Skip to content

[sleep mode] error out with expandable_segments #14189

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 3 commits into from
Mar 4, 2025
Merged

[sleep mode] error out with expandable_segments #14189

Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
06dd1bb
add assert
youkaichao Mar 4, 2025
d6aecca
fix unit test
youkaichao Mar 4, 2025
a18ee2a
fix unit test
youkaichao Mar 4, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
7 changes: 7 additions & 0 deletions vllm/device_allocator/cumem.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -8,6 +8,7 @@
# not sure why, they are created from a different context.
# the only successful approach is to call cuda driver API in C.
import dataclasses
import os
from contextlib import contextmanager
from typing import Any, Callable, Dict, Optional, Tuple, Union

Expand Down Expand Up @@ -140,6 +141,12 @@ def get_instance() -> "CuMemAllocator":
return CuMemAllocator.instance

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.current_tag: str = CuMemAllocator.default_tag
self.allocator_and_pools: Dict[str, Any] = {}
Expand Down