True - and we've been dancing around target busy as well. Basically, if there's an HCIL, any component of the HCIL could potentially be busy. And, stirring the pot, there could be different queuing/resource limits at each point as well. -- james > -----Original Message----- > From: linux-scsi-owner@xxxxxxxxxxxxxxx > [mailto:linux-scsi-owner@xxxxxxxxxxxxxxx]On Behalf Of Jeff Garzik > Sent: Wednesday, November 16, 2005 10:51 AM > To: SCSI Mailing List > Subject: channel busy > > > It would be nice if the SCSI core supported the concept of "channel > busy". SCSI understands host-busy and device-busy. > > Right now libata allocates one scsi host per ATA port largely due to > SCSI implementation details, and I would like to get away from that. > > Jeff > > > - > : 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