On Mon Apr 3 07:07:22 2023 +0000, Florian Will wrote:
I'm not sure if my patch to the test could have possibly introduced this new failure? I mean, sure, the behavior when returning -1 from the callback instead of DIENUM_CONTINUE / DIENUM_STOP is not specified in the docs. So I guess win7 would be free to do something that would end up e.g. blocking the thread. I suspect that behavior would be pretty unlikely though, and [it didn't reproduce when I manually resubmitted the test](https://testbot.winehq.org/JobDetails.pl?Key=131414).
Don't worry, this sometimes happens for various reasons.