RepoMirrors/ffmpeg
1
0
Fork 0
mirror of https://git.ffmpeg.org/ffmpeg.git synced 2025-02-27 01:01:07 +00:00
Code Issues Projects Releases Wiki Activity

swscale: add dithering to yuv2yuvX_altivec_real

Browse Source 
It just does that part in scalar form, I doubt using a vector store
over 2 array would speed it up particularly.

The function should be written to not use a scratch buffer.
This commit is contained in:
Luca Barbato 2011-08-10 02:16:26 +02:00
parent 78622ef362
commit 3304a1e69a
1 changed files with 7 additions and 14 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

21
libswscale/ppc/swscale_altivec.c
View File

@ -92,6 +92,7 @@ altivec_packIntArrayToCharArray(int *val, uint8_t* dest, int dstW)
}
}
//FIXME remove the usage of scratch buffers.
static void
yuv2yuvX_altivec_real(SwsContext *c,
const int16_t *lumFilter, const int16_t **lumSrc,
@ -101,17 +102,13 @@ yuv2yuvX_altivec_real(SwsContext *c,
uint8_t *dest[4], int dstW, int chrDstW)
{
uint8_t *yDest = dest[0], *uDest = dest[1], *vDest = dest[2];
const vector signed int vini = {(1 << 18), (1 << 18), (1 << 18), (1 << 18)};
const uint8_t *lumDither = c->lumDither8, *chrDither = c->chrDither8;
register int i, j;
{
DECLARE_ALIGNED(16, int, val)[dstW];
for (i = 0; i < (dstW -7); i+=4) {
vec_st(vini, i << 2, val);
}
for (; i < dstW; i++) {
val[i] = (1 << 18);
}
for (i=0; i<dstW; i++)
val[i] = lumDither[i & 7] << 12;
for (j = 0; j < lumFilterSize; j++) {
vector signed short l1, vLumFilter = vec_ld(j << 1, lumFilter);
@ -155,13 +152,9 @@ yuv2yuvX_altivec_real(SwsContext *c,
DECLARE_ALIGNED(16, int, u)[chrDstW];
DECLARE_ALIGNED(16, int, v)[chrDstW];
for (i = 0; i < (chrDstW -7); i+=4) {
vec_st(vini, i << 2, u);
vec_st(vini, i << 2, v);
}
for (; i < chrDstW; i++) {
u[i] = (1 << 18);
v[i] = (1 << 18);
for (i=0; i<chrDstW; i++) {
u[i] = chrDither[i & 7] << 12;
v[i] = chrDither[(i + 3) & 7] << 12;
}
for (j = 0; j < chrFilterSize; j++) {