The latest Gaim (0.80) crashes unexpectedly after some time of being unused. I couldn't find the reason why is it so... Here is the backtrace of a core file after emerging it with USE="debug": #0 0x406a8351 in kill () from /lib/libc.so.6 #1 0x40031b81 in pthread_kill () from /lib/libpthread.so.0 #2 0x40031efb in raise () from /lib/libpthread.so.0 #3 0x406a80e4 in raise () from /lib/libc.so.6 #4 0x406a9638 in abort () from /lib/libc.so.6 #5 0x08108cf7 in sighandler (sig=0) at main.c:374 #6 0x400346b5 in __pthread_clock_settime () from /lib/libpthread.so.0 #7 <signal handler called> #8 0x406e2270 in mallopt () from /lib/libc.so.6 #9 0x406e1a3e in mallopt () from /lib/libc.so.6 #10 0x406e0c33 in malloc () from /lib/libc.so.6 #11 0x4050e577 in g_malloc () from /usr/lib/libglib-2.0.so.0 #12 0x000001f8 in ?? () #13 0x08a28ee0 in ?? () #14 0xbfff8b04 in ?? () #15 0x4050e713 in g_free () from /usr/lib/libglib-2.0.so.0 #16 0x08a28ee0 in ?? () #17 0x404a04d4 in ?? () from /usr/lib/libpango-1.0.so.0 Any ideas or sugesstions what to check? ediap Reproducible: Sometimes Steps to Reproduce: Portage 2.0.50-r9 (default-x86-1.4, gcc-3.3.3, glibc-2.3.3.20040420-r0, 2.6.7-gentoo-r11) ================================================================= System uname: 2.6.7-gentoo-r11 i686 Pentium III (Coppermine) Gentoo Base System version 1.4.16 Autoconf: sys-devel/autoconf-2.59-r3 Automake: sys-devel/automake-1.8.3 ACCEPT_KEYWORDS="x86" AUTOCLEAN="yes" CFLAGS="-march=pentium3 -mcpu=pentium3 -O3 -pipe -fomit-frame-pointer" CHOST="i686-pc-linux-gnu" COMPILER="gcc3" CONFIG_PROTECT="/etc /usr/X11R6/lib/X11/xkb /usr/kde/2/share/config /usr/kde/3/share/config /usr/share/config /usr/share/texmf/dvipdfm/config/ /usr/share/texmf/dvips/config/ /usr/share/texmf/tex/generic/config/ /usr/share/texmf/tex/platex/config/ /usr/share/texmf/xdvi/ /var/qmail/control" CONFIG_PROTECT_MASK="/etc/gconf /etc/terminfo /etc/env.d" CXXFLAGS="-march=pentium3 -mcpu=pentium3 -O3 -pipe -fomit-frame-pointer" DISTDIR="/usr/portage/distfiles" FEATURES="autoaddcvs ccache fixpackages sandbox" GENTOO_MIRRORS="http://src.gentoo.pl http://gentoo.zie.pg.gda.pl" MAKEOPTS="-j2" PKGDIR="/usr/portage/packages" PORTAGE_TMPDIR="/var/tmp" PORTDIR="/usr/portage" PORTDIR_OVERLAY="/usr/local/portage" SYNC="rsync://rsync3.pl.gentoo.org/gentoo-portage" USE="X aalib acpi acpi4linux alsa apache2 apm auctex avi berkdb bindist crypt cups dvd emacs encode esd f77 fbcon fftw foomaticdb freetype gdbm gif gpm gtk gtk2 imagemagick imlib jabber java jpeg kadu-modules kadu-voice lcd leim libg++ libwww linguas_pl mad maildir mikmod mmx motif moznocompose moznoirc moznomail mpeg ncurses nls objc oggvorbis opengl oss pam pdflib perl pic png python qt quicktime readline samba sasl sdl sis slang speedo spell sse ssl svga tcltk tcpd tetex tiff tlen truetype type1 usb video_cards_sis x86 xfs xml xml2 xmms xv xvid zlib"
Update: It alsa crashed while talking with someone from GaduGadu network. So not only after some time of being unused... :(
This is dieing at a very low level doing mallocs. I would double check memory and maybe diagnostics on the system. Also no overclocking, if you have any. If this is very reproducable can you try with maybe only one protocol? And to be safe backup and remove ~/.gaim and have it create a new one?
export MALLOC_CHECK_=2 then run gaim in gdb again.
I don't think it is a problem with my hardware... However, after removing my old .gaim config and creating a completely new one, I haven't observed any crash today... So I am going to test it for a few more days and check if it is going to crash ;-)
OK... Finally it crashed again... Here is the new backtrace: (gdb) bt #0 0x406a8351 in kill () from /lib/libc.so.6 #1 0x40031b81 in pthread_kill () from /lib/libpthread.so.0 #2 0x40031efb in raise () from /lib/libpthread.so.0 #3 0x406a80e4 in raise () from /lib/libc.so.6 #4 0x406a9638 in abort () from /lib/libc.so.6 #5 0x406dfb94 in _IO_str_pbackfail () from /lib/libc.so.6 #6 0x406e0d85 in free () from /lib/libc.so.6 #7 0x4122be42 in gg_free_event (e=0x0) at libgg.c:758 #8 0x4122eaee in main_callback (data=0x83bed58, source=1073978912, cond=GAIM_INPUT_READ) at gg.c:656 #9 0x080da4c5 in gaim_gtk_io_invoke (source=0x0, condition=1073970052, data=0x83bed40) at gtkeventloop.c:61 #10 0x4052e7c0 in g_io_channel_unix_get_fd () from /usr/lib/libglib-2.0.so.0 #11 0x0836ac28 in ?? () #12 0x00000001 in ?? () #13 0x083bed40 in ?? () #14 0x40522de7 in g_static_private_get () from /usr/lib/libglib-2.0.so.0 #15 0x4056233c in ?? () from /usr/lib/libglib-2.0.so.0 #16 0x083c62f0 in ?? () #17 0x405620fc in g_idle_funcs () from /usr/lib/libglib-2.0.so.0 #18 0x4050a489 in g_idle_remove_by_data () from /usr/lib/libglib-2.0.so.0 #19 0x083c62f0 in ?? () #20 0x080da480 in gaim_gtk_io_destroy () at gtkeventloop.c:42 Previous frame inner to this frame (corrupt stack?) Any suggestions?
I'm going to ask some upstream folks to take a look and see what they think. It does look like an upstream bug at the moment. Sit tight for now.
I am able to get recreate this bug with GAIM 0.80... all that I have to do is every couple seconds, double click on the same name in the buddy list. Usually letting the tool tip appear in between clicks. After 20 seconds of this torture, Gaim will segfault. (Installing now with debug to compare backtrace)
FYI I've modified the gaim-0.80.ebuild to filter out the unstable flags. Can you rebuild with it (cvs revision 1.7) and try again?
I recompiled the latest CVS revision of gaim: # $Header: /var/cvsroot/gentoo-x86/net-im/gaim/gaim-0.80.ebuild,v 1.8 2004/07/28 06:04:20 eradicator Exp $ and it did not solve the crash problem... Here is the latest backtrace from gdb: (gdb) bt #0 0x406a8351 in kill () from /lib/libc.so.6 #1 0x40031b81 in pthread_kill () from /lib/libpthread.so.0 #2 0x40031efb in raise () from /lib/libpthread.so.0 #3 0x406a80e4 in raise () from /lib/libc.so.6 #4 0x406a9638 in abort () from /lib/libc.so.6 #5 0x406dfb94 in _IO_str_pbackfail () from /lib/libc.so.6 #6 0x406e0d85 in free () from /lib/libc.so.6 #7 0x412261d2 in gg_free_event (e=0x0) at libgg.c:758 #8 0x41228cce in main_callback (data=0x82ce1f0, source=1073978912, cond=GAIM_INPUT_READ) at gg.c:656 #9 0x080d15d5 in gaim_gtk_io_invoke (source=0x0, condition=1073970052, data=0x82e1bd8) at gtkeventloop.c:61 #10 0x4052e7c0 in g_io_channel_unix_get_fd () from /usr/lib/libglib-2.0.so.0 #11 0x082e19f0 in ?? () #12 0x00000001 in ?? () #13 0x082e1bd8 in ?? () #14 0x40522de7 in g_static_private_get () from /usr/lib/libglib-2.0.so.0 #15 0x4056233c in ?? () from /usr/lib/libglib-2.0.so.0 #16 0x081f9ae8 in ?? () #17 0x405620fc in g_idle_funcs () from /usr/lib/libglib-2.0.so.0 #18 0x4050a489 in g_idle_remove_by_data () from /usr/lib/libglib-2.0.so.0 #19 0x081f9ae8 in ?? () #20 0x080d1590 in gaim_gtk_io_destroy () at gtkeventloop.c:42 Previous frame inner to this frame (corrupt stack?)
Any chance you could attach your config.log from your build?
Is it possible for you to provide output from gaim -d when the crash occurs as well? The guy handling the gg prpl would like some more details.
Created attachment 36796 [details] Here is the config.log from the crashing gaim-0.80 compilation
Don: I will try to obtain a few more details using "gaim -d", however it seems that today the GG plugin does not connect to the Gadu-Gadu network. But as soon as I get someting, I will put it here... Thanks for your patience...
Just popping in to see if this occurs with 0.81 as well.
Created attachment 37649 [details] Debug of the crashing gaim 0.81-r2
Unfortunatelly, version 0.81-r2 also crashes in similar way. Here is the latest bugtrace: (gdb) bt #0 0x406a2351 in kill () from /lib/libc.so.6 #1 0x40032b81 in pthread_kill () from /lib/libpthread.so.0 #2 0x40032efb in raise () from /lib/libpthread.so.0 #3 0x406a20e4 in raise () from /lib/libc.so.6 #4 0x406a3638 in abort () from /lib/libc.so.6 #5 0x080fb2e5 in sighandler (sig=0) at main.c:369 #6 0x400356b5 in __pthread_clock_settime () from /lib/libpthread.so.0 #7 <signal handler called> #8 0x406dc270 in mallopt () from /lib/libc.so.6 #9 0x406dba3e in mallopt () from /lib/libc.so.6 #10 0x406dac33 in malloc () from /lib/libc.so.6 #11 0x40508127 in g_malloc () from /usr/lib/libglib-2.0.so.0 #12 0x000001f8 in ?? () #13 0x083c4550 in ?? () #14 0xbfff8b34 in ?? () #15 0x405082c3 in g_free () from /usr/lib/libglib-2.0.so.0 #16 0x083c4550 in ?? () #17 0x4049ae14 in ?? () from /usr/lib/libpango-1.0.so.0 Please also look at the attached output of the gaim session using "gaim -d > gaim_debug.txt 2>&1"
Adam did you `export MALLOC_CHECK_=2` before getting that latest backtrace?
Sorry, I forgot about the "MALLOC_CHECK_=2" ;-) Here is the latest backtrace (for gaim 0.82): (gdb) bt #0 0x406a2351 in kill () from /lib/libc.so.6 #1 0x40032b81 in pthread_kill () from /lib/libpthread.so.0 #2 0x40032efb in raise () from /lib/libpthread.so.0 #3 0x406a20e4 in raise () from /lib/libc.so.6 #4 0x406a3638 in abort () from /lib/libc.so.6 #5 0x406d9b94 in _IO_str_pbackfail () from /lib/libc.so.6 #6 0x406dad85 in free () from /lib/libc.so.6 #7 0x412161d2 in gg_free_event (e=0x0) at libgg.c:755 #8 0x41218c7e in main_callback (data=0x82e5150, source=1073983008, cond=GAIM_INPUT_READ) at gg.c:656 #9 0x080cf4f5 in gaim_gtk_io_invoke (source=0x0, condition=1073974148, data=0x82ef330) at gtkeventloop.c:46 #10 0x40528400 in g_io_channel_unix_get_fd () from /usr/lib/libglib-2.0.so.0 #11 0x082ee870 in ?? () #12 0x00000001 in ?? () #13 0x082ef330 in ?? () #14 0x4051ca27 in g_static_private_get () from /usr/lib/libglib-2.0.so.0 #15 0x4055b33c in ?? () from /usr/lib/libglib-2.0.so.0 #16 0x080cf4b0 in gaim_gtkdialogs_remove_chat (chat=0x80cf4b0) at gtkdialogs.c:882 Previous frame inner to this frame (corrupt stack?) I also attach the output of the "gaim -d > gaim_debug.txt 2>&1" command... Hope this help, ediap
Created attachment 38277 [details] Debug of the crashing gaim 0.82-r1
Other Crash: gaim-encryption: Finding conversation: xxx@hotmail.com, 0x84eddc0 gaim-encryption: set_rx_encryption 'xxx@hotmail.com': 0x8429d18 : 0 : 0 gaim-encryption: No header: <FONT FACE="arial Bold"><B><FONT COLOR="#ff6d00">als je bij het lettertype gaat verandere de times now roman special 1 aanklikt ga je off</FONT></B></FONT> msn: S: SB 001: MSG xxx@hotmail.com (#)MiChAeL(#) 423 msn: Got User Display: 3O82OnQQS0w4JxaoNbQB8C5j+Rg= The application 'gaim' lost its connection to the display :0.0; most likely the X server was shut down or you killed/destroyed the application.
Aditional information: Other user was using Gaim too (Windows version), with a not resized buddy icon, lots bigger than 100x100. I resume a rescaling function for any buddy icon would do the trick.
Don't post other crashes here that are obviously completely unrelated. That is just garbage.
Adam is this still a problem in gaim-1.0.3?
Unfortunately, I can not say so, because I stopped using it a few weeks ago and switched to Psi. Sorry, guys.
Alright I'm going to close this as WORKSFORME then. The GG developer and I have both looked at that code and can't see why it would get down that far.
