On Fri, Jan 29, 2010 at 11:31:06AM -0800, Robert Love wrote: > On Thu, 2010-01-28 at 08:01 -0800, Hannes Reinecke wrote: > > Robert Love wrote: [...] > > > # ls /sys/devices/virtual/net/eth3.101/fcport1/fcfport_2001000deca31e81/fcfabric_0/ > > > fabric_name fcvport_0 fcvport_1 max_npiv_vports npiv_vports_inuse power uevent vport_create vport_delete > > > > > Consistent numbering. Either use '_' as a separator always or for none. > > This mix-up doesn't make sense. > > > > And, why doesn't the fcfport doesn't show up under the fabric? > > One would assume that the fcfport is part of the fabric, too. > > > > And what about the remote ports? Do they not participate in the fabric? > > > > If the 'fcfabric' element is _not_ the actual fabric I would get rid of it altogether as it'll only serve to confuse > > the users. Like just has been happened with me :-) > > > Yeah, they should be under the fcfabric. This work is just incomplete in > that area. Besides the fabric, FC also has the arbitrated loop and point-to-point support. If there is a fcfabric element, would the "point-to-point remote port" be placed under the fcfabric? Or might this be another point for not having the fcfabric element? [...] > > > 9) Is this change worth all of the changes that will need to happen? > > > > > > This is probably the most difficult question. I view this change as > > > positive overall, but it's very heavy lifting. Not only does the > > > FC transport need an overhaul, but it looks like the scsi transport > > > and AC code might need some additions. Every HBA will need to be > > > modified and every HBA's HBAAPI vendor library will need to change > > > too. (I hope this would be a catalyst to move to an OS library that > > > everyone shares) > > > > > > I'd like to hear other opinions about this because this really is > > > going to be something that all HBAs will need to be involved in. The changes required for "all HBAs" also include the zfcp driver, i am looking at this from zfcp's point of view. With zfcp, we only support FCP/SCSI with point-to-point or fabric attached storage, so we don't need as much flexibility as other HBAs. I assume the LLD driver will report the available hardware resources with fc_fcport_add, fc_fcpinit_add or whatever the entry points will be called in the final version. Reporting the new FC elements instead of FC host and SCSI host should be doable for zfcp. -- Christof -- 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
