Shuffle fix for gfx950 (#3491)

* solve compiler issue

* solve the gfx950 mfma shuffle regression

* refactor jenkinsfile to handle arch name better

* [CK TILE] set divisor to count of thread along k dimension

* fix the compiler error

* solve degradation

* Finish the multiplies fix

* fix the scales

* solve compilation error

* solve the composes

* solve the error of tile sweeper

* fix the test and example

* fix for gfx950

---------

Co-authored-by: Max Podkorytov <4273004+tenpercent@users.noreply.github.com>
Co-authored-by: illsilin_amdeng <Illia.Silin@amd.com>
Co-authored-by: Cong Ma <congma13@amd.com>
This commit is contained in:
Thomas Ning
2026-01-14 01:21:29 +08:00
committed by GitHub
parent 9908a87c31
commit 00c46785a8
33 changed files with 161 additions and 152 deletions

View File

@@ -1351,8 +1351,8 @@ fwd_result fmha_fwd_run(mode_enum mode,
auto oacc_element_func = [&]() {
if constexpr(std::is_same_v<ODataType, ck_tile::fp8_t> && supports_qscale)
return ck_tile::composes(ck_tile::saturates<ck_tile::fp8_t>{},
ck_tile::scales{scale_o_host});
return ck_tile::make_composes(ck_tile::saturates<ck_tile::fp8_t>{},
ck_tile::scales{scale_o_host});
else if constexpr(supports_qscale)
return ck_tile::scales{scale_o_host};
else