kernel 2.6.9-rc2-mm2 vs glxgears

[Dave Airlie]

Compiling Xorg from source isn't always the best idea, I'd just use
vendor supplied RPMS  for Fedora (I'm nearly sure Mike did some for
FC2)....

The problem with xfs is that the FC2 init script is bascially broken,
I've reported this as a bug, but it is vendor specific, the problem is
the new security code for sockets is broken by the rm -rf
/tmp/.font-unix line in the /etc/init.d/xfs script..

Dave.


On Wed, 29 Sep 2004 20:10:08 -0400, Gene Heskett
<gene.heskett at verizon.net> wrote:
> On Wednesday 29 September 2004 06:21, Felix Kühling wrote:
> >On Tue, 28 Sep 2004 23:56:15 -0400
> >
> >Gene Heskett <gene.heskett at verizon.net> wrote:
> >> On Tuesday 28 September 2004 18:21, Felix Kühling wrote:
> >> >On Tue, 28 Sep 2004 15:34:36 -0400
> >
> >[snip]
> >
> >> >Sounds like it could be a problem with frame throttling. My guess
> >> > is that interrupts get delayed somehow. The driver uses them to
> >> > wait for previous frames to finish rendering. You can try
> >> > different settings that don't use interrupts using the
> >> > environment variable fthrottle_mode, e.g.
> >> >
> >> >fthrottle_mode=0 glxgears (busy waiting)
> >> >fthrottle_mode=1 glxgears (usleeps)
> >> >fthrottle_mode=2 glxgears (interrupts, the default)
> >>
> >> Humm, interesting results:
> >
> >Indeed. Frame throttling is obviously not the problem. Low CPU usage
> >even with busy waiting indicates that the driver gives away the CPU
> > when it shouldn't. There was a discussion about proper use of
> > sched_yield and 2.6 kernels some time ago. Was this issue resolved?
> >
> I saw that go by on lkml, but I've NDI if it was resolved, sorry.
> 
> >> [root at coyote linux-2.6.9-rc2]# fthrottle_mode=0 glxgears
> >> ATTENTION: default value of option fthrottle_mode overridden by
> >> environment.
> >> IRQ's not enabled, falling back to busy waits: 11 0 5
> >> 826 frames in 5.0 seconds = 165.200 FPS
> >> 992 frames in 5.0 seconds = 198.400 FPS
> >> 991 frames in 5.0 seconds = 198.200 FPS
> >> 821 frames in 5.0 seconds = 164.200 FPS
> >
> >[snip]
> >
> >What are you doing differently when you get 10fps? (Different
> 
> running any -mm2-3-4 kernel in the 2.6.9-rc2 series, 2.6.9-rc2 bare
> seems to work (as shown above but I'm not running it at the moment as
> its an unknown, untested after the last dri patch this morning)
> although today has been hell trying to make X run at all.
> 
> Yum, in its infinite stupidity, over-wrote my 6.8.1 install which had
> been checkinstalled and rpm reported it correctly, with the latest
> 6.7-something or other, totally screwing up the system the instant I
> tried a restart.  Somewhere along the line Xft disappeared, and that
> meant the /var/log/Xorg.0.log ended with a can't find the fixed font,
> and the machine goes out to lala land until the reset key is tapped.
> It should revert to the VT I started from in that case, but does not,
> going south till the universe runs down or you find the reset button.
> 
> Shouldn't that get a bug filed?
> 
> Using my firewall boxes old mozilla, I got on the wiki page and
> finally found what that error represents, so now I have about 20
> lines of FontPath = in my /etc/X11/xorg.conf.  I've NDI where the
> hell Xft went, and I am *not* a happy camper with yum.  As of about
> 10 minutes ago now, its gone from the init.d dir.  As in mv S98yum to
> K02yum in rc3.d.  And I find it odd that 1) yum was running, and 2)
> chkconfig --list yum returned a full line of off's *before* I did
> that.
> 
> Anyway, after about 12 hours of seducing the canine over this, I'm
> finally running, albeit probably crippled somewhere mozilla will find
> immediately without the Xft server running.
> 
> Here's what I did when I found yum had pissed in my cherrios.
> 
> from a non-x shell: rpm -e --nodeps --force `rpm -qa |grep xorg`
> cd'ing to my /usr/src/X-6.8.1 build directory and:
> #>make World && make install (several times)
> #>checkinstall -R -fstrans=no
> #>cd /usr/src/redhat/RPMS/i386
> #>rpm -Uvh --nodeps --force X-6.8.1.i386.rpm
> 
> I've been around this loop 4 or 5 times, also ran a 6 pack of fancy
> cement blocks I'm making for a retaining wall, so I didn't spend all
> my time on this, that takes about 5 hours & 3 ea 80 lb bags of
> ready-mix.  Frankly Felix, I was too pissed to trust myself to attack
> it with any great amount of patience.
> 
> So my next pressing concern is to restore the Xft function, and the
> man pages seem to be relatively silent about what to do if it doesn't
> run.
> 
> Help?  Please...
> 
> > kernel?) What effect does fthrottle_mode have in that
> > configuration?
> 
> In the 10fps config, zip.
> >
> >| Felix Kühling <fxkuehl at gmx.de>
> >| http://fxk.de.vu | PGP Fingerprint: 6A3C 9566 5B30 DDED 73C3  B152
> >| 151C 5CC1 D888 E595 |
> 
> --
> Cheers Felix, Gene
> 
> 
> "There are four boxes to be used in defense of liberty:
>  soap, ballot, jury, and ammo. Please use in that order."
> -Ed Howdershelt (Author)
> 99.26% setiathome rank, not too shabby for a WV hillbilly
> Yahoo.com attorneys please note, additions to this message
> by Gene Heskett are:
> Copyright 2004 by Maurice Eugene Heskett, all rights reserved.
> _______________________________________________
> xorg mailing list
> xorg at freedesktop.org
> http://freedesktop.org/mailman/listinfo/xorg
>