[Bug fortran/105924] New: false floating point exception when evaluating exponential function

[Bug fortran/105924] New: false floating point exception when evaluating exponential function

[yelinhui at hotmail dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=105924

            Bug ID: 105924
           Summary: false floating point exception when evaluating
                    exponential function
           Product: gcc
           Version: 7.5.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: fortran
          Assignee: unassigned at gcc dot gnu.org
          Reporter: yelinhui at hotmail dot com
  Target Milestone: ---

$ gfortran --version
GNU Fortran (Ubuntu 7.5.0-3ubuntu1~18.04) 7.5.0
Copyright (C) 2017 Free Software Foundation, Inc.

$ cat test.f
      program test
      implicit none
      real*8 :: e
      e=-4.d2      
      print *, e, exp(-e*e)
      end program test

$ gfortran -O0 -g -fbacktrace
-ffpe-trap=zero,overflow,underflow,invalid,denormal test.f

$./a.out

Program received signal SIGFPE: Floating-point exception - erroneous arithmetic
operation.

Backtrace for this error:
#0  0x1554d03f82ed in ???
#1  0x1554d03f7503 in ???
#2  0x1554cfc8cf0f in ???
#3  0x1554d00c5ea4 in ???
#4  0x1554d004e1fe in ???
#5  0x55b38e68a9d5 in test
        at /tmp/test.f:7
#6  0x55b38e68aa54 in main
        at /tmp/test.f:9
Floating point exception (core dumped)

[Bug fortran/105924] false floating point exception when evaluating exponential function

[kargl at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=105924

kargl at gcc dot gnu.org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |RESOLVED
                 CC|                            |kargl at gcc dot gnu.org
         Resolution|---                         |INVALID

--- Comment #1 from kargl at gcc dot gnu.org ---
Why do you thing that you should not get an exception?

e = -400
e*e = 160000
-e*e = -160000
exp(-e*e) = exp(-160000)  <-- This is going to underflow to zero.

You specifically asked gfortran to signal an exception if
underflow occurs with the -ffpe-trap=underflow option.  The
underflow threshold occurs at x = -745 for exp(x).

[Bug fortran/105924] false floating point exception when evaluating exponential function

[yelinhui at hotmail dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=105924

Lin-Hui Ye <yelinhui at hotmail dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|RESOLVED                    |CLOSED

--- Comment #2 from Lin-Hui Ye <yelinhui at hotmail dot com> ---
Sorry I didn't know this is an underflow. I was expecting exp(-160000) to give
a value close to zero. Thanks for the explanation.

Linhui

(In reply to kargl from comment #1)
> Why do you thing that you should not get an exception?
> 
> e = -400
> e*e = 160000
> -e*e = -160000
> exp(-e*e) = exp(-160000)  <-- This is going to underflow to zero.
> 
> You specifically asked gfortran to signal an exception if
> underflow occurs with the -ffpe-trap=underflow option.  The
> underflow threshold occurs at x = -745 for exp(x).