SBP-2 src field of status block

Wink Saville wink at saville.com
Thu Oct 23 13:08:37 PDT 1997


* From the T10 (formerly SCSI) Reflector (t10 at symbios.com), posted by:
* "Wink Saville" <wink at saville.com>
*
I think there might be a mistake in section 9.3 or 5.4. According to section
5.4 when the status block src field is 1 the next_ORB "was null" when this
just completed ORB was read. From my perspective when src in the status
block field is 1, the state of the fetch engine may be suspended and that
the agent.orb_pointer field may still pointing at the just completed ORB. To
me this implies, that the next field of the ORB just completed may be
re-read in the future if and when new ORB's are appended to this task set
and the doorbell register is written.

If the above is true, then when an task completes with src equal to one it
may not be reused and if it's zero it may be reused. This is the contrary to
the last sentence of the last paragraph of section 9.3 which reads: "if the
src field has a value of one, the initiator may reuse or deallocate the
system memory occupied by an ORB." I believe it should read: "If the src
field has a value of zero, the initiator may reuse or deallocate the system
memory occupied by an ORB."

I may be totally screwed up, in which case I'd like some help or I may have
found a minor mistake in the documentation.


Wink


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




More information about the T10 mailing list