[06:55]  * smb -> BOD
[07:32] <ppisati> moin
[07:45]  * ppisati -> reboot
[09:59]  * ppisati -> out
[14:11]  * henrix reboots
[14:20] <Eimann> hmm, I still need to troubleshoot why 3.4.0-030400rc5-generic works fine on my thinkpad x220 but 3.4.4-030404-generic and 3.5 do not
[14:34] <jsalisbury> pgraner, I posed a new kernel to bug 1018020  This should be the final patch
[14:34] <ubot2> Launchpad bug 1018020 in linux "Logitech webcam not working" [High,In progress] https://launchpad.net/bugs/1018020
[14:44] <pgraner> jsalisbury, ok otp will be in about 20 min
[14:45] <jsalisbury> pgraner, thanks.  just wanted to let you know the kernel is available in case you had some video meetings today.
[15:02] <pgraner> jsalisbury, I have them all day
[15:02] <jsalisbury> pgraner, cool, your a great test subject then :-)
[15:12]  * ogasawara back in 20
[15:18] <jokerdino> hey guys. i recently upgraded to quantal from precise and i had graphics issues.
[15:18] <jokerdino> i then had to install the inux-image-extra-3.5.0-3-generic to fix the issues.
[15:19] <jokerdino> i am wondering why i have to manually install it when it seems quite necessary?
[15:28] <smb> jokerdino, Upgrading with either update-manager or do-release-upgrade should pull in that package automatically. Though update-manager had some other issues yesterday (at least). do-release-upgrade did pull in extras though.
[15:31] <jokerdino> smb: thanks for your response. i upgraded earlier and i didn't have it installed through do-release-upgrade
[15:35] <jokerdino> i mean, i upgraded through the command line and it didn't work as you say
[15:36] <smb> jokerdino, If neither update-manager nor do-release-upgrade would work (though those are preferred as they do a bit more than just switch to the new sources), then the best way to ensure the kernel packages are complete is to install a meta package. Like "apt-get install linux-generic". Those should give you header files, the kernel itself and anything else
[15:38] <jokerdino> i understand that. but keeping QA testing in mind, this is not ideal is it?
[15:39] <smb> Well QA testing should upgrade using the supported methods. Though if you say, you used do-release-upgrade and it did not work, that would have been a bug, but it seems at least as of yesterday it was ok
[15:40] <jokerdino> i upgraded on Tuesday IIRC
[15:40] <jokerdino> http://chat.stackexchange.com/transcript/201?m=5297694#5297694 
[15:40] <jokerdino> i am quite sure it wasn't auto installed.
[15:41] <smb> And just to be sure, you say "using the command line" and that was do-release-upgrade?
[15:42] <jokerdino> yes that.
[15:42] <jokerdino> can i look for logs that might be useful for you?
[15:43] <smb> Hm, yeah. you should probably file a bug then as well. 
[15:44] <jokerdino> you do suggest it is not a problem anymore though?
[15:45] <jokerdino> and this answer on Ask Ubuntu suggests installing extra is not being done for quantal? http://askubuntu.com/a/153033/25798
[15:45] <jokerdino> "As far as I know, the above approach has not been taken for the Quantal kernels -- only -virtual is affected as usual. "
[15:45] <rtg> ogra_, whats the normal way to install a new kernel on an ARM board if it can u-boot ? the usual 'dpkg -i linux-image*' doesn't update /boot/uImage
[15:46] <ogra_> rtg, hmm,. that would be a bug 
[15:46] <smb> jokerdino, For quantal the virtual and the generic kernel are the same packages. Only the virtual meta-package does not install extras while the generic one does.
[15:46] <rtg> ogra_, well, its a linaro kernel deb
[15:46] <ogra_> if you used a proper image the bootloader setup should take care
[15:46] <ogra_> oh, linaor
[15:46] <ogra_> well, try to run sudo flash-kernel manually
[15:46] <ogra_> see what happens
[15:47] <rtg> ogra_, that won't wreck the SPI NOR ?
[15:47] <ogra_> might be that linaro (once again) broke the subarch naming convention
[15:47] <jokerdino> smb: alright then. i'll chat with balloons regarding this then.
[15:47] <ppisati> rtg: ogra_ there's today discussion about incompatibilities between our flash-kernel and their kernels
[15:47] <ogra_> ppisati, if they follow the debian and ubuntu naming scheme its fine 
[15:47] <ppisati> rtg: ogra_ Re: Problem with flash-kernel script on PandaBoard with Ubuntu image and Linaro kernel
[15:47] <ppisati> rtg: on ubuntu-devel
[15:47] <ogra_> i.e. the mx5 kernel which does this, just works 
[15:47] <ppisati> ogra_: they actually wanted to patch our flash-kernel to support their stuff
[15:47] <ogra_> ppisati, yes, i saw that
[15:48] <rtg> ogra_, this is the mx6 on a sabrelite board
[15:48] <ogra_> they should patch debians
[15:48] <ogra_> rtg, well, then it most likely isnt even in the flash-kernel database yet 
[15:48] <ogra_> is that quantal ?
[15:48] <rtg> oneiric
[15:48]  * ogra_ prays that it is
[15:48] <ogra_> OMG !
[15:49] <rtg> ogra_, I'm working on a 3.2 kernel to bring it up to precise
[15:49] <ogra_> k, so you need to hack up flash-kernel ... that version had all HW data for all boards hardcoded directly in the flash-kernel script
[15:49] <ogra_> take a look :)
[15:49] <smb> jokerdino, Ok. If the upgrade path still fails to install extras for you, then please file a bug. And in that case the installer logs hopefully help to find out why it did not happen.
[15:49] <ogra_> it will not be an easy task (compared to the quantal flash-kernel) to add all the functions you need
[15:50] <jokerdino> smb: okay. thanks for your help
[15:50] <ogra_> rtg, any chance you could use a quantal userspace ? 
[15:50] <rtg> ogra_, eventually, but I thought I'd get the kernel working first.
[15:50] <rtg> imx_flash_kernel()  ?
[15:51] <ogra_> well, no idea how that board actually boots without having played with it ...
[15:51] <ogra_> imx_flash_kernel is for the mx5 
[15:51] <ogra_> if the mx6 didnt change wrt booting, that should work
[15:51] <ogra_> but you need the match for the "hardware" entry from /proc/cpuinfo
[15:51] <rtg> ogra_, it boots from u-boot by loading /boot/uImage, which looks kind of like the mx5
[15:52] <ogra_> (must be somewhere in the middle of the code)
[15:52] <ogra_> k, sounds good
[15:52] <ogra_> so try to add a line for the cpuinfo output and make it match for the imx_flash_kernel function and you should be good
[15:58] <rtg> ogra_, I'm assuming if the u-boot command is 'ext2load mmc ${disk}:1 10800000 /boot/uImage && bootm 10800000 ;', then I should set IMX_KERNEL_ADDR= 10800000 in flash-kernel ?
[15:59] <ogra_> heh, thats a good question
[16:00] <ogra_> i dont think this directly translates to the address from boot.scr, but i'm not sure
[16:01] <rtg> well, the worst it can is not boot. I've got an image of the original SD card.
[16:01] <rtg> can do*
[16:01] <ogra_> ah
[16:01] <ogra_> original wasnt linaro i suppose ?
[16:01] <ogra_> (else you should be able to look up that address)
[16:02] <ogra_>        "Freescale i.MX6 Quad (Device Tree)")
[16:02] <ogra_>                 check_subarch "mx6"
[16:02] <ogra_>                 IMX_KERNEL_ADDR=0x10008000
[16:02] <ogra_>                 imx_flash_kernel
[16:02] <ogra_> heh
[16:02] <ogra_> my copy of flash-kernel already has an entry for the mx6 here 
[16:02] <ogra_> (flash-kernel-2.28ubuntu42 is the last copy of the old cruft i have around here)
[16:02] <rtg> ogra_, so, where would I find a Quantal image ?
[16:03] <ogra_> no idea, linaro might have one already 
[16:03] <ogra_> oh, quantal
[16:03] <ogra_> sorry
[16:03] <rtg> I haven't been able to find anything newer then oneiric
[16:03] <ogra_> ubuntu-core would be a good start probably
[16:03] <ogra_> but that needs some manual tinkering (networking and a rootpw at least)
[16:04] <rtg> maybe I should hassle Eric Miao who has done all the work on this
[16:04] <ogra_> well, if it boots you should at least see a kernel panic now (not finding a rootfs)
[16:05] <ogra_> if you got that far, just creating a second partition on your SD and unpacking ubuntu-core to it should get you a booting system (no login though without passwd indeed)
[16:09] <rtg> ogra_, bummer, what ever flash-kernel did trashed u-boot. 
[16:09] <ogra_> ignore flash-kernel then, do it manually 
[16:10] <ogra_> do you have a boot.scr from the original SD ?
[16:10] <rtg> ogra_, I'm not even getting the u-boot serial startuop. how do I recover from that ?
[16:10] <ogra_> ugh, no idea
[16:10] <rtg> shit
[16:10] <ogra_> we dont support any boards in ubuntu that dont boot from SD :)
[16:11] <ogra_> i.e. all boards we support atm are unbrickable
[16:11] <ogra_> hard to tell if the mx6 has any option to force boot from SD only 
[16:11] <rtg> well, this does boot from SD, but I think the initial startup is in SPI NOR which then loads u-boot from SD
[16:11] <ogra_> ah, then its fine 
[16:12] <ogra_> does the original SD have two partitions ? 
[16:12] <rtg> not taht I know of, but I didn't look too clase. lemme reimage it.
[16:12] <ogra_> yeah, take a look
[16:13] <ogra_> usually uboot systems have an ext2 or vfat partition as the first part on SD and read their bootloader stuff from there
[16:13] <ogra_> (though indeed not all of them :P )
[16:15] <rtg> re-imaging, could take awhile...
[16:36] <bjf> bug 1023214
[16:36] <ubot2> Launchpad bug 1023214 in linux "bogus utime and stime in /proc/<PID/stat" [High,Confirmed] https://launchpad.net/bugs/1023214
[17:00]  * apw notes there is a s/r fix coming in v3.5-rc7
[17:00] <apw> commit dc332fdf9f373a87b1e2f423b5b004b2a3c37e1a
[17:00] <apw> Author: Jonathan Nieder <jrnieder@gmail.com>
[17:00] <apw> Date:   Sun Jul 8 21:55:14 2012 +0200
[17:00] <apw>     ACPI / PM: Leave Bus Master Arbitration enabled for suspend/resume
[17:07] <smb> bjf, jsalisbury I pushed git trees (which you can look at for reference) to lucid-* directories on gomeisa and started builds. This basically reverts one (I believe rather SAUCE though it does not say so) patch and replaces it with two upstream changes. We always were thinking of backporting but since it is so hard to verify never actually did.
[17:07] <bjf> smb, thanks
[17:12] <smb> bjf, jsalisbury Just assuming you will be longer around today than me to have the bug updated. Btw, I looked at the Maverick repo again and that already seems to carry the other two and not the one reverted. So it seem it was about that time it got changed. I believe there also has been something else concerning load values but much later...
[17:13] <bjf> smb, ack
[17:39]  * rtg -> lunch
[17:50] <jsalisbury> smb, ack
[18:01] <ogasawara> jsalisbury: for your patches for 1018020, following the alsa-devel thread it sounds like the first patch isn't needed.  Did you want to re-test really quick with only the second patch before I apply.
[18:02] <jsalisbury> ogasawara, sure, sounds like a good idea.  I can test here, and post a new kernel to the bug.
[18:02] <ogasawara> jsalisbury: ack
[18:11]  * smb -> EOD
[18:30] <pgraner> jsalisbury, all is good with that kernel on two boxes been in use for a few hours and still chugging
[18:31] <cnd> arges: I just replied on https://bugs.launchpad.net/ubuntu-advantage/+bug/1015824
[18:31] <ubot2> cnd: Error: <Bugtracker.plugin.Launchpad instance at 0xa2d980c> bug 1015824 not found
[18:31] <cnd> will you be able to follow up as I suggested?
[18:31] <jsalisbury> pgraner, great.  I'm actually posting one more test kernel.  It shouldn't have any issues, just some code removed that is not needed.
[18:32] <pgraner> jsalisbury, cool, I'll watch the bug and test it once it comes out
[18:33] <jsalisbury> pgraner, I'm uploading them now.  Should be here in a minute: http://people.canonical.com/~jsalisbury/lp1018020/
[18:33]  * ogasawara lunch
[18:34] <jsalisbury> pgraner, they are there now.  The have 'v3' in the .deb file name.
[18:52] <jsalisbury> pgraner, when you test the latest v3 kernel, can you see if there is a delay at the login screen?  
[20:05]  * rtg -> EOD
[20:06]  * cking --> EOD
[23:52] <arges> cnd, yes
[23:52] <cnd> arges: thanks!
[23:52] <arges> : )