Streaming Clarification

George_Penokie/Tivoli_Systems%TIVOLI_SYSTEMS at us.ibm.com George_Penokie/Tivoli_Systems%TIVOLI_SYSTEMS at us.ibm.com
Fri Nov 10 12:16:34 PST 2000


* From the T10 Reflector (t10 at t10.org), posted by:
* George_Penokie/Tivoli_Systems%TIVOLI_SYSTEMS at us.ibm.com
*
Sriram,
To answer your original questions:
It is mandatory for the target to use P_CRCA if read streaming is active
and it is not going to disconnect or go bus free after the read data IU. In
the case where there is a stream of read data IUs which is followed by an
L_Q IU without the P_CRCA signal the initiator would not know the L_Q IU
was an L_Q IU. That is why P_CRC is mandatory when doing read streaming.
If the target is going to disconnect or go bus free after the read data IU
then it is not required to use P_CRCA but it could. It depends on the
implementation, but it  works regardless.

As for; your second question on write flow control:
If the target agrees to use write flow control through PPR then it is
required to use P_CRCA to indicate the last IU in the write stream before
the end of the IU. This is to give the initiator an early warning that the
current I_T_L_Q nexus is about to end. As I understand it some initiators
can take advantage of this early warning to improve system throughput
performance. And yes write streaming always ends with a phase change or a
bus free. In packetized there is no other way to go from a single write
data IU or a stream of write data IUs to any other type of IU.

Bye for now,
George Penokie

Dept 2C6  114-2 N212
E-Mail:    gpenokie at tivoli.com
Internal:  553-5208
External: 507-253-5208   FAX: 507-253-2880


Sriram Srinivasan <srirams at lsil.com>@t10.org on 11/09/2000 04:50:00 PM

Please respond to Sriram Srinivasan <srirams at lsil.com>

Sent by:  owner-t10 at t10.org


To:   t10 at t10.org
cc:
Subject:  Streaming Clarification



* From the T10 Reflector (t10 at t10.org), posted by:
* Sriram Srinivasan <srirams at lsil.com>
*
   I have a question on "how to end a read stream".

   00-378r0 (SPI-4 rev1) states:

   1) in the last paragraph (and NOTE-5) of section 4.9.3.3 DT_DATA_IN
      flow control is implied by the term "read streaming"

   2) in section 16.3.10.1 that RD_STRM must be negotiated to a 1 so
      that read streaming can be used.

   So putting 1) and 2) together can it be deduced that the initiator
will EXPECT "DT_DATA_IN flow control" (using P_CRCA) to happen during
read streams?  In other words:

   a) Is it mandatory for the target to use the P_CRCA to end the read
stream?  Or

   b) can it do a phase change at the end of a data stream IU and end
the stream (as stated in paragraph 6 of section 14.3.4 in 00-378r0)?

   c) Or is the target allowed to do either a) or b) stated above?

   thanx,
   \Sriram\


----------------------------------------------------------------------

 Sriram Srinivasan                       Sriram.Srinivasan at lsil.com
 ASIC Design Engineer, LSI Logic,
 2001 Danfield Ct.,                      Phone: 970-206-5847
 Fort Collins, CO 80525                  FAX  : 970-206-5244
----------------------------------------------------------------------

*
* 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