[05:29] <psusi> what the crap?  git-unpack-objects says it is unpacking 187,127 objects.... only it doesn't.... the objects aren't there
[05:30] <psusi> shouldn't git-unpack-objects < .git/objects/pack/pack-xxxx.pack undo a git-repack?
[05:51] <dolson> if there is a driver that I want to see get into Ubuntu, how do I make that request? do I open a bug on Malone?
[05:53] <psusi> if the driver is in the linux kernel, it should be in ubuntu already
[05:53] <dolson> it isn't
[05:54] <psusi> it isn't in the stock kernel or it isn't built into the ubuntu build?
[05:55] <dolson> I don't think it's in either. it's for Logitech Quickcam Communicate. I don't think it works with the `quickcam` module, and I did some searching, and I found http://home.mag.cx/messenger/ which has another driver
[05:56] <dolson> unless I'm a moron and can't figure it out for myself.
[05:56] <psusi> if it isn't in the stock kernel, it won't go into ubuntu AFAIK
[05:56] <dolson> well there is a big list of add-on drivers in the kernel source tree, in the debian folder
[05:57] <psusi> hrm.... interesting....
[05:57] <dolson> podxtpro is one of them that I can recall, and I don't think it's in a stock kernel
[05:58] <psusi> it is open source right?
[05:59] <dolson> yes, well, it looks like it's based on the qce code. I am going to look more into it, because I am not sure why it looks like the same stuff as the standard quickcam module.. modinfo quickcam doesn't show an alias for my vendor+device ID that I get from lsusb though
[06:04] <dolson> hrmm. that's not good. it makes a module called `quickcam`
[07:53] <Mithrandir> kernel bugs should be assigned to kernel-source-$version or ?
[07:54] <fabbione> linux
[07:54] <fabbione> afaik
[07:54] <fabbione> otherwise linux-source-$version
[07:56] <Mithrandir> yeah, linux worked.  It kinda sucks that the "choose" dialog just gives you some semi-random ordering rather than anything sensible.
[11:09] <Tonio_> hi everyone
[11:09] <netzmeister> hi
[11:10] <Tonio_> just noticed there is no "build" simlink in /lib/modules/2.6.15-16-386 with linux-headers version 2.6.15-16-386
[11:10] <Tonio_> the resulting is about al kernel modules compilations are failing... is that normal ?
[11:12] <netzmeister> i have the same problem too
[11:12] <netzmeister> i have installed the ......-686 Kernel. And all works fine
[11:12] <Tonio_> hum, probematic :)
[11:12] <netzmeister> *g*
[11:12] <Tonio_> s/b/bl
[11:13] <netzmeister> why is this problematic?
[11:14] <Tonio_> well much people will install directly there uname -r linux header versions
[11:18] <netzmeister> yes thats right..
[11:19] <netzmeister> let's wait..
[11:19] <netzmeister> the maintainers are sleeping ;-)
[11:19] <netzmeister> -are
[02:31] <fabbione> BenC: ping?
[02:32] <fabbione> hmm it's a bit early...
[02:33] <fabbione> BenC: please pull from my tree and push to kernel org.
[02:33] <fabbione> Changes:
[02:33] <fabbione> More SUN4V love and fixed a few problems on old sparc hw
[02:33] <fabbione>   * Update redhat cluster suite to CVS20060222.
[02:33] <fabbione>   * Stop spca5xx to redefine PUD_SHIFT that's arch specific mem management.
[02:33] <fabbione>   * Import OCFS2 and configfs bug fixes from 2.6.16-rc5. This commit must be
[02:33] <fabbione>     reverted before we will sync 2.6.16 directly from git or they will
[02:33] <fabbione>     conflict badly.
[02:34] <fabbione> if you can also push to kernel.org it would be lovely
[02:34] <fabbione> i also checked the git purge stuff
[02:34] <fabbione> it works just fine
[02:34] <fabbione> it basically removes from the tree al unrequired objects
[02:34] <fabbione> making the tree WAYYY smaller
[02:35] <fabbione> i meant.. git prine
[02:35] <fabbione> ARGH
[02:35] <fabbione> GIT PRUNE <-
[02:35] <fabbione> ^^^^^
[02:35] <fabbione> that one
[02:59] <BenC> fabbione: getting it in just a minute
[04:05] <fabbione> BenC: cool
[04:06] <BenC> git-pruned the kernel.org and people repo's, finishing the repack on people right now
[04:12] <fabbione> nice
[04:21] <BenC> git-repack needs to be done before git-prune (/me makes a note)
[04:21] <BenC> have to run prune after repack anyway
[04:26] <fabbione> ok
[04:31] <BenC> ok, both repo's done, and my local repo is nice and tidy :)
[04:31] <fabbione> BenC: if you did pull/push i will be glad to trash my local one and clone again...
[04:31] <BenC> I've got it, so yeah, go ahead and ditch your local repo
[04:32] <BenC> push in progress
[04:33] <fabbione> BenC: note that the OCFS2 update has not been done via git-cherry
[04:33] <fabbione> we will need to revert that commit before we sync 2.6.16
[04:33] <fabbione> it was way too complicate to do it with cherry
[04:36] <BenC> ok
[04:36] <fabbione> nice.. pack rsyncing here :)
[04:36] <fabbione> much much better :)
[04:38] <desrt> turns out my firewire woes are fixed by a patch that you personally signed off on for inclusion upstream :)
[04:39] <BenC> desrt: really? :)
[04:39] <BenC> point me to it, and I'll include it
[04:39] <desrt> BenC; i've been annoying the heck out of you on launchpad :p
[04:39] <BenC> is it Al Viro's patch?
[04:40] <desrt> BenC; it seems like launchpad reduces the effectiveness of communication by a good 80%
[04:40] <BenC> my bug emails are backlogged, so I guess I'll get to it soon
[04:40] <desrt> fwiw, this one: https://launchpad.net/malone/bugs/28174
[04:41] <desrt> it's "from the scsi maintainers"
[04:41] <BenC> ok
[05:12] <davyd_> ok
[05:12] <davyd_> on #30557
[05:12] <davyd_> at the moment, on 2.6.15-16
[05:12] <davyd_> when my machine boots up, it is putting itself into C2
[05:13] <davyd_> and the idle counter goes whack, and it clocks the CPU up
[05:13] <davyd_> if I force the machine into C1, the idle counter starts working
[05:13] <davyd_> but my CPU can't idle properly
[05:14] <BenC> davyd_: what CPU type is this?
[05:15] <davyd_> Pentium <
[05:15] <davyd_> M
[05:15] <davyd_> on the i915 chipset
[05:15] <davyd_> incidently, and possibly related to this, is that my laptop is running very warm
[05:15] <BenC> which cpufreq module is being used?
[05:15] <davyd_> and the fan isn't engaging
[05:15] <davyd_> (it hard locked once because of this)
[05:15] <davyd_> BenC: userspace
[05:16] <BenC> you're not using speedstep or anything?
[05:16] <BenC> powernowd uses a kernel module
[05:16] <davyd_> oh, right, let me look
[05:16] <davyd_> where is that kept now, it is no longer in proc
[05:16] <BenC> lsmod | grep cpu
[05:17] <davyd_> speedstep_centrino
[05:17] <davyd_> but yeah, it counts idle ticks properly in C1, but not C2, never seems to go into C3
[05:17] <davyd_> the fan is also not engaging reliably
[05:18] <davyd_> my machine was at 64C
[05:18] <davyd_> since my options were C1 or dodgy idle times causing the CPU to clock up
[05:18] <BenC> was there a prior 2.6.15 kernel that worked ok?
[05:19] <davyd_> the fan trip would appear to be 114C
[05:19] <davyd_> (as a result the machine hard locked)
[05:19] <davyd_> BenC: I've not have a lot of success with Ubuntu 2.6.15
[05:19] <davyd_> 2.6.12 worked perfectly
[05:19] <BenC> what was the first 2.6.15 kernel you used?
[05:19] <davyd_> I didn't personally try, but someone said Debian 2.6.15 was working on this
[05:19] <BenC> but that doesn't help me much :)
[05:20] <davyd_> two secs
[05:20] <davyd_> the first 2.6.15 I have installed is -15
[05:20] <BenC> trying to find out if there was a point where this went bad with our kernel
[05:20] <davyd_> I only upgraded this machine to dapper about a fortnight ago
[05:20] <BenC> ok, so you've only used very recent ones
[05:20] <BenC> any chance of you trying flight 1/2/3 liveCD's to see if they fix anything?
[05:20] <davyd_> someone with an ASUS has had this problem forever
[05:21] <BenC> I assume you are using the -686 kernel, have you tested the -386 kernel?
[05:21] <davyd_> I haven't tested -386, I can
[05:21] <BenC> yeah, that would help a little
[05:22] <davyd_> interestingly, forcing C1 seems to stop cpufreq from working, scrollkeeper update didn't clock my CPU up from 600Mhz
[05:22] <davyd_> ok, I'm installing linux-image-2.6.15-16-386
[05:31] <davyd_> ok, testing -386, back in a bit
[05:38] <davyd> BenC: positive results with -386
[05:39] <davyd> machine has put itself into C3
[05:39] <davyd> getting 100 idle ticks/s (~)
[05:39] <davyd> cpufreq is clocking up and down as required
[05:39] <davyd> fan cut in and out to cool the machine down
[05:41] <davyd> sky2 still seems busted though :-(
[05:41] <BenC> ok, so we have an SMP issue
[05:41] <BenC> the 686 kernel is compiled with SMP support, it's really the only difference from the 686 kernel
[05:41] <BenC> 386 kernel
[05:45] <BenC> Please annotate the bug about this...I would stick with the 386 version (it's really not a performance hit as most people would think)
[05:45] <bluefoxicy> the 686 kernel can see more than 900M of RAM
[05:46] <BenC> so can 386
[05:46] <bluefoxicy> no it can't
[05:46] <bluefoxicy> unless I missed something new
[05:46] <BenC> 386 kernel has highmem
[05:46] <bluefoxicy> in Dapper?
[05:46] <BenC> yes
[05:46] <bluefoxicy> but not breezy
[05:47] <BenC> we aren't talking about breezy :)
[05:47] <bluefoxicy> ok just making sure
[05:47] <bluefoxicy> I know the 386 has never had highmem before
[05:50] <davyd> hmm, on the network front sk98lin is also dysfunctional
[05:50] <davyd> have you had any reports on ethernet usefulness with dapper and the R200?
[05:50] <davyd> (you can tell I use wireless all the time, but that I perhaps forgot to install the -restricted package)
[05:51] <BenC> not that I recall
[05:52] <davyd> ok, I just booted into Windows to check the hardware was actually working
[05:53] <davyd> sky2 gives me an interface, but does no link detection or dhcp
[05:53] <davyd> sk98lin, which used to work, doesn't even give me an interface
[05:53] <davyd> I suppose I should file this as a separate issue
[05:53] <BenC> yeah, I probably need to update the sky2 driver
[06:08] <davyd> ok, -386 seems sane
[06:08] <davyd> do you want me to file on sky2?
[06:15] <BenC> I marked it on my todo, so no need
[06:17] <davyd> ok
[06:17] <davyd> thanks for your help
[06:20] <BenC> no, thanks for checking in
[06:20] <BenC> hopefully I'll get this squared away for you
[06:22] <davyd> no probs
[07:53] <BenC> mjg59: ping
[09:26] <bluefoxicy> psusi:  the via dri in 2.6.15 seems horridly broken, do any of the devs have via chipsets, or do you know if it's part of mainline?
[09:26] <bluefoxicy> if it's part of mainline I'll hit the lkml with the bug.
[09:27] <psusi> via dri?  I have a via chipset, but my video card is an ati radeon 9800 pro
[09:27] <bluefoxicy> my via chipset has a via video adaptor in its north bridge
[09:27] <psusi> ewww.... don't use it... integrated video sucks ;)
[09:28] <BenC> I have a VIA EPIA, I can try that some time
[09:28] <psusi> it robs memory space and time from the cpu
[09:28] <bluefoxicy> 0000:01:00.0 VGA compatible controller: VIA Technologies, Inc. S3 Unichrome Pro VGA Adapter (rev 01)
[09:28] <bluefoxicy> psusi:  I don't need heavy video
[09:28] <bluefoxicy> I'm running VESA right now, and if I play any animated videos the CPU hits 100% and they lag and X drags and shit.
[09:28] <psusi> bluefoxicy: that's fine... it's still slowing down processing because of the stollen memory cycles
[09:29] <BenC> that doesn't sound like a kernel issue
[09:29] <psusi> what exactly does "horrbily broken" mean?
[09:29] <bluefoxicy> with the Via driver, I can actually seek back and forth while playing videos, and I get decent GLX gears performance, and the CPU usage is lower.
[09:29] <BenC> My VIA EPIA M10k dies completely in windows if I startup any opengl apps :)
[09:29] <bluefoxicy> psusi:  um, well, if I use the via driver in X
[09:29] <bluefoxicy> my kernel oopses several times
[09:30] <bluefoxicy> and somewhere between an hour and 5 hours in I get a hard freeze (kernel level memory corruption?)
[09:30] <BenC> ah, that's a different story :)
[09:30] <psusi> oopsesses are not good ;)
[09:30] <BenC> yeah, dem oopsses is bad tings
[09:31] <bluefoxicy> at first I thought it was because of rhythmbox so I filed a bug against rhythmbox, then i changed the bug saying it was video related, and worked down :)
[09:31] <bluefoxicy> (I thought it was rhythmbox because it seemed playing shit in rhythmbox made my kernel freeze in about 5 minutes)
[09:31] <BenC> kernel freezes are always the kernels fault
[09:32] <bluefoxicy> https://launchpad.net/distros/ubuntu/+source/linux/+bug/29586
[09:32] <BenC> just because it happened when app Foo was running, doesn't make it the apps fault
[09:32] <BenC> reassign to linux-source-2.6.15
[09:32] <bluefoxicy> yeah
[09:32] <BenC> don't create a new target, just rename the old one
[09:32] <bluefoxicy> It's targeted at kernel
[09:33] <bluefoxicy> also when 2.6.16 is out will there be a universe package for it?  ^o.o^
[09:35] <zul> no 
[09:40] <BenC> bluefoxicy: target it at "linux-source-2.6.15", else I will never see it
[09:43] <bluefoxicy> BenC:  what's the priority
[09:44] <BenC> Major
[09:44] <bluefoxicy> priority low med high?
[09:44] <BenC> Is it a regression from breezy, or did 2.6.12 work ok?
[09:44] <bluefoxicy> Back in the day of 2.6.12 there wasn't even a driver IIRC
[09:44] <bluefoxicy> I KNOW there was no X driver in breezy
[09:47] <BenC> well, if it's not a regression, and there exists a workaround, and it affects few people (this is the first I have heard of it), then mark it Normal, High
[10:11] <bluefoxicy> reassigned
[10:43] <peterz> BenC: I got 2.6.15-16-686 and it still won't boot a UML kernel.
[10:50] <BenC> was that related to the 1G/3G change?
[10:50] <BenC> right now it's 2G/2G
[10:51] <peterz> might be; I'll try and boot one of my own kernels for a change and see what they do with a 2/2 split
[11:21] <mjg59> BenC: Hi