Mirrors/yuzu
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

Shaders: Fix multiplications between 0.0 and inf

Browse source 
The PICA200 semantics for multiplication are so that when multiplying
inf by exactly 0.0, the result is 0.0, instead of NaN, as defined by
IEEE. This is relied upon by games.

Fixes #1024 (missing OoT interface items)
This commit is contained in:
Yuri Kunde Schlesner 2015-08-24 01:48:15 -03:00
parent 082b74fa24
commit 630a850d4d
3 changed files with 60 additions and 42 deletions
Download patch file Download diff file Expand all files Collapse all files

14
src/video_core/pica.h
View file

@ -1021,12 +1021,20 @@ struct float24 {
return ret;
}
static float24 Zero() {
return FromFloat32(0.f);
}
// Not recommended for anything but logging
float ToFloat32() const {
return value;
}
float24 operator * (const float24& flt) const {
if ((this->value == 0.f && flt.value == flt.value) ||
(flt.value == 0.f && this->value == this->value))
// PICA gives 0 instead of NaN when multiplying by inf
return Zero();
return float24::FromFloat32(ToFloat32() * flt.ToFloat32());
}
@ -1043,7 +1051,11 @@ struct float24 {
}
float24& operator *= (const float24& flt) {
value *= flt.ToFloat32();
if ((this->value == 0.f && flt.value == flt.value) ||
(flt.value == 0.f && this->value == this->value))
// PICA gives 0 instead of NaN when multiplying by inf
*this = Zero();
else value *= flt.ToFloat32();
return *this;
}