[01:27] <Philip5> fabbione: are you awake?
[01:58] <desrt> BenC; i wonder if you'd take a totally crack patch to the kernel that no "real user" would ever use but would make developer's lives nicer?
[02:15] <BenC> desrt: depends on how invasive it is
[02:16] <desrt> BenC; very minimally
[02:16] <BenC> I can review it if you email it to me
[02:16] <desrt> BenC; it adds 5 fields to struct task_struct
[02:16] <BenC> oooh
[02:16] <desrt> BenC; and uses them to keep track of how many times poll/select/epoll/itimer/schedule_timeout timeouts fire
[02:16] <desrt> BenC; then makes the info available in /proc
[02:16] <neuralis> desrt: for future discussions of this type, just stick the patch up somewhere, http://pastie.caboo.se or something
[02:16] <BenC> I've heard task_struct is very peculiar about it's size right now
[02:16] <desrt> oh.  really?
[02:17] <BenC> and that sort of tracking sounds like it has some overhead unless it's disabled by default
[02:17] <desrt> BenC; it's a ++ in the poll syscall
[02:17] <BenC> desrt: if task_struct gets > PAGE_SIZE it wastes mem, and costs ticks
[02:17] <desrt> ya.  task_struct is pretty big, too
[02:18] <desrt> i'll find out.  hold a sec.
[02:18] <BenC> email me your findings
[02:18] <desrt> ok.
[02:18] <BenC> redskins vs. dallas is about to start, and I can't miss that :)
[02:18] <desrt> :)
[02:20] <desrt> 1360 bytes with my patch on x86
[03:08] <neuralis> BenC: ping me when you get back from the game
[05:10] <infinity> BenC_: Pong?
[08:59] <pdr> anyone know why saa7146 driver was removed form the edgy kernel?  is there a problem with it?
[12:01] <`ph8> no daily image yet boys and goyles?
[01:38] <jbailey> BenC_: ping
[01:51] <`ph8> idle 8 hours jbailey :)
[01:51] <jbailey> ph8: That's a full night's sleep, he should wake up soon ;)
[01:52] <`ph8> He's overslept in fact. He should wire his alarm clock up to the internet for situations like this. Too much sleep can be bad for you
[01:53] <`ph8> :o)
[02:27] <BenC_> jbailey: pong
[02:28] <jbailey> BenC_: Heya.  I'm back in Canada now.  ISTR that there was a git tool to make doing a binary search for a particular change in the git tree easier.
[02:28] <mjg59> jbailey: git bisect
[02:29] <jbailey> BenC: 2.6.18 is working fine with sata_foo (I don't remember off hand which one), so I'd like to find the change that fixed and ask for it in edgy.
[02:29] <jbailey> mjg59: Thanks.
[02:29] <BenC> yep, that's the one
[02:29] <BenC> the man page will make it easy to get started
[02:29] <jbailey> Fabulous. =)
[02:32] <mjg59> BenC: Any ETA on the edgy upload?
[02:32] <BenC> this morning, I hope
[02:32] <mjg59> Cool
[02:32] <mjg59> Are you merging the stuff I sent, or is that for the next one?
[02:33] <zul> BenC: do you want me to keep an eye on the dapper-proposed-updates tree?
[02:33] <neuralis> BenC: what's our policy on updates to the dapper kernel? security-only? apparently all the newer hp servers ship with iLO2, which kills the dapper installer. description and patch here: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=384202
[02:39] <`ph8> BenC: what timezone are you in? ;)
[02:42] <BenC> zul: if you want to, that'd be great
[02:42] <BenC> ph8: EST -04
[02:43] <BenC> neuralis: right now, security only for user visible updates, but patches are going into -proposed, and being migrated with security fixes after testing
[02:44] <BenC> I wonder if we can get cd's built with the -proposed kernel
[02:45] <zul> neuralis: ill see if i can get it included
[02:45] <neuralis> zul: thanks; currently we're not installing on many hp servers, and that's not good
[02:45] <`ph8> BenC: once the image is up would you mind mentioning it in here/ubuntu+1? If you want to throw in my nick as well that would be *really* useful. Hopefully I can transform my expensive paperweight into a machine with ubuntu on today! :)
[02:48] <`ph8> i'll come and molest you when the patch doesn't work - sods law dictates it won't work first time
[02:48] <BenC> $ git-cherry-pick 0d8fc83fd11b85e917269dce896740e0875c685c
[02:48] <BenC> First trying simple merge strategy to cherry-pick.
[02:48] <BenC> Finished one cherry-pick.
[02:48] <BenC> nothing to commit
[02:48] <BenC> neuralis: seems that patch is already in dapper
[02:49] <neuralis> BenC: the hcs folks tried this yesterday, and the installation failed. ignore it for now; i'll verify with them.
[02:53] <desrt> kernel dudes!!!
[02:54] <desrt> the kernel is like the POPCORN of the movie theatre, really
[02:55] <desrt> it is EXPENSIVE and if you eat too much you will be MONOLITHIC
[02:55] <`ph8> i'd say it's more like the floor
[02:55] <`ph8> or the structural supports for the roof
[02:55] <desrt> where's the fun in rational metaphor?
[02:56] <jbailey> Is there a way of mounting a partition more than once?  I thought I read at some point that Linux was growing that ability
[02:57] <desrt> jbailey; -o bind
[02:57] <jbailey> desrt: Thanks
[02:57] <desrt> use the first mountpoint as the 'device'
[02:57] <`ph8> isn't it a similie desrt ? :)
[02:57] <jbailey> Oh.  See, the problem is that I've lost the original mount point. =)
[02:57] <desrt> ph8; similies are metaphors with 'like' or 'as'
[02:57] <`ph8> lol
[02:57] <desrt> you might be screwed, then :)
[02:58] <jbailey> I move mounted something onto /, and forgot to get my HD partition out of it first. =)
[03:01] <Mithrandir> jbailey: see if you can move-mount / out of the way to get the old one back?
[03:02] <desrt> Mithrandir; that would involve mounting / inside of itself
[04:00] <zul> Mithrandir: have you sent me your nvidia stuff yet?
[04:00] <Mithrandir> zul: no, I'm a slacker.
[04:00] <zul> ok...good to know that there are more of us out there
[04:00] <`ph8> heh
[04:02] <Mithrandir> I need to push some of my local changes to the tree too.
[04:02] <Mithrandir> zul: for auto-update-initramfs we need to patch kernel-package.
[04:02] <zul> Mithrandir: thats what i thought ;)
[04:04] <Mithrandir> it's probably easy enough; make /usr/share/kernel-package/pkg/virtual/xen/postinst a symlink to /usr/share/kernel-package/pkg/image/postinst (and similar for the other pre/post scripts)
[04:06] <Mithrandir> I haven't tested that, though, so we might want to test it properly first.
[04:06] <`ph8> ah zul, i meant to ask - with the fix for jmicron chipsets/IDE, were you aware of issues detecting SATA ports as well? Because once i'd got around the IDE problems that's the wall I hit..
[04:07] <zul> no i wasnt
[04:07] <`ph8> do you think the fix will cover that as well?
[04:07] <`ph8> or was it very specifically ide
[04:07] <zul> probably not in this upload
[04:07] <`ph8> we'll see :o(
[04:07] <`ph8> i'm going to have to use fedora if it doesn't work in this upload
[04:08] <`ph8> that's how horribly horribly twisted the situation has become
[04:13] <`ph8> zul: did you mean the fix won't be in this upload back then? or just that sata may not be covered?
[04:19] <Nafallo> BenC: I just noticed the new kernel doesn't close bug #58117, is this an error in the changelog or did you forget it?
[04:21] <zul> ph8: well have to see
[04:25] <BenC> Nafallo: just didn't make it into this upload
[04:25] <Nafallo> oki. but the plan is still to have it in edgy?
[04:35] <`ph8> latest changelog's out?
[04:35] <`ph8> oh no, old one
[04:35] <Nafallo> `ph8: edgy-changes
[04:35] <`ph8> you can all go about your business :)
[04:43] <zul> right im off for the ultrasound
[04:50] <infinity> BenC: *poke*
[05:05] <BenC> crappy xchat topic editor
[05:06] <mjg59> BenC: So this is just the ACPI stuff, not the more recent stuff I sent you?
[05:06] <BenC> yeah, minus the stuff you sent me in the past day or so
[05:06] <BenC> fans and "more acpi crack" are not in there, but will be in a few days
[05:06] <mjg59> BenC: So you didn't add the patch you accidently dropped?
[05:06] <mjg59> Ah. That has the downside of meaning that fans won't work on HP laptops.
[05:07] <BenC> I can do another upload real quick
[05:07] <BenC> there is no lrm or linux-meta yet
[05:07] <mjg59> The 0003 one that I resent (or was it 0004? Anyway) is the important one, though the other fan ones would also be helpful
[05:07] <mjg59> Sorry, I should have pointed that out
[05:08] <infinity> BenC: Can I get you to mangle debian/control in the dapper branch so I don't have to do manual archive overrides on the next ABI bump?
[05:09] <infinity> BenC: We always send ufs-modules-* to universe in dapper.
[05:09] <BenC> infinity: -security or -proposed?
[05:09] <infinity> BenC: Yes? :)  Just in your git branch for the next upload(s).
[05:09] <infinity> BenC: Just got reminded of this when I did queue/new processing for the latest dapper security update.
[05:10] <BenC> ufs-modules package stuff is done by kernel-wedge
[05:11] <mjg59> Is the ti acx firmware in l-r-m?
[05:11] <infinity> BenC: And that can't easily be mangled?  If not, nevermind.
[05:11] <BenC> mjg59: yeah
[05:11] <mjg59> BenC: Ah, ok
[05:11] <mjg59> BenC: Have we mangled it so a more useful version is the default for PCI chipsets?
[05:11] <BenC> infinity: I think I can override it. I'll give it a shot
[05:11] <BenC> mjg59: not yet, let me do that for the edgy lrm upload
[05:12] <mjg59> (Amusingly, I lost my acx card about a day after getting it. I think it may be under my bed somewhere)
[05:12] <mjg59> BenC: Ok, cool
[05:14] <BenC> mjg59: I don't see anything about a missing patch, I have "more acpi crack" tarball, and the fan stuff
[05:14] <mjg59> BenC: I mentioned on IRC that you'd applied all the patches from the previous tarball except one, and then resent that with the fan stuff
[05:14] <mjg59> (I believe)
[05:15] <BenC> can you check current git to see if I got it, because I don't see it in my inbox
[05:19] <mjg59> BenC: It was called 0003-Fix-fan-control-on-HP-hardware.txt and was originally in the pm-fixes.tar.gz tarball
[05:20] <BenC> "The fan on/off code is currently a touch flaky. This patch from kernel 
[05:20] <BenC> bugzilla 7122 improves that.
[05:20] <BenC> "
[05:20] <BenC> that it?
[05:20] <mjg59> No
[05:20] <mjg59> Hm.
[05:20] <mjg59> It's possible that I accidently failed to resend it.
[05:20] <BenC> I probably don't have it then
[05:21] <mjg59> You applied all the other patches from that tarball
[05:21] <BenC> weird that I missed it, I used git-applymbox to apply them all
[05:21] <mjg59> Well, checking the tree it's certainly not there
[05:22] <BenC> if you can resend it, I'll add it now and reupload
[05:22] <mjg59> Ok
[05:22] <BenC> infinity: Oh, that reminds me, I need to ask about my grand scheme to make sure things aren't going to break
[05:23] <BenC> infinity: If I have 2.6.15-50 in dapper-proposed, and 2.6.15-27 in dapper-security, the next upload to -security (which will likely be -27 still) wont be rejected because it is older than the -50 in -proposed, right?
[05:26] <mjg59> BenC: I'll send you the stuff I have outstanding now
[05:27] <BenC> ok
[05:31] <mjg59> Ok, done
[05:43] <BenC> mjg59: Weird, I remember that patch
[05:43] <BenC> applied now
[05:44] <mjg59> BenC: Thanks
[05:44] <mjg59> BenC: Just that one, or the others as well?
[05:46] <BenC> all of them, I had the others already
[05:46] <mjg59> Ok, cool
[05:46] <mjg59> Thanks!
[05:52] <dantolini> hi everybody
[05:53] <dantolini> sb present?
[05:54] <dantolini> i'm looking for someone who has experience withe the problem that you get the error message: "access beyond end of device"
[05:55] <BenC> sounds like your filesystem is reporting that it is bigger than the device that it's one
[05:55] <BenC> is this on a CD?
[05:55] <dantolini> exactly
[05:55] <dantolini> file system
[05:56] <BenC> or maybe a USB drive?
[05:56] <dantolini> ext3 partition (root file system, unluckily)
[05:56] <mjg59> dantolini: What hardware?
[05:56] <dantolini> i did a little research, but i have no idea where this problem comes from
[05:56] <BenC> how was the fs created?
[05:56] <dantolini> it's a hard drive
[05:56] <dantolini> fs = fstab?
[05:56] <mjg59> dantolini: On what hardware?
[05:57] <dantolini> mjg59, you wanna know the producer`of the hard drive?
[05:57] <mjg59> dantolini: No, I want to know what it's plugged into
[05:58] <dantolini> well, i don't understand very well.. but i have 2 hard drives on one IDE-cable...
[05:58] <dantolini> ?
[05:58] <BenC> ide chipset, scsi chipset, usb, firewire, what?
[05:58] <dantolini> its IDE
[05:59] <BenC> lspci | grep -i IDE
[05:59] <dantolini> what's that for?
[05:59] <BenC> what do you get from that?
[05:59] <dantolini> (currently i booted with windows, since it's not possible with ubuntu)
[05:59] <BenC> run that command so we can see what IDE controller you have
[06:00] <dantolini> ah, can i get this information in windows as well?
[06:01] <BenC> dantolini: what version of ubuntu did you install, and how did the install process create the filesystem?
[06:01] <BenC> probably, but don't ask me how, because I can't remember
[06:01] <`ph8> BenC: Do you happen to know if the daily is on its way old bean?
[06:01] <`ph8> (NB: I have no confirmed sources for implying that you are old, or a bean)
[06:02] <BenC> e.g. did you use automatic filesystem creation where it did everything itself, or did you use existing partitions, or create your own partitions?
[06:02] <dantolini> first insalled ubuntu 5.0x, then upgraded a few month ago... and a few days ago i got this error message while booting up (in the recovery mode, in normal mode, it just hangs itself)
[06:02] <BenC> ph8: read topic
[06:02] <dantolini> upgraded to 6.06
[06:02] <BenC> I just uploaded edgy kernel
[06:03] <dantolini> the file system was created automatically..
[06:03] <`ph8> ah. cool - will that work into an install-cd image release today?
[06:03] <`ph8> or can i update my install cd somehow?
[06:03] <dantolini> after i created partitions for it....
[06:04] <BenC> dantolini: what changed when the problem started occuring? Some security update, nothing?
[06:04] <dantolini> i have : Standard-IDE-ATA/ATAPI-Controller
[06:04] <BenC> so when you say "upgraded" you really you freshly installed dapper 6.06?
[06:04] <BenC> windows isn't descriptive enough
[06:05] <dantolini> well, i was in holidays for a month, and i had to do a major update (all of august)... but, i don't think that the problem occurred directly whilst the next bootup
[06:06] <dantolini> no, i made an upgrade from the last version to the actual ubuntu..as described somewhere on ubuntu.com
[06:06] <BenC> it's very hard to follow what happened
[06:06] <dantolini> yeah, i think...
[06:06] <dantolini> so:
[06:06] <dantolini> installed ubuntu 5.10 with dvd
[06:06] <dantolini> then upgraded directly to 6.06 lts without cd...internet connection
[06:06] <BenC> if you made an upgrade, then how did you "create partitions"?
[06:07] <BenC> upgrades don't involve creating new partitions
[06:07] <dantolini> then problem, after upgrades of all august
[06:07] <dantolini> sorry, updates of august
[06:07] <dantolini> the partitions were already created..
[06:08] <dantolini> so i guess this "access beyond end of device" problem has not a very specific source...
[06:08] <BenC> at what point did the problem start occuring?
[06:08] <dantolini> i did an fsck.ext3 of the partition if the results of that helped..
[06:09] <dantolini> well, for me it was a little out of the blue.. thats why i am so clueless
[06:09] <dantolini> i'd say after the update session a few days ago
[06:09] <BenC> no, what I need to know is two things...1) What IDE chipset you have (get this from the lspci output, boot from the CD if you have to), and 2) What event during your install/upgrade happened right before you started having problems
[06:10] <dantolini> ok.. that will take a while..
[06:10] <BenC> do you have an older kernel that you can boot from?
[06:10] <dantolini> nah, just live dvd
[06:10] <dantolini> all the kernels have the same problem
[06:11] <BenC> unless you removed kernels, then grub should have an older kernel in the menu on boot
[06:12] <dantolini> grub lists me at least 8-10 kernels..., but i'm going to do it with the live cd...
[06:12] <BenC> then you need to tell me what kernels you have installed (versions listed in grub will work)
[06:12] <BenC> if all the kernels have the problem, then this isn't something new, it just took awhile for it to appear (maybe something is finally being written to the problem area)
[06:12] <BenC> dantolini: No, I need to know if any of the older kernels work
[06:12] <BenC> this is very important
[06:13] <dantolini> i've already tried
[06:13] <dantolini> not a single works..
[06:13] <BenC> no need to try the rescue options, just  the main kernels
[06:13] <BenC> so they all failed?
[06:13] <dantolini> yep
[06:13] <BenC> then this isn't a new problem, it's existed on your machine since it was installed
[06:13] <BenC> you started off with a 5.0x install?
[06:14] <dantolini> yep
[06:14] <dantolini> i think 5.10
[06:14] <dantolini> yes, exactly 5.10
[06:14] <BenC> I'd say your best bet is to use the livecd to mount your drive, copy off any data you need, and do a fresh install
[06:14] <dantolini> but that was more than six months ago
[06:15] <BenC> I don't know if there's any way to recover (easily) from a filesystem problem like that
[06:15] <dantolini> mhm, i kinda tought that.. :)
[06:16] <`ph8> BenC: There doesn't appear to be an entry for today at bcollins/kernels-daily/. Also, would you mind giving me a yes/no as to whether this will make it into a new 'daily' (although there hasn't been one for days) install-cd image?
[06:16] <BenC> ph8: I uploaded it to edgy, not kernels-daily
[06:16] <BenC> ph8: I don't handle cd images, so I wouldn' t know
[06:16] <`ph8> ah.
[06:16] <dantolini> and in your opinion there is no chance that this could be a BIOS problem?
[06:16] <`ph8> do you not know how long it usually takes your kernel to get into them or anything?
[06:17] <BenC> dantolini: what makes you think that?
[06:17] <BenC> ph8: No idea
[06:17] <`ph8> :o( cheerse then
[06:17] <`ph8> i wish i could type :/
[06:17] <`ph8> i'm going to have to go to fedora at this rate
[06:17] <BenC> dantolini: is there something you are not telling me about your setup? :)
[06:18] <dantolini> because, i read about a 1000 different things of people who had this problem... and one said to another that sometimes if the battery gets weak, settings of partitions could be deleted
[06:18] <BenC> dantolini: you could always redo your BIOS settings to see
[06:18] <dantolini> BenC, well, i have read a lotta things, but i wanted to know what an "expert" thinks unindependently
[06:19] <dantolini> if i redo my BIOS settings such parameters would be recovered?
[06:19] <BenC> ph8: there's more to this kernel than just the JMicron patches, there'a hundreds of other patches that need adding and testing
[06:19] <dantolini> unluckily i have no idea of bios
[06:19] <BenC> ph8: Not to mention that the cd builds are dependent on thousands of packages besides the kernel, any of which could hold it back from creating a working image
[06:20] <BenC> ph8: Have some patience, we are not all here to serve your one patch for your one server
[06:20] <BenC> dantolini: I can't help you there
[06:21] <dantolini> ok..
[06:21] <dantolini> i just also like to know what kink of information you expect from the lspci
[06:21] <dantolini> i mean, whats to know about the ide?
[06:22] <BenC> ~$ lspci | grep -i IDE
[06:22] <BenC> 00:1f.2 IDE interface: Intel Corporation Enterprise Southbridge SATA IDE (rev 09)
[06:22] <BenC> the exact chipset that you are using
[06:22] <BenC> different ones use different drivers
[06:23] <dantolini> Silicon Integrated System
[06:23] <dantolini> ?
[06:23] <BenC> narrowing down which driver helps us understand what the problem might be
[06:23] <BenC> there are at least 2 drivers for SiS, I believe
[06:23] <BenC> so more info is needed
[06:25] <dantolini> ok :)
[06:25] <`ph8> BenC: It's been widely documented for a long while now, there's at least three forum threads like this one -> http://www.ubuntuforums.org/showthread.php?t=258010&page=2 - it's not purely selfish. This is something that's already working on things like redhat, obviously it works with windows because they must have thrown money at it at some point.
[06:25] <`ph8> I don't think it's especially selfish
[06:25] <`ph8> I've spoken to quite a few people sitting next to expensive paperweights
[06:25] <dantolini> so i'll boot from the live dvd now. you'll be here for another 20 mins?
[06:25] <BenC> Reported on:
[06:25] <BenC> 2006-08-23 
[06:25] <BenC> that's not that long
[06:25] <dantolini> don't underestimate me ! :D
[06:25] <dantolini> a plus
[06:25] <BenC> dantolini: ok
[06:25] <`ph8> considering the fix was already in the kernel.org kernel..?
[06:26] <`ph8> well, if you say so
[06:26] <`ph8> i imagine it takes a while to backport things like that
[06:26] <BenC> ph8: the patch is in the kernel git, and the tarballs are uploaded, it's now a matter of hours before you can use it
[06:26] <BenC> so lighten up
[06:26] <`ph8> well if i knew that i wouldn't be bothered
[06:26] <`ph8> !!
[06:26] <`ph8> when you say use it?
[06:26] <`ph8> you mean a cd image will be available?
[06:26] <BenC> I told you that it was uploaded
[06:27] <`ph8> or i'll have to do some jiggery pokery replacing my install cd kernel?
[06:27] <BenC> no, the kernel .deb's
[06:27] <BenC> I have no idea when the CD's will be ready
[06:27] <`ph8> can i replace my install cd kernel with your image do you know?
[06:27] <`ph8> with your kernel do you know even
[06:27] <BenC> no idea, I've never messed with install cd's
[06:27] <`ph8> ok
[06:27] <BenC> other than booting them
[06:27] <`ph8> If you fancy using your ubuntu team member status toi find the person who makes the install cds...
[06:28] <`ph8> fancy asking them what happened to dailies? :)
[06:28] <BenC> ask in #ubuntu-devel
[06:28] <Mithrandir> we build daily installation CD daily.
[06:28] <BenC> someone there will know
[06:28] <Mithrandir> I just forgot to turn the cronjobs on this morning, but they're there now.
[06:29] <`ph8> lo Mithrandir 
[06:29] <`ph8> not @ http://cdimage.ubuntu.com/daily/ ?
[06:29] <Mithrandir> yes
[06:29] <Mithrandir> that is, there probably aren't any for today built
[06:29] <Mithrandir> since the cronjobs were off this morning
[06:29] <`ph8> i see
[06:29] <`ph8> so there will be one built tomorrow?
[06:30] <`ph8> (or can you build one for today? ;))
[06:30] <Mithrandir> they'll be built tomorrow morning; I can build one for today, but if the kernel with the stuff you want haven't hit the archive yet it'll be more or less equivalent with the just-release knot-3
[06:31] <`ph8> i see, and the archive is what Ben just referred to as what it will reach in a few hours?
[06:31] <Mithrandir> yes
[06:32] <`ph8> right, i'll sit and wait for it tomorrow then i guess :o)
[06:32] <`ph8> will get it installed eventually!
[06:32] <`ph8> thx
[06:32] <`ph8> i'll post on the forums so the others know
[06:37] <dantolini> BenC, 0000:00:02.5 IDE interface: Silicon Integrated Systems [SiS]  5513 [IDE] 
[06:44] <dantolini> does that help? :)
[07:05] <dantolini> hullo??
[07:08] <dantolini> well, BenC if you have an idea, contact me stupid_homer [at]  gmx [dot]  net
[07:08] <dantolini> thanks anyway
[07:32] <Mithrandir> zul: is there any reason why you make -amd64-generic rather than just -generic for amd64 too?
[08:26] <Mithrandir> renamed to -generic
[08:29] <gnomefreak> ok ty
[09:08] <Mithrandir> BenC: 2.6.17-8.21 ftbfs
[09:08] <BenC> yeah, working on it now
[09:09] <BenC> Mithrandir: it is -generic on amd64
[09:09] <Mithrandir> that broke all the others too?
[09:09] <BenC> -amd64-k8 and -amd64-xeon are gone
[09:10] <BenC> there is only -generic for amd64 now
[09:10] <BenC> and -server
[10:05] <fractalmind> Can anyone here help me get Edgy to boot my custom AKPM kernel?  I test Andrew Morton's unstable kernel tree, and since upgrading to Edgy, I find that my custom kernels won't boot.  This seems to be due to Edgy's GRUB support of UUID.  My custom kernel can't open the root device.
[10:07] <fractalmind> Also, when I try to make an initrd, I am getting an error.  I am running "mkinitrd -o initrd.img 2.6.18-rc6-mm2" and get back "/usr/sbin/mkinitrd: 253:0: Cannot find LVM device."
[10:07] <fractalmind> I am not using an LVM device or RAID.  
[10:10] <Mithrandir> you might want to run mkinitramfs instead.
[10:10] <Mithrandir> zul: you broke the amd64 builds. :-P
[10:12] <fractalmind> Mithrandir: Would not having an initrd have something to do with root=<UUID> not working?  I have all the requisite drivers built into the monolithic vmlinuz for it to access the boot partition.   
[10:13] <Mithrandir> fractalmind: you need udev there too, to make the by-uuid symlinks
[10:14] <fractalmind> Mithrandir: In the kernel?  I already have Edgy's udev installed, and the Ubuntu Edgy kernel boots fine.
[10:14] <Mithrandir> fractalmind: in the initramfs.
[10:16] <fractalmind> Mithrandir: Is there an Ubuntu HOWTO that explains putting together the necessary bits in the initramfs?
[10:16] <fractalmind> Mithrandir: I'll google for it.
[10:16] <mjg59> fractalmind: mkinitramfs
[10:16] <mjg59> Seriously. Our boot system heavily depends on doing it that way
[10:17] <fractalmind> mjg59: I just built one for my custom kernel "mkinitramfs -o initrd.img 2.6.18-rc6-mm2".  
[10:17] <makx> fractalmind: check out initramfs-tools manpage fore details
[10:19] <fractalmind> mjg59: I'll copy the resulting initrd.img to initrd.img-2.6.18-rc6-mm2 and make the necessary changes to GRUB's menu.lst.  I'll go read up on whether I need to do more to get the udev stuff in the initrd.img file.
[10:19] <fractalmind> makx: thx
[10:21] <mjg59> fractalmind: The udev stuff is copied in automatically
[10:22] <fractalmind> mjg59: Cool.  Yeah, I am reading the manpage about it now.
[10:22] <makx> mjg59: latest usplash bar jumps back and forth for me on debian
[10:22] <mjg59> makx: "jumps back and forth"?
[10:22] <makx> you call usplash_write from gdm or?
[10:23] <mjg59> No...
[10:24] <makx> debian/usplash.init has in the stop target usplash_write "TIMEOUT 15"
[10:25] <mjg59> The stop target doesn't stop usplash
[10:25] <mjg59> It's, uh, counterintuitive
[10:25] <makx> i know they are inverted :)
[10:25] <mjg59> Anyway, I didn't write that stuff
[10:25] <makx> but when i shutdown the stop gets invoked
[10:25] <mjg59> Oh, it's supposed to be launched by gdm
[10:25] <mjg59> On shutdown
[10:26] <makx> yes in debian i tried to workaround with "usplash -c &"
[10:26] <makx> but since latest that doesn't make usplash happy
[10:26] <mjg59> Scott's probably a better person to ask
[10:26] <makx> ok
[10:26] <makx> progressbar makes funny jumps :)
[10:44] <zul> Mithrandir: ill unbreak it tonight
[10:44] <Mithrandir> zul: I'll probably beat you to it, since I have something which works locally.
[10:44] <Mithrandir> (and almost works for lrm)
[10:45] <zul> sweet...commit it to git and send me the crack :)
[10:45] <ajmitch> good, zul has users clamouring for packages :)
[10:45] <mjg59> zul: Looks like i386 is b0rked too
[10:45] <ajmitch> seent hat bug that xen-ioemu-3.0 is empty?
[10:46] <zul> i just got back from the ultra sound so no
[10:47] <zul> grr...ill take a look at it
[10:49] <mjg59> zul: linux-source-2.6.17, that is (re: my comment)
[10:49] <mjg59> The asus_acpi code is malformatted in some way
[10:50] <zul> gah...cant get a break can i?
[10:50] <zul> im just catching up
[10:52] <zul> mjg59: i dont see your asus_acpi code comment
[10:53] <mjg59> zul: The i386 build being broken
[10:55] <kinema> what meta package should be installed on a 686 smp box (quad ppro) when running edgy?
[10:55] <kinema> are the linux-generic compiled for SMP?
[10:56] <Mithrandir> zul: dpkg-deb: Bygger pakken xen-restricted-modules-2.6.16-11.2-generic i ../xen-restricted-modules-2.6.16-11.2-generic_2.6.16-1_amd64.deb.
[10:56] <zul> mjg59: looks like it fixed it git
[10:57] <mjg59> Ah, yes
[11:02] <zul> kinema: generic
[11:07] <ajmitch> Mithrandir: great, I can drop vmware now :)
[11:07] <ajmitch> mjg59: do you remember when the last big batch of ACPI updates landed in the ubuntu kernels? was it for dapper?
[11:08] <ajmitch> I just noticed that I have no decent ACPI info on the 2.6.16 xen kernel
[11:11] <mjg59> ajmitch: There was a set for dapper, and there's a set for edgy
[11:11] <mjg59> I have no idea what the interaction between xen and acpi status reporting is
[11:13] <Mithrandir> making sure something actually ends up in the deb == good
[11:20] <kinema> zul: I checked and I am running linux-image-2.6.17-7-generic via the linux-generic meta-package but when I look at /proc/cpuinfo I only see a single cpu when i should see four.  are you sure that the generic kernel packages are smp capable?
[11:21] <kinema> Linux lorax 2.6.17-7-386 #2 Wed Sep 6 17:53:03 UTC 2006 i686 GNU/Linux
[11:22] <mjg59>  Linux lorax 2.6.17-7-386
[11:22] <mjg59> No, that's not the generic kernel
[11:23] <kinema> hmmm...
[11:23] <kinema> the only image package dpkg is listing as installed is linux-image-2.6.17-7-generic 
[11:24] <crimsun> Linux garnish 2.6.17-7-generic #2 SMP Wed Sep 6 17:56:40 UTC 2006 i686 GNU/Linux
[11:24] <crimsun> that's what generic reports, and it is SMP-aware
[11:24] <kinema> maybe not.....
[11:24] <mjg59> If you were running 2.6.17-7-generic, it'd say 2.6.17-7-generic
[11:24] <mjg59> So something's gone wrong somewhere
[11:26] <kinema> linux-image-2.6.17-7-386 is installed as well...  it's been selected for removal but it's currently being used.
[11:27] <kinema> how do i tell grub to boot the other kernel?  i don't have physical access to the box right now so i can't select it when i reboot.
[11:36] <Mithrandir> zul: you might want to pull from http://git.err.no/xen-ubuntu-2.6.16-tfheen/ and use that as a basis; I can't be bothered to find my laptop with ssh key on it.
[11:38] <Mithrandir> zul: as well as http://err.no/tmp/xen-restricted-modules-2.6.16-2.6.16.1.tar.gz which is just where I am.  ETA ~1 hour or so.
[12:06] <zul> mdz: looking now
[12:09] <zul> oops...sorry Mithrandir looking now i meant