[14:53] <AStachowski> Hello guys. Could anyone point me in the right direction when it comes to build a recent Lucid kernel for my ppa? I tried it with debuild -S, -S -sd and -S -sa. But none of these seem to work. Is there a special script I need to prepare it?
[14:56] <tgardner> AStachowski, have you read https://wiki.ubuntu.com/KernelTeam/KernelMaintenance ?
[14:58] <AStachowski> I think I've forgotten that part if it is necessary: dpkg-buildpackage -S -rfakeroot -I.git -I.gitignore -i'\.git.*' -sa
[14:59] <AStachowski> Though I thought it was only necessary for git pulled releases.
[15:02] <AStachowski> I think I already read https://wiki.ubuntu.com/KernelTeam/KernelMaintenance and thought it would only apply to GIT. Guess I'll try it this evening. ;)
[15:03] <JFo> AStachowski, how did you get your source?
[15:03] <JFo> did you not use git to pull it?
[15:04] <AStachowski> nope
[15:04] <AStachowski> I got it by apt-get source linux-image-$(uname -r)
[15:05] <AStachowski> if git pull is the way to go I don't mind, too.
[15:05] <JFo> ah, I see
[15:06] <JFo> AStachowski, whatever works for you
[15:06] <JFo> but I think I see a need for another set of instructions for folks wanting to pull from there
[15:06]  * JFo adds a note to his 'wiki improvements' doc
[15:07] <AStachowski> ;D
[15:07]  * peterz fixed /sbin/installkernel and uses make install form the kernel tree ;-)
[15:08] <peterz> all this package nonsense is just that ;-)
[15:08] <JFo> :-)
[15:08] <AStachowski> I thought I'd finally give something back to the community by building a subflavor of linux-image-generic for SCST -> SRP Target for the Infiniband folks.
[15:09] <JFo> AStachowski, a noble gesture ;)
[15:09] <persia> Are drivers for that not generally available, or just not built-ins?
[15:09] <AStachowski> Infiniband sadly is a bit more complicated than that.
[15:10] <amitk> peterz: don't we all wish that all users could compile their own kernels :)
[15:10] <AStachowski> there is a standard target daemon in alls linux distros but it genuinely sucks when it comes to larger setups.
[15:10] <JFo> amitk, I don't :-D
[15:11]  * JFo looks at the bug volume as it is
[15:11] <AStachowski> too many cooks spoil...you know the deal ;)
[15:11] <JFo> yep
[15:11] <JFo> great minds AStachowski 
[15:35] <MTecknology> I thoguht apparmor was in the mainline kernel now
[15:36] <abogani> peterz: What is the fix for /sbin/installkernel?
[15:37] <peterz> abogani: add to the tail: mkinitramfs -o /tmp/initrd.img ${ver}; updatever initrd.img /tmp/initrd.img; update-grub
[15:38] <peterz> sometimes I need to hack update-grub too, depending on the crazy ass ideas it gets
[15:39] <achiang> peterz: nice. i figured out that i could issue a 'make deb-pkg' from kernel directory, install the deb, and then issue a mkinitramfs, but your way seems much nicer
[15:42] <peterz> achiang: make sure to do 'make INSTALL_MOD_STRIP=1 modules_install' if you use modules 
[15:42] <peterz> because for some reason mkinitramfs doesn't strip the modules
[15:42] <peterz> and you end up with a 100M+ initrd
[15:43] <peterz> of course the sane thing to do is not use modules :-)
[15:43] <MTecknology> peterz: That REALLY hard to do sometimes though
[15:44] <peterz> MTecknology: nah, just don't do silly things like use lvm/crypto/raid root paritions
[15:44]  * MTecknology grumbles at one particular module
[15:45] <MTecknology> peterz: I need moduel support for vbox but having module support forces another config to be module instead of built-in
[15:45] <achiang> peterz: eh?
[15:45] <achiang> achiang@aspen:/boot$ ls -l initrd*34-rc5* | awk '{print $5}'
[15:45] <achiang> 3381926
[15:46] <MTecknology> 34 hates my computer it seems
[15:46] <achiang> peterz: hm, maybe part of the make deb-pkg magic strips the modules for me, since i don't issue make modules_install directly
[15:46] <peterz> MTecknology: ah, out of tree stuff, I don't acknowledge the existance of such
[15:47] <peterz> achiang: yeah, I figure it would, also you of course need to start out with CONFIG_DEBUG_INFO=y
[15:48] <achiang> peterz: heh. achiang@aspen:~/kernels/linux-2.6$ grep DEBUG_INFO .config
[15:48] <achiang> # CONFIG_DEBUG_INFO is not set
[15:49] <MTecknology> peterz: that would be awesome if they could have their modules in there
[15:49] <peterz> MTecknology: not a chance, the code is alike waaaay ugleh
[15:50] <MTecknology> probably
[15:55] <cnd> peterz: sorry about the kernel-team list bounces
[15:56] <cnd> I think we've got that taken care of, so it shouldn't happen in the future
[15:56] <peterz> cnd: awesome
[15:56] <cnd> peterz: btw, any more thoughts on the ILB load avg implementation?
[15:57] <peterz> cnd: right, so the reason I wante dto share infrastructure with the ILB is because people were looking at making the ILB scale better and its all about NOHZ muck, so it seemed like a nice fit
[15:57] <cnd> ahh, ok
[15:57] <peterz> cnd: just need to get a spare cycle to sort through all the details, hopefully somewhere later this week
[15:58] <cnd> peterz: ok, thanks
[21:19] <osmosis> running fresh lucid daily build, I get random lockups. Can't drop to a terminal or anything.
[22:01] <osmosis> Im getting an hard freeze on two different laptops. one with 64, other x86. Both running fresh installs of lucid daily builds.
[22:05] <bjf> osmosis, please file a _new_ bug against the issue
[22:06] <osmosis> bjf, im not sure how to gather any info about the cause of the lockups.
[22:07] <bjf> osmosis, will it stay up long enought to run ubuntu-bug on it?
[22:07] <osmosis> bjf, cursor stops moving and spinning.  CTRL-ALT-F1 for a terminal doesnt work either.
[22:07] <bjf> osmosis, had you been running a previous daily build?
[22:09] <osmosis> bjf, i had beta2 installed briefly, but did a full format clean install of lucid daily build yesterday.
[22:09] <osmosis> bjf, i thought maybe the crash was hardware specific...but I installed on two separate laptops and they both are getting the same issue. And they are very different hardware.
[22:11] <bjf> osmosis, what graphics do the two systems have?
[22:12] <bjf> osmosis, was beta2 running ok?
[22:15] <osmosis> bjf, both are ati ...thought different models.
[22:17] <osmosis> bjf, ATI Radeon Express 200  and   ATI Mobility Radeon HD 4670
[22:17] <bjf> osmosis, have you tried booting without "quiet" or "splash" on the kernel boot line?
[22:18] <osmosis> bjf, machines come up and are usable. The crash is a randomly triggered lockup...occurring on average every hour or two.
[22:19] <bjf> osmosis, then they are up long enough to run ubuntu-bug on and capture the hardware information
[22:19] <bjf> osmosis, ubuntu-bug will capture everything you need
[22:20] <osmosis> bjf, ok
[22:22] <bjf> osmosis, do two separate bugs, one for each system and then let me know the bug numbers
[22:22] <bjf> osmosis, thanks
[22:30] <cnd> osmosis: you might want to try booting with "pci=nomsi" on the kernel command line
[22:30] <cnd> that has fixed some ATI instabilities, and we can add quirks in the kernel for your motherboard if it helps
[22:31] <anoteng> Where can I find a detailed list of changes between ubuntu kernels? I'm trying to locate the source of a bug..
[22:50] <osmosis> bjf, here is one, https://bugs.launchpad.net/ubuntu/+source/xorg/+bug/568103
[22:50] <ubot3> Malone bug 568103 in xorg "Xorg freeze lucid daily build ATI Mobility Radeon HD 4670" [Undecided,New] 
[22:53] <bjf> osmosis, did you see the comment from cnd about booting with "pci=nomsi" ?
[22:54] <osmosis> bjf, yah
[22:54] <bjf> osmosis, that would be good to try
[22:54] <osmosis> bjf, cnd: im gonna see if i can get the crash to occur a couple more times first.
[23:00] <osmosis> bjf, here is the second, https://bugs.launchpad.net/ubuntu/+source/xorg/+bug/568111
[23:00] <ubot3> Malone bug 568111 in xorg "Xorg freeze dell vostro 1000 lucid daily build ATI Radeon Xpress 200" [Undecided,New] 
[23:00] <bjf> osmosis, thanks for those
[23:01] <osmosis> bjf, was previously on karmic with no such issue.
[23:01] <osmosis> on both laptops
[23:01] <bjf> osmosis, lots of changes to graphics drivers this round