IVF-Flat: fix irrelevant assert in the fused kernel mode#1890
Closed
achirkin wants to merge 0 commit intorelease/26.04from
Closed
IVF-Flat: fix irrelevant assert in the fused kernel mode#1890achirkin wants to merge 0 commit intorelease/26.04from
achirkin wants to merge 0 commit intorelease/26.04from
Conversation
divyegala
requested changes
Mar 7, 2026
Member
divyegala
left a comment
There was a problem hiding this comment.
As we are maintaining 2 versions of the kernel temporarily due to JIT LTO, can you please mirror the changes in the JIT version of this kernel?
Ludu-nuvai
added a commit
to Nuvai/cuvs
that referenced
this pull request
Mar 23, 2026
Cherry-picked from upstream PR rapidsai#1890. Removes incorrect assertion that triggered crashes in valid IVF-Flat fused kernel use cases. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
7 tasks
divyegala
approved these changes
Mar 24, 2026
f414d39 to
0545fa2
Compare
rapids-bot bot
pushed a commit
that referenced
this pull request
Mar 24, 2026
Variable `max_samples` inside the `interleaved_scan_kernel` is only used in the non-fused codepath and tells how many distances to compute. It's not set in the fused-kernel mode, yet checked by an `assert`, causing assertion failures incorrectly. This PR fixes the problem by guarding the assert using a constexpr condition on the fused/unfused mode. This PR is open to fix incorrectly closed and already reviewed PR #1890 Authors: - Artem M. Chirkin (https://github.com/achirkin) Approvers: - Tamas Bela Feher (https://github.com/tfeher) URL: #1941
jrbourbeau
pushed a commit
to jrbourbeau/cuvs
that referenced
this pull request
Mar 25, 2026
Variable `max_samples` inside the `interleaved_scan_kernel` is only used in the non-fused codepath and tells how many distances to compute. It's not set in the fused-kernel mode, yet checked by an `assert`, causing assertion failures incorrectly. This PR fixes the problem by guarding the assert using a constexpr condition on the fused/unfused mode. This PR is open to fix incorrectly closed and already reviewed PR rapidsai#1890 Authors: - Artem M. Chirkin (https://github.com/achirkin) Approvers: - Tamas Bela Feher (https://github.com/tfeher) URL: rapidsai#1941
jrbourbeau
pushed a commit
to jrbourbeau/cuvs
that referenced
this pull request
Mar 25, 2026
Variable `max_samples` inside the `interleaved_scan_kernel` is only used in the non-fused codepath and tells how many distances to compute. It's not set in the fused-kernel mode, yet checked by an `assert`, causing assertion failures incorrectly. This PR fixes the problem by guarding the assert using a constexpr condition on the fused/unfused mode. This PR is open to fix incorrectly closed and already reviewed PR rapidsai#1890 Authors: - Artem M. Chirkin (https://github.com/achirkin) Approvers: - Tamas Bela Feher (https://github.com/tfeher) URL: rapidsai#1941
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
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.
Variable
max_samplesinside theinterleaved_scan_kernelis only used in the non-fused codepath and tells how many distances to compute. It's not set in the fused-kernel mode, yet checked by anassert, causing assertion failures incorrectly. This PR fixes the problem by guarding the assert using a constexpr condition on the fused/unfused mode.