Rework barrier and event code (#761)

* Rework barrier and event code

* Switch to inline asm

[ROCm/rccl commit: 5a38ff192b]

projects/rccl/src/misc/strongstream.cc

@@ -237,8 +237,6 @@ ncclResult_t ncclStrongStreamRelease(struct ncclCudaGraph graph, struct ncclStro
       }
     }
   #endif
-  CUDACHECK(cudaEventRecord(ss->scratchEvent, ss->cudaStream));
-  CUDACHECK(cudaStreamWaitEvent(ss->cudaStream, ss->scratchEvent, 0));
   return ncclSuccess;
 }