You're correct that debuggers can be sensitive to the timing of the initial breakpoint.
Adding a conformance test would let us confirm your observation, and ensure such regression won't happen in the future. For reference, see 2b40969c58a099f958c58909b666915991a42979.
I'm still approving since the existing tests did not fail and I cannot find a particular reason `process_breakpoint();` must be placed exactly there in the commit history.
--
https://gitlab.winehq.org/wine/wine/-/merge_requests/4655#note_56412
DMUS_PMSG::dwType is the type of the message, but we are comparing it to
curve types. We should be using DMUS_CURVE_PMSG::bType instead.
* * *
(I tried to write some test cases for this. I created a IDirectMusicPort implementation that checks the generated midi commands, but instead I cannot get Windows to generate any midi commands for CCCURVEs. )
--
v2: dmime: Fix handling of curve PMSG.
https://gitlab.winehq.org/wine/wine/-/merge_requests/4703
Not really needed for 16k pages, but since the comment says "one page" the logic should reflect that.
Not 100% sure about that since I don't know why it should use exactly one page, so please double check.
--
https://gitlab.winehq.org/wine/wine/-/merge_requests/4740