Hi, > >>> + glBegin(GL_QUADS); > >>> + glTexCoord2f(0, 1); glVertex3f(-1, -1, 0); > >>> + glTexCoord2f(0, 0); glVertex3f(-1, 1, 0); > >>> + glTexCoord2f(1, 0); glVertex3f(1, 1, 0); > >>> + glTexCoord2f(1, 1); glVertex3f(1, -1, 0); > >>> + glEnd(); > >> I've been trained to hate direct mode, but it should be fine for just > >> this quad. > > --verbose please. Guess for longer sequences it would be much more > > efficient to compile this into a shader program? > > Well, again, I'm used to OpenGL 3/4 Core now which doesn't have the > immediate mode any more. [ ... ]
> [ ... ] and maybe for some > reason there are people which want to use qemu with OpenGL acceleration > on a pre OpenGL 2 machine. For virtio-gpu we'll need OPENGL 3 anyway, so I don't feel like caring too much about old opengl versions. How would the opengl 3/4 version of the above would look like? > >> Using glTexSubImage2D() would give us the advantage of being able to > >> perform partial updates on the texture; but it seems to fit pretty bad > >> into the existing code. To make it fit, I'd call glTexSubImage2D() > >> directly in sdl2_gl_update() and just draw the quad here. > > Yes, that should work. Done, also factoring this into helper functions so gtk can use this too some day, new series sent out, please have a look. thanks, Gerd
