xinyun/vllm
1
0
Fork 0
mirror of https://git.datalinker.icu/vllm-project/vllm.git synced 2026-07-09 08:17:09 +08:00
Code Issues Packages Projects Releases Wiki Activity

[sleep mode] error out with expandable_segments (#14189)

Browse Source 
Signed-off-by: youkaichao <youkaichao@gmail.com>
This commit is contained in:
youkaichao 2025-03-04 18:54:19 +08:00 committed by GitHub
parent ac65bc92df
commit 71c4b40562
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 7 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

7
vllm/device_allocator/cumem.py
View File

@ -8,6 +8,7 @@
# not sure why, they are created from a different context. # not sure why, they are created from a different context.
# the only successful approach is to call cuda driver API in C. # the only successful approach is to call cuda driver API in C.
import dataclasses import dataclasses
import os
from contextlib import contextmanager from contextlib import contextmanager
from typing import Any, Callable, Dict, Optional, Tuple, Union from typing import Any, Callable, Dict, Optional, Tuple, Union
@ -140,6 +141,12 @@ class CuMemAllocator:
return CuMemAllocator.instance return CuMemAllocator.instance
def __init__(self): def __init__(self):
conf = os.environ.get("PYTORCH_CUDA_ALLOC_CONF", "")
assert "expandable_segments:True" not in conf, \
("Expandable segments are not compatible with memory pool. "
"Please track https://github.com/pytorch/pytorch/issues/147851 "
"for the latest updates.")
self.pointer_to_data: Dict[int, AllocationData] = {} self.pointer_to_data: Dict[int, AllocationData] = {}
self.current_tag: str = CuMemAllocator.default_tag self.current_tag: str = CuMemAllocator.default_tag
self.allocator_and_pools: Dict[str, Any] = {} self.allocator_and_pools: Dict[str, Any] = {}