=== horstle_ is now known as horstle
=== ajmitch__ is now known as ajmitch
[00:57] <RAOF> cnd: Whoops!  I didn't notice before, but your xserver-xorg-input-synaptics upload is a native package.  Generally when we take upstream snapshots we ‘make dist’ a tarball to get an orig.tar.gz.
[00:57] <cnd> RAOF, ahh yes, I didn't think to do that :(
[00:58] <cnd> RAOF, have you already uploaded?
[00:58] <RAOF> Yup, we're pending publication.
[00:58] <cnd> well, we can upload a fix
[00:58] <RAOF> Yeah.
[00:59] <cnd> it's not a show stopper I wouldn't think
[00:59] <RAOF> Just next time we touch it.
[00:59] <cnd> ok
[00:59] <RAOF> It's not.  It's a bit annoying, but nothing terrible.
[01:06] <cnd> RAOF, it's alive!
[01:06] <cnd> we now have multitouch, upstream, and in ubuntu!
[01:06] <RAOF> :)
[01:06]  * cnd is so happy!
[01:06] <RAOF> Rock on!
[01:11] <Prf_Jakob> Cool
[01:14] <RAOF> cnd: You've failed to add 0004-xi22-ubuntu.patch to git :)
[01:14] <cnd> gah...
[01:14] <RAOF> Oh, to xserver-xorg-input-evdev git.
[01:15] <cnd> RAOF, I'll fix it up
[01:15] <RAOF> Ta.
[01:17] <cnd> RAOF, synced up
[01:17] <RAOF> Thanks muchly.
[01:18] <RAOF> Um, that fails to apply in configure.ac?
[01:21] <cnd> grr.
[01:21] <RAOF> You appear to have a copy-number mutation in the patch :)
[01:21] <cnd> how did this build before?
[01:21] <RAOF> Dunno.
[01:21] <RAOF> I can push the fixed patch if you like.
[01:22] <cnd> I'll figure it out
[01:22] <cnd> hmmm, something is pretty wrong in the patch I added
[01:23] <cnd> looks like it does the same thing twice for some reason
[01:23] <RAOF> It's duplicated.
[01:23] <RAOF> You've probably >>ed the patch to itself at some point, or something.
[01:23] <cnd> yeah
[01:25] <RAOF> Now to hit up my new xserver
[01:26] <cnd> RAOF, evdev should be all fixed up
[01:26] <cnd> it builds properly once again
[01:33] <RAOF> Woot
[01:36] <RAOF> Oh, man.  I'm awesome like the fox.
[02:06] <bryce> nice, congrats on the landing guys
[02:40] <bryce> hmm, X starts but doesn't render anything.  odd
[02:54]  * cnd is glad he doesn't do graphics :)
[03:08] <RAOF> bryce: What GPU?
[03:20] <bryce> 945
[03:20] <bryce> RAOF, its the little netbook I was using during the rally
[03:20] <bryce> RAOF, startx works to boot it into 2d mode
[03:23] <bryce> odd, and now it's working fine again
[03:26] <bryce> just wanted another reboot
[03:45] <RAOF> Huh.
[04:38] <tjaalton> bryce: check out the old logfile..
[04:39] <tjaalton> though I doubt it has anything to do with purging the xkb cache..
[06:04] <RAOF> Huh.  For those playing at home, sqrt(dx*dx) isn't actually the quickest way of doing what I was after :)
[06:55] <RAOF> Oh, balls.  Pointer movement can be restricted by multiple barriers.  Arse.
[09:51] <tjaalton> huh, did some defaults change or why is git mergetool failing/refusing to merge anything
[09:52] <tjaalton> have to pull every change one by one, at least when using meld..
[11:58] <tjaalton> RAOF: still around? looks like we ship libwayland-egl in libegl1-mesa, and debian in libegl1-mesa-drivers. maybe ok to go with debian?
[12:04] <tjaalton> also, I don't think there's much point in providing the classic radeon drivers in -dri-experimental
[12:04] <tjaalton> if they even build
[12:06] <tjaalton> hmm dricore is apparently upstream?
[15:04] <tjaalton> galliumcore needs further work, and wayland needs an update
[15:04] <tjaalton> kibi said he'd take care of the latter tomorrow
[15:19] <Sarvatt> tjaalton: what's wrong with galliumcore?
[15:20] <tjaalton> Sarvatt: building dri/swrast fails, since -lgallium is everywhere
[15:20] <tjaalton> didn't look closer, need to run now :)
[15:21] <tjaalton> it's in git, have a look
[15:21] <Sarvatt> yeah got it building now
[15:30] <Sarvatt> +        DRI_LIB_DEPS='-L$(TOP)/$(LIB_DIR) -Wl,-R$(DRI_DRIVER_INSTALL_DIR) -lgallium -ldricore -lglsl'
[15:35] <Sarvatt> could probably even ditch galliumcore with all the space gained from dropping dri1 drivers
=== yofel_ is now known as yofel
[16:55] <Sarvatt> ricotz: damn, almost there
[16:55] <Sarvatt> *** Warning: Linking the executable i965_symbols_test against the loadable module
[16:55] <Sarvatt> *** i965_dri.so is not portable!
[16:55] <Sarvatt> ./.libs/i965_dri.so: undefined reference to `vbo_get_minmax_indices'
[16:55] <Sarvatt> down to 1 undefined reference on mesa master instead of pages full of them
[17:00] <ricotz> Sarvatt, hi, nice and yeah these fuzzing patches could do some harm
[17:03] <Sarvatt> ricotz: tempted to upload an 8.0 branch snapshot for now since mesa master hasn't built since jan 10th
[17:04] <Sarvatt> oh yeah master needs to be 8.0+ now too
[17:06] <ricotz> Sarvatt, i see, since the current version number permits that it should be fine
[17:07] <ricotz> to do a 8.0 branch snapshot
[17:07] <ricotz> but if you are this close ;)
[18:32] <hallyn> hi - I was wondering whether anyone might be willing to upload a change to xserver-xorg-video-qxl for bug 913314
[18:32] <ubot4`> Launchpad bug 913314 in xserver-xorg-video-qxl (Debian) (and 3 other projects) "Corrupted display with Spice (affects: 2) (dups: 2) (heat: 23)" [Unknown,New] https://launchpad.net/bugs/913314
[18:33] <tjaalton> hallyn: is the fix in upstream master or some release?
[18:34] <hallyn> the fix is upstream
[18:34] <hallyn> it's a straight cherrypick
[18:39] <tjaalton> I could push that to debian
[18:39] <hallyn> that'd be great
[18:40] <tjaalton> looks like there is no new release available since 0.16
[18:40] <tjaalton> ..which is seven months old
[18:40] <tjaalton> no, six
[18:42] <tjaalton> hallyn: btw, there is ddebs.u.c which has -dbgsym packages for the drivers. not that discoverable though
[18:44] <hallyn> d'oh.  i thought that was obsolete :)
[18:44] <tjaalton> :)
[18:44] <tjaalton> Sarvatt: d'oh, how did I skip those patches
[18:45] <tjaalton> and why
[18:46] <Sarvatt> tjaalton: can ya pastebin ls -al /usr/lib/x86_64-linux-gnu/dri/ ?
[18:46] <Sarvatt> curious to see the size difference without libgallium
[18:47] <tjaalton> http://paste.ubuntu.com/815632/
[18:48] <Sarvatt> think libdrm's build-dep should be bumped to 2.4.30?
[18:48] <Sarvatt> libdrm-intel requires it
[18:48] <tjaalton> ok
[18:48] <tjaalton> bad upstream :)
[18:49] <Sarvatt> LIBDRM_REQUIRED=2.4.24
[18:49] <Sarvatt> LIBDRM_RADEON_REQUIRED=2.4.24
[18:49] <Sarvatt> LIBDRM_INTEL_REQUIRED=2.4.30
[18:49] <tjaalton> oh ok
[18:49] <Sarvatt> i'm unsure because intel isn't built on all arches and those dont need 2.4.30
[18:52] <tjaalton> could build-deps have 'libdrm-dev (>= 2.4.30) [amd64 i386]'?
[18:52] <tjaalton> and kfreebsd-*
[18:52] <tjaalton> i mean two lines for libdrm-dev
[18:52] <tjaalton> maybe it doesn't matter
[18:52] <tjaalton> just bump it :)
[18:53] <tjaalton> most of them have .30
[18:53] <tjaalton> just hppa, powerpcspe, avr32 behind
[19:03] <Sarvatt> 7th mesa build of the day, poor macbook air
[19:05] <jcristau> hallyn: get upstream to release it?
[19:08] <tjaalton> that would be optimal
[19:15] <Sarvatt> argh, forgot mesa 8.0 branch doesn't compile against newer wayland
[19:16] <tjaalton> sounds great
[19:18] <Sarvatt> only bug i've seen so far against the newer x https://bugs.launchpad.net/ubuntu/+source/nvidia-graphics-drivers-173/+bug/921023
[19:18] <ubot4`> Launchpad bug 921023 in nvidia-graphics-drivers-173 (Ubuntu) "Crash at startup (affects: 1) (heat: 6)" [Undecided,New]
[19:19] <tjaalton> is there a newer version even available?
[19:20] <Sarvatt> .31 but doesn't say it has 1.11 support
[19:21] <tjaalton> right
[19:21] <Sarvatt> might need to add IgnoreABI via jockey again for that package
[19:21] <Sarvatt> if it even works
[19:23] <tjaalton> whee, so we get to drop all the legacy ones :)
[19:23] <Sarvatt> cnd: might be a synaptics problem after the updates - https://bugs.launchpad.net/ubuntu/+source/xorg/+bug/921082
[19:23] <ubot4`> Launchpad bug 921082 in xorg (Ubuntu) "after touchpad movement and stop, pointer continues in same direction at a steady pace (affects: 1) (heat: 6)" [Undecided,New]
[19:23] <cnd> hmm
[19:25] <cnd> it sounds like the intended behavior if it's near the edge of the touchpad
[19:26] <cnd> though it may be new in the latest synaptics version
[20:04] <bryce> uploaded intel-gpu-tools 1.1 just now
[21:13] <RAOF> tjaalton: We've got the go-ahead from the techboard for --enable-texture-float, by the way.  Please feel free to flip that switch.
[21:17] <tjaalton> RAOF: it's flipped already :)
[21:17] <tjaalton> on the debian branch
[21:18] <RAOF> Superb
[21:18] <tjaalton> but good to know that it's ok for them, was about to ask around what the situation was
[21:20] <tjaalton> Sarvatt: how's the dri size comparison doing, got any results?
[21:21] <RAOF> How hard is refreshing galliumcore?  That was a ~10MB win, likely to be more now that llvmpipe is default swrast.
[21:21] <Sarvatt> tjaalton: nope, hit the wayland problem then got poked about some hotkey bugs
[21:23]  * Sarvatt has edgers installed atm and downgrading away from it takes about 30 minutes worth of manual apt-get crap
[21:23] <tjaalton> ah, ok
[21:24] <Sarvatt> oh scratch that, mesa doesn't explicitly depend on the exact version anymore!
[21:24] <Sarvatt> downgraded and got it building again now
[21:24] <tjaalton> RAOF: well, my branch had all sort of fuzzy patches applied, so could be that me "fixing" the patch to apply might've actually broken it further
[21:46] <tjaalton> are we going to use the dri-alternates path anymore, now that the classic drivers are gone?
[21:47] <Sarvatt> i915g would be the only consumer
[21:47] <tjaalton> so it seems
[21:47] <RAOF> And we're not exactly lining up to support i915g :)
[21:48] <Sarvatt> :)
[21:48] <Sarvatt> tjaalton: oh hell
[21:48] <RAOF> We can drop it; it doesn't prevent us from shipping i915g in -experimental if we want to.
[21:48] <Sarvatt> soo it doesn't build with the wayland in precise, and it doesn't build with wayland from git
[21:48] <Sarvatt> going to need to make a special wayland checkout just for mesa 8.0 :(
[21:48] <tjaalton> Sarvatt: yep..
[21:49] <tjaalton> RAOF: ok, I'll clean up rules a bit
[21:52] <Sarvatt> well theres a weird input bug https://bugs.launchpad.net/ubuntu/+source/xorg/+bug/921236
[21:52] <ubot4`> Launchpad bug 921236 in xorg (Ubuntu) "[12.04 Xorg] Dual monitor, after entering passoword, mouse pointer stuck on LHS of screen, no desktop. (affects: 1) (heat: 6)" [Critical,New]
[21:52] <Sarvatt> [    20.232] (II) intel(0): Allocated new frame buffer 3840x1080 stride 15360, tiled
[21:52] <Sarvatt> [    20.436] BUG: triggered 'if (dev->valuator->numAxes < 2)'
[21:52] <Sarvatt> BUG: ../../dix/getevents.c:848 in scale_to_desktop()
[21:54] <RAOF> Awesome.
[21:54] <RAOF> I wonder how you can get there without an at least two-axis valuator.
[21:54] <Sarvatt> already fixed in evdev git
[21:54] <Sarvatt> http://cgit.freedesktop.org/xorg/driver/xf86-input-evdev/commit/?id=5c5b2c8db851df7921cedd888222a6630a007fd8
[22:31] <Sarvatt> anyone mind sponsoring http://sarvatt.com/downloads/merges/evdev-precise/ ? it's been confirmed to fix it by the bug reporter
[22:31] <Sarvatt> (and is in git)
[22:35] <RAOF> Looking
[22:38] <Prf_Jakob> Hello Ubuntu peeps!
[22:39] <Prf_Jakob> How goes the mesa packageing?
[22:43] <RAOF> Prf_Jakob: tjaalton's on it; it sounds like there's significant progress.
[22:44] <tjaalton> missing pieces are getting the wayland bits building, and then adding libxatracker etc
[22:45] <Prf_Jakob> Cool
[22:46] <tjaalton> kibi from debian wants to do the wayland stuff
[22:46] <Prf_Jakob> You need to redo the video-vmware package as well, to pick up xatracker.
[22:46] <tjaalton> not just a rebuild with a suitable option?
[22:46] <Prf_Jakob> That should do it.
[22:47] <tjaalton> ok good
[22:47] <tjaalton> I'll play with it tomorrow..
[22:47] <Prf_Jakob> cool
[22:49] <tjaalton> but the wayland situation is a tricky one, Sarvatt says 8.0 doesn't build with wayland git due to this http://cgit.freedesktop.org/wayland/wayland/commit/?id=151ca457b4384c385c0062716b55595e22fef7ea
[22:49] <Sarvatt> and it doesn't build with our current wayland checkout either
[22:49] <tjaalton> right
[23:10] <Sarvatt> looks like nvidia-graphics-drivers-updates and fglrx-updates didn't get updated