--- simple_idct.c 2003/01/21 12:51:16 1.1 +++ simple_idct.c 2003/02/15 15:22:18 1.2 @@ -0,0 +1,258 @@ +/* + * Simple IDCT + * + * Copyright (c) 2001 Michael Niedermayer + * + * This library is free software; you can redistribute it and/or + * modify it under the terms of the GNU Lesser General Public + * License as published by the Free Software Foundation; either + * version 2 of the License, or (at your option) any later version. + * + * This library is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public + * License along with this library; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + */ +/* + based upon some outcommented c code from mpeg2dec (idct_mmx.c + written by Aaron Holtzman ) + */ +#include "../portab.h" +#include "idct.h" + +#if 0 +#define W1 2841 /* 2048*sqrt (2)*cos (1*pi/16) */ +#define W2 2676 /* 2048*sqrt (2)*cos (2*pi/16) */ +#define W3 2408 /* 2048*sqrt (2)*cos (3*pi/16) */ +#define W4 2048 /* 2048*sqrt (2)*cos (4*pi/16) */ +#define W5 1609 /* 2048*sqrt (2)*cos (5*pi/16) */ +#define W6 1108 /* 2048*sqrt (2)*cos (6*pi/16) */ +#define W7 565 /* 2048*sqrt (2)*cos (7*pi/16) */ +#define ROW_SHIFT 8 +#define COL_SHIFT 17 +#else +#define W1 22725 //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5 +#define W2 21407 //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5 +#define W3 19266 //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5 +#define W4 16383 //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5 +#define W5 12873 //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5 +#define W6 8867 //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5 +#define W7 4520 //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5 +#define ROW_SHIFT 11 +#define COL_SHIFT 20 // 6 +#endif + +#if defined(ARCH_IS_PPC) + +/* signed 16x16 -> 32 multiply add accumulate */ +#define MAC16(rt, ra, rb) \ + asm ("maclhw %0, %2, %3" : "=r" (rt) : "0" (rt), "r" (ra), "r" (rb)); + +/* signed 16x16 -> 32 multiply */ +#define MUL16(rt, ra, rb) \ + asm ("mullhw %0, %1, %2" : "=r" (rt) : "r" (ra), "r" (rb)); + +#else + +/* signed 16x16 -> 32 multiply add accumulate */ +#define MAC16(rt, ra, rb) rt += (ra) * (rb) + +/* signed 16x16 -> 32 multiply */ +#define MUL16(rt, ra, rb) rt = (ra) * (rb) + +#endif + +static __inline void idctRowCondDC (int16_t * const row) +{ + int a0, a1, a2, a3, b0, b1, b2, b3; +#ifdef FAST_64BIT + uint64_t temp; +#else + uint32_t temp; +#endif + +#ifdef FAST_64BIT +#ifdef ARCH_IS_BIG_ENDIAN +#define ROW0_MASK 0xffff000000000000LL +#else +#define ROW0_MASK 0xffffLL +#endif + if ( ((((uint64_t *)row)[0] & ~ROW0_MASK) | + ((uint64_t *)row)[1]) == 0) { + temp = (row[0] << 3) & 0xffff; + temp += temp << 16; + temp += temp << 32; + ((uint64_t *)row)[0] = temp; + ((uint64_t *)row)[1] = temp; + return; + } +#else + if (!(((uint32_t*)row)[1] | + ((uint32_t*)row)[2] | + ((uint32_t*)row)[3] | + row[1])) { + temp = (row[0] << 3) & 0xffff; + temp += temp << 16; + ((uint32_t*)row)[0]=((uint32_t*)row)[1] = + ((uint32_t*)row)[2]=((uint32_t*)row)[3] = temp; + return; + } +#endif + + a0 = (W4 * row[0]) + (1 << (ROW_SHIFT - 1)); + a1 = a0; + a2 = a0; + a3 = a0; + + /* no need to optimize : gcc does it */ + a0 += W2 * row[2]; + a1 += W6 * row[2]; + a2 -= W6 * row[2]; + a3 -= W2 * row[2]; + + MUL16(b0, W1, row[1]); + MAC16(b0, W3, row[3]); + MUL16(b1, W3, row[1]); + MAC16(b1, -W7, row[3]); + MUL16(b2, W5, row[1]); + MAC16(b2, -W1, row[3]); + MUL16(b3, W7, row[1]); + MAC16(b3, -W5, row[3]); + +#ifdef FAST_64BIT + temp = ((uint64_t*)row)[1]; +#else + temp = ((uint32_t*)row)[2] | ((uint32_t*)row)[3]; +#endif + if (temp != 0) { + a0 += W4*row[4] + W6*row[6]; + a1 += - W4*row[4] - W2*row[6]; + a2 += - W4*row[4] + W2*row[6]; + a3 += W4*row[4] - W6*row[6]; + + MAC16(b0, W5, row[5]); + MAC16(b0, W7, row[7]); + + MAC16(b1, -W1, row[5]); + MAC16(b1, -W5, row[7]); + + MAC16(b2, W7, row[5]); + MAC16(b2, W3, row[7]); + + MAC16(b3, W3, row[5]); + MAC16(b3, -W1, row[7]); + } + + row[0] = (a0 + b0) >> ROW_SHIFT; + row[7] = (a0 - b0) >> ROW_SHIFT; + row[1] = (a1 + b1) >> ROW_SHIFT; + row[6] = (a1 - b1) >> ROW_SHIFT; + row[2] = (a2 + b2) >> ROW_SHIFT; + row[5] = (a2 - b2) >> ROW_SHIFT; + row[3] = (a3 + b3) >> ROW_SHIFT; + row[4] = (a3 - b3) >> ROW_SHIFT; +} + + +static __inline void idctSparseCol (int16_t * const col) +{ + int a0, a1, a2, a3, b0, b1, b2, b3; + + /* XXX: I did that only to give same values as previous code */ + a0 = W4 * (col[8*0] + ((1<<(COL_SHIFT-1))/W4)); + a1 = a0; + a2 = a0; + a3 = a0; + + a0 += + W2*col[8*2]; + a1 += + W6*col[8*2]; + a2 += - W6*col[8*2]; + a3 += - W2*col[8*2]; + + MUL16(b0, W1, col[8*1]); + MUL16(b1, W3, col[8*1]); + MUL16(b2, W5, col[8*1]); + MUL16(b3, W7, col[8*1]); + + MAC16(b0, + W3, col[8*3]); + MAC16(b1, - W7, col[8*3]); + MAC16(b2, - W1, col[8*3]); + MAC16(b3, - W5, col[8*3]); + + if(col[8*4]){ + a0 += + W4*col[8*4]; + a1 += - W4*col[8*4]; + a2 += - W4*col[8*4]; + a3 += + W4*col[8*4]; + } + + if (col[8*5]) { + MAC16(b0, + W5, col[8*5]); + MAC16(b1, - W1, col[8*5]); + MAC16(b2, + W7, col[8*5]); + MAC16(b3, + W3, col[8*5]); + } + + if(col[8*6]){ + a0 += + W6*col[8*6]; + a1 += - W2*col[8*6]; + a2 += + W2*col[8*6]; + a3 += - W6*col[8*6]; + } + + if (col[8*7]) { + MAC16(b0, + W7, col[8*7]); + MAC16(b1, - W5, col[8*7]); + MAC16(b2, + W3, col[8*7]); + MAC16(b3, - W1, col[8*7]); + } + + col[0 ] = ((a0 + b0) >> COL_SHIFT); + col[8 ] = ((a1 + b1) >> COL_SHIFT); + col[16] = ((a2 + b2) >> COL_SHIFT); + col[24] = ((a3 + b3) >> COL_SHIFT); + col[32] = ((a3 - b3) >> COL_SHIFT); + col[40] = ((a2 - b2) >> COL_SHIFT); + col[48] = ((a1 - b1) >> COL_SHIFT); + col[56] = ((a0 - b0) >> COL_SHIFT); +} + +void simple_idct_c(int16_t * const block) +{ + int i; + for(i=0; i<8; i++) + idctRowCondDC(block + i*8); + + for(i=0; i<8; i++) + idctSparseCol(block + i); +} + + +/* Input permutation for the simple_idct_mmx */ +static const uint8_t simple_mmx_permutation[64]={ + 0x00, 0x08, 0x04, 0x09, 0x01, 0x0C, 0x05, 0x0D, + 0x10, 0x18, 0x14, 0x19, 0x11, 0x1C, 0x15, 0x1D, + 0x20, 0x28, 0x24, 0x29, 0x21, 0x2C, 0x25, 0x2D, + 0x12, 0x1A, 0x16, 0x1B, 0x13, 0x1E, 0x17, 0x1F, + 0x02, 0x0A, 0x06, 0x0B, 0x03, 0x0E, 0x07, 0x0F, + 0x30, 0x38, 0x34, 0x39, 0x31, 0x3C, 0x35, 0x3D, + 0x22, 0x2A, 0x26, 0x2B, 0x23, 0x2E, 0x27, 0x2F, + 0x32, 0x3A, 0x36, 0x3B, 0x33, 0x3E, 0x37, 0x3F, +}; + +#if defined(ARCH_IS_IA32) +/* wrapper function, as simple_idct_mmx expects data to be permutated */ +void simple_idct_mmx2(int16_t * const block) +{ + int16_t tmp[64]; + int i; + + for(i=0;i<64;i++) tmp[simple_mmx_permutation[i]] = block[i]; + simple_idct_mmx(tmp); + for(i=0;i<64;i++) block[i] = tmp[i]; +} +#endif