[12:19] <Jerub> 2.2.9a is basically trunk, and is the package in ubuntu
[12:26] <lifeless> so, please file a bug on the kernel source package (linux-source in launchpad)
[12:26] <lifeless> BenC: ^
[12:27] <lifeless> Jerub: I was hoping BenC would be around, but hes apparently not, and infinity is ill I think
[12:27] <BenC> linux-source-2.6.X
[12:27] <lifeless> ah-ha, da man is here
[12:27] <lifeless> BenC: any chance of including this ?
[12:28] <BenC> Jerub: send me the patch, or paste a link to it here
[12:28] <Jerub> http://www.netfilter.org/projects/ipset/index.html
[12:29] <Jerub> weird.
[12:29] <BenC> I don't see a patch
[12:30] <lifeless> Jerub: its in patch-o-matic
[12:31] <Jerub> the main site for it looks down now
[12:31] <Jerub> was up yesterday I thought
[12:31] <lifeless> its in p-o-m base
[12:31] <lifeless> I think
[12:31] <BenC> http://www.netfilter.org/projects/patch-o-matic/index.html ?
[12:31] <lifeless> http://www.netfilter.org/projects/patch-o-matic/pom-base.html
[12:31] <BenC> which patch?
[12:32] <BenC> iptbles "set" match?
[12:32] <Jerub> that's the one
[12:33] <BenC> maybe I'm being dense, but I don't see a patch, just a desc
[12:33] <Jerub> you're right, I don't see a downloadable patch either
[12:33] <lifeless> I'm digging a little
[12:34] <BenC> brb
[12:34] <Jerub> I have the patch we apply to our own kernels, but I'd rather make sure to use the current upstream on
[12:35] <lifeless> is it not http://svn.netfilter.org/cgi-bin/viewcvs.cgi/trunk/ipset/ ?
[12:36] <kylem> lifeless, that's the userspace tool
[12:36] <lifeless> kylem: dang. I was hoping it wasn't ;)
[12:36] <kylem> :)
[12:37] <lifeless> Jerub: so I'm pretty sure that its in 'base' - but I dont know what tha tmeans
[12:38] <Jerub> lifeless: I think it's accessable via the 'patch-o-match-ng' program
[12:38] <kylem> this is by far the most convoluted piece of crap i've ever seen.
[12:38] <Jerub> but as I'm not the kernel developer on my team, and he's in england, I'm not sure how to use it.
[12:39] <lifeless> http://www.netfilter.org/documentation/HOWTO//netfilter-extensions-HOWTO-2.html#ss2.2
[12:39] <lifeless> garhfuckle
[12:39] <lifeless> http://svn.netfilter.org/cgi-bin/viewcvs.cgi/trunk/patch-o-matic-ng/patchlets/set/
[12:40] <kylem> so, uhm, i don't see anything in here that requires patching.
[12:40] <lifeless> theres a pom-to-patch program that may be useful
[12:40] <kylem> it seems to be entirely standalone.
[12:41] <lifeless> yay for modular code
[12:41] <lifeless> kylem: you're looking at the netfilter kernel module now ?
[12:42] <lifeless> Jerub: what you do is svn checkout the patch-o-matic-ng svn tree
[12:42] <lifeless> Jerub: then run the pom tools to get a patch out, or apply it or whatever
[12:42] <lifeless> Jerub: so I suggest doing that in pom2patch mode
[12:42] <kylem> lifeless, as far as i can see, they would be able to provide an external module and just build against headers like others do.
[12:44] <lifeless> kylem: you're the kernel guy ;). I'm just facilitating.
[12:48] <BenC> [    0.072411]  Booting paravirtualized kernel on bare hardware
[12:48] <BenC> sweet
[12:53] <kylem> hot.
[12:57] <Jerub> BenC: I'll get that patch for you shortly.
[02:38] <zul> hmm?
[02:42] <zul> BenC: ping...which paravirt are you running?
[02:43] <BenC> the stock 2.6.20 paravirt, but with vmware's VMI patches
[02:43] <BenC> for some reason my vmware isn't causing this kernel to show that it's using vmi paravirt though
[02:44] <zul> cool
[02:45] <zul> there should be some xen-paravirt patches being sent to akpm soon
[02:48] <zul> macd: i dont have the problem if im not using ioemu
[03:13] <Jerub> okay, I've built the patch, took a bit of wedging
[03:14] <Jerub> http://shiny.thorne.id.au/~stephen/set-2.6.20.patch
[03:18] <BenC> Jerub: Hmm...not sure about this one...File a wishlist bug on linux-source-2.6.20 and attach this patch to it, and I'll review it more in the next week
[03:20] <BenC> Jerub: Why would people need to build a new kernel for this?
[03:20] <BenC> It doesn't touch kernel code, just adds a module, which can be built just using the linux-headers packages
[03:22] <Jerub> BenC: I honestly have no idea.
[03:22] <BenC> Jerub: Save yourself the trouble of the bug report, this can be packages and/or built separately from the kernel
[03:22] <BenC> s/packages/packaged/
[03:22] <Jerub> excellent.
[03:25] <lifeless> Jerub: the motu team may be willing to do that for you
[03:25] <Jerub> okay, how do I ask this of them? 
[03:25] <Jerub> :)
[03:26] <BenC> #ubuntu-motu? I think that's the channel
[03:27] <zul> yep
[03:27] <lifeless> thts the channel
[04:01] <BenC> crimsun: ping
[05:29] <crimsun> BenC: pong
[05:43] <BenC> crimsun: I have small issue that I've been meaning to ask you about
[05:44] <BenC> my HP dv5000 laptop with HDA Intel sound doesn't have very high volume
[05:44] <BenC> I checked alsamixer and everything's max'd
[05:44] <BenC> was the same with 2.6.17, and still the same with 2.6.20 latest git
[05:45] <crimsun> BenC: realtek or conexant codec?
[05:45] <crimsun> (head -3 /proc/asound/card0/codec* )
[05:45] <BenC> Conexant
[05:45] <crimsun> ok, that's the issue that Rich Johnson (nixternal) reported
[05:46] <crimsun> it's supposed to be fixed in alsa-driver 1.0.14rc1, but alsa's git/hg trees lack sound/pci/hda/patch_conexant.c
[05:46] <crimsun> so what needs to be done is to pull in that file and any necessary changes to sound/pci/hda/hda*.c
[05:47] <crimsun> sorry, alsa hg -does- contain those fixes
[05:47] <BenC> is there a quick way I can get the alsa tree?
[05:47] <crimsun> but what takashi and jaroslav have pushed to linus does not contain them
[05:47] <crimsun> hg clone http://hg-mirror.alsa-project.org/alsa-kernel alsa-kernel
[05:47] <crimsun> and 
[05:47] <BenC> I have mercurial
[05:48] <crimsun> hg clone http://hg-mirror.alsa-project.org/alsa-driver alsa-driver
[05:48] <BenC> I need also driver updates too?
[05:48] <BenC> alsa
[05:48] <BenC> was hoping it was just kernel :)
[05:48] <crimsun> one sec
[05:49] <crimsun> just -kernel
[05:49] <BenC> ok
[05:57] <BenC> crimsun: what do I need to shutdown to release the ref counts on the snd modules?
[05:58] <crimsun> this should do it: kill $(lsof -t /dev/dsp* /dev/snd/*) && sudo modprobe -r $(lsmod |grep ^snd |awk '{print $1}') && sudo modprobe snd-hda-intel
[05:58] <lifeless> fugly
[06:49] <BenC> crimsun: Yay, I have loud sound!
[06:49] <BenC> problem is, I rebuilt all of alsa to install it
[06:49] <BenC> crimsun: Anyway you can isolate the patches needed for patch_conexant.c?
[06:51] <crimsun> BenC: err, all of alsa (meaning -libs, -utils, too)?
[06:53] <crimsun> BenC: well, you need patch_conexant.c for sure. I'll dig through some history and see if sound/pci/hda/hda* need to be patched (I vaguely remember some additional changes there)
[06:53] <BenC> all of alsa-kernel
[06:54] <crimsun> ah
[06:55] <BenC> had to reset my mixer settings, so hopefully just the conexant changes wont affect that
[06:55] <crimsun> ok, yeah, I'm positive the changes you need are restricted to sound/pci/hda/{hda*,patch_conexant.c}
[06:56] <crimsun> git pull over a 56kbps is fun!
[06:56] <BenC> hehe, I can get just the hda subdir for you if you want
[06:56] <BenC> you might need the history though
[07:02] <crimsun> hmm, some workqueue changes I'm not familiar w/
[07:02] <crimsun> is
[07:02] <crimsun> destroy_workqueue(bus->unsol->workq);
[07:02] <crimsun> deprecated in favor of:
[07:03] <crimsun> flush_scheduled_work();
[07:12] <fabbione> BenC: booting ppc now
[07:19] <fabbione> BenC: http://people.ubuntu.com/~fabbione/IMG_4054.JPG
[07:19] <fabbione> this is the best shot i can get
[07:19] <fabbione> it's a bit out of focus
[07:21] <BenC> that's a hairy looking oops
[07:21] <BenC> udevtrigger, with only two modules loaded, failed to allocate memory
[07:22] <fabbione> i get that on -1- too now
[07:22] <fabbione> WTF
[07:23] <fabbione> this is random
[07:23] <fabbione> now -5- is booting
[07:23] <fabbione> after 10 attempts
[07:23] <BenC> are the oopses always the same?
[07:23] <fabbione> yes
[07:41] <BenC> you might have to pass that one by benh
[07:41] <BenC> is it stable once you get it booted?
[07:41] <crimsun> BenC: 3 patches against current ubuntu-2.6.git sent to kernel-team@ ; the last (patch_conexant.c) is awaiting moderation
[07:45] <BenC> ok, thanks
[07:45] <fabbione> BenC: looks like it
[07:45] <BenC> I'll let you know if that by itself fixes things up
[07:46] <crimsun> ok, thanks
[07:46] <BenC> fabbione: Must be a race on initialization for one of the kmem_caches to get setup
[07:46] <fabbione> BenC: pushing to benh
[07:54] <lifeless> hmm, I have no sound :(
[07:55] <fabbione> lifeless: did you wash your hears? :P
[07:55] <lifeless> new motherboard
[07:55] <lifeless> ah, fixed it
[07:56] <lifeless> reload the sound modules. 
[07:56] <lifeless> possibly it would not have broken if I had rebooted when changing the motherboard
[08:06] <fabbione>   * ubuntu/mol: Fixes to build again. Fabio owes me a beer.
[08:06] <fabbione>     - GIT-SHA f57afa67e6410d9c3a21d3bf64fb4274c8d83bd2
[08:06] <fabbione> BenC: you bet.. if it works :)
[08:27] <fabbione> BenC: 
[08:27] <fabbione> ubuntu/fs/gfs/ops_file.c: In function 'gfs_aio_read':
[08:27] <fabbione> ubuntu/fs/gfs/ops_file.c:452: error: 'buf' undeclared (first use in this function)
[08:27] <fabbione> ubuntu/fs/gfs/ops_file.c:452: error: (Each undeclared identifier is reported only once
[08:27] <fabbione> ubuntu/fs/gfs/ops_file.c:452: error: for each function it appears in.)
[08:27] <fabbione> ubuntu/fs/gfs/ops_file.c:452: error: 'count' undeclared (first use in this function)
[08:27] <fabbione> ubuntu/fs/gfs/ops_file.c: At top level:
[08:27] <fabbione> ubuntu/fs/gfs/ops_file.c:1739: warning: initialization from incompatible pointer type
[08:28] <fabbione> the gfs_aio_read change you did needs to be propagated somehow to __gfs_read
[08:28] <BenC> shut, I didn't mean for that to go in
[08:28] <BenC> shit
[08:28] <BenC> I'll fix it...it will compile, but it will still be broken
[08:29] <fabbione> actually.. ok
[08:29] <fabbione> is it enough to revert that bit?
[08:29] <fabbione> or should i propagate the iovec.iobase as buf?
[08:29] <fabbione> yeah broken i know
[08:31] <fabbione> BenC: if it is the only line to revert, please leave it to me
[08:31] <fabbione> i have a bunch of updates in progress
[08:41] <BenC> sure
[08:41] <fabbione> thanks
[08:41] <fabbione> can i just push to people.. right?
[08:41] <fabbione> it's only gfs1 stuff.. so don't worry.. it won't clash with other changes
[08:42] <BenC> Sure, just follow the commit templates so the changelog gets update right
[08:43] <fabbione> yeps
[08:43] <fabbione> i doubt it will be ready before the meeting
[08:43] <fabbione> i am hoping to give it at least a test
[10:20] <fabbione> BenC: .20-6 looks good on PPC.. mol too
[04:19] <fabbione> BenC: read the scrollback :)
[04:19] <BenC> fabbione: No oopses from -6?
[04:19] <fabbione> BenC: do you know if it's possible to run mol in a window instead of full screen? i can't get to switch back to X
[04:19] <fabbione> BenC: apparently no
[04:20] <fabbione> and still fighting with the network.. but that's probably just my setup
[04:20] <BenC> fabbione: Somewhere in /etc/mol/ you can set it up to use an X window instead of framebuffer
[04:20] <fabbione> BenC: ok cool
[04:21] <BenC> I was switching between fullscreen framebuffer and X when I was using it though...not sure why it wouldn't work for you
[04:21] <fabbione> how? alt+ctrl+f7 ?
[04:27] <BenC> yeah
[04:27] <fabbione> hmm
[04:27] <fabbione> i will have to test that again
[05:10] <zul> hey
[05:49] <doko> hmm, my notebook doesn't boot anymore after installation of l-r-m :-(
[05:49] <thom> doko: atheros? 
[05:49] <doko> thom: yes
[05:49] <thom> launchpad #76294
[05:50] <doko> great!
[07:01] <IDOIT> hi all
[07:02] <IDOIT> i know that /dev/mem is an image of the physical memory ..
[07:02] <IDOIT> but i couldn't cat /dev/mem > file
[07:02] <IDOIT> it says Operationg not permitted .. why ?
[07:03] <IDOIT> more specifically .. i can copy the first 1.1 MB of it (using dd)
[07:03] <IDOIT> i'm using 2.6.15-27-amd64-generic
[07:05] <IDOIT> sorry .. is it a specific ubunto issue or it's a Linux kernel issue .. googling for it didn't help
[07:32] <pmjdebruijn> IDOIT, the Ubuntu kernel isn't heavily modified, so it's probably a generic Linux issue (but this is just an educated guess)
[07:33] <IDOIT> pmjdebruijn: this is my guess too .. but cause googling for this issue showed that this operation could be done normally "cat /dev/mem" .. 
[07:34] <pmjdebruijn> IDOIT, how is the example article dated?
[07:34] <pmjdebruijn> IDOIT, wait, do your have root priviledges?
[07:34] <IDOIT> yes
[07:34] <kylem> what you can do with /dev/mem varies by architecture.
[07:35] <IDOIT> i know that it's an image of physical mem
[07:35] <pmjdebruijn> kylem, any clue why AMD64 is limited?
[07:36] <IDOIT> and i got the same result from intel 32 and AMD64
[07:36] <kylem> pmjdebruijn, probably so that you can only read the BIOS, but that is just a guess.
[08:49] <BenC> wow, vmware has a mac beta
[09:07] <BenC> crimsun: Those conexant patches worked for me
[09:11] <lifeless> BenC: mac intel ?
[09:11] <BenC> lifeless: I would bet it isn't for ppc
[09:26] <tumbleweed> I'm getting no sound on ati-ixp audio after hibernation resume (in feisty)
[09:30] <Nafallo> BenC: would it be helpful if I recompile the current git and use rt2x00 instead of rt2500-legacy?
[09:30] <BenC> You can try, but it doesn't work for my rt2500 cards
[09:31] <Nafallo> ah, you've already tried.
[09:31] <Nafallo> damn. I was hoping they would have came farther with it :-/
[09:32] <Nafallo> thanks anyway. I'll put it down a bit on the prioritylist then :-)
[12:05] <BenC> VMI: Found VMware, Inc. Hypervisor OPROM, API version 3.0, ROM version 0.3
[12:05] <BenC> VMI: ROM license 'Proprietary' taints kernel... inlining disabled
[12:05] <BenC> Booting paravirtualized kernel on vmi
[12:05] <BenC> sweetness
[12:06] <kylem> no more crashes?
[12:07] <kylem> ah.
[12:07] <kylem> i see.
[12:07] <kylem> :)
[12:07] <BenC> had to add "acpi=off noapic" to get it to boot