/* ------------------------------------------------------------------ * Copyright (C) 1998-2009 PacketVideo * * 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. * ------------------------------------------------------------------- */ #include "mp4def.h" #include "idct.h" #include "motion_comp.h" #ifdef FAST_IDCT /**************************************************************** * vca_idct.c : created 6/1/99 for several options * of hard-coded reduced idct function (using nz_coefs) ******************************************************************/ /*****************************************************/ //pretested version void idctrow0(int16 *, uint8 *, uint8 *, int) { return ; } void idctcol0(int16 *) { return ; } __attribute__((no_sanitize("signed-integer-overflow"))) void idctrow1(int16 *blk, uint8 *pred, uint8 *dst, int width) { /* shortcut */ int tmp; int i = 8; uint32 pred_word, dst_word; int res, res2; /* preset the offset, such that we can take advantage pre-offset addressing mode */ width -= 4; dst -= width; pred -= 12; blk -= 8; while (i--) { tmp = (*(blk += 8) + 32) >> 6; *blk = 0; pred_word = *((uint32*)(pred += 12)); /* read 4 bytes from pred */ res = tmp + (pred_word & 0xFF); CLIP_RESULT(res); res2 = tmp + ((pred_word >> 8) & 0xFF); CLIP_RESULT(res2); dst_word = (res2 << 8) | res; res = tmp + ((pred_word >> 16) & 0xFF); CLIP_RESULT(res); dst_word |= (res << 16); res = tmp + ((pred_word >> 24) & 0xFF); CLIP_RESULT(res); dst_word |= (res << 24); *((uint32*)(dst += width)) = dst_word; /* save 4 bytes to dst */ pred_word = *((uint32*)(pred += 4)); /* read 4 bytes from pred */ res = tmp + (pred_word & 0xFF); CLIP_RESULT(res); res2 = tmp + ((pred_word >> 8) & 0xFF); CLIP_RESULT(res2); dst_word = (res2 << 8) | res; res = tmp + ((pred_word >> 16) & 0xFF); CLIP_RESULT(res); dst_word |= (res << 16); res = tmp + ((pred_word >> 24) & 0xFF); CLIP_RESULT(res); dst_word |= (res << 24); *((uint32*)(dst += 4)) = dst_word; /* save 4 bytes to dst */ } return; } void idctcol1(int16 *blk) { /* shortcut */ blk[0] = blk[8] = blk[16] = blk[24] = blk[32] = blk[40] = blk[48] = blk[56] = blk[0] << 3; return; } __attribute__((no_sanitize("signed-integer-overflow"))) void idctrow2(int16 *blk, uint8 *pred, uint8 *dst, int width) { int32 x0, x1, x2, x4, x5; int i = 8; uint32 pred_word, dst_word; int res, res2; /* preset the offset, such that we can take advantage pre-offset addressing mode */ width -= 4; dst -= width; pred -= 12; blk -= 8; while (i--) { /* shortcut */ x4 = blk[9]; blk[9] = 0; x0 = ((*(blk += 8)) << 8) + 8192; *blk = 0; /* for proper rounding in the fourth stage */ /* first stage */ x5 = (W7 * x4 + 4) >> 3; x4 = (W1 * x4 + 4) >> 3; /* third stage */ x2 = (181 * (x4 + x5) + 128) >> 8; x1 = (181 * (x4 - x5) + 128) >> 8; /* fourth stage */ pred_word = *((uint32*)(pred += 12)); /* read 4 bytes from pred */ res = (x0 + x4) >> 14; ADD_AND_CLIP1(res); res2 = (x0 + x2) >> 14; ADD_AND_CLIP2(res2); dst_word = (res2 << 8) | res; res = (x0 + x1) >> 14; ADD_AND_CLIP3(res); dst_word |= (res << 16); res = (x0 + x5) >> 14; ADD_AND_CLIP4(res); dst_word |= (res << 24); *((uint32*)(dst += width)) = dst_word; /* save 4 bytes to dst */ pred_word = *((uint32*)(pred += 4)); /* read 4 bytes from pred */ res = (x0 - x5) >> 14; ADD_AND_CLIP1(res); res2 = (x0 - x1) >> 14; ADD_AND_CLIP2(res2); dst_word = (res2 << 8) | res; res = (x0 - x2) >> 14; ADD_AND_CLIP3(res); dst_word |= (res << 16); res = (x0 - x4) >> 14; ADD_AND_CLIP4(res); dst_word |= (res << 24); *((uint32*)(dst += 4)) = dst_word; /* save 4 bytes to dst */ } return ; } __attribute__((no_sanitize("signed-integer-overflow"))) void idctcol2(int16 *blk) { int32 x0, x1, x3, x5, x7;//, x8; x1 = blk[8]; x0 = ((int32)blk[0] << 11) + 128; /* both upper and lower*/ x7 = W7 * x1; x1 = W1 * x1; x3 = x7; x5 = (181 * (x1 - x7) + 128) >> 8; x7 = (181 * (x1 + x7) + 128) >> 8; blk[0] = (x0 + x1) >> 8; blk[8] = (x0 + x7) >> 8; blk[16] = (x0 + x5) >> 8; blk[24] = (x0 + x3) >> 8; blk[56] = (x0 - x1) >> 8; blk[48] = (x0 - x7) >> 8; blk[40] = (x0 - x5) >> 8; blk[32] = (x0 - x3) >> 8; return ; } __attribute__((no_sanitize("signed-integer-overflow"))) void idctrow3(int16 *blk, uint8 *pred, uint8 *dst, int width) { int32 x0, x1, x2, x3, x4, x5, x6, x7, x8; int i = 8; uint32 pred_word, dst_word; int res, res2; /* preset the offset, such that we can take advantage pre-offset addressing mode */ width -= 4; dst -= width; pred -= 12; blk -= 8; while (i--) { x2 = blk[10]; blk[10] = 0; x1 = blk[9]; blk[9] = 0; x0 = ((*(blk += 8)) << 8) + 8192; *blk = 0; /* for proper rounding in the fourth stage */ /* both upper and lower*/ /* both x2orx6 and x0orx4 */ x4 = x0; x6 = (W6 * x2 + 4) >> 3; x2 = (W2 * x2 + 4) >> 3; x8 = x0 - x2; x0 += x2; x2 = x8; x8 = x4 - x6; x4 += x6; x6 = x8; x7 = (W7 * x1 + 4) >> 3; x1 = (W1 * x1 + 4) >> 3; x3 = x7; x5 = (181 * (x1 - x7) + 128) >> 8; x7 = (181 * (x1 + x7) + 128) >> 8; pred_word = *((uint32*)(pred += 12)); /* read 4 bytes from pred */ res = (x0 + x1) >> 14; ADD_AND_CLIP1(res); res2 = (x4 + x7) >> 14; ADD_AND_CLIP2(res2); dst_word = (res2 << 8) | res; res = (x6 + x5) >> 14; ADD_AND_CLIP3(res); dst_word |= (res << 16); res = (x2 + x3) >> 14; ADD_AND_CLIP4(res); dst_word |= (res << 24); *((uint32*)(dst += width)) = dst_word; /* save 4 bytes to dst */ pred_word = *((uint32*)(pred += 4)); /* read 4 bytes from pred */ res = (x2 - x3) >> 14; ADD_AND_CLIP1(res); res2 = (x6 - x5) >> 14; ADD_AND_CLIP2(res2); dst_word = (res2 << 8) | res; res = (x4 - x7) >> 14; ADD_AND_CLIP3(res); dst_word |= (res << 16); res = (x0 - x1) >> 14; ADD_AND_CLIP4(res); dst_word |= (res << 24); *((uint32*)(dst += 4)) = dst_word; /* save 4 bytes to dst */ } return ; } __attribute__((no_sanitize("signed-integer-overflow"))) void idctcol3(int16 *blk) { int32 x0, x1, x2, x3, x4, x5, x6, x7, x8; x2 = blk[16]; x1 = blk[8]; x0 = ((int32)blk[0] << 11) + 128; x4 = x0; x6 = W6 * x2; x2 = W2 * x2; x8 = x0 - x2; x0 += x2; x2 = x8; x8 = x4 - x6; x4 += x6; x6 = x8; x7 = W7 * x1; x1 = W1 * x1; x3 = x7; x5 = (181 * (x1 - x7) + 128) >> 8; x7 = (181 * (x1 + x7) + 128) >> 8; blk[0] = (x0 + x1) >> 8; blk[8] = (x4 + x7) >> 8; blk[16] = (x6 + x5) >> 8; blk[24] = (x2 + x3) >> 8; blk[56] = (x0 - x1) >> 8; blk[48] = (x4 - x7) >> 8; blk[40] = (x6 - x5) >> 8; blk[32] = (x2 - x3) >> 8; return; } __attribute__((no_sanitize("signed-integer-overflow"))) void idctrow4(int16 *blk, uint8 *pred, uint8 *dst, int width) { int32 x0, x1, x2, x3, x4, x5, x6, x7, x8; int i = 8; uint32 pred_word, dst_word; int res, res2; /* preset the offset, such that we can take advantage pre-offset addressing mode */ width -= 4; dst -= width; pred -= 12; blk -= 8; while (i--) { x2 = blk[10]; blk[10] = 0; x1 = blk[9]; blk[9] = 0; x3 = blk[11]; blk[11] = 0; x0 = ((*(blk += 8)) << 8) + 8192; *blk = 0; /* for proper rounding in the fourth stage */ x4 = x0; x6 = (W6 * x2 + 4) >> 3; x2 = (W2 * x2 + 4) >> 3; x8 = x0 - x2; x0 += x2; x2 = x8; x8 = x4 - x6; x4 += x6; x6 = x8; x7 = (W7 * x1 + 4) >> 3; x1 = (W1 * x1 + 4) >> 3; x5 = (W3 * x3 + 4) >> 3; x3 = (- W5 * x3 + 4) >> 3; x8 = x1 - x5; x1 += x5; x5 = x8; x8 = x7 - x3; x3 += x7; x7 = (181 * (x5 + x8) + 128) >> 8; x5 = (181 * (x5 - x8) + 128) >> 8; pred_word = *((uint32*)(pred += 12)); /* read 4 bytes from pred */ res = (x0 + x1) >> 14; ADD_AND_CLIP1(res); res2 = (x4 + x7) >> 14; ADD_AND_CLIP2(res2); dst_word = (res2 << 8) | res; res = (x6 + x5) >> 14; ADD_AND_CLIP3(res); dst_word |= (res << 16); res = (x2 + x3) >> 14; ADD_AND_CLIP4(res); dst_word |= (res << 24); *((uint32*)(dst += width)) = dst_word; /* save 4 bytes to dst */ pred_word = *((uint32*)(pred += 4)); /* read 4 bytes from pred */ res = (x2 - x3) >> 14; ADD_AND_CLIP1(res); res2 = (x6 - x5) >> 14; ADD_AND_CLIP2(res2); dst_word = (res2 << 8) | res; res = (x4 - x7) >> 14; ADD_AND_CLIP3(res); dst_word |= (res << 16); res = (x0 - x1) >> 14; ADD_AND_CLIP4(res); dst_word |= (res << 24); *((uint32*)(dst += 4)) = dst_word; /* save 4 bytes to dst */ } return ; } __attribute__((no_sanitize("signed-integer-overflow"))) void idctcol4(int16 *blk) { int32 x0, x1, x2, x3, x4, x5, x6, x7, x8; x2 = blk[16]; x1 = blk[8]; x3 = blk[24]; x0 = ((int32)blk[0] << 11) + 128; x4 = x0; x6 = W6 * x2; x2 = W2 * x2; x8 = x0 - x2; x0 += x2; x2 = x8; x8 = x4 - x6; x4 += x6; x6 = x8; x7 = W7 * x1; x1 = W1 * x1; x5 = W3 * x3; x3 = -W5 * x3; x8 = x1 - x5; x1 += x5; x5 = x8; x8 = x7 - x3; x3 += x7; x7 = (181 * (x5 + x8) + 128) >> 8; x5 = (181 * (x5 - x8) + 128) >> 8; blk[0] = (x0 + x1) >> 8; blk[8] = (x4 + x7) >> 8; blk[16] = (x6 + x5) >> 8; blk[24] = (x2 + x3) >> 8; blk[56] = (x0 - x1) >> 8; blk[48] = (x4 - x7) >> 8; blk[40] = (x6 - x5) >> 8; blk[32] = (x2 - x3) >> 8; return ; } void idctrow0_intra(int16 *, PIXEL *, int) { return ; } void idctrow1_intra(int16 *blk, PIXEL *comp, int width) { /* shortcut */ int32 tmp; int i = 8; int offset = width; uint32 word; comp -= offset; while (i--) { tmp = ((blk[0] + 32) >> 6); blk[0] = 0; CLIP_RESULT(tmp) word = (tmp << 8) | tmp; word = (word << 16) | word; *((uint32*)(comp += offset)) = word; *((uint32*)(comp + 4)) = word; blk += B_SIZE; } return; } __attribute__((no_sanitize("signed-integer-overflow"))) void idctrow2_intra(int16 *blk, PIXEL *comp, int width) { int32 x0, x1, x2, x4, x5, temp; int i = 8; int offset = width; int32 word; comp -= offset; while (i--) { /* shortcut */ x4 = blk[1]; blk[1] = 0; x0 = ((int32)blk[0] << 8) + 8192; blk[0] = 0; /* for proper rounding in the fourth stage */ /* first stage */ x5 = (W7 * x4 + 4) >> 3; x4 = (W1 * x4 + 4) >> 3; /* third stage */ x2 = (181 * (x4 + x5) + 128) >> 8; x1 = (181 * (x4 - x5) + 128) >> 8; /* fourth stage */ word = ((x0 + x4) >> 14); CLIP_RESULT(word) temp = ((x0 + x2) >> 14); CLIP_RESULT(temp) word = word | (temp << 8); temp = ((x0 + x1) >> 14); CLIP_RESULT(temp) word = word | (temp << 16); temp = ((x0 + x5) >> 14); CLIP_RESULT(temp) word = word | (temp << 24); *((int32*)(comp += offset)) = word; word = ((x0 - x5) >> 14); CLIP_RESULT(word) temp = ((x0 - x1) >> 14); CLIP_RESULT(temp) word = word | (temp << 8); temp = ((x0 - x2) >> 14); CLIP_RESULT(temp) word = word | (temp << 16); temp = ((x0 - x4) >> 14); CLIP_RESULT(temp) word = word | (temp << 24); *((int32*)(comp + 4)) = word; blk += B_SIZE; } return ; } __attribute__((no_sanitize("signed-integer-overflow"))) void idctrow3_intra(int16 *blk, PIXEL *comp, int width) { int32 x0, x1, x2, x3, x4, x5, x6, x7, x8, temp; int i = 8; int offset = width; int32 word; comp -= offset; while (i--) { x2 = blk[2]; blk[2] = 0; x1 = blk[1]; blk[1] = 0; x0 = ((int32)blk[0] << 8) + 8192; blk[0] = 0;/* for proper rounding in the fourth stage */ /* both upper and lower*/ /* both x2orx6 and x0orx4 */ x4 = x0; x6 = (W6 * x2 + 4) >> 3; x2 = (W2 * x2 + 4) >> 3; x8 = x0 - x2; x0 += x2; x2 = x8; x8 = x4 - x6; x4 += x6; x6 = x8; x7 = (W7 * x1 + 4) >> 3; x1 = (W1 * x1 + 4) >> 3; x3 = x7; x5 = (181 * (x1 - x7) + 128) >> 8; x7 = (181 * (x1 + x7) + 128) >> 8; word = ((x0 + x1) >> 14); CLIP_RESULT(word) temp = ((x4 + x7) >> 14); CLIP_RESULT(temp) word = word | (temp << 8); temp = ((x6 + x5) >> 14); CLIP_RESULT(temp) word = word | (temp << 16); temp = ((x2 + x3) >> 14); CLIP_RESULT(temp) word = word | (temp << 24); *((int32*)(comp += offset)) = word; word = ((x2 - x3) >> 14); CLIP_RESULT(word) temp = ((x6 - x5) >> 14); CLIP_RESULT(temp) word = word | (temp << 8); temp = ((x4 - x7) >> 14); CLIP_RESULT(temp) word = word | (temp << 16); temp = ((x0 - x1) >> 14); CLIP_RESULT(temp) word = word | (temp << 24); *((int32*)(comp + 4)) = word; blk += B_SIZE; } return ; } __attribute__((no_sanitize("signed-integer-overflow"))) void idctrow4_intra(int16 *blk, PIXEL *comp, int width) { int32 x0, x1, x2, x3, x4, x5, x6, x7, x8, temp; int i = 8; int offset = width; int32 word; comp -= offset; while (i--) { x2 = blk[2]; blk[2] = 0; x1 = blk[1]; blk[1] = 0; x3 = blk[3]; blk[3] = 0; x0 = ((int32)blk[0] << 8) + 8192; blk[0] = 0;/* for proper rounding in the fourth stage */ x4 = x0; x6 = (W6 * x2 + 4) >> 3; x2 = (W2 * x2 + 4) >> 3; x8 = x0 - x2; x0 += x2; x2 = x8; x8 = x4 - x6; x4 += x6; x6 = x8; x7 = (W7 * x1 + 4) >> 3; x1 = (W1 * x1 + 4) >> 3; x5 = (W3 * x3 + 4) >> 3; x3 = (- W5 * x3 + 4) >> 3; x8 = x1 - x5; x1 += x5; x5 = x8; x8 = x7 - x3; x3 += x7; x7 = (181 * (x5 + x8) + 128) >> 8; x5 = (181 * (x5 - x8) + 128) >> 8; word = ((x0 + x1) >> 14); CLIP_RESULT(word) temp = ((x4 + x7) >> 14); CLIP_RESULT(temp) word = word | (temp << 8); temp = ((x6 + x5) >> 14); CLIP_RESULT(temp) word = word | (temp << 16); temp = ((x2 + x3) >> 14); CLIP_RESULT(temp) word = word | (temp << 24); *((int32*)(comp += offset)) = word; word = ((x2 - x3) >> 14); CLIP_RESULT(word) temp = ((x6 - x5) >> 14); CLIP_RESULT(temp) word = word | (temp << 8); temp = ((x4 - x7) >> 14); CLIP_RESULT(temp) word = word | (temp << 16); temp = ((x0 - x1) >> 14); CLIP_RESULT(temp) word = word | (temp << 24); *((int32*)(comp + 4)) = word; blk += B_SIZE; } return ; } #endif