wangbo/ComfyUI
1
0
Fork 0
mirror of https://github.com/comfyanonymous/ComfyUI.git synced 2026-07-03 13:19:23 +08:00
Code Issues Actions 21 Packages Projects Releases Wiki Activity

Gate static pin_memory on the host-RAM budget (fixes #13730)

Browse Source 
On AMD/ROCm a clean launch stalls at "Requested to load LTXAV" while system RAM
fills and spills to swap, even though VRAM sits at ~65%. It is host-side
pinned-memory exhaustion, not VRAM pressure.

partially_load() pins every offloaded weight via the static path pin_memory,
which ignores ensure_pin_registerable()'s result and unconditionally
cudaHostRegisters up to MAX_PINNED_MEMORY (0.90*RAM on Linux). Those pins are
only reclaimable from is_dynamic() models by free_registrations, and dynamic
VRAM is off by default on AMD, so they are never reclaimable. Page-locked RAM
is not swappable, so the loader exhausts RAM and thrashes.

The dynamic-VRAM pin path (comfy/pinned_memory.py) already guards this with
ensure_pin_budget/ensure_pin_registerable; only the static path was missing it.
Gate pin_memory the same way and skip pinning when the budget cannot be met
(the weight stays in pageable RAM, still correct, just not page-locked).
Behavior is unchanged when RAM is ample and under --high-ram.
This commit is contained in:
liminfei-amd 2026-06-17 22:08:49 +08:00
parent a590d60bb1
commit 8d8e0c68b0
1 changed files with 8 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

9
comfy/model_management.py
View File

@ -1534,7 +1534,14 @@ def pin_memory(tensor):
size = tensor.nbytes
comfy.memory_management.extra_ram_release(comfy.memory_management.RAM_CACHE_HEADROOM)
ensure_pin_registerable(size)
# Respect the host-RAM budget like the dynamic-VRAM pin path (comfy/pinned_memory.py)
# already does. Without this gate the static load path keeps cudaHostRegister-ing
# offloaded weights toward MAX_PINNED_MEMORY (90% of RAM on Linux) regardless of how
# little RAM is actually free, so unswappable pages fill RAM+swap and large model
# loads stall (issue #13730). When the budget cannot be met, skip pinning and leave
# the weight in pageable RAM (still correct, just not pinned).
if not ensure_pin_budget(size) or not ensure_pin_registerable(size):
return False
ptr = tensor.data_ptr()
if ptr == 0: