mirror of
https://github.com/mpv-player/mpv
synced 2025-01-19 05:41:16 +00:00
Remove the query_format_cache, it is extra complexity that makes improvements
difficult while the speed advantage in normal use cases is not measurable. git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@30408 b3059339-0415-0410-9bf9-f77b7e298cf2
This commit is contained in:
parent
f1714582b2
commit
86f1430a59
@ -30,7 +30,6 @@ static struct vf_priv_s {
|
||||
int interlaced;
|
||||
int noup;
|
||||
int accurate_rnd;
|
||||
int query_format_cache[64];
|
||||
} const vf_priv_dflt = {
|
||||
-1,-1,
|
||||
0,
|
||||
@ -102,11 +101,7 @@ static unsigned int find_best_out(vf_instance_t *vf){
|
||||
// find the best outfmt:
|
||||
for(i=0; outfmt_list[i]; i++){
|
||||
const int format= outfmt_list[i];
|
||||
int ret= vf->priv->query_format_cache[i]-1;
|
||||
if(ret == -1){
|
||||
ret= vf_next_query_format(vf, outfmt_list[i]);
|
||||
vf->priv->query_format_cache[i]= ret+1;
|
||||
}
|
||||
int ret = vf_next_query_format(vf, format);
|
||||
|
||||
mp_msg(MSGT_VFILTER,MSGL_DBG2,"scale: query(%s) -> %d\n",vo_format_name(format),ret&3);
|
||||
if(ret&VFCAP_CSP_SUPPORTED_BY_HW){
|
||||
|
Loading…
Reference in New Issue
Block a user