Thanks a lot, Bill. That was the implication of what I'd read, that an
instruction called SIGA, did the I/O. They didn't use the word 'command' as
would normally be used with a CCW. There's also this diagram that shows that
it's quite a different flow:


[image: Non-QDIO versus QDIO data paths]

Regards,
John.


On 4 November 2010 20:04, Bill Fairchild <[email protected]> wrote:
> I didn't mean to imply that SIGA is not a machine instruction.  All I said
was that SSCH is a machine instruction and is not a CCW operation.
>
>
>
> The following also is evidence that SIGA is a machine instruction which
causes an I/O operation to start in the Channel Subsystem, just as will
happen with SSCH:
>
>
http://publib.boulder.ibm.com/infocenter/zos/v1r9/index.jsp?topic=/com.ibm.zos.r9.ieav100/stio.htm
>
>
>
> This page shows the format of system trace entries for the CSCH, HSCH,
..., SSCH, SIGA, ... trace entries, confirming that SIGA is a machine
instruction.  It is also mentioned in the Diagnosis Tools and Service Aids
book under System Trace but not in the section on GTF.
>
>
>
> The opcode and some operands are shown on this page:
>
> http://www.cbttape.org/~jjaeger/opcodes.html#OSA
>
>
>
> That's all I know about it.  This should give you a good place to start
for further research.
>
>
>
> Bill Fairchild
>
> Rocket Softare
>
>
>
> -----Original Message-----
> From: IBM Mainframe Discussion List [mailto:[email protected]] On
Behalf Of John Blythe Reid
> Sent: Thursday, November 04, 2010 12:27 PM
> To: [email protected]
> Subject: Re: QDIO question
>
>
>
> This is the bit that made me think SIGA was a machine instruction:
>
>
>
> --------------------------Extract from networking on
>
> z/OS-------------------------------------------------------------
>
> The advantages of using QDIO mode are:
>
>
>
> --A 20% improvement in performance versus non-QDIO mode
>
> --The reduction of system assist processor (SAP) utilization
>
> --Improved response time
>
> --Server cycle reduction
>
> How is this all accomplished? Instead of attaching an OSA-Express card
>
> using a channel, and hence utilizing a channel or I/O program (IOP)
>
> combined with CCW operations, the OSA-Express card attaches using an
>
> STI bus. STI stands for Self-Timed Interface. The OSA-Express card is
>
> still connected within the I/O cage, but the STI bus is directly
>
> connected to the memory bus of the CPC.
>
>
>
> What is used instead of a CCW operation to signify that I/O needs to
>
> be completed? With an OSA-Express card running in QDIO mode, I/O
>
> operations are effected using a signal adapter instruction, or SIGA.
>
> The SIGA is still processed by the SAP, similar to the way a CCW is
>
> processed by the SAP. However, the SIGA effectively passes a pointer
>
> to the data because the data already occupies internal storage.
>
> -----------------------End of extract from networking on
>
> z/OS-------------------------------------------------------------
>
>
>
> From this I assume that a SSCH instruction is issued for a QDIO
>
> sub-channel and this will cause one CCW to be executed with a CCW
>
> command code of SIGA. Am I on the right track here ?
>
>
>
> Thanks.
>
>
>
> Regards,
>
> John.
>
>
>
> ------------------------------
>
> For IBM-MAIN subscribe / signoff / archive access instructions,
>
> send email to [email protected] with the message: GET IBM-MAIN INFO
>
> Search the archives at http://bama.ua.edu/archives/ibm-main.html
>
> ----------------------------------------------------------------------
> For IBM-MAIN subscribe / signoff / archive access instructions,
> send email to [email protected] with the message: GET IBM-MAIN INFO
> Search the archives at http://bama.ua.edu/archives/ibm-main.html
>



-- 
John Blythe Reid,
Técnico de Sistemas de z/OS y de Sistemas Transaccionales,
Barcelona,
España.

----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Reply via email to