SAS rate matching - ALIGNs in OPEN

Seto, Pak-lung pak-lung.seto at intel.com
Thu Feb 20 10:16:29 PST 2003


* From the T10 Reflector (t10 at t10.org), posted by:
* "Seto, Pak-lung" <pak-lung.seto at intel.com>
*
Processing OPEN Address Frame is a store and forward type of operation (it
may be possible to do cut-thru), why rate matching is required?
As Rob said, the ALIGNs are deleted and added back later (I don't think it
need to add in this case).  I don't see what is the reason to add these
unnecessary restrictions into the protocol.

Pak

-----Original Message-----
From: Elliott, Robert (Server Storage) [mailto:Elliott at hp.com]
Sent: Wednesday, February 19, 2003 5:55 PM
To: Seto, Pak-lung; Hoglund, Tim; T10 Reflector
Subject: RE: SAS rate matching - ALIGNs in OPEN


The important rule is that rate matching begins sometime during the first
idle dwords - it must already be rate matching when the expander forwards an
OPEN_ACCEPT.

There's nothing special about rate matching ALIGNs vs. other ALIGNs.  They
don't get forwarded through an expander like normal dwords - they're deleted
and added back as needed.

The January SAS WG thought it would be better to specify exactly when the
rate matching starts to reduce testing variety.

Nobody is commenting on whether I should remove that sentence about rate
matching during the OPEN or not.

If I go with Brian's wording:
Immediately after transmitting the EOAF of the OPEN address frame, the
source phy shall start rate matching at the selected connection rate.

George will complain about "immediately" and this prohibits rate matching
during the OPEN itself (do we want to prohibit that?).

--
Rob Elliott, elliott at hp.com
Hewlett-Packard Industry Standard Server Storage Advanced Technology
https://ecardfile.com/id/RobElliott




> -----Original Message-----
> From: Seto, Pak-lung [mailto:pak-lung.seto at intel.com] 
> Sent: Wednesday, February 19, 2003 4:22 PM
> To: 'Hoglund, Tim'; T10 Reflector
> Subject: RE: SAS rate matching - ALIGNs in OPEN
> 
> 
> * From the T10 Reflector (t10 at t10.org), posted by:
> * "Seto, Pak-lung" <pak-lung.seto at intel.com>
> *
> I agree and I don't know what is the reason to put in such 
> requirements?
> 
> Pak
> 
> 
> 
> -----Original Message-----
> From: Hoglund, Tim [mailto:thoglund at lsil.com]
> Sent: Wednesday, February 19, 2003 12:19 PM
> To: T10 Reflector
> Subject: Re: SAS rate matching - ALIGNs in OPEN
> 
> 
> * From the T10 Reflector (t10 at t10.org), posted by:
> * "Hoglund, Tim" <thoglund at lsil.com>
> *
> I do not agree with imposing a requirement to send a 
> rate-matching ALIGN immediately following EOAF - this is a 
> needless constraint on implementations.
> 
> The OPEN address frame is not delivered through the topology, 
> rather is store/forward through expanders, therefore it 
> should not matter whether the rate matching ALIGNs begin as 
> EOAF/ALIGN/idle dword or as EOAF/idle dword/ALIGN.
> 
> TimH
> 
> ~~~~~~~~~~~~~~~~~~~~~~~
> Tim.Hoglund at lsil.com
> 719/533-7450 (Voice)
> 719/533-7480 (Fax)
> ~~~~~~~~~~~~~~~~~~~~~~~
> 
> 
> 
> -----Original Message-----
> From: Elliott, Robert (Server Storage) [mailto:Elliott at hp.com]
> Sent: Tuesday, February 18, 2003 2:40 PM
> To: t10 at t10.org
> Subject: SAS rate matching - ALIGNs in OPEN
> 
> 
> * From the T10 Reflector (t10 at t10.org), posted by:
> * "Elliott, Robert (Server Storage)" <Elliott at hp.com>
> *
> sas-r03 7.15 Rate matching says:
>     The OPEN address frame itself may or may not have ALIGNs 
> interspersed.
> 
> A letter ballot comment asked that we be very precise in 
> mandating that the first dword after the OPEN address frame 
> shall be an ALIGN,
> yielding:
> 
>     After transmitting the OPEN address frame, the source phy 
> shall start rate matching at the selected connection rate. 
> The first dword after the EOF shall be an ALIGN or NOTIFY for 
> rate matching. 
> 
> Should I keep or delete the sentence about ALIGNs (and 
> NOTIFYs) in the OPEN address frame?  It's true, but might 
> lead someone to think they should start rate matching in the 
> OPEN (there's probably no reason to do so).
> 
>  
> --
> Rob Elliott, elliott at hp.com
> Hewlett-Packard Industry Standard Server Storage Advanced 
> Technology https://ecardfile.com/id/RobElliott
> 
> 
> 
> *
> * 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
> *
> * 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