# Copyright (c) 2017 - 2024 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
# SPDX-License-Identifier: BSD-3-Clause
#
# Redistribution and use in source and binary forms, with or without
# modification, are permitted provided that the following conditions are met:
#
# 1. Redistributions of source code must retain the above copyright notice, this
# list of conditions and the following disclaimer.
#
# 2. Redistributions in binary form must reproduce the above copyright notice,
# this list of conditions and the following disclaimer in the documentation
# and/or other materials provided with the distribution.
#
# 3. Neither the name of the copyright holder nor the names of its
# contributors may be used to endorse or promote products derived from
# this software without specific prior written permission.
#
# THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
# AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
# IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
# DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE
# FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
# DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
# SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
# CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
# OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
# OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
 
list(SORT CUTLASS_NVCC_ARCHS_ENABLED)
set(CUTLASS_NVCC_ARCHS_ENABLED_REVERSED ${CUTLASS_NVCC_ARCHS_ENABLED})
list(REVERSE CUTLASS_NVCC_ARCHS_ENABLED_REVERSED)
list(GET CUTLASS_NVCC_ARCHS_ENABLED_REVERSED 0 CUTLASS_NVCC_MAX_ARCH)

add_custom_target(
  cutlass_test_unit_conv_device
  DEPENDS
  cutlass_test_unit_conv_device_simt
)

 add_custom_target(
  test_unit_conv_device
  DEPENDS
  test_unit_conv_device_simt
)

if (CUTLASS_NVCC_MAX_ARCH GREATER_EQUAL 70)

  add_dependencies(
    cutlass_test_unit_conv_device
    cutlass_test_unit_conv_device_tensorop_f32_sm70
  )

  add_dependencies(
    test_unit_conv_device
    test_unit_conv_device_tensorop_f32_sm70
  )

endif()

if (CUTLASS_NVCC_MAX_ARCH GREATER_EQUAL 75)

  add_dependencies(
    cutlass_test_unit_conv_device
    cutlass_test_unit_conv_device_tensorop_f32_sm75
    cutlass_test_unit_conv_device_tensorop_s32
    cutlass_test_unit_conv_device_tensorop_s32_interleaved
  )

  add_dependencies(
    test_unit_conv_device
    test_unit_conv_device_tensorop_f32_sm75
    test_unit_conv_device_tensorop_s32
    test_unit_conv_device_tensorop_s32_interleaved
  )

endif()

if (CUTLASS_NVCC_MAX_ARCH GREATER_EQUAL 80)

  add_dependencies(
    cutlass_test_unit_conv_device
    cutlass_test_unit_conv_device_tensorop_f16_sm80
    cutlass_test_unit_conv_device_tensorop_f32_sm80
    cutlass_test_unit_conv_device_tensorop_f32_tf32_sm80
  )

  add_dependencies(
    test_unit_conv_device
    test_unit_conv_device_tensorop_f16_sm80
    test_unit_conv_device_tensorop_f32_sm80
    test_unit_conv_device_tensorop_f32_tf32_sm80
  )

endif()

#
# OpClassSimt (CUDA cores)
#

cutlass_test_unit_add_executable(
  cutlass_test_unit_conv_device_simt
  
  # F32  
  conv2d_fprop_implicit_gemm_f32nhwc_f32nhwc_f32nhwc_simt_f32_sm50.cu

  # CF32
  conv2d_fprop_implicit_gemm_cf32nhwc_cf32nhwc_cf32nhwc_simt_f32_sm50.cu
  conv2d_dgrad_implicit_gemm_cf32nhwc_cf32nhwc_cf32nhwc_simt_f32_sm50.cu
  conv2d_wgrad_implicit_gemm_cf32nhwc_cf32nhwc_cf32nhwc_simt_f32_sm50.cu 

  # F16
  conv2d_fprop_implicit_gemm_f16nhwc_f16nhwc_f16nhwc_simt_f16_sm60.cu
  depthwise_conv2d_fprop_implicit_gemm_f16nhwc_f16nhwc_f16nhwc_simt_f16_sm60.cu
  depthwise_conv2d_fprop_direct_conv_f16nhwc_f16nhwc_f16nhwc_simt_f16_sm60.cu
  depthwise_conv2d_fprop_direct_conv_fixed_stride_dilation_f16nhwc_f16nhwc_f16nhwc_simt_f16_sm60.cu
)

if (CUTLASS_NVCC_MAX_ARCH GREATER_EQUAL 80)

  cutlass_target_sources(
    cutlass_test_unit_conv_device_simt
    PRIVATE
    conv2d_fprop_implicit_gemm_f32nhwc_f32nhwc_f32nhwc_simt_f32_sm80.cu
    conv2d_dgrad_implicit_gemm_f32nhwc_f32nhwc_f32nhwc_simt_f32_sm80.cu
    conv2d_wgrad_implicit_gemm_f32nhwc_f32nhwc_f32nhwc_simt_f32_sm80.cu
    conv2d_fprop_implicit_gemm_cf32nhwc_cf32nhwc_cf32nhwc_simt_f32_sm80.cu
    conv2d_dgrad_implicit_gemm_cf32nhwc_cf32nhwc_cf32nhwc_simt_f32_sm80.cu
    conv2d_wgrad_implicit_gemm_cf32nhwc_cf32nhwc_cf32nhwc_simt_f32_sm80.cu
  )

endif()

#
# OpClassTensorOp (Tensor cores)
#

# Conv - F16 input, F32 output, F32 accumulation
cutlass_test_unit_add_executable(
  cutlass_test_unit_conv_device_tensorop_f32_sm70
  conv2d_fprop_with_broadcast_sm70.cu 
  conv2d_fprop_implicit_gemm_f16nhwc_f16nhwc_f32nhwc_tensor_op_f32_sm70.cu
  conv2d_dgrad_implicit_gemm_f16nhwc_f16nhwc_f32nhwc_tensor_op_f32_sm70.cu
  conv2d_wgrad_implicit_gemm_f16nhwc_f16nhwc_f32nhwc_tensor_op_f32_sm70.cu
)

# Conv - F16 input, F32 output, F32 accumulation - SM75
cutlass_test_unit_add_executable(
  cutlass_test_unit_conv_device_tensorop_f32_sm75

  conv2d_fprop_implicit_gemm_f16nhwc_f16nhwc_f32nhwc_tensor_op_f32_sm75.cu
  conv2d_dgrad_implicit_gemm_f16nhwc_f16nhwc_f32nhwc_tensor_op_f32_sm75.cu
  conv2d_wgrad_implicit_gemm_f16nhwc_f16nhwc_f32nhwc_tensor_op_f32_sm75.cu

  conv2d_fprop_with_broadcast_sm75.cu
  conv2d_fprop_with_reduction_sm75.cu

  conv3d_wgrad_implicit_gemm_f16ndhwc_f16ndhwc_f32ndhwc_tensor_op_f32_sm75.cu
)

if (CUTLASS_NVCC_MAX_ARCH GREATER_EQUAL 80)
  
  # Conv - F16 input, F16 output, F16 accumulation 
  cutlass_test_unit_add_executable(
    cutlass_test_unit_conv_device_tensorop_f16_sm80
  
    conv2d_fprop_implicit_gemm_f16nhwc_f16nhwc_f16nhwc_tensor_op_f16_sm80.cu
    conv2d_dgrad_implicit_gemm_f16nhwc_f16nhwc_f16nhwc_tensor_op_f16_sm80.cu 
    conv2d_wgrad_implicit_gemm_f16nhwc_f16nhwc_f16nhwc_tensor_op_f16_sm80.cu 
  )

  # Conv - F16 input, F32 output, F32 accumulation
  cutlass_test_unit_add_executable(
    cutlass_test_unit_conv_device_tensorop_f32_sm80

    # Conv2d
    conv2d_fprop_implicit_gemm_f16nhwc_f16nhwc_f32nhwc_tensor_op_f32_sm80.cu
    conv2d_dgrad_implicit_gemm_f16nhwc_f16nhwc_f32nhwc_tensor_op_f32_sm80.cu
    conv2d_wgrad_implicit_gemm_f16nhwc_f16nhwc_f32nhwc_tensor_op_f32_sm80.cu

    # Conv2d (small channel count specializations)
    conv2d_fprop_fixed_channels_f16nhwc_f16nhwc_f16nhwc_tensor_op_f32_sm80.cu
    conv2d_fprop_few_channels_f16nhwc_f16nhwc_f16nhwc_tensor_op_f32_sm80.cu

    # Conv2d (Strided Dgrad)
    conv2d_strided_dgrad_implicit_gemm_f16nhwc_f16nhwc_f32nhwc_tensor_op_f32_sm80.cu
    conv2d_strided_dgrad_implicit_gemm_tf32nhwc_tf32nhwc_f32nhwc_tensor_op_f32_sm80.cu
    conv2d_strided_dgrad_implicit_gemm_swizzling4_sm80.cu

    # Conv3d
    conv3d_wgrad_implicit_gemm_f16ndhwc_f16ndhwc_f32ndhwc_tensor_op_f32_sm80.cu

    # Group Conv2d
    group_conv2d_fprop_implicit_gemm_f16nhwc_f16nhwc_f16nhwc_tensor_op_f32_sm80.cu
  )

  # Conv - TF32 input, F32 output, F32 accumulation
  cutlass_test_unit_add_executable(
    cutlass_test_unit_conv_device_tensorop_f32_tf32_sm80
  
    conv2d_fprop_implicit_gemm_tf32nhwc_tf32nhwc_f32nhwc_tensor_op_f32_sm80.cu
    conv2d_dgrad_implicit_gemm_tf32nhwc_tf32nhwc_f32nhwc_tensor_op_f32_sm80.cu
    conv2d_wgrad_implicit_gemm_tf32nhwc_tf32nhwc_f32nhwc_tensor_op_f32_sm80.cu
  
    conv3d_fprop_implicit_gemm_tf32ndhwc_tf32ndhwc_f32ndhwc_tensor_op_f32_sm80.cu
    conv3d_dgrad_implicit_gemm_tf32ndhwc_tf32ndhwc_f32ndhwc_tensor_op_f32_sm80.cu
    conv3d_wgrad_implicit_gemm_tf32ndhwc_tf32ndhwc_f32ndhwc_tensor_op_f32_sm80.cu
  )

endif()

if (CUTLASS_NVCC_MAX_ARCH GREATER_EQUAL 75)

  # Conv2d - S8 input, S32 output, S32 accumulation
  cutlass_test_unit_add_executable(
    cutlass_test_unit_conv_device_tensorop_s32
    conv2d_fprop_implicit_gemm_s8nhwc_s8nhwc_s32nhwc_tensor_op_s32_sm75.cu
    conv2d_fprop_implicit_gemm_s4nhwc_s4nhwc_s32nhwc_tensor_op_s32_sm75.cu
  )
  
  # Conv2d - S8 interleaved input, S8 interleaved output, S32 accumulation
  cutlass_test_unit_add_executable(
    cutlass_test_unit_conv_device_tensorop_s32_interleaved  
    conv2d_fprop_implicit_gemm_s8ncxhwx_s8cxrskx_s8ncxhwx_tensor_op_s32_sm75.cu
    conv2d_fprop_implicit_gemm_s4ncxhwx_s4cxrskx_s4ncxhwx_tensor_op_s32_sm75.cu
    )

  if (CUTLASS_NVCC_MAX_ARCH GREATER_EQUAL 80)

    cutlass_target_sources(
      cutlass_test_unit_conv_device_tensorop_s32
      PRIVATE
      conv2d_fprop_implicit_gemm_s8nhwc_s8nhwc_s32nhwc_tensor_op_s32_sm80.cu
      conv2d_fprop_implicit_gemm_s4nhwc_s4nhwc_s32nhwc_tensor_op_s32_sm80.cu
    )
    
    # Conv2d - S8 interleaved input, S8 interleaved output, S32 accumulation
    cutlass_target_sources(
      cutlass_test_unit_conv_device_tensorop_s32_interleaved
      PRIVATE
      conv2d_fprop_implicit_gemm_s8ncxhwx_s8cxrskx_s8ncxhwx_tensor_op_s32_sm80.cu
      conv2d_fprop_implicit_gemm_s4ncxhwx_s4cxrskx_s4ncxhwx_tensor_op_s32_sm80.cu
    )

  endif()

endif()