This message is from the T13 list server.

In this revision, there are a few unresolved technical issues:

1. T,UI is still incorrect (670.12 should be 670.23) (volume 3 section
14.4.10 page 62 table 11)
2. The HOLDA 20 dword response rule doesn't properly account for
outgoing ALIGNs (volume 3 section 15.4.8.1 page 124)
3. The HOLDA 20 dword response rule doesn't properly account for
outgoing HOLDs (volume 3 section 15.4.8.1 page 124)
4. The HOLD/HOLDA transmit 20 DWORDs/receive 20 DWORDs discrepancy still
exists (volume 3 section 15.4.8.1 page 124)


More details (all of these have been posted to the T13 reflector):
1. SATA 1.0 defined the frequency skew as +350/-2650 ppm, noting it
included +350/-150 ppm for the crystal and -0.25% [= -2500 ppm] fixed
skew for SSC but excluded +/-0.25% [= +2500/-2500 ppm] AC modulation
from SSC. That adds up to +2850/-5150 ppm. 

SATA 1.0a defines the frequency skew as +/- 350 ppm for DC skew and
+0/-5000 ppm for AC skew (adding up to +350/-5350 ppm). The minimum T,UI
is not based on +2850, it's based on +350 ppm:
666.666 * (1 - .00285) = 664.76666
666.666 * (1 - .000350) = 666.4333 
so the SATA 1.0a change makes the minimum agree.

The maximum of 670.12 ps is based on the old SATA 1.0 -5150 ppm rather
than the 1.0a -5350 ppm:
666.666 * 1.005150 = 670.10 (SATA 1.0 maximum, with 2 fs rounding
discrepancy)
666.666 * 1.005350 = 670.23 (SATA 1.0a maximum)

For SATA 1.0a and ATA/ATAPI-7, it should be 670.23 ps.


2. ALIGNs sent by the frame transmitter are not supposed to count
against this budget, although they do consume a "DWORD time." Sending 10
data dwords, 2 ALIGNs, and 10 data dwords would technically violate the
above statement but not really be an error.


3. HOLDs sent by the frame transmitter cannot always count against this
budget. Assume the frame transmitter is in LT4:L_SendData and it finds
it has no more data to send, so it moves to LT6:L_SendHOLD and starts
transmitting HOLDs. It remains in that state as long as it has no more
data to send. If a HOLD shows up, it does not leave the state and switch
to transmitting HOLDAs; it continues sending HOLDs. Only when it decides
it has no more data to send does it switch states; it goes to
LT5:L_RcvrHold until the incoming HOLDs go away.

Unlike ALIGNs, the HOLDs cannot just be excluded from the budget. The
frame transmitter won't ever leave the HOLD state to send a few more
data dwords before sending the HOLDA in reply to a HOLD it received
while it was transmitting HOLD itself; it always goes directly to
sending HOLDA. 


4. The 20/20 transmit/receive rule, based on a mythical wire delay of 0,
doesn't make sense. It should either be 19/20 or 20/21 (I prefer 19/20;
Knut preferred 20/21). If a transmitter takes n dwords to reply to HOLD
with HOLDA, the receiver is going to see the reply n + 1 dwords after it
sent the HOLD, not n dwords later.

Incorporating #2, #3, and #4 (assuming 20/21 is chosen), I suggest this
text replace most of 15.4.8.1: 
"Not including ALIGNs, the frame transmitter shall transmit no more
than: 
a) 20 data dwords; or 
b) 20 or fewer data dwords following by one or more HOLD primitives, 
after receiving a HOLD primitive before transmitting HOLDA primitives. 

Not including ALIGNs, the frame receiver shall accept at least: 
a) 21 data dwords; or 
b) 21 or fewer data dwords following by one or more HOLD primitives, 
after transmitting a HOLD primitive before expecting HOLDA primitives." 

-- 
Rob Elliott, [EMAIL PROTECTED] 
Hewlett-Packard Industry Standard Server Storage Advanced Technology 
https://ecardfile.com/id/RobElliott 


-----Original Message-----
From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of John
Masiewicz
Sent: Monday, April 05, 2004 3:30 PM
To: [EMAIL PROTECTED]
Subject: [t13] ATA-7 rev 4a posted - Letter Ballot Comments
incorporation completed


T13, 
As Jim has already noted, the ATA-7 Rev 4a has been posted under the
drafts section on the T13 website.  This version has all of the Letter
Ballot Responses incorporated into all three volumes. There were
hundreds of editorial changes (formatting, links, punctuation, etc), as
well as a lot of clean-up to improve the indexes, tables, and annexes.
The PDF versions are now getting very good, with usable links and
indexes as requested. The PDF files are also a lot smaller. I think that
this is due to better formatting and PDF option settings. As far as I
can tell, nothing is lost and it seems to work much better. Please let
me know if there are any issues.
The list of contributors is the initial list provided by Dan Colgrove,
Both Dan and I have received additional names to add to the list. If
there is another revision, I will add them at that time, otherwise, they
will go to the ANSI editor for addition prior to the final publication.
I also incorporated the additional front pages content that the ANSI
editors want, as provided by Dan. 
If there are new comments, please use the Adobe comment tool to make
comments in each document. The documents can the be exported into a
comment file. This is the method we used in the last plenary meeting,
and it worked rather well. 
We are hoping this is the last version for ATA-7. 


John Masiewicz 
T13 Editor 
Western Digital 
949-672-7686 

Reply via email to