8bed5c041023249ae9484489ddb385e3ea23fbf7
[libav.git] / libavcodec / bfin / hpel_pixels_bfin.S
1 /*
2 * Blackfin Pixel Operations
3 * Copyright (C) 2007 Marc Hoffman <marc.hoffman@analog.com>
4 *
5 * This file is part of Libav.
6 *
7 * Libav is free software; you can redistribute it and/or
8 * modify it under the terms of the GNU Lesser General Public
9 * License as published by the Free Software Foundation; either
10 * version 2.1 of the License, or (at your option) any later version.
11 *
12 * Libav is distributed in the hope that it will be useful,
13 * but WITHOUT ANY WARRANTY; without even the implied warranty of
14 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
15 * Lesser General Public License for more details.
16 *
17 * You should have received a copy of the GNU Lesser General Public
18 * License along with Libav; if not, write to the Free Software
19 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
20 */
21
22 #include "config_bfin.h"
23
24 /*
25 motion compensation
26 primitives
27
28 * Halfpel motion compensation with rounding (a+b+1)>>1.
29 * This is an array[4][4] of motion compensation funcions for 4
30 * horizontal blocksizes (8,16) and the 4 halfpel positions<br>
31 * *pixels_tab[ 0->16xH 1->8xH ][ xhalfpel + 2*yhalfpel ]
32 * @param block destination where the result is stored
33 * @param pixels source
34 * @param line_size number of bytes in a horizontal line of block
35 * @param h height
36
37 */
38
39 DEFUN(put_pixels8uc,mL1,
40 (uint8_t *block, const uint8_t *s0, const uint8_t *s1,
41 int dest_size, int line_size, int h)):
42 i3=r0; // dest
43 i0=r1; // src0
44 i1=r2; // src1
45 r0=[sp+12]; // dest_size
46 r2=[sp+16]; // line_size
47 p0=[sp+20]; // h
48 [--sp] = (r7:6);
49 r0+=-4;
50 m3=r0;
51 r2+=-8;
52 m0=r2;
53 LSETUP(pp8$0,pp8$1) LC0=P0;
54 DISALGNEXCPT || R0 = [I0++] || R2 =[I1++];
55
56 pp8$0: DISALGNEXCPT || R1 = [I0++] || R3 =[I1++];
57 R6 = BYTEOP1P(R1:0,R3:2) || R0 = [I0++M0]|| R2 =[I1++M0];
58 R7 = BYTEOP1P(R1:0,R3:2)(R) || R0 = [I0++] || [I3++] = R6 ;
59 pp8$1: DISALGNEXCPT || R2 = [I1++] || [I3++M3] = R7;
60
61 (r7:6) = [sp++];
62 RTS;
63 DEFUN_END(put_pixels8uc)
64
65 DEFUN(put_pixels16uc,mL1,
66 (uint8_t *block, const uint8_t *s0, const uint8_t *s1,
67 int dest_size, int line_size, int h)):
68 link 0;
69 [--sp] = (r7:6);
70 i3=r0; // dest
71 i0=r1; // src0
72 i1=r2; // src1
73 r0=[fp+20]; // dest_size
74 r2=[fp+24]; // line_size
75 p0=[fp+28]; // h
76
77
78 r0+=-12;
79 m3=r0; // line_size
80 r2+=-16;
81 m0=r2;
82
83 LSETUP(pp16$0,pp16$1) LC0=P0;
84 DISALGNEXCPT || R0 = [I0++] || R2 =[I1++];
85
86 pp16$0: DISALGNEXCPT || R1 = [I0++] || R3 =[I1++];
87 R6 = BYTEOP1P(R1:0,R3:2) || R0 = [I0++] || R2 =[I1++];
88 R7 = BYTEOP1P(R1:0,R3:2)(R) || R1 = [I0++] || R3 =[I1++];
89 [I3++] = R6;
90 R6 = BYTEOP1P(R1:0,R3:2) || R0 = [I0++M0] || R2 =[I1++M0];
91 R7 = BYTEOP1P(R1:0,R3:2)(R) || R0 = [I0++] || [I3++] = R7 ;
92 [I3++] = R6;
93 pp16$1: DISALGNEXCPT || R2 = [I1++] || [I3++M3] = R7;
94
95 (r7:6) = [sp++];
96 unlink;
97 RTS;
98 DEFUN_END(put_pixels16uc)
99
100
101
102
103
104
105 DEFUN(put_pixels8uc_nornd,mL1,
106 (uint8_t *block, const uint8_t *s0, const uint8_t *s1,
107 int line_size, int h)):
108 i3=r0; // dest
109 i0=r1; // src0
110 i1=r2; // src1
111 r2=[sp+12]; // line_size
112 p0=[sp+16]; // h
113 [--sp] = (r7:6);
114 r2+=-4;
115 m3=r2;
116 r2+=-4;
117 m0=r2;
118 LSETUP(pp8$2,pp8$3) LC0=P0;
119 DISALGNEXCPT || R0 = [I0++] || R2 =[I1++];
120
121 pp8$2: DISALGNEXCPT || R1 = [I0++] || R3 =[I1++];
122 R6 = BYTEOP1P(R1:0,R3:2)(T) || R0 = [I0++M0]|| R2 =[I1++M0];
123 R7 = BYTEOP1P(R1:0,R3:2)(T,R) || R0 = [I0++] || [I3++] = R6 ;
124 pp8$3: DISALGNEXCPT || R2 = [I1++] || [I3++M3] = R7;
125
126 (r7:6) = [sp++];
127 RTS;
128 DEFUN_END(put_pixels8uc_nornd)
129
130 DEFUN(put_pixels16uc_nornd,mL1,
131 (uint8_t *block, const uint8_t *s0, const uint8_t *s1,
132 int line_size, int h)):
133 i3=r0; // dest
134 i0=r1; // src0
135 i1=r2; // src1
136 r2=[sp+12]; // line_size
137 p0=[sp+16]; // h
138
139 [--sp] = (r7:6);
140 r2+=-12;
141 m3=r2; // line_size
142 r2+=-4;
143 m0=r2;
144
145 LSETUP(pp16$2,pp16$3) LC0=P0;
146 DISALGNEXCPT || R0 = [I0++] || R2 =[I1++];
147
148 pp16$2:
149 DISALGNEXCPT || R1 = [I0++] || R3 =[I1++];
150 R6 = BYTEOP1P(R1:0,R3:2)(T) || R0 = [I0++] || R2 =[I1++];
151 R7 = BYTEOP1P(R1:0,R3:2)(T,R) || R1 = [I0++] || R3 =[I1++];
152 [I3++] = R6;
153
154 R6 = BYTEOP1P(R1:0,R3:2)(T) || R0 = [I0++M0] || R2 =[I1++M0];
155 R7 = BYTEOP1P(R1:0,R3:2)(T,R) || R0 = [I0++] || [I3++] = R7 ;
156 [I3++] = R6;
157 pp16$3: DISALGNEXCPT || R2 = [I1++] || [I3++M3] = R7;
158
159 (r7:6) = [sp++];
160
161 RTS;
162 DEFUN_END(put_pixels16uc_nornd)
163
164 DEFUN(z_put_pixels16_xy2,mL1,
165 (uint8_t *block, const uint8_t *s0,
166 int dest_size, int line_size, int h)):
167 link 0;
168 [--sp] = (r7:4);
169 i3=r0; // dest
170 i0=r1; // src0--> pixels
171 i1=r1; // src1--> pixels + line_size
172 r2+=-12;
173 m2=r2; // m2=dest_width-4
174 r2=[fp+20];
175 m3=r2; // line_size
176 p0=[fp+24]; // h
177 r2+=-16;
178 i1+=m3; /* src1 + line_size */
179 m0=r2; /* line-size - 20 */
180
181 B0 = I0;
182 B1 = I1;
183 B3 = I3;
184
185 DISALGNEXCPT || R0 = [I0++] || R2 =[I1++];
186
187 LSETUP(LS$16E,LE$16E) LC0=P0;
188 LS$16E: DISALGNEXCPT || R1 = [I0++] || R3 =[I1++];
189 R4 = BYTEOP2P (R3:2,R1:0) (RNDL) || R0 = [I0++] || R2 =[I1++];
190 R5 = BYTEOP2P (R3:2,R1:0) (RNDL,R) || R1 = [I0++] || [I3++] = R4 ;
191 DISALGNEXCPT || R3 = [I1++] || [I3++] = R5;
192 R4 = BYTEOP2P (R3:2,R1:0) (RNDL) || R0 = [I0++M0]|| R2 = [I1++M0];
193 R5 = BYTEOP2P (R3:2,R1:0) (RNDL,R) || R0 = [I0++] || [I3++] = R4 ;
194 LE$16E: DISALGNEXCPT || R2 = [I1++] || [I3++M2] = R5;
195
196 M1 = 1;
197 I3 = B3;
198 I1 = B1;
199 I0 = B0;
200
201 I0 += M1;
202 I1 += M1;
203
204 DISALGNEXCPT || R0 = [I0++] || R2 =[I1++];
205 LSETUP(LS$16O,LE$16O) LC0=P0;
206 LS$16O: DISALGNEXCPT || R1 = [I0++] || R3 =[I1++];
207 R4 = BYTEOP2P (R3:2,R1:0) (RNDH) || R0 = [I0++] || R2 =[I1++];
208 R5 = BYTEOP2P (R3:2,R1:0) (RNDH,R) || R1 = [I0++] || R6 =[I3++];
209 R4 = R4 +|+ R6 || R7 = [I3--];
210 R5 = R5 +|+ R7 || [I3++] = R4;
211 DISALGNEXCPT || R3 =[I1++] || [I3++] = R5;
212 R4 = BYTEOP2P (R3:2,R1:0) (RNDH) || R0 = [I0++M0]|| R2 = [I1++M0];
213 R5 = BYTEOP2P (R3:2,R1:0) (RNDH,R) || R0 = [I0++] || R6 = [I3++];
214 R4 = R4 +|+ R6 || R7 = [I3--];
215 R5 = R5 +|+ R7 || [I3++] = R4;
216 LE$16O: DISALGNEXCPT || R2 = [I1++] || [I3++M2] = R5;
217
218 (r7:4) = [sp++];
219 unlink;
220 rts;
221 DEFUN_END(z_put_pixels16_xy2)
222
223 DEFUN(z_put_pixels8_xy2,mL1,
224 (uint8_t *block, const uint8_t *s0,
225 int dest_size, int line_size, int h)):
226 link 0;
227 [--sp] = (r7:4);
228 i3=r0; // dest
229 i0=r1; // src0--> pixels
230 i1=r1; // src1--> pixels + line_size
231 r2+=-4;
232 m2=r2; // m2=dest_width-4
233 r2=[fp+20];
234 m3=r2; // line_size
235 p0=[fp+24]; // h
236 r2+=-8;
237 i1+=m3; /* src1 + line_size */
238 m0=r2; /* line-size - 20 */
239
240 b0 = I0;
241 b1 = I1;
242 b3 = I3;
243
244 LSETUP(LS$8E,LE$8E) LC0=P0;
245 DISALGNEXCPT || R0 = [I0++] || R2 =[I1++];
246 LS$8E: DISALGNEXCPT || R1 = [I0++] || R3 =[I1++];
247 R4 = BYTEOP2P (R3:2,R1:0) (RNDL) || R0 = [I0++M0] || R2 =[I1++M0];
248 R5 = BYTEOP2P (R3:2,R1:0) (RNDL,R) || R0 = [I0++] || [I3++] = R4 ;
249 LE$8E: DISALGNEXCPT || R2 = [I1++] || [I3++M2] = R5;
250
251 M1 = 1;
252 I3 = b3;
253 I1 = b1;
254 I0 = b0;
255
256 I0 += M1;
257 I1 += M1;
258
259 LSETUP(LS$8O,LE$8O) LC0=P0;
260 DISALGNEXCPT || R0 = [I0++] || R2 =[I1++];
261 LS$8O: DISALGNEXCPT || R1 = [I0++] || R3 =[I1++];
262 R4 = BYTEOP2P (R3:2,R1:0) (RNDH) || R0 = [I0++M0] || R2 =[I1++M0];
263 R5 = BYTEOP2P (R3:2,R1:0) (RNDH,R) || R0 = [I0++] || R6 =[I3++];
264 R4 = R4 +|+ R6 || R7 = [I3--];
265 R5 = R5 +|+ R7 || [I3++] = R4;
266 LE$8O: DISALGNEXCPT || R2 =[I1++] || [I3++M2] = R5;
267
268 (r7:4) = [sp++];
269 unlink;
270 rts;
271 DEFUN_END(z_put_pixels8_xy2)