ITS: slab allocator idea - #15678
Open
f3sch wants to merge 1 commit into
Open
Conversation
f3sch
requested review from
fprino,
mconcas,
mcoquet642 and
shahor02
as code owners
August 13, 2026 16:17
Signed-off-by: Felix Schlepper <felix.schlepper@cern.ch>
Collaborator
|
Error while checking build/O2/fullCI_slc9 for 69d6252 at 2026-08-13 20:18: Full log here. |
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
2 participants
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
@mpuccio this what I think is a pretty neat idea (after Gabriele reminded me to revive this dead code and codex looked over it and wrote reasonable test cases).
We carry the size estimates of the buffer + margin and hand out slices of that pre-allocated buffer (slabs).
If the memory does not suffice there additional spill vectors transparently created and merged at the end.
Keeping the same output and running with 8 threads:
48 kHz Pb-Pb 23% faster (here the size estimates are spilling more often due to charged-particle multiplicity fluctuations)
750 kHz pp 30% faster
The memory profile also barely moved (<2%) upwards in my test case and the size estimates stabilise quickly.