Sylvain Jeaugey 8dcf8e8720 2.17.1-1 ...

Add new NVLS algorithm for allreduce using NVLink SHARP (intra-node only).
Add new config options: cgaClusterSize, minCTAs, maxCTAs, netName.
Enable LL128 when we use PXN to close rings.
NVTX3 includes update.
Fix crash when one CollNet (SHARP) rail fails to initialize.


[ROCm/rccl commit: 5d3ab08b69]

2023-03-01 00:39:04 -08:00

..

all_gather.cu

2.8.3-1

2020-11-17 11:08:52 -08:00

all_gather.h

2.17.1-1

2023-03-01 00:39:04 -08:00

all_reduce.cu

2.8.3-1

2020-11-17 11:08:52 -08:00

all_reduce.h

2.17.1-1

2023-03-01 00:39:04 -08:00

broadcast.cu

2.8.3-1

2020-11-17 11:08:52 -08:00

broadcast.h

2.13.4-1

2022-07-11 08:10:34 -07:00

common_kernel.h

2.17.1-1

2023-03-01 00:39:04 -08:00

common.h

2.17.1-1

2023-03-01 00:39:04 -08:00

functions.cu

2.17.1-1

2023-03-01 00:39:04 -08:00

gen_rules.sh

address review comments

2022-09-20 11:58:33 +02:00

Makefile

Fix intermittent 11.6 builds: generate unique .cu file for each object file

2022-09-20 11:58:33 +02:00

onerank_reduce.cu

2.17.1-1

2023-03-01 00:39:04 -08:00

op128.h

2.17.1-1

2023-03-01 00:39:04 -08:00

primitives.h

2.17.1-1

2023-03-01 00:39:04 -08:00

prims_ll128.h

2.17.1-1

2023-03-01 00:39:04 -08:00

prims_ll.h

2.17.1-1

2023-03-01 00:39:04 -08:00

prims_simple.h

2.17.1-1

2023-03-01 00:39:04 -08:00

reduce_kernel.h

2.17.1-1

2023-03-01 00:39:04 -08:00

reduce_scatter.cu

2.8.3-1

2020-11-17 11:08:52 -08:00

reduce_scatter.h

2.17.1-1

2023-03-01 00:39:04 -08:00

reduce.cu

2.8.3-1

2020-11-17 11:08:52 -08:00

reduce.h

2.13.4-1

2022-07-11 08:10:34 -07:00

sendrecv.cu

2.8.3-1

2020-11-17 11:08:52 -08:00

sendrecv.h

2.17.1-1

2023-03-01 00:39:04 -08:00