/srv/irclogs.ubuntu.com/2011/08/12/#ubuntu-kernel.txt

=== rsalveti` is now known as rsalveti
jayTI'm trying to build a module, but everytime I get the error: ERROR: Kernel configuration is invalid.03:28
jayThowever, I can build my kernel just fine03:28
jayTthe module is outside the kernel tree03:28
jayTjust a simple hello test module03:28
jayTmodules inside the kernel source tree are built fine03:28
jayTI have tried make oldconfig, make prepare, etc, and re-built the whole kernel,03:28
jayTbut always got that error when I tried to compile a module outside03:29
jayTcould anyone give a hint what's going wrong?03:29
jayTusing ubuntu 11.0403:30
philipballewHow do I put a the kernel time in my terminal when i issue comands?05:50
jk-philipballew: what do you mean by the kernel time?06:00
philipballewjk-, I have seen it by my kernal dev friends where on the side of the terminal the system time is there and shows next to each command and each line of output. 06:02
jk-so just the time of the day?06:02
jk-^ philipballew 06:03
amitkphilipballew: export PS1="[\d \t \u@\h:\w ] $ "06:03
jk-yep, what amitk said :)06:04
jk-but I'm a bit puzzled about the date appearing next to each line of output.06:04
amitkphilipballew: not really kernel related, but here are the keywords to google for: PS1 linux shell prompt time06:05
amitkjk-: you could remove the '06:06
amitk\d 06:06
jk-amitk: no, more that you're only setting the prompt, which won't affect terminal output.06:06
jk-philipballew mentioned "shows next to each command *and each line of output*"06:07
amitkjk-: PS2 is your friend for that trick06:07
amitkI suspect06:08
jk-PS2 is for the secondary prompt06:08
jk-not output06:08
jk-ie, continuations06:08
amitkaah, i see what you mean now. I doubt it is on every line of output06:08
jk-yeah, me too :)06:08
amitkmaybe he just saw dmesg with printk times06:08
jk-maybe.06:09
amitkwhich might explain "kernel time"06:09
jk-philipballew: show us a screenshot if you can.06:09
jk-</offtopic>06:09
amitktotally :)06:09
philipballewamitk, jk-  sorry, Had to step away06:24
philipballew export PS1="[\d \t \u@\h:\w ] $ " should be pasted where06:25
jk-philipballew: in to the terminal, and it will change the prompt temporarily06:27
philipballewalright, ill see what happenes06:27
jk-if you want to change it permanently, then add that line to ~/.bashrc06:27
philipballewoh, alright. Im ive seen it where it shows the system time. so you can see exactly when a partictular event happens06:29
amitkphilipballew: type dmesg and tell me if you saw that?06:29
jk-philipballew: I'm still not clear on what you mean by 'system time', and how that differs from the current "clock" time.06:30
philipballewdid like I would like when i tyoe dmesg i can have a system time stamp next to each line on the terminal output06:31
philipballew*type06:31
jk-that's what the numbers in square-brackets are - seconds since boot06:31
philipballewis that a common thing to have in your terminal?06:32
philipballewhttp://pastebin.ubuntu.com/663993/06:32
jk-no, it's specific to dmesg06:33
philipballewhum. haha. alright. Ill look around06:34
jk-you could add `| awk '{print strftime("%H:%M:%S")" "$0}'` after every command you type, but I don't think that's very convenient :D06:35
philipballewvery true. im trying to debug a possible kernel power management error to see if an error is the kernal or grub related. thought it might help06:36
philipballewmight be to much hassle06:37
jk-it's probably not a grub issue :)06:37
philipballewthe desktop will not restart06:38
philipballewive seen the issue be a boot loader issue, 06:38
jk-ok, that doesn't sound like a power management issue then06:38
philipballewhaha. it hangs on reboot, but shutdown -h now works fine. 06:39
jk-riight06:39
philipballewseemes weird. but i think I have the wrong boot parameters myself06:40
* philipballew is now off topic06:40
jk-where does it hang? do you see the BIOS screen?06:40
jk-no, this is now *on* topic :)06:40
philipballewwhenever I type sudo reboot or just click reboot the comp shuts down all processes then says will now reboot and just sits there06:41
apwand what kernel command line options do you have07:15
philipballewnone07:18
philipballewwell easily07:18
* philipballew is not a kernal know-it-all07:18
jk-philipballew: cat /proc/cmdline07:20
philipballewBOOT_IMAGE=/boot/vmlinuz-2.6.35-30-generic root=UUID=3ccd7388-34b9-40fe-960b-15371e1fb72a ro quiet splash07:20
* philipballew is doing this from sshing to his desktop (if that matters)07:21
jk-philipballew: yeah, that's cool. nothing unusual there.07:21
philipballewmaybe it needs something unusual to work?07:22
apwthere is a reboot=<letter> which switches reboot mode07:23
* apw finds the documentation07:23
philipballewit might be a bug for my hardware but the desktop is 6 years old so a big bug like this should mave been fixed07:24
apwphilipballew, not necessarily it may be unsusual h/w to have linx on07:25
apwphilipballew, ok i've pm'd you the documentation for the reboot=<x> command line option07:26
philipballewtrue. but i never had this problem with ubuntu and only xubuntu gives it. haha07:26
apwhmmm07:27
apwwell thats odd as i would expect those to have the same kernels, our kernels, and so to behave the same07:27
apwwhich ubuntu worked (version) and which of xubuntu does not07:27
philipballewi would to. i had ubuntu server on there and nothing07:27
philipballewboth 10.10 and 11.0407:27
philipballewbut ubuntu 10.10 didnt07:28
philipballewi ran ubuntu on it for years from 7.10 up to 10.10 and no problems ever07:28
apwso i can ignore the xubuntu thing07:28
apwas its a new version which doesn't work07:28
philipballewubuntu 10.10 reboot works. xubuntu 10.10 reboot = bad07:29
apwand which kernle does xubuntu 10.10 have in it07:29
apwcat /proc/version_signature tells you07:29
philipballewUbuntu 2.6.35-30.56-generic 2.6.35.1307:30
apwand have you tested the same version in ubuntu 10.10 ?07:32
apwie. how are you comparing to vanilla 10.1007:32
philipballewi can run a live cd you think?07:33
sijiapw, goodafter noon07:37
apwwell you are telling me a fact about ubuntu 10.10 so i am assuming you have tested it, i am asking how and when you _did_ test it.  so i can work oout if that would have had a similar or different kernel07:37
apwhi07:37
sijiam trying here to fix the bug which we discussed yesterday07:37
sijihttp://marcin.juszkiewicz.com.pl/2010/10/19/how-to-cross-compile-arm-kernel-under-ubuntu-10-10/07:38
sijican you pls tell  me whether this doc is capable of building kernel for beagle 07:38
sijiwhich is based on xdeb07:39
siji(till now i tried build the kernel directly in beagle , which consuming time) 07:39
philipballewapw it was 10.10 about 3 months ago. i tested it by knowing I restarted it and it worked just fine07:41
apwsiji, it looks viable07:41
sijiok07:41
apwphilipballew, ok so that had a lot older kernel so its entirly possible its an upstream stable change which has affected the kernel here ... you have a vanilla ubuntu kernel in your xubuntu image, so i suspect a real ubuntu install will also be affected the same way07:42
sijiinstead of linux-linaro have to use linux-omap3 right ?07:42
apwphilipballew, so the correct thing to do is now is to test those reboot= options.  as the kernel includes a table to select the appropriate reboot option.  let us know which if a07:42
apwany work for you07:42
apwsiji, i would guess so, i have obviously never actually tried it07:43
philipballewill do it right now07:43
sijiapw, ok will give a try07:43
amitksiji: you might find the author of the blogpost on #ubuntu-arm or #linaro as 'hrw'07:43
sijiok07:43
philipballewforce did not work. onto next07:51
sijiamitk, thanks i got him @ ubuntu-arm 07:52
amitksiji: I know, I'm there too :)07:54
siji:)07:54
sijiapw, amitk can you pls suggest some other method to cross compile the kernel 08:09
siji(am familiar with OE)08:10
apwsiji, i do all my cross-compiles in a regular chroot with the cross tools installed as in that post08:11
sijiregular chroot  = x86 chroot | Cross tools=xdeb & gcc-arm-linux-gnueabi ?08:13
amitksiji: I haven't used xdeb either08:15
amitkI probably should try again, it was broken last time I tried08:15
sijiamitk, i installed it with out any errors 08:16
amitksiji: it is the kernel compilation bit that was broken for xdeb08:16
sijiamitk, ok08:21
=== medberry is now known as med_out
apwsiji, regular chroot with the cross compiler installed10:19
sijiapw, ya understod 10:21
Kaleohi11:03
apwhi11:09
tjaaltonhum, how to build an amd64.deb mainline kernel with mainline-build? it defaults to i38613:10
tgardnerapw, ^^13:12
tjaaltonlooks like it's hardcoded in mainline-build-one, I'll try creating one for amd6413:16
ogasawaratjaalton: I hand edit the mainline-build-one script to generate amd64 debs13:17
tjaaltonogasawara: heh, okay13:17
ogasawaratjaalton: http://pastebin.ubuntu.com/664241/13:18
tjaaltonthanks, that saved me some time13:18
apwtjaalton, does it not make both by default13:19
tjaaltonapw: nope13:20
apwdon't understnad why not for you, it does normally13:20
apwhttp://kernel.ubuntu.com/~kernel-ppa/mainline/v3.0.1-oneiric/13:20
tjaaltonthis is the -one script13:20
* apw thinks, is that something someone else added13:21
ogasawaraapw: or maybe changes didn't get checked into kteam-tools?13:21
tjaaltonmainline-build sends email to the list, so running that directly doesn't feel like a good idea :)13:21
apwogasawara, no whats used daily is whats in the tree13:21
apwmainline-build-one builds the amd64 first i think13:22
apwit does clean, binary-indep, clean, binary-generic (in <release>) and then binary-generic (in <release-i386>)13:23
apwso it defiantly makes both13:23
apwtjaalton, ogasawara, ^^13:23
tjaaltonwell it didn't here, though I did run it against linux.git HEAD13:24
apwwell ... it must have said something when it tried.  perhaps your chroots are not the right names13:25
apwtgardner, in case you hadn't noticed i've pushed up an empty LBM13:29
apwit even builds as of just now13:29
tgardnerapw, ah ,and this tidbit of information implies I should upload it ?13:30
apwtgardner, i was hoping you'd glance over it for now13:30
apwas it only produced a single .udeb i am unsure if we care if it is uploaded or not yet13:30
apwogasawara, t13:30
apwogasawara, those lbm net packages, i think you added them13:31
apwogasawara, do you know where the source was from, ie. does oneiiric need them?13:31
* ogasawara tries to remember the lbm net bits13:31
ogasawaraapw: was that the updated e1000e driver?13:31
apwthat and igp or something13:32
apwigb13:32
apwsomething like that13:32
tgardnerthat stuff should all be upstream by now13:32
ogasawaraapw: if I remember correctly, those were picked from upstream so O shouldn't need them13:32
apwok then off is correct for them13:32
ogasawaraapw: but let me double check to make sure13:32
apwgawd they have changed the login screen to be all mad and off centre, really jars in my OCD world13:33
ogasawaraapw: yep, don't need them for O13:35
=== smb` is now known as smb
tgardnerapw, ogasawara: I think we can drop the LBM alsa, input, and media directories (and assorted rules and control.d cruft) since they aren't even enabled in Natty.13:40
ogasawaratgardner: works for me13:40
* ogasawara wonders if bjf is really here or in transit13:54
bjfunfortunately, really here, i just couldn't make myself sleep in13:55
bjfflight doesn't leave unti 13:3013:55
bjfbut i'm _not_ going through ATL13:56
tgardnerbjf, JFK instead ?13:56
ogasawarabjf: I'm going to change the format for this weeks IRC meeting.  Rather than me repeating myself every week with the delta review blueprint, I'm instead going to harp about outstanding work items that need done for the upcoming beta-1 milestone13:56
bjftgardner, PDX -> AMS13:56
tgardnerogasawara, hmm, will that make you a 'harpie' or 'harpette' ?13:57
ogasawaraheh13:57
bjfogasawara, cool13:57
* ogasawara back in 2013:58
brendandhi sconklin14:29
brendandsconklin - do you reckon we'll need to fully retest the respun kernel, or will the change be minor enough to make our results from the current valid?14:30
sconklinwell,  on one hand, since he tests didn't catch this, being a graphics issue and all, there's not much to be added by retesting. On the other hand, these are probably invasive changes so I think we'll have to completely retest.14:35
sconklinBut capping that off, we really don't have a fix yet, so we don't know what's required14:35
sconklinSo I think we'll probably have to recertify plus run some additional tests to see if the graphics work on those platforms14:35
tgardnerapw, ogasawara: uploaded linux-backports-modules-3.0.0_3.0.0-8.114:39
ogasawaratgardner: thanks!14:39
brendandsconklin - is it only affecting particular gfx cards?14:39
apwtgardner, heh, oh i'll take that as being ok :).  /me orders ram for pete14:40
tgardnerapw, oh, it was fine. I ripped out some obsolete stuff though.14:40
apwsconklin, did we have a display regression on 38-11 in natty ?14:41
sconklinbrendand: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/814325 We know it affects Intel graphics, I don't know whether we have a subset of models. 14:41
ubot2Ubuntu bug 814325 in linux "fuzzy and corrupted display with update in natty-proposed" [High,Triaged]14:41
apwtgardner, yeah i half heartedly stripped it15:15
tgardnerapw, I got out the bob nailed boots....15:15
tgardnerhob*15:15
* apw was liking the thought of bobs hammered into my boots15:15
* tgardner is going back to learning cloud infrastructure15:16
* apw notes that unity is the biggest source of bugs in oneiric again this week15:17
* smb is glad he only needs to cope with the olds ones this week15:20
* apw finds 5 bugs in the first 5 minutes as usual15:20
smbapw, You are using it in an unusual using manner15:21
apwsmb, yeah i am using it rather than osx15:22
smbAnd trying to get some work done with it instead of looking how nice it looks... ;-P15:23
apwjust managed to get compiz to coredump15:24
apwafter trying to restart it cause it hung15:24
apwquality s/w15:24
tgardnerapw, yeah, I've been having compiz issues this morning also15:25
apwluckily i only updated my scrapper today15:25
* apw checks, beta 1 is definatly later than alpha 115:26
* smb reminds apw about optimism15:28
apwi stand my interpretation, stupid (n)15:28
* smb keeps silent about apw having upgraded any machine15:29
* jjohansen thanks apw for the subtle reminder not upgrade while traveling15:31
* ogasawara feels compelled to upload today15:32
apwogasawara, :)15:32
ogasawaratgardner, apw: anything else you want to shove onto master-next before I kick some test builds and upload15:32
smbogasawara, And then run for the weekend. :)15:32
ogasawarasmb: of course!15:32
apwogasawara, nothing i know of here15:32
* ogasawara goes back to opening P15:33
tgardnerogasawara, the only thing pending is perhaps jjohansen's AA patches.15:33
ogasawaratgardner: ack, I'll maybe wait till about lunch time to kick off builds to give him time to respond15:34
ogasawaratgardner: if not, they can go up in the next round15:34
tgardnerogasawara, ack15:34
mdeslaurherton: think you can build me a test kernel with ROAF's two patches from comment #27?15:37
hertonmdeslaur: yep, I'm about to copy built kernels here, hang on15:38
mdeslaurherton: ah, cool :)15:38
jjohansenogasawara: just upload my patches are not worth holding off for15:40
ogasawarajjohansen: ack, I'm in no big hurry to upload so if they're not in by noon-ish I'll pull the trigger.15:42
jjohansenogasawara: okay, but it may be worth holding off on them and waiting to hash it out next week after I have talked to jdstrand15:42
hertonmdeslaur: hmm, had to restart build here because an error, will ping you when I finish15:43
ogasawarajjohansen: ah, in that case then we can wait and I'll go ahead an upload15:43
mdeslaurherton: thanks15:43
hertonmdeslaur: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/814325/comments/4415:57
ubot2Ubuntu bug 814325 in linux "fuzzy and corrupted display with update in natty-proposed" [High,Confirmed]15:57
mdeslaurherton: it only takes you 14 minutes to build a kernel? :P15:58
mdeslaurherton: thanks15:58
hertonyes, I use one of our build machines, it's pretty fast15:58
mdeslaurkernel team gets all the cool toys :)15:59
tgardnermdeslaur, the kernel team has this really annoying guy that makes lots of demands16:00
mdeslaurtgardner: hehe :)16:00
* herton -> lunch16:12
=== _LibertyZero is now known as LibertyZero
georghi fellows! had an GPF using dm raid5 with dm crypt on top. Trace: http://pastebin.com/QK15DkaU (actual kernel: ubuntu natty 2.6.38-8-generic #42-Ubuntu SMP x86_64 GNU/Linux). is this known, should i file a bug?16:38
=== jsalisbury is now known as jsalisbury_otp
=== lag` is now known as lag
=== lag is now known as Guest26716
tgardnerogasawara, do you remember how I added a macro to chromium? For example, I can enter 'bug 1' in the URL window which automatically expands to https://bugs.launchpad.net/ubuntu/+bug/1. The Lucid version in -proposed is wedging on some searches, so I wanted to make sure my browser is mostly vanilla.18:00
ubot2tgardner: https://bugs.launchpad.net/ubuntu/+bug/1 (Not reporting large bug)18:00
ogasawaratgardner: hrm, I don't know how to do it with chromium.  I did for firefox18:01
tgardneris ther e a common extensions directory?18:01
ogasawarano idea18:01
tgardnerhow did you do it for FF ?18:01
apwtgardner, you do that in the search section18:01
tgardnerI must have imported something18:01
* ogasawara digs through notes to remember18:01
apwtgardner, hit the spanner, preferences, then on the basics tab there is a "Manage Search Engines"18:02
apwadd a newone at the bottom with a keyword and you are good18:02
apwogasawara, ^^18:03
tgardnerapw, indeed, thats where it is18:03
ogasawaraapw: awesome, thanks!18:03
tgardnerso, thats unlikely to be the root of my lockup18:03
* tgardner --> lunch18:07
apwBAH pidgin just ate a bunch of my disk space ... it seems18:28
apwany suggestion where it put it18:28
apwi had that 2G left message, cleared 8, and got it back in minutes ... and found pidgin consuming a whole CPU18:31
apwahhh .xession-errors ... 15GB18:34
jjohansenapw: its beer o'clock, its just telling you to start the weekend19:08
=== jsalisbury_otp is now known as jsalisbury
kamalhow does one build the linux-tools-{version} binary package?   I discovered "fakeroot debian/rules binary-tools" which seems to try but eventually bombs out with this (on tangerine):19:41
kamal  dpkg-gencontrol: error: package linux-image-3.0.0-8-tools not in control info19:41
kamal... along the way, the build yields warnings about "newt not found, disables TUI support" and "/usr/bin/python-config is not an executable".19:43
kamalapw, tgardner, ogasawara: is tangerine set up to build the binary-tools target?  am I doing it wrong?19:43
ogasawarakamal: I can't say I've specifically ever tried fdr binary-tools19:44
* ogasawara tries19:44
tgardnerkamal, I don't remember that as being a target19:44
kamaltgardner: it is a target :-)    ogasawara, thanks!19:45
tgardnerkamal, I think its binary-perarch19:46
ogasawarakamal: I just did a fdr binary tools (but on tyler) and it seems to be building19:46
ogasawarakamal: ah, fails19:46
kamalogasawara: you did "binary<space>tools"  or "binary-tools"?19:47
ogasawarakamal: binary-tools19:47
ogasawarakamal: fails in the end though19:47
ogasawarakamal: same error19:47
kamalogasawara: ok, thanks for confirmation19:47
kamaltgardner: I'll try binary-perarch19:48
tgardnerkamal, the only reason I say that is because of this: 'debian/rules.d/2-binary-arch.mk:binary-perarch: toolspkg = $(tools_pkg_name)'19:48
kamaltgardner, ogasawara: yes, binary-perarch builds the pkg fine19:48
kamaltgardner: thanks!19:49
tgardnerkamal, you also need linux-tools-common so that it installs the wrappers19:50
kamaltgardner: understood, thanks19:52
tgardnerkamal, I presume you're looking at the x86 power tools ?19:53
kamaltgardner: my purpose here is to figure out whether the new "tools/power/cpupower/..." stuff is going to be built and packaged in linux-tools-* or not.  do you know?19:54
tgardnerkamal, indeed I do, 'UBUNTU: [Config] Package x86_energy_perf_policy and turbostat'19:55
kamaltgardner: I saw that, but was unclear what effect that would have on the tools packaging.19:55
tgardnerit adds wrapperd ala perf so that the correct ABI related tool is launched19:56
tgardnerwrappers*19:56
kamaltgardner: I think I'm talking about something else... I'm talking about the (very) new tools/power/cpupower/... stuff that just landed in the v3.1-rc1 tree (its the userspace utility that used to be called 'cpufrequtils' but is now called 'cpupower').20:01
tgardnerkamal, ah, I've not seen that yet20:02
* tgardner --> EOW20:45
* herton -> eow20:56
asabilhi all22:13
asabilis it normal that I don't have any pdflush process running ?22:14
khanxyep22:15
khanxits perfectly normal22:15
=== yofel_ is now known as yofel
milehigh2.6.38-10-generic-pae kernel, is CONFIG_NR_CPUS set to 7 for a reason? How can I change it while still using packaged kernels?22:29
=== emma is now known as em

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