lshiftc.asm revision 1.1.1.2 1 dnl PowerPC-32 mpn_lshiftc.
2
3 dnl Copyright 1995, 1998, 2000, 2002-2005, 2010 Free Software Foundation, Inc.
4
5 dnl This file is part of the GNU MP Library.
6 dnl
7 dnl The GNU MP Library is free software; you can redistribute it and/or modify
8 dnl it under the terms of either:
9 dnl
10 dnl * the GNU Lesser General Public License as published by the Free
11 dnl Software Foundation; either version 3 of the License, or (at your
12 dnl option) any later version.
13 dnl
14 dnl or
15 dnl
16 dnl * the GNU General Public License as published by the Free Software
17 dnl Foundation; either version 2 of the License, or (at your option) any
18 dnl later version.
19 dnl
20 dnl or both in parallel, as here.
21 dnl
22 dnl The GNU MP Library is distributed in the hope that it will be useful, but
23 dnl WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY
24 dnl or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
25 dnl for more details.
26 dnl
27 dnl You should have received copies of the GNU General Public License and the
28 dnl GNU Lesser General Public License along with the GNU MP Library. If not,
29 dnl see https://www.gnu.org/licenses/.
30
31 include(`../config.m4')
32
33 C cycles/limb
34 C 603e: ?
35 C 604e: 3.0
36 C 75x (G3): 3.0
37 C 7400,7410 (G4): 3.0
38 C 7445,7455 (G4+): 2.5
39 C 7447,7457 (G4+): 2.25
40 C power4/ppc970: 2.5
41 C power5: 2.5
42
43 C INPUT PARAMETERS
44 C rp r3
45 C up r4
46 C n r5
47 C cnt r6
48
49 ASM_START()
50 PROLOGUE(mpn_lshiftc)
51 cmpwi cr0, r5, 30 C more than 30 limbs?
52 slwi r0, r5, 2
53 add r4, r4, r0 C make r4 point at end of s1
54 add r7, r3, r0 C make r7 point at end of res
55 bgt L(BIG) C branch if more than 12 limbs
56
57 mtctr r5 C copy size into CTR
58 subfic r8, r6, 32
59 lwzu r11, -4(r4) C load first s1 limb
60 srw r3, r11, r8 C compute function return value
61 bdz L(end1)
62
63 L(oop): lwzu r10, -4(r4)
64 slw r9, r11, r6
65 srw r12, r10, r8
66 nor r9, r9, r12
67 stwu r9, -4(r7)
68 bdz L(end2)
69 lwzu r11, -4(r4)
70 slw r9, r10, r6
71 srw r12, r11, r8
72 nor r9, r9, r12
73 stwu r9, -4(r7)
74 bdnz L(oop)
75
76 L(end1):
77 slw r0, r11, r6
78 nor r0, r0, r0
79 stw r0, -4(r7)
80 blr
81 L(end2):
82 slw r0, r10, r6
83 nor r0, r0, r0
84 stw r0, -4(r7)
85 blr
86
87 L(BIG):
88 stwu r1, -48(r1)
89 stmw r24, 8(r1) C save registers we are supposed to preserve
90 lwzu r9, -4(r4)
91 subfic r8, r6, 32
92 srw r3, r9, r8 C compute function return value
93 slw r0, r9, r6
94 addi r5, r5, -1
95
96 andi. r10, r5, 3 C count for spill loop
97 beq L(e)
98 mtctr r10
99 lwzu r28, -4(r4)
100 bdz L(xe0)
101
102 L(loop0):
103 slw r12, r28, r6
104 srw r24, r28, r8
105 lwzu r28, -4(r4)
106 nor r24, r0, r24
107 stwu r24, -4(r7)
108 mr r0, r12
109 bdnz L(loop0) C taken at most once!
110
111 L(xe0): slw r12, r28, r6
112 srw r24, r28, r8
113 nor r24, r0, r24
114 stwu r24, -4(r7)
115 mr r0, r12
116
117 L(e): srwi r5, r5, 2 C count for unrolled loop
118 addi r5, r5, -1
119 mtctr r5
120 lwz r28, -4(r4)
121 lwz r29, -8(r4)
122 lwz r30, -12(r4)
123 lwzu r31, -16(r4)
124
125 L(loopU):
126 slw r9, r28, r6
127 srw r24, r28, r8
128 lwz r28, -4(r4)
129 slw r10, r29, r6
130 srw r25, r29, r8
131 lwz r29, -8(r4)
132 slw r11, r30, r6
133 srw r26, r30, r8
134 lwz r30, -12(r4)
135 slw r12, r31, r6
136 srw r27, r31, r8
137 lwzu r31, -16(r4)
138 nor r24, r0, r24
139 stw r24, -4(r7)
140 nor r25, r9, r25
141 stw r25, -8(r7)
142 nor r26, r10, r26
143 stw r26, -12(r7)
144 nor r27, r11, r27
145 stwu r27, -16(r7)
146 mr r0, r12
147 bdnz L(loopU)
148
149 slw r9, r28, r6
150 srw r24, r28, r8
151 slw r10, r29, r6
152 srw r25, r29, r8
153 slw r11, r30, r6
154 srw r26, r30, r8
155 slw r12, r31, r6
156 srw r27, r31, r8
157 nor r24, r0, r24
158 stw r24, -4(r7)
159 nor r25, r9, r25
160 stw r25, -8(r7)
161 nor r26, r10, r26
162 stw r26, -12(r7)
163 nor r27, r11, r27
164 stw r27, -16(r7)
165 nor r12, r12, r12
166 stw r12, -20(r7)
167 lmw r24, 8(r1) C restore registers
168 addi r1, r1, 48
169 blr
170 EPILOGUE()
171