Fix infinite spin with pthread_cond_timedwait - #875
Conversation
For non-threaded targets the refactoring in WebAssembly#851 accidentally led this function to spin infinitely instead of trapping as expected. Closes WebAssembly#874
|
Yes, this fixes it. |
|
Are other such cases going to be proactively adjusted (compare internal IDs, not the outer pointers)? Or are we just treating this as an exception? |
|
I don't personally know the rationale for having these be structures/pointers vs integers like they are on other platforms. To that extent I do not know myself the fallout and consequences of such a decision other than what comes up in issues like this. Basically I would not be confident myself in proactively changing things without having test cases to back things up. Not because I don't want to but because I don't understand enough of the space myself. If you've got other cases though that don't work or similar I can try to help take a look. |
|
It's just the one for now. For context, I ran into this updating DRuntime to Ultimately, it is not a blocker for us because we have to work around the subsequent trap anyway. |
3 participants
For non-threaded targets the refactoring in #851 accidentally led this function to spin infinitely instead of trapping as expected.
Closes #874