RE: alarm() not working as expected

errNo is set when an error occurs, but is not set when no error occurs. You
can (and should, IMO) initialize it. The following procedure code works just
as the errNo function works in C, allowing you to specify errNo(0) to
initialize the storage.

 

     D errNo           PR            10I 0

     D                               10I 0 Value Options(*NOPASS)

 

     D errNo_Builtin   PR              *   ExtProc('__errno')

 

 

     P errNo           B                   Export


     D errNo           PI            10I 0

     D setErrNoValue                 10I 0 Value Options(*NOPASS)

 

 

     D errNoBase       S               *

     D errNoValue      S             10I 0 Based(errNoBase)

 

 

      /Free

 

       If %Parms = 1;

          errNoBase = errNo_BuiltIn();

          errNoValue = setErrNoValue;

       EndIf;

 

       errNoBase = errNo_BuiltIn();

 

       Return errNoValue;

 

      /End-Free

 

     P errNo           E

 

       
Duane Christen


-----Original Message-----
From: Scott Klement [mailto:rpg400-l@xxxxxxxxxxxxxxxx]
Sent: Tuesday, January 17, 2006 3:49 PM
To: RPG programming on the AS400 / iSeries
Subject: RE: alarm() not working as expected



> I think the problem is actually the errno did not get reset.  Does the
> program suppose to reset the errno after receviced the alarm timeout
> before doing any recv again?

As far as I know, the system never resets errno.  The only time the value 
of errno changes is when the next error occurs.

Therefore, you should only look at the value of errno when an API returns 
-1 to indicate that an error has occurred.  Don't use errno to determine 
IF an error occurred, only to determine WHICH errror has occurred.