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
349
350
351
352
353
354
355
356
357
358
359
360
361
|
///*****************************************************************************
//*
//* Copyright (C) 2012 Ittiam Systems Pvt Ltd, Bangalore
//*
//* Licensed under the Apache License, Version 2.0 (the "License");
//* you may not use this file except in compliance with the License.
//* You may obtain a copy of the License at:
//*
//* http://www.apache.org/licenses/LICENSE-2.0
//*
//* Unless required by applicable law or agreed to in writing, software
//* distributed under the License is distributed on an "AS IS" BASIS,
//* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
//* See the License for the specific language governing permissions and
//* limitations under the License.
//*
//*****************************************************************************/
///**
//*******************************************************************************
//* @file
//* ihevc_intra_pred_chroma_horz_neon.s
//*
//* @brief
//* contains function definition for intra prediction interpolation filters
//*
//*
//* @author
//* parthiban v
//*
//* @par list of functions:
//* - ihevc_intra_pred_luma_horz()
//*
//* @remarks
//* none
//*
//*******************************************************************************
//*/
//
///**
//*******************************************************************************
//*
//* @brief
//* intra prediction interpolation filter for horizontal luma variable.
//*
//* @par description:
//* horizontal intraprediction(mode 10) with.extern samples location
//* pointed by 'pu1_ref' to the tu block location pointed by 'pu1_dst' refer
//* to section 8.4.4.2.6 in the standard (special case)
//*
//* @param[in] pu1_src
//* uword8 pointer to the source
//*
//* @param[out] pu1_dst
//* uword8 pointer to the destination
//*
//* @param[in] src_strd
//* integer source stride
//*
//* @param[in] dst_strd
//* integer destination stride
//*
//* @param[in] nt
//* integer transform block size
//*
//* @param[in] mode
//* integer intraprediction mode
//*
//* @returns
//*
//* @remarks
//* none
//*
//*******************************************************************************
//*/
//void ihevc_intra_pred_chroma_horz(uword8 *pu1_ref,
// word32 src_strd,
// uword8 *pu1_dst,
// word32 dst_strd,
// word32 nt,
// word32 mode)
//**************variables vs registers*****************************************
//x0 => *pu1_ref
//x1 => src_strd
//x2 => *pu1_dst
//x3 => dst_strd
.text
.align 4
.include "ihevc_neon_macros.s"
.globl ihevc_intra_pred_chroma_horz_av8
.type ihevc_intra_pred_chroma_horz_av8, %function
ihevc_intra_pred_chroma_horz_av8:
// stmfd sp!, {x4-x12, x14} //stack stores the values of the arguments
push_v_regs
stp x19, x20,[sp,#-16]!
lsl x6,x4,#2 //four_nt
add x12,x0,x6 //*pu1_ref[four_nt]
cmp x4,#4 //if nt == 4
beq core_loop_4
cmp x4,#8 //if nt == 8
beq core_loop_8
//cmp x4,#16 @if nt == 16
//beq core_loop_16
sub x12,x12,#16 //move to 16th value pointer
add x9,x2,#16
core_loop_16:
ld1 { v0.8h},[x12] //load 16 values. d1[7] will have the 1st value.
sub x12,x12,#16
ld1 { v10.8h},[x12] //load 16 values. d1[7] will have the 1st value.
dup v2.8h, v0.4h[7] //duplicate the i value.
dup v4.8h, v0.4h[6] //duplicate the ii value.
dup v6.8h, v0.4h[5] //duplicate the iii value.
st1 { v2.8h},[x2],x3 //store in 1st row 0-16 columns
st1 { v2.8h},[x9],x3 //store in 1st row 16-32 columns
dup v8.8h, v0.4h[4]
st1 { v4.8h},[x2],x3
st1 { v4.8h},[x9],x3
dup v2.8h, v0.4h[3]
st1 { v6.8h},[x2],x3
st1 { v6.8h},[x9],x3
dup v4.8h, v0.4h[2]
st1 { v8.8h},[x2],x3
st1 { v8.8h},[x9],x3
dup v6.8h, v0.4h[1]
st1 { v2.8h},[x2],x3
st1 { v2.8h},[x9],x3
dup v8.8h, v0.4h[0]
st1 { v4.8h},[x2],x3
st1 { v4.8h},[x9],x3
dup v2.8h, v10.4h[7]
st1 { v6.8h},[x2],x3
st1 { v6.8h},[x9],x3
dup v4.8h, v10.4h[6]
st1 { v8.8h},[x2],x3
st1 { v8.8h},[x9],x3
dup v6.8h, v10.4h[5]
st1 { v2.8h},[x2],x3
st1 { v2.8h},[x9],x3
dup v8.8h, v10.4h[4]
st1 { v4.8h},[x2],x3
st1 { v4.8h},[x9],x3
dup v2.8h, v10.4h[3]
st1 { v6.8h},[x2],x3
st1 { v6.8h},[x9],x3
dup v4.8h, v10.4h[2]
st1 { v8.8h},[x2],x3
st1 { v8.8h},[x9],x3
dup v6.8h, v10.4h[1]
st1 { v2.8h},[x2],x3
st1 { v2.8h},[x9],x3
sub x12,x12,#16 //move to 16th value pointer
dup v8.8h, v10.4h[0]
st1 { v4.8h},[x2],x3
st1 { v4.8h},[x9],x3
subs x4,x4,#16 //decrement the loop count by 16
st1 { v6.8h},[x2],x3
st1 { v6.8h},[x9],x3
st1 { v8.8h},[x2],x3
st1 { v8.8h},[x9],x3
bgt core_loop_16
// ldmfd sp!,{x4-x12,x15} //reload the registers from sp
ldp x19, x20,[sp],#16
pop_v_regs
ret
b endloop
core_loop_8:
ldrb w14,[x12],#1 //pu1_ref[two_nt]
sxtw x14,w14
//vld1.8 {q15},[x12] @pu1_ref[two_nt + 1 + col]
dup v28.8b,w14
sub x12,x12,#17
ld1 { v0.16b},[x12]
sub x12,x12,#16
// ld1 { v30.16b},[x12]
dup v10.8h, v0.4h[7]
//vmovl.u8 q13,d26
dup v2.8h, v0.4h[6]
//vsubl.u8 q12,d30,d28
dup v4.8h, v0.4h[5]
//vshr.s16 q12,q12,#1
dup v6.8h, v0.4h[4]
//vqadd.s16 q11,q13,q12
dup v8.8h, v0.4h[3]
//vqmovun.s16 d22,q11
st1 { v10.8h},[x2],x3
dup v10.8h, v0.4h[2]
//vsubl.u8 q12,d31,d28
dup v12.8h, v0.4h[1]
//vshr.s16 q12,q12,#1
dup v14.8h, v0.4h[0]
//vqadd.s16 q11,q13,q12
dup v16.8h, v0.4h[3]
//vqmovun.s16 d22,q11
st1 { v2.8h},[x2],x3
//sub x2,x2,#8
st1 { v4.8h},[x2],x3
st1 { v6.8h},[x2],x3
st1 { v8.8h},[x2],x3
st1 { v10.8h},[x2],x3
//vdup.8 q1,d0[2]
st1 { v12.8h},[x2],x3
//vdup.8 q2,d0[1]
st1 { v14.8h},[x2],x3
//vdup.8 q3,d0[0]
//vst1.8 {q7},[x2],x3
//vdup.8 q4,d0[3]
//vst1.8 {q8},[x2],x3
//vdup.8 q5,d0[2]
//vst1.8 {q1},[x2],x3
//vdup.8 q6,d0[1]
//vst1.8 {q2},[x2],x3
//vdup.8 q7,d0[0]
//vst1.8 {q3},[x2],x3
//vst1.8 {q4},[x2],x3
//vst1.8 {q5},[x2],x3
//vst1.8 {q6},[x2],x3
//vst1.8 {q7},[x2],x3
// ldmfd sp!,{x4-x12,x15} //reload the registers from sp
ldp x19, x20,[sp],#16
pop_v_regs
ret
b endloop
core_loop_4:
ldrb w14,[x12] //pu1_ref[two_nt]
sxtw x14,w14
add x12,x12,#1 //pu1_ref[two_nt + 1]
//vld1.8 {d30},[x12] @pu1_ref[two_nt + 1 + col]
sub x12,x12,#9
ld1 {v0.8b},[x12]
sub x12,x12,#8
ld1 {v30.8b},[x12]
dup v26.4h, v0.4h[3]
dup v28.8b,w14
dup v3.4h, v0.4h[2]
uxtl v26.8h, v26.8b
dup v4.4h, v0.4h[1]
usubl v24.8h, v30.8b, v28.8b
dup v5.4h, v0.4h[0]
sshr v24.8h, v24.8h,#1
dup v6.4h, v0.4h[3]
sqadd v22.8h, v26.8h , v24.8h
dup v7.4h, v0.4h[2]
sqxtun v22.8b, v22.8h
st1 {v6.8b},[x2],x3
st1 {v3.8b},[x2],x3
dup v8.4h, v0.4h[1]
st1 {v4.8b},[x2],x3
st1 {v5.8b},[x2],x3
dup v9.4h, v0.4h[0]
//vst1.8 {d6},[x2],x3
//vst1.8 {d7},[x2],x3
//vst1.8 {d8},[x2],x3
//vst1.8 {d9},[x2],x3
// ldmfd sp!,{x4-x12,x15} //reload the registers from sp
ldp x19, x20,[sp],#16
pop_v_regs
ret
b endloop
//core_loop_4
ldrb w14,[x12] //pu1_ref[two_nt]
sxtw x14,w14
add x12,x12,#1 //pu1_ref[two_nt + 1]
ld1 {v30.8b},[x12] //pu1_ref[two_nt + 1 + col]
sub x12,x12,#5
ld1 {v0.8b},[x12]
dup v28.8b,w14
dup v26.8b, v0.8b[3]
uxtl v26.8h, v26.8b
dup v3.8b, v0.8b[2]
usubl v24.8h, v30.8b, v28.8b
dup v4.8b, v0.8b[1]
sshr v24.8h, v24.8h,#1
dup v5.8b, v0.8b[0]
sqadd v22.8h, v26.8h , v24.8h
sqxtun v22.8b, v22.8h
st1 {v22.s}[0],[x2],x3
st1 {v3.s}[0],[x2],x3
st1 {v4.s}[0],[x2],x3
st1 {v5.s}[0],[x2],x3
// ldmfd sp!,{x4-x12,x15} //reload the registers from sp
ldp x19, x20,[sp],#16
pop_v_regs
ret
endloop:
|