/srv/irclogs.ubuntu.com/2010/09/28/#ubuntu-kernel.txt

=== ivoks is now known as ivoks-afk
bullgard4Does the following relation hold? 2.6.35-22-generic = linux+v2.6.35.606:09
RAOFbullgard4: No; I don't even think the weaker relation 2.6.35-22-generic >= linux+v2.6.35.6 holds.06:17
bullgard4RAOF: So what linx kernel number comes next to Ubuntu's kernel version 2.6.35-22-generic?06:30
bullgard4s/linx/linux/06:30
ikepanhcbullgard4: IIRC .35-22 kernel based on 2.6.35.406:38
ikepanhcmaybe "based on" is not the best word..06:40
bullgard4ikepanhc: Thank you very much for your help.06:40
RAOFYou could also use git to determine that.  Although my git is currently taking the opportunity to garbage collect.06:40
lucentI haven't written the list or filed a bug report yet, I'm still confused about where my problems are coming from, with firewire and the new firewire stack07:22
=== ericm_ is now known as ericm|ubuntu
=== jjohansen is now known as jj-afk
=== smb` is now known as smb
=== smb is now known as smb`
=== smb` is now known as smb
=== smb is now known as smb`
sander^workSeems like someone deleted a file in yours ppa repository: http://pastebin.com/e8RRyxUW getting: W: Failed to fetch http://ppa.launchpad.net/kernel-ppa/ppa/ubuntu/dists/karmic/main/binary-amd64/Packages.gz  403  Forbidden09:20
sander^workAre anyone able to fix it?.. or know who can fix it?09:21
smb`hm, karmic. there should be nothing newer than proposed for that anyways...09:23
smb`apw, Did you do some cleanups yesterday09:23
smb`sander^work, Give me a bit. I will check on that09:25
smb`sander^work, Actually I don't think the generic PPA there ever had Karmic packages... What are you looking for?09:30
sander^worksmb`, W: Failed to fetch http://ppa.launchpad.net/kernel-ppa/ppa/ubuntu/dists/lucid/main/binary-amd64/Packages.gz  403  Forbidden09:35
sander^worksmb`, I tested on karmic instead of lucid.. but same thing happens.09:36
smb`sander^work, Hm when I wget the link you just posted I succeed. Does that fail for you as well?09:38
sander^worksmb`, works to wget09:38
smb`sander^work, Do you have any proxy defined for apt?09:39
=== smb` is now known as smb
apwsmb`,  no i didn't do anything to the PPA, though yes we did talk about cleaning it up.  if the package lists were affected that isn't anything i'd have control over or access to though09:39
sander^worksmb, I think so.. How do I remove it?09:39
smbapw, Right. I thought it might be related to not having any package for that release09:40
apwsmb, sander^work, i can wget that URL no problem so its not LP09:40
smbapw, btw, clean up pre-proposed just now09:40
apwsmb, was there lots of hidden cruft in there (as rtg suggested) ?09:40
smbsander^work, it might be defined in /etc/apt/apt.conf or even encoded in the url in the sources.list09:41
smbapw, Yes, you had even to go to delete to actually see them09:41
apwheh talk about making it difficult to see ... we should switch the PPAs to a group ...09:41
apwwill try and find out how one might do that at the sprint09:41
smbRight, that makes us close enough to bend an arm of is... Though that probably is not very well advised09:42
apw:)09:42
apwbest way to get ones hand bitten off for sure09:42
smbor ripped :-P09:43
sander^worksmb, thanks! :-).. is there any reason to use a proxy?09:43
smbsander^work, Depends. Usually I use a proxy like apt-cacher-ng to have packages cached locally. So it does not need to download it all the time.09:44
apwsander^work, most of us use them to use a caching proxy to reduce our downloads when we ahve many machine09:44
apwjinx09:44
* smb tries to remember what he had to say. Or was it nothing at all09:44
sander^workWondring why it failed now then.09:44
smbThat would depend on where your proxy is and whether it is running09:45
araHello all12:02
araIn the last couple of days, after using two USB sticks in maverick, they are not recognized anymore12:02
arawhen I insert any of those two, no activity shows up in dmesg and lsusb does not show  them12:03
apwara, when you insert them in the same maverick machine ?12:18
=== ara_ is now known as ara
araJFo, morning! are you around?12:35
arawhere are the kernel people when you need them? :)12:37
ikepanhcara: I just have a test, two usb key inserted into maverick machine 'almost' the same time, they all recognized.12:39
ikepanhcara: you insert them one-by-one or in the same time?12:40
araikepanhc, they were recognized before12:40
araikepanhc, and some of them are still recognized12:41
araikepanhc, but, these two, no longer recognized12:41
aradmesg shows no activity12:41
ikepanhcara: is the output of lsusb different before and after usb key inserted?12:42
JFoara, I am :)12:42
JFosorry for the delay, was working on another machine for the moment12:43
araikepanhc, no, already tried that as well12:43
ikepanhcJFo: good morning :)12:43
JFoikepanhc, good morning :)12:43
araJFo, I am just having some USB key issues :)12:43
JFouh oh :)12:43
arathey are not recognized anymore (two of them)12:44
JFointeresting12:44
JFono new mount info?12:44
JFoafter you insert them?12:44
aranothing, no dmesg activity, no lsusb, niente12:45
aradead12:45
arakaput12:45
JFohuh12:46
araand it had happened already with two of them12:46
JFoboth the same make or different?12:46
arathe same make?12:47
JFosame manufacturer and possibly model12:47
JFosorry :)12:47
arathey are not same model12:48
araas for the manufacturer, no idea, both are marketing keys12:48
JFook12:48
JFohmmm12:48
araone is Ubuntu, if you know their maker12:49
JFoI'm not sure what we should do here.12:53
JFoapw, any thoughts about ara's USB issues?12:53
apwseems she missed my reply:12:54
apw<ara> when I insert any of those two, no activity shows up in dmesg and lsusb does not show  them12:54
apw<apw> ara, when you insert them in the same maverick machine ?12:54
apw* ara_ is now known as ara12:54
apwi assume she was bounced, and i didn't notice12:54
araapw, when they started not being recognized, they are not recognize in two differnt maverick machines12:55
apwara will you be at release sprint ?12:55
apwthey sound like dead keys, but it is odd you lost two in a day12:55
apwnothing in dmesg at all is not normal for sure12:55
araapw, if maverick is breaking keys... 12:58
apwara, well indeed, that is a possibility, though i have little idea how that could be so ... we don't have that low a level of access to them12:58
apwhow old are these keys, are they ones from our shop ?12:59
apwi use sticks pretty often with my maverick installs and have not yet had anything similar occur12:59
araone of them is the classic ubuntu 4gb one, yes, I won it in the last all hands ;-)12:59
apwcking, you use sticks endlessly .... anything like this occured on maverick for you?12:59
apwara, i have a number of those ones too, obtained somewhere or other13:00
ckingapw, not seen this issue13:05
smblsusb would be sort of second stage show up. if there is not even errors or anything in dmesg when the sticks are plugged in it seems not even the controller reacts13:10
smbara, just to confirm some other than these two work on the same usb ports, just the two show no reaction at all13:11
arasmb, yes, that's it13:13
arawell, maybe it is just a strange coincidence. both were destined to die 10 days apart :)13:15
smbThe accumulation of two in a row is a bit odd. Hm, its not the time of year with increased static discharges... 13:15
smbyeah13:15
aralunch time for me13:16
smbOne thing to try may be to put them in and press them down or up a little....13:16
smbStrange enough that helped on one of the usb wireless keys I got13:16
apwyeah that one was an oddie13:34
ograhmm13:35
ograis there any way to compile a kernel in a awy that init= doesnt work ?13:35
ogra*way13:35
* ogra has a weird device with android kernel and its neither possible to use a shellscript as /init nor can i override init with the cmdline param13:37
smbNobody could prevent someone from modifying the kernel code not to parse for init I guess13:38
ograah, damned13:38
jk-ogra: is ti running from a ramdisk? maybe try rdinit= ?13:38
jk-s/ti/it/13:38
* ogra tries13:38
apwogra, indeed, they have likely removed the option for security13:39
ogra"security" ... pfft 13:39
ckingwhere there is a will, there is a way..13:39
ogralocking out users to use the HW with their desired OS i'd say13:40
ograthey dont provinde the kernel source either13:40
ogra*provide13:40
apwhrm, now i am supprised they can do that13:40
ckingthey should provide source for sure13:41
ograyes13:41
* ogra ponders to report to http://www.gnu.org/licenses/gpl-violation.html13:41
ograbut the device is brandnew so i dont want to be harsh and give them some days13:42
ograits basically just an android kernel but there are some HW specific patches without which you cant move on13:43
tgardnerbjf, how come you uploaded Lucid mvl-dove again? as far as I can tell you only bumped the ABI, yet the previous build appeared to be OK.14:17
smbtgardner, uploaded again compared to what?14:18
smbthe last upload would include a rebase to latest master and some updates from eric14:18
tgardnersmb, he uploaded yesterday with Ubuntu-2.6.32-210.27, then again this morning with Ubuntu-2.6.32-211.2714:19
tgardnerI guess there must have been an FTBS since the the diffs jump from 209.25 to 211.2714:20
smbtgardner, I believe the 210-27 had the abi bump from a previous upload but since then the patches from eric required another bump14:20
tgardnerbjf, lemme know when your iso builder is done so I can bounce tangerine14:35
bjftgardner, got for it, looks like it ran into problems15:17
apwogasawara, /me is starting to put together the skeleton blueprints for natty, just so we don't duplicate effort15:44
ogasawaraapw: cool15:45
ogasawaraapw: once you're done, we can start picking the one's off that need some prep work (eg the delta blueprint, etc)15:45
apwogasawara, indeed so15:45
bjf##16:08
bjf## Ubuntu Kernel Team Meeting - Today @ 17:00 UTC - #ubuntu-meeting16:08
bjf##      agenda: https://wiki.ubuntu.com/KernelTeam/Meeting16:08
bjf##16:08
ogasawaraapw: your patch to reduce disk usage on buildd's, should we throw that in the day-0 kernel?16:55
apwogasawara, it doesn't really need uploading as such, as i understand things it only really impacts the PPA builders16:56
apwogasawara, so it needs to be in our tree, but not urgently in an upload16:56
ogasawaraapw: ack, I'll put it in the holding pattern then16:57
apwheh do we have a branch called that ? :)16:57
ogasawaraheh, no16:57
JFoapw, there was something we wanted to discuss in the AOB bit of today's meeting... do you remember what it was?17:05
apwJFo, there was?  hrm, where were we when we talked about it ?b17:05
JFoTaipei17:05
JFoI'm drawing a blank as well17:05
apwhelpful!17:05
JFoheh indeed17:05
JFowe were in the office17:05
JFohmm17:06
JFowonder if it was about the kernel-input tagged bugs for review17:06
* apw is still blank17:07
JFoah well17:08
JFono matter17:08
JFoif it is important17:08
JFowe will remember eventually17:08
=== ivoks-afk is now known as ivoks
* jj-afk needs to run his kids to school, back in 1517:18
jj-afkhe oops, forgot to change my nick17:18
bjf##17:39
bjf## Kernel team meeting in 20 minutes17:40
bjf##17:40
bjf##17:55
bjf## Kernel team meeting in 5 minutes17:55
bjf##17:55
=== jj-afk is now known as jjohansen
lamontapw: if you could post how to limit flavors, or drop ddebs to bug 645653, that would be very wonderful as a workaround, I expect (hope?)...18:03
ubot2Launchpad bug 645653 in linux (Ubuntu) "kernel is unbuildably large (affects: 1) (heat: 8)" [High,In progress] https://launchpad.net/bugs/64565318:03
apwlamont, the patch to fix that is on our kernel-team list18:03
lamontyeah - It's just that it'd be good to have a URL to point people at when I kill their builds that are tying up ppa builders, and the bug seems the cleanest place to point them18:05
apwlamont, ok so probabally attaching the patch is the cleanest soln then18:05
lamontta18:05
apwlamont, done18:12
avinash_hmhi, i have a program generating seg fault ... i wan't to dump the core file .. it's not getting dumped...'ulimit -c' is unlimited .. anyone noticed this issue18:20
avinash_hmi am using 10.04 ubuntu18:20
vanhoofavinash_hm: where are you dumping it, local fs?18:24
avinash_hmvanhoof, its in my home directory18:25
vanhoofavinash_hm: is it suid?18:25
avinash_hmvanhoof, no .. shall i log as root and try18:25
vanhoofavinash_hm: nah, was wondering if you needed to fiddle w/ suid_dumpable18:26
vanhoofand i've seen weird problems in the past dumping core on a nfs based fs18:26
vanhoofavinash_hm: can you reproduce this with something other than your app18:26
vanhoofsay something like this ...18:26
=== maco2 is now known as maco
vanhoofavinash_hm: http://paste.ubuntu.com/502225/18:27
avinash_hmvanhoof, let me try18:28
vanhoofmake sure you done have /proc/sys/kernel/core_pattern set to something that would dump outside of your cwd18:29
vanhoofand if its an app thats stared at boot, you should check its cwd as well18:29
vanhoofor set core_pattern to somewhere constant18:30
avinash_hmvanhoof, http://pastebin.com/XTwHxdmQ .. not generating .. core_pattern is core .. so cool on that side18:32
kamalvanhoof, avinash_hm: can I butt in?18:33
vanhoofkamal: sure18:33
kamalavinash_hm: see whether you can make anything core dump...  Try putting this file into /tmp, compiling it, and running it:18:33
avinash_hmsure18:33
kamal  /* crashy.c */  int main() { int *p=0; return *p; }18:34
kamaltry it in /tmp specifically, so as to avoid any potential issue with the directory you're working in there.18:34
kamaldoes it dump core?18:34
* vanhoof is curious what /media/windows-pc is ... and what mount options are associated w/ it18:34
kamalvanhoof: yes -- my thinking exactly ;-)18:35
avinash_hmkamal, core dumped ... 18:35
avinash_hmvanhoof, /media/windows-pc .. this is my windows drive .. its 200+ gb .. so mounted it and started accessing it18:36
kamalavinash_hm: ok, try copying the your built binary into /tmp, and run it there.  dump core?18:36
avinash_hmkamal, no .. didn't dump .. but more or less its your code .. http://pastebin.com/sjFMiEK318:38
kamalavinash_hm: when I compile that (gcc -o avi avi.c) it does dump core for me.18:40
avinash_hmkamal, u compiled as a sudoer .. ?? i logged as su and core dumped :-(18:42
kamalavinash_hm: I don't think it should matter how it was compiled18:43
kamalavinash_hm: what happens if you change your code to be more like mine?  take out the printf, and try just "return *c".   (if you can make my code dump, yours should too, lets find the difference).18:44
avinash_hmlet me try ur code in my home .. [ old dir [18:45
avinash_hmkamal, again no luck .. http://pastebin.com/dqRYbYnq .. may be somthing to do with permissions18:47
kamalavinash_hm: but you're back in that /media/windows-pc directory again!  I think *that* is the problem.  try your code in /tmp.18:47
kamalactually just try running the *binary* in /tmp18:48
kamalavinash_hm: note also that this should even work:18:50
kamal   cd /tmp18:50
kamal   /media/windows-pc/linux-home/avi-projects/learn-c-programming/AOD/crash18:50
avinash_hmkamal, yeah .. running binary .. [ mine, yours ] works in /tmp ... doesnt work in AOD18:50
kamal(that should leave you with /tmp/core)18:50
avinash_hmyep ...18:50
avinash_hmlet me go ahead with this dump generated for now ... 18:51
avinash_hmthanks kamal, vanhoof for the help ... 18:51
kamalok, so the problem is something about the mount options with that mounted directory -- I don't know how to "fix" that -- but yes, I think your best bet is to just arrange to dump core elsewhere.  :-)   glad we could help!18:51
avinash_hmpossible .. avinash@avinash-laptop:/tmp$ mount | grep windows18:52
avinash_hm/dev/sda3 on /media/windows-pc type fuseblk (rw,nosuid,nodev,allow_other,blksize=4096)18:52
avinash_hmthanks again kamal :-) .. have fun18:53
* tgardner lunches19:15
savvas0JFo: hi, I was told that you can help include a post-update regarding maverick -- is it possible to include the new carl9170 which would improve performance of some atheros based usb wireless devices? bug 54082719:42
ubot2Launchpad bug 540827 in linux (Ubuntu) (and 1 other project) "ar9170usb wifi module crash + instabile connection (affects: 16) (dups: 2) (heat: 84)" [Undecided,Confirmed] https://launchpad.net/bugs/54082719:42
JFohi savvas0 19:42
* JFo looks at the bug19:42
savvas0thank you for your time :)19:43
JFosavvas0, no problem, but let's see how much help I am :)19:43
* savvas0 crosses fingers heh19:44
JFotgardner, do you know anything about the firmware mentioned in that bug? ^^19:44
JFohmmm, he is at lunch19:45
JFosavvas0, let me check with tgardner when he returns to see what has been done/needs to be done19:45
JFosound good?19:45
savvas0sure, works for me -- if not for maverick, then for ubuntu+1+1 :)19:46
savvas0thanks again19:46
JFomy pleasure19:46
JFoI'll ping you once he gets back in case we need to discuss :)19:47
savvas0alright, I'll stick around then19:48
hackeronanyone able to get linux-crashdump working on maverick? - I installed it, rebooted, induced a panic but can't see anything in /var/crash - any ideas?19:54
* ogasawara lunch20:07
JFohackeron, I vaguely recall some conversation around that20:11
JFobut I don't recall specifics20:11
* jjohansen -> lunch20:17
hackeronJFo: heh, hmmmmm20:19
=== ivoks is now known as ivoks-afk
JFohackeron, if I had to guess I'd say there was a reason we don't have it in by default, but I want to say we had some wiki info on it20:21
JFoI am looking20:21
hackeronJFo: thanks20:22
JFohackeron, my pleasure20:22
smoserDaviey, what 'apt-get' do i need ?20:22
smoseroops20:22
smoserwrong window20:22
JFoheh20:23
JFono sweat smoser 20:23
JFoapw, if you are still around, do you recall the kernel-crashdump conversation?20:24
JFoI can't seem to find a page on it hackeron 20:24
JFostill looking20:24
JFohmmm, may have found some stuff hackeron 20:25
JFoone sec...20:25
* JFo reads it20:25
JFoyou saw this already hackeron? https://wiki.ubuntu.com/Kernel/CrashdumpRecipe20:26
JFonot really much there20:26
hackeronJFo: yeh, doesn't work on maverick :(20:26
JFo:-(20:27
JFoeverything else I see is old20:27
JFoogasawara, any memory about kernel-crashdump?20:28
JFooh, she is probably at lunch20:28
JFohmmm20:28
* JFo digs further20:29
JFoso hackeron you get E:Unable to locate package kernel-crashdump also?20:33
JFothat is what I see on my Mav machine20:33
=== ivoks-afk is now known as ivoks
scrllockdoes anyone know which kernel versions include pv-on-hvm drivers?21:06
ogasawaraJFo: I haven't looked at crashdump for a few releases so probably can't add anything of use21:15
JFoogasawara, cool21:15
JFothanks21:15
JFoscrllock, I do not21:15
JFosorry :-/21:15
scrllocki'm reading stuff (meeting notes, mainly) indicating that jjohansen was looking into putting something into a PPA, but i haven't found anything other than that21:16
apwi don't recall anything specific occuring with those, once the pv-ops kernel started working right21:17
apwwern't they a fall back?21:17
scrllockafaik, it lets you set up a hvm guest with paravirtualized i/o drivers, for performance reasons21:17
apwthey may know more specifically on #ubuntu-server, as they are the main drivers of the server side functionality such as that21:20
scrllockthank you, i'll look there21:20
hackeronJFo: no, it installs, just nothing appears in /var/crash when I induce a panic21:25
JFohmmm21:25
JFowonder why I am having issues21:25
JFok hackeron thanks for the update. I'll keep looking21:25
jjohansenscrllock: the pv-on-hvm drivers have been on hold behind maverick items21:32
jjohansenscrllock: the ppa should be up in the next couple weeks21:34
scrllockjjohansen: will that be announced somewhere?21:41
jjohansenscrllock: yeah, server team list, and meeting21:42
scrllockjjohansen: excellent, thank you very much, i appreciate it21:43
=== lifeless_ is now known as lifeless
hackeronJFo: any luck?21:59
JFohackeron, sadly no22:00
JFogot puled of on something else22:00
JFobut I still don't see anything22:00
JFojjohansen, any memories about kernel-crashdump?22:00
JFoI recall a recent conversation22:01
JFobut I haven't found anything on it yet22:01
JFoand I seem to be unable to install it in the normal way22:01
JFoon Mav22:01
jjohansenJFo: paging ...22:01
jjohansenhrmm, no nothing atm22:03
JFok22:03
JFothanks 22:03
jjohansenJFo: what do you mean it want install in the normal way22:03
JFohackeron, may have to ask tgardner tomorrow22:03
JFojjohansen, from what I found it should be an apt-get22:04
JFobut it fails22:04
JFoworks on Lucid22:04
JFohackeron, has it installing on Maverick22:04
jjohansenhow does it fail?  What is the message?22:04
JFobut not leaving the crashdump22:04
jjohansenhrmmm strange22:04
hackeronJFo: yep22:04
JFo E:Unable to locate package kernel-crashdump22:05
JFothat from my maverick box22:05
JFowonder if the package just got overlooked22:05
JFoor smth22:05
JFowell, that can't be right22:06
JFoif hackeron got it to install on Maverick22:06
jjohansenJFo: so you are getting E:Unable to locate package kernel-crashdump22:07
JFoyep22:08
jjohansenhrmm, did you do an update before trying to install22:08
jjohansensudo apt-get update22:08
hackeronJFo: I'm using deb http://gb.archive.ubuntu.com/ubuntu/ maverick including partner, universe and multiverse22:08
jjohansenI just installed it no problems22:08
JFoah22:08
JFowait a tick22:08
hackeronmaybe it's in universe or something?22:09
JFoI R Dumb22:09
JFoI was trying to do kernel-crashdump22:09
JFoit is linux-crashdump22:09
jjohansen:)22:10
JFoI actually called it properly from my Lucid machine22:10
* JFo headdesks22:10
JFok, that installed22:10
JFonow to why it didn't do the crashdump properly22:11
JFoI've no clue22:11
jjohansenJFo, hackeron: yama maybe22:16
JFoyamo be there?22:17
JFo:)22:17
jjohansengive me a sec I am trying to remember the yama controls22:17
JFok22:18
jjohansenJFo, hackeron: try setting yama ptrace scope22:20
jjohansencat /proc/sys/kernel/yama/ptrace_scope 22:21
JFointeresting22:22
JFoI get 1 on Maverick22:22
JFobut on Lucid:22:22
JFojfo@nightcrawler:~/bzr/kernel-arsenal/contrib/linux$ cat /proc/sys/kernel/yama/ptrace_scope22:22
JFocat: /proc/sys/kernel/yama/ptrace_scope: No such file or directory22:22
jjohansenJFo: yama is new in maverick22:23
JFoI see22:23
* JFo learns22:23
jjohansenit adds some basic system wide restriction on hardlinks, symlinks and ptrace22:23
jjohansen1 is restricting to child only ptraces22:24
jjohansenset it to 0 for classic ptrace behavior22:24
JFok22:31
* JFo finishes up for the day22:34
JFothanks for the information jjohansen :)22:34
jjohansennight JFo22:34
JFonight jjohansen 22:34
=== yofel_ is now known as yofel

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