/srv/irclogs.ubuntu.com/2014/02/28/#ubuntu-kernel.txt

=== fish__ is now known as fish_
=== yofel_ is now known as yofel
=== henrix_ is now known as henrix
rsalvetiapw: let me know once the meta packages are available in the ppa, so we can try to land them today in the archive06:59
apwrsalveti, sorry been having fallout from any upgrade this morning, will be soon09:19
apwrsalveti, ok they are all building or pending publication now09:44
snadgewhats the easiest way to try out the latest kernel/radeon driver in saucy? .. oibaf? edgers?11:41
apwsnadge, i would guess edgers, but #ubuntu-x might have better info11:49
jdstrandso, I've noticed for a little while that if I have VMs running and then I suspend my machine, when I come out of suspend, sometimes they are unresponsive and qemu is pegging the cpu13:26
jdstrandso I decided to look in kern.log this morning, and see: http://paste.ubuntu.com/7010323/13:26
jdstrandwhat's interesting to me is that I suspending at Feb 27 21:36:08 and resumed at Feb 28 07:16:0513:27
jdstrandI don't know how the ordering is supposed to all go, but it seems like all the 'kvm: disabling virtualization on CPU1' is happening way too late (as in, after I resume)13:28
jdstrandmaybe that is just when it hits syslog13:28
apwthe date on the left is a lie, that is "stopped" when syslog is frozed in the freezeing ling13:29
* jdstrand tries to remember the format of the kernel timestamp13:29
apwseconds and microseconds since boot13:29
apwor nanoseconds or something13:29
apwFeb 28 07:16:05 localhost kernel: [260761.679027] ACPI: Low-level resume complete13:30
apwi think that is the first thing it says on the way up, everything before that is before it went down, but cking is more is an expert on those msgs13:30
apwjdstrand, are these 64 or 32 bit guests ?13:31
jdstrandit was a 64 bit bit guest last night13:31
apwnot smb's bug then.  well, i think its time to file a bug, and we'll get smb et al too poke at it13:31
ckingapw, yep the timestamp is a lie at that poinmt13:32
jdstrandI've seen it with more than this guest, but I'm not sure if I've only ever seen it with 64 bit13:32
jdstrandok, that makes sense then13:32
jdstrandI don't see anything particularly wrong from a kernel perspective looking at the kern.log if I consider 'ACPI: Low-level resume complete' as the place to split13:34
jdstrandexcept perhaps the CPU0 is never actually listed as having kvm disabled or enabled13:36
jdstrandbut I'm not convinced that is wrong13:36
jdstrandok thanks. I'll be testing hallyn's new qemu at some point soon, so I'll just put this on the backburner13:37
=== arges_ is now known as arges
rsalvetiapw: thanks!14:04
apwjdstrand, ok if its there, bug us up please14:05
apwrsalveti, let me know if they are any good14:05
rsalvetisure14:05
_bjfarges, bug 127023714:06
ubot2`Launchpad bug 1270237 in linux (Ubuntu Quantal) "prevent the conntrack table from filling up in the kernel" [Medium,Fix committed] https://launchpad.net/bugs/127023714:06
=== _bjf is now known as bjf
bjfarges, you think we can set this to verified?14:06
argesbjf: so far the user that's tested it hasn't hit the problem, but its one of those 'run for a long time and try not to hit it' problems14:07
bjfarges, right. but we have to decide if we are going to revert or not at some point and now would be a good point :-)14:08
lamontwhy does trusty's grub-probe say: grub-probe: error: unknown filesystem.14:09
apwlamont, any idea which one it is whining about ?14:10
lamontthere are several options14:12
lamontwhich is to say, no.14:12
lamonttrusty upgrade threw a fit when memtest86+ and the kernel postinsts failed as a result.14:12
lamontand then the reboot led to a kernel panic (no root fs, iirc - didn't look much), so I'm running trust on a 3.11 kernel atm14:13
lamontapw: I don't mind the unknown filesystem as much as I mind the exit(1)14:13
* lamont gives the reboot the ol' college try14:20
lamontapw: hints on where I might add a --verbose to have grub-probe be more forthright?14:21
* lamont throws strace -ff at the problem14:26
lamontapw: it's complaining about /dev/mapper/sda5_crypt ...14:28
lamontwhich, yes, it shouldn't like so much14:28
lamontbecause it's an lvm pv14:30
apwhurumph, i suppose this is related to the grub2 update14:52
apwcjwatson, ^^ have yu seen any reports of grub getting whiney about encrypted drives ?14:53
* lamont really tries the reboot thing. brb hopefully14:54
lamontapw: after manually moving grub.cfg.new -> grub.cfg, life is happier and I'm running a trusty kernel15:01
=== vmesons is now known as vmeson
apwlamont, well that is something at least15:23
* lamont generates a strace -ff update-grub stream15:27
lamontapw: who should I be pestering to look at chinstrap:~lamont/update.strace?  (is grub2 kernel, or foundation?)15:33
ograapw, bug 128618415:35
ubot2`Launchpad bug 1286184 in linux-flo (Ubuntu) "udev ACLs for sound devices can not be set" [Undecided,New] https://launchpad.net/bugs/128618415:35
ograapw, seems it doesnt happen on mako 15:35
ograchecking manta now 15:35
ograbut it smells like this is flo specific 15:35
ograyup it is ... verified on manta 15:38
apwogra, ack15:55
apwrsalveti, do you want to copy out the current flo before i upload, or shall i replace it15:55
rsalvetiapw: just replace it, I'm publishing the other packages in a CI silo (as we need to rebuild the android package as well)15:56
rsalvetiso don't worry, I'll get  them in the archive via a CI silo15:56
apwrsalveti, so yo have taken them "elsewhere" ok16:00
rsalvetiapw: yeah16:00
ograyeah, so we are able to fish them out of the conatiner at least16:01
ograstill better than having to download the package 16:01
rsalvetiyeah16:04
rsalvetiapw: just ping me once the new flo kernel is available in the ppa16:04
apwrsalveti, ack16:06
=== kamal__ is now known as kamal
hallynhi - i keep getting lockups which i was blaming on unity, but now am wondering whether it's kernel.16:16
hallynhttp://paste.ubuntu.com/7011147/  does this mean anything to anyone?16:16
apwhallyn, not sure i know of them specifally, but it sounds like bad commands in the gpu pipeline, of course that could be mesa16:25
hallynok - yeah i couldn't tell if nouvou was misbehaving, or was getting a bad command.16:26
hallynwas wondering whether it is the same as https://bugs.launchpad.net/ubuntu/+source/unity/+bug/1282342, or a kernel cause for it16:26
ubot2`Launchpad bug 1284536 in compiz (Ubuntu) "duplicate for #1282342 compiz crashed with SIGSEGV in two_way_long_needle()" [High,Confirmed]16:26
hallynbut no i guess it is unrelated.  sigh.16:27
apwi would bring this up on #ubuntu-x someone might recognise it16:27
hallynheh, i try to avoid that chan, but ok will do :)  thanks16:28
sforsheeogasawara, apw, bjf: do we have anyone going through these fedora kernel patch report emails and seeing if there's anything we should consider picking up?16:30
sforsheee.g. https://lkml.org/lkml/2014/2/28/30216:31
bjfsforshee, i'm sure kamal will be happy to look at those :-)16:32
sforsheebjf: I noticed that the most recent one is patches on 3.13, which seems it would definitely be of interest16:34
bjfsforshee, yeah, so someone from the dev kernel team should be looking at those. your on that team aren't you? :_)16:36
apwsforshee, yeah that looks interestnig indeed, if you arn't looking :)  then i can on monday or s16:36
apwso16:36
sforsheebjf: I am, I just didn't want to duplicate effort if someone else is already doing it16:36
henrixsforshee: i usually look at those in my usual review of stable patches, but this one may be of particular interest for T16:37
apwsforshee, then ... we have a plan :)  thanks16:38
henrixmost of the time these are patches that hit the stable trees anyway, so... :)16:38
kamalbjf, sforshee, henrix: I'm not convinced that all of those (or perhaps any of those) are suitable for stable16:38
henrixkamal: well, yeah... i just saw there was an email, i didn't look at the patches16:38
bjfkamal, and i'm ok with that evaluation16:38
kamalbjf, sforshee, apw:  ok,henrix and I will look through them and compare notes16:40
ppisatihey17:05
=== kgunn is now known as Guest30100
=== ogasawara is now known as Guest32383
argeshallyn: hey i'm looking at bug 1278531. there is a patch that fixes 3.13, but are you still seeing issues with a 3.11 kernel?18:27
ubot2`Launchpad bug 1278531 in linux (Ubuntu) "nested kvm fails with trusty and upstream kernels" [Medium,Confirmed] https://launchpad.net/bugs/127853118:27
hallynarges: 3.11 has a idfferent bug,18:28
hallynat least for me.  for some ppl it works fine18:31
hallynfor me, kvm jsut hangs with 100% cpu18:31
argeshallyn: ok is that tracked in a different bug or same one?18:40
hallynarges: iirc that bug originally was for that failure18:40
hallynwe re-used it as i assumed at first it was the same.  but smb found that the 3.11 bug was probably fixed somewhere between 3.11 and 3.1318:41
argeshallyn: ahh here's the other bug 1208455 to track it18:44
ubot2`Launchpad bug 1208455 in linux (Ubuntu Saucy) "general protection fault running apt-get inside double nested kvm VM" [High,In progress] https://launchpad.net/bugs/120845518:44
hallynthat one gets more convoluted :)18:45
hallynarges: what exactliy are you trying to nail down?18:46
hallynyou're having a failure on saucy?18:46
argeshallyn: just want to make sure 3.11 work as well18:46
hallynok.  yeah, i'm kinda letting 3.11 suffer until trusty is golden18:46
sconklin#rake18:50
=== ogasawara_ is now known as ogasawara
apwrsalveti, ok uploaded flo, should be about an hour all things being equal19:54
rsalvetiapw: awesome, thanks!19:54
cetexACPI. Is it polling-based or event based? or a combination?23:59

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!