Skip to content

Conversation

@ktoso
Copy link
Contributor

@ktoso ktoso commented Apr 30, 2025

This was a regression and now even if we had a queue set in
job->SchedulerPrivate[Job::DispatchQueueIndex].

Arguably the way we're using that storage needs to be fixed somehow,
since if we're opening this storage to other executors then it means
we could have OTHER things in there, not just a dispatch queueue,
and therefore this cast would be bad.

For now though we can bring back the existing behavior while we work out
with Alastair what the right way to handle this will be -- some marker
bit somewhere that indeed this is a queue stored here?

This was accidentally removed in 8b15b05#diff-d90051c83a2c096660ba71209df0910583b57e96b23f0135533f425c49a9baacL384-L390

resolves rdar://150213107

ktoso added 2 commits April 30, 2025 21:34
…ecutorForEnqueuedJob

This was a regression and now even if we had a queue set in
job->SchedulerPrivate[Job::DispatchQueueIndex].

Arguably the way we're using that storage needs to be fixed somehow,
since if we're opening this storage to other executors then it means
we could have OTHER things in there, not just a dispatch queueue,
and therefore this cast would be bad.

For now though we can bring back the existing behavior while we work out
with Alastair what the right way to handle this will be -- some marker
bit somewhere that indeed this is a queue stored here?

resolves rdar://150213107
@ktoso ktoso marked this pull request as draft April 30, 2025 13:23
@ktoso
Copy link
Contributor Author

ktoso commented May 1, 2025

Replaced by #81201

@ktoso ktoso closed this May 1, 2025
@ktoso ktoso deleted the wip-bring-back-queue-executorForEnqueuedJob branch May 1, 2025 12:50
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant