/* * Copyright (c) 2017, 2018 Arm Limited. * * SPDX-License-Identifier: MIT * * Permission is hereby granted, free of charge, to any person obtaining a copy * of this software and associated documentation files (the "Software"), to * deal in the Software without restriction, including without limitation the * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or * sell copies of the Software, and to permit persons to whom the Software is * furnished to do so, subject to the following conditions: * * The above copyright notice and this permission notice shall be included in all * copies or substantial portions of the Software. * * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE * SOFTWARE. */ layout(local_size_x = LOCAL_SIZE_X, local_size_y = LOCAL_SIZE_Y, local_size_z = LOCAL_SIZE_Z) in; #include "helpers_cs.h" #if defined(DATA_TYPE_FP16) precision mediump float; #endif // DATA_TYPE_FP16 #define SWAP_ROW_func(u0, l0) \ { \ tmp_swap = u0; \ u0 = l0; \ l0 = tmp_swap; \ } #define SWAP_4x4_func(u0, u1, u2, u3, l0, l1, l2, l3) \ { \ vec4 tmp_swap; \ SWAP_ROW_func(u0, l0); \ SWAP_ROW_func(u1, l1); \ SWAP_ROW_func(u2, l2); \ SWAP_ROW_func(u3, l3); \ } #define TRANSPOSE_4x4_func(u0, u1, u2, u3) \ { \ mat4x4 matin, matout; \ matin[0] = u0; \ matin[1] = u1; \ matin[2] = u2; \ matin[3] = u3; \ matout = transpose(matin); \ u0 = matout[0]; \ u1 = matout[1]; \ u2 = matout[2]; \ u3 = matout[3]; \ } /** This OpenGL ES kernel computes the matrix transposition of input matrix * * @note The data type must be passed at compile time using "#define DATA_TYPE_NAME". e.g. "#define DATA_TYPE_FP32" * @note Optimization name must be passed using "#define OPTIMIZATION_NAME" for F16. e.g. "#define TRANSPOSE_8X8" * * @param[in] src_ptr Pointer to the source matrix. Supported data types: F32/F16 * @param[in] src_attrs The attributes of the source matrix * @param[out] dst_ptr Pointer to the destination matrix Supported data type: same as src_ptr * @param[in] dst_attrs The attributes of the destination matrix */ SHADER_PARAMS_DECLARATION { ImageAttributes src_attrs; ImageAttributes dst_attrs; }; #ifdef DATA_TYPE_FP32 TENSOR_DECLARATION(1, srcBuffer, float, src_ptr, src_shift, 2, readonly); TENSOR_DECLARATION(2, dstBuffer, float, dst_ptr, dst_shift, 2, writeonly); void main(void) { // compute source address ImageIterator src_iter = CONVERT_TO_IMAGE_ITERATOR(src_attrs, src_shift); ImageIterator dst_iter = CONVERT_TO_IMAGE_ITERATOR_NO_STEP(dst_attrs, dst_shift); // load the NxN block at (x, y) vec4 u0 = VLOAD4(vec4, src_ptr, IMAGE_OFFSET(src_iter, 0, 0)); vec4 u1 = VLOAD4(vec4, src_ptr, IMAGE_OFFSET(src_iter, 0, 1)); vec4 u2 = VLOAD4(vec4, src_ptr, IMAGE_OFFSET(src_iter, 0, 2)); vec4 u3 = VLOAD4(vec4, src_ptr, IMAGE_OFFSET(src_iter, 0, 3)); // transpose the block TRANSPOSE_4x4_func(u0, u1, u2, u3); // store the block at (y, x) TENSOR_ITERATOR_ADVANCE_IN_BYTES(dst_iter, uint(16) * uint(gl_GlobalInvocationID.y) + uint(4) * uint(gl_GlobalInvocationID.x) * (dst_attrs.stride_y)); VSTORE4(dst_ptr, IMAGE_OFFSET(dst_iter, 0, 0), u0); VSTORE4(dst_ptr, IMAGE_OFFSET(dst_iter, 0, 1), u1); VSTORE4(dst_ptr, IMAGE_OFFSET(dst_iter, 0, 2), u2); VSTORE4(dst_ptr, IMAGE_OFFSET(dst_iter, 0, 3), u3); } #elif defined(DATA_TYPE_FP16) /* DATA_TYPE_FP16 */ #if defined(TRANSPOSE_4X4) TENSOR_DECLARATION(1, srcBuffer, uvec2, src_ptr, src_shift, 3, readonly); TENSOR_DECLARATION(2, dstBuffer, uvec2, dst_ptr, dst_shift, 3, writeonly); void main(void) { // compute source address ImageIterator src_iter = CONVERT_TO_IMAGE_ITERATOR(src_attrs, src_shift); ImageIterator dst_iter = CONVERT_TO_IMAGE_ITERATOR_NO_STEP(dst_attrs, dst_shift); // load the NxN block at (x, y) vec4 u0 = LOAD_UNPACK4_HALF(src_ptr, IMAGE_OFFSET(src_iter, 0, 0)); vec4 u1 = LOAD_UNPACK4_HALF(src_ptr, IMAGE_OFFSET(src_iter, 0, 1)); vec4 u2 = LOAD_UNPACK4_HALF(src_ptr, IMAGE_OFFSET(src_iter, 0, 2)); vec4 u3 = LOAD_UNPACK4_HALF(src_ptr, IMAGE_OFFSET(src_iter, 0, 3)); // transpose the block TRANSPOSE_4x4_func(u0, u1, u2, u3); // store the block at (y, x) TENSOR_ITERATOR_ADVANCE_IN_BYTES(dst_iter, uint(8) * uint(gl_GlobalInvocationID.y) + uint(gl_GlobalInvocationID.x) * (dst_attrs.step_y)); STORE_PACK4_HALF(dst_ptr, IMAGE_OFFSET(dst_iter, 0, 0), u0); STORE_PACK4_HALF(dst_ptr, IMAGE_OFFSET(dst_iter, 0, 1), u1); STORE_PACK4_HALF(dst_ptr, IMAGE_OFFSET(dst_iter, 0, 2), u2); STORE_PACK4_HALF(dst_ptr, IMAGE_OFFSET(dst_iter, 0, 3), u3); } #elif defined(TRANSPOSE_8X8) /* TRANSPOSE_8X8 */ TENSOR_DECLARATION(1, srcBuffer, uvec4, src_ptr, src_shift, 4, readonly); TENSOR_DECLARATION(2, dstBuffer, uvec4, dst_ptr, dst_shift, 4, writeonly); void main(void) { // compute source address ImageIterator src_iter = CONVERT_TO_IMAGE_ITERATOR(src_attrs, src_shift); ImageIterator dst_iter = CONVERT_TO_IMAGE_ITERATOR_NO_STEP(dst_attrs, dst_shift); vec4 u[8][2]; for(int i = 0; i < 8; i++) { u[i] = LOAD_UNPACK8_HALF(src_ptr, IMAGE_OFFSET(src_iter, 0, i)); } // transpose the block TRANSPOSE_4x4_func(u[0][0], u[1][0], u[2][0], u[3][0]); TRANSPOSE_4x4_func(u[0][1], u[1][1], u[2][1], u[3][1]); TRANSPOSE_4x4_func(u[4][0], u[5][0], u[6][0], u[7][0]); TRANSPOSE_4x4_func(u[4][1], u[5][1], u[6][1], u[7][1]); SWAP_4x4_func(u[0][1], u[1][1], u[2][1], u[3][1], u[4][0], u[5][0], u[6][0], u[7][0]); // store the block at (y, x) TENSOR_ITERATOR_ADVANCE_IN_BYTES(dst_iter, uint(16) * uint(gl_GlobalInvocationID.y) + uint(gl_GlobalInvocationID.x) * (dst_attrs.step_y)); for(int i = 0; i < 8; i++) { STORE_PACK8_HALF(dst_ptr, IMAGE_OFFSET(dst_iter, 0, i), u[i]); } } #elif defined(TRANSPOSE_8X8_SQUARE) /* TRANSPOSE_8x8_SQUARE */ TENSOR_DECLARATION(1, srcBuffer, uvec4, src_ptr, src_shift, 4, readonly); TENSOR_DECLARATION(2, dstBuffer, uvec4, dst_ptr, dst_shift, 4, writeonly); void main(void) { ImageIterator src_iter = CONVERT_TO_IMAGE_ITERATOR(src_attrs, src_shift); ImageIterator dst_iter = CONVERT_TO_IMAGE_ITERATOR_NO_STEP(dst_attrs, dst_shift); if(gl_GlobalInvocationID.x <= gl_GlobalInvocationID.y) { uint blk1_offset_in_bytes = CURRENT_ITEM_OFFSET_IN_BYTES(src_iter); TENSOR_ITERATOR_ADVANCE_IN_BYTES(dst_iter, uint(16) * uint(gl_GlobalInvocationID.y) + uint(gl_GlobalInvocationID.x) * (dst_attrs.step_y)); uint blk2_offset_in_bytes = CURRENT_ITEM_OFFSET_IN_BYTES(dst_iter); // load block1 vec4 u1[8][2]; SET_TENSOR_ITERATOR_OFFSET_IN_BYTES(src_iter, blk1_offset_in_bytes); for(int i = 0; i < 8; i++) { u1[i] = LOAD_UNPACK8_HALF(src_ptr, IMAGE_OFFSET(src_iter, 0, i)); } // transpose block1 TRANSPOSE_4x4_func(u1[0][0], u1[1][0], u1[2][0], u1[3][0]); TRANSPOSE_4x4_func(u1[0][1], u1[1][1], u1[2][1], u1[3][1]); TRANSPOSE_4x4_func(u1[4][0], u1[5][0], u1[6][0], u1[7][0]); TRANSPOSE_4x4_func(u1[4][1], u1[5][1], u1[6][1], u1[7][1]); SWAP_4x4_func(u1[0][1], u1[1][1], u1[2][1], u1[3][1], u1[4][0], u1[5][0], u1[6][0], u1[7][0]); // write to block2 SET_TENSOR_ITERATOR_OFFSET_IN_BYTES(dst_iter, blk2_offset_in_bytes); for(int i = 0; i < 8; i++) { STORE_PACK8_HALF(dst_ptr, IMAGE_OFFSET(dst_iter, 0, i), u1[i]); } // load block2 vec4 u2[8][2]; SET_TENSOR_ITERATOR_OFFSET_IN_BYTES(src_iter, blk2_offset_in_bytes); for(int i = 0; i < 8; i++) { u2[i] = LOAD_UNPACK8_HALF(src_ptr, IMAGE_OFFSET(src_iter, 0, i)); } // transpose block2 TRANSPOSE_4x4_func(u2[0][0], u2[1][0], u2[2][0], u2[3][0]); TRANSPOSE_4x4_func(u2[0][1], u2[1][1], u2[2][1], u2[3][1]); TRANSPOSE_4x4_func(u2[4][0], u2[5][0], u2[6][0], u2[7][0]); TRANSPOSE_4x4_func(u2[4][1], u2[5][1], u2[6][1], u2[7][1]); SWAP_4x4_func(u2[0][1], u2[1][1], u2[2][1], u2[3][1], u2[4][0], u2[5][0], u2[6][0], u2[7][0]); // write to block1 SET_TENSOR_ITERATOR_OFFSET_IN_BYTES(dst_iter, blk1_offset_in_bytes); for(int i = 0; i < 8; i++) { STORE_PACK8_HALF(dst_ptr, IMAGE_OFFSET(dst_iter, 0, i), u2[i]); } } } #endif /* TRANSPOSE_4X4 */ #endif /* DATA_TYPE_FP32 */