[d3d8] question about "bogus" code

[Christoph Frick]

hiho

as with newer versions more and more games i try no longer work (which
worked with wine around 0.9.0) i try to find the reasons. i stumbled
over this code in dlls/d3d8/surface.c:

date: 2006-02-14 17:05:35 +0000;  author: julliard;  state: Exp;  lines: +202 -319
Roderick Colenbrander [email protected]
d3d8: Move d3d8 surface over to wined3d (based on the work of Oliver Stieber).

149 1.48         (julliard 14-Feb-06):     if (This == D3D8_SURFACE_GET_DEVICE(This)->backBuffer || This == D3D8_SURFACE_GET_DEVICE(This)->renderTarget || This == D3D8_SURFACE_GET_DEVICE(This)->frontBuffer || D3D8_SURFACE_GET_DEVICE(This)->depthStencilBuffer) {
150 1.48         (julliard 14-Feb-06):       if (This == D3D8_SURFACE_GET_DEVICE(This)->backBuffer) {
151 1.48         (julliard 14-Feb-06):      TRACE("(%p, backBuffer) : rect@%p flags(%08lx), output lockedRect@%p, memory@%p\n", This, pRect, Flags, pLockedRect, D3D8_SURFACE(This)->resource.allocatedMemory);
152 1.48         (julliard 14-Feb-06):       } else if (This == D3D8_SURFACE_GET_DEVICE(This)->frontBuffer) {
153 1.48         (julliard 14-Feb-06):      TRACE("(%p, frontBuffer) : rect@%p flags(%08lx), output lockedRect@%p, memory@%p\n", This, pRect, Flags, pLockedRect, D3D8_SURFACE(This)->resource.allocatedMemory);
154 1.48         (julliard 14-Feb-06):       } else if (This == D3D8_SURFACE_GET_DEVICE(This)->renderTarget) {
155 1.48         (julliard 14-Feb-06):      TRACE("(%p, renderTarget) : rect@%p flags(%08lx), output lockedRect@%p, memory@%p\n", This, pRect, Flags, pLockedRect, D3D8_SURFACE(This)->resource.allocatedMemory);
156 1.48         (julliard 14-Feb-06):       } else if (This == D3D8_SURFACE_GET_DEVICE(This)->depthStencilBuffer) {
157 1.48         (julliard 14-Feb-06):      TRACE("(%p, stencilBuffer) : rect@%p flags(%08lx), output lockedRect@%p, memory@%p\n", This, pRect, Flags, pLockedRect, D3D8_SURFACE(This)->resource.allocatedMemory);
158 1.9          (julliard 17-May-03):       }
159 1.9          (julliard 17-May-03):     } else {
160 1.48         (julliard 14-Feb-06):       TRACE("(%p) : rect@%p flags(%08lx), output lockedRect@%p, memory@%p\n", This, pRect, Flags, pLockedRect, D3D8_SURFACE(This)->resource.allocatedMemory);
161 1.9          (julliard 17-May-03):     }

why is in line 149 for the depthStencilBuffer check not comparing against This
like in three checks before? if there is a reason, if would be good to have a
comment there - otherwise it looks like an error.

[H. Verbeet]

On 15/02/06, Christoph Frick [email protected] wrote:
> as with newer versions more and more games i try no longer work (which
> worked with wine around 0.9.0)

In general, are those d3d8 or d3d9 games?

[Christoph Frick]

On Wed, Feb 15, 2006 at 12:42:51PM +0100, H. Verbeet wrote:
> > as with newer versions more and more games i try no longer work (which
> > worked with wine around 0.9.0)
>
> In general, are those d3d8 or d3d9 games?

well for the tests i did i have problems with out-of-memory in games
programmed in java (il2 sturmovik + pacfic fighters), problems with GLX
(BadMatch) with pacfic fighters and nascar racing 2003 and a segfault
related to my mail i sent there with life for speed,which is actually
d3d8 (although i though it would use d3d9). the only three "games" that
use d3d9 on my disk are ankh and richard burns ralley (which never
worked from the first day) and the ogre 3d engine demos (many demos
worked with wine a few days ago).

long story short - i have the general impression that lots of stuff that
worked with earlier wine versions no longer works for me with different
reasons and theses are not all d3d related (as i dont use d3d if there
is an ogl alterantive).

[Roderick Colenbrander]

> 149 1.48         (julliard 14-Feb-06):     if (This ==
> D3D8_SURFACE_GET_DEVICE(This)->backBuffer || This ==
> D3D8_SURFACE_GET_DEVICE(This)->renderTarget || This ==
> D3D8_SURFACE_GET_DEVICE(This)->frontBuffer ||
> D3D8_SURFACE_GET_DEVICE(This)->depthStencilBuffer) { 150 1.48        
> why is in line 149 for the depthStencilBuffer check not comparing against
> This like in three checks before? if there is a reason, if would be good to
> have a comment there - otherwise it looks like an error.

The patch for the biggest part makes 'old' code working using a new wined3d 
based d3d8 surface. In order to make the old code reasable I added those 
D3D8_SURFACE / D3D8_SURFACE_GET_DEVICE macros to get access to data hidden in 
the wined3d surface which is wrapped inside a d3d8 surface. The LockRect code 
to which you are refering is just the old surface code moved over to the new 
surface structure using the macros I mentioned. I didn't change much in the 
old code except for macro changes and sometimes a little more. I have checked 
the old d3d8 code and the line looks the same. I think the missing is a bug 
but I don't know for sure. Note that code will be moved to wined3d soon.

Roderick