Lines Matching +full:non +full:- +full:zero
43 .Nd floating-point environment control
80 routines manipulate the floating-point environment,
82 .St -ieee754 .
84 Exception flags are set as side-effects of floating-point arithmetic
89 representing the five standard floating-point exceptions.
90 .Bl -tag -width ".Dv FE_DIVBYZERO"
92 A divide-by-zero exception occurs when the
95 For example, dividing a finite non-zero number by zero or computing
97 raises a divide-by-zero exception.
105 For instance, subtraction of like-signed infinities, division of zero by zero,
109 In contrast with divide-by-zero,
119 because it is too close to zero.
120 The result is a subnormal number or zero.
126 architecture-specific flags.
135 processor's floating-point exception flags, respectively.
148 .St -ieee754
152 floating-point variables.
154 .Bl -tag -width ".Dv FE_TOWARDZERO"
158 values, the value whose last binary digit is even (zero) is chosen.
165 Results are rounded towards zero.
178 functions save and restore the floating-point environment,
180 the rounding mode, and possibly other implementation-specific
188 .Em non-stop
190 In non-stop mode, floating-point operations will set exception flags
194 Non-stop mode is the default, but it may be altered by
202 but it also re-raises any floating-point exceptions from the old
215 .Bd -literal -offset indent
228 while (fabs((x * x) - n) > DBL_EPSILON * 2 * x)
258 .St -isoC-99 .
270 It supersedes the non-standard routines defined in
280 This lexically-scoped annotation tells the compiler that the program
281 may access the floating-point environment, so optimizations that would
282 violate strict IEEE-754 semantics are disabled.
285 is off, the floating-point environment will become undefined.
290 However, non-constant expressions generally produce the correct
291 side-effects at low optimization levels.