Use addressed-ordered first fit for the pinned memory pool #9989
Conversation
This file contains 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
rongou
added
3 - Ready for Review
Codecov Report
@@ Coverage Diff @@
## branch-22.02 #9989 +/- ##
================================================
- Coverage 10.49% 10.45% -0.04%
================================================
Files 119 119
Lines 20305 20417 +112
================================================
+ Hits 2130 2134 +4
- Misses 18175 18283 +108
Continue to review full report at Codecov.
|
jlowe
reviewed
Jan 7, 2022
revans2
approved these changes
Jan 7, 2022
jlowe
approved these changes
Jan 7, 2022
|
@gpucibot merge |
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Outdated suggestions cannot be applied.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.
The current
PinnedMemoryPoolalways allocate from the largest free buffer, which might cause more fragmentation than necessary. Address-ordered first-fit has shown to perform slightly better than best-fit when it comes to memory fragmentation*, and slightly cheaper to implement. It is also used by some popular allocators such as jemalloc.This seems like a low risk change, and I do see a few percentage of performance improvement on my desktop with Q50, although the effect is less obvious on the DGX-2.