| // Copyright 2021 Google LLC |
| // |
| // This source code is licensed under the BSD-style license found in the |
| // LICENSE file in the root directory of this source tree. |
| // |
| // Auto-generated file. Do not edit! |
| // Specification: test/x64-transpose.yaml |
| // Generator: tools/generate-transpose-test.py |
| |
| |
| #include <gtest/gtest.h> |
| |
| #include <xnnpack/common.h> |
| #include <xnnpack/isa-checks.h> |
| |
| #include <xnnpack/transpose.h> |
| #include "transpose-microkernel-tester.h" |
| |
| |
| TEST(X64_TRANSPOSE__1X2_SCALAR_INT, bh_1_bw_2) { |
| TransposeMicrokernelTester() |
| .input_stride(2) |
| .output_stride(1) |
| .block_width(2) |
| .block_height(1) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__1x2_scalar_int); |
| } |
| |
| TEST(X64_TRANSPOSE__1X2_SCALAR_INT, bh_1_2_bw_1_4) { |
| for(size_t i = 1; i <= 2; ++i){ |
| for(size_t j = 1; j <= 4; ++j){ |
| TransposeMicrokernelTester() |
| .input_stride(j) |
| .output_stride(i) |
| .block_width(j) |
| .block_height(i) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__1x2_scalar_int); |
| } |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__1X2_SCALAR_INT, bh_1_bw_4) { |
| TransposeMicrokernelTester() |
| .input_stride(4) |
| .output_stride(1) |
| .block_width(4) |
| .block_height(1) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__1x2_scalar_int); |
| } |
| |
| TEST(X64_TRANSPOSE__1X2_SCALAR_INT, bh_1_bw_3_4) { |
| for(size_t i = 3; i < 4; ++i){ |
| TransposeMicrokernelTester() |
| .input_stride(i) |
| .output_stride(1) |
| .block_width(i) |
| .block_height(1) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__1x2_scalar_int); |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__1X2_SCALAR_INT, bh_2_bw_3_4) { |
| for(size_t i = 3; i < 4; ++i){ |
| TransposeMicrokernelTester() |
| .input_stride(i) |
| .output_stride(2) |
| .block_width(i) |
| .block_height(2) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__1x2_scalar_int); |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__1X2_SCALAR_INT, bh_2_bw_2) { |
| TransposeMicrokernelTester() |
| .input_stride(2) |
| .output_stride(2) |
| .block_width(2) |
| .block_height(2) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__1x2_scalar_int); |
| } |
| |
| TEST(X64_TRANSPOSE__1X2_SCALAR_INT, bh_2_2_bw_2){ |
| for(size_t i = 2; i < 2; ++i){ |
| TransposeMicrokernelTester() |
| .input_stride(2) |
| .output_stride(i) |
| .block_width(2) |
| .block_height(i) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__1x2_scalar_int); |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__1X2_SCALAR_INT, bh_2_2_bw_4){ |
| for(size_t i = 2; i < 2; ++i){ |
| TransposeMicrokernelTester() |
| .input_stride(4) |
| .output_stride(i) |
| .block_width(4) |
| .block_height(i) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__1x2_scalar_int); |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__1X2_SCALAR_INT, bh_2_2_bw_3_4) { |
| for(size_t i = 2; i < 2; ++i){ |
| for(size_t j = 3; j < 4; ++j){ |
| TransposeMicrokernelTester() |
| .input_stride(j) |
| .output_stride(i) |
| .block_width(j) |
| .block_height(i) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__1x2_scalar_int); |
| } |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__1X2_SCALAR_INT, bh_1_bw_2_is_4) { |
| TransposeMicrokernelTester() |
| .input_stride(4) |
| .output_stride(1) |
| .block_width(2) |
| .block_height(1) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__1x2_scalar_int); |
| } |
| |
| TEST(X64_TRANSPOSE__1X2_SCALAR_INT, bh_1_bw_2_os_2) { |
| TransposeMicrokernelTester() |
| .input_stride(2) |
| .output_stride(2) |
| .block_width(2) |
| .block_height(1) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__1x2_scalar_int); |
| } |
| |
| TEST(X64_TRANSPOSE__1X2_SCALAR_INT, bh_1_bw_2_is_4_os_2) { |
| TransposeMicrokernelTester() |
| .input_stride(4) |
| .output_stride(2) |
| .block_width(2) |
| .block_height(1) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__1x2_scalar_int); |
| } |
| |
| TEST(X64_TRANSPOSE__2X1_SCALAR_INT, bh_2_bw_1) { |
| TransposeMicrokernelTester() |
| .input_stride(1) |
| .output_stride(2) |
| .block_width(1) |
| .block_height(2) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x1_scalar_int); |
| } |
| |
| TEST(X64_TRANSPOSE__2X1_SCALAR_INT, bh_1_4_bw_1_2) { |
| for(size_t i = 1; i <= 4; ++i){ |
| for(size_t j = 1; j <= 2; ++j){ |
| TransposeMicrokernelTester() |
| .input_stride(j) |
| .output_stride(i) |
| .block_width(j) |
| .block_height(i) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x1_scalar_int); |
| } |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__2X1_SCALAR_INT, bh_2_bw_2) { |
| TransposeMicrokernelTester() |
| .input_stride(2) |
| .output_stride(2) |
| .block_width(2) |
| .block_height(2) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x1_scalar_int); |
| } |
| |
| TEST(X64_TRANSPOSE__2X1_SCALAR_INT, bh_2_bw_2_2) { |
| for(size_t i = 2; i < 2; ++i){ |
| TransposeMicrokernelTester() |
| .input_stride(i) |
| .output_stride(2) |
| .block_width(i) |
| .block_height(2) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x1_scalar_int); |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__2X1_SCALAR_INT, bh_4_bw_2_2) { |
| for(size_t i = 2; i < 2; ++i){ |
| TransposeMicrokernelTester() |
| .input_stride(i) |
| .output_stride(4) |
| .block_width(i) |
| .block_height(4) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x1_scalar_int); |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__2X1_SCALAR_INT, bh_4_bw_1) { |
| TransposeMicrokernelTester() |
| .input_stride(1) |
| .output_stride(4) |
| .block_width(1) |
| .block_height(4) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x1_scalar_int); |
| } |
| |
| TEST(X64_TRANSPOSE__2X1_SCALAR_INT, bh_3_4_bw_1){ |
| for(size_t i = 3; i < 4; ++i){ |
| TransposeMicrokernelTester() |
| .input_stride(1) |
| .output_stride(i) |
| .block_width(1) |
| .block_height(i) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x1_scalar_int); |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__2X1_SCALAR_INT, bh_3_4_bw_2){ |
| for(size_t i = 3; i < 4; ++i){ |
| TransposeMicrokernelTester() |
| .input_stride(2) |
| .output_stride(i) |
| .block_width(2) |
| .block_height(i) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x1_scalar_int); |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__2X1_SCALAR_INT, bh_3_4_bw_2_2) { |
| for(size_t i = 3; i < 4; ++i){ |
| for(size_t j = 2; j < 2; ++j){ |
| TransposeMicrokernelTester() |
| .input_stride(j) |
| .output_stride(i) |
| .block_width(j) |
| .block_height(i) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x1_scalar_int); |
| } |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__2X1_SCALAR_INT, bh_2_bw_1_is_2) { |
| TransposeMicrokernelTester() |
| .input_stride(2) |
| .output_stride(2) |
| .block_width(1) |
| .block_height(2) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x1_scalar_int); |
| } |
| |
| TEST(X64_TRANSPOSE__2X1_SCALAR_INT, bh_2_bw_1_os_4) { |
| TransposeMicrokernelTester() |
| .input_stride(1) |
| .output_stride(4) |
| .block_width(1) |
| .block_height(2) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x1_scalar_int); |
| } |
| |
| TEST(X64_TRANSPOSE__2X1_SCALAR_INT, bh_2_bw_1_is_2_os_4) { |
| TransposeMicrokernelTester() |
| .input_stride(2) |
| .output_stride(4) |
| .block_width(1) |
| .block_height(2) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x1_scalar_int); |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_SCALAR_INT, bh_2_bw_2) { |
| TransposeMicrokernelTester() |
| .input_stride(2) |
| .output_stride(2) |
| .block_width(2) |
| .block_height(2) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_scalar_int); |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_SCALAR_INT, bh_1_4_bw_1_4) { |
| for(size_t i = 1; i <= 4; ++i){ |
| for(size_t j = 1; j <= 4; ++j){ |
| TransposeMicrokernelTester() |
| .input_stride(j) |
| .output_stride(i) |
| .block_width(j) |
| .block_height(i) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_scalar_int); |
| } |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_SCALAR_INT, bh_2_bw_4) { |
| TransposeMicrokernelTester() |
| .input_stride(4) |
| .output_stride(2) |
| .block_width(4) |
| .block_height(2) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_scalar_int); |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_SCALAR_INT, bh_2_bw_3_4) { |
| for(size_t i = 3; i < 4; ++i){ |
| TransposeMicrokernelTester() |
| .input_stride(i) |
| .output_stride(2) |
| .block_width(i) |
| .block_height(2) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_scalar_int); |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_SCALAR_INT, bh_4_bw_3_4) { |
| for(size_t i = 3; i < 4; ++i){ |
| TransposeMicrokernelTester() |
| .input_stride(i) |
| .output_stride(4) |
| .block_width(i) |
| .block_height(4) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_scalar_int); |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_SCALAR_INT, bh_4_bw_2) { |
| TransposeMicrokernelTester() |
| .input_stride(2) |
| .output_stride(4) |
| .block_width(2) |
| .block_height(4) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_scalar_int); |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_SCALAR_INT, bh_3_4_bw_2){ |
| for(size_t i = 3; i < 4; ++i){ |
| TransposeMicrokernelTester() |
| .input_stride(2) |
| .output_stride(i) |
| .block_width(2) |
| .block_height(i) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_scalar_int); |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_SCALAR_INT, bh_3_4_bw_4){ |
| for(size_t i = 3; i < 4; ++i){ |
| TransposeMicrokernelTester() |
| .input_stride(4) |
| .output_stride(i) |
| .block_width(4) |
| .block_height(i) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_scalar_int); |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_SCALAR_INT, bh_3_4_bw_3_4) { |
| for(size_t i = 3; i < 4; ++i){ |
| for(size_t j = 3; j < 4; ++j){ |
| TransposeMicrokernelTester() |
| .input_stride(j) |
| .output_stride(i) |
| .block_width(j) |
| .block_height(i) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_scalar_int); |
| } |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_SCALAR_INT, bh_2_bw_2_is_4) { |
| TransposeMicrokernelTester() |
| .input_stride(4) |
| .output_stride(2) |
| .block_width(2) |
| .block_height(2) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_scalar_int); |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_SCALAR_INT, bh_2_bw_2_os_4) { |
| TransposeMicrokernelTester() |
| .input_stride(2) |
| .output_stride(4) |
| .block_width(2) |
| .block_height(2) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_scalar_int); |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_SCALAR_INT, bh_2_bw_2_is_4_os_4) { |
| TransposeMicrokernelTester() |
| .input_stride(4) |
| .output_stride(4) |
| .block_width(2) |
| .block_height(2) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_scalar_int); |
| } |
| |
| TEST(X64_TRANSPOSE__4X1_SCALAR_INT, bh_4_bw_1) { |
| TransposeMicrokernelTester() |
| .input_stride(1) |
| .output_stride(4) |
| .block_width(1) |
| .block_height(4) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__4x1_scalar_int); |
| } |
| |
| TEST(X64_TRANSPOSE__4X1_SCALAR_INT, bh_1_8_bw_1_2) { |
| for(size_t i = 1; i <= 8; ++i){ |
| for(size_t j = 1; j <= 2; ++j){ |
| TransposeMicrokernelTester() |
| .input_stride(j) |
| .output_stride(i) |
| .block_width(j) |
| .block_height(i) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__4x1_scalar_int); |
| } |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__4X1_SCALAR_INT, bh_4_bw_2) { |
| TransposeMicrokernelTester() |
| .input_stride(2) |
| .output_stride(4) |
| .block_width(2) |
| .block_height(4) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__4x1_scalar_int); |
| } |
| |
| TEST(X64_TRANSPOSE__4X1_SCALAR_INT, bh_4_bw_2_2) { |
| for(size_t i = 2; i < 2; ++i){ |
| TransposeMicrokernelTester() |
| .input_stride(i) |
| .output_stride(4) |
| .block_width(i) |
| .block_height(4) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__4x1_scalar_int); |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__4X1_SCALAR_INT, bh_8_bw_2_2) { |
| for(size_t i = 2; i < 2; ++i){ |
| TransposeMicrokernelTester() |
| .input_stride(i) |
| .output_stride(8) |
| .block_width(i) |
| .block_height(8) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__4x1_scalar_int); |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__4X1_SCALAR_INT, bh_8_bw_1) { |
| TransposeMicrokernelTester() |
| .input_stride(1) |
| .output_stride(8) |
| .block_width(1) |
| .block_height(8) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__4x1_scalar_int); |
| } |
| |
| TEST(X64_TRANSPOSE__4X1_SCALAR_INT, bh_5_8_bw_1){ |
| for(size_t i = 5; i < 8; ++i){ |
| TransposeMicrokernelTester() |
| .input_stride(1) |
| .output_stride(i) |
| .block_width(1) |
| .block_height(i) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__4x1_scalar_int); |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__4X1_SCALAR_INT, bh_5_8_bw_2){ |
| for(size_t i = 5; i < 8; ++i){ |
| TransposeMicrokernelTester() |
| .input_stride(2) |
| .output_stride(i) |
| .block_width(2) |
| .block_height(i) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__4x1_scalar_int); |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__4X1_SCALAR_INT, bh_5_8_bw_2_2) { |
| for(size_t i = 5; i < 8; ++i){ |
| for(size_t j = 2; j < 2; ++j){ |
| TransposeMicrokernelTester() |
| .input_stride(j) |
| .output_stride(i) |
| .block_width(j) |
| .block_height(i) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__4x1_scalar_int); |
| } |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__4X1_SCALAR_INT, bh_4_bw_1_is_2) { |
| TransposeMicrokernelTester() |
| .input_stride(2) |
| .output_stride(4) |
| .block_width(1) |
| .block_height(4) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__4x1_scalar_int); |
| } |
| |
| TEST(X64_TRANSPOSE__4X1_SCALAR_INT, bh_4_bw_1_os_8) { |
| TransposeMicrokernelTester() |
| .input_stride(1) |
| .output_stride(8) |
| .block_width(1) |
| .block_height(4) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__4x1_scalar_int); |
| } |
| |
| TEST(X64_TRANSPOSE__4X1_SCALAR_INT, bh_4_bw_1_is_2_os_8) { |
| TransposeMicrokernelTester() |
| .input_stride(2) |
| .output_stride(8) |
| .block_width(1) |
| .block_height(4) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__4x1_scalar_int); |
| } |
| |
| TEST(X64_TRANSPOSE__4X2_SCALAR_INT, bh_4_bw_2) { |
| TransposeMicrokernelTester() |
| .input_stride(2) |
| .output_stride(4) |
| .block_width(2) |
| .block_height(4) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__4x2_scalar_int); |
| } |
| |
| TEST(X64_TRANSPOSE__4X2_SCALAR_INT, bh_1_8_bw_1_4) { |
| for(size_t i = 1; i <= 8; ++i){ |
| for(size_t j = 1; j <= 4; ++j){ |
| TransposeMicrokernelTester() |
| .input_stride(j) |
| .output_stride(i) |
| .block_width(j) |
| .block_height(i) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__4x2_scalar_int); |
| } |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__4X2_SCALAR_INT, bh_4_bw_4) { |
| TransposeMicrokernelTester() |
| .input_stride(4) |
| .output_stride(4) |
| .block_width(4) |
| .block_height(4) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__4x2_scalar_int); |
| } |
| |
| TEST(X64_TRANSPOSE__4X2_SCALAR_INT, bh_4_bw_3_4) { |
| for(size_t i = 3; i < 4; ++i){ |
| TransposeMicrokernelTester() |
| .input_stride(i) |
| .output_stride(4) |
| .block_width(i) |
| .block_height(4) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__4x2_scalar_int); |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__4X2_SCALAR_INT, bh_8_bw_3_4) { |
| for(size_t i = 3; i < 4; ++i){ |
| TransposeMicrokernelTester() |
| .input_stride(i) |
| .output_stride(8) |
| .block_width(i) |
| .block_height(8) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__4x2_scalar_int); |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__4X2_SCALAR_INT, bh_8_bw_2) { |
| TransposeMicrokernelTester() |
| .input_stride(2) |
| .output_stride(8) |
| .block_width(2) |
| .block_height(8) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__4x2_scalar_int); |
| } |
| |
| TEST(X64_TRANSPOSE__4X2_SCALAR_INT, bh_5_8_bw_2){ |
| for(size_t i = 5; i < 8; ++i){ |
| TransposeMicrokernelTester() |
| .input_stride(2) |
| .output_stride(i) |
| .block_width(2) |
| .block_height(i) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__4x2_scalar_int); |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__4X2_SCALAR_INT, bh_5_8_bw_4){ |
| for(size_t i = 5; i < 8; ++i){ |
| TransposeMicrokernelTester() |
| .input_stride(4) |
| .output_stride(i) |
| .block_width(4) |
| .block_height(i) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__4x2_scalar_int); |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__4X2_SCALAR_INT, bh_5_8_bw_3_4) { |
| for(size_t i = 5; i < 8; ++i){ |
| for(size_t j = 3; j < 4; ++j){ |
| TransposeMicrokernelTester() |
| .input_stride(j) |
| .output_stride(i) |
| .block_width(j) |
| .block_height(i) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__4x2_scalar_int); |
| } |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__4X2_SCALAR_INT, bh_4_bw_2_is_4) { |
| TransposeMicrokernelTester() |
| .input_stride(4) |
| .output_stride(4) |
| .block_width(2) |
| .block_height(4) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__4x2_scalar_int); |
| } |
| |
| TEST(X64_TRANSPOSE__4X2_SCALAR_INT, bh_4_bw_2_os_8) { |
| TransposeMicrokernelTester() |
| .input_stride(2) |
| .output_stride(8) |
| .block_width(2) |
| .block_height(4) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__4x2_scalar_int); |
| } |
| |
| TEST(X64_TRANSPOSE__4X2_SCALAR_INT, bh_4_bw_2_is_4_os_8) { |
| TransposeMicrokernelTester() |
| .input_stride(4) |
| .output_stride(8) |
| .block_width(2) |
| .block_height(4) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__4x2_scalar_int); |
| } |
| |
| TEST(X64_TRANSPOSE__1X2_SCALAR_FLOAT, bh_1_bw_2) { |
| TransposeMicrokernelTester() |
| .input_stride(2) |
| .output_stride(1) |
| .block_width(2) |
| .block_height(1) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__1x2_scalar_float); |
| } |
| |
| TEST(X64_TRANSPOSE__1X2_SCALAR_FLOAT, bh_1_2_bw_1_4) { |
| for(size_t i = 1; i <= 2; ++i){ |
| for(size_t j = 1; j <= 4; ++j){ |
| TransposeMicrokernelTester() |
| .input_stride(j) |
| .output_stride(i) |
| .block_width(j) |
| .block_height(i) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__1x2_scalar_float); |
| } |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__1X2_SCALAR_FLOAT, bh_1_bw_4) { |
| TransposeMicrokernelTester() |
| .input_stride(4) |
| .output_stride(1) |
| .block_width(4) |
| .block_height(1) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__1x2_scalar_float); |
| } |
| |
| TEST(X64_TRANSPOSE__1X2_SCALAR_FLOAT, bh_1_bw_3_4) { |
| for(size_t i = 3; i < 4; ++i){ |
| TransposeMicrokernelTester() |
| .input_stride(i) |
| .output_stride(1) |
| .block_width(i) |
| .block_height(1) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__1x2_scalar_float); |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__1X2_SCALAR_FLOAT, bh_2_bw_3_4) { |
| for(size_t i = 3; i < 4; ++i){ |
| TransposeMicrokernelTester() |
| .input_stride(i) |
| .output_stride(2) |
| .block_width(i) |
| .block_height(2) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__1x2_scalar_float); |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__1X2_SCALAR_FLOAT, bh_2_bw_2) { |
| TransposeMicrokernelTester() |
| .input_stride(2) |
| .output_stride(2) |
| .block_width(2) |
| .block_height(2) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__1x2_scalar_float); |
| } |
| |
| TEST(X64_TRANSPOSE__1X2_SCALAR_FLOAT, bh_2_2_bw_2){ |
| for(size_t i = 2; i < 2; ++i){ |
| TransposeMicrokernelTester() |
| .input_stride(2) |
| .output_stride(i) |
| .block_width(2) |
| .block_height(i) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__1x2_scalar_float); |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__1X2_SCALAR_FLOAT, bh_2_2_bw_4){ |
| for(size_t i = 2; i < 2; ++i){ |
| TransposeMicrokernelTester() |
| .input_stride(4) |
| .output_stride(i) |
| .block_width(4) |
| .block_height(i) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__1x2_scalar_float); |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__1X2_SCALAR_FLOAT, bh_2_2_bw_3_4) { |
| for(size_t i = 2; i < 2; ++i){ |
| for(size_t j = 3; j < 4; ++j){ |
| TransposeMicrokernelTester() |
| .input_stride(j) |
| .output_stride(i) |
| .block_width(j) |
| .block_height(i) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__1x2_scalar_float); |
| } |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__1X2_SCALAR_FLOAT, bh_1_bw_2_is_4) { |
| TransposeMicrokernelTester() |
| .input_stride(4) |
| .output_stride(1) |
| .block_width(2) |
| .block_height(1) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__1x2_scalar_float); |
| } |
| |
| TEST(X64_TRANSPOSE__1X2_SCALAR_FLOAT, bh_1_bw_2_os_2) { |
| TransposeMicrokernelTester() |
| .input_stride(2) |
| .output_stride(2) |
| .block_width(2) |
| .block_height(1) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__1x2_scalar_float); |
| } |
| |
| TEST(X64_TRANSPOSE__1X2_SCALAR_FLOAT, bh_1_bw_2_is_4_os_2) { |
| TransposeMicrokernelTester() |
| .input_stride(4) |
| .output_stride(2) |
| .block_width(2) |
| .block_height(1) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__1x2_scalar_float); |
| } |
| |
| TEST(X64_TRANSPOSE__2X1_SCALAR_FLOAT, bh_2_bw_1) { |
| TransposeMicrokernelTester() |
| .input_stride(1) |
| .output_stride(2) |
| .block_width(1) |
| .block_height(2) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x1_scalar_float); |
| } |
| |
| TEST(X64_TRANSPOSE__2X1_SCALAR_FLOAT, bh_1_4_bw_1_2) { |
| for(size_t i = 1; i <= 4; ++i){ |
| for(size_t j = 1; j <= 2; ++j){ |
| TransposeMicrokernelTester() |
| .input_stride(j) |
| .output_stride(i) |
| .block_width(j) |
| .block_height(i) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x1_scalar_float); |
| } |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__2X1_SCALAR_FLOAT, bh_2_bw_2) { |
| TransposeMicrokernelTester() |
| .input_stride(2) |
| .output_stride(2) |
| .block_width(2) |
| .block_height(2) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x1_scalar_float); |
| } |
| |
| TEST(X64_TRANSPOSE__2X1_SCALAR_FLOAT, bh_2_bw_2_2) { |
| for(size_t i = 2; i < 2; ++i){ |
| TransposeMicrokernelTester() |
| .input_stride(i) |
| .output_stride(2) |
| .block_width(i) |
| .block_height(2) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x1_scalar_float); |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__2X1_SCALAR_FLOAT, bh_4_bw_2_2) { |
| for(size_t i = 2; i < 2; ++i){ |
| TransposeMicrokernelTester() |
| .input_stride(i) |
| .output_stride(4) |
| .block_width(i) |
| .block_height(4) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x1_scalar_float); |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__2X1_SCALAR_FLOAT, bh_4_bw_1) { |
| TransposeMicrokernelTester() |
| .input_stride(1) |
| .output_stride(4) |
| .block_width(1) |
| .block_height(4) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x1_scalar_float); |
| } |
| |
| TEST(X64_TRANSPOSE__2X1_SCALAR_FLOAT, bh_3_4_bw_1){ |
| for(size_t i = 3; i < 4; ++i){ |
| TransposeMicrokernelTester() |
| .input_stride(1) |
| .output_stride(i) |
| .block_width(1) |
| .block_height(i) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x1_scalar_float); |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__2X1_SCALAR_FLOAT, bh_3_4_bw_2){ |
| for(size_t i = 3; i < 4; ++i){ |
| TransposeMicrokernelTester() |
| .input_stride(2) |
| .output_stride(i) |
| .block_width(2) |
| .block_height(i) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x1_scalar_float); |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__2X1_SCALAR_FLOAT, bh_3_4_bw_2_2) { |
| for(size_t i = 3; i < 4; ++i){ |
| for(size_t j = 2; j < 2; ++j){ |
| TransposeMicrokernelTester() |
| .input_stride(j) |
| .output_stride(i) |
| .block_width(j) |
| .block_height(i) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x1_scalar_float); |
| } |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__2X1_SCALAR_FLOAT, bh_2_bw_1_is_2) { |
| TransposeMicrokernelTester() |
| .input_stride(2) |
| .output_stride(2) |
| .block_width(1) |
| .block_height(2) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x1_scalar_float); |
| } |
| |
| TEST(X64_TRANSPOSE__2X1_SCALAR_FLOAT, bh_2_bw_1_os_4) { |
| TransposeMicrokernelTester() |
| .input_stride(1) |
| .output_stride(4) |
| .block_width(1) |
| .block_height(2) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x1_scalar_float); |
| } |
| |
| TEST(X64_TRANSPOSE__2X1_SCALAR_FLOAT, bh_2_bw_1_is_2_os_4) { |
| TransposeMicrokernelTester() |
| .input_stride(2) |
| .output_stride(4) |
| .block_width(1) |
| .block_height(2) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x1_scalar_float); |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_SCALAR_FLOAT, bh_2_bw_2) { |
| TransposeMicrokernelTester() |
| .input_stride(2) |
| .output_stride(2) |
| .block_width(2) |
| .block_height(2) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_scalar_float); |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_SCALAR_FLOAT, bh_1_4_bw_1_4) { |
| for(size_t i = 1; i <= 4; ++i){ |
| for(size_t j = 1; j <= 4; ++j){ |
| TransposeMicrokernelTester() |
| .input_stride(j) |
| .output_stride(i) |
| .block_width(j) |
| .block_height(i) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_scalar_float); |
| } |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_SCALAR_FLOAT, bh_2_bw_4) { |
| TransposeMicrokernelTester() |
| .input_stride(4) |
| .output_stride(2) |
| .block_width(4) |
| .block_height(2) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_scalar_float); |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_SCALAR_FLOAT, bh_2_bw_3_4) { |
| for(size_t i = 3; i < 4; ++i){ |
| TransposeMicrokernelTester() |
| .input_stride(i) |
| .output_stride(2) |
| .block_width(i) |
| .block_height(2) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_scalar_float); |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_SCALAR_FLOAT, bh_4_bw_3_4) { |
| for(size_t i = 3; i < 4; ++i){ |
| TransposeMicrokernelTester() |
| .input_stride(i) |
| .output_stride(4) |
| .block_width(i) |
| .block_height(4) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_scalar_float); |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_SCALAR_FLOAT, bh_4_bw_2) { |
| TransposeMicrokernelTester() |
| .input_stride(2) |
| .output_stride(4) |
| .block_width(2) |
| .block_height(4) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_scalar_float); |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_SCALAR_FLOAT, bh_3_4_bw_2){ |
| for(size_t i = 3; i < 4; ++i){ |
| TransposeMicrokernelTester() |
| .input_stride(2) |
| .output_stride(i) |
| .block_width(2) |
| .block_height(i) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_scalar_float); |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_SCALAR_FLOAT, bh_3_4_bw_4){ |
| for(size_t i = 3; i < 4; ++i){ |
| TransposeMicrokernelTester() |
| .input_stride(4) |
| .output_stride(i) |
| .block_width(4) |
| .block_height(i) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_scalar_float); |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_SCALAR_FLOAT, bh_3_4_bw_3_4) { |
| for(size_t i = 3; i < 4; ++i){ |
| for(size_t j = 3; j < 4; ++j){ |
| TransposeMicrokernelTester() |
| .input_stride(j) |
| .output_stride(i) |
| .block_width(j) |
| .block_height(i) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_scalar_float); |
| } |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_SCALAR_FLOAT, bh_2_bw_2_is_4) { |
| TransposeMicrokernelTester() |
| .input_stride(4) |
| .output_stride(2) |
| .block_width(2) |
| .block_height(2) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_scalar_float); |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_SCALAR_FLOAT, bh_2_bw_2_os_4) { |
| TransposeMicrokernelTester() |
| .input_stride(2) |
| .output_stride(4) |
| .block_width(2) |
| .block_height(2) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_scalar_float); |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_SCALAR_FLOAT, bh_2_bw_2_is_4_os_4) { |
| TransposeMicrokernelTester() |
| .input_stride(4) |
| .output_stride(4) |
| .block_width(2) |
| .block_height(2) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_scalar_float); |
| } |
| |
| TEST(X64_TRANSPOSE__4X1_SCALAR_FLOAT, bh_4_bw_1) { |
| TransposeMicrokernelTester() |
| .input_stride(1) |
| .output_stride(4) |
| .block_width(1) |
| .block_height(4) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__4x1_scalar_float); |
| } |
| |
| TEST(X64_TRANSPOSE__4X1_SCALAR_FLOAT, bh_1_8_bw_1_2) { |
| for(size_t i = 1; i <= 8; ++i){ |
| for(size_t j = 1; j <= 2; ++j){ |
| TransposeMicrokernelTester() |
| .input_stride(j) |
| .output_stride(i) |
| .block_width(j) |
| .block_height(i) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__4x1_scalar_float); |
| } |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__4X1_SCALAR_FLOAT, bh_4_bw_2) { |
| TransposeMicrokernelTester() |
| .input_stride(2) |
| .output_stride(4) |
| .block_width(2) |
| .block_height(4) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__4x1_scalar_float); |
| } |
| |
| TEST(X64_TRANSPOSE__4X1_SCALAR_FLOAT, bh_4_bw_2_2) { |
| for(size_t i = 2; i < 2; ++i){ |
| TransposeMicrokernelTester() |
| .input_stride(i) |
| .output_stride(4) |
| .block_width(i) |
| .block_height(4) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__4x1_scalar_float); |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__4X1_SCALAR_FLOAT, bh_8_bw_2_2) { |
| for(size_t i = 2; i < 2; ++i){ |
| TransposeMicrokernelTester() |
| .input_stride(i) |
| .output_stride(8) |
| .block_width(i) |
| .block_height(8) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__4x1_scalar_float); |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__4X1_SCALAR_FLOAT, bh_8_bw_1) { |
| TransposeMicrokernelTester() |
| .input_stride(1) |
| .output_stride(8) |
| .block_width(1) |
| .block_height(8) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__4x1_scalar_float); |
| } |
| |
| TEST(X64_TRANSPOSE__4X1_SCALAR_FLOAT, bh_5_8_bw_1){ |
| for(size_t i = 5; i < 8; ++i){ |
| TransposeMicrokernelTester() |
| .input_stride(1) |
| .output_stride(i) |
| .block_width(1) |
| .block_height(i) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__4x1_scalar_float); |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__4X1_SCALAR_FLOAT, bh_5_8_bw_2){ |
| for(size_t i = 5; i < 8; ++i){ |
| TransposeMicrokernelTester() |
| .input_stride(2) |
| .output_stride(i) |
| .block_width(2) |
| .block_height(i) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__4x1_scalar_float); |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__4X1_SCALAR_FLOAT, bh_5_8_bw_2_2) { |
| for(size_t i = 5; i < 8; ++i){ |
| for(size_t j = 2; j < 2; ++j){ |
| TransposeMicrokernelTester() |
| .input_stride(j) |
| .output_stride(i) |
| .block_width(j) |
| .block_height(i) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__4x1_scalar_float); |
| } |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__4X1_SCALAR_FLOAT, bh_4_bw_1_is_2) { |
| TransposeMicrokernelTester() |
| .input_stride(2) |
| .output_stride(4) |
| .block_width(1) |
| .block_height(4) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__4x1_scalar_float); |
| } |
| |
| TEST(X64_TRANSPOSE__4X1_SCALAR_FLOAT, bh_4_bw_1_os_8) { |
| TransposeMicrokernelTester() |
| .input_stride(1) |
| .output_stride(8) |
| .block_width(1) |
| .block_height(4) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__4x1_scalar_float); |
| } |
| |
| TEST(X64_TRANSPOSE__4X1_SCALAR_FLOAT, bh_4_bw_1_is_2_os_8) { |
| TransposeMicrokernelTester() |
| .input_stride(2) |
| .output_stride(8) |
| .block_width(1) |
| .block_height(4) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__4x1_scalar_float); |
| } |
| |
| TEST(X64_TRANSPOSE__4X2_SCALAR_FLOAT, bh_4_bw_2) { |
| TransposeMicrokernelTester() |
| .input_stride(2) |
| .output_stride(4) |
| .block_width(2) |
| .block_height(4) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__4x2_scalar_float); |
| } |
| |
| TEST(X64_TRANSPOSE__4X2_SCALAR_FLOAT, bh_1_8_bw_1_4) { |
| for(size_t i = 1; i <= 8; ++i){ |
| for(size_t j = 1; j <= 4; ++j){ |
| TransposeMicrokernelTester() |
| .input_stride(j) |
| .output_stride(i) |
| .block_width(j) |
| .block_height(i) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__4x2_scalar_float); |
| } |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__4X2_SCALAR_FLOAT, bh_4_bw_4) { |
| TransposeMicrokernelTester() |
| .input_stride(4) |
| .output_stride(4) |
| .block_width(4) |
| .block_height(4) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__4x2_scalar_float); |
| } |
| |
| TEST(X64_TRANSPOSE__4X2_SCALAR_FLOAT, bh_4_bw_3_4) { |
| for(size_t i = 3; i < 4; ++i){ |
| TransposeMicrokernelTester() |
| .input_stride(i) |
| .output_stride(4) |
| .block_width(i) |
| .block_height(4) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__4x2_scalar_float); |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__4X2_SCALAR_FLOAT, bh_8_bw_3_4) { |
| for(size_t i = 3; i < 4; ++i){ |
| TransposeMicrokernelTester() |
| .input_stride(i) |
| .output_stride(8) |
| .block_width(i) |
| .block_height(8) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__4x2_scalar_float); |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__4X2_SCALAR_FLOAT, bh_8_bw_2) { |
| TransposeMicrokernelTester() |
| .input_stride(2) |
| .output_stride(8) |
| .block_width(2) |
| .block_height(8) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__4x2_scalar_float); |
| } |
| |
| TEST(X64_TRANSPOSE__4X2_SCALAR_FLOAT, bh_5_8_bw_2){ |
| for(size_t i = 5; i < 8; ++i){ |
| TransposeMicrokernelTester() |
| .input_stride(2) |
| .output_stride(i) |
| .block_width(2) |
| .block_height(i) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__4x2_scalar_float); |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__4X2_SCALAR_FLOAT, bh_5_8_bw_4){ |
| for(size_t i = 5; i < 8; ++i){ |
| TransposeMicrokernelTester() |
| .input_stride(4) |
| .output_stride(i) |
| .block_width(4) |
| .block_height(i) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__4x2_scalar_float); |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__4X2_SCALAR_FLOAT, bh_5_8_bw_3_4) { |
| for(size_t i = 5; i < 8; ++i){ |
| for(size_t j = 3; j < 4; ++j){ |
| TransposeMicrokernelTester() |
| .input_stride(j) |
| .output_stride(i) |
| .block_width(j) |
| .block_height(i) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__4x2_scalar_float); |
| } |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__4X2_SCALAR_FLOAT, bh_4_bw_2_is_4) { |
| TransposeMicrokernelTester() |
| .input_stride(4) |
| .output_stride(4) |
| .block_width(2) |
| .block_height(4) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__4x2_scalar_float); |
| } |
| |
| TEST(X64_TRANSPOSE__4X2_SCALAR_FLOAT, bh_4_bw_2_os_8) { |
| TransposeMicrokernelTester() |
| .input_stride(2) |
| .output_stride(8) |
| .block_width(2) |
| .block_height(4) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__4x2_scalar_float); |
| } |
| |
| TEST(X64_TRANSPOSE__4X2_SCALAR_FLOAT, bh_4_bw_2_is_4_os_8) { |
| TransposeMicrokernelTester() |
| .input_stride(4) |
| .output_stride(8) |
| .block_width(2) |
| .block_height(4) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__4x2_scalar_float); |
| } |
| |
| #if XNN_ARCH_X86 || XNN_ARCH_X86_64 |
| TEST(X64_TRANSPOSE__2X2_MULTI_MOV_SSE2, bh_2_bw_2) { |
| TEST_REQUIRES_X86_SSE2; |
| TransposeMicrokernelTester() |
| .input_stride(2) |
| .output_stride(2) |
| .block_width(2) |
| .block_height(2) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_multi_mov_sse2); |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_MULTI_MOV_SSE2, bh_1_4_bw_1_4) { |
| TEST_REQUIRES_X86_SSE2; |
| for(size_t i = 1; i <= 4; ++i){ |
| for(size_t j = 1; j <= 4; ++j){ |
| TransposeMicrokernelTester() |
| .input_stride(j) |
| .output_stride(i) |
| .block_width(j) |
| .block_height(i) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_multi_mov_sse2); |
| } |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_MULTI_MOV_SSE2, bh_2_bw_4) { |
| TEST_REQUIRES_X86_SSE2; |
| TransposeMicrokernelTester() |
| .input_stride(4) |
| .output_stride(2) |
| .block_width(4) |
| .block_height(2) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_multi_mov_sse2); |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_MULTI_MOV_SSE2, bh_2_bw_3_4) { |
| TEST_REQUIRES_X86_SSE2; |
| for(size_t i = 3; i < 4; ++i){ |
| TransposeMicrokernelTester() |
| .input_stride(i) |
| .output_stride(2) |
| .block_width(i) |
| .block_height(2) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_multi_mov_sse2); |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_MULTI_MOV_SSE2, bh_4_bw_3_4) { |
| TEST_REQUIRES_X86_SSE2; |
| for(size_t i = 3; i < 4; ++i){ |
| TransposeMicrokernelTester() |
| .input_stride(i) |
| .output_stride(4) |
| .block_width(i) |
| .block_height(4) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_multi_mov_sse2); |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_MULTI_MOV_SSE2, bh_4_bw_2) { |
| TEST_REQUIRES_X86_SSE2; |
| TransposeMicrokernelTester() |
| .input_stride(2) |
| .output_stride(4) |
| .block_width(2) |
| .block_height(4) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_multi_mov_sse2); |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_MULTI_MOV_SSE2, bh_3_4_bw_2){ |
| TEST_REQUIRES_X86_SSE2; |
| for(size_t i = 3; i < 4; ++i){ |
| TransposeMicrokernelTester() |
| .input_stride(2) |
| .output_stride(i) |
| .block_width(2) |
| .block_height(i) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_multi_mov_sse2); |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_MULTI_MOV_SSE2, bh_3_4_bw_4){ |
| TEST_REQUIRES_X86_SSE2; |
| for(size_t i = 3; i < 4; ++i){ |
| TransposeMicrokernelTester() |
| .input_stride(4) |
| .output_stride(i) |
| .block_width(4) |
| .block_height(i) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_multi_mov_sse2); |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_MULTI_MOV_SSE2, bh_3_4_bw_3_4) { |
| TEST_REQUIRES_X86_SSE2; |
| for(size_t i = 3; i < 4; ++i){ |
| for(size_t j = 3; j < 4; ++j){ |
| TransposeMicrokernelTester() |
| .input_stride(j) |
| .output_stride(i) |
| .block_width(j) |
| .block_height(i) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_multi_mov_sse2); |
| } |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_MULTI_MOV_SSE2, bh_2_bw_2_is_4) { |
| TEST_REQUIRES_X86_SSE2; |
| TransposeMicrokernelTester() |
| .input_stride(4) |
| .output_stride(2) |
| .block_width(2) |
| .block_height(2) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_multi_mov_sse2); |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_MULTI_MOV_SSE2, bh_2_bw_2_os_4) { |
| TEST_REQUIRES_X86_SSE2; |
| TransposeMicrokernelTester() |
| .input_stride(2) |
| .output_stride(4) |
| .block_width(2) |
| .block_height(2) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_multi_mov_sse2); |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_MULTI_MOV_SSE2, bh_2_bw_2_is_4_os_4) { |
| TEST_REQUIRES_X86_SSE2; |
| TransposeMicrokernelTester() |
| .input_stride(4) |
| .output_stride(4) |
| .block_width(2) |
| .block_height(2) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_multi_mov_sse2); |
| } |
| #endif // XNN_ARCH_X86 || XNN_ARCH_X86_64 |
| |
| |
| #if XNN_ARCH_X86 || XNN_ARCH_X86_64 |
| TEST(X64_TRANSPOSE__2X2_MULTI_MULTI_SSE2, bh_2_bw_2) { |
| TEST_REQUIRES_X86_SSE2; |
| TransposeMicrokernelTester() |
| .input_stride(2) |
| .output_stride(2) |
| .block_width(2) |
| .block_height(2) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_multi_multi_sse2); |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_MULTI_MULTI_SSE2, bh_1_4_bw_1_4) { |
| TEST_REQUIRES_X86_SSE2; |
| for(size_t i = 1; i <= 4; ++i){ |
| for(size_t j = 1; j <= 4; ++j){ |
| TransposeMicrokernelTester() |
| .input_stride(j) |
| .output_stride(i) |
| .block_width(j) |
| .block_height(i) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_multi_multi_sse2); |
| } |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_MULTI_MULTI_SSE2, bh_2_bw_4) { |
| TEST_REQUIRES_X86_SSE2; |
| TransposeMicrokernelTester() |
| .input_stride(4) |
| .output_stride(2) |
| .block_width(4) |
| .block_height(2) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_multi_multi_sse2); |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_MULTI_MULTI_SSE2, bh_2_bw_3_4) { |
| TEST_REQUIRES_X86_SSE2; |
| for(size_t i = 3; i < 4; ++i){ |
| TransposeMicrokernelTester() |
| .input_stride(i) |
| .output_stride(2) |
| .block_width(i) |
| .block_height(2) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_multi_multi_sse2); |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_MULTI_MULTI_SSE2, bh_4_bw_3_4) { |
| TEST_REQUIRES_X86_SSE2; |
| for(size_t i = 3; i < 4; ++i){ |
| TransposeMicrokernelTester() |
| .input_stride(i) |
| .output_stride(4) |
| .block_width(i) |
| .block_height(4) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_multi_multi_sse2); |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_MULTI_MULTI_SSE2, bh_4_bw_2) { |
| TEST_REQUIRES_X86_SSE2; |
| TransposeMicrokernelTester() |
| .input_stride(2) |
| .output_stride(4) |
| .block_width(2) |
| .block_height(4) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_multi_multi_sse2); |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_MULTI_MULTI_SSE2, bh_3_4_bw_2){ |
| TEST_REQUIRES_X86_SSE2; |
| for(size_t i = 3; i < 4; ++i){ |
| TransposeMicrokernelTester() |
| .input_stride(2) |
| .output_stride(i) |
| .block_width(2) |
| .block_height(i) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_multi_multi_sse2); |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_MULTI_MULTI_SSE2, bh_3_4_bw_4){ |
| TEST_REQUIRES_X86_SSE2; |
| for(size_t i = 3; i < 4; ++i){ |
| TransposeMicrokernelTester() |
| .input_stride(4) |
| .output_stride(i) |
| .block_width(4) |
| .block_height(i) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_multi_multi_sse2); |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_MULTI_MULTI_SSE2, bh_3_4_bw_3_4) { |
| TEST_REQUIRES_X86_SSE2; |
| for(size_t i = 3; i < 4; ++i){ |
| for(size_t j = 3; j < 4; ++j){ |
| TransposeMicrokernelTester() |
| .input_stride(j) |
| .output_stride(i) |
| .block_width(j) |
| .block_height(i) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_multi_multi_sse2); |
| } |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_MULTI_MULTI_SSE2, bh_2_bw_2_is_4) { |
| TEST_REQUIRES_X86_SSE2; |
| TransposeMicrokernelTester() |
| .input_stride(4) |
| .output_stride(2) |
| .block_width(2) |
| .block_height(2) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_multi_multi_sse2); |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_MULTI_MULTI_SSE2, bh_2_bw_2_os_4) { |
| TEST_REQUIRES_X86_SSE2; |
| TransposeMicrokernelTester() |
| .input_stride(2) |
| .output_stride(4) |
| .block_width(2) |
| .block_height(2) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_multi_multi_sse2); |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_MULTI_MULTI_SSE2, bh_2_bw_2_is_4_os_4) { |
| TEST_REQUIRES_X86_SSE2; |
| TransposeMicrokernelTester() |
| .input_stride(4) |
| .output_stride(4) |
| .block_width(2) |
| .block_height(2) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_multi_multi_sse2); |
| } |
| #endif // XNN_ARCH_X86 || XNN_ARCH_X86_64 |
| |
| |
| #if XNN_ARCH_X86 || XNN_ARCH_X86_64 |
| TEST(X64_TRANSPOSE__2X2_MULTI_SWITCH_SSE2, bh_2_bw_2) { |
| TEST_REQUIRES_X86_SSE2; |
| TransposeMicrokernelTester() |
| .input_stride(2) |
| .output_stride(2) |
| .block_width(2) |
| .block_height(2) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_multi_switch_sse2); |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_MULTI_SWITCH_SSE2, bh_1_4_bw_1_4) { |
| TEST_REQUIRES_X86_SSE2; |
| for(size_t i = 1; i <= 4; ++i){ |
| for(size_t j = 1; j <= 4; ++j){ |
| TransposeMicrokernelTester() |
| .input_stride(j) |
| .output_stride(i) |
| .block_width(j) |
| .block_height(i) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_multi_switch_sse2); |
| } |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_MULTI_SWITCH_SSE2, bh_2_bw_4) { |
| TEST_REQUIRES_X86_SSE2; |
| TransposeMicrokernelTester() |
| .input_stride(4) |
| .output_stride(2) |
| .block_width(4) |
| .block_height(2) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_multi_switch_sse2); |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_MULTI_SWITCH_SSE2, bh_2_bw_3_4) { |
| TEST_REQUIRES_X86_SSE2; |
| for(size_t i = 3; i < 4; ++i){ |
| TransposeMicrokernelTester() |
| .input_stride(i) |
| .output_stride(2) |
| .block_width(i) |
| .block_height(2) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_multi_switch_sse2); |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_MULTI_SWITCH_SSE2, bh_4_bw_3_4) { |
| TEST_REQUIRES_X86_SSE2; |
| for(size_t i = 3; i < 4; ++i){ |
| TransposeMicrokernelTester() |
| .input_stride(i) |
| .output_stride(4) |
| .block_width(i) |
| .block_height(4) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_multi_switch_sse2); |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_MULTI_SWITCH_SSE2, bh_4_bw_2) { |
| TEST_REQUIRES_X86_SSE2; |
| TransposeMicrokernelTester() |
| .input_stride(2) |
| .output_stride(4) |
| .block_width(2) |
| .block_height(4) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_multi_switch_sse2); |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_MULTI_SWITCH_SSE2, bh_3_4_bw_2){ |
| TEST_REQUIRES_X86_SSE2; |
| for(size_t i = 3; i < 4; ++i){ |
| TransposeMicrokernelTester() |
| .input_stride(2) |
| .output_stride(i) |
| .block_width(2) |
| .block_height(i) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_multi_switch_sse2); |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_MULTI_SWITCH_SSE2, bh_3_4_bw_4){ |
| TEST_REQUIRES_X86_SSE2; |
| for(size_t i = 3; i < 4; ++i){ |
| TransposeMicrokernelTester() |
| .input_stride(4) |
| .output_stride(i) |
| .block_width(4) |
| .block_height(i) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_multi_switch_sse2); |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_MULTI_SWITCH_SSE2, bh_3_4_bw_3_4) { |
| TEST_REQUIRES_X86_SSE2; |
| for(size_t i = 3; i < 4; ++i){ |
| for(size_t j = 3; j < 4; ++j){ |
| TransposeMicrokernelTester() |
| .input_stride(j) |
| .output_stride(i) |
| .block_width(j) |
| .block_height(i) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_multi_switch_sse2); |
| } |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_MULTI_SWITCH_SSE2, bh_2_bw_2_is_4) { |
| TEST_REQUIRES_X86_SSE2; |
| TransposeMicrokernelTester() |
| .input_stride(4) |
| .output_stride(2) |
| .block_width(2) |
| .block_height(2) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_multi_switch_sse2); |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_MULTI_SWITCH_SSE2, bh_2_bw_2_os_4) { |
| TEST_REQUIRES_X86_SSE2; |
| TransposeMicrokernelTester() |
| .input_stride(2) |
| .output_stride(4) |
| .block_width(2) |
| .block_height(2) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_multi_switch_sse2); |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_MULTI_SWITCH_SSE2, bh_2_bw_2_is_4_os_4) { |
| TEST_REQUIRES_X86_SSE2; |
| TransposeMicrokernelTester() |
| .input_stride(4) |
| .output_stride(4) |
| .block_width(2) |
| .block_height(2) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_multi_switch_sse2); |
| } |
| #endif // XNN_ARCH_X86 || XNN_ARCH_X86_64 |
| |
| |
| #if XNN_ARCH_X86 || XNN_ARCH_X86_64 |
| TEST(X64_TRANSPOSE__2X2_REUSE_MOV_SSE2, bh_2_bw_2) { |
| TEST_REQUIRES_X86_SSE2; |
| TransposeMicrokernelTester() |
| .input_stride(2) |
| .output_stride(2) |
| .block_width(2) |
| .block_height(2) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_reuse_mov_sse2); |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_REUSE_MOV_SSE2, bh_1_4_bw_1_4) { |
| TEST_REQUIRES_X86_SSE2; |
| for(size_t i = 1; i <= 4; ++i){ |
| for(size_t j = 1; j <= 4; ++j){ |
| TransposeMicrokernelTester() |
| .input_stride(j) |
| .output_stride(i) |
| .block_width(j) |
| .block_height(i) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_reuse_mov_sse2); |
| } |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_REUSE_MOV_SSE2, bh_2_bw_4) { |
| TEST_REQUIRES_X86_SSE2; |
| TransposeMicrokernelTester() |
| .input_stride(4) |
| .output_stride(2) |
| .block_width(4) |
| .block_height(2) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_reuse_mov_sse2); |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_REUSE_MOV_SSE2, bh_2_bw_3_4) { |
| TEST_REQUIRES_X86_SSE2; |
| for(size_t i = 3; i < 4; ++i){ |
| TransposeMicrokernelTester() |
| .input_stride(i) |
| .output_stride(2) |
| .block_width(i) |
| .block_height(2) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_reuse_mov_sse2); |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_REUSE_MOV_SSE2, bh_4_bw_3_4) { |
| TEST_REQUIRES_X86_SSE2; |
| for(size_t i = 3; i < 4; ++i){ |
| TransposeMicrokernelTester() |
| .input_stride(i) |
| .output_stride(4) |
| .block_width(i) |
| .block_height(4) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_reuse_mov_sse2); |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_REUSE_MOV_SSE2, bh_4_bw_2) { |
| TEST_REQUIRES_X86_SSE2; |
| TransposeMicrokernelTester() |
| .input_stride(2) |
| .output_stride(4) |
| .block_width(2) |
| .block_height(4) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_reuse_mov_sse2); |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_REUSE_MOV_SSE2, bh_3_4_bw_2){ |
| TEST_REQUIRES_X86_SSE2; |
| for(size_t i = 3; i < 4; ++i){ |
| TransposeMicrokernelTester() |
| .input_stride(2) |
| .output_stride(i) |
| .block_width(2) |
| .block_height(i) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_reuse_mov_sse2); |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_REUSE_MOV_SSE2, bh_3_4_bw_4){ |
| TEST_REQUIRES_X86_SSE2; |
| for(size_t i = 3; i < 4; ++i){ |
| TransposeMicrokernelTester() |
| .input_stride(4) |
| .output_stride(i) |
| .block_width(4) |
| .block_height(i) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_reuse_mov_sse2); |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_REUSE_MOV_SSE2, bh_3_4_bw_3_4) { |
| TEST_REQUIRES_X86_SSE2; |
| for(size_t i = 3; i < 4; ++i){ |
| for(size_t j = 3; j < 4; ++j){ |
| TransposeMicrokernelTester() |
| .input_stride(j) |
| .output_stride(i) |
| .block_width(j) |
| .block_height(i) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_reuse_mov_sse2); |
| } |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_REUSE_MOV_SSE2, bh_2_bw_2_is_4) { |
| TEST_REQUIRES_X86_SSE2; |
| TransposeMicrokernelTester() |
| .input_stride(4) |
| .output_stride(2) |
| .block_width(2) |
| .block_height(2) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_reuse_mov_sse2); |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_REUSE_MOV_SSE2, bh_2_bw_2_os_4) { |
| TEST_REQUIRES_X86_SSE2; |
| TransposeMicrokernelTester() |
| .input_stride(2) |
| .output_stride(4) |
| .block_width(2) |
| .block_height(2) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_reuse_mov_sse2); |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_REUSE_MOV_SSE2, bh_2_bw_2_is_4_os_4) { |
| TEST_REQUIRES_X86_SSE2; |
| TransposeMicrokernelTester() |
| .input_stride(4) |
| .output_stride(4) |
| .block_width(2) |
| .block_height(2) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_reuse_mov_sse2); |
| } |
| #endif // XNN_ARCH_X86 || XNN_ARCH_X86_64 |
| |
| |
| #if XNN_ARCH_X86 || XNN_ARCH_X86_64 |
| TEST(X64_TRANSPOSE__2X2_REUSE_MULTI_SSE2, bh_2_bw_2) { |
| TEST_REQUIRES_X86_SSE2; |
| TransposeMicrokernelTester() |
| .input_stride(2) |
| .output_stride(2) |
| .block_width(2) |
| .block_height(2) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_reuse_multi_sse2); |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_REUSE_MULTI_SSE2, bh_1_4_bw_1_4) { |
| TEST_REQUIRES_X86_SSE2; |
| for(size_t i = 1; i <= 4; ++i){ |
| for(size_t j = 1; j <= 4; ++j){ |
| TransposeMicrokernelTester() |
| .input_stride(j) |
| .output_stride(i) |
| .block_width(j) |
| .block_height(i) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_reuse_multi_sse2); |
| } |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_REUSE_MULTI_SSE2, bh_2_bw_4) { |
| TEST_REQUIRES_X86_SSE2; |
| TransposeMicrokernelTester() |
| .input_stride(4) |
| .output_stride(2) |
| .block_width(4) |
| .block_height(2) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_reuse_multi_sse2); |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_REUSE_MULTI_SSE2, bh_2_bw_3_4) { |
| TEST_REQUIRES_X86_SSE2; |
| for(size_t i = 3; i < 4; ++i){ |
| TransposeMicrokernelTester() |
| .input_stride(i) |
| .output_stride(2) |
| .block_width(i) |
| .block_height(2) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_reuse_multi_sse2); |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_REUSE_MULTI_SSE2, bh_4_bw_3_4) { |
| TEST_REQUIRES_X86_SSE2; |
| for(size_t i = 3; i < 4; ++i){ |
| TransposeMicrokernelTester() |
| .input_stride(i) |
| .output_stride(4) |
| .block_width(i) |
| .block_height(4) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_reuse_multi_sse2); |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_REUSE_MULTI_SSE2, bh_4_bw_2) { |
| TEST_REQUIRES_X86_SSE2; |
| TransposeMicrokernelTester() |
| .input_stride(2) |
| .output_stride(4) |
| .block_width(2) |
| .block_height(4) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_reuse_multi_sse2); |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_REUSE_MULTI_SSE2, bh_3_4_bw_2){ |
| TEST_REQUIRES_X86_SSE2; |
| for(size_t i = 3; i < 4; ++i){ |
| TransposeMicrokernelTester() |
| .input_stride(2) |
| .output_stride(i) |
| .block_width(2) |
| .block_height(i) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_reuse_multi_sse2); |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_REUSE_MULTI_SSE2, bh_3_4_bw_4){ |
| TEST_REQUIRES_X86_SSE2; |
| for(size_t i = 3; i < 4; ++i){ |
| TransposeMicrokernelTester() |
| .input_stride(4) |
| .output_stride(i) |
| .block_width(4) |
| .block_height(i) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_reuse_multi_sse2); |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_REUSE_MULTI_SSE2, bh_3_4_bw_3_4) { |
| TEST_REQUIRES_X86_SSE2; |
| for(size_t i = 3; i < 4; ++i){ |
| for(size_t j = 3; j < 4; ++j){ |
| TransposeMicrokernelTester() |
| .input_stride(j) |
| .output_stride(i) |
| .block_width(j) |
| .block_height(i) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_reuse_multi_sse2); |
| } |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_REUSE_MULTI_SSE2, bh_2_bw_2_is_4) { |
| TEST_REQUIRES_X86_SSE2; |
| TransposeMicrokernelTester() |
| .input_stride(4) |
| .output_stride(2) |
| .block_width(2) |
| .block_height(2) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_reuse_multi_sse2); |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_REUSE_MULTI_SSE2, bh_2_bw_2_os_4) { |
| TEST_REQUIRES_X86_SSE2; |
| TransposeMicrokernelTester() |
| .input_stride(2) |
| .output_stride(4) |
| .block_width(2) |
| .block_height(2) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_reuse_multi_sse2); |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_REUSE_MULTI_SSE2, bh_2_bw_2_is_4_os_4) { |
| TEST_REQUIRES_X86_SSE2; |
| TransposeMicrokernelTester() |
| .input_stride(4) |
| .output_stride(4) |
| .block_width(2) |
| .block_height(2) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_reuse_multi_sse2); |
| } |
| #endif // XNN_ARCH_X86 || XNN_ARCH_X86_64 |
| |
| |
| #if XNN_ARCH_X86 || XNN_ARCH_X86_64 |
| TEST(X64_TRANSPOSE__2X2_REUSE_SWITCH_SSE2, bh_2_bw_2) { |
| TEST_REQUIRES_X86_SSE2; |
| TransposeMicrokernelTester() |
| .input_stride(2) |
| .output_stride(2) |
| .block_width(2) |
| .block_height(2) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_reuse_switch_sse2); |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_REUSE_SWITCH_SSE2, bh_1_4_bw_1_4) { |
| TEST_REQUIRES_X86_SSE2; |
| for(size_t i = 1; i <= 4; ++i){ |
| for(size_t j = 1; j <= 4; ++j){ |
| TransposeMicrokernelTester() |
| .input_stride(j) |
| .output_stride(i) |
| .block_width(j) |
| .block_height(i) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_reuse_switch_sse2); |
| } |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_REUSE_SWITCH_SSE2, bh_2_bw_4) { |
| TEST_REQUIRES_X86_SSE2; |
| TransposeMicrokernelTester() |
| .input_stride(4) |
| .output_stride(2) |
| .block_width(4) |
| .block_height(2) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_reuse_switch_sse2); |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_REUSE_SWITCH_SSE2, bh_2_bw_3_4) { |
| TEST_REQUIRES_X86_SSE2; |
| for(size_t i = 3; i < 4; ++i){ |
| TransposeMicrokernelTester() |
| .input_stride(i) |
| .output_stride(2) |
| .block_width(i) |
| .block_height(2) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_reuse_switch_sse2); |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_REUSE_SWITCH_SSE2, bh_4_bw_3_4) { |
| TEST_REQUIRES_X86_SSE2; |
| for(size_t i = 3; i < 4; ++i){ |
| TransposeMicrokernelTester() |
| .input_stride(i) |
| .output_stride(4) |
| .block_width(i) |
| .block_height(4) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_reuse_switch_sse2); |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_REUSE_SWITCH_SSE2, bh_4_bw_2) { |
| TEST_REQUIRES_X86_SSE2; |
| TransposeMicrokernelTester() |
| .input_stride(2) |
| .output_stride(4) |
| .block_width(2) |
| .block_height(4) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_reuse_switch_sse2); |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_REUSE_SWITCH_SSE2, bh_3_4_bw_2){ |
| TEST_REQUIRES_X86_SSE2; |
| for(size_t i = 3; i < 4; ++i){ |
| TransposeMicrokernelTester() |
| .input_stride(2) |
| .output_stride(i) |
| .block_width(2) |
| .block_height(i) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_reuse_switch_sse2); |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_REUSE_SWITCH_SSE2, bh_3_4_bw_4){ |
| TEST_REQUIRES_X86_SSE2; |
| for(size_t i = 3; i < 4; ++i){ |
| TransposeMicrokernelTester() |
| .input_stride(4) |
| .output_stride(i) |
| .block_width(4) |
| .block_height(i) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_reuse_switch_sse2); |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_REUSE_SWITCH_SSE2, bh_3_4_bw_3_4) { |
| TEST_REQUIRES_X86_SSE2; |
| for(size_t i = 3; i < 4; ++i){ |
| for(size_t j = 3; j < 4; ++j){ |
| TransposeMicrokernelTester() |
| .input_stride(j) |
| .output_stride(i) |
| .block_width(j) |
| .block_height(i) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_reuse_switch_sse2); |
| } |
| } |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_REUSE_SWITCH_SSE2, bh_2_bw_2_is_4) { |
| TEST_REQUIRES_X86_SSE2; |
| TransposeMicrokernelTester() |
| .input_stride(4) |
| .output_stride(2) |
| .block_width(2) |
| .block_height(2) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_reuse_switch_sse2); |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_REUSE_SWITCH_SSE2, bh_2_bw_2_os_4) { |
| TEST_REQUIRES_X86_SSE2; |
| TransposeMicrokernelTester() |
| .input_stride(2) |
| .output_stride(4) |
| .block_width(2) |
| .block_height(2) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_reuse_switch_sse2); |
| } |
| |
| TEST(X64_TRANSPOSE__2X2_REUSE_SWITCH_SSE2, bh_2_bw_2_is_4_os_4) { |
| TEST_REQUIRES_X86_SSE2; |
| TransposeMicrokernelTester() |
| .input_stride(4) |
| .output_stride(4) |
| .block_width(2) |
| .block_height(2) |
| .iterations(1) |
| .Test(xnn_x64_transpose_ukernel__2x2_reuse_switch_sse2); |
| } |
| #endif // XNN_ARCH_X86 || XNN_ARCH_X86_64 |