Skip to content

Add SVE2 Reorder32bit optimization - #381

Merged
ermig1979 merged 1 commit into
devfrom
cursor/sve2-reorder32bit-c87a
Jun 9, 2026
Merged

Add SVE2 Reorder32bit optimization#381
ermig1979 merged 1 commit into
devfrom
cursor/sve2-reorder32bit-c87a

Conversation

@ermig1979

@ermig1979 ermig1979 commented Jun 8, 2026

Copy link
Copy Markdown
Owner

Summary

  • Add Sve2::Reorder32bit using SVE2 byte-reversal loads/stores with predicated tail handling.
  • Route SimdReorder32bit through the SVE2 implementation when available.
  • Add SVE2 Reorder32bit auto-test coverage and update 7.2.163 release notes.

Testing

  • cmake ./prj/cmake -B build -DCMAKE_BUILD_TYPE=Release -DCMAKE_CXX_COMPILER=g++ -DCMAKE_C_COMPILER=gcc -DSIMD_TOOLCHAIN="g++" -DSIMD_TARGET="" -DSIMD_AVX512VNNI=ON -DSIMD_AMXBF16=ON -DSIMD_TEST_FLAGS="-march=native" -DSIMD_SHARED=ON
  • cmake --build build --parallel$(nproc)
  • export LD_LIBRARY_PATH="$(pwd)/build:${LD_LIBRARY_PATH}" && ./Test "-r=.." -fi=Reorder32bit -tt=1 -ts=1
  • aarch64-linux-gnu-g++ -march=armv8.2-a+sve2 -std=c++17 -fsyntax-only -I./src -I./build src/Simd/SimdSve2Reorder.cpp
  • aarch64-linux-gnu-g++ -march=armv8.2-a+sve2 -std=c++17 -O2 -c -I./src -I./build src/Simd/SimdSve2Reorder.cpp -o /tmp/SimdSve2Reorder.o
Open in Web Open in Cursor 

Co-authored-by: Ihar Yermalayeu <ermig1979@gmail.com>
@ermig1979
ermig1979 marked this pull request as ready for review June 9, 2026 06:12
@ermig1979
ermig1979 merged commit 8dc9867 into dev Jun 9, 2026
1 check passed
@ermig1979
ermig1979 deleted the cursor/sve2-reorder32bit-c87a branch July 1, 2026 11:51
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants