=== panda is now known as Guest7646 === dimmortal1 is now known as dimmortal [06:53] hello! I want compile a custom kernel with generic-pae config but the https://help.ubuntu.com/community/Kernel/Compile doesn't help :( [06:54] how do I do this? I only compile a custom kernel same config of generic-pae [06:54] for start [06:55] after I want apply a custom patch but I need apply with patch program or is there any way more organized? [06:55] I see that kernel-package has the --apply-patches but I don't know if it works with custom patch [06:58] I tried extract the /usr/src/linux-source-2.6.38.tar.bz2, copied the config of generic-pae in /boot and compile with make-kpkg, but I always get kernel panic === smb` is now known as smb [08:27] mfilipe, https://wiki.ubuntu.com/Kernel/BuildYourOwnKernel might help more. [08:29] smb, it is very confuse :( [08:30] I see tree ways to compile kernel: (1) using commands of kernel vanilla, (2) debian/rules and (3) kernel-package [08:30] some people say that kernel-package is the better solution [08:31] but when I compile my kernel with it, I get kernel panic [08:31] I will try with debian/rules [08:32] Not if you want a kernel compiled as the provided one. Then dpkg / debian is the way to go. Everything else either has not the whole packaging or is not too well tested (we don'T use kernel-package) [08:40] smb, the better way to apply a patch is in kernel source with patch command or is there a way more organized? [08:42] mfilipe, For just applying the change patch is the way to do it. And the simplest [08:42] thanks a lot for your help! :) [08:43] Usually "patch -p1 < " === yofel_ is now known as yofel === lag` is now known as lag === lag is now known as Guest77582 === maks_ is now known as maks === maks is now known as maks_ [12:33] rebooting === Guest77582 is now known as lag === lag is now known as Guest52725 [14:00] hello! I want apply a patch in generic-pae. So, how do I do to create a kernel called i915patch using debian/rules based in generic-pae? [14:01] mfilipe, You can modify the version number in debian.master/changelog [14:01] So on the first line [14:01] mfilipe, everything you need to know is somewhere in here: https://wiki.ubuntu.com/Kernel [14:01] change 2.6.x-y.z -> 2.6.x-y.6+i915patch [14:02] err last 6 should be z [14:04] ogasawara, did you ever find the armel build problem with oneiric? I rebased to -rc6 last night and pushed to master-next. [14:06] tgardner: I've finally been able to reproduce it on kakadu. It's basically because the compiler is optimizing a loop in megaraid_sas into a divmod call, but the error message isn't verbose enough to point out where, so I'm still hunting it down. [14:08] ogasawara, has is struck you that kakadu is vaguely scatological ? [14:08] tgardner: hehe, not till you mentioned it :) [14:09] ogasawara, now I've ruined your mind with an idea :) [14:09] * smb thinks he might ask tgardner whatever that means over a beer... [14:10] smb, giyf [14:12] wtfmgiyf [14:12] heh [14:12] smb: google is your friend [14:13] ogasawara, Heh, ok. Thanks. :) [14:14] Usually I ask leo whenever our leader comes up with a even more mystical attribute for a release... [14:16] Ok, so wikipedia knew it... :-P [14:16] tgardner, armel doesnt build atm, gcc issues that cause apt to segfault [14:16] dont even try :) [14:17] Bug 774175 [14:17] Launchpad bug 774175 in apt "apt segfaults on armel in oneiric" [Critical,Confirmed] https://launchpad.net/bugs/774175 [14:18] smb, thanks man! :) [14:18] ogra_, ok. oneiric _is_ getting quite aways into the build but we're encountering a missing symbol error. We're not gonna worry about it too much for awhile. [14:18] tgardner, ah [14:19] well, i dont expect image builds before A1 [14:19] tgardner, btw, would you mind bringing my ultra small wlan stick to UDS ? i forgot to claim it back last time [14:21] ogra_, hmm, its so small that I haven't seen it in awhile. [14:21] I'll look around for it [14:21] that the problem with thos itty bitty gizmos. they are easily lost [14:21] hehe, or forgotten :) [14:21] well, that too === jjohansen is now known as jj-afk [15:18] tgardner: don't suppose we'll ever get armel chroots on tangerine for oneiric? [15:18] ogasawara, hrmph. I can't even get them working reliably for natty. [15:18] not on a lucid host, that is. [15:19] tgardner: I figured as much, I won't hold my breath [15:19] ogasawara, I guess we'll just have to stick with cross compiling for now [15:21] JFo, i'm going to edit: https://wiki.ubuntu.com/Kernel/BugTriage/BugStates, feel free to revert my changes if you feel them inappropriate [15:24] ogasawara, one thing we could consider is upgrading tangerine to natty. that would at least give you a natty armel schroot. alternatively, I could spend some serious time with slangasek whilst in Budapest to see if we can figure out the Lucid host issues. [15:28] JFo, i take that back, i made no changes, i'm unhappy with the page but not sure how to "fix" it [15:29] tgardner: nah, cross compiling should be fine, it's just unfortunately not triggering the gcc error at the moment [15:36] * ogasawara back in 20 [15:40] bjf, I feel the same [15:40] :-/ [15:40] been trying to figure out how best to address it [15:40] but not making much in the way of progress [15:40] JFo, i'll send email with my thoughts, mostly what I don't like about it [15:41] sounds good [15:41] I'll respond with some of what I have been thinking of as well [15:41] maybe we can figure it out [16:11] hi, is it possible to get 32+64 bit mainline kernel for 39rc6 [16:11] since rc5 32 bit is not available [16:12] it can not be that compilcated to change the config in order to skip that stupid module that does not build [16:13] since 20th april only 64 bit is there [16:13] 2 weeks for a config change? [16:22] Kano, whats the config option that isn't building. I guess we don't check the success of the daily crack builds very often. [16:23] maybe look at the log [16:24] I would have expected you had done that given your statement [16:26] olpc_dcon [16:26] fails, the option should not be that hard to find [16:27] CONFIG_FB_OLPC_DCON ? [16:28] looks reasonable [16:32] tgardner, turning off those things is normally done via the 'adhoc' patches [16:32] which are based on the bust version of the file [16:33] there is one for another of the files in that driver [16:33] apw, looks like we've a patch in the ubuntu repo that include delay.h [16:33] JFo, wiki page rant sent [16:33] adding the checksum to the patch for that additioanl file should do the trick i think [16:34] tgardner, ok i'll look at switching the adhoc to that then [16:34] apw, there is already a 0004-DISABLE-olpc.patch [16:34] yeah thats dependant on a file which seems to have gotten fixed [16:35] tgardner, i'll widen it to all the files in that driver or something [16:35] apw, cool. its your baby now. [16:46] anyone has the processor intel iX? If answer yes, what is the temperature when it compiling kernel? [16:46] it is compiling kernel* [16:47] there are iX cpus with 3 or 4 numbers [16:48] I'm using intel i5 560M and the temperature is 84C when compile kernel [16:48] mfilipe, The kernel compile surely tests how well your cooling works as it uses all cores [16:49] 84 is surely not good [16:50] it is a mobile version [16:50] i'm compiling with --jobs=4 (make-kpkg) [16:52] smb, i use a thinkpad with cooler control through software (thinkfan) [16:52] From personal experience compiling on a laptop usually causes high temperature. I would probably try to limit the upper frequence though it takes longer then [16:53] bjf, I think your explanation clarifies what I was worried about and provides a better way forward than what I had. I agree completely. [16:53] the cooler is running with disengaged level [16:53] hehe [16:54] JFo, i'm wondering if i should pass that around to a larger audience as a general comment on our wiki pages [16:54] I think so. It would definitely benefit imo [16:54] mfilipe, Well disabling fans and running cpu intensive tasks is a good recipe for meltdown.. :-P [16:55] lol [16:55] but disengaged level is the extreme level of cooler... it is running with ~5800 RPM [16:56] ah, so the wording is a bit misleading === jj-afk is now known as jjohansen [17:21] tgardner, ok rebuilding the last mainline tag to see if that fixes things [17:22] apw, kano thanks you [17:22] would be nice for oss testing [17:22] a new kernel [17:23] for userspace i have got already scripts [17:26] I installed a new kernel .deb (and the header.debs) Now none of my modules are working. for the life of me I can't recall how to complie the modules for the new kernel [17:27] if you try .39 kernel you may need to patch em [17:27] which modules do you need [17:28] ndiswrapper, plymouth for srue [17:28] plymouth has no module [17:28] but ndiswrapper was a kernel patch, logically you need that now external [17:28] jgould, it would depend which kernel you are installing [17:28] 2.6.39-999 [17:29] jgould, yeah so thats mainline and doesn't include ndiswrapper, thats an ubuntu add on and those are not ubuntu kernels [17:29] I can't use the stock .38-8-generic [17:30] jgould, why so ? [17:30] It contains the bug that affects the i915 graphic chipset [17:30] which bug is that? [17:30] is the 38-9 kernel in -proposed affected ? [17:30] https://bugzilla.redhat.com/show_bug.cgi?id=669489 [17:30] bugzilla.redhat.com bug 669489 in kernel "Include ACPI _DSM index and label support" [High,New] [17:32] apw: you lost me with -proposed [17:32] I'm just coming back to Ubuntu after about 10 years away from linux [17:32] there is a proposed pocket, this contains the likely next updates, and does currently contain a kernel update [17:33] jgould, but either way you won't find an ndiswrapper for the mainline kernels, as they don't include ubuntu modules [17:34] damn [17:37] fucking intel hardware [17:37] jgould, worth testing both the proposed kernel and the preproposed on as both have a lot of stable updates [17:37] https://wiki.ubuntu.com/Testing/EnableProposed [17:37] ^^ for the proposed kernels [17:38] https://launchpad.net/~kernel-ppa/+archive/pre-proposed [17:38] ^^ for pre-proposed [17:42] jgould, if there is no ubuntu bug for that filed, please file one and link it to the redhat one [17:43] tgardner, am i expecting brcm80211 to work at all? i just noticed i am using it on this machine, rather unexpectedly [17:43] bye [17:43] apw, 2.6.39 ? [17:43] no i seem to be running a 2.6.38-9 kernel [17:43] it works in most cases. I think it still has issues with WPA2 enterprise. [17:44] its still staging [17:44] hmmm, ok, i'll let it play this week and see how/if it copes [17:45] apw, I've never found one, but I may not have the right terms [17:46] i wonder when it switched over [17:46] jgould, then do file one with ubuntu-bug [17:47] smb, I could compile the kernel with make-kpkg!!! :) [17:47] the commend was very big: make-kpkg --rootcmd=fakeroot --initrd --append-to-version=-i915patch --overlay-dir=/home/mfilipe/Workspace/linux-2.6.38-i915patch/linux-2.6.38 --revision=1 --jobs=4 kernel_image kernel_headers [17:47] hehehe [17:51] bjf: I just ran SRU tests against the karmix EC2 kernel; but the versions being reported are weird [17:51] hggdh, weird as in ? [17:52] bjf: /proc/version_signature reports Ubuntu 2.6.31-308.28-ec2, but dpkg shows 2.6.31-308.29 [17:52] sorry, slow typing [17:55] hggdh, not sure what to say about that [17:55] sconklin, thoughts? ^ [17:56] reading [17:56] strange, let me look at that branch of the repo [18:00] sconklin, looking at the .changes for the package i uploaded, it looks right to me [18:03] bjf, the repo shows Ubuntu-2.6.31-308.29 [18:03] tgardner, yes [18:04] bjf, this might be a CONFIG_VERSION_SIGNATURE issue [18:07] hggdh, are you sure you booted the right kernel ? [18:09] tgardner: sure? No. But please see http://pastebin.ubuntu.com/603352/ [18:09] * ppisati -> gym [18:12] hggdh, hmm, definitely bizarre. [18:13] hum. Let me do a quick sanity check with smoser [18:15] tgardner: some ec2 versions have peculiarities... [18:15] Bug has been filed [18:16] hggdh, well, they shouldn't have version peculiarities. [18:18] hggdh, I downloaded the package and groveled vmlinuz-2.6.31-308-ec2 for strings:Linux version 2.6.31-308-ec2 (buildd@crested) (gcc version 4.4.1 (Ubuntu 4.4.1-4ubuntu9) ) #29-Ubuntu SMP Fri Mar 18 22:02:21 UTC 2011 (Ubuntu 2.6.31-308.29-ec2) [18:21] tgardner, i did some comparison between what is in our ppa and what is in the pocket, and they match [18:21] so I really need to check with smoser [18:21] hggdh, yep. good catch on the version though. [18:21] tgardner: the peculiarities come in play on how one boots the original instance... [18:21] hggdh, something seems wonky on your end [18:22] indeed [18:32] Ok. Back to the MacOS on this other machine for a while... [18:37] <-grabbing food... back soon [18:54] * tgardner --> lunch [20:16] * jjohansen -> lunch === tgardner is now known as tgardner-afk [20:58] hggdh, i would assume that the kernel is reporting itself wrong in that case. [21:00] smoser, its can't be. that string is only stored in one place. I think the wrong kernel got installed. [21:02] tgardner-afk, hggdh so, i suspect that what happened here is that hggdh booted a karmic instance and apt-get upgrade (or perhaps apt-get dist-upgrade) [21:02] the result is that the kernel in dpkg was upgraded, but karmic will forever boot the kernel that it booted with [21:05] smoser, that seems likely === smb` is now known as smb [21:42] hggdh, smoser Be careful when looking at versions as it is sometimes easy to catch the version of the meta-package instead of the binary [21:47] * smb -> off for real [21:48] smb indeed, will keep it in mind