round.s

VXWORKS源代码

	jeq 	ms_clr		| branch if no bits were set
	cmpw	d3,d0		| of X>Y
	jmi 	greater		| then exp will go past 0 (neg) if
|				| it is just shifted
	bsrl	__x_nrm_set		/* | else exp won't go past 0 */
	moveml	a7@+,d2/d3/d5/d6
	rts
greater:
	movel	d2,d6		| save ls mant in d6
	lsll	d0,d2		| shift ls mant by count
	lsll	d0,d1		| shift ms mant by count
	movel	#32,d5
	subl	d0,d5		| make op a denorm by shifting bits
	lsrl	d5,d6		| by the number in the exp, then
|				| set exp = 0.
	orl	d6,d1		| shift the ls mant bits into the ms mant
	movel	#0,d0		| same as if decremented exp to 0
|				| while shifting
	movew	d0,a0@(LOCAL_EX)
	movel	d1,a0@(LOCAL_HI)
	movel	d2,a0@(LOCAL_LO)
	moveml	a7@+,d2/d3/d5/d6
	rts
ms_clr:
	bfffo	d2{#0:#32},d3	| check if any bits set in ls mant
	jeq 	all_clr		| branch if none set
	addw	#32,d3
	cmpw	d3,d0		| if X>Y
	jmi 	greater		| then branch
	bsrl	__x_nrm_set		/* | else exp won't go past 0 */
	moveml	a7@+,d2/d3/d5/d6
	rts
all_clr:
	movew	#0,a0@(LOCAL_EX)	| no mantissa bits set. Set exp = 0.
	moveml	a7@+,d2/d3/d5/d6
	rts
|
|	NRM_SET
|
	.globl	__x_nrm_set
__x_nrm_set:
	movel	d7,a7@-
	bfffo	a0@(LOCAL_HI){#0:#32},d7 | find first 1 in ms mant to d7)
	jeq 	lower		/* | branch if ms mant is all 0's */

	movel	d6,a7@-

	subw	d7,a0@(LOCAL_EX)	| sub exponent by count
	movel	a0@(LOCAL_HI),d0	| d0 has ms mant
	movel	a0@(LOCAL_LO),d1 | d1 has ls mant

	lsll	d7,d0		| shift first 1 to j bit position
	movel	d1,d6		| copy ls mant into d6
	lsll	d7,d6		| shift ls mant by count
	movel	d6,a0@(LOCAL_LO)	| store ls mant into memory
	moveql	#32,d6
	subl	d7,d6		| continue shift
	lsrl	d6,d1		| shift off all bits but those that will
|				| be shifted into ms mant
	orl	d1,d0		| shift the ls mant bits into the ms mant
	movel	d0,a0@(LOCAL_HI)	| store ms mant into memory
	moveml	a7@+,d7/d6	| restore registers
	rts

|
| We get here if ms mant was = 0, and we assume ls mant has bits
| set (otherwise this would have been tagged a zero not a denorm).
|
lower:
	movew	a0@(LOCAL_EX),d0	| d0 has exponent
	movel	a0@(LOCAL_LO),d1	| d1 has ls mant
	subw	#32,d0		| account for ms mant being all zeros
	bfffo	d1{#0:#32},d7	| find first 1 in ls mant to d7)
	subw	d7,d0		| subtract shift count from exp
	lsll	d7,d1		| shift first 1 to integer bit in ms mant
	movew	d0,a0@(LOCAL_EX)	| store ms mant
	movel	d1,a0@(LOCAL_HI)	| store exp
	clrl	a0@(LOCAL_LO)	| clear ls mant
	movel	a7@+,d7
	rts
|
|	__x_denorm --- denormalize an intermediate result
|
|	Used by underflow.
|
| Input:
|	a0	 points to the operand to be denormalized
|		 (in the internal extended format)
|
|	d0: 	 rounding precision
| Output:
|	a0	 points to the denormalized result
|		 (in the internal extended format)
|
|	d0 	is guard,round,sticky
|
| d0 comes into this routine with the rounding precision. It
| is then loaded with the denormalized exponent threshold for the
| rounding precision.
|

	.globl	__x_denorm
__x_denorm:
	btst	#6,a0@(LOCAL_EX)	| check for exponents between 0x7fff-0x4000
	jeq 	no_sgn_ext
	bset	#7,a0@(LOCAL_EX)	| sign extend if it is so
no_sgn_ext:

	cmpib	#0,d0		| if 0 then extended precision
	jne 	not_ext		| else branch

	clrl	d1		| load d1 with ext threshold
	clrl	d0		| clear the sticky flag
	bsrl	__x_dnrm_lp		| denormalize the number
	tstb	d1		| check for inex
	jeq 	no_inex		| if clr, no inex
	jra 	dnrm_inex	| if set, set inex

not_ext:
	cmpil	#1,d0		| if 1 then single precision
	jeq 	load_sgl	| else must be 2, double prec

load_dbl:
	movew	#dbl_thresh,d1	| put copy of threshold in d1
	movel	d1,d0		| copy d1 into d0
	subw	a0@(LOCAL_EX),d0	| diff = threshold - exp
	cmpw	#67,d0		| if diff > 67 (mant + grs bits)
	jpl 	chk_stky	| then branch (all bits would be
|				|  shifted off in denorm routine)
	clrl	d0		| else clear the sticky flag
	bsrl	__x_dnrm_lp		| denormalize the number
	tstb	d1		| check flag
	jeq 	no_inex		| if clr, no inex
	jra 	dnrm_inex	| if set, set inex

load_sgl:
	movew	#sgl_thresh,d1	| put copy of threshold in d1
	movel	d1,d0		| copy d1 into d0
	subw	a0@(LOCAL_EX),d0	| diff = threshold - exp
	cmpw	#67,d0		| if diff > 67 (mant + grs bits)
	jpl 	chk_stky	| then branch (all bits would be
|				|  shifted off in __x_denorm routine)
	clrl	d0		| else clear the sticky flag
	bsrl	__x_dnrm_lp		| denormalize the number
	tstb	d1		| check flag
	jeq 	no_inex		| if clr, no inex
	jra 	dnrm_inex	| if set, set inex

chk_stky:
	tstl	a0@(LOCAL_HI)	| check for any bits set
	jne 	set_stky
	tstl	a0@(LOCAL_LO)	| check for any bits set
	jne 	set_stky
	jra 	clr_mant
set_stky:
	orl	#inx2a_mask,a6@(USER_FPSR) | set inex2/ainex
	movel	#0x20000000,d0	| set sticky bit in return value
clr_mant:
	movew	d1,a0@(LOCAL_EX)		| load exp with threshold
	movel	#0,a0@(LOCAL_HI) 	| set d1 = 0 (ms mantissa)
	movel	#0,a0@(LOCAL_LO)		| set d2 = 0 (ms mantissa)
	rts
dnrm_inex:
	orl	#inx2a_mask,a6@(USER_FPSR) | set inex2/ainex
no_inex:
	rts

|
|	__x_dnrm_lp --- normalize exponent/mantissa to specified threshhold
|
| Input:
|	a0		points to the operand to be denormalized
|	d0{31:29} 	initial guard,round,sticky
|	d1{15:0}	denormalization threshold
| Output:
|	a0		points to the denormalized operand
|	d0{31:29}	final guard,round,sticky
|	d1b		inexact flag:  all ones means inexact result
|
| The LOCAL_LO and LOCAL_GRS parts of the value are copied to FP_SCR2
| so that bfext can be used to extract the new low part of the mantissa.
| Dnrm_lp can be called with a0 pointing to ETEMP or WBTEMP and there
| is no LOCAL_GRS scratch word following it on the fsave frame.
|
	.globl	__x_dnrm_lp
__x_dnrm_lp:
	movel	d2,a7@-		| save d2 for temp use
	btst	#E3,a6@(E_BYTE)		| test for type E3 exception
	jeq 	not_E3			| not type E3 exception
	clrl	d0			| guard,round,sticky init.
	bfextu	a6@(WBTEMP_GRS){#6:#3},d2	| extract guard,round, sticky  bit

|  The following bit shift specifies a larger number of bits than actually
|  allowed (8).  So, it was divided into a series of smaller shifts.
| 	lsll	#29,d2			| original shift g,r,s to their postions
|
|
|	lsll	#8,d2			| shift g,r,s 8 bits
|	lsll	#8,d2			| shift g,r,s 8 bits (total 16)
|	lsll	#8,d2			| shift g,r,s 8 bits (total 24)
|	lsll	#5,d2			| shift g,r,s 5 bits (total 29)
|	orl	d2,d0			| in d0

	movel	#29,d0
	lsll	d0,d2			| shift g,r,s to their postions
	movel	d2,d0

not_E3:
	movel	a7@+,d2		| restore d2
	movel	a0@(LOCAL_LO),a6@(FP_SCR2+LOCAL_LO)
	movel	d0,a6@(FP_SCR2+LOCAL_GRS)
	movel	d1,d0			| copy the denorm threshold
	subw	a0@(LOCAL_EX),d1		| d1 = threshold - uns exponent
	jle 	no_lp			| d1 <= 0
	cmpw	#32,d1
	jlt 	case_1			| 0 = d1 < 32
	cmpw	#64,d1
	jlt 	case_2			| 32 <= d1 < 64
	jra 	case_3			| d1 >= 64
|
| No normalization necessary
|
no_lp:
	clrb	d1			| set no inex2 reported
	movel	a6@(FP_SCR2+LOCAL_GRS),d0	| restore original g,r,s
	rts
|
| case (0<d1<32)
|
case_1:
	movel	d2,a7@-
	movew	d0,a0@(LOCAL_EX)		| exponent = denorm threshold
	movel	#32,d0
	subw	d1,d0			| d0 = 32 - d1
	bfextu	a0@(LOCAL_EX){d0:#32},d2
	bfextu	d2{d1:d0},d2		| d2 = new LOCAL_HI
	bfextu	a0@(LOCAL_HI){d0:#32},d1	| d1 = new LOCAL_LO
	bfextu	a6@(FP_SCR2+LOCAL_LO){d0:#32},d0	| d0 = new G,R,S
	movel	d2,a0@(LOCAL_HI)		| store new LOCAL_HI
	movel	d1,a0@(LOCAL_LO)		| store new LOCAL_LO
	clrb	d1
	bftst	d0{#2:#30}
	jeq 	c1nstky
	bset	#rnd_stky_bit,d0
	st	d1
c1nstky:
	movel	a6@(FP_SCR2+LOCAL_GRS),d2	| restore original g,r,s
	andil	#0xe0000000,d2		| clear all but G,R,S
	tstl	d2			| test if original G,R,S are clear
	jeq 	grs_clear
	orl	#0x20000000,d0		| set sticky bit in d0
grs_clear:
	andil	#0xe0000000,d0		| clear all but G,R,S
	movel	a7@+,d2
	rts
|
| case (32<=d1<64)
|
case_2:
	movel	d2,a7@-
	movew	d0,a0@(LOCAL_EX)		| unsigned exponent = threshold
	subw	#32,d1			| d1 now between 0 and 32
	movel	#32,d0
	subw	d1,d0			| d0 = 32 - d1
	bfextu	a0@(LOCAL_EX){d0:#32},d2
	bfextu	d2{d1:d0},d2		| d2 = new LOCAL_LO
	bfextu	a0@(LOCAL_HI){d0:#32},d1	| d1 = new G,R,S
	bftst	d1{#2:#30}
	jne 	c2_sstky		| jra  if sticky bit to be set
	bftst	a6@(FP_SCR2+LOCAL_LO){d0:#32}
	jne 	c2_sstky		| jra  if sticky bit to be set
	movel	d1,d0
	clrb	d1
	jra 	end_c2
c2_sstky:
	movel	d1,d0
	bset	#rnd_stky_bit,d0
	st	d1
end_c2:
	clrl	a0@(LOCAL_HI)		| store LOCAL_HI = 0
	movel	d2,a0@(LOCAL_LO)		| store LOCAL_LO
	movel	a6@(FP_SCR2+LOCAL_GRS),d2	| restore original g,r,s
	andil	#0xe0000000,d2		| clear all but G,R,S
	tstl	d2			| test if original G,R,S are clear
	jeq 	clear_grs
	orl	#0x20000000,d0		| set sticky bit in d0
clear_grs:
	andil	#0xe0000000,d0		| get rid of all but G,R,S
	movel	a7@+,d2
	rts
|
| d1 >= 64 Force the exponent to be the denorm threshold with the
| correct sign.
|
case_3:
	movew	d0,a0@(LOCAL_EX)
	tstw	a0@(LOCAL_SGN)
	jge 	c3con
c3neg:
	orl	#0x80000000,a0@(LOCAL_EX)
c3con:
	cmpw	#64,d1
	jeq 	sixty_four
	cmpw	#65,d1
	jeq 	sixty_five
|
| Shift value is out of range.  Set d1 for inex2 flag and
| return a zero with the given threshold.
|
	clrl	a0@(LOCAL_HI)
	clrl	a0@(LOCAL_LO)
	movel	#0x20000000,d0
	st	d1
	rts

sixty_four:
	movel	a0@(LOCAL_HI),d0
	bfextu	d0{#2:#30},d1
	andil	#0xc0000000,d0
	jra 	c3com

sixty_five:
	movel	a0@(LOCAL_HI),d0
	bfextu	d0{#1:#31},d1
	andil	#0x80000000,d0
	lsrl	#1,d0			| shift high bit into R bit

c3com:
	tstl	d1
	jne 	c3ssticky
	tstl	a0@(LOCAL_LO)
	jne 	c3ssticky
	tstb	a6@(FP_SCR2+LOCAL_GRS)
	jne 	c3ssticky
	clrb	d1
	jra 	c3end

c3ssticky:
	bset	#rnd_stky_bit,d0
	st	d1
c3end:
	clrl	a0@(LOCAL_HI)
	clrl	a0@(LOCAL_LO)
	rts

|	end