-
Notifications
You must be signed in to change notification settings - Fork 0
/
neon_scale3x.S
349 lines (291 loc) · 11.7 KB
/
neon_scale3x.S
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
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
@@
@@ Copyright (C) 2012 Roman Pauer
@@
@@ Permission is hereby granted, free of charge, to any person obtaining a copy of
@@ this software and associated documentation files (the "Software"), to deal in
@@ the Software without restriction, including without limitation the rights to
@@ use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies
@@ of the Software, and to permit persons to whom the Software is furnished to do
@@ so, subject to the following conditions:
@@
@@ The above copyright notice and this permission notice shall be included in all
@@ copies or substantial portions of the Software.
@@
@@ THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
@@ IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
@@ FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
@@ AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
@@ LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
@@ OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
@@ SOFTWARE.
@@
.arm
.include "neon_scale3x.Sinc"
.include "neon_normalxx.Sinc"
.global neon_scale3x_8_8
.global neon_scale3x_16_16
.global neon_scale3x_8_16
.align 4
neon_scale3x_8_8:
@ r0 = const uint8_t *src
@ r1 = uint8_t *dst
@ r2 = unsigned int width (pixels)
@ r3 = unsigned int srcstride (bytes)
@ [sp] = unsigned int dststride (bytes)
@ [sp+4] = unsigned int height
@ lr = return address
ldr ip, [sp] @ ip = dststride
push {r4-r11}
ldr r9, [sp, #(9*4)] @ r9 = height
sub r4, r0, r3 @ r4 = src - srcstride
mov r11, sp @ oldsp = sp
add r5, r0, r3 @ r5 = src + srcstride
bic sp, sp, #31 @ align sp to 32 bytes
add r6, r1, ip @ r6 = dst + dststride
sub sp, sp, #64 @ sp -= 64
sub r3, r3, r2 @ r3 = srcstride - width
vst1.64 {d8-d11}, [sp:256] @ save q4,q5
add r7, r1, ip, lsl #1 @ r7 = dst + 2 * dststride
add r8, sp, #32 @ r8 = sp + 32
sub ip, ip, r2 @ ip = dststride - width
vst1.64 {d12-d15}, [r8:256] @ save q6,q7
add ip, ip, ip, lsl #1 @ ip = 3 * dststride - 3 * width
mov r8, r2 @ r8 = width
sub r9, r9, #2 @ r9 = height - 2
@ r0 = src
@ r1 = dst
@ r2 = width
@ r3 = srcdiff (srcstride - width)
@ r4 = src - srcstride
@ r5 = src + srcstride
@ r6 = dst + dststride
@ r7 = dst + 2 * dststride
@ r8 = counter
@ r9 = height
@ r10 = tmpreg
@ r11 = oldsp
@ ip = dstdiff (3 * dststride - 3 * width)
@ first line
neon_scale3x_8_8_line first, r4, r0, r5, r8, r1, r6, r7, r10, 0, 0
add r0, r0, r3
add r4, r4, r3
add r5, r5, r3
add r1, r1, ip
add r6, r6, ip
add r7, r7, ip
@ middle lines
101:
mov r8, r2
neon_scale3x_8_8_line middle, r4, r0, r5, r8, r1, r6, r7, r10, 0, 0
subS r9, r9, #1
add r0, r0, r3
add r4, r4, r3
add r5, r5, r3
add r1, r1, ip
add r6, r6, ip
add r7, r7, ip
bne 101b
@ last line
mov r8, r2
neon_scale3x_8_8_line last, r4, r0, r5, r8, r1, r6, r7, r10, 0, 0
add ip, sp, #32 @ ip = sp + 32
vld1.64 {d8-d11}, [sp:256] @ restore q4,q5
mov sp, r11 @ sp = oldsp
vld1.64 {d12-d15}, [ip:256] @ restore q6,q7
pop {r4-r11}
bx lr
@ end procedure neon_scale3x_8_8
neon_scale3x_16_16:
@ r0 = const uint16_t *src
@ r1 = uint16_t *dst
@ r2 = unsigned int width (pixels)
@ r3 = unsigned int srcstride (bytes)
@ [sp] = unsigned int dststride (bytes)
@ [sp+4] = unsigned int height
@ lr = return address
ldr ip, [sp] @ ip = dststride
push {r4-r11}
ldr r9, [sp, #(9*4)] @ r9 = height
sub r4, r0, r3 @ r4 = src - srcstride
mov r11, sp @ oldsp = sp
add r5, r0, r3 @ r5 = src + srcstride
bic sp, sp, #31 @ align sp to 32 bytes
add r6, r1, ip @ r6 = dst + dststride
sub sp, sp, #64 @ sp -= 64
sub r3, r3, r2, lsl #1 @ r3 = srcstride - 2 * width
vst1.64 {d8-d11}, [sp:256] @ save q4,q5
add r7, r1, ip, lsl #1 @ r7 = dst + 2 * dststride
add r8, sp, #32 @ r8 = sp + 32
sub ip, ip, r2, lsl #1 @ ip = dststride - 2 * width
vst1.64 {d12-d15}, [r8:256] @ save q6,q7
add ip, ip, ip, lsl #1 @ ip = 3 * dststride - 6 * width
mov r8, r2 @ r8 = width
sub r9, r9, #2 @ r9 = height - 2
@ r0 = src
@ r1 = dst
@ r2 = width
@ r3 = srcdiff (srcstride - 2 * width)
@ r4 = src - srcstride
@ r5 = src + srcstride
@ r6 = dst + dststride
@ r7 = dst + 2 * dststride
@ r8 = counter
@ r9 = height
@ r10 = tmpreg
@ r11 = oldsp
@ ip = dstdiff (3 * dststride - 6 * width)
@ first line
neon_scale3x_16_16_line first, r4, r0, r5, r8, r1, r6, r7, r10, 0, 0
add r0, r0, r3
add r4, r4, r3
add r5, r5, r3
add r1, r1, ip
add r6, r6, ip
add r7, r7, ip
@ middle lines
101:
mov r8, r2
neon_scale3x_16_16_line middle, r4, r0, r5, r8, r1, r6, r7, r10, 0, 0
subS r9, r9, #1
add r0, r0, r3
add r4, r4, r3
add r5, r5, r3
add r1, r1, ip
add r6, r6, ip
add r7, r7, ip
bne 101b
@ last line
mov r8, r2
neon_scale3x_16_16_line last, r4, r0, r5, r8, r1, r6, r7, r10, 0, 0
add ip, sp, #32 @ ip = sp + 32
vld1.64 {d8-d11}, [sp:256] @ restore q4,q5
mov sp, r11 @ sp = oldsp
vld1.64 {d12-d15}, [ip:256] @ restore q6,q7
pop {r4-r11}
bx lr
@ end procedure neon_scale3x_16_16
neon_scale3x_8_16:
@ r0 = const uint8_t *src
@ r1 = uint8_t *dst
@ r2 = const uint32_t *palette
@ r3 = unsigned int width (pixels)
@ [sp] = unsigned int srcstride (bytes)
@ [sp+4] = unsigned int dststride (bytes)
@ [sp+8] = unsigned int height
@ lr = return address
@ three temporary lines
ldr ip, [sp] @ ip = srcstride
push {r4-r11,lr}
ldr r4, [sp, #(4*10)] @ r4 = dststride
ldr r5, [sp, #(4*11)] @ r5 = height
mov r6, sp @ r6 = sp
sub ip, ip, r3 @ ip = srcstride - width
bic sp, sp, #31 @ align sp to 32 bytes
sub r7, r4, r3, lsl #1 @ r7 = dststride - 2 * width
sub sp, sp, r3, lsl #1 @ sp -= 2 * width
sub r5, r5, #2 @ height -= 2
mov r10, sp @ tmpline3 = sp
add r7, r7, r7, lsl #1 @ r7 = 3 * dststride - 6 * width
bic sp, sp, #31 @ align sp to 32 bytes
sub sp, sp, r3, lsl #1 @ sp -= 2 * width
mov r11, sp @ tmpline2 = sp
bic sp, sp, #31 @ align sp to 32 bytes
sub sp, sp, r3, lsl #1 @ sp -= 2 * width
mov lr, sp @ tmpline1 = sp
bic sp, sp, #31 @ align sp to 32 bytes
sub r8, sp, #64 @ r8 = sp - 64
vst1.64 {d8-d11}, [r8:256] @ save q4,q5
sub r9, sp, #32 @ r9 = sp - 32
vst1.64 {d12-d15}, [r9:256] @ save q6,q7
sub sp, sp, #(36 + 64) @ sp -= (36 + 64)
str r6, [sp] @ oldsp = r6
str r5, [sp, #4] @ height = r5
str ip, [sp, #8] @ srcdiff = ip
str r7, [sp, #12] @ dstdiff = r7
str r4, [sp, #16] @ dststride = r4
str lr, [sp, #20] @ tmpline1 = lr
str r11, [sp, #24] @ tmpline2 = r11
str r10, [sp, #28] @ tmpline3 = r10
str r3, [sp, #32] @ width = r3
@ r0 = src
@ r1 = dst
@ r2 = palette
@ r3 = counter
@ r4 = dst2
@ r5 = dst3
@ r11 = bufptr1
@ ip = bufptr2
@ lr = bufptr3
@ [sp] = oldsp
@ [sp, #4] = height
@ [sp, #8] = srcdiff (srcstride - width)
@ [sp, #12] = dstdiff (3 * dststride - 6 * width)
@ [sp, #16] = dststride
@ [sp, #20] = tmpline1
@ [sp, #24] = tmpline2
@ [sp, #28] = tmpline3
@ [sp, #32] = width
@ lr = tmpline1
@ r3 = counter
@ first line
neon_normal1x_8_16_line r0, lr, r2, r3, r4, r5, r6, r7, r8, r9, r10, r11, ip
ldr r7, [sp, #8] @ r7 = srcdiff
ldr r3, [sp, #32] @ counter = width
ldr lr, [sp, #24] @ bufptr3 = tmpline2
add r0, r0, r7 @ src += srcdiff
@ second line
neon_normal1x_8_16_line r0, lr, r2, r3, r4, r5, r6, r7, r8, r9, r10, r11, ip
ldr r9, [sp, #16] @ r9 = dststride
ldr r3, [sp, #32] @ counter = width
ldr ip, [sp, #20] @ bufptr2 = tmpline1
ldr lr, [sp, #24] @ bufptr3 = tmpline2
add r4, r1, r9 @ dst2 = dst + dststride
add r5, r1, r9, lsl #1 @ dst3 = dst + 2 * dststride
@ first temporary line
neon_scale3x_16_16_line first, r11, ip, lr, r3, r1, r4, r5, r6, 1, 0
ldr r7, [sp, #8] @ r7 = srcdiff
ldr r8, [sp, #12] @ r8 = dstdiff
ldr r3, [sp, #32] @ counter = width
ldr lr, [sp, #28] @ bufptr3 = tmpline3
add r0, r0, r7 @ src += srcdiff
add r1, r1, r8 @ dst += dstdiff
100:
@ line n+1
neon_normal1x_8_16_line r0, lr, r2, r3, r4, r5, r6, r7, r8, r9, r10, r11, ip
ldr r9, [sp, #16] @ r9 = dststride
ldr r11, [sp, #20] @ bufptr1 = tmpline1
ldr ip, [sp, #24] @ bufptr2 = tmpline2
ldr lr, [sp, #28] @ bufptr3 = tmpline3
add r4, r1, r9 @ dst2 = dst + dststride
ldr r3, [sp, #32] @ counter = width
add r5, r1, r9, lsl #1 @ dst3 = dst + 2 * dststride
str r11, [sp, #28] @ tmpline3 = bufptr1
str ip, [sp, #20] @ tmpline1 = bufptr2
str lr, [sp, #24] @ tmpline2 = bufptr3
@ temporary line n
neon_scale3x_16_16_line middle, r11, ip, lr, r3, r1, r4, r5, r6, 1, 0
ldr r6, [sp, #4] @ r6 = height
ldr r7, [sp, #8] @ r7 = srcdiff
ldr r8, [sp, #12] @ r8 = dstdiff
ldr r3, [sp, #32] @ counter = width
subS r6, r6, #1 @ height--
ldr lr, [sp, #28] @ bufptr3 = tmpline3
add r0, r0, r7 @ src += srcdiff
add r1, r1, r8 @ dst += dstdiff
str r6, [sp, #4] @ height = r6
bne 100b
ldr r9, [sp, #16] @ r9 = dststride
ldr r11, [sp, #20] @ bufptr1 = tmpline1
ldr ip, [sp, #24] @ bufptr2 = tmpline2
add r4, r1, r9 @ dst2 = dst + dststride
add r5, r1, r9, lsl #1 @ dst3 = dst + 2 * dststride
@ last temporary line
neon_scale3x_16_16_line last, r11, ip, lr, r3, r1, r4, r5, r6, 1, 0
add r6, sp, #36 @ r6 = sp + 36
ldr sp, [sp] @ sp = oldsp
vld1.64 {d8-d11}, [r6:256] @ restore q4,q5
add ip, r6, #32 @ ip = r6 + 32
vld1.64 {d12-d15}, [ip:256] @ restore q6,q7
pop {r4-r11,lr}
bx lr
@ end procedure neon_scale3x_8_16