Hi Xunlei, Xunlei Pang writes:
Add cond_resched() at the upper shrink_node_memcgs() to solve this issue, and any other possible issue like meomry.min protection.
I understand the general intent of the patch, but could you clarify your concern around memory protection?
