simpler GPU-compositing. will make it easier to grab neighbor 2D/3D pixels individually for filtering.

This commit is contained in:
Arisotura
2019-05-25 19:36:47 +02:00
parent 8c93a45574
commit 70a3243714
6 changed files with 266 additions and 287 deletions

View File

@ -216,7 +216,7 @@ bool GLScreen_Init()
glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_WRAP_T, GL_CLAMP_TO_EDGE);
glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_MIN_FILTER, GL_NEAREST);
glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_MAG_FILTER, GL_NEAREST);
glTexImage2D(GL_TEXTURE_2D, 0, GL_RGBA8UI, 1024, 1536, 0, GL_RGBA_INTEGER, GL_UNSIGNED_BYTE, NULL);
glTexImage2D(GL_TEXTURE_2D, 0, GL_RGBA8UI, 256*3 + 1, 192*2, 0, GL_RGBA_INTEGER, GL_UNSIGNED_BYTE, NULL);
GL_ScreenSizeDirty = true;
@ -317,31 +317,31 @@ void GLScreen_DrawScreen()
switch (ScreenRotation)
{
case 0:
s0 = 0; t0 = 768;
s1 = scwidth; t1 = 768;
s2 = 0; t2 = 768+scheight;
s3 = scwidth; t3 = 768+scheight;
s0 = 0; t0 = 192;
s1 = scwidth; t1 = 192;
s2 = 0; t2 = 192+scheight;
s3 = scwidth; t3 = 192+scheight;
break;
case 1:
s0 = 0; t0 = 768+scheight;
s1 = 0; t1 = 768;
s2 = scwidth; t2 = 768+scheight;
s3 = scwidth; t3 = 768;
s0 = 0; t0 = 192+scheight;
s1 = 0; t1 = 192;
s2 = scwidth; t2 = 192+scheight;
s3 = scwidth; t3 = 192;
break;
case 2:
s0 = scwidth; t0 = 768+scheight;
s1 = 0; t1 = 768+scheight;
s2 = scwidth; t2 = 768;
s3 = 0; t3 = 768;
s0 = scwidth; t0 = 192+scheight;
s1 = 0; t1 = 192+scheight;
s2 = scwidth; t2 = 192;
s3 = 0; t3 = 192;
break;
case 3:
s0 = scwidth; t0 = 768;
s1 = scwidth; t1 = 768+scheight;
s2 = 0; t2 = 768;
s3 = 0; t3 = 768+scheight;
s0 = scwidth; t0 = 192;
s1 = scwidth; t1 = 192+scheight;
s2 = 0; t2 = 192;
s3 = 0; t3 = 192+scheight;
break;
}
@ -381,14 +381,14 @@ void GLScreen_DrawScreen()
{
glTexSubImage2D(GL_TEXTURE_2D, 0, 0, 0, 256, 192, GL_RGBA_INTEGER,
GL_UNSIGNED_BYTE, GPU::Framebuffer[frontbuf][0]);
glTexSubImage2D(GL_TEXTURE_2D, 0, 0, 768, 256, 192, GL_RGBA_INTEGER,
glTexSubImage2D(GL_TEXTURE_2D, 0, 0, 192, 256, 192, GL_RGBA_INTEGER,
GL_UNSIGNED_BYTE, GPU::Framebuffer[frontbuf][1]);
}
else
{
glTexSubImage2D(GL_TEXTURE_2D, 0, 0, 0, 256*3 + 2, 192, GL_RGBA_INTEGER,
glTexSubImage2D(GL_TEXTURE_2D, 0, 0, 0, 256*3 + 1, 192, GL_RGBA_INTEGER,
GL_UNSIGNED_BYTE, GPU::Framebuffer[frontbuf][0]);
glTexSubImage2D(GL_TEXTURE_2D, 0, 0, 768, 256*3 + 2, 192, GL_RGBA_INTEGER,
glTexSubImage2D(GL_TEXTURE_2D, 0, 0, 192, 256*3 + 1, 192, GL_RGBA_INTEGER,
GL_UNSIGNED_BYTE, GPU::Framebuffer[frontbuf][1]);
}
}
@ -2241,11 +2241,13 @@ void DestroyMainWindow()
void RecreateMainWindow(bool opengl)
{
int winX, winY;
int winX, winY, maxi;
uiWindowPosition(MainWindow, &winX, &winY);
maxi = uiWindowMaximized(MainWindow);
DestroyMainWindow();
CreateMainWindow(opengl);
uiWindowSetPosition(MainWindow, winX, winY);
uiWindowSetMaximized(MainWindow, maxi);
}

View File

@ -66,121 +66,81 @@ void main()
{
ivec4 pixel = ivec4(texelFetch(ScreenTex, ivec2(fTexcoord), 0));
// bit0-13: BLDCNT
// bit14-15: DISPCNT display mode
// bit16-20: EVA
// bit21-25: EVB
// bit26-30: EVY
ivec4 ctl = ivec4(texelFetch(ScreenTex, ivec2(256*3, int(fTexcoord.y)), 0));
ivec4 mbright = ivec4(texelFetch(ScreenTex, ivec2(256*3 + 1, int(fTexcoord.y)), 0));
int dispmode = (ctl.g >> 6) & 0x3;
ivec4 mbright = ivec4(texelFetch(ScreenTex, ivec2(256*3, int(fTexcoord.y)), 0));
int dispmode = mbright.b & 0x3;
if (dispmode == 1)
{
int eva = ctl.b & 0x1F;
int evb = (ctl.b >> 5) | ((ctl.a & 0x03) << 3);
int evy = ctl.a >> 2;
ivec4 val1 = pixel;
ivec4 val2 = ivec4(texelFetch(ScreenTex, ivec2(fTexcoord) + ivec2(256,0), 0));
ivec4 val3 = ivec4(texelFetch(ScreenTex, ivec2(fTexcoord) + ivec2(512,0), 0));
ivec4 top = pixel;
ivec4 mid = ivec4(texelFetch(ScreenTex, ivec2(fTexcoord) + ivec2(256,0), 0));
ivec4 bot = ivec4(texelFetch(ScreenTex, ivec2(fTexcoord) + ivec2(512,0), 0));
int compmode = val3.a & 0xF;
int eva, evb, evy;
int winmask = top.b >> 7;
if ((top.a & 0xC0) == 0x40)
if (compmode == 4)
{
float xpos = top.r + fract(fTexcoord.x);
float ypos = mod(fTexcoord.y, 768);
// 3D on top, blending
float xpos = val3.r + fract(fTexcoord.x);
float ypos = mod(fTexcoord.y, 192);
ivec4 _3dpix = ivec4(texelFetch(_3DTex, ivec2(vec2(xpos, ypos)*u3DScale), 0).bgra
* vec4(63,63,63,31));
if (_3dpix.a > 0) { top = _3dpix; top.a |= 0x40; bot = mid; }
else top = mid;
if (_3dpix.a > 0)
{
eva = (_3dpix.a & 0x1F) + 1;
evb = 32 - eva;
val1 = ((_3dpix * eva) + (val1 * evb)) >> 5;
if (eva <= 16) val1 += ivec4(1,1,1,0);
val1 = min(val1, 0x3F);
}
else
val1 = val2;
}
else if ((mid.a & 0xC0) == 0x40)
else if (compmode == 1)
{
float xpos = mid.r + fract(fTexcoord.x);
float ypos = mod(fTexcoord.y, 768);
// 3D on bottom, blending
float xpos = val3.r + fract(fTexcoord.x);
float ypos = mod(fTexcoord.y, 192);
ivec4 _3dpix = ivec4(texelFetch(_3DTex, ivec2(vec2(xpos, ypos)*u3DScale), 0).bgra
* vec4(63,63,63,31));
if (_3dpix.a > 0) { bot = _3dpix; bot.a |= 0x40; }
}
else
{
// conditional texture fetch no good for performance, apparently
//texelFetch(_3DTex, ivec2(0, fTexcoord.y*2), 0);
bot = mid;
}
top.b &= 0x3F;
bot.b &= 0x3F;
int target2;
if ((bot.a & 0x80) != 0) target2 = 0x10;
else if ((bot.a & 0x40) != 0) target2 = 0x01;
else target2 = bot.a;
bool t2pass = ((ctl.g & target2) != 0);
int coloreffect = 0;
if ((top.a & 0x80) != 0 && t2pass)
{
// sprite blending
coloreffect = 1;
if ((top.a & 0x40) != 0)
if (_3dpix.a > 0)
{
eva = top.a & 0x1F;
evb = 16 - eva;
eva = val3.g;
evb = val3.b;
val1 = ((val1 * eva) + (_3dpix * evb)) >> 4;
val1 = min(val1, 0x3F);
}
else
val1 = val2;
}
else if ((top.a & 0x40) != 0 && t2pass)
else if (compmode <= 3)
{
// 3D layer blending
// 3D on top, normal/fade
coloreffect = 4;
eva = (top.a & 0x1F) + 1;
evb = 32 - eva;
}
else
{
if ((top.a & 0x80) != 0) top.a = 0x10;
else if ((top.a & 0x40) != 0) top.a = 0x01;
float xpos = val3.r + fract(fTexcoord.x);
float ypos = mod(fTexcoord.y, 192);
ivec4 _3dpix = ivec4(texelFetch(_3DTex, ivec2(vec2(xpos, ypos)*u3DScale), 0).bgra
* vec4(63,63,63,31));
if ((ctl.r & top.a) != 0 && winmask != 0)
if (_3dpix.a > 0)
{
int effect = ctl.r >> 6;
if ((effect != 1) || t2pass) coloreffect = effect;
evy = val3.g;
val1 = _3dpix;
if (compmode == 2) val1 += ((ivec4(0x3F,0x3F,0x3F,0) - val1) * evy) >> 4;
else if (compmode == 3) val1 -= (val1 * evy) >> 4;
}
else
val1 = val2;
}
if (coloreffect == 0)
{
pixel = top;
}
else if (coloreffect == 1)
{
pixel = ((top * eva) + (bot * evb)) >> 4;
pixel = min(pixel, 0x3F);
}
else if (coloreffect == 2)
{
pixel = top;
pixel += ((ivec4(0x3F,0x3F,0x3F,0) - pixel) * evy) >> 4;
}
else if (coloreffect == 3)
{
pixel = top;
pixel -= (pixel * evy) >> 4;
}
else
{
pixel = ((top * eva) + (bot * evb)) >> 5;
if (eva <= 16) pixel += ivec4(1,1,1,0);
pixel = min(pixel, 0x3F);
}
pixel = val1;
}
if (dispmode != 0)