For those of you who missed my previous announcement (it was hidden
behind a rather noninformative subject line), I've packaged up the
newest ATI driver (8.40.4) as wip/ati-xfree86-bin.

These drivers unfortunately seem to have an issue with proper
initialization of some display cards: on my machine, using the
8.40.4 driver immediately upon boot causes the machine to lock up
solid.

A work around is to start up an X server using an earlier release.
Then running the X server with the 8.40.4 drivers will be fine.

I've put together a simple command, /usr/pkg/sbin/iniapi, starts
up an X server using an older version of the driver, and then kills
it off.  It's available in a package called wip/ati-xfree86-bin-init.

This has been reported on the unofficial ATI proprietary bugzilla:

    http://ati.cchtml.com/show_bug.cgi?id=784

which is monitored by some of the ATI guys.  Should hopefully be
fixed in the next version!

Brian.

On 2007.08.16 13:30:56 -0600, Brian de Alwis wrote:
> On 2007.08.15 18:00:14 -0600, Brian de Alwis wrote:
> > On a related note, I finally made a package for installing the ATI
> > binary drivers, which is currently in wip/ati-xfree86-bin.  This
> > package installs the latest and greatest -- which unfortunately
> > still has this same performance issue!
> 
> On the bright side, the drivers now properly display the mouse
> pointer on the second screen.
> 
> I had some problems running the resulting X server -- the machine
> appeared to lock solid.  Running X directly with a specified layout
> first seems to have magically (puzzlingly) fixed this.  For example:
> 
>     $ X -layout 'your layout identifier'
> 
> YMMV.
> 
> Brian.
> 
> -- 
>   Brian de Alwis | Software Practices Lab | UBC | http://www.cs.ubc.ca/~bsd/
>       "Amusement to an observing mind is study." - Benjamin Disraeli

-- 
  Brian de Alwis | Software Practices Lab | UBC | http://www.cs.ubc.ca/~bsd/
      "Amusement to an observing mind is study." - Benjamin Disraeli