RepoMirrors
/
mpv
mirror of https://github.com/mpv-player/mpv
1
0
Fork 0
Code Issues Releases Wiki Activity

vo_opengl: fix bicubic_fast in ES mode 

GLES shaders disallow implicit conversion from int to float.

This has been broken for quite a while.
This commit is contained in:
wm4 2016-05-16 12:10:47 +02:00
parent e047cc0931
commit 51603e472d
1 changed files with 1 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
video/out/opengl/video_shaders.c
View File

@ -177,7 +177,7 @@ static void bicubic_calcweights(struct gl_shader_cache *sc, const char *t, const
GLSLF("%s = %s * %s + vec4(0, 0, -0.5, 0.5);\n", t, t, s);
GLSLF("%s = %s * %s + vec4(-0.6666, 0, 0.8333, 0.1666);\n", t, t, s);
GLSLF("%s.xy *= vec2(1, 1) / vec2(%s.z, %s.w);\n", t, t, t);
GLSLF("%s.xy += vec2(1 + %s, 1 - %s);\n", t, s, s);
GLSLF("%s.xy += vec2(1.0 + %s, 1.0 - %s);\n", t, s, s);
}
void pass_sample_bicubic_fast(struct gl_shader_cache *sc)