=== emma is now known as em === Ming is now known as Guest92879 === smb` is now known as smb [07:14] moin [07:23] morning === amitk is now known as amitk-afk [07:58] smb: Stef, i've a problem with maint-startnewrelease [07:58] smb: 100% i screwed the packaging [07:58] smb: but i don't get why... [07:58] smb: so, basically [07:59] * smb does not know either... [07:59] hold on... :) [08:01] smb: http://paste.ubuntu.com/1146380/ [08:44] smb: morning [08:44] cooloney, good morning [08:44] these days i'm fighting with the building test on my imx6 board [08:45] i was asked to test building kdepim, gcc-4.6, chromium-browser, firefox and webkit [08:45] cooloney, how is sbuild working out for you there [08:46] i only got kdepim build successfully, but other four packages fail for some reason. [08:46] sometimes out-of-memory [08:46] cooloney, but they build on your babbage ? [08:46] (or whatever the current buildds are?) [08:46] oh, i never tried that on babbage before. [08:46] babbage may be the wrong name, i get very confused [08:47] they are supposed to be build without any error [08:47] i probabally mean panda [08:47] since we build them on panda [08:47] you have a panda right? so you could shift the sbuild containing disk over there and test [08:47] OK, if they build there with the same compiler it sounds a bit like some corruption going on on the other [08:47] and sometime the system will lockup without any output after a very long building time like 8 hours. [08:47] to see if its your environment or the machine [08:48] apw: yeah, i'm trying to setup the environment with my PandaES. bs [08:48] it might be good to literally take the same disk over there [08:48] then you'd know it wasn't that exact sbuild environment [08:51] right, i'm trying to do these on my PandaES. i tried another USB harddisk before, it looks like not very stable. [08:51] cooloney, if they are not reliable over time, they are next to useless as that is exactly what we would want to do with them [08:51] i think i need a USB disk with external power supply. i just found one, and going to do that [08:51] cooloney, ok [08:53] apw and smb, thanks for understanding. it is quite annoying about the build failure. and I will post the build logs after i tested it on my PandaES [08:54] apw: and for the lxc /root prefix issue. based on your idea, i think i can post a patch to LP for you guys discussion [08:55] cooloney, ok sounds good, point me at it when it is there, i will forget otherwise [08:56] cooloney, As apw said I think it is a good idea (if possible) to take the environment to the Panda and see what happens there. It would rule out this as a source of error. And yes, if you have something, post it [08:57] cooloney, I assume it is a change to the lxc setup? [08:57] smb: yeah, that's what i'm going to do. i got an SATA-TO-USB cable and an external SATA HD power supply now [08:58] ppisati, Try the getabis posted and let us know how that works :) [08:59] smb: i'm finishing another kernel than i'll test ot [08:59] it [08:59] apw, Yes, some message/warning would do ok. Hopefully you never place a package of the exact name there, but anyway. [09:00] smb, indeed but if you do it'll be a big confusing supprise that you cannot get it to work, then i ask you to test and it works and i pull all my (remaining) hair out [09:01] apw, No we would not want you to do that... :) [09:01] smb, exactly, who would look like bill the cat then [09:02] bill the cat himself? [09:27] apw: i posted the patch in bug https://bugs.launchpad.net/ubuntu/+source/linux/+bug/959352 [09:27] Ubuntu bug 959352 in lxc "Ephemeral containers have "/rootfs" prefix in /proc/self/maps entries" [High,Confirmed] [09:29] cooloney, thanks [09:35] cooloney, i have also made comment that as only the rootfs part of the name leaks into the container, there may be something we can do with the way the unions are placed to get rid of it should they really not be able to cope with the prefix [09:35] perhaps placing two unions instead of one or something [09:38] + # echo "rming rootfs softlink for overlayfs" >&2 [09:38] Removing/Deleting [09:38] dileks: that's is just a comment, i followed others [09:39] and I thought its rming's rootfs [09:39] apw: hmmm, i don't fully understand here. [09:40] apw: do you mean we move the unions directory to some other place instead of "rootfs"? [09:41] cooloney, me either i am just suggesting they may be able to play with mounting differnetly [09:41] i have not looked to see how its mounted [09:53] brb [10:09] * smb -> munch === yofel_ is now known as yofel [11:41] todays quantal server with LVM fails to reboot after installation [11:41] bug 1036612 [11:41] Launchpad bug 1036612 in linux "Quantal Server failed to install with LVM: VFS: Cannot open root device "mapper/ubuntu-root" or unknown-block(0,0): error -6" [Undecided,New] https://launchpad.net/bugs/1036612 [12:01] jibel, Repeatable or maybe just some failure with the initrd (which it quite often is) [12:02] smb, repeatable, all the automated tests that use LVM failed and reproduced manually [12:03] smb, alternate, which also runs a test with LVM but doesn't use a squashfs for installation passes. And it uses the same kernel [12:04] Ok, hm. mapper/ubuntu-root... would be good to know what grub.cfg looks like and what the vg / lv is named, but I assume it uses ubuntu as hostname, too... [12:05] I'll try to reproduce it here... [12:08] apw: are you aware of any network changes on tangerine or gomeisa? [12:09] apw: i'm not able to run rmadison anymore [12:09] apw: for ex., 'rmadison --architecture=source linux' seems to hang forever [12:10] well... no forever. it eventually exit with a 'couldn't connect to host' error [12:10] henrix, firewall? did that work before? [12:11] smb: yes, it did work before [12:11] smb: the makefile for the meta packages actually use rmadison [12:11] smb: that's how i found something's changed [12:12] henrix, Ok, so that really sounds like something changed in between [12:12] (not on the hosts) [12:12] yep [12:17] smb: it works [12:17] * smb assumes it is the getabis script and not the rmadison call [12:18] smb: yep, getabi [12:18] ppisati, Cool, then you can ack it [12:19] * smb closes both eyes on tangerines dmesg... [12:20] hm gomeisa as well... [12:20] ouch! :) [12:20] apw, non-accessible hardlink creation was attempted by: git [12:26] i've never seen this error before, but it looks like its yama related [12:26] sysctl kernel.yama.protected_nonaccess_hardlinks [12:28] henrix, Yes I think it is some effect of hardening [12:38] henrix, It might be we need to talk to sforshee ... ;) [12:39] ... and Joe and Colin... [12:41] ... he and I have to talk to me as well... [12:41] heh [12:42] smb, so when is that happening? when pusing or locally? [12:42] apw, Cannot say, its just entries that happen at some point [12:42] smb, not that it matters probabally ... as git will fall back to copying [12:43] Must be something I cannot write to... hm, maybe --reference /usr3/... [12:44] apw, Yeah, both are on the same disk... I guess that is what happens [12:45] yeah makes sense. and git is hiding it from you probabally [12:45] as it probally just does ln, and if that fails its copies [12:45] assuming ln failure will be cross disk [12:45] sounds reasonable... Just trying it, so my id should produce a new msg [12:45] it did [12:47] apw, And no error from git. Only scary thing is a double entry in alternates (which likely does not matter either) [12:49] oh that is odd [12:50] but I suspect it might be completely unrelated [12:51] * smb wonders whether hardlink protection might go a step too far, though [12:52] it is doing what is intended, it is preventing you linking to things you cannot write to [12:52] so that you cannot drop links to things you cannot modify in random places [12:52] to be overwritten by accident [12:52] hm, thought git would create only readable objects... [12:53] though at some point they probably are writeable, otherwise it should not be possible to create them... [12:58] i may be able to write to the directory but not link the files as they are yours [13:05] apw, Well I guess, given that my objects appear to be writeable to me (on gomeisa, my local repo objects seem to be read-only) that hardlink failure is correct [13:06] apw: what about the rmadison failure i referred above? any idea? [13:07] henrix, seems to work for me here now [13:07] rmadison -a source linux [13:07] hmm... i'm unable to run it on tangerine or gomeisa, but it runs locally on my laptop [13:08] And it still does not on gomeisa for me [13:08] I asked on #is but got no answer yet [13:08] on gomeisa i get an error; on tangerine it seems to hang forever === lynxman- is now known as lynxman [14:50] * ogasawara back in 20 [15:10] sforshee: just a Q -- my touchpad when south with the new kernel on quantal, is it still an option to try your dkms? [15:15] hggdh, that dkms package was only for testing. All the changes should be in quantal, so if it worked with the dkms package and doesn't work with quantal file a bug and assign it to me. [15:15] sforshee: will do, thanks. [15:19] mjg59, would you be agreeable to picking up the gmux switcheroo changes now, before the graphics driver issues are sorted out? That will at least get the mux state restored correctly after s3. [15:20] sforshee: Yeah, I think so [15:20] sforshee: It shouldn't break any existing config [15:20] mjg59, no it shouldn't. I'll rebase on top of your patches for the retina then and send the patches later today. [15:21] sforshee: Brilliant, thanks [15:28] live hangout on status update for single zimage for arm http://www.youtube.com/channel/UCIVqQKxCyQLJS6xvSmfndLA?v=4Lpzc_dkh9E [16:37] ** [16:37] ** Ubuntu Kernel Team Meeting - Today @ 17:00 UTC - #ubuntu-meeting [16:37] ** === jsalisbury changed the topic of #ubuntu-kernel to: Home: https://wiki.ubuntu.com/Kernel/ || Ubuntu Kernel Team Meeting - Tues August 21th, 2012 - 17:00 UTC || If you have a question just ask, and do wait around for an answer! [17:23] * ppisati -> out for a bit [18:40] * henrix -> EOD === lifeless_ is now known as lifeless [19:19] * smb -> EOD [20:02] ogra_: bug 1008713, from linux-base/linux, but guess it affects more the images depending on flash-kernel [20:02] Launchpad bug 1008713 in linux "[STAGING] package linux-tools-common (not installed) failed to install/upgrade: trying to overwrite '/usr/share/man/man1/perf.1.gz', which is also in package linux-base 3.4ubuntu2" [Medium,Confirmed] https://launchpad.net/bugs/1008713 [20:02] as flash-kernel depends on linux-base, which then doesn't let the user to install the linux-tools-common [20:03] which is a dependency from any specific linux-tools package, that covers perf === jdstrand_ is now known as jdstrand [20:18] Does anyone know if there's a way to tell if a USB port is USB 3.0 compatible ? [20:19] Of course; $ lsusb. === JanC_ is now known as JanC === bjf is now known as bjf[afk]