jbeckers: hi everyone
jbeckers: I'm having trouble getting the radeonhd driver to work with my x1600
jbeckers: I've built and installed the driver, but xorg refuses to use it and falls back to vesa
jbeckers: i'm on ubuntu hardy with xserver 1.4 and a minimal xorg.conf, but even with a full xorg.conf to won't work
jbeckers: (minimal is: no driver specified and other things ripped out)
jbeckers: (full is: radeonhd as driver specified)
jbeckers: also: my onboard x1250 works perfectly with this same install
Dioxin: http://pastebin.com/d2051ba67 is my xorg.conf, and http://pastebin.com/dfaa5a39 is my logfile
Dioxin: can anyone give me a hand to get an x2300 working please
Dioxin: I think I just need to put the correct settings into xorg.conf but I'm not sure what they are
Dioxin: I've downloaded the radeonhd drivers from adept package manager using the 320 x 200 display that I can actually get working
wirechief: I have a ATI1300 pro 512mb, this is my xorg.conf for comparisons to your x2300's http://paste.debian.net/48203
wirechief_: Dioxin: hope this is helpful to you ^^^
libv: Dioxin: 320x200?
libv: how did that happen?
Dioxin: the only way I can get a display in xserver is to use vga and it defaults to 320x200
libv: vesa?
Dioxin: fails
libv: hrm...
Dioxin: wirechief the problem is that I'm also using a laptop .. so its a x2300 mobility radeon
wirechief_: ah ok
wirechief_: well i use intel x3100 and a ati9200 mobility not much with them.
wirechief_: but the screen section in your xorg does look flakey
wirechief_: and your log is complaining about it ;)
wirechief_: you have tried to dpkg-reconfigure -phigh xserver-xorg ???
wirechief_: not sure with your kernel version if xrandr -s 0 will help to restore it...
Dioxin: it doesnt detect my display nor my graphics card
Dioxin: I can try the dpkg-reconfigure thing tho
Dioxin: what does xrandr -s 0 do?
Dioxin: I've got Kubuntu gutsy gibon 7.10 installed
wirechief_: Dioxion basically resets it.
wirechief_: heres more info on xrandr http://bgoglin.livejournal.com/10423.html
Dioxin: well I've done all this from a fresh install
wirechief_: hmm kubuntu was it a upgrade from earlier version or new install ?
wirechief_: ooops ok
wirechief_: well the dpkg thing might help, just use default answers
Dioxin: fresh install and then suffer with 320x200 resolution to get adept manager to install radeonHD from net
Dioxin: ok I'll try that
wirechief_: i had problems like that with my ati1300
Dioxin: is it configure or reconfigure?
Dioxin: nvm its reconfigure
wirechief_: reconfigure i believe
Dioxin: brb 10 mins then
wirechief_: but i did a fresh install of ubuntu 7.10 with it messing up the resolution
udovdh: back home!
wirechief_: udovdh: a real home or home:home
udovdh: real
libv: Kano: versioning everybody?
libv: anybody else get this?
Kano: hi libv, yesterday agd5f told me that in src/r5xx_exa.c a #include
libv: oh, right, memset
libv: right :)
libv: i cannot go and do it directly, but it will be the first thing i do on the driver.
Kano: also i dont get why the version is still 1.1.0 when you compile it, cant you name it 1.1.0+git or whatever?
Kano: you really dont know if you use the old or new driver
libv: Kano: did you try reading the log?
Kano: the log?
Kano: sure
libv: because it has the start of the sha-id in the log
Kano: well i clean all .git* from a debian source package
Kano: if you parse that thats stupid
libv: oh?
libv: in case this is missing, we cannot know what version this is.
Kano: then add just another pre version
libv: how?
libv: Kano: do you expect us to adjust these version numbers all the time?
libv: how should scripts know what to do towards version?
Kano: no, but at least after a final to 1.1.0+git or so
libv: they have nothing to work from
libv: ...
libv: maybe we should stop version.sh from running under certain circumstances
libv: so that you keep original information
libv: Kano: besides.. our driver is one of the only drivers out there that tries to handle things sanely
Kano: and one driver i can not test due to missing hardware. and when i look into logs from others i can not see what driver is really used...
libv: Kano: how do you deal with this with all those other drivers (besides unichrome)?
Kano: i dont update em. but as etch has no xorg 7.1.1 i only could use radeonhd as update
Kano: not the standard ati driver
libv: right, we're pretty much the only ones doing that sortof thing too :)
Kano: of course i can already install fglrx very easyly but i would prefer a wrong oss driver too
Kano: well kanotix has not only 1 user
Kano: and i dislike to boot and have got only vesa
Kano: fglrx is 50 mb download by default
Kano: stripped down 11 mb
Kano: maybe you add support for rv410 too then i could try that driver myself
rmh3093: libv, whats up, anything new with the driver?
libv: xaa/exa
libv: Kano: totally different hardware
Kano: i think atombios would help?
libv: Kano: not really.
Kano: well i dont spend money to buy new ati hw
libv: 30EUR on ebay go a long way :)
libv: or 50 in the shops :)
Kano: i prefer getting a new nv one for 200+
Kano: the x700se i got at least for free
Kano: of course it did not work directly would be really too easy
Kano: mesa needed a fix
Honk: this may be a bit offtopic, but is radeonhd supposed to be faster than fglrx (newest) in 2d (e.g. while i'm moving windows around)? :]
Kano: it cant be slower ;)
udovdh: so finally I made the switch from ananlog to dvi-d cabling!
udovdh: thanks to xrandr
udovdh: xrandr --output DVI-I_2/digital --auto
udovdh: and voila
udovdh: so that also works. :-)
libv: Honk: no, we only use a limited part of the features
udovdh: just as long as the implemented part works well, is relatively bugfree and fast enough...
end-schranz: hi, i am using the latest git version of the driver + a x1950 pro + dvi2vga , there is only one problem that image of the monitor is placed to far left
udovdh: re-adjust the monitor?
egbert: end-schranz: crt of panel?
udovdh: or tweak the video mode?
end-schranz: crt
egbert: so you want to follow udovdh's suggestion probably.
end-schranz: re-adjust wont work cause i have to use it wiht other oses to
udovdh: then you might redjust the mode
udovdh: I don't know what tool could be used nowadays
egbert: it should store the values for the specific mode.
udovdh: but I rmember from a few years back
udovdh: that there was a way to display the current mode
udovdh: and ajust parameters on the fly to move the picture etc
egbert: yeah, thru the xf86vidmod extension.
end-schranz: udovdh: ok
udovdh: when stuff was good it would spit out the new mode
udovdh: extension you say
udovdh: does that work on radeonhd?
egbert: this won't work here.
udovdh: so what is the right way now?
egbert: it only had support for a single head.
udovdh: I\ll try dual tomorrow perhaps
egbert: in theory the mode should be ok because it comes from the monitor/is standard/ is caluclated thru cvt
egbert: so i haven't run into any case recently where the mode that was picked required some readjustment.
egbert: end-schranz: are you using a 'hand made' mode in xorg.conf?
end-schranz: egbert: yes
end-schranz: egbert: restarting X brb
egbert: ok, maybe you want to comment it out.
egbert: end-schranz: did you comment out this mode?
end-schranz: comment out what mode?
egbert: in the config file
end-schranz: egbert: i didn't set anymode
Kano: libv: it seems still some ids are missing like rv635
libv: Kano: we are still working on rv620/635 support.
Kano: did you compare with common/lib/modules/fglrx/build_mod/fglrxko_pci_ids.h from fglrx?
libv: Kano: yes, sndirsch spams us with more ids from time to time
libv: Kano: as a reward he gets a glare and some remarks from me :)
Kano: well thats not really hard to use that file
Tigerchen: libv: sorry to bother you again, but i'm curious if you were able to reproduce the whiteout-bug
libv: Tigerchen: too busy, sorry
libv: the laptop is right beside me
Tigerchen: no problem with that, just curious
libv: Kano: yes, we know, stefan needs to track new releases of fglrx, and each time one happens, i roll my eyes as our id list becomes more and more insane :)
Death_Syn: i'm going to try radeonhd on my rs690 soon
Honk:
udovdh: stuff will improve over time...
Honk: huh? radeonhd is working way better than fglrx already for me :P
Honk: that's what i was trying to say
udovdh: dunno fglrx
udovdh: got the card vy accident
udovdh: got the driver
udovdh: started using it
udovdh: only missing 2d/3d
udovdh: that is my limited experience
wirechief_: is 3d in the near future or is it still a dream?
udovdh: 2d for r5xx is there in git
udovdh: rest might follow?
Honk:
Honk: cool :D
udovdh: yes
udovdh: just pull
udovdh: can't test it with my rv630
Honk: me neither ;)
Honk: it's cool anyway
wirechief_: i tried it and seems to work
wirechief_: ati1300
mstrobert: I need to buy some video cards for our workstations. If I want the cards to Just Work with Ubuntu and the open-source radeonhd driver sometime in the coming months, is it more important to purchase R500 or R600, or are they both equally good choices? I'm not concerned with hardware video decoding, but I want them to drive dual LCDs and Compiz. Right now I'm considering the Radeon X1650PRO.
glisse: wirechief_: well for 3d r500 might have a quick hack so r300 driver work for it
glisse: but r600 will take times
glisse: and proper|clean 3d support too
wirechief_: hmm well, i was just wondering if it was going to happen, like 2008 or 09 (no date needed)
wirechief_: or is there too many obsticals at this point to even venture a guess.
glisse: wirechief_: well i would say that by this summer a proper 3d for r300/r400/r500 should exist but i don't think any distrib will ship it
wirechief_: ok, well thats good enough..
glisse: for r600 i can't properly guess but i would say by end of this year
wirechief_: hmm well that sounds good, look forward to seeing that.
sonne: glisse, I would already be happy if macbookpro's were working with xrandr ... but yes I should not complain
glisse: sonne: hey i should really find few hour to powerup my macbook and look at this things
sonne: glisse, if you want to be loved (platonically!) then yes pleeeeeaaaase.
glisse: sonne: to refresh my memory the things is that it doesn't properly detect the dvi output right ?
sonne: ddc does only randomly work on both vga and dvi
sonne: (externally)
sonne: and after resume the colors are garbled, which can be fixed by either xgamme -f 1.0 or switching to the console and back
sonne: glisse, for some monitors/displays ddc never works and for some it works sometimes and sometimes a suspend to ram / resume cycle helps
sonne: (hope that information is useful)
libv: sonne: right, i'm not sure what is happening there
libv: sonne: still need to go and find this out... like so many things
libv: but we're rather strechted thinly at the moment
libv: stretched even
libv: so stretched that my brain finger fifo is busted :p
glisse: libv: well apple changed the fake atombios and they likely doesn't init this things properly anymore
sonne: it is just so sad that I am not able to be of help here...
libv: glisse: the gamma issue is not just apple
glisse: or at least in the same way they use to do
glisse: libv: just thinking of i2c issue here
libv: glisse: it's RandR1.2 code and our driver no longer working together
glisse: i love weird gamma :)
sonne: all I can do is provide logs and beg you people
libv: the i2c stuff, we need to get this hardware in our hands and test for ages :)
libv: or we can fallback to gpio in this case, and add a card specific flag to our cardid table
sonne: glisse, you mean quite safe to type in clear text passwords ;-)
glisse: libv: well i will test this i2c things with avivotool there is code their to use the i2c engine like i think you do in radeonhd
glisse: ie not using gpio
libv: glisse: sounds great, but don't let it hurt your phd :)
sonne: I guess glisse does this withouth specs during the night :)
glisse: speaking of that anyone familiar with array & operation in postgres ? :)
rbmorse: Dioxin: In case you're reviewing the logs, this is an xorg.conf that has worked for me on Fiesty with the radeonhd driver:
rbmorse: # If you have edited this file but would like it to be automatically updated
rbmorse: # again, run the following command:
rbmorse: # sudo dpkg-reconfigure -phigh xserver-xorg
rbmorse: Section "Device"
rbmorse: Identifier "ATI X1950XTX"
rbmorse: Driver "radeonhd"
rbmorse: EndSection
rbmorse: unquote
rbmorse: Oops...wrong item in clipboard. Anyway, you get the idea