Repository navigation
perf: speed up PrimitiveArray hashing for nullable inputs - #26194
Closed
Rich-T-kid wants to merge 2 commits into
Closed
Rich-T-kid wants to merge 2 commits into
Rich-T-kid wants to merge 2 commits into
Conversation
Walk the null bitmap in 64-row chunks instead of calling `valid_indices()` per row. Dense chunks (`mask == u64::MAX`) hash 8 rows at a time with `std::array::from_fn`, so the independent hashes overlap in the pipeline instead of serializing on the store back to `hashes`. Mixed chunks bit-walk only the set bits via `trailing_zeros`, so a valid lane costs one shift + one ctz instead of a per-row null check and bounds-checked `hashes_buffer[i]`. The no-null path (`array.null_count() == 0`) is unchanged. Benchmarks (`datafusion/common/benches/with_hashes.rs`, Apple M4 Max, int64 at 3% null density, cool-cache A/B in one session): single, nulls : 5.99 µs -> 5.33 µs (-12%) multiple, nulls : 17.44 µs -> 15.51 µs (-11%) No-null benches stay within noise. The dense-path 8-wide unroll for the no-null case is deferred to a follow-up PR.
Codecov Report✅ All modified and coverable lines are covered by tests. Additional details and impacted files@@ Coverage Diff @@
## main #26194 +/- ##
==========================================
+ Coverage 82.75% 82.80% +0.05%
==========================================
Files 1147 1149 +2
Lines 449886 452300 +2414
Branches 449886 452300 +2414
==========================================
+ Hits 372303 374545 +2242
- Misses 54911 54957 +46
- Partials 22672 22798 +126 ☔ View full report in Codecov by Harness. 🚀 New features to boost your workflow:
|
Contributor
Author
|
run benchmark with_hashes |
1 similar comment
Contributor
Author
|
run benchmark with_hashes |
|
🤖 Benchmark running (GKE) | trigger CPU Details (lscpu)Comparing rich-T-kid/primitive-hash-null-chunks (25fdea2) to 409e3b8 (merge-base) diff Run configurationrun benchmark with_hashes
env:
BENCH_FILTER: "int64"Results will be posted here when complete File an issue against this benchmark runner |
|
🤖 Benchmark running (GKE) | trigger CPU Details (lscpu)Comparing rich-T-kid/primitive-hash-null-chunks (25fdea2) to 409e3b8 (merge-base) diff Run configurationrun benchmark with_hashes
env:
BENCH_FILTER: "int64"Results will be posted here when complete File an issue against this benchmark runner |
|
🤖 Benchmark completed (GKE) | trigger Instance: Comparing rich-T-kid/primitive-hash-null-chunks (25fdea2) to 409e3b8 (merge-base) diff Run configurationrun benchmark with_hashes
env:
BENCH_FILTER: "int64"CPU Details (lscpu)Details
Resource Usagewith_hashes — base (merge-base)
with_hashes — branch
File an issue against this benchmark runner |
|
🤖 Benchmark completed (GKE) | trigger Instance: Comparing rich-T-kid/primitive-hash-null-chunks (25fdea2) to 409e3b8 (merge-base) diff Run configurationrun benchmark with_hashes
env:
BENCH_FILTER: "int64"CPU Details (lscpu)Details
Resource Usagewith_hashes — base (merge-base)
with_hashes — branch
File an issue against this benchmark runner |
Contributor
Author
|
results are not significant, closing the PR |
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.
Which issue does this PR close?
Rationale for this change
Hashing a nullable
PrimitiveArraypreviously callednulls().valid_indices()andhashed one row at a time. We can speed this up
What changes are included in this PR?
Rewrite the nullable branch of
hash_array_primitiveto walk the null bufferin 64-row chunks via
NullBuffer::inner().bit_chunks():mask == u64::MAX): hash 8 rows per inner iteration usingstd::array::from_fn::<8>so the independent hashes stay in flight across thepipeline before the stores commit.
remaining & (remaining - 1)+trailing_zeros, one hash per set bit, no branching on cleared ones.remainder_len().Both paths share one
hash_prim_one::<_, _, REHASH>helper, so therehash=true(multi-column) andrehash=false(single-column) variantscollapse to one body with the const-generic picking the hash expression.
The no-null branches are intentionally left alone in this PR ! done in #26191
What is the testing strategy for this PR?
Existing
hash_utils::testsAre there any user-facing changes?
no