[06:31] <ppisati> it's monday! :)
[07:32] <ppisati> 3.13 is out. rejoy!
[08:13]  * smb re... err enjoys his Monday cup #1
[09:58]  * apw yawns extravigantly
[10:00] <brendand> cking, hi
[10:01] <cking> brendand, hi there
[10:01] <brendand> cking, i've got a little fwts conundrum
[10:01] <cking> what's the problem?
[10:02] <brendand> cking, if a test has some subtests that are skipped, the overall status will be skipped
[10:03] <brendand> cking, at the moment we ignore skipped tests, even though some of the sub tests may be FAIL as well
[10:03] <cking> brendand, ok, which particular test(s) are that give you that issue?
[10:03] <brendand> Test           |Pass |Fail |Abort|Warn |Skip |Info |
[10:03] <brendand> ---------------+-----+-----+-----+-----+-----+-----+
[10:03] <brendand> method         |  286|   18|     |     |  121|     |
[10:03] <brendand> ---------------+-----+-----+-----+-----+-----+-----+
[10:03] <brendand> Total:         |  286|   18|    0|    0|  121|    0|
[10:03] <brendand> ---------------+-----+-----+-----+-----+-----+-----+
[10:04] <brendand> cking, can we ignore the Fails because the overall status was Skip?
[10:04] <cking> brendand, give me a an example of a test so I can get some better context
[10:05] <brendand> cking, the 'method' test - as here
[10:06] <cking> so why can't you just check for FAILs first as these take precedence
[10:07] <brendand> cking, we use --stdout-summary
[10:07] <brendand> cking, that says SKIPPED
[10:07] <cking> OK, that sounds like a bug, I'll look at that today
[10:09] <cking> yep, I'll fix that so it reports FAILED if any fails are found, it looks like it picks up an inherited status from the last subtest
[10:09] <cking> brendand, thanks for spotting that
[10:11] <cking> ok, figured out the bug, I'll send a patch out today
[10:12] <brendand> cking, i can create a subtask to our checkbox bug and could you update that when it's fixed?
[10:13] <cking> brendand, sure
[10:14] <brendand> cking, https://bugs.launchpad.net/fwts/+bug/1252186
[10:14] <ubot2> Launchpad bug 1252186 in checkbox "Some HIGH/CRITICAL failures in fwts related tests cannot be detected by checkbox" [Medium,In progress]
[10:41] <apw> cking, you are gnuplot savvy, my old plots which plot to gnuplots own interactive mode "click to dismiss" stylee, have stopped working after upgrades, you seen that ?
[10:42] <cking> apw, no, I've not used it on trusty
[10:43] <cking> normally I have to grok the docs or google for stuff in gnuplot as I can never remember how to use it 
[10:43] <apw> its almost like the default terminal is not 'x-window' any more
[10:43] <cking> hrm, that is annoy
[10:43] <cking> ing
[10:45] <apw> whe you start gnuplot on saucy what does it say terminal is set to ?
[10:46] <cking> apw, i really can't recall (useless eh?)
[10:46] <henrix> apw: Terminal type set to 'wxt'
[10:46] <apw> ahh ta
[10:47] <cking> wxt? what's that?
[10:47] <henrix> no idea!  haven't been using gnuplot for ages :p
[10:47] <apw> probally WindowXTerm or something
[10:48] <apw> and 'wxt' is not an option any longer
[10:48] <apw> bah
[10:50] <smb> The world is constantly changing to annoy us
[10:55] <ppisati> don't worry, i heard there's a comet going struck us.. it won't be for long...
[11:09] <apw> smb, this raid56 thing you did.  you know it can't do anything but offset 0... can we use device mapper to fix where its not at offset 0?
[11:09] <apw> smb, ie, lay down a linear mappnig at offset N which makes a device we can use at offset 0
[11:10] <smb> apw, raid45 but yes it probably might be done. Technically no problem more a question how well that fits into userspace
[11:10] <apw> smb, like my md raid0 inside lvm things
[11:10] <apw> smb, welll i thnk you make up that table for the new type in userspace?  so i think its a matter of makng the table bigger in the same place ?
[11:10] <smb> apw, And I did not want to spend too much effort if there would in practice be no offsets != 0
[11:11] <smb> apw, Currently the userspace does not create linear tables for the raid drives
[11:11] <apw> indeed, and i do wonder (in my email reply) if we should have a call-for-testing on it, to try and find out
[11:11] <smb> apw, So it is a matter of adding that functionality
[11:11] <apw> as yes, if there are none, it is moot
[11:12] <smb> apw, All true... actually I probably should email out the code oto ubuntu-devel as I have not heard anythng aback from xnox 
[11:12] <apw> right, the linears mappings would be supurflous if there is no offset, just popped into my head when replying and thought i would put it in yours so i can forget :0
[11:15] <smb> apw, Yeah
[11:23] <apw> henrix, beat me to it with the cves :)  nice
[11:26] <smb> Oh yeah, that scary one
[11:30] <henrix> apw: :)
[11:30] <henrix> apw: it was tagged for stable, so it was easy to spot
[11:36] <apw> ahh i found it the a different way, but when i went to paste it in, there it was :) and the autotraiger had started chewing on it 2m before
[12:31] <apw> henrix, ugg, this huge patch is erm huge.  I think we might need another fix patch which I have pointed to in my reply.
[12:32] <henrix> apw: looking...
[12:35] <apw> henrix, otherwise it looks to be ported as correctly as any review can see anyhow
[12:35] <henrix> apw: yeah, you're right.  it looks like we should also apply f81152e3
[12:36] <henrix> hmm... i'll need to go throught the stable trees as well to make sure all kernels have this fix
[12:37] <henrix> apw: thanks ;)
[12:37] <apw> henrix, thank you indeed
[12:38]  * henrix goes have some food before looking at this
[12:39] <apw> a plan indeed
[12:50] <Faux> My Trusty HP EliteBook 8570w with an GK107GLM [Quadro K2000M] hardlocks going into X on 3.13.0-4, only gets vesa-y resolutions (800, 1024) on 3.13.0-2 (and sometimes X just vanishes), but works fine on 3.13.0-1.  How do I report this?
[12:53] <apw> ubuntu-bug linux to file a bug against the kernel
[12:59] <Faux> Okay, cheers. :)
[13:03] <Faux> Hah, it refuses as I'm booted into a kernel that's no-longer in the repo.
[13:04] <henrix> apw: ok, looks like all the stable trees already contain the additional fix.  shall i update the cvetracker to include this additional sha1?
[13:20] <Faux> #1270818 is my bug report, if anyone has any questions.
[13:27] <apw> henrix, yeah i recon so
[13:27] <apw> bug #1270818
[13:27] <ubot2> Launchpad bug 1270818 in linux (Ubuntu) "nouveau(?) breaks or hangs in 3.13.0-4, -2, but not -1 on GK107GLM [Quadro K2000M]" [Undecided,New] https://launchpad.net/bugs/1270818
[13:27] <henrix> apw: ack, will do
[14:09] <cjwatson> I could use help in tracking down bug 1270228, if anyone has some time
[14:09] <ubot2> Launchpad bug 1270228 in linux (Ubuntu) ""Loading partman-xfs failed for unknown reasons. Aborting" error in trusty server installations" [Undecided,New] https://launchpad.net/bugs/1270228
[14:17] <smb> apw, Are you looking at this ^?
[14:28] <apw> smb, i am not currently, are you interested in poking that ?
[14:44] <smb> apw, Yeah I can have a look. Just wanted to avoid us going off both on the same
[14:47] <xnox> I am back from holliday and rebooted my desktop.... which did not end well.
[14:47] <xnox> linux-image-3.13.0-1-generic boots, while linux-image-3.13.0-4-generic does not.
[14:47] <xnox> i get bad screen resolution, kernel tracebacks, dropped into busybox which says that my rootfs device by uuid is not present =)
[14:47] <cjwatson> smb: one thing I forgot to say: to reproduce the out-of-memory part of it, you'll need kvm -m 512 or otherwise limit memory to match what the CI testing is doing
[14:47]  * xnox is puzzled.
[14:47] <xnox> (otherwise it's fully up to date)
[14:48] <smb> cjwatson, Ok, ack. 
[14:49] <apw> xnox, nvidia graphics ?
[14:50] <xnox> apw: intel hd / motherboard built in, I may however have inappropriate amount of *gl* packages installed which pulls in all sorts.
[14:50] <smb> xnox, I think I saw someone else having issues ... but that was with vmware gfx. Maybe check whether you see some issues with the drm driver and simple-framebuffer..
[14:50] <BenC> apw: I'm rebasing on master-next for trusty, if that will be of any help
[14:50] <xnox> apw: i'll wait for nvidia upload, and test that kernel, to shoot down one thing at a time.
[14:53] <apw> BenC, thanks, thats on my list for 'next' :)
[14:53] <BenC> apw: Ok, it's done
[15:32] <smb> apw, Hm, would you know from the top of your head whether rootfs might be closely related to tmpfs?
[15:36] <smb> apw, Or iow I wonder whether the "maddening" part of d1969a84dd6a44d375aa82bba7d6c38713a429c3 could not be true for rootfs
[15:38] <cjwatson> I think rootfs is a tmpfs instance if you have tmpfs enabled and aren't using rootfstype=.  See init/do_mounts.c
[15:39] <smb> cjwatson, So that might lead to the above commit which is not yet in 3.13.0-4 to be a possible culprit
[15:40] <cjwatson> I certainly wouldn't have got there from the commitdiff, but the description looks plausible
[15:40] <smb> Yeah, the change looks like and OMG but not directly would lead to the  effect
[15:41] <cjwatson> tmpfs does indeed use percpu counters to keep track of block usage
[15:42] <smb> So I would add that to the bug report and I am sure someone is already preparing a rebase to 3.13 release version
[15:53]  * smb is away for a bit
[16:01] <apw> Faux, test kernel in your bug
[16:02] <apw> smb, ok that commit is in the tree now, there is a test kernel by luck with it in on http://people.canonical.com/~apw/lp1270818-trusty/ if that helps
[16:02] <apw> though i was proposing to upload it today anyooo
[16:09] <apw> smb, and as you correctly say the existing -4 does not have that fix
[16:15] <Faux> apw: Sorry, I don't follow.
[16:57] <apw> Faux, i have built a test kernel which may fix your issue and posted the link to it in your bug
[17:01] <Faux> Ah, I shall have a look tomorrow (work machine).
[17:01] <Faux> Thanks.
[17:43] <apw> ogasawara, i am thinking about wrapping up a new trusty with 3.13 in it, anything pennding from you ?
[17:45] <apw> smb, did you bottom out on whether that percpu fix was your issue ?
[17:47] <cjwatson> it looks very likely to me ...
[17:48] <apw> ok i will get this thing tied up with string
[18:17] <SpamapS> Anybody hear reports about ext4 resize problems with 3.11.0-15-generic on i386?
[18:29] <apw> SpamapS, not that has come to my notice
[18:30] <SpamapS> apw: I may have had some corruption.. trying to reproduce
[18:36] <smb> apw, Yeah as cjw said, it looks quite likely. And since it needs to be tied up into an installer iso and we are very likely doing the 3.13 final soonish anyway...
[18:39] <smb> SpamapS, There was something that I remember only very vaguely. Though it may have been an unusually large resize and I cannot remember exactly whether that was at that point just not done before or broken. Nor can I remember whether it got solved or was lost to workstack overflow...
[18:46] <smb> Hm rather looks like forgotten... bug 1233075
[18:46] <ubot2> Launchpad bug 1233075 in linux (Ubuntu) "resizefs failure with raring" [High,Confirmed] https://launchpad.net/bugs/1233075
[18:53] <apw> smb, hmmm maybe indeed
[18:54] <apw> though i thought we saw some fixes for things like that recently
[18:54] <smb> apw, Which one? the resizefs?
[18:58] <apw> yeah particularly ext4 extending i thought ... hnmm
[19:00] <smb> apw, Not sure, somewhere in the bug report it was said that resize to 20G might be enough to reproduce, so I would try that tomorrow with T (if I can remember that long)
[19:06] <apw> smb, sounds like a plan
[21:41] <pkern> http://www.heise.de/newsticker/meldung/Kuenftiges-Init-System-der-Linux-Distribution-Debian-Zwischen-Patt-und-allgemeiner-Abstimmung-2088684.html
[21:41] <pkern> Oops, sorry. Not meant to paste that. -_-
[22:42] <tilgovi> apologies if this is not a good channel for a support question like this... i just got the brand new x1 carbon (i7) from lenovo and I've got it all set up and dual booting fine, but resume from suspend doesn't work and I was hoping someone could help me debug it.
[22:43] <tilgovi> the big issue is that this model has no leds for hd activity and pressing buttons seems to do nothing at all to resume
[22:43] <tilgovi> so the debugging pages about resume are not helpful. there's little info to collect. the suspend log looks normal, but i don't even know how to begin looking at the resume
[22:44] <tilgovi> again, I know this isn't a support channel, but since i'm able and willing to dive deeply in here, and there are probably few people with this hardware right now, i thuoght i'd come straight to the kernel folks