Closed ghost closed 9 years ago
Thanks for the report. What architecture and steam client version are you running? Could you post the output of steam_idle?
I was able to reproduce this once. Steam client crashed with segfault and installed an update afterwards (version 1440016726). Now I'm unable to reproduce it again. Could you please re-check with the latest stable steam client version?
I was also using 1440016726 on Ubuntu 15.04 64 bit. I tried it again just and it seems to work fine now.
The only difference is that all the crashes occurred when idling more than 1 game for the initial 2 hour period. Launching multiple games at once seemed to be the only difference so far. I'm only idling 1 game right now since I only have 1 game which requires the initial 2 hour sleep period. I'll keep an eye on it, and see if I can reproduce it once I have multiple games that need to sleep. Thanks!
It breaks on "multi-idle" for me too (again now, but only once and after that it's running fine). The output looks as if the idle child's don't spawn correctly:
3 games with a total of 8 card drops left:
Sleeping for 1:54:00 till Mon Aug 24 12:16:06 2015
When it's working we see the idle childs spawn (for two games in my case):
3 games with a total of 8 card drops left:
Sleeping for 1:54:00 till Mon Aug 24 12:20:46 2015
Idle-2-[232770](9290): Ideling appid 232770
Idle-1-[262830](9289): Ideling appid 262830
I've added a debug switch -d
to master and a new pre-python3 branch that is based on 2d6ff2bbc3292bc91a8002c2695f6850f77f289f. That enables some extra output and disables the redirection of stderr and stdout when loading the steam API. I hope we're able to catch the problem than.
For some reason it's working fine now for multiple game idling. I'll update to latest and test again.
And pre-python3 branch seems to be working fine for now.
While experimenting with Qt's threading this occurred again. This is what the steam client dumps to stderr in case of the crash:
/home/buildbot/buildslave/steam_rel_client_linux64/build/src/clientdll/../common/pipes.cpp (735) : CClientPipe::BWriteAndReadResult: BWaitResult failed, disconnected
/home/buildbot/buildslave/steam_rel_client_linux64/build/src/clientdll/../common/pipes.cpp (735) : Fatal assert failed: /home/buildbot/buildslave/steam_rel_client_linux64/build/src/clientdll/../common/pipes.cpp, line 735. Application exiting.
/home/buildbot/buildslave/steam_rel_client_linux64/build/src/clientdll/../common/pipes.cpp (735) : CClientPipe::BWriteAndReadResult: BWaitResult failed, disconnected
/home/buildbot/buildslave/steam_rel_client_linux64/build/src/clientdll/../common/pipes.cpp (735) : Fatal assert failed: /home/buildbot/buildslave/steam_rel_client_linux64/build/src/clientdll/../common/pipes.cpp, line 735. Application exiting.
Assert( CClientPipe::BWriteAndReadResult: BWaitResult failed, disconnected ):/home/buildbot/buildslave/steam_rel_client_linux64/build/src/clientdll/../common/pipes.cpp:735
Assert( CClientPipe::BWriteAndReadResult: BWaitResult failed, disconnected ):/home/buildbot/buildslave/steam_rel_client_linux64/build/src/clientdll/../common/pipes.cpp:735
No minidump written, nothing to upload.
_ExitOnFatalAssert
This has occurred again for me several times, but it only occurs on the first time after logging into Steam and so it's kind of hard for me to replicate.
Ok, I was able to catch a crash with a fresh batch of games. Here's what steam spit at me:
https://gist.github.com/xHN35RQ/a6f5ca8d915262c82e39
Let me know if you want /tmp/dumps/assert_20150902205511_1.dmp
Meanwhile steam-idle
said this during the crash:
3 games with a total of 15 card drops left:
Sleeping for 2:00:00 till Wed Sep 2 22:55:25 2015
Idle-2-[284850](4879): Ideling appid 284850
EDIT - just realized this is not helpful, as I forgot to use the debug mode to catch the output. argh....
The breakpad file (*.dmp) is unfortunately pretty useless because we only have the steam client libs with debugging info stripped of :disappointed:.
But I can reproduce this now with the following steps (in multi-idle mode only!):
This crashed the master branch and the pre-python3 branch for me. So at least that is consistent (I would have been really surprised if the branched would behave differently).
Thanks for your help! :+1: Would like to know if this is "fixed" for you too.
Steam segfaults and crashed when launching
steam_idle
- even when launching with Python3. This commit seems to have caused some problems: a87e0f8e7I rolled back to 2d6ff2bbc and have no issues. Let me know what more information I can provide.
Ubuntu 15.04