1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
|
/* strrchr -- find the last occurence of C in a nul-terminated string
Copyright (C) 2013-2020 Free Software Foundation, Inc.
This file is part of the GNU C Library.
The GNU C 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 C 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 C Library. If not, see
<https://www.gnu.org/licenses/>. */
#include <sysdep.h>
.syntax unified
.text
ENTRY (strrchr)
@ r0 = start of string
@ r1 = character to match
@ returns NULL for no match, or a pointer to the match
mov r3, r0
mov r0, #0
uxtb r1, r1
@ Loop a few times until we're aligned.
tst r3, #7
beq 2f
1: ldrb r2, [r3], #1
cmp r2, r1 @ Find the character
it eq
subeq r0, r3, #1
cmp r2, #0 @ Find EOS
it eq
bxeq lr
tst r3, #7 @ Find the aligment point
bne 1b
@ So now we're aligned. Now we actually need a stack frame.
2: push { r4, r5, r6, r7 }
cfi_adjust_cfa_offset (16)
cfi_rel_offset (r4, 0)
cfi_rel_offset (r5, 4)
cfi_rel_offset (r6, 8)
cfi_rel_offset (r7, 12)
orr r1, r1, r1, lsl #8 @ Replicate C to all bytes
#ifdef ARCH_HAS_T2
movw ip, #0x0101
movt ip, #0x0101
#else
ldr ip, =0x01010101
#endif
orr r1, r1, r1, lsl #16
mov r2, #0 @ No found bits yet
@ Loop searching for EOS and C, 8 bytes at a time.
@ Any time we find a match in a word, we copy the address of
@ the word to r0, and the found bits to r2.
3: ldrd r4, r5, [r3], #8
@ Subtracting (unsigned saturating) from 1 means result of 1 for
@ any byte that was originally zero and 0 otherwise. Therefore
@ we consider the lsb of each byte the "found" bit.
uqsub8 r6, ip, r4 @ Find EOS
uqsub8 r7, ip, r5
eor r4, r4, r1 @ Convert C bytes to 0
eor r5, r5, r1
uqsub8 r4, ip, r4 @ Find C
uqsub8 r5, ip, r5
cmp r6, #0 @ Found EOS, first word
bne 4f
cmp r4, #0 @ Handle C, first word
itt ne
subne r0, r3, #8
movne r2, r4
cmp r7, #0 @ Found EOS, second word
bne 5f
cmp r5, #0 @ Handle C, second word
itt ne
subne r0, r3, #4
movne r2, r5
b 3b
@ Found EOS in second word; fold to first word.
5: add r3, r3, #4 @ Dec pointer to 2nd word, with below
mov r4, r5 @ Overwrite first word C found
mov r6, r7 @ Overwrite first word EOS found
@ Found EOS. Zap found C after EOS.
4: sub r3, r3, #8 @ Decrement pointer to first word
#ifdef __ARMEB__
@ Byte swap to be congruent with LE, which is easier from here on.
rev r6, r6 @ Byte swap found EOS,
rev r4, r4 @ ... this found C
rev r2, r2 @ ... prev found C
#endif
sub r7, r6, #1 @ Toggle EOS lsb and below
eor r6, r6, r7 @ All bits below and including lsb
ands r4, r4, r6 @ Zap C above EOS
itt ne
movne r2, r4 @ Copy to result, if still non-zero
movne r0, r3
pop { r4, r5, r6, r7 }
cfi_adjust_cfa_offset (-16)
cfi_restore (r4)
cfi_restore (r5)
cfi_restore (r6)
cfi_restore (r7)
@ Adjust the result pointer if we found a word containing C.
cmp r2, #0
clz r2, r2 @ Find the bit offset of the last C
itt ne
rsbne r2, r2, #32 @ Convert to a count from the right
addne r0, r0, r2, lsr #3 @ Convert to byte offset and add.
bx lr
END (strrchr)
weak_alias (strrchr, rindex)
libc_hidden_builtin_def (strrchr)
|
