Skip to content

GitLab

Commit 1b616990 authored by aska-0096's avatar aska-0096
Browse files

Merge branch 'develop' of https://github.com/ROCm/composable_kernel into update_cka8w8_uc

parents af30d6b6 800cf897
Hide whitespace changes
Inline Side-by-side
Showing with 961 additions and 485 deletions
include/ck_tile/ops/gemm/pipeline/gemm_universal_pipeline_ag_bg_cr_policy.hpp
View file @ 1b616990
// SPDX-License-Identifier: MIT // SPDX-License-Identifier: MIT
// Copyright (c) 2024, Advanced Micro Devices, Inc. All rights reserved. // Copyright (c) 2024-2025, Advanced Micro Devices, Inc. All rights reserved.
#pragma once #pragma once
#include "ck_tile/core.hpp" #include "ck_tile/core.hpp"
#include "ck_tile/ops/gemm/warp/warp_gemm_dispatcher.hpp" #include "ck_tile/ops/gemm/warp/warp_gemm_dispatcher.hpp"
#include "ck_tile/ops/common/tensor_layout.hpp"
namespace ck_tile { namespace ck_tile {
// UniversalGemm Policy // UniversalGemm Policy
struct UniversalGemmPipelineAgBgCrPolicy struct UniversalGemmPipelineAgBgCrPolicy
{ {
static constexpr auto I0 = number<0>{}; static constexpr auto I0 = number<0>{};
static constexpr auto I1 = number<1>{}; static constexpr auto I1 = number<1>{};
static constexpr auto I2 = number<2>{}; static constexpr auto I2 = number<2>{};
static constexpr bool TransposeC = true; static constexpr auto ATileAccessPattern = tile_distribution_pattern::thread_raked;
static constexpr auto BTileAccessPattern = tile_distribution_pattern::thread_raked;
template <typename Problem, typename DataType, index_t MNPerBlock>
CK_TILE_HOST_DEVICE static constexpr auto GetVectorLoadSize() /**
* @brief Get the maximum global memory vector load size.
*
* @tparam Problem The UniversalGemmPipelineProblem object.
* @tparam DataType The tensor data type we're considering.
* @tparam MNPerBlock The MPerBlock or NPerBlock value depending on tensor (A/B).
* @tparam XPerTile The contiguous Tile dimension size.
* @return Maximum DRAM vector load size.
*/
template <typename Problem, typename DataType, index_t MNPerBlock, index_t XPerTile>
CK_TILE_HOST_DEVICE static constexpr auto GetGlobalVectorLoadSize()
{ {
constexpr index_t BlockSize = Problem::kBlockSize; constexpr index_t BlockSize = Problem::kBlockSize;
constexpr index_t KPerBlock = Problem::BlockGemmShape::kK; constexpr index_t KPerBlock = Problem::BlockGemmShape::kK;
constexpr index_t elements_per_thread = MNPerBlock * KPerBlock / BlockSize; constexpr index_t elements_per_thread = MNPerBlock * KPerBlock / BlockSize;
if constexpr(elements_per_thread % (16 / sizeof(DataType)) == 0) // Assume DataType is even!
if constexpr(XPerTile % (16 / sizeof(DataType)) == 0 &&
elements_per_thread % (16 / sizeof(DataType)) == 0)
{ {
return (16 / sizeof(DataType)); return (16 / sizeof(DataType));
} }
else if constexpr(elements_per_thread % (8 / sizeof(DataType)) == 0) else if constexpr(XPerTile % (8 / sizeof(DataType)) == 0 &&
elements_per_thread % (8 / sizeof(DataType)) == 0)
{ {
return (8 / sizeof(DataType)); return (8 / sizeof(DataType));
} }
else if constexpr(elements_per_thread % (4 / sizeof(DataType)) == 0 && else if constexpr(sizeof(DataType) >= 4 && XPerTile % (4 / sizeof(DataType)) == 0 &&
sizeof(DataType) >= 4) elements_per_thread % (4 / sizeof(DataType)) == 0)
{ {
return (4 / sizeof(DataType)); return (4 / sizeof(DataType));
} }
else if constexpr(elements_per_thread % (2 / sizeof(DataType)) == 0 && else if constexpr(sizeof(DataType) >= 2 && XPerTile % (2 / sizeof(DataType)) == 0 &&
sizeof(DataType) >= 2) elements_per_thread % (2 / sizeof(DataType)) == 0)
{ {
return (2 / sizeof(DataType)); return (2 / sizeof(DataType));
} }
...@@ -49,6 +62,126 @@ struct UniversalGemmPipelineAgBgCrPolicy ...@@ -49,6 +62,126 @@ struct UniversalGemmPipelineAgBgCrPolicy
} }
} }
template <typename Problem>
CK_TILE_HOST_DEVICE static constexpr auto GetVectorSizeA()
{
using ALayout = remove_cvref_t<typename Problem::ALayout>;
using ADataType = remove_cvref_t<typename Problem::ADataType>;
constexpr index_t MPerBlock = Problem::BlockGemmShape::kM;
constexpr index_t KPerBlock = Problem::BlockGemmShape::kK;
if constexpr(std::is_same_v<ALayout, ck_tile::tensor_layout::gemm::RowMajor>)
{
return GetGlobalVectorLoadSize<Problem, ADataType, MPerBlock, KPerBlock>();
}
else
{
return GetGlobalVectorLoadSize<Problem, ADataType, MPerBlock, MPerBlock>();
}
}
template <typename Problem>
CK_TILE_HOST_DEVICE static constexpr auto GetVectorSizeB()
{
using BLayout = remove_cvref_t<typename Problem::BLayout>;
using BDataType = remove_cvref_t<typename Problem::BDataType>;
constexpr index_t NPerBlock = Problem::BlockGemmShape::kN;
constexpr index_t KPerBlock = Problem::BlockGemmShape::kK;
if constexpr(std::is_same_v<BLayout, ck_tile::tensor_layout::gemm::RowMajor>)
{
return GetGlobalVectorLoadSize<Problem, BDataType, NPerBlock, NPerBlock>();
}
else
{
return GetGlobalVectorLoadSize<Problem, BDataType, NPerBlock, KPerBlock>();
}
}
/**
* @brief Get the vector store size for C tensor.
*
* @tparam Problem - Gemm pipeline problem class.
*
* @note The vector store size for output C tensor would depend on multiple factors
* like its data layout and warp gemm C transposition. In general it would
* be the number of consecutive elements in contiguous C dimension hold by
* single thread.
*
* @return The vector store size for C tensor.
*/
template <typename Problem>
CK_TILE_HOST_DEVICE static constexpr auto GetVectorSizeC()
{
using BlockGemm = remove_cvref_t<decltype(GetBlockGemm<Problem>())>;
using WG = typename BlockGemm::WarpGemm;
constexpr bool TransposeC = Problem::TransposeC;
using CLayout = typename Problem::CLayout;
using CWarpDstr = typename WG::CWarpDstr;
// N is contiguous dimension
if constexpr(std::is_same_v<CLayout, tensor_layout::gemm::RowMajor>)
{
if constexpr(TransposeC)
{
// In this case each thread has multiple consecutive elements in
// N dimension, however consecutive threads' elements have stride.
constexpr index_t NDimY = CWarpDstr::NDimY;
constexpr auto c_warp_y_lengths =
CWarpDstr{}.get_ys_to_d_descriptor().get_lengths();
static_assert(WG::WarpGemmAttribute::Impl::kCM1PerLane ==
c_warp_y_lengths.get(number<NDimY - 1>{}));
return c_warp_y_lengths.get(number<NDimY - 1>{});
}
else
{
// In this case each thread has just a single item in Ndim
return WG::WarpGemmAttribute::Impl::kCNLane / WG::kN;
}
}
// M is contiguous dimension
else if constexpr(std::is_same_v<CLayout, tensor_layout::gemm::ColumnMajor>)
{
if constexpr(TransposeC)
{
// In this case each thread has just a single item in Mdim
return WG::WarpGemmAttribute::Impl::kCNLane / WG::kN;
}
else
{
// In this case each thread has multiple consecutive elements in
// M dimension, however consecutive threads' elements have stride.
constexpr index_t NDimY = CWarpDstr::NDimY;
constexpr auto c_warp_y_lengths =
CWarpDstr{}.get_ys_to_d_descriptor().get_lengths();
static_assert(WG::WarpGemmAttribute::Impl::kCM1PerLane ==
c_warp_y_lengths.get(number<NDimY - 1>{}));
return c_warp_y_lengths.get(number<NDimY - 1>{});
}
}
else
{
static_assert(false, "Unsupported CLayout!");
}
}
template <typename Problem>
CK_TILE_HOST_DEVICE static constexpr auto GetSmemPackA()
{
using BlockGemm = decltype(GetBlockGemm<Problem>());
constexpr index_t KPack = BlockGemm::Traits::KPack;
return KPack;
}
template <typename Problem>
CK_TILE_HOST_DEVICE static constexpr auto GetSmemPackB()
{
using BlockGemm = decltype(GetBlockGemm<Problem>());
constexpr index_t KPack = BlockGemm::Traits::KPack;
return KPack;
}
template <typename Problem> template <typename Problem>
CK_TILE_HOST_DEVICE static constexpr auto MakeALdsBlockDescriptor() CK_TILE_HOST_DEVICE static constexpr auto MakeALdsBlockDescriptor()
{ {
...@@ -57,7 +190,7 @@ struct UniversalGemmPipelineAgBgCrPolicy ...@@ -57,7 +190,7 @@ struct UniversalGemmPipelineAgBgCrPolicy
constexpr index_t MPerBlock = Problem::BlockGemmShape::kM; constexpr index_t MPerBlock = Problem::BlockGemmShape::kM;
constexpr index_t KPerBlock = Problem::BlockGemmShape::kK; constexpr index_t KPerBlock = Problem::BlockGemmShape::kK;
constexpr index_t KPack = GetVectorLoadSize<Problem, ADataType, MPerBlock>(); constexpr index_t KPack = GetSmemPackA<Problem>();
constexpr auto DataTypeSize = sizeof(ADataType); constexpr auto DataTypeSize = sizeof(ADataType);
constexpr auto MLdsLayer = constexpr auto MLdsLayer =
...@@ -100,54 +233,193 @@ struct UniversalGemmPipelineAgBgCrPolicy ...@@ -100,54 +233,193 @@ struct UniversalGemmPipelineAgBgCrPolicy
return a_lds_block_desc; return a_lds_block_desc;
} }
/**
* @brief Create LDS block descriptor for B tensor.
*
* @tparam Problem Gemm pipeline problem.
* @return B tensor LDS block descriptor.
*/
template <typename Problem> template <typename Problem>
CK_TILE_HOST_DEVICE static constexpr auto MakeBLdsBlockDescriptor() CK_TILE_HOST_DEVICE static constexpr auto MakeBLdsBlockDescriptor()
{ {
// using BLayout = remove_cvref_t<typename Problem::BLayout>;
using BDataType = remove_cvref_t<typename Problem::BDataType>; using BDataType = remove_cvref_t<typename Problem::BDataType>;
constexpr index_t NPerBlock = Problem::BlockGemmShape::kN; constexpr index_t NPerBlock = Problem::BlockGemmShape::kN;
constexpr index_t KPerBlock = Problem::BlockGemmShape::kK; constexpr index_t KPerBlock = Problem::BlockGemmShape::kK;
constexpr index_t KPack = GetVectorLoadSize<Problem, BDataType, NPerBlock>();
constexpr auto DataTypeSize = sizeof(BDataType);
constexpr auto NLdsLayer =
(32 * 4 / KPerBlock / DataTypeSize) < 1 ? 1 : (32 * 4 / KPerBlock / DataTypeSize);
constexpr auto b_lds_block_desc_0 = make_naive_tensor_descriptor(
make_tuple(number<KPerBlock / KPack * NLdsLayer>{},
number<NPerBlock / NLdsLayer>{},
number<KPack>{}),
make_tuple(number<KPack>{}, number<KPerBlock * NLdsLayer>{}, number<1>{}),
number<KPack>{},
number<1>{});
constexpr auto b_lds_block_desc_permuted = transform_tensor_descriptor( #if 1
b_lds_block_desc_0, // if constexpr(std::is_same_v<BLayout, ck_tile::tensor_layout::gemm::ColumnMajor>)
make_tuple(make_xor_transform(make_tuple(number<NPerBlock / NLdsLayer>{}, {
number<KPerBlock / KPack * NLdsLayer>{})), constexpr index_t KPack = GetSmemPackB<Problem>();
make_pass_through_transform(number<KPack>{})), constexpr auto BK0 = number<KPerBlock / KPack>{};
make_tuple(sequence<1, 0>{}, sequence<2>{}), constexpr auto DataTypeSize = sizeof(BDataType);
make_tuple(sequence<1, 0>{}, sequence<2>{})); constexpr auto NLdsLayer =
(32 * 4 / KPerBlock / DataTypeSize) < 1 ? 1 : (32 * 4 / KPerBlock / DataTypeSize);
constexpr auto b_lds_block_desc_xk0_mnldslayer_mn_xk1 = transform_tensor_descriptor(
b_lds_block_desc_permuted, constexpr auto b_lds_block_desc_0 = make_naive_tensor_descriptor(
make_tuple(make_unmerge_transform( make_tuple(
make_tuple(number<KPerBlock / KPack>{}, number<NLdsLayer>{})), BK0 * number<NLdsLayer>{}, number<NPerBlock / NLdsLayer>{}, number<KPack>{}),
make_pass_through_transform(number<NPerBlock / NLdsLayer>{}), make_tuple(number<KPack>{}, number<KPerBlock * NLdsLayer>{}, number<1>{}),
make_pass_through_transform(number<KPack>{})), number<KPack>{},
make_tuple(sequence<0>{}, sequence<1>{}, sequence<2>{}), number<1>{});
make_tuple(sequence<0, 2>{}, sequence<1>{}, sequence<3>{}));
constexpr auto b_lds_block_desc_permuted = transform_tensor_descriptor(
constexpr auto b_lds_block_desc = transform_tensor_descriptor( b_lds_block_desc_0,
b_lds_block_desc_xk0_mnldslayer_mn_xk1, make_tuple(make_xor_transform(make_tuple(number<NPerBlock / NLdsLayer>{},
make_tuple(make_merge_transform_v3_division_mod( BK0 * number<NLdsLayer>{})),
make_tuple(number<NPerBlock / NLdsLayer>{}, number<NLdsLayer>{})), make_pass_through_transform(number<KPack>{})),
make_merge_transform_v3_division_mod( make_tuple(sequence<1, 0>{}, sequence<2>{}),
make_tuple(number<KPerBlock / KPack>{}, number<KPack>{}))), make_tuple(sequence<1, 0>{}, sequence<2>{}));
make_tuple(sequence<1, 2>{}, sequence<0, 3>{}),
make_tuple(sequence<0>{}, sequence<1>{})); constexpr auto b_lds_block_desc_bk0_nldslayer_n_bk1 = transform_tensor_descriptor(
return b_lds_block_desc; b_lds_block_desc_permuted,
make_tuple(make_unmerge_transform(make_tuple(BK0, number<NLdsLayer>{})),
make_pass_through_transform(number<NPerBlock / NLdsLayer>{}),
make_pass_through_transform(number<KPack>{})),
make_tuple(sequence<0>{}, sequence<1>{}, sequence<2>{}),
make_tuple(sequence<0, 2>{}, sequence<1>{}, sequence<3>{}));
constexpr auto b_lds_block_desc = transform_tensor_descriptor(
b_lds_block_desc_bk0_nldslayer_n_bk1,
make_tuple(make_merge_transform_v3_division_mod(
make_tuple(number<NPerBlock / NLdsLayer>{}, number<NLdsLayer>{})),
make_merge_transform_v3_division_mod(make_tuple(BK0, number<KPack>{}))),
make_tuple(sequence<1, 2>{}, sequence<0, 3>{}),
make_tuple(sequence<0>{}, sequence<1>{}));
return b_lds_block_desc;
}
#else
else // B is Row Major
{
constexpr index_t BlockSize = Problem::kBlockSize;
constexpr index_t VecLoadSize = GetVectorSizeB<Problem>();
using TileEncodingPattern = TileDistributionEncodingPattern2D<BlockSize,
KPerBlock,
NPerBlock,
VecLoadSize,
BTileAccessPattern>;
constexpr auto BK0 = number<TileEncodingPattern::X1>{};
constexpr auto BK1 = number<TileEncodingPattern::Y0>{};
// constexpr auto N0 = BBlockTransferThreadClusterLengths_BK0_N_BK1{}.At(I1);
constexpr auto N0 = TileEncodingPattern::X0;
constexpr auto N1 = NPerBlock / N0;
using WarpTile = typename Problem::BlockGemmShape::WarpTile;
constexpr auto NPerXdl = number<WarpTile::at(I1)>{};
// constexpr auto KThreadWrite =
// BBlockTransferThreadClusterLengths_BK0_N_BK1{}.At(I0);
constexpr auto KThreadWrite = TileEncodingPattern::Y2;
constexpr auto K0PerThreadWrite = BK0 / KThreadWrite;
constexpr auto KThreadRead = 64 / NPerXdl;
constexpr auto K0PerThreadRead = BK0 / KThreadRead;
constexpr auto kfold =
(BK1 * N0 * sizeof(BDataType) > 128) ? 1 : 128 / (BK1 * N0 * sizeof(BDataType));
constexpr auto KThreadReadPerm =
(kfold * K0PerThreadWrite / K0PerThreadRead) > 1
? KThreadRead / (kfold * K0PerThreadWrite / K0PerThreadRead)
: KThreadRead;
// 1<=npair<=n0
constexpr auto npair = (BK1 * NPerXdl * sizeof(BDataType) > 128)
? 1
: ((128 / (BK1 * NPerXdl * sizeof(BDataType))) > N0
? N0
: 128 / (BK1 * NPerXdl * sizeof(BDataType)));
constexpr auto b_lds_block_desc = make_naive_tensor_descriptor_packed(
make_tuple(number<KThreadWrite / kfold / KThreadReadPerm>{},
number<K0PerThreadWrite>{},
number<KThreadReadPerm * N1>{},
number<kfold * N0 / npair>{},
number<npair>{},
BK1));
constexpr auto b_lds_block_desc_permuted = transform_tensor_descriptor(
b_lds_block_desc,
make_tuple(
make_pass_through_transform(number<KThreadWrite / kfold / KThreadReadPerm>{}),
make_pass_through_transform(number<K0PerThreadWrite>{}),
make_xor_transform(
make_tuple(number<KThreadReadPerm * N1>{}, number<kfold * N0 / npair>{})),
make_pass_through_transform(number<npair>{}),
make_pass_through_transform(BK1)),
make_tuple(
sequence<0>{}, sequence<1>{}, sequence<2, 3>{}, sequence<4>{}, sequence<5>{}),
make_tuple(
sequence<0>{}, sequence<1>{}, sequence<2, 3>{}, sequence<4>{}, sequence<5>{}));
constexpr auto b_lds_block_desc_unmerged = transform_tensor_descriptor(
b_lds_block_desc_permuted,
make_tuple(
make_pass_through_transform(number<KThreadWrite / kfold / KThreadReadPerm>{}),
make_pass_through_transform(number<K0PerThreadWrite>{}),
make_unmerge_transform(make_tuple(number<KThreadReadPerm>{}, number<N1>{})),
make_unmerge_transform(make_tuple(number<kfold>{}, number<N0 / npair>{})),
make_pass_through_transform(number<npair>{}),
make_pass_through_transform(BK1)),
make_tuple(sequence<0>{},
sequence<1>{},
sequence<2>{},
sequence<3>{},
sequence<4>{},
sequence<5>{}),
make_tuple(sequence<1>{},
sequence<2>{},
sequence<0, 3>{},
sequence<4, 5>{},
sequence<6>{},
sequence<7>{}));
// constexpr auto b_lds_block_desc_bk0_n_bk1 = transform_tensor_descriptor(
// b_lds_block_desc_unmerged,
// make_tuple(make_merge_transform_v3_division_mod(
// make_tuple(number<KThreadReadPerm>{},
// number<KThreadWrite / kfold / KThreadReadPerm>{},
// number<kfold>{},
// number<K0PerThreadWrite>{})),
// make_merge_transform_v3_division_mod(
// make_tuple(number<N0 / npair>{}, number<npair>{}, number<N1>{})),
// make_pass_through_transform(BK1)),
// make_tuple(sequence<0, 1, 4, 2>{}, sequence<5, 6, 3>{}, sequence<7>{}),
// make_tuple(sequence<0>{}, sequence<1>{}, sequence<2>{}));
constexpr auto b_lds_block_desc_kn = transform_tensor_descriptor(
b_lds_block_desc_unmerged,
make_tuple(make_merge_transform_v3_division_mod(
make_tuple(number<KThreadReadPerm>{},
number<KThreadWrite / kfold / KThreadReadPerm>{},
number<kfold>{},
number<K0PerThreadWrite>{},
BK1)),
make_merge_transform_v3_division_mod(
make_tuple(number<N0 / npair>{}, number<npair>{}, number<N1>{}))),
make_tuple(sequence<0, 1, 4, 2, 7>{}, sequence<5, 6, 3>{}),
make_tuple(sequence<1>{}, sequence<0>{}));
// return b_lds_block_desc_bk0_n_bk1;
return b_lds_block_desc_kn;
// constexpr auto b_lds_block_desc_bk0_n_bk1 = make_naive_tensor_descriptor(
// make_tuple(BK0, number<NPerBlock>{}, number<KPack>{}),
// make_tuple(number<KPack>{}, number<KPerBlock>{}, number<1>{}),
// number<KPack>{},
// number<1>{});
// constexpr auto b_lds_block_desc = transform_tensor_descriptor(
// b_lds_block_desc_bk0_n_bk1,
// make_tuple(make_pass_through_transform(number<NPerBlock>{}),
// make_merge_transform_v3_division_mod(make_tuple(BK0,
// number<KPack>{}))),
// make_tuple(sequence<1>{}, sequence<0, 2>{}),
// make_tuple(sequence<0>{}, sequence<1>{}));
// return b_lds_block_desc;
}
#endif
} }
template <typename Problem> template <typename Problem>
...@@ -180,289 +452,121 @@ struct UniversalGemmPipelineAgBgCrPolicy ...@@ -180,289 +452,121 @@ struct UniversalGemmPipelineAgBgCrPolicy
template <typename Problem> template <typename Problem>
CK_TILE_HOST_DEVICE static constexpr auto MakeADramTileDistribution() CK_TILE_HOST_DEVICE static constexpr auto MakeADramTileDistribution()
{ {
using ADataType = remove_cvref_t<typename Problem::ADataType>; using ALayout = remove_cvref_t<typename Problem::ALayout>;
using ALayout = remove_cvref_t<typename Problem::ALayout>;
constexpr index_t BlockSize = Problem::kBlockSize;
constexpr index_t MPerBlock = Problem::BlockGemmShape::kM; constexpr index_t BlockSize = Problem::kBlockSize;
constexpr index_t KPerBlock = Problem::BlockGemmShape::kK; constexpr index_t MPerBlock = Problem::BlockGemmShape::kM;
constexpr index_t KPerBlock = Problem::BlockGemmShape::kK;
constexpr index_t VecLoadSize = GetVectorSizeA<Problem>();
if constexpr(std::is_same_v<ALayout, ck_tile::tensor_layout::gemm::ColumnMajor>) // Tile: MPerBlock X KPerBlock
if constexpr(std::is_same_v<ALayout, ck_tile::tensor_layout::gemm::RowMajor>)
{ {
constexpr index_t M1 = Problem::VectorLoadSize / sizeof(ADataType); using TileEncodingPattern = TileDistributionEncodingPattern2D<BlockSize,
constexpr index_t M0 = MPerBlock / M1; MPerBlock,
constexpr index_t total_pixels = MPerBlock * KPerBlock / BlockSize; KPerBlock,
static_assert(total_pixels % M1 == 0); VecLoadSize,
constexpr index_t K3 = total_pixels / M1; ATileAccessPattern>;
constexpr index_t KPack = GetVectorLoadSize<Problem, ADataType, MPerBlock>(); return TileEncodingPattern::Make2DStaticTileDistribution();
static_assert(KPack % K3 == 0);
constexpr index_t K2 = KPack / K3;
if constexpr(get_warp_size() % (K2 * M0) == 0)
{
constexpr index_t K1 = get_warp_size() / (K2 * M0);
constexpr index_t K0 = BlockSize / get_warp_size();
static_assert(KPerBlock == K0 * K1 * K2 * K3);
return make_static_tile_distribution(
tile_distribution_encoding<sequence<1>,
tuple<sequence<M0, M1>, sequence<K0, K1, K2, K3>>,
tuple<sequence<2>, sequence<2, 1, 2>>,
tuple<sequence<0>, sequence<1, 0, 2>>,
sequence<2, 1>,
sequence<3, 1>>{});
}
else
{
constexpr index_t K1 = (K2 * M0) / get_warp_size();
constexpr index_t K2_m = K2 / K1;
constexpr index_t K0 = BlockSize / get_warp_size() / K1;
static_assert(KPerBlock == K0 * K1 * K2_m * K3);
return make_static_tile_distribution(
tile_distribution_encoding<sequence<1>,
tuple<sequence<M0, M1>, sequence<K0, K1, K2_m, K3>>,
tuple<sequence<2, 2>, sequence<1, 2>>,
tuple<sequence<0, 1>, sequence<0, 2>>,
sequence<2, 1>,
sequence<3, 1>>{});
}
} }
// Tile: KPerBlock X MPerBlock
else else
{ {
constexpr index_t K1 = Problem::VectorLoadSize / sizeof(ADataType); using TileEncodingPattern = TileDistributionEncodingPattern2D<BlockSize,
constexpr index_t K0 = KPerBlock / K1; KPerBlock,
constexpr index_t M2 = get_warp_size() / K0; MPerBlock,
if constexpr(get_warp_size() % (M2 * K0) == 0) VecLoadSize,
{ ATileAccessPattern>;
constexpr index_t M1 = BlockSize / get_warp_size(); return TileEncodingPattern::Make2DStaticTileDistribution();
static_assert(M2 != 0, "M2 is zero, which will lead to a division by zero error.");
static_assert(M1 != 0, "M1 is zero, which will lead to a division by zero error.");
constexpr index_t M0 = MPerBlock / (M2 * M1);
return make_static_tile_distribution(
tile_distribution_encoding<sequence<1>,
tuple<sequence<M0, M1, M2>, sequence<K0, K1>>,
tuple<sequence<1>, sequence<1, 2>>,
tuple<sequence<1>, sequence<2, 0>>,
sequence<1, 2>,
sequence<0, 1>>{});
}
else
{
constexpr index_t M0 = BlockSize / get_warp_size();
constexpr index_t M1 = MPerBlock / (M2 * M0);
return make_static_tile_distribution(
tile_distribution_encoding<sequence<1>,
tuple<sequence<M0, M1, M2>, sequence<K0, K1>>,
tuple<sequence<1>, sequence<1, 2>>,
tuple<sequence<0>, sequence<2, 0>>,
sequence<1, 2>,
sequence<1, 1>>{});
}
} }
} }
template <typename Problem> template <typename Problem>
CK_TILE_HOST_DEVICE static constexpr auto MakeBDramTileDistribution() CK_TILE_HOST_DEVICE static constexpr auto MakeBDramTileDistribution()
{ {
using BDataType = remove_cvref_t<typename Problem::BDataType>; using BLayout = remove_cvref_t<typename Problem::BLayout>;
using BLayout = remove_cvref_t<typename Problem::BLayout>;
constexpr index_t BlockSize = Problem::kBlockSize;
constexpr index_t NPerBlock = Problem::BlockGemmShape::kN; constexpr index_t BlockSize = Problem::kBlockSize;
constexpr index_t KPerBlock = Problem::BlockGemmShape::kK; constexpr index_t NPerBlock = Problem::BlockGemmShape::kN;
constexpr index_t KPerBlock = Problem::BlockGemmShape::kK;
constexpr index_t VecLoadSize = GetVectorSizeB<Problem>();
// Tile: KPerBlock X NPerBlock
if constexpr(std::is_same_v<BLayout, ck_tile::tensor_layout::gemm::RowMajor>) if constexpr(std::is_same_v<BLayout, ck_tile::tensor_layout::gemm::RowMajor>)
{ {
constexpr index_t N1 = Problem::VectorLoadSize / sizeof(BDataType); using TileEncodingPattern = TileDistributionEncodingPattern2D<BlockSize,
constexpr index_t N0 = NPerBlock / N1; KPerBlock,
constexpr index_t total_pixels = NPerBlock * KPerBlock / BlockSize; NPerBlock,
static_assert(total_pixels % N1 == 0); VecLoadSize,
constexpr index_t K3 = total_pixels / N1; BTileAccessPattern>;
constexpr index_t KPack = GetVectorLoadSize<Problem, BDataType, NPerBlock>(); return TileEncodingPattern::Make2DStaticTileDistribution();
static_assert(KPack % K3 == 0);
constexpr index_t K2 = KPack / K3;
if constexpr(get_warp_size() % (K2 * N0) == 0)
{
constexpr index_t K1 = get_warp_size() / (K2 * N0);
constexpr index_t K0 = BlockSize / get_warp_size();
static_assert(KPerBlock == K0 * K1 * K2 * K3);
return make_static_tile_distribution(
tile_distribution_encoding<sequence<1>,
tuple<sequence<N0, N1>, sequence<K0, K1, K2, K3>>,
tuple<sequence<2>, sequence<2, 1, 2>>,
tuple<sequence<0>, sequence<1, 0, 2>>,
sequence<2, 1>,
sequence<3, 1>>{});
}
else
{
constexpr index_t K1 = (K2 * N0) / get_warp_size();
constexpr index_t K2_m = K2 / K1;
constexpr index_t K0 = BlockSize / get_warp_size() / K1;
static_assert(KPerBlock == K0 * K1 * K2_m * K3);
return make_static_tile_distribution(
tile_distribution_encoding<sequence<1>,
tuple<sequence<N0, N1>, sequence<K0, K1, K2_m, K3>>,
tuple<sequence<2, 2>, sequence<1, 2>>,
tuple<sequence<0, 1>, sequence<0, 2>>,
sequence<2, 1>,
sequence<3, 1>>{});
}
} }
// Tile: NPerBlock X KPerBlock
else else
{ {
using TileEncodingPattern = TileDistributionEncodingPattern2D<BlockSize,
constexpr index_t K1 = Problem::VectorLoadSize / sizeof(BDataType); NPerBlock,
constexpr index_t K0 = KPerBlock / K1; KPerBlock,
constexpr index_t N2 = get_warp_size() / K0; VecLoadSize,
// coalesce reading for each blocks BTileAccessPattern>;
if constexpr(get_warp_size() % (N2 * K0) == 0) return TileEncodingPattern::Make2DStaticTileDistribution();
{
constexpr index_t N1 = BlockSize / get_warp_size();
static_assert(N2 != 0, "N2 is zero, which will lead to a division by zero error.");
static_assert(N1 != 0, "N1 is zero, which will lead to a division by zero error.");
constexpr index_t N0 = NPerBlock / (N2 * N1);
return make_static_tile_distribution(
tile_distribution_encoding<sequence<1>,
tuple<sequence<N0, N1, N2>, sequence<K0, K1>>,
tuple<sequence<1>, sequence<1, 2>>,
tuple<sequence<1>, sequence<2, 0>>,
sequence<1, 2>,
sequence<0, 1>>{});
}
// coalesce reading for each warps
else
{
constexpr index_t N0 = BlockSize / get_warp_size();
constexpr index_t N1 = NPerBlock / (N2 * N0);
return make_static_tile_distribution(
tile_distribution_encoding<sequence<1>,
tuple<sequence<N0, N1, N2>, sequence<K0, K1>>,
tuple<sequence<1>, sequence<1, 2>>,
tuple<sequence<0>, sequence<2, 0>>,
sequence<1, 2>,
sequence<1, 1>>{});
}
} }
} }
template <typename Problem> template <typename Problem>
CK_TILE_HOST_DEVICE static constexpr auto MakeShuffledARegBlockDescriptor() CK_TILE_HOST_DEVICE static constexpr auto MakeShuffledARegTileDistribution()
{ {
using ALayout = remove_cvref_t<typename Problem::ALayout>; using ALayout = remove_cvref_t<typename Problem::ALayout>;
using ADataType = remove_cvref_t<typename Problem::ADataType>;
static_assert(std::is_same_v<ALayout, ck_tile::tensor_layout::gemm::ColumnMajor>); static_assert(std::is_same_v<ALayout, ck_tile::tensor_layout::gemm::ColumnMajor>);
constexpr index_t BlockSize = Problem::kBlockSize; constexpr index_t BlockSize = Problem::kBlockSize;
constexpr index_t MPerBlock = Problem::BlockGemmShape::kN; constexpr index_t MPerBlock = Problem::BlockGemmShape::kN;
constexpr index_t KPerBlock = Problem::BlockGemmShape::kK; constexpr index_t KPerBlock = Problem::BlockGemmShape::kK;
constexpr index_t VecLoadSize = GetVectorSizeA<Problem>();
constexpr index_t M1 = Problem::VectorLoadSize / sizeof(ADataType);
constexpr index_t M0 = MPerBlock / M1; using TileEncodingPattern = TileDistributionEncodingPattern2D<BlockSize,
constexpr index_t total_pixels = MPerBlock * KPerBlock / BlockSize; KPerBlock,
static_assert(total_pixels % M1 == 0); MPerBlock,
constexpr index_t K3 = total_pixels / M1; VecLoadSize,
constexpr index_t kKPack = GetVectorLoadSize<Problem, ADataType, MPerBlock>(); ATileAccessPattern>;
static_assert(kKPack % K3 == 0); return TileEncodingPattern::MakeShuffled2DStaticTileDistribution();
constexpr index_t K2 = kKPack / K3; // TODO: this dimention could be outside single wave
constexpr index_t warp_size = get_warp_size();
if constexpr(warp_size % (K2 * M0) == 0)
{
constexpr index_t K1 = warp_size / (K2 * M0);
constexpr index_t K0 = BlockSize / warp_size;
return make_static_tile_distribution(
tile_distribution_encoding<sequence<1>,
tuple<sequence<M0, M1>, sequence<K0, K1, K2, K3>>,
tuple<sequence<2>, sequence<2, 1, 2>>,
tuple<sequence<0>, sequence<1, 0, 2>>,
sequence<1, 2>,
sequence<1, 3>>{});
}
else
{
constexpr index_t K1 = (K2 * M0) / get_warp_size();
constexpr index_t K2_m = K2 / K1;
constexpr index_t K0 = BlockSize / get_warp_size() / K1;
static_assert(KPerBlock == K0 * K1 * K2_m * K3);
return make_static_tile_distribution(
tile_distribution_encoding<sequence<1>,
tuple<sequence<M0, M1>, sequence<K0, K1, K2_m, K3>>,
tuple<sequence<2, 2>, sequence<1, 2>>,
tuple<sequence<0, 1>, sequence<0, 2>>,
sequence<1, 2>,
sequence<1, 3>>{});
}
} }
template <typename Problem> template <typename Problem>
CK_TILE_HOST_DEVICE static constexpr auto MakeShuffledBRegBlockDescriptor() CK_TILE_HOST_DEVICE static constexpr auto MakeShuffledBRegTileDistribution()
{ {
using BLayout = remove_cvref_t<typename Problem::BLayout>; using BLayout = remove_cvref_t<typename Problem::BLayout>;
using BDataType = remove_cvref_t<typename Problem::BDataType>;
static_assert(std::is_same_v<BLayout, ck_tile::tensor_layout::gemm::RowMajor>); static_assert(std::is_same_v<BLayout, ck_tile::tensor_layout::gemm::RowMajor>);
constexpr index_t BlockSize = Problem::kBlockSize; constexpr index_t BlockSize = Problem::kBlockSize;
constexpr index_t NPerBlock = Problem::BlockGemmShape::kN; constexpr index_t NPerBlock = Problem::BlockGemmShape::kN;
constexpr index_t KPerBlock = Problem::BlockGemmShape::kK; constexpr index_t KPerBlock = Problem::BlockGemmShape::kK;
constexpr index_t VecLoadSize = GetVectorSizeB<Problem>();
constexpr index_t N1 = Problem::VectorLoadSize / sizeof(BDataType);
constexpr index_t N0 = NPerBlock / N1; using TileEncodingPattern = TileDistributionEncodingPattern2D<BlockSize,
constexpr index_t total_pixels = NPerBlock * KPerBlock / BlockSize; KPerBlock,
static_assert(total_pixels % N1 == 0); NPerBlock,
constexpr index_t K3 = total_pixels / N1; VecLoadSize,
constexpr index_t kKPack = GetVectorLoadSize<Problem, BDataType, NPerBlock>(); BTileAccessPattern>;
static_assert(kKPack % K3 == 0); return TileEncodingPattern::MakeShuffled2DStaticTileDistribution();
constexpr index_t K2 = kKPack / K3; // TODO: this dimention could be outside single wave
constexpr index_t warp_size = get_warp_size();
if constexpr(warp_size % (K2 * N0) == 0)
{
constexpr index_t K1 = warp_size / (K2 * N0);
constexpr index_t K0 = BlockSize / warp_size;
return make_static_tile_distribution(
tile_distribution_encoding<sequence<1>,
tuple<sequence<N0, N1>, sequence<K0, K1, K2, K3>>,
tuple<sequence<2>, sequence<2, 1, 2>>,
tuple<sequence<0>, sequence<1, 0, 2>>,
sequence<1, 2>,
sequence<1, 3>>{});
}
else
{
constexpr index_t K1 = (K2 * N0) / get_warp_size();
constexpr index_t K2_m = K2 / K1;
constexpr index_t K0 = BlockSize / get_warp_size() / K1;
static_assert(KPerBlock == K0 * K1 * K2_m * K3);
return make_static_tile_distribution(
tile_distribution_encoding<sequence<1>,
tuple<sequence<N0, N1>, sequence<K0, K1, K2_m, K3>>,
tuple<sequence<2, 2>, sequence<1, 2>>,
tuple<sequence<0, 1>, sequence<0, 2>>,
sequence<1, 2>,
sequence<1, 3>>{});
}
} }
template <typename Problem> template <typename Problem>
CK_TILE_HOST_DEVICE static constexpr auto GetBlockGemm() CK_TILE_HOST_DEVICE static constexpr auto GetBlockGemm()
{ {
using AccDataType = float;
using BlockWarps = typename Problem::BlockGemmShape::BlockWarps; using BlockWarps = typename Problem::BlockGemmShape::BlockWarps;
using WarpTile = typename Problem::BlockGemmShape::WarpTile; using WarpTile = typename Problem::BlockGemmShape::WarpTile;
using WarpGemm = WarpGemmMfmaDispatcher<typename Problem::ADataType, using WarpGemm = WarpGemmMfmaDispatcher<typename Problem::ADataType,
typename Problem::BDataType, typename Problem::BDataType,
AccDataType, typename Problem::CDataType,
WarpTile::at(I0), WarpTile::at(I0),
WarpTile::at(I1), WarpTile::at(I1),
WarpTile::at(I2), WarpTile::at(I2),
TransposeC>; Problem::TransposeC>;
using BlockGemmPolicy = BlockGemmASmemBSmemCRegV1CustomPolicy<typename Problem::ADataType, using BlockGemmPolicy = BlockGemmASmemBSmemCRegV1CustomPolicy<typename Problem::ADataType,
typename Problem::BDataType, typename Problem::BDataType,
typename Problem::CDataType, typename Problem::CDataType,
BlockWarps, BlockWarps,
WarpGemm>; WarpGemm>;
return BlockGemmASmemBSmemCRegV1<Problem, BlockGemmPolicy>{}; return BlockUniversalGemmAsBsCr<Problem, BlockGemmPolicy>{};
} }
}; };
......
include/ck_tile/ops/gemm/pipeline/tile_gemm_traits.hpp
View file @ 1b616990
...@@ -19,11 +19,34 @@ struct TileGemmTraits ...@@ -19,11 +19,34 @@ struct TileGemmTraits
static constexpr bool kPadN = kPadN_; static constexpr bool kPadN = kPadN_;
static constexpr bool kPadK = kPadK_; static constexpr bool kPadK = kPadK_;
// TODO this can't be hardcoded here! Should be in policy!
static constexpr int _VectorSize = 16; static constexpr int _VectorSize = 16;
using ALayout = ALayout_; using ALayout = ALayout_;
using BLayout = BLayout_; using BLayout = BLayout_;
using CLayout = CLayout_; using CLayout = CLayout_;
static constexpr bool TransposeC = false;
};
template <bool kPadM_,
bool kPadN_,
bool kPadK_,
typename ALayout_,
typename BLayout_,
typename CLayout_,
bool TransposeC_ = false>
struct TileGemmUniversalTraits
{
static constexpr bool kPadM = kPadM_;
static constexpr bool kPadN = kPadN_;
static constexpr bool kPadK = kPadK_;
using ALayout = ALayout_;
using BLayout = BLayout_;
using CLayout = CLayout_;
static constexpr bool TransposeC = TransposeC_;
}; };
} // namespace ck_tile } // namespace ck_tile
include/ck_tile/ops/image_to_column.hpp
View file @ 1b616990
// SPDX-License-Identifier: MIT // 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 #pragma once
......
include/ck_tile/ops/layernorm2d.hpp
View file @ 1b616990
// SPDX-License-Identifier: MIT // 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 #pragma once
......
include/ck_tile/ops/layernorm2d/kernel/layernorm2d_fwd_kernel.hpp
View file @ 1b616990
// SPDX-License-Identifier: MIT // 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 #pragma once
...@@ -14,7 +14,8 @@ struct Layernorm2dFwdHostArgs ...@@ -14,7 +14,8 @@ struct Layernorm2dFwdHostArgs
{ {
const void* p_x; // [m ,n], input, fp16/bf16 const void* p_x; // [m ,n], input, fp16/bf16
const void* p_x_residual; // [m ,n], shortcut input, prec same as input, nullptr if not used const void* p_x_residual; // [m ,n], shortcut input, prec same as input, nullptr if not used
const void* p_x_scale; // [1 ,n], smooth scale input, fp32, nullptr if not used const void* p_sm_scale; // [1 ,n], smooth scale input, fp32, nullptr if not used
const void* p_x_bias; // [1, n], bias, prec same as input
const void* p_gamma; // [1, n], gamma, prec same as input const void* p_gamma; // [1, n], gamma, prec same as input
const void* p_beta; // [1, n], beta, prec same as input const void* p_beta; // [1, n], beta, prec same as input
...@@ -42,15 +43,16 @@ struct Layernorm2dFwd ...@@ -42,15 +43,16 @@ struct Layernorm2dFwd
using Epilogue = remove_cvref_t<Epilogue_>; using Epilogue = remove_cvref_t<Epilogue_>;
using Problem = typename Pipeline::Problem; using Problem = typename Pipeline::Problem;
using XDataType = remove_cvref_t<typename Problem::XDataType>; using XDataType = remove_cvref_t<typename Problem::XDataType>;
using GammaDataType = remove_cvref_t<typename Problem::GammaDataType>; using XBiasDataType = remove_cvref_t<typename Problem::XBiasDataType>;
using BetaDataType = remove_cvref_t<typename Problem::BetaDataType>; using GammaDataType = remove_cvref_t<typename Problem::GammaDataType>;
using ComputeDataType = remove_cvref_t<typename Problem::ComputeDataType>; using BetaDataType = remove_cvref_t<typename Problem::BetaDataType>;
using YDataType = remove_cvref_t<typename Problem::YDataType>; using ComputeDataType = remove_cvref_t<typename Problem::ComputeDataType>;
using MeanDataType = remove_cvref_t<typename Problem::MeanDataType>; using YDataType = remove_cvref_t<typename Problem::YDataType>;
using InvStdDataType = remove_cvref_t<typename Problem::InvStdDataType>; using MeanDataType = remove_cvref_t<typename Problem::MeanDataType>;
using XScaleDataType = remove_cvref_t<typename Problem::XScaleDataType>; using InvStdDataType = remove_cvref_t<typename Problem::InvStdDataType>;
using YScaleDataType = remove_cvref_t<typename Problem::YScaleDataType>; using SmoothScaleDataType = remove_cvref_t<typename Problem::SmoothScaleDataType>;
using YScaleDataType = remove_cvref_t<typename Problem::YScaleDataType>;
// for simplicity, shortcut input/output type is same as X // for simplicity, shortcut input/output type is same as X
using XResidualDataType = XDataType; using XResidualDataType = XDataType;
...@@ -67,6 +69,7 @@ struct Layernorm2dFwd ...@@ -67,6 +69,7 @@ struct Layernorm2dFwd
static constexpr bool kPadM = false; // always no need to pad along M static constexpr bool kPadM = false; // always no need to pad along M
static constexpr bool kPadN = Problem::Traits::kPadN; static constexpr bool kPadN = Problem::Traits::kPadN;
static constexpr bool kTwoPass = Problem::Traits::kTwoPass; static constexpr bool kTwoPass = Problem::Traits::kTwoPass;
static constexpr auto kXbias = Problem::Traits::kXbias;
static constexpr auto kFusedAdd = Problem::Traits::kFusedAdd; static constexpr auto kFusedAdd = Problem::Traits::kFusedAdd;
static constexpr auto kFusedQuant = Problem::Traits::kFusedQuant; static constexpr auto kFusedQuant = Problem::Traits::kFusedQuant;
...@@ -81,7 +84,8 @@ struct Layernorm2dFwd ...@@ -81,7 +84,8 @@ struct Layernorm2dFwd
{ {
const void* p_x; // [m ,n], input, fp16/bf16 const void* p_x; // [m ,n], input, fp16/bf16
const void* p_x_residual; // [m ,n], shortcut input, prec same as input, nullptr if not used const void* p_x_residual; // [m ,n], shortcut input, prec same as input, nullptr if not used
const void* p_x_scale; // [1 ,n], smooth scale input, fp32, nullptr if not used const void* p_sm_scale; // [1 ,n], smooth scale input, fp32, nullptr if not used
const void* p_x_bias; // [1, n], bias, prec same as input
const void* p_gamma; // [1, n], gamma, prec same as input const void* p_gamma; // [1, n], gamma, prec same as input
const void* p_beta; // [1, n], beta, prec same as input const void* p_beta; // [1, n], beta, prec same as input
...@@ -107,7 +111,8 @@ struct Layernorm2dFwd ...@@ -107,7 +111,8 @@ struct Layernorm2dFwd
{ {
return Kargs{hargs.p_x, return Kargs{hargs.p_x,
hargs.p_x_residual, hargs.p_x_residual,
hargs.p_x_scale, hargs.p_sm_scale,
hargs.p_x_bias,
hargs.p_gamma, hargs.p_gamma,
hargs.p_beta, hargs.p_beta,
hargs.p_y, hargs.p_y,
...@@ -152,6 +157,7 @@ struct Layernorm2dFwd ...@@ -152,6 +157,7 @@ struct Layernorm2dFwd
using S_ = typename Problem::BlockShape; using S_ = typename Problem::BlockShape;
auto surfix = [&] () { auto surfix = [&] () {
std::string n; std::string n;
if (kXbias != Layernorm2dXBiasEnum::NO_BIAS) n += _SS_("_") + Layernorm2dXBiasEnumName<kXbias>::name;
if (kFusedAdd != Layernorm2dFusedAddEnum::NO_ADD) n += _SS_("_") + Layernorm2dFusedAddEnumName<kFusedAdd>::name; if (kFusedAdd != Layernorm2dFusedAddEnum::NO_ADD) n += _SS_("_") + Layernorm2dFusedAddEnumName<kFusedAdd>::name;
if (kFusedQuant != Layernorm2dFusedQuantEnum::NO_SWEEP) n += _SS_("_") + Layernorm2dFusedQuantEnumName<kFusedQuant>::name; if (kFusedQuant != Layernorm2dFusedQuantEnum::NO_SWEEP) n += _SS_("_") + Layernorm2dFusedQuantEnumName<kFusedQuant>::name;
if (kPadN) n += "_pn"; if (kPadN) n += "_pn";
...@@ -165,7 +171,7 @@ struct Layernorm2dFwd ...@@ -165,7 +171,7 @@ struct Layernorm2dFwd
base_str += _SS_("_") + _SS_(t2s<YDataType>::name); base_str += _SS_("_") + _SS_(t2s<YDataType>::name);
} }
if (kFusedQuant == Layernorm2dFusedQuantEnum::SMOOTH_DYNAMIC_QUANT) { if (kFusedQuant == Layernorm2dFusedQuantEnum::SMOOTH_DYNAMIC_QUANT) {
base_str += _SS_("_sx") + _SS_(t2s<XScaleDataType>::name); base_str += _SS_("_sx") + _SS_(t2s<SmoothScaleDataType>::name);
base_str += _SS_("_sy") + _SS_(t2s<YScaleDataType>::name); base_str += _SS_("_sy") + _SS_(t2s<YScaleDataType>::name);
} }
if (kFusedQuant == Layernorm2dFusedQuantEnum::DYNAMIC_QUANT) { if (kFusedQuant == Layernorm2dFusedQuantEnum::DYNAMIC_QUANT) {
...@@ -228,6 +234,27 @@ struct Layernorm2dFwd ...@@ -228,6 +234,27 @@ struct Layernorm2dFwd
} }
}(); }();
const auto x_bias_window = [&]() {
if constexpr(kXbias == Layernorm2dXBiasEnum::ADD_BIAS)
{
const auto tmp_ = make_naive_tensor_view<address_space_enum::global>(
static_cast<const XBiasDataType*>(kargs.p_x_bias),
make_tuple(kargs.n),
make_tuple(1),
number<Vector_N>{},
number<1>{});
const auto tmp2_ =
pad_tensor_view(tmp_, make_tuple(number<Block_N>{}), sequence<false>{});
return make_tile_window(tmp2_, make_tuple(number<Block_N>{}), {0});
}
else
{
return make_null_tile_window(make_tuple(number<Block_N>{}));
}
}();
const auto gamma_window = [&]() { const auto gamma_window = [&]() {
const auto tmp_ = make_naive_tensor_view<address_space_enum::global>( const auto tmp_ = make_naive_tensor_view<address_space_enum::global>(
static_cast<const GammaDataType*>(kargs.p_gamma), static_cast<const GammaDataType*>(kargs.p_gamma),
...@@ -329,18 +356,18 @@ struct Layernorm2dFwd ...@@ -329,18 +356,18 @@ struct Layernorm2dFwd
return make_null_tile_window(make_tuple(number<Block_M>{})); return make_null_tile_window(make_tuple(number<Block_M>{}));
}(); }();
auto x_scale_window = [&]() { auto sm_scale_window = [&]() {
if constexpr(kFusedQuant == Layernorm2dFusedQuantEnum::SMOOTH_DYNAMIC_QUANT) if constexpr(kFusedQuant == Layernorm2dFusedQuantEnum::SMOOTH_DYNAMIC_QUANT)
{ {
const auto win_ = [&]() { const auto win_ = [&]() {
const auto tmp_0_ = make_naive_tensor_view_packed<address_space_enum::global>( const auto tmp_0_ = make_naive_tensor_view_packed<address_space_enum::global>(
static_cast<const XScaleDataType*>(kargs.p_x_scale), static_cast<const SmoothScaleDataType*>(kargs.p_sm_scale),
make_tuple(kargs.n), make_tuple(kargs.n),
number<Vector_N>{}); number<Vector_N>{});
return pad_tensor_view(tmp_0_, return pad_tensor_view(tmp_0_,
make_tuple(number<Block_N>{}), make_tuple(number<Block_N>{}),
sequence<false>{}); // x_scale no need pad sequence<false>{}); // sm_scale no need pad
}(); }();
return make_tile_window(win_, make_tuple(number<Block_N>{}), {0}); return make_tile_window(win_, make_tuple(number<Block_N>{}), {0});
} }
...@@ -371,13 +398,14 @@ struct Layernorm2dFwd ...@@ -371,13 +398,14 @@ struct Layernorm2dFwd
Pipeline{}(x_window, Pipeline{}(x_window,
x_residual_window, x_residual_window,
x_bias_window,
gamma_window, gamma_window,
beta_window, beta_window,
y_window, y_window,
y_residual_window, y_residual_window,
mean_window, mean_window,
inv_std_window, inv_std_window,
x_scale_window, sm_scale_window,
y_scale_window, y_scale_window,
static_cast<const ComputeDataType>(kargs.epsilon), static_cast<const ComputeDataType>(kargs.epsilon),
kargs.n, kargs.n,
......
include/ck_tile/ops/layernorm2d/pipeline/layernorm2d_fwd_pipeline_default_policy.hpp
View file @ 1b616990
...@@ -4,8 +4,8 @@ ...@@ -4,8 +4,8 @@
#pragma once #pragma once
#include "ck_tile/core.hpp" #include "ck_tile/core.hpp"
#include "ck_tile/ops/welford/block/block_welford_problem.hpp" #include "ck_tile/ops/norm_reduce/block/block_norm_reduce_problem.hpp"
#include "ck_tile/ops/welford/block/block_welford.hpp" #include "ck_tile/ops/norm_reduce/block/block_norm_reduce.hpp"
namespace ck_tile { namespace ck_tile {
...@@ -43,36 +43,38 @@ struct Layernorm2dFwdPipelineDefaultPolicy ...@@ -43,36 +43,38 @@ struct Layernorm2dFwdPipelineDefaultPolicy
} }
template <typename Problem> template <typename Problem>
CK_TILE_HOST_DEVICE static constexpr auto GetBlockWelford() CK_TILE_HOST_DEVICE static constexpr auto GetBlockNormReduce()
{ {
using P_ = BlockWelfordProblem<typename Problem::ComputeDataType, using P_ = BlockNormReduceProblem<typename Problem::ComputeDataType,
typename Problem::ComputeDataType, typename Problem::ComputeDataType,
typename Problem::BlockShape, typename Problem::BlockShape,
Problem::Traits::kFastFDiv>; Problem::Traits::kFastFDiv,
Problem::Traits::kWelford>;
return BlockWelford<P_>{}; return BlockNormReduce<P_>{};
} }
template <typename Problem> template <typename Problem>
CK_TILE_HOST_DEVICE static constexpr auto GetBlockWelfordSync() CK_TILE_HOST_DEVICE static constexpr auto GetBlockNormReduceSync()
{ {
using P_ = BlockWelfordProblem<typename Problem::ComputeDataType, using P_ = BlockNormReduceProblem<typename Problem::ComputeDataType,
typename Problem::ComputeDataType, typename Problem::ComputeDataType,
typename Problem::BlockShape, typename Problem::BlockShape,
Problem::Traits::kFastFDiv>; Problem::Traits::kFastFDiv,
Problem::Traits::kWelford>;
return BlockWelfordSync<P_>{}; return BlockNormReduceSync<P_>{};
} }
template <typename Problem> template <typename Problem>
CK_TILE_HOST_DEVICE static constexpr auto GetBlockWelfordCrossWarpSync() CK_TILE_HOST_DEVICE static constexpr auto GetBlockNormReduceCrossWarpSync()
{ {
using P_ = BlockWelfordProblem<typename Problem::ComputeDataType, using P_ = BlockNormReduceProblem<typename Problem::ComputeDataType,
typename Problem::ComputeDataType, typename Problem::ComputeDataType,
typename Problem::BlockShape, typename Problem::BlockShape,
Problem::Traits::kFastFDiv>; Problem::Traits::kFastFDiv,
Problem::Traits::kWelford>;
return BlockWelfordCrossWarpSync<P_>{}; return BlockNormReduceCrossWarpSync<P_>{};
} }
template <typename Problem> template <typename Problem>
...@@ -80,19 +82,20 @@ struct Layernorm2dFwdPipelineDefaultPolicy ...@@ -80,19 +82,20 @@ struct Layernorm2dFwdPipelineDefaultPolicy
{ {
if constexpr(Problem::kNeedCrossWarpSync) if constexpr(Problem::kNeedCrossWarpSync)
{ {
using P_ = BlockWelfordProblem<typename Problem::ComputeDataType, using P_ = BlockNormReduceProblem<typename Problem::ComputeDataType,
typename Problem::ComputeDataType, typename Problem::ComputeDataType,
typename Problem::BlockShape, typename Problem::BlockShape,
Problem::Traits::kFastFDiv>; Problem::Traits::kFastFDiv,
Problem::Traits::kWelford>;
using block_welford = BlockWelford<P_>; using block_welford = BlockNormReduce<P_>;
using x_block_tile = using x_block_tile =
decltype(make_static_distributed_tensor<typename Problem::ComputeDataType>( decltype(make_static_distributed_tensor<typename Problem::ComputeDataType>(
MakeXBlockTileDistribution<Problem>())); MakeXBlockTileDistribution<Problem>()));
using mean_var_block_tile = using mean_var_block_tile =
decltype(block_welford::template MakeMeanVarBlockTile<x_block_tile>()); decltype(block_welford::template MakeMeanVarBlockTile<x_block_tile>());
return GetBlockWelfordCrossWarpSync<Problem>() return GetBlockNormReduceCrossWarpSync<Problem>()
.template GetSmemSize<mean_var_block_tile>(); .template GetSmemSize<mean_var_block_tile>();
} }
else else
......
include/ck_tile/ops/layernorm2d/pipeline/layernorm2d_fwd_pipeline_one_pass.hpp
View file @ 1b616990
// SPDX-License-Identifier: MIT // 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 #pragma once
...@@ -18,6 +18,7 @@ struct Layernorm2dFwdPipelineOnePass ...@@ -18,6 +18,7 @@ struct Layernorm2dFwdPipelineOnePass
using Policy = ck_tile::remove_cvref_t<Policy_>; using Policy = ck_tile::remove_cvref_t<Policy_>;
using XDataType = ck_tile::remove_cvref_t<typename Problem::XDataType>; using XDataType = ck_tile::remove_cvref_t<typename Problem::XDataType>;
using XBiasDataType = ck_tile::remove_cvref_t<typename Problem::XBiasDataType>;
using GammaDataType = ck_tile::remove_cvref_t<typename Problem::GammaDataType>; using GammaDataType = ck_tile::remove_cvref_t<typename Problem::GammaDataType>;
using BetaDataType = ck_tile::remove_cvref_t<typename Problem::BetaDataType>; using BetaDataType = ck_tile::remove_cvref_t<typename Problem::BetaDataType>;
using ComputeDataType = ck_tile::remove_cvref_t<typename Problem::ComputeDataType>; using ComputeDataType = ck_tile::remove_cvref_t<typename Problem::ComputeDataType>;
...@@ -37,6 +38,8 @@ struct Layernorm2dFwdPipelineOnePass ...@@ -37,6 +38,8 @@ struct Layernorm2dFwdPipelineOnePass
static constexpr bool kPadM = false; // TODO - BlockLayernorm2dFwdProblem::kPadM static constexpr bool kPadM = false; // TODO - BlockLayernorm2dFwdProblem::kPadM
static constexpr bool kPadN = Problem::Traits::kPadN; static constexpr bool kPadN = Problem::Traits::kPadN;
static constexpr bool kFastFDiv = Problem::Traits::kFastFDiv; static constexpr bool kFastFDiv = Problem::Traits::kFastFDiv;
static constexpr bool kWelford = Problem::Traits::kWelford;
static constexpr auto kXbias = Problem::Traits::kXbias;
static constexpr auto kFusedAdd = Problem::Traits::kFusedAdd; static constexpr auto kFusedAdd = Problem::Traits::kFusedAdd;
static constexpr auto kFusedQuant = Problem::Traits::kFusedQuant; static constexpr auto kFusedQuant = Problem::Traits::kFusedQuant;
...@@ -54,24 +57,26 @@ struct Layernorm2dFwdPipelineOnePass ...@@ -54,24 +57,26 @@ struct Layernorm2dFwdPipelineOnePass
template <typename XWindow, template <typename XWindow,
typename XResidualWindow, typename XResidualWindow,
typename XBiasWindow,
typename GammaWindow, typename GammaWindow,
typename BetaWindow, typename BetaWindow,
typename YWindow, typename YWindow,
typename YResidualWindow, typename YResidualWindow,
typename MeanWindow, typename MeanWindow,
typename InvStdWindow, typename InvStdWindow,
typename XScaleWindow, typename SmoothScaleWindow,
typename YScaleWindow, typename YScaleWindow,
typename Epilogue> typename Epilogue>
CK_TILE_DEVICE auto operator()(const XWindow& x_window_, CK_TILE_DEVICE auto operator()(const XWindow& x_window_,
const XResidualWindow& x_residual_window_, const XResidualWindow& x_residual_window_,
const XBiasWindow& x_bias_window_,
const GammaWindow& gamma_window_, const GammaWindow& gamma_window_,
const BetaWindow& beta_window_, const BetaWindow& beta_window_,
YWindow& y_window_, YWindow& y_window_,
const YResidualWindow& y_residual_window_, const YResidualWindow& y_residual_window_,
MeanWindow& mean_window, MeanWindow& mean_window,
InvStdWindow& inv_std_window, InvStdWindow& inv_std_window,
const XScaleWindow& x_scale_window_, const SmoothScaleWindow& sm_scale_window_,
YScaleWindow& y_scale_window, YScaleWindow& y_scale_window,
ComputeDataType epsilon, ComputeDataType epsilon,
ck_tile::index_t row_size, ck_tile::index_t row_size,
...@@ -80,6 +85,8 @@ struct Layernorm2dFwdPipelineOnePass ...@@ -80,6 +85,8 @@ struct Layernorm2dFwdPipelineOnePass
{ {
const auto x_window = const auto x_window =
make_tile_window(x_window_, Policy::template MakeXBlockTileDistribution<Problem>()); make_tile_window(x_window_, Policy::template MakeXBlockTileDistribution<Problem>());
const auto x_bias_window = make_tile_window(
x_bias_window_, Policy::template MakeGammaBetaBlockTileDistribution<Problem>());
const auto gamma_window = make_tile_window( const auto gamma_window = make_tile_window(
gamma_window_, Policy::template MakeGammaBetaBlockTileDistribution<Problem>()); gamma_window_, Policy::template MakeGammaBetaBlockTileDistribution<Problem>());
const auto beta_window = make_tile_window( const auto beta_window = make_tile_window(
...@@ -89,23 +96,38 @@ struct Layernorm2dFwdPipelineOnePass ...@@ -89,23 +96,38 @@ struct Layernorm2dFwdPipelineOnePass
auto y_residual_window = make_tile_window( auto y_residual_window = make_tile_window(
y_residual_window_, Policy::template MakeXBlockTileDistribution<Problem>()); y_residual_window_, Policy::template MakeXBlockTileDistribution<Problem>());
auto x = load_tile(x_window); auto x = load_tile(x_window);
auto x_resi = load_tile(x_residual_window); auto x_resi = load_tile(x_residual_window);
const auto x_bias = load_tile(x_bias_window);
int cur_count = 0; int cur_count = 0;
int max_count = int max_count =
block_tile_welford_calculate_max_count<typename Problem::BlockShape>(row_size); block_tile_welford_calculate_max_count<typename Problem::BlockShape>(row_size);
auto block_welford = Policy::template GetBlockWelford<Problem>(); auto block_norm_reduce = Policy::template GetBlockNormReduce<Problem>();
auto block_welford_sync = Policy::template GetBlockWelfordSync<Problem>(); auto block_norm_reduce_sync = Policy::template GetBlockNormReduceSync<Problem>();
auto block_welford_cross_warp_sync = auto block_norm_reduce_cross_warp_sync =
Policy::template GetBlockWelfordCrossWarpSync<Problem>(); Policy::template GetBlockNormReduceCrossWarpSync<Problem>();
using XTensorType = decltype(cast_tile<ComputeDataType>(x));
auto mean = block_norm_reduce.template MakeMeanVarBlockTile<XTensorType>();
auto var = block_norm_reduce.template MakeMeanVarBlockTile<XTensorType>();
clear_tile(mean);
clear_tile(var);
// load gamma/beta (TODO: support no gamma/beta?) // load gamma/beta (TODO: support no gamma/beta?)
const auto gamma = load_tile(gamma_window); const auto gamma = load_tile(gamma_window);
const auto beta = load_tile(beta_window); const auto beta = load_tile(beta_window);
auto acc = cast_tile<ComputeDataType>(x); auto acc = cast_tile<ComputeDataType>(x);
if constexpr(kXbias == Layernorm2dXBiasEnum::ADD_BIAS)
{
sweep_tile(x, [&](auto idx) {
// compute x = bias + x
constexpr auto j_idx = make_tuple(idx[number<1>{}]);
acc(idx) = type_convert<ComputeDataType>(x_bias[j_idx]) + acc(idx);
});
}
if constexpr(kFusedAdd == Layernorm2dFusedAddEnum::PRE_ADD_STORE || if constexpr(kFusedAdd == Layernorm2dFusedAddEnum::PRE_ADD_STORE ||
kFusedAdd == Layernorm2dFusedAddEnum::PRE_ADD) kFusedAdd == Layernorm2dFusedAddEnum::PRE_ADD)
{ {
...@@ -117,12 +139,21 @@ struct Layernorm2dFwdPipelineOnePass ...@@ -117,12 +139,21 @@ struct Layernorm2dFwdPipelineOnePass
store_tile(y_residual_window, cast_tile<YResidualDataType>(acc)); store_tile(y_residual_window, cast_tile<YResidualDataType>(acc));
} }
// compute welford each-thread->cross-lane->cross-warp // compute reduce each-thread->cross-lane->cross-warp
auto [mean, var] = block_welford(acc, cur_count, max_count); block_norm_reduce(acc, mean, var, cur_count, max_count);
block_welford_sync(mean, var, cur_count); block_norm_reduce_sync(mean, var, cur_count);
block_welford_cross_warp_sync(mean, var, cur_count, smem); block_norm_reduce_cross_warp_sync(mean, var, cur_count, smem);
block_tile_welford_post_scale_var(var, cur_count, constant<kFastFDiv>{}); if(kWelford)
{
block_tile_welford_post_scale_var(var, cur_count, constant<kFastFDiv>{});
}
else
{
sweep_tile(mean, [&](auto idx) {
mean(idx) = mean(idx) / type_convert<MeanDataType>(row_size);
var(idx) = var(idx) / type_convert<MeanDataType>(row_size) - mean(idx) * mean(idx);
});
}
// compute inv-std // compute inv-std
auto inv_std = tile_elementwise_in( auto inv_std = tile_elementwise_in(
[&](const auto& v_) { [&](const auto& v_) {
...@@ -153,14 +184,13 @@ struct Layernorm2dFwdPipelineOnePass ...@@ -153,14 +184,13 @@ struct Layernorm2dFwdPipelineOnePass
const auto beta_ = type_convert<ComputeDataType>(beta[j_idx]); const auto beta_ = type_convert<ComputeDataType>(beta[j_idx]);
auto ln_ = (acc[idx] - mean_[i_idx]) * inv_std[i_idx] * gamma_ + beta_; auto ln_ = (acc[idx] - mean_[i_idx]) * inv_std[i_idx] * gamma_ + beta_;
ln(idx) = ln_;
ln(idx) = ln_;
}); });
if constexpr(kFusedQuant == Layernorm2dFusedQuantEnum::DYNAMIC_QUANT || if constexpr(kFusedQuant == Layernorm2dFusedQuantEnum::DYNAMIC_QUANT ||
kFusedQuant == Layernorm2dFusedQuantEnum::SMOOTH_DYNAMIC_QUANT) kFusedQuant == Layernorm2dFusedQuantEnum::SMOOTH_DYNAMIC_QUANT)
{ {
Epilogue{}(y_window_, x_scale_window_, y_scale_window, ln, smem); Epilogue{}(y_window_, sm_scale_window_, y_scale_window, ln, smem);
} }
else else
Epilogue{}(y_window_, ln); Epilogue{}(y_window_, ln);
......
include/ck_tile/ops/layernorm2d/pipeline/layernorm2d_fwd_pipeline_problem.hpp
View file @ 1b616990
// SPDX-License-Identifier: MIT // 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 #pragma once
...@@ -8,28 +8,30 @@ ...@@ -8,28 +8,30 @@
namespace ck_tile { namespace ck_tile {
template <typename XDataType_, template <typename XDataType_,
typename XBiasDataType_,
typename GammaDataType_, typename GammaDataType_,
typename BetaDataType_, typename BetaDataType_,
typename ComputeDataType_, typename ComputeDataType_,
typename YDataType_, typename YDataType_,
typename MeanDataType_, typename MeanDataType_,
typename InvStdDataType_, typename InvStdDataType_,
typename XScaleDataType_, typename SmoothScaleDataType_,
typename YScaleDataType_, typename YScaleDataType_,
typename BlockShape_, typename BlockShape_,
typename Traits_> typename Traits_>
struct Layernorm2dFwdPipelineProblem struct Layernorm2dFwdPipelineProblem
{ {
using XDataType = remove_cvref_t<XDataType_>; using XDataType = remove_cvref_t<XDataType_>;
using GammaDataType = remove_cvref_t<GammaDataType_>; using XBiasDataType = remove_cvref_t<XBiasDataType_>;
using BetaDataType = remove_cvref_t<BetaDataType_>; using GammaDataType = remove_cvref_t<GammaDataType_>;
using ComputeDataType = remove_cvref_t<ComputeDataType_>; using BetaDataType = remove_cvref_t<BetaDataType_>;
using YDataType = remove_cvref_t<YDataType_>; using ComputeDataType = remove_cvref_t<ComputeDataType_>;
using MeanDataType = remove_cvref_t<MeanDataType_>; using YDataType = remove_cvref_t<YDataType_>;
using InvStdDataType = remove_cvref_t<InvStdDataType_>; using MeanDataType = remove_cvref_t<MeanDataType_>;
using XScaleDataType = remove_cvref_t<XScaleDataType_>; using InvStdDataType = remove_cvref_t<InvStdDataType_>;
using YScaleDataType = remove_cvref_t<YScaleDataType_>; using SmoothScaleDataType = remove_cvref_t<SmoothScaleDataType_>;
using BlockShape = remove_cvref_t<BlockShape_>; using YScaleDataType = remove_cvref_t<YScaleDataType_>;
using BlockShape = remove_cvref_t<BlockShape_>;
static constexpr bool kNeedCrossLaneSync = BlockShape::ThreadPerWarp_N > 1; static constexpr bool kNeedCrossLaneSync = BlockShape::ThreadPerWarp_N > 1;
static constexpr bool kNeedCrossWarpSync = BlockShape::WarpPerBlock_N > 1; static constexpr bool kNeedCrossWarpSync = BlockShape::WarpPerBlock_N > 1;
......
include/ck_tile/ops/layernorm2d/pipeline/layernorm2d_fwd_pipeline_two_pass.hpp
View file @ 1b616990
// SPDX-License-Identifier: MIT // 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 #pragma once
...@@ -17,6 +17,7 @@ struct Layernorm2dFwdPipelineTwoPass ...@@ -17,6 +17,7 @@ struct Layernorm2dFwdPipelineTwoPass
using Policy = ck_tile::remove_cvref_t<Policy_>; using Policy = ck_tile::remove_cvref_t<Policy_>;
using XDataType = ck_tile::remove_cvref_t<typename Problem::XDataType>; using XDataType = ck_tile::remove_cvref_t<typename Problem::XDataType>;
using XBiasDataType = ck_tile::remove_cvref_t<typename Problem::XBiasDataType>;
using GammaDataType = ck_tile::remove_cvref_t<typename Problem::GammaDataType>; using GammaDataType = ck_tile::remove_cvref_t<typename Problem::GammaDataType>;
using BetaDataType = ck_tile::remove_cvref_t<typename Problem::BetaDataType>; using BetaDataType = ck_tile::remove_cvref_t<typename Problem::BetaDataType>;
using ComputeDataType = ck_tile::remove_cvref_t<typename Problem::ComputeDataType>; using ComputeDataType = ck_tile::remove_cvref_t<typename Problem::ComputeDataType>;
...@@ -36,6 +37,8 @@ struct Layernorm2dFwdPipelineTwoPass ...@@ -36,6 +37,8 @@ struct Layernorm2dFwdPipelineTwoPass
static constexpr bool kPadM = false; // TODO - BlockLayernorm2dFwdProblem::kPadM static constexpr bool kPadM = false; // TODO - BlockLayernorm2dFwdProblem::kPadM
static constexpr bool kPadN = Problem::Traits::kPadN; static constexpr bool kPadN = Problem::Traits::kPadN;
static constexpr bool kFastFDiv = Problem::Traits::kFastFDiv; static constexpr bool kFastFDiv = Problem::Traits::kFastFDiv;
static constexpr bool kWelford = Problem::Traits::kWelford;
static constexpr auto kXbias = Problem::Traits::kXbias;
static constexpr auto kFusedAdd = Problem::Traits::kFusedAdd; static constexpr auto kFusedAdd = Problem::Traits::kFusedAdd;
static constexpr auto kFusedQuant = Problem::Traits::kFusedQuant; static constexpr auto kFusedQuant = Problem::Traits::kFusedQuant;
...@@ -53,32 +56,37 @@ struct Layernorm2dFwdPipelineTwoPass ...@@ -53,32 +56,37 @@ struct Layernorm2dFwdPipelineTwoPass
template <typename XWindow, template <typename XWindow,
typename XResidualWindow, typename XResidualWindow,
typename XBiasWindow,
typename GammaWindow, typename GammaWindow,
typename BetaWindow, typename BetaWindow,
typename YWindow, typename YWindow,
typename YResidualWindow, typename YResidualWindow,
typename MeanWindow, typename MeanWindow,
typename InvStdWindow, typename InvStdWindow,
typename XScaleWindow, typename SmoothScaleWindow,
typename YScaleWindow, typename YScaleWindow,
typename Epilogue> typename Epilogue>
CK_TILE_DEVICE auto operator()(const XWindow& x_window_, CK_TILE_DEVICE auto operator()(const XWindow& x_window_,
const XResidualWindow& x_residual_window_, const XResidualWindow& x_residual_window_,
const XBiasWindow& x_bias_window_,
const GammaWindow& gamma_window_, const GammaWindow& gamma_window_,
const BetaWindow& beta_window_, const BetaWindow& beta_window_,
YWindow& y_window, YWindow& y_window,
const YResidualWindow& y_residual_window_, const YResidualWindow& y_residual_window_,
MeanWindow& mean_window, MeanWindow& mean_window,
InvStdWindow& inv_std_window, InvStdWindow& inv_std_window,
const XScaleWindow& /*x_scale_window*/, const SmoothScaleWindow& /*sm_scale_window*/,
YScaleWindow& /*y_scale_window*/, YScaleWindow& /*y_scale_window*/,
ComputeDataType epsilon, ComputeDataType epsilon,
ck_tile::index_t row_size, ck_tile::index_t row_size,
void* smem, void* smem,
Epilogue) const Epilogue) const
{ {
static_assert(kWelford == true, "2 pass only supports welford merge");
auto x_window = auto x_window =
make_tile_window(x_window_, Policy::template MakeXBlockTileDistribution<Problem>()); make_tile_window(x_window_, Policy::template MakeXBlockTileDistribution<Problem>());
auto x_bias_window = make_tile_window(
x_bias_window_, Policy::template MakeGammaBetaBlockTileDistribution<Problem>());
auto gamma_window = make_tile_window( auto gamma_window = make_tile_window(
gamma_window_, Policy::template MakeGammaBetaBlockTileDistribution<Problem>()); gamma_window_, Policy::template MakeGammaBetaBlockTileDistribution<Problem>());
auto beta_window = make_tile_window( auto beta_window = make_tile_window(
...@@ -102,24 +110,35 @@ struct Layernorm2dFwdPipelineTwoPass ...@@ -102,24 +110,35 @@ struct Layernorm2dFwdPipelineTwoPass
int max_count = int max_count =
(num_n_tile_iteration - 1) * count_per_iter + (num_n_tile_iteration - 1) * count_per_iter +
block_tile_welford_calculate_max_count<typename Problem::BlockShape>(last_iter_n); block_tile_welford_calculate_max_count<typename Problem::BlockShape>(last_iter_n);
auto block_welford = Policy::template GetBlockWelford<Problem>(); auto block_norm_reduce = Policy::template GetBlockNormReduce<Problem>();
auto block_welford_sync = Policy::template GetBlockWelfordSync<Problem>(); auto block_norm_reduce_sync = Policy::template GetBlockNormReduceSync<Problem>();
auto block_welford_cross_warp_sync = auto block_norm_reduce_cross_warp_sync =
Policy::template GetBlockWelfordCrossWarpSync<Problem>(); Policy::template GetBlockNormReduceCrossWarpSync<Problem>();
using XTensorType = decltype(cast_tile<ComputeDataType>(load_tile(x_window))); using XTensorType = decltype(cast_tile<ComputeDataType>(load_tile(x_window)));
auto mean = block_welford.template MakeMeanVarBlockTile<XTensorType>(); auto mean = block_norm_reduce.template MakeMeanVarBlockTile<XTensorType>();
auto var = block_welford.template MakeMeanVarBlockTile<XTensorType>(); auto var = block_norm_reduce.template MakeMeanVarBlockTile<XTensorType>();
for(int iN = __builtin_amdgcn_readfirstlane(0); iN < num_n_tile_iteration; ++iN) for(int iN = __builtin_amdgcn_readfirstlane(0); iN < num_n_tile_iteration; ++iN)
{ {
auto x = load_tile(x_window); auto x = load_tile(x_window);
auto x_resi = load_tile(x_residual_window); auto x_resi = load_tile(x_residual_window);
const auto x_bias = load_tile(x_bias_window);
move_tile_window(x_window, {0, Block_N}); move_tile_window(x_window, {0, Block_N});
move_tile_window(x_residual_window, {0, Block_N}); move_tile_window(x_residual_window, {0, Block_N});
move_tile_window(x_bias_window, {Block_N});
auto acc = cast_tile<ComputeDataType>(x); auto acc = cast_tile<ComputeDataType>(x);
if constexpr(kXbias == Layernorm2dXBiasEnum::ADD_BIAS)
{
sweep_tile(x, [&](auto idx) {
// compute x = bias + x
constexpr auto j_idx = make_tuple(idx[number<1>{}]);
acc(idx) = type_convert<ComputeDataType>(x_bias[j_idx]) + acc(idx);
});
}
if constexpr(kFusedAdd == Layernorm2dFusedAddEnum::PRE_ADD_STORE || if constexpr(kFusedAdd == Layernorm2dFusedAddEnum::PRE_ADD_STORE ||
kFusedAdd == Layernorm2dFusedAddEnum::PRE_ADD) kFusedAdd == Layernorm2dFusedAddEnum::PRE_ADD)
{ {
...@@ -133,11 +152,11 @@ struct Layernorm2dFwdPipelineTwoPass ...@@ -133,11 +152,11 @@ struct Layernorm2dFwdPipelineTwoPass
move_tile_window(y_residual_window, {0, Block_N}); move_tile_window(y_residual_window, {0, Block_N});
} }
} }
block_welford(acc, mean, var, cur_count, max_count); block_norm_reduce(acc, mean, var, cur_count, max_count);
} }
block_welford_sync(mean, var, cur_count); block_norm_reduce_sync(mean, var, cur_count);
block_welford_cross_warp_sync(mean, var, cur_count, smem); block_norm_reduce_cross_warp_sync(mean, var, cur_count, smem);
block_tile_welford_post_scale_var(var, cur_count, constant<kFastFDiv>{}); block_tile_welford_post_scale_var(var, cur_count, constant<kFastFDiv>{});
// compute inv-std // compute inv-std
...@@ -165,6 +184,7 @@ struct Layernorm2dFwdPipelineTwoPass ...@@ -165,6 +184,7 @@ struct Layernorm2dFwdPipelineTwoPass
move_tile_window(x_window, {0, -Block_N}); move_tile_window(x_window, {0, -Block_N});
move_tile_window(x_residual_window, {0, -Block_N}); move_tile_window(x_residual_window, {0, -Block_N});
move_tile_window(x_bias_window, {-Block_N});
move_tile_window(gamma_window, {stride_to_right_most_window}); move_tile_window(gamma_window, {stride_to_right_most_window});
move_tile_window(beta_window, {stride_to_right_most_window}); move_tile_window(beta_window, {stride_to_right_most_window});
move_tile_window(y_window, {0, stride_to_right_most_window}); move_tile_window(y_window, {0, stride_to_right_most_window});
...@@ -172,9 +192,19 @@ struct Layernorm2dFwdPipelineTwoPass ...@@ -172,9 +192,19 @@ struct Layernorm2dFwdPipelineTwoPass
// layernorm computation // layernorm computation
for(int iN = __builtin_amdgcn_readfirstlane(0); iN < num_n_tile_iteration; ++iN) for(int iN = __builtin_amdgcn_readfirstlane(0); iN < num_n_tile_iteration; ++iN)
{ {
auto x = load_tile(x_window); auto x = load_tile(x_window);
auto x_resi = load_tile(x_residual_window); auto x_resi = load_tile(x_residual_window);
auto acc = cast_tile<ComputeDataType>(x); const auto x_bias = load_tile(x_bias_window);
auto acc = cast_tile<ComputeDataType>(x);
if constexpr(kXbias == Layernorm2dXBiasEnum::ADD_BIAS)
{
sweep_tile(x, [&](auto idx) {
// compute x = bias + x
constexpr auto j_idx = make_tuple(idx[number<1>{}]);
acc(idx) = type_convert<ComputeDataType>(x_bias[j_idx]) + acc(idx);
});
}
if constexpr(kFusedAdd == Layernorm2dFusedAddEnum::PRE_ADD_STORE || if constexpr(kFusedAdd == Layernorm2dFusedAddEnum::PRE_ADD_STORE ||
kFusedAdd == Layernorm2dFusedAddEnum::PRE_ADD) kFusedAdd == Layernorm2dFusedAddEnum::PRE_ADD)
...@@ -207,6 +237,7 @@ struct Layernorm2dFwdPipelineTwoPass ...@@ -207,6 +237,7 @@ struct Layernorm2dFwdPipelineTwoPass
move_tile_window(x_window, {0, -Block_N}); move_tile_window(x_window, {0, -Block_N});
move_tile_window(x_residual_window, {0, -Block_N}); move_tile_window(x_residual_window, {0, -Block_N});
move_tile_window(x_bias_window, {-Block_N});
move_tile_window(gamma_window, {-Block_N}); move_tile_window(gamma_window, {-Block_N});
move_tile_window(beta_window, {-Block_N}); move_tile_window(beta_window, {-Block_N});
move_tile_window(y_window, {0, -Block_N}); move_tile_window(y_window, {0, -Block_N});
......
include/ck_tile/ops/layernorm2d/pipeline/layernorm2d_fwd_traits.hpp
View file @ 1b616990
...@@ -7,6 +7,19 @@ ...@@ -7,6 +7,19 @@
namespace ck_tile { namespace ck_tile {
enum class Layernorm2dXBiasEnum
{
NO_BIAS = 0,
// add bias before fused add
ADD_BIAS = 1,
};
// clang-format off
template<Layernorm2dXBiasEnum> struct Layernorm2dXBiasEnumName;
template<> struct Layernorm2dXBiasEnumName<Layernorm2dXBiasEnum::NO_BIAS> { static constexpr const char * name = "no"; };
template<> struct Layernorm2dXBiasEnumName<Layernorm2dXBiasEnum::ADD_BIAS> { static constexpr const char * name = "xbias"; };
// clang-format on
enum class Layernorm2dFusedAddEnum enum class Layernorm2dFusedAddEnum
{ {
NO_ADD = 0, NO_ADD = 0,
...@@ -40,7 +53,9 @@ template<> struct Layernorm2dFusedQuantEnumName<Layernorm2dFusedQuantEnum::SMOOT ...@@ -40,7 +53,9 @@ template<> struct Layernorm2dFusedQuantEnumName<Layernorm2dFusedQuantEnum::SMOOT
template <bool kPadN_, template <bool kPadN_,
bool kSaveMeanInvStd_, bool kSaveMeanInvStd_,
bool kFastFDiv_, bool kFastFDiv_,
bool kWelford_,
bool kTwoPass_, bool kTwoPass_,
Layernorm2dXBiasEnum kXbias_,
Layernorm2dFusedAddEnum kFusedAdd_, Layernorm2dFusedAddEnum kFusedAdd_,
Layernorm2dFusedQuantEnum kFusedQuant_> Layernorm2dFusedQuantEnum kFusedQuant_>
struct Layernorm2dFwdTraits struct Layernorm2dFwdTraits
...@@ -48,7 +63,9 @@ struct Layernorm2dFwdTraits ...@@ -48,7 +63,9 @@ struct Layernorm2dFwdTraits
static constexpr bool kPadN = kPadN_; static constexpr bool kPadN = kPadN_;
static constexpr bool kSaveMeanInvStd = kSaveMeanInvStd_; static constexpr bool kSaveMeanInvStd = kSaveMeanInvStd_;
static constexpr bool kFastFDiv = kFastFDiv_; static constexpr bool kFastFDiv = kFastFDiv_;
static constexpr bool kWelford = kWelford_;
static constexpr bool kTwoPass = kTwoPass_; static constexpr bool kTwoPass = kTwoPass_;
static constexpr Layernorm2dXBiasEnum kXbias = kXbias_;
static constexpr Layernorm2dFusedAddEnum kFusedAdd = kFusedAdd_; static constexpr Layernorm2dFusedAddEnum kFusedAdd = kFusedAdd_;
static constexpr Layernorm2dFusedQuantEnum kFusedQuant = kFusedQuant_; static constexpr Layernorm2dFusedQuantEnum kFusedQuant = kFusedQuant_;
}; };
......
include/ck_tile/ops/norm_reduce.hpp 0 → 100644
View file @ 1b616990
// SPDX-License-Identifier: MIT
// Copyright (c) 2018-2025, Advanced Micro Devices, Inc. All rights reserved.
#pragma once
#include "ck_tile/ops/norm_reduce/block/block_norm_reduce.hpp"
#include "ck_tile/ops/norm_reduce/block/block_norm_reduce_problem.hpp"
#include "ck_tile/ops/norm_reduce/thread/thread_welford.hpp"
#include "ck_tile/ops/common/generic_2d_block_shape.hpp"
#include "ck_tile/ops/common/tensor_layout.hpp"
include/ck_tile/ops/welford/block/block_welford.hpp → include/ck_tile/ops/norm_reduce/block/block_norm_reduce.hpp
View file @ 1b616990
...@@ -4,22 +4,23 @@ ...@@ -4,22 +4,23 @@
#pragma once #pragma once
#include "ck_tile/core.hpp" #include "ck_tile/core.hpp"
#include "ck_tile/ops/welford/thread/thread_welford.hpp" #include "ck_tile/ops/norm_reduce/thread/thread_welford.hpp"
namespace ck_tile { namespace ck_tile {
template <typename Problem_, typename Policy_ = void> template <typename Problem_, typename Policy_ = void>
struct BlockWelford struct BlockNormReduce
{ {
using Problem = remove_cvref_t<Problem_>; using Problem = remove_cvref_t<Problem_>;
using XDataType = typename Problem::XDataType; using XDataType = typename Problem::XDataType;
using ComputeDataType = typename Problem::ComputeDataType; using ComputeDataType = typename Problem::ComputeDataType;
static constexpr bool kFastFDiv = Problem::kFastFDiv; static constexpr bool kFastFDiv = Problem::kFastFDiv;
static constexpr bool kWelford = Problem::kWelford;
CK_TILE_DEVICE constexpr BlockWelford() {} CK_TILE_DEVICE constexpr BlockNormReduce() {}
// [CAUSION] - max_count_ is to deal with the padding problem // [CAUSION] - max_count_ is to deal with the padding problem
// max_count_ is depend on caller, eg: naive and splitN welford will have different // max_count_ is depend on caller, eg: naive and splitN norm_reduce will have different
// calculation of max_count_ // calculation of max_count_
// -> use block_welford_calculate_max_count to compute // -> use block_welford_calculate_max_count to compute
template <typename XDistributedTensor_, template <typename XDistributedTensor_,
...@@ -40,18 +41,24 @@ struct BlockWelford ...@@ -40,18 +41,24 @@ struct BlockWelford
if(cur_count_ < max_count_) if(cur_count_ < max_count_)
{ {
++cur_count_; ++cur_count_;
sweep_tile_span(spans[I0], [&](auto dstr_idx_i0) { sweep_tile_span(spans[I0], [&](auto dstr_idx_i0) {
constexpr auto in_dstr_idx = make_tuple(dstr_idx_i0, dstr_idx_i1); constexpr auto in_dstr_idx = make_tuple(dstr_idx_i0, dstr_idx_i1);
constexpr auto out_dstr_idx = make_tuple(dstr_idx_i0); constexpr auto out_dstr_idx = make_tuple(dstr_idx_i0);
auto x = ck_tile::type_convert<ComputeDataType>(x_tensor[in_dstr_idx]); auto x = ck_tile::type_convert<ComputeDataType>(x_tensor[in_dstr_idx]);
if(kWelford)
welford_update(mean_tensor(out_dstr_idx), {
var_tensor(out_dstr_idx), welford_update(mean_tensor(out_dstr_idx),
x, var_tensor(out_dstr_idx),
cur_count_, x,
constant<kFastFDiv>{}); cur_count_,
constant<kFastFDiv>{});
}
else
{
mean_tensor(out_dstr_idx) += x;
var_tensor(out_dstr_idx) += x * x;
}
}); });
} }
}); });
...@@ -91,10 +98,11 @@ struct BlockWelford ...@@ -91,10 +98,11 @@ struct BlockWelford
}; };
template <typename Problem_, typename Policy_ = void> template <typename Problem_, typename Policy_ = void>
struct BlockWelfordSync struct BlockNormReduceSync
{ {
using Problem = remove_cvref_t<Problem_>; using Problem = remove_cvref_t<Problem_>;
static constexpr bool kFastFDiv = Problem::kFastFDiv; static constexpr bool kFastFDiv = Problem::kFastFDiv;
static constexpr bool kWelford = Problem::kWelford;
template <typename MeanDistributedTensor_, typename VarDistributedTensor_> template <typename MeanDistributedTensor_, typename VarDistributedTensor_>
CK_TILE_DEVICE void CK_TILE_DEVICE void
...@@ -152,36 +160,48 @@ struct BlockWelfordSync ...@@ -152,36 +160,48 @@ struct BlockWelfordSync
(number<lid_over_rid_derivative << istage.value>{}.value); (number<lid_over_rid_derivative << istage.value>{}.value);
// pull data from remote lane // pull data from remote lane
const auto v_remote_mean = warp_shuffle(v_local_mean, src_lane); const auto v_remote_mean = warp_shuffle(v_local_mean, src_lane);
const auto v_remote_var = warp_shuffle(v_local_var, src_lane); const auto v_remote_var = warp_shuffle(v_local_var, src_lane);
const auto v_remote_count = warp_shuffle(v_local_count, src_lane); if(kWelford)
{
// welford merge const auto v_remote_count = warp_shuffle(v_local_count, src_lane);
welford_merge(v_local_mean,
v_local_var, // norm_reduce merge
v_local_count, welford_merge(v_local_mean,
v_remote_mean, v_local_var,
v_remote_var, v_local_count,
v_remote_count, v_remote_mean,
constant<kFastFDiv>{}); v_remote_var,
v_remote_count,
constant<kFastFDiv>{});
}
else
{
v_local_mean += v_remote_mean;
v_local_var += v_remote_var;
}
}); });
} }
}); });
mean_tensor.get_thread_buffer()(i) = v_local_mean; mean_tensor.get_thread_buffer()(i) = v_local_mean;
var_tensor.get_thread_buffer()(i) = v_local_var; var_tensor.get_thread_buffer()(i) = v_local_var;
if(kWelford)
count = v_local_count; {
count = v_local_count;
}
}); });
} }
}; };
template <typename Problem_, typename Policy_ = void> template <typename Problem_, typename Policy_ = void>
struct BlockWelfordCrossWarpSync struct BlockNormReduceCrossWarpSync
{ {
using Problem = remove_cvref_t<Problem_>; using Problem = remove_cvref_t<Problem_>;
using BlockShape = typename Problem::BlockShape; using BlockShape = typename Problem::BlockShape;
static constexpr bool kFastFDiv = Problem::kFastFDiv; static constexpr bool kFastFDiv = Problem::kFastFDiv;
static constexpr bool kWelford = Problem::kWelford;
using smem_dtype = std::conditional_t<kWelford, fp32x4_t, fp32x2_t>;
template <typename MeanDistributedTensor_> template <typename MeanDistributedTensor_>
CK_TILE_DEVICE static constexpr index_t GetReduceWarps() CK_TILE_DEVICE static constexpr index_t GetReduceWarps()
...@@ -252,7 +272,7 @@ struct BlockWelfordCrossWarpSync ...@@ -252,7 +272,7 @@ struct BlockWelfordCrossWarpSync
static_assert(thread_buf_size == VarDistributedTensor_::get_thread_buffer_size()); static_assert(thread_buf_size == VarDistributedTensor_::get_thread_buffer_size());
// Note: we always pack everything into fp32x4 // Note: we always pack everything into fp32x4
fp32x4_t* smem_ptr = reinterpret_cast<fp32x4_t*>(smem); smem_dtype* smem_ptr = reinterpret_cast<smem_dtype*>(smem);
const index_t lane_id = get_lane_id(); const index_t lane_id = get_lane_id();
const index_t warp_id = get_warp_id(); const index_t warp_id = get_warp_id();
constexpr auto num_reduce_warps = GetReduceWarps<MeanDistributedTensor_>(); constexpr auto num_reduce_warps = GetReduceWarps<MeanDistributedTensor_>();
...@@ -267,11 +287,13 @@ struct BlockWelfordCrossWarpSync ...@@ -267,11 +287,13 @@ struct BlockWelfordCrossWarpSync
if(lane_id == 0) if(lane_id == 0)
{ {
static_for<0, thread_buf_size, 1>{}([&](auto i) { static_for<0, thread_buf_size, 1>{}([&](auto i) {
fp32x4_t local_scratch_; smem_dtype local_scratch_;
local_scratch_[0] = bit_cast<float>(mean_tensor.get_thread_buffer()[i]); local_scratch_[0] = bit_cast<float>(mean_tensor.get_thread_buffer()[i]);
local_scratch_[1] = bit_cast<float>(var_tensor.get_thread_buffer()[i]); local_scratch_[1] = bit_cast<float>(var_tensor.get_thread_buffer()[i]);
local_scratch_[2] = bit_cast<float>(count); if(kWelford)
{
local_scratch_[2] = bit_cast<float>(count);
}
smem_ptr[smem_offset + i * num_warps] = local_scratch_; smem_ptr[smem_offset + i * num_warps] = local_scratch_;
}); });
} }
...@@ -280,7 +302,7 @@ struct BlockWelfordCrossWarpSync ...@@ -280,7 +302,7 @@ struct BlockWelfordCrossWarpSync
// load from smem. here we let everythread to do compute :) // load from smem. here we let everythread to do compute :)
index_t local_warp_id = warp_id / num_reduce_warps; index_t local_warp_id = warp_id / num_reduce_warps;
index_t local_smem_os = local_warp_id * num_reduce_warps; index_t local_smem_os = local_warp_id * num_reduce_warps;
fp32x4_t all_scratch[thread_buf_size * num_reduce_warps]; smem_dtype all_scratch[thread_buf_size * num_reduce_warps];
static_for<0, thread_buf_size, 1>{}([&](auto i_0) { static_for<0, thread_buf_size, 1>{}([&](auto i_0) {
static_for<0, num_reduce_warps, 1>{}([&](auto i_1) { static_for<0, num_reduce_warps, 1>{}([&](auto i_1) {
all_scratch[i_0 * num_reduce_warps + i_1] = all_scratch[i_0 * num_reduce_warps + i_1] =
...@@ -293,32 +315,40 @@ struct BlockWelfordCrossWarpSync ...@@ -293,32 +315,40 @@ struct BlockWelfordCrossWarpSync
static_for<0, thread_buf_size, 1>{}([&](auto i_0) { static_for<0, thread_buf_size, 1>{}([&](auto i_0) {
// TODO: use descriptor for this // TODO: use descriptor for this
auto v_local = all_scratch[i_0 * num_reduce_warps]; auto v_local = all_scratch[i_0 * num_reduce_warps];
auto v_local_mean = bit_cast<DataType>(v_local[0]); auto v_local_mean = bit_cast<DataType>(v_local[0]);
auto v_local_var = bit_cast<DataType>(v_local[1]); auto v_local_var = bit_cast<DataType>(v_local[1]);
auto v_local_count = bit_cast<int>(v_local[2]); int v_local_count = kWelford ? bit_cast<int>(v_local[2]) : 0;
// further reduce mean/var // further reduce mean/var
static_for<0, num_reduce_warps - 1, 1>{}([&](auto i_1_n1) { static_for<0, num_reduce_warps - 1, 1>{}([&](auto i_1_n1) {
constexpr auto i_1 = number<i_1_n1 + 1>{}; constexpr auto i_1 = number<i_1_n1 + 1>{};
const fp32x4_t v_remote = all_scratch[i_0 * num_reduce_warps + i_1]; const smem_dtype v_remote = all_scratch[i_0 * num_reduce_warps + i_1];
const auto v_remote_mean = bit_cast<DataType>(v_remote[0]); const auto v_remote_mean = bit_cast<DataType>(v_remote[0]);
const auto v_remote_var = bit_cast<DataType>(v_remote[1]); const auto v_remote_var = bit_cast<DataType>(v_remote[1]);
const auto v_remote_count = bit_cast<int>(v_remote[2]); if(kWelford)
{
welford_merge(v_local_mean, const auto v_remote_count = bit_cast<int>(v_remote[2]);
v_local_var,
v_local_count, welford_merge(v_local_mean,
v_remote_mean, v_local_var,
v_remote_var, v_local_count,
v_remote_count, v_remote_mean,
constant<kFastFDiv>{}); v_remote_var,
v_remote_count,
constant<kFastFDiv>{});
}
else
{
v_local_mean += v_remote_mean;
v_local_var += v_remote_var;
}
}); });
mean_tensor.get_thread_buffer()(i_0) = v_local_mean; mean_tensor.get_thread_buffer()(i_0) = v_local_mean;
var_tensor.get_thread_buffer()(i_0) = v_local_var; var_tensor.get_thread_buffer()(i_0) = v_local_var;
if(kWelford)
count = v_local_count; count = v_local_count;
}); });
} }
}; };
......
include/ck_tile/ops/welford/block/block_welford_problem.hpp → include/ck_tile/ops/norm_reduce/block/block_norm_reduce_problem.hpp
View file @ 1b616990
...@@ -7,13 +7,18 @@ ...@@ -7,13 +7,18 @@
namespace ck_tile { namespace ck_tile {
template <typename XDataType_, typename ComputeDataType_, typename BlockShape_, bool kFastFDiv_> template <typename XDataType_,
struct BlockWelfordProblem typename ComputeDataType_,
typename BlockShape_,
bool kFastFDiv_,
bool kWelford_>
struct BlockNormReduceProblem
{ {
using XDataType = remove_cvref_t<XDataType_>; using XDataType = remove_cvref_t<XDataType_>;
using ComputeDataType = remove_cvref_t<ComputeDataType_>; using ComputeDataType = remove_cvref_t<ComputeDataType_>;
using BlockShape = remove_cvref_t<BlockShape_>; using BlockShape = remove_cvref_t<BlockShape_>;
static constexpr bool kFastFDiv = kFastFDiv_; static constexpr bool kFastFDiv = kFastFDiv_;
static constexpr bool kWelford = kWelford_;
}; };
} // namespace ck_tile } // namespace ck_tile
include/ck_tile/ops/permute.hpp
View file @ 1b616990
// SPDX-License-Identifier: MIT // 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 #pragma once
......
include/ck_tile/ops/reduce.hpp
View file @ 1b616990
// SPDX-License-Identifier: MIT // 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 #pragma once
......
include/ck_tile/ops/rmsnorm2d.hpp
View file @ 1b616990
// SPDX-License-Identifier: MIT // 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 #pragma once
...@@ -8,5 +8,6 @@ ...@@ -8,5 +8,6 @@
#include "ck_tile/ops/rmsnorm2d/pipeline/rmsnorm2d_fwd_pipeline_one_pass.hpp" #include "ck_tile/ops/rmsnorm2d/pipeline/rmsnorm2d_fwd_pipeline_one_pass.hpp"
#include "ck_tile/ops/rmsnorm2d/pipeline/rmsnorm2d_fwd_pipeline_problem.hpp" #include "ck_tile/ops/rmsnorm2d/pipeline/rmsnorm2d_fwd_pipeline_problem.hpp"
#include "ck_tile/ops/rmsnorm2d/pipeline/rmsnorm2d_fwd_pipeline_two_pass.hpp" #include "ck_tile/ops/rmsnorm2d/pipeline/rmsnorm2d_fwd_pipeline_two_pass.hpp"
#include "ck_tile/ops/rmsnorm2d/pipeline/rmsnorm2d_fwd_traits.hpp"
#include "ck_tile/ops/common/generic_2d_block_shape.hpp" #include "ck_tile/ops/common/generic_2d_block_shape.hpp"
#include "ck_tile/ops/common/tensor_layout.hpp" #include "ck_tile/ops/common/tensor_layout.hpp"
include/ck_tile/ops/rmsnorm2d/kernel/rmsnorm2d_fwd_kernel.hpp
View file @ 1b616990
// SPDX-License-Identifier: MIT // 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 #pragma once
#include "ck_tile/core.hpp" #include "ck_tile/core.hpp"
#include "ck_tile/ops/common.hpp" #include "ck_tile/ops/common.hpp"
#include "ck_tile/ops/rmsnorm2d/pipeline/rmsnorm2d_fwd_traits.hpp"
namespace ck_tile { namespace ck_tile {
// host side args // host side args
struct Rmsnorm2dFwdHostArgs struct Rmsnorm2dFwdHostArgs
{ {
const void* p_x; // [m ,n], input, fp16/bf16 const void* p_x; // [m ,n], input, fp16/bf16
const void* p_gamma; // [1, n], gamma, prec same as input const void* p_x_residual; // [m ,n], shortcut input, prec same as input, nullptr if not used
const void* p_sm_scale; // [1 ,n], smooth scale input, fp32, nullptr if not used
const void* p_gamma; // [1, n], gamma, prec same as input
void* p_y; // [m, n], output, fp16/bf16 void* p_y; // [m, n], output, fp16/bf16
void* p_invRms; // [m, 1], output inv-rms, prec same as input, nullptr if not used void* p_y_residual; // [m, n], shortcut output, prec same as input, nullptr if not used
void* p_y_scale; // [m, 1], output a dynamic quant per row, nullptr if not used
void* p_invRms; // [m, 1], output inv-rms, prec same as input, nullptr if not used
float epsilon; float epsilon;
index_t m; index_t m;
index_t n; index_t n;
index_t stride; // row_stride index_t x_stride; // x row_stride
index_t xr_stride; // x residule row stride
index_t y_stride; // y row stride
index_t yr_stride; // y residule row stride
}; };
// TODO: Extract some type to wrapper class // TODO: Extract some type to wrapper class
template <typename Pipeline_> template <typename Pipeline_, typename Epilogue_>
struct Rmsnorm2dFwd struct Rmsnorm2dFwd
{ {
using Pipeline = remove_cvref_t<Pipeline_>; using Pipeline = remove_cvref_t<Pipeline_>;
using Epilogue = remove_cvref_t<Epilogue_>;
using Problem = typename Pipeline::Problem; using Problem = typename Pipeline::Problem;
using XDataType = remove_cvref_t<typename Problem::XDataType>; using XDataType = remove_cvref_t<typename Problem::XDataType>;
using GammaDataType = remove_cvref_t<typename Problem::GammaDataType>; using GammaDataType = remove_cvref_t<typename Problem::GammaDataType>;
using ComputeDataType = remove_cvref_t<typename Problem::ComputeDataType>; using ComputeDataType = remove_cvref_t<typename Problem::ComputeDataType>;
using YDataType = remove_cvref_t<typename Problem::YDataType>; using YDataType = remove_cvref_t<typename Problem::YDataType>;
using InvRmsDataType = remove_cvref_t<typename Problem::InvRmsDataType>; using InvRmsDataType = remove_cvref_t<typename Problem::InvRmsDataType>;
using SmoothScaleDataType = remove_cvref_t<typename Problem::SmoothScaleDataType>;
using YScaleDataType = remove_cvref_t<typename Problem::YScaleDataType>;
// for simplicity, shortcut input/output type is same as X
using XResidualDataType = XDataType;
using YResidualDataType = XDataType;
static constexpr bool kHasGamma = !std::is_same_v<GammaDataType, null_type>; static constexpr bool kHasGamma = !std::is_same_v<GammaDataType, null_type>;
static constexpr bool kSaveInvRms = Problem::kSaveInvRms; static constexpr bool kSaveInvRms = Problem::Traits::kSaveInvRms;
static constexpr index_t Block_M = Problem::BlockShape::Block_M; static constexpr index_t Block_M = Problem::BlockShape::Block_M;
static constexpr index_t Block_N = Problem::BlockShape::Block_N; static constexpr index_t Block_N = Problem::BlockShape::Block_N;
static constexpr bool kPadM = false; // always no need to pad along M static constexpr bool kPadM = false; // always no need to pad along M
static constexpr bool kPadN = Problem::kPadN; static constexpr bool kPadN = Problem::Traits::kPadN;
static constexpr bool kTwoPass = Problem::kTwoPass; static constexpr bool kTwoPass = Problem::Traits::kTwoPass;
static constexpr auto kFusedAdd = Problem::Traits::kFusedAdd;
static constexpr auto kFusedQuant = Problem::Traits::kFusedQuant;
static constexpr index_t ThreadPerWarp_N = Problem::BlockShape::ThreadPerWarp_N; static constexpr index_t ThreadPerWarp_N = Problem::BlockShape::ThreadPerWarp_N;
static constexpr index_t Vector_N = Problem::BlockShape::Vector_N; static constexpr index_t Vector_N = Problem::BlockShape::Vector_N;
...@@ -56,29 +73,43 @@ struct Rmsnorm2dFwd ...@@ -56,29 +73,43 @@ struct Rmsnorm2dFwd
struct Kargs struct Kargs
{ {
const void* p_x; const void* p_x;
const void* p_x_residual;
const void* p_sm_scale;
const void* p_gamma; const void* p_gamma;
void* p_y; void* p_y;
void* p_y_residual;
void* p_y_scale;
void* p_invRms; void* p_invRms;
float epsilon; float epsilon;
index_t m; index_t m;
index_t n; index_t n;
index_t stride; // row_stride index_t x_stride; // x row_stride
index_t xr_stride; // x residule row stride
index_t y_stride; // y row stride
index_t yr_stride; // y residule row stride
}; };
using Hargs = Rmsnorm2dFwdHostArgs; using Hargs = Rmsnorm2dFwdHostArgs;
CK_TILE_HOST static constexpr Kargs MakeKargs(const Hargs& hargs) CK_TILE_HOST static constexpr Kargs MakeKargs(const Hargs& hargs)
{ {
return Kargs{hargs.p_x, return Kargs{hargs.p_x,
hargs.p_x_residual,
hargs.p_sm_scale,
hargs.p_gamma, hargs.p_gamma,
hargs.p_y, hargs.p_y,
hargs.p_y_residual,
hargs.p_y_scale,
hargs.p_invRms, hargs.p_invRms,
hargs.epsilon, hargs.epsilon,
hargs.m, hargs.m,
hargs.n, hargs.n,
hargs.stride}; hargs.x_stride,
hargs.xr_stride,
hargs.y_stride,
hargs.yr_stride};
} }
CK_TILE_HOST static constexpr auto GridSize(const Hargs& hargs) CK_TILE_HOST static constexpr auto GridSize(const Hargs& hargs)
...@@ -95,6 +126,7 @@ struct Rmsnorm2dFwd ...@@ -95,6 +126,7 @@ struct Rmsnorm2dFwd
template <> struct t2s<ck_tile::bf16_t> { static constexpr const char * name = "bf16"; }; template <> struct t2s<ck_tile::bf16_t> { static constexpr const char * name = "bf16"; };
template <> struct t2s<ck_tile::fp8_t> { static constexpr const char * name = "fp8"; }; template <> struct t2s<ck_tile::fp8_t> { static constexpr const char * name = "fp8"; };
template <> struct t2s<ck_tile::bf8_t> { static constexpr const char * name = "bf8"; }; template <> struct t2s<ck_tile::bf8_t> { static constexpr const char * name = "bf8"; };
template <> struct t2s<ck_tile::int8_t> { static constexpr const char * name = "int8"; };
// clang-format on // clang-format on
// in byte // in byte
...@@ -102,24 +134,41 @@ struct Rmsnorm2dFwd ...@@ -102,24 +134,41 @@ struct Rmsnorm2dFwd
CK_TILE_HOST static std::string GetName() CK_TILE_HOST static std::string GetName()
{ {
#define _SS_ std::string
#define _TS_ std::to_string
// clang-format off // clang-format off
using S_ = typename Problem::BlockShape; using S_ = typename Problem::BlockShape;
auto surfix = [&] () { auto surfix = [&] () {
std::string n; std::string n;
if (kFusedAdd != Rmsnorm2dFusedAddEnum::NO_ADD) n += _SS_("_") + Rmsnorm2dFusedAddEnumName<kFusedAdd>::name;
if (kFusedQuant != Rmsnorm2dFusedQuantEnum::NO_SWEEP) n += _SS_("_") + Rmsnorm2dFusedQuantEnumName<kFusedQuant>::name;
if (kPadN) n += "_pn"; if (kPadN) n += "_pn";
if (kSaveInvRms) n += "_rms"; if (kSaveInvRms) n += "_rms";
if (kTwoPass) n += "_2p"; if (kTwoPass) n += "_2p";
return n; }(); return n; }();
#define _SS_ std::string auto prec_str = [&] () {
#define _TS_ std::to_string std::string base_str = _SS_(t2s<XDataType>::name);
return _SS_("rmsnorm2d_fwd_") + _SS_(t2s<XDataType>::name) + "_" + if (!std::is_same_v<XDataType, YDataType>) {
base_str += _SS_("_") + _SS_(t2s<YDataType>::name);
}
if (kFusedQuant == Rmsnorm2dFusedQuantEnum::SMOOTH_DYNAMIC_QUANT) {
base_str += _SS_("_sx") + _SS_(t2s<SmoothScaleDataType>::name);
base_str += _SS_("_sy") + _SS_(t2s<YScaleDataType>::name);
}
if (kFusedQuant == Rmsnorm2dFusedQuantEnum::DYNAMIC_QUANT) {
base_str += _SS_("_sy") + _SS_(t2s<YScaleDataType>::name);
}
return base_str;
}();
return _SS_("rmsnorm2d_fwd_") + _SS_(prec_str) + "_" +
_TS_(S_::Block_M) + "x" + _TS_(S_::Block_N) + "_" + _TS_(S_::WarpPerBlock_M) + "x" + _TS_(S_::WarpPerBlock_N) + "_" + _TS_(S_::Block_M) + "x" + _TS_(S_::Block_N) + "_" + _TS_(S_::WarpPerBlock_M) + "x" + _TS_(S_::WarpPerBlock_N) + "_" +
_TS_(S_::Warp_M) + "x" + _TS_(S_::Warp_N) + "_" + _TS_(S_::Vector_M) + "x" + _TS_(S_::Vector_N) + "_" + _TS_(S_::Warp_M) + "x" + _TS_(S_::Warp_N) + "_" + _TS_(S_::Vector_M) + "x" + _TS_(S_::Vector_N) + "_" +
_SS_(Pipeline::name) + surfix; _SS_(Pipeline::name) + surfix;
#undef _SS_
#undef _TS_
// clang-format on // clang-format on
#undef _SS_
#undef _TS_
} }
CK_TILE_DEVICE void operator()(Kargs kargs) const CK_TILE_DEVICE void operator()(Kargs kargs) const
...@@ -130,7 +179,7 @@ struct Rmsnorm2dFwd ...@@ -130,7 +179,7 @@ struct Rmsnorm2dFwd
const auto tmp_ = make_naive_tensor_view<address_space_enum::global>( const auto tmp_ = make_naive_tensor_view<address_space_enum::global>(
static_cast<const XDataType*>(kargs.p_x), static_cast<const XDataType*>(kargs.p_x),
make_tuple(kargs.m, kargs.n), make_tuple(kargs.m, kargs.n),
make_tuple(kargs.stride, 1), make_tuple(kargs.x_stride, 1),
number<Vector_N>{}, number<Vector_N>{},
number<1>{}); number<1>{});
...@@ -140,6 +189,29 @@ struct Rmsnorm2dFwd ...@@ -140,6 +189,29 @@ struct Rmsnorm2dFwd
tmp2_, make_tuple(number<Block_M>{}, number<Block_N>{}), {iM, 0}); tmp2_, make_tuple(number<Block_M>{}, number<Block_N>{}), {iM, 0});
}(); }();
const auto x_residual_window = [&]() {
if constexpr(kFusedAdd == Rmsnorm2dFusedAddEnum::PRE_ADD ||
kFusedAdd == Rmsnorm2dFusedAddEnum::PRE_ADD_STORE)
{
const auto tmp_ = make_naive_tensor_view<address_space_enum::global>(
static_cast<const XResidualDataType*>(kargs.p_x_residual),
make_tuple(kargs.m, kargs.n),
make_tuple(kargs.xr_stride, 1),
number<Vector_N>{},
number<1>{});
const auto tmp2_ = pad_tensor_view(tmp_,
make_tuple(number<Block_M>{}, number<Block_N>{}),
sequence<kPadM, kPadN>{});
return make_tile_window(
tmp2_, make_tuple(number<Block_M>{}, number<Block_N>{}), {iM, 0});
}
else
{
return make_null_tile_window(make_tuple(number<Block_M>{}, number<Block_N>{}));
}
}();
const auto gamma_window = [&]() { const auto gamma_window = [&]() {
const auto tmp_ = make_naive_tensor_view<address_space_enum::global>( const auto tmp_ = make_naive_tensor_view<address_space_enum::global>(
static_cast<const GammaDataType*>(kargs.p_gamma), static_cast<const GammaDataType*>(kargs.p_gamma),
...@@ -158,7 +230,7 @@ struct Rmsnorm2dFwd ...@@ -158,7 +230,7 @@ struct Rmsnorm2dFwd
auto tmp_ = make_naive_tensor_view<address_space_enum::global>( auto tmp_ = make_naive_tensor_view<address_space_enum::global>(
static_cast<YDataType*>(kargs.p_y), static_cast<YDataType*>(kargs.p_y),
make_tuple(kargs.m, kargs.n), make_tuple(kargs.m, kargs.n),
make_tuple(kargs.stride, 1), make_tuple(kargs.y_stride, 1),
number<Vector_N>{}, number<Vector_N>{},
number<1>{}); number<1>{});
...@@ -168,6 +240,28 @@ struct Rmsnorm2dFwd ...@@ -168,6 +240,28 @@ struct Rmsnorm2dFwd
tmp2_, make_tuple(number<Block_M>{}, number<Block_N>{}), {iM, 0}); tmp2_, make_tuple(number<Block_M>{}, number<Block_N>{}), {iM, 0});
}(); }();
auto y_residual_window = [&]() {
if constexpr(kFusedAdd == Rmsnorm2dFusedAddEnum::PRE_ADD_STORE)
{
auto tmp_ = make_naive_tensor_view<address_space_enum::global>(
static_cast<YResidualDataType*>(kargs.p_y_residual),
make_tuple(kargs.m, kargs.n),
make_tuple(kargs.yr_stride, 1),
number<Vector_N>{},
number<1>{});
auto tmp2_ = pad_tensor_view(tmp_,
make_tuple(number<Block_M>{}, number<Block_N>{}),
sequence<kPadM, kPadN>{});
return make_tile_window(
tmp2_, make_tuple(number<Block_M>{}, number<Block_N>{}), {iM, 0});
}
else
{
return make_null_tile_window(make_tuple(number<Block_M>{}, number<Block_N>{}));
}
}();
auto inv_rms_window = [&]() { auto inv_rms_window = [&]() {
if constexpr(kSaveInvRms) if constexpr(kSaveInvRms)
{ {
...@@ -187,15 +281,62 @@ struct Rmsnorm2dFwd ...@@ -187,15 +281,62 @@ struct Rmsnorm2dFwd
return make_null_tile_window(make_tuple(number<Block_M>{})); return make_null_tile_window(make_tuple(number<Block_M>{}));
}(); }();
auto sm_scale_window = [&]() {
if constexpr(kFusedQuant == Rmsnorm2dFusedQuantEnum::SMOOTH_DYNAMIC_QUANT)
{
const auto win_ = [&]() {
const auto tmp_0_ = make_naive_tensor_view_packed<address_space_enum::global>(
static_cast<const SmoothScaleDataType*>(kargs.p_sm_scale),
make_tuple(kargs.n),
number<Vector_N>{});
return pad_tensor_view(tmp_0_,
make_tuple(number<Block_N>{}),
sequence<false>{}); // sm_scale no need pad
}();
return make_tile_window(win_, make_tuple(number<Block_N>{}), {0});
}
else
{
return make_null_tile_window(make_tuple(number<Block_N>{}));
}
}();
auto y_scale_window = [&]() {
if constexpr(kFusedQuant == Rmsnorm2dFusedQuantEnum::SMOOTH_DYNAMIC_QUANT ||
kFusedQuant == Rmsnorm2dFusedQuantEnum::DYNAMIC_QUANT)
{
const auto win_ = [&]() {
const auto tmp_0_ = make_naive_tensor_view_packed<address_space_enum::global>(
static_cast<YScaleDataType*>(kargs.p_y_scale),
make_tuple(kargs.m),
number<1>{});
return pad_tensor_view(
tmp_0_, make_tuple(number<Block_M>{}), sequence<kPadM>{});
}();
return make_tile_window(win_, make_tuple(number<Block_M>{}), {iM});
}
else
{
return make_null_tile_window(make_tuple(number<Block_M>{}));
}
}();
__shared__ char smem[GetSmemSize()]; __shared__ char smem[GetSmemSize()];
Pipeline{}(x_window, Pipeline{}(x_window,
x_residual_window,
gamma_window, gamma_window,
y_window, y_window,
y_residual_window,
inv_rms_window, inv_rms_window,
sm_scale_window,
y_scale_window,
static_cast<const ComputeDataType>(kargs.epsilon), static_cast<const ComputeDataType>(kargs.epsilon),
kargs.n, kargs.n,
smem); smem,
Epilogue{});
} }
}; };
......
include/ck_tile/ops/rmsnorm2d/pipeline/rmsnorm2d_fwd_pipeline_default_policy.hpp
View file @ 1b616990
...@@ -45,7 +45,7 @@ struct Rmsnorm2dFwdPipelineDefaultPolicy ...@@ -45,7 +45,7 @@ struct Rmsnorm2dFwdPipelineDefaultPolicy
template <typename Problem> template <typename Problem>
CK_TILE_HOST_DEVICE static constexpr auto GetBlockReduce2d() CK_TILE_HOST_DEVICE static constexpr auto GetBlockReduce2d()
{ {
using P_ = BlockReduce2dProblem<typename Problem::XDataType, using P_ = BlockReduce2dProblem<typename Problem::ComputeDataType,
typename Problem::ComputeDataType, typename Problem::ComputeDataType,
typename Problem::BlockShape>; typename Problem::BlockShape>;
return BlockReduce2d<P_>{}; return BlockReduce2d<P_>{};
...@@ -54,7 +54,7 @@ struct Rmsnorm2dFwdPipelineDefaultPolicy ...@@ -54,7 +54,7 @@ struct Rmsnorm2dFwdPipelineDefaultPolicy
template <typename Problem> template <typename Problem>
CK_TILE_HOST_DEVICE static constexpr auto GetBlockReduce2dSync() CK_TILE_HOST_DEVICE static constexpr auto GetBlockReduce2dSync()
{ {
using P_ = BlockReduce2dProblem<typename Problem::XDataType, using P_ = BlockReduce2dProblem<typename Problem::ComputeDataType,
typename Problem::ComputeDataType, typename Problem::ComputeDataType,
typename Problem::BlockShape>; typename Problem::BlockShape>;
return BlockReduce2dSync<P_>{}; return BlockReduce2dSync<P_>{};
...@@ -63,7 +63,7 @@ struct Rmsnorm2dFwdPipelineDefaultPolicy ...@@ -63,7 +63,7 @@ struct Rmsnorm2dFwdPipelineDefaultPolicy
template <typename Problem> template <typename Problem>
CK_TILE_HOST_DEVICE static constexpr auto GetBlockReduce2dCrossWarpSync() CK_TILE_HOST_DEVICE static constexpr auto GetBlockReduce2dCrossWarpSync()
{ {
using P_ = BlockReduce2dProblem<typename Problem::XDataType, using P_ = BlockReduce2dProblem<typename Problem::ComputeDataType,
typename Problem::ComputeDataType, typename Problem::ComputeDataType,
typename Problem::BlockShape>; typename Problem::BlockShape>;
return BlockReduce2dCrossWarpSync<P_>{}; return BlockReduce2dCrossWarpSync<P_>{};
...@@ -74,13 +74,13 @@ struct Rmsnorm2dFwdPipelineDefaultPolicy ...@@ -74,13 +74,13 @@ struct Rmsnorm2dFwdPipelineDefaultPolicy
{ {
if constexpr(Problem::kNeedCrossWarpSync) if constexpr(Problem::kNeedCrossWarpSync)
{ {
using P_ = BlockReduce2dProblem<typename Problem::XDataType, using P_ = BlockReduce2dProblem<typename Problem::ComputeDataType,
typename Problem::ComputeDataType, typename Problem::ComputeDataType,
typename Problem::BlockShape>; typename Problem::BlockShape>;
using block_reduce2d = BlockReduce2d<P_>; using block_reduce2d = BlockReduce2d<P_>;
using x_block_tile = using x_block_tile =
decltype(make_static_distributed_tensor<typename Problem::XDataType>( decltype(make_static_distributed_tensor<typename Problem::ComputeDataType>(
MakeXBlockTileDistribution<Problem>())); MakeXBlockTileDistribution<Problem>()));
using y_block_tile = decltype(block_reduce2d::template MakeYBlockTile<x_block_tile>()); using y_block_tile = decltype(block_reduce2d::template MakeYBlockTile<x_block_tile>());
......
include/ck_tile/ops/rmsnorm2d/pipeline/rmsnorm2d_fwd_pipeline_one_pass.hpp
View file @ 1b616990
// SPDX-License-Identifier: MIT // 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 #pragma once
...@@ -22,12 +22,17 @@ struct Rmsnorm2dFwdPipelineOnePass ...@@ -22,12 +22,17 @@ struct Rmsnorm2dFwdPipelineOnePass
using YDataType = ck_tile::remove_cvref_t<typename Problem::YDataType>; using YDataType = ck_tile::remove_cvref_t<typename Problem::YDataType>;
using InvRmsDataType = ck_tile::remove_cvref_t<typename Problem::InvRmsDataType>; using InvRmsDataType = ck_tile::remove_cvref_t<typename Problem::InvRmsDataType>;
using XResidualDataType = XDataType;
using YResidualDataType = XDataType;
static constexpr bool kHasGamma = !std::is_same_v<GammaDataType, ck_tile::null_type>; static constexpr bool kHasGamma = !std::is_same_v<GammaDataType, ck_tile::null_type>;
static constexpr bool kSaveInvRms = Problem::kSaveInvRms; static constexpr bool kSaveInvRms = Problem::Traits::kSaveInvRms;
static constexpr bool kNeedCrossWarpSync = Problem::kNeedCrossWarpSync; static constexpr bool kNeedCrossWarpSync = Problem::kNeedCrossWarpSync;
static constexpr bool kPadM = false; // TODO - BlockRmsnorm2dFwdProblem::kPadM static constexpr bool kPadM = false; // TODO - BlockRmsnorm2dFwdProblem::kPadM
static constexpr bool kPadN = Problem::kPadN; static constexpr bool kPadN = Problem::Traits::kPadN;
static constexpr auto kFusedAdd = Problem::Traits::kFusedAdd;
static constexpr auto kFusedQuant = Problem::Traits::kFusedQuant;
static constexpr const char* name = []() { static constexpr const char* name = []() {
if constexpr(kNeedCrossWarpSync) if constexpr(kNeedCrossWarpSync)
...@@ -41,19 +46,36 @@ struct Rmsnorm2dFwdPipelineOnePass ...@@ -41,19 +46,36 @@ struct Rmsnorm2dFwdPipelineOnePass
return Policy::template GetSmemSize<Problem>(); return Policy::template GetSmemSize<Problem>();
} }
template <typename XWindow, typename GammaWindow, typename YWindow, typename InvRmsWindow> template <typename XWindow,
typename XResidualWindow,
typename GammaWindow,
typename YWindow,
typename YResidualWindow,
typename InvRmsWindow,
typename SmoothScaleWindow,
typename YScaleWindow,
typename Epilogue>
CK_TILE_DEVICE auto operator()(const XWindow& x_window_, CK_TILE_DEVICE auto operator()(const XWindow& x_window_,
const XResidualWindow& x_residual_window_,
const GammaWindow& gamma_window_, const GammaWindow& gamma_window_,
YWindow& y_window, YWindow& y_window_,
const YResidualWindow& y_residual_window_,
InvRmsWindow& inv_rms_window, InvRmsWindow& inv_rms_window,
const SmoothScaleWindow& sm_scale_window_,
YScaleWindow& y_scale_window_,
ComputeDataType epsilon, ComputeDataType epsilon,
ck_tile::index_t row_size, ck_tile::index_t row_size,
void* smem) const void* smem,
Epilogue) const
{ {
const auto x_window = const auto x_window =
make_tile_window(x_window_, Policy::template MakeXBlockTileDistribution<Problem>()); make_tile_window(x_window_, Policy::template MakeXBlockTileDistribution<Problem>());
const auto gamma_window = make_tile_window( const auto gamma_window = make_tile_window(
gamma_window_, Policy::template MakeGammaBlockTileDistribution<Problem>()); gamma_window_, Policy::template MakeGammaBlockTileDistribution<Problem>());
const auto x_residual_window = make_tile_window(
x_residual_window_, Policy::template MakeXBlockTileDistribution<Problem>());
auto y_residual_window = make_tile_window(
y_residual_window_, Policy::template MakeXBlockTileDistribution<Problem>());
auto reduce_square_sum_func = ReduceOp::SquareAdd{}; auto reduce_square_sum_func = ReduceOp::SquareAdd{};
auto reduce_sum_func = ReduceOp::Add{}; auto reduce_sum_func = ReduceOp::Add{};
...@@ -62,13 +84,31 @@ struct Rmsnorm2dFwdPipelineOnePass ...@@ -62,13 +84,31 @@ struct Rmsnorm2dFwdPipelineOnePass
auto block_reduce2d_cross_warp_sync = auto block_reduce2d_cross_warp_sync =
Policy::template GetBlockReduce2dCrossWarpSync<Problem>(); Policy::template GetBlockReduce2dCrossWarpSync<Problem>();
const auto x = load_tile(x_window); auto x = load_tile(x_window);
auto x_resi = load_tile(x_residual_window);
// load gamma (TODO: support no gamma?) // load gamma (TODO: support no gamma?)
const auto gamma = load_tile(gamma_window); const auto gamma = load_tile(gamma_window);
auto acc = cast_tile<ComputeDataType>(x);
if constexpr(kFusedAdd == Rmsnorm2dFusedAddEnum::PRE_ADD ||
kFusedAdd == Rmsnorm2dFusedAddEnum::PRE_ADD_STORE)
{
sweep_tile(x_resi, [&](auto idx) {
// compute x = x_resi + x
acc(idx) = type_convert<ComputeDataType>(x_resi(idx)) + acc(idx);
});
if constexpr(kFusedAdd == Rmsnorm2dFusedAddEnum::PRE_ADD_STORE)
{
store_tile(y_residual_window, cast_tile<YResidualDataType>(acc));
}
}
// compute mean square each-thread->cross-lane->cross-warp // compute mean square each-thread->cross-lane->cross-warp
auto square_sum = block_reduce2d( auto square_sum = block_reduce2d(acc,
x, reduce_square_sum_func.GetIdentityValue<ComputeDataType>(), reduce_square_sum_func); reduce_square_sum_func.GetIdentityValue<ComputeDataType>(),
reduce_square_sum_func);
block_reduce2d_sync(square_sum, reduce_sum_func); block_reduce2d_sync(square_sum, reduce_sum_func);
block_reduce2d_cross_warp_sync(square_sum, smem, reduce_sum_func); block_reduce2d_cross_warp_sync(square_sum, smem, reduce_sum_func);
...@@ -83,19 +123,30 @@ struct Rmsnorm2dFwdPipelineOnePass ...@@ -83,19 +123,30 @@ struct Rmsnorm2dFwdPipelineOnePass
store_tile(inv_rms_window, cast_tile<InvRmsDataType>(inv_rms)); store_tile(inv_rms_window, cast_tile<InvRmsDataType>(inv_rms));
// rmsnorm computation // rmsnorm computation
auto y = make_static_distributed_tensor<YDataType>(x.get_tile_distribution()); auto rmsn = make_static_distributed_tensor<ComputeDataType>(x.get_tile_distribution());
sweep_tile(y, [&, inv_rms_ = inv_rms](auto idx) { sweep_tile(rmsn, [&, inv_rms_ = inv_rms](auto idx) {
constexpr auto i_idx = make_tuple(idx[number<0>{}]); constexpr auto i_idx = make_tuple(idx[number<0>{}]);
constexpr auto j_idx = make_tuple(idx[number<1>{}]); constexpr auto j_idx = make_tuple(idx[number<1>{}]);
const auto gamma_ = type_convert<ComputeDataType>(gamma[j_idx]); const auto gamma_ = type_convert<ComputeDataType>(gamma[j_idx]);
const auto x_ = type_convert<ComputeDataType>(x[idx]); auto rmsn_ = acc[idx] * inv_rms_[i_idx] * gamma_;
auto y_ = x_ * inv_rms_[i_idx] * gamma_;
y(idx) = type_convert<YDataType>(y_); rmsn(idx) = rmsn_;
}); });
store_tile(y_window, y);
if constexpr(kFusedQuant == Rmsnorm2dFusedQuantEnum::SMOOTH_DYNAMIC_QUANT)
{
Epilogue{}(y_window_, sm_scale_window_, y_scale_window_, rmsn, smem);
}
else if constexpr(kFusedQuant == Rmsnorm2dFusedQuantEnum::DYNAMIC_QUANT)
{
Epilogue{}(y_window_, y_scale_window_, rmsn, smem);
}
else
{
Epilogue{}(y_window_, rmsn);
}
} }
}; };
} // namespace ck_tile } // namespace ck_tile
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment