Nightwulf|work: hi all
shepherd: Can any devs help me with a bug?
shepherd: actually, Be right back.
rindolf: Hi shepherd
navygator: hi guys
navygator: any guys can help me?
adamk_: Only if you ask your question.
navygator: excuse for my bad english
libv: navygator: that doesn't matter, just ask your question
navygator: I have a video card ati radeon hd3470 on notebook asus
navygator: i installed radeonhd drive
navygator: on kubuntu
navygator: but don't started on resolution 1440x900
navygator: system start on 1280x800
navygator: when i start systemsetting change automatic at 1440x900
navygator: any idea?
udovdh: navygator, just configure a resolution or let ubuntu know a default resolution (I can/could in Fedora)
navygator: defaul resolution
navygator: resolutio I want is in the list
navygator: but sistem start on bad resolution
SR_: tries to remember who she was talking to the other day, about the hanging when using Xv after suspend+resume
libv: SR_: has this been filed as a bug yet?
libv: SR_: it must have been either ndim or yangman
SR_: umm, someone else said they were going to try and reproduce it, I think it was yangman
SR_: I tried to google for the relevant IRC logs but failed
SR_: because they thought it was the same as an issue they were having
libv: SR_: on www.radeonhd.org
SR_: libv: I found /some/ logs, but was trying to google for things I know I said
libv: ok... well, i am currently too swamped with suse work to go and try sadly
SR_: no no, it's fair enough, I wasn't trying to cause trouble...
libv: nah, it's no trouble, just need to make sure that it will not fall off the planet completely, so file a bug please
weissj: can someone suggest steps for debugging a suspend hang with radeonhd on a thinkpad t60, fedora 10? I am not 100% sure the video is what's hanging, but that's usually the culprit.
libv: weissj: same as SR_ was talking about a few minutes ago?
weissj: libv: sorry i just joined, i missed that
weissj: can someone pastie it for me or something
SR_: libv: ah, I needed to google for "SR_", not "SR", because the latter matches "SR||" but not "SR_"
SR_: http://www.radeonhd.org/?page=archive_display&c=radeonhd&m=11&y=2008&d=2008-11-27 < was this
SR_: I'll try to write a bug report for it in the next few days
dancor: i have X1400 which has M54 chipset. is R500(RHD) the column for me for radeonhd on http://www.x.org/wiki/RadeonFeature ? and R500 is what works in the radeon driver?
ndim: dancor: yes, and yes, and R500 is what works in the radeonhd driver.
dancor: thanks, sorry for noobery
ndim: should really test his xorg-x11-drv-radeonhd F-10 packages on his own T60.
dancor: also rocking a T60
SR_: *heart* her T60
weissj: SR_: i think i'm having the same suspend problem as you. i havea T60 with X1400. when i suspend the "moon" led flashes and it just stays that way. screen backlight is still on but screen is blank
SR_: weissj: um, that's not my problem
SR_: mine suspends and resumes fine
ndim: dancor, SR_: Your Fedora radeonhd package maintainer is still using F-9 on his T60. I won't be able to update or test until some time within the next few weeks.
SR_: but using Xv after resume causes the laptop to lock hard
SR_: (not responding even to Alt+SysRQ+B)
weissj: SR_: you using F10?
ndim: weissj: So, it is a "suspend" problem, not a "resume" problem? Then it is new...
SR_: no, Debian Lenny, with the radeonhd packages from experimental
weissj: ndim, yeah it hangs on suspend
weissj: where's a good place to look to figure out what's happening?
ndim: SR_: I haven't looked in a few weeks, but the Debian packages have not set any records for being current during the last few months.
ndim: I just remember a hang-on-resume bug on F-9 when the update from 2.6.26 to 2.6.27 was going on.
SR_: ndim: heh.
SR_: ndim: releases? or git versions?
SR_: ii xserver-xorg-v 1.2.3-1 X.Org X server -- AMD/ATI r5xx, r6xx display
ndim: SR_: If the packages are not tracking git, they are outdated.
SR_: (II) RADEONHD: version 1.2.3, built from dist of git branch master, commit 57aca005
SR_: ^ no idea how old that is, but that's what I'm running.
ndim: git describe
ndim: That is 73 commits behind.
SR_: the reason I asked originally rather than filing a bug, was because I suspected my packages were old, but didn't see the point in compiling from source if the issue was either A: not fixed, or B: not radeonhd's fault (eg, my fault)
ndim: SR_: Might have been fixed in radeonhd, or might be something in the kernel/dri part which is possibly relatively old as well
SR_: [ 83.540004] [drm] Initialized drm 1.1.0 20060810
SR_: [ 83.562633] [drm] Initialized radeon 1.29.0 20080528 on minor 0
SR_: kernel 2.6.26-1-amd64
shepherd: Hello people... I have a problem, when adding option "DRI" to section "device" of my xorg.conf I get the error: "(EE) RADEONHD(0): RHDDRMCPBuffer: drmDMA returned -9" upon restarting the X server. I'm using rs690m with 126.96.36.199-117.fc10.i686 kernel
yangman: SR_: 57aca005 is the tagging for 1.2.3. A resume fix went in at 29cd73824 6 commits later
Arrow: Hi all
Arrow: anyone here that can answer a question?
ndim: Yes. Done.
Arrow: I'm trying openSUSE 11 here
Thargor: Arrow: Best way to find that out would probably be to ask that question
Arrow: I have a 3870x2 card
Arrow: it loads the radionhd driver but I get errors...(ee) no video memory found and (ee) screen(s) detected but not usable
Arrow: I don't know what version of radionhd shipped with 11.0
Arrow: but it not working even on one gpu
Arrow: I have a Kubuntu loaded on a harddrive..and I have to use a vesa driver...but that atlest works
Arrow: is there some sorta howto on the wiki about the 3870x2 or later 4870x2?
Arrow: well that's my issue :/
yangman: Arrow: pastebin your Xorg.0.log somewhere. that'd be more useful
Arrow: sure I was just trying the suse live cd and it just burped ...I'll go and try and save it to something
Arrow: dli hey
shepherd: um.. im getting "(EE) RADEONHD(0): RHDDRMCPBuffer: drmDMA returned -9" when adding option "DRI" to my xorg.conf, Ive got a rs690m, any ideas why?
Jayku1: Hey, just checking... for R620 (HD 3450) the tv-out still isn't supported with radeonhd?
yangman: Jayku1: it's not supported by any of the OSS ATI drives
Jayku1: ugh. And thanks.
libv: Jayku1: there was a lot of noise about it working early on, but that was rather bogus and based on a very limited result.
Jayku1: Yeah, I did some searchign and thought it should work ...
libv: since then, quite opposite to the early statements, ATI has put tv way back on the importance list
Jayku1: but I've blown a few hours on it, and I can't make it work at all
libv: Jayku1: with enough knowledge and perseverence it should be possible to dump registers from fglrx with a working mode and to make it work in your case
Jayku1: and the fglrx stuff works on the tv, but it's like ... black and whit
libv: Jayku1: are you sure you wired up everything correctly?
libv: Jayku1: are you sure you're not mixing composite and s-video?
Jayku1: Reasonably sure ...
Jayku1: but no
Jayku1: The fact is, it has 7PIN (composite) out
Jayku1: so I'm still wrestling
libv: s-video is a pair of signals, luminance and chrominance. composite is a single signal with everything encoded in one
libv: if you swap one for the other somewhere along the line, you will get black and white
libv: this goes from hw config, to connector, to cabling to even the input on your tv
libv: and even the setting of your tv
Jaykul: Well, I had the cabling working with a different card
libv: ok, then that can be ruled out
libv: this 7 pin thing can carry full component video, but afaik, you get those yellow adapters to composite with most ati graphics cards
Jaykul: I got a saphire ...
yangman: there's a splitter from the 7-pin out to component and s-video
Jaykul: and the adapter only has the rgb
Jaykul: i have another adapter lying around that has component + s-video, but it's a 9Pin
Jaykul: crazy annoyance
yangman: 7-pin is kinda unusual. I always figured they were all 9-pin
libv: Jaykul: the really crazy stuff is from a driver pov, knowing which of those wires is actually connected and then driving it accordingly
yangman: or maybe it's the other way around
yangman: my 4850 is still on backorder along with my LGA1366 bracket :(
Jaykul: libv: yeah ... I've no desire to rewrite the driver :-/
Jaykul: almost ready to install XP and buy BeyondTV or some such.
libv: after digging around in our "gathered connector box" i find a 7 pin to composite/s-video connector
libv: so they do exist
libv: there's even 8 and 9 ones in here
Jaykul: Maybe I'll go have another look in my spare cables box
libv: the stumps are all 8
libv: the s-video + composite one is 9: the one with 7 is a cable that goes only to composite
Jaykul: yeah, see ... that's what I've got
Jaykul: strange thing is... the manual shows an adapter with composite + s-video
Jaykul: *sigh* (this is the sound of me, loosing faith in AMD/ATI)
yangman: FWIW, stuff like this is left up to the manufacturer
Jaykul: yeah, but not the (sucky) drivers
Jaykul: nor the part-time support for OSS
Jaykul: do you know, does fgrlx not support xranr?
yangman: no idea. I haven't used fglrx in over a year
Jaykul: well, the whole point of this box is the tv-out, so ... I guess I gotta
libv: Jaykul: well, i would've loved to have been able to point you to a register doc so that you could try to dump fglrx registers and play with them inside the radeonhd driver, but no rv620/635 register specs were made available
Jaykul: Hmm. really?
libv: yeah... 610/630 was the last
libv: i don't even know how much of rv620 we got under nda here
libv: but i know that there was nothing released to the public, which kind of goes in against the whole thing
Jaykul: Bah, I just turned the fglrx back on, and now I remember what made me switch in the first place ... I had gotten the fglrx to show on tv out, but I couldn't get it to come back to the monitor :-p