On Wed Oct 1 06:11:42 2025 +0000, Rémi Bernon wrote:
The problem with retrieving an fd during thread init, and the reason I moved it to create_thread request, is that it triggers a deadlock with some kernel32 test terminating a thread quickly after having created it. I had opened https://gitlab.winehq.org/wine/wine/-/merge_requests/8878 to fix it, which I can reopen, but the discussions there seemed to indicate that it was unlikely to be accepted.
I don't think that this should constrain the implementation. We'll want to address it one way or another, but for now it would be OK to comment out the offending test.