On 10/21/2013 1:11 AM, Martin Trübner wrote:
Be very very carefull. My POP say about MVST: The CPU-determined
number is at least one.
Good catch! Never ASS-U-ME anything! :-)
--
Edward E Jaffe
Phoenix Software International, Inc
831 Parkview Drive North
El Segundo, CA 90245
http://www.phoenixs
Bernd,
>> I believe that the RC = 2 was found to be not necessary in the case of
TRTE, because this case (nonzero function byte in the last byte of the string
to be tested)
Absolutly- If I reshuffle my code such that I maintain position within
source (and target of course) I could easy forget abo
> I don't think I've ever truly leveraged RC=2 from TRT. I usually test
> the length remaining at the top of the loop via DOEXIT.
Never leverage a short word where a long one will do.
Best regards, Steve Hobson
CICS Strategy, HLASM Development, Master Inventor
Hursley Laboratories, MP 189, Room A
On Mon, 21 Oct 2013 10:11:50 +0200 Martin Trübner
wrote:
:>>> Also, these instructions have the benefit of a guaranteed 256-byte minimum
before
:>>> they are interrupted,
:>and
:>>> if you know you are working with "short" data (<256 bytes) you
:>>> don't need to check for RC=3 at all...
Ed,
>> Also, these instructions have the benefit of a guaranteed 256-byte minimum
>> before
>> they are interrupted,
and
>> if you know you are working with "short" data (<256 bytes) you
>> don't need to check for RC=3 at all...
Be very very carefull.
My POP say about MVST: The CPU-deter
From: "John Gilmore"
Sent: Monday, October 21, 2013 4:47 AM
TRT and TRTE are different, but TRTE and TRTRE provide you with
everything you need.
o RC=0, no [new] non-zero function byte
found (and input exhausted)
o RC=1, a non-zero function byte has
been found and input is not exhausted
On Sun, 20 Oct 2013 13:44:53 -0700 Ed Jaffe
wrote:
:>On 10/20/2013 7:07 AM, Martin Truebner wrote:
:>> Does anyone have insights (or ideas) why CC 2 (xH) was omitted from
:>> TRTE.
:>> It is there in TRT and it does come handy when determine if there is a
:>> remainder to process after the speci
On 10/20/2013 3:28 PM, Bernd Oppolzer wrote:
... (but different from MVCL, the interruption is
signalled to the program by means of CC = 3).
Which saves a *LOT* of silicon by offloading the restart to software.
All of the new interruptible instructions work this way. Also, these
instructions ha
No problem at all.
My point was, that the RC = 2 to me seems not important any more,
because, as I read it, the fact that the nonzero function byte was found
when processing the last byte can be derived from the fact that the
remaining length in R1 +1 is 1 - if I understand the PoOp correctly.
A
On 10/20/2013 3:09 PM, John Gilmore wrote:
I think that Berhard's [sic] point was that the requirement that
"the remaining length must always be computed and checked at the top
of the loop"
when a TRT is used is eliminated when a TRT[R]E is used instead.
Agreed. That's a nice/convenient featu
"Bernd' somehow became "Bernhard' in my last post. Anothert senior
moment. My apologies to Bernd.
John Gilmore, Ashland, MA 01721 - USA
I think that Berhard's point was that the requirement that
"the remaining length must always be computed and checked at the top
of the loop"
when a TRT is used is eliminated when a TRT[R]E is used instead.
John Gilmore, Ashland, MA 01721 - USA
On 10/20/2013 1:54 PM, Bernd Oppolzer wrote:
With TRT, there was no easy way to test the remaining length; it first
had to be computed by subtracting R1 from the end address.
True, but I've almost never had a situation in which the work I had to
do after finding the byte was so trivial that the
With TRT, there was no easy way to test the remaining length; it first
had to
be computed by subtracting R1 from the end address.
But with TRTE, the remaining length is in R1+1, so that makes a
difference IMHO.
Kind regards
Bernd
Am 20.10.2013 22:44, schrieb Ed Jaffe:
On 10/20/2013 7:07 AM,
On 10/20/2013 7:07 AM, Martin Truebner wrote:
Does anyone have insights (or ideas) why CC 2 (xH) was omitted from
TRTE.
It is there in TRT and it does come handy when determine if there is a
remainder to process after the special-char was found.
I don't think I've ever truly leveraged RC=2 fro
Sorry, to make my posting more precise, the wording in the parantheses
should be:
(nonzero function byte RESULTING FROM the last byte of the string to be
tested)
even if my english is not perfect, because I'm no native english speaker
(same as
Martin), we should take care about the meaning of ou
Martin,
I believe that the RC = 2 was found to be not necessary in the case of
TRTE,
because this case (nonzero function byte in the last byte of the string
to be tested)
can be examined easily by looking at the remaining length in R1 + 1
after RC = 1
(when TRTE finds a nonzero function byte).
A
Martin,
Let me try again.
Paraphrased only a little, see page 7-337, the current z/Architecture PROP says:
o RC = 0, all function bytes [examined]
are zero,
o RC = 1, non-zero function byte found,
input IS NOT exhausted,
o RC = 2, non-zero function byte found,
input IS exhausted.
For
John,
let me rephrase it in HLASM (english is not my native language)
CC can only be zero, one, two or three (only two bits)
0=complete
1=found something
2=not set
3=the usual suspect (not complete yet)
how can I test foun d something in last?
TRTE R10,R2
JO *-4
* it could be
I'm currently out of the office.
I'll reply to your note as soon as I return to the office.
Have a great day,
Steve
Martin,
TRT and TRTE are different, but TRTE and TRTRE provide you with
everything you need.
o RC=0, no [new] non-zero function byte
found (and input exhausted)
o RC=1, a non-zero function byte has
been found and input is not exhausted
o RC=1, a non-zero function byte has
been found an
I will be out of the office from 10/20 to 10/25. If you need immediate
assistance, Please contact the Help Desk. Thank You...Guy Gates
I am currently out of the office.
I will return Monday 10/21
If you need a faster response please call Frank Cunningham or Ron Root.
Frank's Phone: +1 512 795 6910
Ron's Phone +1 512 795 6923 (Ron will only be here Thursday, not Friday)
Thanks,
Joey
Does anyone have insights (or ideas) why CC 2 (xH) was omitted from
TRTE.
It is there in TRT and it does come handy when determine if there is a
remainder to process after the special-char was found.
--
Martin
Pi_cap_CPU - all you ever need around MWLC/SCRT/CMT in z/VSE
more at http://www.picapc
Je suis actuellement en congés et je ne serais de retour que le 2013/10/04.
Durant cette période merci de contacter DSI-PROD-SYSTEME-CENTRAUX. Cordialement.
Cette notification ne vous sera adressée qu'une seule fois.
I'm currently out of the office and will not return until 2013/10/04. In the
me
25 matches
Mail list logo
