b10167
ggml-org/llama.cppb10167Jul 28, 2026by github-actions[bot]same day as b10166
What changed since b10166
b10167 is a single internal server refactor that routes llama_memory calls through a shared common_memory abstraction.
Improvements
- Server memory-management calls now go through a common_memory layer instead of calling llama_memory directly.
- The shared abstraction reduces duplication and should make future memory-related server changes easier to reuse.
Upgrade notes
This is a pure internal refactor of the server's memory handling, with no config, API, or CLI changes called out. A drop-in replacement of b10166 with b10167 should be safe; smoke-test server sessions if you rely on custom memory or context-shift behavior.
Analyzed Nov 1, 2026
Upgrading?
Release notes summary
Refactored memory management calls by abstracting `llama_memory` to a common `common_memory` function to improve code structure and potential reuse.
Key Highlights
- Abstracted llama_memory calls to common_memory
- Improved code architecture and maintainability
Release Notes (excerpt)
<details open> server: abstract llama_memory calls to common_memory (#26221) </details> **Website:** - <https://llama.app> **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:**…
Published by ggml-org on GitHub.
Read the full release notesRelated Repositories
:electron: Build cross-platform desktop apps with JavaScript, HTML, and CSS
Godot Engine – Multi-platform 2D and 3D game engine
ClickHouse® is a real-time analytics database management system
Self-hosted game stream host for Moonlight.
DuckDB is a high-performance analytical database system