mirror of
https://github.com/ROCm/composable_kernel.git
synced 2026-04-19 22:39:03 +00:00
* [FMHA] Enable page size 16 for batch prefill kernel * Refactor batch prefill KV offset logic to simplify template arguments - Remove redundant `kLog2PageSize` and `kIsVTileFitsInPage` from template args. - Add static assert to forbid `page_size=1` with vectorized layout.