| commit | 1f8e221a21119a95c8975fa30044d34b35c25b32 | |
| author | Stefan Becker <chemobejk@gmail.com> | |
| Fri, 31 Aug 2012 21:46:36 +0000 (1 00:46 +0300) | ||
| committer | Stefan Becker <chemobejk@gmail.com> | |
| Sun, 9 Sep 2012 12:20:06 +0000 (9 15:20 +0300) | ||
| tree | 3a95f13ffc7b53a952cf3f697fabf49339298960 | treesnapshot (tar.gz zip) |
| parent | 9668dea53413cda4e3b763df0b52963a6ab8d0dc | commitdiff |
telepathy: fix crash from read_completed during flush
When a transport has the do_flush flag set then none of the outstanding
async operations gets cancelled, because they all share the same cancel
object. So it can happen that read_completed is called when the SIPE
core has already been deallocated.
Simply ignore the read result in this case.
When a transport has the do_flush flag set then none of the outstanding
async operations gets cancelled, because they all share the same cancel
object. So it can happen that read_completed is called when the SIPE
core has already been deallocated.
Simply ignore the read result in this case.
| src/telepathy/telepathy-transport.c | diffblobblamehistory |