microsoft/mscclpp
1
0
Fork 0
mirror of https://github.com/microsoft/mscclpp.git synced 2026-06-08 15:30:41 +00:00
Code Issues Packages Projects Releases Wiki Activity

Add CI pipeline for no-IB environment testing (#755)

Browse Source 
## Summary

Add CI pipeline support for testing in environments without InfiniBand
(IB) hardware.

## Changes

### IB stubs for no-IB builds (`src/core/ib.cc`)
- Added stub implementations for `IbMr` and `IbQp` classes in the `#else
// !defined(USE_IBVERBS)` block so the library links successfully when
built with `-DMSCCLPP_USE_IB=OFF`.

### Environment variable to disable IB tests
(`MSCCLPP_DISABLE_IB_TESTS`)
- Added `disableIbTests` field to the `Env` class
(`include/mscclpp/env.hpp`, `src/core/env.cpp`), reading from
`MSCCLPP_DISABLE_IB_TESTS` env var.
- Exposed as `disable_ib_tests` in Python bindings
(`python/csrc/env_py.cpp`).
- Updated `python/test/test_mscclpp.py` to skip IB-dependent tests
(`create_group_and_connection` with IB transport, `test_h2h_semaphores`,
`test_h2h_semaphores_gil_release`) when `env().disable_ib_tests` is
true.

### CI pipeline (`ut-no-ib-env.yaml`, `ut.yml`)
The no-IB environment pipeline runs two phases:

1. **No-IB build phase**: Build with `-DMSCCLPP_USE_IB=OFF`, deploy, run
unit tests, multi-process unit tests, and pytests (with
`MSCCLPP_DISABLE_IB_TESTS=1`).
2. **IB build phase**: Rebuild with IB enabled (default), stop the
existing container, redeploy, and run pytests (with
`MSCCLPP_DISABLE_IB_TESTS=1`) — verifying that the full IB-enabled build
works correctly in a non-IB environment when IB tests are skipped.

Also increased the job timeout from 40 to 60 minutes to accommodate the
two-phase pipeline.
This commit is contained in:
Binyang Li
2026-02-24 15:55:59 -08:00
committed by GitHub
parent 7738603d63
commit 184dcbf9d7
4 changed files with 140 additions and 9 deletions
Download Patch File Download Diff File Expand all files Collapse all files

13
python/test/test_mscclpp.py
View File

@@ -162,13 +162,10 @@ def create_connection(group: CommGroup, connection_type: str):
def create_group_and_connection(mpi_group: MpiGroup, connection_type: str):
if (connection_type == "NVLink" or connection_type == "NVLS") and all_ranks_on_the_same_node(mpi_group) is False:
pytest.skip("cannot use nvlink/nvls for cross node")
if connection_type == "IB" and os.environ.get("MSCCLPP_DISABLE_IB_TESTS", "0") != "0":
pytest.skip("IB tests are disabled via MSCCLPP_DISABLE_IB_TESTS=1")
group = CommGroup(mpi_group.comm)
try:
connection = create_connection(group, connection_type)
except Error as e:
if connection_type == "IB" and e.args[0] == ErrorCode.InvalidUsage:
pytest.skip("IB not supported on this node")
raise
connection = create_connection(group, connection_type)
return group, connection
@@ -281,6 +278,8 @@ def test_connection_write_and_signal(mpi_group: MpiGroup, connection_type: str,
@parametrize_mpi_groups(2, 4, 8, 16)
def test_h2h_semaphores(mpi_group: MpiGroup):
if os.environ.get("MSCCLPP_DISABLE_IB_TESTS", "0") != "0":
pytest.skip("IB tests are disabled via MSCCLPP_DISABLE_IB_TESTS=1")
group = CommGroup(mpi_group.comm)
tran = group.my_ib_device(group.my_rank % 8)
endpoint = EndpointConfig(tran, Device(DeviceType.CPU))
@@ -301,6 +300,8 @@ def test_h2h_semaphores(mpi_group: MpiGroup):
@parametrize_mpi_groups(2, 4, 8, 16)
def test_h2h_semaphores_gil_release(mpi_group: MpiGroup):
if os.environ.get("MSCCLPP_DISABLE_IB_TESTS", "0") != "0":
pytest.skip("IB tests are disabled via MSCCLPP_DISABLE_IB_TESTS=1")
group = CommGroup(mpi_group.comm)
tran = group.my_ib_device(group.my_rank % 8)
endpoint = EndpointConfig(tran, Device(DeviceType.CPU))