Re: Expensive op codes

Mark,

The big difference is caused by the setll that accompanies the READE.  
You cannot block a file that has a setll against it.   You could probably 
get similar results between READ and READE if you specify BLOCK(*YES) 
and do your SETLL against another file that shares the access path of the 
file that you READE against.  I never thought to try this, but I did find a 
significant improvement between setll/reade and setll file1/read file2.  
We use I/O modules on all of our files which made it fairly easy to 
implement a pseudo reade.  I ran benchmarks with live data and 
found about a 33% improvement.  

David Morris

>>> Mark Lazarus <mlazarus@ttec.com> 10/18/99 09:09PM >>>
Peter,

At 12:01 AM 10/18/99 -0700, you wrote:

>As I mentioned in an earlier post, I fail to see why READE and READ/compare
>are so different. They are both doing sequential reads (last time I checked
><g>) and then comparing some key fields to determine whether to use the
>record or not. I would like to think that it's possible for the READE code
>to read a block of records just as READ does, and look at each record within
>the block to do it's compare before either passing the record to the program
>or setting the = indicator. This is something that could be tweaked by IBM
>at any time.

 1) Functionally, one way READE works differently is that the data does not
populate the input fields if the key is not equal.  Code has to be
generated for that.

2)  When the EOF indicator comes on, this is considered an "error"
condition to that opcode.  I/O error conditions tend to be expensive.

 -mark


+---
| This is the RPG/400 Mailing List!
| To submit a new message, send your mail to RPG400-L@midrange.com.
| To subscribe to this list send email to RPG400-L-SUB@midrange.com.
| To unsubscribe from this list send email to RPG400-L-UNSUB@midrange.com.
| Questions should be directed to the list owner/operator: david@midrange.com
+---