1 /*
2  * Linux/PA-RISC Project (http://www.parisc-linux.org/)
3  *
4  * Floating-point emulation code
5  *  Copyright (C) 2001 Hewlett-Packard (Paul Bame) <bame@debian.org>
6  *
7  *    This program is free software; you can redistribute it and/or modify
8  *    it under the terms of the GNU General Public License as published by
9  *    the Free Software Foundation; either version 2, or (at your option)
10  *    any later version.
11  *
12  *    This program is distributed in the hope that it will be useful,
13  *    but WITHOUT ANY WARRANTY; without even the implied warranty of
14  *    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
15  *    GNU General Public License for more details.
16  *
17  *    You should have received a copy of the GNU General Public License
18  *    along with this program; if not, write to the Free Software
19  *    Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
20  */
21 /*
22  * BEGIN_DESC
23  *
24  *  File:
25  *	@(#)	pa/spmath/dfsqrt.c		$Revision: $
26  *
27  *  Purpose:
28  *	Double Floating-point Square Root
29  *
30  *  External Interfaces:
31  *	dbl_fsqrt(srcptr,nullptr,dstptr,status)
32  *
33  *  Internal Interfaces:
34  *
35  *  Theory:
36  *	<<please update with a overview of the operation of this file>>
37  *
38  * END_DESC
39 */
40 
41 
42 #include "float.h"
43 #include "dbl_float.h"
44 
45 /*
46  *  Double Floating-point Square Root
47  */
48 
49 /*ARGSUSED*/
50 unsigned int
dbl_fsqrt(dbl_floating_point * srcptr,unsigned int * nullptr,dbl_floating_point * dstptr,unsigned int * status)51 dbl_fsqrt(
52 	    dbl_floating_point *srcptr,
53 	    unsigned int *nullptr,
54 	    dbl_floating_point *dstptr,
55 	    unsigned int *status)
56 {
57 	register unsigned int srcp1, srcp2, resultp1, resultp2;
58 	register unsigned int newbitp1, newbitp2, sump1, sump2;
59 	register int src_exponent;
60 	register boolean guardbit = FALSE, even_exponent;
61 
62 	Dbl_copyfromptr(srcptr,srcp1,srcp2);
63         /*
64          * check source operand for NaN or infinity
65          */
66         if ((src_exponent = Dbl_exponent(srcp1)) == DBL_INFINITY_EXPONENT) {
67                 /*
68                  * is signaling NaN?
69                  */
70                 if (Dbl_isone_signaling(srcp1)) {
71                         /* trap if INVALIDTRAP enabled */
72                         if (Is_invalidtrap_enabled()) return(INVALIDEXCEPTION);
73                         /* make NaN quiet */
74                         Set_invalidflag();
75                         Dbl_set_quiet(srcp1);
76                 }
77                 /*
78                  * Return quiet NaN or positive infinity.
79 		 *  Fall thru to negative test if negative infinity.
80                  */
81 		if (Dbl_iszero_sign(srcp1) ||
82 		    Dbl_isnotzero_mantissa(srcp1,srcp2)) {
83                 	Dbl_copytoptr(srcp1,srcp2,dstptr);
84                 	return(NOEXCEPTION);
85 		}
86         }
87 
88         /*
89          * check for zero source operand
90          */
91 	if (Dbl_iszero_exponentmantissa(srcp1,srcp2)) {
92 		Dbl_copytoptr(srcp1,srcp2,dstptr);
93 		return(NOEXCEPTION);
94 	}
95 
96         /*
97          * check for negative source operand
98          */
99 	if (Dbl_isone_sign(srcp1)) {
100 		/* trap if INVALIDTRAP enabled */
101 		if (Is_invalidtrap_enabled()) return(INVALIDEXCEPTION);
102 		/* make NaN quiet */
103 		Set_invalidflag();
104 		Dbl_makequietnan(srcp1,srcp2);
105 		Dbl_copytoptr(srcp1,srcp2,dstptr);
106 		return(NOEXCEPTION);
107 	}
108 
109 	/*
110 	 * Generate result
111 	 */
112 	if (src_exponent > 0) {
113 		even_exponent = Dbl_hidden(srcp1);
114 		Dbl_clear_signexponent_set_hidden(srcp1);
115 	}
116 	else {
117 		/* normalize operand */
118 		Dbl_clear_signexponent(srcp1);
119 		src_exponent++;
120 		Dbl_normalize(srcp1,srcp2,src_exponent);
121 		even_exponent = src_exponent & 1;
122 	}
123 	if (even_exponent) {
124 		/* exponent is even */
125 		/* Add comment here.  Explain why odd exponent needs correction */
126 		Dbl_leftshiftby1(srcp1,srcp2);
127 	}
128 	/*
129 	 * Add comment here.  Explain following algorithm.
130 	 *
131 	 * Trust me, it works.
132 	 *
133 	 */
134 	Dbl_setzero(resultp1,resultp2);
135 	Dbl_allp1(newbitp1) = 1 << (DBL_P - 32);
136 	Dbl_setzero_mantissap2(newbitp2);
137 	while (Dbl_isnotzero(newbitp1,newbitp2) && Dbl_isnotzero(srcp1,srcp2)) {
138 		Dbl_addition(resultp1,resultp2,newbitp1,newbitp2,sump1,sump2);
139 		if(Dbl_isnotgreaterthan(sump1,sump2,srcp1,srcp2)) {
140 			Dbl_leftshiftby1(newbitp1,newbitp2);
141 			/* update result */
142 			Dbl_addition(resultp1,resultp2,newbitp1,newbitp2,
143 			 resultp1,resultp2);
144 			Dbl_subtract(srcp1,srcp2,sump1,sump2,srcp1,srcp2);
145 			Dbl_rightshiftby2(newbitp1,newbitp2);
146 		}
147 		else {
148 			Dbl_rightshiftby1(newbitp1,newbitp2);
149 		}
150 		Dbl_leftshiftby1(srcp1,srcp2);
151 	}
152 	/* correct exponent for pre-shift */
153 	if (even_exponent) {
154 		Dbl_rightshiftby1(resultp1,resultp2);
155 	}
156 
157 	/* check for inexact */
158 	if (Dbl_isnotzero(srcp1,srcp2)) {
159 		if (!even_exponent && Dbl_islessthan(resultp1,resultp2,srcp1,srcp2)) {
160 			Dbl_increment(resultp1,resultp2);
161 		}
162 		guardbit = Dbl_lowmantissap2(resultp2);
163 		Dbl_rightshiftby1(resultp1,resultp2);
164 
165 		/*  now round result  */
166 		switch (Rounding_mode()) {
167 		case ROUNDPLUS:
168 		     Dbl_increment(resultp1,resultp2);
169 		     break;
170 		case ROUNDNEAREST:
171 		     /* stickybit is always true, so guardbit
172 		      * is enough to determine rounding */
173 		     if (guardbit) {
174 			    Dbl_increment(resultp1,resultp2);
175 		     }
176 		     break;
177 		}
178 		/* increment result exponent by 1 if mantissa overflowed */
179 		if (Dbl_isone_hiddenoverflow(resultp1)) src_exponent+=2;
180 
181 		if (Is_inexacttrap_enabled()) {
182 			Dbl_set_exponent(resultp1,
183 			 ((src_exponent-DBL_BIAS)>>1)+DBL_BIAS);
184 			Dbl_copytoptr(resultp1,resultp2,dstptr);
185 			return(INEXACTEXCEPTION);
186 		}
187 		else Set_inexactflag();
188 	}
189 	else {
190 		Dbl_rightshiftby1(resultp1,resultp2);
191 	}
192 	Dbl_set_exponent(resultp1,((src_exponent-DBL_BIAS)>>1)+DBL_BIAS);
193 	Dbl_copytoptr(resultp1,resultp2,dstptr);
194 	return(NOEXCEPTION);
195 }
196