On Monday 06 August 2007 20:36:34 Cedric BAIL wrote:
> On Saturday 04 August 2007 09:03:45 Carsten Haitzler wrote:
> > On Fri, 3 Aug 2007 13:52:22 +0200 Cedric BAIL <cedric.bail@...>
> > > On Thursday 02 August 2007 18:23:00 Cedric BAIL wrote:
> > > > Caro noticed a bug with gl_common. It segv when doing some gradient
> > > > stuff. I did some more testing on Linux with valgrind on
> > > > evas_gl_test. It did some wrong write before having malloc going
> > > > crazy. Here is the call stack :
> > >
> > > Ok, I found the source of this error. It's in gl_x11 engine in fact. In
> > > evas_engine.c at line 424, evas_common_gradient_fill_angle_set is
> > > called instead of evas_gl_common_gradient_fill_angle_set. That's solve
> > > the segv.
> > >
> > > But we have still some issue with gradient and gl. When it is
> > > resized, I can see some garbage around. I don't know exactly why, but I
> > > suspect the surface is not correctly initialized or copied in the GL
> > > texture during the gradient draw. But as I don't really understand the
> > > GL code, I don't know how to fix the issue.
> > do you have an example of this i can use to test with?
> I just came back from the weekend and you already fixed all the bug. All
> the previous garbage are gone.
> I only noticed something strange with test 33 of evas_gl_test. When you
> cover the window, some part of the gradient became black (like on the
> picture attached).
Hum, the picture didn't get to the mailing-list, you can see it at