ROCm/composable_kernel
1
0
Fork 0
mirror of https://github.com/ROCm/composable_kernel.git synced 2026-04-20 06:49:15 +00:00
Code Issues Packages Projects Releases Wiki Activity

Extend XDL kernel to Support RDNA3/4 - Part 4 (#2724)

Browse Source 
* Fix example

* fix build error

* update pk_i4 & moe test case

* fix all instance build (examples)

* fix batched_gemm_gemm (example)

* disable example_gemm_bias_softmax_gemm_permute on gfx11

* remove unnecessary disable gfx11

* update tests

* update tests2
This commit is contained in:
linqunAMD
2025-09-12 23:17:07 +08:00
committed by GitHub
parent bca99a499d
commit 321627aec5
123 changed files with 848 additions and 574 deletions
Download Patch File Download Diff File Expand all files Collapse all files

8
include/ck/tensor_operation/gpu/thread/threadwise_tensor_slice_transfer_v7r3.hpp
View File

@@ -1,5 +1,5 @@
// SPDX-License-Identifier: MIT
// Copyright (c) 2018-2024, Advanced Micro Devices, Inc. All rights reserved.
// Copyright (c) 2018-2025, Advanced Micro Devices, Inc. All rights reserved.
#pragma once
@@ -48,8 +48,6 @@ struct ThreadwiseTensorSliceTransfer_v7r3
{
static constexpr auto I0 = Number<0>{};
static constexpr auto SrcScalarPerVector = SrcScalarPerVectors{}[I0];
static constexpr index_t nDim = SliceLengths::Size();
static constexpr index_t nSrc = SrcDescs::Size();
@@ -67,6 +65,10 @@ struct ThreadwiseTensorSliceTransfer_v7r3
Number<Descs::Size()>{});
}
static constexpr auto SrcScalarPerVector =
reduce_on_sequence(SrcScalarPerVectors{},
math::minimize<index_t>{},
Number<1>{}); // GetMinSrcScalarPerVector(); SrcScalarPerVectors{}[I0];
using SrcCoords = decltype(MakeCoordinates(SrcDescs{}, StaticallyIndexedArray<Index, nSrc>{}));
using DstCoords = decltype(MakeCoordinates(DstDescs{}, StaticallyIndexedArray<Index, nDst>{}));