Linjun-AMD 08792e0b31 [rocm-libraries] ROCm/rocm-libraries#5504 (commit 47f86c7) ...

[CK Tile] Add sink token gradient support in FMHA backward
 pass (#5504)

## Motivation

Adds sink token support to the FMHA backward kernel (dot_do_o pipeline):

## Technical Details

- Extend BlockFmhaBwdOGradDotOPipelineProblem with LSEDataType
- Add sink_ptr/d_sink_ptr/lse_ptr/nhead to FmhaBwdOGradDotOCommonKargs
- Compute per-head sink gradient via atomic accumulation in the pipeline
- Update example runner with reference validation for sink gradient

## Test Plan

Add new test case

## Test Result

WIP

## Submission Checklist

- [ ] Look over the contributing guidelines at
https://github.com/ROCm/ROCm/blob/develop/CONTRIBUTING.md#pull-requests.

2026-04-02 03:17:45 +00:00

..

batched_contraction

Implement batched gemm bias permute for RDNA4 (#3534)

2026-01-17 08:30:27 +01:00

batched_gemm

[rocm-libraries] ROCm/rocm-libraries#4415 (commit b3b4af7)

2026-02-25 23:23:02 +00:00

batched_gemm_b_scale

chore(copyright) update library wide CMakeLists.txt copyright header template (#3313)

2025-11-28 13:49:54 -08:00

batched_gemm_gemm

[rocm-libraries] ROCm/rocm-libraries#4415 (commit b3b4af7)

2026-02-25 23:23:02 +00:00

batched_gemm_multi_d

[CI, CK examples] Disable time_kernel for CI tests and examples (#3464)

2026-01-07 16:30:57 +01:00

batched_gemm_multiple_d_gemm_multiple_d

Implement batched gemm add relu gemm add for rdna4 (#3391)

2026-01-20 13:06:59 -08:00

batched_gemm_reduce

[rocm-libraries] ROCm/rocm-libraries#5921 (commit 032ac1b)

2026-03-30 14:20:20 +00:00

batched_gemm_softmax_gemm

chore(copyright) update library wide CMakeLists.txt copyright header template (#3313)

2025-11-28 13:49:54 -08:00

batched_gemm_softmax_gemm_permute

chore(copyright) update library wide CMakeLists.txt copyright header template (#3313)

2025-11-28 13:49:54 -08:00

batchnorm

chore(copyright) update library wide CMakeLists.txt copyright header template (#3313)

2025-11-28 13:49:54 -08:00

block_swizzle_test

chore(copyright): update copyright header for test directory

2025-11-19 17:43:28 -07:00

block_to_ctile_map

chore(copyright) update library wide CMakeLists.txt copyright header template (#3313)

2025-11-28 13:49:54 -08:00

ck_tile

[rocm-libraries] ROCm/rocm-libraries#5504 (commit 47f86c7)

2026-04-02 03:17:45 +00:00

common

Add grouped convnd dataset tests for bwd_data, bwd_weight and make them parallel (#3380)

2025-12-15 13:38:25 +01:00

contraction

Add support to fp16 + compute fp16 and bf16 + compute bf16 contractions (#3598)

2026-01-20 09:39:57 -08:00

conv_tensor_rearrange

chore(copyright) update library wide CMakeLists.txt copyright header template (#3313)

2025-11-28 13:49:54 -08:00

conv_util

chore(copyright) update library wide CMakeLists.txt copyright header template (#3313)

2025-11-28 13:49:54 -08:00

convnd_bwd_data

[CK tests] Extend conv GPU reference (#3539)

2026-01-27 09:49:42 +01:00

convnd_fwd

[CK tests] Extend conv GPU reference (#3539)

2026-01-27 09:49:42 +01:00

data_type

chore(copyright) update library wide CMakeLists.txt copyright header template (#3313)

2025-11-28 13:49:54 -08:00

device_memory

[CK_BUILDER] Integrate CKB validation with CK verification (#3649)

2026-01-28 17:41:02 +01:00

elementwise_normalization

chore(copyright) update library wide CMakeLists.txt copyright header template (#3313)

2025-11-28 13:49:54 -08:00

gemm

[CI, CK examples] Disable time_kernel for CI tests and examples (#3464)

2026-01-07 16:30:57 +01:00

gemm_ab_scale

Wmma support for gemm_ab_scale (#3314)

2025-12-11 09:06:20 +01:00

gemm_add

[rocm-libraries] ROCm/rocm-libraries#4415 (commit b3b4af7)

2026-02-25 23:23:02 +00:00

gemm_b_scale

[rocm-libraries] ROCm/rocm-libraries#4415 (commit b3b4af7)

2026-02-25 23:23:02 +00:00

gemm_bias_add_reduce

Wmma support for gemm_bias_add_reduce (#3316)

2026-01-07 10:27:16 -08:00

gemm_blockscale_wp

[rocm-libraries] ROCm/rocm-libraries#4372 (commit 738ffd7)

2026-02-07 00:09:58 +00:00

gemm_layernorm

chore(copyright) update library wide CMakeLists.txt copyright header template (#3313)

2025-11-28 13:49:54 -08:00

gemm_multi_abd

[rocm-libraries] ROCm/rocm-libraries#4415 (commit b3b4af7)

2026-02-25 23:23:02 +00:00

gemm_multiply_multiply_wp

Wmma support for gemm_multiply_multiply_wp (#3278)

2025-12-03 07:38:23 -08:00

gemm_mx

chore(copyright) update library wide CMakeLists.txt copyright header template (#3313)

2025-11-28 13:49:54 -08:00

gemm_reduce

chore(copyright) update library wide CMakeLists.txt copyright header template (#3313)

2025-11-28 13:49:54 -08:00

gemm_split_k

chore(copyright) update library wide CMakeLists.txt copyright header template (#3313)

2025-11-28 13:49:54 -08:00

gemm_universal

[rocm-libraries] ROCm/rocm-libraries#4415 (commit b3b4af7)

2026-02-25 23:23:02 +00:00

gemm_universal_preshuffle

Implement device_gemm_universal_preshuffle_instance for RDNA4 (#3429)

2026-01-15 07:19:31 -08:00

gemm_universal_reduce

chore(copyright) update library wide CMakeLists.txt copyright header template (#3313)

2025-11-28 13:49:54 -08:00

gemm_universal_streamk

chore(copyright) update library wide CMakeLists.txt copyright header template (#3313)

2025-11-28 13:49:54 -08:00

gpu_reference

[CK tests] Extend conv GPU reference (#3539)

2026-01-27 09:49:42 +01:00

gpu_verification

[CK_BUILDER] Integrate CKB validation with CK verification (#3649)

2026-01-28 17:41:02 +01:00

grouped_convnd_bwd_data

[rocm-libraries] ROCm/rocm-libraries#5856 (commit 2d9a0a1)

2026-03-27 03:58:37 +00:00

grouped_convnd_bwd_weight

[rocm-libraries] ROCm/rocm-libraries#4872 (commit ca623f7)

2026-02-25 20:11:01 +00:00

grouped_convnd_fwd

[rocm-libraries] ROCm/rocm-libraries#4407 (commit adde219)

2026-02-11 13:43:01 +00:00

grouped_convnd_fwd_activation

[rocm-libraries] ROCm/rocm-libraries#5856 (commit 2d9a0a1)

2026-03-27 03:58:37 +00:00

grouped_gemm

[rocm-libraries] ROCm/rocm-libraries#5921 (commit 032ac1b)

2026-03-30 14:20:20 +00:00

grouped_gemm_tile_loop

Implement grouped gemm tile loop for RDNA4 (#3304)

2026-01-13 07:14:23 +01:00

magic_number_division

chore(copyright) update library wide CMakeLists.txt copyright header template (#3313)

2025-11-28 13:49:54 -08:00

mx_mfma_op

chore(copyright) update library wide CMakeLists.txt copyright header template (#3313)

2025-11-28 13:49:54 -08:00

normalization_bwd_data

chore(copyright) update library wide CMakeLists.txt copyright header template (#3313)

2025-11-28 13:49:54 -08:00

normalization_bwd_gamma_beta

chore(copyright) update library wide CMakeLists.txt copyright header template (#3313)

2025-11-28 13:49:54 -08:00

normalization_fwd

chore(copyright) update library wide CMakeLists.txt copyright header template (#3313)

2025-11-28 13:49:54 -08:00

permute_scale

chore(copyright) update library wide CMakeLists.txt copyright header template (#3313)

2025-11-28 13:49:54 -08:00

pool

chore(copyright) update library wide CMakeLists.txt copyright header template (#3313)

2025-11-28 13:49:54 -08:00

position_embedding

[Compiler] Addressing new compiler warnings (#3640)

2026-02-02 09:39:48 -08:00

quantization

chore(copyright) update library wide CMakeLists.txt copyright header template (#3313)

2025-11-28 13:49:54 -08:00

reduce

fix some minor error (#3409)

2025-12-16 19:50:49 -08:00

reference_conv_fwd

chore(copyright) update library wide CMakeLists.txt copyright header template (#3313)

2025-11-28 13:49:54 -08:00

s_prefetch_op

chore(copyright) update library wide CMakeLists.txt copyright header template (#3313)

2025-11-28 13:49:54 -08:00

scatter_gather

chore(copyright) update library wide CMakeLists.txt copyright header template (#3313)

2025-11-28 13:49:54 -08:00

smfmac_op

chore(copyright) update library wide CMakeLists.txt copyright header template (#3313)

2025-11-28 13:49:54 -08:00

softmax

chore(copyright) update library wide CMakeLists.txt copyright header template (#3313)

2025-11-28 13:49:54 -08:00

space_filling_curve

chore(copyright) update library wide CMakeLists.txt copyright header template (#3313)

2025-11-28 13:49:54 -08:00

threadwise_transfer_helper

[rocm-libraries] ROCm/rocm-libraries#4673 (commit ec385da)

2026-03-06 16:27:59 +00:00

transpose

chore(copyright) update library wide CMakeLists.txt copyright header template (#3313)

2025-11-28 13:49:54 -08:00

util

[rocm-libraries] ROCm/rocm-libraries#4828 (commit 7de19bb)

2026-02-28 20:11:11 +00:00

wmma_op

chore(copyright) update library wide CMakeLists.txt copyright header template (#3313)

2025-11-28 13:49:54 -08:00

wrapper

[CI, CK examples] Disable time_kernel for CI tests and examples (#3464)

2026-01-07 16:30:57 +01:00

CMakeLists.txt

[rocm-libraries] ROCm/rocm-libraries#4368 (commit 17f7dfc)

2026-03-11 10:00:52 +00:00