lib1funcs.asm

Mac OS X 10.4.9 for x86 Source Code gcc 实现源代码

@ libgcc routines for ARM cpu.
@ Division routines, written by Richard Earnshaw, ([e-mail])

/* Copyright 1995, 1996, 1998, 1999, 2000, 2003, 2004, 2005
   Free Software Foundation, Inc.

This file is free software; you can redistribute it and/or modify it
under the terms of the GNU General Public License as published by the
Free Software Foundation; either version 2, or (at your option) any
later version.

In addition to the permissions in the GNU General Public License, the
Free Software Foundation gives you unlimited permission to link the
compiled version of this file into combinations with other programs,
and to distribute those combinations without any restriction coming
from the use of this file.  (The General Public License restrictions
do apply in other respects; for example, they cover modification of
the file, and distribution when not linked into a combine
executable.)

This file is distributed in the hope that it will be useful, but
WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
General Public License for more details.

You should have received a copy of the GNU General Public License
along with this program; see the file COPYING.  If not, write to
the Free Software Foundation, 59 Temple Place - Suite 330,
Boston, MA 02111-1307, USA.  */
/* ------------------------------------------------------------------------ */

/* We need to know what prefix to add to function names.  */

#ifndef __USER_LABEL_PREFIX__
#error  __USER_LABEL_PREFIX__ not defined
#endif

/* ANSI concatenation macros.  */

#define CONCAT1(a, b) CONCAT2(a, b)
#define CONCAT2(a, b) a ## b

/* Use the right prefix for global labels.  */

#define SYM(x) CONCAT1 (__USER_LABEL_PREFIX__, x)

#ifdef __ELF__
#ifdef __thumb__
#define __PLT__  /* Not supported in Thumb assembler (for now).  */
#else
#define __PLT__ (PLT)
#endif
#define TYPE(x) .type SYM(x),function
#define SIZE(x) .size SYM(x), . - SYM(x)
#define LSYM(x) .x
#else
#define __PLT__
#define TYPE(x)
#define SIZE(x)
#define LSYM(x) x
#endif

/* Function end macros.  Variants for interworking.  */

@ This selects the minimum architecture level required.
#define __ARM_ARCH__ 3

#if defined(__ARM_ARCH_3M__) || defined(__ARM_ARCH_4__) \
	|| defined(__ARM_ARCH_4T__)
/* We use __ARM_ARCH__ set to 4 here, but in reality it's any processor with
   long multiply instructions.  That includes v3M.  */
# undef __ARM_ARCH__
# define __ARM_ARCH__ 4
#endif
	
#if defined(__ARM_ARCH_5__) || defined(__ARM_ARCH_5T__) \
	|| defined(__ARM_ARCH_5E__) || defined(__ARM_ARCH_5TE__) \
	|| defined(__ARM_ARCH_5TEJ__)
# undef __ARM_ARCH__
# define __ARM_ARCH__ 5
#endif

#if defined(__ARM_ARCH_6__) || defined(__ARM_ARCH_6J__) \
	|| defined(__ARM_ARCH_6K__) || defined(__ARM_ARCH_6Z__) \
	|| defined(__ARM_ARCH_6ZK__)
# undef __ARM_ARCH__
# define __ARM_ARCH__ 6
#endif

/* How to return from a function call depends on the architecture variant.  */

#if (__ARM_ARCH__ > 4) || defined(__ARM_ARCH_4T__)

# define RET		bx	lr
# define RETc(x)	bx##x	lr

/* Special precautions for interworking on armv4t.  */
# if (__ARM_ARCH__ == 4)

/* Always use bx, not ldr pc.  */
#  if (defined(__thumb__) || defined(__THUMB_INTERWORK__))
#    define __INTERWORKING__
#   endif /* __THUMB__ || __THUMB_INTERWORK__ */

/* Include thumb stub before arm mode code.  */
#  if defined(__thumb__) && !defined(__THUMB_INTERWORK__)
#   define __INTERWORKING_STUBS__
#  endif /* __thumb__ && !__THUMB_INTERWORK__ */

#endif /* __ARM_ARCH == 4 */

#else

# define RET		mov	pc, lr
# define RETc(x)	mov##x	pc, lr

#endif

/* Don't pass dirn, it's there just to get token pasting right.  */

.macro	RETLDM	regs=, cond=, dirn=ia
#if defined (__INTERWORKING__)
	.ifc "\regs",""
	ldr\cond	lr, [sp], #4
	.else
	ldm\cond\dirn	sp!, {\regs, lr}
	.endif
	bx\cond	lr
#else
	.ifc "\regs",""
	ldr\cond	pc, [sp], #4
	.else
	ldm\cond\dirn	sp!, {\regs, pc}
	.endif
#endif
.endm


.macro ARM_LDIV0
LSYM(Ldiv0):
	str	lr, [sp, #-4]!
	bl	SYM (__div0) __PLT__
	mov	r0, #0			@ About as wrong as it could be.
	RETLDM
.endm


.macro THUMB_LDIV0
LSYM(Ldiv0):
	push	{ lr }
	bl	SYM (__div0)
	mov	r0, #0			@ About as wrong as it could be.
#if defined (__INTERWORKING__)
	pop	{ r1 }
	bx	r1
#else
	pop	{ pc }
#endif
.endm

.macro FUNC_END name
	SIZE (__\name)
.endm

.macro DIV_FUNC_END name
LSYM(Ldiv0):
#ifdef __thumb__
	THUMB_LDIV0
#else
	ARM_LDIV0
#endif
	FUNC_END \name
.endm

.macro THUMB_FUNC_START name
	.globl	SYM (\name)
	TYPE	(\name)
	.thumb_func
SYM (\name):
.endm

/* Function start macros.  Variants for ARM and Thumb.  */

#ifdef __thumb__
#define THUMB_FUNC .thumb_func
#define THUMB_CODE .force_thumb
#else
#define THUMB_FUNC
#define THUMB_CODE
#endif
	
.macro FUNC_START name
	.text
	.globl SYM (__\name)
	TYPE (__\name)
	.align 0
	THUMB_CODE
	THUMB_FUNC
SYM (__\name):
.endm

/* Special function that will always be coded in ARM assembly, even if
   in Thumb-only compilation.  */

#if defined(__INTERWORKING_STUBS__)
.macro	ARM_FUNC_START name
	FUNC_START \name
	bx	pc
	nop
	.arm
/* A hook to tell gdb that we've switched to ARM mode.  Also used to call
   directly from other local arm routines.  */
_L__\name:		
.endm
#define EQUIV .thumb_set
/* Branch directly to a function declared with ARM_FUNC_START.
   Must be called in arm mode.  */
.macro  ARM_CALL name
	bl	_L__\name
.endm
#else
.macro	ARM_FUNC_START name
	.text
	.globl SYM (__\name)
	TYPE (__\name)
	.align 0
	.arm
SYM (__\name):
.endm
#define EQUIV .set
.macro  ARM_CALL name
	bl	__\name
.endm
#endif

.macro	FUNC_ALIAS new old
	.globl	SYM (__\new)
#if defined (__thumb__)
	.thumb_set	SYM (__\new), SYM (__\old)
#else
	.set	SYM (__\new), SYM (__\old)
#endif
.endm

.macro	ARM_FUNC_ALIAS new old
	.globl	SYM (__\new)
	EQUIV	SYM (__\new), SYM (__\old)
#if defined(__INTERWORKING_STUBS__)
	.set	SYM (_L__\new), SYM (_L__\old)
#endif
.endm

#ifdef __thumb__
/* Register aliases.  */

work		.req	r4	@ XXXX is this safe ?
dividend	.req	r0
divisor		.req	r1
overdone	.req	r2
result		.req	r2
curbit		.req	r3
#endif
#if 0
ip		.req	r12
sp		.req	r13
lr		.req	r14
pc		.req	r15
#endif

/* ------------------------------------------------------------------------ */
/*		Bodies of the division and modulo routines.		    */
/* ------------------------------------------------------------------------ */	
.macro ARM_DIV_BODY dividend, divisor, result, curbit

#if __ARM_ARCH__ >= 5 && ! defined (__OPTIMIZE_SIZE__)

	clz	\curbit, \dividend
	clz	\result, \divisor
	sub	\curbit, \result, \curbit
	rsbs	\curbit, \curbit, #31
	addne	\curbit, \curbit, \curbit, lsl #1
	mov	\result, #0
	addne	pc, pc, \curbit, lsl #2
	nop
	.set	shift, 32
	.rept	32
	.set	shift, shift - 1
	cmp	\dividend, \divisor, lsl #shift
	adc	\result, \result, \result
	subcs	\dividend, \dividend, \divisor, lsl #shift
	.endr

#else /* __ARM_ARCH__ < 5 || defined (__OPTIMIZE_SIZE__) */
#if __ARM_ARCH__ >= 5

	clz	\curbit, \divisor
	clz	\result, \dividend
	sub	\result, \curbit, \result
	mov	\curbit, #1
	mov	\divisor, \divisor, lsl \result
	mov	\curbit, \curbit, lsl \result
	mov	\result, #0
	
#else /* __ARM_ARCH__ < 5 */

	@ Initially shift the divisor left 3 bits if possible,
	@ set curbit accordingly.  This allows for curbit to be located
	@ at the left end of each 4 bit nibbles in the division loop
	@ to save one loop in most cases.
	tst	\divisor, #0xe0000000
	moveq	\divisor, \divisor, lsl #3
	moveq	\curbit, #8
	movne	\curbit, #1

	@ Unless the divisor is very big, shift it up in multiples of
	@ four bits, since this is the amount of unwinding in the main
	@ division loop.  Continue shifting until the divisor is 
	@ larger than the dividend.
1:	cmp	\divisor, #0x10000000
	cmplo	\divisor, \dividend
	movlo	\divisor, \divisor, lsl #4
	movlo	\curbit, \curbit, lsl #4
	blo	1b

	@ For very big divisors, we must shift it a bit at a time, or
	@ we will be in danger of overflowing.
1:	cmp	\divisor, #0x80000000
	cmplo	\divisor, \dividend
	movlo	\divisor, \divisor, lsl #1
	movlo	\curbit, \curbit, lsl #1
	blo	1b

	mov	\result, #0

#endif /* __ARM_ARCH__ < 5 */

	@ Division loop
1:	cmp	\dividend, \divisor
	subhs	\dividend, \dividend, \divisor
	orrhs	\result,   \result,   \curbit
	cmp	\dividend, \divisor,  lsr #1
	subhs	\dividend, \dividend, \divisor, lsr #1
	orrhs	\result,   \result,   \curbit,  lsr #1
	cmp	\dividend, \divisor,  lsr #2
	subhs	\dividend, \dividend, \divisor, lsr #2
	orrhs	\result,   \result,   \curbit,  lsr #2
	cmp	\dividend, \divisor,  lsr #3
	subhs	\dividend, \dividend, \divisor, lsr #3
	orrhs	\result,   \result,   \curbit,  lsr #3
	cmp	\dividend, #0			@ Early termination?
	movnes	\curbit,   \curbit,  lsr #4	@ No, any more bits to do?
	movne	\divisor,  \divisor, lsr #4
	bne	1b

#endif /* __ARM_ARCH__ < 5 || defined (__OPTIMIZE_SIZE__) */

.endm
/* ------------------------------------------------------------------------ */	
.macro ARM_DIV2_ORDER divisor, order

#if __ARM_ARCH__ >= 5

	clz	\order, \divisor
	rsb	\order, \order, #31

#else

	cmp	\divisor, #(1 << 16)
	movhs	\divisor, \divisor, lsr #16
	movhs	\order, #16
	movlo	\order, #0

	cmp	\divisor, #(1 << 8)
	movhs	\divisor, \divisor, lsr #8
	addhs	\order, \order, #8

	cmp	\divisor, #(1 << 4)
	movhs	\divisor, \divisor, lsr #4
	addhs	\order, \order, #4

	cmp	\divisor, #(1 << 2)
	addhi	\order, \order, #3
	addls	\order, \order, \divisor, lsr #1

#endif

.endm
/* ------------------------------------------------------------------------ */
.macro ARM_MOD_BODY dividend, divisor, order, spare

#if __ARM_ARCH__ >= 5 && ! defined (__OPTIMIZE_SIZE__)

	clz	\order, \divisor
	clz	\spare, \dividend
	sub	\order, \order, \spare
	rsbs	\order, \order, #31
	addne	pc, pc, \order, lsl #3
	nop
	.set	shift, 32
	.rept	32
	.set	shift, shift - 1
	cmp	\dividend, \divisor, lsl #shift
	subcs	\dividend, \dividend, \divisor, lsl #shift
	.endr

#else /* __ARM_ARCH__ < 5 || defined (__OPTIMIZE_SIZE__) */
#if __ARM_ARCH__ >= 5

	clz	\order, \divisor
	clz	\spare, \dividend
	sub	\order, \order, \spare
	mov	\divisor, \divisor, lsl \order
	
#else /* __ARM_ARCH__ < 5 */

	mov	\order, #0

	@ Unless the divisor is very big, shift it up in multiples of
	@ four bits, since this is the amount of unwinding in the main
	@ division loop.  Continue shifting until the divisor is 
	@ larger than the dividend.
1:	cmp	\divisor, #0x10000000
	cmplo	\divisor, \dividend
	movlo	\divisor, \divisor, lsl #4
	addlo	\order, \order, #4
	blo	1b

	@ For very big divisors, we must shift it a bit at a time, or
	@ we will be in danger of overflowing.
1:	cmp	\divisor, #0x80000000
	cmplo	\divisor, \dividend
	movlo	\divisor, \divisor, lsl #1
	addlo	\order, \order, #1
	blo	1b

#endif /* __ARM_ARCH__ < 5 */

	@ Perform all needed substractions to keep only the reminder.
	@ Do comparisons in batch of 4 first.
	subs	\order, \order, #3		@ yes, 3 is intended here
	blt	2f

1:	cmp	\dividend, \divisor
	subhs	\dividend, \dividend, \divisor
	cmp	\dividend, \divisor,  lsr #1
	subhs	\dividend, \dividend, \divisor, lsr #1
	cmp	\dividend, \divisor,  lsr #2
	subhs	\dividend, \dividend, \divisor, lsr #2
	cmp	\dividend, \divisor,  lsr #3
	subhs	\dividend, \dividend, \divisor, lsr #3
	cmp	\dividend, #1
	mov	\divisor, \divisor, lsr #4
	subges	\order, \order, #4
	bge	1b

	tst	\order, #3
	teqne	\dividend, #0
	beq	5f

	@ Either 1, 2 or 3 comparison/substractions are left.
2:	cmn	\order, #2
	blt	4f
	beq	3f
	cmp	\dividend, \divisor
	subhs	\dividend, \dividend, \divisor
	mov	\divisor,  \divisor,  lsr #1
3:	cmp	\dividend, \divisor
	subhs	\dividend, \dividend, \divisor
	mov	\divisor,  \divisor,  lsr #1
4:	cmp	\dividend, \divisor
	subhs	\dividend, \dividend, \divisor
5:

#endif /* __ARM_ARCH__ < 5 || defined (__OPTIMIZE_SIZE__) */

.endm
/* ------------------------------------------------------------------------ */
.macro THUMB_DIV_MOD_BODY modulo
	@ Load the constant 0x10000000 into our work register.
	mov	work, #1
	lsl	work, #28
LSYM(Loop1):
	@ Unless the divisor is very big, shift it up in multiples of
	@ four bits, since this is the amount of unwinding in the main
	@ division loop.  Continue shifting until the divisor is 
	@ larger than the dividend.
	cmp	divisor, work
	bhs	LSYM(Lbignum)
	cmp	divisor, dividend
	bhs	LSYM(Lbignum)
	lsl	divisor, #4
	lsl	curbit,  #4
	b	LSYM(Loop1)
LSYM(Lbignum):
	@ Set work to 0x80000000
	lsl	work, #3
LSYM(Loop2):
	@ For very big divisors, we must shift it a bit at a time, or
	@ we will be in danger of overflowing.
	cmp	divisor, work
	bhs	LSYM(Loop3)
	cmp	divisor, dividend
	bhs	LSYM(Loop3)
	lsl	divisor, #1
	lsl	curbit,  #1
	b	LSYM(Loop2)
LSYM(Loop3):
	@ Test for possible subtractions ...
  .if \modulo
	@ ... On the final pass, this may subtract too much from the dividend, 
	@ so keep track of which subtractions are done, we can fix them up 
	@ afterwards.
	mov	overdone, #0
	cmp	dividend, divisor
	blo	LSYM(Lover1)
	sub	dividend, dividend, divisor
LSYM(Lover1):
	lsr	work, divisor, #1
	cmp	dividend, work
	blo	LSYM(Lover2)
	sub	dividend, dividend, work
	mov	ip, curbit
	mov	work, #1
	ror	curbit, work
	orr	overdone, curbit
	mov	curbit, ip
LSYM(Lover2):
	lsr	work, divisor, #2
	cmp	dividend, work
	blo	LSYM(Lover3)
	sub	dividend, dividend, work
	mov	ip, curbit
	mov	work, #2
	ror	curbit, work
	orr	overdone, curbit
	mov	curbit, ip
LSYM(Lover3):
	lsr	work, divisor, #3
	cmp	dividend, work
	blo	LSYM(Lover4)
	sub	dividend, dividend, work
	mov	ip, curbit
	mov	work, #3
	ror	curbit, work
	orr	overdone, curbit
	mov	curbit, ip
LSYM(Lover4):
	mov	ip, curbit
  .else
	@ ... and note which bits are done in the result.  On the final pass,
	@ this may subtract too much from the dividend, but the result will be ok,
	@ since the "bit" will have been shifted out at the bottom.
	cmp	dividend, divisor
	blo	LSYM(Lover1)
	sub	dividend, dividend, divisor
	orr	result, result, curbit
LSYM(Lover1):
	lsr	work, divisor, #1
	cmp	dividend, work
	blo	LSYM(Lover2)
	sub	dividend, dividend, work
	lsr	work, curbit, #1
	orr	result, work
LSYM(Lover2):
	lsr	work, divisor, #2
	cmp	dividend, work
	blo	LSYM(Lover3)
	sub	dividend, dividend, work
	lsr	work, curbit, #2
	orr	result, work
LSYM(Lover3):
	lsr	work, divisor, #3
	cmp	dividend, work
	blo	LSYM(Lover4)
	sub	dividend, dividend, work
	lsr	work, curbit, #3
	orr	result, work
LSYM(Lover4):
  .endif
	
	cmp	dividend, #0			@ Early termination?
	beq	LSYM(Lover5)
	lsr	curbit,  #4			@ No, any more bits to do?
	beq	LSYM(Lover5)
	lsr	divisor, #4
	b	LSYM(Loop3)
LSYM(Lover5):
  .if \modulo
	@ Any subtractions that we should not have done will be recorded in
	@ the top three bits of "overdone".  Exactly which were not needed
	@ are governed by the position of the bit, stored in ip.
	mov	work, #0xe
	lsl	work, #28
	and	overdone, work
	beq	LSYM(Lgot_result)
	
	@ If we terminated early, because dividend became zero, then the 
	@ bit in ip will not be in the bottom nibble, and we should not
	@ perform the additions below.  We must test for this though
	@ (rather relying upon the TSTs to prevent the additions) since
	@ the bit in ip could be in the top two bits which might then match
	@ with one of the smaller RORs.
	mov	curbit, ip
	mov	work, #0x7
	tst	curbit, work
	beq	LSYM(Lgot_result)
	
	mov	curbit, ip
	mov	work, #3
	ror	curbit, work
	tst	overdone, curbit
	beq	LSYM(Lover6)
	lsr	work, divisor, #3
	add	dividend, work
LSYM(Lover6):
	mov	curbit, ip
	mov	work, #2