I've been having the occasional crash while playing titans usually in larger 6 AI team matches where it closes to desktop. I decided to run the game in terminal to see if it gave any info, but the crashes seemed to stop. I decided to give devmode a try to compare some units. 20 minutes in, while turning off an army's vision it crashed to desktop again. The only mod I was running was legion and its dependencies. Also I run titans without steam, if that makes a difference. This is the first crash I have gotten since. Where are the crash reports located? I'm running Ubuntu 18.04.1 The terminal has some info, don't know if it is useful mind you. [11:41:58.873] INFO Message from client manicvoid : {"message_type":"change_vision_flags","payload":{"vision_flags":[false,true]}} [11:41:58.911] INFO setVisionFlags [11:41:58.911] INFO 0 0 [11:41:58.911] INFO 1 1 [11:41:58.911] INFO setVisionFlagsMask [11:41:58.911] INFO 0 0 [11:41:58.911] INFO 1 1 [11:41:58.911] INFO ClientWorld::updateObservability - Updating [21372] malloc(): memory corruption [11:41:59.117] ERROR HistoryServer::Connection 127.0.0.1:46118: control pipe ended! [11:41:59.117] INFO 127.0.0.1:46118: recvCmd() [11:41:59.117] INFO 127.0.0.1:46118: command channel closed [11:41:59.137] INFO manicvoid disconnected. 0/1 clients remaining It's probably a software problem but just in case: The Computer I'm using has: 1800X Processor 2x8gb ram gtx650 I'm kind of curious about what the cause could be, and if its the same as my earlier crashes.
You can find the whole logs in ~/.local/Uber Entertainment/Planetary Annihilation/log", but those are the same as what you see on the terminal. It may also make crashdumps that could be useful for PA Inc to examine. Those get written to /tmp/. The file names are similar to /tmp/2c97844a-06aa-fa29-167f801b-0b709e40.dmp
Thanks for the help. Checked /tmp then searched whole computer for .dmp files, no PA ones... odd. I've found the logs of a couple of my crashes, though. If they could be useful, should i upload them here? or is there a specific way of sending them to PA inc?
Files in /tmp usually get deleted on every reboot. For log files, you can use the "Upload a File" button on this forum.
I decided to manually install the latest nvidia drivers, instead of what ubuntu recommended. Then set up a match that seems to crash often (4mla AI v 4legion AI on Styx) to see if I got another crash. Crashed around 15 minutes in. Got a .dmp so hopefully more useful. An 8 AI free for all on bedlam lasted over an hour with no problem. I'm starting to wonder if it's a specific unit or planet thing. I guess this is all that can be done for now. Thanks for all the help. I always enjoy an excuse to dig through files and data.
I ran minidump-2-core and then gdb on it. Looks like a crash in the particle parsing. So this may actually be a bug in Legion. There's also an error in your log about not finding teardrop.papa particle.
I tried running a debugger on the .dmp, but pretty much have no idea what i'm doing. I've verified integrity through steam, and searched my installation for teardrop.papa. I have tearDrop.papa. Could the uppercase d be messing it up?
Yes, definitely. That was a problem with the XOV commander for the longest time too. You can try making a copy of tearDrop.papa as teardrop.papa and see if the crashing persists.
Just FYI, the case bug was fixed in Legion, so making a copy of tearDrop.pap is no longer required. See https://forums.planetaryannihilatio...expansion-released.71682/page-10#post-1148977
That seems to have been the main problem. I did get another crash today. but it seems like I need to update the game through steam, so my mistake... probably. it did produce a checkpoint file. what is it? I'll put the dmp and log here, if you/ anyone's curious.
A checkpoint file keeps track of how many times you crashed. Not sure it does anything useful these days. And I just updated to build 112314 which is newer than what you had when you crashed, so I can't read the dmp file myself anymore. At least you don't get the memory corruption anymore, so this seems to have been an unrelated crash. Please do get the dmp files coming if you do crash again with the new build.