llama.cpp b12102：ggml 视图按 stride 实际字节做边界检查
原标题：b12102
AI 摘要
llama.cpp 发布 b12102 版本，核心改动在 ggml 后端：视图（view）的边界检查改为按 stride 实际触及的字节数计算，修复了行重叠视图被误判、宽 stride 视图越界却通过检查的问题；同时要求带 stride 的视图起始元素在 dim 0 上连续。delta_net 的卷积窗口改为一次性按全部行创建。该版本同步提供 macOS、Linux、Windows、Android 等多平台及 CUDA、Vulkan、ROCm、SYCL、OpenVINO 等后端构建包。
正文节选
<details open> ggml: bound views by the bytes their strides reach (#31584) * ggml: bound views by the bytes their strides reach A view was checked against its source with the size of a contiguous tensor of the same shape, computed before its strides were set. Views whose rows overlap failed that check although they stay inside their source, and views with wider strides passed it while reaching past the end. The view functions now pass their strides to the tensor constructor, which checks the