schbench: Avoid zero-page backed memory for matrices#12
Open
punitagrawal wants to merge 1 commit intomasoncl:mainfrom
Open
schbench: Avoid zero-page backed memory for matrices#12punitagrawal wants to merge 1 commit intomasoncl:mainfrom
punitagrawal wants to merge 1 commit intomasoncl:mainfrom
Conversation
Schbench uses matrix multiplication to model the cache footprint of cpu-bound phase of a query processing workload. The matrix buffers are allocated during setup and then used uninitialised. The kernel initially backs such pages with the global read-only zero page until they are faulted in and written. In the workload, the result matrix is read and written, but the operand matrices are read-only. As a result, the operand pages remain mapped to the same zero-page, collapsing the working set and producing a much smaller cache footprint than requested. Fix this by allocating matrix memory with mmap() and prefaulting it using `MAP_POPULATE` so operand pages are instantiated and the cache footprint matches the configured size. Signed-off-by: Punit Agrawal <punit.agrawal@oss.qualcomm.com>
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
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Schbench uses matrix multiplication to model the cache footprint of cpu-bound phase of a query processing workload.
The matrix buffers are allocated during setup and then used uninitialised. The kernel initially backs such pages with the global read-only zero page until they are faulted in and written.
In the workload, the result matrix is read and written, but the operand matrices are read-only. As a result, the operand pages remain mapped to the same zero-page, collapsing the working set and producing a much smaller cache footprint than requested.
Fix this by allocating matrix memory with mmap() and prefaulting it using
MAP_POPULATEso operand pages are instantiated and the cache footprint matches the configured size.