/srv/irclogs.ubuntu.com/2010/06/10/#ubuntu-kernel.txt

=== sconklin is now known as sconklin-gone
=== bjf[afk] is now known as bjf
bjfmanjo, i've looked at uck, what about it?00:43
=== bjf is now known as bjf[afk]
* apw yawns08:32
cooloneyapw: morning08:34
apwmorning08:34
=== lag_ is now known as lag
jk-brb, rewiring house09:34
RAOFAnd you'll be right back? :)09:36
jk-back09:42
krautmoin09:51
apwikepanhc, hey you about ?11:05
ikepanhcapw: yes11:05
ikepanhcapw: anything?11:05
apwikepanhc, is the current version of the karmic netbook branch well current11:05
apwikepanhc, in order to apply debian commonisation to it i need to 1) modify it and 2) rebase it to the current proposed kernel11:06
apwikepanhc, so i want to be sure you've not got a modified version11:06
ikepanhcapw: I do not aware who is using the branch, can you give me a hint?11:07
apwikepanhc, the karmic netbook branch is the karmic equivalent of hardy netbook-lpia ... i thought you sort of looked after those for OEM11:07
ikepanhcapw: as I know, we have oem projects using jaunty and hardy netbook branch11:08
apwikepanhc, ok so its likely as anything a dead branch11:08
apwikepanhc, and if you don't have any interest in it i can just 'do it' ...11:08
apwthanks11:08
ikepanhcapw: can wait for US guys awake? so that I can ask them11:09
apwikepanhc, i can, though if you arn't maintaining it for them, and i am not, then i suspect they arn't using it11:10
apworigin/master        Ubuntu-2.6.31-22.6111:10
apworigin/netbook       Ubuntu-2.6.31-14.4711:10
ikepanhcapw: yeah, that's what I am going to ask, if anyone need a rebase on karmic netbook-lpia branch11:10
apwoh dear, its 8 ABI bumps behind ... perhaps it really is _dead_11:10
apwikepanhc, ok thanks man if you could ask, don't stay up till midnight to find out, i can wait till tomorrow no bother11:11
ikepanhcapw: thanks, I need to finish hardy netbook branch rebase today, otherwise too much action item queue for me11:12
apwgood luck with that :)11:13
ikepanhcapw: almost, just find out I forget to update debian/control11:14
apwikepanhc, we removed those generated files from the git repository on the other branches11:14
apwthat way they are missing rather than wrong if you don't clean before building, and the source package build fails then11:15
ikepanhcapw: lrm/lum is removed, but not lbm :(11:15
apwurgle, rip them out so you don't hit it again11:16
ikepanhcapw: good idea11:16
ikepanhcapw: btw, can take some time from you reviewing the removing LGUEST_GUEST config patch? Steve gave me a green light on removing it11:17
apwikepanhc, will have a look yep11:18
ikepanhcapw: thanks11:18
apwikepanhc, done11:21
ikepanhcapw: thanks a lot..11:22
=== ericm|ubuntu is now known as ericm-afk
* apw wanders out to get some parts12:11
=== lifeless_ is now known as lifeless
apwamazing ... success13:20
nessitatgardner: good morning! I performed the Lucid server install, yesterday, and then installed the backport of the maverick kernel. Got oops-free dmesg outputs showing the e1000e issue13:51
ubot2https://lp-oops.canonical.com/oops.py/?oopsid=free13:51
nessitatgardner: both dmesg output (lucid kernel, backported kernel) are attached to the bug report (#591707)13:52
nessitaubot2: link please?13:53
ubot2Factoid 'link please?' not found13:53
nessitabummer13:53
nessitatgardner: I guess I don't need to tell that I still have no eth0 :-)13:54
tgardnernessita, ack.13:56
tgardnernessita, ok, one last test. Lets try a vanilla upstream kernel (http://kernel.ubuntu.com/~kernel-ppa/mainline/v2.6.35-rc2-maverick/linux-image-2.6.35-020635rc2-generic_2.6.35-020635rc2_amd64.deb) to make sure there isn't an Ubuntu patch causing problems.14:04
nessitatgardner: bien sur, monsier. Would that installs in lucid?14:05
tgardnernessita, yes, it should work fine in your server flavour installation14:05
nessitaok, I'll try as soon as I finish some code reviews14:06
tgardnernessita, np, thanks for your time14:06
* abogani2 waves14:38
abogani2apw: ping14:39
abogani2apw: Only for information I'm keeping up-to-date lowlatency git tree on Zinc and I'm also doing build test on my PPA: https://launchpad.net/~abogani/+archive/fix-436043/+packages14:49
nessitatgardner: test done, no eth0 with the vanilla kernel, dmesg ouput attached to https://bugs.launchpad.net/ubuntu/maverick/+source/linux/+bug/59170715:03
ubot2Launchpad bug 591707 in linux (Ubuntu Maverick) (and 1 other project) "After upgrade lucid -> maverick eth0 interface is gone (affects: 1) (dups: 1) (heat: 16)" [High,In progress]15:03
tgardnernessita, ok, I've sent an email to the e1000e Intel developers list15:10
nessitatgardner: yey!15:12
nessitatgardner: keep me posted, please. The connection through the USB dongle sucks :-/15:12
tgardnernessita, if you really need Maverick user space, its quite likely the Lucid kernel would work just fine. its worth a try.15:14
nessitatgardner: video is messed up in my maverick installation using lucid kernel15:14
tgardnernessita, I guess thats life on the bleeding edge :)15:15
* nessita loves adrenaline15:15
ogasawaraapw: that check-alias build failure is odd.  I expected the ports to fail in the latest upload, but they built just fine.15:29
tgardnerogasawara, are the tools versions changing?15:30
ogasawaratgardner: hrm, I'd have to check15:30
apwogasawara, yeah its 'dependant' on the state of the tree whether it triggers15:30
apwi've taken the liberty of direct pushing a fix for it to your tree15:30
ogasawaraapw: cool15:31
apwthat allowed me to test it on the mainline builds, which trigger it every time15:31
apwogasawara, which in better news the 35-rc2 build has built correctly for the first time15:39
=== bjf[afk] is now known as bjf
phunge0achiang: https://bugzilla.kernel.org/show_bug.cgi?id=1609016:42
phunge0achiang: https://bugzilla.kernel.org/show_bug.cgi?id=1616116:42
ubot2bugzilla.kernel.org bug 16090 in PCI "sysfs: cannot create duplicate filename" [Normal,New]16:42
ubot2bugzilla.kernel.org bug 16161 in Video(Other) "[2.6.35-rc1 regression] sysfs: cannot create duplicate filename ... XVR-600 related?" [Normal,New]16:42
achiangphunge0: ugh, thanks for reminding me. i've been busy lately. :(16:43
phunge0np, understood16:44
achiangphunge0: the issue, which i haven't figured out a nice way to fix yet, is that pci_sysfs_init() calls pci_create_sysfs_dev_files(), which then calls pci_create_slot_links()16:46
achiangphunge0: preventing pci_create_slot_links() from being called during init would fix it, but i wonder why some of the other things like pci_create_capabilities_sysfs() seem to be fine16:47
achiangphunge0: if you want to explore this area, i'd be happy to give you pointers. :)16:47
phunge0achiang: i can take a quick look16:49
phunge0achiang: but given that 2.6.35 is -rc2 people are starting to notice16:49
achiangnod16:50
tgardnernessita, inbound email re: e1000e16:50
nessitatgardner: ok... I'll check. I hate flashing the BIOS :-(16:52
tgardnernessita, yeah, its always a bit risky.16:52
achiangphunge0: i just asked jesse to revert it. i'll try and find some time in the next merge window to get it working properly, i guess16:52
achiangphunge0: especially since i have a reproducer16:53
achiangphunge0: thanks for your help16:53
phunge0achiang: np, makes sense to me16:53
nessitatgardner: other than updating the BIOS, there is nothing else I can try?16:53
tgardnernessita, so far thats the only suggestion that I have from Intel16:53
achiangphunge0: if you want to keep an eye out for the revert and help manage the bugzillas though, that would be awesome...16:54
nessitatgardner: ok, thank you, I'll keep you posted16:54
tgardnernessita, if the BIOS upgrade doesn't fix it, then I'll have to instrument the driver and give you a custom kernel 16:54
nessitaright...16:55
phunge0achiang: sure, i keep an eye on lkml... though i'm not comfortable doing the revert request myself :)16:55
achiangphunge0: oh, i just pinged jesse on irc, so that part's taken care of.16:55
achiangphunge0: when it hits linus's tree, if you could close out the bz's that would be lovely16:56
achiangor at least remind me and i can go do it16:56
=== tgardner is now known as tgardner-afk
jjohansenapw: is there a per flavour way of overriding the config enforcer17:39
=== pgraner is now known as pgraner-afk
apwjjohansen, not currently it would be possible to add pretty easily i suspect ... wahts your usecase ?17:56
jjohansenwell one of the configs I need if I can get this stupid thing to work is CONFIG_SYSFS_DEPRECATED_V217:57
jjohansenwhich currently is enforced off17:58
apwso for a branch you you can just edit the file and remove it18:01
apwif its coming back to master, we'll need to add support for flavours in the file18:01
apwjjohansen, i lied, seems i do have a flavour pred18:02
apwso you should be able to say18:03
apw(flavour ec2; value CONFIG_SYSFS_DEPRECATED_V2 y) | value CONFIG_SYSFS_DEPRECATED_V2 n18:04
apwor similar18:04
jjohansenokay, thanks18:05
voslhi18:22
* bjf is having email problems and has no x on laptop after this a.m. update... been trying to get to a setup where I can get something done all a.m.18:33
jjohansenouch18:34
* jjohansen has been avoiding updates since robbies email18:34
bjfbut my btrfs is *FABULOUS* :-)18:35
sconklinrobbie's email to which list?18:36
voslhello18:37
bjfsconklin, i'd tell you which list but I can't get to my email18:40
voslguys i just wanted some help on your kernel version numbering system18:40
voslhow does it usually wwork18:40
vosli read the wiki but couldnt make sense of it18:40
bjfvosl, what is your question?18:41
vosli assume for debian and ubuntu its the same kinda versioning18:41
voslliunx kernel version number18:41
voslhow is the numbering done?18:41
bjfvosl, via 'uname -r' = 2.6.32-22-generic18:43
bjfvosl, we have made 22 abi bumps on a 2.6.32 kernel18:43
voslare you talking in general or this is specific to ubuntu kernels?18:43
bjfvosl, this is a ubuntu lucid install18:44
voslok how does the numbering work?18:45
vosl2.6.32-2218:45
bjfvosl, it is telling you that it is based on the 2.6.32 kernel from upstream (linus' tree)18:46
voslah18:47
voslso this is the same case with vanilla too?18:47
voslthe versioning doesnt differ per distro does it?18:47
dupondjehttp://bugzilla.intellinuxwireless.org/show_bug.cgi?id=2208 / https://patchwork.kernel.org/patch/103568/18:48
ubot2bugzilla.intellinuxwireless.org bug 2208 in others "Driver hang while running auto test cases" [Critical,Verified: fixed]18:48
dupondjeis this something we can get into maverick kernel bit earlier ?18:48
bjfvosl, i don't know how other distros report versioning but it's probably similar18:48
voslah18:48
voslwhat does the -22 stand for?18:48
bjfvosl, as I said previously, there have been 22 abi bumps18:49
voslooh ok18:49
=== bjf is now known as bjf[afk]
voslbjf[afk]: which is better18:51
vosl.30 or a .30.1?18:51
vosldepends on which type of patches are allowed for each versions18:51
voslright?18:51
voslnew features, regressions or bugfixes only18:51
ckingmanjo, you around?18:59
manjocking, yes18:59
ckingok to mumble for 5 mins?18:59
manjosure 19:00
dupondjeWhen will maverick kernel be rebased on the upstream kernel ?19:00
dupondjethe bug seems fixed in it today http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commitdiff;h=63a07cb64ccc3ceae619d3298545d602ab5ecd38;hp=b95a56809343fb727c818ad1b9da14a17fa92ef619:01
ogasawaradupondje: it's already been rebased to 2.6.35-rc2.19:01
dupondjewill it get rebased to rc3 ?19:02
dupondjecause that bug is killing me :)19:02
ogasawaradupondje: yes, when it's released upstream we'll rebase again19:02
ogasawaradupondje: wash, rinse, repeat for the entire 2.6.35 cycle19:02
dupondje:)19:02
dupondjeis there a chance to get the patch in kernel faster, or you say, wait for rebase ? :)19:03
ogasawaradupondje: seems we should just wait for the rebase as it'll be in -rc319:03
ogasawaradupondje: and rc3 is just around the corner19:04
dupondjelets hope so. its crashing daily atm :( anyway its good to see its in the main git now, so it will be picked up for sure now19:04
ogasawaradupondje:  as an interim solution you could use the latest mainline daily build19:05
ogasawaradupondje: well you may need to wait till tomorrow for the build to include the patch that landed today19:05
dupondjei'll grab it tomorrow19:05
=== tgardner-afk is now known as tgardner
tgardnerkees, what does dumpable mean wrt to ptrace?19:50
keestgardner: it means core-dumpable.  as in, "has this process transitioned uid?"19:51
keestgardner: you can't ptrace a process that started setuid, e.g.19:51
keestgardner: but this flag can be set by a process itself using prctl.19:52
keestgardner: this is what well-behaved processes that handle sensitive memory do (e.g. ssh-agent)19:52
tgardnerkees, I was just reviewing your email that includes the 10-ptrace.conf verbage. It occurred to me that dumpable didn't mean anything to me.19:52
keestgardner: see "man prctl" and search for PR_SET_DUMPABLE19:52
keestgardner: I was trying to be as accurate as possible while hopefully not being too opaque.  :P19:53
=== bjf[afk] is now known as bjf
jjohansen->Lunch20:21
MTecknologyjjohansen: enjoy20:22
Laibschmanjo: please be more careful when closing bugs20:22
MaximLevitskywhat is the policy on EXPEREMENTAL in the ubuntu kernel?20:22
manjoLaibsch, which one ?20:22
Laibschyou should have received mail about it20:23
Laibschplease be careful with any bug you close20:23
manjoLaibsch, checking 20:23
MaximLevitskyI have 2 drivers in 2.6.3520:23
Laibschyou seem to be new at this20:23
rippsWhat's the likelyhood of an updated wacom driver getting into the kernel? I'm constantly rebuilding the module because the default one doesn't work with my wacom bamboo ctl-46020:23
MaximLevitskyone of which I labeled experemental20:23
MaximLevitskynow I say it is more or less well tested20:24
manjoLaibsch, bug# ?20:24
dupondjeIs there a way I can build my own daily ? really need the patch that got in today :)20:24
MaximLevitskyit still contains a disclaimer about how it can eat your cat & dog though :-)20:24
MaximLevitskyI had no reports of this though :-)20:25
ogasawaramjg59: hi, we've been carrying an Ubuntu kernel SAUCE patch that was originally authored by yourself years ago.  I wanted to get your feedback on it and am curious if you prefer I use your redhat.com email address or a different one?20:25
MTecknologyTransaction rate:        7.31 trans/sec20:25
MTecknologyOUCH20:25
MTecknologysorry... wrong channel20:25
mjg59ogasawara: Which one is it?20:25
MaximLevitskymjg59: hi20:25
manjoLaibsch, if you can't tell me what bug number it is I can't help 20:26
Laibschmanjo: bug 52736120:26
ubot2Launchpad bug 527361 in linux (Ubuntu) (and 1 other project) "hotplug interferes with ethernet card (affects: 1) (heat: 8)" [High,Triaged] https://launchpad.net/bugs/52736120:26
Laibschmanjo: is it too much to ask that you check your mail?20:26
ogasawaramjg59: the patch is "UBUNTU: SAUCE: hostap: send events on data interface as well as master interface" from back in 200720:26
Laibschgeez20:26
manjoLaibsch, sorry did not see your mail yet not in my inbox yet20:26
Laibschmanjo: just don't close bugs if you don't understand what it means, OK?20:26
MaximLevitskyHow can I ask ubuntu to enable it in marvic kernel?20:27
mjg59ogasawara: That was because network manager was only listening for association events on the data interface, and hsotap was only sending them on the master interface20:27
mjg59ogasawara: I have no idea whether that's still an issue20:27
ogasawaramjg59: I'm inclined to drop it and see who screams20:27
manjoLaibsch, I don't think that bug is SRU'able 20:28
mjg59ogasawara: Yeah, I'd go for it20:28
Laibschmanjo: please learn about LP before making any further embarassing comments20:28
ogasawaramjg59: perfect, thanks20:28
dupondjeogasawara: can I build my own daily ? Its crashing to much :(20:29
MaximLevitskyanybody ?20:29
ogasawaradupondje: sure, go for it20:29
dupondjeogasawara: is there some easy script or so for it ? or just pull from the git or ? :)20:30
ogasawaradupondje: I don't have a link for building the upstream kernel off the top of my head, but should be easily found via google20:31
dupondjethe script you guys use for the dailies isn't availible ? :)20:31
MaximLevitskymjg59: btw I am bisecting another ACPI bug20:31
ogasawaraapw: ^^?20:31
mjg59MaximLevitsky: Excellent20:32
MaximLevitskymjg59: I see that on boot the EC GPE is disabled20:32
MaximLevitskymjg59: so no notify from battery, ac, ....20:32
MaximLevitskymjg59: suspend/resume 'fixes' this. I use the acpi-test branch now. Going to reboot20:33
manjotgardner, on the phone 20:40
tgardnermanjo, ack20:40
MaximLevitskyIs it possible to enable a driver that depends on EXPREMENTAL in ubuntu kernel?20:45
MaximLevitskymy driver is CONFIG_SM_FTL (sm_ftl.ko)20:46
tgardnerMaximLevitsky, its possible. which driver(s) are you considering?20:46
MaximLevitskytgardner: ^^^20:46
tgardnerlemme look20:46
MaximLevitskythis is FTL (flash translation layer)20:46
MaximLevitskyI admit that I overdue the Kconfig option for it 20:47
tgardnerMaximLevitsky,  likely to have little impact on the general PC class of installs. Are you building for an embedded device?20:47
MaximLevitskyAll reports I recieve untill now were positive (minus typical user errors)20:47
MaximLevitskytgardner: this is for xD card reader20:47
tgardnerah, that thing.20:48
MaximLevitskyRicoh R852 xD card reader20:48
tgardnerMaximLevitsky, so, do you think it works better then what you've stated in the Kconfig description? 20:49
MaximLevitskytgardner: I think yes20:49
MaximLevitskytgardner: I didn't update it from first merge. In general especially modern xD cards (type M) should be prefectly safe because they use emulated interface20:50
MaximLevitskyIn addition to that I need new udev rule20:51
MaximLevitskyto load the FTL20:51
MaximLevitskyand to tell devicekit to treat the device normally20:52
tgardnerMaximLevitsky, I was just gonna ask what loads the driver20:52
MaximLevitskyconverting whole mtd system to proper bus interface it too big job for me.20:53
tgardnerMaximLevitsky, are you getting any traction with the udev developers? I'm not sure I can help you get a new udev rule.20:54
MaximLevitskytgardner: I didn't yet posted anything to udev list20:54
MaximLevitskytgardner: however I also need some modifications to devicekit rules20:54
MaximLevitskythe 'udisk' I forgot20:54
tgardnerMaximLevitsky, send a patch to the kernel team list enabling the config option. I'm not opposed to it. You're on your own for components outside the kernel.20:55
MaximLevitskybecause it treats all devices as system unless they are in white list20:55
MaximLevitskytgardner: beeing lazy, whats the address :-)20:56
tgardnerkernel-team@lists.ubuntu.com20:56
MaximLevitskytgardner: thanks!20:56
MaximLevitskytgardner: last question, the debian.master/config is the place I am looking for to change the config option ?21:03
matumbahello, i guess i should report bugs with ubuntu's mainline kernel to upstream rather than to launchpad?21:05
MaximLevitskymatumba: its is always better to report to upstream21:05
matumbaMaximLevitsky, well then... thx ;-)21:08
rippsWhat's the likelyhood of an updated wacom driver getting into the kernel? I'm constantly rebuilding the module because the default one doesn't work with my wacom bamboo ctl-46021:08
tgardnerMaximLevitsky, yeah, change it to CONFIG_SM_FTL=m in debian.master/config/config.common.ubuntu, then run 'fakeroot debian/rules updateconfigs'21:09
manjotgardner, I am looking at SHAID bc9d24a3aeb1532fc3e234907a8b6d671f7ed68f21:10
MaximLevitskytgardner: thanks!21:10
manjowhich is the 2nd one mentioned 21:10
tgardnermanjo, so I'm not sure where you came up with bc9d24a3aeb1532fc3e234907a8b6d671f7ed68f. Its not mentioned in either the BS or LP reports, is it? Though it does seem reasonable for a stable update.21:17
tgardnerBS->BZ21:17
manjotgardner, right I looked at the git logs for 2.6 upstream and I see eeepc-laptop: check wireless hotplug events with SHAID bc9d24a3aeb1532fc3e234907a8b6d671f7ed68f by Alan Jenkins <alan-jenkins@tuffmail.co.uk> Date:   Mon Feb 22 16:03:58 2010 +000021:19
manjotgardner, the bz report mentions this patch from a private tree 21:19
manjothe last comment in BZ21:19
tgardnerah21:19
manjothat is what I have been talking all along 21:20
manjotalking about21:20
tgardnermanjo, ok, then see if you can get that one included for 2.6.32 stable21:20
tgardnermanjo, you're gonna have to do a backport of that commit since it doesn't apply cleanly.21:34
manjoah crap21:34
manjotgardner, will do thanks21:34
MaximLevitsky1I have another probably newbie question21:42
MaximLevitsky1I downloaded the marvic kernel from git://kernel.ubuntu.com/ubuntu/ubuntu-maverick.git21:42
MaximLevitsky1I do 'fakeroot debian/rules updateconfigs21:42
MaximLevitsky1dh_testdir: cannot read debian/control: No such file or directory21:42
tgardnerMaximLevitsky1, fakeroot debian/rules clean  updateconfigs21:43
MaximLevitsky1I can copy the file there, but I think I need to start some init script21:43
tgardnerMaximLevitsky1, do the clean first21:43
MaximLevitsky1tgardner: works now21:43
MaximLevitsky1tgardner: maybe add that to wiki?21:44
MaximLevitsky1didn't find that at https://wiki.ubuntu.com/Kernel/Dev21:44
tgardnerMaximLevitsky1, I think its there somewhere. we're in the process of rewriting a lot of that info21:44
MaximLevitsky1tgardner: ok, thanks21:44
MaximLevitsky1tgardner: should I now build the package or the 'updateconfigs' is enough?21:47
MaximLevitsky1I hate waiting 2 hours for that :-(21:47
tgardnerthe updateconfigs should be sufficient.21:47
MaximLevitsky1tgardner: thanks again21:48
=== sconklin is now known as sconklin-gone
MTecknologyThe only ban in here is a bot - does that still need to exist?23:04
manjoemerald not responding 23:08

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