M_M: please, anybody knows if I can make the hd-2700 with the radeonhd driver?
libvde: never heard of that one :)
M_M: it is mobile :)
M_M: and as far as I know, pretty new one
M_M: till now I have no choise but ussing the binary drivers :(
libvde: right... hrm, released in december
libvde: do you have a pci id for this card?
Tigerchen: can someone point me to the bugzilla? i have found a VERY strange behaviour with the newest version
libvde: Tigerchen: what's happening?
Tigerchen: starting an urxvt with -depth 32 results in a sig11 to the x-server
M_M: 01:00.0 VGA compatible controller: ATI Technologies Inc Unknown device 9583 (this one???)
libvde: RHD_DEVICE_MATCH( 0x9583, RHD_M76 ), /* Mobility Radeon HD 2600 XT */
libvde: M_M: so our driver should work
Temujin_: M_M: update-pciids ;)
M_M: i will give it a try again... last time was one month ago and it didnt work.
libvde: M_M: ok... what failed?
M_M: uff... I dont have the logs anymore, but basically didn't recognize the card
M_M: and please, something I don't understand well... what is this device?
M_M: 01:00.1 Audio device: ATI Technologies Inc Unknown device aa08
libvde: the mobility you have its id was added on 2007-10-15
libvde: M_M: hdaudio
Temujin_: HDMI audio port
libvde: M_M: goes to hdmi
M_M: I seeeee...
M_M: Well... I suposed... but not sure about it
Temujin_: M_M: what distro are you using?
Temujin_: ah, i know there;s an update-pciids command to fill in the unknown devices in Debian/Ubuntu, but I'm not sure if there's a Gentoo equivalent
M_M: yeap, the same command
M_M: this is after it: 01:00.0 VGA compatible controller: ATI Technologies Inc Unknown device 9583 and 01:00.1 Audio device: ATI Technologies Inc RV630/M76 audio device [Radeon HD 2600 Series]
M_M: so, new driver is installed... i'll give it a try
Temujin_: senses another satisfied customer
Tigerchen: so pug filed
M_M: so.... it is... "working".. but there are some distortion lines in the screen
M_M: and from console to X (alt + F7) part of the screen goes black and distorted for some seconds
M_M: direct rendering is not working either
Tigerchen: M_M: dri doesn't work at all atm
M_M: yes, I was reading that right now
Tigerchen: OpenGL renderer string: Mesa GLX Indirect
Tigerchen: and this is r500 here
M_M: but this line... what could be the problem?
M_M: lines I mean...
M_M: it is like out of sync in some parts of the screen
M_M: do you think maybe updating mesa?
libvde: Tigerchen: your bug is finally going to get some attention today :)
libvde: well, after the rather horrible issue yaneti is seeing is debunked :)
egbert: Tigerchen: your segfault only happened after the update?
Tigerchen: egbert: yes
egbert: Tigerchen: this is strange as we didn't touch anything in the code that's running at that time.
egbert: Tigerchen: would you be able to use git-bisect or something to track down where it was introduced?
egbert: could be that something is corrupting memory.
Tigerchen: i think it should be possible, if you aid me in using git-bisect
the-me: any news on this backpanel gamma issue on some notebooks?
egbert: Tigerchen: i haven't used it much either. but maybe i can help you.
Tigerchen: ok, first i should get this ssh to a real shell
egbert: Tigerchen: my collegues are forcing me to go to lunch with them.
egbert: can we do this in an hour or so?
Tigerchen: guten hunger ^^
egbert: emmes is starving...
egbert: libv isn't in yet.
libv: Tigerchen: booting up the m56 laptop to verify the new issue :)
Tigerchen: libv: ^^
Tigerchen: libv: verified?
libv: Tigerchen: egbert will take this one
libv: Tigerchen: please talk to him
Tigerchen: well, let's see when he'll be back from lunch
egbert: Tigerchen: i'm back
Death_Syn: yay lunch!
Death_Syn: er, sorry
Tigerchen: emmes: p i n g
Tigerchen: well, ehre were we
Tigerchen: egbert: we wanted to git-bisect right?
Tigerchen: keksdose-mobile xf86-video-radeonhd # git-bisect start -v
Tigerchen: won't bisect on seeked tree
Tigerchen: which was the first commit before the giant-block today?
libv: me adding ids to conntest
libv: that's the last we did upstream before the big batch went out :)
Tigerchen: ok, I'll search that for bisectiong
Tigerchen: libv: did you mean commit 252491e4a1b97b2ad020ebf45862d75c99da561b
libv: 00e1b0250 is the first of the new commits
Tigerchen: Bisecting: -1 revisions left to test after this
Tigerchen: ok better now
egbert: Tigerchen: i was just called out for a meeting. i was trying to reproduce your problem before.
egbert: so far no success...
egbert: Tigerchen: i cannot reproduce the problem here.
egbert: Tigerchen: did you get anywhere doing a bisect?
libv: Tigerchen: the restore issue is about ordering the restore apparently
libv: or at least the connector ordering
libv: if the VGA connector is listed first, then restore fails
libv: if the lvds connector is first, the restore is ok
libv: hrm... not about ordering the restore... it's about the CRTCs that get assigned to each output !
Tigerchen: egbert: I'm still going, since atm happyli skipping since many commits wont work
rmh3093: egbert, any dvi2hdmi updates :D
libv: rmh3093: should just work :)
libv: rmh3093: but you are probably talking about rs690 and the DDIA block :)
rmh3093: i have an r500
rmh3093: what ever u want to call it
libv: ah, right, you had some other issue than hdmi no?
libv: because hdmi as such is not available on r500
rmh3093: well radeonhd wouldnt detect my TV when it was connected
libv: no hdmi audio device there :)
libv: ah, right, that thing...
libv: we had two leads we could follow there
libv: rmh3093: let me get back to you on that one when i have found out what is wrong with the m56 panel restoration :)
rmh3093: ok.... im cant wait to get rid for fglrx for good ;)
Tigerchen: egbert: finished bisect
egbert: Tigerchen: ok, any result?
Tigerchen: egbert: opne of these, but a can't test them because they either produce a black screen (two of them) or segfault during startup
egbert: Tigerchen: hrm...
Tigerchen: need anything else? .Xdefaults versions of something -verbose 7 run-logs?
Tigerchen: why do i always stumble upon bugs noone else has
egbert: Tigerchen: your fate? :p
Tigerchen: next job: bug-finder ?
Tigerchen: empirically proven: every program containing more than 2 lines of code has at least 1 bug
udovdh: congrats with the new commits!
udovdh: driver still works for rv630 :-)
udovdh: even in dual monitor setup
egbert: Tigerchen: still around?
Tigerchen: egbert: yes
Obscene_X2: egbert could we define _RHDRegRead and _RHDRegWrite as macros in rhd.h and comment the functions out in rhd_driver.c ?
Obscene_X2: the macros would be
Obscene_X2: #define _RHDRegRead(scrnIndex,offset) *(volatile CARD32 *)((CARD8 *) RHDPTR(xf86Screens[scrnIndex])->MMIOBase + offset)
Obscene_X2: #define _RHDRegWrite(scrnIndex,offset,value) (*(volatile CARD32 *)((CARD8 *) RHDPTR(xf86Screens[scrnIndex])->MMIOBase + offset) =value)
egbert: Tigerchen: sorry. i'm distracted and i may have to run soon.
Tigerchen: egbert: np
egbert: Tigerchen: so you said you had the server segfault at startup on some of the commits in the list?
Tigerchen: egbert: yes, all but two of them
Tigerchen: the other two showed only a black screen and i had to blindly swich to console and reboot because i won't get a picture
egbert: Tigerchen: this was without starting urxvt? does it also happen when you start a 'naked' xserver?
Tigerchen: egbert: the segfaults came directly when X was started, before it even got a chance to display anything but a black screen
Tigerchen: ahh, got a log
yaneti: so there was a new Xorg snapshot in rawhide today, together with the latest radeonhd git, the result is quite different from yesterday http://www.declera.com/~yaneti/radeonhd/Xorg.0.log.13-2.00f182d6.randr