From: Andrew Apted <ajapted@netspace.net.au>
  To  : ggi-develop@eskimo.com
  Date: Sat, 27 Feb 1999 13:11:09 +1100

Re: ggiSetOrigin with fbdev...

Andy writes:

>  > I have some problems getting ggiSetOrigin to work with the fbdev target.
>  > What does the ff->fix.ywrapstep variable do?
>  
>  Hmm - from the name I'd say it give the stepping in which warping, aka
>  splitline can be set.
>  
>  This is the scanline number at which the memory counter warps back to the
>  logical start of the graphics memory and used to fake a "circular" video
>  buffer which is _very_ useful for scrolling textmodes.
  
This is my understanding of it as well.  The exact way it is supposed to
be used (and how it interacts with ypan and vyres>yres), that I don't
know.

>  > GGI_fbdev_setorigin in fbcon/pan.c contains:
>  
>  > if ((ff->fix.ypanstep == 0) || (ff->fix.ywrapstep == 0)) {
>  > 	y = 0;
>  > }
>  
>  > and ypanstep is set to 1 and ywrapstep is set to 0 in fbcon-kgi.c, so the
>  > setorigin function will never work, is this correct or should the driver
>  > change the values somehow?
>  
>  Hmm - try commenting out the check for the warpstep. IMHO this looks wrong.

Yes it does look wrong, I don't know why I wrote that (or even *if* I
wrote that :->), but I'll remove the bogus check (the LibGGI API doesn't
handle wrapping anyway).

Cheers,
___________________________________________________
                                               \  /
  Andrew Apted   <andrew@ggi-project.org>       \/