ACA handling for "temporary initiators

Charles Monia, SHR3-2/W3, 237-6757 14-Jun-1994 1408 monia at
Tue Jun 14 11:06:21 PDT 1994

Jeff Williams wrote:
I would like to avoid changing the rules to accommodate this command.

To me, this is an issue analogous to a drive getting a read command and
having some error on the mechanism.  It is up to the target controller
to clean up any error in the mechanism, not up to the initiator.  I would
use the same approach here.  The initiator gave the target/temp initiator
a command to execute which just so happens to propogate into another SCSI
command.  From the initiator's point of view, it is one command.  The 
target/temp initiator is responsible for any subsequent clean up on the
XPWRITE.  The initiator shouldn't even get involved.  If it wanted to be
involved, it should have issued XDWRITE, XDREAD, and XPWRITE by itself.

I agree with this position.


More information about the T10 mailing list