Added differentiation between Non power of two textures when mip mapped vs
when not mipped mapped to get round the issue of incomplete support under ATI cards.
This commit is contained in:
@@ -189,8 +189,11 @@ void Texture1D::applyTexImage1D(GLenum target, Image* image, State& state, GLsiz
|
||||
bool compressed = isCompressedInternalFormat(_internalFormat);
|
||||
|
||||
//Rescale if resize hint is set or NPOT not supported or dimension exceeds max size
|
||||
if( _resizeNonPowerOfTwoHint || !extensions->isNonPowerOfTwoTextureSupported() || inwidth > extensions->maxTextureSize() )
|
||||
if( _resizeNonPowerOfTwoHint || !extensions->isNonPowerOfTwoTextureSupported(_min_filter) || inwidth > extensions->maxTextureSize() )
|
||||
{
|
||||
// this is not thread safe... should really create local image data and rescale to that as per Texture2D.
|
||||
image->ensureValidSizeForTexturing(extensions->maxTextureSize());
|
||||
}
|
||||
|
||||
glPixelStorei(GL_UNPACK_ALIGNMENT,image->getPacking());
|
||||
|
||||
|
||||
Reference in New Issue
Block a user