Pack buffer-backed tensors correctly when moving into and out of staging #4673
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Outdated suggestions cannot be applied.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.
Summary:
Context
Previously, #4594 was landed which removed zero padding for buffer-backed tensors. As part of this change, the staging <-> GPU Buffer shaders were implemented as a direct buffer to buffer copy. However, the direct buffer to buffer copy does not account for the GPU memory layout argument; essentially, buffer backed tensors would always be packed with a contiguous memory layout.
This diff adds proper staging <-> buffer shaders such that buffer-backed tensors will have the correct memory layout on the GPU as specified by the provided memory layout.
Reviewed By: copyrightly
Differential Revision: D61150844