[00:37] <Sarvatt> cnd: seing lots of bugs crashing in NewInputDeviceRequest that are very obviously something frankenserver specific, are you absolutely sure backporting the input options stuff and really using the 1.12 input abi isnt the way to go?
[00:37] <cnd> Sarvatt, I know there are bugs out there, but I haven't had any time to look at them yet :(
[00:37] <cnd> so I don't have a good answer
[00:38] <bryceh> list 'o crashes:  https://bugs.launchpad.net/ubuntu/+source/xorg-server?field.tag=precise&field.tags_combinator=ALL
[00:39] <bryceh> list Sarvatt said, most of these are crashing in or through input-related functionality
[00:39] <Sarvatt> things like invalid evdev options crashing the server
[00:39] <Sarvatt> and they work with 1.12, wasting whot's time forwarding them upstream :)
[00:40] <cnd> Sarvatt, do you have a list of them handy?
[00:40] <cnd> even just a couple
[00:41] <Sarvatt> give me a few to look it up, i lost the tabs i had open for a few weeks meaning to mention it
[00:41] <bryceh> 931397 
[00:41] <bryceh> 939023 maybe too
[00:42] <bryceh> cnd, I can repro both of those
[00:43] <Sarvatt> bryceh: thank you, the second one specifically
[00:43] <Sarvatt> https://bugs.freedesktop.org/show_bug.cgi?id=46588
[00:43] <Sarvatt> that was the one i was trying to dig up
[00:43] <cnd> bug 939023 is more important
[00:44] <cnd> the only bug I was aware of was 931397
[00:44] <cnd> which was a low priority because disabling auto add of devices doesn't seem to have a valid use case
[00:44] <cnd> I was chatting with the bug reporter
[00:44] <cnd> and he said he was following some instructions somewhere
[00:45] <cnd> something to do with hybrid graphics, but auto add of devices is only for input devices, IIUC
[00:45] <cnd> anyways, I will try to get to those bugs soonish
[00:45] <cnd> but if anyone can lend a hand, I would very much appreciate it
[00:46] <bryceh> cnd, what sort of help is needed?
[00:46] <cnd> bryceh, fixing the bug?
[00:47] <cnd> I don't know what's wrong yet
[00:47] <Sarvatt> cnd: would just plain doing an actual input abi backport from 1.12 at all be an option?
[00:47] <cnd> Sarvatt, we have an input abi backport
[00:47] <cnd> the problem is that there is an option abi backport
[00:47] <Sarvatt> yeah but its not compatible, things built against 1.12 dont work because of the options
[00:47] <cnd> well, there could be
[00:48] <cnd> the problem is that the options abi isn't contained to input drivers
[00:48] <cnd> it's possible that other drivers, like video drivers, would use the abi
[00:48] <Sarvatt> ohhh, didn't know that
[00:48] <cnd> if a proprietary driver did that, we would be incompatible with the upstream 1.11 video abi
[00:48] <cnd> so we really need to figure out why it's crashing
[00:48] <RAOF> Which would be *awesome fun*
[00:48] <cnd> and fix it
[00:49] <cnd> it probably wouldn't be difficult to figure out what's wrong
[00:49] <cnd> firing up the server under gdb
[00:49] <cnd> I'm just lacking for time :(
[00:56] <Sarvatt> cnd: It could be worse than coding style changes breaking everything, drivers could be in the server :)
[00:57]  * Sarvatt is so glad that idea stalled
[00:58] <bryceh> Sarvatt, maybe only stalled until next UDS?
[00:58] <bryceh> er, XDC
[05:05] <tjaalton> having the drivers in the server would make backports much easier, that's for sure :)
[07:17] <tjaalton> green light for wacom 0.14
[08:19] <tjaalton> so mesa 8.0.2?
[08:20] <tjaalton> Sarvatt: did you test the new mesa? maybe I'll hold off on uploading it until you're back online..
[11:58] <apw> bryceh, did you see my updated edid patched kernels, with the latest upstream contribution
[16:31] <tjaalton> Sarvatt: were we supposed to get the newer libdrm?
[16:31] <tjaalton> for precise
[16:32] <tjaalton> intel-gpu-tools could be synced
[16:34] <Sarvatt> i gave up on it at feature freeze because it wasnt in debian and nothing needed it, we'd have to revert efd6e81e2ba112105457887ae18a58dfa4bbc8ef and 82c6938d232327233caac743a07639ac91bceb7e or else plymouth would be busted
[16:34] <tjaalton> ah, ok
[16:34] <Sarvatt> no response on the plymouth bug i filed asking for backporting the year old commit to make drm backends optional
[16:35] <tjaalton> oh, libxcb
[16:36] <jcristau> Sarvatt: that's been fixed upstream in plymouth for a while
[16:36] <Sarvatt> jcristau: yea our plymouth is just ancient
[16:36] <jcristau> and there was finally a plymouth release recently
[16:37] <Sarvatt> definitely not going to be updated until 12.10, plymouth is too fragile :)
[16:38] <tjaalton> synced libxcb 1.8.1
[16:39] <Sarvatt> 0.8.2-2ubuntu28, its the new initramfs-tools
[16:39] <Sarvatt> 2 year old release with lots of backports because noone wants to break it :P
[16:40] <tjaalton> there's a bug against MoM to show the age of the merge
[16:40] <tjaalton> there are/were lot of those that noone wanted to touch..
[16:41] <tjaalton> I merged nss & nspr with 4+ years of accumulated diff
[16:41] <tjaalton> turns out the diff was very small, just the history was messy
[16:42] <tjaalton> but the plumbing components _are_ harder..
[16:43] <tjaalton> Sarvatt: x-x-v-vmware, syncable?
[16:43] <Sarvatt> tjaalton: yes but no point
[16:43] <Sarvatt> only xserver 1.12 changes past what we have
[16:43] <Sarvatt> well i better test build to make sure :)
[16:43] <tjaalton> yeah but the version number is ugly :)
[16:44] <tjaalton> vmmouse too
[16:45] <Sarvatt> vmware builds and is only bug fixes past what we have, ship it! :)
[16:46]  * Sarvatt grumbles about util-macros verboseness
[16:47] <Sarvatt> xf86-video-vmware hates -Wredundant-decls
[16:52] <Prf_Jakob> Sarvatt: sorry...
[16:52] <Sarvatt> nothing to be sorry about, the newer macros are noisy on absolutely everything :)
[16:53] <Prf_Jakob> Oh I released 8.0.2 yesteraday.
[16:54] <Sarvatt> http://paste.ubuntu.com/895321/ is my favorite example
[16:54] <Sarvatt> Prf_Jakob: yeah we saw, already in ubuntu \o/ thanks for that
[16:55] <Prf_Jakob> Sarvatt: np
[16:56] <Prf_Jakob> also \o/
[16:59] <jcristau> btw, *poke* about merging src:xorg :)
[17:03] <bryceh> apw, yeah do you need me to re-test?  I've got a few other bugs I'm backed up on (and am piloting today) but can squeeze in time if its urgent
[17:05] <tjaalton> jcristau: yup
[17:08] <bryceh> apw, oh wow this patch is quite a bit different isn't it
[17:19] <tjaalton> umm, what about -intel?
[17:19] <tjaalton> we're one release behind
[17:20] <bryceh> tjaalton, what's in it?
[17:21] <tjaalton> bryceh: no idea, but it's part of the Q1 release?
[17:22] <apw> bryceh, quite different and yet not really, it works ish for me, but to propose it for P i would like a valid test like your 'fixed my config test you reported' would be good
[17:22] <bryceh> are we frozen now?  too late for a merge?
[17:22] <bryceh> apw, on it
[17:23] <tjaalton> not frozen until 3-4h from now
[17:23] <bryceh> tjaalton, plenty of time!  ;-)
[17:23] <tjaalton> bryceh: my thoughts exactly! :)
[17:23] <tjaalton> i'll merge xorg
[17:23] <tjaalton> that one _is_ important
[17:24] <tjaalton> intel is mostly sna changes i guess
[17:24] <bryceh> yeah the ddx has gotten pretty tame
[17:27] <tjaalton> ah no, 2.17.0 is the 12.02 release component
[17:31] <bryceh> heh http://cgit.freedesktop.org/xorg/driver/xf86-video-intel/commit/?id=3640a0d4cb9e0f115fda9ea36212670f6ccafb22
[17:32] <tjaalton> hah
[17:36] <Sarvatt> got intel merged locally and testing it out now
[17:36] <Sarvatt> just incase someone gives the go-ahead :)
[17:37] <tjaalton> git log --oneline 2.17.0..2.18.0 |grep -v " sna", could filter even more
[17:44] <bryceh> tjaalton, grep -v glamor ?
[17:44] <tjaalton> bryceh: yeah that too
[17:44] <bryceh> looks like there's a few uxa bits that might be worthwhile but a lot of the remainder here feels code refactory
[17:45] <bryceh> if worse came to worse we could probably just cherrypick out the uxa bits that look useful
[17:46] <bryceh> https://bugs.launchpad.net/ubuntu/+source/xserver-xorg-video-intel?field.tag=precise&field.tags_combinator=ALL
[17:47] <bryceh> pretty much all our open bugs for precise look like drm / kms type issues.
[17:47] <bryceh> tjaalton, guess an argument could be made "ain't broke, don't fix"
[17:51] <Sarvatt> see any bugs with intel_uxa_prepare_access: bo map failed: No space left on device in xorg.0.log?
[17:52] <Sarvatt> i've hit that before but i dont see any bugs, argh :)
[17:52] <bryceh> https://bugs.launchpad.net/ubuntu/+source/xserver-xorg-video-intel/+bug/539898
[17:53] <bryceh> wait no, that's bo map failed
[17:53] <bryceh> hmm, same thing with #468923
[17:53] <bryceh> stupid launchpad search
[17:53] <Sarvatt> thats too old to possibly be relevant i'd imagine
[17:54] <bryceh> yeah karmic
[17:54] <bryceh> not spotting anything in precise offhand
[17:55] <tjaalton> bryceh: ok then :)
[17:55] <tjaalton> xorg merge uploaded
[17:56] <tjaalton> i'll discuss -qxl sync with hallyn tomorrow
[17:57] <Sarvatt> ah i see
[17:57] <Sarvatt> it happens under OOM conditions, and people file the bugs against the kernel
[17:57] <Sarvatt> so no xorg.0.log
[17:58] <Sarvatt> https://bugs.launchpad.net/ubuntu/+source/linux/+bug/891300 is one that needs http://cgit.freedesktop.org/xorg/driver/xf86-video-intel/commit/?id=5b5cd6780ef7cae8f49d71d7c8532597291402d8
[18:00] <bryceh> cherrypickable?
[18:03] <bryceh> apw, this edid patch totally rocks :-)
[18:04] <apw> bryceh, cool ... its a direct cherry of upsteams merge, so it also makes sense to use i
[18:04] <apw> it
[18:05] <bryceh> apw, will shoot an email back with results
[18:05] <apw> bryceh, and for UDS-Q we need to start thinking abnout some userspace for it
[18:05] <apw> bryceh, sounds good
[18:57] <LLStarks> this isn't good. i'm getting a lot of cursor death with the precise input drivers. cursor will get stuck in the grab mode and can't release to click on anything else.
[19:00] <LLStarks> any way to unstick other than a sysreq k?
[19:02] <tjaalton> "input drivers".. which ones?
[19:06] <Sarvatt> bryceh: huzzah https://bugs.freedesktop.org/attachment.cgi?id=58726
[19:08] <Sarvatt> fixes the intel_miptree_release segfaults
[19:08] <tjaalton> phew
[19:13] <Sarvatt> great, cant even load https://bugs.launchpad.net/ubuntu/+source/unity/+bug/926379
[19:17] <LLStarks> tjaalton: i've observed it starting with -mouse
[19:18] <LLStarks> and persists with -synaptics
[19:18] <tjaalton> LLStarks: you most certainly aren't using mouse
[19:19] <LLStarks> what's the default for usb mice then, -evdev?
[19:19] <tjaalton> yes
[19:19] <LLStarks> that's what i meant
[19:20] <LLStarks> trying to find a goodway to reproduce on command
[19:20] <tjaalton> what do you mean by "grab mode"?
[19:20] <tjaalton> some app has the focus and won't let go?
[19:23] <tjaalton> Sarvatt: should we push another mesa with that commit?
[19:23] <Sarvatt> for sure
[19:23] <Sarvatt> i've been trying to interact with that bug for 10 minutes now, about to give up
[19:23] <Sarvatt> want me to push it?
[19:24] <tjaalton> either way is fine
[19:24] <tjaalton> you found it, maybe best if you commit it :)
[19:25] <Sarvatt> okie pushed
[19:26] <Sarvatt> https://bugs.freedesktop.org/show_bug.cgi?id=46303
[19:26] <tjaalton> uploaded
[19:28] <LLStarks> tjaalton: pretty much, but moreso its a problem of the window drag mode can't be released
[19:29] <LLStarks> so, i get the fist icon and can't click anything
[19:29] <LLStarks> keyboard works
[19:30] <tjaalton> LLStarks: first icon?
[19:32] <LLStarks> fist
[19:32] <LLStarks> the grab icon
[19:33] <tjaalton> oh fist, hehe
[19:34] <tjaalton> this is with unity?
[19:35] <LLStarks> shell
[19:36] <tjaalton> right, well I'd say it's a bug there then
[19:36] <LLStarks> don't use unity long enough to encounter i
[19:36] <LLStarks> *it
[19:36] <LLStarks> not sure what packages to file against
[19:37] <tjaalton> gnome-shell?
[19:37] <LLStarks> yeah
[19:37] <tjaalton> that then :)
[19:37] <LLStarks> ok
[19:37] <LLStarks> thanks
[19:44] <tjaalton> i could be wrong, but that's where to look for possible dupes
[21:49] <chandler> Hello folks. I just updated to the most recent xserver-xorg-video-vmware (12.0.1) in precise, and it looks like XA is not enabled in this version of the package. Was this disabled for some reason, or should I file a bug? (Or none of the above?)
[21:57] <bryceh> chandler, looks like -vmware got sync'd from debian, which dropped Robert's change to enable XA in the build, from 1:11.99.901-0ubuntu1 
[21:58] <bryceh> chandler, so yes, please file a bug report, and assign to the canonical-x team if you could.
[21:58] <Sarvatt> ah heck, sorry about that! bryceh would you be able to help me out with uploading the merge? i'll get it all ready
[21:59] <Sarvatt> if we dont get reenabled in beta 2 there won't be 3D accelerated unity on the image, thats a pretty big feature to miss by mistake
[21:59] <bryceh> Sarvatt, sure thing
[21:59] <Sarvatt> (in vmware)
[22:00] <bryceh> do we have -vmware in git?
[22:01] <Sarvatt> yup
[22:02] <Sarvatt> http://anonscm.debian.org/gitweb/?p=pkg-xorg/driver/xserver-xorg-video-vmware.git;a=summary
[22:02] <Sarvatt> i'm just updating it now
[22:02] <bryceh> cool
[22:02] <Sarvatt> i didnt realize xatracker was disabled in debian but it makes sense because its in experimental still
[22:03] <tjaalton> right, missed that
[22:06] <Sarvatt> tjaalton: oh you're around still? ok if I push it UNRELEASED and you do the close/upload since thats so much easier? :)
[22:06] <chandler> I created a bug report, but I can't assign it: https://bugs.launchpad.net/ubuntu/+source/xserver-xorg-video-vmware/+bug/962599
[22:07] <Sarvatt> chandler: thanks for that!
[22:07] <Sarvatt> was just about to do that to add it to the changelog
[22:08] <Sarvatt> well i'll just get it ready for sponsorship then, its late there :)
[22:10] <Sarvatt> bryceh: http://kernel.ubuntu.com/~sarvatt/merges/vmware/ and its in git
[22:10] <Sarvatt> bryceh: scratch that!!
[22:11] <tjaalton> yeah still here
[22:12] <Sarvatt> test building now
[22:12] <Sarvatt> just pushed another change
[22:13] <bryceh> pulling
[22:13] <bryceh> 3df8bb3e ?
[22:14] <Sarvatt> ok http://kernel.ubuntu.com/~sarvatt/merges/vmware/ is up to date and has a build log, so is git
[22:14] <Sarvatt> bryceh: yep
[22:16] <Sarvatt> tjaalton: i even test built it before, but the schroot i used wasn't clean, had libxatracker-dev installed already :( sorry about that
[22:17] <bryceh> Sarvatt, uploaded
[22:17] <Sarvatt> bryceh: thanks a million! that would have really sucked, glad chandler caught it :)
[22:18] <bryceh> chandler, if you wouldn't mind please do an update in a couple hours (once it's done building and distributed to your mirror), install version 12.0.1-1ubuntu1, and verify it fixes the issue you saw
[22:18] <bryceh> Sarvatt, yeah totally.  thanks chandler 
[22:21] <chandler> Sure, of course.
[22:22] <tjaalton> Sarvatt: heh, no worries.. sorry for not noticing that change myself