| // Copyright 2019 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/x32-packx.yaml |
| // Generator: tools/generate-pack-test.py |
| |
| |
| #include <gtest/gtest.h> |
| |
| #include <xnnpack/common.h> |
| #include <xnnpack/isa-checks.h> |
| |
| #include <xnnpack/packx.h> |
| #include "pack-microkernel-tester.h" |
| |
| |
| #if XNN_ARCH_ARM || XNN_ARCH_ARM64 |
| TEST(X32_PACKX_4X__NEON_ST4, k_eq_4) { |
| TEST_REQUIRES_ARM_NEON; |
| PackMicrokernelTester() |
| .mr(4) |
| .m(4) |
| .k(4) |
| .Test(xnn_x32_packx_ukernel_4x__neon_st4); |
| } |
| |
| TEST(X32_PACKX_4X__NEON_ST4, k_eq_4_subtile) { |
| TEST_REQUIRES_ARM_NEON; |
| for (size_t m = 1; m <= 4; m++) { |
| PackMicrokernelTester() |
| .mr(4) |
| .m(m) |
| .k(4) |
| .Test(xnn_x32_packx_ukernel_4x__neon_st4); |
| } |
| } |
| |
| TEST(X32_PACKX_4X__NEON_ST4, k_lt_4) { |
| TEST_REQUIRES_ARM_NEON; |
| for (size_t k = 1; k < 4; k++) { |
| PackMicrokernelTester() |
| .mr(4) |
| .m(4) |
| .k(k) |
| .Test(xnn_x32_packx_ukernel_4x__neon_st4); |
| } |
| } |
| |
| TEST(X32_PACKX_4X__NEON_ST4, k_lt_4_subtile) { |
| TEST_REQUIRES_ARM_NEON; |
| for (size_t k = 1; k < 4; k++) { |
| for (size_t m = 1; m <= 4; m++) { |
| PackMicrokernelTester() |
| .mr(4) |
| .m(m) |
| .k(k) |
| .Test(xnn_x32_packx_ukernel_4x__neon_st4); |
| } |
| } |
| } |
| |
| TEST(X32_PACKX_4X__NEON_ST4, k_gt_4) { |
| TEST_REQUIRES_ARM_NEON; |
| for (size_t k = 5; k < 8; k++) { |
| PackMicrokernelTester() |
| .mr(4) |
| .m(4) |
| .k(k) |
| .Test(xnn_x32_packx_ukernel_4x__neon_st4); |
| } |
| } |
| |
| TEST(X32_PACKX_4X__NEON_ST4, k_gt_4_subtile) { |
| TEST_REQUIRES_ARM_NEON; |
| for (size_t k = 5; k < 8; k++) { |
| for (size_t m = 1; m <= 4; m++) { |
| PackMicrokernelTester() |
| .mr(4) |
| .m(m) |
| .k(k) |
| .Test(xnn_x32_packx_ukernel_4x__neon_st4); |
| } |
| } |
| } |
| |
| TEST(X32_PACKX_4X__NEON_ST4, k_div_4) { |
| TEST_REQUIRES_ARM_NEON; |
| for (size_t k = 8; k < 40; k += 4) { |
| PackMicrokernelTester() |
| .mr(4) |
| .m(4) |
| .k(k) |
| .Test(xnn_x32_packx_ukernel_4x__neon_st4); |
| } |
| } |
| |
| TEST(X32_PACKX_4X__NEON_ST4, k_div_4_subtile) { |
| TEST_REQUIRES_ARM_NEON; |
| for (size_t k = 8; k < 40; k += 4) { |
| for (size_t m = 1; m <= 4; m++) { |
| PackMicrokernelTester() |
| .mr(4) |
| .m(m) |
| .k(k) |
| .Test(xnn_x32_packx_ukernel_4x__neon_st4); |
| } |
| } |
| } |
| |
| TEST(X32_PACKX_4X__NEON_ST4, strided_x) { |
| TEST_REQUIRES_ARM_NEON; |
| for (size_t k = 1; k <= 20; k += 5) { |
| PackMicrokernelTester() |
| .mr(4) |
| .m(4) |
| .k(k) |
| .x_stride(23) |
| .Test(xnn_x32_packx_ukernel_4x__neon_st4); |
| } |
| } |
| #endif // XNN_ARCH_ARM || XNN_ARCH_ARM64 |
| |
| |
| #if XNN_ARCH_X86 || XNN_ARCH_X86_64 |
| TEST(X32_PACKX_4X__SSE, k_eq_4) { |
| TEST_REQUIRES_X86_SSE; |
| PackMicrokernelTester() |
| .mr(4) |
| .m(4) |
| .k(4) |
| .Test(xnn_x32_packx_ukernel_4x__sse); |
| } |
| |
| TEST(X32_PACKX_4X__SSE, k_eq_4_subtile) { |
| TEST_REQUIRES_X86_SSE; |
| for (size_t m = 1; m <= 4; m++) { |
| PackMicrokernelTester() |
| .mr(4) |
| .m(m) |
| .k(4) |
| .Test(xnn_x32_packx_ukernel_4x__sse); |
| } |
| } |
| |
| TEST(X32_PACKX_4X__SSE, k_lt_4) { |
| TEST_REQUIRES_X86_SSE; |
| for (size_t k = 1; k < 4; k++) { |
| PackMicrokernelTester() |
| .mr(4) |
| .m(4) |
| .k(k) |
| .Test(xnn_x32_packx_ukernel_4x__sse); |
| } |
| } |
| |
| TEST(X32_PACKX_4X__SSE, k_lt_4_subtile) { |
| TEST_REQUIRES_X86_SSE; |
| for (size_t k = 1; k < 4; k++) { |
| for (size_t m = 1; m <= 4; m++) { |
| PackMicrokernelTester() |
| .mr(4) |
| .m(m) |
| .k(k) |
| .Test(xnn_x32_packx_ukernel_4x__sse); |
| } |
| } |
| } |
| |
| TEST(X32_PACKX_4X__SSE, k_gt_4) { |
| TEST_REQUIRES_X86_SSE; |
| for (size_t k = 5; k < 8; k++) { |
| PackMicrokernelTester() |
| .mr(4) |
| .m(4) |
| .k(k) |
| .Test(xnn_x32_packx_ukernel_4x__sse); |
| } |
| } |
| |
| TEST(X32_PACKX_4X__SSE, k_gt_4_subtile) { |
| TEST_REQUIRES_X86_SSE; |
| for (size_t k = 5; k < 8; k++) { |
| for (size_t m = 1; m <= 4; m++) { |
| PackMicrokernelTester() |
| .mr(4) |
| .m(m) |
| .k(k) |
| .Test(xnn_x32_packx_ukernel_4x__sse); |
| } |
| } |
| } |
| |
| TEST(X32_PACKX_4X__SSE, k_div_4) { |
| TEST_REQUIRES_X86_SSE; |
| for (size_t k = 8; k < 40; k += 4) { |
| PackMicrokernelTester() |
| .mr(4) |
| .m(4) |
| .k(k) |
| .Test(xnn_x32_packx_ukernel_4x__sse); |
| } |
| } |
| |
| TEST(X32_PACKX_4X__SSE, k_div_4_subtile) { |
| TEST_REQUIRES_X86_SSE; |
| for (size_t k = 8; k < 40; k += 4) { |
| for (size_t m = 1; m <= 4; m++) { |
| PackMicrokernelTester() |
| .mr(4) |
| .m(m) |
| .k(k) |
| .Test(xnn_x32_packx_ukernel_4x__sse); |
| } |
| } |
| } |
| |
| TEST(X32_PACKX_4X__SSE, strided_x) { |
| TEST_REQUIRES_X86_SSE; |
| for (size_t k = 1; k <= 20; k += 5) { |
| PackMicrokernelTester() |
| .mr(4) |
| .m(4) |
| .k(k) |
| .x_stride(23) |
| .Test(xnn_x32_packx_ukernel_4x__sse); |
| } |
| } |
| #endif // XNN_ARCH_X86 || XNN_ARCH_X86_64 |
| |
| |
| #if !XNN_ARCH_ASMJS && !XNN_ARCH_WASM && !XNN_COMPILER_MSVC && !XNN_COMPILER_ICC |
| TEST(X32_PACKX_4X__PSIMD, k_eq_4) { |
| TEST_REQUIRES_PSIMD; |
| PackMicrokernelTester() |
| .mr(4) |
| .m(4) |
| .k(4) |
| .Test(xnn_x32_packx_ukernel_4x__psimd); |
| } |
| |
| TEST(X32_PACKX_4X__PSIMD, k_eq_4_subtile) { |
| TEST_REQUIRES_PSIMD; |
| for (size_t m = 1; m <= 4; m++) { |
| PackMicrokernelTester() |
| .mr(4) |
| .m(m) |
| .k(4) |
| .Test(xnn_x32_packx_ukernel_4x__psimd); |
| } |
| } |
| |
| TEST(X32_PACKX_4X__PSIMD, k_lt_4) { |
| TEST_REQUIRES_PSIMD; |
| for (size_t k = 1; k < 4; k++) { |
| PackMicrokernelTester() |
| .mr(4) |
| .m(4) |
| .k(k) |
| .Test(xnn_x32_packx_ukernel_4x__psimd); |
| } |
| } |
| |
| TEST(X32_PACKX_4X__PSIMD, k_lt_4_subtile) { |
| TEST_REQUIRES_PSIMD; |
| for (size_t k = 1; k < 4; k++) { |
| for (size_t m = 1; m <= 4; m++) { |
| PackMicrokernelTester() |
| .mr(4) |
| .m(m) |
| .k(k) |
| .Test(xnn_x32_packx_ukernel_4x__psimd); |
| } |
| } |
| } |
| |
| TEST(X32_PACKX_4X__PSIMD, k_gt_4) { |
| TEST_REQUIRES_PSIMD; |
| for (size_t k = 5; k < 8; k++) { |
| PackMicrokernelTester() |
| .mr(4) |
| .m(4) |
| .k(k) |
| .Test(xnn_x32_packx_ukernel_4x__psimd); |
| } |
| } |
| |
| TEST(X32_PACKX_4X__PSIMD, k_gt_4_subtile) { |
| TEST_REQUIRES_PSIMD; |
| for (size_t k = 5; k < 8; k++) { |
| for (size_t m = 1; m <= 4; m++) { |
| PackMicrokernelTester() |
| .mr(4) |
| .m(m) |
| .k(k) |
| .Test(xnn_x32_packx_ukernel_4x__psimd); |
| } |
| } |
| } |
| |
| TEST(X32_PACKX_4X__PSIMD, k_div_4) { |
| TEST_REQUIRES_PSIMD; |
| for (size_t k = 8; k < 40; k += 4) { |
| PackMicrokernelTester() |
| .mr(4) |
| .m(4) |
| .k(k) |
| .Test(xnn_x32_packx_ukernel_4x__psimd); |
| } |
| } |
| |
| TEST(X32_PACKX_4X__PSIMD, k_div_4_subtile) { |
| TEST_REQUIRES_PSIMD; |
| for (size_t k = 8; k < 40; k += 4) { |
| for (size_t m = 1; m <= 4; m++) { |
| PackMicrokernelTester() |
| .mr(4) |
| .m(m) |
| .k(k) |
| .Test(xnn_x32_packx_ukernel_4x__psimd); |
| } |
| } |
| } |
| |
| TEST(X32_PACKX_4X__PSIMD, strided_x) { |
| TEST_REQUIRES_PSIMD; |
| for (size_t k = 1; k <= 20; k += 5) { |
| PackMicrokernelTester() |
| .mr(4) |
| .m(4) |
| .k(k) |
| .x_stride(23) |
| .Test(xnn_x32_packx_ukernel_4x__psimd); |
| } |
| } |
| #endif // !XNN_ARCH_ASMJS && !XNN_ARCH_WASM && !XNN_COMPILER_MSVC && !XNN_COMPILER_ICC |
| |
| |
| #if XNN_ARCH_WASMSIMD |
| TEST(X32_PACKX_4X__WASMSIMD, k_eq_4) { |
| PackMicrokernelTester() |
| .mr(4) |
| .m(4) |
| .k(4) |
| .Test(xnn_x32_packx_ukernel_4x__wasmsimd); |
| } |
| |
| TEST(X32_PACKX_4X__WASMSIMD, k_eq_4_subtile) { |
| for (size_t m = 1; m <= 4; m++) { |
| PackMicrokernelTester() |
| .mr(4) |
| .m(m) |
| .k(4) |
| .Test(xnn_x32_packx_ukernel_4x__wasmsimd); |
| } |
| } |
| |
| TEST(X32_PACKX_4X__WASMSIMD, k_lt_4) { |
| for (size_t k = 1; k < 4; k++) { |
| PackMicrokernelTester() |
| .mr(4) |
| .m(4) |
| .k(k) |
| .Test(xnn_x32_packx_ukernel_4x__wasmsimd); |
| } |
| } |
| |
| TEST(X32_PACKX_4X__WASMSIMD, k_lt_4_subtile) { |
| for (size_t k = 1; k < 4; k++) { |
| for (size_t m = 1; m <= 4; m++) { |
| PackMicrokernelTester() |
| .mr(4) |
| .m(m) |
| .k(k) |
| .Test(xnn_x32_packx_ukernel_4x__wasmsimd); |
| } |
| } |
| } |
| |
| TEST(X32_PACKX_4X__WASMSIMD, k_gt_4) { |
| for (size_t k = 5; k < 8; k++) { |
| PackMicrokernelTester() |
| .mr(4) |
| .m(4) |
| .k(k) |
| .Test(xnn_x32_packx_ukernel_4x__wasmsimd); |
| } |
| } |
| |
| TEST(X32_PACKX_4X__WASMSIMD, k_gt_4_subtile) { |
| for (size_t k = 5; k < 8; k++) { |
| for (size_t m = 1; m <= 4; m++) { |
| PackMicrokernelTester() |
| .mr(4) |
| .m(m) |
| .k(k) |
| .Test(xnn_x32_packx_ukernel_4x__wasmsimd); |
| } |
| } |
| } |
| |
| TEST(X32_PACKX_4X__WASMSIMD, k_div_4) { |
| for (size_t k = 8; k < 40; k += 4) { |
| PackMicrokernelTester() |
| .mr(4) |
| .m(4) |
| .k(k) |
| .Test(xnn_x32_packx_ukernel_4x__wasmsimd); |
| } |
| } |
| |
| TEST(X32_PACKX_4X__WASMSIMD, k_div_4_subtile) { |
| for (size_t k = 8; k < 40; k += 4) { |
| for (size_t m = 1; m <= 4; m++) { |
| PackMicrokernelTester() |
| .mr(4) |
| .m(m) |
| .k(k) |
| .Test(xnn_x32_packx_ukernel_4x__wasmsimd); |
| } |
| } |
| } |
| |
| TEST(X32_PACKX_4X__WASMSIMD, strided_x) { |
| for (size_t k = 1; k <= 20; k += 5) { |
| PackMicrokernelTester() |
| .mr(4) |
| .m(4) |
| .k(k) |
| .x_stride(23) |
| .Test(xnn_x32_packx_ukernel_4x__wasmsimd); |
| } |
| } |
| #endif // XNN_ARCH_WASMSIMD |
| |
| |
| TEST(X32_PACKX_2X__SCALAR, k_eq_1) { |
| PackMicrokernelTester() |
| .mr(2) |
| .m(2) |
| .k(1) |
| .Test(xnn_x32_packx_ukernel_2x__scalar); |
| } |
| |
| TEST(X32_PACKX_2X__SCALAR, k_eq_1_subtile) { |
| for (size_t m = 1; m <= 2; m++) { |
| PackMicrokernelTester() |
| .mr(2) |
| .m(m) |
| .k(1) |
| .Test(xnn_x32_packx_ukernel_2x__scalar); |
| } |
| } |
| |
| TEST(X32_PACKX_2X__SCALAR, k_gt_1) { |
| for (size_t k = 2; k < 10; k++) { |
| PackMicrokernelTester() |
| .mr(2) |
| .m(2) |
| .k(k) |
| .Test(xnn_x32_packx_ukernel_2x__scalar); |
| } |
| } |
| |
| TEST(X32_PACKX_2X__SCALAR, k_gt_1_subtile) { |
| for (size_t k = 2; k < 10; k++) { |
| for (size_t m = 1; m <= 2; m++) { |
| PackMicrokernelTester() |
| .mr(2) |
| .m(m) |
| .k(k) |
| .Test(xnn_x32_packx_ukernel_2x__scalar); |
| } |
| } |
| } |
| |
| TEST(X32_PACKX_2X__SCALAR, strided_x) { |
| for (size_t k = 1; k <= 5; k += 2) { |
| PackMicrokernelTester() |
| .mr(2) |
| .m(2) |
| .k(k) |
| .x_stride(7) |
| .Test(xnn_x32_packx_ukernel_2x__scalar); |
| } |
| } |
| |
| TEST(X32_PACKX_3X__SCALAR, k_eq_1) { |
| PackMicrokernelTester() |
| .mr(3) |
| .m(3) |
| .k(1) |
| .Test(xnn_x32_packx_ukernel_3x__scalar); |
| } |
| |
| TEST(X32_PACKX_3X__SCALAR, k_eq_1_subtile) { |
| for (size_t m = 1; m <= 3; m++) { |
| PackMicrokernelTester() |
| .mr(3) |
| .m(m) |
| .k(1) |
| .Test(xnn_x32_packx_ukernel_3x__scalar); |
| } |
| } |
| |
| TEST(X32_PACKX_3X__SCALAR, k_gt_1) { |
| for (size_t k = 2; k < 10; k++) { |
| PackMicrokernelTester() |
| .mr(3) |
| .m(3) |
| .k(k) |
| .Test(xnn_x32_packx_ukernel_3x__scalar); |
| } |
| } |
| |
| TEST(X32_PACKX_3X__SCALAR, k_gt_1_subtile) { |
| for (size_t k = 2; k < 10; k++) { |
| for (size_t m = 1; m <= 3; m++) { |
| PackMicrokernelTester() |
| .mr(3) |
| .m(m) |
| .k(k) |
| .Test(xnn_x32_packx_ukernel_3x__scalar); |
| } |
| } |
| } |
| |
| TEST(X32_PACKX_3X__SCALAR, strided_x) { |
| for (size_t k = 1; k <= 5; k += 2) { |
| PackMicrokernelTester() |
| .mr(3) |
| .m(3) |
| .k(k) |
| .x_stride(7) |
| .Test(xnn_x32_packx_ukernel_3x__scalar); |
| } |
| } |
| |
| TEST(X32_PACKX_4X__SCALAR, k_eq_1) { |
| PackMicrokernelTester() |
| .mr(4) |
| .m(4) |
| .k(1) |
| .Test(xnn_x32_packx_ukernel_4x__scalar); |
| } |
| |
| TEST(X32_PACKX_4X__SCALAR, k_eq_1_subtile) { |
| for (size_t m = 1; m <= 4; m++) { |
| PackMicrokernelTester() |
| .mr(4) |
| .m(m) |
| .k(1) |
| .Test(xnn_x32_packx_ukernel_4x__scalar); |
| } |
| } |
| |
| TEST(X32_PACKX_4X__SCALAR, k_gt_1) { |
| for (size_t k = 2; k < 10; k++) { |
| PackMicrokernelTester() |
| .mr(4) |
| .m(4) |
| .k(k) |
| .Test(xnn_x32_packx_ukernel_4x__scalar); |
| } |
| } |
| |
| TEST(X32_PACKX_4X__SCALAR, k_gt_1_subtile) { |
| for (size_t k = 2; k < 10; k++) { |
| for (size_t m = 1; m <= 4; m++) { |
| PackMicrokernelTester() |
| .mr(4) |
| .m(m) |
| .k(k) |
| .Test(xnn_x32_packx_ukernel_4x__scalar); |
| } |
| } |
| } |
| |
| TEST(X32_PACKX_4X__SCALAR, strided_x) { |
| for (size_t k = 1; k <= 5; k += 2) { |
| PackMicrokernelTester() |
| .mr(4) |
| .m(4) |
| .k(k) |
| .x_stride(7) |
| .Test(xnn_x32_packx_ukernel_4x__scalar); |
| } |
| } |