Bruno Kestemont Lv 1
The other client crashed
The other client crashed
Another crash after several hours running only 1 client
Another crash after some time, 2 clients (on separate logs)
I tried various Linux distros, cleanly installed them and tried running the client. After an update from the client I always get the already mentioned:
Foldit Error: The game has previously not started up correctly. If this problem persists, please install the latest version. terminate called after throwing an instance of 'std::logic_error' what(): basic_string::_S_construct null not valid Aborted (core dumped)
Since I cannot attach a coredump (unless renamed) I can only attach information about the crash:
PID: 2597 (Foldit)
UID: 1000 (nert)
GID: 1000 (nert)
Signal: 6 (ABRT)
Timestamp: Tue 2020-06-02 19:06:24 EDT (10h ago)
Executable: /home/nert/Foldit/Foldit Control Group: /user.slice/user-1000.slice/session-2.scope
Unit: session-2.scope
Slice: user-1000.slice
Session: 2
Owner UID: 1000 (nert)
Boot ID: 385011df38b14256973e9136af0d3527
Machine ID: 18e52608328c2bdd3cb158741e4b366a
Hostname: foldit.virt
Storage: /var/lib/systemd/coredump/core.Foldit.1000.385011df38b14256973e9136af0d3527.2597.1591139184000000000000.lz4
Message: Process 2597 (Foldit) of user 1000 dumped core.
Stack trace of thread 2597:
#0 0x00007fc1e9691625 raise (libc.so.6)
#1 0x00007fc1e967a8d9 abort (libc.so.6)
#2 0x00007fc1e98d8756 _ZN9__gnu_cxx27__verbose_terminate_handlerEv.cold (libstdc++.so.6)
#3 0x00007fc1e98e46dc _ZN10__cxxabiv111__terminateEPFvvE (libstdc++.so.6)
#4 0x00007fc1e98e4747 _ZSt9terminatev (libstdc++.so.6)
#5 0x00007fc1e98e49b9 __cxa_throw (libstdc++.so.6)
#6 0x00007fc1e98dae2e _ZSt19__throw_logic_errorPKc (libstdc++.so.6)
#7 0x00007fc1e992a9c1 _ZNSs12_S_constructIPKcEEPcT_S3_RKSaIcESt20forward_iterator_tag (libstdc++.so.6)
#8 0x00007fc1e992ae05 _ZNSsC2EPKcRKSaIcE (libstdc++.so.6)
#9 0x00007fc1e3cdd0ee n/a (/home/nert/Foldit/cmp-binary-db581aba9f1cb3cbef43044cbea3c66a/game_library.so)
#10 0x00007ffebb705a43 n/a (n/a)
This is from a clean install of Fedora 31 which should be very easy to reproduce.
If someone needs the coredump and is not able to reproduce that let me know where I can upload it.
Your to first sentences are normal if you don't wait untill a puzzle is open before to start a new client. In normal situation, this has no effect (I didn't try for Linux).
It seems that the memory allocation is the cause of many crashes on Linux.
I started again a second client for exactly the same puzzle 1843, same share, on another exact copy of foldit folder.
No recipe running on any of the 2 clients.
The memory use reported in the System Monitor gives:
% CPU is the same for both clients (13-14% each which is quite high fonon active clients)
Since my total memory avaiability is 8 Go for max 4 virtual cores, I can suspect that crashes are linked to memory accumulation.
Based on OWM3's comments and reports of NUMEROUS Seg Fault crashes, I'm inclined to agree with Bruno's observation on problems with Memory Allocation on Linux.
I'm also inclined to think perhaps it's also still applicable in a small way to at least Windows since I run it (as such, I can't say about MacOS). I recently had a two clients crash and I genuinely suspect that the one's crash caused the other, despite both operating out of different directories, and on different hard drives.
Impossible to open crashed client (copy of Foldit)
It's possible to run separate clients from separate copies of Foldit. Another Foldit is running with memory reachin 1.4 Go. This one does't open any more.
Unable to start (one other foldit version is running on another map with 1.1 Gb). After crash see above.
Unable to start playing a puzzle with new version: it crashes after several seconds.
Same "old" computer early i7.