BUG: FOR2932 Error in Round or Truncate of Extra Digits

MORE INFORMATION

The maximal values of single precision negative numbers is in the range of -3.4028235E+38 to -1.1754944E-38 and for single-precision positive numbers is in the range of +1.1754944E-38 to +3.4028235E+38.

The following code demonstrates the problem.

Sample Code

C Compile options needed: none

      real*4 x1

      x1 = huge(1.0)
      write (*,100) x1
      x1 = 3.402823466385288787198009999e38      ! in range here.
      write (*,100) x1
      x1 = 3.4028234663852887871980100e38        ! out of range here.
      write (*,100) x1
      x1 = 3.4028234663852887871989999e38        ! to here.
      write (*,100) x1
      x1 = 3.4028234663852887871990e38           ! in range here
      write (*,100) x1
      x1 = 3.4028234663852887871990099999999e38  ! to here.
      write (*,100) x1
      x1 = 3.40282346638528878719901e38          ! then out of range here.
      write (*,100) x1
100   format (e16.9)
      end