Commen=
t # 13
on bug 97305<=
/a>
from Matias N. Goldberg
I've added a "Proof of the bug".
There IS a bug. I don't know if it's software or in the hardware. The attac=
hed
sample does the following:
1. Create a 1MB buffer.
2. Fill the first 4kb with the pattern:
data[i+0] =3D 1.0f;
data[i+1] =3D 0.5f;
data[i+2] =3D 0.15f;
data[i+3] =3D 0.0f;
3. Attach it via:
glActiveTexture( GL_TEXTURE0 );
glBindTexture( GL_TEXTURE_BUFFER, texName );
glTexBufferRange( GL_TEXTURE_BUFFER, GL_RGBA32F, vboName,
elementStart*sizeof(float),
1024*4*sizeof(float) );
++elementStart;
4. Render a triangle showing the colours.
If the alignment is not a problem, I should see Orange-Cream, then Carmine,
then blue-violetish, then green; repeat.
Instead I **ALWAYS** see Orange-Cream; which is clearly incorrect.
After running the sample long enough, elementStart will be > 1024 (thus =
1024*4
exceeding 4kb) and begins to display black. (and if you leave it long enough
running it could crash or throw an error, that's not the point)