fix: stage direct NUDFT output in a contiguous buffer - #37
Open
DiamonDinoia wants to merge 1 commit into
Open
Conversation
do_direct_bitwise and do_direct_prime write through fiw_vec.data() and index that pointer contiguously. fiw_vec is an nda::array_view and may be a strided slice. On a strided view the kernels overwrite neighbouring elements and drop the accumulation. measures/M_iw.cpp builds exactly such a view, M_data(bl)(range::all, i, j). Both kernels now write into fk_vec, the contiguous staging buffer the type3 path already uses. do_direct zeroes fk_vec, runs the kernel, then adds the result into fiw_vec stride-aware. The constructor sizes fk_vec for every non-uniform type instead of only for type3. Direct_Strided_1D covers do_direct_bitwise and Direct_Strided_2D covers do_direct_prime, both through a stride-16 slice, against a contiguous type3 reference. Direct_Strided_1D also asserts that no element outside the slice changes. Assisted-by: Claude Opus 5 <noreply@anthropic.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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
do_direct_bitwise and do_direct_prime write through fiw_vec.data() and index that pointer contiguously. fiw_vec is an nda::array_view and may be a strided slice. On a strided view the kernels overwrite neighbouring elements and drop the accumulation. measures/M_iw.cpp builds exactly such a view, M_data(bl)(range::all, i, j).
Both kernels now write into fk_vec, the contiguous staging buffer the type3 path already uses. do_direct zeroes fk_vec, runs the kernel, then adds the result into fiw_vec stride-aware. The constructor sizes fk_vec for every non-uniform type instead of only for type3.
Direct_Strided_1D covers do_direct_bitwise and Direct_Strided_2D covers do_direct_prime, both through a stride-16 slice, against a contiguous type3 reference. Direct_Strided_1D also asserts that no element outside the slice changes.
Assisted-by: Claude Opus 5 noreply@anthropic.com