Re: [PATCH 1/5] ddraw: Move clipper handling to ddraw.

2011-12-13 Thread Stefan Dösinger
Am Montag, 12. Dezember 2011, 20:53:55 schrieb Henri Verbeet:
 --- a/dlls/wined3d/clipper.c
 +++ /dev/null
Assuming that msdn[1] is correct, implies that we'll have to move overlays to 
ddraw as well. I'm not opposed to that, just mentioning it. In fact I'm 
actually inclined to move the overlay blits to ddraw, but I'm not sure yet 
because they have some similarities to d3d8/9 software cursors.

If we move overlays to ddraw this also means the ddraw shadow frontbuffer is 
here to stay.

1: http://msdn.microsoft.com/en-
us/library/windows/desktop/gg426213(v=VS.85).aspx

signature.asc
Description: This is a digitally signed message part.



Re: [PATCH 2/5] ddraw: Add support for clipped blits.

2011-12-13 Thread Stefan Dösinger
Am Montag, 12. Dezember 2011, 20:53:56 schrieb Henri Verbeet:
 +if (!dst_surface-clipper)
I think some tests would be a good idea, e.g. if clippers work on offscreen 
surfaces or backbuffers, or what happens when a D3D render target has a clipper 
attached. Msdn also suggests that BltFast fails if the surface has a clipper 
attached.


signature.asc
Description: This is a digitally signed message part.



Re: [PATCH 2/4] d3dx9: Implement ID3DXBaseEffect::SetVector().

2011-12-13 Thread Matteo Bruni
Hi Rico,

2011/12/13 Rico Schüller kgbric...@web.de:
 ---
  dlls/d3dx9_36/effect.c |   52
 +--
  1 files changed, 49 insertions(+), 3 deletions(-)

 +static void set_vector(struct d3dx_parameter *param, CONST D3DXVECTOR4 
 *vector)
 +{
 +set_number((float *)param-data, param-type, vector, D3DXPT_FLOAT);
 +if (param-columns  1) set_number((FLOAT *)param-data + 1, 
 param-type, (FLOAT *)vector + 1, D3DXPT_FLOAT);
 +if (param-columns  2) set_number((FLOAT *)param-data + 2, 
 param-type, (FLOAT *)vector + 2, D3DXPT_FLOAT);
 +if (param-columns  3) set_number((FLOAT *)param-data + 3, 
 param-type, (FLOAT *)vector + 3, D3DXPT_FLOAT);
 +}

Can't you use a for loop instead? While you're at it, you should also
pick one between float and FLOAT.




Re: [PATCH 4/4] d3dx9: Implement ID3DXBaseEffect::SetValue().

2011-12-13 Thread Matteo Bruni
2011/12/13 Rico Schüller kgbric...@web.de:
 ---
  dlls/d3dx9_36/effect.c |   42 --
  1 files changed, 40 insertions(+), 2 deletions(-)

 +TRACE(Copy %u bytes\n, param-bytes);
 +memcpy(param-data, data, bytes);

That trace is not correct.




Re: [PATCH 2/4] d3dx9: Implement ID3DXBaseEffect::SetVector().

2011-12-13 Thread Rico Schüller

Am 13.12.2011 16:10, schrieb Matteo Bruni:

Hi Rico,

2011/12/13 Rico Schüllerkgbric...@web.de:

---
  dlls/d3dx9_36/effect.c |   52
+--
  1 files changed, 49 insertions(+), 3 deletions(-)

+static void set_vector(struct d3dx_parameter *param, CONST D3DXVECTOR4 *vector)
+{
+set_number((float *)param-data, param-type, vector, D3DXPT_FLOAT);
+if (param-columns  1) set_number((FLOAT *)param-data + 1, param-type, 
(FLOAT *)vector + 1, D3DXPT_FLOAT);
+if (param-columns  2) set_number((FLOAT *)param-data + 2, param-type, 
(FLOAT *)vector + 2, D3DXPT_FLOAT);
+if (param-columns  3) set_number((FLOAT *)param-data + 3, param-type, 
(FLOAT *)vector + 3, D3DXPT_FLOAT);
+}

Can't you use a for loop instead? While you're at it, you should also
pick one between float and FLOAT.

Yeah, I'll send a new version with a loop. That way the float vs 
FLOAT will also be fixed.


Thanks
Rico




Debugging Wine with Lightroom 3.5

2011-12-13 Thread Roland Baudin

Hi,

let me first congratulate all the people who work hard on the Wine 
project : their work is much appreciated here...


Now, my question : I was able to install and run Adobe Photoshop 
Lightroom 3.5 under wine 1.3.33 or 1.3.34. While it runs quite well and 
is fairly stable I have two major issues :


- with Windows version set in winecfg to Windows XP (or below) the 
menu bar is hidden. However, clicking on the empty place of the menu bar 
partially reveals the menu items and they are functional. Moreover, 
other context menus (for example, right clicking on a picture and 
selecting Delete brings up the deletion dialog) are all working OK.


- with Windows version set in winecfg to Vista (or higher) the menu 
bar is now visible and fully functional, but many context menus are not 
working anymore (for example, right clicking on a picture and selecting 
Delete does nothing now).


[Note this problem is half known (see bug 18635 
http://bugs.winehq.org/show_bug.cgi?id=18635).]


So the dilemma is : either set Windows version to XP and get fully 
functional context menu but a hidden menu bar, or set version to Vista 
and see the menu bar but get many context menus broken.


I'd like to debug this situation so it would be nice if someone could 
point me to the right direction :


- first I don't understand the relation between Windows versions and 
menu bar or dialogs. Is it related to different versions of the dlls or 
any other mechanism?


- second, I noticed with XP version a lot of fixme:GetMenuInfo when 
running Lightroom. These messages don't appear with Vista version. I've 
tried to set up a patch (found on the Internet) that implement the 
GetMenuInfo function but this doesn't help. The patch seems to work but 
the menu bar is still hidden.


- third, with Vista version, when invoking a broken menu, there is no 
particular message I can see on the terminal that could tell me what is 
not working.


I've tried to play with various dll replacement but with no success. 
After googling and reading I suspect the problem is in the user32.dll 
and perhaps the DrawMenuBar implementation but I'm not sure...


Any help (at least some clue to start debugging) would be greatly 
appreciated.


Thanks,
RB




--
X File Explorer http://roland65.free.fr/xfe
Toutes Choses http://roland65.free.fr/ttc




Re: [PATCH 1/5] ddraw: Move clipper handling to ddraw.

2011-12-13 Thread Henri Verbeet
On 13 December 2011 15:31, Stefan Dösinger stefandoesin...@gmx.at wrote:
 Assuming that msdn[1] is correct, implies that we'll have to move overlays to
 ddraw as well.
It probably makes sense to do overlays completely in ddraw, but I
don't think this patch requires that.