rF2 keeps crashing during LFM multiplayer, after the qualifying session. Trace file in the attachment. There was a private qualifying session, which worked fine. The problems started when switching to the next session (warm up): rF2 seemed stuck in the qualifying session. When the 2 min warmup expired, and the next session was supposed to be the race, the game crashed. Some message logs from the trace file: 1228.75s slot.cpp 1121: Exited Slot::Restart() 1228.75s slot.cpp 949: Entered Slot::Restart(8) 1228.75s slot.cpp 1036: vi(1) 1228.76s slot.cpp 1040: sr(1,1) 1228.76s slot.cpp 1121: Exited Slot::Restart() 1228.76s slot.cpp 583: Reloading skin 1228.76s vehgfx.cpp 517: Entered AdjustVehicleLiveryMaterials 1228.78s Masfile.cpp 2581: Error opening MAS file matsusaka_east.mas 1228.78s Masfile.cpp 2581: Error opening MAS file matsusaka_gp.mas 1228.78s Masfile.cpp 2581: Error opening MAS file matsusaka_international.mas 1228.78s Masfile.cpp 2581: Error opening MAS file matsusaka_intl_old_chicane.mas 1228.78s Masfile.cpp 2581: Error opening MAS file matsusaka_intl_west_chicane.mas 1228.78s Masfile.cpp 2581: Error opening MAS file matsusaka_motorcycle.mas 1228.78s Masfile.cpp 2581: Error opening MAS file matsusaka_west.mas 1229.21s main.cpp 343: onFatalError 11 Edit: related thread about "onFatalError 11" : https://forum.studio-397.com/index.php?threads/onfatalerror-11-in-trace-file.66308/ Edit2: added two mini dump files.
These crashes might be related to the lags/freezes when switching from a session to another. The recently added warmup session in LFM is 2 minutes, so possibly the crash happens because the warmup session time is over before the game has finished loading/synchronizing all the players data for that session, and it has to load a new session (race) while still in the middle of loading the previous one. These delays/lags (when joining a session, when someone leaves the session, etc.). are proportional to the number of players: for instance, when joining, it might take 3s to load each driver in the session, and this happens sequentially (so for instance for 30 drivers it's easy to reach 1.5 minutes in delay). Also, while driving in a multiplayer session, the CPU load seems also related to the number of players in the session, even if they are in totally different and distant parts of the track. This might be due to the fact that rF2 multiplayer code is in the same thread as the graphics, physics and general gameplay code. rF2 gameplay seems to be synchronous with networking, so any networking delay cause the game to block/freeze. A solution might be to move the networking code to a separate thread (as it was done for sound in the Q3 build), making it asynchronous from the rest of the gameplay.
onFatalError 11 just means something went wrong. Often a useful error will precede it (immediately, or some number of lines earlier), other times whatever has stopped working isn't mentioned in the trace file. (The former would be things like files not being found, audio device not working, etc) All the errors in that trace look normal. I would recommend cleaning out old plugins (eg Damplugin and MotecPlugin do the same thing, the latter is old) and that would suggest there could be heaps of old files present, but might not have anything to do with crashing. @coasting&cruising the dmp file seems to indicate rF2 was trying to get information from the workshop when it crashed (rFactor2!getWorkshopInfo+0x00000000003123f7). Might need a dev to look into that, assuming it's not coincidental with a GPU driver crash or something else in the background.
adding a fatal erro 11 here (on top post theres othe trace files for crash to desktop on other reasons) https://forum.studio-397.com/index....nging-sessions-escaping-to-monitor-etc.74593/
