lamikr: Hi, does anybody know should the AT HD 3200 be supported by the radeonhd drivers?
tigerchen: should be i think
Zhenech: tigerchen, did you see the updated brightness patch?
tigerchen: nope, but since the normal one "WORKSFORME" (tm) i don't need to
Zhenech: the second version works for me, even without suspend :P
tigerchen: my subversion won't build somehow it tries to use icc
lamikr: tigerchen I tried to install latest mandriva 2008.1 rc2 which contained radeonhd-1.1.1-0-20080314 drivers and they give me error that no display device found.
lamikr: I tried in x86_64 architecture.
Zhenech: lamikr, any logs?
lamikr: Zhenech I found out /var/log/Xorg.9.log which has following lines in the end following
lamikr: (II) Primary device is: PCI 01:05:0
lamikr: (--) Assigning device section with no busID to primary device
lamikr: (EE) No devices detected
lamikr: Fatal server error:
lamikr: no screens found
lamikr: Thats all...
lamikr: Earlier the log lists a lot of radeon devices like RV620: Radeon HD3400, but I could not find HD3200 mentioned in there at all.
Zhenech: could you post the WHOLE log somewhere?
Zhenech: anyways, off for an hour or so, need to go shopping
tulcod: lamikr: rule 1 of irc: don't ever, EVER consider pasting more than 2 lines in chat. rule 2 of irc: stick to rule 1.
lamikr: tucod No pasteing, I typed them with keyboard from another computer because also the atheros network card has some problems... :-(
tulcod: lamikr: usb stick?
tulcod: a floppy?
tulcod: I mean, just transfer the hdd, for god's sake :p
tulcod: lamikr: also, fix your network first, then
lamikr: tucold yes, usb stick is the way I can get it. :-) I am working with network card problem on mandriva forum
tulcod: not tucod, or tucold
tulcod: lern @ read
tulcod: lern 2 reed
lamikr: xorg conf coming but giving first the "lspci -v" pasted for Mandriva by using usb stick tranfer metod :-) http://mandriva.pastebin.com/m1b3a7f12
rmh3093: egbert_away, ping
rmh3093: libv, ping
lamikr: btulcod Zhenech Xorg.0.conf: http://mandriva.pastebin.com/m4e1db938
tulcod: tucod, tucold and btulcod
tulcod: none of them spelled correctly
libv: rmh3093: hi :)
|_emming: could someone explain me the following error message: (II) Attempted to read BIOS 128KB from /sys/bus/pci/devices/0000:02:00.0/rom: got 60KB
|_emming: i have got 2 2600xt cards installed and am using them in a multi-gfx setup
|_emming: (without crosslink cable)
|_emming: so there are 2 xservers on the same machine
libv: could this be the kernel messing up? because this vga rom reading through the kernel is not really that great
lamikr: tulcod I will learn :-)
|_emming: this error happens only on the second gfx, the primary one works fine
|_emming: even with fglrx and DRI
tulcod: lamikr: ah, good, finally! :)
|_emming: but DRI on two heads simultaneously seems to be not possible
|_emming: kernel is 22.214.171.124
|_emming: dri is of course not loaded
rmh3093: libv, did you or egbert get anywhere with that dvi2hdmi issue I had
libv: rmh3093: no, not yet
libv: and i'm currently stuck with a rather nasty unichrome bug that has been sitting in my bugs list for 8months already
libv: so no radeonhd time for me atm
rmh3093: nice u work on that driver also :)
rmh3093: i will have to pester you when i start working on my carpc again
libv: rmh3093: i cut my teeth on unichrome btw, -unichrome is largely my own code :)
rmh3093: just the X driver or do you work on the kernel/fb driver
Zhenech: |_emming, dri on two heads as in "on two displays and same xserver" or as in "two xservers on same card"?
libv: rmh3093: x driver was enough work already, especially since i pioneered quite a few things in there
libv: rmh3093: there is no halfdecent fb driver for unichrome afaik too
|_emming: Zhenech: well, as one head/xserver with DRI for each card
Zhenech: ah, you have two cards
|_emming: i head that it is possible to have dri on more than one card, but it seems to fail if it is two cards of the same kind like in my setup
|_emming: Zhenech: even the fglrx kernel module detects and sets up just one of the cards
|_emming: ..for dri
|_emming: whatever :P
Zhenech: afaik it's just linux' drm being bad designed and needs a rework (which is worked on now afaik=
libv: Zhenech: we do not depend on the drm yet
|_emming: ah.. i thought it could be a hardware problem related to the 790X chipset or something
|_emming: maybe not correctly POSTing the second gfx card
Zhenech: libv, yeah, but other drivers cannot do multi-master too
|_emming: doesn't fglrx use its own proprietary DRM implementation? so the one to blame would be amd
fabian__: Hey guys! Anyone there?
nakal: hi, im trying to compile xf86-driver-radeonhd on freebsd. i get: configure.ac:254: error: possibly undefined macro: XORG_MANPAGE_SECTIONS
nakal: any idea what to do?
ndim: nakal: RTFM
ndim: nakal: README
egbert_away: rmh3093: this dvi2hdmi thing is really nasty. i've seen reports of issues that might be related.
egbert_away: i haven't been able to reproduce a single one of them.
nakal: ndim: oh, thanks. i only looked at wiki. i thought that when 1.1.0 installs without problems, i wouldnt need any other dependencies for the latest snapshot
dmb: libv: you do good old mach64 stuffs also?
ndim: dmb: Not in this driver.
libv: dmb: used to, i kickstarted splitting off mach64
libv: dmb: was the only one with the dangly thingies who dared go in there :)
dmb: libv: i have one on an old ibook g3 ppc
dmb: drm/dri seems to fail though, but thats another story
dmb: tigerchen: can you post on that bug that you still experience the issues (so egbert doesn't think it was just me :D)
libv: dmb: yeah, never went near that on mach64 :)
dmb: libv: surprisingly, i had it work before with an earlier driver, it worked surprisingly well
dmb: was able to play a snes emulator
libv: before the whole pci-id cleanup and split-off?
libv: have you filed a bug for this btw?
dmb: nope :D
dmb: don't know if i should even bother, don't know if anyone actually uses them accept as toys
libv: george sapountzis did a lot of work there over the last few years
dmb: it just says something like cannot create rgb visual
udovdh: I started glxgears
udovdh: and the desktop hang
dmb: no other error messages
dmb: udovdh: with what driver?
udovdh: on 2600pro
udovdh: could ctrl-alt-nkspace
udovdh: mousepointer was movable
dmb: ah, at least it isn't a hard lock
udovdh: can Xorg.0.log tell something?
dmb: doesn't know
drac: Certain GL screensavers from xscreensaver distribution does same here (radeonhd, 2600pro)
drac: 1.1 release to be exact..
udovdh: I use the git version
udovdh: but I remember it ran
udovdh: or am I mistaken?
libv: drac: it probably has nothing to do with the driver, but more with some mesa software path that's buggy
libv: drac: what happens if you try the same thing with vesa?
udovdh: it runs ok on my epia
libv: udovdh: with indirect rendering?
udovdh: how do I know that libv?
drac: libv: good point. testing, sec.
dmb: udovdh: glxinfo
dmb: i think
udovdh: OpenGL renderer string: Mesa GLX Indirect
udovdh: # glxinfo |grep -i render
udovdh: direct rendering: No (If you want to find out why, try setting LIBGL_DEBUG=verbose)
udovdh: OpenGL renderer string: Mesa GLX Indirect
drac: ok that was phun, hardlocked..
drac: with vesa, too
dmb: drac: so i guess its not the driver :D
dmb: should file a mesa bug i guess
drac: with vesa + xscreensaver "Fireflies" and Mesa 6.5.2.. yeah it's not the driver, just my horked old "distribution stable" system
dmb: isn't mesa 7.0.3 the latest?
dmb: goes to class
drac: should try with newer Mesa, I guess
rmh3093: egbert, let me know if u can think of anything i can do to help figure this out
rmh3093: im not afraid of making src code changes
rmh3093: if you have ideas
tigerchen: dmb: will do
egbert: rmh3093: ok, sounds great!
egbert: i've got a few ideas what to do. so i might send you some patches. but not today.
dmb: wow, lots of commits
dmb: egbert: any of those commits related to that bug?
dmb: (besides commiting the mc thing for 515)
egbert: rmh3093: https://bugs.freedesktop.org/show_bug.cgi?id=15105 might be related to your issue. although it's a different chipset.
egbert: dmb: you mean the 'color issue'?
Zhenech: teh new firefox ssl warnings are annoying
dmb: egbert: well, libv said it was the same issue as the color issue
dmb: where when i switch to the vt sometimes its horizontally corrupt
egbert: dmb: i don't think it's related really.
dmb: egbert: should i file a different bug then?
egbert: this issue is not only transient, is it?
egbert: well, the original bug is still there.
Zhenech: egbert, libv, did anyone of you had time to look at the patch from http://bugs.freedesktop.org/show_bug.cgi?id=13984?
egbert: it doesn't go away after a few seconds.
dmb: nope, it stays until i switch to x then back
egbert: Zhenech: i will asap. i had planned to do it today.
egbert: it's great.
Zhenech: egbert, ah, fine
Zhenech: it works fine with xbrightness
Zhenech: however, I would love it more if radeonhd would not block the HW-keys (as vesa or radeon do)
dmb: egbert: also, sometimes in x i see a simular issue where i can see the scrambled lines when typing
dmb: i still think it has to do with the memory restore stuff
egbert: Zhenech: in general i don't like the system bios doing things behind the driver's back.
egbert: it would be nicer to receive an acpi event and deal with this from the driver.
egbert: dmb: so far i failed to play any of the movies.
dmb: i can try to convert it
dmb: its some format my cell phone outputed
dmb: vidbug2.mpg: ISO Media, MPEG v4 system, 3GPP
Zhenech: egbert, that works fine on different laptops with different chips, and it's "just" the brightness
egbert: Zhenech: i've got an hp laptop with an intel chip which locks solid as soon as i close the lid.
egbert: the firmware action on the lid switch does something that causes this.
Zhenech: how that? bad acpi-foo on lodclose
egbert: i haven't had time to investigate.
Zhenech: that's fine here
egbert: Zhenech: no acpi event is ever seen.
Zhenech: lid close shutsoff lcd, and sends acpi (to trigger suspend)
egbert: appearantly it happens entirely transparently :(
Zhenech: lid open even resumes from suspend
Zhenech: without any hacks
egbert: Zhenech: lucky you ;)
Zhenech: thinkpads <3
Zhenech: a friend of mine has a samsung with intel, works as fine
airlied: Zhenech: pot luck.
airlied: Zhenech: my thinkpad corrupts the video on lid close unless we disasble it
Zhenech: airlied, saw this on early radeonhd versions, and some time later on radeon, but nothing like this for current gits
dmb: egbert: converted it to theora, http://output.da4.org/vidbug1.ogg and http://output.da4.org/vidbug2.ogg
Zhenech: and your t60p/m56 should be that different from my z61m/m54
airlied: Zhenech: as I said its pot luck.. the BIOS can do what it likes, its easier to just switch it off.
egbert: dmb: ok, so what i see is line images flashing between the text mode lines.
dmb: i think so
egbert: it's less a color issue as far as i can see.
egbert: dmb: thanks for converting btw.
dmb: so should i create a separate bug?
egbert: dmb: leaving this one open should be fine.
Soul_keeper: "AMD plans to axe off 1,600 employees before the close of September -- this would represent a 10% reduction in AMD's current workforce of 16,000 employees."
dmb: Soul_keeper: less fgrlx development :/
GerbilSoft: something screwed up with my laptop's video card (FireGL V5200)
GerbilSoft: the VGA port doesn't seem to output any working HSync/VSync signal, but the RGB lines and DDC work
GerbilSoft: is this a driver error, or should i get the system board replaced
airlied: GerbilSoft: you should probalby test with fglrx or windows.
GerbilSoft: did, same problem :(
airlied: sounds like hw problem then
dmb: GerbilSoft: does vesa and friends work?
dmb: does console work?
GerbilSoft: the internal LCD works fine
GerbilSoft: it's the external VGA port that's broken
GerbilSoft: and vesa as far as i know won't let me switch to VGA
yangman: GerbilSoft: that's odd. I'm having the same problems. I chalked it up to hardware failure,but haven't had it repaired yet
GerbilSoft: weird thing is, it was working yesterday
yangman: it's been... a few months
yangman: yeah, mine went from working to not working in the span of a reboot
GerbilSoft: so i wonder, is it a hardware malfunction or a bad register somewhere
GerbilSoft: (is there any way to "zero" the registers on the GPU?)
yangman: registers should be reset properly when you reboot, though
yangman: have you tried setting the BIOS to boot with the VGA? that's the one thing I haven't tried
GerbilSoft: yeah, i've tried
GerbilSoft: no image with VGA+LCD or VGA Only
airlied_: turn if off, atke the batery out?? running out of idea.s
GerbilSoft: will try that later
GerbilSoft: working on something else right now
yangman: the kicker is, monitor detection works fine
yangman: so, there's IO. just no image
airlied_: I'[ve nodea how H/V would break though.
airlied_: maybe there is some static issue.
GerbilSoft: yangman: yeah
GerbilSoft: i tested it via a voltmeter
GerbilSoft: RGB lines measured ~0.9 V with a white scren, ~0.1 V with black
GerbilSoft: HV measured 0
yangman: that's interesting
yangman: so, there's signal, but the monitor is not being modeset
GerbilSoft: so the monitor doesn't bother doing anything
yangman: I'm beginning to think there's a quirk in the HW that only manifested under radeonhd that the windows drivers or fglrx knows about
yangman: s/know/don't know/
airlied_: yangman: hence why I'd suggeset removing power and battery
airlied_: and leaving it off for 5 mins
GerbilSoft: i'll try that now
yangman: airlied_: well, at least in my case, the laptop's cold-rebooted several times
airlied_: yangman: no use.. laptop has battery.
yangman: I have a bad habit of draining the battery accidentally
airlied_: yangman: you need to fully remove it
airlied_: the residual may hold up parts of the GPU.
airlied_: even if it can boot the machine
airlied_: can't .
yangman: well, lets see what GerbilSoft comes back with
GerbilSoft: unfortunately that didn't help :(
airlied_: damn sounds like busted ass hardware
GerbilSoft: i'll leave it unplugged overnight once i'm finished homework
yangman: bummer. that's too bad
yangman: hm.. I'm gonna see if the VGA port on the dock works...
yangman: hm.. nothing
GerbilSoft: does DVI work on the dock?