Re: X11 server on alpha -- update

[Michael Cree <mcree@orcon.net.nz>]

On 27/10/2009, at 5:09 AM, Bob Tracy wrote:

> On Mon, Oct 26, 2009 at 08:18:05AM -0500, Bob Tracy wrote:
> > On Mon, Oct 26, 2009 at 10:05:10PM +1300, Michael Cree wrote:
> > > Bob Tracy wrote:
> > > > Long story short: 1.7.0.902 fixed the problems with modules not  
> > > > loading
> > > > because of undefined symbols, but the end result is no better  
> > > > than what
> > > > I reported for the earlier 1.7.0 version.  The mouse cursor is  
> > > > normal
> > > > brightness, but the display is otherwise extremely dim:
> > >
> > > Maybe remove as much as possible from your xorg.conf?  Xserver 1.7
> > > should start with a very minimal xorg.conf.  I am not quite sure  
> > > why you
> > > need the NoInt10 and DefaultConnectorTable options.
> >
> > The reason for "NoInt10" escapes me at the moment...
>
> I actually commented the change in my xorg.conf file :-).  If  
> "NoInt10"
> isn't "true", there's a probe for the video BIOS which fails on the
> Alpha (or at least, that *used* to be the case).  Since I can't access
> the video BIOS, I have to use the default connector table too.  Are  
> you
> claiming that the current X server on Alpha can now do something  
> useful
> with a x86 video BIOS?  If that's true, I *am* impressed.

Yes, I am.  There is an x86 emulator in the Xserver and it can read  
and use the video BIOS. On my Radeon 9200 the int10 code loads fine  
and I get the message:

(II) RADEON(0): Legacy BIOS detected

The Xserver can now POST video cards that the SRM fails to POST, and  
with the most recent kernels with KMS I presume even the kernel can  
POST Radeon cards that the SRM fails to POST (though I haven't tested  
that yet).

Cheers
Michael.