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

=== bjf is now known as bjf[afk]
=== jibel_ is now known as jibel
=== smb` is now known as smb
smbmorning08:41
lynxmansmb: moin moin09:10
smblynxman, hey :)09:11
lynxmansmb: :)09:16
brendandcan anyone think of a good reason why iwlist wouldn't be able to scan properly where network manager can? wouldn't they use the same interfaces? (in Oneiric)09:27
brendandhttp://paste.ubuntu.com/763594/09:28
brendandapw - this is another Broadcom system, but this time the wl driver is installed09:29
brendand(from jockey-text -l) kmod:wl - Broadcom STA wireless driver (Proprietary, Enabled, In use) [auto-install]09:30
apwbrendand, often the act of connecting to and aquiring a channel affects the ability of the card to scan09:35
apwso you find with some cards that if you run ilist constantly it will show the selected channel sometimes and a fuller listing others09:35
apwas nm cycles the card through the two modes09:35
brendandapw - no connections created on this system (/etc/NetworkManager/system-connections is empty)09:36
* smb wonders whether that actually mean nm cannot scan either...09:37
brendandsmb - i thought i got a scan out of nmcli a few minutes ago, but now that's not happening either09:38
smbis by chance some rfkill actiuve?09:38
apwon my brcm based systems i see the list come and go a lot from the command line09:38
apwand nm always has a list, i suspect nm i simply caching the result for some time09:39
brendandhttp://paste.ubuntu.com/763601/09:39
brendandapw - i'll just check, i certainly didn't activate it myself09:39
brendandcrap, yes there is a hard block (i don't have physical access so no idea if the switch is actually flicked)09:40
smbThere also was once a problem of two laptop drivers fighting over rfkill on Lenovo's09:41
brendandhttp://paste.ubuntu.com/763604/09:41
smb(somehow it went away for me and I think there has been a propoer fix too)09:42
brendanddell-wifi and brcmwl-0?09:42
smbone for what the hw thinks and the other often showing bios setting09:43
brendandi guess i'll have to wait for the lab engineer to confirm the state of the switch09:45
smbcould be physical switch or hotkey09:46
brendandheh, i rebooted and now it's not hardblocked. not according to rfkill09:49
cariboumorning; Do we provide a kernel module for the Realtek 8190 wireless adapter ?09:49
brendandbut still iwlist doesn't work09:50
apwcaribou, what are the PCI ids for it09:50
caribouapw: hold on ...09:50
brendandand nmcli now does09:50
brendand(i have no idea how dell-wifi got hardblocked before)09:50
caribouapw: 0a:02.0 Network controller [0280]: Realtek Semiconductor Co., Ltd. Device [10ec:8190]09:50
caribouapw: Subsystem: Realtek Semiconductor Co., Ltd. Device [10ec:8190] 09:50
apwcaribou, sounding like it may be in a staging driver, at least in precise, no idea if it works, being realtek your milage may vary09:56
apwseems to have some support in oneiric too, but ... quality, from a realtek driver ... erm09:57
apwyet to be proven, you would need to test for sure09:57
caribouapw: yes, this is what I read from a few LP bugs.09:57
caribouapw: ok, thanks for checking09:57
caribouapw: may I ask what is a "staging driver" ?09:58
caribouand expect an answer ;-)09:58
apwnew drivers which are too crap to be accepted into the kernel as they are09:58
caribouapw: ok, thanks. That's how I understood it09:59
apwthat may be functionally, stylistically, structually09:59
ogra_philosophically :)09:59
brendandapw - so if iwlist isn't working and network-manager is, does anyone care?10:01
apwbrendand, i think i heard yesterday that iwlist was using an old interface which was going to be deprecated10:01
apwand that we should demote the iw* commands10:01
brendandapw - i didn't expect it to be the case in Oneiric though. and that too on only one system10:02
apwbrendand, on O am not sure 10:02
brendandor one network card as it were (BCM4322 802.11a/b/g/n)10:02
brendandtgardner said it could break in P anytime soon10:02
brendandis network manager the only authoritative tool, or what's replacing iw tools?10:03
* apw would have to defer to tim, he mentioned it yesterday for the first tim to my ears10:05
brendandone last question before i go. who do i talk to about getting an item on the agenda for the kernel team meeting?10:14
smbbrendand, joe (the one that sends out the meeting notes) or you just raise your hand when there is the question about open discussion10:18
dupondje_sforshee: thanks for replying to the bug.10:21
=== dupondje_ is now known as dupondje
apwppisati, fyi that 3.2 ti-omap4 kernel is in the archive now, it'd better work :)11:24
ogra_apw, yeah, else the masses of pand users will start compleining 11:25
* apw prepares for both of them to complain at once11:27
ogra_i'll just give them paolos address ;)11:27
apwogra_, heh good plan11:34
ppisatiit boots here :)11:40
ppisatiand it drives my lcd, so it's good11:40
ppisatirelease it! :)11:41
apwppisati, oh i already did ... :) ... meta will be in the archive shortly, and the screaming will begin11:44
apwppisati, ok all published, let the fun begine12:19
ppisatiapw: cool, just found a problem12:45
apwppisati, bad ?12:51
ogra_campfire on the SoC surface ? :)12:52
ppisati:)12:52
ppisatinope12:52
ppisatihangs on reboot in there's a usb hdd mounted (musb problem)12:52
ppisatididn't notice cause i was running off the sd card12:52
ppisatii'm testing a workaround now12:52
ppisatiand perhaps i found the guilty commit12:53
smbcrashtime...13:15
* ppisati -> lunch13:15
=== yofel_ is now known as yofel
* ogasawara back in 2015:22
ckingit's getting really stormy here, if I dropoff line it's 'cos my power has gone15:26
ppisaticking: i've 14C here, do you want some? :)15:30
ppisatiseems the winter is really late this year...15:31
ckingppisati, I will swap you the warmth with the UK wind + rain15:31
=== bjf[afk] is now known as bjf
hertonGrueMaster, about bug 893190, it's needing verification. I think you will want to verify it while doing qa of current maverick ti-omap4? In this case we can mark that as verification-done15:52
ubot2Launchpad bug 893190 in linux-ti-omap4 "Qa-testing failures for 2.6.35-903.27" [Undecided,Fix committed] https://launchpad.net/bugs/89319015:52
GrueMasterherton, yes I plan on it.  I have been working with the security team to get the QRT suite working on arm.  That came up as a "needs kernel fix", hence the bug report.15:53
GrueMasterAnd actually, I will test it as part of the "Regression-testing".15:55
hertonGrueMaster, ok, I'll mark that bug as verified then, so the tracking bug of current maverick ti-omap4 can go to Testing15:55
GrueMasterok15:55
tgardnerarges, mumble ?16:06
argestgardner, ahh was just about to call 16:06
argeslet me unmute16:06
argestgardner, ok hold on getting mic16:06
argestgardner, https://bugs.launchpad.net/ubuntu/+source/network-manager/+bug/83625016:08
ubot2Launchpad bug 836250 in linux "[Oneiric] [Regression] Intel Corporation Centrino Ultimate-N 6300 poor networking, packet loss and very slow Lenovo X201 and T500 laptops" [Critical,In progress]16:09
argeshttps://bugs.launchpad.net/bugs/87614716:09
ubot2Launchpad bug 876147 in ubuntu "Intel Corporation Centrino Wireless-N + WiMAX 6150 (rev 67) not working after upgrade to 11.10 (dup-of: 836250)" [Critical,Confirmed]16:09
ogasawaraarges: just fyi, I've pinged our intel contact about bug 836250.  they were unfortunately unable to reproduce.  I've got jsalisbury working with the original bug reporter to narrow down the regression window.16:14
ubot2Launchpad bug 836250 in linux "[Oneiric] [Regression] Intel Corporation Centrino Ultimate-N 6300 poor networking, packet loss and very slow Lenovo X201 and T500 laptops" [Critical,In progress] https://launchpad.net/bugs/83625016:14
argesogasawara, yea i've been speaking with him too16:17
brendandrtg_ - hi16:59
rtg_brendand, dude16:59
brendandrtg_ - quick question, how reliable is iperf going to be in a high noise environment?17:00
brendandrtg_ - say i have 3-4 systems hitting the iperf server simultaneously?17:00
rtg_brendand, its as reliable as the underlying TCP connection.17:00
rtg_I presume the iperf server is at the end of an ethernet cable ?17:01
brendandrtg_ - it will be. right now it's not because i'm testing it in my home office17:01
rtg_brendand, so I'm not sure what you mean by reliable.17:02
brendandrtg_-  in terms of the bandwidth figures given by iperf17:28
brendandrtg_-  in my current setup i get a figure of between 5 and 8 Mbits/sec17:29
rtg_brendand, with multiple simultaneous wifi clients?17:29
brendandrtg_-  would that go down significantly given a few systems trying to hit the same server?17:30
brendandrtg_-  this is with just one17:30
rtg_brendand, try it with an ethernet client17:30
brendandrtg_-  the client is always going to be wireless in these tests though17:31
brendandrtg_-  the server could well be wired to the router17:31
brendandso the scenario would be 1+ clients connected to a router with a server wired to that router17:31
rtg_brendand, _first_ establish the stability of iperf wrt consistent bandwidth measurements.17:31
brendandrtg_-  are you suggesting to wire both systems to the router?17:32
AwsoonnHi all, Can I get some feedback on my line of thought in bug 901770 plase? Thank you~17:33
ubot2Launchpad bug 901770 in linux "Kernel Bug in bluetoothd?" [Undecided,New] https://launchpad.net/bugs/90177017:33
rtg_brendand, not to the router, but to the switch so that both are in the same subnet and have not extraneous HW between them.17:33
rtg_s/not/no/17:34
brendandrtg_-  right, i'll try that if i can17:35
brendandrtg_-  are there any other ways we could detect these wireless issues?17:36
rtg_brendand, there will be connect/disconnect messages in dmesg if you've a troubled connection.17:37
brendandrtg_- i only wish i had access to a system troubled by problems like this17:37
brendandrtg_-  what phrases might i look for?17:39
rtg_brendand, yank the plug on your AP and look in your dmesg17:39
brendandrtg_-  ok. i was wondering if that would be a realistic test. i'll try it17:40
rtg_brendand, there is also prolly some dbus traffic that goes to NM when a connection fails.17:40
brendandi'll see if i can try both those things. thanks17:40
smosersmb, ping18:01
smoserhttps://bugs.launchpad.net/ubuntu/+source/linux/+bug/87695618:01
ubot2Launchpad bug 876956 in linux "starting interactive atop causes kernel alignment check" [Undecided,Incomplete]18:01
Awsoonnanother newb question, If I'm tracking down a kernel bug, which version of the kernel source shoudl I be making modifications to? The latest Linus release, or the latest ubuntu release, or something else altogether?18:11
bjfAwsoonn: if you intend to submit a patch to upstream, you should be moding Linus' tree18:14
Awsoonnbjf: Thanks! Will I be able to use the latest kernel with 11.10 without much issue? Or is it luck of the draw?18:16
bjfAwsoonn: you should be able to use it18:16
bjfAwsoonn: note, the current upstream kernel is currently at rc4 meaning it could be problematic18:17
bjfAwsoonn: it's probably safest to start with our Oniric sources, determine your patch then port it to Linus' tree (which shouldn't be much of a port depending on what you are doing)18:18
AwsoonnI'll be working in the bluetooth stack, do you think it would be reasonable to use the latest release rather than -RCx?18:18
bjfAwsoonn: i don't know how much churn that area of the kernel is getting18:19
Awsoonnnot much by the looks of it, last modification was on Oct 24 if I'm reading things correctly...18:19
bjfAwsoonn: then i'd do as i suggested and hack on the oneiric sources and then work with the upstream maintainer to get it upstream (they'll want patches against linus' tree)18:21
Awsoonnok, that's great! :) Now I've got someplace to start.18:22
Awsoonnok, I just did a little bit more digging and found that upstream has already changed the function I was about to modify. I'm wondering if my bug is already fixed, so is there a prefered method for installing a more recent kernel?18:27
vanhoofherton: looks like lts-backport-oneiric should be released soon, yeah?18:33
vanhoofherton: (based on the tracking bug)18:33
hertonvanhoof, it's released on updates18:33
herton-updates/-security18:33
hertonat least rmadison already shows it18:34
=== arges is now known as arges_home
bjfAwsoonn: just download the appropriate .deb for Precise and install it18:42
bjfAwsoonn: are you running amd64 or i386 on a desktop ?18:44
Awsoonnbjf: fair enough, I wondered if that would work or not, thanks! AMD6418:44
Awsoonnand it's on a laptop if that matters at all to you18:45
bjfAwsoonn: https://launchpad.net/ubuntu/precise/amd64/linux-image-3.2.0-3-generic/3.2.0-3.918:45
Awsoonndoes the version string 3.2.0-3 and no '-ubuntuX' indicate that this is RC3 and there is no Ubuntu sause added yet?18:47
* rtg_ -> lunch18:48
apwAwsoonn, https://wiki.ubuntu.com/Kernel/FAQ#Kernel.2BAC8-FAQ.2BAC8-GeneralVersionMeaning.What_does_a_specific_Ubuntu_kernel_version_number_mean.3F18:50
Awsoonnthat's quite the URL, thanks!18:50
apwyeah the wiki is a bit pants18:50
bjfAwsoonn: our kernels always come with sauce :-)19:04
brendandrtg_-  all i get in dmesg that looks useful is:19:33
brendand[ 1295.068497] cfg80211: All devices are disconnected, going to restore regulatory settings19:33
jsalisburybjf, I'll be updating bug 836250 shortly.19:36
ubot2Launchpad bug 836250 in linux "[Oneiric] [Regression] Intel Corporation Centrino Ultimate-N 6300 poor networking, packet loss and very slow Lenovo X201 and T500 laptops" [Critical,In progress] https://launchpad.net/bugs/83625019:36
vanhoofherton: perfect, thanks!19:36
bjfjsalisbury: thanks, sorry to bother you19:36
jsalisburybjf, heh, no problem.  Thanks for keeping up with the bug ;-)19:37
brendandrtg_-  it looks like that bug above should affect a few systems we have. i can test my test on that19:42
=== dereck is now known as awsoonn
awsoonnwell, the percise kernel still contains my bluetooth bug, onward to vanila-RC4 I suppose.19:45
rtg_apw, I keep getting Mainline Build v3.2-rc4 messages20:11
* cking kicks off an overnight soak test and calls it a day...20:22
chrisccoulsoni've just updated my precise machine for the first time since i switched to precise, and i can't boot the current precise kernel21:08
chrisccoulsoni just get a whole bunch of these messages flashing rapidly by on screen:21:08
chrisccoulsonhttp://paste.ubuntu.com/764247/21:08
chrisccoulsonit doesn't even get as far as plymouth21:09
chrisccoulsoni left it for a few minutes, but had to switch it off when my laptop felt like it was going to burst in to flames21:09
Awsoonnchrisccoulson: I'm no expert, but it looks like you have a HDD that is failing. :( (I could be very wrong but that's what it looks like to me)21:12
chrisccoulsonAwsoonn, how come it only fails with the latest kernel? i can boot and use 3.1 without any problems21:12
rtg_chrisccoulson, start a bug and attach your apport info. I guess you'll have to use 3.1 to do it.21:13
rtg_what device is 03:00.0 ? Is that the display ?21:13
* ogasawara lunch21:14
ogasawarachrisccoulson: bug 89407021:15
ubot2Launchpad bug 894070 in linux "IOMMU error loop early in boot" [High,Triaged] https://launchpad.net/bugs/89407021:15
chrisccoulsonrtg_, 03:00.0 CardBus bridge: Ricoh Co Ltd Device e476 (rev 02)21:16
chrisccoulsonogasawara, thanks!21:16
ogasawarachrisccoulson: there's a test kernel there, or alternatively just try booting with "intel_iommu=off"21:16
ogasawarachrisccoulson: it's on my list of bugs to get to21:16
chrisccoulsonexcellent, thanks21:16
rtg_ogasawara, good eye. I knew I'd seen something like that21:17
* ogasawara really goes to eat really quick21:17
* rtg_ -> EOD21:29
=== bladernr_ is now known as bladernr_afk
AwsoonnI followed the instructions on https://wiki.ubuntu.com/Kernel/CrashdumpRecipe but my /var/crash dir is still empty, even after the sysrq test. Why might this be happening?22:30
Awsoonn_How can I get a crash dump or a stack trace if wiki:Kernel/CrashDumpRecipe leaves me with an empty /var/crash dir?22:59
=== bjf is now known as bjf[afk]
hggdhshould installation of linux-image-extra-virtual kick update-initramfs?23:32

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