[00:03] darkxst: & cinnamon? [00:05] also, please apply for at least ubuntu-gnome-dev [00:05] Laney, http://launchpadlibrarian.net/146225058/cinnamon_1.7.4-2ubuntu3_1.7.4-2ubuntu4.diff.gz [00:06] Laney, am planning doing that soon, just need to get application sorted [00:06] oh ok [00:07] Probably have to look at it tomorrow [00:08] happyaron, hey, could you look at https://bugs.launchpad.net/ubuntu/+source/ibus/+bug/1221593 ? [00:08] Launchpad bug 1221593 in ibus (Ubuntu) "ibus-ui-gtk3 crashed with SIGABRT in _g_log_abort()" [High,Confirmed] === m_conley_away is now known as m_conley === m_conley is now known as m_conley_away === duflu_ is now known as duflu [05:53] morning [06:08] mlankhorst: Good morning! [06:12] mlankhorst: You wouldn't know, off the top of your head, how to get a radeon GPU to wait for rendering to a bo before copying from it, other than radeon_bo_wait? [06:16] RAOF: copying on the cpu, or gpu? [06:16] GPU [06:16] oh you just need to annotate you use it, and then it waits automatically [06:17] Well, should be on the GPU. Copying via EXA :) [06:18] well if you use existing methods that annotation should already be there [06:27] mlankhorst: Hm. Where is that annotation done? === Trevinho_ is now known as Trevinho [06:46] RAOF: probably shows up as relocation, lets just say it's hard in radeon to do it wrongly [06:47] only newer cards that support vm don't need relocations [06:47] Ah. [06:48] So, if a radeon_bo_wait() in copy_to_mir fixes some corruption, it's likely that this is not actually a fix, but merely slowing things down enough to avoid some other problem. [06:49] indeed [06:49] well [06:49] you need to wait on the target, probably :P [06:49] or at least flush the command stream [06:50] command stream doesn't need to be completed, but it has to be submitted before mir tries to use the buffer [06:50] Which would be done by radeon_cs_flush_indirect, if I'm not mistaken? [06:51] Which we indeed do before sending the buffer to Mir. [06:58] RAOF: what do you see then when not waiting? [07:08] mlankhorst: Roughly speaking, this: [07:08] https://bugs.launchpad.net/xmir/+bug/1233545 [07:08] Launchpad bug 1233545 in xserver-xorg-video-ati (Ubuntu) "XMir screen corruption on radeon chipset" [Critical,Confirmed] [07:13] meh I'll give up on the glamor-egl bug and just file an upstream one, damned corruption.. I'll see if I can reproduce the radeon one [07:13] RAOF: lets just say this, native mir clients don't wait for the bo to finish either right? [07:14] mlankhorst: I'm actually not sure; they *do* call flush() before submitting. [07:14] and can you accept glamor-egl to saucy? It fixes running default configuration saucy with the binary drivers. [07:14] Sure. [07:14] Also, ???! [07:15] https://bugs.launchpad.net/ubuntu/+source/glamor-egl/+bug/1232658 [07:15] Launchpad bug 1232658 in glamor-egl (Ubuntu) "glamoregl fails to load with binary drivers" [Critical,Fix released] [07:15] turns out ordering is important ;) [07:15] - dh $@ --with autoreconf,xsf,quilt --builddirectory=build/ [07:15] + dh $@ --with quilt,autoreconf,xsf --builddirectory=build/ [07:15] Heh. [07:18] mlankhorst: Going to add a saucy task for that bug? Also, some bare-bones testing instructions? [07:18] oops added raring lol [07:19] barebones testing: install any type of binary driver, try to start x [07:19] done [08:05] RAOF: just out of curiosity do you have a debug option to always count the entire screen as damaged? === vrruiz_ is now known as rvr [09:12] vila: seems to be the recordmydesktop thing crashing in conjunction with autopilot tests [09:17] if running a separate "/usr/bin/recordmydesktop --no-sound --no-frame -o /dev/null" before calling autopilot run -v autopilot.tests.functional it crashes almost right away.. [09:18] mlankhorst: I don't have such a debug option, no. It might be worth adding one. [09:20] I want to run Ubuntu Desktop on the device which is having Android ...Please help me [09:23] mlankhorst: not sure I follow, recordmydesktop is crashing or does it make the x server crashing ? [09:24] I have IFC6410 development board which is preloaded with Android. I want to run Ubuntu Desktop on this. No ubuntu is officially available for this board. Can someone please help me in creating the boot.img Ubuntu image for this board\ [09:26] vila: something recordmydesktop does makes xserver crash.. [09:26] mlankhorst: good, so you mean you've reduced the reproducing recipe even further ? [09:27] a little [09:29] but only running recordmydesktop doesn't trigger it, so a second component is needed.. [09:33] Please help me in creating the Ubuntu Desktop image for the ARM board [09:38] :q [09:39] vila: well it crashes on the first test if i run with recordmydesktop [09:39] which was autopilot.tests.functional.test_dbus_query.DbusQueryTests.test_select_single_on_object_with_param [09:42] hm minimal testcase: start recordmydesktop with compiz enabled for compositing, run window-mocker, kill window-mocker window [09:47] mlankhorst: great, please add those to the bug report and I'll try to followup with whoever is responsible for the tests once we have a fix [09:48] it's probably 'trying to read the contents of the window after being removed from compositing, or something.. [09:48] I added the testcase for the bug anyway [09:50] mlankhorst: great [09:51] I fear that this is something upstream has to resolve, though.. [10:06] mlankhorst: can you mail a short summary to didrocks, I start losing track of what the next step is :-/ [10:08] poke amd devs about the bug, I guess :) [10:08] or glamor-egl devs, dno [10:11] mlankhorst: right, I have no idea who those guys are :) So, didrocks should know better since he knows the impact on ci [10:21] but yeah seems running that small test crashes every time [10:38] hm found a possible fix anyway [10:43] ugh how is this even working to begin with :s === MacSlow is now known as MacSlow|lunch [12:16] tseliot, so... when will we get proper opencl support on ubuntu? :) === MacSlow|lunch is now known as MacSlow [13:01] derp [13:03] mhr3: err... what do you mean? [13:04] vila: seems to be caused by PIXMAP_PRIVATE re-use :P [13:06] mlankhorst: is this good or bad ? ;) [13:08] good, I hope [13:30] but no idea really, meh :/ [13:33] hm indeed [14:00] vila: I think glamor-egl is not allowed to use PIXMAP_PRIVATE [14:22] mlankhorst: sorry, I have no idea what this means :-/ [14:28] vila: i think glamor is re-using a key that xorg uses internally, would explain the valgrind error [14:31] mlankhorst: ha, a glamor coding bug then ? I thought it means glamor design was not supporting some higher level feature [14:34] possibly, no diea yet [14:38] hm, seems theoretically allowed, weird [14:47] RAOF: how do you keep up with Xorg? It really really needs less layers of obfuscation === alan_g is now known as alan_g|tea === alan_g|tea is now known as alan_g === Laney is now known as KruSha666 [15:47] morning desktoppers [15:52] 900D /\/\0r|\|1|\|9 UBU|\|7U d3$|<70P [15:54] hello [15:58] hello KruSha666 [15:59] 900D 3\/3|\|1|\|9 |_/·\|\|3Y :) [16:00] 7|-|3 L3373$7 d3$|<70P 4r0U|\|D [16:01] * KruSha666 coughs === KruSha666 is now known as Laney [16:01] hey [16:01] I'm watching you guys!!! [16:04] krusha? seriously? [16:07] mhr3: what do you mean? [16:10] tseliot, our nvidia pkgs aren't overly opencl-friendly [16:10] mhr3: any bug reports in particular? [16:10] I guess ICD support for mesa is missing [16:12] tseliot, the problem is that every opencl implementation ships its own libOpenCL.so and we're not spliting those into separate pkgs [16:12] because yea, if you want to have multiple icds, it's just not possible to install all [16:13] tseliot, debian is going the right way these days, they split all the icds into separate pkgs [16:13] tseliot, of course that would mean we'd have like 5 pkgs for each nvidia driver version :/ [16:14] and if there wasn't enough of those :P [16:14] mhr3: debian has the nvidia packages split into several packages, which is something I'm trying to avoid unless it's really necessary [16:14] tseliot, right, but as i said, by doing that they have full icd support [16:15] you can mix and match any icd with any libopencl [16:16] mhr3: I can probably split cuda and openCL in 14.04. Backporting drivers to 13.10 or 12.04 would then be a bit of a hell though... [16:17] tseliot, well you know, with all the available icds out there it would be nice to be able to have more than one [16:18] tseliot, as in pretty much every laptop these days can have 3 opencl platforms - cpu, integrated gpu, discrete gpu [16:18] and our packaging is limiting this to just one [16:19] mhr3: I'll probably do this in the transition from 319 to 331. We're going to have a separate kernel module anyway for cuda... [16:20] tseliot, would be nice, the minimal thing that would be needed is just to move the libOpenCL.so into a separate pkg [16:21] meaning the .so that's current in nvidia-[ver] [16:21] and fglrx-... [16:22] mhr3: right [16:22] and installing the icd into /etc/OpenCL/vendors... (at least i think that was still missing last time i checked) [16:23] mhr3: well, currently there are links in that directory which point to the actual files [16:23] but yes, I'll make sure it's all in the right place [16:24] tseliot, i wrote http://mhr3.blogspot.co.uk/2013/06/opencl-on-ubuntu-1304.html some time ago, so feel free to ping me if you'd need me to test a new layout or something :) [16:26] mhr3: thanks. I'll also have a look at the Debian packages. This time I think I'll fix things for good [16:27] tseliot, awesomeness :) [16:27] :) [16:35] darkxst: here? [16:38] RAOF: no glitches here.. [16:38] ati just worksforme on trusty [16:38] except if I use glamor, unsurprisingly [16:40] darkxst: I've updated the media-keys patch to just always start the keygrabber except for 'gnome' [16:41] seb128: meeting time? [16:42] mlankhorst, oh, sorry, we are skipping this week since most of us are in Oakland, technically it was also an hour ago (DST change) [16:42] oh right, anyway.. android to dma-fence porting, more drm-dkms backporting, mesa 9.2.2 upload, working on glamor-egl bugs [16:42] gone :P [16:43] mlankhorst, thanks ;-) [16:53] ChrisTownsend, hi, busy ? question . wrt bugs.launchpad.net/bugs/1199015 . Any suggestions on trying to isolate the fix ? [16:53] seb128: hey, i didn't switch the dupes around this time! it was "Shaun P" the reporter of the one bug that i marked won't fix, apparently [16:56] ritz: Hi, I just looked at the bug and off the top of my head, I don't have any ideas. Is this only specific to Precise? [16:56] anything before 0.9.10 [16:56] fixed in saucy [16:58] ritz: Ok, that's good info. I will need to debug to have some idea, but I have a pretty big backlog atm. I will try to keep it on my radar and see if I can at least see about getting this fixed in Precise. [16:59] hmm, thank you [17:03] ritz: I'll assign myself to the bug to make sure I remember it:) [17:03] dobey, hey, don't worry, I'm going to keep it this way until I get free slot to look at that, I'm going to reopen/file a new bug then [17:05] seb128: ah, ok. just didn't want you to think i was trying to be a pain. :) [17:06] ChrisTownsend, thanks. I am kindda of tired of bisecting compiz, rebuilding compiz and unity [17:06] what is the usual process here ? [17:07] ritz: Umm, you're doing it the way I would:) And yes, it is very tedious. [17:35] Laney, https://bugs.launchpad.net/ubuntu/+source/gnome-settings-daemon/+bug/1232419 [17:35] Launchpad bug 1232419 in gnome-settings-daemon (Ubuntu) "[xsettings]: gnome-settings-daemon crashed with SIGSEGV in notify_have_shell()" [High,Confirmed] [17:35] dobey, thanks ;-) [18:11] seb128: I will have a look at the bug tomorrow and I'm going to bed now.. [18:11] happyaron, hey, thanks [18:12] maybe bug #1245925 also needs comments from people working on it. [18:12] Launchpad bug 1245925 in maliit-framework (Ubuntu) "Troublesome export in /etc/profile.d/maliit-framework.sh" [Undecided,New] https://launchpad.net/bugs/1245925 [19:32] Laney, could you commit https://launchpad.net/ubuntu/+source/gnome-control-center/1:3.4.2-0ubuntu0.12 and https://launchpad.net/ubuntu/+source/gnome-control-center/1:3.4.2-0ubuntu0.13 to lp:~ubuntu-desktop/gnome-control-center/precise/? [19:34] oh that exists? [19:34] YES! YES I CAN! === m_conley_away is now known as m_conley === m_conley is now known as m_conley_away === ChrisTownsend1 is now known as ChrisTownsend [20:34] desrt, online? [20:40] cyphermox: can you pkg ack http://10.97.0.1:8080/view/cu2d/view/Head/view/Unity/job/cu2d-unity-head-3.0publish/lastSuccessfulBuild/artifact/packaging_changes_libunity_7.1.3+14.04.20131029.1-0ubuntu1.diff ? [20:42] attente, https://bugzilla.gnome.org/show_bug.cgi?id=697002 [20:42] Gnome bug 697002 in general "Grab and emit a signal when XK_ISO_Next_Group is pressed" [Normal,Resolved: fixed] [20:52] cyphermox: nevermind [20:59] Laney, ok, guess I will need to rebase bug 1232419 then [20:59] Launchpad bug 1232419 in gnome-settings-daemon (Ubuntu) "[xsettings]: gnome-settings-daemon crashed with SIGSEGV in notify_have_shell()" [High,Confirmed] https://launchpad.net/bugs/1232419 [20:59] seb128, why is pitivi in main? [21:00] mterry, we were talking about installing by default, back in the day [21:00] I wonder if we put it to supported/dvd by then and let it there [21:01] seb128, seeded-in-ubuntu isn't showing anything... [21:05] mterry, I can't find anything either, weird ... I guess it's worth asking on #ubuntu-devel [21:11] Mirv: so no need for an ack of that? [21:13] cyphermox: ahum, I thought it was a core dev approving the commit but actually yes please do ack [21:14] ok, just a second [21:14] cyphermox: but it's this adding of two symbols http://bazaar.launchpad.net/~unity-team/libunity/trunk/revision/305 [21:15] plus ABI bump [21:15] but I fumbled since I thought it was acked by a core-dev already so I pushed the button already [21:16] Mirv: I don't see the thing now [21:16] right [21:17] well, yeah, it looks fine [21:17] it's a little non-obvious at first because that's just exposing generated getters/setters [21:18] yeah I saw that but it seemed correct then [21:18] yeah [21:18] it looks right, no worries [21:19] thanks [21:23] Mirv: I'm a bit stuck w.r.t VPN atm, can you please check if address-book-app was rebuild recently to get rev 109? [21:24] cyphermox: just a sec [21:25] cyphermox: 53 minutes ago, but bzr 108 still. do you want me to rerun it? apparently 109 came around 15mins after that [21:33] well rerunning anyhow [21:35] thanks [22:04] cyphermox: address-book-app bzr109 in PPA (built) [22:23] Mirv: thanks! [23:23] * Mirv added lp:ubuntu-settings-components to our TODO list, but didn't assign myself to it yet. it's free to take if you have time [23:23] ie daily releasing that [23:23] thanks [23:57] Laney, still planning to commit those precise gnome-control-center SRUs to the vcs? (/me wants to do another SRU there) [23:57] did [23:57] or did I? [23:58] oh that's amusing [23:58] I did gsd [23:59] Laney, hehe ;-) === jjardon_ is now known as jjardon [23:59] are you SRUing that sleep thing? [23:59] Laney, I was going to, if you want to do it though feel free [23:59] not especially