[encoding] Fix path_reduced_scan buffer size #551
Merged
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
This is the intermediate TagMonoid buffer that is used by the two-stage scan in the "use_large_path_scan" case. This needs to be the same size as the overall path_reduce output buffer, which gets rounded up to a multiple of the path_reduce workgroup size. This hasn't been a problem until now because src/wgpu_engine.rs allocates Buffers that are generally larger than the entries returned in the BufferSizes structure, due its quantized size class / pooling strategy. The bindings get their view size assigned based on the whole size of the buffer. Skia uses a similar allocation strategy but assigns view size to be the precise value from BufferSizes. This causes incorrect behavior as WGSL clamps buffer accesses to the view size. The now corrected buffer size fixes this issue.
raphlinus
approved these changes
Apr 19, 2024
There was a problem hiding this comment.
Yes, this is definitely a problem in the original - pathtag_scan1 writes in full workgroup increments. I'm assuming the observed problem only manifests with more than 256k pathtags? I did look at the code in the small case and it seems fine, but it's possible I've missed something.
In any case, good catch.
Yes, I've only observed this in the large scan case. |
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Outdated suggestions cannot be applied.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 is the intermediate TagMonoid buffer that is used by the two-stage scan in the "use_large_path_scan" case. This needs to be the same size as the overall path_reduce output buffer, which gets rounded up to a multiple of the path_reduce workgroup size.
This hasn't been a problem until now because src/wgpu_engine.rs allocates Buffers that are generally larger than the entries returned in the BufferSizes structure, due its quantized size class / pooling strategy. The bindings get their view size assigned based on the whole size of the buffer.
Skia uses a similar allocation strategy but assigns view size to be the precise value from BufferSizes. This causes incorrect behavior as WGSL clamps buffer accesses to the view size.
The now corrected buffer size fixes this issue.