f77 floating point exceptions opn SS1 4.0.3

[marty at puppsr.princeton.edu]

In article <1990Aug16.000008.7986 at rice.edu>, preston at ra.src.umd.edu
(James Preston) writes:
|> We are using f77 on Sparcstations (OS4.0.3) and are quite often
getting
|> the message:
|> 	Floating point errors ocurred...
|> 	Inexact; Underflow;

You have to look over your code carefully to find out whether you have a
real problem.  The compiler usually does the right thing (like exp(large
neg.  number) becomes zero but will raise flags).  If the code is
thoroughly debugged :-) it can be ignored.  Try inserting the line

	idummy = ieee_handler('set','invalid',SIGFPE_ABORT)

near the top of the program to make it crash where the error is occurring,
and use dbx to examine the pieces...  Good luck.

Marty Ryba
Princeton University
marty at pulsar.princeton.edu