AI/rocm-systems
2
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Merge remote-tracking branch 'nccl/master' into develop

Browse Source 
[ROCm/rccl commit: a79d9e3586]
This commit is contained in:
Wenkai Du
2022-09-09 01:20:52 +00:00
parent fe99249cde c9a50a9ec5
commit 7874a99c75
90 changed files with 5517 additions and 3115 deletions
Download Patch File Download Diff File Expand all files Collapse all files
projects/rccl/tools/topo_expl/model.cpp
+8 -7
View File
@@ -179,10 +179,10 @@ ncclResult_t netSendSetup(struct ncclComm* comm, struct ncclTopoGraph* graph, st
NCCLCHECK(ncclTopoCheckGdr(comm->topo, myInfo->busId, req.netDev, 1, &req.useGdr));
if (proxyRank == myInfo->rank) {
INFO(NCCL_INIT|NCCL_NET,"Channel %02d/%d : %d[%lx] -> %d[%lx] [send] via NET/%s/%d%s%s", channelId, connIndex, myInfo->rank, myInfo->busId, peerInfo->rank, peerInfo->busId, ncclNetName(), req.netDev,
INFO(NCCL_INIT|NCCL_NET,"Channel %02d/%d : %d[%lx] -> %d[%lx] [send] via NET/%s/%d%s%s", channelId, connIndex, myInfo->rank, myInfo->busId, peerInfo->rank, peerInfo->busId, ncclNetName(comm), req.netDev,
req.useGdr ? "/GDRDMA" : "", req.shared ? "/Shared" : "");
} else {
INFO(NCCL_INIT|NCCL_NET,"Channel %02d/%d : %d[%lx] -> %d[%lx] [send] via NET/%s/%d(%d)%s%s", channelId, connIndex, myInfo->rank, myInfo->busId, peerInfo->rank, peerInfo->busId, ncclNetName(), req.netDev,
INFO(NCCL_INIT|NCCL_NET,"Channel %02d/%d : %d[%lx] -> %d[%lx] [send] via NET/%s/%d(%d)%s%s", channelId, connIndex, myInfo->rank, myInfo->busId, peerInfo->rank, peerInfo->busId, ncclNetName(comm), req.netDev,
proxyRank, req.useGdr ? "/GDRDMA" : "", req.shared ? "/Shared" : "");
}
*((int*)connectInfo) = proxyRank;
@@ -205,7 +205,7 @@ ncclResult_t netRecvSetup(struct ncclComm* comm, struct ncclTopoGraph* graph, st
if (req.netDev < 0) NCCLCHECK(ncclTopoGetNetDev(comm, myInfo->rank, graph, channelId, myInfo->rank, &req.netDev, &proxyRank));
NCCLCHECK(ncclTopoCheckGdr(comm->topo, myInfo->busId, req.netDev, 0, &req.useGdr));
INFO(NCCL_INIT|NCCL_NET,"Channel %02d/%d : %d[%lx] -> %d[%lx] [receive] via NET/%s/%d%s%s", channelId, connIndex, peerInfo->rank, peerInfo->busId, myInfo->rank, myInfo->busId, ncclNetName(), req.netDev,
INFO(NCCL_INIT|NCCL_NET,"Channel %02d/%d : %d[%lx] -> %d[%lx] [receive] via NET/%s/%d%s%s", channelId, connIndex, peerInfo->rank, peerInfo->busId, myInfo->rank, myInfo->busId, ncclNetName(comm), req.netDev,
req.useGdr ? "/GDRDMA" : "", req.shared ? "/Shared" : "");
return ncclSuccess;
}
@@ -250,8 +250,9 @@ struct ncclTransport collNetTransport = {
{ collNetRecvSetup, NULL, NULL, NULL }
};
struct ncclTransport ncclTransports[NTRANSPORTS] = {
p2pTransport,
shmTransport,
netTransport,
struct ncclTransport* ncclTransports[] = {
&p2pTransport,
&shmTransport,
&netTransport,
&collNetTransport,
};