From 7dd33f36752ec02eaf9cfd8290e27f4925abe8f0 Mon Sep 17 00:00:00 2001
From: xmarre <mmquant1@gmail.com>
Date: Wed, 15 Apr 2026 05:25:27 +0200
Subject: [PATCH] Address PR review feedback

---
 nodes.py   | 13 +++++++++++--
 patches.py | 27 +++++++++++----------------
 2 files changed, 22 insertions(+), 18 deletions(-)

diff --git a/nodes.py b/nodes.py
index 97713ff..f136806 100644
--- a/nodes.py
+++ b/nodes.py
@@ -1,6 +1,7 @@
 from __future__ import annotations
 
 import json
+import logging
 from typing import Any
 
 import comfy.model_management as model_management
@@ -16,6 +17,8 @@ from .kj_loader import (
 )
 from .residency import REGISTRY
 
+_LOG = logging.getLogger(__name__)
+
 
 def _entry_report_json(obj: Any) -> str:
     entry = REGISTRY.entry_for_object(obj)
@@ -63,14 +66,20 @@ def _evict_patcher(patcher, *, unpatch_weights: bool) -> bool:
         loaded_model = getattr(loaded, "model", None)
         if loaded_model is None:
             continue
-        if loaded_model is patcher or loaded_model.is_clone(patcher):
+        safe_is_clone = False
+        if loaded_model is not patcher:
+            try:
+                safe_is_clone = loaded_model.is_clone(patcher)
+            except Exception as exc:
+                _LOG.warning("GPU Resident Loader: failed to evaluate clone state during eviction: %s", exc)
+        if loaded_model is patcher or safe_is_clone:
             fully_unloaded = unload_loaded_model(
                 loaded,
                 active_device=getattr(loaded, "device", None),
                 force_offload_to_cpu=True,
                 unpatch_weights=unpatch_weights,
             )
-            if fully_unloaded:
+            if fully_unloaded and unpatch_weights:
                 try:
                     model_management.current_loaded_models.remove(loaded)
                 except ValueError:
diff --git a/patches.py b/patches.py
index e5023e8..6a11468 100644
--- a/patches.py
+++ b/patches.py
@@ -11,7 +11,11 @@ from typing import Any, Callable
 import torch
 from safetensors import safe_open
 
-from .cleanup import unload_loaded_model
+from .cleanup import (
+    _device_matches as _shared_device_matches,
+    _should_force_cpu_offload,
+    unload_loaded_model,
+)
 from .residency import (
     KIND_CHECKPOINT,
     KIND_CLIP,
@@ -222,18 +226,10 @@ def _device_string(device: torch.device | None) -> str:
 
 
 def _devices_match(device_a: Any | None, device_b: Any | None) -> bool:
-    normalized_a = _normalize_device(device_a)
-    normalized_b = _normalize_device(device_b)
-    if normalized_a is not None and normalized_b is not None:
-        return normalized_a == normalized_b
-    return device_a == device_b
+    return _shared_device_matches(device_a, device_b)
 
 
 def _cpu_offload_required(model: Any, loaded_device: Any | None) -> bool:
-    offload_device = _normalize_device(getattr(model, "offload_device", None))
-    if offload_device is None or offload_device.type == "cpu":
-        return False
-
     current_device = None
     if hasattr(model, "current_loaded_device"):
         try:
@@ -242,10 +238,7 @@ def _cpu_offload_required(model: Any, loaded_device: Any | None) -> bool:
             current_device = None
     if current_device is None:
         current_device = _normalize_device(loaded_device)
-    if current_device is None or current_device.type == "cpu":
-        return False
-
-    return _devices_match(offload_device, current_device)
+    return _should_force_cpu_offload(model, active_device=current_device)
 
 
 @contextlib.contextmanager
@@ -643,8 +636,10 @@ def _wrap_load_models_gpu(func: Callable[..., Any]) -> Callable[..., Any]:
                 try:
                     if not requested.is_clone(loaded_model):
                         continue
-                except Exception:
-                    continue
+                except Exception as exc:
+                    raise RuntimeError(
+                        "GPU Resident Loader: failed to evaluate clone-conflict state before replacement"
+                    ) from exc
                 clone_conflicts.append(loaded)
                 seen_loaded_ids.add(id(loaded))