b9877

ggml-org/llama.cppb9877Jul 5, 2026by github-actions[bot]same day as b9876

What changed since b9876

medium risk

b9877 adds a hard abort when a multi-buffer configuration is detected, on top of b9876's tensor-parallel MoE crash fix.

Breaking changes

  • Execution now aborts if a multi-buffer setup is detected, so configurations that previously ran (possibly with undefined behavior) will terminate instead.

Improvements

  • Adds a defensive safety check around backend buffer handling to catch multi-buffer states early rather than letting them cause undefined behavior.

Upgrade notes

The only change relative to b9876 is the new multi-buffer abort; the tensor-parallel plus -ncmoe crash fix from b9876 is retained. If you use custom or multi-backend buffer setups, test before rolling out, since a detected multi-buffer now terminates the process rather than continuing. No config or migration steps are described.

Analyzed Oct 31, 2026

Upgrading?

Release notes summary

This release introduces a safety check that aborts the program execution if a multi-buffer configuration is detected, likely to prevent undefined behavior or crashes in specific backend scenarios.

Key Highlights

  • Abort execution on multi-buffer detection
  • Safety check for backend buffer handling

Release Notes (excerpt)

<details open> abort if we see a multi buffer (#25276) </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