0001677: Sometimes vcmi crashes during advancing to next scenario

View Revisions: Issue #1677		[ Back to Issue ]
Summary	0001677: Sometimes vcmi crashes during advancing to next scenario

Revision	2014-02-01 12:05 by Ivan
Description	This is very random crash that happens quite rarely when advancing to next scenario in campaign. May be duplicate of some of crashes reported here. Not sure what's going on so I took two stacktraces from gdb: gdb-short.txt - just stacktraces from all threads (bt) gdb-full.txt - full stacktraces including locals (bt full) Two most "interesting threads are: 1) One that causes crash, failure to unlock a mutex: 0000005 0x00000000009f3e4b in boost::unique_lock<boost::recursive_mutex>::~unique_lock (this=0x7fffdf065d00, __in_chrg=<optimized out>) at /usr/include/boost/thread/lock_types.hpp:329 0000006 0x0000000000ba4d0f in CPlayerInterface::update (this=0x7fffb9e01f88) at code/client/CPlayerInterface.cpp:1615 0000007 0x0000000000a3933a in CGuiHandler::run (this=0x11d1d20 <GH>) at code/client/gui/CGuiHandler.cpp:388 2) Thread that handles connection, currently in middle of de-serialization of campaign state (may be unrelated): 0000014 0x0000000000b51553 in Bonus::serialize<CISer<CConnection> > ( this=0x7fffb4d0f580, h=..., version=0) at code/lib/HeroBonus.h:330 ... 0000085 0x0000000000b15315 in CGHeroInstance::serialize<CISer<CConnection> > ( this=0x7fffb57b8220, h=..., version=743) at code/lib/CObjectHandler.h:397 ... 0000140 0x00007ffff7392184 in UpdateCampaignState::serialize<CISer<CConnection> > (this=0x7fffb5556f90, h=..., version=743) at code/lib/NetPacks.h:442 ... 0000148 0x0000000000b74a2a in CClient::run (this=0x7fffb8eb50b0) at code/client/Client.cpp:142

Revision	2014-02-01 12:05 by Ivan
Description	This is very random crash that happens quite rarely when advancing to next scenario in campaign. May be duplicate of some of crashes reported here. Not sure what's going on so I took two stacktraces from gdb: gdb-short.txt - just stacktraces from all threads (bt) gdb-full.txt - full stacktraces including locals (bt full) Two most "interesting threads are: 1) One that causes crash, failure to unlock a mutex: 5 0x00000000009f3e4b in boost::unique_lock<boost::recursive_mutex>::~unique_lock (this=0x7fffdf065d00, __in_chrg=<optimized out>) at /usr/include/boost/thread/lock_types.hpp:329 6 0x0000000000ba4d0f in CPlayerInterface::update (this=0x7fffb9e01f88) at code/client/CPlayerInterface.cpp:1615 7 0x0000000000a3933a in CGuiHandler::run (this=0x11d1d20 <GH>) at code/client/gui/CGuiHandler.cpp:388 2) Thread that handles connection, currently in middle of de-serialization of campaign state (may be unrelated): 14 0x0000000000b51553 in Bonus::serialize<CISer<CConnection> > ( this=0x7fffb4d0f580, h=..., version=0) at code/lib/HeroBonus.h:330 ... 85 0x0000000000b15315 in CGHeroInstance::serialize<CISer<CConnection> > ( this=0x7fffb57b8220, h=..., version=743) at code/lib/CObjectHandler.h:397 ... 140 0x00007ffff7392184 in UpdateCampaignState::serialize<CISer<CConnection> > (this=0x7fffb5556f90, h=..., version=743) at code/lib/NetPacks.h:442 ... 148 0x0000000000b74a2a in CClient::run (this=0x7fffb8eb50b0) at code/client/Client.cpp:142

Site | Forums | Wiki | Slack | GitHub