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
130
131
132
133
134
135
136
137
138
139
140
141
142
|
;******************************************************************************
;* x86 optimizations for PNG decoding
;*
;* Copyright (c) 2008 Loren Merritt <lorenm@u.washington.edu>
;*
;* This file is part of Libav.
;*
;* Libav 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.
;*
;* Libav 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 Libav; if not, write to the Free Software
;* 51, Inc., Foundation Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
;******************************************************************************
%include "x86inc.asm"
%include "x86util.asm"
SECTION_RODATA
cextern pw_255
section .text align=16
; %1 = nr. of xmm registers used
%macro ADD_BYTES_FN 1
cglobal add_bytes_l2, 4, 6, %1, dst, src1, src2, wa, w, i
%if ARCH_X86_64
movsxd waq, wad
%endif
xor iq, iq
; vector loop
mov wq, waq
and waq, ~(mmsize*2-1)
jmp .end_v
.loop_v:
mova m0, [src1q+iq]
mova m1, [src1q+iq+mmsize]
paddb m0, [src2q+iq]
paddb m1, [src2q+iq+mmsize]
mova [dstq+iq ], m0
mova [dstq+iq+mmsize], m1
add iq, mmsize*2
.end_v:
cmp iq, waq
jl .loop_v
; scalar loop for leftover
jmp .end_s
.loop_s:
mov wab, [src1q+iq]
add wab, [src2q+iq]
mov [dstq+iq], wab
inc iq
.end_s:
cmp iq, wq
jl .loop_s
REP_RET
%endmacro
INIT_MMX mmx
ADD_BYTES_FN 0
%macro ADD_PAETH_PRED_FN 1
cglobal add_png_paeth_prediction, 5, 7, %1, dst, src, top, w, bpp, end, cntr
%if ARCH_X86_64
movsxd bppq, bppd
movsxd wq, wd
%endif
lea endq, [dstq+wq-(mmsize/2-1)]
sub topq, dstq
sub srcq, dstq
sub dstq, bppq
pxor m7, m7
movh m0, [dstq]
movh m1, [topq+dstq]
punpcklbw m0, m7
punpcklbw m1, m7
add dstq, bppq
.loop:
mova m2, m1
movh m1, [topq+dstq]
mova m3, m2
punpcklbw m1, m7
mova m4, m2
psubw m3, m1
psubw m4, m0
mova m5, m3
paddw m5, m4
%if cpuflag(ssse3)
pabsw m3, m3
pabsw m4, m4
pabsw m5, m5
%else ; !cpuflag(ssse3)
psubw m7, m5
pmaxsw m5, m7
pxor m6, m6
pxor m7, m7
psubw m6, m3
psubw m7, m4
pmaxsw m3, m6
pmaxsw m4, m7
pxor m7, m7
%endif ; cpuflag(ssse3)
mova m6, m4
pminsw m6, m5
pcmpgtw m3, m6
pcmpgtw m4, m5
mova m6, m4
pand m4, m3
pandn m6, m3
pandn m3, m0
movh m0, [srcq+dstq]
pand m6, m1
pand m2, m4
punpcklbw m0, m7
paddw m0, m6
paddw m3, m2
paddw m0, m3
pand m0, [pw_255]
mova m3, m0
packuswb m3, m3
movh [dstq], m3
add dstq, bppq
cmp dstq, endq
jle .loop
REP_RET
%endmacro
INIT_MMX mmx2
ADD_PAETH_PRED_FN 0
INIT_MMX ssse3
ADD_PAETH_PRED_FN 0
|
