mirror of
https://github.com/FFmpeg/FFmpeg.git
synced 2025-01-13 21:28:01 +02:00
Merge remote-tracking branch 'qatar/master'
* qatar/master: altivec: perform an explicit unaligned load Merged-by: Michael Niedermayer <michaelni@gmx.at>
This commit is contained in:
commit
d7ed473d5c
@ -84,14 +84,12 @@ static int32_t scalarproduct_int16_altivec(const int16_t *v1, const int16_t *v2,
|
|||||||
{
|
{
|
||||||
int i;
|
int i;
|
||||||
LOAD_ZERO;
|
LOAD_ZERO;
|
||||||
const vec_s16 *pv;
|
|
||||||
register vec_s16 vec1;
|
register vec_s16 vec1;
|
||||||
register vec_s32 res = vec_splat_s32(0), t;
|
register vec_s32 res = vec_splat_s32(0), t;
|
||||||
int32_t ires;
|
int32_t ires;
|
||||||
|
|
||||||
for(i = 0; i < order; i += 8){
|
for(i = 0; i < order; i += 8){
|
||||||
pv = (const vec_s16*)v1;
|
vec1 = vec_unaligned_load(v1);
|
||||||
vec1 = vec_perm(pv[0], pv[1], vec_lvsl(0, v1));
|
|
||||||
t = vec_msum(vec1, vec_ld(0, v2), zero_s32v);
|
t = vec_msum(vec1, vec_ld(0, v2), zero_s32v);
|
||||||
res = vec_sums(t, res);
|
res = vec_sums(t, res);
|
||||||
v1 += 8;
|
v1 += 8;
|
||||||
|
Loading…
Reference in New Issue
Block a user