Hi all,

  I'm using DirectFB on the TI dm355 with the Montavista kernel 2.6.10. Or 
should I say, I'm using it successfully until I try to make it use the Davinci 
gfx driver for devmem.

The driver was created for the 64x, so I've had to remove any DSP calls and 
references to the 64x, as well as make a few tweaks to make it play nicely with 
the Montavista distribution of the kernel (vs TI's).

Anyway, the problem at hand: The solution may be trivial, not sure.

When directfb is initializing, it fails in the call to FBIOPAN_DISPLAY because 
within my kernel's fbmem.c "fb_pan_display" routine the test :
"Yoffset + info->var.yres  >  info->var.yres_virtual" is true

In this case,
Yoffset = 12147
info->var.yres = 480
info.yres_virtual = 960
So of course it fails

I'm wondering if this is just a matter of changing a setting in my directfbrc 
file, or the fb.modes, or...? I don't think the solution is just increasing my 
yres_virtual, I suspect something is wrong with the value of Yoffset. I've been 
trying to trace the creation of the Yoffset down into the surface pool creation 
where it is created (chunk), but I feel like I'm barking up the wrong tree.

Thanks!
Paul
_______________________________________________
directfb-users mailing list
directfb-users@directfb.org
http://mail.directfb.org/cgi-bin/mailman/listinfo/directfb-users

Reply via email to