tags 707758 + confirmed block 707758 376463 block 714499 376463 thanks On Fri, May 10, 2013 at 07:32:12PM -0400, Antoine Beaupré wrote: > You need to successfully authenticate against a server, then /load > otr, then /quit. > > Here's a gdb backtrace, notice how the debugging symbols don't seem to load > properly: > > warning: > "/usr/lib/debug/.build-id/1f/7c15953b6cc4fd3351a760e909e55291de093a.debug": > separate debug info file has no debug info > > warning: > "/usr/lib/debug/.build-id/34/f0ef4dae462d0c77467e0b7a70418967b2f6d2.debug": > separate debug info file has no debug info > > warning: > "/usr/lib/debug/.build-id/83/71de0e29582dd873bc1942962ce76a64ac3f71.debug": > separate debug info file has no debug info > Core was generated by `irssi --config=.irssi/config.proxy'. > Program terminated with signal 6, Aborted. > #0 0x00007f2b58758475 in *__GI_raise (sig=<optimized out>) at > ../nptl/sysdeps/unix/sysv/linux/raise.c:64 > 64 ../nptl/sysdeps/unix/sysv/linux/raise.c: Aucun fichier ou dossier de > ce type. > (gdb) bt > #0 0x00007f2b58758475 in *__GI_raise (sig=<optimized out>) at > ../nptl/sysdeps/unix/sysv/linux/raise.c:64 > #1 0x00007f2b5875b6f0 in *__GI_abort () at abort.c:92 > #2 0x00007f2b5879352b in __libc_message (do_abort=<optimized out>, > fmt=<optimized out>) at ../sysdeps/unix/sysv/linux/libc_fatal.c:189 > #3 0x00007f2b5879cd76 in malloc_printerr (action=3, str=0x7f2b58871b9c > "free(): invalid pointer", ptr=<optimized out>) at malloc.c:6283 > #4 0x00007f2b587a1aac in *__GI___libc_free (mem=<optimized out>) at > malloc.c:3738 > #5 0x00007f2b568d2a9e in ?? () from /lib/x86_64-linux-gnu/libgcrypt.so.11 > #6 0x00007f2b56d81064 in ?? () from /usr/lib/x86_64-linux-gnu/libgnutls.so.26 > #7 0x00007f2b56d8b615 in ?? () from /usr/lib/x86_64-linux-gnu/libgnutls.so.26 > #8 0x00007f2b56d93b74 in gnutls_deinit () from > /usr/lib/x86_64-linux-gnu/libgnutls.so.26 > #9 0x00007f2b57671139 in _lm_ssl_close (ssl=0x1220620) at lm-ssl-gnutls.c:300 > #10 0x00007f2b5766be78 in connection_do_close (connection=0x1225a40) at > lm-connection.c:514 > #11 connection_do_close (connection=0x1225a40) at lm-connection.c:494 > #12 0x00007f2b5766db32 in lm_connection_close (connection=0x1225a40, > error=0x0) at lm-connection.c:1375 > #13 0x00007f2b578863bd in server_cleanup () from > /usr/lib/irssi/modules/libxmpp_core.so > #14 0x000000000048e0ba in ?? () > #15 0x000000000048e56d in signal_emit () > #16 0x0000000000488ed0 in server_disconnect () > #17 0x0000000000490645 in ?? () > #18 0x0000000000490c84 in ?? () > #19 0x000000000048e0ba in ?? () > #20 0x000000000048e56d in signal_emit () > #21 0x000000000047a36e in ?? () > #22 0x000000000048e0ba in ?? () > #23 0x000000000048e56d in signal_emit () > #24 0x000000000041c37b in ?? () > #25 0x000000000048e0ba in ?? () > #26 0x000000000048e56d in signal_emit () > #27 0x000000000044e9e7 in ?? () > #28 0x000000000048e0ba in ?? () > #29 0x000000000048e56d in signal_emit () > #30 0x000000000044f3b3 in key_pressed () > #31 0x000000000041bc2e in ?? () > #32 0x000000000048e0ba in ?? () > #33 0x000000000048e56d in signal_emit () > #34 0x000000000041d05e in ?? () > #35 0x0000000000480f99 in ?? () > #36 0x00007f2b59588355 in g_main_context_dispatch () from > /lib/x86_64-linux-gnu/libglib-2.0.so.0 > #37 0x00007f2b59588688 in ?? () from /lib/x86_64-linux-gnu/libglib-2.0.so.0 > #38 0x00007f2b59588744 in g_main_context_iteration () from > /lib/x86_64-linux-gnu/libglib-2.0.so.0 > #39 0x000000000041905c in main ()
In my reading, this is just a regular shutdown of an xmpp connection over ssl, which works in the normal case but leads to a segfault when the otr module (irssi-plugin-otr) is loaded as well. With a few more debugging symbols, the first few lines of the backtrace look like this: #0 0x00007f825cec11e5 in __GI_raise (sig=sig@entry=6) at ../nptl/sysdeps/unix/sysv/linux/raise.c:56 #1 0x00007f825cec4398 in __GI_abort () at abort.c:90 #2 0x00007f825cefc7cb in __libc_message (do_abort=do_abort@entry=2, fmt=fmt@entry=0x7f825cff8be8 "*** Error in `%s': %s: 0x%s ***\n") at ../sysdeps/unix/sysv/linux/libc_fatal.c:199 #3 0x00007f825cf06a26 in malloc_printerr (action=3, str=0x7f825cff4b8a "free(): invalid pointer", ptr=<optimized out>) at malloc.c:4902 #4 0x00007f825cf077a3 in _int_free (av=<optimized out>, p=0x25eb5c8, have_lock=0) at malloc.c:3758 #5 0x00007f825b666ea6 in _gcry_free (p=0x25eb5e0) at global.c:879 #6 0x00007f825b904bf4 in _gnutls_cipher_deinit (handle=handle@entry=0x25e1208) at gnutls_cipher_int.c:147 #7 0x00007f825b90f1b5 in free_record_state (read=1, state=0x25e11d8) at gnutls_constate.c:786 #8 _gnutls_epoch_free (session=<optimized out>, params=0x25e11c0) at gnutls_constate.c:797 #9 0x00007f825b917704 in gnutls_deinit (session=0x25e3a20) at gnutls_state.c:399 #10 0x00007f825c40d139 in _lm_ssl_close (ssl=0x22e7ec0) at lm-ssl-gnutls.c:300 #11 0x00007f825c407e78 in connection_do_close (connection=0x22e8b30) at lm-connection.c:514 #12 connection_do_close (connection=0x22e8b30) at lm-connection.c:494 #13 0x00007f825c409b32 in lm_connection_close (connection=0x22e8b30, error=error@entry=0x0) at lm-connection.c:1375 #14 0x00007f825c621ffb in server_cleanup (server=0x22ed670) at xmpp-servers.c:102 #15 0x00000000004879e2 in ?? () #16 0x0000000000487e4d in signal_emit () #17 0x0000000000482a9e in server_disconnect () When running under valgrind, the valgrind log shows libotr doing a few invalid reads/writes just there, so I thought maybe it hooks into gnutls/gcrypt and wrecks havock there: ==10086== Invalid read of size 8 ==10086== at 0xDD5BCE5: ??? (in /usr/lib/libotr.so.5.0.0) ==10086== by 0x8C1FEA5: _gcry_free (in /lib/x86_64-linux-gnu/libgcrypt.so.11.8.2) ==10086== by 0x897FBF3: _gnutls_cipher_deinit (gnutls_cipher_int.c:147) ==10086== by 0x898A1B4: _gnutls_epoch_free (gnutls_constate.c:786) ==10086== by 0x8992703: gnutls_deinit (gnutls_state.c:399) ==10086== by 0x7EE3138: _lm_ssl_close (lm-ssl-gnutls.c:300) ==10086== by 0x7EDDE77: connection_do_close (lm-connection.c:514) ==10086== by 0x7EDFB31: lm_connection_close (lm-connection.c:1375) ==10086== by 0x7CC5FFA: server_cleanup (in /usr/lib/irssi/modules/libxmpp_core.so) ==10086== by 0x4879E1: ??? (in /usr/bin/irssi) ==10086== by 0x487E4C: signal_emit (in /usr/bin/irssi) ==10086== by 0x482A9D: server_disconnect (in /usr/bin/irssi) ==10086== Address 0xb70a9e8 is 8 bytes before a block of size 1,167 alloc'd ==10086== at 0x4C2935B: malloc (vg_replace_malloc.c:270) ==10086== by 0x8C1F0F0: do_malloc.constprop.4 (in /lib/x86_64-linux-gnu/libgcrypt.so.11.8.2) ==10086== by 0x8C1FF96: _gcry_calloc (in /lib/x86_64-linux-gnu/libgcrypt.so.11.8.2) ==10086== by 0x8C27EB6: _gcry_cipher_open (in /lib/x86_64-linux-gnu/libgcrypt.so.11.8.2) ==10086== by 0x89EE663: wrap_gcry_cipher_init (cipher.c:46) ==10086== by 0x897FA0E: _gnutls_cipher_init (gnutls_cipher_int.c:64) ==10086== by 0x89898BD: _gnutls_init_record_state.isra.2 (gnutls_constate.c:299) ==10086== by 0x8989D6D: _gnutls_epoch_set_keys (gnutls_constate.c:431) ==10086== by 0x898A37D: _gnutls_write_connection_state_init (gnutls_constate.c:602) ==10086== by 0x89754C3: _gnutls_send_handshake_final (gnutls_handshake.c:2888) ==10086== by 0x897841B: _gnutls_handshake_common (gnutls_handshake.c:3121) ==10086== by 0x8979BC9: gnutls_handshake (gnutls_handshake.c:2690) ==10086== ==10086== Invalid write of size 4 ==10086== at 0x4C2EBEF: memset (mc_replace_strmem.c:1007) ==10086== by 0xDD5BCF2: ??? (in /usr/lib/libotr.so.5.0.0) ==10086== by 0x8C1FEA5: _gcry_free (in /lib/x86_64-linux-gnu/libgcrypt.so.11.8.2) ==10086== by 0x897FBF3: _gnutls_cipher_deinit (gnutls_cipher_int.c:147) ==10086== by 0x898A1B4: _gnutls_epoch_free (gnutls_constate.c:786) ==10086== by 0x8992703: gnutls_deinit (gnutls_state.c:399) ==10086== by 0x7EE3138: _lm_ssl_close (lm-ssl-gnutls.c:300) ==10086== by 0x7EDDE77: connection_do_close (lm-connection.c:514) ==10086== by 0x7EDFB31: lm_connection_close (lm-connection.c:1375) ==10086== by 0x7CC5FFA: server_cleanup (in /usr/lib/irssi/modules/libxmpp_core.so) ==10086== by 0x4879E1: ??? (in /usr/bin/irssi) ==10086== by 0x487E4C: signal_emit (in /usr/bin/irssi) ==10086== Address 0xb70a9e8 is 8 bytes before a block of size 1,167 alloc'd ==10086== at 0x4C2935B: malloc (vg_replace_malloc.c:270) ==10086== by 0x8C1F0F0: do_malloc.constprop.4 (in /lib/x86_64-linux-gnu/libgcrypt.so.11.8.2) ==10086== by 0x8C1FF96: _gcry_calloc (in /lib/x86_64-linux-gnu/libgcrypt.so.11.8.2) ==10086== by 0x8C27EB6: _gcry_cipher_open (in /lib/x86_64-linux-gnu/libgcrypt.so.11.8.2) ==10086== by 0x89EE663: wrap_gcry_cipher_init (cipher.c:46) ==10086== by 0x897FA0E: _gnutls_cipher_init (gnutls_cipher_int.c:64) ==10086== by 0x89898BD: _gnutls_init_record_state.isra.2 (gnutls_constate.c:299) ==10086== by 0x8989D6D: _gnutls_epoch_set_keys (gnutls_constate.c:431) ==10086== by 0x898A37D: _gnutls_write_connection_state_init (gnutls_constate.c:602) ==10086== by 0x89754C3: _gnutls_send_handshake_final (gnutls_handshake.c:2888) ==10086== by 0x897841B: _gnutls_handshake_common (gnutls_handshake.c:3121) ==10086== by 0x8979BC9: gnutls_handshake (gnutls_handshake.c:2690) But just now I found #376463, especially Message #10: On 7/3/06, Ian Goldberg <i...@cypherpunks.ca> wrote: > This is a known problem; it occurs when you've got another plugin active > that also uses libgcrypt. libgcrypt keeps global state, and has no way > to gracefully handle being initialized and removed by two separate > shared objects. Over on the libgcrypt list, they were talking about how > to fix this a little while back. Since irssi-plugin-otr is using libgcrypt, and irssi-plugin-xmpp is using loudmouth which is using gnutls for ssl connections which is using libgcrypt, I think we can conclude that until #376463 is fixed, using jabber-over-ssl together with otr in irssi is just not possible. Florian -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
