cuda_malloc: add missing GPUs to blacklist and fix bare except clauses #12333
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Comfy-Org#940)\n\nAdd GPU models reported in issue Comfy-Org#940 that are missing from the\ncudaMallocAsync blacklist:\n- Tesla P40 (Pascal datacenter GPU)\n- Tesla P4 (Pascal datacenter GPU)\n- NVIDIA L4 (reported by users in cloud environments)\n- NVIDIA A10G (AWS GPU, confirmed fix by user)\n- GRID A800D (vGPU mode, reported by user)\n\nAlso replace bare `except:` with `except Exception:` in three locations\nto follow Python best practices and avoid accidentally catching\nKeyboardInterrupt and SystemExit.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Summary
Fixes #940 - Adds GPU models reported by users in the issue that are missing from the
cudaMallocAsyncblacklist, and replaces bareexcept:withexcept Exception:for better Python practices.Changes
1. Added missing GPUs to the blacklist
Several users reported
CUDA error: operation not supportedon GPUs not yet in the blacklist. These have been added:--disable-cuda-mallocfixed itThe blacklist uses substring matching (
if b in x), so these entries correctly match the full GPU name strings returned by bothnvidia-smi -L(Linux) andEnumDisplayDevicesA(Windows).2. Replaced bare
except:withexcept Exception:Three locations in the file used bare
except:which catches all exceptions includingKeyboardInterruptandSystemExit. This is a Python anti-pattern (PEP 8, B001/E722). Changed toexcept Exception:which still catches all operational errors (OSError,FileNotFoundError,RuntimeError,CalledProcessError, etc.) but allows clean process termination.Locations changed:
cuda_malloc_supported()- catches failures fromget_gpu_names()(nvidia-smi/ctypes)Testing
except Exception:catchesRuntimeError,OSError,FileNotFoundError,CalledProcessError; correctly does NOT catchKeyboardInterruptorSystemExitpy_compilepasses