LR> GFP_KERNEL is not allowed here, since from->lock is locked. Not LR> sure that GFP_ATOMIC is acceptable though. Perhaps it would be LR> better to temporarily move the queue to a local head, copy it (no LR> spinlock needed), and then push it again. This would need to LR> block concurrent senders/receivers during this operation, unless LR> it's guaranteed that they are all frozen.
Ah, yeah, good catch. I'll have to think about that a little. Thanks! -- Dan Smith IBM Linux Technology Center email: [email protected] _______________________________________________ Containers mailing list [email protected] https://lists.linux-foundation.org/mailman/listinfo/containers _______________________________________________ Devel mailing list [email protected] https://openvz.org/mailman/listinfo/devel
