URL
https://opencores.org/ocsvn/openrisc/openrisc/trunk
Subversion Repositories openrisc
[/] [openrisc/] [trunk/] [gnu-old/] [newlib-1.17.0/] [newlib/] [libm/] [mathfp/] [s_pow.c] - Rev 866
Go to most recent revision | Compare with Previous | Blame | View Log
/* @(#)z_pow.c 1.0 98/08/13 */ /* FUNCTION <<pow>>, <<powf>>---x to the power y INDEX pow INDEX powf ANSI_SYNOPSIS #include <math.h> double pow(double <[x]>, double <[y]>); float pow(float <[x]>, float <[y]>); TRAD_SYNOPSIS #include <math.h> double pow(<[x]>, <[y]>); double <[x]>, <[y]>; float pow(<[x]>, <[y]>); float <[x]>, <[y]>; DESCRIPTION <<pow>> and <<powf>> calculate <[x]> raised to the exponent <[y]>. @tex (That is, $x^y$.) @end tex RETURNS On success, <<pow>> and <<powf>> return the value calculated. When the argument values would produce overflow, <<pow>> returns <<HUGE_VAL>> and set <<errno>> to <<ERANGE>>. If the argument <[x]> passed to <<pow>> or <<powf>> is a negative noninteger, and <[y]> is also not an integer, then <<errno>> is set to <<EDOM>>. If <[x]> and <[y]> are both 0, then <<pow>> and <<powf>> return <<1>>. You can modify error handling for these functions using <<matherr>>. PORTABILITY <<pow>> is ANSI C. <<powf>> is an extension. */ #include <float.h> #include "fdlibm.h" #include "zmath.h" #ifndef _DOUBLE_IS_32BITS double pow (double x, double y) { double d, k, t, r = 1.0; int n, sign, exponent_is_even_int = 0; __uint32_t px; GET_HIGH_WORD (px, x); k = modf (y, &d); if (k == 0.0) { /* Exponent y is an integer. */ if (modf (ldexp (y, -1), &t)) { /* y is odd. */ exponent_is_even_int = 0; } else { /* y is even. */ exponent_is_even_int = 1; } } if (x == 0.0) { if (y <= 0.0) errno = EDOM; } else if ((t = y * log (fabs (x))) >= BIGX) { errno = ERANGE; if (px & 0x80000000) { /* x is negative. */ if (k) { /* y is not an integer. */ errno = EDOM; x = 0.0; } else if (exponent_is_even_int) x = z_infinity.d; else x = -z_infinity.d; } else { x = z_infinity.d; } } else if (t < SMALLX) { errno = ERANGE; x = 0.0; } else { if ( !k && fabs(d) <= 32767 ) { n = (int) d; if ((sign = (n < 0))) n = -n; while ( n > 0 ) { if ((unsigned int) n % 2) r *= x; x *= x; n = (unsigned int) n / 2; } if (sign) r = 1.0 / r; return r; } else { if ( px & 0x80000000 ) { /* x is negative. */ if ( k ) { /* y is not an integer. */ errno = EDOM; return 0.0; } } x = exp (t); if (!exponent_is_even_int) { if (px & 0x80000000) { /* y is an odd integer, and x is negative, so the result is negative. */ GET_HIGH_WORD (px, x); px |= 0x80000000; SET_HIGH_WORD (x, px); } } } } return x; } #endif _DOUBLE_IS_32BITS
Go to most recent revision | Compare with Previous | Blame | View Log