From 0d0028ca25a47dd51260e2555b336fc9f09d1df1 Mon Sep 17 00:00:00 2001 From: Manuel Bottini Date: Tue, 2 Oct 2018 16:41:52 +0100 Subject: COMPMID-1298: Fuse ReLu activation in CLWinogradOutputTransform Change-Id: I9e6e43a5839d04c2e4b4552c05446efb0a5074cf Reviewed-on: https://review.mlplatform.org/232 Tested-by: Arm Jenkins Reviewed-by: Georgios Pinitas --- .../CL/cl_kernels/winograd_output_transform.cl | 203 +++++++++++++-------- 1 file changed, 130 insertions(+), 73 deletions(-) (limited to 'src/core/CL/cl_kernels/winograd_output_transform.cl') diff --git a/src/core/CL/cl_kernels/winograd_output_transform.cl b/src/core/CL/cl_kernels/winograd_output_transform.cl index f52b027420..e979978fa2 100644 --- a/src/core/CL/cl_kernels/winograd_output_transform.cl +++ b/src/core/CL/cl_kernels/winograd_output_transform.cl @@ -23,7 +23,15 @@ */ #include "helpers.h" +#if defined(FUSED_ACTIVATION) +#include "activation_layer.cl" +#define ACTIVATION_FUNC(x) ACTIVATION_OP(FUSED_ACTIVATION, x) +#else /* defined(FUSED_ACTIVATION) */ +#define ACTIVATION_FUNC(x) (x) +#endif /* defined(FUSED_ACTIVATION) */ + #if defined(NUM_TILES_X) && defined(OUTPUT_TILE_W) && defined(OUTPUT_TILE_H) +#if defined(VEC_SIZE) && VEC_SIZE == 2 /** This OpenCL kernel performs Winograd output transform when the output tile is 2x2/2x1 or 1x2, the filter size 3x3/3x1 or 1x3 and the data layout is NCHW * * @note The number of tiles along the X direction must be passed at compile time using -DNUM_TILES_X: e.g. -DNUM_TILES_X=16 @@ -32,6 +40,10 @@ * @note If this kernel is used to perform Winograd output transform 3x1, -DWINOGRAD_OUTPUT_TRANSFORM_HORIZONTAL has to be passed at compile time * @note If this kernel is used to perform Winograd output transform 1x3, -DWINOGRAD_OUTPUT_TRANSFORM_VERTICAL has to be passed at compile time * @note The data type must be passed at compile time using -DDATA_TYPE e.g. -DDATA_TYPE=float. Supported data types: float/half. + * @note It is possible to select the activation function to apply using -DFUSED_ACTIVATION e.g. -DFUSED_ACTIVATION=relu + * @note A, B variables required by some activation functions are set using -DA_VAL= and -DB_VAL= respectively. + * @note Vector size should be given as a preprocessor argument using -DVEC_SIZE=size. Accepted values are -DVEC_SIZE=2 (for output_tile_size 2x2, 2x1, 1x2) and -DVEC_SIZE=4 (for output_tile_size 4x4, 4x1, 1x4) + * @note Select data type should be given too with -DSELECT_DATA_TYPE e.g -DSELECT_DATA_TYPE=int * * @param[in] src_ptr Pointer to the source tensor. Supported data types: F32/F16 * @param[in] src_stride_x Stride of the source tensor in X dimension (in bytes) @@ -86,6 +98,7 @@ __kernel void winograd_output_transform_2x2_3x3_nchw( float out00 = d00 + d01 + d02; float out01 = d01 - d02 - d03; #else // defined(WINOGRAD_OUTPUT_TRANSFORM_HORIZONTAL) || defined(WINOGRAD_OUTPUT_TRANSFORM_VERTICAL) + DATA_TYPE d10 = *((__global DATA_TYPE *)(src_addr + 4 * src_stride_z)); DATA_TYPE d11 = *((__global DATA_TYPE *)(src_addr + 5 * src_stride_z)); DATA_TYPE d12 = *((__global DATA_TYPE *)(src_addr + 6 * src_stride_z)); @@ -150,10 +163,12 @@ __kernel void winograd_output_transform_2x2_3x3_nchw( // Store the output tile #if defined(WINOGRAD_OUTPUT_TRANSFORM_VERTICAL) - *((__global DATA_TYPE *)(dst_addr + 0 * dst_stride_y)) = (DATA_TYPE)out00; - *((__global DATA_TYPE *)(dst_addr + 1 * dst_stride_y)) = (DATA_TYPE)out01; + const const VEC_DATA_TYPE(DATA_TYPE, 2) + out0_dt = ACTIVATION_FUNC(CONVERT((VEC_DATA_TYPE(float, 2))(out00, out01), VEC_DATA_TYPE(DATA_TYPE, 2))); + *((__global DATA_TYPE *)(dst_addr + 0 * dst_stride_y)) = out0_dt.s0; + *((__global DATA_TYPE *)(dst_addr + 1 * dst_stride_y)) = out0_dt.s1; #else // defined(WINOGRAD_OUTPUT_TRANSFORM_VERTICAL) - vstore2((VEC_DATA_TYPE(DATA_TYPE, 2))(out00, out01), 0, (__global DATA_TYPE *)(dst_addr + 0 * dst_stride_y)); + vstore2(ACTIVATION_FUNC(CONVERT((VEC_DATA_TYPE(float, 2))(out00, out01), VEC_DATA_TYPE(DATA_TYPE, 2))), 0, (__global DATA_TYPE *)(dst_addr + 0 * dst_stride_y)); #endif // defined(WINOGRAD_OUTPUT_TRANSFORM_VERTICAL) #if !defined(WINOGRAD_OUTPUT_TRANSFORM_HORIZONTAL) && !defined(WINOGRAD_OUTPUT_TRANSFORM_VERTICAL) @@ -162,11 +177,12 @@ __kernel void winograd_output_transform_2x2_3x3_nchw( out10 += (DATA_TYPE)b; out11 += (DATA_TYPE)b; #endif // defined(HAS_BIAS) - - vstore2((VEC_DATA_TYPE(DATA_TYPE, 2))((DATA_TYPE)out10, (DATA_TYPE)out11), 0, (__global DATA_TYPE *)(dst_addr + 1 * dst_stride_y)); + vstore2(ACTIVATION_FUNC(CONVERT((VEC_DATA_TYPE(float, 2))(out10, out11), VEC_DATA_TYPE(DATA_TYPE, 2))), 0, (__global DATA_TYPE *)(dst_addr + 1 * dst_stride_y)); #endif // !defined(WINOGRAD_OUTPUT_TRANSFORM_HORIZONTAL) && !defined(WINOGRAD_OUTPUT_TRANSFORM_VERTICAL) } +#endif // defined(VEC_SIZE) && VEC_SIZE == 2 +#if defined(VEC_SIZE) && VEC_SIZE == 4 /** This OpenCL kernel performs Winograd output transform when the output tile is 4x4, the filter size 3x3 and the data layout is NCHW * * @note The number of tiles along the X direction must be passed at compile time using -DNUM_TILES_X: e.g. -DNUM_TILES_X=16 @@ -230,6 +246,7 @@ __kernel void winograd_output_transform_4x4_3x3_nchw( float out02 = d01 + d02 + 4.0f * d03 + 4.0f * d04; float out03 = d01 - d02 + 8.0f * d03 - 8.0f * d04 + d05; #else // defined(WINOGRAD_OUTPUT_TRANSFORM_HORIZONTAL) || defined(WINOGRAD_OUTPUT_TRANSFORM_VERTICAL) + DATA_TYPE d10 = *((__global DATA_TYPE *)(src_addr + 6 * src_stride_z)); DATA_TYPE d11 = *((__global DATA_TYPE *)(src_addr + 7 * src_stride_z)); DATA_TYPE d12 = *((__global DATA_TYPE *)(src_addr + 8 * src_stride_z)); @@ -351,12 +368,14 @@ __kernel void winograd_output_transform_4x4_3x3_nchw( // Store the output tile #if defined(WINOGRAD_OUTPUT_TRANSFORM_VERTICAL) - *((__global DATA_TYPE *)(dst_addr + 0 * dst_stride_y)) = (DATA_TYPE)out00; - *((__global DATA_TYPE *)(dst_addr + 1 * dst_stride_y)) = (DATA_TYPE)out01; - *((__global DATA_TYPE *)(dst_addr + 2 * dst_stride_y)) = (DATA_TYPE)out02; - *((__global DATA_TYPE *)(dst_addr + 3 * dst_stride_y)) = (DATA_TYPE)out03; + VEC_DATA_TYPE(DATA_TYPE, 4) + out0_dt = ACTIVATION_FUNC(CONVERT((VEC_DATA_TYPE(float, 4))(out00, out01, out02, out03), VEC_DATA_TYPE(DATA_TYPE, 4))); + *((__global DATA_TYPE *)(dst_addr + 0 * dst_stride_y)) = out0_dt.s0; + *((__global DATA_TYPE *)(dst_addr + 1 * dst_stride_y)) = out0_dt.s1; + *((__global DATA_TYPE *)(dst_addr + 2 * dst_stride_y)) = out0_dt.s2; + *((__global DATA_TYPE *)(dst_addr + 3 * dst_stride_y)) = out0_dt.s3; #else // defined(WINOGRAD_OUTPUT_TRANSFORM_VERTICAL) - vstore4((VEC_DATA_TYPE(DATA_TYPE, 4))((DATA_TYPE)out00, (DATA_TYPE)out01, (DATA_TYPE)out02, (DATA_TYPE)out03), 0, (__global DATA_TYPE *)(dst_addr + 0 * dst_stride_y)); + vstore4(ACTIVATION_FUNC(CONVERT((VEC_DATA_TYPE(float, 4))(out00, out01, out02, out03), VEC_DATA_TYPE(DATA_TYPE, 4))), 0, (__global DATA_TYPE *)(dst_addr + 0 * dst_stride_y)); #endif // defined(WINOGRAD_OUTPUT_TRANSFORM_VERTICAL) #if !defined(WINOGRAD_OUTPUT_TRANSFORM_HORIZONTAL) && !defined(WINOGRAD_OUTPUT_TRANSFORM_VERTICAL) @@ -377,9 +396,9 @@ __kernel void winograd_output_transform_4x4_3x3_nchw( out32 += (float)b; out33 += (float)b; #endif // defined(HAS_BIAS) - vstore4((VEC_DATA_TYPE(DATA_TYPE, 4))((DATA_TYPE)out10, (DATA_TYPE)out11, (DATA_TYPE)out12, (DATA_TYPE)out13), 0, (__global DATA_TYPE *)(dst_addr + 1 * dst_stride_y)); - vstore4((VEC_DATA_TYPE(DATA_TYPE, 4))((DATA_TYPE)out20, (DATA_TYPE)out21, (DATA_TYPE)out22, (DATA_TYPE)out23), 0, (__global DATA_TYPE *)(dst_addr + 2 * dst_stride_y)); - vstore4((VEC_DATA_TYPE(DATA_TYPE, 4))((DATA_TYPE)out30, (DATA_TYPE)out31, (DATA_TYPE)out32, (DATA_TYPE)out33), 0, (__global DATA_TYPE *)(dst_addr + 3 * dst_stride_y)); + vstore4(ACTIVATION_FUNC(CONVERT((VEC_DATA_TYPE(float, 4))(out10, out11, out12, out13), VEC_DATA_TYPE(DATA_TYPE, 4))), 0, (__global DATA_TYPE *)(dst_addr + 1 * dst_stride_y)); + vstore4(ACTIVATION_FUNC(CONVERT((VEC_DATA_TYPE(float, 4))(out20, out21, out22, out23), VEC_DATA_TYPE(DATA_TYPE, 4))), 0, (__global DATA_TYPE *)(dst_addr + 2 * dst_stride_y)); + vstore4(ACTIVATION_FUNC(CONVERT((VEC_DATA_TYPE(float, 4))(out30, out31, out32, out33), VEC_DATA_TYPE(DATA_TYPE, 4))), 0, (__global DATA_TYPE *)(dst_addr + 3 * dst_stride_y)); #endif // !defined(WINOGRAD_OUTPUT_TRANSFORM_HORIZONTAL) && !defined(WINOGRAD_OUTPUT_TRANSFORM_VERTICAL) } @@ -579,25 +598,29 @@ __kernel void winograd_output_transform_4x4_3x3_nhwc( #if defined(WINOGRAD_OUTPUT_TRANSFORM_VERTICAL) #if defined(SRC_DEPTH) int4 offset = (int4)(dst_offset_first_element_in_bytes + x_out * sizeof(DATA_TYPE) + y_out * dst_stride_y + z_out * dst_stride_z + batch * dst_stride_w); -#else /* defined(SRC_DEPTH) */ +#else /* defined(SRC_DEPTH) */ int4 offset = (int4)(dst_offset_first_element_in_bytes + x_out * sizeof(DATA_TYPE) + y_out * dst_stride_y + z_out * dst_stride_z); -#endif /* defined(SRC_DEPTH) */ +#endif /* defined(SRC_DEPTH) */ offset = min(offset + (int4)(0, 1, 2, 3) * (int4)dst_stride_z, (int4)dst_size); // If address is beyond the last plane, clamp it to dst_size (which points to the last padding). // Store the 1x4 output tile - *((__global DATA_TYPE *)(dst_ptr + offset.s0)) = (DATA_TYPE)out00; - *((__global DATA_TYPE *)(dst_ptr + offset.s1)) = (DATA_TYPE)out01; - *((__global DATA_TYPE *)(dst_ptr + offset.s2)) = (DATA_TYPE)out02; - *((__global DATA_TYPE *)(dst_ptr + offset.s3)) = (DATA_TYPE)out03; + VEC_DATA_TYPE(DATA_TYPE, 4) + out0_dt = ACTIVATION_FUNC(CONVERT((VEC_DATA_TYPE(float, 4))(out00, out01, out02, out03), VEC_DATA_TYPE(DATA_TYPE, 4))); + *((__global DATA_TYPE *)(dst_ptr + offset.s0)) = out0_dt.s0; + *((__global DATA_TYPE *)(dst_ptr + offset.s1)) = out0_dt.s1; + *((__global DATA_TYPE *)(dst_ptr + offset.s2)) = out0_dt.s2; + *((__global DATA_TYPE *)(dst_ptr + offset.s3)) = out0_dt.s3; #elif defined(WINOGRAD_OUTPUT_TRANSFORM_HORIZONTAL) // Store the 4x1 output tile int offset = dst_offset_first_element_in_bytes + x_out * sizeof(DATA_TYPE) + y_out * dst_stride_y + z_out * dst_stride_z; int mult_y = min(dst_size - offset, 1); - *((__global DATA_TYPE *)(dst_ptr + mult_y * 0 * dst_stride_y + offset)) = (DATA_TYPE)out00; - *((__global DATA_TYPE *)(dst_ptr + mult_y * 1 * dst_stride_y + offset)) = (DATA_TYPE)out01; - *((__global DATA_TYPE *)(dst_ptr + mult_y * 2 * dst_stride_y + offset)) = (DATA_TYPE)out02; - *((__global DATA_TYPE *)(dst_ptr + mult_y * 3 * dst_stride_y + offset)) = (DATA_TYPE)out03; + VEC_DATA_TYPE(DATA_TYPE, 4) + out0_dt = ACTIVATION_FUNC(CONVERT((VEC_DATA_TYPE(float, 4))(out00, out01, out02, out03), VEC_DATA_TYPE(DATA_TYPE, 4))); + *((__global DATA_TYPE *)(dst_ptr + mult_y * 0 * dst_stride_y + offset)) = out0_dt.s0; + *((__global DATA_TYPE *)(dst_ptr + mult_y * 1 * dst_stride_y + offset)) = out0_dt.s1; + *((__global DATA_TYPE *)(dst_ptr + mult_y * 2 * dst_stride_y + offset)) = out0_dt.s2; + *((__global DATA_TYPE *)(dst_ptr + mult_y * 3 * dst_stride_y + offset)) = out0_dt.s3; #else // defined(WINOGRAD_OUTPUT_TRANSFORM_HORIZONTAL) // Get output address #if defined(SRC_DEPTH) @@ -609,22 +632,30 @@ __kernel void winograd_output_transform_4x4_3x3_nhwc( int4 mult_y = min((int4)dst_size - offset, (int4)1); // If out of bound, we don't want to increase dst_stride_y, so we set the multiplier to 0. It will be 1 otherwise. // Store the 4x4 output tile - *((__global DATA_TYPE *)(dst_ptr + mult_y.s0 * 0 * dst_stride_y + offset.s0)) = (DATA_TYPE)out00; - *((__global DATA_TYPE *)(dst_ptr + mult_y.s0 * 1 * dst_stride_y + offset.s0)) = (DATA_TYPE)out01; - *((__global DATA_TYPE *)(dst_ptr + mult_y.s0 * 2 * dst_stride_y + offset.s0)) = (DATA_TYPE)out02; - *((__global DATA_TYPE *)(dst_ptr + mult_y.s0 * 3 * dst_stride_y + offset.s0)) = (DATA_TYPE)out03; - *((__global DATA_TYPE *)(dst_ptr + mult_y.s1 * 0 * dst_stride_y + offset.s1)) = (DATA_TYPE)out10; - *((__global DATA_TYPE *)(dst_ptr + mult_y.s1 * 1 * dst_stride_y + offset.s1)) = (DATA_TYPE)out11; - *((__global DATA_TYPE *)(dst_ptr + mult_y.s1 * 2 * dst_stride_y + offset.s1)) = (DATA_TYPE)out12; - *((__global DATA_TYPE *)(dst_ptr + mult_y.s1 * 3 * dst_stride_y + offset.s1)) = (DATA_TYPE)out13; - *((__global DATA_TYPE *)(dst_ptr + mult_y.s2 * 0 * dst_stride_y + offset.s2)) = (DATA_TYPE)out20; - *((__global DATA_TYPE *)(dst_ptr + mult_y.s2 * 1 * dst_stride_y + offset.s2)) = (DATA_TYPE)out21; - *((__global DATA_TYPE *)(dst_ptr + mult_y.s2 * 2 * dst_stride_y + offset.s2)) = (DATA_TYPE)out22; - *((__global DATA_TYPE *)(dst_ptr + mult_y.s2 * 3 * dst_stride_y + offset.s2)) = (DATA_TYPE)out23; - *((__global DATA_TYPE *)(dst_ptr + mult_y.s3 * 0 * dst_stride_y + offset.s3)) = (DATA_TYPE)out30; - *((__global DATA_TYPE *)(dst_ptr + mult_y.s3 * 1 * dst_stride_y + offset.s3)) = (DATA_TYPE)out31; - *((__global DATA_TYPE *)(dst_ptr + mult_y.s3 * 2 * dst_stride_y + offset.s3)) = (DATA_TYPE)out32; - *((__global DATA_TYPE *)(dst_ptr + mult_y.s3 * 3 * dst_stride_y + offset.s3)) = (DATA_TYPE)out33; + VEC_DATA_TYPE(DATA_TYPE, 4) + out0_dt = ACTIVATION_FUNC(CONVERT((VEC_DATA_TYPE(float, 4))(out00, out01, out02, out03), VEC_DATA_TYPE(DATA_TYPE, 4))); + VEC_DATA_TYPE(DATA_TYPE, 4) + out1_dt = ACTIVATION_FUNC(CONVERT((VEC_DATA_TYPE(float, 4))(out10, out11, out12, out13), VEC_DATA_TYPE(DATA_TYPE, 4))); + VEC_DATA_TYPE(DATA_TYPE, 4) + out2_dt = ACTIVATION_FUNC(CONVERT((VEC_DATA_TYPE(float, 4))(out20, out21, out22, out23), VEC_DATA_TYPE(DATA_TYPE, 4))); + VEC_DATA_TYPE(DATA_TYPE, 4) + out3_dt = ACTIVATION_FUNC(CONVERT((VEC_DATA_TYPE(float, 4))(out30, out31, out32, out33), VEC_DATA_TYPE(DATA_TYPE, 4))); + *((__global DATA_TYPE *)(dst_ptr + mult_y.s0 * 0 * dst_stride_y + offset.s0)) = out0_dt.s0; + *((__global DATA_TYPE *)(dst_ptr + mult_y.s0 * 1 * dst_stride_y + offset.s0)) = out0_dt.s1; + *((__global DATA_TYPE *)(dst_ptr + mult_y.s0 * 2 * dst_stride_y + offset.s0)) = out0_dt.s2; + *((__global DATA_TYPE *)(dst_ptr + mult_y.s0 * 3 * dst_stride_y + offset.s0)) = out0_dt.s3; + *((__global DATA_TYPE *)(dst_ptr + mult_y.s1 * 0 * dst_stride_y + offset.s1)) = out1_dt.s0; + *((__global DATA_TYPE *)(dst_ptr + mult_y.s1 * 1 * dst_stride_y + offset.s1)) = out1_dt.s1; + *((__global DATA_TYPE *)(dst_ptr + mult_y.s1 * 2 * dst_stride_y + offset.s1)) = out1_dt.s2; + *((__global DATA_TYPE *)(dst_ptr + mult_y.s1 * 3 * dst_stride_y + offset.s1)) = out1_dt.s3; + *((__global DATA_TYPE *)(dst_ptr + mult_y.s2 * 0 * dst_stride_y + offset.s2)) = out2_dt.s0; + *((__global DATA_TYPE *)(dst_ptr + mult_y.s2 * 1 * dst_stride_y + offset.s2)) = out2_dt.s1; + *((__global DATA_TYPE *)(dst_ptr + mult_y.s2 * 2 * dst_stride_y + offset.s2)) = out2_dt.s2; + *((__global DATA_TYPE *)(dst_ptr + mult_y.s2 * 3 * dst_stride_y + offset.s2)) = out2_dt.s3; + *((__global DATA_TYPE *)(dst_ptr + mult_y.s3 * 0 * dst_stride_y + offset.s3)) = out3_dt.s0; + *((__global DATA_TYPE *)(dst_ptr + mult_y.s3 * 1 * dst_stride_y + offset.s3)) = out3_dt.s1; + *((__global DATA_TYPE *)(dst_ptr + mult_y.s3 * 2 * dst_stride_y + offset.s3)) = out3_dt.s2; + *((__global DATA_TYPE *)(dst_ptr + mult_y.s3 * 3 * dst_stride_y + offset.s3)) = out3_dt.s3; #endif // defined(WINOGRAD_OUTPUT_TRANSFORM_HORIZONTAL) } @@ -690,6 +721,7 @@ __kernel void winograd_output_transform_4x4_5x5_nchw( Tensor4D src = CONVERT_TO_TENSOR4D_STRUCT(src, SRC_DEPTH); const __global uchar *src_addr = tensor4D_offset(&src, 0, 0, 0, 0); #else /* defined(SRC_DEPTH) */ + Tensor3D src = CONVERT_TO_TENSOR3D_STRUCT(src); const __global uchar *src_addr = tensor3D_offset(&src, 0, 0, 0); #endif /* defined(SRC_DEPTH) */ @@ -706,6 +738,7 @@ __kernel void winograd_output_transform_4x4_5x5_nchw( #if defined(SRC_DEPTH) __global uchar *dst_addr = dst_ptr + dst_offset_first_element_in_bytes + x_out * sizeof(DATA_TYPE) + y_out * dst_stride_y + z_out * dst_stride_z + batch * dst_stride_w; #else /* defined(SRC_DEPTH) */ + __global uchar *dst_addr = dst_ptr + dst_offset_first_element_in_bytes + x_out * sizeof(DATA_TYPE) + y_out * dst_stride_y + z_out * dst_stride_z; #endif /* defined(SRC_DEPTH) */ @@ -740,15 +773,18 @@ __kernel void winograd_output_transform_4x4_5x5_nchw( // Store the output tile #if defined(WINOGRAD_OUTPUT_TRANSFORM_VERTICAL) - *((__global DATA_TYPE *)(dst_addr + 0 * dst_stride_y)) = out00; - *((__global DATA_TYPE *)(dst_addr + 1 * dst_stride_y)) = out01; - *((__global DATA_TYPE *)(dst_addr + 2 * dst_stride_y)) = out02; - *((__global DATA_TYPE *)(dst_addr + 3 * dst_stride_y)) = out03; + VEC_DATA_TYPE(DATA_TYPE, 4) + out0_dt = ACTIVATION_FUNC(CONVERT((VEC_DATA_TYPE(float, 4))(out00, out01, out02, out03), VEC_DATA_TYPE(DATA_TYPE, 4))); + *((__global DATA_TYPE *)(dst_addr + 0 * dst_stride_y)) = out0_dt.s0; + *((__global DATA_TYPE *)(dst_addr + 1 * dst_stride_y)) = out0_dt.s1; + *((__global DATA_TYPE *)(dst_addr + 2 * dst_stride_y)) = out0_dt.s2; + *((__global DATA_TYPE *)(dst_addr + 3 * dst_stride_y)) = out0_dt.s3; #else // defined(WINOGRAD_OUTPUT_TRANSFORM_VERTICAL) - vstore4((VEC_DATA_TYPE(DATA_TYPE, 4))(out00, out01, out02, out03), 0, (__global DATA_TYPE *)(dst_addr)); + vstore4(ACTIVATION_FUNC(CONVERT((VEC_DATA_TYPE(float, 4))(out00, out01, out02, out03), VEC_DATA_TYPE(DATA_TYPE, 4))), 0, (__global DATA_TYPE *)(dst_addr)); #endif // defined(WINOGRAD_OUTPUT_TRANSFORM_VERTICAL) #else // defined(WINOGRAD_OUTPUT_TRANSFORM_HORIZONTAL) || defined(WINOGRAD_OUTPUT_TRANSFORM_VERTICAL) + DATA_TYPE d10 = *((__global DATA_TYPE *)(src_addr + 8 * src_stride_z)); DATA_TYPE d11 = *((__global DATA_TYPE *)(src_addr + 9 * src_stride_z)); DATA_TYPE d12 = *((__global DATA_TYPE *)(src_addr + 10 * src_stride_z)); @@ -859,10 +895,10 @@ __kernel void winograd_output_transform_4x4_5x5_nchw( #endif // defined(HAS_BIAS) // Store the output tile - vstore4((VEC_DATA_TYPE(DATA_TYPE, 4))((DATA_TYPE)out_col0.s0, (DATA_TYPE)out_col1.s0, (DATA_TYPE)out_col2.s0, (DATA_TYPE)out_col3.s0), 0, (__global DATA_TYPE *)(dst_addr + 0 * dst_stride_y)); - vstore4((VEC_DATA_TYPE(DATA_TYPE, 4))((DATA_TYPE)out_col0.s1, (DATA_TYPE)out_col1.s1, (DATA_TYPE)out_col2.s1, (DATA_TYPE)out_col3.s1), 0, (__global DATA_TYPE *)(dst_addr + 1 * dst_stride_y)); - vstore4((VEC_DATA_TYPE(DATA_TYPE, 4))((DATA_TYPE)out_col0.s2, (DATA_TYPE)out_col1.s2, (DATA_TYPE)out_col2.s2, (DATA_TYPE)out_col3.s2), 0, (__global DATA_TYPE *)(dst_addr + 2 * dst_stride_y)); - vstore4((VEC_DATA_TYPE(DATA_TYPE, 4))((DATA_TYPE)out_col0.s3, (DATA_TYPE)out_col1.s3, (DATA_TYPE)out_col2.s3, (DATA_TYPE)out_col3.s3), 0, (__global DATA_TYPE *)(dst_addr + 3 * dst_stride_y)); + vstore4(ACTIVATION_FUNC((VEC_DATA_TYPE(DATA_TYPE, 4))(out_col0.s0, out_col1.s0, out_col2.s0, out_col3.s0)), 0, (__global DATA_TYPE *)(dst_addr + 0 * dst_stride_y)); + vstore4(ACTIVATION_FUNC((VEC_DATA_TYPE(DATA_TYPE, 4))(out_col0.s1, out_col1.s1, out_col2.s1, out_col3.s1)), 0, (__global DATA_TYPE *)(dst_addr + 1 * dst_stride_y)); + vstore4(ACTIVATION_FUNC((VEC_DATA_TYPE(DATA_TYPE, 4))(out_col0.s2, out_col1.s2, out_col2.s2, out_col3.s2)), 0, (__global DATA_TYPE *)(dst_addr + 2 * dst_stride_y)); + vstore4(ACTIVATION_FUNC((VEC_DATA_TYPE(DATA_TYPE, 4))(out_col0.s3, out_col1.s3, out_col2.s3, out_col3.s3)), 0, (__global DATA_TYPE *)(dst_addr + 3 * dst_stride_y)); #endif // !defined(WINOGRAD_OUTPUT_TRANSFORM_HORIZONTAL) && !defined(WINOGRAD_OUTPUT_TRANSFORM_VERTICAL) } @@ -960,18 +996,21 @@ __kernel void winograd_output_transform_4x4_5x5_nhwc( #endif /* defined(SRC_DEPTH) */ offset = min(offset + (int4)(0, 1, 2, 3) * (int4)dst_stride_z, (int4)dst_size); // If address is beyond the last plane, clamp it to dst_size (which points to the last padding). - *(__global DATA_TYPE *)(dst_ptr + offset.s0) = (DATA_TYPE)out00; - *(__global DATA_TYPE *)(dst_ptr + offset.s1) = (DATA_TYPE)out01; - *(__global DATA_TYPE *)(dst_ptr + offset.s2) = (DATA_TYPE)out02; - *(__global DATA_TYPE *)(dst_ptr + offset.s3) = (DATA_TYPE)out03; + VEC_DATA_TYPE(DATA_TYPE, 4) + out0_dt = ACTIVATION_FUNC(CONVERT((VEC_DATA_TYPE(float, 4))(out00, out01, out02, out03), VEC_DATA_TYPE(DATA_TYPE, 4))); + *(__global DATA_TYPE *)(dst_ptr + offset.s0) = out0_dt.s0; + *(__global DATA_TYPE *)(dst_ptr + offset.s1) = out0_dt.s1; + *(__global DATA_TYPE *)(dst_ptr + offset.s2) = out0_dt.s2; + *(__global DATA_TYPE *)(dst_ptr + offset.s3) = out0_dt.s3; #else // defined(WINOGRAD_OUTPUT_TRANSFORM_VERTICAL) // Get output address int offset = dst_offset_first_element_in_bytes + x_out * sizeof(DATA_TYPE) + y_out * dst_stride_y + z_out * dst_stride_z; - - *(__global DATA_TYPE *)(dst_ptr + 0 * dst_stride_y + offset) = (DATA_TYPE)out00; - *(__global DATA_TYPE *)(dst_ptr + 1 * dst_stride_y + offset) = (DATA_TYPE)out01; - *(__global DATA_TYPE *)(dst_ptr + 2 * dst_stride_y + offset) = (DATA_TYPE)out02; - *(__global DATA_TYPE *)(dst_ptr + 3 * dst_stride_y + offset) = (DATA_TYPE)out03; + VEC_DATA_TYPE(DATA_TYPE, 4) + out0_dt = ACTIVATION_FUNC(CONVERT((VEC_DATA_TYPE(float, 4))(out00, out01, out02, out03), VEC_DATA_TYPE(DATA_TYPE, 4))); + *(__global DATA_TYPE *)(dst_ptr + 0 * dst_stride_y + offset) = out0_dt.s0; + *(__global DATA_TYPE *)(dst_ptr + 1 * dst_stride_y + offset) = out0_dt.s1; + *(__global DATA_TYPE *)(dst_ptr + 2 * dst_stride_y + offset) = out0_dt.s2; + *(__global DATA_TYPE *)(dst_ptr + 3 * dst_stride_y + offset) = out0_dt.s3; #endif // defined(WINOGRAD_OUTPUT_TRANSFORM_VERTICAL) #else // defined(WINOGRAD_OUTPUT_TRANSFORM_HORIZONTAL) || defined(WINOGRAD_OUTPUT_TRANSFORM_VERTICAL) @@ -1094,26 +1133,37 @@ __kernel void winograd_output_transform_4x4_5x5_nhwc( int4 mult_y = min((int4)dst_size - offset, (int4)1); // If out of bound, we don't want to increase dst_stride_y, so we set the multiplier to 0. It will be 1 otherwise. // Store the output tile - *(__global DATA_TYPE *)(dst_ptr + mult_y.s0 * 0 * (int)dst_stride_y + offset.s0) = (DATA_TYPE)out_col0.s0; - *(__global DATA_TYPE *)(dst_ptr + mult_y.s0 * 1 * (int)dst_stride_y + offset.s0) = (DATA_TYPE)out_col1.s0; - *(__global DATA_TYPE *)(dst_ptr + mult_y.s0 * 2 * (int)dst_stride_y + offset.s0) = (DATA_TYPE)out_col2.s0; - *(__global DATA_TYPE *)(dst_ptr + mult_y.s0 * 3 * (int)dst_stride_y + offset.s0) = (DATA_TYPE)out_col3.s0; - *(__global DATA_TYPE *)(dst_ptr + mult_y.s1 * 0 * (int)dst_stride_y + offset.s1) = (DATA_TYPE)out_col0.s1; - *(__global DATA_TYPE *)(dst_ptr + mult_y.s1 * 1 * (int)dst_stride_y + offset.s1) = (DATA_TYPE)out_col1.s1; - *(__global DATA_TYPE *)(dst_ptr + mult_y.s1 * 2 * (int)dst_stride_y + offset.s1) = (DATA_TYPE)out_col2.s1; - *(__global DATA_TYPE *)(dst_ptr + mult_y.s1 * 3 * (int)dst_stride_y + offset.s1) = (DATA_TYPE)out_col3.s1; - *(__global DATA_TYPE *)(dst_ptr + mult_y.s2 * 0 * (int)dst_stride_y + offset.s2) = (DATA_TYPE)out_col0.s2; - *(__global DATA_TYPE *)(dst_ptr + mult_y.s2 * 1 * (int)dst_stride_y + offset.s2) = (DATA_TYPE)out_col1.s2; - *(__global DATA_TYPE *)(dst_ptr + mult_y.s2 * 2 * (int)dst_stride_y + offset.s2) = (DATA_TYPE)out_col2.s2; - *(__global DATA_TYPE *)(dst_ptr + mult_y.s2 * 3 * (int)dst_stride_y + offset.s2) = (DATA_TYPE)out_col3.s2; - *(__global DATA_TYPE *)(dst_ptr + mult_y.s3 * 0 * (int)dst_stride_y + offset.s3) = (DATA_TYPE)out_col0.s3; - *(__global DATA_TYPE *)(dst_ptr + mult_y.s3 * 1 * (int)dst_stride_y + offset.s3) = (DATA_TYPE)out_col1.s3; - *(__global DATA_TYPE *)(dst_ptr + mult_y.s3 * 2 * (int)dst_stride_y + offset.s3) = (DATA_TYPE)out_col2.s3; - *(__global DATA_TYPE *)(dst_ptr + mult_y.s3 * 3 * (int)dst_stride_y + offset.s3) = (DATA_TYPE)out_col3.s3; + VEC_DATA_TYPE(DATA_TYPE, VEC_SIZE) + out_col0_dt = ACTIVATION_FUNC(CONVERT(out_col0, VEC_DATA_TYPE(DATA_TYPE, VEC_SIZE))); + VEC_DATA_TYPE(DATA_TYPE, VEC_SIZE) + out_col1_dt = ACTIVATION_FUNC(CONVERT(out_col1, VEC_DATA_TYPE(DATA_TYPE, VEC_SIZE))); + VEC_DATA_TYPE(DATA_TYPE, VEC_SIZE) + out_col2_dt = ACTIVATION_FUNC(CONVERT(out_col2, VEC_DATA_TYPE(DATA_TYPE, VEC_SIZE))); + VEC_DATA_TYPE(DATA_TYPE, VEC_SIZE) + out_col3_dt = ACTIVATION_FUNC(CONVERT(out_col3, VEC_DATA_TYPE(DATA_TYPE, VEC_SIZE))); + + *(__global DATA_TYPE *)(dst_ptr + mult_y.s0 * 0 * (int)dst_stride_y + offset.s0) = out_col0_dt.s0; + *(__global DATA_TYPE *)(dst_ptr + mult_y.s0 * 1 * (int)dst_stride_y + offset.s0) = out_col1_dt.s0; + *(__global DATA_TYPE *)(dst_ptr + mult_y.s0 * 2 * (int)dst_stride_y + offset.s0) = out_col2_dt.s0; + *(__global DATA_TYPE *)(dst_ptr + mult_y.s0 * 3 * (int)dst_stride_y + offset.s0) = out_col3_dt.s0; + *(__global DATA_TYPE *)(dst_ptr + mult_y.s1 * 0 * (int)dst_stride_y + offset.s1) = out_col0_dt.s1; + *(__global DATA_TYPE *)(dst_ptr + mult_y.s1 * 1 * (int)dst_stride_y + offset.s1) = out_col1_dt.s1; + *(__global DATA_TYPE *)(dst_ptr + mult_y.s1 * 2 * (int)dst_stride_y + offset.s1) = out_col2_dt.s1; + *(__global DATA_TYPE *)(dst_ptr + mult_y.s1 * 3 * (int)dst_stride_y + offset.s1) = out_col3_dt.s1; + *(__global DATA_TYPE *)(dst_ptr + mult_y.s2 * 0 * (int)dst_stride_y + offset.s2) = out_col0_dt.s2; + *(__global DATA_TYPE *)(dst_ptr + mult_y.s2 * 1 * (int)dst_stride_y + offset.s2) = out_col1_dt.s2; + *(__global DATA_TYPE *)(dst_ptr + mult_y.s2 * 2 * (int)dst_stride_y + offset.s2) = out_col2_dt.s2; + *(__global DATA_TYPE *)(dst_ptr + mult_y.s2 * 3 * (int)dst_stride_y + offset.s2) = out_col3_dt.s2; + *(__global DATA_TYPE *)(dst_ptr + mult_y.s3 * 0 * (int)dst_stride_y + offset.s3) = out_col0_dt.s3; + *(__global DATA_TYPE *)(dst_ptr + mult_y.s3 * 1 * (int)dst_stride_y + offset.s3) = out_col1_dt.s3; + *(__global DATA_TYPE *)(dst_ptr + mult_y.s3 * 2 * (int)dst_stride_y + offset.s3) = out_col2_dt.s3; + *(__global DATA_TYPE *)(dst_ptr + mult_y.s3 * 3 * (int)dst_stride_y + offset.s3) = out_col3_dt.s3; #endif // defined(WINOGRAD_OUTPUT_TRANSFORM_HORIZONTAL) || defined(WINOGRAD_OUTPUT_TRANSFORM_VERTICAL) } +#endif // defined(VEC_SIZE) && VEC_SIZE == 4 #if defined(WINOGRAD_OUTPUT_TRANSFORM_HORIZONTAL) +#if defined(VEC_SIZE) && VEC_SIZE == 2 /** This OpenCL kernel performs Winograd output transform when the output tile is 2x1, the filter size 3x1 and the data layout is NCHW * * @note The number of tiles along the X direction must be passed at compile time using -DNUM_TILES_X: e.g. -DNUM_TILES_X=16 @@ -1181,7 +1231,9 @@ __kernel void winograd_output_transform_2x1_3x1_nchw( #endif // defined(HAS_BIAS) ); } +#endif // defined(VEC_SIZE) && VEC_SIZE == 2 +#if defined(VEC_SIZE) && VEC_SIZE == 4 /** This OpenCL kernel performs Winograd output transform when the output tile is 4x1, the filter size 3x1 and the data layout is NCHW * * @note The number of tiles along the X direction must be passed at compile time using -DNUM_TILES_X: e.g. -DNUM_TILES_X=16 @@ -1449,9 +1501,11 @@ __kernel void winograd_output_transform_4x1_5x1_nhwc( #endif // defined(HAS_BIAS) dst_size); } +#endif // defined(VEC_SIZE) && VEC_SIZE == 4 #endif // defined(WINOGRAD_OUTPUT_TRANSFORM_HORIZONTAL) #if defined(WINOGRAD_OUTPUT_TRANSFORM_VERTICAL) +#if defined(VEC_SIZE) && VEC_SIZE == 2 /** This OpenCL kernel performs Winograd output transform when the output tile is 1x2, the filter size 1x3 and the data layout is NCHW * * @note The number of tiles along the X direction must be passed at compile time using -DNUM_TILES_X: e.g. -DNUM_TILES_X=16 @@ -1519,7 +1573,9 @@ __kernel void winograd_output_transform_1x2_1x3_nchw( #endif // defined(HAS_BIAS) ); } +#endif // defined(VEC_SIZE) && VEC_SIZE == 2 +#if defined(VEC_SIZE) && VEC_SIZE == 4 /** This OpenCL kernel performs Winograd output transform when the output tile is 1x4, the filter size 1x3 and the data layout is NCHW * * @note The number of tiles along the X direction must be passed at compile time using -DNUM_TILES_X: e.g. -DNUM_TILES_X=16 @@ -1787,5 +1843,6 @@ __kernel void winograd_output_transform_1x4_1x5_nhwc( #endif // defined(HAS_BIAS) dst_size); } +#endif // defined(VEC_SIZE) && VEC_SIZE == 4 #endif // defined(WINOGRAD_OUTPUT_TRANSFORM_VERTICAL) #endif // defined(NUM_TILES_X) && defined(OUTPUT_TILE_W) && defined(OUTPUT_TILE_H) -- cgit v1.2.1