Bug#700612: i3-wm: i3wm crashed corrupted double-linked list
Hi Ian, Ian Campbell i...@hellion.org.uk writes: This is bug #653327, not in i3. BTW, I wonder what you think of including an i3-debug.desktop in the i3-dbg package with the shmlogging enabled? Might be convenient? At first glance, that is a good idea. However, since in the next version we will have the ability to enable shared memory and debug logging at runtime, this is not necessary. Also, bug #653327 makes it much more complicated than just shipping another .desktop file, because to circumvent it, we would need to ship a shell script and have a manpage for it, etc. Therefore, I think it’s not worth it. -- Best regards, Michael -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#700612: i3-wm: i3wm crashed corrupted double-linked list
On Sat, 2013-07-06 at 12:10 +0200, Michael Stapelberg wrote: Hi Ian, Ian Campbell i...@hellion.org.uk writes: This is bug #653327, not in i3. BTW, I wonder what you think of including an i3-debug.desktop in the i3-dbg package with the shmlogging enabled? Might be convenient? At first glance, that is a good idea. However, since in the next version we will have the ability to enable shared memory and debug logging at runtime, this is not necessary. Also, bug #653327 makes it much more complicated than just shipping another .desktop file, because to circumvent it, we would need to ship a shell script and have a manpage for it, etc. Therefore, I think it’s not worth it. Yes, fair enough. Ian. -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#700612: i3-wm: i3wm crashed corrupted double-linked list
On Sun, 2013-06-30 at 11:06 +0200, Michael Stapelberg wrote: Could you upgrade to 4.5.1? You can do that without logging out. FYI I am now running 4.5.1-2 with the --shmlog-size=26214400, will let you/the bug know if I see this crash again. BTW, the problem I had before which prevented me from enabling --shmlog-size is that after editing i3.desktop to have Exec=i3 --shmlog-size=26214400 gdm3 would fail with Cannot exec 'i3 --shmlog-size=26214400' (or words to that affect. I solved this by creating /usr/local/bin/i3-debug with the obvious contents and launching that via Exec=. I don't know if this is a bug in the i3 docs or a bug in gdm3's handling of Exec= in session.desktop files. Normal application.desktop files will correctly handle arguments given in the Exec= line (at least judging from the examples in /usr/share/applications. What do you think? Ian. -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#700612: i3-wm: i3wm crashed corrupted double-linked list
Hi Ian, Ian Campbell i...@hellion.org.uk writes: BTW, the problem I had before which prevented me from enabling --shmlog-size is that after editing i3.desktop to have Exec=i3 --shmlog-size=26214400 gdm3 would fail with Cannot exec 'i3 --shmlog-size=26214400' (or words to that affect. I solved this by creating /usr/local/bin/i3-debug with the obvious contents and launching that via Exec=. I don't know if this is a bug in the i3 docs or a bug in gdm3's handling of Exec= in session.desktop files. Normal application.desktop files will correctly handle arguments given in the Exec= line (at least judging from the examples in /usr/share/applications. What do you think? This is bug #653327, not in i3. -- Best regards, Michael -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#700612: i3-wm: i3wm crashed corrupted double-linked list
On Thu, 2013-07-04 at 22:47 +0200, Michael Stapelberg wrote: Hi Ian, Ian Campbell i...@hellion.org.uk writes: BTW, the problem I had before which prevented me from enabling --shmlog-size is that after editing i3.desktop to have Exec=i3 --shmlog-size=26214400 gdm3 would fail with Cannot exec 'i3 --shmlog-size=26214400' (or words to that affect. I solved this by creating /usr/local/bin/i3-debug with the obvious contents and launching that via Exec=. I don't know if this is a bug in the i3 docs or a bug in gdm3's handling of Exec= in session.desktop files. Normal application.desktop files will correctly handle arguments given in the Exec= line (at least judging from the examples in /usr/share/applications. What do you think? This is bug #653327, not in i3. Thanks for the tip! BTW, I wonder what you think of including an i3-debug.desktop in the i3-dbg package with the shmlogging enabled? Might be convenient? Ian. -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#700612: i3-wm: i3wm crashed corrupted double-linked list
Hi Ian, Just FYI: I cannot act on the updated information you provided, since there still is no log. Debugging an i3 problem without a log is a super-tedious thing which I am not willing to do. Especially if it’s for an ancient version, such as 4.2. Could you upgrade to 4.5.1? You can do that without logging out. If you are willing to use the git version (see http://i3wm.org/docs/repositories.html), you can even enable shared memory logging without logging it (debug level will soon be adjustable at run time, too). -- Best regards, Michael -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#700612: i3-wm: i3wm crashed corrupted double-linked list
On Sun, 2013-06-30 at 11:06 +0200, Michael Stapelberg wrote: Hi Ian, Just FYI: I cannot act on the updated information you provided, since there still is no log. Understood. I figured it was better to give whatever information I had than nothing, I quite understand if this bug languishes until I manage to get some better logging. BTW I tried enabling logging but it failed (for reasons I can't recall) and since this is my work machine I didn't have much time to poke at it so ended up reverting the command line. I'm in the middle of setting up a fresh machine, so I will try again on that one. Debugging an i3 problem without a log is a super-tedious thing which I am not willing to do. Especially if it’s for an ancient version, such as 4.2. Ancient or not it is the version in Wheezy, although I mostly track testing so I'm quite happy to upgrade. Could you upgrade to 4.5.1? You can do that without logging out. Sure. If you are willing to use the git version (see http://i3wm.org/docs/repositories.html), you can even enable shared memory logging without logging it (debug level will soon be adjustable at run time, too). I think I'll wait for 4.6 for that functionality. Ian. signature.asc Description: This is a digitally signed message part
Bug#700612: i3-wm: i3wm crashed corrupted double-linked list
On Sun, 2013-05-26 at 08:00 +0100, Ian Campbell wrote: On Sat, 2013-05-25 at 17:34 +0200, Michael Stapelberg wrote: Hi Ian, As you mentioned, without i3-wm-dbg installed, the backtrace is not going to help. Can you still reproduce that problem/does it still happen? It still happens occasionally but not regularly. I had an instance last week but I stupidly forgot to gather a backtrace. I'll see if I can provoke it with vncviewer next week when I'm back at my desk. I forgot to poke but I just got a spontaneous instance of the issue. In .xsession-errors I get: 03/06/13 14:50:59 - ERROR: X11 Error received! sequence 0x34e7, error_code = 3 03/06/13 14:50:59 - ERROR: X11 Error received! sequence 0x34e8, error_code = 3 03/06/13 14:51:01 - ERROR: X11 Error received! sequence 0x358f, error_code = 3 03/06/13 14:51:01 - ERROR: X11 Error received! sequence 0x3590, error_code = 3 03/06/13 14:51:18 - ERROR: No state found i3: src/x.c:79: state_for_frame: Assertion `0' failed. This isn't exactly the same as the previous linked-list corruption, although I suppose memory corruption is also plausible in this instance. Attaching with gdb I got: (gdb) bt #0 0x7f94057a31a0 in __poll_nocancel () at ../sysdeps/unix/syscall-template.S:81 #1 0x7f94070391c2 in _xcb_conn_wait (c=c@entry=0xb036e0, cond=cond@entry=0xb03720, vector=vector@entry=0x0, count=count@entry=0x0) at ../../src/xcb_conn.c:400 #2 0x7f940703a93f in xcb_wait_for_event (c=0xb036e0) at ../../src/xcb_in.c:517 #3 0x00429344 in handle_signal (sig=optimized out, info=optimized out, data=optimized out) at src/sighandler.c:181 #4 signal handler called #5 0x7f94056fb295 in __GI_raise (sig=sig@entry=6) at ../nptl/sysdeps/unix/sysv/linux/raise.c:56 #6 0x7f94056fe438 in __GI_abort () at abort.c:90 #7 0x7f94056f4322 in __assert_fail_base (fmt=0x7f9405831a40 %s%s%s:%u: %s%sAssertion `%s' failed.\n%n, assertion=assertion@entry=0x4537b2 0, file=file@entry=0x44eafa src/x.c, line=line@entry=79, function=function@entry=0x44f0c0 state_for_frame) at assert.c:92 #8 0x7f94056f43d2 in __GI___assert_fail (assertion=assertion@entry=0x4537b2 0, file=file@entry=0x44eafa src/x.c, line=line@entry=79, function=function@entry=0x44f0c0 state_for_frame) at assert.c:101 #9 0x00423ef2 in state_for_frame (window=optimized out) at src/x.c:79 #10 state_for_frame (window=optimized out) at src/x.c:71 #11 0x0042523c in x_push_node (con=con@entry=0xc35490) at src/x.c:567 #12 0x00419f6c in drag_window_callback (con=0xc35490, old_rect=optimized out, new_x=optimized out, new_y=optimized out, extra=optimized out) at src/floating.c:368 #13 0x0041aa9d in drag_pointer (con=con@entry=0xc35490, event=event@entry=0xc2e830, confine_to=confine_to@entry=0, border=border@entry=BORDER_TOP, callback=callback@entry=0x419f40 drag_window_callback, extra=extra@entry=0xc2e830) at src/floating.c:571 #14 0x0041abf5 in floating_drag_window (con=con@entry=0xc35490, event=event@entry=0xc2e830) at src/floating.c:390 #15 0x00428bc7 in route_click (con=0xc34c00, event=0xc2e830, mod_pressed=true, dest=CLICK_INSIDE) at src/click.c:256 #16 0x00428def in handle_button_press (event=event@entry=0xc2e830) at src/click.c:325 #17 0x0042b020 in handle_event (type=optimized out, event=event@entry=0xc2e830) at src/handlers.c:1068 #18 0x0042bcac in xcb_check_cb (loop=optimized out, w=optimized out, revents=optimized out) at src/main.c:137 #19 0x7f9405ebd66f in ev_invoke_pending () from /usr/lib/libev.so.4 #20 0x7f9405ec06f5 in ev_run () from /usr/lib/libev.so.4 #21 0x0040894d in ev_loop (flags=0, loop=optimized out) at /usr/include/ev.h:820 #22 main (argc=optimized out, argv=optimized out) at src/main.c:836 Also see http://i3wm.org/docs/debugging.html for how to enable logging, that might be helpful in figuring out what is going on. Done, although I'd have to logout for it to take affect I don't think this had applied, since I hadn't logged out. Since I restarted with 'r' in the crash dialog window I expect it hasn't taken hold yet either. Ian. -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#700612: i3-wm: i3wm crashed corrupted double-linked list
On Sat, 2013-05-25 at 17:34 +0200, Michael Stapelberg wrote: Hi Ian, As you mentioned, without i3-wm-dbg installed, the backtrace is not going to help. Can you still reproduce that problem/does it still happen? It still happens occasionally but not regularly. I had an instance last week but I stupidly forgot to gather a backtrace. I'll see if I can provoke it with vncviewer next week when I'm back at my desk. Also see http://i3wm.org/docs/debugging.html for how to enable logging, that might be helpful in figuring out what is going on. Done, although I'd have to logout for it to take affect -- is there a way to enable at runtime? Thanks, Ian. signature.asc Description: This is a digitally signed message part
Bug#700612: i3-wm: i3wm crashed corrupted double-linked list
Hi Ian, Ian Campbell i...@hellion.org.uk writes: Done, although I'd have to logout for it to take affect -- is there a way to enable at runtime? No, not for the release versions. Debug versions (anything built from git) have shared memory logging enabled by default. -- Best regards, Michael -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#700612: i3-wm: i3wm crashed corrupted double-linked list
Hi Ian, As you mentioned, without i3-wm-dbg installed, the backtrace is not going to help. Can you still reproduce that problem/does it still happen? Also see http://i3wm.org/docs/debugging.html for how to enable logging, that might be helpful in figuring out what is going on. -- Best regards, Michael -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#700612: i3-wm: i3wm crashed corrupted double-linked list
Package: i3-wm Version: 4.2-1 Severity: important Dear Maintainer, i3-wm has crashed on me a couple of times over the last day. The most recent time I thought to look in .xsession-errors and found: *** glibc detected *** i3: corrupted double-linked list: 0x00c31530 *** === Backtrace: = /lib/x86_64-linux-gnu/libc.so.6(+0x76d76)[0x7f82384b7d76] /lib/x86_64-linux-gnu/libc.so.6(+0x7a6a4)[0x7f82384bb6a4] /lib/x86_64-linux-gnu/libc.so.6(__libc_malloc+0x70)[0x7f82384bcb90] /lib/x86_64-linux-gnu/libc.so.6(+0x2dff7)[0x7f823846eff7] /lib/x86_64-linux-gnu/libc.so.6(+0x2de9e)[0x7f823846ee9e] /lib/x86_64-linux-gnu/libc.so.6(+0x2d510)[0x7f823846e510] /lib/x86_64-linux-gnu/libc.so.6(__assert_fail+0x88)[0x7f823846c5b8] i3[0x42ba02] i3[0x42cd1c] i3[0x41a97c] i3[0x41b4e0] i3[0x41b645] i3[0x411b8b] i3[0x41cb60] i3[0x420d7c] /usr/lib/libev.so.4(ev_invoke_pending+0x8f)[0x7f8238c1566f] /usr/lib/libev.so.4(ev_run+0x835)[0x7f8238c186f5] i3[0x4088f3] /lib/x86_64-linux-gnu/libc.so.6(__libc_start_main+0xfd)[0x7f823845fead] i3[0x40901d] I'm unsure what exactly I did at the moment it crashed but I was starting vncviewer (on a remote Xen virtual machine) and pressing the hot key to make the window floating. When a Xen VM boots it changes mode which causes the vncviewer window to come and go a bit, which might be relevant? I was running 4.2-1 but I have upgraded to 4.2-2 (which doesn't look relevant) and installed i3-wm-dbg so that in the future the backtrace might be more useful. Thanks, Ian. -- System Information: Debian Release: 7.0 APT prefers testing APT policy: (990, 'testing'), (500, 'unstable'), (500, 'stable') Architecture: amd64 (x86_64) Kernel: Linux 3.2.0-4-amd64 (SMP w/2 CPU cores) Locale: LANG=en_GB.UTF-8, LC_CTYPE=en_GB.UTF-8 (charmap=UTF-8) Shell: /bin/sh linked to /bin/bash Versions of packages i3-wm depends on: ii libc6 2.13-37 ii libev41:4.11-1 ii libpcre3 1:8.30-5 ii libstartup-notification0 0.12-1 ii libx11-6 2:1.5.0-1 ii libxcb-icccm4 0.3.9-2 ii libxcb-keysyms1 0.3.9-1 ii libxcb-randr0 1.8.1-2 ii libxcb-util0 0.3.8-2 ii libxcb-xinerama0 1.8.1-2 ii libxcb1 1.8.1-2 ii libxcursor1 1:1.1.13-1 ii libyajl2 2.0.4-2 ii perl 5.14.2-16 ii x11-utils 7.7~1 Versions of packages i3-wm recommends: ii xfonts-base 1:1.0.3 Versions of packages i3-wm suggests: ii gnome-terminal [x-terminal-emulator] 3.4.1.1-2 ii xterm [x-terminal-emulator] 278-4 -- no debconf information -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
