xinyun/vllm
1
0
Fork 0
mirror of https://git.datalinker.icu/vllm-project/vllm.git synced 2026-03-20 23:49:14 +08:00
Code Issues Packages Projects Releases Wiki Activity

[Bugfix][plugin] fla crash on plugin (#27322)

Browse Source
This commit is contained in:
Hank_ 2025-11-04 05:27:03 +08:00 committed by GitHub
parent 01baefe674
commit ccd3e55e51
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 3 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

5
vllm/model_executor/layers/fla/ops/utils.py
View File

@ -17,6 +17,7 @@ from typing import Any, Literal
import torch
from vllm.platforms import current_platform
from vllm.triton_utils import triton
logger = logging.getLogger(__name__)
@ -137,8 +138,8 @@ def _check_platform() -> Literal["nvidia", "amd", "intel", "musa"]:
# For AMD GPUs, the triton backend is 'hip', while for Nvidia GPUs, the triton backend is 'cuda'.
# However, the torch backend is 'cuda' for both Nvidia and AMD GPUs.
# Therefore, we need to check the triton backend to determine the actual GPU vendor.
device = get_available_device() if get_available_device() != "hip" else "cuda"
device_torch_lib = getattr(torch, device)
device = "cuda" if current_platform.is_cuda_alike() else get_available_device()
device_torch_lib = getattr(torch, device, None)
device_platform = _check_platform()
is_amd = device_platform == "amd"