https://bugs.winehq.org/show_bug.cgi?id=53475
--- Comment #6 from ilkka.prusi@gmail.com --- (In reply to ilkka.prusi from comment #5)
(In reply to Rafał Mużyło from comment #3)
...I strongly suspect this is a dupe of my bug 50779.
The exact conditions needed for a ruby maker game to trigger this bug are somewhat unclear.
The mentioned recent pulseaudio flag should workaround the bug, but only in a 'paper over' sense.
And it likely doesn't solve the thing when using Pipewire instead of Pulseaudio?
I speculate that with synchronization bugs there will happen memory leaks and other issues as well, which can manifest as running out of memory at some point. Differences in latency (kernel scheduling options, Pipewire vs. Pulseaudio) might also change how contested the critical sections are and contribute to how synchronization bugs appear.