On Sun, 2006-03-12 at 10:06 -0600, James Bottomley wrote: > This patch moves the discovery and event processing entirely over to > being workqueue based. Additionally, it tries to ensure discovery gets > time to proceed before finishing the module load. There's still a > non-determinism in there since it starts with a port interrupt for a > connected phy after phys are enabled, so there's a ssleep(1) in there to > capture this before the scsi_flush_work() waits for it to complete. Just wanted to let you know that this patch has passed an overnight reboot test on a x366. The ssleep(1) seems to allow for enough time in order to verify that sas_process_events has been called before waiting on discovery. > > James > Regards, Alexis > > > > - > : send the line "unsubscribe linux-scsi" in > the body of a message to majordomo@xxxxxxxxxxxxxxx > More majordomo info at http://vger.kernel.org/majordomo-info.html - : send the line "unsubscribe linux-scsi" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
