=== sconklin is now known as sconklin-afk
[00:45] <Kano> hi apw , is 38-rc1 ready soon?
[00:45] <Kano> mainline i mean
=== BenC_ is now known as BenC
=== sconklin-afk is now known as sconklin
=== sconklin is now known as sconklin-gone
=== czr is now known as 92AAB6NJT
=== TeTeT_ is now known as TeTeT
[11:06] <apw> smb, https://bugs.launchpad.net/ubuntu/+source/linux/+bug/681083 looks to be a xen issue in the lucid -proposed kernel (i think)
[11:06] <ubot2> Launchpad bug 681083 in linux (Ubuntu) "Ubuntu Crashes/Freeze on XenMotion (affects: 3) (heat: 26)" [Undecided,Triaged]
[11:07] <smb> apw, Just one more I need. :-P Thanks for the pointer though
[11:08] <apw> smb, i subbed the stable team, but thought you might have insite as you have more xen knowledge
[11:11] <smb> apw, The initial output looks more like generic blockage though the others hit some BUGON in spinlock. And there seems to be a patch referenced as well
[11:11] <apw> smb, yeah thats why i saw it, patch piloting over the patch
[11:17] <smb> apw, So it seems Natty is safe already as that went into 2.6.37. And as there is no xen in .32 we need our topic branch pull that in
[11:18] <apw> the patch in the bug is already upstream?  got a commit id ?
[11:18] <smb> yes it is. bah, just closed the window. A sec
[11:19] <smb> I believe that one: 6903591f314b8947d0e362bda7715e90eb9df75e though the title is slightly different
[11:20] <apw> yes that is deffo the patch
[11:21] <apw> smb, will you write that up for the bug
[11:21] <smb> yup
[11:21] <apw> smb, ta muchly
[14:04]  * tgardner --> server room
[14:26] <apw> sconklin-gone, bjf, we seem to be missing some kernel dependant packages in the latest uploads, ie. that meta has been uploaded before lum/lbm/lrm etc are built
[14:26] <apw> this has been reported on our kernel-team list for hardy
[14:26] <tgardner> apw, actually, I'm just checking that. I think its LPIA thats missing
[14:27] <apw> tgardner, as in lum for lpia only ?
=== sconklin-gone is now known as sconklin
[14:49] <smoser> smb, just trying to remember. i386 natty kernels on m1.small fail like : http://paste.ubuntu.com/555805/
[14:49] <Warlock072> hey guys 
[14:49] <Warlock072> ive got ubuntu lucid server showing high cpu usuage on this command ksoftirqd
[14:51] <smoser> smb, never mind . i found bug 699828
[14:51] <ubot2> Launchpad bug 699828 in linux (Ubuntu) "natty 2.6.37-11-virtual doesn't boot on EC2 in 32bits (affects: 1) (heat: 444)" [High,Confirmed] https://launchpad.net/bugs/699828
[14:52] <apw> Warlock072, that is interrupt related, does any of the lines in /proc/interrupts show high numbers/high change between two consecutive cats ?
=== smb` is now known as smb
[15:07] <Warlock072> ps -aux | grep ksoftirqd root        22 73.1  0.0      0     0 ?        S    15:11  84:25 [ksoftirqd/6]
[15:09] <Warlock072> Local timer interrupts
[15:09] <Warlock072> TLB shootdowns
[15:09] <Warlock072> changes alot
[15:11] <Warlock072> IO-APIC-fasteoi   megasas
[15:11] <Warlock072> this one seems to jump alot
[16:33] <hallyn> ppa:kernel-ppa/ppa only has natty images.  Where should I point a user for maverick daily builds?  Or should they just use the natty one?
[16:49] <cking> kentb, so how's it going?
[16:50] <kentb> cking, well, macallan with 10.10 server + your kernel and no iommu settings survived 100 s4 cycles
[16:50] <cking> perhaps I should have asked somewhere else. hold on
[16:51] <kentb> cking,  ick...just noticed that, too ;)
[16:52] <mjg59> Hm. You people took the Dell patch that changes the touchpad from basic PS/2 to IMPS/2?
[16:58] <hallyn> jjohansen: ppa:kernel-ppa/ppa only has natty images.  Where should I point a user for maverick daily builds?  Or should they just use the natty one?
[16:59] <jjohansen> hallyn: for daily builds just natty
[16:59] <jjohansen> hallyn: for most recent maverick -proposed kernels
[16:59] <jjohansen> just check the -proposed update option in software sources
[17:00] <hallyn> so the natty one will run ok in maverick for a test?
[17:00] <jjohansen> yeah, I actually use the natty kernel on maverick on a couple machines
[17:00] <hallyn> awesome, thanks
[17:00] <apw> mjg59, i think we have something like that in our tree at the moment
[17:01] <apw> commit f233403def40bee6ae6bd05c6f998a5f820bf106
[17:01] <apw> Author: Rezwanul Kabir <Rezwanul_Kabir@dell.com>
[17:01] <apw> Date:   Tue Nov 23 20:15:14 2010 +0000
[17:01] <apw>     Add support for Intellimouse Mode in ALPS touchpad on Dell E2 series Laptops
[17:01] <apw>     
[17:01] <apw>  
[17:01] <apw> mjg59, that one?  i applied that to natty to to get some coverage on it before deciding if it was safe for older releases
[17:02] <apw> as everyone expects to get borked boxes on natty anyhow :)
=== cmagina is now known as cmagina-lunch
[17:13] <mjg59> apw: I'm somewhat worried about it. Dell seem to be pushing it hard, but it just turns one broken configuration into another
[17:14] <mjg59> apw: ie, you get scrolling and tap to click, but now you can't turn them off...
[17:14] <apw> mjg59, ok that sounds most undesirable 
[17:15] <apw> mjg59, that doesn't sound very acceptable and pushing back on it sounds appropriate
[17:15] <mjg59> They're trying to get us to take it as well, but I'm pushing back there
[17:15] <apw> mjg59, cirtainly taking it in natty does not consitute 'acceptance' here
[17:15] <apw> the purpose of that is to determine if there is regressions for non-dell h/w
[17:15] <mjg59> Really need to have support for putting it into absolute mode
[17:16] <apw> well saying that seems completely appropriate
[17:18] <mjg59> Ok - just wanted to make sure we didn't end up in a situation where their argument is "Ubuntu's taken it, so it must be fine" :)
[17:19] <apw> mjg59, definatly i intended the application to natty to be for regression testing on non-dell h/w as we had applied it once before and all hell had broken loose
[17:19] <apw> (a previous version)
[17:20] <mjg59> Awesome
[17:20] <mjg59> I think this one's DMI-limited to Dell?
[17:20] <apw> mjg59, indeed so, but i got badly burnt on the last iteration so i am suspicious of it
[17:23] <apw> mjg59, now you mention we cannot turn of click-to-tap i'll make sure someone goes test that
[17:23] <mjg59> apw: I suspect it simply won't appear as a touchpad to X, so the control panel won't appear
[17:28] <apw> mjg59, https://bugs.launchpad.net/ubuntu/+source/linux/+bug/632884 ... the intent is loosly worded but in that bug
[17:28] <ubot2> Launchpad bug 632884 in linux (Ubuntu Maverick) (and 2 other projects) " Add support for Intellimouse Mode in ALPS touchpad on Dell Latitude series Laptops (affects: 1) (heat: 44)" [High,Triaged]
[17:29] <mjg59> apw: Yeah, I think the upstream answer to that is going to be "no"
[17:31] <apw> mjg59, heads up ... x86: Make relocatable kernel work with new binutils ... is triggering hard early boot failures for me (v2.6.38-rc1 based kernels)
[17:32] <mjg59> Yes, we've been hitting that
[17:33] <apw> mjg59, i don't see it reported on lkml yet, so i guess i'll whine
=== sforshee is now known as sforshee-lunch
[17:36] <ilmari> is the intel-drm-next kernel package built on a regular schedule?
[17:36] <ilmari> drm-intel-next, even
[17:37] <apw> ilmari, that is one of the ones which is built when the tree changes, in theory at least
[17:39] <bjf> smb, isn't there a wiki page which says what in lucid are our "franken" kernel bits?
[17:40] <apw> kees, you might want to look at this: [REGRESSION] S3 resume on SandyBridge doesn't work with NX protection (5bd5a45)
[17:40] <ilmari> apw: it seems to have been rebased today, so I guess there should be a new package shortly?
[17:40] <apw> ilmari, its checked daily at 10am UTC
[17:40] <apw> so yes i'd expect it to pop out tommorrow in that case
[17:41] <ilmari> cheers
[17:45] <apw> kees, thats your NX patches triggering all that :)
=== cmagina-lunch is now known as cmagina
[17:54] <kees> apw: upstream! :)
[17:54] <kees> apw: I can poke at it, but we'll need upstream to help with it. since it's now in the .28 tree, it shouldn't be too insane.
[17:55] <apw> well knowing linus it might not be in upstream if it doesn't get fixed and quick
[17:55] <apw> cking, i note that natty is now reporting dmi information on boot ... not sure when it arrived:
[17:55] <apw> [    0.000000] DMI 2.6 present.
[17:55] <apw> [    0.000000] DMI: 3653/HP Mini 5102, BIOS 68PGU Ver. F.05 03/11/2010
[17:55] <apw> might be useful for the fwts stuff too
[17:56] <cking> nice
[17:58] <kees> apw: yeah, true. I'll try to take a look at it, but we might need to dig deeper with it. I don't understand why they did it the way they did.
[18:04] <apw> cking, were you telling me something about acpi_init performance yesterday
[18:36] <jjohansen> rebooting
=== sforshee-lunch is now known as sforshee
=== 92AAB6NJT is now known as czr_
[19:55] <bjf> apw, git://git.kernel.org/pub/scm/linux/kernel/git/longterm/linux-2.6.35.y.git
[19:55] <apw> :wq
[19:55] <apw> grr
[20:01]  * jjohansen -> lunch
[20:02] <cking> apw shows off his vi skills
[20:02] <apw> cking, shows off his "i want apw to call my wife skills"
[20:03] <cking> dude, I've been offline for last 2 hours - honest!
[20:03] <vanhoof> apw: can you begin calling my fiance as well?
[20:03] <vanhoof> she would appreciate it :D
[20:03] <apw> sure can, where do you want her to kick you ?
[20:03] <vanhoof> lulz
[20:04]  * vanhoof makes mental note to not joke like this w/ apw as he'll likely make a call :)
[20:04] <apw> vanhoof, sure will ... :)
[20:08] <cking> vanhoof, he will do - that's for sure
[20:31] <Kano> hi, why is there still no .38 mainline kernel?
=== sconklin is now known as sconklin-afk
[20:44] <htorque> hello, everyone! are there known issues with 2.6.38-rc1and the ubuntu kernel config? like dying right after grub?
=== sconklin-afk is now known as sconklin
[21:55] <apw> htorque, thats an upstream error
[21:56] <apw> htorque, revert "x86: Make relocatable kernel work with new binutils"
[21:56] <apw> Kano, timing ... it will be built in the next hour or so
[21:57] <Kano> and it will not work?
[21:57] <apw> Kano, it is actually possible for you to make your own if you need them that urgently
[21:57] <apw> rc1 won't boot indeed, its broken
[21:57] <Kano> ah, is the script available
[21:57] <Kano> :)
[21:57] <Kano> would be no problem, my pc is fast enough
[21:58] <Kano> i just dont like to upload kernel which are unpatched, thats wasted time
[21:59] <htorque> apw, thanks :-)
[22:02] <apw> htorque, np
[22:03] <tgardner> apw, I was able to get that zillion core machine to boot by using maxcpus=0
[22:04] <apw> tgardner, heh thats a good one and no mistake, tried rebuilding with MAX_CPUS=128 ?
[22:05] <tgardner> apw, not yet, I had to leave before I could get it built.
[22:05] <apw> tgardner, doh of course ... i wonder if we should be upping that regardless
[22:06] <tgardner> apw, I'm gonna research that option to see what side effect it might have. it could be appropriate for a Lucid SRU
[22:28] <Kano> apw: where is the script used?
[22:29] <Kano> it could be optimizied too because rc kernels are in front of final kernels
[22:29] <Kano> you should use ~rc
[22:31] <sconklin> apw: I had the wrong patch implicated earlier. It's fa7e0b "drm/radeon/kms: properly compute group_size on 6xx/7xx"
[22:35] <sconklin> http://web.archiveorange.com/archive/v/hhSc5lXjWhzzBSGwGzzM
[22:53] <smb> bjf, https://wiki.ubuntu.com/Kernel/Dev/KernelDriverDeviations in case you are stll looking
[22:54] <bjf> smb, thanks
[22:54] <smb> Though by now, we also got the deviations.txt in the tree itself
[22:54] <bjf> smb, forgot about that
=== sconklin is now known as sconklin-away
[23:16]  * apw notes that v2.6.38-rc1 doesn't boot on any of his 32bit machines ... sigh