=== RAOF_ is now known as RAOF === 92AAAUIMO is now known as RAOF === smb` is now known as smb [07:02] reboot [07:02] login === fmasi_afk is now known as fmasi [09:01] oh my ... who decided to make the framebuffer console builtin on grouper ... [09:01] * ogra_ sighs [09:04] ogra_: this one? - git://kernel.ubuntu.com/ubuntu/ubuntu-nexus7.git [09:04] well, the last upload [09:05] it was a module in the former upload, that i could at least rm at build time [09:05] (imge build time that is) [09:06] we dont want anything to initialize the framebuffer before surfaceflinger or Mir do that ... fbcon somehow did get enabled (as a module) as part of the "we need a /dev/console for upstart" change ... [09:06] while we want a /dev/console we dont want the framebuffer console ... [09:08] * ppisati ponders... it sucks to be a framebuffer console nowadays... [09:08] haha [09:08] on android for sure [09:11] argh, and all the other kernels were switched to "quiet" [09:11] GRRRRR ! [09:11] so we wont have any info on the ramconsole for debugging anymore [09:12] ogra_: you wanna talk with rtg as soon as he wakes up (~3hrs) [09:13] yes, just writing a mail, i think that was a request from slangasek (only for grouper though) [09:38] bjf, henrix: hey, I have a last-minute change or two for the ubuntu-quantal kernel :/ (for -35.5x) [09:38] tjaalton: i believe sconklin started prep'ing the Q kernel yesterday [09:39] henrix: yeah I noticed -35.56 got tagged.. [09:40] tjaalton: i guess you'll need to convince him to respin the kernel :) [09:41] henrix: right.. so in 4-5h maybe? [09:42] tjaalton: yep. maybe you can start by sending the patches to the mailing list so that we may get the ACKs before that [09:42] sure [09:42] I'll prep the second one [09:42] tjaalton: cool. thanks [09:57] ppisati, ... [09:57] ogra@anubis:~/datengrab/bastel/nx7/linux-grouper-3.1.10$ fakeroot debian/rules binary-grouper [09:57] make: *** No rule to make target `binary-grouper'. Stop. [09:57] shouldnt that work ? (i dont want to build the tools) [09:58] ogra_: should be, hold on [09:59] ppisati, heh, false alarm, ignore me [09:59] i forgot to do the dpkg-architecture export [10:00] DOH! [10:01] since i upgraded this box to saucy, i lost gcc4.6 [10:01] and that kernel explicitely requires 4.6 [10:01] AKA i can't build it anymore [10:02] that cant be ... the package gets built on saucy in the archive [10:02] ogra_: i'm cross compiling here [10:03] ah, got it [10:46] henrix: ok, pushed the branch and email sent [10:49] tjaalton: great, thanks. i'll have a look at them in a bit. [10:54] thans [10:54] +k [10:54] it's kinda backwards though to add the pci-id's before they are even on saucy, but oh well .) [10:54] :) [11:03] tjaalton: ugh, i guess the 2nd patch generation is automated, right? :p [11:06] henrix: the drm_mm* madness? [11:06] tjaalton: yep [11:07] well I did double-check against the old version for correctness, should be right :) [11:07] also built fine, dkms too [11:07] tjaalton: yeah, and it will be a pain to maintain :) [11:08] so it should match 1:1 to the old one, a650b0fdeda61 shows how to rename next time if such trickery is needed again [11:08] sure is [11:08] tjaalton: that's why i was asking if you had some script to automate that renaming [11:08] but, I think it wont receive many updates anymore [11:08] not to the old names [11:08] old-old names that is, the recent rebase was easy [11:10] tjaalton: cool. anyway, we're already maintaining that ubuntu/i915 thing, so maintenance will always be painful [11:10] I just wish we'd be able to just suck it up and pull all of drm/* as sru.. [11:10] tjaalton: one of the things missing for that is a bug # :p [11:10] henrix: 3.8.13.x won't receive that many updates, I hope. and the old version was kinda left behind anyway [11:10] bug for? [11:11] A little while back I was testing a kernel bug with jsalisbury and so installed a specific kernel, 3.8.0-030800-generic. I'd like to stop using that and get back to running the most recent kernel from raring. How do I do this? I am frightened of apt-get removing the kernel that I am currently running ;) [11:11] the rename? maybe reuse 1188305? [11:11] tjaalton: oh, sorry. i missed the BugLink in the 1st patch [11:12] yeah, so "some oem's" want us to have all these new pci-id's in the archive by end of next week.. [11:13] userspace is already being dealt of, uploaded to -proposed just not acked yet [11:13] tjaalton: ok, cool. so lets just wait for sconklin and bjf and see what they say. maybe they are already waiting for these patches...? [11:13] but if the kernel one could be squeezed in it would roughly fit that timeframe as well [11:13] henrix: nah, I was just late to the party.. [11:14] tjaalton: heh, so lets hope they had a good night sleep :) [11:14] indeed :) [11:15] I did mention this late last night (my time) but I guess they were gone already [11:32] iam getting a amd-vi event logged io_page_fault device=0a:00.0 domain=0x001f message on the x64 kernel 3.8.0-25-generic on a asus sabertooh 990fx r2.0 (this disables my nic) (i think i found related bug report here :http://linux-kernel.2935.n7.nabble.com/AMD-Vi-error-and-lost-networking-with-r8169-td630152.html )(but in that bug report the device and domain are different) [13:03] * henrix -> back in 15 [13:20] gompa, sounds like an iommu getting all humpy, i would file a but, at least that way people can track what h/w etc you have [13:20] s/but/bug [13:23] * ogra_ hugs rtg_ [13:24] rtg_, yeah, theh quiet thing is simply useless (we dont (and likely wont) set quiet on the cmdline but its a patch we dont need to carry [13:24] ogra_, do the rest of the Nexus kernels have the same FRAMEBUFFER_CONSOLE issue ? [13:24] android explicitly disabled fbcon, so they shouldnt [13:25] that came in with the console fixes for boting grouper in the flipped setup ... [13:25] ogra_, why was slangasek hot for the 'quiet' patch ? [13:25] all other HW seems to properly just use the ramconsole, its a tegra thing [13:25] because he didnt understand that we wont have any console output ever [13:26] unless we actually patch the kernels, but that would mean a lot of extra work for porters (exynos for example doesnt boot if you enable fbcon) [13:26] ogra_, ok. I won't bother reverting for now since its a harmless patch as long as 'quiet' isn't on the boot command line. [13:26] yeah [13:27] ogra_, upload grouper then ? [13:27] yup, i tested it locally and it works fine here [13:27] ogra_, ok, will do [13:27] thanks :) [14:24] tjaalton, you have made me very sad this a.m. I will accept your nasty commits and we will respin. this _really_ should have come in last week. [14:25] bjf: :/ sorry.. and huge thanks [14:25] tjaalton, shouldn't we apply the pci id commit to raring as well? [14:25] bjf: it's there via 3.8.13.2 [14:26] tjaalton, ok, i'm not awake enough to check [14:28] infinity, ^ respinning Q [14:34] btw, the ReleaseInterlock page is old, shouldn't saucy have one? [14:35] hmm that's not what I wanted.. kernel-sru-workflow.html instead [14:54] * ppisati steps away for ~10mins [15:08] apw, ogasawara: anything you want in saucy before I upload ? there are a couple of d-i fixes that need to get propagated. [15:10] rtg_: nope, I'm good here [15:10] rtg_: I assume the dkms issues tseliot was sorting are done? [15:10] ogasawara, heck if I know. [15:11] ogasawara, I just annoy apw periodically who then annoys tseliot [15:11] rtg_: I guess I should clarify, this is the 3.10 kernel you're going to upload right, not 3.9 [15:12] ogasawara, don't you know ? I'm a nexus hacker these days. who thought it was a good idea to do 4 of those dang things ? [15:12] ogasawara, no, 3.9 for now [15:12] rtg_: heh. ok then nevermind my babbling about dkms bits. that's only an issue for 3.10 [15:13] ogasawara, right. IIRC tseliot said he hoped to be ready by end of the week [15:13] rtg_: yep, I recall the same [15:14] aiui only fglrx left [15:15] Daviey: ^^ it would be good to get your formal list of dkms packages you guys care about if you want us to perform advanced testing before we upload. I sense an impending upload soon. [15:15] ogasawara: yes, I'm working on fglrx [15:15] an impending dkms explosion :) [15:15] heh [15:16] ogasawara: Ah, i thought my team were doing the advance testing this time.. so i held off on sending a list. [15:17] Daviey: ok cool, in that case...heads up, we're probably going to upload a 3.10 kernel soon. [15:17] Daviey: would probably be good for you guys to test sooner rather than later [15:21] ogasawara: I was looking to get some stuff validated against the PPA tomorrow. Which means that we'll know of issues then. [15:21] I wouldn't like you to block on this tho. [15:22] Daviey: ok great, tomorrow should be good. I don't think we'd upload until EOW anyways at best. === rtg_ is now known as rtg-afk [16:12] bjf: Ack on the Q respin. === rtg-afk is now known as rtg_ === fmasi is now known as fmasi_afk [18:03] * rtg_ -> lunch [18:06] * henrix -> EOD [18:09] apw, rtg_: is there a wiki for doing test builds of the phablet kernels [18:28] jjohansen, yeah you want er [18:28] apw: yes please [18:29] jjohansen, https://wiki.ubuntu.com/Kernel/Dev/AndroidKernel [18:29] jjohansen, i cross build my kernels to make them for that use [18:30] yeah [18:42] jjohansen, what apw says. for what you are doing you can just cross build them [18:43] rtg_: yep thanks [18:50] * apw calls it a day ... i hate hyper-v and i love beer, so the decision is easy === kees_ is now known as kees [20:08] * rtg_ -> EOD === yofel_ is now known as yofel