DeviceBuffer<RVec> d_rVecInput;
allocateDeviceBuffer(&d_rVecInput, numElements, deviceContext);
- copyToDeviceBuffer(&d_rVecInput, h_rVecInput.data(), 0, numElements, deviceStream,
- GpuApiCallBehavior::Sync, nullptr);
+ copyToDeviceBuffer(
+ &d_rVecInput, h_rVecInput.data(), 0, numElements, deviceStream, GpuApiCallBehavior::Sync, nullptr);
DeviceBuffer<float3> d_float3Output;
allocateDeviceBuffer(&d_float3Output, numElements * DIM, deviceContext);
kernelLaunchConfig.sharedMemorySize = 0;
auto kernelPtr = convertRVecToFloat3OnDevice_kernel;
- const auto kernelArgs = prepareGpuKernelArguments(kernelPtr, kernelLaunchConfig,
- &d_float3Output, &d_rVecInput, &numElements);
- launchGpuKernel(kernelPtr, kernelLaunchConfig, deviceStream, nullptr,
- "convertRVecToFloat3OnDevice_kernel", kernelArgs);
-
- copyFromDeviceBuffer(h_float3Output.data(), &d_float3Output, 0, numElements, deviceStream,
- GpuApiCallBehavior::Sync, nullptr);
+ const auto kernelArgs = prepareGpuKernelArguments(
+ kernelPtr, kernelLaunchConfig, &d_float3Output, &d_rVecInput, &numElements);
+ launchGpuKernel(kernelPtr,
+ kernelLaunchConfig,
+ deviceStream,
+ nullptr,
+ "convertRVecToFloat3OnDevice_kernel",
+ kernelArgs);
+
+ copyFromDeviceBuffer(
+ h_float3Output.data(), &d_float3Output, 0, numElements, deviceStream, GpuApiCallBehavior::Sync, nullptr);
saveFloat3InRVecFormat(h_rVecOutput, h_float3Output.data(), numElements);