Models & Labs
llama.cpp b9491 release addresses PDL race conditions
llama.cpp Releaseshigh confidence
Why it matters
- →Addresses compatibility issues with PDL, improving stability.
- →Retains performance on older architectures through new directives.
- →Simplifies code maintenance with macro-based 'restrict' usage.
The b9491 release of llama.cpp addresses PDL race conditions by removing 'restrict' from PDL kernel headers due to incompatibility issues. The update introduces preprocessor directives to retain performance on older architectures and simplifies 'restrict' usage via macros. This release also includes a fix for the PDL restrict issue on Hopper architectures. These changes enhance compatibility and performance across multiple platforms, benefiting developers using llama.cpp.
Read originalThe story around this
Earlier coverage that leads up to this article, and what followed. Lines connect each piece to the closest one after it, converging here.
What happened next
- 1llama.cpp b9691 Release Enhances Compiler Compatibility
llama.cpp Releases · June 18, 2026 · Same story
- 2llama.cpp b9950 release focuses on platform fixes
llama.cpp Releases · July 11, 2026 · Same story
More from llama.cpp Releases
Coding Toolscoding
llama.cpp b12022 adds ROCm 10 and Snapdragon support
This release quietly cements llama.cpp as the universal inference runtime by finally bringing ROCm 10.0 to Linux and Windows alongside CUDA 13.4, effectively closing the hardware gap for AMD users who were previously left behind. The inclusion of native Snapdragon support on Linux marks a significant step toward running local LLMs on ARM-based mobile chips without emulation overhead. While KleidiAI on Apple Silicon is disabled in this build, the broader expansion to diverse accelerators ensures that developers aren't locked into NVIDIA ecosystems anymore.
llama.cpp Releases
Coding Toolscoding
llama.cpp b12023 fixes shared prefix bug
This release addresses a specific inference artifact where decision prefixes were incorrectly bleeding into output embeddings. By capping shared prefixes at the first output position, llama.cpp ensures that option tokens remain distinct for parent and child tasks. This is a targeted fix for multi-step reasoning or voting workflows rather than a performance boost. It stabilizes complex generation patterns without changing the underlying model architecture.
