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
|
/* SPDX-License-Identifier: GPL-2.0 */
#include <linux/linkage.h>
#include <asm/asm.h>
ENTRY(__memmove)
WEAK(memmove)
/*
* Here we determine if forward copy is possible. Forward copy is
* preferred to backward copy as it is more cache friendly.
*
* If a0 >= a1, t0 gives their distance, if t0 >= a2 then we can
* copy forward.
* If a0 < a1, we can always copy forward. This will make t0 negative,
* so a *unsigned* comparison will always have t0 >= a2.
*
* For forward copy we just delegate the task to memcpy.
*/
sub t0, a0, a1
bltu t0, a2, 1f
tail __memcpy
1:
/*
* Register allocation for code below:
* a0 - end of uncopied dst
* a1 - end of uncopied src
* t0 - start of uncopied dst
*/
mv t0, a0
add a0, a0, a2
add a1, a1, a2
/*
* Use bytewise copy if too small.
*
* This threshold must be at least 2*SZREG to ensure at least one
* wordwise copy is performed. It is chosen to be 16 because it will
* save at least 7 iterations of bytewise copy, which pays off the
* fixed overhead.
*/
li a3, 16
bltu a2, a3, .Lbyte_copy_tail
/*
* Bytewise copy first to align t0 to word boundary.
*/
andi a2, a0, ~(SZREG-1)
beq a0, a2, 2f
1:
addi a1, a1, -1
lb a5, 0(a1)
addi a0, a0, -1
sb a5, 0(a0)
bne a0, a2, 1b
2:
/*
* Now a0 is word-aligned. If a1 is also word aligned, we could perform
* aligned word-wise copy. Otherwise we need to perform misaligned
* word-wise copy.
*/
andi a3, a1, SZREG-1
bnez a3, .Lmisaligned_word_copy
/* Wordwise copy */
addi t0, t0, SZREG-1
bleu a0, t0, 2f
1:
addi a1, a1, -SZREG
REG_L a5, 0(a1)
addi a0, a0, -SZREG
REG_S a5, 0(a0)
bgtu a0, t0, 1b
2:
addi t0, t0, -(SZREG-1)
.Lbyte_copy_tail:
/*
* Bytewise copy anything left.
*/
beq a0, t0, 2f
1:
addi a1, a1, -1
lb a5, 0(a1)
addi a0, a0, -1
sb a5, 0(a0)
bne a0, t0, 1b
2:
mv a0, t0
ret
.Lmisaligned_word_copy:
/*
* Misaligned word-wise copy.
* For misaligned copy we still perform word-wise copy, but we need to
* use the value fetched from the previous iteration and do some shifts.
* This is safe because we wouldn't access more words than necessary.
*/
/* Calculate shifts */
slli t3, a3, 3
sub t4, x0, t3 /* negate is okay as shift will only look at LSBs */
/* Load the initial value and align a1 */
andi a1, a1, ~(SZREG-1)
REG_L a5, 0(a1)
addi t0, t0, SZREG-1
/* At least one iteration will be executed here, no check */
1:
sll a4, a5, t4
addi a1, a1, -SZREG
REG_L a5, 0(a1)
srl a2, a5, t3
or a2, a2, a4
addi a0, a0, -SZREG
REG_S a2, 0(a0)
bgtu a0, t0, 1b
/* Update pointers to correct value */
addi t0, t0, -(SZREG-1)
add a1, a1, a3
j .Lbyte_copy_tail
END(__memmove)
|