Change8

b8061

Breaking Changes
📦 llama-cppView on GitHub →
2 breaking🐛 1 fixes🔧 1 symbols

Summary

This release primarily addresses a build issue related to the KleidiAI backend when compiling multiple CPU backends by correcting the use of CMake's FetchContent functions. It also provides updated pre-compiled binaries for numerous platforms.

⚠️ Breaking Changes

  • The build process for KleidiAI backend when building multiple cpu backends is fixed. Previously, FetchContent_Populate was used instead of FetchContent_MakeAvailable, causing issues because FetchContent_Populate is not idempotent.
  • If you were relying on the previous behavior of FetchContent_Populate in multi-CPU backend builds, you might see changes in how dependencies are handled during the build.

Migration Steps

  1. No specific migration steps are required unless you were directly interacting with the CMake logic that was modified (commit 3a00c98584e42a20675b6569d81beadb282b0952).

🐛 Bug Fixes

  • Fixed a build issue with the KleidiAI backend when building multiple cpu backends by ensuring FetchContent_MakeAvailable is used instead of FetchContent_Populate where appropriate.

Affected Symbols