about the SMP functions behavior on SAS-1.1
Elliott, Robert (Server Storage)
elliott at hp.com
Fri Mar 26 12:50:01 PST 2004
* From the T10 Reflector (t10 at t10.org), posted by:
* "Elliott, Robert (Server Storage)" <elliott at hp.com>
> * From the T10 Reflector (t10 at t10.org), posted by:
> * Shinji Ugawa <ugawa-sxa at necst.nec.co.jp>
> When should the expander initialize or forget the following
> parameters those are set by the SMP PHY CONTROL function?
> - PROGRAMMED MINIMUM PHYSICAL LINK RATE
> - PROGRAMMED MAXIMUM PHYSICAL LINK RATE
> - PARTIAL PATHWAY TIMEOUT VALUE
> -- Initialize events ---------+--- e.g. PROG. MIN. PHYS. LINK RATE ---
> Power on reset | initialize
> HARD_RESET primitive received | initialize, forget the previous
> PHY OPERATION = 'LINK RESET' | still keep the value
> PHY OPERATION = 'HARD RESET' | initialize, forget the previous
> PHY OPERATION = 'DISABLE' | still keep the values
The programmed physical link rates and partial pathway
timeout value are only reset back to their defaults by
Expanders themselves don't care about HARD RESETs. See 4.5:
"If the port is an expander port, the expander function
and other expander ports in the expander device shall
not be affected by hard reset. "
> /// comments on the SMP state machine descriptions.
> A figure is drawn based on SAS-1.1 r03 for the SMP target port.
> So, this figure has a few arrows without a destination (requests and
> confirmation). My comments are as the followings.
> 1. SMP_TP2: Frame Transmitted confirmation
> The Frame Transmitted confirmation send from SMP_TP2 is in
> the Figure 106. But I could not find a description for this
> confirmation in 18.104.22.168.3.
You are correct, it should be in the text too but is missing. It
gets sent once a Frame Transmitted message is received from the
SMP transmitter. I will correct that in sas1r05.
> 2. PL_PM3: SMP Transmit Break request
> The SMP Transmit Break request send from PL_PM3 is in the
> Figure 110. But I could not find a description for this request
> in 22.214.171.124.1.
The PL_OC state machine actually handles that, not PL_PM. Section
126.96.36.199.5 notes that PL_OC2 passes along the SMP Transmit Break
|from the SMP transport layer to the SMP link layer.
> 3. in SAS-1.1 r03 page 257, in SAS-1.1 r04 page 259.
> 188.8.131.52.5 PL_OC2:Overall_Control state connection management
> If this state receives an Accept_Reject Opens request,
> then this state shall send an Accept_Reject Opens message
> to all phys in the port.
> I guess that the "Accept_Reject Opens message to all phys"
> is misprint for the "Accept_Reject Opens request to all link layers".
> I referred to a table J.11 in an annex J.
Phy is the architectural term for the entity that contains link layer
state machines and phy layer state machines.
See figure 23 (State machines for SAS devices) and figure 30
(State machines and SAS device, SAS port, and SAS phy objects).
> 4. in SAS-1.1 r03 page 259, in SAS-1.1 r04 page 261.
> If this state is in an SMP target port, then this state
> shall send the Tx Frame message containing the SMP REQUEST
> frame to the PL_PM state machine on which the connection was
> established for the Tx Open message.
> I thought that the "SMP REQUEST frame" should change to
> the "SMP RESPONSE".
Correct. I will fix that in sas1r05.
> Shinji Ugawa
> NEC System Technologies, Ltd.
Rob Elliott, elliott at hp.com
Hewlett-Packard Industry Standard Server Storage Advanced Technology
* 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