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
This commit is contained in:
Wenkai Du
2022-04-18 11:14:51 -07:00
parent fd2f1b3b88 9bfc1c6e35
commit d28e1cb44f
81 changed files with 7358 additions and 4468 deletions
Download Patch File Download Diff File Expand all files Collapse all files
src/include/transport.h
+15 -5
View File
@@ -1,5 +1,6 @@
/*************************************************************************
* Copyright (c) 2016-2021, NVIDIA CORPORATION. All rights reserved.
* Copyright (c) 2016-2022, NVIDIA CORPORATION. All rights reserved.
* Modifications Copyright (c) 2019-2022 Advanced Micro Devices, Inc. All rights reserved.
*
* See LICENSE.txt for license information
************************************************************************/
@@ -11,12 +12,14 @@
#include "graph.h"
#include "nvmlwrap.h"
#include "core.h"
#include "proxy.h"
#define NTRANSPORTS 3
#define NTRANSPORTS 4
#define TRANSPORT_P2P 0
#define TRANSPORT_SHM 1
#define TRANSPORT_NET 2
#define TRANSPORT_COLLNET 3
#include "proxy.h"
extern struct ncclTransport ncclTransports[];
@@ -28,12 +31,15 @@ struct ncclComm;
struct ncclPeerInfo {
int rank;
int cudaDev;
int netDev;
int gdrSupport;
bool hasFineGrain;
uint64_t hostHash;
uint64_t pidHash;
dev_t shmDev;
int64_t busId;
struct ncclComm* comm;
int cudaCompCap;
};
#define CONNECT_SIZE 128
@@ -44,8 +50,12 @@ struct ncclConnect {
struct ncclTransportComm {
ncclResult_t (*setup)(struct ncclComm* comm, struct ncclTopoGraph* graph, struct ncclPeerInfo*, struct ncclPeerInfo*, struct ncclConnect*, struct ncclConnector*, int channelId, int connIndex);
ncclResult_t (*connect)(struct ncclComm* comm, struct ncclConnect*, int nranks, int rank, struct ncclConnector*);
ncclResult_t (*free)(void*);
ncclResult_t (*proxy)(struct ncclProxyArgs*);
ncclResult_t (*free)(struct ncclConnector*);
ncclResult_t (*proxySharedInit)(struct ncclProxyConnection* connection, struct ncclComm* comm, int nChannels);
ncclResult_t (*proxySetup)(struct ncclProxyConnection* connection, struct ncclComm* comm, void* reqBuff, int reqSize, void* respBuff, int respSize, int* done);
ncclResult_t (*proxyConnect)(struct ncclProxyConnection* connection, struct ncclComm* comm, void* reqBuff, int reqSize, void* respBuff, int respSize, int* done);
ncclResult_t (*proxyFree)(struct ncclProxyConnection* connection, struct ncclComm* comm);
ncclResult_t (*proxyProgress)(struct ncclComm* comm, struct ncclProxyArgs*);
};
struct ncclTransport {