Re: [AD] OpenGL broken on Windows

[ Thread Index | Date Index | More lists.liballeg.org/allegro-developers Archives ]


On Sat, 19 May 2012 23:49:21 -0600
Trent Gamblin <trent@xxxxxxxxxx> wrote:

> It produces no output on iOS either. Just black screen.
> 

Did your patch fix it? I believe the only change should be to call
vt->update_render_state from within vt->set_current_display. Same in
Android if that shares any code (else Android simply does not support
the function for now).

> 
> On 2012-05-19, at 11:48 PM, Peter Wang wrote:
> 
> > Hi Elias,
> > 
> > GL on Windows produces no output.  From the log file:
> > 
> > opengl   D   ogl_render_state.c:28
> > _al_ogl_update_render_state      [   0.36928] alpha_test: 0
> > opengl   D   ogl_render_state.c:29
> > _al_ogl_update_render_state      [   0.36934] depth_test: 0
> > opengl   D   ogl_render_state.c:30
> > _al_ogl_update_render_state      [   0.36939] write_mask: 0
> > 
> > Probably something minor.
> > 

Yes, I had it working, but I shuffled code around and broke it
before committing. I think it will work as long as each driver calls it
in vt->set_current_display. That should be the only time it has to be
restored without the user calling al_set_render_state.

I also removed those three logs again as I'm calling
al_set_render_state too often. If it's still important to have it we
can add it back of course, maybe under a special debug channel shared
with the target bitmap changes. So then those two could be filtered out
easily with allegro5.cfg. For the bitmap changes I get like 4 lines per
frame which I can live with though, but for the render state like 1000
lines (so 60 000 per second) so I run out of disk space :P




Mail converted by MHonArc 2.6.19+ http://listengine.tuxfamily.org/