[00:02] <arand> rebooting now, will be back here.
[00:23] <Dink> Hello, I have been trying to use launchpad to only build a specific flavour kernel ie, netbook. I am now getting "make: *** No rule to make target `build-netbook', needed by `build-arch'.  Stop." in my build log. Any help would be appreciated. I have tried to remove all non configs etc for x86_64 as well as other flavours for x86. Did I remove too much ?
[00:24] <Dink> Sorry for my ignorance I have never done this as a .deb before.
[00:47] <kamalm> achiang, cnd: hey, I figured out my problem with my LKML post never showing up!  apparently, the vger.kernel.org mailing lists silently reject any message without a Message-Id: header (my message file, the edited output from git-format-patch, had no Message-Id: header). ...
[00:49] <kamalm> I was feeding my message file directly to msmtp, but I eventually guessed that might be the problem (after noticing that git-send-email *does* generate a Message-Id) so I just resubmitted it using git-send-email directly and it appeared on LKML immediately (yay).
[00:50] <kamalm> I still don't like git-send-email much, but I'll either use it anyway, or else come up with a way to generate a Message-Id: for msmtp-direct use.
[01:18] <achiang> kamalm: stg mail
[01:18] <achiang> :)
[01:19] <kamalm> achiang: stg?
[01:19] <achiang> stacked git
[01:19] <achiang> it's quite nice
[01:20] <kamalm> I guess I don't see why I need to fiddle with git at all just to send a piece of email ;-)   -- I know git-send-email (etc.) can do more fancy stuff than that, but I (foolishly) wanted to do it manually to see how it worked (or didn't as the case may be).  that'll teach me.
[01:21] <achiang> kamalm: oh, i use stacked git to manage patch series on top of git
[01:21] <achiang> kinda like quilt, but you know, not stupid like. :)
[01:22] <kamalm> ;-)  my patch was sooo simple, I thought, surely I don't need a *tool* to send the email for me.   I'm not even ready to think about stacked patches yet.
[02:01] <johanbr> has anyone else noticed broken sound on some HP laptops with the lucid -21 kernel (32 bits)?
[02:01] <johanbr> -19 works fine
[07:30] <psurbhi> good morning #ubuntu-kernel
[07:36] <jjohansen> morning psurbhi
[08:35]  * psurbhi devel machines scsi controller is almost dead :( got back to my old slow laptop now
[10:32] <tgardner> apw, rsync -zhhP rsync://cdimage.ubuntu.com/cdimage/daily-live/current/${RELEASE}-desktop-${i}.iso ${RELEASE}-desktop-${i}-daily.iso
[10:44] <tseliot> apw: I've compiled a kernel from the Linus git branch but I'm getting kernel panic VFS: unable to mount root fs on unknown-block(0,0). Any ideas on what this could be? (support for ext4 is built into the kernel)
[10:44] <apw> unable to mount the initramfs perhaps ?
[10:45] <tseliot> ah
[10:48] <tseliot> apw: this also happen with a mainline build of 2.6.34-rc5 here
[10:48] <smb> tseliot, I had this sometimes when something failed to do update-grub correctly...
[10:49] <tseliot> smb: it could be that. As the same kernel on the same machine (different installation of Lucid) works
[11:34] <elmo> seriously, people
[11:34] <elmo> a hardy kernel release TODAY?
[11:36] <smb> elmo, That can only be one moved from proposed. I would not have expected this to happen exactly today. But that is not really in our hands
[11:45] <tseliot> smb, apw: it looks like it didn't generate any initrd file even though I built it with --initrd
[11:46] <tseliot> and CONFIG_BLK_DEV_INITRD=y while CONFIG_INITRAMFS_SOURCE isn't set to anything
[12:03] <fabbione> hmmm guys who is in charge of karmic kernel and all the CPU speed step stuff?
[12:03] <fabbione> (like setting ondemand vs performance cpu bits)
[12:22] <AnAnt> Hello, why is there a rt3070sta.ko in linux-rt but not in linux ?
[12:45] <AnAnt> ah, linux has rt3090sta
[13:00] <smb> fabbione, Wassup?
[13:00] <fabbione> smb: i am trying to figure out why setting ondemand cpu governor on karmic -> filesystem corruption (a bad one)
[13:01] <fabbione> smb: and if there is somekind of blacklist table to make sure ondemand is never EVER used on a similar system
[13:01] <smb> fabbione, repeatable?
[13:01] <fabbione> smb: yes absolutely. it's 10 days I am debugging this thing
[13:02] <fabbione> smb: started from swapping one disk at a time in the raid(s), cables, controllers, etc. etc.
[13:02] <fabbione> you know.. usual drill trying to understand what the hell is wrong
[13:02] <fabbione> till i figured that in init 1 i didn't see the corruption
[13:02] <fabbione> init 2 did
[13:03] <fabbione> different kernels.. etc.
[13:03] <fabbione> smb: note that ondemand seems the default in karmic (at least it's been pulled in when updating from jaunty)
[13:04] <fabbione> not sure how many people are affected potentially by this problem, but it's not easy to spot at first
[13:04] <smb> fabbione, It would be the first time I heard this. Yes it is. It only performance during a small time during init when I remember that correctly
[13:05] <smb> Is that a special raid setup of just normal md raid?
[13:05] <fabbione> yes right.. performance at init then at the end of the boot goes to ondemand
[13:05] <fabbione> smb: 2 md raid -> different controllers -> different disks -> same behaviour (one shows it faster tho)
[13:05] <fabbione> how to reproduce:
[13:05] <fabbione> get a folder of random file that's at least twice the size of system RAM (so it can't be cached and you need to I/O)
[13:06] <fabbione> for i in $(seq 1 100); do md5sum /folder; done
[13:06] <fabbione> depending how lucky you are it will eventually fail
[13:06] <fabbione> at somepoint I was reading the same 7GB iso file, and I was getting a different md5sum each run
[13:06] <fabbione> the data on disk look ok
[13:06] <soren> Which filesystem?
[13:06] <fabbione> but then.. copy/backup/burn to dvd
[13:07] <fabbione> all bad
[13:07] <fabbione> soren: ext3
[13:07] <fabbione> and removing that writeback option youdid turn on between 2.6.31-16 and -17 doesn't change a thing
[13:07] <fabbione> soren: you are welcome to pass by and look at it yourself
[13:07] <fabbione> :)
[13:07] <smb> fabbione, Well that writeback only matters on crash
[13:08] <soren> fabbione: Heh.. Don't wait up :)
[13:08] <fabbione> smb: right, but i wanted to exclude every possibility
[13:08] <smb> fabbione, sure
[13:08] <fabbione> i even started modprobing one module at a time in case it was a conflict of somekind
[13:08] <smb> Especially with something that looks rather odd
[13:08] <fabbione> exactly
[13:08] <fabbione> the raids are:
[13:08] <fabbione> raid5 -> SATA -> 4x1.5TB 
[13:09] <fabbione> raid1 -> IDE -> 2xsomethingGB
[13:09] <fabbione> so different controllers, technologies and disks
[13:09] <fabbione> all show the same issue
[13:09] <soren> If you've got tons of space somewhere, you could try copying the data somewhere insteaed of just md5sum'ing it.
[13:09] <soren> That way, you could see what the actual difference is.
[13:09] <fabbione> i come to the conclusion that's something to do with either a buggy CPU speedstep
[13:10] <soren> ...and see if it's random data, data leaked from somewhere else, or just the correct data in the wrong order or whatnot.
[13:10] <fabbione> soren: i did that test too. even if you copy and make a difference you obtain nothing useful. it's always a random different data that gets corrupted
[13:10] <fabbione> and yes I did run memtest for > 24h
[13:10] <fabbione> soren: that would require also performing a kdump and compare.
[13:11] <fabbione> soren: not enough processing power for that
[13:11] <smb> fabbione, what type of cpu? amd or intel?
[13:11] <fabbione> smb: amd
[13:11] <soren> Either that or copy it to a remote place.
[13:11] <soren> Oh..
[13:11] <fabbione> vendor_id       : AuthenticAMD
[13:11] <fabbione> cpu family      : 15
[13:11] <fabbione> model           : 47
[13:11] <fabbione> model name      : AMD Athlon(tm) 64 Processor 3000+
[13:11] <fabbione> stepping        : 0
[13:11] <fabbione> cpu MHz         : 1800.000
[13:11] <soren> right, I see what you mean.
[13:11] <fabbione> cache size      : 512 KB
[13:11] <soren> Well, you might be able to make sense of it just looking at a hexdump of it.
[13:11] <fabbione> soren: it's doable, but it means comparing TB of data at each run...
[13:12] <soren> fabbione: Jesus, how much RAM do you have?!?
[13:12] <fabbione> soren: i don't really care how it gets corrupted.. it just shouldn't happen in the first place
[13:12] <soren> fabbione: Of course.
[13:12] <smb> fabbione, Hm, I just heard today that there might be a bug that might make it not scale up correctly (but nothing about corruption)
[13:12] <fabbione> soren: it's 3GB of RAM, but then you need several copies of a 2xRAMsize file etc.
[13:13] <soren> fabbione: I'm just saying that the contents might give you a hint about where it came from 
[13:13] <fabbione> smb: the corruption could be caused by more than one bug...
[13:13] <fabbione> soren: it's probably the same data swapped or something.
[13:13] <fabbione> soren: i doubt there is a leak because everything that is loaded before switching ondemand (application and OS) are deadly stable
[13:14] <fabbione> i suspect simple byteswaps or something
[13:14] <fabbione> smb: i suspect the CPU can even scale correctly, but other bits on the mobo don't.. 
[13:15] <fabbione> anyway... now you know
[13:15] <fabbione> keep an eye
[13:15] <fabbione> smb: another thing... dvb related to karmic
[13:15] <fabbione> http://fabbione.fedorapeople.org/dvb.patch
[13:15] <fabbione> you want to revert that patch from 2.6.31
[13:15] <fabbione> i think it's already removed in dvb-trunk and later linus' trees
[13:16] <fabbione> (not sure about the git sha1 tho)
[13:16] <smb> fabbione, Yeah, thanks. I got a similar system which I might use to test. 
[13:16] <fabbione> the patch as you see it went in at some point, but it breaks the HVR-3000 dvb-t
[13:16] <smb> fabbione, I guess I can find that out
[13:16] <fabbione> i already spoke to Darron and he agreed to revert it
[13:16] <fabbione> (that was a few months back)
[13:17] <fabbione> but yeah you get the idea
[13:17] <fabbione> once that patch is in karmic i can finally stop building my own kernelks
[13:17] <fabbione> getting tired of gcc :P
[13:19] <smb> fabbione, Well, to get it into Karmic, there might be a slightly harder need to it that tiring you. Especially when Lucid is released.
[13:19] <smb> fabbione, Don't we do that as a module? /me cannot remember that
[13:20] <fabbione> smb: yeah it's a module, but i am pretty sure i am not the only HVR-3000 user out there
[13:20] <fabbione> and doesn't have to go in yesterday.. just when it fits in the next update
[13:20] <fabbione> lucid.. that's going to take me at least 4/6 months before I upgrade those systems
[13:21] <cmug> Is it out yet
[13:21] <fabbione> i'd guess shortly
[13:21] <smb> fabbione, Likely not. Just that the SRU process is not really allowing for less critical problems. However annoying they are. You probably know. ;-)
[13:22] <fabbione> smb: you can consider that a regression from Jaunty if that pleases you :)
[13:22] <fabbione> in Jaunty it used to work
[13:23] <fabbione> and yeah I can imagine that people have been tighting the SRU process..
[13:23] <smb> fabbione, It would likely be more pleasing. But not if it would cause different regressions. 
[13:23] <fabbione> smb: it doesn't.
[13:23] <fabbione>     This update also fixes analogue tuning on the hvr-1300 when in blackbird mode.
[13:23] <fabbione> analog doesn't exists anymore
[13:23] <fabbione> it wasn't working before, doesn't work now
[13:24] <smb> Would sound good then
[13:24] <fabbione> hvr4000 works with or without the patch
[13:24] <fabbione> Darron has a 4k and a 1300
[13:24] <smb> fabbione, Would you happen to have a lp bug filed or you rather want to leave the fun part to me?
[13:24] <fabbione> fun part is all your :)
[13:24] <fabbione> feel free to copy/paste from IRC
[13:25] <fabbione> i already have to deal with bugzilla
[13:25] <fabbione> that's plenty for me :P
[13:25]  * smb thanks fabbione so much. :-P
[13:25] <fabbione> want to swap for sometime?
[13:25] <fabbione> you work with my BZ account
[13:25] <fabbione> and i work with your LP account
[13:25] <fabbione> ;)
[13:26] <smb> heh. :) Not sure I'll enjoy that
[13:26] <fabbione> at least I have a bit of kernel experience (even if minimal).. but i'd love to see you banging on cluster stacks
[13:26]  * fabbione feels a bit sadist today
[13:26] <fabbione> s/a bit/extremely/
[13:27] <fabbione> picture this.. you debug an issue in the kernel.. that's local to one machine
[13:27] <fabbione> it's there, right in front of you
[13:27] <cmug> fabbione, define shortly :)
[13:27] <fabbione> with cluster, it scales across N nodes that are part of the clsuter
[13:27] <fabbione> because you don't know what's going to trigger what (so to speak)
[13:31] <apw> https://edge.launchpad.net/ubuntu/+source/linux
[13:36] <fabbione> cmug: shortly?
[13:39] <cmug> :)
[15:10] <JFo> this bug looks interesting: bug 568782
[15:10] <ubot3> Malone bug 568782 in linux "cannot install two linux-backports-modules-wireless packages" [Undecided,Incomplete] https://launchpad.net/bugs/568782
[15:12] <BenC> crimsun: ping
[15:15] <tgardner> JFo, having a look
[15:16] <JFo> tgardner, ta
[15:28] <fabbione> hey BenC 
[15:28] <BenC> fabbione: hey bro, how's things?
[15:28] <fabbione> BenC: not too bad and you?
[15:59] <JFo> apw, you around?
[16:00] <apw> hi
[16:00] <JFo> this looks rather serious: bug 559888
[16:00] <ubot3> Malone bug 559888 in linux "Lucid installation bug: blank screen" [Undecided,New] https://launchpad.net/bugs/559888
[16:00] <JFo> not sure if it is us or X or some other
[16:00] <apw> does nomodeset sort it out?
[16:00] <JFo> bryceh, care to look ^^ as well?
[16:00] <JFo> apw, good question, I don't believe they tried that
[16:03] <JFo> apw, looks to be happening during install
[16:03] <JFo> is it possible to add the nomodeset during boot from CD?
[16:04] <manjo> apw, what hr is the release or is it already announced ?
[16:05] <amitk> still testing...
[16:06] <manjo> amitk, dual boot installer bug ? 
[16:06] <amitk> yup
[16:11]  * bjf has to run my son to school
[16:29] <JFo> hey internalkernel 
[16:30] <internalkernel> hey hey... been awhile since I signed in...
[16:30] <internalkernel> JFo: how's life?
[16:30] <JFo> not bad
[16:30] <JFo> tons of bugs :)
[16:44] <JFo> morning kamalm 
[16:45] <kamalm> JFo: good morning
[16:46] <kamalm> is there a "release party" channel of any sort?
[16:46] <azop> kamalm: #ubuntu-release-party
[16:46] <kamalm> azop: thanks
[16:47] <azop> kamalm: if you enjoy spam :P
[16:47] <kamalm> azop: wow, you're not kidding, that's a firehose!
[16:47] <azop> yap
[16:47] <azop> with 'is it our yet?!' or 'it's broke!'
[16:54] <psurbhi> its fun to read the messages on the #ubuntu-release-party
[16:57] <JFo> <-lunch
[17:43]  * apw cracks a cold one
[17:45] <vanhoof> wow #ubuntu-release is wild!
[17:46] <vanhoof> -party .. ;)
[17:46] <azop> it's somethin'
[17:48] <akgraner> vanhoof, you didn't know what all you were missing did ya :-P
[17:49] <vanhoof> haha :)
[17:59] <bryceh> hi JFo, looking
[18:02] <manjo> akgraner, guess you forgot about me? 
[18:03] <akgraner> no
[18:03] <manjo> :)
[18:03] <akgraner> open week just just got finish
[18:03] <bryceh> JFo, not enough information to say
[18:03] <akgraner> I have your stuff in front of me now :-P
[18:03] <akgraner> just got finished rather
[18:03] <akgraner> typing fail today :-(
[18:04] <bryceh> JFo, generally, if they're using a KMS driver then probably should be handled as a kms drm bug
[18:30] <vanhoof> lucid is live :)
[18:30] <vanhoof> https://lists.ubuntu.com/archives/ubuntu-announce/2010-April/000133.html
[18:33] <manjo> apw, congrats on the great job on the kernel 
[18:51] <akgraner> Thank you all for all your hard work!  You all rock!!
[18:52]  * akgraner is <3'ing Lucid :-)
[19:13] <apw> manjo, yo
[19:13] <akgraner> manjo, take a look
[19:13] <manjo> apw, congrats dude
[19:14] <apw> thanks ... see ya soon its release party time
[19:15] <manjo> apw, cool have fun
[19:15] <manjo> akgraner, can you add some testimonial for me :P
[19:16] <akgraner> yep
[19:16] <akgraner> did you read my comments  - 
[19:17] <manjo> :) cool! will buy you a beer next time we meet! yes the comments looks great...  much better than what I had 
[19:17] <akgraner> manjo, you will also need to look at what I commented out - I made a few suggestions, but I don't have that information so you'll have to add it
[19:17] <manjo> akgraner, they announce membership meetings adhoc-ly .. I was on vacation last time they had meetings
[19:19] <akgraner> manjo, keep checking the wiki  - right now regional membership boards just accepted nominations - I would guess, but don't know, there will be a meeting sometime after the new membership board members are a announced