2.11.4-1

Add new API for creating a reduction operation which multiplies the input by a rank-specific scalar before doing an inter-rank summation (see: ncclRedOpCreatePreMulSum).
Improve CollNet (SHARP) performance of ncclAllReduce when captured in a CUDA Graph via user buffer registration.
Add environment variable NCCL_NET_PLUGIN="<suffix>" to allow user to choose among multiple NCCL net plugins by substituting into "libnccl-net-<suffix>.so".
Fix memory leak of NVB connections.
Fix topology detection of IB Virtual Functions (SR-IOV).


[ROCm/rccl commit: e11238b302]

projects/rccl/src/debug.cc

@@ -138,7 +138,7 @@ void ncclDebugLog(ncclDebugLogLevel level, unsigned long flags, const char *file
   int cudaDev;
   cudaGetDevice(&cudaDev);
   int pid = getpid();
-  int tid = gettid();
+  int tid = syscall(SYS_gettid);
 
   char buffer[1024];
   size_t len = 0;