-
Notifications
You must be signed in to change notification settings - Fork 288
MVP for disabling nvtx ranges for thrust::seq
#6415
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
base: main
Are you sure you want to change the base?
Conversation
bernhardmgruber
left a comment
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Please fix the remaining two comments, otherwise LGTM.
…cy is not thrust::seq when applicable
|
@bernhardmgruber @miscco now that I am thinking. Is a test expected for these facilities? And should it be under libcu++ better of maybe? Could be in |
Good point! Please add a test if you come up with a good idea on how to test it. |
🥳 CI Workflow Results🟩 Finished in 3h 34m: Pass: 100%/70 | Total: 3d 15h | Max: 3h 30m | Hits: 19%/115311See results here. |
fixes #6298
by disabling NVTX ranges for
thrust::seq. This PR specifically tracks the thrust algorithms involved inhost_vector::push_back()operations and disables NVTX ranges just for these.The disabling basically evaluates at compile time whether the policy passed to the thrust algorithm (and subsequently to
should_enable_nvtx_for_policy()) isthrust::seqor any derivatives using CRTP.Upon approval of the utility provided here, I will add all the remaining thrust algorithms.
I did isolate and run profiles on the cudf code that pointed to the problem and after the changes I got zero nvtx ranges for these lines in the profile.