42b488507d
rccl-prim-test: print GPU info and set iterations ( #77 )
2019-06-05 15:16:33 -07:00
b7a6307371
Implement HDP flush when transfer data over PCIe P2P ( #75 )
...
* Implement HDP flush when transfer data over PCIe P2P
* Add some descriptions to HDP flushing
* Fix for review comments
2019-06-03 16:29:55 -07:00
8c974f1f50
Merge pull request #76 from ROCmSoftwarePlatform/fix-indirect-call
...
Make ncclFuncs static
2019-05-29 12:04:58 -07:00
5827a4f616
Make ncclFuncs static
...
This is necessary to constant propagate the function pointers
to eliminate the indirect function call.
2019-05-29 10:50:13 -04:00
c85d629355
Merge pull request #74 from saadrahim/readmeUpdate
...
Readme update
2019-05-24 20:27:08 -07:00
0c0a8ed86f
Adding NVIDIA copyright
2019-05-24 15:05:00 -07:00
02ef2d27e6
Fixing whitespace
2019-05-24 14:49:12 -07:00
fac7ef9370
Adding link to readthedocs
2019-05-24 14:48:24 -07:00
9a0ac849fa
Merge pull request #72 from wenkaidu/default_rings
...
Increase number of rings with XGMI connection
2019-05-24 14:42:54 -07:00
bb7542c1d9
Readthedocs documentation support ( #71 )
2019-05-24 15:03:56 -06:00
f45566a8bd
Increase number of rings with XGMI connection
...
Improve throughput for about 20%. Also remove P2P over PCIe which was
left enabled at initial release.
Signed-off-by: Wenkai Du <wenkai.du@amd.com >
2019-05-24 20:58:51 +00:00
b921279a21
Fix build failure for hip-clang ( #69 )
2019-05-23 16:53:25 -06:00
1bb6d2104c
Add RCCL primitive testing ( #70 )
2019-05-23 16:52:17 -06:00
6d8b2421bc
Update debian dependencies in README ( #228 )
...
'fakeroot' is needed for building deb packages
2019-05-22 21:19:36 -07:00
4c4351673b
Jenkinsfile ( #65 )
...
* Changing Jenkinsfile to support runs without docker
* Updating install file for build options
* Fixing command execution
* Fixing Jenkinsfile
* fixing test execution
* Removing junit search
2019-05-22 15:32:32 -06:00
42c3e4b93d
Updating readme for 2.5 release ( #67 )
2019-05-22 15:31:12 -06:00
ffe2054ed2
Test combined calls ( #64 )
...
* Adding test for queueing multiple different collectives, 1 device per thread
2019-05-22 15:30:37 -06:00
6e8f40eb22
Update README to note install rocm-cmake ( #68 )
2019-05-22 15:29:59 -06:00
a115f577dd
Adding fix for unsufficient devices / better logging for skipped tests ( #63 )
2019-05-21 14:34:20 -06:00
afa945d6e6
Renaming jenkinsfile
2019-05-21 15:54:41 +00:00
4bfa506a6b
Remove extra compiler path setting
2019-05-21 00:08:42 +00:00
e517dbed5c
By default will not build test program
2019-05-20 18:37:58 +00:00
c57ab960ff
Merge pull request #62 from gilbertlee-amd/AlignmentTests
...
Adding support for alignment tests via sub-datasets
2019-05-18 10:54:52 -06:00
a50c852851
Adding support for alignment tests via sub-datasets
...
Added sample alignment test for AllGather
Datasets no longer free memory on destruction so Release() must be used
2019-05-18 00:04:03 +00:00
0d6a5a3d25
Update README.md
...
Adding mention of requirement for chrpath for unit tests.
2019-05-17 11:42:14 -06:00
28f8dcb7c9
Merge pull request #61 from gilbertlee-amd/master
...
Reducing scope of linker flags to rccl target to avoid warnings.
Changing to tagged version of googletests
Checking for chrpath
2019-05-17 11:36:23 -06:00
248e1d61cc
Adding check for chrpath for unit tests
2019-05-17 17:13:40 +00:00
08fcce5ec9
Fixing GoogleTest to 1.8.1 and making changes to tests to support older API
2019-05-16 23:13:49 +00:00
11f78df04d
Reducing scope of linker flags to rccl target to avoid warnings
2019-05-16 21:39:39 +00:00
55a4b22ad7
Updating RCCL based on NCCL 2.3.7
...
- Contains modifications to support AMD hardware
- Adds unit tests
2019-05-16 16:16:18 +00:00
0ceaec9cee
NCCL 2.4.7-1
...
Performance tweaks for PowerPC builds only;
Set default NCCL_MIN_NRINGS to 4
Disable PCI-E NUMA distance detection
2019-05-10 13:52:16 -07:00
60a586ded9
Allow CUDA runtime library selection ( #220 )
...
Makes a change to allow the user to select between the static CUDA
runtime library (default) and the dynamic CUDA runtime library. Does
this by allowing `CUDARTLIB` to be overridden.
2019-05-07 17:35:14 -07:00
9db4b1d801
Add pkgconfig file ( #190 )
2019-04-08 09:16:54 -07:00
f40ce73e89
NCCL 2.4.6-1
...
Added detection of IBM/Power NVLink bridge device.
Add NUMA support to PCI distance calculations.
Added NCCL_IGNORE_CPU_AFFINITY env var.
Fix memory leaks; GithubIssue#180
Compiler warning fix; GithubIssue#178
Replace non-standard variable length arrays. GithubIssue#171
Fix Tree+Shared Memory crash. GithubPR#185
Fix LL cleanup hang during long running DL jobs.
Fix NCCL_RINGS environment variable handling.
Added extra checks to catch repeat calls to ncclCommDestroy() GithubIssue#191
Improve bootstrap socket connection reliability at scale.
Fix hostname hashing issue. GithubIssue#187
Code cleanup to rename all non device files from *.cu to *.cc
2019-04-05 13:05:45 -07:00
161763aab2
Fix share memory collision in multi-communicator case.
...
Current SHM object name would only use pidHash and ranks as
identification, which would collide each other when program runs with
multiple communicators. Here we added commId info into pidHash, it makes
'pidHash'es of different communicators keeping in same process will be
distincted with each other.
2019-03-15 12:50:32 +08:00
14e0cf644b
Fix crash during shared memory creation ( #185 )
...
The shared memory filename was only based on the destination. While
this was OK for rings since only one rank would send data to a given
rank, it would crash with trees because they communicate in both
directions.
Co-authored-by: Rong Ou <rong.ou@gmail.com >
2019-03-04 11:42:47 -08:00
1450d42675
2.4.2-1
...
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.
2019-01-29 15:19:27 -08:00
4861e197fd
Fix memory leak in bootstrapRoot()
2019-01-07 14:18:46 -08:00
c244b51ae7
Replace CUDA_VERSION by CUDART_VERSION
2018-12-13 15:22:17 -08:00
3e6afef473
Qualify nullptr_t with std::
2018-12-13 14:18:09 -08:00
346fc49514
Two temporary workarounds for cuda-clang issues.
2018-12-13 14:17:58 -08:00
d08e9b5279
Change __CUDACC_VER_*__ preprocessor directives to CUDA_VERSION because clang doesn't define the former.
2018-12-13 14:17:46 -08:00
469b69a5d0
Fix #163 : remove warnings
2018-12-11 09:19:16 -08:00
8606cdb8b2
Fix dummy plugin
2018-12-05 17:25:23 -08:00
57368189e1
Remove error logging from a normal path
...
When initNet fails, we should not print the backtrace as it is
supposed to be normal operation (falling back to sockets)
2018-12-04 14:47:41 -08:00
4b39a4cf91
Fix GPU Direct RDMA detection.
...
Whether the network supported GPU Direct RDMA or not was ignored,
causing sockets to break when cards were local enough that NCCL
tried to use it.
2018-12-04 14:42:28 -08:00
b8a9a32ccb
Add NCCL_NET flag to many debug lines.
2018-12-04 13:10:19 -08:00
cdae05b277
Improve INFO message when external network is not found.
...
Fix #162
2018-12-04 12:10:58 -08:00
5fe2618c0e
Fixed some compilation errors when TRACE=1 set
2018-11-29 14:12:14 -08:00
eed8218e17
Rework shared memory code to use SYSCHECK macros.
...
This is to handle EINTR/EGAIN properly (issue #137 ), and also
make the code consistent with the rest.
Unfortunately posix_fallocate and mmap do not follow the classic
return code/errno pattern, so we need to write wrappers around those
functions.
2018-11-29 12:52:13 -08:00
Wenkai Du
Yaxun Sam Liu
Saad Rahim
saadrahim
Wenkai Du
Rajat Chopra
gilbertlee-amd
Aaron Enye Shi
Stanley Tsang
Gilbert Lee
David Addison
jakirkham
Gustavo Alvarez
Cao Zongyan
Rong Ou
Sylvain Jeaugey
Christian Sigg
Christian Sigg
Ke Wen