[thirdparty/glibc.git] / sysdeps / sparc / sparc32 / lshift.S

! Sparc __mpn_lshift --
!
! Copyright (C) 1995-2014 Free Software Foundation, Inc.
!
! This file is part of the GNU MP Library.
!
! The GNU MP Library is free software; you can redistribute it and/or modify
! it under the terms of the GNU Lesser General Public License as published by
! the Free Software Foundation; either version 2.1 of the License, or (at your
! option) any later version.
!
! The GNU MP Library 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 Lesser General Public
! License for more details.
!
! You should have received a copy of the GNU Lesser General Public License
! along with the GNU MP Library; see the file COPYING.LIB.  If not,
! see <http://www.gnu.org/licenses/>.


! INPUT PARAMETERS
! RES_PTR	%o0
! SRC_PTR	%o1
! SIZE		%o2
! CNT		%o3

#include <sysdep.h>

ENTRY(__mpn_lshift)
	sll	%o2,2,%g1
	add	%o1,%g1,%o1	! make %o1 point at end of src
	ld	[%o1-4],%g2	! load first limb
	sub	%g0,%o3,%o5	! negate shift count
	add	%o0,%g1,%o0	! make %o0 point at end of res
	add	%o2,-1,%o2
	andcc	%o2,4-1,%g4	! number of limbs in first loop
	srl	%g2,%o5,%g1	! compute function result
	be	LOC(0)		! if multiple of 4 limbs, skip first loop
	st	%g1,[%sp+80]

	sub	%o2,%g4,%o2	! adjust count for main loop

LOC(loop0):
	ld	[%o1-8],%g3
	add	%o0,-4,%o0
	add	%o1,-4,%o1
	addcc	%g4,-1,%g4
	sll	%g2,%o3,%o4
	srl	%g3,%o5,%g1
	mov	%g3,%g2
	or	%o4,%g1,%o4
	bne	LOC(loop0)
	 st	%o4,[%o0+0]

LOC(0):	tst	%o2
	be	LOC(end)
	 nop

LOC(loop):
	ld	[%o1-8],%g3
	add	%o0,-16,%o0
	addcc	%o2,-4,%o2
	sll	%g2,%o3,%o4
	srl	%g3,%o5,%g1

	ld	[%o1-12],%g2
	sll	%g3,%o3,%g4
	or	%o4,%g1,%o4
	st	%o4,[%o0+12]
	srl	%g2,%o5,%g1

	ld	[%o1-16],%g3
	sll	%g2,%o3,%o4
	or	%g4,%g1,%g4
	st	%g4,[%o0+8]
	srl	%g3,%o5,%g1

	ld	[%o1-20],%g2
	sll	%g3,%o3,%g4
	or	%o4,%g1,%o4
	st	%o4,[%o0+4]
	srl	%g2,%o5,%g1

	add	%o1,-16,%o1
	or	%g4,%g1,%g4
	bne	LOC(loop)
	 st	%g4,[%o0+0]

LOC(end):
	sll	%g2,%o3,%g2
	st	%g2,[%o0-4]
	retl
	ld	[%sp+80],%o0

END(__mpn_lshift)
Commit	Line	Data
f41c8091 UD	1	! Sparc __mpn_lshift --
f41c8091 UD	2	!
d4697bc9	3	! Copyright (C) 1995-2014 Free Software Foundation, Inc.
f41c8091	4	!
ba848785	5	! This file is part of the GNU MP Library.
f41c8091	6	!
ba848785	7	! The GNU MP Library is free software; you can redistribute it and/or modify
6d84f89a AJ	8	! it under the terms of the GNU Lesser General Public License as published by
6d84f89a AJ	9	! the Free Software Foundation; either version 2.1 of the License, or (at your
ba848785	10	! option) any later version.
f41c8091	11	!
ba848785 RM	12	! The GNU MP Library is distributed in the hope that it will be useful, but
ba848785 RM	13	! WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY
6d84f89a	14	! or FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public
ba848785	15	! License for more details.
f41c8091	16	!
6d84f89a	17	! You should have received a copy of the GNU Lesser General Public License
59ba27a6 PE	18	! along with the GNU MP Library; see the file COPYING.LIB. If not,
59ba27a6 PE	19	! see <http://www.gnu.org/licenses/>.
ba848785 RM	20
	21
	22	! INPUT PARAMETERS
f41c8091 UD	23	! RES_PTR %o0
	24	! SRC_PTR %o1
	25	! SIZE %o2
	26	! CNT %o3
ba848785	27
f41c8091	28	#include <sysdep.h>
ba848785	29
f41c8091	30	ENTRY(__mpn_lshift)
ba848785 RM	31	sll %o2,2,%g1
	32	add %o1,%g1,%o1 ! make %o1 point at end of src
	33	ld [%o1-4],%g2 ! load first limb
	34	sub %g0,%o3,%o5 ! negate shift count
	35	add %o0,%g1,%o0 ! make %o0 point at end of res
	36	add %o2,-1,%o2
	37	andcc %o2,4-1,%g4 ! number of limbs in first loop
	38	srl %g2,%o5,%g1 ! compute function result
f41c8091	39	be LOC(0) ! if multiple of 4 limbs, skip first loop
ba848785 RM	40	st %g1,[%sp+80]
	41
	42	sub %o2,%g4,%o2 ! adjust count for main loop
	43
f41c8091 UD	44	LOC(loop0):
f41c8091 UD	45	ld [%o1-8],%g3
ba848785 RM	46	add %o0,-4,%o0
	47	add %o1,-4,%o1
	48	addcc %g4,-1,%g4
	49	sll %g2,%o3,%o4
	50	srl %g3,%o5,%g1
	51	mov %g3,%g2
	52	or %o4,%g1,%o4
f41c8091	53	bne LOC(loop0)
ba848785 RM	54	st %o4,[%o0+0]
ba848785 RM	55
f41c8091 UD	56	LOC(0): tst %o2
f41c8091 UD	57	be LOC(end)
ba848785 RM	58	nop
ba848785 RM	59
f41c8091 UD	60	LOC(loop):
f41c8091 UD	61	ld [%o1-8],%g3
ba848785 RM	62	add %o0,-16,%o0
	63	addcc %o2,-4,%o2
	64	sll %g2,%o3,%o4
	65	srl %g3,%o5,%g1
	66
	67	ld [%o1-12],%g2
	68	sll %g3,%o3,%g4
	69	or %o4,%g1,%o4
	70	st %o4,[%o0+12]
	71	srl %g2,%o5,%g1
	72
	73	ld [%o1-16],%g3
	74	sll %g2,%o3,%o4
	75	or %g4,%g1,%g4
	76	st %g4,[%o0+8]
	77	srl %g3,%o5,%g1
	78
	79	ld [%o1-20],%g2
	80	sll %g3,%o3,%g4
	81	or %o4,%g1,%o4
	82	st %o4,[%o0+4]
	83	srl %g2,%o5,%g1
	84
	85	add %o1,-16,%o1
	86	or %g4,%g1,%g4
f41c8091	87	bne LOC(loop)
ba848785 RM	88	st %g4,[%o0+0]
ba848785 RM	89
f41c8091 UD	90	LOC(end):
f41c8091 UD	91	sll %g2,%o3,%g2
ba848785 RM	92	st %g2,[%o0-4]
	93	retl
	94	ld [%sp+80],%o0
f41c8091 UD	95
f41c8091 UD	96	END(__mpn_lshift)