Re: misc/compat6x port no longer sufficient for DRI under head?

On Thu, 2009-09-17 at 20:13 -0500, Sean C. Farley wrote:
> On Thu, 17 Sep 2009, David Wolfskill wrote:
> 
> > On Thu, Sep 17, 2009 at 06:37:46PM -0500, Sean C. Farley wrote:
> >> On Thu, 17 Sep 2009, David Wolfskill wrote:
> >>> On Thu, Sep 17, 2009 at 08:32:09AM -0700, Freddie Cash wrote:
> >>>> ...
> >>>> Have you tried re-enabling hald and dbus and configuring X to use 
> >>>> those?
> >>>
> >>> No, I haven't.
> >>>
> >>> One of the reasons I stopped trying to use hald(8) was that when I 
> >>> booted head, hald went into a 100% CPU busy loop.  (Mind, this was 
> >>> with hald built inder stable/6.)
> >>
> >> How about building your system without HAL?  I have been quite happy 
> >> without it.  As long as I have the following in my Xorg.conf, 
> >> everything (that I need) seems to be detected nicely:
> >>
> >> Section "ServerLayout"
> >>     ...
> >>     Option "AllowEmptyInput" "off"
> >>     ...
> >> EndSection
> >
> > Well, ref. the above "I stopped trying to use hald(8)" -- I did this by:
> 
> You expected me to actually read pertinent information.  ;)  Sorry about 
> that.
> 
> > Section "ServerFlags"
> >        Option      "AutoAddDevices" "False"
> > EndSection
> >
> > And then disabling the hald & dbus startup.
> 
> I built all of my ports using WITHOUT_HAL as opposed to disabling hald 
> startup.  Did you try this variation?
> 
> > It had been working in stable/6, stable/7, and head before the 
> > (apparently DRI-related) issues arose in stable/7 and (mnore recently) 
> > head.
> 
> WITHOUT_HAL throughout ports, dbus running and the option I listed above 
> works well for me with stable/7 (r196739).

Just to be clear... hald has a number of build time features and tweaks
depending on current OS support, so I'm not really surprised that hald
doesn't work well or correctly in David's case.

hald has no interaction with drm though, other than if hald is running
it will try to open drm and get the driver version.  hald certainly
wouldn't have any impact of X functioning or not based on the DRI
option.

The evidence so far seems to be pointing to some issue with setting
MTRR, though I'm not exactly certain how or what is failing.

robert.

> Sean
-- 
Robert Noland <rnoland_at_FreeBSD.org>
FreeBSD