=== Jay-laptop_ is now known as Jay-laptop
GrayFoxhey guys01:36
GrayFoxanyone who can help?01:36
cjwatsonhey, does anyone know what happened to Ingo Molnar's patch to make relatime smarter? http://people.redhat.com/mingo/relatime-patches/improve-relatime.patch09:58
cjwatsonI don't really care about the mount options to a huge extent, but I'm turning on relatime by default in Ubuntu and it would be good if we had the "Is the previous atime value older than a day? If yes, update atime" bit09:59
cjwatsonwhich was at Linus' suggestion to avoid regressions in things like tmpwatch/tmpreaper, so I'm not sure why it hasn't been included10:00
cjwatsonhttp://lwn.net/Articles/244829/ <- context10:02
cjwatsonwould also avoid regressing popularity-contest data10:03
amitkcjwatson: I'm look at it now10:04
makscjwatson: afaik it is merged in fedora but not mainlined yet10:06
=== maks is now known as maks_
maks_fedora always ahead :D10:07
cjwatsonIngo works for RH, hardly surprising :P10:07
cjwatsonis the Fedora kernel in git anywhere? having no luck finding it10:08
maks_good old cvs10:08
maks_you can see kyle busily working :) np cjwatson 10:09
maks_need to watch that too for debian10:09
cjwatsonamitk: want a bug?10:11
firashello ---- how  save Kernel IP routing table rules10:12
amitkcjwatson: bug is appreciated10:12
firasI lost them after reboot10:13
maks_keep them in your pocket firas, wrong channel ask on #ubuntu for user support10:14
* soren discovers that it says "relatime", not "realtime" and now finds that the discussion is starting to make sense.10:16
aboganisoren: :-)10:16
firasthank you MAKS   -- I will not do becuas  your hand some times visit my pocket10:16
maks_firas you must mistake me for someone else :D10:17
cjwatsonamitk: bug 19942710:18
ubotuLaunchpad bug 199427 in linux "smarter relatime updates" [Undecided,New] https://launchpad.net/bugs/19942710:18
firasMAKS is the thief of route rules be careful poeple 10:20
tjaaltonabogani: hey, there's a bug against lrm that needs some rt-love, bug 197130 :)10:24
ubotuLaunchpad bug 197130 in linux-restricted-modules-2.6.24 "[hardy] Black X screen and system halt after installing linux-rt" [Undecided,New] https://launchpad.net/bugs/19713010:24
aboganitjaalton: Ok, thanks10:24
aboganiSeems to me that this patch is a good for inclusion in -rt...10:43
amitkcjwatson: it seems to me from reviewing the patch that we ought to enable CONFIG_DEFAULT_RELATIME in our kernel to force all filesystems use relatime by default. Do you see any drawbacks in that?10:56
sivanhi all11:03
=== sivan is now known as sivang
aboganiamitk: IMHO dump would be affected, at least11:05
sivangcould anybody tell me if splash boot option is absolutely neccessary to have splashy doing it's thing when booting ?11:05
cjwatsonamitk: I was less convinced about that; that's more painful for users to disable since it's not so fine-grained11:05
sivangI have it set up from the ubuntu package but nothing happens11:05
sivangI tried highring it's start prio in rcS.d and rc1.d but nothing11:06
sivangcjwatson: hey Colin11:06
sivangamitk: fellow from Israel? :)11:06
amitkcjwatson: so you plan on adding relatime to /etc/fstab?11:08
amitksivang: no, amit is a very common name in India :)11:09
cjwatsonamitk: already done :-)11:10
cjwatsonat least for new installs11:10
cjwatsonI sort of question adding it to existing installs11:10
sivangamitk: Ah, heh pleased to meet you then11:11
amitkcjwatson: that leaves all the upgraders out...11:12
amitksivang: you too11:12
cjwatsonamitk: your call11:13
amitkabogani: does dump use atime instead of ctime and mtime?11:14
aboganiamitk: No, i wrong. Spoken before think. Sorry.11:17
=== cjwatson_ is now known as cjwatson
cjwatsonamitk: IIRC you were working with ogra on making CONFIG_USB_SUSPEND a boot parameter, since some hardware likes it one way and some the other; did anything come of that?12:15
amitkcjwatson: the key is to enable USB_PERSIST and make it runtime configurable. I think it is very much doable and I intend to work on it later today12:17
BenCGood morning folks13:38
aboganiBenC: Good morning, sir!13:42
BenCcking_, amitk: I am prepping the next upload (-12.18), so if there is anything you want in, let me know now14:07
cking_BenC: Nothing to add... fine by me.14:07
cjwatsonif either of the relatime thing I mentioned earlier or the USB suspend/persist thing are straightforward, I'd appreciate having those in beta14:10
cjwatsondid I remember to ask if anyone had picked up mjg59's VT font restoration patch?14:11
cjwatsonI confirmed that it fixed the font restoration problems from the hardy-console spec for me14:14
amitkcjwatson: mjg59's patch went in14:20
amitkso did relatime14:20
amitkworking on USB persist now14:20
amitkBenC: how long do I have? I am reviewing the USB persist config option.14:20
cjwatsonamitk: it did? awesome14:22
BenCamitk: probably can do it14:23
BenCamitk: I'm doing a test build now...make sure you pull before messing with config changes (had some updates to config files just pushed)14:24
tjaaltonBenC: check bugs 199037 and 18918514:25
ubotuLaunchpad bug 199037 in linux "Null bytes in files access by 2 or more NFS clients" [High,Triaged] https://launchpad.net/bugs/19903714:25
amitkBenC: yeah.. looks like enabling CONFIG_USB_PERSIST in i386/generic should be enough for the needs of ogra and eee pc guys.14:25
ubotuLaunchpad bug 189185 in linux "Thinkpad X61 hangs when removing from dock" [Unknown,Fix released] https://launchpad.net/bugs/18918514:25
BenCtjaalton: do they have tested patches?14:25
BenCtjaalton: if not, I can't do much with them for this upload14:25
tjaaltonBenC: both are applied upstream14:25
amitkBenC: great! so you decided to enable RELATIME by default :)14:26
BenCtjaalton: well I assumed that the relatime patch was only useful if we did enable it :)14:26
cjwatsonheh, I'd already turned it on in the installer, but feel free to do it in the kernel too14:26
BenCwell the patch presented a new option, so I figured it should be turned on...is it safe to do so right before beta?14:27
amitkBenC: the patch is fairly straight-forward and I am of the opinion that everyone (including upgraders) should benefit from relatime I/O savings.14:28
BenCThat's good enough for me, so I'll leave it on14:29
BenCtjaalton: Just in time...I'll get both of those in, assuming they pass some trivial review and testing14:31
BenCtjaalton: 199037, that commit is already in our kernel14:33
BenCtjaalton: looks like it got cherry picked from some other report...so I'll mark this bug14:34
BenCamitk: so with your config changes, is usb-persist enabled by default?14:50
BenCsmb: Tim had already bumped the ABI to -12, and the last one was -11...not sure what sort of ABI related failure you would have had though...are you sure it wasn't a module issue?14:52
smbBenC: Not completely, I just saw messages about abi check failed and by reflex bumped the abi. :-(14:53
BenCI'm doing a build now (reverted to -12), so I'll see what happens14:56
smbBenC: Ok, I will do the same on one of the porters and on my build machine.14:57
BenCamitk: I assumed that the USB_PERSIST thing was supposed to be enabled in config only, and we would add a mechanism to even then make it not enabled by default unless a proper kernel param was passed14:57
amitkBenC: on a closer look that feature is already implemented, as a power/persist file for each USB device.15:02
amitkso whatever device needs to enable it will just write to the 'persist' file from userspace15:03
BenCamitk: ah, ok15:05
BenCamitk: then maybe we should be consistent and enable it across the board?15:06
amitkBenC: I have mixed feelings about it. It would allow more users to potentially shoot themselves in the foot...but on the other hand everyone running -generic (i386) has already been given the gun.15:09
=== amitk is now known as amitk-afk
BenCamitk-afk: yeah, which is like 90%15:29
BenCsmb: ABI/module checks passed for me with -12 ABI set15:30
BenCsmb: perhaps the -11 ABI files weren't present when you tried the compile the first time15:31
=== amitk-afk is now known as amitk
amitkBenC: so I should enabled it across the board?15:31
smbBenC: Hm, I really have to look closer next time. :(15:32
BenCamitk: sure, can't hurt any more than enabling for -generic x8615:33
BenCit's not like people will install -386 to get around the problem :)15:34
amitkheh, true15:34
amitkBenC: and I surely don't want to suggest creating another flavour :)15:34
BenCamitk: let me know when you have that ready...test build went fine, so I'm ready to upload15:41
amitkBenC: done15:43
krautone general question, perhaps anyone could help me with this here: is it safe to give an user the possibility to use chroot /new/chroot via sudo as root? i am using a grsec-patched kernel, so it's not possible to mknod devices within the chroot.15:44
BenCkraut: you're better off using schroot package15:47
BenCamitk: thanks15:52
krautBenC: what's schrott?15:52
krauttranslated in german the word sounds not really trustable15:53
krautah, i see.15:55
krautthanks BenC!15:55
BenCOk, -12.18 is away16:07
=== BenC changed the topic of #ubuntu-kernel to: Ubuntu kernel development discussion ONLY | Kernel Wiki: https://wiki.ubuntu.com/KernelTeam | Latest kernel upload: 2.6.24-12.18 | Latest news: 2.6.24 Release in Hardy | Next meeting: Mar 11, 17:00 UTC | Kernel Team machine: http://kernel.ubuntu.com
BenCI'm going to sign off for awhile to do some maint on my laptop...if anyone needs me, my mobile is on me16:08
=== kylem_ is now known as kylem
amitksmb: when Ben comes back online, could you tell him we need to respin the kernel. I forgot to push a lpia patch to solve a FTBFS. *sigh* I might do it over the weekend.17:04
smbamitk: Ok, I will tell him. Might be I found my problem as well. Missing modules in virtual...17:05
amitksmb: I mean the patch is pushed, but I might as well wait for all builds to be done before uploading another one17:06
amitksmb: thanks17:06
smbamitk: np. 17:08
smbsoren: should kvm modules be build for virtual?17:11
* smb goes for some food17:13
sorensmb: Nope.17:14
=== thegodfather is now known as fabbione
amitksmb: Ben forgot to enable relatime on the custom flavours, I just pushed patches18:35
smbsoren: Hm, ok. It looks a bit like have been built before and this is being complained. Ok, then I try to fix that.18:41
smbamitk: Ok, I will tell him later.18:41
=== asac_ is now known as asac
amitkheya kylem_ 21:19
=== clever[rev] is now known as clever
xhakernoticed the new kernel update. just ftbs. could you include https://bugs.edge.launchpad.net/ubuntu/+source/linux/+bug/18712123:11
ubotuLaunchpad bug 187121 in linux "Udma not fully available in Acer 1694 Wlmi" [Medium,Confirmed] 23:11

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