blob: 1958c90f685f273b09baca4ae64c6d19f8e910c1 [file] [log] [blame]
// 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/f32-vmulcaddc.yaml
// Generator: tools/generate-vmulcaddc-test.py
#include <gtest/gtest.h>
#include <xnnpack/common.h>
#include <xnnpack/isa-checks.h>
#include <xnnpack/vmulcaddc.h>
#include "vmulcaddc-microkernel-tester.h"
#if XNN_ARCH_ARM || XNN_ARCH_ARM64
TEST(F32_VMULCADDC_C4__NEONFMA_2X, channels_eq_4) {
TEST_REQUIRES_ARM_NEON_FMA;
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(4)
.rows(2)
.Test(xnn_f32_vmulcaddc_ukernel_c4__neonfma_2x);
}
TEST(F32_VMULCADDC_C4__NEONFMA_2X, channels_div_4) {
TEST_REQUIRES_ARM_NEON_FMA;
for (size_t channels = 8; channels < 40; channels += 4) {
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(channels)
.rows(2)
.Test(xnn_f32_vmulcaddc_ukernel_c4__neonfma_2x);
}
}
TEST(F32_VMULCADDC_C4__NEONFMA_2X, channels_lt_4) {
TEST_REQUIRES_ARM_NEON_FMA;
for (size_t channels = 1; channels < 4; channels++) {
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(channels)
.rows(2)
.Test(xnn_f32_vmulcaddc_ukernel_c4__neonfma_2x);
}
}
TEST(F32_VMULCADDC_C4__NEONFMA_2X, channels_gt_4) {
TEST_REQUIRES_ARM_NEON_FMA;
for (size_t channels = 5; channels < 8; channels++) {
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(channels)
.rows(2)
.Test(xnn_f32_vmulcaddc_ukernel_c4__neonfma_2x);
}
}
TEST(F32_VMULCADDC_C4__NEONFMA_2X, rows_lt_2) {
TEST_REQUIRES_ARM_NEON_FMA;
for (size_t rows = 1; rows < 2; rows++) {
for (size_t channels = 1; channels <= 20; channels += 3) {
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(channels)
.rows(rows)
.Test(xnn_f32_vmulcaddc_ukernel_c4__neonfma_2x);
}
}
}
TEST(F32_VMULCADDC_C4__NEONFMA_2X, rows_div_2) {
TEST_REQUIRES_ARM_NEON_FMA;
for (size_t rows = 4; rows <= 8; rows += 2) {
for (size_t channels = 1; channels <= 20; channels += 3) {
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(channels)
.rows(rows)
.Test(xnn_f32_vmulcaddc_ukernel_c4__neonfma_2x);
}
}
}
TEST(F32_VMULCADDC_C4__NEONFMA_2X, rows_gt_2) {
TEST_REQUIRES_ARM_NEON_FMA;
for (size_t rows = 3; rows < 4; rows++) {
for (size_t channels = 1; channels <= 20; channels += 3) {
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(channels)
.rows(rows)
.Test(xnn_f32_vmulcaddc_ukernel_c4__neonfma_2x);
}
}
}
TEST(F32_VMULCADDC_C4__NEONFMA_2X, input_stride) {
TEST_REQUIRES_ARM_NEON_FMA;
for (size_t rows = 1; rows <= 6; rows += 1) {
for (size_t channels = 1; channels <= 20; channels += 3) {
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(channels)
.rows(rows)
.input_stride(23)
.Test(xnn_f32_vmulcaddc_ukernel_c4__neonfma_2x);
}
}
}
TEST(F32_VMULCADDC_C4__NEONFMA_2X, output_stride) {
TEST_REQUIRES_ARM_NEON_FMA;
for (size_t rows = 1; rows <= 6; rows += 1) {
for (size_t channels = 1; channels <= 20; channels += 3) {
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(channels)
.rows(rows)
.output_stride(23)
.Test(xnn_f32_vmulcaddc_ukernel_c4__neonfma_2x);
}
}
}
TEST(F32_VMULCADDC_C4__NEONFMA_2X, inplace) {
TEST_REQUIRES_ARM_NEON_FMA;
for (size_t rows = 1; rows <= 6; rows += 1) {
for (size_t channels = 1; channels <= 20; channels += 3) {
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(channels)
.rows(rows)
.inplace(true)
.Test(xnn_f32_vmulcaddc_ukernel_c4__neonfma_2x);
}
}
}
TEST(F32_VMULCADDC_C4__NEONFMA_2X, qmin) {
TEST_REQUIRES_ARM_NEON_FMA;
for (size_t rows = 1; rows <= 6; rows += 1) {
for (size_t channels = 1; channels <= 20; channels += 3) {
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(channels)
.rows(rows)
.qmin(128)
.Test(xnn_f32_vmulcaddc_ukernel_c4__neonfma_2x);
}
}
}
TEST(F32_VMULCADDC_C4__NEONFMA_2X, qmax) {
TEST_REQUIRES_ARM_NEON_FMA;
for (size_t rows = 1; rows <= 6; rows += 1) {
for (size_t channels = 1; channels <= 20; channels += 3) {
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(channels)
.rows(rows)
.qmax(128)
.Test(xnn_f32_vmulcaddc_ukernel_c4__neonfma_2x);
}
}
}
#endif // XNN_ARCH_ARM || XNN_ARCH_ARM64
#if XNN_ARCH_ARM || XNN_ARCH_ARM64
TEST(F32_VMULCADDC_C8__NEONFMA_2X, channels_eq_8) {
TEST_REQUIRES_ARM_NEON_FMA;
VMulCAddCMicrokernelTester()
.channel_tile(8)
.channels(8)
.rows(2)
.Test(xnn_f32_vmulcaddc_ukernel_c8__neonfma_2x);
}
TEST(F32_VMULCADDC_C8__NEONFMA_2X, channels_div_8) {
TEST_REQUIRES_ARM_NEON_FMA;
for (size_t channels = 16; channels < 80; channels += 8) {
VMulCAddCMicrokernelTester()
.channel_tile(8)
.channels(channels)
.rows(2)
.Test(xnn_f32_vmulcaddc_ukernel_c8__neonfma_2x);
}
}
TEST(F32_VMULCADDC_C8__NEONFMA_2X, channels_lt_8) {
TEST_REQUIRES_ARM_NEON_FMA;
for (size_t channels = 1; channels < 8; channels++) {
VMulCAddCMicrokernelTester()
.channel_tile(8)
.channels(channels)
.rows(2)
.Test(xnn_f32_vmulcaddc_ukernel_c8__neonfma_2x);
}
}
TEST(F32_VMULCADDC_C8__NEONFMA_2X, channels_gt_8) {
TEST_REQUIRES_ARM_NEON_FMA;
for (size_t channels = 9; channels < 16; channels++) {
VMulCAddCMicrokernelTester()
.channel_tile(8)
.channels(channels)
.rows(2)
.Test(xnn_f32_vmulcaddc_ukernel_c8__neonfma_2x);
}
}
TEST(F32_VMULCADDC_C8__NEONFMA_2X, rows_lt_2) {
TEST_REQUIRES_ARM_NEON_FMA;
for (size_t rows = 1; rows < 2; rows++) {
for (size_t channels = 1; channels <= 40; channels += 7) {
VMulCAddCMicrokernelTester()
.channel_tile(8)
.channels(channels)
.rows(rows)
.Test(xnn_f32_vmulcaddc_ukernel_c8__neonfma_2x);
}
}
}
TEST(F32_VMULCADDC_C8__NEONFMA_2X, rows_div_2) {
TEST_REQUIRES_ARM_NEON_FMA;
for (size_t rows = 4; rows <= 8; rows += 2) {
for (size_t channels = 1; channels <= 40; channels += 7) {
VMulCAddCMicrokernelTester()
.channel_tile(8)
.channels(channels)
.rows(rows)
.Test(xnn_f32_vmulcaddc_ukernel_c8__neonfma_2x);
}
}
}
TEST(F32_VMULCADDC_C8__NEONFMA_2X, rows_gt_2) {
TEST_REQUIRES_ARM_NEON_FMA;
for (size_t rows = 3; rows < 4; rows++) {
for (size_t channels = 1; channels <= 40; channels += 7) {
VMulCAddCMicrokernelTester()
.channel_tile(8)
.channels(channels)
.rows(rows)
.Test(xnn_f32_vmulcaddc_ukernel_c8__neonfma_2x);
}
}
}
TEST(F32_VMULCADDC_C8__NEONFMA_2X, input_stride) {
TEST_REQUIRES_ARM_NEON_FMA;
for (size_t rows = 1; rows <= 6; rows += 1) {
for (size_t channels = 1; channels <= 40; channels += 7) {
VMulCAddCMicrokernelTester()
.channel_tile(8)
.channels(channels)
.rows(rows)
.input_stride(43)
.Test(xnn_f32_vmulcaddc_ukernel_c8__neonfma_2x);
}
}
}
TEST(F32_VMULCADDC_C8__NEONFMA_2X, output_stride) {
TEST_REQUIRES_ARM_NEON_FMA;
for (size_t rows = 1; rows <= 6; rows += 1) {
for (size_t channels = 1; channels <= 40; channels += 7) {
VMulCAddCMicrokernelTester()
.channel_tile(8)
.channels(channels)
.rows(rows)
.output_stride(43)
.Test(xnn_f32_vmulcaddc_ukernel_c8__neonfma_2x);
}
}
}
TEST(F32_VMULCADDC_C8__NEONFMA_2X, inplace) {
TEST_REQUIRES_ARM_NEON_FMA;
for (size_t rows = 1; rows <= 6; rows += 1) {
for (size_t channels = 1; channels <= 40; channels += 7) {
VMulCAddCMicrokernelTester()
.channel_tile(8)
.channels(channels)
.rows(rows)
.inplace(true)
.Test(xnn_f32_vmulcaddc_ukernel_c8__neonfma_2x);
}
}
}
TEST(F32_VMULCADDC_C8__NEONFMA_2X, qmin) {
TEST_REQUIRES_ARM_NEON_FMA;
for (size_t rows = 1; rows <= 6; rows += 1) {
for (size_t channels = 1; channels <= 40; channels += 7) {
VMulCAddCMicrokernelTester()
.channel_tile(8)
.channels(channels)
.rows(rows)
.qmin(128)
.Test(xnn_f32_vmulcaddc_ukernel_c8__neonfma_2x);
}
}
}
TEST(F32_VMULCADDC_C8__NEONFMA_2X, qmax) {
TEST_REQUIRES_ARM_NEON_FMA;
for (size_t rows = 1; rows <= 6; rows += 1) {
for (size_t channels = 1; channels <= 40; channels += 7) {
VMulCAddCMicrokernelTester()
.channel_tile(8)
.channels(channels)
.rows(rows)
.qmax(128)
.Test(xnn_f32_vmulcaddc_ukernel_c8__neonfma_2x);
}
}
}
#endif // XNN_ARCH_ARM || XNN_ARCH_ARM64
#if XNN_ARCH_ARM || XNN_ARCH_ARM64
TEST(F32_VMULCADDC_C4__NEON_2X, channels_eq_4) {
TEST_REQUIRES_ARM_NEON;
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(4)
.rows(2)
.Test(xnn_f32_vmulcaddc_ukernel_c4__neon_2x);
}
TEST(F32_VMULCADDC_C4__NEON_2X, channels_div_4) {
TEST_REQUIRES_ARM_NEON;
for (size_t channels = 8; channels < 40; channels += 4) {
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(channels)
.rows(2)
.Test(xnn_f32_vmulcaddc_ukernel_c4__neon_2x);
}
}
TEST(F32_VMULCADDC_C4__NEON_2X, channels_lt_4) {
TEST_REQUIRES_ARM_NEON;
for (size_t channels = 1; channels < 4; channels++) {
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(channels)
.rows(2)
.Test(xnn_f32_vmulcaddc_ukernel_c4__neon_2x);
}
}
TEST(F32_VMULCADDC_C4__NEON_2X, channels_gt_4) {
TEST_REQUIRES_ARM_NEON;
for (size_t channels = 5; channels < 8; channels++) {
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(channels)
.rows(2)
.Test(xnn_f32_vmulcaddc_ukernel_c4__neon_2x);
}
}
TEST(F32_VMULCADDC_C4__NEON_2X, rows_lt_2) {
TEST_REQUIRES_ARM_NEON;
for (size_t rows = 1; rows < 2; rows++) {
for (size_t channels = 1; channels <= 20; channels += 3) {
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(channels)
.rows(rows)
.Test(xnn_f32_vmulcaddc_ukernel_c4__neon_2x);
}
}
}
TEST(F32_VMULCADDC_C4__NEON_2X, rows_div_2) {
TEST_REQUIRES_ARM_NEON;
for (size_t rows = 4; rows <= 8; rows += 2) {
for (size_t channels = 1; channels <= 20; channels += 3) {
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(channels)
.rows(rows)
.Test(xnn_f32_vmulcaddc_ukernel_c4__neon_2x);
}
}
}
TEST(F32_VMULCADDC_C4__NEON_2X, rows_gt_2) {
TEST_REQUIRES_ARM_NEON;
for (size_t rows = 3; rows < 4; rows++) {
for (size_t channels = 1; channels <= 20; channels += 3) {
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(channels)
.rows(rows)
.Test(xnn_f32_vmulcaddc_ukernel_c4__neon_2x);
}
}
}
TEST(F32_VMULCADDC_C4__NEON_2X, input_stride) {
TEST_REQUIRES_ARM_NEON;
for (size_t rows = 1; rows <= 6; rows += 1) {
for (size_t channels = 1; channels <= 20; channels += 3) {
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(channels)
.rows(rows)
.input_stride(23)
.Test(xnn_f32_vmulcaddc_ukernel_c4__neon_2x);
}
}
}
TEST(F32_VMULCADDC_C4__NEON_2X, output_stride) {
TEST_REQUIRES_ARM_NEON;
for (size_t rows = 1; rows <= 6; rows += 1) {
for (size_t channels = 1; channels <= 20; channels += 3) {
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(channels)
.rows(rows)
.output_stride(23)
.Test(xnn_f32_vmulcaddc_ukernel_c4__neon_2x);
}
}
}
TEST(F32_VMULCADDC_C4__NEON_2X, inplace) {
TEST_REQUIRES_ARM_NEON;
for (size_t rows = 1; rows <= 6; rows += 1) {
for (size_t channels = 1; channels <= 20; channels += 3) {
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(channels)
.rows(rows)
.inplace(true)
.Test(xnn_f32_vmulcaddc_ukernel_c4__neon_2x);
}
}
}
TEST(F32_VMULCADDC_C4__NEON_2X, qmin) {
TEST_REQUIRES_ARM_NEON;
for (size_t rows = 1; rows <= 6; rows += 1) {
for (size_t channels = 1; channels <= 20; channels += 3) {
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(channels)
.rows(rows)
.qmin(128)
.Test(xnn_f32_vmulcaddc_ukernel_c4__neon_2x);
}
}
}
TEST(F32_VMULCADDC_C4__NEON_2X, qmax) {
TEST_REQUIRES_ARM_NEON;
for (size_t rows = 1; rows <= 6; rows += 1) {
for (size_t channels = 1; channels <= 20; channels += 3) {
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(channels)
.rows(rows)
.qmax(128)
.Test(xnn_f32_vmulcaddc_ukernel_c4__neon_2x);
}
}
}
#endif // XNN_ARCH_ARM || XNN_ARCH_ARM64
#if XNN_ARCH_ARM || XNN_ARCH_ARM64
TEST(F32_VMULCADDC_C8__NEON_2X, channels_eq_8) {
TEST_REQUIRES_ARM_NEON;
VMulCAddCMicrokernelTester()
.channel_tile(8)
.channels(8)
.rows(2)
.Test(xnn_f32_vmulcaddc_ukernel_c8__neon_2x);
}
TEST(F32_VMULCADDC_C8__NEON_2X, channels_div_8) {
TEST_REQUIRES_ARM_NEON;
for (size_t channels = 16; channels < 80; channels += 8) {
VMulCAddCMicrokernelTester()
.channel_tile(8)
.channels(channels)
.rows(2)
.Test(xnn_f32_vmulcaddc_ukernel_c8__neon_2x);
}
}
TEST(F32_VMULCADDC_C8__NEON_2X, channels_lt_8) {
TEST_REQUIRES_ARM_NEON;
for (size_t channels = 1; channels < 8; channels++) {
VMulCAddCMicrokernelTester()
.channel_tile(8)
.channels(channels)
.rows(2)
.Test(xnn_f32_vmulcaddc_ukernel_c8__neon_2x);
}
}
TEST(F32_VMULCADDC_C8__NEON_2X, channels_gt_8) {
TEST_REQUIRES_ARM_NEON;
for (size_t channels = 9; channels < 16; channels++) {
VMulCAddCMicrokernelTester()
.channel_tile(8)
.channels(channels)
.rows(2)
.Test(xnn_f32_vmulcaddc_ukernel_c8__neon_2x);
}
}
TEST(F32_VMULCADDC_C8__NEON_2X, rows_lt_2) {
TEST_REQUIRES_ARM_NEON;
for (size_t rows = 1; rows < 2; rows++) {
for (size_t channels = 1; channels <= 40; channels += 7) {
VMulCAddCMicrokernelTester()
.channel_tile(8)
.channels(channels)
.rows(rows)
.Test(xnn_f32_vmulcaddc_ukernel_c8__neon_2x);
}
}
}
TEST(F32_VMULCADDC_C8__NEON_2X, rows_div_2) {
TEST_REQUIRES_ARM_NEON;
for (size_t rows = 4; rows <= 8; rows += 2) {
for (size_t channels = 1; channels <= 40; channels += 7) {
VMulCAddCMicrokernelTester()
.channel_tile(8)
.channels(channels)
.rows(rows)
.Test(xnn_f32_vmulcaddc_ukernel_c8__neon_2x);
}
}
}
TEST(F32_VMULCADDC_C8__NEON_2X, rows_gt_2) {
TEST_REQUIRES_ARM_NEON;
for (size_t rows = 3; rows < 4; rows++) {
for (size_t channels = 1; channels <= 40; channels += 7) {
VMulCAddCMicrokernelTester()
.channel_tile(8)
.channels(channels)
.rows(rows)
.Test(xnn_f32_vmulcaddc_ukernel_c8__neon_2x);
}
}
}
TEST(F32_VMULCADDC_C8__NEON_2X, input_stride) {
TEST_REQUIRES_ARM_NEON;
for (size_t rows = 1; rows <= 6; rows += 1) {
for (size_t channels = 1; channels <= 40; channels += 7) {
VMulCAddCMicrokernelTester()
.channel_tile(8)
.channels(channels)
.rows(rows)
.input_stride(43)
.Test(xnn_f32_vmulcaddc_ukernel_c8__neon_2x);
}
}
}
TEST(F32_VMULCADDC_C8__NEON_2X, output_stride) {
TEST_REQUIRES_ARM_NEON;
for (size_t rows = 1; rows <= 6; rows += 1) {
for (size_t channels = 1; channels <= 40; channels += 7) {
VMulCAddCMicrokernelTester()
.channel_tile(8)
.channels(channels)
.rows(rows)
.output_stride(43)
.Test(xnn_f32_vmulcaddc_ukernel_c8__neon_2x);
}
}
}
TEST(F32_VMULCADDC_C8__NEON_2X, inplace) {
TEST_REQUIRES_ARM_NEON;
for (size_t rows = 1; rows <= 6; rows += 1) {
for (size_t channels = 1; channels <= 40; channels += 7) {
VMulCAddCMicrokernelTester()
.channel_tile(8)
.channels(channels)
.rows(rows)
.inplace(true)
.Test(xnn_f32_vmulcaddc_ukernel_c8__neon_2x);
}
}
}
TEST(F32_VMULCADDC_C8__NEON_2X, qmin) {
TEST_REQUIRES_ARM_NEON;
for (size_t rows = 1; rows <= 6; rows += 1) {
for (size_t channels = 1; channels <= 40; channels += 7) {
VMulCAddCMicrokernelTester()
.channel_tile(8)
.channels(channels)
.rows(rows)
.qmin(128)
.Test(xnn_f32_vmulcaddc_ukernel_c8__neon_2x);
}
}
}
TEST(F32_VMULCADDC_C8__NEON_2X, qmax) {
TEST_REQUIRES_ARM_NEON;
for (size_t rows = 1; rows <= 6; rows += 1) {
for (size_t channels = 1; channels <= 40; channels += 7) {
VMulCAddCMicrokernelTester()
.channel_tile(8)
.channels(channels)
.rows(rows)
.qmax(128)
.Test(xnn_f32_vmulcaddc_ukernel_c8__neon_2x);
}
}
}
#endif // XNN_ARCH_ARM || XNN_ARCH_ARM64
#if XNN_ARCH_X86 || XNN_ARCH_X86_64
TEST(F32_VMULCADDC_C4__SSE_2X, channels_eq_4) {
TEST_REQUIRES_X86_SSE;
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(4)
.rows(2)
.Test(xnn_f32_vmulcaddc_ukernel_c4__sse_2x);
}
TEST(F32_VMULCADDC_C4__SSE_2X, channels_div_4) {
TEST_REQUIRES_X86_SSE;
for (size_t channels = 8; channels < 40; channels += 4) {
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(channels)
.rows(2)
.Test(xnn_f32_vmulcaddc_ukernel_c4__sse_2x);
}
}
TEST(F32_VMULCADDC_C4__SSE_2X, channels_lt_4) {
TEST_REQUIRES_X86_SSE;
for (size_t channels = 1; channels < 4; channels++) {
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(channels)
.rows(2)
.Test(xnn_f32_vmulcaddc_ukernel_c4__sse_2x);
}
}
TEST(F32_VMULCADDC_C4__SSE_2X, channels_gt_4) {
TEST_REQUIRES_X86_SSE;
for (size_t channels = 5; channels < 8; channels++) {
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(channels)
.rows(2)
.Test(xnn_f32_vmulcaddc_ukernel_c4__sse_2x);
}
}
TEST(F32_VMULCADDC_C4__SSE_2X, rows_lt_2) {
TEST_REQUIRES_X86_SSE;
for (size_t rows = 1; rows < 2; rows++) {
for (size_t channels = 1; channels <= 20; channels += 3) {
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(channels)
.rows(rows)
.Test(xnn_f32_vmulcaddc_ukernel_c4__sse_2x);
}
}
}
TEST(F32_VMULCADDC_C4__SSE_2X, rows_div_2) {
TEST_REQUIRES_X86_SSE;
for (size_t rows = 4; rows <= 8; rows += 2) {
for (size_t channels = 1; channels <= 20; channels += 3) {
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(channels)
.rows(rows)
.Test(xnn_f32_vmulcaddc_ukernel_c4__sse_2x);
}
}
}
TEST(F32_VMULCADDC_C4__SSE_2X, rows_gt_2) {
TEST_REQUIRES_X86_SSE;
for (size_t rows = 3; rows < 4; rows++) {
for (size_t channels = 1; channels <= 20; channels += 3) {
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(channels)
.rows(rows)
.Test(xnn_f32_vmulcaddc_ukernel_c4__sse_2x);
}
}
}
TEST(F32_VMULCADDC_C4__SSE_2X, input_stride) {
TEST_REQUIRES_X86_SSE;
for (size_t rows = 1; rows <= 6; rows += 1) {
for (size_t channels = 1; channels <= 20; channels += 3) {
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(channels)
.rows(rows)
.input_stride(23)
.Test(xnn_f32_vmulcaddc_ukernel_c4__sse_2x);
}
}
}
TEST(F32_VMULCADDC_C4__SSE_2X, output_stride) {
TEST_REQUIRES_X86_SSE;
for (size_t rows = 1; rows <= 6; rows += 1) {
for (size_t channels = 1; channels <= 20; channels += 3) {
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(channels)
.rows(rows)
.output_stride(23)
.Test(xnn_f32_vmulcaddc_ukernel_c4__sse_2x);
}
}
}
TEST(F32_VMULCADDC_C4__SSE_2X, inplace) {
TEST_REQUIRES_X86_SSE;
for (size_t rows = 1; rows <= 6; rows += 1) {
for (size_t channels = 1; channels <= 20; channels += 3) {
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(channels)
.rows(rows)
.inplace(true)
.Test(xnn_f32_vmulcaddc_ukernel_c4__sse_2x);
}
}
}
TEST(F32_VMULCADDC_C4__SSE_2X, qmin) {
TEST_REQUIRES_X86_SSE;
for (size_t rows = 1; rows <= 6; rows += 1) {
for (size_t channels = 1; channels <= 20; channels += 3) {
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(channels)
.rows(rows)
.qmin(128)
.Test(xnn_f32_vmulcaddc_ukernel_c4__sse_2x);
}
}
}
TEST(F32_VMULCADDC_C4__SSE_2X, qmax) {
TEST_REQUIRES_X86_SSE;
for (size_t rows = 1; rows <= 6; rows += 1) {
for (size_t channels = 1; channels <= 20; channels += 3) {
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(channels)
.rows(rows)
.qmax(128)
.Test(xnn_f32_vmulcaddc_ukernel_c4__sse_2x);
}
}
}
#endif // XNN_ARCH_X86 || XNN_ARCH_X86_64
#if XNN_ARCH_X86 || XNN_ARCH_X86_64
TEST(F32_VMULCADDC_C8__SSE_2X, channels_eq_8) {
TEST_REQUIRES_X86_SSE;
VMulCAddCMicrokernelTester()
.channel_tile(8)
.channels(8)
.rows(2)
.Test(xnn_f32_vmulcaddc_ukernel_c8__sse_2x);
}
TEST(F32_VMULCADDC_C8__SSE_2X, channels_div_8) {
TEST_REQUIRES_X86_SSE;
for (size_t channels = 16; channels < 80; channels += 8) {
VMulCAddCMicrokernelTester()
.channel_tile(8)
.channels(channels)
.rows(2)
.Test(xnn_f32_vmulcaddc_ukernel_c8__sse_2x);
}
}
TEST(F32_VMULCADDC_C8__SSE_2X, channels_lt_8) {
TEST_REQUIRES_X86_SSE;
for (size_t channels = 1; channels < 8; channels++) {
VMulCAddCMicrokernelTester()
.channel_tile(8)
.channels(channels)
.rows(2)
.Test(xnn_f32_vmulcaddc_ukernel_c8__sse_2x);
}
}
TEST(F32_VMULCADDC_C8__SSE_2X, channels_gt_8) {
TEST_REQUIRES_X86_SSE;
for (size_t channels = 9; channels < 16; channels++) {
VMulCAddCMicrokernelTester()
.channel_tile(8)
.channels(channels)
.rows(2)
.Test(xnn_f32_vmulcaddc_ukernel_c8__sse_2x);
}
}
TEST(F32_VMULCADDC_C8__SSE_2X, rows_lt_2) {
TEST_REQUIRES_X86_SSE;
for (size_t rows = 1; rows < 2; rows++) {
for (size_t channels = 1; channels <= 40; channels += 7) {
VMulCAddCMicrokernelTester()
.channel_tile(8)
.channels(channels)
.rows(rows)
.Test(xnn_f32_vmulcaddc_ukernel_c8__sse_2x);
}
}
}
TEST(F32_VMULCADDC_C8__SSE_2X, rows_div_2) {
TEST_REQUIRES_X86_SSE;
for (size_t rows = 4; rows <= 8; rows += 2) {
for (size_t channels = 1; channels <= 40; channels += 7) {
VMulCAddCMicrokernelTester()
.channel_tile(8)
.channels(channels)
.rows(rows)
.Test(xnn_f32_vmulcaddc_ukernel_c8__sse_2x);
}
}
}
TEST(F32_VMULCADDC_C8__SSE_2X, rows_gt_2) {
TEST_REQUIRES_X86_SSE;
for (size_t rows = 3; rows < 4; rows++) {
for (size_t channels = 1; channels <= 40; channels += 7) {
VMulCAddCMicrokernelTester()
.channel_tile(8)
.channels(channels)
.rows(rows)
.Test(xnn_f32_vmulcaddc_ukernel_c8__sse_2x);
}
}
}
TEST(F32_VMULCADDC_C8__SSE_2X, input_stride) {
TEST_REQUIRES_X86_SSE;
for (size_t rows = 1; rows <= 6; rows += 1) {
for (size_t channels = 1; channels <= 40; channels += 7) {
VMulCAddCMicrokernelTester()
.channel_tile(8)
.channels(channels)
.rows(rows)
.input_stride(43)
.Test(xnn_f32_vmulcaddc_ukernel_c8__sse_2x);
}
}
}
TEST(F32_VMULCADDC_C8__SSE_2X, output_stride) {
TEST_REQUIRES_X86_SSE;
for (size_t rows = 1; rows <= 6; rows += 1) {
for (size_t channels = 1; channels <= 40; channels += 7) {
VMulCAddCMicrokernelTester()
.channel_tile(8)
.channels(channels)
.rows(rows)
.output_stride(43)
.Test(xnn_f32_vmulcaddc_ukernel_c8__sse_2x);
}
}
}
TEST(F32_VMULCADDC_C8__SSE_2X, inplace) {
TEST_REQUIRES_X86_SSE;
for (size_t rows = 1; rows <= 6; rows += 1) {
for (size_t channels = 1; channels <= 40; channels += 7) {
VMulCAddCMicrokernelTester()
.channel_tile(8)
.channels(channels)
.rows(rows)
.inplace(true)
.Test(xnn_f32_vmulcaddc_ukernel_c8__sse_2x);
}
}
}
TEST(F32_VMULCADDC_C8__SSE_2X, qmin) {
TEST_REQUIRES_X86_SSE;
for (size_t rows = 1; rows <= 6; rows += 1) {
for (size_t channels = 1; channels <= 40; channels += 7) {
VMulCAddCMicrokernelTester()
.channel_tile(8)
.channels(channels)
.rows(rows)
.qmin(128)
.Test(xnn_f32_vmulcaddc_ukernel_c8__sse_2x);
}
}
}
TEST(F32_VMULCADDC_C8__SSE_2X, qmax) {
TEST_REQUIRES_X86_SSE;
for (size_t rows = 1; rows <= 6; rows += 1) {
for (size_t channels = 1; channels <= 40; channels += 7) {
VMulCAddCMicrokernelTester()
.channel_tile(8)
.channels(channels)
.rows(rows)
.qmax(128)
.Test(xnn_f32_vmulcaddc_ukernel_c8__sse_2x);
}
}
}
#endif // XNN_ARCH_X86 || XNN_ARCH_X86_64
#if !XNN_ARCH_ASMJS && !XNN_ARCH_WASM
TEST(F32_VMULCADDC_C4__PSIMD_2X, channels_eq_4) {
TEST_REQUIRES_PSIMD;
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(4)
.rows(2)
.Test(xnn_f32_vmulcaddc_ukernel_c4__psimd_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
TEST(F32_VMULCADDC_C4__PSIMD_2X, channels_div_4) {
TEST_REQUIRES_PSIMD;
for (size_t channels = 8; channels < 40; channels += 4) {
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(channels)
.rows(2)
.Test(xnn_f32_vmulcaddc_ukernel_c4__psimd_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
TEST(F32_VMULCADDC_C4__PSIMD_2X, channels_lt_4) {
TEST_REQUIRES_PSIMD;
for (size_t channels = 1; channels < 4; channels++) {
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(channels)
.rows(2)
.Test(xnn_f32_vmulcaddc_ukernel_c4__psimd_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
TEST(F32_VMULCADDC_C4__PSIMD_2X, channels_gt_4) {
TEST_REQUIRES_PSIMD;
for (size_t channels = 5; channels < 8; channels++) {
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(channels)
.rows(2)
.Test(xnn_f32_vmulcaddc_ukernel_c4__psimd_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
TEST(F32_VMULCADDC_C4__PSIMD_2X, rows_lt_2) {
TEST_REQUIRES_PSIMD;
for (size_t rows = 1; rows < 2; rows++) {
for (size_t channels = 1; channels <= 20; channels += 3) {
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(channels)
.rows(rows)
.Test(xnn_f32_vmulcaddc_ukernel_c4__psimd_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
}
TEST(F32_VMULCADDC_C4__PSIMD_2X, rows_div_2) {
TEST_REQUIRES_PSIMD;
for (size_t rows = 4; rows <= 8; rows += 2) {
for (size_t channels = 1; channels <= 20; channels += 3) {
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(channels)
.rows(rows)
.Test(xnn_f32_vmulcaddc_ukernel_c4__psimd_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
}
TEST(F32_VMULCADDC_C4__PSIMD_2X, rows_gt_2) {
TEST_REQUIRES_PSIMD;
for (size_t rows = 3; rows < 4; rows++) {
for (size_t channels = 1; channels <= 20; channels += 3) {
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(channels)
.rows(rows)
.Test(xnn_f32_vmulcaddc_ukernel_c4__psimd_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
}
TEST(F32_VMULCADDC_C4__PSIMD_2X, input_stride) {
TEST_REQUIRES_PSIMD;
for (size_t rows = 1; rows <= 6; rows += 1) {
for (size_t channels = 1; channels <= 20; channels += 3) {
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(channels)
.rows(rows)
.input_stride(23)
.Test(xnn_f32_vmulcaddc_ukernel_c4__psimd_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
}
TEST(F32_VMULCADDC_C4__PSIMD_2X, output_stride) {
TEST_REQUIRES_PSIMD;
for (size_t rows = 1; rows <= 6; rows += 1) {
for (size_t channels = 1; channels <= 20; channels += 3) {
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(channels)
.rows(rows)
.output_stride(23)
.Test(xnn_f32_vmulcaddc_ukernel_c4__psimd_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
}
TEST(F32_VMULCADDC_C4__PSIMD_2X, inplace) {
TEST_REQUIRES_PSIMD;
for (size_t rows = 1; rows <= 6; rows += 1) {
for (size_t channels = 1; channels <= 20; channels += 3) {
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(channels)
.rows(rows)
.inplace(true)
.Test(xnn_f32_vmulcaddc_ukernel_c4__psimd_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
}
TEST(F32_VMULCADDC_C4__PSIMD_2X, qmin) {
TEST_REQUIRES_PSIMD;
for (size_t rows = 1; rows <= 6; rows += 1) {
for (size_t channels = 1; channels <= 20; channels += 3) {
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(channels)
.rows(rows)
.qmin(128)
.Test(xnn_f32_vmulcaddc_ukernel_c4__psimd_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
}
TEST(F32_VMULCADDC_C4__PSIMD_2X, qmax) {
TEST_REQUIRES_PSIMD;
for (size_t rows = 1; rows <= 6; rows += 1) {
for (size_t channels = 1; channels <= 20; channels += 3) {
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(channels)
.rows(rows)
.qmax(128)
.Test(xnn_f32_vmulcaddc_ukernel_c4__psimd_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
}
#endif // !XNN_ARCH_ASMJS && !XNN_ARCH_WASM
#if !XNN_ARCH_ASMJS && !XNN_ARCH_WASM
TEST(F32_VMULCADDC_C8__PSIMD_2X, channels_eq_8) {
TEST_REQUIRES_PSIMD;
VMulCAddCMicrokernelTester()
.channel_tile(8)
.channels(8)
.rows(2)
.Test(xnn_f32_vmulcaddc_ukernel_c8__psimd_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
TEST(F32_VMULCADDC_C8__PSIMD_2X, channels_div_8) {
TEST_REQUIRES_PSIMD;
for (size_t channels = 16; channels < 80; channels += 8) {
VMulCAddCMicrokernelTester()
.channel_tile(8)
.channels(channels)
.rows(2)
.Test(xnn_f32_vmulcaddc_ukernel_c8__psimd_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
TEST(F32_VMULCADDC_C8__PSIMD_2X, channels_lt_8) {
TEST_REQUIRES_PSIMD;
for (size_t channels = 1; channels < 8; channels++) {
VMulCAddCMicrokernelTester()
.channel_tile(8)
.channels(channels)
.rows(2)
.Test(xnn_f32_vmulcaddc_ukernel_c8__psimd_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
TEST(F32_VMULCADDC_C8__PSIMD_2X, channels_gt_8) {
TEST_REQUIRES_PSIMD;
for (size_t channels = 9; channels < 16; channels++) {
VMulCAddCMicrokernelTester()
.channel_tile(8)
.channels(channels)
.rows(2)
.Test(xnn_f32_vmulcaddc_ukernel_c8__psimd_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
TEST(F32_VMULCADDC_C8__PSIMD_2X, rows_lt_2) {
TEST_REQUIRES_PSIMD;
for (size_t rows = 1; rows < 2; rows++) {
for (size_t channels = 1; channels <= 40; channels += 7) {
VMulCAddCMicrokernelTester()
.channel_tile(8)
.channels(channels)
.rows(rows)
.Test(xnn_f32_vmulcaddc_ukernel_c8__psimd_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
}
TEST(F32_VMULCADDC_C8__PSIMD_2X, rows_div_2) {
TEST_REQUIRES_PSIMD;
for (size_t rows = 4; rows <= 8; rows += 2) {
for (size_t channels = 1; channels <= 40; channels += 7) {
VMulCAddCMicrokernelTester()
.channel_tile(8)
.channels(channels)
.rows(rows)
.Test(xnn_f32_vmulcaddc_ukernel_c8__psimd_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
}
TEST(F32_VMULCADDC_C8__PSIMD_2X, rows_gt_2) {
TEST_REQUIRES_PSIMD;
for (size_t rows = 3; rows < 4; rows++) {
for (size_t channels = 1; channels <= 40; channels += 7) {
VMulCAddCMicrokernelTester()
.channel_tile(8)
.channels(channels)
.rows(rows)
.Test(xnn_f32_vmulcaddc_ukernel_c8__psimd_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
}
TEST(F32_VMULCADDC_C8__PSIMD_2X, input_stride) {
TEST_REQUIRES_PSIMD;
for (size_t rows = 1; rows <= 6; rows += 1) {
for (size_t channels = 1; channels <= 40; channels += 7) {
VMulCAddCMicrokernelTester()
.channel_tile(8)
.channels(channels)
.rows(rows)
.input_stride(43)
.Test(xnn_f32_vmulcaddc_ukernel_c8__psimd_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
}
TEST(F32_VMULCADDC_C8__PSIMD_2X, output_stride) {
TEST_REQUIRES_PSIMD;
for (size_t rows = 1; rows <= 6; rows += 1) {
for (size_t channels = 1; channels <= 40; channels += 7) {
VMulCAddCMicrokernelTester()
.channel_tile(8)
.channels(channels)
.rows(rows)
.output_stride(43)
.Test(xnn_f32_vmulcaddc_ukernel_c8__psimd_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
}
TEST(F32_VMULCADDC_C8__PSIMD_2X, inplace) {
TEST_REQUIRES_PSIMD;
for (size_t rows = 1; rows <= 6; rows += 1) {
for (size_t channels = 1; channels <= 40; channels += 7) {
VMulCAddCMicrokernelTester()
.channel_tile(8)
.channels(channels)
.rows(rows)
.inplace(true)
.Test(xnn_f32_vmulcaddc_ukernel_c8__psimd_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
}
TEST(F32_VMULCADDC_C8__PSIMD_2X, qmin) {
TEST_REQUIRES_PSIMD;
for (size_t rows = 1; rows <= 6; rows += 1) {
for (size_t channels = 1; channels <= 40; channels += 7) {
VMulCAddCMicrokernelTester()
.channel_tile(8)
.channels(channels)
.rows(rows)
.qmin(128)
.Test(xnn_f32_vmulcaddc_ukernel_c8__psimd_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
}
TEST(F32_VMULCADDC_C8__PSIMD_2X, qmax) {
TEST_REQUIRES_PSIMD;
for (size_t rows = 1; rows <= 6; rows += 1) {
for (size_t channels = 1; channels <= 40; channels += 7) {
VMulCAddCMicrokernelTester()
.channel_tile(8)
.channels(channels)
.rows(rows)
.qmax(128)
.Test(xnn_f32_vmulcaddc_ukernel_c8__psimd_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
}
#endif // !XNN_ARCH_ASMJS && !XNN_ARCH_WASM
#if XNN_ARCH_WASM
TEST(F32_VMULCADDC_C1__WASM_2X, channels_eq_1) {
VMulCAddCMicrokernelTester()
.channel_tile(1)
.channels(1)
.rows(2)
.Test(xnn_f32_vmulcaddc_ukernel_c1__wasm_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
TEST(F32_VMULCADDC_C1__WASM_2X, channels_gt_1) {
for (size_t channels = 2; channels < 10; channels++) {
VMulCAddCMicrokernelTester()
.channel_tile(1)
.channels(channels)
.rows(2)
.Test(xnn_f32_vmulcaddc_ukernel_c1__wasm_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
TEST(F32_VMULCADDC_C1__WASM_2X, rows_lt_2) {
for (size_t rows = 1; rows < 2; rows++) {
for (size_t channels = 1; channels <= 5; channels += 1) {
VMulCAddCMicrokernelTester()
.channel_tile(1)
.channels(channels)
.rows(rows)
.Test(xnn_f32_vmulcaddc_ukernel_c1__wasm_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
}
TEST(F32_VMULCADDC_C1__WASM_2X, rows_div_2) {
for (size_t rows = 4; rows <= 8; rows += 2) {
for (size_t channels = 1; channels <= 5; channels += 1) {
VMulCAddCMicrokernelTester()
.channel_tile(1)
.channels(channels)
.rows(rows)
.Test(xnn_f32_vmulcaddc_ukernel_c1__wasm_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
}
TEST(F32_VMULCADDC_C1__WASM_2X, rows_gt_2) {
for (size_t rows = 3; rows < 4; rows++) {
for (size_t channels = 1; channels <= 5; channels += 1) {
VMulCAddCMicrokernelTester()
.channel_tile(1)
.channels(channels)
.rows(rows)
.Test(xnn_f32_vmulcaddc_ukernel_c1__wasm_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
}
TEST(F32_VMULCADDC_C1__WASM_2X, input_stride) {
for (size_t rows = 1; rows <= 6; rows += 1) {
for (size_t channels = 1; channels <= 5; channels += 1) {
VMulCAddCMicrokernelTester()
.channel_tile(1)
.channels(channels)
.rows(rows)
.input_stride(7)
.Test(xnn_f32_vmulcaddc_ukernel_c1__wasm_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
}
TEST(F32_VMULCADDC_C1__WASM_2X, output_stride) {
for (size_t rows = 1; rows <= 6; rows += 1) {
for (size_t channels = 1; channels <= 5; channels += 1) {
VMulCAddCMicrokernelTester()
.channel_tile(1)
.channels(channels)
.rows(rows)
.output_stride(7)
.Test(xnn_f32_vmulcaddc_ukernel_c1__wasm_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
}
TEST(F32_VMULCADDC_C1__WASM_2X, inplace) {
for (size_t rows = 1; rows <= 6; rows += 1) {
for (size_t channels = 1; channels <= 5; channels += 1) {
VMulCAddCMicrokernelTester()
.channel_tile(1)
.channels(channels)
.rows(rows)
.inplace(true)
.Test(xnn_f32_vmulcaddc_ukernel_c1__wasm_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
}
TEST(F32_VMULCADDC_C1__WASM_2X, qmin) {
for (size_t rows = 1; rows <= 6; rows += 1) {
for (size_t channels = 1; channels <= 5; channels += 1) {
VMulCAddCMicrokernelTester()
.channel_tile(1)
.channels(channels)
.rows(rows)
.qmin(128)
.Test(xnn_f32_vmulcaddc_ukernel_c1__wasm_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
}
TEST(F32_VMULCADDC_C1__WASM_2X, qmax) {
for (size_t rows = 1; rows <= 6; rows += 1) {
for (size_t channels = 1; channels <= 5; channels += 1) {
VMulCAddCMicrokernelTester()
.channel_tile(1)
.channels(channels)
.rows(rows)
.qmax(128)
.Test(xnn_f32_vmulcaddc_ukernel_c1__wasm_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
}
#endif // XNN_ARCH_WASM
#if XNN_ARCH_WASM
TEST(F32_VMULCADDC_C2__WASM_2X, channels_eq_2) {
VMulCAddCMicrokernelTester()
.channel_tile(2)
.channels(2)
.rows(2)
.Test(xnn_f32_vmulcaddc_ukernel_c2__wasm_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
TEST(F32_VMULCADDC_C2__WASM_2X, channels_div_2) {
for (size_t channels = 4; channels < 20; channels += 2) {
VMulCAddCMicrokernelTester()
.channel_tile(2)
.channels(channels)
.rows(2)
.Test(xnn_f32_vmulcaddc_ukernel_c2__wasm_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
TEST(F32_VMULCADDC_C2__WASM_2X, channels_lt_2) {
for (size_t channels = 1; channels < 2; channels++) {
VMulCAddCMicrokernelTester()
.channel_tile(2)
.channels(channels)
.rows(2)
.Test(xnn_f32_vmulcaddc_ukernel_c2__wasm_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
TEST(F32_VMULCADDC_C2__WASM_2X, channels_gt_2) {
for (size_t channels = 3; channels < 4; channels++) {
VMulCAddCMicrokernelTester()
.channel_tile(2)
.channels(channels)
.rows(2)
.Test(xnn_f32_vmulcaddc_ukernel_c2__wasm_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
TEST(F32_VMULCADDC_C2__WASM_2X, rows_lt_2) {
for (size_t rows = 1; rows < 2; rows++) {
for (size_t channels = 1; channels <= 10; channels += 1) {
VMulCAddCMicrokernelTester()
.channel_tile(2)
.channels(channels)
.rows(rows)
.Test(xnn_f32_vmulcaddc_ukernel_c2__wasm_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
}
TEST(F32_VMULCADDC_C2__WASM_2X, rows_div_2) {
for (size_t rows = 4; rows <= 8; rows += 2) {
for (size_t channels = 1; channels <= 10; channels += 1) {
VMulCAddCMicrokernelTester()
.channel_tile(2)
.channels(channels)
.rows(rows)
.Test(xnn_f32_vmulcaddc_ukernel_c2__wasm_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
}
TEST(F32_VMULCADDC_C2__WASM_2X, rows_gt_2) {
for (size_t rows = 3; rows < 4; rows++) {
for (size_t channels = 1; channels <= 10; channels += 1) {
VMulCAddCMicrokernelTester()
.channel_tile(2)
.channels(channels)
.rows(rows)
.Test(xnn_f32_vmulcaddc_ukernel_c2__wasm_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
}
TEST(F32_VMULCADDC_C2__WASM_2X, input_stride) {
for (size_t rows = 1; rows <= 6; rows += 1) {
for (size_t channels = 1; channels <= 10; channels += 1) {
VMulCAddCMicrokernelTester()
.channel_tile(2)
.channels(channels)
.rows(rows)
.input_stride(13)
.Test(xnn_f32_vmulcaddc_ukernel_c2__wasm_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
}
TEST(F32_VMULCADDC_C2__WASM_2X, output_stride) {
for (size_t rows = 1; rows <= 6; rows += 1) {
for (size_t channels = 1; channels <= 10; channels += 1) {
VMulCAddCMicrokernelTester()
.channel_tile(2)
.channels(channels)
.rows(rows)
.output_stride(13)
.Test(xnn_f32_vmulcaddc_ukernel_c2__wasm_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
}
TEST(F32_VMULCADDC_C2__WASM_2X, inplace) {
for (size_t rows = 1; rows <= 6; rows += 1) {
for (size_t channels = 1; channels <= 10; channels += 1) {
VMulCAddCMicrokernelTester()
.channel_tile(2)
.channels(channels)
.rows(rows)
.inplace(true)
.Test(xnn_f32_vmulcaddc_ukernel_c2__wasm_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
}
TEST(F32_VMULCADDC_C2__WASM_2X, qmin) {
for (size_t rows = 1; rows <= 6; rows += 1) {
for (size_t channels = 1; channels <= 10; channels += 1) {
VMulCAddCMicrokernelTester()
.channel_tile(2)
.channels(channels)
.rows(rows)
.qmin(128)
.Test(xnn_f32_vmulcaddc_ukernel_c2__wasm_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
}
TEST(F32_VMULCADDC_C2__WASM_2X, qmax) {
for (size_t rows = 1; rows <= 6; rows += 1) {
for (size_t channels = 1; channels <= 10; channels += 1) {
VMulCAddCMicrokernelTester()
.channel_tile(2)
.channels(channels)
.rows(rows)
.qmax(128)
.Test(xnn_f32_vmulcaddc_ukernel_c2__wasm_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
}
#endif // XNN_ARCH_WASM
#if XNN_ARCH_WASM
TEST(F32_VMULCADDC_C4__WASM_2X, channels_eq_4) {
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(4)
.rows(2)
.Test(xnn_f32_vmulcaddc_ukernel_c4__wasm_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
TEST(F32_VMULCADDC_C4__WASM_2X, channels_div_4) {
for (size_t channels = 8; channels < 40; channels += 4) {
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(channels)
.rows(2)
.Test(xnn_f32_vmulcaddc_ukernel_c4__wasm_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
TEST(F32_VMULCADDC_C4__WASM_2X, channels_lt_4) {
for (size_t channels = 1; channels < 4; channels++) {
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(channels)
.rows(2)
.Test(xnn_f32_vmulcaddc_ukernel_c4__wasm_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
TEST(F32_VMULCADDC_C4__WASM_2X, channels_gt_4) {
for (size_t channels = 5; channels < 8; channels++) {
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(channels)
.rows(2)
.Test(xnn_f32_vmulcaddc_ukernel_c4__wasm_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
TEST(F32_VMULCADDC_C4__WASM_2X, rows_lt_2) {
for (size_t rows = 1; rows < 2; rows++) {
for (size_t channels = 1; channels <= 20; channels += 3) {
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(channels)
.rows(rows)
.Test(xnn_f32_vmulcaddc_ukernel_c4__wasm_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
}
TEST(F32_VMULCADDC_C4__WASM_2X, rows_div_2) {
for (size_t rows = 4; rows <= 8; rows += 2) {
for (size_t channels = 1; channels <= 20; channels += 3) {
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(channels)
.rows(rows)
.Test(xnn_f32_vmulcaddc_ukernel_c4__wasm_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
}
TEST(F32_VMULCADDC_C4__WASM_2X, rows_gt_2) {
for (size_t rows = 3; rows < 4; rows++) {
for (size_t channels = 1; channels <= 20; channels += 3) {
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(channels)
.rows(rows)
.Test(xnn_f32_vmulcaddc_ukernel_c4__wasm_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
}
TEST(F32_VMULCADDC_C4__WASM_2X, input_stride) {
for (size_t rows = 1; rows <= 6; rows += 1) {
for (size_t channels = 1; channels <= 20; channels += 3) {
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(channels)
.rows(rows)
.input_stride(23)
.Test(xnn_f32_vmulcaddc_ukernel_c4__wasm_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
}
TEST(F32_VMULCADDC_C4__WASM_2X, output_stride) {
for (size_t rows = 1; rows <= 6; rows += 1) {
for (size_t channels = 1; channels <= 20; channels += 3) {
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(channels)
.rows(rows)
.output_stride(23)
.Test(xnn_f32_vmulcaddc_ukernel_c4__wasm_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
}
TEST(F32_VMULCADDC_C4__WASM_2X, inplace) {
for (size_t rows = 1; rows <= 6; rows += 1) {
for (size_t channels = 1; channels <= 20; channels += 3) {
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(channels)
.rows(rows)
.inplace(true)
.Test(xnn_f32_vmulcaddc_ukernel_c4__wasm_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
}
TEST(F32_VMULCADDC_C4__WASM_2X, qmin) {
for (size_t rows = 1; rows <= 6; rows += 1) {
for (size_t channels = 1; channels <= 20; channels += 3) {
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(channels)
.rows(rows)
.qmin(128)
.Test(xnn_f32_vmulcaddc_ukernel_c4__wasm_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
}
TEST(F32_VMULCADDC_C4__WASM_2X, qmax) {
for (size_t rows = 1; rows <= 6; rows += 1) {
for (size_t channels = 1; channels <= 20; channels += 3) {
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(channels)
.rows(rows)
.qmax(128)
.Test(xnn_f32_vmulcaddc_ukernel_c4__wasm_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
}
#endif // XNN_ARCH_WASM
TEST(F32_VMULCADDC_C1__SCALAR_2X, channels_eq_1) {
VMulCAddCMicrokernelTester()
.channel_tile(1)
.channels(1)
.rows(2)
.Test(xnn_f32_vmulcaddc_ukernel_c1__scalar_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
TEST(F32_VMULCADDC_C1__SCALAR_2X, channels_gt_1) {
for (size_t channels = 2; channels < 10; channels++) {
VMulCAddCMicrokernelTester()
.channel_tile(1)
.channels(channels)
.rows(2)
.Test(xnn_f32_vmulcaddc_ukernel_c1__scalar_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
TEST(F32_VMULCADDC_C1__SCALAR_2X, rows_lt_2) {
for (size_t rows = 1; rows < 2; rows++) {
for (size_t channels = 1; channels <= 5; channels += 1) {
VMulCAddCMicrokernelTester()
.channel_tile(1)
.channels(channels)
.rows(rows)
.Test(xnn_f32_vmulcaddc_ukernel_c1__scalar_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
}
TEST(F32_VMULCADDC_C1__SCALAR_2X, rows_div_2) {
for (size_t rows = 4; rows <= 8; rows += 2) {
for (size_t channels = 1; channels <= 5; channels += 1) {
VMulCAddCMicrokernelTester()
.channel_tile(1)
.channels(channels)
.rows(rows)
.Test(xnn_f32_vmulcaddc_ukernel_c1__scalar_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
}
TEST(F32_VMULCADDC_C1__SCALAR_2X, rows_gt_2) {
for (size_t rows = 3; rows < 4; rows++) {
for (size_t channels = 1; channels <= 5; channels += 1) {
VMulCAddCMicrokernelTester()
.channel_tile(1)
.channels(channels)
.rows(rows)
.Test(xnn_f32_vmulcaddc_ukernel_c1__scalar_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
}
TEST(F32_VMULCADDC_C1__SCALAR_2X, input_stride) {
for (size_t rows = 1; rows <= 6; rows += 1) {
for (size_t channels = 1; channels <= 5; channels += 1) {
VMulCAddCMicrokernelTester()
.channel_tile(1)
.channels(channels)
.rows(rows)
.input_stride(7)
.Test(xnn_f32_vmulcaddc_ukernel_c1__scalar_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
}
TEST(F32_VMULCADDC_C1__SCALAR_2X, output_stride) {
for (size_t rows = 1; rows <= 6; rows += 1) {
for (size_t channels = 1; channels <= 5; channels += 1) {
VMulCAddCMicrokernelTester()
.channel_tile(1)
.channels(channels)
.rows(rows)
.output_stride(7)
.Test(xnn_f32_vmulcaddc_ukernel_c1__scalar_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
}
TEST(F32_VMULCADDC_C1__SCALAR_2X, inplace) {
for (size_t rows = 1; rows <= 6; rows += 1) {
for (size_t channels = 1; channels <= 5; channels += 1) {
VMulCAddCMicrokernelTester()
.channel_tile(1)
.channels(channels)
.rows(rows)
.inplace(true)
.Test(xnn_f32_vmulcaddc_ukernel_c1__scalar_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
}
TEST(F32_VMULCADDC_C1__SCALAR_2X, qmin) {
for (size_t rows = 1; rows <= 6; rows += 1) {
for (size_t channels = 1; channels <= 5; channels += 1) {
VMulCAddCMicrokernelTester()
.channel_tile(1)
.channels(channels)
.rows(rows)
.qmin(128)
.Test(xnn_f32_vmulcaddc_ukernel_c1__scalar_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
}
TEST(F32_VMULCADDC_C1__SCALAR_2X, qmax) {
for (size_t rows = 1; rows <= 6; rows += 1) {
for (size_t channels = 1; channels <= 5; channels += 1) {
VMulCAddCMicrokernelTester()
.channel_tile(1)
.channels(channels)
.rows(rows)
.qmax(128)
.Test(xnn_f32_vmulcaddc_ukernel_c1__scalar_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
}
TEST(F32_VMULCADDC_C2__SCALAR_2X, channels_eq_2) {
VMulCAddCMicrokernelTester()
.channel_tile(2)
.channels(2)
.rows(2)
.Test(xnn_f32_vmulcaddc_ukernel_c2__scalar_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
TEST(F32_VMULCADDC_C2__SCALAR_2X, channels_div_2) {
for (size_t channels = 4; channels < 20; channels += 2) {
VMulCAddCMicrokernelTester()
.channel_tile(2)
.channels(channels)
.rows(2)
.Test(xnn_f32_vmulcaddc_ukernel_c2__scalar_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
TEST(F32_VMULCADDC_C2__SCALAR_2X, channels_lt_2) {
for (size_t channels = 1; channels < 2; channels++) {
VMulCAddCMicrokernelTester()
.channel_tile(2)
.channels(channels)
.rows(2)
.Test(xnn_f32_vmulcaddc_ukernel_c2__scalar_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
TEST(F32_VMULCADDC_C2__SCALAR_2X, channels_gt_2) {
for (size_t channels = 3; channels < 4; channels++) {
VMulCAddCMicrokernelTester()
.channel_tile(2)
.channels(channels)
.rows(2)
.Test(xnn_f32_vmulcaddc_ukernel_c2__scalar_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
TEST(F32_VMULCADDC_C2__SCALAR_2X, rows_lt_2) {
for (size_t rows = 1; rows < 2; rows++) {
for (size_t channels = 1; channels <= 10; channels += 1) {
VMulCAddCMicrokernelTester()
.channel_tile(2)
.channels(channels)
.rows(rows)
.Test(xnn_f32_vmulcaddc_ukernel_c2__scalar_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
}
TEST(F32_VMULCADDC_C2__SCALAR_2X, rows_div_2) {
for (size_t rows = 4; rows <= 8; rows += 2) {
for (size_t channels = 1; channels <= 10; channels += 1) {
VMulCAddCMicrokernelTester()
.channel_tile(2)
.channels(channels)
.rows(rows)
.Test(xnn_f32_vmulcaddc_ukernel_c2__scalar_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
}
TEST(F32_VMULCADDC_C2__SCALAR_2X, rows_gt_2) {
for (size_t rows = 3; rows < 4; rows++) {
for (size_t channels = 1; channels <= 10; channels += 1) {
VMulCAddCMicrokernelTester()
.channel_tile(2)
.channels(channels)
.rows(rows)
.Test(xnn_f32_vmulcaddc_ukernel_c2__scalar_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
}
TEST(F32_VMULCADDC_C2__SCALAR_2X, input_stride) {
for (size_t rows = 1; rows <= 6; rows += 1) {
for (size_t channels = 1; channels <= 10; channels += 1) {
VMulCAddCMicrokernelTester()
.channel_tile(2)
.channels(channels)
.rows(rows)
.input_stride(13)
.Test(xnn_f32_vmulcaddc_ukernel_c2__scalar_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
}
TEST(F32_VMULCADDC_C2__SCALAR_2X, output_stride) {
for (size_t rows = 1; rows <= 6; rows += 1) {
for (size_t channels = 1; channels <= 10; channels += 1) {
VMulCAddCMicrokernelTester()
.channel_tile(2)
.channels(channels)
.rows(rows)
.output_stride(13)
.Test(xnn_f32_vmulcaddc_ukernel_c2__scalar_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
}
TEST(F32_VMULCADDC_C2__SCALAR_2X, inplace) {
for (size_t rows = 1; rows <= 6; rows += 1) {
for (size_t channels = 1; channels <= 10; channels += 1) {
VMulCAddCMicrokernelTester()
.channel_tile(2)
.channels(channels)
.rows(rows)
.inplace(true)
.Test(xnn_f32_vmulcaddc_ukernel_c2__scalar_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
}
TEST(F32_VMULCADDC_C2__SCALAR_2X, qmin) {
for (size_t rows = 1; rows <= 6; rows += 1) {
for (size_t channels = 1; channels <= 10; channels += 1) {
VMulCAddCMicrokernelTester()
.channel_tile(2)
.channels(channels)
.rows(rows)
.qmin(128)
.Test(xnn_f32_vmulcaddc_ukernel_c2__scalar_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
}
TEST(F32_VMULCADDC_C2__SCALAR_2X, qmax) {
for (size_t rows = 1; rows <= 6; rows += 1) {
for (size_t channels = 1; channels <= 10; channels += 1) {
VMulCAddCMicrokernelTester()
.channel_tile(2)
.channels(channels)
.rows(rows)
.qmax(128)
.Test(xnn_f32_vmulcaddc_ukernel_c2__scalar_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
}
TEST(F32_VMULCADDC_C4__SCALAR_2X, channels_eq_4) {
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(4)
.rows(2)
.Test(xnn_f32_vmulcaddc_ukernel_c4__scalar_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
TEST(F32_VMULCADDC_C4__SCALAR_2X, channels_div_4) {
for (size_t channels = 8; channels < 40; channels += 4) {
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(channels)
.rows(2)
.Test(xnn_f32_vmulcaddc_ukernel_c4__scalar_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
TEST(F32_VMULCADDC_C4__SCALAR_2X, channels_lt_4) {
for (size_t channels = 1; channels < 4; channels++) {
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(channels)
.rows(2)
.Test(xnn_f32_vmulcaddc_ukernel_c4__scalar_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
TEST(F32_VMULCADDC_C4__SCALAR_2X, channels_gt_4) {
for (size_t channels = 5; channels < 8; channels++) {
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(channels)
.rows(2)
.Test(xnn_f32_vmulcaddc_ukernel_c4__scalar_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
TEST(F32_VMULCADDC_C4__SCALAR_2X, rows_lt_2) {
for (size_t rows = 1; rows < 2; rows++) {
for (size_t channels = 1; channels <= 20; channels += 3) {
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(channels)
.rows(rows)
.Test(xnn_f32_vmulcaddc_ukernel_c4__scalar_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
}
TEST(F32_VMULCADDC_C4__SCALAR_2X, rows_div_2) {
for (size_t rows = 4; rows <= 8; rows += 2) {
for (size_t channels = 1; channels <= 20; channels += 3) {
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(channels)
.rows(rows)
.Test(xnn_f32_vmulcaddc_ukernel_c4__scalar_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
}
TEST(F32_VMULCADDC_C4__SCALAR_2X, rows_gt_2) {
for (size_t rows = 3; rows < 4; rows++) {
for (size_t channels = 1; channels <= 20; channels += 3) {
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(channels)
.rows(rows)
.Test(xnn_f32_vmulcaddc_ukernel_c4__scalar_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
}
TEST(F32_VMULCADDC_C4__SCALAR_2X, input_stride) {
for (size_t rows = 1; rows <= 6; rows += 1) {
for (size_t channels = 1; channels <= 20; channels += 3) {
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(channels)
.rows(rows)
.input_stride(23)
.Test(xnn_f32_vmulcaddc_ukernel_c4__scalar_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
}
TEST(F32_VMULCADDC_C4__SCALAR_2X, output_stride) {
for (size_t rows = 1; rows <= 6; rows += 1) {
for (size_t channels = 1; channels <= 20; channels += 3) {
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(channels)
.rows(rows)
.output_stride(23)
.Test(xnn_f32_vmulcaddc_ukernel_c4__scalar_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
}
TEST(F32_VMULCADDC_C4__SCALAR_2X, inplace) {
for (size_t rows = 1; rows <= 6; rows += 1) {
for (size_t channels = 1; channels <= 20; channels += 3) {
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(channels)
.rows(rows)
.inplace(true)
.Test(xnn_f32_vmulcaddc_ukernel_c4__scalar_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
}
TEST(F32_VMULCADDC_C4__SCALAR_2X, qmin) {
for (size_t rows = 1; rows <= 6; rows += 1) {
for (size_t channels = 1; channels <= 20; channels += 3) {
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(channels)
.rows(rows)
.qmin(128)
.Test(xnn_f32_vmulcaddc_ukernel_c4__scalar_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
}
TEST(F32_VMULCADDC_C4__SCALAR_2X, qmax) {
for (size_t rows = 1; rows <= 6; rows += 1) {
for (size_t channels = 1; channels <= 20; channels += 3) {
VMulCAddCMicrokernelTester()
.channel_tile(4)
.channels(channels)
.rows(rows)
.qmax(128)
.Test(xnn_f32_vmulcaddc_ukernel_c4__scalar_2x, VMulCAddCMicrokernelTester::Variant::Scalar);
}
}
}