On Fri, Aug 26, 2011 at 02:17:31PM -0700, Roland Dreier wrote: > Fix this by using queue_work(system_nrt_wq, ...), which uses a > non-reentrant work queue (so the work only runs on one CPU at a time). > Also, while we're working on this code, avoid dropping and reacquiring > the lock to walk the list by splicing the list entries onto a local > list and then operating on that. Please use a local workqueue. Tejun has been trying to get rid of the magic sysem workqueues for a while. I'd also split the splicing into a separate patch - in fact the splice alone should fix the issue for now. -- To unsubscribe from this list: send the line "unsubscribe target-devel" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
