Wenkai Du 3be905ca83 Improve RDMA flushing by write dummy payload with RO=0 (#1570) ...

* Improve RDMA flushing by write dummy payload with RO=0

* Rename env var for disabling this change to RCCL_GDR_FLUSH_GPU_MEM_NO_RELAXED_ORDERING

[ROCm/rccl commit: 60c1264d27]

2025-02-27 16:20:32 -08:00

..

clique

Add support for using contiguous for GPU direct RDMA (#1096)

2024-02-29 10:06:43 -08:00

device

[BUILD] Fix generate.py for gfx950 (#1575)

2025-02-26 18:43:37 -06:00

graph

enable building rccl for gfx950 (#1571)

2025-02-25 16:13:48 -05:00

include

Use bit reversal based mapping for multi-node (#1572)

2025-02-26 09:48:03 -05:00

misc

enable building rccl for gfx950 (#1571)

2025-02-25 16:13:48 -05:00

transport

Improve RDMA flushing by write dummy payload with RO=0 (#1570)

2025-02-27 16:20:32 -08:00

bootstrap.cc

Merge remote-tracking branch 'nccl/master' into develop

2025-01-23 12:08:46 -06:00

channel.cc

2.22.3-1

2024-06-19 01:57:16 -07:00

collectives.cc

Add ncclDataType_t as type to ROCTX (#1512)

2025-01-30 13:46:48 -05:00

debug.cc

Add IB verbs logging and enable traces through install.sh (#1511)

2025-01-31 12:35:39 -05:00

enhcompat.cc

2.12.7-1

2022-03-02 20:48:56 +01:00

enqueue.cc

Use bit reversal based mapping for multi-node (#1572)

2025-02-26 09:48:03 -05:00

group.cc

Merge remote-tracking branch 'nccl/master' into develop

2025-01-23 12:08:46 -06:00

init_nvtx.cc

Merge remote-tracking branch 'nccl/master' into develop

2025-01-23 12:08:46 -06:00

init.cc

Use bit reversal based mapping for multi-node (#1572)

2025-02-26 09:48:03 -05:00

Makefile

2.20.3-1

2024-02-13 04:22:38 -08:00

msccl.cc

Add ncclDataType_t as type to ROCTX (#1512)

2025-01-30 13:46:48 -05:00

nccl.h.in

Merge remote-tracking branch 'nccl/master' into develop

2025-01-23 12:08:46 -06:00

nccl.pc.in

Add pkgconfig file (#190)

2019-04-08 09:16:54 -07:00

net.cc

Merge remote-tracking branch 'nccl/master' into develop

2025-01-23 12:08:46 -06:00

proxy.cc

Make proxy dump print out meaningful information. (#1504)

2025-01-29 16:48:49 -06:00

register.cc

Update MSCCL++ register/deregister (#1523)

2025-02-04 09:09:56 -06:00

transport.cc

Merge remote-tracking branch 'nccl/master' into develop

2025-01-23 12:08:46 -06:00