RepoMirrors
/
ffmpeg
mirror of https://git.ffmpeg.org/ffmpeg.git
1
0
Fork 0
Code Issues Projects Releases Wiki Activity

avcodec/fic: avoid 2 additions per idct row 

before:
5225 decicycles in IDCT, 32756 runs, 12 skips

after:
5057 decicycles in IDCT, 32765 runs, 3 skips

Reviewed-by: Derek Buitenhuis <derek.buitenhuis@gmail.com>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
This commit is contained in:
Michael Niedermayer 2014-04-23 21:06:50 +02:00
parent c26b4b6f55
commit 23cd5cb18f
1 changed files with 7 additions and 6 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

13
libavcodec/fic.c
View File

@ -79,7 +79,7 @@ static const uint8_t fic_header[7] = { 0, 0, 1, 'F', 'I', 'C', 'V' };
#define FIC_HEADER_SIZE 27
static av_always_inline void fic_idct(int16_t *blk, int step, int shift)
static av_always_inline void fic_idct(int16_t *blk, int step, int shift, int rnd)
{
const int t0 = 27246 * blk[3 * step] + 18405 * blk[5 * step];
const int t1 = 27246 * blk[5 * step] - 18405 * blk[3 * step];
@ -91,8 +91,8 @@ static av_always_inline void fic_idct(int16_t *blk, int step, int shift)
const int t7 = t3 - t1;
const int t8 = 17734 * blk[2 * step] - 42813 * blk[6 * step];
const int t9 = 17734 * blk[6 * step] + 42814 * blk[2 * step];
const int tA = (blk[0 * step] - blk[4 * step] << 15) + (1 << shift - 1);
const int tB = (blk[0 * step] + blk[4 * step] << 15) + (1 << shift - 1);
const int tA = (blk[0 * step] - blk[4 * step] << 15) + rnd;
const int tB = (blk[0 * step] + blk[4 * step] << 15) + rnd;
blk[0 * step] = ( t4 + t9 + tB) >> shift;
blk[1 * step] = ( t6 + t7 + t8 + tA) >> shift;
blk[2 * step] = ( t6 - t7 - t8 + tA) >> shift;
@ -109,14 +109,15 @@ static void fic_idct_put(uint8_t *dst, int stride, int16_t *block)
int16_t *ptr;
ptr = block;
for (i = 0; i < 8; i++) {
fic_idct(ptr, 8, 13);
fic_idct(ptr++, 8, 13, (1 << 12) + (1 << 17));
for (i = 1; i < 8; i++) {
fic_idct(ptr, 8, 13, 1 << 12);
ptr++;
}
ptr = block;
for (i = 0; i < 8; i++) {
fic_idct(ptr, 1, 20);
fic_idct(ptr, 1, 20, 0);
ptr += 8;
}