[libcu++] Automatically bump up the release threshold of default mempools#6718
Merged
davebayer merged 3 commits intoNVIDIA:mainfrom Nov 22, 2025
Merged
Conversation
This comment has been minimized.
This comment has been minimized.
pciolkosz
force-pushed
the
bump_default_mempool_release_threshold_to_inf
branch
from
086b0b7 to
7193828
Compare
pciolkosz
force-pushed
the
bump_default_mempool_release_threshold_to_inf
branch
from
7193828 to
2666fad
Compare
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
davebayer
reviewed
Nov 22, 2025
Contributor
|
pre-commit.ci autofix |
davebayer
approved these changes
Nov 22, 2025
Contributor
🥳 CI Workflow Results🟩 Finished in 58m 59s: Pass: 100%/90 | Total: 14h 01m | Max: 52m 38s | Hits: 98%/213937See results here. |
Contributor
|
Successfully created backport PR for |
rapids-bot bot
pushed a commit
to rapidsai/rapidsmpf
that referenced
this pull request
Dec 11, 2025
Before NVIDIA/cccl#6718, the default `release_threshold` was `0`, which defeats the purpose of having a pool. We now set it so the pool never releases unused pinned memory. Also some clean up. Authors: - Mads R. B. Kristensen (https://github.com/madsbk) Approvers: - Lawrence Mitchell (https://github.com/wence-) URL: #735
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
CUDA memory pools by default have 0 release threshold, which means they don't cache any memory. It is considered a mistake and might change in the future. We changed that default for memory pools we create, but we want users to prefer the default memory pools. We should on first access to a default memory pool check if user didn't change that default 0 and if not, we will bump it up to size_t max.
This PR also fixes missing rename of the
__memory_resource_baseto__memory_pool_base.