1 /* $NetBSD: mul.S,v 1.1 2005/12/20 19:28:50 christos Exp $ */
4 * Copyright (c) 1992, 1993
5 * The Regents of the University of California. All rights reserved.
7 * This software was developed by the Computer Systems Engineering group
8 * at Lawrence Berkeley Laboratory under DARPA contract BG 91-66 and
9 * contributed to Berkeley.
11 * Redistribution and use in source and binary forms, with or without
12 * modification, are permitted provided that the following conditions
14 * 1. Redistributions of source code must retain the above copyright
15 * notice, this list of conditions and the following disclaimer.
16 * 2. Redistributions in binary form must reproduce the above copyright
17 * notice, this list of conditions and the following disclaimer in the
18 * documentation and/or other materials provided with the distribution.
19 * 3. Neither the name of the University nor the names of its contributors
20 * may be used to endorse or promote products derived from this software
21 * without specific prior written permission.
23 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
24 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
25 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
26 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
27 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
28 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
29 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
30 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
31 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
32 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
35 * from: Header: mul.s,v 1.5 92/06/25 13:24:03 torek Exp
38 #include <machine/asm.h>
39 #if defined(LIBC_SCCS) && !defined(lint)
41 .asciz "@(#)mul.s 8.1 (Berkeley) 6/4/93"
43 RCSID("$NetBSD: mul.S,v 1.1 2005/12/20 19:28:50 christos Exp $")
45 #endif /* LIBC_SCCS and not lint */
48 * Signed multiply, from Appendix E of the Sparc Version 8
49 * Architecture Manual.
51 * Returns %o0 * %o1 in %o1%o0 (i.e., %o1 holds the upper 32 bits of
52 * the 64-bit product).
54 * This code optimizes short (less than 13-bit) multiplies.
58 mov %o0, %y ! multiplier -> Y
59 andncc %o0, 0xfff, %g0 ! test bits 12..31
60 be Lmul_shortway ! if zero, can do it the short way
61 andcc %g0, %g0, %o4 ! zero the partial product and clear N and V
64 * Long multiply. 32 steps, followed by a final shift step.
66 mulscc %o4, %o1, %o4 ! 1
67 mulscc %o4, %o1, %o4 ! 2
68 mulscc %o4, %o1, %o4 ! 3
69 mulscc %o4, %o1, %o4 ! 4
70 mulscc %o4, %o1, %o4 ! 5
71 mulscc %o4, %o1, %o4 ! 6
72 mulscc %o4, %o1, %o4 ! 7
73 mulscc %o4, %o1, %o4 ! 8
74 mulscc %o4, %o1, %o4 ! 9
75 mulscc %o4, %o1, %o4 ! 10
76 mulscc %o4, %o1, %o4 ! 11
77 mulscc %o4, %o1, %o4 ! 12
78 mulscc %o4, %o1, %o4 ! 13
79 mulscc %o4, %o1, %o4 ! 14
80 mulscc %o4, %o1, %o4 ! 15
81 mulscc %o4, %o1, %o4 ! 16
82 mulscc %o4, %o1, %o4 ! 17
83 mulscc %o4, %o1, %o4 ! 18
84 mulscc %o4, %o1, %o4 ! 19
85 mulscc %o4, %o1, %o4 ! 20
86 mulscc %o4, %o1, %o4 ! 21
87 mulscc %o4, %o1, %o4 ! 22
88 mulscc %o4, %o1, %o4 ! 23
89 mulscc %o4, %o1, %o4 ! 24
90 mulscc %o4, %o1, %o4 ! 25
91 mulscc %o4, %o1, %o4 ! 26
92 mulscc %o4, %o1, %o4 ! 27
93 mulscc %o4, %o1, %o4 ! 28
94 mulscc %o4, %o1, %o4 ! 29
95 mulscc %o4, %o1, %o4 ! 30
96 mulscc %o4, %o1, %o4 ! 31
97 mulscc %o4, %o1, %o4 ! 32
98 mulscc %o4, %g0, %o4 ! final shift
100 ! If %o0 was negative, the result is
101 ! (%o0 * %o1) + (%o1 << 32))
108 ! %o0 was indeed negative; fix upper 32 bits of result by subtracting
109 ! %o1 (i.e., return %o4 - %o1 in %o1).
119 * Short multiply. 12 steps, followed by a final shift step.
120 * The resulting bits are off by 12 and (32-12) = 20 bit positions,
121 * but there is no problem with %o0 being negative (unlike above).
123 mulscc %o4, %o1, %o4 ! 1
124 mulscc %o4, %o1, %o4 ! 2
125 mulscc %o4, %o1, %o4 ! 3
126 mulscc %o4, %o1, %o4 ! 4
127 mulscc %o4, %o1, %o4 ! 5
128 mulscc %o4, %o1, %o4 ! 6
129 mulscc %o4, %o1, %o4 ! 7
130 mulscc %o4, %o1, %o4 ! 8
131 mulscc %o4, %o1, %o4 ! 9
132 mulscc %o4, %o1, %o4 ! 10
133 mulscc %o4, %o1, %o4 ! 11
134 mulscc %o4, %o1, %o4 ! 12
135 mulscc %o4, %g0, %o4 ! final shift
138 * %o4 has 20 of the bits that should be in the low part of the
139 * result; %y has the bottom 12 (as %y's top 12). That is:
142 * +----------------+----------------+
143 * | -12- | -20- | -12- | -20- |
144 * +------(---------+------)---------+
145 * --hi-- ----low-part----
147 * The upper 12 bits of %o4 should be sign-extended to form the
148 * high part of the product (i.e., highpart = %o4 >> 20).
152 sll %o4, 12, %o0 ! shift middle bits left 12
153 srl %o5, 20, %o5 ! shift low bits right 20, zero fill at left
154 or %o5, %o0, %o0 ! construct low part of result
156 sra %o4, 20, %o1 ! ... and extract high part of result