=== smb` is now known as smb [07:33] zequence, yo .. about ? === fmasi_afk is now known as fmasi [08:12] apw: yep [08:12] xequence, i thought you were on holiday [08:12] and ... what happened to your z [08:13] apw: I'm in Switzerland, yes. More like a holiday from my holiday [08:13] apw: I left the z at my server, and cant login to it [08:13] xequence, hehe ... well as you are on holiday i am doing the rebases for Q and R for lowlatency for you [08:13] xequence, ... but it might make sense if you added me to your ubuntu-studio git repo so i can push the result for you [08:14] apw: oh, thanks. Yeah, sure. [08:15] xequence, i am awhitcroft on github [08:15] apw: Yep [08:20] apw: Think I found the right buttons to let you push and pull [08:21] xequence, we shall see shortly thanks [08:31] xequence: Hey, want to add adconrad as a member to that too, just in case? :P [08:36] infinity: Sure. Just a sec [08:38] infinity: Should be done [08:40] xequence: Looks good, thanks. === fmasi is now known as fmasi_afk === fmasi_afk is now known as fmasi [08:53] hey infinity [09:10] apw: hey, do you have a link to your sbuild sources.list magic script? [09:11] apw: I remember infinity mentioning it a couple of times and I had a paste.ubuntu.com link for it but it doesn't work anymore (DB was reset) [09:12] stgraber, it is in the kteam-tools repo, in the apt-pocket directory ... git://kernel.ubuntu.com/ubuntu/kteam-tools [09:12] stgraber, that is also on gitweb on kernel.ubuntu.com if you want to avoid a clone [09:13] stgraber, heh ... so there is a pretty comprehensive readme in there as well [09:13] seems to work pretty well for me, sbuild --arch amd64 -c saucy-proposed+main .dsc stylee [09:33] apw: cool, thanks [09:35] stgraber, if you are on anything newer than P then things are simpler to undo as well [10:17] apw, smb: now that 3.11 is out of proposed, we're seeing different errors (kernel BUGs) wrt the nested kvm issue. See latest details on bug 1208455. [10:17] Launchpad bug 1208455 in linux (Ubuntu Saucy) "general protection fault running apt-get inside double nested kvm VM" [High,In progress] https://launchpad.net/bugs/1208455 [10:18] jodh, you are really trying to hurt yourself arn't you [10:19] apw,smb: I'm going to try running everything locally for now and will report back but could you guys take a look at this again? [10:19] apw: :) [10:19] apw, jodh Can try to reproduce though [10:20] jodh, Have you documented clearly what steps you run _in_ the bug report [10:20] ? [10:20] jodh, so to confirm this is 3.11 in all three cases or somethign else, and it is the host who is blowing chunks [10:20] smb: yes - the latest comments show that it happens running bin/prepare-testbed [10:21] apw: well, I can't as this is via canonistack so I don't know what kernel the host is actually running. [10:22] jodh, and which kernel croaks [10:22] jodh, so ... waht we need is a canonistack recipe, if you can't put that in the bug, email it to smb and i [10:22] and me [10:23] * smb needs to find out whether his credentials still work and if he got them [10:23] apw: the kernel running within the kvm instance that is spun up by prepare-testbed (ie the first kvm created on top of the (possibly kvm) canonistack instance I'm given by euca-run-instances). [10:24] apw: recipe? You mean the steps I took to create the canonistack instance? [10:24] jodh, yeah so ... that is gobblydeegook. we are dumb kernel people here, who are scared of userspace apps. a nice "cut-n-paste these 7 commands and watch the fireworks" [10:25] jodh, Yes, so we can follow that as cloesely as possible [10:25] apw: ok, give me a couple of mins... [10:25] we really are clueless about most things, including canonistack :) [10:25] though i have used it so i do have creds at least [10:26] Well we got machines able to run kvm... so why would we :) [10:31] apw, On unrelated note (before I forget after the context switch): what am I supposed to do with Xen-4.3. All but the xen-api/xcp related package which have dependencies compile ok with the changes. blktap, xen-api-libs ad xen-api do not but due to not being compilable at all anymore in saucy. I can change my xen package to not include blktap which may drop some functionality but then we have not shipped before any devel libs that would all [10:31] ow to compile against anything there... [10:32] all + ow -> allow [10:36] Maybe a question for infinity (if awake or when woken) [10:39] yeah i think that is beyond my current knowledge. we are a bit stuck if xen-api is unhappy i know it stayed unhappy for a while last time [11:36] apw: hey, any hope of getting my small kernel patch reviewed in the near future? ;) [11:36] apw: https://lkml.org/lkml/2013/4/24/518 === fmasi is now known as fmasi_lunch [12:06] stgraber, ok what we should do is get you to resend it, to the same people, and add me to the CC:, with [RESEND] on the subject line (include the Ack from serge as well). then people are reminded and i can review it too, which it looks ok to me [12:07] and i can say "hmmm this has been hanging about for a long time" etc [12:07] as it is from 4 months ago [12:13] stgraber, looking at the patch abstractly i wonder if it might make sense to make it like a prefix, so like %hp was the pid in that space so we could use only one new letter for any number of 'not the one inside the container, but the one outside the container' [12:13] i asume pids have the same issue [12:14] i mean uids [12:29] * henrix -> lunch === ghostcube_ is now known as ghostcube [13:13] * smb notes to jodh that he has not received any notes on setting up a canonistack instance, yet === fmasi_lunch is now known as fmasi [14:37] smb: I've put step-by-step instructions on bug 1208455. [14:37] Launchpad bug 1208455 in linux (Ubuntu Saucy) "general protection fault running apt-get inside double nested kvm VM" [High,In progress] https://launchpad.net/bugs/1208455 [14:38] jodh, How convenient that this is a bug I am not subscribed to, yet. :-P [14:40] Done now [14:40] Probably should have done before [15:01] Hi how do I file a bug on Microsoft Sorface Pro wifi does not work ? [15:03] does running ubuntu-bug linux sufficient ? [15:08] infinity: Is it too loate so SRU the raring lowlatency to the lTS? === fmasi is now known as fmasi_afk [15:38] jodh, So I followed your instructions and seem to have no space left on device on /... [15:39] smb: ok. In which case, "rm -rf /tmp/adt", create a directory in /mnt that user ubuntu can write to (say /mnt/adt) to and then "echo BASEDIR=/mnt/adt > ~/.adtrc". [15:40] smb: autopkgtest should then use /mnt which has lots of space on medium guests. [15:41] jodh, is that what *you* do as well? [15:42] smb: I think something has changed recently in adt. It's rather difficult to document every single step and work-around I've had to use to get this far. sorry :( [15:44] jodh, Just makes it a bit hard to get the same experience... :/ [16:14] apw, have you done a saucy live install lately ? [16:15] rtg, lately, not in the last two weeks no, i did one just before sprint [16:15] rtg, and i did a server install like yesterday on armhf ... [16:15] rtg, if it would be useful i can do a VM one now [16:15] apw, I checked that I wanted updates installed from the net during installtion, but on first boot there were like 100 packages that needed updating. [16:16] rtg, that checkbox is "download updates while installing" [16:16] it doesn't install them i don't believe [16:16] apw, just caches them ? [16:16] yeah i think that is right ... i seem to remember being caught out by the same [16:17] when discussing how a kernel in -updates would avoid something something [16:17] huh, I'm not convinced of the usefulness of that "feature" [16:17] no me either [16:18] apw, though one could argue that if the live CD kernel booted, then you'd at least want that one installed. [16:18] heh yeah you always want the first one [16:30] smb: https://merges.ubuntu.com/libv/libvirt/REPORT [16:37] apw, we're not building an alternate install for saucy anymore, right ? [16:38] we still build alternates yes, d-i images [16:39] apw, where do they live on cdimage.ubuntu.com ? looks like we're only doing server and live [16:39] is not the server image the alternate installer ? [16:40] apw, the alternate used to include desktop whereas the server does not [16:40] oh then ... no :) [16:40] not that I'm really caring. just curious [16:41] Yeah I think that got dropped. Now that the desktop supports LVM... [16:44] zequence: Almost certainly too late to figure out what that would mean and how we'd do it, yeah. [16:44] rtg: No desktop alternates, but you can install desktop fine from a d-i netboot or a server CD (just that you have to hit the interwebs for the packages). [16:45] infinity, my get dailies script was complaining about alternates no longer existing, which is what caught my attention [16:46] infinity, Ni, err I mean Xen MRE, maybe, please? [16:46] rtg: Ahh. We dropped them in Q, I think, your script's been out of date for a while. ;) [16:46] infinity, could be.... [16:47] smb: Did you get someone to sponsor the uploads? [16:47] infinity, No, I am trying to get you to [16:48] infinity, apw Always runs faster than lightning when I mention those three letters [16:48] Heh. [16:50] infinity, what are you doing here ! [16:51] off to bed with you ! [16:51] ogra_: I had a meeting this morning. Blame ogasawara. [16:51] oh, evil girl ! [16:52] Muahahah :) [16:52] :) [16:53] One never knows when infinity sleeps... or which timezone he actually is in... :-P [16:54] smb: Alright, if you can't talk Andy into sponsoring it for you, bug me on the weekend, and I'll see about doing a one-shot sponsor-and-SRU-review. Just need to set aside some time to do so. [17:08] infinity, Usually I would want to avoid bugging people anywhere around weekends, but then I am getting slightly desperate with that, so will try [17:08] smb: Well, I'm happy to do "community" work on weekends occasionally, and Xen currently sits on a bit of a fence there. [17:09] I'm still trying to figure out who I need to convince to make it a first-class citizen in everything we do. [17:10] Yeah that might be a difficult task... [17:12] infinity: I was away for a while, in case you answered. I was asking about SRU of lowlatency for the LTS. Is it too late to do that? [17:13] smb, am i expecting to see crash on the merges list again ? [17:13] apw, For saucy you might as upstream did a 7.x.x [17:14] Actually I had been looking and we had issues with hardening options [17:14] infinity, the issue with xen-4.3 is the xapi stuff whcih all went wrong last time we did xen [17:14] apw, Oh, I asked him about the other Xen [17:14] there is anohter xen [17:15] xequence: 10:44 < infinity> zequence: Almost certainly too late to figure out what that would mean and how we'd do it, yeah. [17:15] Yeah the stable MREs for P, Q and R [17:15] oh ok [17:16] those i don't think i ahve seen anyhow [17:16] infinity, So yeah actually for 4.3 there is this xen-api madness. They got rdepends against xen but are not even buildable right now [17:16] or if i have i have painfully blmked them [17:16] apw, I tried not to "overload" you with xen [17:16] heh [17:25] * rtg -> lunch [17:26] infinity: Ok. I'd like to do one for 12.04.4 though. How long before would it be good to bother you on that? [17:30] * henrix -> EOD (SIGBOOSE) [17:34] xequence: Sometime after next week. [17:34] xequence: Given that you don't actually put lowlatency on your install images (or am I misremembering?), this isn't something that needs to be tied to a point release, really. [17:35] xequence: You could jam an lts-raring kernel in precise and just socialize the idea that it's there for people who want to use it. [17:35] xequence: Assuming you commit to the same cadence as the Canonical SRU people (as you've done for your P/Q/R rebases), I'd be alright with letting that in. [17:45] infinity: We do have lowlatency on the images, so therefore it would be nice to get it on there too. But, I'm not at all against getting it there non timely as well [17:46] Ah, but you meant like using a PPA [17:46] xequence: Right, let's talk about it after .3 is out the door, then. [17:47] xequence: No, no. I don't mean a PPA. I mean in the archive proper. [17:47] infinity: cool [17:47] alright [17:47] * apw relocates away from his mic [17:53] * smb relocates away from work [18:56] * apw goes make food [19:20] rtg: do we file a support ticket to get tyhicks access to zinc and the kernel build machines (tangerine, ...)? [19:25] jjohansen, yes, he needs to be in the LDAP group [19:25] rtg: okay thanks [19:30] jjohansen: should I open it or have you already started? [19:30] tyhicks: feel free to open it, I have not started it yet [19:31] will do [20:13] * rtg -> EOD