[00:01] <hallyn> smb: just fyi, bug 776936 has me worried bc it seems random, i have no idea of the cause, and we don't much test on i386...
[00:02] <ubot2> Launchpad bug 776936 in qemu-kvm "Running KVM guest causes kernel panic on host" [Critical,New] https://launchpad.net/bugs/776936
[00:56] <jjohansen> hallyn: ugh, that is not good
[01:19] <hallyn> jjohansen: yup.  i'll get a simple setup to reproduce in the morning, but i won't have his host storage or network setup.  so we'll see just how easy it is to trigger
[01:28] <jjohansen> hallyn: easy to reproduce would be good, I'll see if can't get to poking at it tonight/tomorrow
[01:58] <hallyn> jjohansen: cool, thanks much
[03:02] <psusi> can someone explain to me why /sys/bus/usb makes no sense?  lsusb shows my keyboard is bus 2, device 3, but /sys/bus/usb/devices has nothing starting with a 2 and having a 3 in it... how are you supposed to find that node?
[03:07] <mjg59> psusi: lsusb is giving the kernel representation, while /sys/bus/usb is the bus representation
[03:07] <mjg59> If you unplug and replug you'll typically get a different device number
[03:08] <psusi> mjg59, why have two representations?  how are you supposed to find the sysfs node for a given device, like the keyboard, so you can enable it to wake the system?
[03:09] <mjg59> If you're looking for the keyboard then start from the keyboard, not from USB
[03:10] <psusi> and if you get a different device number, wouldn't it move to a new sysfs node?
[03:10] <mjg59> No, because the toplogy is the same
[03:11] <mjg59> Walk /sys/class/input, find keyboard, go to the USB node from there
[03:11] <psusi> I'm confused... device number = bus number and an ordinal assigned as devices are detected, no?
[03:11] <mjg59> Yes, which is completely unimportant when it comes to the bus topology
[03:11] <mjg59> A physical port always has the same number
[03:12] <mjg59> The only time you actually care about that port number is when you're using a USB debug cable, because only one port per controller can drive one
[03:12] <mjg59> Otherwise, the relationship is uninteresting
[03:12] <psusi> ohh... wait... so lsusb gives an ordinal assigned in order of detection relative to a bus, but /sys/bus/usb/devices lists them by hub-port?
[03:12] <mjg59> Oh, I guess that's not strictly true. You might care for multihead as well.
[03:12] <mjg59> Yes
[03:13] <mjg59> I've no idea why lsusb behaves the way it does. It's really not useful.
[03:13] <psusi> I was just going to say that
[03:13] <mjg59> But thankfully it's generally not something that you have to care about, because you're more interested in the device and not its usb path
[03:14] <mjg59> (And you can get from device -> USB)
[03:14] <psusi> there isn't a good way ( i.e. user friendly command or gui utility ) to enable devices to wake is there?  I was going to explain to this guy how to go flip on the wakeup setting in the sysfs node then ran into trouble with locating the usb stuff since I don't see anything in class/input other than events for buttons
[03:15] <mjg59> No, there's nothing at present
[03:15] <psusi> didn't think so...
[03:16] <mjg59> If you know it's usb then you can go /sys/bus/usb/drivers/usbhid/whatever/lala/power/wake
[03:17] <psusi> yea... was just doing that... but got 3 usbhid devices now... remind me to kick lsusb for using a useless nomenclature
[05:40] <cshong> I would like to ask questions about bug 662288 ( https://bugs.launchpad.net/ubuntu/+source/linux/+bug/662288 ). What does it mean by "fix committed"?
[05:40] <ubot2> Launchpad bug 662288 in linux "rt3090: freeze on module rt2800pci unload" [Undecided,Fix committed]
[05:40] <ubot2> Launchpad bug 662288 in linux "rt3090: freeze on module rt2800pci unload" [Undecided,Fix committed] https://launchpad.net/bugs/662288
[05:41] <cshong> I would like to ask questions about bug 662288
[05:41] <ubot2> Launchpad bug 662288 in linux "rt3090: freeze on module rt2800pci unload" [Undecided,Fix committed] https://launchpad.net/bugs/662288
[05:42] <cshong> What does it mean by "fix committed" in a bug?
[08:30] <cking> apw, have all the kernel blueprints been renamed again?
[08:46] <jjohansen> cking: I can still accesses them through the wiki
[08:46] <jjohansen> cking: https://launchpad.net/ubuntu/+spec/other-kernel-o-misc
[08:46] <jjohansen> not sure if that has been renamed or not
[08:46] <cking> jjohansen, thanks - they now seem to have an other- prefix
[08:47] <jjohansen> ah
[08:47]  * cking tweaks his URLs
[08:47] <cking> this happens every time, I sort out a list and then the names change
[08:51] <cking> grrr, "Hide talks that aren't for me" *still* does not work
[09:09] <jjohansen> cking: has it ever worked?  I know every time I have tried it at uds something has gone wrong
[09:11] <cking> jjohansen, I'm just hacking a bash script to get the ical data and format it into some plain text that tells me where I should be :-)
[09:11] <jjohansen> hehe, nice
[09:14] <jjohansen> cking: http://arstechnica.com/business/news/2011/05/intel-re-invents-the-microchip.ars
[09:14] <jjohansen> cking: that is a fairly nice explanation of the trigate transitor
[09:15] <cking> I was trying to imagine what they were describing from other articles - so thanks - that does have some useful images to show what's going on
[09:17] <cking> nice explanation of the physics too
[09:28] <smb> morning. Just wondered why it is so peacefully quiet, then remembered I forgot to log in... :-P
[09:29] <ntr0py> Can someone tell me if the linux sata driver for JMicron JMB362 PCIe to SATA bridge chip would support SATA TRIM commands for use with SSD's?
[14:44] <ppisati> why thunderbird keeps reformatting my email? grrr...
[14:45] <soren> ppisati: There's a config option for that.
[14:45] <soren> ppisati: Hang on.
[14:46] <soren> ppisati: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=blob;f=Documentation/email-clients.txt;h=a0b58e29f91171cf8616fd2aaa93cc856c985dd9;hb=HEAD
[14:47] <soren> ppisati: Without those tweaks Thunderbird is dreadful for patches.
[14:49] <ppisati> soren: thanks, i'll take a look
[15:27] <soren> I have a laptop that I installed afresh with Natty. I'm getting stuff like this: [ 3963.082105] EXT4-fs error (device sda5): htree_dirblock_to_tree:586: inode #14162517: block 56631703: comm update-mime-dat: bad entry in directory: rec_len is smaller than minimal - offset=0(4096), inode=4227858432, rec_len=0, name_len=0
[15:29] <soren> Does this sound familiar at all?
[15:30] <smb> does not ring a bell for me at least...
[15:31] <soren> It did exhibit from odd behaviour yesterday. apt crashing rather randomly, and reinstalling apt fixed it. That does suggest some sort of either filesystem, harddrive or memory error.
[15:32] <soren> Erk.. fsck is *not* happy.
[15:32] <smb> Yeah, unfortunately all of it might be possible. John, just had a case of odd behavior that turned out to be his ssd going away
[15:33] <soren> This is a branch new laptop. Old-school hdd.
[15:33] <soren> s/branch/brand/
[15:33] <soren> I actually saw that I had mistyped "brand" as "branch", deleted it, and wrote it again.
[15:33] <smb> heh
[15:34] <smb> at least consistent
[15:34] <soren> \o/
[15:35] <smb> found something?
[15:37]  * ppisati starts packing stuff...
[15:41]  * JFo is so tired of packing
[15:42]  * smb is mostly done with it
[15:42] <JFo> same here, just painful this time. I am bringing a bunch of stuff that I would normally not have to
[15:44] <soren> smb: No, just happy about being consistent.
[15:44] <smb> That can be stressful. Well, somehow it feels a bit like it went too smoothly. Either I forgot half of it, or I just get more relaxed
[15:45] <smb> soren, Ah. :) So I guess I would start with memcheck... (though this had been not so successful with the supermicro boards I tried it recently)
[15:46] <soren> smb: Just finished that. No problems at all.
[15:46] <smb> hmm ok.
[15:46] <soren> smb: To be honest, I didn't let it finish.
[15:47] <smb> Well, wasn't the default to never finish?
[15:47] <soren> smb: ...but it had run a bunch of its tests across all the memory, and the odd behaviour this box was exhibiting was rather consistent, so I sort of assumed it would fail pretty quickly if that was really the problem.
[15:47] <soren> smb: Right, yes, but I didn't let it run its entire test suite. That takes hours.
[15:48] <smb> yeah I know. :)
[15:48] <soren> smb: I'll restart it and let i finish overnight or something.
[15:50] <smb> soren, Yeah, other posiible way would be to use a usb drive for a test installation. That would probably give you a hint whether its somewher rather memory or controller/hdd
[15:51] <soren> smb: Good idea.
[15:52] <soren> smb: Dreadful timing, though. It's my wife's new laptop and I'm heading to UDS on Sunday. I doubt I'll have it production ready by then.
[15:53] <soren> I would *hate* to have to wait for a week before I could get to use my new laptop.
[15:53] <smb> soren, Those things know and resist...
[15:54] <soren> It's taken me years to convince her to upgrade her laptop and now this. Darn it.
[15:54] <soren> Years!
[15:54]  * smb hopes it is not a Toshiba...
[15:55] <smb> Its as bad as saying po...bo... ;-P
[15:56] <soren> It's a Thinkpad. That's how we roll in this house :)
[15:56] <soren> X120e. It's a nice little box.
[16:00]  * sforshee goes to finish packing
[16:37] <JFo> ok, I am off to head to a few errands on the way to the airport. see you guys there.
[16:38] <smb> Safe trip
[16:38] <JFo> same to you smb :-)
[16:38] <smb> Ta :)
[16:54] <ayan> soren: i just ordered an x120e for my wife.  it should be here on the 11th.
[17:11]  * smb wraps up. see who is there at uds
[17:54] <cking> smb, seeya at UDS
[17:55]  * cking considers packing too
[18:19] <Azelphur> Hi, I'm trying to generate info for a bug report on my tablet pc. When I (un)plug the power cable, the kernel panics. I've been following the guide at https://wiki.ubuntu.com/Kernel/Netconsole however when I try to modprobe netconsole, I get "FATAL: Error inserting netconsole (/lib/modules/2.6.38-8-generic/kernel/drivers/net/netconsole.ko): Unknown error 524"
[18:20] <Azelphur> Anyone wanna help me with that so I can get a bug report up? :)
[18:47] <jjohansen> Azelphur: never seen that before, anyhting more in dmesg?
[18:48] <Azelphur> jjohansen: [ 2407.111082] netconsole: eth0 doesn't support polling, aborting.
[18:48] <Azelphur> :(
[18:49] <jjohansen> Azelphur: :( indeed
[18:49] <Azelphur> my wifi is the same
[18:49] <Azelphur> so no usb to ethernet, no wifi, that rules netconsole out I guess?
[18:51] <simonbcn> hi, I've compiled the git kernel 2.6.38-9.43 for Lucid.
[18:51] <simonbcn> Should I upgrade the linux-libc-dev package to ensure conformity with the kernel installed?
[18:52] <simonbcn> In case afirmative, how can I do this?
[18:52] <jjohansen> simonbcn: shouldn't be needed, the kernel tries very hard to keep the abi stable
[18:55] <simonbcn> jjohansen, but this package is directly relationed with the kernel source, no?
[18:55] <simonbcn> and it's used to compile other programs, no?
[18:56] <simonbcn> Does this cause problems when compiling other programs?
[18:56] <simonbcn> (Sorry for my bad english)
[18:56] <jjohansen> simonbcn: not exactly, it pulls from some kernel headers but what it uses is stable
[18:56] <jjohansen> simonbcn: nope, no problems
[18:57] <jjohansen> generally it just doesn't use new features that have been exported by the kernel
[18:57] <simonbcn> ok, then it's right if I have the linux-libc-dev from Ubuntu Lucid (2.6.32) but I use the kernel 2.6.38,
[18:57] <jjohansen> yep
[18:57] <simonbcn> ok, thanks.
[18:57] <simonbcn> Other doubt, this is about compilation parameters 
[18:58] <simonbcn> oh, I forgot to ask: How do you generate the package linux-libc-dev in git kernel?
[19:01] <simonbcn> Another doubt: I compile the kernel with: "AUTOBUILD=1 NOEXTRAS=1 no_dumpfile=true skipabi=true skipmodule=true do_doc_package=false full_build=false do_tools=false fakeroot debian/rules binary-indep"
[19:01] <simonbcn> but it generates the doc and tools package. What is the utility the "do_doc_package" and "do_tools" then?
[19:13] <Azelphur>  jjohansen happen to know any other ways of getting the kernel panic copied out?
[19:14] <jjohansen> Azelphur: usb serial console
[19:14] <jjohansen> maybe crash kernel but I haven't had luck with that
[19:14] <Azelphur> hmm
[20:35] <soren> ayan: It's a very impressive piece of machinery.