Description of problem: gtk-gnutella immediately quits when started. Example: $ gtk-gnutella nm: /usr/bin/gtk-gnutella: no symbols 20-05-11 13:50:51.559 (WARNING): no symbols loaded, trying with pre-computed "/usr/share/gtk-gnutella/x86_64-linux/gtk-gnutella.nm" 20-05-12 01:50:51.575 (WARNING): file "/usr/share/gtk-gnutella/x86_64-linux/gtk-gnutella.nm" not holding symbols for "/usr/bin/gtk-gnutella" Segmentation fault (core dumped) Version-Release number of selected component: gtk-gnutella-1.1.15-3.fc32 Additional info: reporter: libreport-2.12.0 backtrace_rating: 4 cgroup: 0::/user.slice/user-1000.slice/session-2.scope cmdline: gtk-gnutella crash_function: allocate_stack executable: /usr/bin/gtk-gnutella journald_cursor: s=1adee758c59b43d086271441addd8887;i=9da244;b=352a19597b6e400e879b2bccfe97bff4;m=767da532;t=5a55fa0d9f8bf;x=3ae9f18b4e08f9ef kernel: 5.6.10-300.fc32.x86_64 rootdir: / runlevel: N 5 type: CCpp uid: 1000 xsession_errors: Truncated backtrace: Thread no. 1 (10 frames) #0 allocate_stack at allocatestack.c:466 #1 __pthread_create_2_1 at pthread_create.c:660 #2 thread_launch at thread.c:9685 #3 thread_create_full at thread.c:9810 #4 evq_init_once at evq.c:303 #5 once_flag_run_internal at once.c:319 #7 once_flag_run_trace at once.c:348 #8 evq_init at evq.c:324 #9 evq_raw_idle_add at evq.c:1159 #10 once_flag_run_internal at once.c:319
Created attachment 1687341 [details] File: backtrace
Created attachment 1687342 [details] File: core_backtrace
Created attachment 1687343 [details] File: cpuinfo
Created attachment 1687344 [details] File: dso_list
Created attachment 1687345 [details] File: environ
Created attachment 1687346 [details] File: exploitable
Created attachment 1687347 [details] File: limits
Created attachment 1687348 [details] File: maps
Created attachment 1687349 [details] File: mountinfo
Created attachment 1687350 [details] File: open_fds
Created attachment 1687351 [details] File: proc_pid_status
Is it always reproduceable? Could you check whether it is reproduceable on some another F32 host (another hardware/environment/configuration etc.)? Is it exist in F31? Could you try to install f31's package on your F32 system (if it is possible at all) to check whether it is package or system issue?
It crashed every time I tried it today. I tried gtk-gnutella on another PC which is running F32 (upgraded from F31). It also crashes on starting. That PC had not run gtk-gnutella previously. On the same PC I installed the gtk-gnutella package from F31 found on rpmfind.net. This version started and ran and appeared to work normally. I have previously been running it under F31 until upgrading Fedora to F32 in the last couple of days.
Thanks. Just to make sure -- if you still have an access to some F31 system -- is it possible to install F32 package on F31 and check it there?
To check for possible compler issues, I've rebuilt with Clang instead of GCC: https://koji.fedoraproject.org/koji/taskinfo?taskID=44423412 Could you please test the new 1.1.15-4.fc32 package: https://kojipkgs.fedoraproject.org//work/tasks/3412/44423412/gtk-gnutella-1.1.15-4.fc32.x86_64.rpm
Installing the F32 package on a F31 system also results in an immediate segfault in pthread. Using the new -4 package changes the behaviour. It does not crash immediately but has a different buggy behaviour. The program starts what it calls a supervisor process, then runs a "child" process. The child process stalls and never opens the window on the display. But if the child process is killed from another terminal a new child process runs normally. Here is output in the terminal window: $ gtk-gnutella nm: /usr/bin/gtk-gnutella: no symbols 20-05-13 13:30:37.338 (WARNING): no symbols loaded, trying with pre-computed "/usr/share/gtk-gnutella/x86_64-linux/gtk-gnutella.nm" 20-05-14 01:30:37.354 (WARNING): file "/usr/share/gtk-gnutella/x86_64-linux/gtk-gnutella.nm" not holding symbols for "/usr/bin/gtk-gnutella" 20-05-14 01:30:37.365 [52187] (WARNING): turning PID logging for supervisor process 20-05-14 01:30:37.365 [52187] (MESSAGE): use --log-supervise to redirect supervisor logs 20-05-14 01:30:37.365 [52187] (INFO): supervisor starting as PID 52187 20-05-14 01:30:37.365 [52187] (MESSAGE): walloc() size limit set to 312 20-05-14 01:30:37.365 [52187] (WARNING): halloc() could not be disabled 20-05-14 01:30:37.365 [52187] (INFO): callout queue will be running in thread #4 20-05-14 01:30:37.366 [52187] (INFO): program path is /usr/bin/gtk-gnutella 20-05-14 01:30:37.366 [52187] (INFO): will be launching: gtk-gnutella --child 20-05-14 01:30:37.366 [52187] (INFO): launched child #1 as PID 52193 nm: /usr/bin/gtk-gnutella: no symbols 20-05-13 13:30:37.378 (WARNING): no symbols loaded, trying with pre-computed "/usr/share/gtk-gnutella/x86_64-linux/gtk-gnutella.nm" 20-05-14 01:30:37.391 (WARNING): file "/usr/share/gtk-gnutella/x86_64-linux/gtk-gnutella.nm" not holding symbols for "/usr/bin/gtk-gnutella" 20-05-14 01:30:37.402 (INFO): gtk-gnutella/1.1.15 (2019-07-14; GTK2; Linux x86_64) 20-05-14 01:30:37.402 (INFO): using our own malloc() replacement 20-05-14 01:30:37.402 (INFO): running gtk-gnutella --child 20-05-14 01:30:37.402 (INFO): loaded 0 symbols for "/usr/bin/gtk-gnutella" via nm output parsing in 0.015 secs 0.011 secs ago 20-05-14 01:33:31.540 [52187] (MESSAGE): child #1 (PID 52193) killed by SIGTERM after 2m 54s 20-05-14 01:33:31.540 [52187] (INFO): launched child #2 as PID 52710 nm: /usr/bin/gtk-gnutella: no symbols 20-05-13 13:33:31.554 (WARNING): no symbols loaded, trying with pre-computed "/usr/share/gtk-gnutella/x86_64-linux/gtk-gnutella.nm" 20-05-14 01:33:31.567 (WARNING): file "/usr/share/gtk-gnutella/x86_64-linux/gtk-gnutella.nm" not holding symbols for "/usr/bin/gtk-gnutella" 20-05-14 01:33:31.579 (INFO): gtk-gnutella/1.1.15 (2019-07-14; GTK2; Linux x86_64) 20-05-14 01:33:31.580 (INFO): using our own malloc() replacement 20-05-14 01:33:31.580 (INFO): running gtk-gnutella --child 20-05-14 01:33:31.580 (INFO): loaded 0 symbols for "/usr/bin/gtk-gnutella" via nm output parsing in 0.016 secs 0.013 secs ago 20-05-14 01:33:33.645 (INFO): INPUTEVT using customized I/O dispatching with epoll() 20-05-14 01:33:33.647 (INFO): language code: "en" 20-05-14 01:33:33.647 (INFO): using locale character set "UTF-8" 20-05-14 01:33:33.647 (INFO): primary filename character set "UTF-8" 20-05-14 01:33:33.647 (DEBUG-3): ADNS thread started 20-05-14 01:33:33.648 (MESSAGE): D-BUS set up and ready for use. 20-05-14 01:33:33.655 (INFO): malloc() allocated 17488 bytes of heap (17488 remain) 20-05-14 01:33:35.258 (INFO): TLS loaded X.509 certificate: subject `CN=gtk-gnutella/1.1.9', issuer `CN=gtk-gnutella/1.1.9', serial 0xdea811570003ec52, RSA key 3072 bits, signed using RSA-SHA256, activated `2016-04-16 02:52:14 UTC', expires `2037-12-31 23:00:00 UTC', pin-sha256="wXnr+A3+W5IqIYJ82PXUYUAOPcZzxzfZ3bzjESQc8uA=" 20-05-14 01:33:35.416 (WARNING): geo-ip.txt, line 10962: bad country code in "5.206.232.0/22 xk" 20-05-14 01:33:35.417 (WARNING): geo-ip.txt, line 30386: bad country code in "37.35.12.0/23 xk" 20-05-14 01:33:35.422 (WARNING): geo-ip.txt, line 77634: bad country code in "77.46.224.0/22 xk" 20-05-14 01:33:35.422 (WARNING): geo-ip.txt, line 79244: bad country code in "77.247.248.0/24 xk" 20-05-14 01:33:35.422 (WARNING): geo-ip.txt, line 79245: bad country code in "77.247.249.0/24 xk" 20-05-14 01:33:35.422 (WARNING): geo-ip.txt, line 79246: bad country code in "77.247.250.0/23 xk" 20-05-14 01:33:35.422 (WARNING): geo-ip.txt, line 79247: bad country code in "77.247.252.0/23 xk" 20-05-14 01:33:35.422 (WARNING): geo-ip.txt, line 79248: bad country code in "77.247.254.0/24 xk" 20-05-14 01:33:35.422 (WARNING): geo-ip.txt, line 79249: bad country code in "77.247.255.0/24 xk" 20-05-14 01:33:35.422 (WARNING): geo-ip.txt, line 80361: bad country code in "79.101.180.0/24 xk" 20-05-14 01:33:35.423 (WARNING): geo-ip.txt, line 88269: bad country code in "84.22.45.0/24 xk" 20-05-14 01:33:35.423 (WARNING): geo-ip.txt, line 88271: bad country code in "84.22.48.0/21 xk" 20-05-14 01:33:35.423 (WARNING): geo-ip.txt, line 88279: bad country code in "84.22.62.248/32 xk" 20-05-14 01:33:35.425 (WARNING): geo-ip.txt, line 104709: bad country code in "91.240.40.0/22 xk" 20-05-14 01:33:35.425 (WARNING): geo-ip.txt, line 108195: bad country code in "93.87.164.0/22 xk" 20-05-14 01:33:35.425 (WARNING): geo-ip.txt, line 108200: bad country code in "93.87.228.0/24 xk" 20-05-14 01:33:35.434 (WARNING): geo-ip.txt, line 199118: bad country code in "178.132.216.66/32 xk" 20-05-14 01:33:35.434 (WARNING): geo-ip.txt, line 199728: bad country code in "178.221.36.0/22 xk" 20-05-14 01:33:35.434 (WARNING): geo-ip.txt, line 199731: bad country code in "178.221.56.0/22 xk" 20-05-14 01:33:35.434 (WARNING): geo-ip.txt, line 199734: bad country code in "178.221.80.0/22 xk" 20-05-14 01:33:35.434 (WARNING): geo-ip.txt, line 199735: bad country code in "178.221.84.0/23 xk" 20-05-14 01:33:35.434 (WARNING): geo-ip.txt, line 199738: bad country code in "178.221.92.0/22 xk" 20-05-14 01:33:35.434 (WARNING): geo-ip.txt, line 199746: bad country code in "178.223.71.0/24 xk" 20-05-14 01:33:35.434 (WARNING): geo-ip.txt, line 199752: bad country code in "178.223.224.0/20 xk" 20-05-14 01:33:35.434 (WARNING): geo-ip.txt, line 199753: bad country code in "178.223.240.0/21 xk" 20-05-14 01:33:35.435 (WARNING): geo-ip.txt, line 209997: bad country code in "185.75.56.0/24 xk" 20-05-14 01:33:35.435 (WARNING): geo-ip.txt, line 210000: bad country code in "185.75.59.0/24 xk" 20-05-14 01:33:35.435 (WARNING): geo-ip.txt, line 210580: bad country code in "185.82.110.0/24 xk" 20-05-14 01:33:35.436 (WARNING): geo-ip.txt, line 226894: bad country code in "188.246.55.0/24 xk" 20-05-14 01:33:35.440 (WARNING): geo-ip.txt, line 269056: bad country code in "195.252.80.0/21 xk" 20-05-14 01:33:35.445 (WARNING): geo-ip.txt, line 318492: bad country code in "212.200.33.0/24 xk" 20-05-14 01:33:35.453 (DEBUG): loaded 330525 geographical IPv4 ranges (2900521315 hosts) from "/usr/share/gtk-gnutella/geo-ip.txt" 20-05-14 01:33:35.469 (DEBUG): loaded 95090 geographical IPv6 ranges from "/usr/share/gtk-gnutella/geo-ipv6.txt" 20-05-14 01:33:50.223 (WARNING-1): thread_surely_gone(): seen some activity for thread #6, not reusing its ID 20-05-14 01:33:54.562 (MESSAGE): more recent development version of gtk-gnutella: 1.1.15-114-gab450 (2019-07-14) Interestingly there is a switch --no-supervise which skips the supervisor process. As the segfault as reported is very early it may be in the supervisor portion of the program so I may test this switch with the original -3 package.
OK, no joy using the --no-supervise switch with -3 package. $ gtk-gnutella --no-supervise nm: /usr/bin/gtk-gnutella: no symbols 20-05-13 13:53:22.747 (WARNING): no symbols loaded, trying with pre-computed "/usr/share/gtk-gnutella/x86_64-linux/gtk-gnutella.nm" 20-05-14 01:53:22.764 (WARNING): file "/usr/share/gtk-gnutella/x86_64-linux/gtk-gnutella.nm" not holding symbols for "/usr/bin/gtk-gnutella" Segmentation fault (core dumped)
I've rebuilt with the latest upstream source (pre 1.1.16) and still with GCC: https://koji.fedoraproject.org/koji/taskinfo?taskID=44456894 https://kojipkgs.fedoraproject.org//work/tasks/6894/44456894/gtk-gnutella-1.1.15-5.fc32.x86_64.rpm Could you please check this "-5" package, to make sure things is not changed in the latest development?..
Hi. The -5 package crashes immediately. $ gtk-gnutella Segmentation fault (core dumped) $ gtk-gnutella --nosupervise Segmentation fault (core dumped)
Thanks. Reported upstream for more info: https://sourceforge.net/p/gtk-gnutella/bugs/558/
Rebuilt again with system malloc implementation: https://koji.fedoraproject.org/koji/taskinfo?taskID=44495458 https://kojipkgs.fedoraproject.org//work/tasks/5458/44495458/gtk-gnutella-1.1.15-6.fc32.x86_64.rpm It seems "-6" crashes as well, but please try it in your environment too...
Firget about "-6" and try "-7" instead :) . It seems no more crashes, but please test whether things work as expected. https://koji.fedoraproject.org/koji/taskinfo?taskID=44550998 https://kojipkgs.fedoraproject.org//work/tasks/998/44550998/gtk-gnutella-1.1.15-7.fc32.x86_64.rpm
Hi. The -7 package seems to be good. I have run it for several hours and also started and stopped it a few times so it is running at least as well as it did on F31. In fact it is better. There used to be a random crash while using the search function which hasn't happened with this new package. I didn't report the random crash as the program was still mostly usable and random crashes are hard to find. I expected the developers to find it one day. Thanks for the time and effort you have used for finding this bug. Much better than on some reports I have made in the past.
Hi. Bad news sorry. I saw the gui window was gone and the gtk-gnutella processes were still running. Then killing the processes and attempts to restart result in a crash. STDERR output had lots of information and this part looks relevant: 20-05-17 22:13:27.460 (INFO): malloc() allocated 17488 bytes of heap (17488 remain) 20-05-17 22:13:27.834 (INFO): TLS loaded X.509 certificate: subject `CN=gtk-gnutella/1.1.9', issuer `CN=gtk-gnutella/1.1.9', serial 0xdea811570003ec52, RSA key 3072 bits, signed using RSA-SHA256, activated `2016-04-16 02:52:14 UTC', expires `2037-12-31 23:00:00 UTC', pin-sha256="wXnr+A3+W5IqIYJ82PXUYUAOPcZzxzfZ3bzjESQc8uA=" 20-05-17 22:13:28.032 (DEBUG): loaded 330556 geographical IPv4 ranges (3193074021 hosts) from "/usr/share/gtk-gnutella/geo-ip.txt" 20-05-17 22:13:28.051 (DEBUG): loaded 95090 geographical IPv6 ranges from "/usr/share/gtk-gnutella/geo-ipv6.txt" 20-05-17 22:13:28.168 (WARNING): got SIGSEGV for VA=0x0 at PC=0x7f31d28d4854 from /lib64/librsvg-2.so.2 20-05-17 22:13:28.168 (DEBUG): crash_mode(): initial external call, level=basic from thread "main" (#0) 20-05-17 22:13:28.168 (DEBUG): crash_mode(): depth=0, external=y, level=basic, old=none, new=basic 20-05-17 22:13:28.168 (WARNING): filelock_clean(): unlocking /home/richard/.gtk-gnutella/gtk-gnutella.pid 20-05-17 22:13:28.168 (WARNING): filelock_clean(): unlocking /home/richard/gtk-gnutella-downloads/incomplete/.gtk-gnutella.lock 20-05-17 22:13:28.168 (DEBUG): crash_mode(): subsequent internal call, level=exception 20-05-17 22:13:28.168 (DEBUG): crash_mode(): depth=1, external=n, level=exception, old=basic, new=exception 20-05-17 22:13:28.168 (WARNING): disabling locks, PID 4873 now in thread-unsafe mode (11 threads) 20-05-17 22:13:28.168 (DEBUG): crash_mode(): depth=2, external=y, level=basic, old=exception, new=exception 20-05-17 22:13:28.168 (DEBUG): crash_mode(): depth=3, external=y, level=basic, old=exception, new=exception 20-05-17 22:13:28.168 CRASH (pid=4873) for gtk-gnutella/1.1.15 (2019-07-14; GTK2; Linux x86_64) 20-05-17 22:13:28.168 CRASH (pid=4873) by SIGSEGV after 3s -- stack was: 0x44c315 0x6d57ab 0x6d5a4c 0x7f31dfecda90 : libpthread.so.0 0x7f31d28d4854 : librsvg-2.so.2 0x7f31d28d4e38 : librsvg-2.so.2 0x7f31d28dd9a4 : librsvg-2.so.2 0x7f31d28ce587 : librsvg-2.so.2 0x7f31d28ce2a3 : librsvg-2.so.2 0x7f31d28d04a6 : librsvg-2.so.2 0x7f31d28ce2a3 : librsvg-2.so.2 0x7f31d28ceba1 : librsvg-2.so.2 0x7f31d28cc518 : librsvg-2.so.2 0x7f31d289e86b : librsvg-2.so.2 0x7f31d28fec07 : librsvg-2.so.2 0x7f31d28f572b : librsvg-2.so.2 0x7f31d288da85 : librsvg-2.so.2 0x7f31d2af9451 : librsvg-2.so.2 0x7f31d2887d75 : librsvg-2.so.2 0x7f31d272a007 : librsvg-2.so.2 0x7f31d26b439e : librsvg-2.so.2 <rsvg_rust_handle_close+294> : librsvg-2.so.2 0x7f31d2eb5466 : libpixbufloader-svg.so <gdk_pixbuf_loader_close+235> : libgdk_pixbuf-2.0.so.0 0x7f31e0545c03 : libgdk_pixbuf-2.0.so.0 <gdk_pixbuf_new_from_stream_at_scale+127> : libgdk_pixbuf-2.0.so.0 0x7f31e077c182 : libgtk-x11-2.0.so.0 <gtk_icon_info_load_icon+120> : libgtk-x11-2.0.so.0 <gtk_icon_theme_load_icon+186> : libgtk-x11-2.0.so.0 <gtk_icon_set_render_icon+1620> : libgtk-x11-2.0.so.0 <gtk_widget_render_icon+158> : libgtk-x11-2.0.so.0 0x7f31e07914f4 : libgtk-x11-2.0.so.0 0x7f31e0791529 : libgtk-x11-2.0.so.0 <g_closure_invoke+184> : libgobject-2.0.so.0 0x7f31e04b60fa : libgobject-2.0.so.0 <g_signal_emit_valist+4009> : libgobject-2.0.so.0 <g_signal_emit_by_name+527> : libgobject-2.0.so.0 0x7f31e0832f10 : libgtk-x11-2.0.so.0 0x7f31e06eaf5d : libgtk-x11-2.0.so.0 <g_closure_invoke+184> : libgobject-2.0.so.0 0x7f31e04b60fa : libgobject-2.0.so.0 I did change the Icon set a little while ago, so I have just realized that there may be problems handling the Mint-Y-Aqua icons. Changing them back to Adwaita theme icons and the GUI window now runs. I am pleased that I have figured out why it was newly crashing, but there may be a problem with gtk2 libraries loading modern icon themes.
(In reply to Richard Haakma from comment #24) > I saw the gui window was gone and the gtk-gnutella > processes were still running. Probably some gui-related thread is crashed. It is likely that systemd in your F32 system stores coredumps in its journals. If so, try to find the correspond coredump by "coredumpctl" command and provide backtrace etc. here. > Then killing the processes and attempts to restart result in a crash. If it was reproduceable and happens at startup time, it was a good candidate to work with (probably coredumpctl can show such crashes as well). If it is still reproduceable... :) Anyway, it seems that this crash is far deep in the system libraries (librsvg2 under gtk2 etc.) and probably is not related to gtk-gnutella. GTK2 is obsoleting now, people no longer pay much attention to this. So some bugs on an untypical environment is not surprising.
FEDORA-2020-39456d65bc has been submitted as an update to Fedora 32. https://bodhi.fedoraproject.org/updates/FEDORA-2020-39456d65bc
FEDORA-2020-39456d65bc has been pushed to the Fedora 32 testing repository. In short time you'll be able to install the update with the following command: `sudo dnf upgrade --enablerepo=updates-testing --advisory=FEDORA-2020-39456d65bc` You can provide feedback for this update here: https://bodhi.fedoraproject.org/updates/FEDORA-2020-39456d65bc See also https://fedoraproject.org/wiki/QA:Updates_Testing for more information on how to test updates.
FEDORA-2020-39456d65bc has been pushed to the Fedora 32 stable repository. If problem still persists, please make note of it in this bug report.
Richard, I would like to ask you to try to reproduce the last librsvg2-related bug (when gui window was gone, comment #24 ). We have several abort reports about something similar, and upstream has report from Ubuntu as well. IOW, it is a bug which must be fixed, but it seems to be depended by a specific user environment (fe. by non-default/broken icon theme, as probably in your case). If you have a time, please, try to run gtk-gnutella in an environment where those crash had happen, in a hope you then catch the same (or something similar): > I saw the gui window was gone and the gtk-gnutella processes were still running. Then killing the processes and attempts to restart result in a crash. An "attempts to restart result in a crash" would be 1/2 of success. To reproduce the bug by someone else, the crash environment should be provided (ie. all the config files, rpm lists etc.etc.etc.). Since nobody wants to get all the user home directory :) ,it could be fine if you try to reproduce the crash on your system with a newly created user, just copying to its homedir all the appropriate files (ie. ~/.gtk-gnutella, most likely gtk2-related stuff as well, etc.etc.etc.). Then try to minimize the amount of the files, and the result would be a test environment in a form of an archive of user's home dir. If the archive will not contain any private or sencitive data, it can be send for further investigation, fe. by attaching in this bugreport or emailed directly. It seems that only you for now can reproduce the librsvg2-related bug and the same time can provide some useful info about it. Thanks for your time anyway.
Librsvg2 issue is resolved, see https://sourceforge.net/p/gtk-gnutella/bugs/561/ fro more info. No more any reproducion attempts requied. Anyway, thanks for your participation.
FEDORA-2020-25df3af5ee has been submitted as an update to Fedora 32. https://bodhi.fedoraproject.org/updates/FEDORA-2020-25df3af5ee
FEDORA-2020-25df3af5ee has been pushed to the Fedora 32 testing repository. In short time you'll be able to install the update with the following command: `sudo dnf upgrade --enablerepo=updates-testing --advisory=FEDORA-2020-25df3af5ee` You can provide feedback for this update here: https://bodhi.fedoraproject.org/updates/FEDORA-2020-25df3af5ee See also https://fedoraproject.org/wiki/QA:Updates_Testing for more information on how to test updates.
FEDORA-2020-25df3af5ee has been pushed to the Fedora 32 stable repository. If problem still persists, please make note of it in this bug report.