Avoid allocating GPU memory out of RMM managed pool in test #9985
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
Signed-off-by: Ferdinand Xu <[email protected]>
|
build |
|
LGTM |
|
do we know why this just started failing? Seems like this should have been an issue for a while. Were we not using the ARENA allocator before? Or maybe we aren't explicitly setting it and it was always on machine with newer cuda? |
tgravescs
previously approved these changes
Dec 7, 2023
Yes, I think so. To reproduce this, I hardcoded the memory pool to arena. |
|
build |
|
do we need this to be targeted to 23.12, or as this is the test only we just ignore it in 23.12? |
My bad. Should target on 23.12 and merged back to 24.02. Let me re-target this. |
|
build |
pxLi
approved these changes
Dec 8, 2023
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Outdated suggestions cannot be applied.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.
This closes #9982.
The root cause is that within our test, we create a GPU column vector via Spark-Rapids-JNI API
ColumnVector.fromBytesbefore RMM initialization. And we close it within Spark GPU session with RMM initialized. So RMM mistakenly deallocated that column vector via its memory address not visible to RMM (allocated before RMM initialized).