sint.s
上传用户:nvosite88
上传日期:2007-01-17
资源大小:4983k
文件大小:6k
- /* sint.s - Motorola 68040 FP integer routines (EXC) */
- /* Copyright 1991-1993 Wind River Systems, Inc. */
- .data
- .globl _copyright_wind_river
- .long _copyright_wind_river
- /*
- modification history
- --------------------
- 01e,21jul93,kdl added .text (SPR #2372).
- 01d,23aug92,jcf changed bxxx to jxx.
- 01c,26may92,rrr the tree shuffle
- 01b,10jan92,kdl added modification history; general cleanup.
- 01a,15aug91,kdl original version, from Motorola FPSP v2.0.
- */
- /*
- DESCRIPTION
- sintsa 3.1 12/10/90
- The entry point sINT computes the rounded integer
- equivalent of the input argument, sINTRZ computes
- the integer rounded to zero of the input argument.
- Entry points __x_sint and __x_sintrz are called from __x_do_func
- to emulate the fint and fintrz unimplemented instructions,
- respectively. Entry point __x_sintdo is used by __x_bindec.
- Input: (Entry points __x_sint and __x_sintrz) Double-extended
- number X in the ETEMP space in the floating-point
- save stack.
- (Entry point __x_sintdo) Double-extended number X in
- location pointed to by the address register a0.
- (Entry point __x_sintd) Double-extended denormalized
- number X in the ETEMP space in the floating-point
- save stack.
- Output: The function returns int(X) or intrz(X) in fp0.
- Modifies: fp0.
- Algorithm: (sint and __x_sintrz)
- 1. If exp(X) >= 63, return X.
- If exp(X) < 0, return +/- 0 or +/- 1, according to
- the rounding mode.
- 2. (X is in range) set rsc = 63 - exp(X). Unnormalize the
- result to the exponent 0x403e.
- 3. Round the result in the mode given in USER_FPCR. For
- __x_sintrz, force round-to-zero mode.
- 4. Normalize the rounded result| store in fp0.
- For the denormalized cases, force the correct result
- for the given sign and rounding mode.
- Sign(X)
- RMODE + -
- ----- --------
- RN +0 -0
- RZ +0 -0
- RM +0 -1
- RP +1 -0
- Copyright (C) Motorola, Inc. 1990
- All Rights Reserved
- THIS IS UNPUBLISHED PROPRIETARY SOURCE CODE OF MOTOROLA
- The copyright notice above does not evidence any
- actual or intended publication of such source code.
- SINT idnt 2,1 Motorola 040 Floating Point Software Package
- section 8
- NOMANUAL
- */
- #include "fpsp040E.h"
- | xref __x_dnrm_lp
- | xref __x_nrm_set
- | xref __x_round
- | xref __x_t_inx2
- | xref __x_ld_pone
- | xref __x_ld_mone
- | xref __x_ld_pzero
- | xref __x_ld_mzero
- | xref __x_snzrinx
- |
- | FINT
- |
- .text
- .globl __x_sint
- __x_sint:
- bfextu a6@(fpcr_MODE){#2:#2},d1 /* | use user's mode for rounding */
- | | implicity has extend precision
- | | in upper word.
- movel d1,a6@(L_SCR1) | save mode bits
- jra __x_sintexc
- |
- | FINT with extended denorm inputs.
- |
- .globl __x_sintd
- __x_sintd:
- btst #5,a6@(fpcr_MODE)
- jeq __x_snzrinx | if round nearest or round zero, +/- 0
- btst #4,a6@(fpcr_MODE)
- jeq rnd_mns
- rnd_pls:
- btst #sign_bit,a0@(LOCAL_EX)
- jne __x_sintmz
- bsrl __x_ld_pone | if round plus inf & pos, answer is +1
- jra __x_t_inx2
- rnd_mns:
- btst #sign_bit,a0@(LOCAL_EX)
- jeq __x_sintpz
- bsrl __x_ld_mone | if round mns inf and neg, answer is -1
- jra __x_t_inx2
- __x_sintpz:
- bsrl __x_ld_pzero
- jra __x_t_inx2
- __x_sintmz:
- bsrl __x_ld_mzero
- jra __x_t_inx2
- |
- | FINTRZ
- |
- .globl __x_sintrz
- __x_sintrz:
- movel #1,a6@(L_SCR1) | use rz mode for rounding
- | | implicity has extend precision
- | | in upper word.
- jra __x_sintexc
- |
- | SINTDO
- |
- | Input: a0 points to an IEEE extended format operand
- | Output: fp0 has the result
- |
- | Exeptions:
- |
- | If the subroutine results in an inexact operation, the inx2 and
- | ainx bits in the USER_FPSR are set.
- |
- |
- .globl __x_sintdo
- __x_sintdo:
- bfextu a6@(fpcr_MODE){#2:#2},d1 /* | use user's mode for rounding */
- | | implicitly has ext precision
- | | in upper word.
- movel d1,a6@(L_SCR1) | save mode bits
- |
- | Real work of __x_sint is in __x_sintexc
- |
- __x_sintexc:
- bclr #sign_bit,a0@(LOCAL_EX) | convert to internal extended
- | | format
- sne a0@(LOCAL_SGN)
- cmpw #0x403e,a0@(LOCAL_EX) | check if (unbiased) exp > 63
- jgt out_rnge | branch if exp < 63
- cmpw #0x3ffd,a0@(LOCAL_EX) | check if (unbiased) exp < 0
- jgt in_rnge | if 63 >= exp > 0, do calc
- |
- | Input is less than zero. Restore sign, and check for directed
- | rounding modes. L_SCR1 contains the rmode in the lower byte.
- |
- un_rnge:
- btst #1,a6@(L_SCR1+3) | check for rn and rz
- jeq un_rnrz
- tstb a0@(LOCAL_SGN) | check for sign
- jne un_rmrp_neg
- |
- | Sign is +. If rp, load +1.0, if rm, load +0.0
- |
- cmpib #3,a6@(L_SCR1+3) | check for rp
- jeq un_ldpone | if rp, load +1.0
- bsrl __x_ld_pzero | if rm, load +0.0
- jra __x_t_inx2
- un_ldpone:
- bsrl __x_ld_pone
- jra __x_t_inx2
- |
- | Sign is -. If rm, load -1.0, if rp, load -0.0
- |
- un_rmrp_neg:
- cmpib #2,a6@(L_SCR1+3) | check for rm
- jeq un_ldmone | if rm, load -1.0
- bsrl __x_ld_mzero | if rp, load -0.0
- jra __x_t_inx2
- un_ldmone:
- bsrl __x_ld_mone
- jra __x_t_inx2
- |
- | Rmode is rn or rz| return signed zero
- |
- un_rnrz:
- tstb a0@(LOCAL_SGN) | check for sign
- jne un_rnrz_neg
- bsrl __x_ld_pzero
- jra __x_t_inx2
- un_rnrz_neg:
- bsrl __x_ld_mzero
- jra __x_t_inx2
- |
- | Input is greater than 2^63. All bits are significant. Return
- | the input.
- |
- out_rnge:
- bfclr a0@(LOCAL_SGN){#0:#8} | change back to IEEE ext format
- jeq intps
- bset #sign_bit,a0@(LOCAL_EX)
- intps:
- fmovel fpcr,a7@-
- fmovel #0,fpcr
- fmovex a0@(LOCAL_EX),fp0 | if exp > 63
- | | then return X to the user
- | | there are no fraction bits
- fmovel a7@+,fpcr
- rts
- in_rnge:
- | | shift off fraction bits
- clrl d0 | clear d0 - initial g,r,s for
- | | dnrm_lp
- movel #0x403e,d1 | set threshold for __x_dnrm_lp
- | | assumes a0 points to operand
- bsrl __x_dnrm_lp
- | | returns unnormalized number
- | | pointed by a0
- | | output d0 supplies g,r,s
- | | used by round
- movel a6@(L_SCR1),d1 | use selected rounding mode
- |
- |
- bsrl __x_round | round the unnorm based on users
- | | input a0 ptr to ext X
- | | d0 g,r,s bits
- | | d1 PREC/MODE info
- | | output a0 ptr to rounded result
- | | inexact flag set in USER_FPSR
- | | if initial grs set
- |
- | normalize the rounded result and store value in fp0
- |
- bsrl __x_nrm_set | normalize the unnorm
- | | Input: a0 points to operand to
- | | be normalized
- | | Output: a0 points to normalized
- | | result
- bfclr a0@(LOCAL_SGN){#0:#8}
- jeq nrmrndp
- bset #sign_bit,a0@(LOCAL_EX) | return to IEEE extended format
- nrmrndp:
- fmovel fpcr,a7@-
- fmovel #0,fpcr
- fmovex a0@(LOCAL_EX),fp0 | move result to fp0
- fmovel a7@+,fpcr
- rts
- | end