Skip site navigation (1)Skip section navigation (2)
Date:      Fri, 8 Sep 2000 14:54:09 -0700 (PDT)
From:      Matthew Jacob <mjacob@feral.com>
To:        "Kenneth D. Merry" <ken@kdm.org>
Cc:        Wilko Bulte <wkb@freebie.demon.nl>, freebsd-scsi@FreeBSD.ORG
Subject:   Re: suggestion for camcontrol
Message-ID:  <Pine.LNX.4.21.0009081450380.15724-100000@zeppo.feral.com>
In-Reply-To: <20000908134520.A20744@panzer.kdm.org>

next in thread | previous in thread | raw e-mail | index | archive | help

The points about cleanliness are worthy- except the deal here is that the
interfaces have to be clean. The program 'camcontrol devlist' does not.

You should be able to do XPT_PATH_INQ for paths 0 through 0xff to get at the
list of SIMs.

> On Fri, Sep 08, 2000 at 21:19:21 +0200, Wilko Bulte wrote:
> > Would it be difficult for camcontrol to display the initiator ID of the
> > hostadapter in addition to the devices in the output of devlist?
> > 
> > Today:
> > 
> > miata#camcontrol devlist
> > <SEAGATE ST34371W 0702>            at scbus0 target 0 lun 0 (pass0,da0)
> > <IBM DDRS-34560W S97B>             at scbus0 target 1 lun 0 (pass1,da1)
> > <IBM DDRS-39130D DC1B>             at scbus0 target 2 lun 0 (pass2,da2)
> > 
> > Suggested example:
> > 
> > miata#camcontrol devlist
> > <SEAGATE ST34371W 0702>            at scbus0 target 0 lun 0 (pass0,da0)
> > <IBM DDRS-34560W S97B>             at scbus0 target 1 lun 0 (pass1,da1)
> > <IBM DDRS-39130D DC1B>             at scbus0 target 2 lun 0 (pass2,da2)
> > <Qlogic 1040 isp0>		   at scbus0 target 7
> > 
> > I'm currently playing with shared SCSI buses and in situations like that it
> > is really practical if you can get a display of the complete SCSI bus.
> 
> Well, the initiator isn't currently included in the CAM EDT, and I'm not
> sure that it would be the right thing to do that.  You'd need the initiator
> to be in the EDT in order for printing it with 'camcontrol devlist' to just
> work.  (Well, it wouldn't quite "just work", you might need some additional
> printing logic.)
> 
> The most likely place for that information to go would be the Path Inquiry
> CCB.  In fact, the initiator ID is already in that CCB, it just isn't
> displayed by 'camcontrol negotiate da0 -v'.  I've attached a patch to do
> that, here's what you get:
> 
> # ./camcontrol negotiate da0 -v
> Current Parameters:
> (pass0:ahc0:0:0:0): sync parameter: 12
> (pass0:ahc0:0:0:0): frequency: 20.000MHz
> (pass0:ahc0:0:0:0): offset: 8
> (pass0:ahc0:0:0:0): bus width: 16 bits
> (pass0:ahc0:0:0:0): disconnection is enabled
> (pass0:ahc0:0:0:0): tagged queueing is enabled
> ahc0: SIM/HBA version: 1
> ahc0: supports tag queue messages
> ahc0: supports SDTR message
> ahc0: supports 16 bit wide SCSI
> ahc0: HBA engine count: 0
> ahc0: maximum target: 15
> ahc0: maximum LUN: 64
> ahc0: highest path ID in subsystem: 0
> ahc0: initiator ID: 7
> ahc0: SIM vendor: FreeBSD
> ahc0: HBA vendor: Adaptec
> ahc0: bus ID: 0
> ahc0: base transfer speed: 3.300MB/sec
> 
> As far as putting it in the devlist output, we could kludge some of the
> information in there, but the adapter model number isn't currently included
> in the path inquiry.
> 
> It might be possible to add a new node type into the EDT, or perhaps to
> represent the initiator as a target device, but that would require adding
> target matching capability.  We could add the initiator as a regular
> device (lun), but it wouldn't have many of the things associated with a
> lun/device node, like inquiry data.
> 
> Anyway, the bottom line is that it can be done, but doing it in a clean
> manner isn't trivial.  I think it would need some discussion to figure out
> whether there should be a node for the initiator in the EDT, and if so,
> what sort of node it should be.
> 
> Ken
> 



To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-scsi" in the body of the message




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?Pine.LNX.4.21.0009081450380.15724-100000>