Skip to content

GitLab

Commit 542451d8 authored by Po-Yen, Chen's avatar Po-Yen, Chen
Browse files

Remove elementwise op objs from kargs

parent b076a02a
Show whitespace changes
Inline Side-by-side
Showing with 13 additions and 61 deletions
include/ck/tensor_operation/gpu/device/impl/device_gemm_xdl.hpp
View file @ 542451d8
......@@ -245,10 +245,7 @@ struct DeviceGemmXdl : public DeviceGemm<ALayout,
index_t StrideB,
index_t StrideC,
index_t M01,
index_t N01,
AElementwiseOperation a_element_op,
BElementwiseOperation b_element_op,
CElementwiseOperation c_element_op)
index_t N01)
: p_a_grid_{p_a_grid},
p_b_grid_{p_b_grid},
p_c_grid_{p_c_grid},
......@@ -259,9 +256,6 @@ struct DeviceGemmXdl : public DeviceGemm<ALayout,
block_2_ctile_map_{},
M01_{M01},
N01_{N01},
a_element_op_{a_element_op},
b_element_op_{b_element_op},
c_element_op_{c_element_op},
kraw_{K}
{
a_grid_desc_k0_m_k1_ = DeviceGemmXdl::MakeAGridDescriptor_K0_M_K1(M, K, StrideA);
......@@ -293,9 +287,6 @@ struct DeviceGemmXdl : public DeviceGemm<ALayout,
typename GridwiseGemm::DefaultBlock2CTileMap block_2_ctile_map_;
index_t M01_;
index_t N01_;
AElementwiseOperation a_element_op_;
BElementwiseOperation b_element_op_;
CElementwiseOperation c_element_op_;
index_t kraw_;
};
......@@ -347,9 +338,6 @@ struct DeviceGemmXdl : public DeviceGemm<ALayout,
remove_reference_t<DeviceGemmXdl::AGridDesc_K0_M_K1>,
remove_reference_t<DeviceGemmXdl::BGridDesc_K0_N_K1>,
remove_reference_t<typename GridwiseGemm::CGridDesc_M0_N0_M1_N1_M2_M3_M4_N2>,
AElementwiseOperation,
BElementwiseOperation,
CElementwiseOperation,
remove_reference_t<typename GridwiseGemm::DefaultBlock2CTileMap>,
true>;
......@@ -364,9 +352,6 @@ struct DeviceGemmXdl : public DeviceGemm<ALayout,
arg.a_grid_desc_k0_m_k1_,
arg.b_grid_desc_k0_n_k1_,
arg.c_grid_desc_m0_n0_m1_n1_m2_m3_m4_n2_,
arg.a_element_op_,
arg.b_element_op_,
arg.c_element_op_,
arg.block_2_ctile_map_);
}
else
......@@ -378,9 +363,6 @@ struct DeviceGemmXdl : public DeviceGemm<ALayout,
remove_reference_t<DeviceGemmXdl::AGridDesc_K0_M_K1>,
remove_reference_t<DeviceGemmXdl::BGridDesc_K0_N_K1>,
remove_reference_t<typename GridwiseGemm::CGridDesc_M0_N0_M1_N1_M2_M3_M4_N2>,
AElementwiseOperation,
BElementwiseOperation,
CElementwiseOperation,
remove_reference_t<typename GridwiseGemm::DefaultBlock2CTileMap>,
false>;
......@@ -395,9 +377,6 @@ struct DeviceGemmXdl : public DeviceGemm<ALayout,
arg.a_grid_desc_k0_m_k1_,
arg.b_grid_desc_k0_n_k1_,
arg.c_grid_desc_m0_n0_m1_n1_m2_m3_m4_n2_,
arg.a_element_op_,
arg.b_element_op_,
arg.c_element_op_,
arg.block_2_ctile_map_);
}
......@@ -467,24 +446,11 @@ struct DeviceGemmXdl : public DeviceGemm<ALayout,
index_t StrideA,
index_t StrideB,
index_t StrideC,
AElementwiseOperation a_element_op,
BElementwiseOperation b_element_op,
CElementwiseOperation c_element_op)
AElementwiseOperation,
BElementwiseOperation,
CElementwiseOperation)
{
return Argument{p_a,
p_b,
p_c,
M,
N,
K,
StrideA,
StrideB,
StrideC,
1,
1,
a_element_op,
b_element_op,
c_element_op};
return Argument{p_a, p_b, p_c, M, N, K, StrideA, StrideB, StrideC, 1, 1};
}
static auto MakeInvoker() { return Invoker{}; }
......@@ -499,9 +465,9 @@ struct DeviceGemmXdl : public DeviceGemm<ALayout,
index_t StrideA,
index_t StrideB,
index_t StrideC,
AElementwiseOperation a_element_op,
BElementwiseOperation b_element_op,
CElementwiseOperation c_element_op) override
AElementwiseOperation,
BElementwiseOperation,
CElementwiseOperation) override
{
return std::make_unique<Argument>(static_cast<const ADataType*>(p_a),
static_cast<const BDataType*>(p_b),
......@@ -513,10 +479,7 @@ struct DeviceGemmXdl : public DeviceGemm<ALayout,
StrideB,
StrideC,
1,
1,
a_element_op,
b_element_op,
c_element_op);
1);
}
// polymorphic
......
include/ck/tensor_operation/gpu/grid/gridwise_gemm_xdlops_v2r3.hpp
View file @ 542451d8
......@@ -22,9 +22,6 @@ template <typename GridwiseGemm,
typename AGridDesc_K0_M_K1,
typename BGridDesc_K0_N_K1,
typename CGridDesc_M0_N0_M1_N1_M2_M3_M4_N2,
typename AElementwiseOperation,
typename BElementwiseOperation,
typename CElementwiseOperation,
typename Block2CTileMap,
bool HasMainKBlockLoop>
__global__ void
......@@ -38,9 +35,6 @@ __global__ void
const AGridDesc_K0_M_K1 a_grid_desc_k0_m_k1,
const BGridDesc_K0_N_K1 b_grid_desc_k0_n_k1,
const CGridDesc_M0_N0_M1_N1_M2_M3_M4_N2 c_grid_desc_m0_n0_m1_n1_m2_m3_m4_n2,
const AElementwiseOperation a_element_op,
const BElementwiseOperation b_element_op,
const CElementwiseOperation c_element_op,
const Block2CTileMap block_2_ctile_map)
{
#if(!defined(__HIP_DEVICE_COMPILE__) || defined(__gfx908__) || defined(__gfx90a__) || \
......@@ -54,9 +48,6 @@ __global__ void
a_grid_desc_k0_m_k1,
b_grid_desc_k0_n_k1,
c_grid_desc_m0_n0_m1_n1_m2_m3_m4_n2,
a_element_op,
b_element_op,
c_element_op,
block_2_ctile_map);
#else
ignore = p_a_grid;
......@@ -65,9 +56,6 @@ __global__ void
ignore = a_grid_desc_k0_m_k1;
ignore = b_grid_desc_k0_n_k1;
ignore = c_grid_desc_m0_n0_m1_n1_m2_m3_m4_n2;
ignore = a_element_op;
ignore = b_element_op;
ignore = c_element_op;
ignore = block_2_ctile_map;
#endif // end of if (defined(__gfx908__) || defined(__gfx90a__))
}
......@@ -326,9 +314,6 @@ struct GridwiseGemm_k0mk1_k0nk1_mn_xdlops_v2r3
const AGridDesc_K0_M_K1& a_grid_desc_k0_m_k1,
const BGridDesc_K0_N_K1& b_grid_desc_k0_n_k1,
const CGridDesc_M0_N0_M1_N1_M2_M3_M4_N2& c_grid_desc_m0_n0_m1_n1_m2_m3_m4_n2,
const AElementwiseOperation& a_element_op,
const BElementwiseOperation& b_element_op,
const CElementwiseOperation& c_element_op,
const Block2CTileMap& block_2_ctile_map)
{
const auto a_grid_buf = make_dynamic_buffer<AddressSpaceEnum::Global>(
......@@ -338,6 +323,10 @@ struct GridwiseGemm_k0mk1_k0nk1_mn_xdlops_v2r3
auto c_grid_buf = make_dynamic_buffer<AddressSpaceEnum::Global>(
p_c_grid, c_grid_desc_m0_n0_m1_n1_m2_m3_m4_n2.GetElementSpaceSize());
const AElementwiseOperation a_element_op{};
const BElementwiseOperation b_element_op{};
const CElementwiseOperation c_element_op{};
const auto K0 = a_grid_desc_k0_m_k1.GetLength(I0);
// divide block work by [M, N]
......
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