hammer2 - Fix bug in comp:autozero mode
[dragonfly.git] / contrib / openbsd_libm / src / e_fmodf.c
blobcca2f6f63fdb7c57ad4a7f14c637cd5b1cb71828
1 /* e_fmodf.c -- float version of e_fmod.c.
2 * Conversion to float by Ian Lance Taylor, Cygnus Support, ian@cygnus.com.
3 */
5 /*
6 * ====================================================
7 * Copyright (C) 1993 by Sun Microsystems, Inc. All rights reserved.
9 * Developed at SunPro, a Sun Microsystems, Inc. business.
10 * Permission to use, copy, modify, and distribute this
11 * software is freely granted, provided that this notice
12 * is preserved.
13 * ====================================================
16 /*
17 * fmodf(x,y)
18 * Return x mod y in exact arithmetic
19 * Method: shift and subtract
22 #include "math.h"
23 #include "math_private.h"
25 static const float one = 1.0, Zero[] = {0.0, -0.0,};
27 float
28 fmodf(float x, float y)
30 int32_t n,hx,hy,hz,ix,iy,sx,i;
32 GET_FLOAT_WORD(hx,x);
33 GET_FLOAT_WORD(hy,y);
34 sx = hx&0x80000000; /* sign of x */
35 hx ^=sx; /* |x| */
36 hy &= 0x7fffffff; /* |y| */
38 /* purge off exception values */
39 if(hy==0||(hx>=0x7f800000)|| /* y=0,or x not finite */
40 (hy>0x7f800000)) /* or y is NaN */
41 return (x*y)/(x*y);
42 if(hx<hy) return x; /* |x|<|y| return x */
43 if(hx==hy)
44 return Zero[(u_int32_t)sx>>31]; /* |x|=|y| return x*0*/
46 /* determine ix = ilogb(x) */
47 if(hx<0x00800000) { /* subnormal x */
48 for (ix = -126,i=(hx<<8); i>0; i<<=1) ix -=1;
49 } else ix = (hx>>23)-127;
51 /* determine iy = ilogb(y) */
52 if(hy<0x00800000) { /* subnormal y */
53 for (iy = -126,i=(hy<<8); i>=0; i<<=1) iy -=1;
54 } else iy = (hy>>23)-127;
56 /* set up {hx,lx}, {hy,ly} and align y to x */
57 if(ix >= -126)
58 hx = 0x00800000|(0x007fffff&hx);
59 else { /* subnormal x, shift x to normal */
60 n = -126-ix;
61 hx = hx<<n;
63 if(iy >= -126)
64 hy = 0x00800000|(0x007fffff&hy);
65 else { /* subnormal y, shift y to normal */
66 n = -126-iy;
67 hy = hy<<n;
70 /* fix point fmod */
71 n = ix - iy;
72 while(n--) {
73 hz=hx-hy;
74 if(hz<0){hx = hx+hx;}
75 else {
76 if(hz==0) /* return sign(x)*0 */
77 return Zero[(u_int32_t)sx>>31];
78 hx = hz+hz;
81 hz=hx-hy;
82 if(hz>=0) {hx=hz;}
84 /* convert back to floating value and restore the sign */
85 if(hx==0) /* return sign(x)*0 */
86 return Zero[(u_int32_t)sx>>31];
87 while(hx<0x00800000) { /* normalize x */
88 hx = hx+hx;
89 iy -= 1;
91 if(iy>= -126) { /* normalize output */
92 hx = ((hx-0x00800000)|((iy+127)<<23));
93 SET_FLOAT_WORD(x,hx|sx);
94 } else { /* subnormal output */
95 n = -126 - iy;
96 hx >>= n;
97 SET_FLOAT_WORD(x,hx|sx);
98 x *= one; /* create necessary signal */
100 return x; /* exact output */