Searched refs:WarpSize (Results 1 – 6 of 6) sorted by relevance
| /external/llvm-project/openmp/libomptarget/deviceRTLs/nvptx/test/parallel/ |
| D | thread_limit.c | 6 const int WarpSize = 32; variable
7 const int ThreadLimit = 1 * WarpSize;
8 const int NumThreads2 = 2 * WarpSize;
9 const int NumThreads3 = 3 * WarpSize;
|
| D | num_threads.c | 6 const int WarpSize = 32; variable
7 const int NumThreads1 = 1 * WarpSize;
8 const int NumThreads2 = 2 * WarpSize;
9 const int NumThreads3 = 3 * WarpSize;
|
| /external/llvm-project/clang/lib/CodeGen/ |
| D | CGOpenMPRuntimeAMDGCN.cpp | 38 unsigned WarpSize = CGF.getTarget().getGridValue(llvm::omp::GV_Warp_Size); in getGPUWarpSize() local
39 return Bld.getInt32(WarpSize); in getGPUWarpSize()
|
| D | CGOpenMPRuntimeGPU.cpp | 342 unsigned WarpSize = CGF.getTarget().getGridValue(llvm::omp::GV_Warp_Size); in buildRecordForGlobalizedVars() local
349 MappedDeclsFields, WarpSize); in buildRecordForGlobalizedVars()
1642 unsigned WarpSize = CGM.getTarget().getGridValue(llvm::omp::GV_Warp_Size); in emitTeamsOutlinedFunction() local
1651 MappedDeclsFields, WarpSize); in emitTeamsOutlinedFunction()
2486 llvm::Value *WarpSize = in createRuntimeShuffleFunction() local
2491 {ElemCast, Offset, WarpSize}); in createRuntimeShuffleFunction()
2853 unsigned WarpSize = CGF.getTarget().getGridValue(llvm::omp::GV_Warp_Size); in emitInterWarpCopyFunction() local
2855 auto *Ty = llvm::ArrayType::get(CGM.Int32Ty, WarpSize); in emitInterWarpCopyFunction()
|
| /external/llvm-project/openmp/libomptarget/plugins/cuda/src/ |
| D | rtl.cpp | 125 int WarpSize = 0; member
486 int WarpSize; in initDevice() local
488 cuDeviceGetAttribute(&WarpSize, CU_DEVICE_ATTRIBUTE_WARP_SIZE, Device); in initDevice()
491 DeviceData[DeviceId].WarpSize = 32; in initDevice()
493 DP("Using warp size %d\n", WarpSize); in initDevice()
494 DeviceData[DeviceId].WarpSize = WarpSize; in initDevice()
508 DeviceData[DeviceId].ThreadsPerBlock, DeviceData[DeviceId].WarpSize); in initDevice()
871 DP("Adding master warp: +%d threads\n", DeviceData[DeviceId].WarpSize); in runTargetTeamRegion()
872 CudaThreadsPerBlock += DeviceData[DeviceId].WarpSize; in runTargetTeamRegion()
|
| /external/llvm-project/openmp/libomptarget/plugins/amdgpu/src/ |
| D | rtl.cpp | 321 std::vector<int> WarpSize; member in RTLDeviceInfoTy
476 WarpSize.resize(NumberOfDevices); in RTLDeviceInfoTy()
730 DeviceInfo.WarpSize[device_id] = wavefront_size; in __tgt_rtl_init_device()
734 DeviceInfo.WarpSize[device_id] = in __tgt_rtl_init_device()
779 DeviceInfo.WarpSize[device_id], DeviceInfo.ThreadsPerGroup[device_id], in __tgt_rtl_init_device()
|