ShuriZma
/
stella
mirror of https://github.com/stella-emu/stella.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

It seems that GL_BGRA is still the fastest texture format *if* it is available. 

In the process of converting the code to be OpenGL ES compliant, we shouldn't
hurt the performance on systems with more advanced versions.


git-svn-id: svn://svn.code.sf.net/p/stella/code/trunk@2272 8b62c5a3-ac7e-4cc8-8f21-d9a121418aba
This commit is contained in:
stephena 2011-08-20 15:24:35 +00:00
parent 1ab356afe8
commit b45d65fc31
1 changed files with 28 additions and 5 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

33
src/common/FrameBufferGL.cxx
View File

@ -82,7 +82,11 @@ FrameBufferGL::FrameBufferGL(OSystem* osystem)
// since the structure may be needed before any FBSurface's have // since the structure may be needed before any FBSurface's have
// been created // been created
SDL_Surface* s = SDL_CreateRGBSurface(SDL_SWSURFACE, 1, 1, 16, SDL_Surface* s = SDL_CreateRGBSurface(SDL_SWSURFACE, 1, 1, 16,
#if defined(GL_BGRA) && defined(GL_UNSIGNED_SHORT_1_5_5_5_REV)
0x00007c00, 0x000003e0, 0x0000001f, 0x00000000);
#else
0x0000f800, 0x000007c0, 0x0000003e, 0x00000000); 0x0000f800, 0x000007c0, 0x0000003e, 0x00000000);
#endif
myPixelFormat = *(s->format); myPixelFormat = *(s->format);
SDL_FreeSurface(s); SDL_FreeSurface(s);
} }
@ -218,7 +222,12 @@ string FrameBufferGL::about() const
<< " Renderer: " << p_glGetString(GL_RENDERER) << endl << " Renderer: " << p_glGetString(GL_RENDERER) << endl
<< " Version: " << p_glGetString(GL_VERSION) << endl << " Version: " << p_glGetString(GL_VERSION) << endl
<< " Color: " << myDepth << " bit, " << myRGB[0] << "-" << " Color: " << myDepth << " bit, " << myRGB[0] << "-"
<< myRGB[1] << "-" << myRGB[2] << "-" << myRGB[3] << endl << myRGB[1] << "-" << myRGB[2] << "-" << myRGB[3] << ", "
#if defined(GL_BGRA) && defined(GL_UNSIGNED_SHORT_1_5_5_5_REV)
<< "GL_BGRA" << endl
#else
<< "GL_RGBA" << endl
#endif
<< " Filter: " << myFilterParamName << endl << " Filter: " << myFilterParamName << endl
<< " Extensions: "; << " Extensions: ";
if(myVBOAvailable) out << "VBO "; if(myVBOAvailable) out << "VBO ";
@ -531,11 +540,15 @@ FBSurfaceGL::FBSurfaceGL(FrameBufferGL& buffer,
myTexCoordW = (GLfloat) baseWidth / myTexWidth; myTexCoordW = (GLfloat) baseWidth / myTexWidth;
myTexCoordH = (GLfloat) baseHeight / myTexHeight; myTexCoordH = (GLfloat) baseHeight / myTexHeight;
// Based on experimentation, the following is the fastest 16-bit // Based on experimentation, the following are the fastest 16-bit
// format for OpenGL (on all platforms) // formats for OpenGL (on all platforms)
myTexture = SDL_CreateRGBSurface(SDL_SWSURFACE, myTexture = SDL_CreateRGBSurface(SDL_SWSURFACE,
myTexWidth, myTexHeight, 16, myTexWidth, myTexHeight, 16,
#if defined(GL_BGRA) && defined(GL_UNSIGNED_SHORT_1_5_5_5_REV)
0x00007c00, 0x000003e0, 0x0000001f, 0x00000000);
#else
0x0000f800, 0x000007c0, 0x0000003e, 0x00000000); 0x0000f800, 0x000007c0, 0x0000003e, 0x00000000);
#endif
myPitch = myTexture->pitch >> 1; myPitch = myTexture->pitch >> 1;
// Associate the SDL surface with a GL texture object // Associate the SDL surface with a GL texture object
@ -746,7 +759,12 @@ void FBSurfaceGL::update()
p_glActiveTexture(GL_TEXTURE0); p_glActiveTexture(GL_TEXTURE0);
p_glBindTexture(GL_TEXTURE_2D, myTexID); p_glBindTexture(GL_TEXTURE_2D, myTexID);
p_glTexSubImage2D(GL_TEXTURE_2D, 0, 0, 0, myTexWidth, myTexHeight, p_glTexSubImage2D(GL_TEXTURE_2D, 0, 0, 0, myTexWidth, myTexHeight,
GL_RGBA, GL_UNSIGNED_SHORT_5_5_5_1, myTexture->pixels); #if defined(GL_BGRA) && defined(GL_UNSIGNED_SHORT_1_5_5_5_REV)
GL_BGRA, GL_UNSIGNED_SHORT_1_5_5_5_REV,
#else
GL_RGBA, GL_UNSIGNED_SHORT_5_5_5_1,
#endif
myTexture->pixels);
p_glEnableClientState(GL_VERTEX_ARRAY); p_glEnableClientState(GL_VERTEX_ARRAY);
p_glEnableClientState(GL_TEXTURE_COORD_ARRAY); p_glEnableClientState(GL_TEXTURE_COORD_ARRAY);
@ -805,7 +823,12 @@ void FBSurfaceGL::reload()
// Finally, create the texture in the most optimal format // Finally, create the texture in the most optimal format
p_glTexImage2D(GL_TEXTURE_2D, 0, GL_RGBA, p_glTexImage2D(GL_TEXTURE_2D, 0, GL_RGBA,
myTexWidth, myTexHeight, 0, myTexWidth, myTexHeight, 0,
GL_RGBA, GL_UNSIGNED_SHORT_5_5_5_1, myTexture->pixels); #if defined(GL_BGRA) && defined(GL_UNSIGNED_SHORT_1_5_5_5_REV)
GL_BGRA, GL_UNSIGNED_SHORT_1_5_5_5_REV,
#else
GL_RGBA, GL_UNSIGNED_SHORT_5_5_5_1,
#endif
myTexture->pixels);
// Cache vertex and texture coordinates using vertex buffer object // Cache vertex and texture coordinates using vertex buffer object
if(myFB.myVBOAvailable) if(myFB.myVBOAvailable)