That is true if you KNOW that you are going to only be in AR-mode.  In the cases
that I have seen in the past, the program was only conditionally in AR-mode.  In
fact, part of the time it was executing on hardware that did not even have
address registers.

My point was:  avoid problems downstream and code properly.

Lloyd



----- Original Message ----
From: Tom Marchant <[email protected]>
To: [email protected]
Sent: Thu, August 19, 2010 9:12:46 AM
Subject: Re: CDS and alignment question

On Thu, 19 Aug 2010 05:32:15 -0700, Lloyd Fuller wrote:

>And BOTH of the LA commands shown are bad coding practice
>in these days of cross
>memory:  add the "," to make R11 and R12 be base registers
>and not index registers.
>
>From experience that type of problem is difficult to debug
>when a problem occurs
>because you are suddenly running in AR mode.

If the program is going to be running in AR mode, the LA
instructions should probably be LAE.

--
Tom Marchant

>
>
>----- Original Message ----
>From: Paul Gilmartin <[email protected]>
>To: [email protected]
>Sent: Thu, August 19, 2010 1:27:30 AM
>Subject: Re: CDS and alignment question
>
>On Aug 18, 2010, at 17:30, John R. Ehrman (408-463-3543 T/543-) wrote:
>
>
>
>>    (------------------ Referenced Note Follows --------------------)
>> Sent: Wednesday, August 18, 2010 3:51 PM
>>
>> QMGREX01 CSECT ,
>>         STM   R14,R12,12(R13)
>>         LR    R11,R15
>>         USING QMGREX01,R11,R12
>>         LA    R12,4095(R11)
>>         LA    R12,1(R12)   <----------------- noticed that I added 1 to
>> the 4095 bump.
>
>-- gil

Reply via email to