Re: [Linux-fbdev-users] virtual frame buffer

[Geert U. <ge...@li...>]

On Wed, 12 Oct 2005, Yan Seiner wrote:
> OK, this is a long shot since the kernel virtual frame buffer isn't really
> part of the fbdev development....  (that I know of?  Maybe it is....)
> 
> Is it possible to use the virtual frame buffer as a real device?  The one that
> says 'for testing only' in the menuconfig?
> 
> I am trying to set up a graphical system on an embedded system with no video
> hardware at all.  I am using nano-x (http://www.microwindows.com) The output
> is into the virtual frame buffer, and then I am trying to use vnc to remote
> the display to another computer.
> 
> The nano-x client apps can see mouse events from vncviewer, but vncviewer only
> shows total blackness.
> 
> Is the virtual frame buffer just a black hole? Or should vnc work with it?
> IIRC vnc reads the video memory and then sends it to the viewer.  Obviously,
> if the virtual frame buffer is just a black hole that won't work....
> 
> From looking at the source, it seems like the vfb really allocates memory for
> framebuffer use.... And vncserver is mmap'ing it correctly.  So that would
> lead me to believe that it should work....

Vfb behaves like a real frame buffer device, except that you cannot mmap() it
to userspace (resulting behavior is undefined). But that should be fixable by
letting vfb provide its own fb_mmap() routine, if anyone cares to implement.

Gr{oetje,eeting}s,

						Geert

--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- ge...@li...

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
							    -- Linus Torvalds