/srv/irclogs.ubuntu.com/2010/04/30/#ubuntu-kernel.txt

=== kamalm is now known as kamalm-away
arputhx for this https://launchpad.net/~chogydan/+archive/ppa/+packages03:00
arpuworks greate 03:00
arpunow i can use lucid to work :>03:00
jk-arpu: what didn't work previously?03:57
arpujk-,  response time of the gui on file copy 04:07
arpuor apt upgrade ...04:07
mase_wkhi guys, just wondering if there is a reason CONFIG_DMAR and CONFIG_INTR_REMAP is not set for the _server kernel in Lucid ?04:41
mase_wkaccording to http://www.linux-kvm.org/page/How_to_assign_devices_with_VT-d_in_KVM it's required for KVM PCI passthrough04:41
jjohansenmase_wk: no one has asked for it, I can added to the config review of M04:48
mase_wkjjohansen: well i think it was enabled in karmic. i performed an upgrade of a test system yesterday which uses PCI passthrough04:50
mase_wkand it seems PCI passthrough was supported / working in karmic.04:50
mase_wkso if this could be added it would be much appreciated04:50
jjohansenmase_wk: hrmm, I'll have to check on that but we didn't disable almost anything04:51
mase_wkhmm ok04:51
mase_wkthats interesting04:51
jjohansenmase_wk: I am unsure if it will get added to the Lucid kernel, we can look at SRUing it but no promises04:52
mase_wkSRU ?04:52
jjohansenstable release update04:52
mase_wkah ok. well i dunno, maybe it's not required then. i am confused04:52
mase_wksee i'm getting the No IOMMU found04:53
mase_wkand i assumed it was because that config option was disabled in the kernel04:53
mase_wksince it was listed as required on the linux-kvm page04:53
mase_wkbut if your saying that wasn't enabled in karmic either then maybe i'm hitting something else04:54
mase_wksince the PCI passthrough was working on karmic04:54
mase_wkjjohansen: do you know who the most appropriate person to talk to in the ubuntu virt team would be ? I should probably make sure that this is actually required before asking for it to be added to the config review.04:58
jjohansenmase_wk: kirkland knows the most about kvm04:58
persiaIt was set in karmic, it seems.04:58
jjohansenpersia: interesting04:59
mase_wkpersia: thanks for verifying that. I have upgraded all my test boxes so that saves me a reinstall at least04:59
persiaI might be misreading the output of `grep -rin CONFIG_DMAR` though.04:59
mase_wkpersia: is that for the -server kernel ?05:00
persiaSeems to only be for x86_64 and ia64 though, not for i386.  Dunno why.05:00
mase_wkthat makes sense i am using x86_6405:00
jjohansenpersia: I am not seeing that05:01
persiaMaybe I'm reading it wrong.  Maybe it's only enabled in ports.05:01
mase_wkpersia: what do you mean by ports sorry ?05:01
persiaI'm looking at the 2.6.31-21.59 source, but I'm not sure I'm reading it properly05:01
persiamase_wk: sparc/powerpc/ia6405:01
mase_wkok , i understand now.05:01
jjohansenpersia: I am looking at the configs right now for x86_64 karmic and the only one that is enabled is PCI Stub driver05:02
jjohansenthe other 2 options are experimental and not enabled05:02
persiaI'll trust your check more than mine :)05:02
persiamaybe upstream kvm behaviour changed?05:03
mase_wkwell my understanding was that they couldn't  / would never allow software iommu05:03
mase_wkjjohansen: the config you checked was that the -server kernel for x86_64 karmic ?05:06
mase_wki guess i could create a lvm snapshot and install a kernel with it enabled and see if that works05:09
jjohansenmase_wk: yes, double and triple checked05:15
jjohansenmase_wk: yeah I would give that a try05:15
mase_wkk will do :)05:16
mase_wki guess i could reinstall karmic just to make sure i'm not going completely mad05:18
LaibschCan I build a Lucid Kernel on a Debian system (for testing only)?06:43
Laibschor do I need a chroot on that Debian system?06:43
persiaYou'd be better off with a chroot, so you can install all the build-deps easily.  No reason you can't do it in Debian, but I'm not sure it's a well-tested path.06:45
ikepanhcyeah, and if your system need external kernel module, make sure it combiled with the same gcc and reasonable kernel header06:47
LaibschI'll compile on the faster Debian machine natively now07:04
LaibschI merely want to test a patch for a) compilability and b) functionality (things work as expected from the patch)07:05
LaibschiF things break, I'll need to reconsider07:05
=== oubiwann is now known as oubiwann_
=== ogasawara_ is now known as ogasawara
ogasawarahttps://wiki.ubuntu.com/MaverickReleaseSchedule09:56
ogasawaraamitk: https://wiki.ubuntu.com/KernelTeam/Specs/KernelMaverickUbuntuDeltaReview10:29
ogasawaraamitk: that has a list of patches (grouped by whom I thought was the owner) that I think should go upstream this cycle10:29
* psurbhi breaks for lunch11:59
=== psurbhi is now known as psurbhi-afk
cmugHi, 10.04 alternate does not work with my USB cddrive. What gives ?12:14
cmugDesktop and Server installed fine12:15
=== psurbhi-afk is now known as psurbhi
LaibschAnybody I should subscribe to bug 521967 to make sure that patch now lands quickly?12:55
ubot3Malone bug 521967 in linux-backports-modules-2.6.32 "support for new atheros wifi chipset - AR2427/ath9k" [Medium,Triaged] https://launchpad.net/bugs/52196712:55
Laibschtested both compilation and function on-device12:55
fabbionesmb: btw.. i just remembered that there was already a bug related to the DVB HVR patch. but hell if i remember the number13:01
=== oubiwann_ is now known as oubiwann
amitkhey all, we just started gardening our kernel wiki by creating a new namespace wiki.ubuntu.com/Kernel and moving to a more activity-based wiki like wiki.ubuntu.com/X14:03
JFotgardner, is there anything we need to do for bug 521967 or will it land in updates?14:21
ubot3Malone bug 521967 in linux-backports-modules-2.6.32 "support for new atheros wifi chipset - AR2427/ath9k" [Medium,Triaged] https://launchpad.net/bugs/52196714:21
bjfamitk, good for you, let us know how that works out for you :-)14:53
JFoheh14:53
JFomanjo, are you still doing suspend/resume stuff/14:53
JFoerr?14:53
JFoif so bug 44058 looks to be for you14:53
ubot3Malone bug 44058 in gnome-power "suspend when closing laptop lid doesn't work" [Unknown,In progress] https://launchpad.net/bugs/4405814:54
JFoit is HP hardware14:54
JFomorning bjf 14:54
JFomeant to say that first14:54
JFo<-brain still slow14:54
bjfmoin JFo 14:54
JFocoffee brewing as we speak14:54
* psurbhi sees a function definition within a function definition in grub legacy code15:00
amitkbjf: I _started_ the page, I move to newcore next week :-p15:00
bjfamitk, figures15:01
JFoamitk, ship jumper :-P15:02
JFoI bet it is an action item for me :)15:03
JFoamitk, as pete what the rest of "Why be a buddy, when you can..." is ;015:03
amitkJFo: hehe15:03
JFoheh15:03
mjg59psurbhi: That is, by far, not the worst thing about grub15:04
amitkJFo: no actions, just establishing the namespace. We'll leave the fixing to the newbies ;)15:05
JFoheh, either way :)15:05
psurbhimjg59, thats inspiring to know15:05
psurbhi:)15:05
bjfamitk, we're referred to as "the little people"15:08
* JFo has a pot o gold for ye15:08
bjfamitk, isn't there a somehands attendees page somewhere?15:09
amitkbjf: *shrug* us little people dunno15:12
bjfamitk, I thought the lead of the newcore kernel team would be going15:13
tgardnerJFo, I thought I'd wait until Luis releases an official 2.6.34 compat-wireless before backporting t o Lucid.15:14
JFocool15:14
JFojust wanted to check :)15:15
JFothanks tgardner 15:15
tgardnerbjf, can you surf vger.kernel.org ?15:16
bjftgardner, trying to catch a wave right now15:16
bjftgardner, not looking good15:17
bjftgardner, nope15:17
tgardnerbjf, must be down. I've not received any LKML email in several hours15:17
JFomanjo, bug 540643 another HP S/R issue15:18
ubot3Malone bug 540643 in linux "[Hewlett-Packard HP Compaq 6910p (AK677US)] suspend/resume failure [non-free: hsfengine wl]" [Medium,Triaged] https://launchpad.net/bugs/54064315:18
manjoJFo, taking a look15:19
JFocool, thanks man :)15:19
kirklandapw: what can you tell me about this stack trace: http://launchpadlibrarian.net/46615447/kernel_stack_trace15:23
kirklandapw: 10.04 host, 10.04 guest15:23
kirklandapw: virtio networking15:24
kirklandapw: guest continues running, but networking is dead15:24
manjokirkland, looks like you are runing out of swap ?15:32
kirklandmanjo: oh?15:32
kirklandmanjo: checks ...15:32
apwkirkland, looks like we tried to recv a packet, and were not prepared to fail the allocation15:33
apwi'd say i'd expect the allocation to be allowed to fail and the packet simply be dropped15:33
apwas it wasn't and tripped an oom i'd say its a bug15:34
kirklandapw: cool, filed: https://bugs.launchpad.net/bugs/57236815:34
ubot3Malone bug 572368 in linux "virtio networking dies in KVM guest with kernel backtrace" [Undecided,New] 15:34
apwkirkland, so is it reproducible?15:35
apwkirkland, be worth trying to reproduce and monitor /proc/slabinfo to see if we are leaking15:42
kirklandapw: i rebooted;  will post as soon as it happens again15:42
kirklandapw: will monitor that15:42
apwkirkland, thanks, let us know15:43
JFothis looks interesting: bug 57197715:49
ubot3Malone bug 571977 in linux "[Lucid] Possible race condition involving rtc wakealarm when hibernating a system" [Undecided,New] https://launchpad.net/bugs/57197715:49
* bjf heading to the airport, back on in a bit15:50
* psurbhi quits for today...see u all on monday16:23
JFotgardner, don't we have something for bug 521967 going in soon, or am I thinking of a different one?16:31
ubot3Malone bug 521967 in linux-backports-modules-2.6.32 "support for new atheros wifi chipset - AR2427/ath9k" [Medium,Triaged] https://launchpad.net/bugs/52196716:31
JFoheh16:31
JFonevermind16:31
JFothat is a duplicate of the one from earlier16:31
JFoheh16:31
JFo<-brain needs more coffee16:31
JFoyet another interesting bug: bug 57180716:56
ubot3Malone bug 571807 in linux "Lucid fails to spin down hard-drive on shutdown" [Undecided,New] https://launchpad.net/bugs/57180716:56
* jjohansen bets JFo is getting a deluge of interesting bugs16:56
JFojjohansen would be absolutely right :)16:57
JFomost are actually coming in by way of people reassigning them from the ubuntu package16:57
JFoso they may have been filed, the reporter just didn't know the right package to assign16:58
jjohansenmakes sense, its easier to file against17:01
JFoyeah17:04
JFojust sucks that by the time I get them they are stale17:04
jjohansenwell you could triage all the ubuntu package bugs ;)17:05
JFojjohansen, :-|17:12
jjohansenJFo: actually /me thinks it would make the problem worse as the sheer volume of bugs would make it so that on average you would find the kernel bugs even later17:14
manjoJFo, the previous hibernate bug seems to be either NOT a bug or a bug in pm scripts17:28
manjoI added a comment there 17:28
JFocool, thanks manjo :)17:29
JFonice recovery attempt jjohansen, but the suggestion is like a genie once it is out of the bottle. ;-)17:29
JFoyou know what happens when you make suggestions like that? God kills a kitten :-P17:30
JFoj/k17:30
manjoJFo, you should mark 571807 invalid.. there is not any information in that bug to support his claim17:31
* jjohansen will have to do penance and buy JFo a beer17:31
JFojjohansen, I can live with that :-)17:31
JFomanjo, looking17:32
manjosay you are marking invalid coz there is no data in the bug to support the claim. either use ubuntu-bug to attach the logfiles 17:34
JFoI just asked him for more info17:34
manjothere are possibly hundreds of aspire ones and no one has reported data failure 17:35
JFoor odd spin down issues17:36
JFobut we will see what is said17:36
sbeattiejjohansen: your penance is to triage bugs against no package with hints: http://people.canonical.com/~brian/reports/no-package-clues.html :-)17:37
JFosbeattie, :-)17:37
manjoJFo, https://launchpad.net/bugs/44058 is from 4yrs ago ? 17:38
ubot3Malone bug 44058 in gnome-power "suspend when closing laptop lid doesn't work" [Unknown,In progress] 17:38
JFomanjo, yep, but they have provided info from Lucid also 17:39
JFosame issue, they are telling me17:39
manjoJFo, I think I have a HP 5101 let me check... 17:41
JFomanjo, ok17:41
JFocould be an artifact left over from upgrade, but there are several people affected17:41
jjohansensbeattie: argg, no!!!!17:42
manjoJFo, could be a regression in gnome-power-manager17:43
manjochecking 17:43
JFomanjo, k17:43
JFojjohansen, ;P17:43
manjoJFo, looks like its a valid gnome-power-management bug17:46
JFomanjo, I see17:47
JFodo you have a link to a g-p-m bug?17:47
JFoor just in looking through the bug itself?17:47
jjohansensigh, you would hope after 4 years it would have been fixed17:48
JFojjohansen, you know as well as I, things fall through the cracks :-/17:49
JFoas much as we hate for them to17:49
JFoman, this hammock was money well spent :)17:50
jjohansens/cracks/crevices/17:50
JFoyeah17:50
manjoJFo, not a kernel bug17:52
JFomanjo, cool17:53
JFoso I need to set it to g-p-m/17:53
manjoJFo, probably it is already17:53
JFowow, so it is17:54
JFoso the kernel task is invalid?17:54
manjoyep we could call it that 17:55
=== kamalm-away is now known as kamalm
JFok17:58
JFomarked as such18:02
bigcx2hey all18:09
bigcx2i have a question about using make-kpkg18:09
bigcx2if i do something like make-kpkg --append-to-version some-foo18:09
bigcx2that string gets duplicated in the resultant deb that gets created, e.g:18:09
bigcx2linux-image-2.6.32.7-some-foo_2.6.32.7-some-foo-10.00.Custom_i386.deb18:10
bigcx2how do i get rid of that duplication and get rid of the "10.00.Custom_i386" string at the end18:10
krohi18:24
krogot a question18:24
kroI want to use lucid... but plymouth doesn't seem to be compatible with my kernel, it dies on boot-up18:24
kroit's generic non-modular kernel optimised for headless usage.18:25
kronow the question: what does plymouth NEED? So far I tried adding support for framebuffer (VESA), DRM, AGP, ... but doesn't seem to be it18:26
bigcx2kro: do you need a splash screen?18:26
krono18:27
bigcx2try taking the "splash" option out of your grub menu.lst and see what happens18:28
kroit's for dedicated servers... so the output is not needed except some debugging stuff18:28
kroyes, did that already...18:28
bigcx2and nothing?18:29
kroreplaced it by "debug"18:29
krohttp://demo.ovh.com/view/9aa6de29af7a198cac0e9a675c78f23f/0   screenshot of the KVM console18:29
kroI added "--debug" to the /etc/init/plymouth* files18:30
krobut doesn't really bring me further18:30
bjf?18:30
krobjf: init: plymoth main process (1421) killed by SEGV signal18:31
kroinit: plymouth-splash main process (2275) terminated with status 218:31
Sarvattbigcx2: check out /etc/kernel-pkg.conf18:32
bjfkro, the '?' was a typo in the wrong window :-)18:32
krobjf: :)18:32
bigcx2Sarvatt: aha! thanks18:33
bigcx2can these be set via environment variables?18:34
bigcx2before running?18:34
bigcx2or just in the conf, and debian/rules18:34
bigcx2kro: perhaps this applies to your problem? http://www.linode.com/forums/viewtopic.php?t=5522&sid=6a540b7bab2c42cebcddb265752594e618:36
krothx, but already tried that too :-)18:37
bigcx2kro: ok sorry, that was my best shot18:38
bigcx2kro: i'm not a kernel dev, just camping out18:38
kroI'm stuck since nearly two days now, so I hit the google bookmark quite a lot :-)18:38
bigcx2:)18:38
bigcx2heh18:38
* JFo does a long overdue git fetch on his kernel sources18:40
JFoapparently very long overdue18:40
krohow about the "ply_terminal_open:terminal /dev/tty7 is already open" message - is this the fatal one?18:49
bigcx2kro: probably a good guess18:55
bigcx2kro: but then again here's a log of somebody having that exact same error, but w/o failure18:56
bigcx2http://users.csc.calpoly.edu/~dgoyette/no_splash_for_you.log18:57
kroyeah... that shouldn't be the cause for such an ungraceful exit...19:00
bigcx2kro: is this your own kernel, or a stock one19:00
kro"my" kernel (our company kernel)19:01
bigcx2https://bugs.launchpad.net/ubuntu/+source/plymouth/+bug/57125819:01
ubot3Malone bug 571258 in plymouth "plymouth main process (341) killed by SEGV signal" [Undecided,New] 19:01
kro[drm] Initialized drm 1.1.0 20060810   <== my log stops before that line... maybe I should look in that direction19:02
krobigcx2: thx, didn't see that one yet :)19:02
bigcx2np19:03
mvkis that 3d support of the built-in kernel module for radeon buggy?19:42
mvklets see if i can trick this radeon driver into panicing again 19:51
mdzcan anyone recommend a USB wifi dongle which is well supported?20:34
Dandelmdz, look for edimax EW-7717Un wifi adapter, the drivers should be included in 10.04 since 9.10 also has the driver.20:38
mdzDandel, thanks for the suggestion, looking at it20:39
mdzseems to have very bad reviews on http://www.newegg.com/Product/Product.aspx?Item=N82E16833315081&Tpk=EW-7717Un20:40
Dandelpeople on newegg can and will give a bad review because a printer "seems" to have 6 cartridges, when it's really 4 cartridges and a pair of print heads.20:41
Dandelsame rule of thumb applies to usb devices20:41
mdzDandel, true, it's not always reliable. these are "stopped working after 2 months" kind of reviews though20:41
mdzanother one I'm looking at is http://www.newegg.com/Product/Product.aspx?Item=N82E16833127244&cm_re=usb_wifi_adapter-_-33-127-244-_-Product20:42
mdzwhich looks like it should be supported by rt2870sta20:42
mdzbut then there's bug 50737520:43
ubot3Malone bug 507375 in linux "Wireless DLINK Adapter DWA-140 not working" [Undecided,Confirmed] https://launchpad.net/bugs/50737520:43
mdzseems to be one of those devices which has different revs, indistinguishable before purchase20:43
Dandellook at the  EDIMAX EW-7718Un device then.20:45
Dandelthat uses the RT2870 chipset.20:46
mdzDandel, I think I will, thanks20:47
Dandelanyways, i never trust newegg reviews until the device has 300+20:47
mdzDandel, have you tried 802.11n? WPA2?20:48
Dandel802.11g using madwifi tho.20:48
Dandelno, but the router i have is 802.11n with wpa2.20:48
Dandeland yes, i'm using wpa2 right now.20:49
mdzDandel, madwifi? I thought it was an rt2870?20:50
Dandelno, that's what i am using right now.20:50
DandelMy laptop wifi card  is the atheros AR5001.20:51
mdzDandel, oh, I was wondering if you'd had success with WPA2 with the adapter you were recommending20:51
mdzthe edimax20:52
Dandelhaven't checked.20:52
Dandelsince the devices uses the rt2870sta chipset you can go to realtek's website and find out for sure.20:52
Dandeloops... rt2870 is ralink20:53
mdzyes, confusing that20:55
Dandelthe ralink driver that is proprietary does support wpa2 from what i am reading.20:58
Dandeloh and the rt2770 v2.3.0.0 does appear to fall under the gpl license.20:59
kamalmAdvice?... My "fakeroot debian/rules binary-generic" is failing at the final packaging step with:22:43
kamalm  dpkg-deb - error: (upstream) version (`unknown') doesn't contain any digits22:43
kamalmWhat might I have done to get into this state and how might I get out of it?22:43
crimsunmdz: certainly anything atheros ar9285-based is rock-solid IME22:50
crimsunkamalm: interesting; that implies something is fubar in debian/changelog22:51
kamalmcrimsun: its certainly possible that I've fu'd it bar -- but I think this is what I usually do...22:51
kamalm  linux (2.6.32-21.33~kamal~dell_155x_bri) lucid; urgency=low22:51
crimsunnah, that seems fine, but something is overriding it22:54
crimsuni.e., it becomes unknown-foo at some point, which is odd22:54
jjohansenkamalm: hrmm, I have hit that before I trying to remember what caused it22:57
kamalmhmmm.  I just changed it to   linux (2.6.32-21.33~kamal) lucid; urgency=low    and it worked -- sort of ...22:57
kamalmIt built these .debs:22:57
kamalmlinux-headers-2.6.32-21-generic_2.6.32-21.33~kamal_amd64.deb22:57
kamalmlinux-image-2.6.32-21-generic_2.6.32-21.33~kamal_amd64.deb22:57
jjohansenkamalm: okay what is wrong with those?22:58
kamalmI think bigcx2 was asking about this same issue -- the version numbers seem unusually "redundant" there.22:58
kamalmno, I take it back -- i guess that's "normal".22:59
jjohansenyeah23:00
kamalmI still don't know what's wrong with my original version string -- I've used that same pattern  (~kamal~foo_bar_bing) successfully before.23:01
jjohansenhrmm, no idea unfortunately23:02
kamalmok, anyway -- I'm unstuck -- thanks for the help folks :-)23:04
=== achiang is now known as achiang`
=== achiang` is now known as achiang

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