[thirdparty/glibc.git] / sysdeps / sparc / addmul_1.S

! SPARC __mpn_addmul_1 -- Multiply a limb vector with a limb and add
! the result to a second limb vector.

! Copyright (C) 1992, 1993, 1994 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 Library General Public License as published by
! the Free Software Foundation; either version 2 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 Library General Public
! License for more details.

! You should have received a copy of the GNU Library General Public License
! along with the GNU MP Library; see the file COPYING.LIB.  If not, write to
! the Free Software Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.


! INPUT PARAMETERS
! res_ptr	o0
! s1_ptr	o1
! size		o2
! s2_limb	o3

#include "sysdep.h"

.text
	.align 4
	.global	C_SYMBOL_NAME(__mpn_addmul_1)
C_SYMBOL_NAME(__mpn_addmul_1):
	! Make S1_PTR and RES_PTR point at the end of their blocks
	! and put (- 4 x SIZE) in index/loop counter.
	sll	%o2,2,%o2
	add	%o0,%o2,%o4	! RES_PTR in o4 since o0 is retval
	add	%o1,%o2,%o1
	sub	%g0,%o2,%o2

	cmp	%o3,0xfff
	bgu	Large
	nop

	ld	[%o1+%o2],%o5
	mov	0,%o0
	b	L0
	 add	%o4,-4,%o4
Loop0:
	addcc	%o5,%g1,%g1
	ld	[%o1+%o2],%o5
	addx	%o0,%g0,%o0
	st	%g1,[%o4+%o2]
L0:	wr	%g0,%o3,%y
	sra	%o5,31,%g2
	and	%o3,%g2,%g2
	andcc	%g1,0,%g1
	mulscc	%g1,%o5,%g1
 	mulscc	%g1,%o5,%g1
 	mulscc	%g1,%o5,%g1
 	mulscc	%g1,%o5,%g1
	mulscc	%g1,%o5,%g1
	mulscc	%g1,%o5,%g1
	mulscc	%g1,%o5,%g1
	mulscc	%g1,%o5,%g1
	mulscc	%g1,%o5,%g1
	mulscc	%g1,%o5,%g1
	mulscc	%g1,%o5,%g1
	mulscc	%g1,%o5,%g1
	mulscc	%g1,0,%g1
	sra	%g1,20,%g4
	sll	%g1,12,%g1
 	rd	%y,%g3
	srl	%g3,20,%g3
	or	%g1,%g3,%g1

	addcc	%g1,%o0,%g1
	addx	%g2,%g4,%o0	! add sign-compensation and cy to hi limb
	addcc	%o2,4,%o2	! loop counter
	bne	Loop0
	 ld	[%o4+%o2],%o5

	addcc	%o5,%g1,%g1
	addx	%o0,%g0,%o0
	retl
	st	%g1,[%o4+%o2]


Large:	ld	[%o1+%o2],%o5
	mov	0,%o0
	sra	%o3,31,%g4	! g4 = mask of ones iff S2_LIMB < 0
	b	L1
	 add	%o4,-4,%o4
Loop:
	addcc	%o5,%g3,%g3
	ld	[%o1+%o2],%o5
	addx	%o0,%g0,%o0
	st	%g3,[%o4+%o2]
L1:	wr	%g0,%o5,%y
	and	%o5,%g4,%g2
	andcc	%g0,%g0,%g1
	mulscc	%g1,%o3,%g1
	mulscc	%g1,%o3,%g1
	mulscc	%g1,%o3,%g1
	mulscc	%g1,%o3,%g1
	mulscc	%g1,%o3,%g1
	mulscc	%g1,%o3,%g1
	mulscc	%g1,%o3,%g1
	mulscc	%g1,%o3,%g1
	mulscc	%g1,%o3,%g1
	mulscc	%g1,%o3,%g1
	mulscc	%g1,%o3,%g1
	mulscc	%g1,%o3,%g1
	mulscc	%g1,%o3,%g1
	mulscc	%g1,%o3,%g1
	mulscc	%g1,%o3,%g1
	mulscc	%g1,%o3,%g1
	mulscc	%g1,%o3,%g1
	mulscc	%g1,%o3,%g1
	mulscc	%g1,%o3,%g1
	mulscc	%g1,%o3,%g1
	mulscc	%g1,%o3,%g1
	mulscc	%g1,%o3,%g1
	mulscc	%g1,%o3,%g1
	mulscc	%g1,%o3,%g1
	mulscc	%g1,%o3,%g1
	mulscc	%g1,%o3,%g1
	mulscc	%g1,%o3,%g1
	mulscc	%g1,%o3,%g1
	mulscc	%g1,%o3,%g1
	mulscc	%g1,%o3,%g1
	mulscc	%g1,%o3,%g1
	mulscc	%g1,%o3,%g1
	mulscc	%g1,%g0,%g1
	rd	%y,%g3
	addcc	%g3,%o0,%g3
	addx	%g2,%g1,%o0
	addcc	%o2,4,%o2
	bne	Loop
	 ld	[%o4+%o2],%o5

	addcc	%o5,%g3,%g3
	addx	%o0,%g0,%o0
	retl
	st	%g3,[%o4+%o2]
Commit	Line	Data
28f540f4 RM	1	! SPARC __mpn_addmul_1 -- Multiply a limb vector with a limb and add
	2	! the result to a second limb vector.
	3
	4	! Copyright (C) 1992, 1993, 1994 Free Software Foundation, Inc.
	5
	6	! This file is part of the GNU MP Library.
	7
	8	! The GNU MP Library is free software; you can redistribute it and/or modify
	9	! it under the terms of the GNU Library General Public License as published by
	10	! the Free Software Foundation; either version 2 of the License, or (at your
	11	! option) any later version.
	12
	13	! The GNU MP Library is distributed in the hope that it will be useful, but
	14	! WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY
	15	! or FITNESS FOR A PARTICULAR PURPOSE. See the GNU Library General Public
	16	! License for more details.
	17
	18	! You should have received a copy of the GNU Library General Public License
	19	! along with the GNU MP Library; see the file COPYING.LIB. If not, write to
	20	! the Free Software Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
	21
	22
	23	! INPUT PARAMETERS
	24	! res_ptr o0
	25	! s1_ptr o1
	26	! size o2
	27	! s2_limb o3
	28
	29	#include "sysdep.h"
	30
	31	.text
	32	.align 4
	33	.global C_SYMBOL_NAME(__mpn_addmul_1)
	34	C_SYMBOL_NAME(__mpn_addmul_1):
	35	! Make S1_PTR and RES_PTR point at the end of their blocks
	36	! and put (- 4 x SIZE) in index/loop counter.
	37	sll %o2,2,%o2
	38	add %o0,%o2,%o4 ! RES_PTR in o4 since o0 is retval
	39	add %o1,%o2,%o1
	40	sub %g0,%o2,%o2
	41
	42	cmp %o3,0xfff
	43	bgu Large
	44	nop
	45
	46	ld [%o1+%o2],%o5
	47	mov 0,%o0
	48	b L0
	49	add %o4,-4,%o4
	50	Loop0:
	51	addcc %o5,%g1,%g1
	52	ld [%o1+%o2],%o5
	53	addx %o0,%g0,%o0
	54	st %g1,[%o4+%o2]
	55	L0: wr %g0,%o3,%y
	56	sra %o5,31,%g2
	57	and %o3,%g2,%g2
	58	andcc %g1,0,%g1
	59	mulscc %g1,%o5,%g1
	60	mulscc %g1,%o5,%g1
	61	mulscc %g1,%o5,%g1
	62	mulscc %g1,%o5,%g1
	63	mulscc %g1,%o5,%g1
	64	mulscc %g1,%o5,%g1
65	mulscc %g1,%o5,%g1
66	mulscc %g1,%o5,%g1
67	mulscc %g1,%o5,%g1
68	mulscc %g1,%o5,%g1
69	mulscc %g1,%o5,%g1
70	mulscc %g1,%o5,%g1
71	mulscc %g1,0,%g1
72	sra %g1,20,%g4
73	sll %g1,12,%g1
74	rd %y,%g3
75	srl %g3,20,%g3
76	or %g1,%g3,%g1
77
78	addcc %g1,%o0,%g1
79	addx %g2,%g4,%o0 ! add sign-compensation and cy to hi limb
80	addcc %o2,4,%o2 ! loop counter
81	bne Loop0
82	ld [%o4+%o2],%o5
83
84	addcc %o5,%g1,%g1
85	addx %o0,%g0,%o0
86	retl
87	st %g1,[%o4+%o2]
88
89
90	Large: ld [%o1+%o2],%o5
91	mov 0,%o0
92	sra %o3,31,%g4 ! g4 = mask of ones iff S2_LIMB < 0
93	b L1
94	add %o4,-4,%o4
95	Loop:
96	addcc %o5,%g3,%g3
97	ld [%o1+%o2],%o5
98	addx %o0,%g0,%o0
99	st %g3,[%o4+%o2]
100	L1: wr %g0,%o5,%y
101	and %o5,%g4,%g2
102	andcc %g0,%g0,%g1
103	mulscc %g1,%o3,%g1
104	mulscc %g1,%o3,%g1
105	mulscc %g1,%o3,%g1
106	mulscc %g1,%o3,%g1
107	mulscc %g1,%o3,%g1
108	mulscc %g1,%o3,%g1
109	mulscc %g1,%o3,%g1
110	mulscc %g1,%o3,%g1
111	mulscc %g1,%o3,%g1
112	mulscc %g1,%o3,%g1
113	mulscc %g1,%o3,%g1
114	mulscc %g1,%o3,%g1
115	mulscc %g1,%o3,%g1
116	mulscc %g1,%o3,%g1
117	mulscc %g1,%o3,%g1
118	mulscc %g1,%o3,%g1
119	mulscc %g1,%o3,%g1
120	mulscc %g1,%o3,%g1
121	mulscc %g1,%o3,%g1
122	mulscc %g1,%o3,%g1
123	mulscc %g1,%o3,%g1
124	mulscc %g1,%o3,%g1
125	mulscc %g1,%o3,%g1
126	mulscc %g1,%o3,%g1
127	mulscc %g1,%o3,%g1
128	mulscc %g1,%o3,%g1
129	mulscc %g1,%o3,%g1
130	mulscc %g1,%o3,%g1
131	mulscc %g1,%o3,%g1
132	mulscc %g1,%o3,%g1
133	mulscc %g1,%o3,%g1
134	mulscc %g1,%o3,%g1
135	mulscc %g1,%g0,%g1
136	rd %y,%g3
137	addcc %g3,%o0,%g3
138	addx %g2,%g1,%o0
139	addcc %o2,4,%o2
140	bne Loop
141	ld [%o4+%o2],%o5
142
143	addcc %o5,%g3,%g3
144	addx %o0,%g0,%o0
145	retl
146	st %g3,[%o4+%o2]