=== [1]vncm is now known as vncm [08:44] apw: fglrx, nvidia, and bcmwl are all ready for linux 3.18 in vivid now. You might want to let rtg know (apparently I'm never online when he is) [09:22] tseliot, ack thanks, in theory the dkms matrix sort itself out and show that too [09:49] jibel, hey, the dkms testing, when a new dkms package is uploaded, how long after that will a new run for it trigger [09:51] apw, same than the rest, we check once a day. I could schedule it more frequently if you need to. [09:51] apw, for info I filed https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=766668 to add autopkgtest support to all dkms packages [09:51] Debian bug 766668 in autodep8 "Add support for dkms packages" [Normal,Open] [09:52] apw, the patch is attached waiting for a review by the maintainer [09:52] jibel, nice on the debian stuff indeed [09:52] apw, then dkms tests will run each time a dkms test of its deps are uploaded [09:52] and block the migration if they fail [09:53] jibel, nice [09:53] jibel, ok so the fix for the dkms package in question was just 24m ago so thats not a supprise it has not appeared yet [09:54] jibel, is there a button an apw can click to jiggle that one into action, as daily is plenty fast enough on the normal run of things, but it would be nice to be able to jiggle things once in a while [09:55] apw, you should be able to trigger tests manually from http://d-jenkins.ubuntu-ci:8080/view/DKMS/ [09:55] apw, if you are not allowed the CI team can give you the right privileges [09:56] so if i just wanted to run the bcmwl test, where is the button for that for you, i seem to remember having retry buttons so i suspect i have some priviledge at least [09:56] (then i can look see if it is there :)) [09:58] apw, for bcmwl + kernel team ppa: http://d-jenkins.ubuntu-ci:8080/view/DKMS/view/V%20KT-PPA%20-generic/job/dkms-vivid-release_canonical_kernel_team_ppa-generic-bcmwl_kernel_source/ [09:58] apw, for bcmwl + kernel in proposed: http://d-jenkins.ubuntu-ci:8080/view/DKMS/view/V%20-generic/job/dkms-vivid-release-generic-bcmwl_kernel_source/ [09:58] jibel, and i'd expect like a "run me" button on there on that left menu ? [09:59] apw, yes, there should be a "build" button on the left menu. Are you logged in? [10:00] ahh no i am not, and now i am i see a build [10:01] jibel, and that build says "make a new build for the latest kernel/package combo" i assume [10:03] jibel, and ... that worked, excellent [10:03] (the one you tickled for me) [10:03] apw, it'll build with the most recent version of the packages in the archive + proposed + any ppa enabled. [10:04] that, is exactly what i need, thank you [10:04] yw [10:23] tseliot, have all of the below gone in vivid? [10:23] fglrx REMOVED? [10:23] fglrx_updates REMOVED? [10:23] nvidia_173 REMOVED? [10:24] (the first two seem to now have _core versions) [10:36] apw: not really [10:37] apw: fglrx-core is just an addition to fglrx [10:37] fglrx-core is the one with the kernel code [10:37] fglrx MISSING [10:37] fglrx_updates MISSING [10:37] so those two not being tested, is wrong i assume [10:38] apw: right, fglrx-core and fglrx-updates-core are the dkms packages [10:38] nvidia-173 was indeed removed [10:38] tseliot, ok now i am confused has the fgrlx dkms packge name changed then ? [10:39] (I thought you were saying they had been removed from the archive) [10:39] apw: fglrx-core contains only the bare minimum (DKMS stuff, and non X specific libraries), whereas fglrx is specific to X [10:40] ie was the dkms bits in fglrx in U and in fglrx-core in V ? [10:40] both U and V have the fglrx and fglrx-core split [10:41] tseliot, ahh but did the split occur in the developement of U ? [10:41] put simply, we always need to test the -core package [10:41] yep [10:41] ok then _that_ makes my results make sense :) thanks [10:41] :) [10:54] jibel, hey ... during U the fglrx "went away" and was replaced by fglrx-core as a dkms package, i note [10:55] jibel, that during that time the job names went wonky as well (3.16.0-24-generic /2:14.201-0ubuntu2) is that indicative that that result is not relevant perhaps ? [10:57] what does the name before the / there mean anyhow i guess [11:00] if that is the name of the dkms module we _find_ in the package then i guess it is indicative [11:02] apw: BTW, when is the lts-utopic backported stack due in 14.04? [11:04] tseliot, the kernel is already there i think [11:05] linux-lts-utopic | 3.16.0-25.33~14.04.2 | trusty-proposed | source [11:05] apw: in -proposed or in -updates? [11:05] sitting in proposed at the moment it seems [11:05] oh, good [11:05] as that would break my packages [11:05] ok, maybe only bcmwl, IIRC [11:27] apw, if the name changed in U then only results for fglrx-.*-core are relevant. Tests are added automatically but not removed. [11:28] apw, the name before / is PACKAGE_NAME found in dkms.conf of the module [11:28] jibel, ok so if that goes "" then it is very likely that that package can be ignored as a result ... purfect [15:33] jibel, this replication who owns fix that when the two do not match ? [15:54] jibel, do we run any of this tesitng for the lts-backport kernels ? === jsalisbury changed the topic of #ubuntu-kernel to: Home: https://wiki.ubuntu.com/Kernel/ || Ubuntu Kernel Team Meeting - Tues November 18th, 2014 - 17:00 UTC || If you have a question just ask, and do wait around for an answer! If the question is should I file a bug for something, likely you can assume yes. [22:24] What packages are needed to rebuild the kernel? [22:55] j4s0nmchr1st0s, the build dependancies on the kernel are the packages you need [22:57] j4s0nmchr1st0s, https://wiki.ubuntu.com/Kernel/BuildYourOwnKernel