Replace detail::merge_sort::dispatch by CUB's public API#8473
Merged
bernhardmgruber merged 7 commits intoNVIDIA:mainfrom Apr 20, 2026
Merged
Replace detail::merge_sort::dispatch by CUB's public API#8473bernhardmgruber merged 7 commits intoNVIDIA:mainfrom
detail::merge_sort::dispatch by CUB's public API#8473bernhardmgruber merged 7 commits intoNVIDIA:mainfrom
Conversation
oleksandr-pavlyk
approved these changes
Apr 15, 2026
| } | ||
| }; | ||
|
|
||
| struct block_size_compare_t |
Contributor
There was a problem hiding this comment.
This operator assumes that only one CTA is going to be launched. Perhaps an assertion in the if branch should be added to make that assumption explicit
Contributor
Author
There was a problem hiding this comment.
Right. I added atomicMax to reduce the blockDim in case there would be multiple blocks.
This comment has been minimized.
This comment has been minimized.
bernhardmgruber
force-pushed
the
use_merge_sort_public
branch
3 times, most recently
from
6447c37 to
4eb7bcc
Compare
bernhardmgruber
commented
Apr 16, 2026
Comment on lines
+44
to
54
| unsigned int* ptr; | ||
|
|
||
| __device__ int operator()(int a, int b) | ||
| { | ||
| if (threadIdx.x == 0) | ||
| { | ||
| *ptr = blockDim.x; | ||
| // use an atomic operation to write the block dim in case multiple blocks are launched | ||
| atomicMax(ptr, blockDim.x); | ||
| } | ||
| return a + b; | ||
| } |
Contributor
Author
There was a problem hiding this comment.
Remark: This is a drive-by fix for consistency.
bernhardmgruber
commented
Apr 16, 2026
Comment on lines
+515
to
+519
| if constexpr (cuda::std::execution::__queryable_with<env_t, get_expected_allocation_size_t>) | ||
| { | ||
| const size_t expected_bytes_allocated = fixed_env.query(get_expected_allocation_size_t{}); | ||
| REQUIRE(expected_bytes_allocated == bytes_allocated); | ||
| } |
Contributor
Author
There was a problem hiding this comment.
By making the expected_allocation_size check in the launch wrapper optional, we can also use the launch wrappers to test whether e.g. tunings were applied. At least for TEST_LAUNCH != 1.
@gonidelis I think you ran into the same issue, which is why many tests are guarded by #if TEST_LAUNCH == 0. Just FYI.
This comment has been minimized.
This comment has been minimized.
1 task
This comment has been minimized.
This comment has been minimized.
3 tasks
bernhardmgruber
force-pushed
the
use_merge_sort_public
branch
from
323a1d7 to
0aa5860
Compare
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
bernhardmgruber
force-pushed
the
use_merge_sort_public
branch
from
0aa5860 to
0c99200
Compare
miscco
approved these changes
Apr 20, 2026
Contributor
🥳 CI Workflow Results🟩 Finished in 1h 37m: Pass: 100%/323 | Total: 4d 22h | Max: 1h 09m | Hits: 92%/297570See results here. |
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
3 participants
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.
Fixes: #8376
detail::merge::dispatchby CUB's public API #8381thrust.test.sorton SM120cub.bench.merge_sort.pairs.baseon SM120