /lib64/libgobject-2.0.so.0 #14 0x00007ffff6c2cc1e in g_signal_emit_valist () from /lib64/libgobject-2.0.so.0 #15 0x00007ffff6c2d0d2 in g_signal_emit () from /lib64/libgobject-2.0.so.0 #16 0x00007ffff77c0b81 in gtk_widget_event_internal (widget=0x7fe380, event=0x7ca990) at gtkwidget.c:4984 #17 0x00007ffff76a9153 in IA__gtk_propagate_event (widget=0x7fe380, event=0x7ca990) at gtkmain.c:2489 #18 0x00007ffff76a94b3 in IA__gtk_main_do_event (event=0x7ca990) at gtkmain.c:1684 #19 0x00007ffff7320e8c in gdk_event_dispatch (source=<optimized out>, callback=<optimized out>, user_data=<optimized out>) at gdkevents-x11.c:2377 #20 0x00007ffff692b183 in g_main_context_dispatch () from /lib64/libglib-2.0.so.0 #21 0x00007ffff692b4c0 in g_main_context_iterate.isra.23 () from /lib64/libglib-2.0.so.0 #22 0x00007ffff692b8ba in g_main_loop_run () from /lib64/libglib-2.0.so.0 #23 0x00007ffff76a8507 in IA__gtk_main () at gtkmain.c:1256 #24 0x0000000000409d19 in main (argc=1, argv=0x7fffffffd9f8) at main.c:108 (gdb) step 140 switch (value->_d) (gdb) 148 type = GCONF_VALUE_STRING; (gdb) 172 gval = gconf_value_new(type); (gdb) gconf_value_new (type=GCONF_VALUE_STRING) at gconf-value.c:63 63 { (gdb) 67 g_return_val_if_fail(GCONF_VALUE_TYPE_VALID(type), NULL); (gdb) 69 if (!initted) (gdb) 75 value = (GConfValue*) g_slice_new0 (GConfRealValue); (gdb) 77 value->type = type; (gdb) 84 } (gdb) gconf_value_from_corba_value (value=0x9002d0) at gconf-internals.c:174 174 switch (gval->type) (gdb) 172 gval = gconf_value_new(type); (gdb) 174 switch (gval->type) (gdb) 180 if (!g_utf8_validate (value->_u.string_value, -1, NULL)) (gdb) print value $1 = (const ConfigValue *) 0x9002d0 (gdb) print value->_u.string_value $2 = (CORBA_string) 0x0 (gdb) print value->_u.string_value[0] Cannot access memory at address 0x0 (gdb) print value->_u.string_value[1] Cannot access memory at address 0x1 (gdb) print value[0] $3 = {_d = StringVal, _u = {dummy = 0, int_value = 0, string_value = 0x0, float_value = 0, bool_value = 0 '\000', schema_value = {value_type = InvalidVal, value_list_type = InvalidVal, value_car_type = InvalidVal, value_cdr_type = InvalidVal, locale = 0x0, short_desc = 0x0, long_desc = 0x0, owner = 0x0, encoded_default_value = 0x0}, list_value = {seq = {_maximum = 0, _length = 0, _buffer = 0x0, _release = 0 '\000'}, list_type = BInvalidVal}, pair_value = {_maximum = 0, _length = 0, _buffer = 0x0, _release = 0 '\000'}}} (gdb)
I've been having this same problem after cleaning my USE flags and recompiling my system. It's affecting both firefox and gnucash. My firefox backtrace is the same as above. Francesco Riosa's patch fixes it. Commenting out the following line: <entry name="theme_name" mtime="1340736979" schema="/schemas/desktop/gnome/sound/theme_name"/> in /etc/gconf/gconf.xml.defaults/%gconf-tree.xml also prevented firefox from segfaulting. I'm still having problems with gnucash. When I start gnucash, everytime, it displays the window saying "Cannot find default values." It then proceeds to start a configuration druid giving me the option to "Update the search path" or "Install into home directory." Updating the search path segfaults in liborbit. Installing into my home directory calls /usr/bin/update-gnucash-gconf multiple times, each time segfaulting. Also after some time, gconfd-2 begins to use an excessive amount of CPU time, between 60% and 100%. # Relevant packages app-office/gnucash-2.4.10 gnome-base/gconf-2.32.4 gnome-base/orbit-2.14.19-r1 dev-libs/glib-2.30.3::local # Francesco Riosa's patch applied
Created attachment 316361 [details] emerge --info
Created attachment 316363 [details] gnucash backtrace
Created attachment 316365 [details] /usr/bin/update-gnucash-gconf errors
*** Bug 424982 has been marked as a duplicate of this bug. ***
re-add us if needed later.
Have you tried to run: gconftool-2 --unset /desktop/gnome/sound ?
(In reply to comment #61) > Have you tried to run: > gconftool-2 --unset /desktop/gnome/sound > ? Pachos, I'm not a C language expert but it should not crash for an empty string anyway, so a fix is advisable. Never mind that kde users (which use firefox) would never be able to assemble that gconf command. cheers, Francesco
(In reply to comment #62) > (In reply to comment #61) > > Have you tried to run: > > gconftool-2 --unset /desktop/gnome/sound > > ? > > Pachos, I'm not a C language expert but it should not crash for an empty > string anyway, so a fix is advisable. > Never mind that kde users (which use firefox) would never be able to > assemble that gconf command. > > cheers, Francesco Well, the idea is to get it regenerated from scratch as we still don't find what exact value is causing the problem for some people :/
(In reply to comment #63) > (In reply to comment #62) > > (In reply to comment #61) > > > Have you tried to run: > > > gconftool-2 --unset /desktop/gnome/sound > > > ? > > > > Pachos, I'm not a C language expert but it should not crash for an empty > > string anyway, so a fix is advisable. > > Never mind that kde users (which use firefox) would never be able to > > assemble that gconf command. > > > > cheers, Francesco > > Well, the idea is to get it regenerated from scratch as we still don't find > what exact value is causing the problem for some people :/ again, that's not the biggest problem, if the user or whatever edit the config file it's not acceptable that all gnome applications die that way. The problem (as shown in c#54) is that "g_utf8_validate" should (would say MUST) NOT crash if the ConfigValue parameter contains a null string. A better solution could be to print a warning and possibly a trace to find the wrong value, but never ever crash, as demonstrated by the fact the values are not yet known it's not even an informative crash.
(In reply to comment #64) > The problem (as shown in c#54) is that "g_utf8_validate" should (would say > MUST) NOT crash if the ConfigValue parameter contains a null string. No. The creators of glib believe that it's the caller's responsibility to make sure that functions like g_utf8_validate() are never called with an invalid parameter type - and while NULL is a valid character pointer, it's considered to be invalid as a C string. See https://bugzilla.gnome.org/show_bug.cgi?id=652526#c4
(In reply to comment #65) > (In reply to comment #64) > > The problem (as shown in c#54) is that "g_utf8_validate" should (would say > > MUST) NOT crash if the ConfigValue parameter contains a null string. > > No. The creators of glib believe that it's the caller's responsibility to > make sure that functions like g_utf8_validate() are never called with an > invalid parameter type - and while NULL is a valid character pointer, it's > considered to be invalid as a C string. See > https://bugzilla.gnome.org/show_bug.cgi?id=652526#c4 Ok so the resolution of this bug is INVALID or WONTFIX or CANTFIX, but may I request to leave it open anyway, for an easyer access, since it hit a number of people very badly. thanks, francesco
(In reply to comment #65) > (In reply to comment #64) > > The problem (as shown in c#54) is that "g_utf8_validate" should (would say > > MUST) NOT crash if the ConfigValue parameter contains a null string. > > No. The creators of glib believe that it's the caller's responsibility to > make sure that functions like g_utf8_validate() are never called with an > invalid parameter type - and while NULL is a valid character pointer, it's > considered to be invalid as a C string. See > https://bugzilla.gnome.org/show_bug.cgi?id=652526#c4 From a user experience standpoint, would it make sense to just patch g_utf8_validate and then have a blocker for removing that patch when all the offending callers are fixed? If things are breaking for users then we either need to keep this masked or work around it. Sure, maybe the problem is that every other package is broken, but our users just care that things work or they don't, and we can't fix every other package overnight. Otherwise, normally I'd say follow upstream.
(In reply to comment #67) > (In reply to comment #65) > > (In reply to comment #64) > > > The problem (as shown in c#54) is that "g_utf8_validate" should (would say > > > MUST) NOT crash if the ConfigValue parameter contains a null string. > > > > No. The creators of glib believe that it's the caller's responsibility to > > make sure that functions like g_utf8_validate() are never called with an > > invalid parameter type - and while NULL is a valid character pointer, it's > > considered to be invalid as a C string. See > > https://bugzilla.gnome.org/show_bug.cgi?id=652526#c4 > > From a user experience standpoint, would it make sense to just patch > g_utf8_validate and then have a blocker for removing that patch when all the > offending callers are fixed? > > If things are breaking for users then we either need to keep this masked or > work around it. Sure, maybe the problem is that every other package is > broken, but our users just care that things work or they don't, and we can't > fix every other package overnight. > > Otherwise, normally I'd say follow upstream. If running "gconftool-2 --unset /desktop/gnome/sound" and letting it to be regenerated solves the problem, we could add a comment to ebuild to suggest it then :-/
> If running "gconftool-2 --unset /desktop/gnome/sound" and letting it to be > regenerated solves the problem, we could add a comment to ebuild to suggest > it then :-/ which ebuild, of the hundreds failing?
(In reply to comment #69) > > If running "gconftool-2 --unset /desktop/gnome/sound" and letting it to be > > regenerated solves the problem, we could add a comment to ebuild to suggest > > it then :-/ > > which ebuild, of the hundreds failing? gconf one
Alternatively, maybe gnome-base/gconf-3.2.5 could be stabilized. Pacho, are there any problems with using gconf-3.2.5[-orbit] in gnome2?
(In reply to comment #71) > Alternatively, maybe gnome-base/gconf-3.2.5 could be stabilized. > > Pacho, are there any problems with using gconf-3.2.5[-orbit] in gnome2? I am still not using it, but maybe won't be possible to use "-orbit" with old gnome-panel applets :-/, what do you think?
it's possible at least to craft a tool that check *consistency/goodness* of the xml files? mind you all are well formed xml, what the tool should find are the occurences of bad values, for example empty corba string values. Then ship it with gnome-base/gconf upgrading to gnome-base/gconf-2.32.4 I've forgot to apply the patch and wasted time again. Then re-installed it to 3 boxes that stopped working. This bug stopped to be ridiculous too much time ago.
(In reply to comment #61) > Have you tried to run: > gconftool-2 --unset /desktop/gnome/sound > ? Does this solve the issue or not?
(In reply to comment #74) > (In reply to comment #61) > > Have you tried to run: > > gconftool-2 --unset /desktop/gnome/sound > > ? > > Does this solve the issue or not? And if it worked and solved? Would have that changed anything? Pacho, respect for upstream decision is a good thing, but sometimes, you just need to do the "right" thing. IMO the right thing is to NOT let ONE error in hundreds of KB of config files fuck half of a system (if possible), then hurry up to stabilize a "sane" version of glib and forget the hack. monfi ~ # LANG=C gconftool-2 --get /desktop/gnome/sound No value set for `/desktop/gnome/sound' monfi ~ # LANG=C gconftool-2 --unset /desktop/gnome/sound monfi ~ # LANG=C gconftool-2 --get /desktop/gnome/sound No value set for `/desktop/gnome/sound' emerge -1 --usepkgonly dev-libs/glib::gentoo >hom>viv>Imm>work$ firefox Errore di segmentazione emerge -1 --usepkgonly dev-libs/glib::vivo firefox "https://bugs.gentoo.org/show_bug.cgi?id=408857"
BTW I vaguely suspect media-video/cheese may be the remaining offender
(In reply to comment #75) > (In reply to comment #74) > > (In reply to comment #61) > > > Have you tried to run: > > > gconftool-2 --unset /desktop/gnome/sound > > > ? > > > > Does this solve the issue or not? > > And if it worked and solved? Would have that changed anything? I have already explained it: https://bugs.gentoo.org/show_bug.cgi?id=408857#c70 > Pacho, respect for upstream decision is a good thing, but sometimes, you > just need to do the "right" thing. > IMO the right thing is to NOT let ONE error in hundreds of KB of config > files fuck half of a system (if possible), then hurry up to stabilize a > "sane" version of glib and forget the hack. > > monfi ~ # LANG=C gconftool-2 --get /desktop/gnome/sound > No value set for `/desktop/gnome/sound' > monfi ~ # LANG=C gconftool-2 --unset /desktop/gnome/sound > monfi ~ # LANG=C gconftool-2 --get /desktop/gnome/sound > No value set for `/desktop/gnome/sound' > > emerge -1 --usepkgonly dev-libs/glib::gentoo > > >hom>viv>Imm>work$ firefox > Errore di segmentazione > > emerge -1 --usepkgonly dev-libs/glib::vivo > > firefox "https://bugs.gentoo.org/show_bug.cgi?id=408857" You are not unsetting proper value, when running gconftool-2 you are running it for /root home, while you are running firefox in your user account => you need to unset it in affected account
> You are not unsetting proper value, when running gconftool-2 you are running > it for /root home, while you are running firefox in your user account => you > need to unset it in affected account if configs are kept in ~/.gconf those are empty in my home: >hom>viv>.etc$ find .gconf* -ls 5506732 4 drwx------ 2 vivo users 4096 ago 1 22:19 .gconf 5506737 4 drwx------ 2 vivo users 4096 ago 1 22:20 .gconfd 5506494 1304 -rwx------ 1 vivo users 1333301 ago 1 22:20 .gconfd/saved_state I've tryed anyway the --unset run from user "vivo" and result are the same, everything crash. Additionally I've tryed to set a new empty string >hom>viv>Imm>work$ gconftool-2 --type string --set /desktop/gnome/sound "" Errore di segmentazione and a new value of type int $ gconftool-2 --type int --set /desktop/gnome/sound 0 $ echo $? 0 $ gconftool-2 --get /desktop/gnome/sound 0
for completeness a non empty string: $ gconftool-2 --type string --set /desktop/gnome/sound "test" Errore di segmentazione
I now believe that this is another manifestation of an old crossdev bug. If you *do* have sys-devel/crossdev installed, please update to >=crossdev-20120721, and afterwards re-emerge gnome-base/orbit and gnome-base/gconf. If you do *not* have sys-devel/crossdev, please say so!
(In reply to comment #80) > I now believe that this is another manifestation of an old crossdev bug. > > If you *do* have sys-devel/crossdev installed, please update to > >=crossdev-20120721, and afterwards re-emerge gnome-base/orbit and > gnome-base/gconf. > > If you do *not* have sys-devel/crossdev, please say so! I have crossdev installed, I'll check tomorrow ::gentoo versions of the packages mentioned
(In reply to comment #81) > (In reply to comment #80) > > I now believe that this is another manifestation of an old crossdev bug. > > > > If you *do* have sys-devel/crossdev installed, please update to > > >=crossdev-20120721, and afterwards re-emerge gnome-base/orbit and > > gnome-base/gconf. > > > > If you do *not* have sys-devel/crossdev, please say so! > > I have crossdev installed, I'll check tomorrow ::gentoo versions of the > packages mentioned work ok
*** This bug has been marked as a duplicate of bug 427184 ***