Mercurial > libavcodec.hg

comparison alpha/dsputil_alpha_asm.S @ 2967:ef2149182f1c libavcodec

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
COSMETICS: Remove all trailing whitespace.
author diego
date Sat, 17 Dec 2005 18:14:38 +0000
parents 503496800167
children bfabfdf9ce55
comparison
equal deleted inserted replaced
2966:564788471dd4 2967:ef2149182f1c
32 /* Danger: these overlap with the argument list and the return value */ 32 /* Danger: these overlap with the argument list and the return value */
33 #define te a5 33 #define te a5
34 #define tf a4 34 #define tf a4
35 #define tg a3 35 #define tg a3
36 #define th v0 36 #define th v0
37 37
38 .set noat 38 .set noat
39 .set noreorder 39 .set noreorder
40 .arch pca56 40 .arch pca56
41 .text 41 .text
42 42
130 stq t1, 0(t4) 130 stq t1, 0(t4)
131 addq t6, a2, a0 131 addq t6, a2, a0
132 132
133 stq t2, 0(t5) 133 stq t2, 0(t5)
134 stq t3, 0(t6) 134 stq t3, 0(t6)
135 135
136 bne a3, $aligned 136 bne a3, $aligned
137 ret 137 ret
138 .end put_pixels_axp_asm 138 .end put_pixels_axp_asm
139 139
140 /************************************************************************ 140 /************************************************************************
141 * void put_pixels_clamped_mvi_asm(const DCTELEM *block, uint8_t *pixels, 141 * void put_pixels_clamped_mvi_asm(const DCTELEM *block, uint8_t *pixels,
142 * int line_size) 142 * int line_size)
143 */ 143 */
144 .align 6 144 .align 6
145 .globl put_pixels_clamped_mvi_asm 145 .globl put_pixels_clamped_mvi_asm
146 .ent put_pixels_clamped_mvi_asm 146 .ent put_pixels_clamped_mvi_asm
170 170
171 maxsw4 t2, zero, t2 171 maxsw4 t2, zero, t2
172 addq a1, a2, ta 172 addq a1, a2, ta
173 maxsw4 t3, zero, t3 173 maxsw4 t3, zero, t3
174 minsw4 t0, t8, t0 174 minsw4 t0, t8, t0
175 175
176 minsw4 t1, t8, t1 176 minsw4 t1, t8, t1
177 minsw4 t2, t8, t2 177 minsw4 t2, t8, t2
178 minsw4 t3, t8, t3 178 minsw4 t3, t8, t3
179 pkwb t0, t0 179 pkwb t0, t0
180 180
181 pkwb t1, t1 181 pkwb t1, t1
182 pkwb t2, t2 182 pkwb t2, t2
183 pkwb t3, t3 183 pkwb t3, t3
184 stl t0, 0(a1) 184 stl t0, 0(a1)
185 185
186 stl t1, 4(a1) 186 stl t1, 4(a1)
187 addq ta, a2, a1 187 addq ta, a2, a1
188 stl t2, 0(ta) 188 stl t2, 0(ta)
189 stl t3, 4(ta) 189 stl t3, 4(ta)
190 190
191 bne t9, 1b 191 bne t9, 1b
192 ret 192 ret
193 .end put_pixels_clamped_mvi_asm 193 .end put_pixels_clamped_mvi_asm
194 194
195 /************************************************************************ 195 /************************************************************************
196 * void add_pixels_clamped_mvi_asm(const DCTELEM *block, uint8_t *pixels, 196 * void add_pixels_clamped_mvi_asm(const DCTELEM *block, uint8_t *pixels,
197 * int line_size) 197 * int line_size)
198 */ 198 */
199 .align 6 199 .align 6
200 .globl add_pixels_clamped_mvi_asm 200 .globl add_pixels_clamped_mvi_asm
201 .ent add_pixels_clamped_mvi_asm 201 .ent add_pixels_clamped_mvi_asm
234 unpkbw t4, t4 # 1 0 234 unpkbw t4, t4 # 1 0
235 235
236 bic t0, tg, t0 # 0 2 236 bic t0, tg, t0 # 0 2
237 unpkbw t7, t7 # 2 0 237 unpkbw t7, t7 # 2 0
238 and t3, tg, t5 # 1 1 238 and t3, tg, t5 # 1 1
239 addq t0, t1, t0 # 0 3 239 addq t0, t1, t0 # 0 3
240 240
241 xor t0, t2, t0 # 0 4 241 xor t0, t2, t0 # 0 4
242 unpkbw ta, ta # 3 0 242 unpkbw ta, ta # 3 0
243 and t6, tg, t8 # 2 1 243 and t6, tg, t8 # 2 1
244 maxsw4 t0, zero, t0 # 0 5 244 maxsw4 t0, zero, t0 # 0 5
245 245
246 bic t3, tg, t3 # 1 2 246 bic t3, tg, t3 # 1 2
247 bic t6, tg, t6 # 2 2 247 bic t6, tg, t6 # 2 2
248 minsw4 t0, tf, t0 # 0 6 248 minsw4 t0, tf, t0 # 0 6
249 addq t3, t4, t3 # 1 3 249 addq t3, t4, t3 # 1 3
250 250
251 pkwb t0, t0 # 0 7 251 pkwb t0, t0 # 0 7
252 xor t3, t5, t3 # 1 4 252 xor t3, t5, t3 # 1 4
253 maxsw4 t3, zero, t3 # 1 5 253 maxsw4 t3, zero, t3 # 1 5
254 addq t6, t7, t6 # 2 3 254 addq t6, t7, t6 # 2 3
255 255
258 minsw4 t3, tf, t3 # 1 6 258 minsw4 t3, tf, t3 # 1 6
259 bic t9, tg, t9 # 3 2 259 bic t9, tg, t9 # 3 2
260 260
261 maxsw4 t6, zero, t6 # 2 5 261 maxsw4 t6, zero, t6 # 2 5
262 addq t9, ta, t9 # 3 3 262 addq t9, ta, t9 # 3 3
263 stl t0, 0(a1) # 0 8 263 stl t0, 0(a1) # 0 8
264 minsw4 t6, tf, t6 # 2 6 264 minsw4 t6, tf, t6 # 2 6
265 265
266 xor t9, tb, t9 # 3 4 266 xor t9, tb, t9 # 3 4
267 maxsw4 t9, zero, t9 # 3 5 267 maxsw4 t9, zero, t9 # 3 5
268 lda a0, 32(a0) # block += 16; 268 lda a0, 32(a0) # block += 16;
269 pkwb t3, t3 # 1 7 269 pkwb t3, t3 # 1 7
270 270
271 minsw4 t9, tf, t9 # 3 6 271 minsw4 t9, tf, t9 # 3 6
272 subq th, 2, th 272 subq th, 2, th
273 pkwb t6, t6 # 2 7 273 pkwb t6, t6 # 2 7
274 pkwb t9, t9 # 3 7 274 pkwb t9, t9 # 3 7
275 275
277 addq te, a2, a1 # pixels += line_size 277 addq te, a2, a1 # pixels += line_size
278 stl t6, 0(te) # 2 8 278 stl t6, 0(te) # 2 8
279 stl t9, 4(te) # 3 8 279 stl t9, 4(te) # 3 8
280 280
281 bne th, 1b 281 bne th, 1b
282 ret 282 ret
283 .end add_pixels_clamped_mvi_asm 283 .end add_pixels_clamped_mvi_asm