Re: [PATCH 2/7] d3d9/tests: Test that Direct3D9 doesn't modify the pixel format of the window it targets. (try 3)

[Henri Verbeet]

On 14 February 2014 00:31, Ken Thomases  wrote:
> - /* If no opengl32 function has been called, GetPixelFormat() doesn't reflect a
> -
>   prior SetPixelFormat().  So, call wglCreateContext() even though it will fail. */
> - hglrc = wglCreateContext(hdc);
> - ok(hglrc == NULL, "wglCreateContext succeeded when no pixel format has been set\n");
> - if (hglrc) wglDeleteContext(hglrc);

I decided to look into this a bit, but could only reproduce it on the
testbot. It seems just loading opengl32 is enough to make
GetPixelFormat() work. (Also see
https://newtestbot.winehq.org/JobDetails.pl?Key=4908.)

[PATCH 2/7] d3d9/tests: Test that Direct3D9 doesn't modify the pixel format of the window it targets. (try 3)

[Ken Thomases]

On Feb 14, 2014, at 11:53 AM, Henri Verbeet wrote:
> On 14 February 2014 00:31, Ken Thomases  wrote:
> > - /* If no opengl32 function has been called, GetPixelFormat() doesn't reflect a
> > -
> >   prior SetPixelFormat().  So, call wglCreateContext() even though it will fail. */
> > - hglrc = wglCreateContext(hdc);
> > - ok(hglrc == NULL, "wglCreateContext succeeded when no pixel format has been set\n");
> > - if (hglrc) wglDeleteContext(hglrc);
>
> I decided to look into this a bit, but could only reproduce it on the
> testbot. It seems just loading opengl32 is enough to make
> GetPixelFormat() work. (Also see
> https://newtestbot.winehq.org/JobDetails.pl?Key=4908.)

Hmm.  I won't be able to double-check until Monday, but I'm pretty sure I got the opengl32 test_setpixelformat() to fail on a real Win7 machine just by commenting out the wglCreateContext() call (and the other tests) that is done before it.  That is, running the executable from https://newtestbot.winehq.org/JobDetails.pl?Key=4752.

Assuming it works without the wglCreateContext(), would you prefer the dynamic loading approach to the import I had done?

-Ken

[PATCH 2/7] d3d9/tests: Test that Direct3D9 doesn't modify the pixel format of the window it targets. (try 3)

[Henri Verbeet]

On 15 February 2014 03:52, Ken Thomases  wrote:
> Hmm.  I won't be able to double-check until Monday, but I'm pretty sure I got the opengl32 test_setpixelformat() to fail on a real Win7 machine just by commenting out the wglCreateContext() call (and the other tests) that is done before it.  That is, running the executable from https://newtestbot.winehq.org/JobDetails.pl?Key=4752.
>
> Assuming it works without the wglCreateContext(), would you prefer the dynamic loading approach to the import I had done?

Yeah.

[PATCH 2/7] d3d9/tests: Test that Direct3D9 doesn't modify the pixel format of the window it targets. (try 3)

[Ken Thomases]

On Feb 14, 2014, at 11:53 AM, Henri Verbeet wrote:
> On 14 February 2014 00:31, Ken Thomases  wrote:
> > - /* If no opengl32 function has been called, GetPixelFormat() doesn't reflect a
> > -
> >   prior SetPixelFormat().  So, call wglCreateContext() even though it will fail. */
> > - hglrc = wglCreateContext(hdc);
> > - ok(hglrc == NULL, "wglCreateContext succeeded when no pixel format has been set\n");
> > - if (hglrc) wglDeleteContext(hglrc);
>
> I decided to look into this a bit, but could only reproduce it on the
> testbot. It seems just loading opengl32 is enough to make
> GetPixelFormat() work. (Also see
> https://newtestbot.winehq.org/JobDetails.pl?Key=4908.)

It seems that dynamically linked libraries are not initialized until called, while dynamically loaded libraries are initialized immediately.  Or something like that.

If I take your patch and remove the LoadLibraryA()/FreeLibrary() calls but add opengl32 to the IMPORTS, then it fails on a real Windows 7 machine.  Patch and executable are available from https://newtestbot.winehq.org/JobDetails.pl?Key=4980 although, at this writing, the test itself hadn't completed on the testbot.

I guess dynamic loading is good enough for me.  I'll rework the patches and resubmit.

-Ken