Home Phoronix Phoronix Forums X.Org Videos From FOSDEM 2008

Radeon IRC Logs For 2009-11-12

Search This Log:


happycube: airlied - awesome... (too bad it took 8hrs tho)
airlied: glisse: T60 resumes!! it has other issues b it does resume
airlied: glisse: I noticed if I resumed, unloaded radeon, ran atomtools, loaded radeon it all worked
airlied: then I enabled kernel atom debugging and it all worked
airlied: then I bisected kernel atom debugging down to the one printf that made it all work
airlied: realised stupid IIO MC register reading has issues, the printf delay was making it all work
airlied: no idea if the patch is 100% correct but really its not a fastpath I don't think
happycube: gah, timing bugs :P
airlied: EruditeHermit: ping new test patch on the bug
airlied: just a guess coming from fixing the other resume bug
EruditeHermit: airlied, on it
EruditeHermit: airlied, do you want me to get drm-next kernel and patch it?
Nightwulf|work: hi all
dileX: hi Nightwulf|work
Nightwulf|work: hi dileX
soreau: hi dileX, Nightwulf|work
dileX: hi soreau
MostAwesomeDude: hi soreau , dileX , Nightwulf|work
soreau: hi MostAwesomeDude
Nightwulf|work: hi soreau , MostAwesomeDude
dileX: soreau: compiling 2.6.32-rc6-git5 (incl. drm-linus)
dileX: hi MostAwesomeDude
soreau: dileX: Nice :)
dileX: MostAwesomeDude: yesterday, I was discussing about binutils-gold on #grml
soreau: MostAwesomeDude: Any gallium revelations lately?
dileX: MostAwesomeDude: says "A common problem seems to be calls to XOpenDisplay without directly linking to -lX11..."
dileX: MostAwesomeDude: with mesa is "binutils-gold-compatible"
dileX: MostAwesomeDude: shall I send that patch to ML?
MostAwesomeDude: soreau: Not particularly. Been working on some things here and there.
MostAwesomeDude: dileX: Looks good to me, but I have *zero* say in that stuff. Send it to ML, see what they think.
dileX: MostAwesomeDude: the patch is from 2009-05-08 07:27 and IIRC I talked with some guys of debian-x team about it. should look into my backlogs. but anyway, this record on debian QA gives me hope, that this patch was correct.
MostAwesomeDude: dileX: Well, again, you really need to get it to ML.
dileX: hehe, OK
dileX: Nightwulf|work: you recognized what we achieved :-)?
EruditeHermit: hey, I have a question about using the wiki drm-next instructions
Nightwulf|work: dileX: you talk about the 3D abilities for r6xx/r7xx?
EruditeHermit: is the part in "alternatively use drm-next as is" the whole kernel?
dileX: Nightwulf|work: no. much simpler.
Nightwulf|work: dileX: then I perhaps didn't follow the channel properly...what do you mean then?
dileX: Nightwulf|work: it tooks us some hi and hi that others followed :-)
Nightwulf|work: dileX: ahh...ok ;-)
dileX: EruditeHermit: drm-next is a "complete" 2.6.31 linus-tree with drm and especially drm-radeon-kms fixes
dileX: currently, the latest fixes get into drm-next first
glisse: airlied: sweet
EruditeHermit: dileX, I am trying to apply this patch to drm-next. But it seems there is no radeon_combios.c to patchhttp://people.freedesktop.org/~airlied/scratch/combios_readback.diff
dileX: EruditeHermit: come on, thats only two lines :-) change it manually.
EruditeHermit: ah I figured it out
dileX: how did you apply?
EruditeHermit: patch just couldn't find the file
EruditeHermit: it existed
EruditeHermit: specified the location manually
EruditeHermit: don't know why it couldn't find it
dileX: EruditeHermit: whats that patch fixing?
EruditeHermit: just something to test for resume
EruditeHermit: resuming from suspend
dileX: hmm, power-mngt (so long I wanted to test that on my box)
dileX: CONFIG_DEBUG_KERNEL=y
dileX: CONFIG_TIMER_STATS=y
dileX: to follow with powertop
dileX: soreau: if you wanna compile linus-tree: apply
soreau: dileX: Thanks but I don't think I have a need to
soreau: I only have this lowly rv350 :)
glisse: airlied: btw i got a patch to add default 800x600 in face of broken edid
glisse: i will send it to you
glisse: it's attached to bug report
airlied: EruditeHermit: if it applies to whatever kernel you have that'll do
airlied: glisse: do we not do that?
airlied: if we find no modes and its connected we add 800x600
airlied: we are thinking of changing it to 1024x768 at some point
EruditeHermit: its building
EruditeHermit: has been for an hour
EruditeHermit: =p
EruditeHermit: slow machine
airlied: EruditeHermit: cool, no hurry, I'm burned out after fixing one laptop ;-)
EruditeHermit: airlied, seems like the changes were simple
EruditeHermit: one liners
airlied: its always only one line
EruditeHermit: lol
EruditeHermit: airlied, that must be frustrating
airlied: I reckon if I didn't get access to that hw today I'd be still tracking it down in a month
airlied: or glisse would
EruditeHermit: airlied, how did you get access to it?
EruditeHermit: basically I should just mail my laptop to you
EruditeHermit: =p
airlied: EruditeHermit: he just happened to be visiting RH Brisbane for a meeting this week
EruditeHermit: haha
EruditeHermit: al well
EruditeHermit: have to sleep
EruditeHermit: will let you know how it goes in the morning
EruditeHermit: gnight
glisse: airlied: i don't think we do that
glisse: at least i didn't see where it's done
glisse: well what happen if i am not mistaken is that we return connector disconnected if edid is broken
glisse: https://bugzilla.redhat.com/show_bug.cgi?id=533632
airlied: drm_add_modes_noedid
airlied: it gets called if we are connected and have no edid
airlied: or modes
glisse: airlied: thing is it get all only is connector return status connected
glisse: but we don't
glisse: get call
airlied: what do we return?
glisse: disconnected
airlied: if its disconnected there should be no modes
glisse: default ret value
glisse: well it's not disconnected, it's connected but with wrong edid
airlied: I've tested that code on my monitor with broken edid before
airlied: and it always worked
airlied: maybe it an issue in dual-head configs
glisse: airlied: well it doesn't work anymore
glisse: i got report with single config monitor
glisse: i might missunderstand the code
glisse: but from what i grasp
glisse: if edid is null and there is a monitor we return disconnected
glisse: and edid is null if it's broken
glisse: my patch doesn't seem to add mode to non connected output
airlied: ah since we changed how we get edid
airlied: I don't think that patch is the correct way to fix it
glisse: what should i do then ?
glisse: just report connected with no mode
glisse: note that i add mode only if ddc_probe probed a monitor
glisse: so there is somethings connected but it hasn't valid edid
airlied: yeah I think conected with no mode is right
airlied: disconnected is definitely wrnog
glisse: ok will update
airlied: the problem on DVI is we don't know if analog or dig is connected unless we load detect
airlied: glisse: I tink it shuold call the load detect fallback path
airlied: when DDC respsonds but EDID is invalid
airlied: the problem is I've seen on box get DDC response when no monitor exists
glisse: yeah somethings i missed
airlied: stupid servers
glisse: i guess it's somethings designed for windows
airlied: granted the response was all garbage
airlied: another thing we should do is fixing the edid header and redoing the checksum like fb
airlied: dinner &
glisse: airlied: so the dvi detect code is ok
glisse: it was the vga which was wrong
airlied: glisse: oops
glisse: i am testing faking broken edid
glisse: of course my monitor doesn't support 800x600
glisse: airlied: what it's valid to have set_base called with a null fb ?
scarabeus: http://dev.gentoo.org/~scarabeus/render-issue.png <- i have slight issue with current xf86-video-ati head on r600
scarabeus: any ideas if the issue is Qt or the driver?
glisse: it's driver
glisse: so far i never reproduced it
glisse: i don't know why my kde refuse to break i need to retest
taiu: scarabeus: what's you chip and is it dri1 or dri2?
glisse: agd5f, airlied: why do we enable scaler on any lcd output ?
glisse: shouldn't we enable scaler only when the resolution is not the native screen resolution ?
scarabeus: RV670 ;; dri1, no kms :]
glisse: agd5f, airlied: thought on patch at: https://bugzilla.redhat.com/attachment.cgi?id=369204
Zajec: glisse: in "Add default 800x600 mode when getting bad EDID" you dropped printing drm_get_connector_name(connector), is that printed line before or sth?
glisse: yeah will print connector
eosie: hi
Zajec: glisse: do you have some code locally to wait for engine idle? (kms)
Zajec: glisse: eventually could I copy this from KMS?
Zajec: glisse: want engine to be idle to reclock it
glisse: Zajec: suspend path of each asic has the code
glisse: basicly take the cp lock then call the asic function
glisse: Zajec: each asic should have it's pm function
glisse: i think best is one function for each asic in charge of pm
glisse: of course some asic might share a common function
glisse: r100/r200, r300/r400, rv515/r520, ...
Zajec: glisse: will think about it...
Zajec: glisse: so many details about PM there are, blah
Zajec: reclocking, idle, VBLANK, irq, callbacks...
Zajec: timer, schedules
Zajec: :P
Zajec: glisse: should I call resume myself after calling suspend?
Zajec: or will that be called by other code parts?
glisse: don't call resume
Zajec: glisse: ok
glisse: the code for idle is in resume function
glisse: well in each resume function
Zajec: glisse: should I lock rdev->cp.mutex? i can see there is also rdev->cs_mutex, but I guess i should just use rdev->cp.mutex, right?
Zajec: hope it's last question :)
glisse: cp_mutex
Zajec: cp_mutex? in kms?
Zajec: err, moment
Zajec: glisse: did you mean cp.mutex?
glisse: yup
crash2k_: hello?
crash2k_: anyone here?
biotube: you have a problem?
crash2k_: no a question :P
biotube: ask it
crash2k_: is there a tool to change the desktop color settings?
Terman: It looks like there is one outdated docuemnt on http://www.x.org/docs/AMD/
Terman: r600isa.pdf is version 0.31 (may 2007), the version on the AMD site is 0.35, (dec 2008)
crash2k_: huh?
glisse: Terman: yeah such thing happen
biotube: crash2k: multiple conversations happen
biotube: though I'm not sure what color schemes have to do with the driver
crash2k_: no i mean like gamma adjust
biotube: xgamma
Zajec: xgamma --help
crash2k_: thanks
crash2k_: anything for saturation?
Terman: glisse: should I open a bug about that or how can this be fixed?
bridgman: Terman; we kept the older version up because it seemed easier to understand, all the new versions are on the stream site
bridgman: we're going to set up some pages that point to all the relevent areas
bridgman: for docs/code etc
bridgman: the old version of the isa doc might not be needed any more, will check
crash2k_: is there a gui tool for xgamma?
Terman: bridgman: ah, OK. I keep all versions anyway, so I indexed the file with the version and kept it
agd5f: glisse: I think we probably need two vars. rmx_type and rmx_enabled since users can select full/center/aspect/none
glisse: agd5f: i tried to cope with that
glisse: i set rmx off a connector creation and only force to full if there is no other rmx type selected
glisse: thus user selection is not overwritten
Zajec: would like to just see "1920x1080Scaled" mode for PANEL :|
Zajec: solution for dumps, like radeonhd
glisse: agd5f: also should this properites show up in xrandr ?
agd5f: glisse: they should
agd5f: Zajec: kms supports scaled modes. they just don't have "scaled" in the name
agd5f: that's the only way to get non-native modes on lvds
Zajec: agd5f: but it does not offer 1920x1080 for my 1600x900 PANEL, even if I connect my DVI monitor which is 1920x1080
agd5f: Zajec: ah, downscaling. It's trivial to implement, but generally confusing to most users
Zajec: hm, confusing? don't see what can be confusing about this :)
ajax: "why is my text unreadable"
Zajec: also don't remember any report from confused report on radeonhd
Zajec: ajax: it's the same with upscaling
ajax: no, that's "why is my text fuzzy"
Zajec: yeah...
Zajec: both confusing for me ;)
Zajec: well
Zajec: not for me
Ghworg: would just like to have 4:3 modes have the correct aspect on his 16:10 panel
agd5f: Ghworg: already supported. set the scaler mode to aspect
Ghworg: agd5f: Ooh, excellent!
Ghworg: I do that through xrandr?
Zajec: maybe we could standarise that aspect property
agd5f: Ghworg: yes xrandr --outptut --set scaler aspect
agd5f: Zajec: we did for kms I think
Ghworg: agd5f: Thanks
Ghworg: X Error of failed request: BadName (named color or font does not exist)
agd5f: Ghworg: using kms?
agd5f: does xrandr --verbose show the property?
Ghworg: agd5f: Yes, with uptodate git as of yesterday
Ghworg: xrandr --verbose | grep -i scal scaling mode: Full
agd5f: some of the kms properties don't show up in xrandr for some reason.
agd5f: Ghworg: ah, it's xrandr --outptut --set "scaling mode" Aspect
Ghworg: agd5f: Great, that seems to have worked
Zajec: agd5f: i thought standarized properties are perfired with _
roysjosh: on f12 rc, has anyone else had a blank screen after kms kicks in? I can see kernel debug until [drm] messages appear, machine is still up and I can ssh in
agd5f: roysjosh: make sure fbcon is loaded
roysjosh: agd5f, I think it is built in? Module fbcon not found
roysjosh: yep, there's a message in dmesg about it
boris64: Hi folks, does anybody know how i can get the names of my monitors/connectors? I'd like to set custom resolutions via kms and the video= bootparam, but nothing like video=dvi-0:1280x1024 etc. seems to work
agd5f: boris64: they are in sysfs
Ghworg: Okay, xrandr doesn't produce an error. But the 4:3 modes are still stretched into widescreen unfortunatly
agd5f: Ghworg: what chip?
Ghworg: agd5f: m76
Ghworg: agd5f: Or rv630 depending what log I look at :-)
roysjosh: boris64, ls /sys/class/drm/
roysjosh: and remove the card#- prefix
agd5f: Ghworg: should work. file a bug, and I'll take a look
Ghworg: agd5f: Will do
Wood_Warrior: Hi there, does xorg supports KMS driver for radeon x1200 ?
agd5f: Wood_Warrior: yes
glisse: agd5f: i don't think drm take any of the video parameter into account, does it ?
boris64: roysjosh: Do you have an exact address? I didn't find something useful (yet?).
agd5f: glisse: airlied committed code to handle that
roysjosh: boris64, when I do: "ls /sys/class/drm" I see "card0 card0-LVDS-1 card0-SVIDEO-1 card0-VGA-1 ..."
glisse: agd5f: oh missed that
Wood_Warrior: i instaled latest libdrm, mesa and fx86-video-ati from cgit, also kernel 2.6.32-rc6-git5 , but get mess of colors when startx
roysjosh: my dmesg is here, if anyone wants to look: http://pastebin.ca/1668293
boris64: roysjosh: hm strange, i only see "card0","ttm" and "version"
boris64: roysjosh: stupid me
boris64: roysjosh: i just booted without kms, so i'll be back. Thx for help
Wood_Warrior: looks like the problem comes from xf86-video-ati - with older versions no mess of colors, but not working xvideo
Wood_Warrior: with latest cgit version mess of colors, and xvinfo gives something like "textured video"
agd5f: Wood_Warrior: make sure your xf86-video-ati is built with kms support
Wood_Warrior: how can i do it ?
agd5f: Wood_Warrior: see the topic
boris64: airlied: if i disabled one monitor per video=DVI-I-1:d bootparam, is there a way to put it back "on" so i can use it again in X?
boris64: the story behind: i got 2 monitors with different resolutions and i'd like to use the one with max res in framebuffer (second is "off" :d) but i'd also like to use both screens in X
agd5f: boris64: once X has started you can use xrandr
Wood_Warrior: so all i am interested is not kms, but "xvideo overlay" - "man radeon" says that RS6xx supported only 2D and "textured video", but not overlay xvideo and 3D - is it still the case ?
agd5f: Wood_Warrior: all chips support 2D, 3D, and Xv
Wood_Warrior: its my bad english - i ask does the driver supports it for my card radeon x1200 (RS690M)
agd5f: Wood_Warrior: yes
boris64: agd5f: that's what i thougt, but i can't set any modes on this panel anymore ("xrandr: cannot find mode 1280x1024"). Is there an "on" parameter in xrandr i didn't see?
agd5f: boris64: xrandr --output --auto
boris64: agd5f: hm, still no go. xrandr --output DVI-1 --auto gives no output and the monitor stays black. Do i have to restart X for that?
agd5f: boris64: no
agd5f: what does xrandr output?
agd5f: just run xrandr
boris64: "DVI-1 disconnected (normal left inverted right x axis y axis)"
boris64: well, plus the DVI-0 stuff
agd5f: boris64: and it works if you don't specify the video stuff at boot?
boris64: yes, both monitors run at their max resolution then but the framebuffer looks kinda ugly as it is limited to the resolution of my low-res monitor
Wood_Warrior: agd5f: but glxinfo says "direct rendering: Yes", but "OpenGL renderer string: Software Rasterizer" , how can i get hardware ?
agd5f: boris64: sounds like it's breaking something. file a bug
boris64: agd5f: mkay, i'll do that. Thx for your time ;)
Wood_Warrior: do you think the way is tuning more fine ./configure options befeore compilations of libdrm, mesa, and fx86-video-ati ?
agd5f: Wood_Warrior: what are you trying to do?
AstralStorm: any leads on -vo gl:yuv=4 failure?
Wood_Warrior: enable hardware xvideo overlay
AstralStorm: esp. how to debug what causes it (I still bet on the DRM)
AstralStorm: bisection will take far too long and be unwieldy
roysjosh: ok, here's a diff between the output of avivotool regmatch \* ... nokms=nomodeset and working, kms=kms and blank screen. http://fpaste.org/bBTh/
AstralStorm: I now know that it's not mesa
roysjosh: anyone know what those registers do, and is it safe to regset them?
Wood_Warrior: agd5f: i am trying to enable xvideo overlay to get better performance on watching HD videos
agd5f: Wood_Warrior: your chip doesn't have an overlay useable for Xv, so Xv is implemented via the 3D engine
agd5f: Wood_Warrior: the overlay and textured video should perform equally well
agd5f: AstralStorm: I'd say it's a mesa issue unless you can bisect some other commit. but the ddx and drm aren't invovled in shader compilation
agd5f: AstralStorm: try make realclean and rebuild mesa
agd5f: roysjosh: what kernel are you using? Have you tried airlied's drm-next or radeon-testing branches?
roysjosh: agd5f, f12 rc almost latest: 2.6.31.5-122
AstralStorm: agd5f: the mesa is always built clean
AstralStorm: I did a bisect with the help of my package manager (just give it a commit id and it will checkout it and install)
AstralStorm: and always the same error
glisse: roysjosh: do you a bug ?
AstralStorm: sometimes followed with error -22 as in some error in the pipeline
AstralStorm: in dmesg
roysjosh: glisse, no
AstralStorm: something about packet3, I have the pastebin with that one still
roysjosh: glisse, shall I open one on bugs.fdo?
agd5f: AstralStorm: pastebin the dmesg with that error
AstralStorm: I bet it's a bug in drm pipeline validator
AstralStorm: ok
AstralStorm: I've pastebinned the error itself, hope it's enough
AstralStorm: http://dpaste.com/119002/
AstralStorm: this one
glisse: roysjosh: if it's against f12 do it at bugzilla.redhat.com
AstralStorm: that's with older version of mesa
agd5f: AstralStorm: you need a newer mesa
AstralStorm: but still very recent
roysjosh: glisse, ok
AstralStorm: the latest still fails though
AstralStorm: and older fail too
AstralStorm: so I'm in catch22
agd5f: it's a different error
AstralStorm: can't bisect
AstralStorm: not with this kernel
AstralStorm: which leads me to attempt at bisection with an older kernel
AstralStorm: where... latest mesa doesn't fail!
agd5f: AstralStorm: if you are using the lastest mesa, you won't get that error
AstralStorm: yes, I haven't
agd5f: so the problem is elsewhere
AstralStorm: but with older kernel, new mesa... worked
AstralStorm: so wtf
AstralStorm: the problem is likely in drm
AstralStorm: in shader validator
agd5f: AstralStorm: there is no shader validator in the kernel'
AstralStorm: the older kernel has drm-next from 2009-10-10
AstralStorm: agd5f: so whatever it's called
AstralStorm: the part that actually runs these things
agd5f: AstralStorm: what's the actual mesa error
agd5f: with latest mesa
AstralStorm: there's none
AstralStorm: but it fails to run the fragment program
agd5f: and no drm error in dmesg?
AstralStorm: yup, no error there
agd5f: then it's a mesa problem
agd5f: mesa is failing to compile the shader for some reason
AstralStorm: but why does it work with older drm-next?
AstralStorm: ^
AstralStorm: as in, new mesa, new ddx, older drm-next (in the kernel) = work
AstralStorm: new kernel = fail
AstralStorm: which means the bug is *somewhere* in drm
AstralStorm: between 2009-10-10 and today
agd5f: AstralStorm: so bisect the drm
AstralStorm: yeah yeah
AstralStorm: and that's free in my time :)
AstralStorm: why don't someone else with r6xx do that ;)
AstralStorm: *won't
glisse: roysjosh: give me bug # once you done
roysjosh: glisse, ok, almost done
AstralStorm: let me try the old kernel
AstralStorm: with just drm-next update
AstralStorm: it might be some... interference
AstralStorm: bbl
roysjosh: glisse, https://bugzilla.redhat.com/show_bug.cgi?id=537140
AstralStorm: agd5f, you're right, it is a mesa error
AstralStorm: now bisecting on an older kernel to give some meaningful results
glisse: roysjosh: damm you are sadly an owner of rs200 chipset
roysjosh: glisse, yep :\ old laptop that's still kickin
glisse: roysjosh: what is the laptop model ?
roysjosh: glisse, compaq presario 2500 (2598us)
roysjosh: glisse, if you need any other info, lspci, dmidecode, etc, let me know
glisse: roysjosh: i will get back to your bug tomorrow
roysjosh: glisse, ok, thanks
glisse: i am on another one today
glisse: i think reg dump should be helpfull
glisse: and you did attach them so i should have enough info to get started
agd5f: roysjosh: can you dump regs with radeontool?
roysjosh: agd5f, I did, it's all in the bug: https://bugzilla.redhat.com/show_bug.cgi?id=537140
agd5f: avivotool isn't relevant for your chips. different hardware
roysjosh: ok
agd5f: roysjosh: ah, ok. I only saw the avivotool
happycube: blah... the hp dv2-1110us could've had hybrid crossfire if they'd used a newer northbridge. why do amd notebook offerings so often have to fail?
agd5f: roysjosh: does glisse's rmx patch help? https://bugzilla.redhat.com/attachment.cgi?id=369204
roysjosh: agd5f, I think the _SCALER reg was 0 for me
roysjosh: I can try it and see I suppose
agd5f: roysjosh: remember the avivotool regs are irrelevant on your chip
agd5f: glisse: btw, I still think we need rmx_type and rmx_enabled. since we want to maintain the user's preference, but disable it when the requested mode is native
glisse: agd5f: yeah i was going through it again and you right
speps: hey guys what's the gallium radeon development status?Is it possible build mesa with the enable-gallium-radeon flag?Does it support glsl?Thanks
MostAwesomeDude: GLSL support is there but not perfect.
MostAwesomeDude: You can use --enable-gallium-radeon to build Gallium libs.
MostAwesomeDude: I should probably write up a wiki page on how to load and test it.
Dr_Jakob: MostAwesomeDude: do you export GLSL on all r300 cards or only on r500?
MostAwesomeDude: On all cards.
Dr_Jakob: but r300 can't do GLSL
Dr_Jakob: well not all of it.
MostAwesomeDude: I'm still waiting for keithw to split PIPE_CAP_GLSL into a bit more detail.
MostAwesomeDude: r500 can't do all of it either, but that's okay.
Dr_Jakob: oh?
Zajec: MostAwesomeDude: ++howto_pls
Dr_Jakob: whats missing from r500?
MostAwesomeDude: Yeah, there's no way to call subroutines in the frag shader.
Dr_Jakob: Ah so you need to inline them
MostAwesomeDude: And the vert shader call/ret stuff is kind of flaky; I don't think nha has it working.
agd5f: glisse: I'm thinking we store the user's preference in radeon_encoder->rmx_type, and what we actually program in radeon_crtc->rmx_type
MostAwesomeDude: ...That tab-completed
MostAwesomeDude: nha: Ping.
nha: greetings
Dr_Jakob: salut!
nha: sorry for my lack of communications (which may remain flaky through november, as I have a kind of very important deadline coming up)
glisse: agd5f: that would mean we need to check in crtc fix mode rather than in encoder fixmode
nha: and no, I haven't worked on vertex shader call/return stuff
nha: I thought the GLSL compiler inlined that stuff?
agd5f: glisse: yeah, just gotta do it in one place
MostAwesomeDude: Yeah, it should inline those. I expect that in Gallium we'll have to either request inlining or PIPE_CAP it.
MostAwesomeDude: Anyway, could you take a quick peek at the top two commits of my draw-buffers branch? http://cgit.freedesktop.org/~csimpson/mesa/log/?h=r300g-draw-buffers
glisse: agd5f: you are working on a patch right now ?
nha: sure
glisse: i am done with another thing so i could do it
MostAwesomeDude: I think I've got it down and I just need to fix up the actual MRT emit.
agd5f: glisse: yeah
glisse: agd5f: ok i let you on it and i will eat :)
glisse: MostAwesomeDude: i think r500 support calling subfunc
glisse: in ps
glisse: thought long time i haven't read the doc
MostAwesomeDude: glisse: I don't remember seeing that code, but I could look again. I thought it only had branching.
glisse: MostAwesomeDude: page 110, flow control
nha: MostAwesomeDude: in the debugging commit, why did you translate the macro constants into magic numbers?
MostAwesomeDude: glisse: I stand corrected.
nha: MostAwesomeDude: the actual commit looks pretty sweet and pretty much exactly what I had in mind
nha: MostAwesomeDude: but I'd appreciate it if you changed the spaces to tabs in your changes where applicable
MostAwesomeDude: nha: There aren't SHIFT macros for those yet, but I could add them.
MostAwesomeDude: And yeah, I had tabs off. :C
glisse: MostAwesomeDude: JUMP_FUNC
MostAwesomeDude: glisse: I see. JUMP + PUSH or POP.
glisse: well for push pop we have to use temp reg
nha: you can push the return address, but of course there's no stack for local variables or parameters
glisse: yeah it's all a compiler issue
MostAwesomeDude: Icky, icky. Well, still good to know.
MostAwesomeDude: nha: Thanks for the comments, will clean all that up.
glisse: my guess is that best is to inline most of the functions as so far most of glsl shader are split into small function with no complex call pattern
MostAwesomeDude: Agreed.
Zajec: glisse: could you check once more my requesr for help on drm/radeon/kms, engine clocking?
agd5f: glisse: sent a new patch to the list
glisse: agd5f: will take a look
soreau: agd5f: Did you ever get my message WRT dumping regs from fglrx to get S-video 1024x768 timings?
glisse: Zajec: did you try to launch the time from radeon_init rather than from debugfs?
soreau: agd5f: Basically, if it is possible, would you be able to provide me with a complete list of necessary regs to grab required?
glisse: agd5f: your patch looks good
glisse: Zajec: really try moving the timer outside debugfs callback
Zajec: glisse: i tried putting this in int radeon_pm_init(struct radeon_device *rdev)
Zajec: but maybe i didn't try with tohether with my mdelay and for loops...
Zajec: can try
glisse: what is your gpu already ?
Zajec: M82
Zajec: RV620
glisse: doing nothing in the timer callback is fine
glisse: by nothings somethings just like printing a register
Zajec: glisse: i tried reading engine clock in timer and it works fine
Zajec: glisse: however switching from getting (and printing) to setting locks up
glisse: does it works with : msecs_to_jiffies(500)
glisse: instead of 50
glisse: shouldn't matter
glisse: just remembered how this thing works
Zajec: glisse: want me to try this?
Zajec: didn't try yet
glisse: no it's useless
glisse: won't change anythings
Zajec: i'm out of ideas :|
glisse: Zajec: to debug in pm_reclock do the following :
Zajec: listens to
glisse: http://pastebin.com/m5331677a
agd5f: soreau: you want the CRTC, PLL, and TV regs. I don't have a list off hand. basically all the regs in radeon_legacy_tv.c in the drm or radeon_tv.c in the ddx
glisse: Zajec: you should be able to see those 2 regs printed by monitoring log through ssh
Zajec: glisse: ok, thanks, will check after supper
mems: Hey y'all. Can anyone confirm whether it's possible to use two dual-head cards in the same machine (i.e. xorg with 4 monitors)?
agd5f: mems: should be with xserver 1.7.x or kms
mems: agd5f: cheers -- so you reckon it won't work on earlier versions?
agd5f: mems: should work with xserver 1.4 and earlier, just not 1.5 or 1.6
mems: thanks
mems: presumably this is with the radeon driver?
mems: or is this in general?
agd5f: mems: both in theory
mems: cool
AstralStorm: hello again
AstralStorm: seems that my "problem" is neither caused by mesa nor by drm
AstralStorm: hmm
AstralStorm: can ddx cause this?
_moep_: funny. I've a new radeon card (now 4670) and radeon works fine but with my older radeon card only vesa works
Diablo-D3: hey guys
AstralStorm: nah, that shouldn't happen, hmmm
Diablo-D3: does radeon or radeonhd support any 3D acceleration on mt radeon 4850?
Diablo-D3: *my
soreau: AstralStorm: Which problem are you having?
soreau: Diablo-D3: Yes, both do (radeon preferred) with latest code. See the link in the topic
soreau: _moep_: Post the log from the problem session?
_moep_: soreau: yeah I open a bug report but for months nothing happens
Diablo-D3: soreau: how bad is it?
soreau: Diablo-D3: What do you mean?
AstralStorm: soreau: mysteriously now not working -vo gl:yuv=4
AstralStorm: in mplayer
Diablo-D3: soreau: the support
soreau: Diablo-D3: It's in experimental stages, 3D works though
AstralStorm: technically, it *might* be some mplayer bug
soreau: AstralStorm: Do you have direct rendering working and are you using kms/dri2?
Diablo-D3: AstralStorm: yuv=4 uses a shader
Diablo-D3: I think
soreau: AstralStorm: The preferred method would be -vo xv in any event
Diablo-D3: 4: Use a fragment program with additional lookup.
Diablo-D3: Needs the GL_ARB_fragment_program extension and at
Diablo-D3: least four texture units.
Diablo-D3: soreau: xv uses a crappy scaler though
AstralStorm: checking now w/ glxgears
AstralStorm: glxinfo is ok
soreau: AstralStorm: glxgears is meaningless. Use glxinfo|grep renderer
AstralStorm: gears work
AstralStorm: told you it's ok ;)
AstralStorm: OpenGL renderer string: Mesa DRI R600 (RV670 9504) 20090101 TCL
soreau: Yea, use -vo xv
AstralStorm: (that's w/ downgraded mesa for a test, fails too... it didn't before)
AstralStorm: -vo gl w/o yuv=4 works too
AstralStorm: so, wtf does it fail
soreau: In what way does it fail>
mems: anyone know of a table of ATI card marketing names mapped to the chips they have on them?
soreau: mems: There is a wiki page providing that information
MostAwesomeDude: mems: Wikipedia has such a table.
mems: thanks
mems: heads to various wikis
AstralStorm: soreau: http://omploader.org/vMnI3cw
soreau: mems: http://en.wikipedia.org/wiki/Comparison_of_AMD_graphics_processing_units
_moep_: should I enable KMS or not?
soreau: _moep_: yes
_moep_: ok
soreau: AstralStorm: Im sorry, but I am not going to look at that. I will assume the information Diablo-D3 provided is mostly correct
mems: soreau: thanks
mems: MostAwesomeDude: cheers
AstralStorm: Diablo-D3: ...
AstralStorm: yes, and renderer is working
AstralStorm: and GL_ARB_fragment_program is available
AstralStorm: also, RV670 has 4 texture units or more (actually, any fragment program fails)
soreau: AstralStorm: compiz water effect?
AstralStorm: no compiz here.
AstralStorm: anything else I can test that with that doesn't have a zillion of deps?
AstralStorm: http://dpaste.com/119788/ - the excerpt from the log
AstralStorm: some mesa test suite maybe?
adamk: AstralStorm, I believe that arbfplight from the mesa demos requires GL_ARB_fragment_program
AstralStorm: ok
AstralStorm: /usr/lib64/gcc/x86_64-pc-linux-gnu/4.4.1/../../../../x86_64-pc-linux-gnu/bin/ld: cannot find -lGLEW
AstralStorm: hmm, what's that lib?
AstralStorm: well, I can install that
AstralStorm: adamk: extra bonus: arbfplight works
AstralStorm: I'll test all arb there
AstralStorm: all 2 of them, obviously fragment shader one fails :)
AstralStorm: but, this sounds more and more an mplayer problem
AstralStorm: but why? how can I compile and run exactly that fragment program it sends?
AstralStorm: (it's a YUV converter, obviously, so I need some YUV texture)
_moep_: how can I enable KMS in the kernel? I've running 2.6.31.6
AstralStorm: _moep_: no need to, just toggle the radeon.modeset commandline option
chithead: _moep_: modeset=1 parameter to radeon
AstralStorm: the STAGING option just switches the default of it
Luzipher: _moep_: for in kernel, there is a Option in "Staging"
_moep_: Luzipher: ahh ok
Luzipher: _moep_: but as the others said, the radeon.modeset=1 parameter also works (put it in grub)
_moep_: can I use both?
_moep_: so kernel rebuild
AstralStorm: _moep_: yes, both
AstralStorm: but no need to again
AstralStorm: :/
AstralStorm: I can't find the program itself in that vo_gl in mplayer, have to dig deeper
AstralStorm: glGetIntegerv(GL_PROGRAM_ERROR_POSITION, &err);
AstralStorm: it uses that to get the error
AstralStorm: I bet mesa or gl headers fail there and wrong data type is sent
AstralStorm: judging from the numbers
AstralStorm: now, that's glproto and not mesa. nice goose chase
AstralStorm: and here I have glproto 1.4.10
AstralStorm: mplayer compares that error to -1
_moep_: AstralStorm: so I need to grub stuff,too?
AstralStorm: maybe the definition of GLint has silently changed
AstralStorm: _moep_: no, you don't if you toggle the STAGING option
_moep_: ok :)
AstralStorm: but why are you so averse to the command line option?
AstralStorm: really want to rebuild your kernel to toggle the default?
AstralStorm: now, if GLint has changed, that's damned ABI breakage
AstralStorm: whoever has done it, should Die In A Fire
_moep_: how can I test if kms works?
MostAwesomeDude: AstralStorm: You may be way off on the wrong track.
MostAwesomeDude: Does swrast work?
AstralStorm: how would it work? it doesn't do fragment programs.
AstralStorm: and I'm not way off track, since arbfplight works
AstralStorm: now, if only I could... coerce it to try and load the mplayer program w/o failing
AstralStorm: also, the bug is fairly recent
AstralStorm: something unrelated to mesa or drm caused it
AstralStorm: I bet on glproto
AstralStorm: new mesa rebuilt with newer glproto
AstralStorm: mplayer built with older glproto
AstralStorm: if it goes away with mplayer rebuild, I get to bomb whoever changed glproto in abi-incompatible way
AstralStorm: :/
AstralStorm: as old mesa + old drm now fails too
AstralStorm: it didn't before and it doesn't in my snapshot before major update
AstralStorm: so it stands to reason it's something entirely else ;p
MostAwesomeDude: AstralStorm: I mean, if swrast works, then it's a bug in the HW-specific code.
MostAwesomeDude: Try it. Set LIBGL_ALWAYS_SOFTWARE=1.
AstralStorm: ok, trying
AstralStorm: fails as well
AstralStorm: so, I'm about to kill some xorg maintainer
AstralStorm: ;p
AstralStorm: shouldn't have to rebuild all apps after glproto update
AstralStorm: now, I'll better check the build log since Oct 10 on this machine
AstralStorm: I need this unix timestamp translated into human-readable format: 1256442866
AstralStorm: how do I do that?
AstralStorm: oh, done that already, hmm
AstralStorm: 2009-10-25 04:54:26
AstralStorm: that's when mplayer has been built
AstralStorm: that's after glproto update, so that's out
MostAwesomeDude: idr was probably the last person to touch glproto.
MostAwesomeDude: But rebuilding gives you GLX 1.4 in DRI2 contexts. :3
AstralStorm: I'm in DRI1 now.
AstralStorm: still, mplayer is built w/ latest glproto
AstralStorm: and it wasn't failing back then
AstralStorm: so I have one extra thing to bet against
AstralStorm: libdrm.
AstralStorm: I've updated that.
AstralStorm: http://cgit.freedesktop.org/mesa/drm/commit/?id=6eafd1cf386d93bb9e34660227ca6f26aadfeb32 - does it smell like cheese?
AstralStorm: ok, I'll try downgrading libdrm to pure 2.4.15 and test
AstralStorm: if that's it, whoever wrote this patch gets my deepest gratitude
AstralStorm: and that's agd ;)
Ralesk: wonders if it's possible to refer to a xf86-video-ati commit not by its changeset ID
AstralStorm: yes, but it's highly tricky - only if it's before some tag
AstralStorm: alternatively, you can use shortened ids
Ralesk: something where I can use the @{date time} notation, so it's at least roughly incremental, so I can more easily create packages of it
AstralStorm: ?
AstralStorm: bad idea
AstralStorm: there can be many commits on a given date
AstralStorm: now, mercurial has a commit number instead, so does svn
Ralesk: because git keeps telling me something like "the log for branchname goes back only to whenever I inited my local repo"
AstralStorm: but this notion is repo-local
AstralStorm: if you add any commit in your tree, it is broken
Ralesk: I don't care if it's repo-local :)
Ralesk: I only check out
Ralesk: compile
AstralStorm: well, git doesn't support it by design
Ralesk: forget.
Ralesk: I know :/ that's why I thought, hey, if I could ask it to give me 6.12-branch's last commit before @{2009-09-08 00:00:00} then I'd be figgin happy :/
Ralesk: there is git rev-list --before=blah --max-count=1, but I can't tell it which branch I want to ask about!
Ralesk: otherwise it works nice
Ralesk: and for some reason the date notation just doesn't work at all in the local clone, no clue why.
Bigshot_: hi
Bigshot_: soreau:
soreau: yes?
Bigshot_: what to do next after clone
Bigshot_: ./autogen.sh --prefix=/opt/xorg && make && sudo make install
Bigshot_: ??
soreau: Is that what it says in the wiki?
Bigshot_: 1. clone 2. pull 3. autogen?
soreau: Because really, you should probably build libdrm configured with --experimental-radeon-api (or whatever the wiki says it is) first
soreau: Then build mesa, ddx and possibly even X against that libdrm
Bigshot_: soreau: from where to get xorg-ati-master?
Bigshot_: there is no link for it
soreau: what?
soreau: What is xorg-ati-master?
Ralesk: do you mean xf86-video-ati's master branch?
Bigshot_: yeah i pulled it
Bigshot_: * [new tag] xf86-video-ati-6.8.191 -> xf86-video-ati-6.8.191
Bigshot_: * [new tag] xf86-video-ati-6.8.192 -> xf86-video-ati-6.8.192
Bigshot_: * [new tag] xf86-video-ati-6.9.0 -> xf86-video-ati-6.9.0
Bigshot_: * [new tag] xf86-video-ati-6.9.0.91 -> xf86-video-ati-6.9.0.91
MostAwesomeDude: ...
Bigshot_: now what to do?
Ralesk: Bigshot_: I have a feeling you're better off using the git web frontend's download facility :)
Ralesk: http://cgit.freedesktop.org/xorg/driver/xf86-video-ati/log/ <-- click the revision and see the download links
soreau:
Bigshot_: linux-itkq:/home/Download/xf86-video-ati # zypper si xserver-xorg-video-ati Reading installed packages...Loading repository data...Source package 'xserver-xorg-video-ati' not found.Resolving package dependencies...
AstralStorm: Bigshot_: it's called xserver-xorg-video-radeon there
AstralStorm: like in most debian-derived distros
Bigshot_: nope still no
Bigshot_: it's call xf86-video-ati
Bigshot_: ?? still no
chithead: probably you best ask in a distro specific support channel how to create a package from latest git code
lks: Can I somehow use this new page-flip-ioctl to have tear-free-2D-window-moving?
MostAwesomeDude: No, for a couple reasons.
MostAwesomeDude: First, tear-free 2D comes from vsyncing EXA, which is done in the DDX already. (Maybe not for really new chips though, not sure.)
MostAwesomeDude: Second, you need to be full-screen before pageflipping works IIUC.
MostAwesomeDude: Third, this implementation only works for i915 right now; the radeon versions haven't been added yet.
Ralesk: AstralStorm: zypper is Suse :)
Ralesk: I found a solution to my checkout problem, three lines of git...
Ralesk: git init
Ralesk: git pull git://anongit.freedesktop.org/xorg/driver/xf86-video-ati 6.12-branch
Ralesk: git checkout $(git log --pretty=%H --before="2009-09-08" | head -n 1)
woodefec: hi
woodefec: maybe you will know furher. board: radeon 9200. Drriver: radeon. system: opensuse 11.2. hi. i'd been researching this for the past week. I have lost the 3D acceleration. First the circumstances: i used sax2 and activated the dual head mode for a while. Also i did install the radeonhd driver (or it was automatically installed). than I deleted that driver and did sax2 -r -m 0=radeon. now i have only the xorg-x11-video driver.
woodefec: i always used the radeon driver, worked on all suses, fedoras, ubuntus
Zajec: woodefec: maybe just delete xorg.conf
woodefec: now i have no 3D and running glxgears causes total system freeze
Zajec: woodefec: or mv /etc/X11/xorg.conf /etc/X11/xorg.conf.backup
woodefec: Zajec, i tried that
agd5f: woodefec: check your xorg.conf and see if sax added a huge virtual directive
[Enrico]: well imho atm radeon works better on r600 then r200 eheh (at least comparing archlinux radeon with gentoo git radeon eheheh)
woodefec: agd5f, http://pastebin.ca/1669007 here's me xorg.conf, where is the huge virtual directive?
Ghworg: agd5f: Filed bug about the scaling not scaling, 25063. Let me know if I've missed any essential info
woodefec: everything is looking fine in logs. xorgs, the glxinfo, but Warzone 2100, compiz, nexuiz, alienarena, googleearth don't work
[Enrico]: woodefec: glxinfo | grep rendering ?
[Enrico]: | gre render *
[Enrico]: sorry
agd5f: Ghworg: thanks
woodefec: direct rendering: Yes
woodefec: OpenGL renderer string: Mesa DRI R200 (RV280 5961) 20090101 AGP 8x x86/MMX/SSE2 TCL
[Enrico]: woodefec: mhm mine 9250 is an AGP 4
woodefec: ok, so it's set wrongly?
[Enrico]: woodefec: well you should know if it is a 4x or 8x
woodefec: i don't
[Enrico]: woodefec: anyway i have some problem too with r200. i have some screen corruption both with and wothout kms
[Enrico]: woodefec: let me check on wikipedia
[Enrico]: woodefec: according to wikipedia it ia an agp 8
[Enrico]: or pci
[Enrico]: but also the 9250
[Enrico]: but mine is agp 4
[Enrico]: couse the motherboard has an agp 4 only
[Enrico]: woodefec: you should check your motherboard specs
Zajec: woodefec: huge virtual
Zajec: Virtual 3840 1200
woodefec: ok, so what is going on :)
woodefec: it was always working
woodefec: Zajec, i played nexuiz on it, used all googleearth stuff et.
woodefec: but since i upgraded to opensuse 11.2 with the best and newest drivers its kaputt
AstralStorm: Ralesk: heh
AstralStorm: hack, but might work
woodefec: Zajec, what is huge virtual?
woodefec: agd5f, same :) ?
[Enrico]: woodefec: in your screen section you have Virtual 3840 1200
[Enrico]: woodefec: as agd5f suggested try to move your xorg.conf away and startx without it and let's see
woodefec: ok
woodefec: ill give it a go and get back to you
woodefec: th
woodefec: x
agd5f: woodefec: prior to kms and ttm, you have to pre-allocate a framebuffer large enough for dualhead if you want dualhead
woodefec: me back, now running on 1024 X 768
woodefec: renamed xorg.conf to xorg.conf.dupa
Ralesk: AstralStorm: yeah. Well, can replace the head call with --max-count=1. And it should work unless someone silently replaces the public git repo with another one with different commit history
[Enrico]: agd5f: btw if i disable kms on my r600 i can use Option "ForceLowPowerMode" isn't it?
agd5f: [Enrico]: yes
Ralesk: which I see as highly unlikely (and a reason to check our download scripts/variables anyway)
woodefec: [Enrico], what now, should i try some 3D stuff?
AstralStorm: Ralesk: it'll work then still
[Enrico]: good, well i don't want to loose kms but my is a bit hotter then before (not a lot but still....)
AstralStorm: but won't pull the right commit
[Enrico]: woodefec: well it is the only way to check i guess :D
Ralesk: AstralStorm: yeah, that's what I meant by not working :)
woodefec: ok, if i wont be back in 10 sec, itll mean the system froze
[Enrico]: woodefec: good luck
Ralesk: I don't see that happening with a high probability though
Ralesk: and this seems to be the only way to have an incremental version number with git projects unless they tag frequently :)
EruditeHermit: airlied, no luck with your new patch
[Enrico]: if i understand well from here http://www.x.org/wiki/RadeonFeature GLSL support will becamo avaible with gallium driver isn't it? is llvm still the preferred way (i read some times ago that someone was working on a llvm compiler, but generic, not driver specific)
[Enrico]: become*
MostAwesomeDude: [Enrico]: Nope, no LLVM stuff planned anymore. It turned out to be infeasible.
[Enrico]: MostAwesomeDude: oh i see, so it will be a per-driver compiler?
[Enrico]: woodefec: no luck i guess
woodefec: no. but diferent effect
[Enrico]: MostAwesomeDude: well thank you for the info, i hope GLSL will be avaible soon eheheh :D
woodefec: the monitor was cut off signal plus system freeze
[Enrico]: woodefec: wow, sounds strange
MostAwesomeDude: [Enrico]: No, the compiler's already in Mesa core.
[Enrico]: MostAwesomeDude: wow really?
MostAwesomeDude: Yes really.
[Enrico]: MostAwesomeDude: wait so since i'm using mesa git, am i using it?
woodefec: [Enrico], i am reinstalling the whole xorg packages
MostAwesomeDude: Yep.
biotube: [Enrico]: the gallium compiler targets an intermediate, so there is something of a per-driver compiler
woodefec: but i have official versions of all of them
MostAwesomeDude: The catch is that the driver might not do all the opcodes.
[Enrico]: MostAwesomeDude: wow that's explain the very good nexuiz performance i think eheheh, but glxinfo doesn't report it eheh that's why i'm surprised :D
MostAwesomeDude: But those are small driver bugs instead of big driver bugs.
[Enrico]: biotube: i see thanks :D
woodefec: no home cooked stuff, plain from opensuse 11.2 repository
MostAwesomeDude: [Enrico]: Only for Gallium, not classic.
[Enrico]: ohhhhh i see then
[Enrico]: eheheh
MostAwesomeDude: I may have to write up a wiki page explaining why GLSL support isn't such a big hurdle.
[Enrico]: i didn't enabled gallium since i'm on r600 and i think there is nothing for r600 now eheheh
[Enrico]: MostAwesomeDude: lol well it would be nice (and may be append it to the topic). i was just curious anyway, i don't use games normally eheheh, but you know may be some time :D
woodefec: so wtf 's going on with the drivers?
[Enrico]: woodefec: well let me say that, may be there is a bug in the packaged version of opensuse. why don't you try to update all xorg with the xorg repo?
[Enrico]: ok it can be dangerous but....
[Enrico]: woodefec: anyway my 9250 works both on fedora archlinux (and ubuntu ok :D )
woodefec: [Enrico], u mean the xorg repo from opensuse buildservice?
[Enrico]: woodefec: yeah that one
[Enrico]: woodefec: be aware if it b0rk your system don't came to me crying i've warned you
[Enrico]: :D
woodefec: http://download.opensuse.org/repositories/X11:/XOrg/ here's still no opensuse 11.2
[Enrico]: woodefec: so opensuse already provides the lastest?
woodefec: i guess so, they are quite reliable when it comes to the latest
[Enrico]: woodefec: well may be you can try this http://download.opensuse.org/repositories/X11:/Drivers:/Video/ ?
[Enrico]: woodefec: well with lastest i mean laste stable released by upstream
[Enrico]: woodefec: no don't try that one, there is only radeonhd and something else
woodefec: yes
woodefec: there is no xorg-x11-video
woodefec: xorg-x11-driver-video-7.4-87.88.1.i586.rpm - this is the provided version and the one i got
woodefec: [Enrico], it is in the official distribution repository
[Enrico]: no clue sorry
woodefec: so that means no 3D or go back to opensuse 11.1
woodefec: or
woodefec: downgrade that xorg stuff
woodefec: just added the opensuse 11.1 repo or X11
[Enrico]: woodefec: or switch to fedora eheheh
woodefec: [Enrico], i was there, too much hassle i found for me
[Enrico]: woodefec: really? can i ask why?
woodefec: [Enrico], matter of taste, i like the suse approach. YaST, ifup, ifdown, runlevels, build service, package management and plenty of high edge tools and software
[Enrico]: i see
woodefec: with ifup ifdown i also mean yast, there i set all network stuff with couple nice QT clicks
woodefec: kubuntu was disastrous for me. WTF's me frickin YaST??
woodefec: YaST works in cli - great usability. too much poking in scripts and text settings i wouldnt leave out time for me family
woodefec: ok, now some poking with xorg
woodefec: brb
woodefec: hey, still no 3D. But i found xorg is set to huge virtual
woodefec: how can i poke that?
woodefec: 3840 1200
woodefec: and [Enrico] will that help?
[Enrico]: woodefec: well as i understand huge virtual is the bad guy there
woodefec: so u opted for me to start without xorg.conf at all, what i did
[Enrico]: but well i don't use opensuse from a long time, so no idea how to remove it of xorg.conf is not present
[Enrico]: woodefec: may be don't delete xorg.conf but edit it and remove the line which set the virtual
woodefec: now i did a new xorg.conf with the command sax2 -r -m 0=radeon
woodefec: can i change the virtual
woodefec: and what is it for?
biotube: multiple displays
[Enrico]: woodefec: well i don't know but i guess it is for dual head things
[Enrico]: woodefec: but i _think_ you can just use xrandr
woodefec: whats xrandr?
woodefec: ok, brb
[Enrico]: woodefec: http://en.wikipedia.org/wiki/Xrandr
Yewbacca: I hope I don't invoke wrath by asking this. Have any of you heard anything about ATI Eyefinity coming to macs? I found it completely astounding that the official AMD pages talk about Windows and Linux, and doesn't even have the words Apple or Mac on the page. And googling? Try it, it's impossible to get a page that talks about it.
MostAwesomeDude: I don't even know what Eyefinity is.
Yewbacca: "ati eyefinity mac" gives 72k results, and none of which (I am on page 15) have talked about mac support. It's all news sites that had the word mac somewhere on the page.
Yewbacca: Ah, well it's a really really neat new technology on the latest cards, that lets you hook up 4-6 monitors on one card.
Yewbacca: http://www.amd.com/us/products/technologies/eyefinity/Pages/eyefinity.aspx
Yewbacca: The lower end cards are already out, I think they have 3-4 ports
MostAwesomeDude: Oh.
MostAwesomeDude: That's not anything in the drivers, it's a feature of the cards.
MostAwesomeDude: r800 chipsets (HD 5xxx) can do up to 6 monitors.
Yewbacca: Yeah those would be the ones. Hmm. I am completely stumped, short of emailing AMD directly.
MostAwesomeDude: Well, I just told you. It's not anything special in the drivers.
MostAwesomeDude: When the OS X drivers support those cards, they'll support using all 6 monitor controllers.
Yewbacca: Yep but they don't even mention macs on the page. They do have a notice about Linux at the very bottom though. Kinda odd.
Yewbacca: Ahhhh
MostAwesomeDude: AMD cares more about Linux than Apple. I think that's pretty clear. :3
Yewbacca: Yes :-D Probably due to disliking that Apple uses only intel and nvidia hardware
Yewbacca: The only AMD hardware they have is optional ATI cards
Yewbacca: Thanks for clearing it up though, I'll keep waiting and see what happens on the mac front
MostAwesomeDude: Well, it could also be because Linux is awesome and Apple is not. :3
Yewbacca: Haha :-) I like linux and open source, I don't like communism and the GPL :P
Yewbacca: What's your stand on Richy rich? Like or don't like his ideals?
biotube: Yewbacca: I believe that, on a techinical level, the GPL is closer to fascism
MostAwesomeDude: Mediocre movie.
biotube: but this's hardly the place to discus that
raevol: what level of opengl support does the r600 driver currently have?
MostAwesomeDude: raevol: 1.4 IIRC.
biotube: mine says 1.5
raevol: is it moving towards supporting newer versions?
Yewbacca: His GPL3 move to force anyone that uses the Linux kernel or associated apps in a hardware device, to make that device moddable and flashable, threatens commercial use of Linux, and with it, contributions from wealthy organisations. :-P I don't like him.
Yewbacca: But sorry for bringing us back to 2007
Yewbacca: :-D
biotube: Yewbacca: the linux kernel _can't_ be made GPLv3
Yewbacca: Yeah but all the necessary GNU apps already went.
MostAwesomeDude: Yewbacca: I don't like how people are so focused on licenses and not on sitting down and writing code.
MostAwesomeDude: Xorg's all MIT-licensed. Nobody steals it because nobody can understand it.
MostAwesomeDude: That's fine by me.
MostAwesomeDude: Kernel is GPL. Nobody can rip it off and make money without explaining what they changed.
MostAwesomeDude: That's fine by me.
Yewbacca: Well it's a pretty dangerous new license, threatening all vendors that want to use Linux, to either use old GNU v2 tools or write their own
MostAwesomeDude: Licenses are not important. Only code.
Yewbacca: Yeah
soreau: totally agrees with MostAwesomeDude
raevol: i'm thinking about trying the xorg-edgers ppa to try the new -ati on my ubuntu, but i want to be able to play games :E
Yewbacca: I do like the patent clause of GPLv3 though
raevol: and by games i mean heroes of newerth
Yewbacca: If they wanna use our code, don't sue, that's awesome. And Red Hat agreed to it.
MostAwesomeDude: raevol: I wouldn't say that the r600 driver is ready for general use, but if you want to test it, it's there. :3
Yewbacca: It's just the "if you wanna use GNU apps on your hardware device, you must provide all DRM keys, and firmware flashing instructions" that really sucks.
MostAwesomeDude: Not really. Most good devices are pretty easy to flash.
biotube: don't like it? code replacements
raevol: MostAwesomeDude: hmmz. yea i'm trying to find what level of opengl is needed for this game
MostAwesomeDude: Additionally, if you can buy it, you can break it. The lessons of AACS aren't soon forgot.
Ghworg: What use is having the source code if you can't run it because of drm
MostAwesomeDude: raevol: If the game's written correctly, it'll either bitch and moan about missing extensions, or gracefully fallback and run.
Yewbacca: Ghworg: Well, the hardware is owned by the people who made it, the code can be useful for other things, like supporting similar chips in other devices. If they want to have DRM it's their device.
soreau: or crash your X server, fry you gpu and make your cornflakes soggy
Yewbacca: Mmh soggy cornflakes.
raevol: MostAwesomeDude: it's still in beta and their linux support guy is stretched pretty thin. at this time it probably won't be graceful at all. i'm just frustrated because fglrx xinerama is crashing VLC and xfmedia.... ~!??!?!?!? grar
MostAwesomeDude: Yewbacca: I bought it, I should be able to change it, no exceptions.
MostAwesomeDude: It's *my* device.
soreau: no doubt
MostAwesomeDude: raevol: Well, try it out. Sounds like they might be responsive to bug reports. Just keep in mind that r600 might not be doing the Right Thing all the time.
Yewbacca: MostAwesomeDude: But that's not feasible if FOSS is going to compete in areas where the content providers DEMAND DRM.
Yewbacca: Even though I think DRM is completely pointless and can always be broken
Ghworg: DRM is utterly pointless, eventually the content producers will figure this out
MostAwesomeDude: Yewbacca: Someday the content providers will be gone. I intend to see to it personally.
MostAwesomeDude: is card-carrying Pirate Party member
MostAwesomeDude: Until then, we do what we can.
Yewbacca: Yeah, they will. Apples music store took monopoly and then went DRM-free, and after some whining, all music studios agreed. <3 Some day they will learn that DRM is pointless in the first place.
raevol: MostAwesomeDude: i gotta run for now but i may try it in a few days. thanks
MostAwesomeDude: raevol: NP
Yewbacca: MostAwesomeDude: Oh? Swedish?
MostAwesomeDude: Yewbacca: No, American.
Yewbacca: Ah, you have your own Pirate party?
MostAwesomeDude: Yep, pirateparty.us
Yewbacca: Sweet! :-)
Yewbacca: I voted Pirateparty here in sweden, they took 1 seat in the EU.
mzz: most forms of drm are just really peculiar forms of obfuscation. Sometimes it's amusing to see just how peculiar.
Yewbacca: That's the pirate bay people :]
MostAwesomeDude: I'm quite happy about that. Who's filling that seat, by the way?
mzz: (what was that dvd one that involved a virtual machine?)
MostAwesomeDude: mzz: BD+. It got broken by a couple guys that re-implemented the VM as per the instructions in the patent for it.
hnsr: i wish people in my country cared more for privacy and stuff, if I want to get a new id card ot passport I need ot have my fingerprint taken and added to some huge database
hnsr: it pisses me off to no end
hnsr: s/ot/to/
Yewbacca: MostAwesomeDude: Christian Engstrom, an articulate man with a lot of arguments. I heard he's already helped take out some of the draconian bits in the Telecom package that would give ISPs the power to shut people down from the internet for life.
MostAwesomeDude: And it affects us here, in #radeon, too. We can't use all the video-decoding abilities of newer Radeons because of legal issues further up the food chain.
MostAwesomeDude: Yewbacca: Oh, a great pick.
mzz: I read something about that, but I don't remember which radeons it affected
MostAwesomeDude: mzz: All r600+.
mzz: something about an oldish form of copy protection and fear that implementing the support for it would run afoul of the dmca
Yewbacca: hsnr: Yeah, wait for the state-mandated RFID chips. Why do we even need them, other than keeping track of people? Imagine theft in the future, just cut off someones hand and open their car, drive to their home, and open that too.
Yewbacca: MostAwesomeDude: Yeah. Software patents are the tool of the devil, I hate them, you never know when you've stepped on one
hnsr: :(
hnsr: i think i just became depressed
biotube: don't get sad
Yewbacca: :-(
biotube: get MAD
Yewbacca: Sony Ericsson's president (bilderberg member) said that all their phones will have RFID chips by 2010. So that's how they'll get people to accept them, show people how easy it is to unlock the car, the home, etc with your phone. Then soon enough you won't be able to get a plane ticket or board a bus without one, and that's when you get offered to have them implanted. Yep 1984 alright.
Yewbacca: The health care reform bill is talking about giving everyone RFID cards as well, to keep medical history. We need to get mad, not depressed.
biotube: it also says employers have to pay so much and can't negotiate pay cuts
biotube: mass unemployment's plenty to get ticked off about
Yewbacca: Holy hell... They are driving the system down by design. :/
biotube: why not? rome went the same way
Yewbacca: True. The best bit about current society is that so many people are awakening thanks to the internet. That's why they're struggling to take control of that now. In fact, in just 17 days, all of Sweden's internet communications will become wiretapped.
Yewbacca: So it's literally a race against the clock, since reading infowars.com and similar sites, in just 17 days, will get you flagged as a terrorist in sweden
Yewbacca: But at the same time, so many people are aware now, that awareness is spreading fast. So we're living in very interesting times.
Yewbacca: Have a look at this. It's amazing that the Finnish health minister dared speak out: http://www.youtube.com/watch?v=nTgyakGAddM
pdenapo: Hi, I have a question: how may I disable kernel mode setting for ATI Radeon on ubuntu 9.10? (radeon.nomodeset=1 boot parameter does not work)
biotube: radeon.modeset=0
pdenapo: I think I've tried that also, I will try again
pdenapo: I'm experimenting hangs and black screens when switching from X to a virtual console and back. Is there a way to debug this?
biotube: IIRC, passing an invalid option(such as, say, radeon.makegoodpizza=1) should also do the trick
soreau: By default, karmic should have kms disabled
soreau: so eliminating the radeon.modeset line completely should do it
Yewbacca: biotube: That reminds me how hungry I am :-P Gonna cook food. Did you watch the video?
biotube: Yewbacca: no
Yewbacca: Awwright ;-)
pdenapo: If kms is not active for karmic, the problem should be a different one..
soreau: biotube: getmepizza.andabeer=1
soreau: :D
pdenapo: is there some way to know if it is active?
soreau: pdenapo: Your X log should tell you
pdenapo: many thanks for everything !
pdenapo: indeed it says.. KMS inform that modesetting is not supported
pdenapo: I'm having an ATI RS482 [Radeon Express 200M], lspci reports
soreau: pdenapo: Maybe I missed it, but what are you ultimately trying to accomplish?
pdenapo: to see why my notebook hangs when swtching from X to a virtual console and back
pdenapo: I cannot even write a decent bug report, as the hang is a complete one. I cannot even log in to my machine by ssh
soreau: more rs4xx fun :P
EruditeHermit: airlied, hey
AstralStorm: hey, about the gl problem
AstralStorm: the integerv returns Error number: 981206
AstralStorm: wth does that mean?
AstralStorm: 9812064
AstralStorm: glGetIntegerv(GL_PROGRAM_ERROR_POSITION, &err); - that's how mplayer checks for fragment program error
AstralStorm: it returns that value
AstralStorm: highly unlikely that the program is that long
AstralStorm: btw, it only fails... intermittently
AstralStorm: which suggest to me some kind of caching problem
soreau: Hi EruditeHermit
soreau: Any progress with your issue?
EruditeHermit: soreau, we've tried stuff
EruditeHermit: but nothing has worked
EruditeHermit: today another person contributed to the bug report
EruditeHermit: soreau, suspend resume has never worked for you?
AstralStorm: it meant -ENOCOFFEE on my part
soreau: EruditeHermit: It used to work, but like I said I don't really use it since I run my desktop 'always on' and I don't have a lappy worth anything
soreau: The one time I tried it, I believe when I first tried drm-next best my memory serves me, it worked
dottedmag: My server restarts sometimes (sometimes quite often). Nothing suspicious in Xorg.log, and the only line in dmesg I could associate crash with is: [drm] Resetting GPU
dottedmag: Is that harmless or a sign of bug? How to debug if it is?
soreau: checks for any drm-next updates
soreau: Hmm, seemingly 'a lot' of updates
soreau: compiles
soreau: Doesn't look like any changes for my chipset to the untrained eye but who knows? :)
_Groo_: hi/2 all, any dev alive?
EruditeHermit: _Groo_, I would hope they are all alive
_Groo_: hi EruditeHermit
EruditeHermit: =p
_Groo_: i have more info for the googleearth bug and rs485 (aka evil card)
EruditeHermit: I am not a dev
_Groo_: yeah i know :P
EruditeHermit: I was just kidding around with you
EruditeHermit: =p
_Groo_: for the devs, pls read bug 23715, im adding info to it
soreau: is still compiling
soreau: kernel's done, now waiting on rest of system updates to compile
_Groo_: hi soreau
_Groo_: soreau: could you pls take a look at https://bugs.freedesktop.org/attachment.cgi?id=29223
_Groo_: googleearth doest work correctly with rs485 (what does) :P
soreau: _Groo_: In my opinion to the untrained eye, looks like it might be...
soreau: a bug.
soreau: but that is pure speculation
_Groo_: soreau: its a bug :P i opened the bug report...
soreau: At least you already know you have the chipset of doom
_Groo_: soreau: if i enable "disable low-impact fallback" in driconf it works fine but very slooooowwly
_Groo_: soreau: yeah... my luck
_Groo_: radeon engineers were drunk when they did my card...
_Groo_: soreau: since my evil card dont have tcl, i believe is some bug in the non-tcl codepath?
soreau: gotta celebrate somehow I guess
soreau: _Groo_: In case you didn't already know, I am not a radeon developer by any means.
soreau: So I have NFC
_Groo_: soreau: buy you are always around, pls bug dave and alex about this ;) i gotta go , its 3am over here
_Groo_: i mean but not buy
soreau: is confident all will be able to review the logs
_Groo_: ¬¬
_Groo_: gonna sleep now, maybe airlied and agd5f will be benevolent and fix groos evil card
_Groo_: seeya ppl :) bb later
DanaG: hmm, does anyone know where the origin is, when doing a transform?
DanaG: I figured out how to rotate my display by non-90-degree angles... but now I'm trying to figure out how to center it.
DanaG: I see.. the origin is in the upper-left.
soreau: EruditeHermit: With latest updates, it does s/r here but on resume there is nothing but a corrupted screen
EruditeHermit: soreau, what sort of corruption do you see?
EruditeHermit: is it like this picture?
soreau: I tried restarting the WM, using different WM's, even restarting the display manager, nothing helped not even a standalone X session
soreau: Only a reboot 'fixed' it
EruditeHermit: https://bugs.freedesktop.org/attachment.cgi?id=29408
soreau: but audio was working and all functional, just major unusable corruption for the whole screen including cursor (though I could see a corrupted square for the cursor)
EruditeHermit: soreau, does it look like that picture?
soreau: EruditeHermit: Not like that but just random old vram bits splattered across the screen
soreau: and even now that 'everything' is working, the kde cashew in the upper right corner is corrupted even after reboot
soreau: but everything 'works' fine
soreau: Like I said, even after resume, everything was functional including tty (which had no distortion btw), no lock ups or anything like that
soreau: just like something wasn't being 'restored' or reset
soreau: eww, noticing othier corruption in kde panel
EruditeHermit: soreau, for me, everything i s working but the screen isn't turned back on
EruditeHermit: for the laptop screen
soreau: just restarted plasma, still corruption with cashew and in kpanel
EruditeHermit: if I plug in an external screen and try, I get the picture I sent you
soreau: EruditeHermit: I see
soreau: I will try a cold boot now, brb
EruditeHermit: ok
soreau: corruption still here. Looks like this: http://omploader.org/vMnJicg