girrr: agd5f: I'm afraid the patch didn't solve my problem after all.
orkid: hmm. don't know if it's radeon or kde or what, but i start kde4.0.1 and i have no panel no anything 9right click menu works ok ). fading in of menus works too. i wonder why /
orkid: could it be the radeon driver? do i need an option for eXa in the xorg.conf ?
lucas: I'm having trouble setting up the S-video output of an ati X1200 with the radeon driver
lucas: is there some mailing list I could mail to to discuss this, or should I just ask here?
airlied: email@example.com or here..
airlied: we can usually help here...
lucas: ok, I'll try to collect the relevant info first
airlied: lucas: make sure you are using 6.8.0 and have a log from it sosmewhere .
airlied: lucas: or git master
lucas: see http://blop.info/pub/xorg.txt
lucas: and http://blop.info/pub/Xorg.0.log
lucas: X starts fine, but my TV says "no video signal"
lucas: (while it works in console mode)
lucas: I read somewhere that I have to enable the VGA output to get the S-Video output to work. not sure if it's true.
airlied: lucas: can;t see anything obvious.. but I'm not sure how x1200 tv out is. going.
airlied: lucas: agd5f will know more..
airlied: lucas: in theory you cna drop the modeline and the vga force on\
airlied: just having forcetvout + tvstandard should be enough
airlied: I gotta run.. zzz
lucas: agd5f: I'll stay around, just ping me when you have time to have a look
lucas: airlied: yes, I was using that with another ATI card, and it worked perfectly. but not with that x1200.
agd5f: lucas: I don't know that anyone has confirmed tv-out works on the rs690s
girrr: agd5f: I don't know if you have seen what I wrote earlier today, but I'm afraid I still have the gamma problem.
agd5f: girrr: hmmm. your rom looked fine. dac adjust values should by 0x6 and 0xb
girrr: agd5f: Did you see the image I pasted a URL to yesterday?
agd5f: girrr: I didn't see it, my connection must have dropped out when you posted it
girrr: agd5f: It was kind of early before you first got involved..
girrr: it's not a god picture but the test image has two grey fields that are not visable on the monitor on the right
agd5f: and it gets fixed when not using the bios values?
girrr: Yes if I just return false everything works like a charm
girrr: also switching the cables makes the problem to appear on the other monitor and so on
agd5f: girrr: what chip do you have again?
girrr: M18 JN (Radeon Mobility 9800)
girrr: it's part of a Dell Inspiron 9100
agd5f: that's right. I'll put together a patch now
girrr: agd5f: thanks
lucas: agd5f: I wouldn't mind helping debug it, if it's needed
agd5f: lucas: sure. I'll need to look into it more first though
agd5f: girrr: just committed. Option "DefaultTVDACAdj" "TRUE"
agd5f: girrr: I'm not sure what to make the default yet, bios or driver provided
girrr: agd5f: Here we talk about the TV DAC but I only use a VGA and a DVI output, did I miss something?
agd5f: girrr: the TVDAC drives bot tv-out and secondary VGA
girrr: agd5f: I see
girrr: agd5f: It seems to work now, thanks again.
agd5f: girrr: np
PSYCHO___: is there any way to force wine starting games on VGA-0 not on LVDS, think that both screens is utopy
agd5f: PSYCHO___: make is aware of xinerama hints?
PSYCHO___: i have it without xinerama as one large virtual
PSYCHO___: throught xrandr
agd5f: PSYCHO___: xrandr exposes xinerama hints so apps know hot to maximize, etc
PSYCHO___: agd5f: oh ok, so problem is in the wine :(
PSYCHO___: i'll rather stick to sauerbraten then :)
PSYCHO___: could conky cause laggy gl output?
PSYCHO___: cause when i kill it lags disappear
cryos|laptop: Has anyone spotted the radeon r300 driver ignoring color/material changes when drawing triangles?
cryos|laptop: I am trying to debug and it only happens when I am using the open source radeon driver with dri. It just seems to drop the colour changes and draw all the triangles the same colour.
aneas: cryos|laptop, does it work with fglrx dri?
ajax: conky triggers several paths that aren't accelerated well
ajax: so yes, you'll be switching back and forth between software and hardware rendering.
cryos|laptop: aneas: Yes but that is so crashy I would rather work with a dodgy gl..
cryos|laptop: It also works on every other system I have managed to test.
cryos|laptop: It is possible it is a subtle error in our GL code but it looks a lot like a driver bug to me.
cryos|laptop: Just playing with driconfig...
cryos|laptop: It works in software rendering mode too.
aneas: does not sound like a mesa/gl bug
cryos|laptop: Where would the bug be then? I thought it was in the r300_dri stuff.
cryos|laptop: I guess I should try to code a very minimal example to trigger the bug.
aneas: im afraid i cant answer that :/ (because of stupidity, not security ^^)
cryos|laptop: I don't think I have the time right now. It is on an amd64 laptop.
aneas: just didnt want to sound like "thats top secret"
cryos|laptop: You didn't sound that way. It would appear to me the bug is in the r300 dri code but I can't see how to debug it from there.
cryos|laptop: I can work around it for now. I will see about testing 2.6.25 soon...
cryos|laptop: Most stuff is working pretty well, just when rendering GL_TRIANGLES and changing the colour for the vertices... Strange.
Magnade: i dont think a newer kernel would help
cryos|laptop: I think that is where the r300_dri code lives isn't it? In the radeon module.
cryos|laptop: is still learning his way around the new radeon drivers.
Magnade: the kernel just contains the drm which relays info to the card
Magnade: the issue your seeing i would think would be in the r300 driver code in mesa
cryos|laptop: Yes you are right. Still learning my way around...
cryos|laptop: Using mesa-7.0.2, I will see about getting a snapshot to see if it has been fixed in trunk.
cryos|laptop: Thanks for the tips.
rift: the driver radeon can 1600x1200 ?
rift: when i try, i have a "desktop" biger than the screen and xrandr -q say i'm on 1600x1024
rift: it's my xorg config file
Magnade: could try setting virutal to 1600x1200 and see if that helps
agd5f: rift: what modes does xrandr say is available? (output of xrandr)
rift: $ xrandr -q
rift: Screen 0: minimum 320 x 200, current 1280 x 1024, maximum 1600 x 1200
rift: VGA-0 connected 1280x1024+0+0 (normal left inverted right) 352mm x 264mm 1600x1200 70.0 65.0 60.0
rift: 1600x1200 is available
rift: doesn't work with virtual
Magnade: your xrandr output said current res is 1280x1024
Magnade: thats why your desktop is bigger than your screen
rift: now i'm on a correct desktop
Magnade: current 1280 x 1024
rift: now my scren is in 1280x1024 and the desktop too
rift: but later the screen was on 1600x1024
rift: and i dont know why
rift: i have set 1600x1200 on my xorg conf file
Magnade: try removing the modeline you set in your config
rift: the same
rift: i think i have try all the solution :(
Magnade: paste your Xorg log
agd5f: rift: xrandr --output VGA-0 --mode 1600x1200
rift: out of range
rift: the display doesn't like, i dont know why
rift: for the log
agd5f: rift: does your monitor support 1600x1200?
rift: yes of course
agd5f: xrandr --output VGA-0 --mode 1600x1200 --rate 60
rift: i think it's workin
rift: but my wm is too small
rift: yes it work
rift: but howto do this at start ?
agd5f: rift: set Option "PreferredMode" "1600x1200" in teh monitor section of your config, but it looks like you are using xserver 1.3
agd5f: so that probably won't work
rift: yes i'm using 1.3
rift: maybe if i put this command on .xinitrc ?
rift: i will try to install xorg-server 1.4 from pkgsrc/wip
rift: i have put the command on the xinitrc and it works fine
rift: thanks :)
aneas`: ajax, am i right you made a huuuge commit 3 hours ago, or am i just misreading what git tells me?
airlied: aneas`: where? possibly agd5f did it
aneas: airlied, http://gitweb.freedesktop.org/?p=xorg/driver/xf86-video-ati.git;a=commit;h=651fe23f4c650ed91843dec48db24e18e8b91219
airlied: aneas: thats a merge..
aneas: 157 files changed
airlied: aneas: he took the wrong road in git..
aneas: so not a big deal?
orkid: i'm' having problems with git radeon and kde4.0.1.. i have fading, but no kdesktop or plasma toolbar or sanything. no background... i don't know if it's a radeon issue or a kde issue. how can i narrow it down ?
orkid: do i have to choose exa or xaa in xorg.conf or are defaults ok (with extension composite loaded)
orkid: i'm' using an R420
agd5f: orkid: xaa is the default
orkid: and it should work with the current git ?
agd5f: the driver should work
agd5f: what problem are you having?
orkid: nothing in kde4.0.1 after enabling desktop effects. i can only use the right click menu.. but no background 9no kdesktop), etc.. no k menu or panel..
orkid: when i right click and goto logout, i get a window popup, but nothing written in it appears
orkid: but i can see outlines where things should go.. it's really odd.
agd5f: orkid: are you using exa or xaa?
orkid: maybe i'll try a different compositing manager
orkid: i guess xaa, since it's the default, i don't choose one in my xorg.conf
orkid: ktip start too and everything in it is fine (like daily tip.. if you don't know). ... it seems like trasparency works too.
orkid: weird. i just go the background an panel start
orkid: but i click on the kmenu, and the menu does not show up, just the background 9translucent black)
orkid: i clicked it a few times, and got it to show up once. hmn. maybe a kde problem
orkid: i'm almost sure it is.. it's random, and i get some odd messages in the console
orkid: well . prolly not 'random' but you know
funda3: mesa from git keeps failing at radeon_screen.c for me, any ideas on when it might work?
funda3: the problem is with the r200: radeon_screen.c:717: error: ‘RADEON_PARAM_FB_LOCATION’ undeclared
funda3: nm, it worked after getting libdrm from git too :) ignore me
zog-vsl: anyone able to get 4 monitors running on 2 dual head pci-x radeon cards? I can get one or the other but not both