Re: GetCurrentObject

GetCurrentObject returns 0, but as far as it knows, it is
returning successfully.  This is because in the line

case OBJ_BITMAP: ret = dc->hBitmap; break;

dc->hBitmap is 0, although dc is a valid DC.  The HDC that is passed
to GetCurrentObject to get dc is obtained by calling
CWindow::BegingPaint(PAINTSTRUCT *) from a class derived from ATL's
CWindow.  I'm guessing that this is where the problem is probably
coming from - the app is written in C++ with ATL.

-Steve

I've seen this occur when wine frees a gdi handle it shouldn't. Check
that the OBJ_BITMAP handle is really valid.

Also wierd stuff can occur if the app expects a HDIB but we provide it a
HBITMAP... or vice versa.

Mike

[e-mail] wrote:
> My app calls GetCurrentObject with a valid HDC and OBJ_BITMAP.
> It returns 0, however, which causes my app to crash.  What further
> info can I provide to help debug this?
>
> -Steve

GetCurrentObject

[Rein Klazes]

On Fri, 8 Nov 2002 17:56:05 -0500, you wrote:
> GetCurrentObject returns 0, but as far as it knows, it is
> returning successfully.  This is because in the line
>
> case OBJ_BITMAP: ret = dc->hBitmap; break;
>
> dc->hBitmap is 0, although dc is a valid DC.  The HDC that is passed
> to GetCurrentObject to get dc is obtained by calling
> CWindow::BegingPaint(PAINTSTRUCT *) from a class derived from ATL's
> CWindow.  I'm guessing that this is where the problem is probably
> coming from - the app is written in C++ with ATL.
>
> -Steve

Here a patch that is lying around here for quite a while that I use to
prevent an Eudora 4/5 crash with the same symptoms. I wonder if it
helps you.

And yes, GetCurrentObject(..,OBJ_BITMAP) on a freshly created dc
returns != 0 in windows. I'm just not sure this patch is the real
solution, although I haven't seen any negative effects.

Rein.