composable_kernel

mirror of https://github.com/ROCm/composable_kernel.git synced 2026-04-19 22:39:03 +00:00

Files

Jeff Huang 86dd59cd01 [CK_TILE] Add sequence padding and variable length support in fmha (a… (#2851 )

* [CK_TILE] Add sequence padding and variable length support in fmha (and v3)

 - Group Mode Padding: Introduces the `-s_qpad` argument to support
   physically padded layouts. Kernels now use padded start pointers
   (`seqstart_padded_*_ptr`) for memory addressing.

 - Batch Mode Variable Length: Adds `-q_eff_lens` and `-kv_eff_lens`
   arguments for efficient processing of variable-length sequences by
   passing cumulative effective lengths (`cu_seqlen_*_ptr`) to the kernel.

 - FMHA examples: Support padding and variable length both in
   group and batch mode. Dispatcher is updated as well (dispatch to
   kPadSeqLenK enabled pipeline).

 - New padding test cases: Add padding test cases to `smoke_test_fwd.sh`,
   and add benchmarks to `benchmark_fwd.sh` and `benchmark_fwd_v3.sh` as well.
   These test cases and benchmarks that specifically validate/benchmark the
   new padding and variable-length functionalities in both group and batch modes.

* [CK_TILE] Fix build error in fmha unit tests

---------

Co-authored-by: Po Yen Chen <PoYen.Chen@amd.com>
Co-authored-by: Yi DING <yi.ding@amd.com>

2025-09-19 17:36:49 +08:00

fmha_batch_prefill_kernel.hpp

[CK_TILE] Fix fmha_fwd_v3() Default2DEpilogue usage (#2765 )

2025-09-02 09:51:56 -07:00

fmha_bwd_kernel.hpp

[CK_TILE] Fix kname & typo in FMHA BWD (#2809 )

2025-09-09 15:08:00 -07:00

fmha_fwd_appendkv_kernel.hpp

[CK_TILE] FMHA BWD Fix Compilation with Bias (#2682 )

2025-08-22 10:01:10 +08:00

fmha_fwd_appendkv_tile_partitioner.hpp

[CK_TILE] Add PagedAttention kernels (#1387 )