Re: [PATCH 2.6.13 5/14] sas-class: sas_discover.c Discover process (end devices)

[James Bottomley <James.Bottomley@SteelEye.com>]

On Fri, 2005-09-09 at 15:40 -0400, Luben Tuikov wrote:
> +/**
> + * sas_do_lu_discovery -- Discover LUs of a SCSI device
> + * @dev: pointer to a domain device of interest

Aside from all the other problems, this one completely duplicates the
mid-layer infrastructure for handling devices with Logical Units.

> + * Discover logical units present in the SCSI device.  I'd like this
> + * to be moved to SCSI Core, but SCSI Core has no concept of a "SCSI
> + * device with a SCSI Target port".  A SCSI device with a SCSI Target
> + * port is a device which the _transport_ found, but other than that,
> + * the transport has little or _no_ knowledge about the device.
> + * Ideally, a LLDD would register a "SCSI device with a SCSI Target
> + * port" with SCSI Core and then SCSI Core would do LU discovery of
> + * that device.

That would be what scsi_scan_target() actually does.

> + * REPORT LUNS is mandatory.  If a device doesn't support it,
> + * it is broken and you should return it.  Nevertheless, we
> + * assume (optimistically) that the link hasn't been severed and
> + * that maybe we can get to the device anyhow.

That's a surprisingly optimistic statement from someone who claims to
have worked in SCSI for so long.  We have a huge list of heuristics for
devices that violate the standards in one way or another.  We already
have a flag for a SCSI3 device that doesn't respond correctly to
REPORT_LUNS ... and we have a few other reports of potentially more
suspect devices.

Now, if you did this properly and used the mid-layer infrastructure you
wouldn't have to worry about any of this.

> +static int sas_do_lu_discovery(struct domain_device *dev)

Please just handle targets ... scanning beyond targets is best handled
in generic code.

James