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

2.9.6-1

Browse Source 
Add support for CUDA graphs.
Fuse BCM Gen4 switches to avoid suboptimal performance on some platforms. Issue #439.
Fix bootstrap issue caused by connection reordering.
Fix CPU locking block.
Improve CollNet algorithm.
Improve performance on DGX A100 for communicators with only one GPU per node.


[ROCm/rccl commit: a46ea10583]
This commit is contained in:
Sylvain Jeaugey
2021-04-12 16:00:11 -07:00
parent fc7bdb38a5
commit 20da390b96
43 changed files with 2687 additions and 1244 deletions
Download Patch File Download Diff File Expand all files Collapse all files
projects/rccl/src/include/ibvwrap.h
+2 -2
View File
@@ -4,7 +4,7 @@
* Copyright (c) 2005, 2006, 2007 Cisco Systems, Inc. All rights reserved.
* Copyright (c) 2005 PathScale, Inc. All rights reserved.
*
* Copyright (c) 2015-2019, NVIDIA CORPORATION. All rights reserved.
* Copyright (c) 2015-2021, NVIDIA CORPORATION. All rights reserved.
*
* See LICENSE.txt for license information
************************************************************************/
@@ -1089,7 +1089,7 @@ static inline int ibv_post_send(struct ibv_qp *qp, struct ibv_send_wr *wr, struc
static inline ncclResult_t wrap_ibv_post_send(struct ibv_qp *qp, struct ibv_send_wr *wr, struct ibv_send_wr **bad_wr) {
int ret = qp->context->ops.post_send(qp, wr, bad_wr); /*returns 0 on success, or the value of errno on failure (which indicates the failure reason)*/
if (ret != IBV_SUCCESS) {
WARN("ibv_post_send() failed with error %s", strerror(ret));
WARN("ibv_post_send() failed with error %s, Bad WR %p, First WR %p", strerror(ret), wr, *bad_wr);
return ncclSystemError;
}
return ncclSuccess;