[mythtv] release-0-18-fixes branch - segvs
joseph at picsel.com
Thu Sep 1 20:40:39 UTC 2005
I'm getting SEGVs of the mythbackend a few times a week, using the
latest source on the release-0-18-fixes branch.
2005-09-01 20:37:05.986 Changing from None to RecordingOnly
[mpeg4 @ 0xb76c8010]removing common factors from framerate
strange error flushing buffer ...
ASSERT: "rc - result_buff == (int)result_len" in tools/qstring.cpp (1953)
/home/joseph/bin/mythstart: line 7: 9133 Segmentation fault (core
(the segv occurred at 20:53).
The backtrace doesn't look overly helpful:
#0 0xb65fd087 in memcpy () from /lib/tls/libc.so.6
#1 0xb6dccc6d in QString::grow () from /usr/lib/libqt-mt.so.3
#2 0xb6dccd88 in QString::arg () from /usr/lib/libqt-mt.so.3
#3 0xb71fcce6 in MythContext::GetSetting ()
#4 0xb71fd374 in MythContext::GetNumSetting ()
#5 0x080c061b in QPtrList<ProgramInfo>::deleteItem ()
#6 0x080c1741 in QPtrList<ProgramInfo>::deleteItem ()
#7 0xb67a2b63 in start_thread () from /lib/tls/libpthread.so.0
#8 0xb665b18a in clone () from /lib/tls/libc.so.6
I'm quite willing to have a go at tracking this down myself, but would
appreciate some pointers.
I'm presuming this is memory corruption of some kind - if so, what are
my options given it only occurs after the backend has been running for a
while? Presumably valgrind is too slow to run the backend in for this long.
More information about the mythtv-dev