[XPU] Enable sequence parallel support for XPU (#38608)

Signed-off-by: chaojun-zhang <chaojun.zhang@intel.com>
Signed-off-by: Chaojun Zhang <chaojun.zhang@intel.com>
Signed-off-by: Chaojun,Zhang <chaojun.zhang@intel.com>
This commit is contained in:
Chaojun Zhang
2026-06-14 19:26:46 -07:00
committed by GitHub
parent 1801fad0ba
commit 2725c84aae
5 changed files with 126 additions and 17 deletions
+23
View File
@@ -24,6 +24,7 @@ def mock_cuda_platform():
def _mock_platform(is_cuda: bool = True, capability: tuple[int, int] | None = None):
mock_platform = MagicMock()
mock_platform.is_cuda.return_value = is_cuda
mock_platform.is_xpu.return_value = False
device_capability = (
DeviceCapability(*capability) if capability is not None else None
)
@@ -46,3 +47,25 @@ def mock_cuda_platform():
yield mock_platform
return _mock_platform
@pytest.fixture
def mock_xpu_platform():
"""
Fixture that returns a factory for creating mocked XPU platforms.
Usage:
def test_something(mock_xpu_platform):
with mock_xpu_platform():
# test code
"""
@contextmanager
def _mock_platform():
mock_platform = MagicMock()
mock_platform.is_cuda.return_value = False
mock_platform.is_xpu.return_value = True
with patch("vllm.platforms.current_platform", mock_platform):
yield mock_platform
return _mock_platform