Skip to content

GitLab

Commit 69e771f6 authored by Chao Liu's avatar Chao Liu
Browse files

clean up coordinate step hack

parent 491437a7
Show whitespace changes
Inline Side-by-side
Showing with 2 additions and 92 deletions
device_operation/include/device_conv2d_fwd_xdl_bias_activation_add_nhwc_kyxc_nhwk.hpp
View file @ 69e771f6
...@@ -228,46 +228,6 @@ struct DeviceConv2dFwdXdl_Bias_Activation_Add_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Out ...@@ -228,46 +228,6 @@ struct DeviceConv2dFwdXdl_Bias_Activation_Add_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Out
using C0GridDesc_M_N = remove_cvref_t<decltype(ABCGridDescs{}[I3])>; using C0GridDesc_M_N = remove_cvref_t<decltype(ABCGridDescs{}[I3])>;
using C1GridDesc_M_N = remove_cvref_t<decltype(ABCGridDescs{}[I4])>; using C1GridDesc_M_N = remove_cvref_t<decltype(ABCGridDescs{}[I4])>;
// TODO remove these hacks
static constexpr auto a_k0_m_k1_grid_step_hacks = make_tuple(
make_tuple(Sequence<0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0>{}, // 0+: K0
Sequence<0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0>{}, // 1+: M
Sequence<0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0>{}), // 2+: K1
make_tuple(Sequence<0, 0, 0, 0, 0, 0, 0, 0, 0, 2, 0, 0, 0, 0, 0, 0>{}, // 0-: K0
Sequence<0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 2, 0, 0, 0, 0, 0>{}, // 1-: M
Sequence<0, 0, 0, 0, 0, 0, 0, 0, 0, 2, 0, 0, 0, 0, 0, 0>{})); // 2-: K1
static constexpr auto b_k0_n_k1_grid_step_hacks =
make_tuple(make_tuple(Sequence<0, 0, 0, 0, 0>{}, // 0+: K0
Sequence<0, 0, 0, 0, 0>{}, // 1+: N
Sequence<0, 0, 0, 0, 0>{}), // 2+: K1
make_tuple(Sequence<0, 0, 0, 0, 0>{}, // 0-: K0
Sequence<0, 0, 0, 0, 0>{}, // 1-: N
Sequence<0, 0, 0, 0, 0>{})); // 2-: K1
static constexpr auto c_m0_n0_m1_n1_m2_m3_m4_n2_grid_step_hacks =
make_tuple(make_tuple(Sequence<0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0>{}, // 0+: M0
Sequence<0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0>{}, // 1+: N0
Sequence<0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0>{}, // 2+: M1
Sequence<0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0>{}, // 3+: N1
Sequence<0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0>{}, // 4+: M2
Sequence<0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0>{}, // 5+: M3
Sequence<0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0>{}, // 6+: M4
Sequence<0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0>{}), // 7+: N2
make_tuple(Sequence<0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0>{}, // 0-: M0
Sequence<0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0>{}, // 1-: N0
Sequence<0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0>{}, // 2-: M1
Sequence<0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0>{}, // 3-: N1
Sequence<0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0>{}, // 4-: M2
Sequence<0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0>{}, // 5-: M3
Sequence<0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0>{}, // 6-: M4
Sequence<0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0>{})); // 7-: N2
static constexpr auto a_k0_m_k1_grid_move_slice_window_step_hacks =
Sequence<0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 2, 0, 0, 0, 0, 0>{};
static constexpr auto b_k0_n_k1_grid_move_slice_window_step_hacks = Sequence<0, 0, 0, 0, 0>{};
// GridwiseGemm // GridwiseGemm
using GridwiseGemm = GridwiseGemm_k0mk1_k0nk1_mn_xdlops_v2r5< using GridwiseGemm = GridwiseGemm_k0mk1_k0nk1_mn_xdlops_v2r5<
BlockSize, BlockSize,
...@@ -310,11 +270,6 @@ struct DeviceConv2dFwdXdl_Bias_Activation_Add_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Out ...@@ -310,11 +270,6 @@ struct DeviceConv2dFwdXdl_Bias_Activation_Add_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Out
Sequence<2, 3, 0, 1, 7, 5, 4, 6>, // CThreadTransferSrcDstAccessOrder, Sequence<2, 3, 0, 1, 7, 5, 4, 6>, // CThreadTransferSrcDstAccessOrder,
7, // CThreadTransferSrcDstVectorDim, 7, // CThreadTransferSrcDstVectorDim,
CThreadTransferDstScalarPerVector, CThreadTransferDstScalarPerVector,
decltype(a_k0_m_k1_grid_step_hacks), // AGridStepHacks,
decltype(b_k0_n_k1_grid_step_hacks), // BGridStepHacks,
decltype(c_m0_n0_m1_n1_m2_m3_m4_n2_grid_step_hacks), // CGridStepHacks,
decltype(a_k0_m_k1_grid_move_slice_window_step_hacks), // AGridMoveSliceWindowStepHacks,
decltype(b_k0_n_k1_grid_move_slice_window_step_hacks), // BGridMoveSliceWindowStepHacks,
false, // CAccessOrderMRepeatNRepeat, false, // CAccessOrderMRepeatNRepeat,
ABlockLdsAddExtraM, ABlockLdsAddExtraM,
BBlockLdsAddExtraN>; BBlockLdsAddExtraN>;
......
device_operation/include/device_conv2d_fwd_xdl_bias_activation_nhwc_kyxc_nhwk.hpp
View file @ 69e771f6
...@@ -223,46 +223,6 @@ struct DeviceConv2dFwdXdl_Bias_Activation_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_ ...@@ -223,46 +223,6 @@ struct DeviceConv2dFwdXdl_Bias_Activation_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_
using CGridDesc_M_N = remove_cvref_t<decltype(ABCGridDescs{}[I2])>; using CGridDesc_M_N = remove_cvref_t<decltype(ABCGridDescs{}[I2])>;
using C0GridDesc_M_N = remove_cvref_t<decltype(ABCGridDescs{}[I3])>; using C0GridDesc_M_N = remove_cvref_t<decltype(ABCGridDescs{}[I3])>;
// TODO remove these hacks
static constexpr auto a_k0_m_k1_grid_step_hacks = make_tuple(
make_tuple(Sequence<0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0>{}, // 0+: K0
Sequence<0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0>{}, // 1+: M
Sequence<0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0>{}), // 2+: K1
make_tuple(Sequence<0, 0, 0, 0, 0, 0, 0, 0, 0, 2, 0, 0, 0, 0, 0, 0>{}, // 0-: K0
Sequence<0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 2, 0, 0, 0, 0, 0>{}, // 1-: M
Sequence<0, 0, 0, 0, 0, 0, 0, 0, 0, 2, 0, 0, 0, 0, 0, 0>{})); // 2-: K1
static constexpr auto b_k0_n_k1_grid_step_hacks =
make_tuple(make_tuple(Sequence<0, 0, 0, 0, 0>{}, // 0+: K0
Sequence<0, 0, 0, 0, 0>{}, // 1+: N
Sequence<0, 0, 0, 0, 0>{}), // 2+: K1
make_tuple(Sequence<0, 0, 0, 0, 0>{}, // 0-: K0
Sequence<0, 0, 0, 0, 0>{}, // 1-: N
Sequence<0, 0, 0, 0, 0>{})); // 2-: K1
static constexpr auto c_m0_n0_m1_n1_m2_m3_m4_n2_grid_step_hacks =
make_tuple(make_tuple(Sequence<0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0>{}, // 0+: M0
Sequence<0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0>{}, // 1+: N0
Sequence<0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0>{}, // 2+: M1
Sequence<0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0>{}, // 3+: N1
Sequence<0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0>{}, // 4+: M2
Sequence<0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0>{}, // 5+: M3
Sequence<0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0>{}, // 6+: M4
Sequence<0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0>{}), // 7+: N2
make_tuple(Sequence<0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0>{}, // 0-: M0
Sequence<0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0>{}, // 1-: N0
Sequence<0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0>{}, // 2-: M1
Sequence<0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0>{}, // 3-: N1
Sequence<0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0>{}, // 4-: M2
Sequence<0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0>{}, // 5-: M3
Sequence<0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0>{}, // 6-: M4
Sequence<0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0>{})); // 7-: N2
static constexpr auto a_k0_m_k1_grid_move_slice_window_step_hacks =
Sequence<0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 2, 0, 0, 0, 0, 0>{};
static constexpr auto b_k0_n_k1_grid_move_slice_window_step_hacks = Sequence<0, 0, 0, 0, 0>{};
// GridwiseGemm // GridwiseGemm
using GridwiseGemm = GridwiseGemm_k0mk1_k0nk1_mn_xdlops_v2r6< using GridwiseGemm = GridwiseGemm_k0mk1_k0nk1_mn_xdlops_v2r6<
BlockSize, BlockSize,
...@@ -304,11 +264,6 @@ struct DeviceConv2dFwdXdl_Bias_Activation_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_ ...@@ -304,11 +264,6 @@ struct DeviceConv2dFwdXdl_Bias_Activation_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_
Sequence<2, 3, 0, 1, 7, 5, 4, 6>, // CThreadTransferSrcDstAccessOrder, Sequence<2, 3, 0, 1, 7, 5, 4, 6>, // CThreadTransferSrcDstAccessOrder,
7, // CThreadTransferSrcDstVectorDim, 7, // CThreadTransferSrcDstVectorDim,
CThreadTransferDstScalarPerVector, CThreadTransferDstScalarPerVector,
decltype(a_k0_m_k1_grid_step_hacks), // AGridStepHacks,
decltype(b_k0_n_k1_grid_step_hacks), // BGridStepHacks,
decltype(c_m0_n0_m1_n1_m2_m3_m4_n2_grid_step_hacks), // CGridStepHacks,
decltype(a_k0_m_k1_grid_move_slice_window_step_hacks), // AGridMoveSliceWindowStepHacks,
decltype(b_k0_n_k1_grid_move_slice_window_step_hacks), // BGridMoveSliceWindowStepHacks,
false, // CAccessOrderMRepeatNRepeat, false, // CAccessOrderMRepeatNRepeat,
ABlockLdsAddExtraM, ABlockLdsAddExtraM,
BBlockLdsAddExtraN>; BBlockLdsAddExtraN>;
......
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