On Mon, Jun 20, 2011 at 1:37 PM, Christoph Hellwig <hch@xxxxxxxxxxxxx> wrote: > We really shouldn't require workaround for this in every SCSI driver. > > scsi_scan_host_selected already makes sure never to scan above > shost->max_channel. Could it be that you don't have that one set > properly in the srp driver? It not we'll need to debug why it happens > instead of adding hacks like this. What does scsi_device_lookup_by_target() actually do ? As far as I can see the SCSI command REPORT LUNs reports all LUNs that match the pattern H:*:*:* but scsi_device_lookup_by_target() only looks for already existing LUNs in the subrange H:C:I:*. Hence the duplicates that are created during scanning. If my analysis is correct, the bug is in the SCSI scanning code and not in the ib_srp nor in the ibmvscsi driver. The least one can say is that the term "lun" has at least two entirely different meanings inside the SCSI core without clear comments about which meaning it has in which context. Bart. -- To unsubscribe from this list: 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
