http://bugs.freedesktop.org/show_bug.cgi?id=19625
--- Comment #1 from Ian Romanick 2009-01-16 17:25:28
PST ---
Created an attachment (id=22045)
--> (http://bugs.freedesktop.org/attachment.cgi?id=22045)
Generate an error if either the read or draw drawables are not GLXWindow
Proposed fix.
http://bugs.freedesktop.org/show_bug.cgi?id=19625
Summary: Instead of returning and error, glXMakeContextCurrent
crashes
Product: Mesa
Version: CVS
Platform: Other
URL: https://bugzilla.novell.com/show_bug.cgi?id=466806
On Fri, 16 Jan 2009, Timo Aaltonen wrote:
> On Fri, 16 Jan 2009, Brian Paul wrote:
>
>> Owain Ainsworth wrote:
>>> Hey,
>>>
>>> As a companion to the DDX diff I just sent out, this patch rips all of
>>> the broken i9XX pageflipping code from mesa.
>>>
>>> There's one chunk I'm not sure about (whet
On Fri, 16 Jan 2009, Brian Paul wrote:
> Owain Ainsworth wrote:
>> Hey,
>>
>> As a companion to the DDX diff I just sent out, this patch rips all of
>> the broken i9XX pageflipping code from mesa.
>>
>> There's one chunk I'm not sure about (whether the for loops in
>> IntelWindowMoved are still ne
On Fri, 2009-01-16 at 16:27 +0100, Matthias Hopf wrote:
> Full backtrace etc. in https://bugzilla.novell.com/show_bug.cgi?id=466806
>
> In short, during glutCreateWindow() ->.. fbOpenWindow() ->
> MakeContextCurrent() is called, which calls driBindContext() with NULL
> draw and read pointers. The
Full backtrace etc. in https://bugzilla.novell.com/show_bug.cgi?id=466806
In short, during glutCreateWindow() ->.. fbOpenWindow() ->
MakeContextCurrent() is called, which calls driBindContext() with NULL
draw and read pointers. The code looks like this should never happen,
because the draw and rea
Owain Ainsworth wrote:
> Hey,
>
> As a companion to the DDX diff I just sent out, this patch rips all of
> the broken i9XX pageflipping code from mesa.
>
> There's one chunk I'm not sure about (whether the for loops in
> IntelWindowMoved are still needed), but the rest is definitely good to
> die
