Skip to content

[SYCL] uninvoked commands cleanup #2472

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Conversation

cperkinsintel
Copy link
Contributor

@cperkinsintel cperkinsintel commented Sep 14, 2020

releasing of resource from commands that may not have been invoked, or if queue::wait() is not called.
Signed-off-by: Chris Perkins [email protected]

…or if queue::wait() is not called

Signed-off-by: Chris Perkins <[email protected]>
Signed-off-by: Chris Perkins <[email protected]>
@cperkinsintel cperkinsintel requested a review from a team as a code owner September 14, 2020 23:36
@cperkinsintel cperkinsintel changed the title [SYCL[ uninvoked commands cleanup [SYCL] uninvoked commands cleanup Sep 15, 2020
@sergey-semenov
Copy link
Contributor

sergey-semenov commented Sep 15, 2020

Can we kill off commands without memory object dependencies right after they are enqueued with enqueueImp() instead? There's no point in running the regular cleanup for such commands, since they are represented by orphan nodes, nor do we need to keep them in the graph, since they are not used while searching for requirements of a new command.

@cperkinsintel
Copy link
Contributor Author

@sergey-semenov . Maybe we can do what you suggest, but doesn't that presume that there is only one call to enqueueImp ? What if the user makes makes multiple .submits to a queue? Or calls q.single_task and q.submit ?

Signed-off-by: Chris Perkins <[email protected]>
@cperkinsintel
Copy link
Contributor Author

Different approach in a different PR ( #2492 ) . Closing this one.

@cperkinsintel cperkinsintel deleted the cperkins-uninvoked-commands-cleanup branch September 17, 2020 23:54
jsji pushed a commit that referenced this pull request Apr 4, 2024
kbenzie pushed a commit to kbenzie/intel-llvm that referenced this pull request Feb 17, 2025
[L0] For 3-Channel memory, disable Copy engine Usage
Chenyang-L pushed a commit that referenced this pull request Feb 18, 2025
[L0] For 3-Channel memory, disable Copy engine Usage
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.

2 participants