Skip to content

Include number of bytes spilled in nvtx payload #662

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.

Sign up for GitHub

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

Jump to bottom
Draft
wants to merge 11 commits into
base: main
Choose a base branch
from
Draft

Include number of bytes spilled in nvtx payload #662

wants to merge 11 commits into from
+264 −20

Conversation

@TomAugspurger
Copy link
Contributor

@TomAugspurger TomAugspurger commented Nov 13, 2025

No description provided.

@copy-pr-bot
Copy link

copy-pr-bot bot commented Nov 13, 2025

Auto-sync is disabled for draft pull requests in this repository. Workflows must be run manually.

Contributors can view more details about this message here.

@TomAugspurger
Copy link
Contributor Author

TomAugspurger commented Nov 13, 2025

/ok to test ab18e77

@TomAugspurger
Copy link
Contributor Author

TomAugspurger commented Nov 13, 2025

/ok to test 46b1d7b

@pentschev
Copy link
Member

pentschev commented Nov 13, 2025

Is there a benefit in having this always turned on? It might be useful to look at #612 since this can be very verbose and thus difficult to handle in practice for regular builds.

@TomAugspurger TomAugspurger mentioned this pull request Nov 13, 2025
Open
@TomAugspurger
Copy link
Contributor Author

TomAugspurger commented Nov 13, 2025

/ok to test 105c60a

@TomAugspurger
Copy link
Contributor Author

TomAugspurger commented Nov 14, 2025

/ok to test 17bc5e6

@TomAugspurger
Copy link
Contributor Author

TomAugspurger commented Nov 14, 2025

/ok to test 1fa6250

@nirandaperera nirandaperera added improvement Improves an existing functionality non-breaking Introduces a non-breaking change labels Nov 17, 2025
@nirandaperera
Copy link
Contributor

nirandaperera commented Nov 17, 2025

/ok to test

@nirandaperera
Copy link
Contributor

nirandaperera commented Nov 17, 2025

I tested out the HostBuffer with @TomAugspurger's reproducer.

image

Left is with HostBuffer and right is with std::vector.
Now, Buffer::allocate time is insignificant. However, now cudaMemcpyAsync is taking up almost all the gains. The total time for each insert is more or less the same. 😕
Does cudaMemcpyAsync initialize memory before copying?

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

No reviews

Assignees

No one assigned

Labels

improvement Improves an existing functionality non-breaking Introduces a non-breaking change

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

3 participants

@TomAugspurger @pentschev @nirandaperera