b9924

ggml-org/llama.cppb9924Jul 8, 2026by github-actions[bot]same day as b9923

What changed since b9923

low risk

b9924 is a narrow internal refactor of fused operations in llama, with no user-facing changes noted versus b9923.

Improvements

  • Refactored the fused operations implementation in llama for cleaner internal structure.
  • Release notes indicate no behavioral or API changes beyond the fused-ops refactor.

Upgrade notes

This is a small internal refactor of fused ops; no config, API, or migration steps are indicated. Standard rebuild and smoke-test of inference paths is sufficient. The prior b9923 server-stream PImpl and mutex changes remain in effect.

Analyzed Oct 30, 2026

Upgrading?

Release notes summary

Refactors the implementation of fused operations.

Key Highlights

  • Refactoring of fused operations

Release Notes (excerpt)

<details open> llama: refactor fused ops (#24646) </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) - Ubuntu x64 (OpenVINO) - Ubuntu x64 (SYCL FP32) - Ubuntu x64 (SYCL FP16) **Android:** - Android arm64 (CPU) **Windows:** - Windows x64 (CPU)…

Published by ggml-org on GitHub.

Read the full release notes

Related Repositories