Index: [thread] [date] [subject] [author]
  From: Sengan Baring-Gould <sengan@seqnet.net>
  To  : ggi-develop@eskimo.com, core@suntech.com, Andrew Apted <ajapted@netspace.net.au>
  Date: Sun, 23 Aug 1998 18:19:53 -0600

MEDIA GX PATCH

Andrew Apted wrote:

> >  I don't like the bottom line: why 0K?
>
> It looks like the mode isn't being set.  After the mode is set, the mmio
> field of kgifb_display.mode is supposed to have the physical/virtual
> addresses and fb size (I assume the MediaGX driver does this properly).

Thanks to Andrew's help, and his et6000.c code, please add in front of line

971 of chipset/Cyrix/gx.c:

        memcpy(&(dpy->mode), mode, sizeof(struct kgi_mode));

(line 971 is:

        DEBUG1 ("Mode set complete");

of kgim_chipset_set_mode)

This solves the fb1 size problems and gets me a nice block cursor. Should
the Xfb x-server work with this configuration?

Now let me ask why we have 2 modes (one in dpy, and one in the driver's
private space)? Seems a little confusing to me.

Sengan

PS: Manu, pour avoir acces CVS developpeur que dois-je faire?

Index: [thread] [date] [subject] [author]