1dnl  ARM mpn_mod_1_1p
2
3dnl  Contributed to the GNU project by Torbj��rn Granlund.
4
5dnl  Copyright 2012 Free Software Foundation, Inc.
6
7dnl  This file is part of the GNU MP Library.
8dnl
9dnl  The GNU MP Library is free software; you can redistribute it and/or modify
10dnl  it under the terms of either:
11dnl
12dnl    * the GNU Lesser General Public License as published by the Free
13dnl      Software Foundation; either version 3 of the License, or (at your
14dnl      option) any later version.
15dnl
16dnl  or
17dnl
18dnl    * the GNU General Public License as published by the Free Software
19dnl      Foundation; either version 2 of the License, or (at your option) any
20dnl      later version.
21dnl
22dnl  or both in parallel, as here.
23dnl
24dnl  The GNU MP Library is distributed in the hope that it will be useful, but
25dnl  WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY
26dnl  or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License
27dnl  for more details.
28dnl
29dnl  You should have received copies of the GNU General Public License and the
30dnl  GNU Lesser General Public License along with the GNU MP Library.  If not,
31dnl  see https://www.gnu.org/licenses/.
32
33include(`../config.m4')
34
35C	     cycles/limb
36C StrongARM	 -
37C XScale	 ?
38C Cortex-A7	 ?
39C Cortex-A8	 ?
40C Cortex-A9	 7
41C Cortex-A15	 6
42
43define(`ap', `r0')
44define(`n',  `r1')
45define(`d',  `r2')
46define(`cps',`r3')
47
48ASM_START()
49PROLOGUE(mpn_mod_1_1p)
50	push	{r4-r10}
51	add	r0, r0, r1, asl #2
52	ldr	r5, [r0, #-4]!
53	ldr	r12, [r0, #-4]!
54	subs	r1, r1, #2
55	ble	L(4)
56	ldr	r8, [r3, #12]
57	mov	r4, r12
58	mov	r10, r5
59	umull	r7, r5, r10, r8
60	sub	r1, r1, #1
61	b	L(mid)
62
63L(top):	adds	r12, r6, r7
64	adcs	r10, r4, r5
65	sub	r1, r1, #1
66	mov	r6, #0
67	movcs	r6, r8
68	umull	r7, r5, r10, r8
69	adds	r4, r12, r6
70	subcs	r4, r4, r2
71L(mid):	ldr	r6, [r0, #-4]!
72	teq	r1, #0
73	bne	L(top)
74
75	adds	r12, r6, r7
76	adcs	r5, r4, r5
77	subcs	r5, r5, r2
78L(4):	ldr	r1, [r3, #4]
79	cmp	r1, #0
80	beq	L(7)
81	ldr	r4, [r3, #8]
82	umull	r0, r6, r5, r4
83	adds	r12, r0, r12
84	addcs	r6, r6, #1
85	rsb	r0, r1, #32
86	mov	r0, r12, lsr r0
87	orr	r5, r0, r6, asl r1
88	mov	r12, r12, asl r1
89	b	L(8)
90L(7):	cmp	r5, r2
91	subcs	r5, r5, r2
92L(8):	ldr	r0, [r3, #0]
93	umull	r4, r3, r5, r0
94	add	r5, r5, #1
95	adds	r0, r4, r12
96	adc	r5, r3, r5
97	mul	r5, r2, r5
98	sub	r12, r12, r5
99	cmp	r12, r0
100	addhi	r12, r12, r2
101	cmp	r2, r12
102	subls	r12, r12, r2
103	mov	r0, r12, lsr r1
104	pop	{r4-r10}
105	bx	r14
106EPILOGUE()
107
108PROLOGUE(mpn_mod_1_1p_cps)
109	stmfd	sp!, {r4, r5, r6, r14}
110	mov	r5, r0
111	clz	r4, r1
112	mov	r0, r1, asl r4
113	rsb	r6, r0, #0
114	bl	mpn_invert_limb
115	str	r0, [r5, #0]
116	str	r4, [r5, #4]
117	cmp	r4, #0
118	beq	L(2)
119	rsb	r1, r4, #32
120	mov	r3, #1
121	mov	r3, r3, asl r4
122	orr	r3, r3, r0, lsr r1
123	mul	r3, r6, r3
124	mov	r4, r3, lsr r4
125	str	r4, [r5, #8]
126L(2):	mul	r0, r6, r0
127	str	r0, [r5, #12]
128	ldmfd	sp!, {r4, r5, r6, pc}
129EPILOGUE()
130