2.18.1-1

Add support for IB SHARP to NVLS (NVLink SHARP algorithm).
Add NVLS+Tree algorithm.
Add support for memory management using cuMem* functions.
Use all NICs for Send/Receive operations on systems with more than
one NIC per GPU (#804).
Add ncclCommSplit primitive, with resource sharing option in config.
Fix alltoallv hang (#788)
Increase number of channels on H100 when we're not limited by NVLink.
Improve error reporting in case of IB failure, printing local and
remote ID (#779).
Add build option to allow compilation against RDMA includes instead
of dynamically loading IB verbs symbols (#802).
Fix context creation for progress thread (#803).
NET/IB: add option to use multiple QPs in round-robin mode.
Fix tree performance issue when NVB is disabled on HCM topologies.


[ROCm/rccl commit: d97a32fac8]

projects/rccl/src/include/info.h

@@ -25,6 +25,7 @@ typedef enum : uint8_t {
   ncclPatternCollnetChain,
   ncclPatternCollnetDirect,
   ncclPatternNvls,
+  ncclPatternNvlsTree,
   ncclPatternSend,
   ncclPatternRecv
 } ncclPattern_t;
@@ -93,7 +94,6 @@ struct ncclCudaStreamList {
   struct ncclCudaStreamList *next;
   cudaStream_t stream;
 };
-
 struct ncclTasks {
   struct Peer {
     bool sendSeen, recvSeen;
@@ -103,7 +103,8 @@ struct ncclTasks {
   struct ncclIntruQueue<ncclTaskColl, &ncclTaskColl::next> collQueue;
   size_t collBytesTotal;
   struct Peer* peers/*[nRanks]*/;
-  int *p2pSendOrder/*[nRanks]*/, *p2pRecvOrder/*[nRanks]*/;
+  int *p2pSendOrder, *p2pRecvOrder;
+  int p2pOrderSteps;
   int nTasksColl, nTasksP2p;
 
   // The list of user streams aggregated over all tasks present.