--- Denis Oliver Kropp <[EMAIL PROTECTED]> wrote:
> Cool, does 4) include fixes that would still apply
> to embedded-2?
The biggest bug was MakeCurrent(0, 0) was not being
passed into the mesa layer. The others were error
handling.
> I can help you with readding per drawable buffers,
> at least doing
> the changes to the Matrox DRI and DRM driver.
My goal is to get Mesa to a point where each
process/window gets it own buffer. Then a window
manager process will composite these on to the screen.
DirectFB wants this too. Shouldn't we just implement
pbuffers so that we follow the standard?
> I just checked out embedded-1, applied your patch
> and everything
> works as expected. I tested the sample_server with
> glxgears.
> VT switching still seems to be disabled.
__miniglx_StartServer( const char *display_name )
{
Display *dpy;
int use_vt = 0;
Change the 0 to a 1. I should have done that before
posting.
I run XFree on a PCI card for debugging and Mesa uses
the AGP card. If I let Mesa take the VT I lose my
keyboard for the debugger.
I have part of QT working on this driver. With QT I
still don't use a VT but I have modified the kernel to
allow two keyboards and mice.
=====
Jon Smirl
[EMAIL PROTECTED]
__________________________________
Do you Yahoo!?
SBC Yahoo! DSL - Now only $29.95 per month!
http://sbc.yahoo.com
--
Info: To unsubscribe send a mail to [EMAIL PROTECTED] with
"unsubscribe directfb-dev" as subject.
