| // Copyright 2022 The Dawn & Tint Authors | |
| // | |
| // Redistribution and use in source and binary forms, with or without | |
| // modification, are permitted provided that the following conditions are met: | |
| // | |
| // 1. Redistributions of source code must retain the above copyright notice, this | |
| // list of conditions and the following disclaimer. | |
| // | |
| // 2. Redistributions in binary form must reproduce the above copyright notice, | |
| // this list of conditions and the following disclaimer in the documentation | |
| // and/or other materials provided with the distribution. | |
| // | |
| // 3. Neither the name of the copyright holder nor the names of its | |
| // contributors may be used to endorse or promote products derived from | |
| // this software without specific prior written permission. | |
| // | |
| // THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" | |
| // AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE | |
| // IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE | |
| // DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE | |
| // FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL | |
| // DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR | |
| // SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER | |
| // CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, | |
| // OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE | |
| // OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. | |
| #include <vector> | |
| #include "src/dawn/tests/DawnTest.h" | |
| #include "src/dawn/utils/WGPUHelpers.h" | |
| namespace dawn { | |
| namespace { | |
| class Packed4x8IntegerDotProductTests : public DawnTest {}; | |
| TEST_P(Packed4x8IntegerDotProductTests, Dot4x8Packed) { | |
| const char* computeShader = R"( | |
| struct Buf { | |
| data1 : i32, | |
| data2 : u32, | |
| data3 : i32, | |
| data4 : u32, | |
| } | |
| @group(0) @binding(0) var<storage, read_write> buf : Buf; | |
| struct InputData { | |
| a : u32, | |
| b : u32, | |
| c : u32, | |
| } | |
| @group(0) @binding(1) var<storage, read_write> inputBuf : InputData; | |
| @compute @workgroup_size(1) | |
| fn main() { | |
| let a = inputBuf.a; | |
| let b = inputBuf.b; | |
| let c = inputBuf.c; | |
| buf.data1 = dot4I8Packed(a, b); | |
| buf.data2 = dot4U8Packed(a, b); | |
| buf.data3 = dot4I8Packed(a, c); | |
| buf.data4 = dot4U8Packed(a, c); | |
| } | |
| )"; | |
| ASSERT_TRUE( | |
| instance.HasWGSLLanguageFeature(wgpu::WGSLLanguageFeatureName::Packed4x8IntegerDotProduct)); | |
| wgpu::BufferDescriptor bufferDesc; | |
| bufferDesc.size = 4 * sizeof(uint32_t); | |
| bufferDesc.usage = wgpu::BufferUsage::Storage | wgpu::BufferUsage::CopySrc; | |
| wgpu::Buffer bufferOut = device.CreateBuffer(&bufferDesc); | |
| wgpu::Buffer inputBuffer = utils::CreateBufferFromData( | |
| device, | |
| wgpu::BufferUsage::CopySrc | wgpu::BufferUsage::CopyDst | wgpu::BufferUsage::Storage, | |
| {0xFFFEFDFCu, 0xFBFAF9F8u, 0x01020304u}); | |
| wgpu::ComputePipelineDescriptor csDesc; | |
| csDesc.compute.module = utils::CreateShaderModule(device, computeShader); | |
| wgpu::ComputePipeline pipeline = device.CreateComputePipeline(&csDesc); | |
| wgpu::BindGroup bindGroup = utils::MakeBindGroup(device, pipeline.GetBindGroupLayout(0), | |
| { | |
| {0, bufferOut}, | |
| {1, inputBuffer}, | |
| }); | |
| wgpu::CommandEncoder encoder = device.CreateCommandEncoder(); | |
| wgpu::ComputePassEncoder pass = encoder.BeginComputePass(); | |
| pass.SetPipeline(pipeline); | |
| pass.SetBindGroup(0, bindGroup); | |
| pass.DispatchWorkgroups(1); | |
| pass.End(); | |
| wgpu::CommandBuffer commands = encoder.Finish(); | |
| queue.Submit(1, &commands); | |
| int32_t expected[] = {70, 252998, -30, 2530}; | |
| EXPECT_BUFFER_U32_RANGE_EQ(reinterpret_cast<uint32_t*>(expected), bufferOut, 0, | |
| sizeof(expected) / sizeof(int32_t)); | |
| } | |
| TEST_P(Packed4x8IntegerDotProductTests, Pack4x8) { | |
| const char* computeShader = R"( | |
| struct Buf { | |
| data1 : u32, | |
| data2 : u32, | |
| data3 : u32, | |
| data4 : u32, | |
| data5 : u32, | |
| data6 : u32, | |
| data7 : u32, | |
| data8 : u32, | |
| } | |
| @group(0) @binding(0) var<storage, read_write> buf : Buf; | |
| struct InputData { | |
| a : vec4i, | |
| b : vec4i, | |
| c : vec4u, | |
| d : vec4u, | |
| } | |
| @group(0) @binding(1) var<storage, read_write> inputBuf : InputData; | |
| @compute @workgroup_size(1) | |
| fn main() { | |
| buf.data1 = pack4xI8(inputBuf.a); | |
| buf.data2 = pack4xI8(inputBuf.b); | |
| buf.data3 = pack4xU8(inputBuf.c); | |
| buf.data4 = pack4xU8(inputBuf.d); | |
| buf.data5 = pack4xI8Clamp(inputBuf.a); | |
| buf.data6 = pack4xI8Clamp(inputBuf.b); | |
| buf.data7 = pack4xU8Clamp(inputBuf.c); | |
| buf.data8 = pack4xU8Clamp(inputBuf.d); | |
| } | |
| )"; | |
| ASSERT_TRUE( | |
| instance.HasWGSLLanguageFeature(wgpu::WGSLLanguageFeatureName::Packed4x8IntegerDotProduct)); | |
| wgpu::BufferDescriptor bufferDesc; | |
| bufferDesc.size = 8 * sizeof(uint32_t); | |
| bufferDesc.usage = wgpu::BufferUsage::Storage | wgpu::BufferUsage::CopySrc; | |
| wgpu::Buffer bufferOut = device.CreateBuffer(&bufferDesc); | |
| wgpu::Buffer inputBuffer = utils::CreateBufferFromData( | |
| device, | |
| wgpu::BufferUsage::CopySrc | wgpu::BufferUsage::CopyDst | wgpu::BufferUsage::Storage, | |
| {127, 128, -128, -129, -32768, -32769, 32767, 32768, 1, 2, 3, 4, 0, 254, 255, 65535}); | |
| wgpu::ComputePipelineDescriptor csDesc; | |
| csDesc.compute.module = utils::CreateShaderModule(device, computeShader); | |
| wgpu::ComputePipeline pipeline = device.CreateComputePipeline(&csDesc); | |
| wgpu::BindGroup bindGroup = utils::MakeBindGroup(device, pipeline.GetBindGroupLayout(0), | |
| { | |
| {0, bufferOut}, | |
| {1, inputBuffer}, | |
| }); | |
| wgpu::CommandEncoder encoder = device.CreateCommandEncoder(); | |
| wgpu::ComputePassEncoder pass = encoder.BeginComputePass(); | |
| pass.SetPipeline(pipeline); | |
| pass.SetBindGroup(0, bindGroup); | |
| pass.DispatchWorkgroups(1); | |
| pass.End(); | |
| wgpu::CommandBuffer commands = encoder.Finish(); | |
| queue.Submit(1, &commands); | |
| uint32_t expected[] = {0x7F80807F, 0x00FFFF00, 0x04030201, 0xFFFFFE00, | |
| 0x80807F7F, 0x7F7F8080, 0x04030201, 0xFFFFFE00}; | |
| EXPECT_BUFFER_U32_RANGE_EQ(expected, bufferOut, 0, sizeof(expected) / sizeof(uint32_t)); | |
| } | |
| TEST_P(Packed4x8IntegerDotProductTests, Unpack4x8) { | |
| const char* computeShader = R"( | |
| struct Buf { | |
| data1 : vec4i, | |
| data2 : vec4i, | |
| data3 : vec4i, | |
| data4 : vec4i, | |
| data5 : vec4u, | |
| data6 : vec4u, | |
| data7 : vec4u, | |
| data8 : vec4u, | |
| } | |
| @group(0) @binding(0) var<storage, read_write> buf : Buf; | |
| struct InputData { | |
| a : u32, | |
| b : u32, | |
| c : u32, | |
| d : u32, | |
| } | |
| @group(0) @binding(1) var<storage, read_write> inputBuf : InputData; | |
| @compute @workgroup_size(1) | |
| fn main() { | |
| buf.data1 = unpack4xI8(inputBuf.a); | |
| buf.data2 = unpack4xI8(inputBuf.b); | |
| buf.data3 = unpack4xI8(inputBuf.c); | |
| buf.data4 = unpack4xI8(inputBuf.d); | |
| buf.data5 = unpack4xU8(inputBuf.a); | |
| buf.data6 = unpack4xU8(inputBuf.b); | |
| buf.data7 = unpack4xU8(inputBuf.c); | |
| buf.data8 = unpack4xU8(inputBuf.d); | |
| } | |
| )"; | |
| ASSERT_TRUE( | |
| instance.HasWGSLLanguageFeature(wgpu::WGSLLanguageFeatureName::Packed4x8IntegerDotProduct)); | |
| wgpu::BufferDescriptor bufferDesc; | |
| bufferDesc.size = sizeof(uint32_t) * 4 * 8; | |
| bufferDesc.usage = wgpu::BufferUsage::Storage | wgpu::BufferUsage::CopySrc; | |
| wgpu::Buffer bufferOut = device.CreateBuffer(&bufferDesc); | |
| wgpu::Buffer inputBuffer = utils::CreateBufferFromData( | |
| device, | |
| wgpu::BufferUsage::CopySrc | wgpu::BufferUsage::CopyDst | wgpu::BufferUsage::Storage, | |
| {0x01020304u, 0xFFFEFDFCu, 0x05FB06FAu, 0xF907F808u}); | |
| wgpu::ComputePipelineDescriptor csDesc; | |
| csDesc.compute.module = utils::CreateShaderModule(device, computeShader); | |
| wgpu::ComputePipeline pipeline = device.CreateComputePipeline(&csDesc); | |
| wgpu::BindGroup bindGroup = utils::MakeBindGroup(device, pipeline.GetBindGroupLayout(0), | |
| { | |
| {0, bufferOut}, | |
| {1, inputBuffer}, | |
| }); | |
| wgpu::CommandEncoder encoder = device.CreateCommandEncoder(); | |
| wgpu::ComputePassEncoder pass = encoder.BeginComputePass(); | |
| pass.SetPipeline(pipeline); | |
| pass.SetBindGroup(0, bindGroup); | |
| pass.DispatchWorkgroups(1); | |
| pass.End(); | |
| wgpu::CommandBuffer commands = encoder.Finish(); | |
| queue.Submit(1, &commands); | |
| int32_t expected[] = {4, 3, 2, 1, -4, -3, -2, -1, -6, 6, -5, 5, 8, -8, 7, -7, | |
| 4, 3, 2, 1, 252, 253, 254, 255, 250, 6, 251, 5, 8, 248, 7, 249}; | |
| EXPECT_BUFFER_U32_RANGE_EQ(reinterpret_cast<uint32_t*>(expected), bufferOut, 0, | |
| sizeof(expected) / sizeof(int32_t)); | |
| } | |
| // DawnTestBase::CreateDeviceImpl always enables allow_unsafe_apis toggle. | |
| DAWN_INSTANTIATE_TEST(Packed4x8IntegerDotProductTests, | |
| D3D11Backend(), | |
| D3D12Backend(), | |
| D3D12Backend({}, {"use_dxc"}), | |
| D3D12Backend({"polyfill_packed_4x8_dot_product"}), | |
| D3D12Backend({"d3d12_polyfill_pack_unpack_4x8"}), | |
| MetalBackend(), | |
| VulkanBackend(), | |
| VulkanBackend({"polyfill_packed_4x8_dot_product"}), | |
| WebGPUBackend()); | |
| } // anonymous namespace | |
| } // namespace dawn |