=== bladernr_ is now known as bladernr_afk [07:13] BenC: ping [07:18] BenC: About "Reviving Ubuntu PowerPC"... [07:18] BenC: What systems will be used as reference in this effort? [07:18] BenC: Thanks! [08:35] infinity: how is the armhf image going? [08:35] * smb steps in [08:37] ppisati: last I heard well [08:37] cool [11:19] uhm, nice [11:19] if you switch kernel release, it breaks princhanges&c [11:33] hi [12:29] abogani: I'm more interested in server platforms (as opposed to legacy mac systems) === BenC__ is now known as BenC [12:30] abogani: personally, I'm using a freescale e500mc based system (8-core, 32-bit) [12:44] Morning, i'm running xbmc with PVR on oneiric but there is a bug with the 3.0 kernel that prevents me from using my TV card. There is a patch available but hasnt made it upstream I believe. Whats the best way I can go around compiling the patch in to my kernel on my machine? [12:44] or whats the risk with downgrading to a 2.6 kernel? [12:46] BenC: Interesting. Is there a roadmap already? [12:48] Stratisphere, Mostly the risk is that it is not tested at all. So some stuff may break or it works. And probably it will be a hassle to always keep track of updates to those older kernel... [12:49] Stratisphere, But as a work-around... Since it is a different version, at least you can boot back into other kernels with grub [12:51] yeh true. What about downloading the 3.0 source and patching the fix in, if I just grab the source from https://github.com/torvalds/linux/tags is that suitable? [12:51] sorry, what i'm getting at is, does ubuntu have any specific bits in it? [12:51] Stratisphere, Rather go to git://kernel.ubuntu.com/ubuntu/ubuntu-oneiric.git [12:51] ah ha, thats what I was after :) [12:52] That also has all the build stuff to make debs [12:52] oky cool [12:52] is it somewhat stable? [12:52] The master branch is what is the current kernel in proposed which goes into updates [12:53] So fairly stable [12:53] oky cheers [12:53] thanks for your help :) [12:53] np :) [12:53] sucks, had to upgrade from lucid to fix the IR remote, but upgrading meant it broke the Tv card :P [12:54] I know the pain. Though for me it will be the other way round (at least I intend to go to oneric lts backport kernel to enable mine) [12:55] ahh, the struggles of an awesome HTPC :D [12:56] * Stratisphere hopes the wife appreciates it when he's done [12:56] Stratisphere, hehe, depends on the media you put onto it, me thinks. :) [12:57] lol ruhoh. Mostly my scifi stuff! I keep telling her to tell me what to download but she never does lol [12:57] still. She can now play her iphone things on the TV which is kind cool [12:58] You know how it works, you're supposed to know without asking. ;) But having the other thing may help, too. :) [13:13] hmm, lol the changesets are already in that repo [14:01] wow it takes a while to compile the kernel! [14:02] Depends on how powerful the build machine is... But "normal" ones... yes [14:02] out of curiosity, say you were to compile the kernel and then change one .c file... does it just compile that one file and then relink them or does it have to compile everything from scratch? (sorry, c noob here :P) [14:04] * Stratisphere should have started the compile in a screen :( === bladernr_afk is now known as bladernr_ [14:05] Using the normal packaging without tricks, everything. At least when you use dpkg-buildpackage. There is work-arounds [14:06] by only running "fakeroot debian/rules binary-xxx" and removing the build stamp file under debian... [14:06] (xxx = the flavor like generic or generic-pae) [14:07] gotta say, you guys deserve top credits. i'm sure i've just seen the top of the pile but how you guys manage such a massive and complex project is amazing [14:07] btw, whats -pae? [14:07] sconklin, have you been running any of your tests on an ecryptfs mounted FS ? [14:08] The PAE feature enabled (which is physical address extenstion, I believe). This allows to use more than 4G of RAM in i386 [14:08] ahh [14:08] knew I had heard/seen it before === jibel_ is now known as jibel [14:25] hi all [14:26] I have a patch for an ethernet driver and I have to compile the kernel and apply that patch [14:26] could someone help me in that? [14:28] Will this help? https://wiki.ubuntu.com/Kernel/BuildYourOwnKernel [14:29] yes, I have to apply this patch [14:29] http://patchwork.ozlabs.org/patch/125198/ [14:29] how can I do that following the document you linked before? [14:30] ledzgio, I would propose to get the source the git way. Then you can save the raw patch and try "git am " [14:31] could you please explain better whats mean " Then you can save the raw patch and try "git am "" [14:31] ?? [14:32] ledzgio, There is a patch link on that page that will let you download the complete patch. Then you go into the directory containing the source git tree and run the command "git am" with the file as argument [14:33] Or without git you run "patch -p1 < xxx" where xxx is again the patchfile [14:33] ok you mean the patch in the link I have sent to you correct? [14:33] yes [14:34] the command git am will apply the patch to the kernel? [14:34] at that time I only need to build the patched kernel correct? [14:34] im off. Thanks again for your help smb :) [14:34] Stratisphere, welcome [14:35] ledzgio, yes. git am does a bit more if the file is in a special format which is to make it a commit in the repo [14:36] The link you sent looked like it would be the right format. But anyways if either way of applying succeeds you build that source [14:36] I suggest you change debian.master/changelog to add something like +myversion to the version number in the first line [14:37] That way the package and the running kernel have a specific version where you can see that you are really running your kernel [14:37] ok perfect [14:38] thabk you, I'll give it a try later [14:38] if you will be here I'll contact you again if I have problem [14:39] I may or may not be [14:39] But there are always some people that can help around [14:39] ok thank you very much [14:39] your help is really appreciated [14:40] the last thing [14:40] np [14:40] I think that patch is good with kernel 3.0 or 3.1 also [14:40] correct? [14:41] tgardner: no, but there's no reason not to. The ltp lite that's currently being used doesn't do much file system testing. I'll (eventually) try running tests against ecryptfs [14:42] tgardner: I've pretty much completed the inventory of what's in ltp: https://wiki.canonical.com/SteveConklin/LtpTestList [14:42] That depends. I did not read all the details. If its just going upstream the chances are higher. But it was some wireless driver and there could be more change in that area of code [14:43] I told you that because on top of that page are specified the versions and architectures supported(?) [14:43] sconklin: Sync up with me before you do start running tests against eCryptfs. I hope to have a nice way to set up various types of eCryptfs mounts and run the typical filesystems tests against those mounts. [14:44] tyhicks: sweet, OK [14:46] ledzgio, Well that is the usual thing with backports. You may find that the compile breaks because some defines are not present (meaning support was neither) and then you have to tweak things. Also note that on that page the status has been set to superseded. Could mean there is something newer or different [14:46] sconklin, tyhicks: I'd like to get a full suite of ecryptfs tests running well before Beta 1. I'm still getting plenty of hate mail over corrupted disks. [14:47] tgardner: I can walk you through a manual install of ltp and tell you how to run the tests. In a few days, I may have that automated [14:47] (running of ltp in general, not ecryptfs) [14:47] sconklin, I'll wait. I've got other things on my plate right now. [14:48] tgardner: agreed - before Beta 1 would be great [14:52] smb, when you say "you have to tweak things", which kind of tweaks? [14:54] ledzgio, Decide how to modify the code, so it does compile in a way that still does work the same for the remaining code... [14:59] ok I have some basics of C, I think i can do that if those tweaks are not too complicated [15:04] thank you [15:04] smb [15:04] your help is really appreciated [15:04] you're welcome. no worries [15:06] bye [15:29] * ogasawara back in 20 [16:12] smb, around ? [16:13] smoser, at the moment, yes [16:13] http://paste.ubuntu.com/762860/ [16:13] that is get-console-output of a precise hvm instance. [16:14] i'll point out that there is no 'idle=halt'. [16:14] the instance was not reachable. [16:17] smoser, right. Hard to say. Usually we know things get missing. Hm, also let me do a quick version check... [16:17] i dont know if idle=halt would help or not. i tried a quick start, stop, attach, modify /boot/grub/grub.cfg, re-attach, start, [16:17] but then got *no* console output [16:19] smoser, I did not see a using mwait in that one... so probably not... [16:19] smoser, But maybe it was before the k(u)nmap patch, though that normally hit only on higher load... [16:19] smb, that is just the latest precise daily... but i have unfortunately no recollection of if this has worked at all in precise. [16:20] smoser, I think I got precise booting in my environment (I also think in the ec2 alike) but having booted too many variations I get confused [16:20] and I got amd [16:22] smb, i suppose i should open a bug, eh? [16:24] smoser, Can you reproduce it at will? ;-P But ok, we need to have an eye on it... [16:24] it reproduced 2 times out of 2 [16:24] but maybe i'm lucky [16:25] At least you might be lucky with the xen version [16:25] So this one at least is 3.4... [16:28] smb, bug 901305 [16:28] Launchpad bug 901305 in linux "precise fails boot on ec2 hvm" [High,Confirmed] https://launchpad.net/bugs/901305 [16:29] Oh HVM, right... Hm... That one I clearly have been running at least with my Oneiric dom0... at least [16:35] tgardner: is the lts-backport-natty kernel on the 10.04.3 CD/DVD ? [16:37] smoser, The only difference is that I usually go for the "normal" kernels when doing hvm installs... [16:39] nope, only maverick thus far [16:39] bjf, ^^ [16:40] tgardner: thansk [16:41] bjf, the next point release should pick up natty and oneiric [16:44] bjf, I wonder if there is a bug for that yet. [16:45] tgardner: i doubt it but skaet would probably know [16:45] I'm creating one [16:49] bjf, skaet: bug #901317 [16:49] Launchpad bug 901317 in debian-installer "10.04.4 DVD must include Natty and Oneiric kernels" [Undecided,New] https://launchpad.net/bugs/901317 [17:11] * smb ends the day [17:28] hi [17:28] smb are you there? [17:29] I'm going to build my own kernel with a patch for my ethernet card [17:29] is there a way to enable pae support? [17:29] for ram > 4GB [17:29] ? [17:34] i'm going to patch my kernel, but when I type git am patch_file it says "patch format detection failed" [17:34] this is the patch http://patchwork.ozlabs.org/patch/125198/ [17:42] guys [17:48] * ppisati -> eod === arges is now known as arges_lunch [17:49] ogasawara, I think apw's ext4 patch is worthy of an upload [17:49] tgardner: ack [18:38] sforshee: yo [18:38] sforshee: just to be certain, "[PATCH] dell-wmi: Change debug statements from pr_info to pr_debug" [18:38] vanhoof, hey [18:38] sforshee: right thread? [18:38] vanhoof, yep. I was just typing a reply to your email [18:38] sforshee: ok, i'm gonna go ahead and bounce it and send him a mail if that's cool with you [18:39] vanhoof, that would be great, thank you [18:42] sforshee: done, let's see what we turn up :) [18:42] how can I enable pae support when building my own kernel? [18:42] sforshee: also sent him something with a bit of context [18:44] vanhoof, good deal. I hope we can get a response on the thread, but I have no problem with sending the information myself either [18:45] with git [18:45] ledzgio, there is already a flavour with PAE enabled, e.g., generic-pae [18:56] * cking has nearly finished with his wrestling with powertop [19:06] * tgardner -> lunch [19:25] how can I change the kernel name when I try to build my own? === arges_lunch is now known as arges [19:32] thank you guys [19:32] really [20:04] hey, looking at https://bugs.launchpad.net/ubuntu/+source/udev/+bug/842560 . Looks like the bnx2 binary blob is having issues. Does it make sense to suggest upgrading the firmware? and where/how is that done? [20:04] Launchpad bug 842560 in linux "bnx2 firmware missing" [High,Confirmed] [20:06] arges, its not a firmware problem. its a udev race. [20:06] tgardner, ok been trying to understand it reading through the bug. so the fix is in udev? [20:06] or should be [20:07] arges, yep, the driver is working fine. its in the transition from initrd udev to rootfs udev [20:07] arges, apw is more familiar with the details [20:08] tgardner, ok, https://bugs.launchpad.net/ubuntu/+source/udev/+bug/842560/comments/43 made me think it was a kernel issue [20:08] Launchpad bug 842560 in linux "bnx2 firmware missing" [High,Confirmed] [20:08] ok i'll sync up with apw then when he's around [20:09] arges, we had a number of chats with slangasek at UDS about this issue. I'm pretty sure apw found the root of the problem. [20:10] arges, yep we are pretty sure the issue is that firmware loads are not restartable [20:10] tgardner, ok awesome. [20:10] arges, but that only matters because we don't drain waht is in progress when we restart udev [20:11] apw, ok, anything I can help you with on this bug? [20:11] arges, essentially we need to clean up the hand over, probally using the systemd handover [20:11] arges, it near the top of my todo, i need to re-review it in the morning perhaps we can chat tommorrow then after that [20:12] apw, yea sounds good [20:13] arges, you might investigate who has the reproduing machine and see if we have access to it for testing [20:13] somebody alive that takes care of the daily builds ? :) [20:14] apw, volon can reproduce the bug [20:15] dupondje: what is the issue you'd like someone to address ? [20:15] arges, vorlon ? [20:15] Well, the .config needs to be changed a bit [20:15] apw, steve langasek [20:15] arges, indeed [20:15] cause the CONFIG for intel wireless changed. [20:16] dupondje, whats wrong with the dailies ? [20:16] which makes no more intel wireless driver is build :) [20:16] and thats annoying :) [20:16] apw, I think he means mainline [20:16] dupondje: we are kind of detail oriented, not good at guessing [20:16] apw: do you need me to get you access to the reproducing box? [20:16] dupondje, whats the config name change from and to, and i'll get it fixed [20:16] -CONFIG_IWLAGN=m [20:16] +CONFIG_IWLWIFI=m [20:17] slangasek, that may well make sense, i don't think i need it immeidiatly but yes [20:17] * apw adds it to tommorrows todo [20:17] ok, let me see if I can get that sorted [20:19] http://comments.gmane.org/gmane.linux.kernel.wireless.general/81036 FYI [20:19] dupondje: thanks, we'll take a look at that [20:20] thanks :) [20:20] * dupondje was looking why his wireless didn't work in the daily builds :P [20:25] bjf, tgardner - ack. [20:26] skaet, its been so long I've forgotten what this is about. [20:26] slangasek, I've completed the journal-commit test (+readahead) with various applications running and updated bug 900923. Just some more data points to consider [20:26] Launchpad bug 900923 in pm-utils "pm-utils: readahead and journal-commit waste power" [Medium,New] https://launchpad.net/bugs/900923 [20:26] tgardner, backports on DVD [20:26] skaet, ah, right [20:27] cking: thanks :) [20:27] I'm very happy to test different scenarios if you deem it necessary [20:28] cking: I don't imagine so; I'll have a look at the data today [20:28] cool - thanks [20:29] * ogasawara wonders what's bringing zinc to a crawl. taking forever to upload. [20:30] jsalisbury is doing a git merge which is sucking cycles [20:31] nope, it's something else too [20:31] cking, hmm, let me take a look [20:31] ogasawara, prolly gitweb [20:32] reckon so [20:33] cking, on zinc, I'm not doing a git merge that I know of? Maybe I have something that ran away [20:33] jsalisbury, it was busy for some tens of seconds and then popped away.. [20:34] cking, ok [20:35] sforshee: looks like the mail just landed -- hopefully that's helpful [20:36] sforshee: seems your quite active in the platform mailing list. Any chance you can have a look @ my email from some days ago? [20:37] aka 'dell-laptop: bug in rfkill' [20:49] vanhoof, that's great. Thanks for your help. [20:52] dupondje, I found your message. Let me look at it. [20:53] sforshee: np, let me know if you need any other info :) [20:53] dupondje, have you filed a bug? [20:53] in launchpad that is [20:53] sforshee: not yet [20:54] dupondje, please do so. If you use 'ubuntu-bug linux' to file the bug it will attach a bunch of useful information [20:55] dupondje: only thing that comes to mind is this: https://bugs.launchpad.net/ubuntu/+source/network-manager/+bug/887834 [20:55] Launchpad bug 887834 in network-manager "Unable to find wireless networks after toggling wireless" [Undecided,Fix committed] [20:56] dupondje: but that's a total guess [20:58] https://bugs.launchpad.net/ubuntu/+source/linux/+bug/901410 [20:58] Launchpad bug 901410 in linux "dell-laptop: rfkill bug" [Undecided,New] [20:59] When I block the dell-bluetooth, my Wireless LAN gets hard blocked [20:59] which ofc shouldn't happen :) [21:00] dupondje, thanks. I'll set aside some time a little later to take a look, and then I'll comment on the bug. [21:19] * ogasawara lunch [21:19] sforshee: thanks alot :) [22:12] * tgardner -> EOD === bladernr_ is now known as bladernr_afk === bladernr_afk is now known as bladernr_