1/* 2 * Copyright (c) 1986 Regents of the University of California. 3 * All rights reserved. 4 * 5 * This code is derived from software contributed to Berkeley by 6 * Computer Consoles Inc. 7 * 8 * Redistribution and use in source and binary forms are permitted 9 * provided that the above copyright notice and this paragraph are 10 * duplicated in all such forms and that any documentation, 11 * advertising materials, and other materials related to such 12 * distribution and use acknowledge that the software was developed 13 * by the University of California, Berkeley. The name of the 14 * University may not be used to endorse or promote products derived 15 * from this software without specific prior written permission. 16 * THIS SOFTWARE IS PROVIDED ``AS IS'' AND WITHOUT ANY EXPRESS OR 17 * IMPLIED WARRANTIES, INCLUDING, WITHOUT LIMITATION, THE IMPLIED 18 * WARRANTIES OF MERCHANTIBILITY AND FITNESS FOR A PARTICULAR PURPOSE. 19 */ 20 21#if defined(SYSLIBC_SCCS) && !defined(lint) 22 .asciz "@(#)muld.s 1.2 (Berkeley) 02/17/89" 23#endif /* SYSLIBC_SCCS and not lint */ 24 25#include <tahoemath/fp.h> 26#include "DEFS.h" 27 28#define HIDDEN 23 /* here we count from 0 not from 1 as in fp.h */ 29 30XENTRY(muld, R2|R3|R4|R5|R6|R7|R8|R9) 31 clrl r3 /* r3 - sign: 0 for positive,1 for negative. */ 32 movl 4(fp),r0 33 jgeq 1f 34 movl $1,r3 351: movl 12(fp),r2 36 jgeq 2f 37 bbc $0,r3,1f /* seconed operand is negative. */ 38 clrl r3 /* if first was negative, make result positive. */ 39 jmp 2f 401: movl $1,r3 /* if first was positive, make result negative. */ 412: andl2 $EXPMASK,r0 /* compute first 'pure'exponent. */ 42 jeql is_res1 43 shrl $EXPSHIFT,r0,r0 44 subl2 $BIASP1,r0 45 andl2 $EXPMASK,r2 /* compute seconed 'pure'exponent. */ 46 jeql is_res2 47 shrl $EXPSHIFT,r2,r2 48 subl2 $BIASP1,r2 49 addl2 r0,r2 /* add the exponents. */ 50 addl2 $(BIASP1+2),r2 51 jleq underflow 52 cmpl r2,$258 /* normalization can make the exp. smaller. */ 53 jgeq overflow 54 /* 55 * We have the sign in r3,the exponent in r2,now is the time to 56 * perform the multiplication... 57 */ 58 /* fetch first fraction: (r0,r1) */ 59 andl3 $(0!(EXPMASK | SIGNBIT)),4(fp),r0 60 orl2 $(0!CLEARHID),r0 61 movl 8(fp),r1 62 shlq $7,r0,r0 /* leave the sign bit cleared. */ 63 64 /* fetch seconed fraction: (r4,r5) */ 65 andl3 $(0!(EXPMASK | SIGNBIT)),12(fp),r4 66 orl2 $(0!CLEARHID),r4 67 movl 16(fp),r5 68 shlq $7,r4,r4 /* leave the sign bit cleared. */ 69 70 /* in the following lp1 stands for least significant part of operand 1, 71 * lp2 for least significant part of operand 2, 72 * mp1 for most significant part of operand 1, 73 * mp2 for most significant part of operand 2. 74 */ 75 76 clrl r6 77 shrl $1,r1,r1 /* clear the sign bit of the lp1. */ 78 jeql 1f 79 emul r1,r4,$0,r6 /* r6,r7 <-- lp1*mp2 */ 80 shlq $1,r6,r6 /* to compensate for the shift we did to clear the sign bit. */ 811: shrl $1,r5,r5 /* clear the sign bit of the lp2. */ 82 jeql 1f 83 emul r0,r5,$0,r8 /* r8,r9 <-- mp1*lp2 */ 84 shlq $1,r8,r8 85 addl2 r9,r7 /* r6,r7 <-- the sum of the products. */ 86 adwc r8,r6 871: emul r0,r4,$0,r0 /* r0,r1 <-- mp1*mp2 */ 88 addl2 r6,r1 /* add the most sig. part of the sum. */ 89 adwc $0,r0 90 movl r0,r4 /* to see how much we realy need to shift. */ 91 movl $6,r5 /* r5 - shift counter. */ 92 shrl $7,r4,r4 /* dummy shift. */ 931: bbs $HIDDEN,r4,realshift 94 shll $1,r4,r4 95 decl r2 /* update exponent. */ 96 jeql underflow 97 decl r5 /* update shift counter. */ 98 jmp 1b 99realshift: 100 shrq r5,r0,r0 101 bbc $0,r1,shiftmore 102 incl r1 /* rounding. */ 103shiftmore: 104 shrq $1,r0,r0 105comb: 106 andl2 $CLEARHID,r0 107 shll $EXPSHIFT,r2,r4 108 orl2 r4,r0 109 cmpl r2,$256 110 jlss 1f 111 callf $4,fpover 112sign: 1131: bbc $0,r3,done 114 orl2 $SIGNBIT,r0 115done: ret 116 117 118 119is_res1: 120 bbc $31,4(fp),retzero 121 callf $4,fpresop 122 ret 123is_res2: 124 bbc $31,12(fp),retzero 125 callf $4,fpresop 126 ret 127 retzero: 128 clrl r0 129 clrl r1 130 ret 131 overflow: 132 callf $4,fpover 133 jmp sign 134 underflow: 135 callf $4,fpunder 136 ret 137