Re: [PATCH 6/8] d3d10core/tests: Add test for ClearDepthStencilView().

[Henri Verbeet]

On 14 March 2016 at 00:03, Józef Kucia  wrote:
> +static float get_readback_float(struct texture_readback *rb, unsigned int x, unsigned int y)
> +{
> - return rb->texture
> -
>        ? ((float *)rb->mapped_texture.pData)[rb->mapped_texture.RowPitch * y / sizeof(float) + x] : NAN;
>
> +}

Can the texture ever legitimately be NULL?

[PATCH 6/8] d3d10core/tests: Add test for ClearDepthStencilView().

[Józef Kucia]

On Mon, Mar 14, 2016 at 1:41 PM, Henri Verbeet  wrote:
> On 14 March 2016 at 00:03, Józef Kucia  wrote:
> > +static float get_readback_float(struct texture_readback *rb, unsigned int x, unsigned int y)
> > +{
> > - return rb->texture
> > -
> >        ? ((float *)rb->mapped_texture.pData)[rb->mapped_texture.RowPitch * y / sizeof(float) + x] : NAN;
> >
> > +}
>
> Can the texture ever legitimately be NULL?

It can be NULL only when get_texture_readback() fails, in which case a
test will fail anyway. It's inspired by surface_readback from d3d9
tests (which fails gracefully). So yeah, I should drop NULL check and
don't introduce NAN unnecessarily.