=== TheMuso` is now known as TheMuso
[03:45] <dhendrix> does anyone have a pair of kernel-img.conf and kernel-pkg.conf files set up for kernel packages with completely non-interactive setup (using make-kpkg)?
[05:52] <sunnydrake> hi can someone explain me - will ubuntu distro work wihout problems with "original" kernel(i wish to check unified kernel patches on ubuntu liveusb system) ?
[06:00] <lifeless> yes, and there are packages built with the vanilla kernel specifically for testing
[06:00] <lifeless> however, you may find the live system isn't supported by the vanilla kernel, or some hardware 
[08:53] <kdub> hey all, i'm good with kernels, and would like to contribute, where can i start?
[08:58] <smb> kdub, You might want get familiar with the workflow. See Leann's mail to Jeremy at https://lists.ubuntu.com/archives/kernel-team/2009-June/006170.html
[08:58] <smb> If you think you want a go, there is a community section of bugs if you follow the link on the topic
[08:59] <zeroprog> hey guys  im trying to edit the module that colors the ubuntu layout....anybody know what folder those files are located in?
[09:00] <kdub> smb: thanks, i know its annoying when people ask "how can i start" questions, but i really think i can help out a lot if someone helps me get my foot in the door
[09:01] <smb> kdub, No problem. We really can use a few people out in the community helping out with the work
[09:02] <kdub> i'll read the links in that email, and ping back with any questions i have
[09:02] <smb> zeroprog, Are you sure, the kernel is the right place to look for that?
[09:02] <kdub> but i have decent experience with embedded systems, kernels, drivers, all that jazz
[09:03] <smb> kdub, Sure. Thanks for joining in
[09:08] <zeroprog> smb: not at all....i was hoping to write a simple module that says something stupid everytime you delete something but i cant find where that folder is either
[09:08] <zeroprog> the names linux uses are short and sweet but not so much descriptive 
[09:08] <zeroprog> the only things i can figure out are crypto, acpi and a few others 
[09:11] <smb> zeroprog, my problem might just be to get the connection between that and something that colors the ubuntu layout. Heck, I am not even sure I understand what you mean with ubuntu layout. ;-)
[09:13] <zeroprog> ergh...curse my explanatory skills
[09:14] <zeroprog> ive been trying to look for a use for modules that im able to do....drivers are too complicated for some reason plus i have no real devices i wish to use...maybe ill just learn the nokia n800 deal
[09:21] <smb> Hm, not sure I can help there. The modules I look at are either drivers or part of a bigger stack. If I saw something simple I probably forgot it. :-/
=== cjwatson_ is now known as cjwatson
[10:00] <apw> hrm i thought there was a config option for tickless
[10:01] <smb> nohz?
[10:01] <apw> hrm no longer an option for it
[10:02] <kdub> smb: im a bit overwhelmed with all these configuration caveats...
[10:02] <smb> apw, NO_HZ
[10:02] <apw> debian/config/config.common.ports:CONFIG_NO_HZ=y
[10:02] <apw> debian/config/config.common.ubuntu:CONFIG_NO_HZ=y
[10:02] <smb> kdub, which one exactly?
[10:02] <apw> well i assume its on then :)
[10:03] <smb> apw, I guess so
[10:03] <kdub> if i want to build a standard i386 kernel, the .config lives in debian/config/i386?
[10:03] <cking> apw, I suggest asking for powertop output first, e.g. powertop --dump to see if there is anything insane causing CPU heat
[10:04] <smb> kdub, It is split into a generic part and a part for the generic kernel. Karmic will be even more split
[10:05] <smb> kdub, But usually you will not have to care that much if you compile with "fakeroot debian/rules binary-arch" or "debuild -b"
[10:05] <kdub> i'm not so comfortable with fakeroot/deb building stuff just yet...
[10:07] <smb> kdub, For releases up to Jaunty, you get the used config by cat'ing together debian/configs/i386/config{,-generic}
[10:08] <kdub> and jaunty? cat all 3 together
[10:08] <smb> kdub, Jaunty is the same as above. The new stuff will come with Karmic
[10:09] <apw> there are common configs at the top level and at each arch plus config deltas per flavour in karmic
[10:09] <kdub> smb, sorry, i meant karmic
[10:10] <lesshaste> I have been told to enable boot logging and post the output ( https://bugs.launchpad.net/ubuntu/+source/gnome-power-manager/+bug/389930 ) but this feature has not worked for ages and does nothing in jaunty
[10:10] <smb> kdub, As apw said. :)
[10:11] <ubot3> Malone bug 389930 in grub "grub menu skipped after shutdown" [Low,New] 
[10:11] <lesshaste> can anyone tell me, do they secretly mean /var/log/dmesg?
[10:12] <kdub> oh there we go, its 3 configs combined together
[10:14] <apw> cking, i've asked them to identify the exact kernels which are good and bad, then for those that do that i will give them a tragetted set of mainline kernels to test to give us a bisection
[10:14] <apw> i guess thats a start
[10:15] <cking> meanwhile, let me install intrepid + jaunty on one of my spare lappies and see if there is anything obviously different
[10:15] <apw> cking, thanks
[10:15] <smb> apw, I will do a karmic installation on a laptop here, that might give additional data points
[10:15] <kdub> and once this gets done compiling, do you guys usually run in a vm, or as the running kernel?
[10:15] <kdub> eg, what is more useful to do?
[10:15] <apw> yeah kirkland is complaining about a thinkpad
[10:16] <apw> kdub, bit of both depending on the change
[10:16] <cking> apw, 64 or 32 bit machines?  64 will bang more gate transistors :-)
[10:16] <smb> kdub, depends, a vm is good to see it boots at all. real hardware might be the only way to find other things
[10:16] <kdub> apw: alright
[10:16] <apw> cking, hehe always been 64 bit here
[10:16] <smb> kdub, And for some stuff, we have to ask the reportes as we don't have the hw at all
[10:16] <cking> me too. 
[10:17] <apw> 135 wakeups per-second sounds like a lot to me
[10:17] <cking> apw, that sounds typical to me.
[10:17] <smb> apw, Got 240 on hardy here
[10:18] <smb> (given without tweaking)
[10:18] <apw> heh then i guess i shoul be happy :/
[10:18] <kdub> well, i guess i'm up off the ground, so to speak. next up is, what, looking for bugs?
[10:18] <cking> I will also measure power consumption using my little meter gizmo
[10:18] <apw> i guess when your keyboard is showing up ... that you don't have that many ints
[10:19] <apw> kdub, did you see the triaging stuff that leaan emailed out?
[10:19] <cking> turn down the beacon interval on your access point :-)
[10:19] <apw> i _think_ smb may have pointed you at it
[10:19] <smb> I passed the link to her email 
[10:20] <smb> And if that feels ok. The link to the bugday is on our channel topic :)
[10:20] <kdub> apw: i did, it didnt make a whole lot of sense though
[10:20] <cking> 66 wakeups/sec on my karmic dell6400
[10:20] <apw> yeah.  most people who start getting involved in the kernel stuff, start with helping out on the bug triage
[10:21] <apw> that gets you used to the bug handling, which is a little odd
[10:21] <apw> then generally one ends up touching a few which need actual work and can start to make a difference
[10:21] <apw> easing in gentle like
[10:24] <lesshaste> is there any way of disabling a fingerprint reader without doing it through the bios ? I get A USB device is active 100.0% of the time:
[10:24] <lesshaste> in powertop
[10:25] <apw> some devices have a power control separatly
[10:26] <apw> in /sys
[10:26]  * kdub still doesnt really know what to do next
[10:26] <kdub> go through launchpad?
[10:26] <smb> maybe unbinding (for testing purposes) might work as well
[10:27] <apw> kdub, if you want to get a feel for our bug process, then reading the triage doc, then helping out with the community section of the bug day page is a good start
[10:27] <smb> kdub, In the end all goes through launchpad. I don't know, but I believe it helps to go ttps://wiki.ubuntu.com/KernelTeam/BugDay/20090623 there
[10:27] <apw> many of those will just be simple closes, but the ones which are not, will need some investigation
[10:27] <smb> kdub, and then to Bugs, there is a link to a table with bugs
[10:27] <apw> commonly involving finding likely fixes and building test kernels etc
[10:27] <smb> One section is community bugs. 
[10:28] <apw> when you find those too easy, we have plenty of other bugs to move up to
[10:28] <smb> The bug numbers are selectable and take you to launchpad
[10:32] <kdub> alright, so i pick one bug, and see what i can do to help with it, i guess
[10:32] <smb> kdub, Yep, same we do. ;-)
[10:32] <apw> kdub, the process docs give you more idea what states to put things in when you do things with them
[10:32] <lesshaste> kdub, you could start with mine :)
[10:33] <kdub> smb: but unlike you guys, i don't have a lot of experience with this :P
[10:33] <apw> and the bug day page has some help on the sorts of things its reasonable to say with old bugs and which states to use 
[10:33] <apw> so the use knows why we are closing them out etc
[10:33] <apw> then the interesting one you get to make up ways to test, look for patches, etc etc
[10:33] <apw> when you find fixes and stuff come back and we can help you get them in
[10:33] <apw> as in tell you the process there, not to overwhelm you now
[10:34] <smb> kdub, Sure, but we all started at some point. It just takes a bit time to get into it
[10:34] <apw> kdub, heh don't worry if things are too hard, leave them, feel free to pick things you think you can solve, help with, make a difference with
[10:35] <apw> kdub, realise that even if you only did some triage work and helped weed out the old crappy bugs, the ones already fixed, and make sure the right info is obtained, that is all valuable work to do
[10:35] <apw> as it reduces the work the rest of us have to do, so we get to more of the bugs.  if you actually fix bugs as well thats the cherry on top
[10:35] <apw> but either way its appreciated, as it helps everyone get more done
[10:36]  * apw watches the merge window slam shut ...
[10:36] <cking> klunck!
[10:38] <kdub> apw: my general goal is to be less trouble than i'm worth :)
[10:39] <apw> kdub, then you have the right attitude!  really any work you do correctly is a win for us, we are a small team, probabally about the equivalent of 8 people
[10:39] <apw> kdub, welcome aboard.  please do come back here and ask questions on anything you are unsure about, anything that seems mad (and there will be a lot of mad stuff) etc... ask, we are pretty approachable
[10:40] <smb> kdub, mostly. (joking). Yes, welcome
[10:40] <kdub> thanks, everyone
[10:44] <kdub> and mailing list?
[10:45] <apw> there is mailing list assosiated with the ubuntu-kernel-team in launchpad (i believe
[10:46] <kdub> do people use it?
[10:46]  * apw checks its the right one
[10:46] <apw> Ubuntu Kernel Team <kernel-team@lists.ubuntu.com>
[10:46] <apw> is the one we use, we review all patches and there is a fiar bit of discussion on that one
[10:47] <apw> https://lists.ubuntu.com/mailman/listinfo/kernel-team
[10:48] <kdub> alright, thanks. i subscribed
[12:10] <apw> smb, when we rebase across version i assume the -ABI number reverts to -1 ? yes?
[12:20] <Kano> hi apw, do you want to update ndiswrapper toady?
[12:23] <Kano> i got a patch from the s626 author, i am just not fully convided that the patch is needed in that form. i would have guessed adding the id would be all needed
[12:23] <Kano> he added a bit more
[12:27] <[plane]> anybody here??
[13:29] <mok0> I have a regression on my netbook, the latest kernel upgrade to 2.6.28-13-lpia made the wireless go away
[13:50] <AceLan> mok0: whick wireless chip do you use?
[13:50] <mok0> AceLan: bcm4315
[13:52] <mok0> AceLan: I used to have it visible in jockey, but no more
[13:52] <mok0> AceLan: "Broadcom STA wireless driver"
[13:53] <AceLan> mok0: do you remember the version of the original kernel?
[13:53] <mok0> AceLan, 2.6.28-11-lpia
[13:54] <mok0> AceLan: It was a recent update from jaunty-updates
[13:56] <AceLan> mok0: do you know the module name of the wireless driver?
[13:56] <mok0> AceLan: yes, it's "wl.ko"
[13:56] <mok0> wl I guess
[13:57] <Kano> thats in restricted
[13:57] <Kano> well i made a dkms script
[13:57] <mok0> The old driver is in  /lib/modules/2.6.28-11-lpia/volatile/wl.ko
[13:57] <AceLan> mok0: so, the w1 module doesn't exist now?
[13:57] <mok0> Kano: yes it's restricted
[13:58] <mok0> AceLan: no
[13:58] <mok0> AceLan: wasn't included in the upgrade for some reason
[13:58] <mok0> AceLan: ... or something happened during the installation... I guess it is built on installation
[13:59] <smb> mok0, Yes it is. The binary parts come with linux-restricted-modules and it is linked on boot
[13:59] <mok0> OK, makes sense
[14:00] <smb> mok0, Could it be lrm was not upgraded at the same time?
[14:00] <mok0> smb, I can check
[14:00] <mok0> smb, aargh out of battery & no charger here
[14:01] <mok0> I will check when I get home
[14:01] <smb> mok0, bad. ok
[14:01] <mok0> Justed wanted to hear if you guys knew about a problem
[14:02] <smb> Not right of my head. I tested Jaunty wl driver from lrm on amd64 and that worked
[14:03] <mok0> smb: ok, I have not heard from other mini 10 owners about anything either
[14:04] <smb> mok0, So we'll have to find out what went wrong with yours.
[14:04] <mok0> smb: ok, I'll let you know... or perhaps turn up here in panic
[14:04] <mok0> :)
[14:04] <smb> mok0, either way :)
[15:39] <cjwatson> rtg: grub2 --no-floppy> on my list to fix (I think there's an upstream patch for it)
[15:40] <rtg> cjwatson: cool, it bit me on my laptop.
[15:40] <rtg> Is there an LP bug already? I got distracted and didn't file one
[15:54] <cjwatson> rtg: yeah
[15:54] <cjwatson> rtg: bug 391044
[15:54] <ubot3> Malone bug 391044 in grub2 "grub2 update adds --no-floppy to "search " lines" [Undecided,Confirmed] https://launchpad.net/bugs/391044
[16:03]  * hyperair wonders if anybody can boot the 2.6.31 kernel from the kernel-ppa?
[16:26] <cjwatson> rtg: I've uploaded the necessary bits for the installer to use -generic-pae in the same way it used to use -server; will need to be refined I expect, this is just basic support
[16:27] <rtg> cjwatson: noted.
[16:30] <cjwatson> ... and corresponding cdimage and seed changes
[18:53] <apw> hyperair, never tried
[18:54] <hyperair> apw: didn't you compile that, though?
[18:55] <apw> hyperair, the mainline kernels, they come out of my automation yes.  doesn't mean i boot them
[18:55] <hyperair> aha, heheh
[18:55] <hyperair> i se
[18:55] <mok0> smb: I solved my wireless problem
[18:55] <hyperair> +e
[18:55] <hyperair> hello mok0 =)
[18:55] <mok0> hyperair!!!
[18:55] <hyperair> mok0!!!
[18:56] <mok0> hehe
[18:56] <hyperair> =p
[18:56] <rtg> apw: I've booted the mainline 2.6.31-rc1
[18:56] <hyperair> rtg: oh it worked for you? =\
[18:57] <hyperair> it hangs at modprobe for me
[18:57] <rtg> hyperair: amd64, dual quad-core
[18:57] <hyperair> i'm on a dual core
[18:57] <hyperair> amd64
[18:57] <mok0> smb, I had to run lrm-manager and let it form a proper initrd.img file
[18:58] <hyperair> what does lrm-manager have anything to do with initrd? O_o
[18:58] <mok0> hyperair: it builds a new one
[18:58] <mok0> hyperair: after the modules have been installed
[18:58] <hyperair> ah
[18:58] <hyperair> i see
[18:58] <mok0> But it does so for the running kernel
[18:59] <mok0> which means that if you install the restricted modules under another kernel, it will not be the right one
[18:59] <hyperair> that.. doesn't sound nice
[19:00] <mok0> ... therefore the drivers wont take effect when the new kernel starts up, and you become very confused
[19:00] <hyperair> heheh
[19:00] <hyperair> i see
[19:00] <hyperair> i thought that the restricted drivers were loaded after root was mounted though
[19:00] <mok0> I don't understand why that is allowed to happen
[19:01]  * hyperair doesn't either
[19:01] <mok0> hyperair: well, I _think_ this is what has happened in my case...
[19:01] <hyperair> what i do know, however, is that my dvd drive suddenly detected a blank dvd around 30 minutes after i inserted it.
[19:01] <hyperair> heheh
[19:02] <hyperair> which driver were you using?
[19:02] <mok0> hm, at least it came around
[19:02] <hyperair> heh if only it would come around each time
[19:02] <hyperair> it's a blank dvd and i can't burn into it
[19:02] <mok0> hyperair: the wireless driver
[19:02] <hyperair> mok0: i meant which wireless driver 
[19:02] <mok0> :(
[19:02] <hyperair> my dvd drive's whacked up for some reason. it can only detect and burn to cds
[19:02] <hyperair> =(
[19:02] <mok0> ah, its for the broadcom wireless chip
[19:02] <hyperair> ah broadcom
[19:03] <hyperair> nasty one that is X_X
[19:03] <hyperair> i avoided that one like plague when i was shopping for my notebook
[19:03] <mok0> it's the broadcom proprietary dirver, called "wl"
[19:03] <hyperair> wl?
[19:03] <hyperair> not b43?
[19:03] <mok0> no
[19:03] <hyperair> wait, i think i saw that in some guy's notebook during malaysia's osconf
[19:04] <hyperair> except his issue was much simpler--he turned off the rf kill switch and couldn't figure out why it wouldn't work
[19:04] <mok0> it really should use dkms
[19:04] <hyperair> indeed it should
[19:04] <mok0> I might file a bug against it
[19:05] <hyperair> dkms is a seriously cool package =p
[19:05] <mok0> indeed
[19:05] <hyperair> go file it then! =p
[19:05] <mok0> I should learn more about it though
[19:05] <mok0> ... tomorrow
[19:05] <mok0> :-)
[19:06] <mok0> hyperair: did you make uuc btw?
[19:06] <hyperair> mok0: yeah i did =) thanks for your endorsement
[19:07] <mok0> nice, congrats
[19:07] <hyperair> thanks =)
[19:07] <mok0> hyperair: next step is motu
[19:07] <hyperair> indeed
[19:07] <mok0> hyperair: you should start getting your application together
[19:08] <hyperair> =O now?
[19:08] <hyperair> i thought i'd bide my time until i'd gotten more stuff done
[19:08] <mok0> hyperair: yeah, start adding stuff to your journal with stuff you do
[19:08] <hyperair> hmm yeah
[19:09] <mok0> hyperair: so you can apply for the karmic cycle
[19:09] <mok0> karmic+1 
[19:09] <hyperair> yeah
[19:09] <mok0> hyperair: after the summer
[19:09] <hyperair> well i'll see how it goes =)
=== ivoks_ is now known as ivoks
[21:05] <CarlFK1> why does "sudo reboot" sometimes reboot the hardware (back to POST) and sometimes use kexec to reload the current kernel? 
[21:11] <NCommander> CarlFK1, if kexec-tools are installed, it will use it if possible
[21:20] <CarlFK1> NCommander: um... -tools installed, but it is not doing it right now
[21:21] <CarlFK1> juser@dhcp216:~$ apt-cache policy kexec-tools ...   Installed: 20090000-2.0.0ubuntu3.1
=== ivoks_ is now known as ivoks
[21:24] <nick_schembri> ls
[21:30] <CarlFK1> pretty sure this is a firewire module thing: 1041.217340] dvgrab[6370]: segfault at 7f2911245c88 ip 00007f29184700a1 sp 00007f29111a2fa8 error 6 in libc-2.9.so[7f29183ec000+168000]
[21:30] <CarlFK1> trying under gdb now
[21:32] <nick_schembri> cjwatson: Do you have a link to the new live cd remaster process for 9.10 +
[21:35] <ccheney> ogasawara: is bug 392288 a proper kernel backtrace it looks a bit different from what i remember them looking like
[21:35] <ubot3> Malone bug 392288 in linux "dd extremely slow writing to usb key without oflag=dsync" [Undecided,New] https://launchpad.net/bugs/392288
[21:36] <CarlFK1> http://dpaste.com/59783/ #7  0x00007f6027e97fcd in clone ()     at ../sysdeps/unix/sysv/linux/x86_64/clone.S:112 No locals. 
[21:37] <cjwatson> nick_schembri: there's a new process?
[21:40] <nick_schembri> At south east linux fest Creg talked about the aufs being removed from 9.10
[21:42] <nick_schembri> oops Pete Graner
[21:46] <ogasawara> ccheney: probably because it's not an oops/panic message that you're probably used to seeing.  was that an ext4 fs?
[21:47] <ogasawara> ccheney: 2.6.31-rc1 was recently build, would be good if you could test and confirm with that kernel - http://kernel.ubuntu.com/~kernel-ppa/mainline/v2.6.31-rc1/
[21:51] <ccheney> ogasawara: it was writing directly to the device
[21:51] <ccheney> ogasawara: ok will take a look after my system update is done
[22:02] <Sarvatt> is there a reason memory stick support isnt enabled in the kernel? just curious as to what problems it might have that might have led to it being disabled
[22:05] <dtchen> hmph, quite a few things FTBFS against the 2.6.31-rc1 headers, which isn't surprising
[22:05] <dtchen> nvidia, bcmwl, vbox, ...
[22:08] <hyperair> vboxdrv worked
[22:08] <hyperair> not vboxnetflt though
[22:08] <Sarvatt> vbox 3.0 beta 2 compiles fine on 2.6.31-rc1
[22:10] <dtchen> not in a position to test 3.0b2 yet
[22:10] <Sarvatt> http://www.virtualbox.org/ticket/4264
[22:10] <Sarvatt> fix there
[22:11] <Sarvatt> (for 2.2.4 too)
[22:13] <hyperair> cool
[22:13] <hyperair> now if only i could boot that kernel =p
[22:14] <ccheney> ogasawara: i tried the 31 kernel and it wouldn't even boot into gdm for me
[22:14] <ccheney> ogasawara: it just went black after loading the kernel, i think at the point it tried to start gdm
=== ivoks_ is now known as ivoks
[22:14] <ogasawara> ccheney: well that's no good
[22:14] <ccheney> ogasawara: i tried both with and without nomodeset option
[22:15] <ccheney> the current karmic kernel gets my resolution wrong so i use nomodeset on it
[22:15] <ogasawara> ccheney: yah, I've seen similar bugs for the resolution issue
[22:44] <cjwatson> nick_schembri: that doesn't affect customisation
[22:44] <cjwatson> nick_schembri: the bulk of the live CD is still in a squashfs
[22:45] <hyperair> the .31 kernel hangs running modprobe for me =\
[22:46] <cjwatson> nick_schembri: the only difference relevant to customisation that I know of is that the live CD initramfs is now LZMA-compressed - that's just a matter of decompressing initrd.gz, recompressing it with 'lzma -9c', and writing it to initrd.lz
[22:46] <hyperair> it appears to load iwlagn, hda-intel, and some other module, then hangs when it should bel oading i915.
[22:46] <cjwatson> try loading intel-agp first?
[22:46] <hyperair> hmm i should try that.
[22:47] <hyperair> cjwatson: but shouldn't modprobe automatically load intel-agp first?
[22:47] <cjwatson> there was a driver bug that meant it didn't
[22:47] <cjwatson> I believe it's been fixed upstream
[22:47] <hyperair> i see
[22:47] <cjwatson> cf. https://bugs.freedesktop.org/show_bug.cgi?id=22358
[22:47] <ubot3> Freedesktop bug 22358 in Driver/intel "Performance regression:  Sluggish text scrolling, 100%CPU after upgrade from 2.7.1 (UXA bug)" [Major,Resolved: fixed] 
[22:47] <hyperair> so how would i go about getting intel-agp loaded before i915?
[22:47] <cjwatson> stick it in /etc/initramfs-tools/modules?
[22:48] <hyperair> eh..
[22:48] <hyperair> it panicked when i had that
[22:48] <cjwatson> actually, a current karmic initramfs should load intel-agp for you automatically
[22:48] <cjwatson> initramfs-tools 0.92bubuntu32
[22:49] <hyperair> that's my initramfs version, yes
[22:49] <cjwatson> dunno, then
=== btQuark is now known as btQuark2
=== btQuark2 is now known as btQuark