开通博客赚积分
发布资源赚积分
分类

源码开发语言/平台
当前位置：首页 > 源码/资料 > 嵌入式/单片机编程 > Windows CE > 查看源码
mcomp_mips32.c
资源名称：tcpmp.rar　[点击查看]
上传用户：wstnjxml
上传日期：2014-04-03
资源大小：7248k
文件大小：10k
源码类别：

Windows CE

开发平台：

C/C++

mcomp_mips32.c：源码内容
  1. /*****************************************************************************
  2.  *
  3.  * This program is free software ; you can redistribute it and/or modify
  4.  * it under the terms of the GNU General Public License as published by
  5.  * the Free Software Foundation; either version 2 of the License, or
  6.  * (at your option) any later version.
  7.  *
  8.  * This program is distributed in the hope that it will be useful,
  9.  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  10.  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  11.  * GNU General Public License for more details.
  12.  *
  13.  * You should have received a copy of the GNU General Public License
  14.  * along with this program; if not, write to the Free Software
  15.  * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307 USA
  16.  *
  17.  * $Id: mcomp_mips32.c 323 2005-11-01 20:52:32Z picard $
  18.  *
  19.  * The Core Pocket Media Player
  20.  * Copyright (c) 2004-2005 Gabor Kovacs
  21.  *
  22.  ****************************************************************************/
  24. #include "../common.h"
  25. #include "softidct.h"
  27. #if defined(MIPS32)
  29. // $8 src end pointer
  30. // $4 src pointer
  31. // $5 dst pointer
  32. // $6 src stride
  33. // $7 dst stride
  34. // $2,$9 first item lower 4 bytes (in two forms)
  35. // $10,$11 first item upper 4 bytes (in two forms)
  36. // $12,$13 second item lower 4 bytes (in two forms)
  37. // $14,$15 second item upper 4 bytes (in two forms)
  38. // $24 0x0101 0101 - for non horver
  39. // $25 0xFEFE FEFE - for non horver
  40. // $24 rounding - for horver
  41. // $25 temporary - for horver
  42. // $3 0x0303 0303 - for horver
  43. // $1 0xFCFC FCFC - for horver
  44. // $16 temporary - for horver (must be saved/restored)
  46. #define SET_SRCEND8
  47. "sll $8,$6,3;"
  48. "addu $8,$4,$8;"
  50. #define SET_SRCEND16
  51. "sll $8,$6,4;"
  52. "addu $8,$4,$8;"
  54. #define SET_MASKS
  55.     "li $24,0x01010101;"
  56. "nor $25,$24,$0;"
  58. #define SET_MASKS2
  59. ".set noat;"
  60.     "li $3,0x03030303;"
  61. "nor $1,$3,$0;"
  63. #define LOAD_FIRST8(ofs)
  64. "ulw $2, " #ofs "($4);"
  65. "ulw $10," #ofs "+4($4);"
  66. "and $9,$2,$25;"
  67. "and $11,$10,$25;"
  68. "srl $9,$9,1;"
  69. "srl $11,$11,1;"
  71. #define LOAD_SECOND8(ofs)
  72. "ulw $12," #ofs "($4);"
  73. "ulw $14," #ofs "+4($4);"
  74. "and $13,$12,$25;"
  75. "and $15,$14,$25;"
  76. "srl $13,$13,1;"
  77. "srl $15,$15,1;"
  79. #define LOAD_FIRST8_HV
  80. "ulw $2,0($4);"
  81. "ulw $9,1($4);"
  82. "and $16,$2,$1;"
  83. "and $25,$9,$1;"
  84. "and $2,$2,$3;"
  85. "and $9,$9,$3;"
  86. "srl $16,$16,2;"
  87. "srl $25,$25,2;"
  88. "addu $2,$2,$9;"
  89. "addu $9,$16,$25;"
  91. "ulw $10,4($4);"
  92. "ulw $11,5($4);"
  93. "and $16,$10,$1;"
  94. "and $25,$11,$1;"
  95. "and $10,$10,$3;"
  96. "and $11,$11,$3;"
  97. "srl $16,$16,2;"
  98. "srl $25,$25,2;"
  99. "addu $10,$10,$11;"
  100. "addu $11,$16,$25;"
  102. #define LOAD_SECOND8_HV
  103. "ulw $12,0($4);"
  104. "ulw $13,1($4);"
  105. "and $16,$12,$1;"
  106. "and $25,$13,$1;"
  107. "and $12,$12,$3;"
  108. "and $13,$13,$3;"
  109. "srl $16,$16,2;"
  110. "srl $25,$25,2;"
  111. "addu $12,$12,$13;"
  112. "addu $13,$16,$25;"
  114. "ulw $14,4($4);"
  115. "ulw $15,5($4);"
  116. "and $16,$14,$1;"
  117. "and $25,$15,$1;"
  118. "and $14,$14,$3;"
  119. "and $15,$15,$3;"
  120. "srl $16,$16,2;"
  121. "srl $25,$25,2;"
  122. "addu $14,$14,$15;"
  123. "addu $15,$16,$25;"
  125. #define AVG8
  126. "or $2,$2,$12;"
  127. "or $10,$10,$14;"
  128. "and $2,$2,$24;"
  129. "and $10,$10,$24;"
  130. "addu $2,$2,$9;"
  131. "addu $10,$10,$11;"
  132. "addu $2,$2,$13;"
  133. "addu $10,$10,$15;"
  135. #define AVGROUND8
  136. "and $2,$2,$12;"
  137. "and $10,$10,$14;"
  138. "and $2,$2,$24;"
  139. "and $10,$10,$24;"
  140. "addu $2,$2,$9;"
  141. "addu $10,$10,$11;"
  142. "addu $2,$2,$13;"
  143. "addu $10,$10,$15;"
  145. #define SWAPSET8
  146. "move $2,$12;"
  147. "move $9,$13;"
  148. "move $10,$14;"
  149. "move $11,$15;"
  151. #define WRITE8
  152. "sw $2,0($5);"
  153. "sw $10,4($5);"
  154. "addu $5,$5,$7;" 
  156. #define SAVE
  157. "addiu $sp,$sp,-4;"
  158. "sw $16,0(sp);"
  160. #define RESTORE
  161. "lw $16,0(sp);"
  162. "addiu $sp,$sp,4;"
  164. void CopyBlock(unsigned char * Src, unsigned char * Dst, int SrcStride, int DstStride) 
  165. { 
  166. __asm( SET_SRCEND8
  168. "loop:"
  169. "ulw $2,0($4);"
  170. "ulw $10,4($4);" 
  171. "addu $4,$4,$6;"
  172. "sw $2,0($5);"
  173. "sw $10,4($5);"
  174. "addu $5,$5,$7;" 
  175. "bne $4,$8,loop;"); 
  176. } 
  178. void CopyBlockHor(unsigned char * Src, unsigned char * Dst, int SrcStride, int DstStride) 
  179. { 
  180. __asm ( SET_SRCEND8
  181. SET_MASKS
  183. "loophor:"
  184. LOAD_FIRST8(0)
  185. LOAD_SECOND8(1)
  186. "addu $4,$4,$6;" 
  188. AVG8
  189. WRITE8
  191. "bne $4,$8,loophor;");
  192. } 
  194. void CopyBlockHorRound(unsigned char * Src, unsigned char * Dst, int SrcStride, int DstStride) 
  195. { 
  196. __asm ( SET_SRCEND8
  197. SET_MASKS
  199. "loophorround:"
  200. LOAD_FIRST8(0)
  201. LOAD_SECOND8(1)
  203. "addu $4,$4,$6;" 
  205. AVGROUND8
  206. WRITE8
  208. "bne $4,$8,loophorround;");
  209. } 
  211. void CopyBlockVer(unsigned char * Src, unsigned char * Dst, int SrcStride, int DstStride) 
  212. { 
  213. __asm ( SET_SRCEND8
  214. SET_MASKS
  216. LOAD_FIRST8(0)
  218. "loopver:"
  219. "addu $4,$4,$6;" 
  221. LOAD_SECOND8(0)
  223. AVG8
  224. WRITE8
  225. SWAPSET8
  227. "bne $4,$8,loopver;");
  228. } 
  230. void CopyBlockVerRound(unsigned char * Src, unsigned char * Dst, int SrcStride, int DstStride) 
  231. { 
  232. __asm ( SET_SRCEND8
  233. SET_MASKS
  235. LOAD_FIRST8(0)
  237. "loopverround:"
  238. "addu $4,$4,$6;" 
  240. LOAD_SECOND8(0)
  242. AVGROUND8
  243. WRITE8
  244. SWAPSET8
  246. "bne $4,$8,loopverround;");
  247. }
  249. void CopyBlockHorVer(unsigned char * Src, unsigned char * Dst, int SrcStride, int DstStride) 
  250. { 
  251. __asm ( SAVE
  252.     SET_SRCEND8
  253. SET_MASKS2
  255. "sll $24,$3,1;"  
  256. "and $24,$24,$3;" // 0x0202 0202
  258. //preprocessing
  260. LOAD_FIRST8_HV
  262. "loophorver:"
  263. "addu $4,$4,$6;"
  265. LOAD_SECOND8_HV);
  267. __asm ( "addu $2,$2,$12;"
  268. "addu $9,$9,$13;"
  269. "addu $10,$10,$14;"
  270. "addu $11,$11,$15;"
  272. "addu $2,$2,$24;"
  273. "addu $10,$10,$24;"
  275. "and $2,$2,$1;"
  276. "and $10,$10,$1;"
  277. "srl $2,$2,2;"
  278. "srl $10,$10,2;"
  279. "addu $2,$2,$9;"
  280. "addu $10,$10,$11;"
  282. WRITE8
  283. SWAPSET8
  285. "bne $4,$8,loophorver;"
  286. RESTORE);
  287. } 
  289. void CopyBlockHorVerRound(unsigned char * Src, unsigned char * Dst, int SrcStride, int DstStride) 
  290. { 
  291. __asm ( SAVE
  292.     SET_SRCEND8
  293. SET_MASKS2
  295. "srl $24,$3,1;"  
  296. "and $24,$24,$3;" // 0x0101 0101
  298. //preprocessing
  300. LOAD_FIRST8_HV
  302. "loophorverround:"
  303. "addu $4,$4,$6;"
  305. LOAD_SECOND8_HV);
  307. __asm ( "addu $2,$2,$12;"
  308. "addu $9,$9,$13;"
  309. "addu $10,$10,$14;"
  310. "addu $11,$11,$15;"
  312. "addu $2,$2,$24;"
  313. "addu $10,$10,$24;"
  315. "and $2,$2,$1;"
  316. "and $10,$10,$1;"
  317. "srl $2,$2,2;"
  318. "srl $10,$10,2;"
  319. "addu $2,$2,$9;"
  320. "addu $10,$10,$11;"
  322. WRITE8
  323. SWAPSET8
  325. "bne $4,$8,loophorverround;"
  326. RESTORE);
  327. }
  329. void AddBlock8x8(unsigned char * Src, unsigned char * Dst, int SrcStride, int DstStride) 
  330. { 
  331. __asm( SET_SRCEND8
  332. SET_MASKS
  334. "loopadd:"
  335. "lw $2,0($4);" 
  336. "lw $10,4($4);" 
  337. "addu $4,$4,$6;" 
  339. "lw $9,0($5);"
  340. "and $11,$2,$25;"
  341. "or $2,$2,$9;"
  342. "and $2,$2,$24;"
  343. "srl $11,$11,1;"
  344. "addu $2,$2,$11;"
  345. "and $9,$9,$25;"
  346. "srl $9,$9,1;"
  347. "addu $2,$2,$9;"
  349. "lw $11,4($5);"
  350. "and $9,$10,$25;"
  351. "or $10,$10,$11;"
  352. "and $10,$10,$24;"
  353. "srl $9,$9,1;"
  354. "addu $10,$10,$9;"
  355. "and $11,$11,$25;"
  356. "srl $11,$11,1;"
  357. "addu $10,$10,$11;"
  359. "sw $2,0($5);"
  360. "sw $10,4($5);"
  361. "addu $5,$5,$7;" 
  363. "bne $4,$8,loopadd;"); 
  364. } 
  366. void AddBlock4x4(unsigned char * Src, unsigned char * Dst, int SrcStride, int DstStride) 
  367. { 
  368. __asm( SET_MASKS
  370. "lw $2,0($4);" 
  371. "addu $4,$4,$6;" 
  372. "lw $10,0($4);" 
  373. "addu $4,$4,$6;" 
  375. "lw $9,0($5);"
  376. "and $11,$2,$25;"
  377. "or $2,$2,$9;"
  378. "and $2,$2,$24;"
  379. "srl $11,$11,1;"
  380. "addu $2,$2,$11;"
  381. "and $9,$9,$25;"
  382. "srl $9,$9,1;"
  383. "addu $2,$2,$9;"
  384. "sw $2,0($5);"
  385. "addu $5,$5,$7;"
  387. "lw $11,0($5);"
  388. "and $9,$10,$25;"
  389. "or $10,$10,$11;"
  390. "and $10,$10,$24;"
  391. "srl $9,$9,1;"
  392. "addu $10,$10,$9;"
  393. "and $11,$11,$25;"
  394. "srl $11,$11,1;"
  395. "addu $10,$10,$11;"
  396. "sw $10,0($5);"
  397. "addu $5,$5,$7;" 
  399. "lw $2,0($4);" 
  400. "addu $4,$4,$6;" 
  401. "lw $10,0($4);" 
  402. "addu $4,$4,$6;" 
  404. "lw $9,0($5);"
  405. "and $11,$2,$25;"
  406. "or $2,$2,$9;"
  407. "and $2,$2,$24;"
  408. "srl $11,$11,1;"
  409. "addu $2,$2,$11;"
  410. "and $9,$9,$25;"
  411. "srl $9,$9,1;"
  412. "addu $2,$2,$9;"
  413. "sw $2,0($5);"
  414. "addu $5,$5,$7;"
  416. "lw $11,0($5);"
  417. "and $9,$10,$25;"
  418. "or $10,$10,$11;"
  419. "and $10,$10,$24;"
  420. "srl $9,$9,1;"
  421. "addu $10,$10,$9;"
  422. "and $11,$11,$25;"
  423. "srl $11,$11,1;"
  424. "addu $10,$10,$11;"
  425. "sw $10,0($5);"
  426. "addu $5,$5,$7;" ); 
  427. } 
  429. void CopyBlock16x16(unsigned char * Src, unsigned char * Dst, int SrcStride, int DstStride) 
  430. { 
  431. __asm( SET_SRCEND16
  433. "loopcopy16:"
  434. "lw $2,0($4);" 
  435. "lw $10,4($4);" 
  436. "lw $12,8($4);" 
  437. "lw $14,12($4);" 
  438. "addu $4,$4,$6;"
  439. "sw $2,0($5);"
  440. "sw $10,4($5);"
  441. "sw $12,8($5);"
  442. "sw $14,12($5);"
  443. "addu $5,$5,$7;" 
  444. "bne $4,$8,loopcopy16;"); 
  445. } 
  447. void CopyBlock8x8(unsigned char * Src, unsigned char * Dst, int SrcStride, int DstStride) 
  448. { 
  449. __asm( SET_SRCEND8
  451. "loopcopy8:"
  452. "lw $2,0($4);" 
  453. "lw $10,4($4);" 
  454. "addu $4,$4,$6;"
  455. "lw $12,0($4);" 
  456. "lw $14,4($4);" 
  457. "addu $4,$4,$6;"
  458. "sw $2,0($5);"
  459. "sw $10,4($5);"
  460. "addu $5,$5,$7;" 
  461. "sw $12,0($5);"
  462. "sw $14,4($5);"
  463. "addu $5,$5,$7;" 
  464. "bne $4,$8,loopcopy8;"); 
  465. } 
  467. void CopyBlock4x4(unsigned char * Src, unsigned char * Dst, int SrcStride, int DstStride) 
  468. { 
  469. __asm( "lw $2,0($4);" 
  470. "addu $4,$4,$6;"
  471. "lw $10,0($4);" 
  472. "addu $4,$4,$6;"
  473. "lw $12,0($4);" 
  474. "addu $4,$4,$6;"
  475. "lw $14,0($4);" 
  476. "sw $2,0($5);"
  477. "addu $5,$5,$7;" 
  478. "sw $10,0($5);"
  479. "addu $5,$5,$7;" 
  480. "sw $12,0($5);"
  481. "addu $5,$5,$7;" 
  482. "sw $14,0($5);"); 
  483. } 
  485. void CopyBlockM(unsigned char * Src, unsigned char * Dst, int SrcStride, int DstStride) 
  486. { 
  487. __asm( SET_SRCEND16
  489. "loopm:"
  490. "ulw $2,0($4);"
  491. "ulw $10,4($4);" 
  492. "ulw $12,8($4);"
  493. "ulw $14,12($4);" 
  494. "addu $4,$4,$6;"
  495. "sw $2,0($5);"
  496. "sw $10,4($5);"
  497. "sw $12,8($5);"
  498. "sw $14,12($5);"
  499. "addu $5,$5,$7;" 
  500. "bne $4,$8,loopm;"); 
  501. } 
  503. #endif