fppmask — Apollo

2.0;fppmask (floating_point_mask), revision 2.0, 81/07/02
FPPMASK (FLOATING_POINT_MASK) -- Set/display floating-point error mask.
usage:  FPPMASK [-D] [-E] [LOS] [UNDR]


FORMAT

  FPPMASK [options]


  FPPMASK  sets  or  displays the state of the floating-point package error mask
  for a process.    The  error  mask  specifies  some  of  the  conditions  that
  constitute a floating-point exception for the process.


OPTIONS

  If  no  options are specified, the current floating-point error mask condition
  is displayed.

  -D condition ...
                     Disable 'condition' (see below).  Both  conditions  may  be
                     specified.

  -E condition ...
                     Enable  'condition'  (see  below).   Both conditions may be
                     specified.

  'condition' may be either of the following:

  LOS                 Loss  of  significance.    This  condition   occurs   when
                     subtracting floating-point values that are exactly equal.

  UNDR                Underflow.    This  condition occurs when a floating-point
                     result   is   too   small   to   be   represented.      For
                     single-precision   values,   this  error  occurs  at  about
                     0.118E-37.  For double-precision values, the  error  occurs
                     at about 0.223E-307.

  Both conditions are initially disabled when a process is created.


EXAMPLES

  1. $ fppmask                                 Display current settings.
     LOS (loss of significance):   disabled
     UNDR (underflow):             disabled

  2. $ fppmask -e los undr                     Enable both LOS and UNDR
                                                conditions.

  3. $ fppmask -d undr                         Disable UNDR condition.
Museum
