📄 d_support.s
字号:
#ifndef lint#static char *sccsid = "@(#)d_support.s 4.1 (ULTRIX) 7/17/90";#endif lint/************************************************************************ * * * Copyright (c) 1986 by * * Digital Equipment Corporation, Maynard, MA * * All rights reserved. * * * * This software is furnished under a license and may be used and * * copied only in accordance with the terms of such license and * * with the inclusion of the above copyright notice. This * * software or any other copies thereof may not be provided or * * otherwise made available to any other person. No title to and * * ownership of the software is hereby transferred. * * * * This software is derived from software received from the * * University of California, Berkeley, and from Bell * * Laboratories. Use, duplication, or disclosure is subject to * * restrictions under license agreements with University of * * California and with AT&T. * * * * The information in this software is subject to change without * * notice and should not be construed as a commitment by Digital * * Equipment Corporation. * * * * Digital assumes no responsibility for the use or reliability * * of its software on equipment which is not supplied by Digital. * * * ************************************************************************//************************************************************************** Modification History** Jon Reeves, 1989-Nov-28* 003 Set errno in scalb on underflow, too.** Tim N 12-Oct-89* 002 Changed scalb to return HUGE_VAL errno=ERANGE on ovfl (instead of* calling infnan() to drop core).** David Metsky, 12/18/86* 001 Adapted from support.s 1.3 (Berkeley) 8/21/85**************************************************************************//* @(#)support.s 1.3 (Berkeley) 8/21/85 * * copysign(x,y), * logb(x), * scalb(x,N), * finite(x), * drem(x,y), * Coded in vax assembly language by K.C. Ng, 3/14/85. * Revised by K.C. Ng on 4/9/85. *//* * double copysign(x,y) * double x,y; */ .globl _copysign .text .align 1_copysign: .word 0x4 movq 4(ap),r0 # load x into r0 bicw3 $0x807f,r0,r2 # mask off the exponent of x beql Lz # if zero or reserved op then return x bicw3 $0x7fff,12(ap),r2 # copy the sign bit of y into r2 bicw2 $0x8000,r0 # replace x by |x| bisw2 r2,r0 # copy the sign bit of y to xLz: ret/* * double logb(x) * double x; */ .globl _logb .text .align 1_logb: .word 0x0 bicl3 $0xffff807f,4(ap),r0 # mask off the exponent of x beql Ln ashl $-7,r0,r0 # get the bias exponent subl2 $129,r0 # get the unbias exponent cvtld r0,r0 # return the answer in double retLn: movq 4(ap),r0 # r0:1 = x (zero or reserved op) bneq 1f # simply return if reserved op movq $0,r0 # else return zero if zero1: ret/* * long finite(x) * double x; */ .globl _finite .text .align 1_finite: .word 0x0000 bicw3 $0x7f,4(ap),r0 # mask off the mantissa cmpw r0,$0x8000 # to see if x is the reserved op beql 1f # if so, return FALSE (0) movl $1,r0 # else return TRUE (1) ret1: clrl r0 ret/* * double scalb(x,N) * double x; int N; */ .globl _scalb .set ERANGE,34 .text .align 1_scalb: .word 0xc movq 4(ap),r0 bicl3 $0xffff807f,r0,r3 beql ret1 # 0 or reserved operand movl 12(ap),r2 cmpl r2,$0x12c bgeq ovfl cmpl r2,$-0x12c bleq unfl ashl $7,r2,r2 addl2 r2,r3 bleq unfl cmpl r3,$0x8000 bgeq ovfl addl2 r2,r0 retovfl:# calls $1,_infnan # if it returns# bicw3 $0x7fff,4(ap),r2 # get the sign of input arg# bisw2 r2,r0 # re-attach the sign to r0/1 movl $ERANGE,_errno movl $0xffff7fff,r0 movl $0xffffffff,r1 # return HUGE_VAL retunfl: movq $0,r0 movl $ERANGE,_errnoret1: ret/* * DREM(X,Y) * RETURN X REM Y =X-N*Y, N=[X/Y] ROUNDED (ROUNDED TO EVEN IN THE HALF WAY CASE) * DOUBLE PRECISION (VAX D format 56 bits) * CODED IN VAX ASSEMBLY LANGUAGE BY K.C. NG, 4/8/85. */ .globl _drem .set EDOM,33 .text .align 1_drem: .word 0xffc subl2 $12,sp movq 4(ap),r0 #r0=x movq 12(ap),r2 #r2=y jeql Rop #if y=0 then generate reserved op fault bicw3 $0x007f,r0,r4 #check if x is Rop cmpw r4,$0x8000 jeql Ret #if x is Rop then return Rop bicl3 $0x007f,r2,r4 #check if y is Rop cmpw r4,$0x8000 jeql Ret #if y is Rop then return Rop bicw2 $0x8000,r2 #y := |y| movw $0,-4(fp) #-4(fp) = nx := 0 cmpw r2,$0x1c80 #yexp ? 57 bgtr C1 #if yexp > 57 goto C1 addw2 $0x1c80,r2 #scale up y by 2**57 movw $0x1c80,-4(fp) #nx := 57 (exponent field)C1: movw -4(fp),-8(fp) #-8(fp) = nf := nx bicw3 $0x7fff,r0,-12(fp) #-12(fp) = sign of x bicw2 $0x8000,r0 #x := |x| movq r2,r10 #y1 := y bicl2 $0xffff07ff,r11 #clear the last 27 bits of y1loop: cmpd r0,r2 #x ? y bleq E1 #if x <= y goto E1 /* begin argument reduction */ movq r2,r4 #t =y movq r10,r6 #t1=y1 bicw3 $0x807f,r0,r8 #xexp= exponent of x bicw3 $0x807f,r2,r9 #yexp= exponent fo y subw2 r9,r8 #xexp-yexp subw2 $0x0c80,r8 #k=xexp-yexp-25(exponent bit field) blss C2 #if k<0 goto C2 addw2 r8,r4 #t +=k addw2 r8,r6 #t1+=k, scale up t and t1C2: divd3 r4,r0,r8 #x/t cvtdl r8,r8 #n=[x/t] truncated cvtld r8,r8 #float(n) subd2 r6,r4 #t:=t-t1 muld2 r8,r4 #n*(t-t1) muld2 r8,r6 #n*t1 subd2 r6,r0 #x-n*t1 subd2 r4,r0 #(x-n*t1)-n*(t-t1) brb loopE1: movw -4(fp),r6 #r6=nx beql C3 #if nx=0 goto C3 addw2 r6,r0 #x:=x*2**57 scale up x by nx movw $0,-4(fp) #clear nx brb loopC3: movq r2,r4 #r4 = y subw2 $0x80,r4 #r4 = y/2 cmpd r0,r4 #x:y/2 blss E2 #if x < y/2 goto E2 bgtr C4 #if x > y/2 goto C4 cvtdl r8,r8 #ifix(float(n)) blbc r8,E2 #if the last bit is zero, goto E2C4: subd2 r2,r0 #x-yE2: xorw2 -12(fp),r0 #x^sign (exclusive or) movw -8(fp),r6 #r6=nf bicw3 $0x807f,r0,r8 #r8=exponent of x bicw2 $0x7f80,r0 #clear the exponent of x subw2 r6,r8 #r8=xexp-nf bgtr C5 #if xexp-nf is positive goto C5 movw $0,r8 #clear r8 movq $0,r0 #x underflow to zeroC5: bisw2 r8,r0 #put r8 into x's exponent field retRop: #Reserved operand movl $EDOM,_errno# calls $1,_infnan #generate reserved op fault movl $0,r0 #return zero, not Rop retRet: movq $0x8000,r0 #propagate reserved op ret
⌨️ 快捷键说明
复制代码
Ctrl + C
搜索代码
Ctrl + F
全屏模式
F11
切换主题
Ctrl + Shift + D
显示快捷键
?
增大字号
Ctrl + =
减小字号
Ctrl + -