bitlen: performance guesswork#1298
Merged
Merged
Conversation
This is an attempt to address #1296, although unfortunately it doesn't have a self-contained benchmark, and when I tried to run what's available it uses FFI and needs a bunch of external dependencies. I didn't futz with any of that and just tried throwing everything at the wall in this PR hoping one of the changes addresses the problem. First, it sprinkles `inline(always)` on everything. While that's generally an antipattern, these functions are both deeply nested and incredibly trivial: usually just one or two wrapping arithmetic ops in release builds, and they're deep in hot paths, so really they need to get aggressively inlined. Speaking of that, this also completely ditches checks/assertions in release builds for performance, whereas before they were always compiled in. That should keep the hot path incredibly simple. None of this had any major impact on `crypto-bigint`'s own benchmarks. The only change that seemed to be barely escaping the noise threshold was adding `#[inline(always)]` to `UintRef::bits_precision` which seemed to eek out some ~1.8% cross-cutting performance improvements, though I didn't check to see if they're reproducible and that's pretty much within the noise threshold so don't quote me on that. Hopefully one of these changes addresses the situation in #1296.
Codecov Report❌ Patch coverage is
Additional details and impacted files@@ Coverage Diff @@
## master #1298 +/- ##
==========================================
- Coverage 91.31% 91.31% -0.01%
==========================================
Files 190 190
Lines 22544 22546 +2
==========================================
+ Hits 20587 20588 +1
- Misses 1957 1958 +1 ☔ View full report in Codecov by Harness. 🚀 New features to boost your workflow:
|
Merged
tarcieri
added a commit
that referenced
this pull request
Jun 22, 2026
## Added
- `BoxedUint::{from_be_slice_truncated, from_le_slice_truncated}` (#1266)
- `Encoding::{from_bytes, to_bytes}` (#1297)
- `Uint::{from_be_slice_truncated, from_le_slice_truncated}` (#1299)
- `Limb::{from_*_bytes, from_*_slice_truncated, to_*_bytes}` as
`const fn`s (#1300)
- `Encoding::{from_be_slice_truncated, from_le_slice_truncated}` trait
methods (#1301)
- `ArrayEncoding::{from_byte_array, to_byte_array}` (#1302)
## Fixed
- Preserve `NonZero` and `Odd` invariants in `Zeroize` impls (#1287)
- `bitlen` performance (#1298)
- `floor_sqrt` regression (#1304)
- Truncated Karatsuba carry (#1305)
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.
This is an attempt to address #1296, although unfortunately it doesn't have a self-contained benchmark, and when I tried to run what's available it uses FFI and needs a bunch of external dependencies. I didn't futz with any of that and just tried throwing everything at the wall in this PR hoping one of the changes addresses the problem.
First, it sprinkles
inline(always)on everything. While that's generally an antipattern, these functions are both deeply nested and incredibly trivial: usually just one or two wrapping arithmetic ops in release builds, and they're deep in hot paths, so really they need to get aggressively inlined.Speaking of that, this also completely ditches checks/assertions in release builds for performance, whereas before they were always compiled in. That should keep the hot path incredibly simple.
None of this had any major impact on
crypto-bigint's own benchmarks. The only change that seemed to be barely escaping the noise threshold was adding#[inline(always)]toUintRef::bits_precisionwhich seemed to eek out some ~1.8% cross-cutting performance improvements, though I didn't check to see if they're reproducible and that's pretty much within the noise threshold so don't quote me on that.Hopefully one of these changes addresses the situation in #1296.