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
|
PUBLIC Baseline_Add
PUBLIC Baseline_Sub
.CODE
ALIGN 8
Baseline_Add PROC
lea rdx, [rdx+8*rcx]
lea r8, [r8+8*rcx]
lea r9, [r9+8*rcx]
neg rcx ; rcx is negative index
test rcx, 2 ; this clears carry flag
jz $0@Baseline_Add
sub rcx, 2
jmp $1@Baseline_Add
$0@Baseline_Add:
jrcxz $2@Baseline_Add ; loop until rcx overflows and becomes zero
mov rax,[r8+8*rcx]
adc rax,[r9+8*rcx]
mov [rdx+8*rcx],rax
mov rax,[r8+8*rcx+8]
adc rax,[r9+8*rcx+8]
mov [rdx+8*rcx+8],rax
$1@Baseline_Add:
mov rax,[r8+8*rcx+16]
adc rax,[r9+8*rcx+16]
mov [rdx+8*rcx+16],rax
mov rax,[r8+8*rcx+24]
adc rax,[r9+8*rcx+24]
mov [rdx+8*rcx+24],rax
lea rcx,[rcx+4] ; advance index, avoid inc which causes slowdown on Intel Core 2
jmp $0@Baseline_Add
$2@Baseline_Add:
mov rax, 0
setc al ; store carry into rax (return result register)
ret
Baseline_Add ENDP
ALIGN 8
Baseline_Sub PROC
lea rdx, [rdx+8*rcx]
lea r8, [r8+8*rcx]
lea r9, [r9+8*rcx]
neg rcx ; rcx is negative index
test rcx, 2 ; this clears carry flag
jz $0@Baseline_Sub
sub rcx, 2
jmp $1@Baseline_Sub
$0@Baseline_Sub:
jrcxz $2@Baseline_Sub ; loop until rcx overflows and becomes zero
mov rax,[r8+8*rcx]
sbb rax,[r9+8*rcx]
mov [rdx+8*rcx],rax
mov rax,[r8+8*rcx+8]
sbb rax,[r9+8*rcx+8]
mov [rdx+8*rcx+8],rax
$1@Baseline_Sub:
mov rax,[r8+8*rcx+16]
sbb rax,[r9+8*rcx+16]
mov [rdx+8*rcx+16],rax
mov rax,[r8+8*rcx+24]
sbb rax,[r9+8*rcx+24]
mov [rdx+8*rcx+24],rax
lea rcx,[rcx+4] ; advance index, avoid inc which causes slowdown on Intel Core 2
jmp $0@Baseline_Sub
$2@Baseline_Sub:
mov rax, 0
setc al ; store carry into rax (return result register)
ret
Baseline_Sub ENDP
_TEXT ENDS
END
|
