Re: [1/3] d3dx9: Define DDS structures.

[Stefan Dösinger]

Am Donnerstag, 26. April 2012, 11:47:51 schrieb Józef Kucia:
> +    DWORD pitch_or_linear_size;
> Are you aware that lPitch is a (signed) LONG in the original DDSURFACEDESC 
> structure? Honestly I have no idea why, and I have never seen an app that 
> passes a negative pitch, but it might lead to obscure bugs if it is not 
> checked.

[1/3] d3dx9: Define DDS structures.

[Józef Kucia]

On Thu, Apr 26, 2012 at 9:50 PM, Stefan Dösinger <stefandoesinger(a)gmx.at> wrote:
> Am Donnerstag, 26. April 2012, 11:47:51 schrieb Józef Kucia:
> > +    DWORD pitch_or_linear_size;
> > Are you aware that lPitch is a (signed) LONG in the original DDSURFACEDESC
> > structure? Honestly I have no idea why, and I have never seen an app that
> > passes a negative pitch, but it might lead to obscure bugs if it is not
> > checked.

MSDN says that dwPitchOrLinearSize is DWORD in the DDS_HEADER. Also,
the same definition is used in a D3D sample. On the other hand, we can
define it as LONG to be safe.

[1/3] d3dx9: Define DDS structures.

[Stefan Dösinger]

Am 26.04.2012 22:22 schrieb "Józef Kucia" <joseph.kucia(a)gmail.com>:
> MSDN says that dwPitchOrLinearSize is DWORD in the DDS_HEADER. Also,
> the same definition is used in a D3D sample. On the other hand, we can
> define it as LONG to be safe.
> Nah, no need in that case. But I recommend to write tests for invalid
> pitches