Re: [Help-gnunet] gnunetd deadlocked

[David Kuehling]

>>>>> "Christian" == Christian Grothoff <address@hidden> writes:

> Thanks for reporting. As it happens, that one was fixed in SVN on June
> 23rd.  Christian

Really ? :) SVN version of today (June 2) just deadlocked again.  Same
symptoms.  0% CPU and gnunet-stats hangs.

There are some new warning messages output by gnunet-download:

Jul 02 20:17:15 WARNING: Lock aquired for too long (1453 ms) at download.c:1249
Jul 02 20:20:37 WARNING: Lock aquired for too long (1910 ms) at tcpio.c:400
Jul 02 20:20:37 WARNING: Lock aquired for too long (1910 ms) at download.c:1249
Jul 02 20:20:55 WARNING: Lock aquired for too long (2787 ms) at tcpio.c:400
Jul 02 20:20:55 WARNING: Lock aquired for too long (2788 ms) at download.c:1249
Jul 02 20:21:54 WARNING: Lock aquired for too long (2198 ms) at tcpio.c:400
Jul 02 20:21:54 WARNING: Lock aquired for too long (2200 ms) at download.c:1249
Jul 02 20:25:56 WARNING: Lock aquired for too long (4072 ms) at tcpio.c:400
Jul 02 20:25:56 WARNING: Lock aquired for too long (4073 ms) at download.c:1249
Jul 02 20:28:54 WARNING: Lock aquired for too long (1690 ms) at tcpio.c:400
Jul 02 20:28:54 WARNING: Lock aquired for too long (1692 ms) at download.c:1249
Jul 02 20:33:22 WARNING: Lock aquired for too long (1613 ms) at tcpio.c:400
Jul 02 20:33:22 WARNING: Lock aquired for too long (1615 ms) at download.c:1249
Jul 02 20:39:42 WARNING: Lock aquired for too long (12001 ms) at tcpio.c:400
Jul 02 20:39:42 WARNING: Lock aquired for too long (12001 ms) at 
download.c:1249Jul 02 20:39:49 WARNING: Lock aquired for too long (1221 ms) at 
tcpio.c:400
Jul 02 20:39:49 WARNING: Lock aquired for too long (1222 ms) at download.c:1249
Jul 02 20:41:03 WARNING: Lock aquired for too long (5256 ms) at tcpio.c:400
Jul 02 20:41:03 WARNING: Lock aquired for too long (5257 ms) at download.c:1249

Is this expected?

The information from GDB is as follows:

info threads:
  14 Thread -1213138000 (LWP 21986)  0xb7e69189 in __lll_mutex_lock_wait ()
   from /lib/tls/libpthread.so.0
  13 Thread -1225733200 (LWP 21987)  0xb7e69b3c in __nanosleep_nocancel ()
   from /lib/tls/libpthread.so.0
  12 Thread -1242563664 (LWP 21989)  0xb7e69189 in __lll_mutex_lock_wait ()
   from /lib/tls/libpthread.so.0
  11 Thread -1250952272 (LWP 21990)  0xb7df16a7 in select ()
   from /lib/tls/libc.so.6
  10 Thread -1259385936 (LWP 21991)  0xb7e69b3c in __nanosleep_nocancel ()
   from /lib/tls/libpthread.so.0
  9 Thread -1269867600 (LWP 21992)  0xb7e66c01 in 
pthread_cond_wait@@GLIBC_2.3.2 () from /lib/tls/libpthread.so.0
  8 Thread -1278256208 (LWP 21993)  0xb7e69b3c in __nanosleep_nocancel ()
   from /lib/tls/libpthread.so.0
  7 Thread -1293534288 (LWP 21994)  0xb7e69b3c in __nanosleep_nocancel ()
   from /lib/tls/libpthread.so.0
  6 Thread -1310745680 (LWP 21997)  0xb7e69189 in __lll_mutex_lock_wait ()
   from /lib/tls/libpthread.so.0
  5 Thread -1302357072 (LWP 21998)  0xb7e66c01 in 
pthread_cond_wait@@GLIBC_2.3.2 () from /lib/tls/libpthread.so.0
  4 Thread -1319134288 (LWP 21999)  0xb7e69189 in __lll_mutex_lock_wait ()
   from /lib/tls/libpthread.so.0
  3 Thread -1327522896 (LWP 22000)  0xb7e69189 in __lll_mutex_lock_wait ()
   from /lib/tls/libpthread.so.0
  2 Thread -1335911504 (LWP 22001)  0xb7e69189 in __lll_mutex_lock_wait ()
   from /lib/tls/libpthread.so.0
  1 Thread -1212033344 (LWP 21985)  0xb7e66c01 in 
pthread_cond_wait@@GLIBC_2.3.2 () from /lib/tls/libpthread.so.0

Gdb log (backtraces for all threads) attached.  I compiled gnunet with
-Os -fomit-frame-pointer, maybe that's the reason that backtraces aren't
very accurate this time?  I'll try to do better next time.

David
-- 
GnuPG public key: http://user.cs.tu-berlin.de/~dvdkhlng/dk.gpg
Fingerprint: B17A DC95 D293 657B 4205  D016 7DEF 5323 C174 7D40

gdb.log
Description: Binary data