| /openbsd-src/gnu/llvm/llvm/lib/Target/AMDGPU/ |
| H A D | R600Subtarget.cpp | 25 R600Subtarget::R600Subtarget(const Triple &TT, StringRef GPU, StringRef FS, in R600Subtarget() argument 27 : R600GenSubtargetInfo(TT, GPU, /*TuneCPU*/ GPU, FS), AMDGPUSubtarget(TT), in R600Subtarget() 30 TLInfo(TM, initializeSubtargetDependencies(TT, GPU, FS)), in R600Subtarget() 31 InstrItins(getInstrItineraryForCPU(GPU)) { in R600Subtarget() 36 StringRef GPU, in initializeSubtargetDependencies() argument 40 ParseSubtargetFeatures(GPU, /*TuneCPU*/ GPU, FullFS); in initializeSubtargetDependencies()
|
| H A D | R600TargetMachine.cpp | 68 StringRef GPU = getGPUName(F); in getSubtargetImpl() local 71 SmallString<128> SubtargetKey(GPU); in getSubtargetImpl() 80 I = std::make_unique<R600Subtarget>(TargetTriple, GPU, FS, *this); in getSubtargetImpl()
|
| H A D | AMDGPUSubtarget.cpp | 65 StringRef GPU, StringRef FS) { in initializeSubtargetDependencies() argument 95 ParseSubtargetFeatures(GPU, /*TuneCPU*/ GPU, FullFS); in initializeSubtargetDependencies() 169 GCNSubtarget::GCNSubtarget(const Triple &TT, StringRef GPU, StringRef FS, in GCNSubtarget() argument 172 AMDGPUGenSubtargetInfo(TT, GPU, /*TuneCPU*/ GPU, FS), in GCNSubtarget() 176 InstrItins(getInstrItineraryForCPU(GPU)), in GCNSubtarget() 177 InstrInfo(initializeSubtargetDependencies(TT, GPU, FS)), in GCNSubtarget()
|
| H A D | AMDGPUFeatures.td | 46 Value#" GPU generation", Implies>;
|
| H A D | R600Subtarget.h | 82 StringRef GPU, StringRef FS);
|
| H A D | AMDGPUTargetMachine.cpp | 516 static StringRef getGPUOrDefault(const Triple &TT, StringRef GPU) { in getGPUOrDefault() argument 517 if (!GPU.empty()) in getGPUOrDefault() 518 return GPU; in getGPUOrDefault() 817 StringRef GPU = getGPUName(F); in getSubtargetImpl() local 820 SmallString<128> SubtargetKey(GPU); in getSubtargetImpl() 829 I = std::make_unique<GCNSubtarget>(TargetTriple, GPU, FS, *this); in getSubtargetImpl()
|
| /openbsd-src/gnu/llvm/clang/lib/Basic/Targets/ |
| H A D | NVPTX.h | 62 CudaArch GPU; variable 79 Features[CudaArchToString(GPU)] = true; in initFeatureMap() 130 GPU = StringToCudaArch(Name); in setCPU() 131 return GPU != CudaArch::UNKNOWN; in setCPU()
|
| H A D | NVPTX.cpp | 62 GPU = CudaArch::SM_20; in NVPTXTargetInfo() 172 switch (GPU) { in getTargetDefines()
|
| /openbsd-src/gnu/llvm/llvm/lib/TargetParser/ |
| H A D | TargetParser.cpp | 193 AMDGPU::IsaVersion AMDGPU::getIsaVersion(StringRef GPU) { in getIsaVersion() argument 194 AMDGPU::GPUKind AK = parseArchAMDGCN(GPU); in getIsaVersion() 196 if (GPU == "generic-hsa") in getIsaVersion() 198 if (GPU == "generic") in getIsaVersion()
|
| /openbsd-src/gnu/llvm/clang/tools/amdgpu-arch/ |
| H A D | AMDGPUArch.cpp | 123 for (const auto &GPU : GPUs) in main() local 124 printf("%s\n", GPU.c_str()); in main()
|
| /openbsd-src/gnu/llvm/llvm/docs/ |
| H A D | UserGuides.rst | 240 This document describes using the NVPTX backend to compile GPU kernels. 243 This document describes using the AMDGPU backend to compile GPU kernels. 255 This document describes using the SPIR-V target to compile GPU kernels. 258 This document describes using the DirectX target to compile GPU code for the
|
| H A D | CompileCudaWithLLVM.rst | 58 $ clang++ axpy.cu -o axpy --cuda-gpu-arch=<GPU arch> \ 81 * ``<GPU arch>`` -- the `compute capability 82 <https://developer.nvidia.com/cuda-gpus>`_ of your GPU. For example, if you 83 want to run your program on a GPU with compute capability of 3.5, specify 101 GPU hardware allows for more control over numerical operations than most CPUs, 242 * For each GPU architecture ``arch`` that we're compiling for, do: 248 ``S_arch``, containing GPU machine code (SASS) for ``arch``. 262 * For each GPU architecture ``arch`` that we're compiling for, do: 288 host compilation and during device compilation for each GPU architecture.) 503 on a CPU isn't necessarily fast on a GPU. We've made a number of changes to [all …]
|
| H A D | NVPTXUsage.rst | 13 To support GPU programming, the NVPTX back-end supports a subset of LLVM IR 14 along with a defined set of conventions used to represent GPU programming 369 The most common way to execute PTX assembly on a GPU device is to use the CUDA 370 Driver API. This API is a low-level interface to the GPU driver and allows for 371 JIT compilation of PTX code to native GPU machine code. 571 an explicit address space specifier. What is address space 1? NVIDIA GPU 594 program), or a `device` function (callable only from GPU code). You can think 595 of `kernel` functions as entry-points in the GPU program. To mark an LLVM IR 624 a real GPU device? The CUDA Driver API provides a convenient mechanism for 625 loading and JIT compiling PTX to a native GPU device, and launching a kernel. [all …]
|
| /openbsd-src/gnu/llvm/llvm/docs/AMDGPU/ |
| H A D | gfx8_vdata_f2bf57.rst | 15 *Size:* depends on GFX8 GPU revision:
|
| H A D | gfx8_vdata_4f639e.rst | 15 *Size:* depends on GFX8 GPU revision:
|
| H A D | gfx8_vdata_886702.rst | 15 *Size:* depends on GFX8 GPU revision:
|
| H A D | gfx8_vdst_6f591e.rst | 15 *Size:* depends on GFX8 GPU revision:
|
| H A D | gfx8_vdst_c360a5.rst | 15 *Size:* depends on GFX8 GPU revision:
|
| H A D | gfx8_vdst_d809e2.rst | 15 *Size:* depends on GFX8 GPU revision:
|
| H A D | gfx8_vdst_de9309.rst | 15 *Size:* depends on GFX8 GPU revision and :ref:`tfe<amdgpu_synid_tfe>`:
|
| H A D | gfx8_vdst_2d89ba.rst | 15 *Size:* depends on GFX8 GPU revision and :ref:`tfe<amdgpu_synid_tfe>`:
|
| H A D | gfx8_vdst_b61114.rst | 15 *Size:* depends on GFX8 GPU revision and :ref:`tfe<amdgpu_synid_tfe>`:
|
| /openbsd-src/gnu/llvm/clang/docs/HLSL/ |
| H A D | HLSLSupport.rst | 114 with other GPU graphics targets like SPIR-V and possibly other GPU and even CPU 131 An Aside on GPU Languages 134 Due to HLSL being a GPU targeted language HLSL is a Single Program Multiple Data 135 (SPMD) language relying on the implicit parallelism provided by GPU hardware. 140 performance or are not widely supportable on GPU hardware or drivers. As an
|
| /openbsd-src/gnu/llvm/clang/lib/Headers/cuda_wrappers/ |
| H A D | cmath | 33 // headers to parse, but would not allow accidental use of them on a GPU. 71 // which we can't handle on GPU. We need to forward those to CUDA-provided
|
| /openbsd-src/gnu/llvm/llvm/include/llvm/TargetParser/ |
| H A D | TargetParser.h | 154 IsaVersion getIsaVersion(StringRef GPU);
|