Conversation
connortsui20
added
action/benchmark
github-actions
bot
removed
the
action/benchmark
Merging this PR will degrade performance by 15.51%
Performance Changes
Comparing 
Footnotes
|
Contributor
|
Seems like there are some things that needed to be inlined in here? related to bitpacking and between eval |
Contributor
Polar Signals Profiling ResultsLatest Run
Previous Runs (2)
Powered by Polar Signals Cloud |
Contributor
Benchmarks: PolarSignals ProfilingVortex (geomean): 1.030x ➖ datafusion / vortex-file-compressed (1.030x ➖, 0↑ 1↓)
|
Contributor
Benchmarks: TPC-H SF=1 on NVMEVerdict: No clear signal (environment too noisy confidence) datafusion / vortex-file-compressed (0.996x ➖, 1↑ 0↓)
datafusion / vortex-compact (1.010x ➖, 0↑ 0↓)
datafusion / parquet (0.982x ➖, 1↑ 0↓)
datafusion / arrow (0.987x ➖, 1↑ 0↓)
duckdb / vortex-file-compressed (0.900x ✅, 9↑ 0↓)
duckdb / vortex-compact (0.919x ➖, 5↑ 0↓)
duckdb / parquet (0.968x ➖, 5↑ 3↓)
duckdb / duckdb (0.911x ➖, 6↑ 0↓)
Full attributed analysis
|
Contributor
Benchmarks: FineWeb NVMeVerdict: No clear signal (low confidence) datafusion / vortex-file-compressed (0.874x ✅, 7↑ 1↓)
datafusion / vortex-compact (0.919x ➖, 3↑ 1↓)
datafusion / parquet (0.929x ➖, 2↑ 0↓)
duckdb / vortex-file-compressed (0.825x ✅, 8↑ 0↓)
duckdb / vortex-compact (0.884x ✅, 4↑ 0↓)
duckdb / parquet (0.937x ➖, 1↑ 0↓)
Full attributed analysis
|
0ax1
force-pushed
the
ad/clean-up-inline
branch
from
d3bb4ef to
5be4b29
Compare
Contributor
Benchmarks: TPC-DS SF=1 on NVMEVerdict: No clear signal (low confidence) datafusion / vortex-file-compressed (0.939x ➖, 23↑ 0↓)
datafusion / vortex-compact (1.005x ➖, 0↑ 0↓)
datafusion / parquet (0.903x ➖, 40↑ 1↓)
duckdb / vortex-file-compressed (1.019x ➖, 0↑ 1↓)
duckdb / vortex-compact (1.022x ➖, 1↑ 4↓)
duckdb / parquet (1.002x ➖, 0↑ 1↓)
duckdb / duckdb (1.032x ➖, 0↑ 10↓)
Full attributed analysis
|
Contributor
Author
Yeah, I'll double check any regressions for sure. |
Contributor
Benchmarks: TPC-H SF=10 on NVMEVerdict: No clear signal (low confidence) datafusion / vortex-file-compressed (0.976x ➖, 0↑ 0↓)
datafusion / vortex-compact (0.939x ➖, 0↑ 0↓)
datafusion / parquet (0.925x ➖, 8↑ 0↓)
datafusion / arrow (0.941x ➖, 6↑ 0↓)
duckdb / vortex-file-compressed (0.914x ➖, 5↑ 0↓)
duckdb / vortex-compact (0.979x ➖, 1↑ 0↓)
duckdb / parquet (0.944x ➖, 0↑ 0↓)
duckdb / duckdb (0.998x ➖, 0↑ 0↓)
Full attributed analysis
|
Contributor
Benchmarks: TPC-H SF=1 on S3Verdict: No clear signal (environment too noisy confidence) datafusion / vortex-file-compressed (0.908x ➖, 5↑ 1↓)
datafusion / vortex-compact (1.208x ➖, 0↑ 7↓)
datafusion / parquet (1.041x ➖, 0↑ 1↓)
duckdb / vortex-file-compressed (1.049x ➖, 0↑ 2↓)
duckdb / vortex-compact (0.980x ➖, 0↑ 0↓)
duckdb / parquet (1.064x ➖, 0↑ 0↓)
Full attributed analysis
|
Contributor
Benchmarks: FineWeb S3Verdict: No clear signal (low confidence) datafusion / vortex-file-compressed (1.054x ➖, 0↑ 1↓)
datafusion / vortex-compact (1.023x ➖, 0↑ 1↓)
datafusion / parquet (1.050x ➖, 0↑ 0↓)
duckdb / vortex-file-compressed (0.939x ➖, 0↑ 0↓)
duckdb / vortex-compact (0.987x ➖, 0↑ 0↓)
duckdb / parquet (0.997x ➖, 0↑ 0↓)
Full attributed analysis
|
Contributor
Benchmarks: Random AccessVortex (geomean): 0.931x ➖ unknown / unknown (1.001x ➖, 6↑ 15↓)
|
Contributor
Benchmarks: Statistical and Population GeneticsVerdict: No clear signal (low confidence) duckdb / vortex-file-compressed (1.017x ➖, 0↑ 0↓)
duckdb / vortex-compact (0.980x ➖, 1↑ 0↓)
duckdb / parquet (1.008x ➖, 0↑ 0↓)
Full attributed analysis
|
Contributor
Benchmarks: TPC-H SF=10 on S3Verdict: No clear signal (environment too noisy confidence) datafusion / vortex-file-compressed (0.933x ➖, 0↑ 0↓)
datafusion / vortex-compact (0.995x ➖, 0↑ 2↓)
datafusion / parquet (1.078x ➖, 0↑ 3↓)
duckdb / vortex-file-compressed (0.985x ➖, 1↑ 0↓)
duckdb / vortex-compact (0.963x ➖, 0↑ 0↓)
duckdb / parquet (0.989x ➖, 0↑ 1↓)
Full attributed analysis
|
Contributor
Benchmarks: Clickbench on NVMEVerdict: No clear signal (environment too noisy confidence) datafusion / vortex-file-compressed (0.883x ✅, 21↑ 0↓)
datafusion / parquet (0.898x ✅, 17↑ 0↓)
duckdb / vortex-file-compressed (0.953x ➖, 6↑ 1↓)
duckdb / parquet (0.976x ➖, 2↑ 0↓)
duckdb / duckdb (0.980x ➖, 2↑ 0↓)
Full attributed analysis
|
Contributor
Benchmarks: CompressionVortex (geomean): 1.012x ➖ unknown / unknown (1.028x ➖, 0↑ 13↓)
|
0ax1
force-pushed
the
ad/clean-up-inline
branch
3 times, most recently
from
6b439f7 to
1ffe423
Compare
github-actions
bot
removed
the
action/benchmark-sql
0ax1
force-pushed
the
ad/clean-up-inline
branch
2 times, most recently
from
7647ccd to
6f17616
Compare
Remove #[inline] from private, pub(crate), and inner functions where the compiler already has full visibility, and from large (>20 line) functions where code bloat outweighs benefit. Add #[inline] to BitBuffer operator trait impls (BitOr, BitAnd, Not, BitXor) — non-generic impls on a concrete type that are otherwise opaque cross-crate with LTO disabled. Signed-off-by: Alexander Droste <alexander.droste@protonmail.com>
0ax1
force-pushed
the
ad/clean-up-inline
branch
from
6f17616 to
7e978b8
Compare
github-actions
bot
removed
the
action/benchmark
Contributor
Author
|
@connortsui20 I think the current regressions are flaky benchmarks. If you look at the benchmark dashboards you can see how they oscillate back and forth 10% over time. |
0ax1
changed the title
#[inline] for non-generic & public fns
0ax1
changed the title
#[inline] annotations
dimitarvdimitrov
pushed a commit
that referenced
this pull request
Mar 20, 2026
Remove #[inline] from private, pub(crate), and inner functions where the compiler already has full visibility, and from large (>20 line) functions where code bloat outweighs benefit. Add #[inline] to BitBuffer operator trait impls (BitOr, BitAnd, Not, BitXor) — non-generic impls on a concrete type that are otherwise opaque cross-crate with LTO disabled. Signed-off-by: Alexander Droste <alexander.droste@protonmail.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
2 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.
Remove
#[inline]from private, pub(crate), and inner functions where the compiler already has full visibility, and from large (>20 line) functions.Add
#[inline]toBitBufferoperator trait impls (BitOr, BitAnd, Not, BitXor) - non-generic impls on a concrete type that are otherwise opaque cross-crate with LTO disabled.