Re: user32: Add test_CreateIconFromResourceEx().

[Dmitry Timoshkov]

"Andrew Riedi"  wrote:
> changelog: user32: Add test_CreateIconFromResourceEx().

What this patch is supposed to test? Just that CreateIconFromResourceEx
returns a valid icon handle? What about the resulting icon contents?
What about cases when CreateIconFromResourceEx should fail?

user32: Add test_CreateIconFromResourceEx().

[Andrew Riedi]

On Sun, Mar 16, 2008 at 10:05 PM, Dmitry Timoshkov
 wrote:
> "Andrew Riedi"  wrote:
> > changelog: user32: Add test_CreateIconFromResourceEx().
>
> What this patch is supposed to test? Just that CreateIconFromResourceEx
>  returns a valid icon handle? What about the resulting icon contents?
>  What about cases when CreateIconFromResourceEx should fail?
>
> --
>  Dmitry.

Mostly, this patch is just to serve as a check to make sure I don't
horribly screw up CreateIconFromResourceEx().  I need to abstract some
of the functionality out, and make use of it elsewhere.  Right now our
CreateIconFromResourceEx() works flawlessly, as far as I know.
Checking for contents afterwards would be a good idea though - I will
send an updated patch to do exactly that before I touch our
CreateIconFromResourceEx().

As always, thanks for looking over my code, Dmitry.  :)

user32: Add test_CreateIconFromResourceEx().

[Dmitry Timoshkov]

"Andrew Riedi"  wrote:
> Mostly, this patch is just to serve as a check to make sure I don't
> horribly screw up CreateIconFromResourceEx().  I need to abstract some
> of the functionality out, and make use of it elsewhere.  Right now our
> CreateIconFromResourceEx() works flawlessly, as far as I know.
> Checking for contents afterwards would be a good idea though - I will
> send an updated patch to do exactly that before I touch our
> CreateIconFromResourceEx().

I'd suggest to add the tests not only for various parameters passed
to CreateIconFromResourceEx (not just a random one), but also add to
the resources icons/images with different color depth and test how they
are loaded (for instance the bug 10854 mentions an app which has 24-bpp
bitmap or icon in its resources, and that leads to the problem later
on when the app usess the returned handle to draw an image).