RE: Date * 10000.01

["Pytel, Alexei" <pytela1@xxxxxxxxxxxxxxxx>]

While it's true about exception, it's not EAO.
EAO is a different type of exception. Due to historical reasons hardware
segment size (in virtual memory) was smaller than segment size
architectured in MI. So every time program would cross hardware segment
boundary (legitimately staying within software segment), EAO exception
would be generated, internal registers updated and program allowed to
run on.


    Best regards,

          Alexei Pytel



> -----Original Message-----
> From: Danbo [SMTP:danbo@interaccess.com]
> Sent: Friday, November 20, 1998 1:55 AM
> To:   MIDRANGE-L@midrange.com
> Subject:      Re: Date * 10000.01
> 
> I'll add my two cents into here...
> 
> Part of the reason people suggest the data structure approach relates
> to how
> RPG handles truncation of large numbers.  When you multiply a number,
> and
> the result is larger then the result field, it effectively generates
> what is
> called an EAO exception (effective address overflow).  Continued EAO
> exceptions in any of these sorts of calculations result in poor
> performance.
> If you were to use the performance tools to monitor jobs which do
> calculations such as this, you'd notice the job spends a lot of time
> on that
> particular statement.
> 
> Just my note here...
> 
> Daniel R. Boggs
> Software Product Development, Manager
> Lakeview Technology
> (authors of MIMIX)
> 
> -----Original Message-----
> From: Hans Boldt <boldt@ca.ibm.com>
> To: MIDRANGE-L@midrange.com <MIDRANGE-L@midrange.com>
> Date: Wednesday, November 18, 1998 9:37 AM
> Subject: Re: Date * 10000.01
> 
> 
> >Angie Wawrzaszek wrote:
> >>Someone posted a multiplier that can be used to change an 8,0
> >>date from mdy to ymd and vice versa using the same multipier.
> >>I can not find that, and was hoping someone could re-post it.
> >>This is just a huge help for programs that are not yet converted
> >>to RPG1V.
> >
> >I'm not posting the multiplier.  I'm posting to
> >recommend against using a MULT operator to perform
> >date conversions.  I did some performance comparisons
> >a long time ago and found that using MULT to convert
> >a date was 100 to 150 times slower than 3 MOVEs.
> >Furthermore, when you do move to RPG IV, the technique
> >does not work if you use TRUNCNBR(*NO), or if you do
> >the multiplication in an expression.
> >
> >Cheers!  Hans
> >
> >Hans Boldt, ILE RPG Development, IBM Toronto Lab, boldt@ca.ibm.com
> >+---
> >| This is the Midrange System Mailing List!
> >| To submit a new message, send your mail to MIDRANGE-L@midrange.com.
> >| To subscribe to this list send email to
> MIDRANGE-L-SUB@midrange.com.
> >| To unsubscribe from this list send email to
> MIDRANGE-L-UNSUB@midrange.com.
> >| Questions should be directed to the list owner/operator:
> david@midrange.com
> >+---
> >
> 
> +---
> | This is the Midrange System Mailing List!
> | To submit a new message, send your mail to MIDRANGE-L@midrange.com.
> | To subscribe to this list send email to MIDRANGE-L-SUB@midrange.com.
> | To unsubscribe from this list send email to
> MIDRANGE-L-UNSUB@midrange.com.
> | Questions should be directed to the list owner/operator:
> david@midrange.com
> +---
+---
| This is the Midrange System Mailing List!
| To submit a new message, send your mail to MIDRANGE-L@midrange.com.
| To subscribe to this list send email to MIDRANGE-L-SUB@midrange.com.
| To unsubscribe from this list send email to MIDRANGE-L-UNSUB@midrange.com.
| Questions should be directed to the list owner/operator: david@midrange.com
+---