28189e2df8
Add support for CUDA 12.0, drop Kepler (sm_35). Support for H100 features. Make socket code more robust and protected. Solves #555. Improve performance on large CUDA graphs, reducing dependencies. Reduce inter-socket bandwidth on AMD CPUs to favor better paths. Various fixes to ncclCommAbort. Make service thread polling resistant to EINTR. Compile with profiling API by default. Extend NVTX instrumentation with call arguments.
32 lines
1.4 KiB
C++
32 lines
1.4 KiB
C++
/*************************************************************************
|
|
* Copyright (c) 2015-2020, NVIDIA CORPORATION. All rights reserved.
|
|
*
|
|
* See LICENSE.txt for license information
|
|
************************************************************************/
|
|
|
|
#include "enqueue.h"
|
|
#include "collectives.h"
|
|
#include "nccl.h"
|
|
|
|
NCCL_API(ncclResult_t, ncclReduceScatter, const void* sendbuff, void* recvbuff, size_t recvcount,
|
|
ncclDataType_t datatype, ncclRedOp_t op, ncclComm* comm, cudaStream_t stream);
|
|
ncclResult_t ncclReduceScatter(const void* sendbuff, void* recvbuff, size_t recvcount,
|
|
ncclDataType_t datatype, ncclRedOp_t op, ncclComm* comm, cudaStream_t stream) {
|
|
struct NvtxParamsReduceScatter {
|
|
size_t bytes;
|
|
ncclRedOp_t op;
|
|
};
|
|
constexpr nvtxPayloadSchemaEntry_t ReduceScatterSchema[] = {
|
|
{0, NVTX_PAYLOAD_ENTRY_TYPE_SIZE, "Message size [bytes]"},
|
|
{0, NVTX_PAYLOAD_ENTRY_NCCL_REDOP, "Reduction operation", nullptr, 0,
|
|
offsetof(NvtxParamsReduceScatter, op)}
|
|
};
|
|
NvtxParamsReduceScatter payload{recvcount * ncclTypeSize(datatype), op};
|
|
NVTX3_FUNC_WITH_PARAMS(ReduceScatter, ReduceScatterSchema, payload)
|
|
|
|
struct ncclInfo info = { ncclFuncReduceScatter, "ReduceScatter",
|
|
sendbuff, recvbuff, recvcount, datatype, op, 0, comm, stream, /* Args */
|
|
REDUCESCATTER_CHUNKSTEPS, REDUCESCATTER_SLICESTEPS };
|
|
return ncclEnqueueCheck(&info);
|
|
}
|