RepoMirrors/ffmpeg
1
0
Fork 0
mirror of https://git.ffmpeg.org/ffmpeg.git synced 2024-12-24 08:12:44 +00:00
Code Issues Projects Releases Wiki Activity

swscale/output: VSX-optimize 16-bit yuv2plane1

Browse Source 
./ffmpeg_g -f rawvideo -pix_fmt rgb24 -s hd1080 -i /dev/zero -pix_fmt yuv420p16le \
-f null -vframes 100 -v error -nostats -

2120 UNITS in planar1,   65393 runs,    143 skips

-cpuflags 0

19157 UNITS in planar1,   65512 runs,     24 skips

9.03632 speedup, 16be similarly.

Fate passes, each format tested with an image to video conversion.

Signed-off-by: Lauri Kasanen <cand@gmx.com>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
This commit is contained in:
Lauri Kasanen 2018-12-13 14:07:58 +02:00 committed by Michael Niedermayer
parent 5282db5929
commit b4c8c03b00
1 changed files with 59 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

59
libswscale/ppc/swscale_vsx.c
View File

@ -180,6 +180,60 @@ static void yuv2plane1_nbps_vsx(const int16_t *src, uint16_t *dest, int dstW,
yuv2plane1_nbps_u(src, dest, dstW, big_endian, output_bits, i);
}
#undef output_pixel
#define output_pixel(pos, val, bias, signedness) \
if (big_endian) { \
AV_WB16(pos, bias + av_clip_ ## signedness ## 16(val >> shift)); \
} else { \
AV_WL16(pos, bias + av_clip_ ## signedness ## 16(val >> shift)); \
}
static void yuv2plane1_16_u(const int32_t *src, uint16_t *dest, int dstW,
int big_endian, int output_bits, int start)
{
int i;
const int shift = 3;
for (i = start; i < dstW; i++) {
int val = src[i] + (1 << (shift - 1));
output_pixel(&dest[i], val, 0, uint);
}
}
static void yuv2plane1_16_vsx(const int32_t *src, uint16_t *dest, int dstW,
int big_endian, int output_bits)
{
const int dst_u = -(uintptr_t)dest & 7;
const int shift = 3;
const int add = (1 << (shift - 1));
const vector uint32_t vadd = (vector uint32_t) {add, add, add, add};
const vector uint16_t vswap = (vector uint16_t) vec_splat_u16(big_endian ? 8 : 0);
const vector uint32_t vshift = (vector uint32_t) vec_splat_u32(shift);
vector uint32_t v, v2;
vector uint16_t vd;
int i;
yuv2plane1_16_u(src, dest, dst_u, big_endian, output_bits, 0);
for (i = dst_u; i < dstW - 7; i += 8) {
v = vec_vsx_ld(0, (const uint32_t *) &src[i]);
v = vec_add(v, vadd);
v = vec_sr(v, vshift);
v2 = vec_vsx_ld(0, (const uint32_t *) &src[i + 4]);
v2 = vec_add(v2, vadd);
v2 = vec_sr(v2, vshift);
vd = vec_packsu(v, v2);
vd = vec_rl(vd, vswap);
vec_st(vd, 0, &dest[i]);
}
yuv2plane1_16_u(src, dest, dstW, big_endian, output_bits, i);
}
#define yuv2NBPS(bits, BE_LE, is_be, template_size, typeX_t) \
static void yuv2plane1_ ## bits ## BE_LE ## _vsx(const int16_t *src, \
uint8_t *dest, int dstW, \
@ -197,6 +251,8 @@ yuv2NBPS(12, BE, 1, nbps, int16_t)
yuv2NBPS(12, LE, 0, nbps, int16_t)
yuv2NBPS(14, BE, 1, nbps, int16_t)
yuv2NBPS(14, LE, 0, nbps, int16_t)
yuv2NBPS(16, BE, 1, 16, int32_t)
yuv2NBPS(16, LE, 0, 16, int32_t)
#endif /* !HAVE_BIGENDIAN */
@ -240,6 +296,9 @@ av_cold void ff_sws_init_swscale_vsx(SwsContext *c)
case 14:
c->yuv2plane1 = isBE(dstFormat) ? yuv2plane1_14BE_vsx : yuv2plane1_14LE_vsx;
break;
case 16:
c->yuv2plane1 = isBE(dstFormat) ? yuv2plane1_16BE_vsx : yuv2plane1_16LE_vsx;
break;
#endif
}
}