Skip to content

DPL: handle region notifications as required by GPU processing#6878

Merged
ktf merged 2 commits intoAliceO2Group:devfrom
ktf:expected-callbacks
Aug 23, 2021
Merged

DPL: handle region notifications as required by GPU processing#6878
ktf merged 2 commits intoAliceO2Group:devfrom
ktf:expected-callbacks

Conversation

@ktf
Copy link
Copy Markdown
Member

@ktf ktf commented Aug 21, 2021

  • Adds an option, --expected-region-callbacks , to wait for
    enough callback before going to start.
  • Deliver callbacks on the main thread, always.

@ktf ktf requested a review from a team as a code owner August 21, 2021 22:15
@ktf ktf force-pushed the expected-callbacks branch from 2812706 to f54013a Compare August 21, 2021 22:18
@ktf
Copy link
Copy Markdown
Member Author

ktf commented Aug 21, 2021

@davidrohr

@davidrohr
Copy link
Copy Markdown
Collaborator

davidrohr commented Aug 22, 2021

@ktf : thx, will check on Monday. But this seems to break many framework ctests and also the full system test fails in the FullCI. Perhaps it is not yet fully working?

@ktf
Copy link
Copy Markdown
Member Author

ktf commented Aug 22, 2021

I will check. I only tested the diamond topology.

@ktf
Copy link
Copy Markdown
Member Author

ktf commented Aug 22, 2021

Ok, I am missing the default argument. If you pass the option explicitly it should work. Fix coming.

@ktf
DPL: handle region notifications as required by GPU processing
5156107 
* Adds an option, --expected-region-callbacks <N>, to wait for
  enough callback before going to start.
* Deliver callbacks on the main thread, always.
@ktf ktf force-pushed the expected-callbacks branch from f54013a to 5156107 Compare August 22, 2021 09:44
@ktf
Copy link
Copy Markdown
Member Author

ktf commented Aug 22, 2021

I fixed it. Is the fullCI error actually related? The option is not really enabled.

@davidrohr
Copy link
Copy Markdown
Collaborator

davidrohr commented Aug 22, 2021

The FullCI error is no different compared to the one before, I think this one now is not related to this PR.
I have seen it a couple of times recent, it is some segfault in the MFT reconstruction, but yet I have not been able to reproduce it locally.

davidrohr
davidrohr previously approved these changes Aug 23, 2021
View reviewed changes
Copy link
Copy Markdown
Collaborator

@davidrohr davidrohr left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

THx, both the callbacks coming from the main thread as well as the --expected-region-callbacks work as expected

@ktf
Copy link
Copy Markdown
Member Author

ktf commented Aug 23, 2021

Clang format merged.

@ktf ktf merged commit 88efed0 into AliceO2Group:dev Aug 23, 2021
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

@davidrohr davidrohr davidrohr left review comments

Assignees

No one assigned

Labels

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

2 participants

@ktf @davidrohr