Open GoogleCodeExporter opened 8 years ago
Could you reproduce this crash with debug symbols enabled?
Are you sure this is the thread that triggered the SIGSEGV?
Original comment by arvid.no...@gmail.com
on 4 Feb 2010 at 8:55
> Could you reproduce this crash with debug symbols enabled?
you means, with debug symbols in boost libs?
libtorrent-rasterbar itself is built with -g and not stripped,
$ file /usr/lib64/libtorrent-rasterbar.so.5.0.0
/usr/lib64/libtorrent-rasterbar.so.5.0.0: ELF 64-bit LSB shared object, x86-64,
version 1 (SYSV), dynamically linked, not stripped
but it is linked with system boost libs, which are stripped.
> Are you sure this is the thread that triggered the SIGSEGV?
not sure. I am not used to gdb.
It seems I have to read gdb docs about multithreaded debugging,
then I'll give more valuable feedback.
Original comment by ivlase...@gmail.com
on 5 Feb 2010 at 8:15
I'm really only interested in debug symbols for libtorrent. From the gdb
output, it looks like it's only listing
exported symbols, and no line numbers for instance. Please make sure that the
libtorrent.so that's actually used
has debug symbols. To dump the stack of every thread in gdb, run:
(gdb) thread apply all bt full
This will give a verbose state of all threads.
Original comment by arvid.no...@gmail.com
on 5 Feb 2010 at 5:09
Arvid, the original bug report was here:
https://bugs.launchpad.net/qbittorrent/+bug/517062
There is a backtrace with debug symbols (one thread only).
ivlasenko > Please get a backtrace for all threads:
gdb qbittorrent
> run
.. Wait for the crash ..
> thread apply all bt full
.. Will print a good backtrace for all threads ..
Original comment by dch...@gmail.com
on 6 Feb 2010 at 10:19
I am sorry for waiting, here is an example of full backtrace,
I hope to devote time to debugging this evening (GMT+2),
so I will be more useful.
Original comment by ivlase...@gmail.com
on 7 Feb 2010 at 11:53
Attachments:
sorry, forgot the crash message.
Program received signal SIGABRT, Aborted.
[Switching to Thread 140508415039760 (LWP 21198)]
0x00007fcab15f9ff5 in raise () from /lib64/libc.so.6
Original comment by ivlase...@gmail.com
on 7 Feb 2010 at 11:56
This happens when the main timer that drives the session fails. If you build
libtorrent with logging enabled, the
main_session.log will tell you why it failed.
This is considered a fatal error, resuming at this point would not buy you
anything, since the session would be
completely dead and nothing would work.
Please let me know what error message are printed to the log.
Original comment by arvid.no...@gmail.com
on 8 Feb 2010 at 5:05
Here they are, main_session.logs and gdb thread apply all bt full output.
there are two main_session.logs,
libtorrent_logs6881 looks normal,
note that last line in
libtorrent_logs0/main_session.log
is
"cannot bind to interface '[::]:0' Wrong file descriptor"
where last phrase is translated according to locale (ru_RU).
Original comment by ivlase...@gmail.com
on 9 Feb 2010 at 8:27
Attachments:
my network interfaces
$ ifconfig
lo Link encap:Local Loopback
inet addr:127.0.0.1 Mask:255.0.0.0
UP LOOPBACK RUNNING MTU:16436 Metric:1
RX packets:1464 errors:0 dropped:0 overruns:0 frame:0
TX packets:1464 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:0
RX bytes:188122 (183.7 KiB) TX bytes:188122 (183.7 KiB)
net0 Link encap:Ethernet HWaddr 00:22:15:D1:70:91
inet addr:195.--.--.-- Bcast:195.--.--.-- Mask:255.255.255.0
UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
RX packets:1009341 errors:0 dropped:0 overruns:0 frame:0
TX packets:443212 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:1000
RX bytes:1096265844 (1.0 GiB) TX bytes:31141642 (29.6 MiB)
Interrupt:28 Base address:0x2000
tap0 Link encap:Ethernet HWaddr 92:46:72:0C:DC:D6
inet addr:10.100.0.1 Bcast:0.0.0.0 Mask:255.255.255.0
UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
RX packets:0 errors:0 dropped:0 overruns:0 frame:0
TX packets:0 errors:0 dropped:1273 overruns:0 carrier:0
collisions:0 txqueuelen:500
RX bytes:0 (0.0 b) TX bytes:0 (0.0 b)
tap1 Link encap:Ethernet HWaddr 42:6B:CD:2D:55:A4
inet addr:10.100.1.1 Bcast:0.0.0.0 Mask:255.255.255.0
UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
RX packets:0 errors:0 dropped:0 overruns:0 frame:0
TX packets:0 errors:0 dropped:1273 overruns:0 carrier:0
collisions:0 txqueuelen:500
RX bytes:0 (0.0 b) TX bytes:0 (0.0 b)
Original comment by ivlase...@gmail.com
on 9 Feb 2010 at 8:30
These are two different errors. One where the second_timer fails, the other
where there's a crash somewhere
inside asio, possibly because a timer was destructed before the callback
returned
Original comment by arvid.no...@gmail.com
on 11 Feb 2010 at 1:48
Ok, please, notify me when there will be possible fix, to test.
Original comment by ivlase...@gmail.com
on 17 Feb 2010 at 9:25
there is fresh logs for 0.14.9+qbittorrent 2.1.5
gdb backtrace is saved in gdb.log.
Original comment by ivlase...@gmail.com
on 21 Feb 2010 at 4:30
Attachments:
seems to be fixed in 0.14.10.
Original comment by ivlase...@gmail.com
on 21 Mar 2010 at 5:39
not completely. debug build fails on assert
(http://code.google.com/p/libtorrent/issues/detail?id=49)
build without debug code failed with
boost::asio::detail::deadline_timer_service after mass pausing in qbittorrent
gut,
it might be related.
qbittorrent
Peer ID: -qB2210-
HTTP user agent is qBittorrent v2.2.1
Warning: Local domain socket was not closed cleanly, deleting file...
*************************************************************
Catching SIGSEGV, please report a bug at http://bug.qbittorrent.org
and provide the following backtrace:
stack trace:
/lib64/libc.so.6 [0x7fc59a42f070]
/usr/lib64/libtorrent-rasterbar.so.5 :
boost::asio::detail::deadline_timer_service<boost::asio::time_traits<libtorrent:
:ptime>,
boost::asio::detail::epoll_reactor<false>
>::cancel(boost::asio::detail::deadline_timer_service<boost::asio::time_traits<l
ibtorrent::ptime>,
boost::asio::detail::epoll_reactor<false> >::implementation_type&,
boost::system::error_code&)+0x88 [0x7fc59c65a0c6]
/usr/lib64/libtorrent-rasterbar.so.5 : libtorrent::torrent::stop_announcing()+0x34
[0x7fc59c68deba]
/usr/lib64/libtorrent-rasterbar.so.5 : libtorrent::torrent::do_pause()+0x220
[0x7fc59c68e120]
/usr/lib64/libtorrent-rasterbar.so.5 : libtorrent::torrent::pause()+0x39
[0x7fc59c68e95b]
/usr/lib64/libtorrent-rasterbar.so.5 : libtorrent::torrent_handle::pause()
const+0x7b [0x7fc59c6ae781]
/usr/bin/qbittorrent : QTorrentHandle::pause()+0x2e [0x50de5e]
/usr/bin/qbittorrent : TransferListWidget::pauseSelectedTorrents()+0xf6 [0x5e5b26]
...
Original comment by ivlase...@gmail.com
on 21 Mar 2010 at 7:54
Original issue reported on code.google.com by
ivlase...@gmail.com
on 4 Feb 2010 at 5:29