AI/rocm-systems
2
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

2.17.1-1

Browse Source 
Add new NVLS algorithm for allreduce using NVLink SHARP (intra-node only).
Add new config options: cgaClusterSize, minCTAs, maxCTAs, netName.
Enable LL128 when we use PXN to close rings.
NVTX3 includes update.
Fix crash when one CollNet (SHARP) rail fails to initialize.
This commit is contained in:
Sylvain Jeaugey
2023-02-27 02:48:21 -08:00
parent f3d5166783
commit 5d3ab08b69
72 changed files with 4541 additions and 2391 deletions
Download Patch File Download Diff File Expand all files Collapse all files
src/include/cudawrap.h
+23 -1
View File
@@ -73,10 +73,32 @@ DECLARE_CUDA_PFN_EXTERN(cuGetErrorName, 6000);
DECLARE_CUDA_PFN_EXTERN(cuMemGetAddressRange, 3020);
DECLARE_CUDA_PFN_EXTERN(cuCtxCreate, 3020);
DECLARE_CUDA_PFN_EXTERN(cuCtxDestroy, 4000);
DECLARE_CUDA_PFN_EXTERN(cuCtxGetCurrent, 4000);
DECLARE_CUDA_PFN_EXTERN(cuCtxSetCurrent, 4000);
DECLARE_CUDA_PFN_EXTERN(cuCtxGetDevice, 2000);
// cuMem API support
DECLARE_CUDA_PFN_EXTERN(cuMemAddressReserve, 10020);
DECLARE_CUDA_PFN_EXTERN(cuMemAddressFree, 10020);
DECLARE_CUDA_PFN_EXTERN(cuMemCreate, 10020);
DECLARE_CUDA_PFN_EXTERN(cuMemGetAllocationGranularity, 10020);
DECLARE_CUDA_PFN_EXTERN(cuMemExportToShareableHandle, 10020);
DECLARE_CUDA_PFN_EXTERN(cuMemImportFromShareableHandle, 10020);
DECLARE_CUDA_PFN_EXTERN(cuMemMap, 10020);
DECLARE_CUDA_PFN_EXTERN(cuMemRelease, 10020);
DECLARE_CUDA_PFN_EXTERN(cuMemSetAccess, 10020);
DECLARE_CUDA_PFN_EXTERN(cuMemUnmap, 10020);
#if CUDA_VERSION >= 11070
DECLARE_CUDA_PFN_EXTERN(cuMemGetHandleForAddressRange, 11070); // DMA-BUF support
#endif
#if CUDA_VERSION >= 12010
/* NVSwitch Multicast support */
DECLARE_CUDA_PFN_EXTERN(cuMulticastAddDevice, 12010);
DECLARE_CUDA_PFN_EXTERN(cuMulticastBindMem, 12010);
DECLARE_CUDA_PFN_EXTERN(cuMulticastBindAddr, 12010);
DECLARE_CUDA_PFN_EXTERN(cuMulticastCreate, 12010);
DECLARE_CUDA_PFN_EXTERN(cuMulticastGetGranularity, 12010);
DECLARE_CUDA_PFN_EXTERN(cuMulticastUnbind, 12010);
#endif
#endif
/* CUDA Driver functions loaded with dlsym() */
@@ -88,6 +110,7 @@ DECLARE_CUDA_PFN_EXTERN(cuGetProcAddress, 11030);
ncclResult_t ncclCudaLibraryInit(void);
extern int ncclCudaDriverVersionCache;
extern bool ncclCudaLaunchBlocking; // initialized by ncclCudaLibraryInit()
inline ncclResult_t ncclCudaDriverVersion(int* driver) {
int version = __atomic_load_n(&ncclCudaDriverVersionCache, __ATOMIC_RELAXED);
@@ -98,5 +121,4 @@ inline ncclResult_t ncclCudaDriverVersion(int* driver) {
*driver = version;
return ncclSuccess;
}
#endif