Am Sonntag 11 März 2007 03:08 schrieb Erich Hoover:
> I ran all of the d3d9 tests and did not get any errors with my patch
> applied.  I did get an error when I ran your modified visual.c test, but I
> get the same errors with and without my patch.
Yes; The existing d3d9 tests do not check what happens when the surface is 
actually modified, and they do not draw an actual cursor, because the device 
test just checks return values and not the actual drawing. But the test shows 
that the surface can be destroyed without modifying the cursor.

I can try to extend the test cases for modifying the surface if you want to?

Attachment: pgp8kx8TXmMWN.pgp
Description: PGP signature



Reply via email to