[Device] Add dynamic fetch/reduce pipelining for reduction collectives - Simple protocol (#1861)

* Support pipelining codegen and template specialization

* Support ReduceCopy pipelining for AllReduce, ReduceScatter, and Reduce (currently enabled for bfloat16)

* Remove need for FUNC_INDEX_TOTAL

* Add pipeline field to device function key construction logic

* Avoid unneeded codegen for LL/LL64 kernels

* Modify conditions and add pipeline dtypes env

* Optimize selection for both gfx942 and gfx950

* Increase pipeline bitfield width

* Use __forceinline__ for all device functions

* Realign reduceCopy with original form

* Add opt-out option to enable perf debugs

* Remove force-reduce-pipelining option from README

* Update CHANGELOG.md

---------

Co-authored-by: Jeffrey Novotny <jnovotny@amd.com>

[ROCm/rccl commit: 277747c199]

projects/rccl/src/enqueue.cc

@@ -457,7 +457,7 @@ ncclResult_t ncclPrepareTasks(struct ncclComm* comm, bool* algoNeedConnect, bool
       }
 
       NCCLCHECK(getAlgoInfo(comm, &agg, collNetSupport, nvlsSupport, nTasksPerChannel, simInfo));
-      agg.devFuncId = ncclDevFuncId(agg.func, agg.opDev.op, agg.datatype, agg.algorithm, agg.protocol);
+      agg.devFuncId = ncclDevFuncId(agg.func, agg.opDev.op, agg.datatype, agg.algorithm, agg.protocol, agg.pipeline);
       if (agg.devFuncId < 0) {
         WARN("%s: unsupported collective. Please ensure the collective has been enabled in build.", __func__);
         return ncclInvalidUsage;
@@ -480,6 +480,7 @@ ncclResult_t ncclPrepareTasks(struct ncclComm* comm, bool* algoNeedConnect, bool
         struct ncclTaskColl* next = aggBeg->next;
         aggBeg->algorithm = agg.algorithm;
         aggBeg->protocol = agg.protocol;
+        aggBeg->pipeline = agg.pipeline;
         if (aggBeg->protocol == NCCL_PROTO_LL) aggBeg->trafficBytes *= 4;
         aggBeg->nMaxChannels = agg.nMaxChannels;
         aggBeg->nWarps = agg.nWarps;
@@ -1941,6 +1942,7 @@ static ncclResult_t topoGetAlgoInfo(
     return (algoEnv || protoEnv) ? ncclInvalidUsage : ncclInternalError;
   }
   rcclUpdateCollectiveProtocol(comm, nBytes, info);
+  rcclSetPipelining(comm, nBytes, info);
   if (simInfo) simInfo->estimatedTime = time;
   TRACE(NCCL_COLL, "%ld Bytes -> Algo %d proto %d time %f", nBytes, info->algorithm, info->protocol, time);