This bug report was migrated from our old Bugzilla tracker.

These attachments are available in the static archive:

• wgl acceleration patch (sdl.diff, text/plain, 2011-05-02 03:40:47 +0000, 1015 bytes)

Reported in version: HG 1.2
Reported for operating system, platform: Windows (All), x86

Comments on the original bug report:

On 2011-05-02 03:40:00 +0000, Matthias Bentrup wrote:

The current code always passes the WGL_ACCELERATION_ARB attribute with parameter WGL_FULL_ACCELERATION_ARB, and if the user specified the SDL_GL_ACCELERATED_VISUAL attribute it passes another WGL_ACCELERATION_ARB attribute with parameter either WGL_GENERIC_ACCELERATION_ARB or WGL_NO_ACCELERATION_ARB (which are both much slower than WGL_FULL_ACCELERATION_ARB) !

The attached patch fixes this so that the WGL_ACCELERATION_ARB is always set exactly once to either WGL_FULL_ACCELERATION_ARB or WGL_NO_ACCELERATION_ARB and chooses WGL_FULL_ACCELERATION_ARB if the SDL attribute is not set by the user.

On 2011-05-02 03:40:47 +0000, Matthias Bentrup wrote:

Created attachment 609
wgl acceleration patch

On 2011-08-21 09:05:44 +0000, Ryan C. Gordon wrote:

*** Bug 1281 has been marked as a duplicate of this bug. ***

On 2011-08-21 09:14:22 +0000, Ryan C. Gordon wrote:

I'm marking this as a dupe, since Bug # 1254 covers this issue for both 1.2 and 1.3.

--ryan.

*** This bug has been marked as a duplicate of bug 1254 ***

On 2011-08-21 13:12:09 +0000, Thilo Schulz wrote:

I would also like to remark that this patch of yours introduces the behaviour of SDL-1.3, where WGL_FULL_ACCELERATION_ARB is added even if no explicit requirements on hardware acceleration were made. So you will prohibit libSDL from allowing the fallback to software rendering if for some reason hardware rendering is not available.

Maybe I am incorrect in my interpretation how choosing this pixelformat works. Read up on my thinking in my bug that Ryan marked to be the duplicate with this one. Ignore my first comment as it's wrong. Just read the second one and tell me what you think.

On 2011-08-22 00:40:02 +0000, Matthias Bentrup wrote:

Hello Thilo,

your patch looks good. The semantics match the x11 implementation (0 = force accelerated mode, 1 = force software mode, -1 = allow both).

On 2011-08-24 20:32:30 +0000, Ryan C. Gordon wrote:

I went with Thilo's patch.

--ryan.