Nightwulf|work: hi all
Lenin_Cat: is R6xx support good enough to play 3d games yet?
yangman: Lenin_Cat: 3D is handled by mesa. r6xx support is still being worked on
Lenin_Cat: the website says currently this dose extend to r6xx
yangman: it mostly works. just not in a mesa release yet
Lenin_Cat: RS780 is R6xxx right?
Lenin_Cat: yangman, is radeonhd and the experemental 3d in the ubuntu respertory?
yangman: no idea
Terman: any ideas when radeonhd will support KMS?
airlied: not sure if they are planning
airlied: no point using radeonhd with KMS over radeon
udovdh: fwiw: https://bugzilla.redhat.com/show_bug.cgi?id=522936
udovdh: hello btw
airlied: since all the reasons for keeping them separate are in the kernel
udovdh: see the bottom comment
udovdh: stupid Xorg bug, with and without compiz (which I don't use)
udovdh: and now my config isn't simple
udovdh: they even suggest a driver change when Xorg is the issue
airlied: udovdh: its not X.org
airlied: at least I don't think it is
udovdh: if I upgrade just Xorg it crashes
udovdh: if I just downgrade Xorg, it is stable
udovdh: xorg-x11-server-Xorg.x86_64 1.6.3-4.fc11 updates
udovdh: xorg-x11-server-common.x86_64 1.6.3-4.fc11 updates
udovdh: are in my yum queue
udovdh: v 220.127.116.111 is stable
airlied: ah maybe its the same issue with compiz, 1.6.3 seems to have gbroken a few things
udovdh: I do not use compiz.
udovdh: does radeonhd even support it on rv635?
udovdh: or rv630?
airlied: not in the version in F11
udovdh: I use git
udovdh: the f11 version doesn support accelleration
airlied: mesa and kernel
airlied: its not just the driver
udovdh: 18.104.22.168 mesa
udovdh: I odn't need compiz, just stability will do
airlied: don't upgrade X then until someone figures it out,.
airlied: I think ajax was looking at it
udovdh: contact info?
udovdh: I could test
airlied: he's on the bug already under the maintainer
udovdh: so I try to keep this bug open for the non-compiz cases
udovdh: because compiz is partly fixed with an update in testing
airlied: udovdh: have you got /var/log/gdm/:0.log from a crash?
udovdh: this update does not help me
udovdh: lemme see
airlied: it might be worth attaching
udovdh: I have logs there
udovdh: hmm. can reproduce the issue and attach it
airlied: it might be /var/log/gdm/:0.log.1
airlied: if gdm restarts etc
udovdh: doesn't help
udovdh: that is just the same log as on the console
udovdh: when I start Xorg manually
udovdh: it just dies after
udovdh: (==) FontPath set to:
udovdh: (==) ModulePath set to "/usr/lib64/xorg/modules"
udovdh: (eof) is mine
airlied: oh wierd and /var/log/Xorg.0.log has anything?
airlied: that is very wierd
udovdh: Xorg.0.log is similar
udovdh: /var/log/messages shows the segfaults as in the bugreport
b4283: i'm trying to enable the experimental 3d accleration on r6xx
b4283: but i'm a little confused about the dri files generated by mesa
b4283: r600_dri.so and swrast_dri.so, should i put them in /lib/dri or /usr/lib/xorg/modules/dri/ ??
udovdh: UXA? (instead of EXA etc?
Neo_The_User: how do I debug the radeonhd DDX via serial debug and not via gdb?
Neo_The_User: xorg.0.log and gdb say nothing is wrong but my screen is hay-wire
kcodyjr: actually you still want gdb, you just want the login to be via the serial port
kcodyjr: which involves starting a getty on the serial port, look in /etc/inittab, most distros have it there commented out.
kcodyjr: optionally, you could also redirect the kernel console to serial, so you also get the kernel messages
Neo_The_User: i actually rolled this distro entirely from source, purely by hand, and i dont have /etc/inittab
Neo_The_User: i have my kernel set to do that
kcodyjr: you're on your own then; whatever init system you're using has got to have something for starting the logins on each of your vt's
Neo_The_User: ok i just made my own /etc/inittab ;)
Neo_The_User: just typed this up: http://pastebin.ca/1571135
Neo_The_User: memorized the entire /etc/inittab file from archlinux :)
Neo_The_User: MostAwesomeDude: if the leaders of radeon driver development or whatever you want to call them, such as airlied, agd5f, and such, decided to start make an r600g, would you commit to the project?
Neo_The_User: ...if there were to be an r600-r7xx gallium driver. also, do you think there would be one?
ndim: rhd_lut.c: In function 'rhdLUTSetRows':
ndim: rhd_lut.c:193: error: void value not ignored as it ought to be
ndim: I'm getting this while building on PPC, but not on x86_64 or i386...
MAbeeTT: is reported a bug in radeonhd that keyboard, mouse, screen freezes?
MAbeeTT: WW) RADEONHD(0): DRMCPIdle: DRM CP IDLE returned BUSY!
ndim: Oh my.
ndim: That is a very ugly typo with that macro.
ndim: Very ugly.
ndim: bab7c5f9 (Yang Zhao 2009-09-06 15:31:15 -0700 193) RHDRegWrite(LUT, DC_LUT_30_COLOR,
ndim: bab7c5f9 (Yang Zhao 2009-09-06 15:31:15 -0700 194) (colors[index].red << 20) | (colors[index].green) << 10) | (colors[index].blue);
ndim: Not pretty :-)
ndim: Heads up, I have pushed a fix.
ndim: Or what I'd consider one.
Neo_The_User: When I try starting X, i sometimes get MTRR errors with the radeonhd DDX. is that normal?
Neo_The_User: it only happens with my custom kernel though with my custom MTRR clean up values modified