b9916

ggml-org/llama.cppb9916Jul 8, 2026by github-actions[bot]same day as b9915

What changed since b9915

low risk

b9916 is a single bug-fix release correcting A-matrix indexing in the simd_gemm scalar tail-column path.

Bug fixes

  • Corrected an incorrect A-matrix index in simd_gemm's scalar tail-column handling for full row blocks, which could produce wrong GEMM results in that edge case.

Upgrade notes

This is a drop-in patch release with no API, config, or build changes; simply rebuild or swap binaries. Upgrading is recommended if you rely on CPU SIMD GEMM paths where full row blocks with tail columns occur, since prior versions could yield incorrect results there.

Analyzed Oct 30, 2026

Upgrading?

Release notes summary

Fixes incorrect A-matrix indexing in the SIMD GEMM scalar tail-column path for full row blocks.

Key Highlights

  • Fix for A-matrix indexing in simd_gemm scalar tail-column path
  • Correction for full row blocks

Release Notes (excerpt)

<details open> ggml : fix A indexing in simd_gemm scalar tail-column path (#25390) `simd_gemm()` has an incorrect A-matrix index in the scalar tail-column path for full row blocks. </details> **macOS/iOS:** - macOS Apple Silicon (arm64) - macOS Apple Silicon (arm64, KleidiAI enabled) DISABLED - macOS Intel (x64) - iOS XCFramework **Linux:** - Ubuntu x64 (CPU) - Ubuntu arm64 (CPU) - Ubuntu s390x (CPU) - Ubuntu x64 (Vulkan) - Ubuntu arm64 (Vulkan) - Ubuntu x64 (ROCm 7.2)…

Published by ggml-org on GitHub.

Read the full release notes

Related Repositories