FCP-2 correction of incomplete statement required. Please review!

Baldwin, Dave Dave.Baldwin at emulex.com
Wed Nov 15 10:19:36 PST 2000


* From the T10 Reflector (t10 at t10.org), posted by:
* "Baldwin, Dave" <Dave.Baldwin at emulex.com>
*
Bob,

This looks good to me.

Best regards,
Dave Baldwin

Robert Snively wrote:

> *
> * From the fc reflector, posted by:
> * Robert Snively <rsnively at Brocade.COM>
> *
> The problem is that FCP_RESID goes both ways, depending
> on the state of the overrun and underrun bits.  As a result,
> I have the choice of merely stating that it is necessary to
> perform the appropriate calculation or providing a complete
> analysis of all possible states of the calculation.  At
> present, the detailed analysis is actually provided in
> 9.4.8.  I could provide a reference to that section at
> the appropriate place.  The paragraph would then read:
>
>         If the amount of data returned does not match
>
>    (delete: FCP_DL,)
>
>    add:>the number of bytes of data calculated from the
>         value of FCP_DL and the value of FCP_RESID
>           (see 9.4.8), <
>
>         the error detection and recovery procedure described
>         in clause 12 may be invoked or the FCP I/O operation
>         may be terminated with a recovery abort or other failure
>         indication. The mechanism an initiator uses to determine
>         that the correct amount of data has been returned is
>         vendor specific. Data that has been retransmitted and
>         overlaid shall be counted only once for the purposes of
>         calculating residual values.
>
> -----Original Message-----
> From: Baldwin, Dave [mailto:Dave.Baldwin at emulex.com]
> Sent: Tuesday, November 14, 2000 6:05 PM
> To: Robert Snively
> Cc: Fibre Channel T11 reflector (E-mail); T10 Reflector (E-mail)
> Subject: Re: FCP-2 correction of incomplete statement required. Please
> review!
>
> * From the T10 Reflector (t10 at t10.org), posted by:
> * "Baldwin, Dave" <Dave.Baldwin at emulex.com>
> *
> Bob,
>
> How about something a little more specific:
>
> If the amount of data
>
> add:>bytes<
>
> returned does not match
>
>         (delete: FCP_DL,)
>
>  add:>FCP_DL minus FCP_RESID, <
>
>         the error detection and recovery procedure described
>         in clause 12 may be invoked or the FCP I/O operation
>         may be terminated with a recovery abort or other failure
>         indication. The mechanism an initiator uses to determine
>         that the correct amount of data has been returned is
>         vendor specific. Data that has been retransmitted and
>         overlaid shall be counted only once for the purposes of
>         calculating residual values.
>
> Best regards,
> Dave Baldwin
> Emulex Corporation
>
> Robert Snively wrote:
>
> > *
> > * From the fc reflector, posted by:
> > * Robert Snively <rsnively at Brocade.COM>
> > *
> > While performing the final proof reading of FCP-2 and preparing
> > the bi-directional data transfer annex, I found the following
> > discrepancy in clause 9.3 which I believe should be corrected.
> >
> > The next to the last paragraph says:
> >
> >         If the amount of data returned does not match FCP_DL,
> >         the error detection and recovery procedure described
> >         in clause 12 may be invoked or the FCP I/O operation
> >         may be terminated with a recovery abort or other failure
> >         indication. The mechanism an initiator uses to determine
> >         that the correct amount of data has been returned is
> >         vendor specific. Data that has been retransmitted and
> >         overlaid shall be counted only once for the purposes of
> >         calculating residual values.
> >
> > This is really not correct unless the device was expected to
> > transfer FCP_DL amount of data and terminated early
> > because of a link error.  More normally, FCP_DL will not
> > be the amount of data transferred and an appropriate FCP_RSP
> > will be provided with a CHECK CONDITION/INCORRECT LENGTH
> > or with an appropriate overrun or underrun and a proper residual.
> > In those cases, normal operation continues, with a corrected
> > data count calculated from the residual information.
> >
> > I propose that the last paragraph be modified to read:
> >
> >         If the amount of data returned does not match
> >
> >         (delete: FCP_DL,)
> >
> >  add:>the number of bytes of data calculated from the
> >         value of FCP_DL and the value of FCP_RESID, if any,<
> >
> >         the error detection and recovery procedure described
> >         in clause 12 may be invoked or the FCP I/O operation
> >         may be terminated with a recovery abort or other failure
> >         indication. The mechanism an initiator uses to determine
> >         that the correct amount of data has been returned is
> >         vendor specific. Data that has been retransmitted and
> >         overlaid shall be counted only once for the purposes of
> >         calculating residual values.
> >
> > Is this okay with everybody?
> >
> > Bob Snively                        e-mail:    rsnively at brocade.com
> > Brocade Communications Systems     phone:  408 487 8135
> > 1745 Technology Drive
> > San Jose, CA 95110
>
> *
> * For T10 Reflector information, send a message with
> * 'info t10' (no quotes) in the message body to majordomo at t10.org

*
* For T10 Reflector information, send a message with
* 'info t10' (no quotes) in the message body to majordomo at t10.org




More information about the T10 mailing list