trunk/7fe528830802c15f416786ae4c9947ac925a018a: Stop instantiating unreachable cat kernels (#194184)

发生了什么

parallel_cat cleared isAligned and isInOutAligned at runtime under USE_ROCM , because CatArrayBatchedCopy_contig is faster there ( #118685 ). That makes CatArrayBatchedCopy_alignedK_contig and CatArrayBatchedCopy_vectorized unreachable on ROCm, but a runtime flag cannot drop them: __global__ functions are DCE roots, so both families are still instantiated and emitted. A constexpr bool does, as in SoftMax.cu:1109 . T…

摘要按规则整理自下方来源原文

来源