[00:27] <LLStarks> sup raof
[00:29] <RAOF> Tiredness, apparently.
[00:29] <RAOF> Also me accidentally suspending my IRC bouncer.
[00:49] <bryceh> thanks RAOF!
[03:18] <bryceh> interesting; my nouveau box OOPSed with mesa 8.0.2.  With 8.0.4 I still get an EQ overflow in X and X lockup, but no kernel OOPS
[03:18] <bryceh> just a lot of
[03:18] <bryceh> [  120.491370] [drm] nouveau 0000:05:00.0: Ctxprog is still running
[03:19] <RAOF> Odd.
[03:19] <RAOF> On the other hand, yay progress!
[03:35] <Sarvatt> nouveau is getting worse instead of better over time :(
[03:35] <ajmitch> nvidia as well, it seems
[03:35] <Sarvatt> really?
[03:36]  * ajmitch has a new system running quantal, gets multi-second pauses & occasional full X hang when running games
[03:36] <Sarvatt> i'm still on 295.xx on my rarely used nvidia box
[03:37] <ajmitch> from what I've read, it's a known problem & I'm on the latest drivers also
[03:38] <ajmitch> there's probably an LP bug about it, though it's not dumping anything useful to logs
[06:16] <mlankhorst> morning
[06:22] <tjaalton> yo
[06:47] <mlankhorst> oh great prime work was upstreamed, so I did miss something ;)
[06:49] <tjaalton> xrandr 1.4 isn't released yet though
[06:49] <mlankhorst> tjaalton: there's a patch we could use from airlied though, it automatically binds things
[06:54] <tjaalton> yeah I heard fedora has some
[06:54] <tjaalton> but we need ddx updates too
[06:55] <mlankhorst> do we really want to at this point?
[07:05] <tjaalton> if we want that stuff?
[07:05] <tjaalton> I don't know
[07:06] <tjaalton> intel 2.20.7 has it, so why not
[07:06] <tjaalton> no nouveau or ati releases
[07:06] <tjaalton> our ati is actually rather old
[07:11] <tjaalton> bug 1046759 is fun
[07:12] <tjaalton> -modesetting fail
[07:25] <mlankhorst> hm, modesetting should init as a matter of fact, but it should just pisss off afterwards
[07:26] <tjaalton> right
[07:26] <tjaalton> there are a couple of commits past 0.4 but dunno if they'd make a difference
[07:27] <mlankhorst> maybe it will though, platform bus stuff
[07:28] <tjaalton> oh we have that
[07:29] <mlankhorst> that's post 0.4.0 though
[07:29] <mlankhorst> would probably get rid of [     3.798] (WW) Falling back to old probe method for modesetting
[07:29] <tjaalton> yes, but rsalveti wanted it
[07:30] <tjaalton> hum
[07:30] <tjaalton> so we're missing something on x86?
[07:30] <mlankhorst> dno, likely modesetting is just destructing weirdly
[07:31] <mlankhorst> get a more complete backtrace?
[07:33] <mlankhorst> I'm surprised ati doesn't show up in the log, wonder why though
[07:59] <dj_ryan> so i am trying to make this intel q77 mobo do triple monitors, but its extremely resistent to google'n
[07:59] <dj_ryan> just pluggin in the Displayport cable doesnt do it alas
[08:00] <dj_ryan> (this board had 2*dvi, 1*dp on board, and apparently the intel 4000 hd can do triple monitor, and the intel xorg driver can support it, maybe?)
[08:00] <mlankhorst> intel's doing a some complete reworking of relevant code in their kernel git tree, so if you feel comfortable
[08:00] <mlankhorst> try drm-intel-next
[08:01] <dj_ryan> i dont think ican do that alas
[08:02] <dj_ryan> i need to be on a specific version of kernel so i can install zst
[08:02] <mlankhorst> would be just for testing
[08:03] <dj_ryan> well i'd also like to have my triple displays back
[08:03] <dj_ryan> maybe i'll co-install that ati 5000 card i have
[08:05] <dj_ryan> we had to ditch the ATI 4xxx-based desktops and go with these intels due to stability caused by what we think are xorg crashes
[08:07] <dj_ryan> oh my coworker is getting me a dvi splitter so the 3rd monitor can be connected via VGA... 
[12:03]  * mlankhorst somehow read linux-image-3.4.0-1486-omap4 as linux-image-3.4.0-i486-omap4 and wtf'd for a bit until he realized it :)
[12:20] <mlankhorst> tjaalton: oh I think I know why it fails at least, it tries to add the second screen with prime, but modesetting doesn't handle that
[12:27] <mdeslaur> So...is there any way to get Quantal working with nouveau right now?
[12:27] <mlankhorst> mdeslaur: prime?
[12:28] <mdeslaur> is there a bug for the INVALID_OPCODE failures resulting in me being unable to login?
[12:28] <mdeslaur> mlankhorst: what do you mean by "prime"?
[12:28] <mlankhorst> optimus or general?
[12:29] <mdeslaur> general, it's a thinkpad T61 that worked fine with precise and all other previous releases
[12:30] <mlankhorst> unsure, not that I heard of at least
[12:32] <mdeslaur> mlankhorst: ok, let me try and figure out how to open a bug if I can't log in
[12:32] <mlankhorst> apport-bug ;)
[12:41] <mdeslaur> mlankhorst: LP: #1049088
[12:41] <mdeslaur> yeah, apport-bug is awesome until you get to the part where it wants to open a browser :)
[12:41] <mlankhorst> tjaalton: hm, did you push the 20.6 changes?
[12:43] <tjaalton> mlankhorst: oops, did now
[13:28] <mlankhorst> tjaalton: prime display failed here on that ;) should we push 20.7 and nouveau-ddx git?
[13:53] <tjaalton> mlankhorst: failed the same way?
[13:55] <mlankhorst> nah in a different way, addscreen failed
[13:55] <mlankhorst> but the bug is probably that modesetting didn't expect to be removed early or something
[14:26] <tjaalton> mlankhorst: sure, why not. intel has a release and nouveau not that many commits post 1.0.1
[14:28] <mlankhorst> tjaalton: yeah I asked darktama if i could release current git as 1.0.2
[14:29] <tjaalton> cool
[14:33] <jcristau> oh, chase left canonical?  /me catches up on email
[14:35] <tjaalton> yup, dunno where
[14:46] <mlankhorst> hm.. I wonder if at this point we should remove the nomodeset from recovery mode and attempt to use modesetting instead of fbdev
[15:02] <tjaalton> yes
[15:05] <mlankhorst> requires changing grub-common
[19:16] <bryceh> anyone got thoughts on if we might put mesa 9.1 into x-updates?  good idea / bad idea ?
[21:20] <mlankhorst> bryceh: hm couldn' t we do the same as we will do for mesa in quantal?
[22:19] <bryceh> mlankhorst, yeah probably.  Not sure what's in 9.1 vs. 9.0
[22:49] <RAOF> mlankhorst: re: nomodeset for recovery mode.
[22:51] <RAOF> mlankhorst: I think that VGA text mode is more likely to work, and we want a recovery mode that's as likely as possible to actually work.
[23:12] <bryceh> agreed
[23:14] <bryceh> btw I've set up a PPA for mesa 9 - https://launchpad.net/~ubuntu-x-swat/+archive/mesa9
[23:14] <bryceh> jordan at Intel took care of the packaging.  I've added him as member of ubuntu-x-swat, and invited him to do updates of it
[23:15] <RAOF> Ah, cool.
[23:15] <bryceh> I'm also thinking for x-updates we should think about carrying a mesa update as well.  The DDX's alone really aren't that interesting anymore
[23:15] <RAOF> +1
[23:15] <bryceh> RAOF, great
[23:16] <bryceh> figure we should include for precise the same package we'll behaving in quantal
[23:16] <bryceh> then after 12.04.2 is released we can up that to 9.1 or whatever if we want
[23:55] <Sarvatt> bryceh: problem is carrying a mesa update involves many other packages too ~1 month after a stable release releases.. wayland weston libxkbcommon are guaranteed, maybe llvm and libffi it needs, then tricky things like out of tree builds being broken for months or nouveau libdrm abi breaks. its always easiest right now where theres hardly any divergence but it will get stuck and stable updates will probably be better off for intel during that time be
[23:55] <Sarvatt> cause they backport so much to them
[23:56] <Sarvatt> just having it be drivers is a huge benefit for nvidia people