Repository navigation
GPU Clusterizer: synchronize the cluster transfer to the host before sorting the clusters - #16168
Merged
Merged
Conversation
…sorting the clusters In deterministic mode (or with debugLevel >= 4), the clusters are copied from the GPU to the host on the last stream, sorted on the host, and copied back to the GPU. The copy to the host was only waited for with synchronizeOutput, which RunTPCClusterizer gets as false, so the host sorted a partially transferred buffer, still holding the previous time frame's clusters, and uploaded the result: the clusters seen by the tracking, and the track counts, changed from run to run.
Member
Author
|
This is part of the reproducibility issue on macOS with the Phase 2 development PR. The issue itself predates the development, though, so I guess we can merge it as a separate PR. I will open another PR with the two other related fixes for reproducibility. |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
In deterministic mode (or with debugLevel >= 4), the clusters are copied
from the GPU to the host on the last stream, sorted on the host, and
copied back to the GPU. The copy to the host was only waited for with
synchronizeOutput, which RunTPCClusterizer gets as false, so the host
sorted a partially transferred buffer, still holding the previous time
frame's clusters, and uploaded the result: the clusters seen by the
tracking, and the track counts, changed from run to run.