[08:22] <alkisg> Hi, there's a 5-10 second delay in the kernel in Trusty that wasn't there in previous versions, as shown in  `dmesg`, e.g.:
[08:22] <alkisg> [    3.740100] Switched to clocksource tsc
[08:22] <alkisg> [   14.208118] PM: Hibernation image not present or could not be loaded.
[08:22] <alkisg> [   14.208885] Freeing unused kernel memory: 864K (c19ac000 - c1a84000)
[08:22] <alkisg> The exact messages there don't matter, the delay happens between different messages on each boot or on different machines... any idea why the delay exists, or how to avoid it?
[08:23] <alkisg> I can reproduce it even with: sudo kvm -m 768 -cdrom trusty-desktop-i386.iso
[08:27] <apw> alkisg, is that is 3.12 kernels?
[08:28] <alkisg> apw: yes
[08:28] <alkisg> apw: a person "reported" it at http://askubuntu.com/questions/371777/dmesg-mounting-takes-longer-in-lubuntu-than-in-debian for previous kernels too, but I haven't seen that, I've been using Precise till now
[08:30] <apw> alkisg, so the first thing would be to test a 3.13 kernel as those will be replacing the kernel you have fairly soon
[08:31] <apw> 64bit: https://launchpad.net/~canonical-kernel-team/+archive/ppa/+build/5311593
[08:31] <apw> 32bit: https://launchpad.net/~canonical-kernel-team/+archive/ppa/+build/5311596
[08:33] <apw> alkisg, the other thing one could try is taking a boot trace, adding initcall_debug on the kernel command line and then the result is in dmesg
[08:35] <alkisg> apw: thanks, I'll wait until it comes to trusty via -updates and then ask again here, but if the problem started with 3.8, it might be there for 3.13 too...
[08:38] <apw> alkisg, it might, but your statement was it was new in trusty and not in precise, so you at least odnt know if your problem is the same or when it appears in fact.  you should file your own bug so we can try and figure out if it is the same or not
[08:47] <alkisg> apw: thanks, I'll do so when 3.13 arrives in trusty
[08:48] <alkisg> It should need only 1 minute to reproduce it with the kvm command though :)
[08:48] <alkisg> (or, in any ubuntu system...)
[08:48] <alkisg> If it's not something that affects all ubuntu users, then it's not significant, I don't care about my particular installation...
[08:49] <infinity> I see a similar "pause" in my ringbuffer here on saucy's 3.11
[08:49] <infinity> Can't say what's going on there, but you're not alone.
[08:50] <infinity> There's probably a sleep(5) someone accidentally left in there for debugging. :P
[08:50] <infinity> (And man, if that were true...)
[08:51] <apw> heh ... alkisg just file a bug, then i have something to hand to someone to look at
[08:56] <alkisg> apw, infinity: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1259861
[08:56] <ubot2> Launchpad bug 1259861 in linux (Ubuntu) "5-10 second delay in kernel boot" [Undecided,New]
[10:27] <zequence-vacatio> infinity: I don't have access to my gpg key until Saturday, So, can't do any updating until then, I'm afraid.
[17:41] <xnox> apw: around? do you think you can upload goldfish patch from bug #1260023 ?
[17:41] <ubot2> Launchpad bug 1260023 in linux-goldfish (Ubuntu) "enable CONFIG_UNPUT_UINPUT" [Critical,In progress] https://launchpad.net/bugs/1260023
[17:43] <xnox> ogasawara: ^ or maybe you can / know who is in current working hours timezone
[17:58] <apw> xnox, i can do that indeed
[17:59] <xnox> apw: would be much appreciated. I'll be back later to respin android package.
[18:41] <ripthejacker> Hi everyone I am trying to build a kernel and I'm getting this error during build, "install: cannot change owner and permissions of ‘/home/ripthejacker/builds/linux-3.11.0/debian/linux-image-3.11.7-11-12-13/etc/kernel/postinst.d’: Operation not permitted
[18:41] <ripthejacker> What can be the problem?
[18:43] <apw> xnox, fire in the hole
[18:44] <apw> ripthejacker, not running as root / using fakeroot ?
[18:44] <ripthejacker> so fakeroot make-kpkp?
[18:45] <ripthejacker> apw: fakeroot make-kpkg
[18:45] <ripthejacker> ?
[18:45] <apw> make-kpkg --rootcmd fakeroot
[18:45] <apw> maybe
[18:46] <ripthejacker> apw: ok thanks, will try that
[18:47] <rsalveti> xnox: did you test to see if enabling uinput is already enough for autopilot?
[18:48] <apw> rsalveti, the bug says he did
[18:48] <apw> (though that might not make it true)
[18:48] <rsalveti> right
[18:48] <xnox> rsalveti: /dev/uinput device is created. I haven't tested that autopilot itself runs uses /dev/uinput correctly.
[18:48] <rsalveti> right, that's what I wanted to know
[18:48] <xnox> rsalveti: i presume that /dev/uinput would be the same as elsewhere.
[18:49] <rsalveti> right, but would be good to confirm as the emulator might be special
[18:49] <rsalveti> as you probably have the kernel around, just try to run any autopilot test
[18:50] <xnox> rsalveti: i was going to ping you / sergious to test it once the kernel is build. Or i can scp vmlinuz to people.c.c, such that you can just drop it into out/target/product/generic/ubuntu/kernel/vmlinuz to try out.
[18:51] <xnox> rsalveti: sergiusens: http://people.canonical.com/~xnox/vmlinuz-uinput
[18:51] <rsalveti> I need to be off in a few, sergiusens, can you give that a try?
[18:51] <sergiusens> xnox, rsalveti sure, I'll try
[18:51] <rsalveti> sergiusens: awesome, thanks
[18:51] <xnox> sergiusens: thanks.
[18:52] <sergiusens> xnox, give me 15'
[19:50] <xnox> rsalveti: sergiusens: kernel published, uploaded no-change rebuild of android. I'm afk for now, if there are further changes required will debug tomorrow. EOD.
[19:50] <xnox> apw: thanks a lot for the upload! =)
[21:56] <sergiusens> xnox, I couldn't really download your kernel to test
[21:56] <sergiusens> permission denied on server
[21:57] <xnox> darn.
[21:58] <xnox> sergiusens: just apt-get install android-emulator, it's rebuild and published now.
[21:58] <xnox> sergiusens: i wonder if you could have scp'ed if off p.c.c.
[21:58] <xnox> (and thus has the updated kernel built natively)
[21:59] <sergiusens> xnox, -rw------- 1 xnox warthogs 3108016 Dec 11 18:51 vmlinuz-uinput
[21:59] <xnox> sergiusens: corrected now.
[22:00] <xnox> sergiusens: i'm confused how it gets such restrictive permissions by default. well never mind.
[22:00] <sergiusens> xnox, do you have a custom umask?
[22:00] <xnox> i don't believe i do.