From: SourceForge.net <no...@so...> - 2007-02-14 00:28:22
|
Bugs item #1653199, was opened at 2007-02-06 03:23 Message generated for change (Comment added) made by nobody You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=109655&aid=1653199&group_id=9655 Please note that this message will contain a full copy of the comment thread, including the initial issue submission, for this request, not just the latest update. Category: gxine Group: v0.5.10 Status: Open Resolution: None Priority: 5 Private: No Submitted By: Nobody/Anonymous (nobody) Assigned to: Darren Salt (dsalt) Summary: 0.5.10, 0.5.11 crash at startup Initial Comment: I'm the lead developer for Puppy Linux, wanted to upgrade Gxine from 0.5.9. However, both 0.5.10 and 0.5.11 segfault at startup. I don't know anything about gdb, but had a go. Starting 0.5.11, get this: ---------------------------- (gdb) run Starting program: /usr/local/bin/gxine Failed to read a valid object file image from memory. (no debugging symbols found) (no debugging symbols found) .... Program received signal SIG32, Real-time event 32. 0xb7aef5ac in pthread_create () from /lib/libpthread.so.0 ------------------------- 0.5.9 and earlier run fine, so what has changed that causes this crash? Note: gtk 2.8.17. Regards, Barry Kauler www.puppyos.com ---------------------------------------------------------------------- Comment By: Nobody/Anonymous (nobody) Date: 2007-02-13 16:28 Message: Logged In: NO I tried the 'cset_1804_partial_reversion.patch', it still crashes: Program received signal SIGSEGV, Segmentation fault. 0xb7c79859 in gdk_window_set_user_data () from /usr/lib/libgdk-x11-2.0.so.0 Note, yesterday, on this same system, I was using Gxine 0.5.9 no problems. I took the precaution of deleting all of its files before running 'make install' for the new version. Configure options the same for compiling both: ./configure --sysconfdir=/etc --localstatedir=/var --build=i486-t2-linux-gnu --disable-lirc --with-x --disable-integration-wizard --without-dbus --disable-ffjs ...note, I need to put in the --disable-ffjs as have the SeaMonkey js lib and cannot find it unless have that option. Regards, Barry Kauler ---------------------------------------------------------------------- Comment By: Darren Salt (dsalt) Date: 2007-02-12 13:18 Message: Logged In: YES user_id=294680 Originator: NO Right... I've done a bisection and found the buggy changeset. It turns out that only part of it is buggy; a patch which reverts that part of it is attached to this bug report. The changeset in question is this one: http://zap.tartarus.org/~ds/hg/gxine/?cmd=changeset;node=1804;style=gitweb File Added: cset_1804_partial_reversion.patch ---------------------------------------------------------------------- Comment By: Nobody/Anonymous (nobody) Date: 2007-02-10 06:44 Message: Logged In: NO I tried the patch on the 0.5.11 source, with debugging, just ran it, got this: -------------------------------- gtkvideo: couldn't get a GDK handle for the video window! gxine has suffered a fatal internal error. -------------------------------- Compile with debug symbols, run with gdb: -------------------------------- (gdb) handle SIG32 noprint nostop Signal Stop Print Pass to program Description SIG32 No No Yes Real-time event 32 (gdb) run Starting program: /usr/local/bin/gxine Failed to read a valid object file image from memory. gtkvideo: couldn't get a GDK handle for the video window! Program received signal SIGSEGV, Segmentation fault. 0xb7cb7e95 in gdk_window_destroy_notify () from /usr/lib/libgdk-x11-2.0.so.0 ------------------------------------- Regards, Barry Kauler ---------------------------------------------------------------------- Comment By: Darren Salt (dsalt) Date: 2007-02-09 11:01 Message: Logged In: YES user_id=294680 Originator: NO That's actually a good match for the bug which I found. If it is the same bug, then the calling function is gtk_video_realize (src/gtkvideo.c, line 1045) and, if you use "up" to get to the appropriate stack frame (should be once), "print widget->window" will return 0 - in which case try the patch with 0.5.11 (expect one reject for ChangeLog). ---------------------------------------------------------------------- Comment By: Nobody/Anonymous (nobody) Date: 2007-02-08 21:39 Message: Logged In: NO Okay, I'm testing on the released 0.5.11 for now, compiled with debugging info. This is what I get: ------------------------- (gdb) handle SIG32 noprint nostop Signal Stop Print Pass to program Description SIG32 No No Yes Real-time event 32 (gdb) run Starting program: /usr/local/bin/gxine Failed to read a valid object file image from memory. Program received signal SIGSEGV, Segmentation fault. 0xb7c4d859 in gdk_window_set_user_data () from /usr/lib/libgdk-x11-2.0.so.0 (gdb) ---------------------------- Gxine gets as far as displaying the splash window, then immediately the segfault. It doesn't have anything to do with the splash screen as I tried with 'gdb --args gxine --no-splash' and got the same crash. Regards, Barry Kauler ---------------------------------------------------------------------- Comment By: Darren Salt (dsalt) Date: 2007-02-07 12:12 Message: Logged In: YES user_id=294680 Originator: NO Additionally, could you test this tarball: http://zap.tartarus.org/~ds/hg/gxine/?ca=3a99ad1de4d7;type=bz2 And again, with this patch: http://zap.tartarus.org/~ds/hg/gxine/?cmd=changeset;node=fda818f21cef72582722af10f24baa3080fd64d0;style=raw I suspect that that patch will fix your problem (not the SIG32, which is expected) - it certainly appears to fix a problem which I'm seeing in a chroot environment... ---------------------------------------------------------------------- Comment By: Darren Salt (dsalt) Date: 2007-02-06 11:13 Message: Logged In: YES user_id=294680 Originator: NO Please rebuild xine/gxine/xine-lib with debug information. This will help us identify where and why it crashed. To rebuild xine-lib, xine-ui or gxine with debug information, unpack the tarball and execute the following commands: $ ./configure $ make clean $ make debug If using xine-ui, install it: $ make install-debug Then run xine or gxine from gdb: $ gdb xine (gdb) handle SIG32 noprint nostop (gdb) run $ ./gxine-test --gdb (gdb) handle SIG32 noprint nostop (gdb) run Try to reproduce the crash. Program received signal SIGSEGV, Segmentation fault. [....] Type: (gdb) thread apply all bt This will print the stack trace for all threads. Paste it into the comment area for the bug. ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=109655&aid=1653199&group_id=9655 |