LVD SCSI Question

Tony Priborsky Tony_Priborsky at maxtor.com
Tue Oct 15 07:51:46 PDT 1996


* From the SCSI Reflector (scsi at symbios.com), posted by:
* "Tony Priborsky" <Tony_Priborsky at maxtor.com>
*
George and all:

Your note about hot-plugging and managing the states of the dual-mode LVD 
transceivers was timely.  I was going to post a similar question myself.

I'd had a discussion with Bill Ham about this, and I too was wondering how 
to handle the hot-plugging of a legacy HVD or SE device.  This causes the 
DIFFSENS line to transition in a manner that affects the entire bus.  The 
scenario you mentioned about a device device having negotiated for FAST-40 
transfers suddenly being pulled to single-ended mode was the question I 
posed to Bill.   After thinking about it a bit, it became clear to me that 
device (and probably host adaptor) firmware must actively monitor the 
state of the DIFFSENS signal and react to changes in it.   SPI-2 does 
specify some debounce intervals, but doesn't say what to do if it changes 
after initial power up.  I was going to suggest that one way to handle it 
would be to treat the event as a kind of SCSI bus reset, and have the 
devices affected process it as such (i.e., reverting to 8-bit, asynch 
mode).  A new Unit Attention ASC/ASCQ pair would indicate that the bus 
configuration had changed.

It may be possible to silently change modes under some circumstances, but 
the debounce intervals make it difficult to think about doing it with 
concurrent I/O processes.

I was also wondering how terminators would deal with this event.  If there 
are going to be dual-mode terminators on the drives, firmware may need to 
change their mode or protect themselves too (unless they are crafted to do 
so automatically), and standalone terminators need to change or protect 
themselves automatically.

Can SPI-2 be considered complete until this issue is addressed, especially 
with respect to this hot-plugging scenario?

Cheers,
Tony Priborsky
Sr. Staff Engineer
Maxtor Corp
2121 Miller Dr.
Longmont, CO  80501
303/682-4880

----------------------[Reply - Original Message]----------------------

Sent by:"George Penokie" <GOP at rchvmp3.vnet.ibm.COM>
 * From the SCSI Reflector (scsi at symbios.com), posted by:
* "George Penokie" <GOP at RCHVMP3.VNET.IBM.COM>
*
In SCSI-2 and SPI the differential dirver protection is a simple circuit
that outputs directly into the differential drivers enable pin.  The
wording and the circuit (SPI only) are very clear as to how the protection
works.  It is basicly an on/off switch and when off the device will just
not be able to access the bus.

Now enter LVD where the on/off switch become an on/on/off switch and there
is no longer a clear indication as to what to do with the three resulting
states.

One can only assume the hardware would have logic that would continuously
monitor the three signals as follows.
-The HVD signal would cause all the drivers to be disabled in the
same fashion as in SCSI-2 and SPI.
-The LVD signal would cause the LVD drivers to be enabled and the single
ended driver to be disabled.
-The single ended signal would cause a the single ended drivers to be 
enabled
and the LVD drivers to be disabled.

So far so good. One could argure that, based on the descriptions in SPI,
that is the only logical conclusion.

But there is more: To support LVD and single ended operations the software
must be involved.  In SCSI-2 and SPI how the software handled the mixing
was not importance because the drivers where turned off so the device
could not receive or send anything on the bus.  This not the case for
LVD, however, where the device must operated properly in both a
single ended mode and a LVD mode.  In other words the devices software
must immediatly know when there is a mode change so it can transfer
information correcly on the bus.

The problem is the current description in SPI-2 could lead you to think
these mode changes will only occur after a reset or power on.  This is
NOT the case.  A device could be hot-plugged without a corresponded
reset or POR occurring. If this occured and the new device was single 
ended
the drivers would switch but the software would not and you could end up
transfering fast-40 data on a single ended bus (not good).

I belived that SPI-2 should make it clear that the DIFFSENS line
can change at any time and the devices hardware and software must 
immediately
respond to any of those changes.

Any comments???

Bye for now,
George Penokie
*
* For SCSI Reflector information, send a message with
* 'info scsi' (no quotes) in the message body to majordomo at symbios.com

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




More information about the T10 mailing list