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: 0b2062c560]
This commit is contained in:
BertanDogancay
2025-03-27 12:51:55 -05:00
parent f0c853438c 60240fec77
commit 8ed27fde74
92 changed files with 7322 additions and 2168 deletions
Download Patch File Download Diff File Expand all files Collapse all files
projects/rccl/src/include/graph.h
+4 -3
View File
@@ -34,16 +34,17 @@ ncclResult_t ncclTopoComputeCommCPU(struct ncclComm* comm);
// Query topology
ncclResult_t ncclTopoGetNetDev(struct ncclComm* comm, int rank, struct ncclTopoGraph* graph, int channelId, int peerRank, int64_t* id, int* dev, int* proxyRank);
ncclResult_t ncclTopoCheckP2p(struct ncclTopoSystem* system, int64_t id1, int64_t id2, int* p2p, int *read, int* intermediateRank);
ncclResult_t ncclTopoCheckP2p(struct ncclTopoSystem* system, int rank1, int rank2, int* p2p, int *read, int* intermediateRank);
ncclResult_t ncclTopoCheckMNNVL(struct ncclTopoSystem* system, struct ncclPeerInfo* info1, struct ncclPeerInfo* info2, int* ret);
ncclResult_t ncclTopoCheckGdr(struct ncclTopoSystem* topo, int64_t busId, int64_t netId, int read, int* useGdr);
#define MAX_XGMI_INTER_GPUS 4
ncclResult_t ncclTopoGetIntraNetDev(struct ncclTopoSystem* system, int rank, struct ncclTopoGraph* graph, int channelId, int type, int64_t* id, int* dev);
ncclResult_t ncclTopoGetLinkType(struct ncclTopoSystem* system, int cudaDev1, int cudaDev2, bool* isXGMI, int maxInter=MAX_XGMI_INTER_GPUS, int nInter=0, int *inter=nullptr);
ncclResult_t ncclTopoNeedFlush(struct ncclTopoSystem* system, int64_t busId, int* flush);
ncclResult_t ncclTopoCheckNet(struct ncclTopoSystem* system, int64_t id1, int64_t id2, int* net);
ncclResult_t ncclTopoCheckNet(struct ncclTopoSystem* system, int rank1, int rank2, int* net);
int ncclPxnDisable(struct ncclComm* comm);
ncclResult_t ncclTopoGetPxnRanks(struct ncclComm* comm, int** intermediateRanks, int* nranks);
ncclResult_t ncclGetLocalCpu(struct ncclTopoSystem* system, int gpu, int* retCpu);
// Find CPU affinity
ncclResult_t ncclTopoGetCpuAffinity(struct ncclTopoSystem* system, int rank, cpu_set_t* affinity);
@@ -82,7 +83,7 @@ ncclResult_t ncclTopoSearchInit(struct ncclTopoSystem* system);
#define NCCL_TOPO_PATTERN_COLLNET_DIRECT 6 // Collnet Direct
struct ncclTopoGraph {
// Input / output
int id; // ring : 0, tree : 1, collnet : 2
int id; // ring : 0, tree : 1, collnet : 2, nvls : 3, collnetDirect : 4
int pattern;
int crossNic;
int collNet;