Repository navigation
AArch64: Improve sequence decoding - #4813
Open
davidbolvansky wants to merge 2 commits into
Open
davidbolvansky wants to merge 2 commits into
davidbolvansky wants to merge 2 commits into
Conversation
Assisted-by: OpenAI Codex
Compute FSE bit positions from a common starting position to avoid serial bitsConsumed updates and retain GCC throughput with the new length-read layout. Assisted-by: OpenAI Codex
davidbolvansky
changed the title
pps83
pushed a commit
to pps83/zstd
that referenced
this pull request
Oct 6, 2026
The windows-11-arm runner label now resolves to the windows-11-vs2026-arm64 image (20260924), which ships Visual Studio 2026 and CMake 4.4, but no longer Visual Studio 2022. CMake fails at configure time with "could not find any instance of Visual Studio", so the job fails on every PR opened since 2026-10-01 (e.g. meta#4806, meta#4811, meta#4813). Switch that matrix entry to the "Visual Studio 18 2026" generator. The x64 and Win32 entries run on windows-2022 and are unaffected. Drop "VS2022" from the job name template, which would now be wrong for the ARM64 entry: jobs become "CMake Windows (<entry>)". No branch protection rule requires these checks by name.
This branch has not been deployed
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Improve Apple Clang 21 decompression throughput on Apple M1 Pro by 11.3% on a 3.2 MB bundle of actual zstd C/header files, representative of compressed source archives, and 12.9% on synthetic JSON records (level-1 inputs; median of five runs). GCC 16 throughput stays within 0.4% of baseline on both inputs.
Use one shared AArch64 implementation: separate the uncommon additional literal-length-bit case, calculate the three FSE bit positions from one starting position, and update
bitsConsumedonce. Also extend the existing 64-bit symbol-load idiom to Clang.Assisted-by: OpenAI Codex