Info on segfaults? Getting a lot on RPI4

[ninjakitty844]

Some background info: I'm turning my raspberry pi 4b (8gb of ram and overclocked to 2.3ghz cpu, 900hz gpu) into a retro gaming laptop. I use both Box86(with dynarec) and Wine(i386) to play many games that would normally only run on Windows and not ARM architecture. Examples include: Quake III (more accurately, a standalone mod for it called World of Padman), Undertale, and Portal.

The problem: Upon running "wine program.exe" I get a segmentation fault error sometimes. Not all the time, only sometimes. If I am able to successfully launch a game, it will not segfault while running. The problem only occurs while starting up Wine, but once the program is running it's fine. I find wine 5.13 development branch to be stable and it rarely segfaults when i launch a program with it, but it is clearly outdated by now. I want to use a higher version of Wine, but the higher I go the more often I get segfaults. With Wine 7 and above I can type "wine program.exe" upwards of 10 times and only get 1 successful launch.

Most times when I get a segfault, it looks identical to this one:

Code: Select all

/usr/local/bin/wine: line 2:  2952 Segmentation fault      setarch linux32 -L /home/pi/wine/bin/wine "$@"

But I'll also very rarely get errors about services not being started, etc:

Code: Select all

double free or corruption (fasttop)
0060:err:seh:NtRaiseException Exception frame is not in stack limits => unable to dispatch exception.
003c:err:service:process_send_command service protocol error - failed to read pipe r = 0  count = 0!
003c:fixme:service:scmdatabase_autostart_services Auto-start service L"MountMgr" failed to start: 1053
0078:fixme:hid:handle_IRP_MN_QUERY_ID Unhandled type 00000005
0078:fixme:hid:handle_IRP_MN_QUERY_ID Unhandled type 00000005
0078:fixme:hid:handle_IRP_MN_QUERY_ID Unhandled type 00000005
0078:fixme:hid:handle_IRP_MN_QUERY_ID Unhandled type 00000005
Box86 with Dynarec v0.2.8 891593c built on Nov 26 2022 02:46:15
0058:err:service:process_send_command receiving command result timed out
0050:err:ole:start_rpcss Failed to start RpcSs service

If I start a persistant wineserver manually ("wineserver -p") and wait a bit before starting my program, Wine will segfault less often.

So what might be causing these segfaults? What kind of debug info should I collect? And how can I better avoid segfaults without staying back on old Wine 5.13?
Is my wineserver really taking too long to start up? Should I just always have a persistent wineserver running? (I'm looking for a different solution ideally)

[ninjakitty844]

Correction: I just now had a few segfaults even with wineserver already started up. So it helps still, but it does not fully fix the problem.