ROCm/composable_kernel
1
0
Fork 0
mirror of https://github.com/ROCm/composable_kernel.git synced 2026-04-19 22:39:03 +00:00
Code Issues Packages Projects Releases Wiki Activity

[rocm-libraries] ROCm/rocm-libraries#4313 (commit 080ac66)

Browse Source 
[CK] Fix gptoss sink

## Motivation

This PR removes conditional logic for handling infinity values in the
sink mechanism across multiple FMHA pipeline implementations, defaulting
sink_size to 0 and adding a constraint in the kernel selection logic.

## Technical Details

Changes:

Removed __builtin_isinf_sign(sink_v) checks and conditional
initialization of LSE accumulators across 7 pipeline files
Added default initialization (= 0) for sink_size in 4 argument structs
Added F_sink == "f" constraint to kernel compatibility checking

## Test Plan

Local test

## Test Result

passed

## Submission Checklist

- [x] Look over the contributing guidelines at
https://github.com/ROCm/ROCm/blob/develop/CONTRIBUTING.md#pull-requests.
This commit is contained in:
Linjun-AMD
2026-03-02 01:54:46 +00:00
committed by assistant-librarian[bot]
parent d32d515f64
commit 78ae3835a6
7 changed files with 8 additions and 70 deletions
Download Patch File Download Diff File Expand all files Collapse all files

1
example/ck_tile/01_fmha/codegen/ops/fmha_fwd.py
View File

@@ -1254,6 +1254,7 @@ def get_product(receipt: int) -> Product:
cond &= kernel_ctx.pipeline.F_bias in ["no", "alibi"]
cond &= kernel_ctx.pipeline.F_qscale == "no"
cond &= kernel_ctx.pipeline.F_skip == "f"
cond &= kernel_ctx.pipeline.F_sink == "f"
return cond
return Product(name="Flash attention integration", rule=fit)