2.17.1-1

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.


[ROCm/rccl commit: 5d3ab08b69]

projects/rccl/src/include/enqueue.h

@@ -15,8 +15,7 @@
 #define NCCL_MIN_CHANNEL_SIZE (NCCL_LL_THREAD_THRESHOLD*64)
 #define NCCL_AGG_CHANNEL_SIZE (1LL << 21) /* 2 MiB, ideal per-channel size to fully utilize bandwidth */
 
-size_t ncclKernMaxLocalSize();
-ncclResult_t ncclKernSetSharedMemoryCarveout(int carveOut);
+ncclResult_t ncclInitKernelsForDevice(int cudaArch, size_t* maxStackSize);
 ncclResult_t ncclEnqueueCheck(struct ncclInfo* info);
 ncclResult_t ncclLaunchPrepare(struct ncclComm* comm);
 ncclResult_t ncclLaunchKernelBefore_NoUncapturedCuda(struct ncclComm* comm, struct ncclKernelPlan* plan);