Random Crashes with "double free or corruption (!prev)" and other error messages #923
Comments
|
It just happened again, ending with a different error: |
lasers
added
bug
|
Again... so, I thought I can maybe reproduce this by stressing the CPU and/or memory. The conky had already been up and running for a few hours when I tried this. An Archlinux forum search for "conky AND crash" does not provide any insight. |
|
And one more time. Should I compile conky with debug symbols so I can run it with gdb? I have no other problems with my system whatsoever. |
Yeah, why not? Go for it, sport. 👍 We need more collaborators, testers, volunteers, etc. |
ohnonot
changed the title
|
OK, I compiled the latest release with "Debug" option. $ gdb conky
$ run -c /path/to/conkyrc
...
reproduce crash by stressing the memory
...
free(): double free detected in tcache 2
Thread 1 "conky" received signal SIGABRT, Aborted.
0x00007ffff77b0f25 in raise () from /usr/lib/libc.so.6
(gdb) bt full
#0 0x00007ffff77b0f25 in raise () from /usr/lib/libc.so.6
No symbol table info available.
#1 0x00007ffff779a897 in abort () from /usr/lib/libc.so.6
No symbol table info available.
#2 0x00007ffff77f4258 in __libc_message () from /usr/lib/libc.so.6
No symbol table info available.
#3 0x00007ffff77fb77a in malloc_printerr () from /usr/lib/libc.so.6
No symbol table info available.
#4 0x00007ffff77fd59d in _int_free () from /usr/lib/libc.so.6
No symbol table info available.
#5 0x00007ffff77ff805 in _int_realloc () from /usr/lib/libc.so.6
No symbol table info available.
#6 0x00007ffff7800996 in realloc () from /usr/lib/libc.so.6
No symbol table info available.
#7 0x000055555559e2d9 in new_graph (obj=0x5555557c0780, buf=0x5555557c7a50 "\001", buf_max_size=15872, val=0) at /home/data/git/conky-1.11.5/conky-1.11.5/src/specials.cc:559
graph = 0x7fffffffda20
s = 0x555555822640
g = 0x5555557bb060
graph_id = 93
#8 0x000055555556e991 in generate_text_internal (p=0x5555557c7a50 "\001", p_max_size=15872, root=...) at /home/data/git/conky-1.11.5/conky-1.11.5/src/conky.cc:703
obj = 0x5555557c0780
a = 1
#9 0x000055555556ec37 in generate_text () at /home/data/git/conky-1.11.5/conky-1.11.5/src/conky.cc:762
p = 0x5555557c7850 "\001\001CPU\001\001\001\001stress \001 3.25%\n\001\001stress \001 3.12%\n\001\001stress \001 3.12%\n\001\001stress \001 3.12%\n\001\001\001\001\001\001\001\n\001\001\001Core 0\001\064\070°C\n\001\001\001\061\060\060%\001\063.70GHz\n\001\001\061\060\060%\001\063.70GHz\n\001\001\001\001\001\001\n\001\001\001\001Core 1\001\065\061°C\n\001\001\001\061\060\060"...
i = 21845
j = 4294957856
k = 32767
mw = 4152802959
tbs = 32767
ui = 0
time = 0
#10 0x000055555557238a in update_text () at /home/data/git/conky-1.11.5/conky-1.11.5/src/conky.cc:1811
No locals.
#11 0x0000555555572962 in main_loop () at /home/data/git/conky-1.11.5/conky-1.11.5/src/conky.cc:1912
fdsr = {fds_bits = {0 <repeats 16 times>}}
tv = {tv_sec = 0, tv_usec = 0}
s = 0
terminate = 0
t = 1.6068912189998628
inotify_config_wd = 1
inotify_buff = '\000' <repeats 104 times>, "StVUUU", '\000' <repeats 34 times>, "\360\335\377\377\377\177\000\000\000\000\000\000\000\000\b@\300~yUUU\000\000\340\246cUUU\000\000\020\336\377\377\377\177\000\000\300~yUUU\000\000\300~yUUU\000\000\000\000\000\000"...
#12 0x0000555555565472 in main (argc=3, argv=0x7fffffffe138) at /home/data/git/conky-1.11.5/conky-1.11.5/src/main.cc:354
No locals.
(gdb) I hope we can get to the bottom of this. |
|
@su8 (?) |
|
I'm hitting exactly the same issue (ubuntu 19.10 + XFCE4/MATE/Cinnamon) using latest conky (github source or github relese). More info:
|
|
I did some wget / cmake / run iterations. And so far I believe that the bug was introduced between 1.11.3 and 1.11.4/5. [Update]: Although the crash frequency is lower in 1.11.4 than in 1.11.5 but is not zero. So far moving down to 1.11.3 passed all stress tests. @ohnonot - can You test 1.11.3 to see if You are seeing this bug? |
|
Setting I will compile 1.11.3 with default values now, and use it for a few days. We will see. |
|
1.11.3 has been running super stable on my desktop for +2 weeks now. Haven't yet tested on my laptop, but since that one is running an almost identical setup, I am confident tha this helps. With this information, is it possible to find the issue & fix it for newer versions? |
|
The 1.11.3 release has been working reliably on 4 installs across 2 machines - all of them Archlinux. The current release is still 1.11.5 and Arch is still on 1.11.5. The current release appears to be stable with simpler config files; I realize the next step would be to hunt down the culprit in my larger config file (see first post). This could take ages due to the "randomness" of the issue. Maybe somebody already did part of that and has a hint where the problem might be? I have now also built & installed conky with this PKGBUILD - that's a different download link and a different build method. It, too, seems perfectly stable. |
|
This issue is stale because it has been open 365 days with no activity. Remove stale label or comment, or this issue will be closed in 30 days. |
github-actions
bot
added
the
Stale
|
This issue was closed because it has been stalled for 30 days with no activity. |
Issue
A conky with graphs disappears from the screen, as well as its process.
I have had this problem for a long time, and checked the issues here several times.
At first it seemed to be sth like #855 or #866 - but for almost half a year Archlinux has been using the release that includes the supposed fix, yet one of my conkys (the one with cpu graphs) crashes - rarely, not every day, but it does happen evtl., reliably.
I'm not aware of conky using up too much memory before that.
Information
Archlinux with LTS kernel (currently 4.19.90-1-lts), conky --version: 1.11.5_pre compiled Thu 29 Aug 2019 (from [extra])
Intel desktop, Intel GPU HD Graphics 530
I have not tried with gdb - it said that this version of conky is compiled without it.
I started the conky like this:
Today it finally crashed, and this was at the end of the log:
The text was updated successfully, but these errors were encountered: