Sylvain Jeaugey
1450d42675
Add tree algorithms for allreduce to improve performance at scale. Add ncclCommAbort() and ncclCommGetAsyncError() to properly handle network errors and be permit recover. Detect initial CPU affinity and no longer escape it.
19 lines
874 B
Plaintext
19 lines
874 B
Plaintext
/*************************************************************************
|
|
* Copyright (c) 2015-2018, NVIDIA CORPORATION. All rights reserved.
|
|
*
|
|
* See LICENSE.txt for license information
|
|
************************************************************************/
|
|
|
|
#include "enqueue.h"
|
|
#include "collectives.h"
|
|
|
|
NCCL_API(ncclResult_t, ncclReduce, const void* sendbuff, void* recvbuff, size_t count,
|
|
ncclDataType_t datatype, ncclRedOp_t op, int root, ncclComm_t comm, cudaStream_t stream);
|
|
ncclResult_t ncclReduce(const void* sendbuff, void* recvbuff, size_t count,
|
|
ncclDataType_t datatype, ncclRedOp_t op, int root, ncclComm_t comm, cudaStream_t stream) {
|
|
struct ncclInfo info = { ncclCollReduce, "Reduce",
|
|
sendbuff, recvbuff, count, datatype, op, root, comm, stream, /* Args */
|
|
REDUCE_CHUNKSTEPS, REDUCE_SLICESTEPS };
|
|
return ncclEnqueueCheck(&info);
|
|
}
|