unnatural log

Below is a little demo program that produces what are, to me, 
unpredictable results.  This is a exercise in using C library 
functions.  The program was compiled under v3r7 using activation 
group *caller and binding directory qc2le.  

The first and second calls execute correctly. The third call returns 
a value that appears to be ln(ln(2)).  It's not an artifact of using 
2 for the function argument--it returns the same value for other 
numbers also.  The same code works fine if the function argument is 
initialized each time using the %float() function.

I'm sure there's something I'm missing here.  Any suggestions.


JEFTST/QRPGLESRC 
 SEU==>
NATLOG3 
        *************** Beginning of data
**********************************************************************
**
***************   
0004.00 Dlogc             pr             8f   extproc('log')
0005.00 Darg                             8f
0006.00 D
0007.00 Dlog              s              8f
0008.00 Ddog              s              8f
0013.00 D*
0014.00 C                   eval      dog = 2.0
0015.00 C                   eval      log = logc(dog)
0016.00 C
0021.00 C                   eval      dog = %float(2.0)
0023.00 C                   eval      log = logc(dog)
0024.00 C
0030.00 C                   eval      log = logc(dog)
0031.00 C
0034.00 C                   eval      *inlr = *on
                 






* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
* This is the RPG/400 Discussion Mailing List!  To submit a new         *
* message, send your mail to "RPG400-L@midrange.com".  To unsubscribe   *
* from this list send email to MAJORDOMO@midrange.com and specify       *
* 'unsubscribe RPG400-L' in the body of your message.  Questions should *
* be directed to the list owner / operator: david@midrange.com          *
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *