llama.cpp b11090:修复 CUDA sm_70 tile 编译错误
原标题:b11090
AI 摘要
llama.cpp 发布 b11090 版本,主要修复了 CUDA 后端在 sm_70(Volta 架构)上的 tile 编译错误。此前提交 1884824fd 引入的五参数 load_ldmatrix 仅定义了 tile<16,8>,导致 Volta 的 tile<8,4> 无法匹配;本次将 tile 形状泛化为 <I,J>,使非 swizzle 分支可转发到三参数加载器。该修复由 DeepSeek V4.1 Flash 辅助完成,并同步发布了覆盖 macOS、Linux、Windows、Android 等多平台多后端的预编译包。
正文节选
<details open> cuda: fix sm_70 tile compilation error (#29224) The 5-argument load_ldmatrix added in 1884824fd only defines tile<16,8>, so the Volta tile<8,4> does not match. See https://github.com/ggml-org/llama.cpp/issues/29222 for details. Building on 1884824fd, generalize the tile shape of the 5-argument load_ldmatrix from <16,8> to <I,J>, so the non-swizzle branch forwards to the 3-argument loader for any shape. Local compilation and testing passed. Assisted-by: DeepSeek V4.1 Flash (Open