[05:34] <mlankhorst> morning
[05:37] <RAOF> Howdie.
[05:46] <mlankhorst> still talking to danvet, before I really woke up this time, not a good idea :)
[06:09] <mlankhorst> but the alternatives are worse, heatwave here :/
[06:18] <mlankhorst> Anything i can work on that doesn't require much logical thought today?
[06:25] <mlankhorst> oh right, could look at vdpau blobs again :)
[08:58] <tjaalton> oh duh, libglsl.so is merged into libdricore.so..
[08:59] <mlankhorst> :>
[09:03] <tjaalton> hmm, debian hasn't installed it
[09:12] <mlankhorst> tjaalton: oh hey we're again a intel version behind
[09:13] <tjaalton> yup
[09:14] <mlankhorst> git merge debian-experimental seems to be enough
[09:14] <mlankhorst> I guess jcristau beat us this time :)
[09:14] <jcristau> haven't tested at all
[09:14] <tjaalton> we can do that for you :)
[09:15] <tjaalton> or the quantal test puppies can
[09:15] <jcristau> i wanted to remove /etc/modprobe.d/i915-kms.conf in this version
[09:15] <jcristau> since it's not needed post wheezy
[09:15] <mlankhorst> hm does that mean sync to debian again?
[09:15] <tjaalton> nah the apport hook is the only dif
[09:15] <tjaalton> f
[09:15] <tjaalton> after that change
[09:15] <tjaalton> but close
[09:16] <jcristau> i'm going to need some maintainer scripts love to get the conffile removed properly
[09:16] <tjaalton> those are always icky..
[09:17] <mlankhorst> yeah
[09:18] <jcristau> it's written, just not tested yet :)
[09:18] <mlankhorst> blegh I need some work that's not dma-buf related for a change :)
[09:18] <mlankhorst> anything i could work on for a while?
[09:20] <tjaalton> well..
[09:20] <tjaalton> there's the crasher on precise due to synaptics upgrade in -updates
[09:21] <tjaalton> so figuring out why going to 1.6.2 triggered it would be cool
[09:21] <tjaalton> or is the server missing something
[09:22] <mlankhorst> likely server related sadly
[09:23] <mlankhorst> and i don't have the faintest clue what is differing between x1.12 and 1.11 in ubuntu
[09:27] <mlankhorst> tjaalton: it's not a problem in quantal right?
[09:31] <tjaalton> mlankhorst: some dupes of 956071 at least have been reported on quantal
[09:31] <tjaalton> bug 956071
[09:31] <tjaalton> there's also a similar bug on debian
[09:31] <tjaalton> and upstream
[09:32] <tjaalton> well, forwarded upstream by the same guy
[09:33] <mlankhorst> quantal added on 2012-07-18, did we already have the x1.12 stack then? *looks*
[09:33] <tjaalton> there's one filed 7h ago
[09:34] <mlankhorst> aw looks that way, ok I'll take a look
[09:45] <mlankhorst> tjaalton: I think I might have seen that bug when I was playing with valgrind :)
[09:50] <tjaalton> oh good
[09:50] <mlankhorst> or at least some bug during synaptics probing, but it was a weird one and didn't make much sense, maybe it will show up again under more paranoid settings
[09:51] <tjaalton> suspend/resume testing should reveal that before too long
[09:52] <tjaalton> also, it can crash seconds after typing the password, which is kinda late in the resume cycle..
[09:55] <mlankhorst> yeah I get some weird erorr during a log output
[09:55] <mlankhorst> which didn't make sense
[09:55] <mlankhorst> I'll just increase the tracking that valgrind does
[09:56] <tjaalton> the sigsafe logging patches should be in 1.13 though..
[09:56] <tjaalton> if that's related
[09:57] <mlankhorst> valgrind worked fine without on x1.11
[09:59] <tjaalton> k
[09:59] <tjaalton> -synaptics was probably the old version too
[10:00] <mlankhorst> but I'll update quantal first :)
[10:21] <seb128> tseliot, tjaalton, mlankhorst, RAOF_: hey, what's the status on nvidia binary drivers on quantal then? did the driver update work?
[10:22] <mlankhorst> seb128: from ubuntu-devel, it is updated to document it's not compatible, unless there has beeen an update since :)
[10:22] <seb128> do we have an ETA on compatible driver?
[10:23]  * mlankhorst doesn't know nvidia's plans
[10:23] <tjaalton> no eta
[10:24] <tjaalton> "next release"
[10:24] <seb128> :-(
[10:24] <mlankhorst> but at least it's not fglrx
[10:24] <tseliot> seb128: the update doesn't work therefore I simply made sure that the driver conflicts with the new X abi
[10:24] <seb128> well, ati works fine so fglrx is less of an issue
[10:24] <seb128> nouveau is suboptimal still though :-(
[10:25] <mlankhorst> sadly
[10:25] <seb128> tseliot, ok, thanks
[10:25] <mlankhorst> maintainer has spent a lot of time turning every line of code upside down..
[10:25] <mlankhorst> hopefully we'll start seeing fixes again soon
[10:25] <tseliot> :)
[10:26] <mlankhorst> (there's even some lines of code that haven't changed \o/)
[10:27] <tjaalton> single-threaded mesa build on top of a nfs share via wifi is.. surprisingly slow :)
[10:28] <mlankhorst> no what's surprisingly slow is apt-get install over nfs on a gigabet network with a ssd backing..
[10:28] <mlankhorst> gigabit*
[10:28] <mlankhorst> what you described is unsurprisingly slow :)
[10:28] <tjaalton> hehe
[10:28] <tjaalton> yeah
[10:29] <tjaalton> preseeded netboot install over gigabit is like 15min here
[10:29] <tjaalton> full desktop
[10:29] <mlankhorst> I just use debootstra plocally
[10:29] <mlankhorst> usually for apt-get updates in chroot too
[10:30] <tjaalton> oh well, I know the build passes, now it's just finding the right files for dh_install
[10:31] <mlankhorst> brb
[10:55] <mlankhorst> tjaalton: oh great.. same bug as another one I was hitting
[10:55] <mlankhorst> updatetouchstate overflows
[10:56] <mlankhorst> cnd: ^ seems to happen on quantal 1.13rc4 too
[10:57] <tjaalton> mlankhorst: woo, thanks
[10:58] <mlankhorst> well I fear it could be a dup of the other synaptics bug then
[11:03] <tjaalton> hmm libdricore is now versioned
[11:03] <tjaalton> libdricore8.1.0.so.1.0.0
[11:04] <mlankhorst> great..
[11:04] <tjaalton> probably no point in splitting it from libgl1-mesa-dri though
[11:05] <mlankhorst> true just leave it in :)
[11:13] <tjaalton> wha.. the gallium dri's are huge, over 41MB apiece
[11:14] <mlankhorst> debug symbols enabled?
[11:14] <mlankhorst> tjaalton: oh btw https://bugs.launchpad.net/oem-priority/+bug/956071/comments/59
[11:16] <mlankhorst> the crasher backtrace i hadnt seen before
[11:23] <tjaalton> ahh
[11:28] <tjaalton> right, the binaries haven't been stripped yet, oops
[11:29] <tjaalton> no libgallium.so then, the patch would need to be reworked
[11:29] <phillw> Hi folks, could some one please advise as to against what this bug should be reported...http://pastebin.com/SRLnr5fB
[11:30] <tjaalton> the kernel
[11:30] <phillw> tjaalton: thanks!
[11:31] <tjaalton> although
[11:32] <tjaalton> what does "get live session working" mean
[11:32] <tjaalton> ?
[11:32] <phillw> it means that the tester has to insert nomode to get it to run?
[11:32] <phillw> In my distinctly amateur opinion I'd suspect nvidia-graphics-drivers since it could well be fallout from the new X stack. 
[11:33] <tjaalton> no
[11:33] <tjaalton> you don't have that on a live-cd
[11:33] <Sarvatt> tjaalton: --with-llvm-shared-libs saves a bunch of space but needs the patch on https://bugs.freedesktop.org/show_bug.cgi?id=52167 still
[11:34] <tjaalton> Sarvatt: i'll whine again after seeing what they look like after dh_strip :)
[11:34] <phillw> okies, so it is a kernel regression?
[11:34] <tjaalton> phillw: not necessarily
[11:34] <tjaalton> is there a boot splash? does X start but hang later?
[11:36] <phillw> I'm surmising from the tester that he needs to alter the boot sequence to get X to actually run. Once he knows where to file the bug, he will be able to give more information. He's an experienced tester, just unsure as to what to report this one against.
[11:37] <tjaalton> ubuntu-bug xorg
[11:37] <phillw> thanks.
[11:37] <tjaalton> that's a good place to start..
[11:37] <phillw> it can always be re-assigned later.
[11:56]  * mlankhorst recovers from brain thermal cutout
[12:04] <mlankhorst> bryceh: you added 217_revert_bgnonevisitwindow.patch, is it still needed?
[12:13] <mlankhorst> similar to some other patches, for example some old workarounds in xorg-server should probably be upstreamed or dropped entirely..
[12:15] <tjaalton> yes, there was a task about it but it got marked as done, though I think we should be more aggressive about them..
[12:16] <mlankhorst> agreed
[12:16] <mlankhorst> i dropped a 0->null change now, since it really doesn't add anything to keep it
[12:18] <mlankhorst> RAOF_: ping?
[13:23] <tjaalton> head hurts trying to fix 116_use_shared_galliumcore.diff
[13:35] <tjaalton> got to give up for now
[13:39] <mlankhorst> next person to make a cd can fix it? :p
[15:42] <tjaalton> mlankhorst: hehe, exactly. or maybe RAOF_ :)
[16:15] <mlankhorst> brr, fighting ttm is no fun :(
[16:16] <mlankhorst> screw it, most common case will be re-using local reservation object anyhow *does ugly hack*
[16:21] <tjaalton> btw, virtualbox is broken with the new server :)
[16:21] <tjaalton> reportedly
[16:22] <mlankhorst> sigh :s
[16:42] <jcristau> when is vbox not broken
[16:46] <tjaalton> whemn it's not installed i guess..
[16:58] <bryceh> mlankhorst, 217_revert_bgnonevisitwindow.patch is not actually enabled, it was dropped a year ago because it was included in 1.11.0.  But looks like raof left the patch itself in place for some reason
[16:59] <mlankhorst> hm suppose we ought to get rid of all the patches not in series to clean up confusion then :)
[17:01] <bryceh> definitely
[17:07]  * mlankhorst just created a very scary patch
[17:10] <mlankhorst> http://cgit.freedesktop.org/~mlankhorst/linux/log/?h=v10-wip makes reservations sort of global, scared to test if it boots..
[18:38] <tjaalton> wth, "configure: error: --with-driver=dri is deprecated"
[18:39] <tjaalton> built just fine earlier today
[18:41] <tjaalton> oh, with --enable-osmesa it fails that way.. nice
[18:41] <tjaalton> if it's forced
[18:42] <tjaalton> ahah, --enable-dri
[19:08] <mlankhorst> tjaalton: yay obsolete packages gone :)
[19:09] <tjaalton> mlankhorst: they got removed the same time when the stack got copied
[19:10] <tjaalton> oh bloody osmesa
[19:14] <tjaalton> doesn't find glapi/glapitable.h
[19:20] <tjaalton>         -I$(top_srcdir)/src/mapi \
[19:20] <tjaalton> probably should be top_builddir
[19:21] <tjaalton> no
[19:23] <tjaalton> well, needs both :)
[19:24] <tjaalton> 726f534bbb3f is similar for glx
[21:07] <tjaalton> ffs, "libdricore*" is not enough to get all the links to be copied to the package dir
[21:15] <barry> howdy folks.  over in #ubuntu-devel i've been chatting with tjaalton and bryceh about today's big problems with 12.10 in a vmware fusion 4.1 host.  is there a vmware dev here who would like to help me debug the problem?
[21:15] <barry> bug 1039157
[21:17] <bryceh> pst Prf_Jakob ^
[21:17] <tjaalton> Prf_Jakob: ^ vmware-tools needs some love for newer kernels
[21:17] <tjaalton> so that it would build :)
[21:20] <barry> Prf_Jakob: hi.  i'm using 12.10 in fusion 4.1.3 (730298).  i know i'm bleeding edge but perhaps i can help test vmware-tools for 12.10?
[21:57] <tjaalton> right, helps if you copy from the right directory, *sigh*
[22:18] <tjaalton> at long last.. mesa debs created, without libgallium.so though
[22:57] <tjaalton> RAOF_: updated mesa git to build, if libgallium.so is commented out from libgl1-mesa-dri.install.linux.in.. if you have time to look at the patch and what it needs to make it work again, that would be great
[22:58] <tjaalton> I started but after the trivial bits it was unclear what was needed, since some bits have been automakeified
[23:00] <tjaalton> like libdricore build
[23:01] <RAOF_> Yay.
[23:02] <tjaalton> "have fun" :)
[23:03] <Sarvatt> "thank you so much" :)
[23:03] <tjaalton> if you're up for it that is ;)
[23:54] <Prf_Jakob> tjaalton: ugh seems like a lot of things are broken in 12.10
[23:55] <Prf_Jakob> tjaalton: have you guys forgotten to enable the vmwgfx kernel module?
[23:56] <Prf_Jakob> barry: still around
[23:56] <Prf_Jakob> ?
[23:56] <barry> Prf_Jakob: yep!
[23:56] <Prf_Jakob> lsmod | grep vmwgfx ?
[23:57] <Prf_Jakob> yeah its using the old driver for some reason.
[23:57] <barry> @resist[~:1001]% lsmod | grep vmwgfx
[23:57] <barry> @resist[~:1002]% 
[23:57] <barry>  
[23:57] <Prf_Jakob> I'm guessing its because vmwgfx isn't loaded.
[23:57] <barry> Prf_Jakob: yeah, i've tried installing both open-vm-tools and building vmware-tools from the "cd"
[23:58] <Prf_Jakob> We don't ship vmwgfx kernel driver or X drivers in tools for newer distros.
[23:59] <Prf_Jakob> cat /boot/config-`uname -r` | grep VMWGFX
[23:59] <barry> CONFIG_DRM_VMWGFX=m
[23:59] <barry>  
[23:59] <Prf_Jakob> okay
[23:59] <Prf_Jakob> modprobe vmwgfx
[23:59] <Prf_Jakob> sudo service lightdm restart
[23:59] <Prf_Jakob> should sudo modprobe as well.