1133310850 M * Bertl well, back now ... 1133311413 J * iprone ~iprone@adsl-065-012-167-027.sip.asm.bellsouth.net 1133312977 J * roeb[] ~roeb@dslb-084-059-192-069.pools.arcor-ip.net 1133313243 M * Bertl welcome roeb[]! 1133314058 M * bragon night Bertl :) 1133314072 M * bragon learning C langage is very hard 1133314095 M * bragon +alone 1133315234 M * Bertl hmm ... 1133315344 M * Bertl bragon: you already know a programming language? 1133315380 M * bragon good night Bertl 1133315387 M * bragon it's to late for me 1133315389 M * bragon sorry 1133315392 M * bragon see u latter 1133315397 M * Bertl k, night! 1133315773 M * Bertl okay, off to bed now ... cya 1133315790 N * Bertl Bertl_zZ 1133318112 Q * Wolfwood^DC Read error: Connection reset by peer 1133320534 J * sebi_ ~sebi@Fcc25.f.strato-dslnet.de 1133320626 J * ryker ~ryker@c-67-176-240-74.hsd1.in.comcast.net 1133320641 Q * sebi Ping timeout: 480 seconds 1133320914 J * Wolfwood nicholasd@d60-65-131-235.col.wideopenwest.com 1133323217 J * menomc ~amery@200.75.27.89 1133323326 Q * mnemoc Ping timeout: 480 seconds 1133323326 N * menomc mnemoc 1133323704 J * stefani ~stefani@c-24-19-46-211.hsd1.wa.comcast.net 1133325035 Q * ryker Quit: Leaving 1133331180 Q * iprone Quit: Leaving 1133331407 P * stefani parting (is such sweet sorrow) 1133335517 N * Bertl_zZ Bertl 1133335524 M * Bertl morning folks! 1133335529 M * daniel_hozac morning! 1133335569 M * Bertl ah, daniel! how are you? 1133335584 M * daniel_hozac fine. how's princeton? 1133335661 M * Bertl pretty nice ... and somewhat chaotic :) 1133335694 M * Bertl i.e. works for me :) 1133335706 M * daniel_hozac hehe 1133335760 M * Bertl well, I spent most of the time with fixing hardware and getting permissions (parking, room access, network ...) 1133335828 M * Bertl compared to the university location, the hotel network access was pretty painless *G* 1133335872 M * daniel_hozac lol 1133335915 M * Hollow morning Bertl 1133335934 M * Bertl hey Hollow! everything fine? 1133335961 M * Hollow yeah, except i should write my physics test atm, but i have absolutely no idea ;) 1133335979 M * Hollow how long do you stay in princeton? 1133336006 M * Bertl till 17th of december, so three weeks total 1133336052 M * Hollow ic.. i read about help for ngnet..? 1133336193 M * Bertl you mean from princeton? 1133336236 M * Hollow yup 1133336286 M * Bertl yeah, well, I'm not sure about that, but there will be definitely some input as the planetlab folks have some ngnet expectations 1133336370 M * Bertl some guy here has for example modified the scheduler (hard cpu) to support guarantees and do work preserving fair scheduling 1133336421 M * Bertl will try to get some structure into the different effords and cherry pick stuff/ideas for mainline 1133336435 M * Hollow i don't understand both of them, but sounds good :P 1133336505 M * Bertl Hollow: so what's the problem with physics? 1133336514 M * Hollow meanwhile i finally have a working chbind replacement now, and therefore added the persistant feature to network contexts, and fixed a bug in /proc/virtnet 1133336538 M * Hollow it's about astronomy, and i simply didn't do anything.. ;) 1133336541 M * Bertl ah, great, that reminds me to finally look at your patch ... 1133336550 M * Hollow so, to save my abitur i called the doctor :P 1133336573 M * Bertl Hollow: you know: "cheaters will never win"? 1133336606 M * Bertl (but maybe this can be considered a timeout) 1133336620 M * Hollow oh, school is not about won or lost... it's about getting the fckn abitur, not how you got it :P 1133336649 M * Bertl hmm, hmm ... well, guess I disagree here ... 1133336678 M * Hollow and you're probably right, but i'm not a friend of school, really. 1133336783 M * Hollow anyway.. the dynamic id patch is borked, kernel bug if you're trying to create a dynamic context id.. rest should be fine 1133336841 M * Bertl non scolae sed vitae discimus ... 1133336864 M * Hollow too bad, i only had french :P 1133336889 M * Bertl hehe, means: "Not for school, but for life we learn" 1133336903 M * Hollow sure we do, but not in school, haha 1133336922 M * Bertl well, will look into it, should be on 13thfloor? 1133336944 M * Hollow my patches? 1133336960 M * Bertl ah, no, sorry 1133336978 M * Bertl do you have the url at hand? 1133336991 M * Hollow http://home.xnull.de/work/vserver/patches/ 1133336995 M * Bertl tx 1133337102 M * Bertl hmm, we probably want to extend the allowed range too (i.e. give the full range of 2-65535) 1133337142 M * Hollow yeah, would be great 1133337572 M * Hollow Bertl: what's your timeline for the 2.1.0 release? 1133337615 M * Bertl asap .. probably before 2.6.15 1133337657 M * Bertl similar for 2.0.1 btw 1133337708 M * Hollow ic.. i'd like to release a dev version of the new util-vserver for 2.1.0, since all low-level tools are implemented now (iirc only vrsetup is missing) 1133337732 M * Hollow i'll also try to add a build command till then 1133337750 M * Bertl okay, any plans regarding naming? 1133337774 M * Hollow naming of the package? 1133337782 M * Bertl (to avoid clashes and confusion with util-vserver) 1133337809 M * Loki|muh_ vserver-utils 1133337809 M * Hollow dunno.. i had vserver-utils some time ago.. 1133337811 A * Loki|muh_ runs 1133337835 M * Loki|muh_ what about vserver-userspace? 1133337842 M * Hollow god, no :P 1133337851 N * Loki|muh_ Loki|muh 1133337865 M * Loki|muh vserver-tools 1133337866 M * Hollow it reminds me of usermode-utils ;) 1133337879 M * Hollow there was apckage with that name already.. 1133337885 M * Hollow (vserver-tools) 1133337902 M * Loki|muh :( 1133337911 M * Hollow you see, it's not that easy ;) 1133337916 M * Bertl Hollow: which one? 1133337927 M * Hollow vserver-tools iirc 1133337937 M * Loki|muh tha-real-vserver-utils[tm] 1133337948 M * Hollow *g* 1133337987 M * Hollow well, vserver-utils would be ok for me 1133337991 M * Bertl hmm, well, probably better than hollow-tools :) 1133338021 M * Hollow lol 1133338089 M * Bertl ah, well, not my problem, just wanted to point it out 1133338101 M * daniel_hozac hollows-stuff ;) 1133338113 M * Hollow heh, i don't care that much too.. i'm also fine with util-vserver *shrug* 1133338138 M * Hollow Bertl: is the persistant patch ok? 1133338158 M * Bertl well, I'd avoid that like hell, we had similar with jack's vs. enrico's tools 1133338170 M * Bertl and it was more than confusing ... 1133338191 M * Hollow ic.. vserver-utils sounds fine then.. 1133338231 Q * FireEgl Ping timeout: 480 seconds 1133338423 M * Bertl okay, will try to get some more sleep now .. cya later 1133338440 M * Hollow night Bertl 1133338442 N * Bertl Bertl_zZ 1133339210 J * FireEgl Atlantica@Atlantica.US.TO 1133341030 Q * shedi Quit: Leaving 1133343188 M * harry Bertl_zZ: you going to 22c3 too? 1133343616 J * Milf ~Miranda@ipsio352.ipsi.fraunhofer.de 1133343630 M * Milf Hello 1133343665 Q * Aiken Ping timeout: 480 seconds 1133346208 M * cemil hi there 1133346214 M * cemil how can i hard stop a vserver? :) 1133346218 M * cemil not "vserver stop foo" 1133346224 M * cemil i want to do something like "power off" 1133346586 M * daniel_hozac chcontext --xid /sbin/killall5 -9 ? 1133346955 M * cemil doesnt work 1133347164 Q * TheSeer Remote host closed the connection 1133347340 J * TheSeer ~theseer@border.office.salesemotion.net 1133347372 M * cemil with -ctx it workl 1133347760 J * shedi ~siggi@tolvudeild-197.lhi.is 1133348381 Q * Loki|muh Read error: Connection reset by peer 1133348687 J * Loki|muh loki@satanix.de 1133348995 M * roeb[] hi all 1133349711 J * krakan ~krakan@jinx.usbe.umu.se 1133350020 M * krakan when using gentoo vserver kernel i get this when tying to SSH to either the vserver root server och guest: ssh_exchange_identification: Connection closed by remote host 1133350084 M * ag- Hollow: to my mind, it isn't school which sucks, but the majority of teachers who did that job, because they weren't able to do what they really like 1133350162 M * Hollow ag-: indeed 1133351153 Q * krakan Quit: Leaving 1133352471 Q * shedi Quit: Leaving 1133352721 J * _nokoya young@hi-230-82.tm.net.org.my 1133352858 Q * Hunger jupiter.oftc.net quasar.oftc.net 1133352858 Q * dhansen jupiter.oftc.net quasar.oftc.net 1133352858 Q * nox jupiter.oftc.net quasar.oftc.net 1133352858 Q * mugwump jupiter.oftc.net quasar.oftc.net 1133352858 Q * MostlyHarmless jupiter.oftc.net quasar.oftc.net 1133352858 Q * AndrewLee jupiter.oftc.net quasar.oftc.net 1133352858 Q * jkl jupiter.oftc.net quasar.oftc.net 1133352858 Q * nokoya jupiter.oftc.net quasar.oftc.net 1133352858 Q * daniel_hozac jupiter.oftc.net quasar.oftc.net 1133352858 Q * mountie jupiter.oftc.net quasar.oftc.net 1133352858 Q * alexx jupiter.oftc.net quasar.oftc.net 1133352858 Q * bragon jupiter.oftc.net quasar.oftc.net 1133352858 Q * harry jupiter.oftc.net quasar.oftc.net 1133352858 Q * derjohn jupiter.oftc.net quasar.oftc.net 1133352858 Q * sizo jupiter.oftc.net quasar.oftc.net 1133352858 Q * FaUl jupiter.oftc.net quasar.oftc.net 1133352858 Q * DaCa jupiter.oftc.net quasar.oftc.net 1133352858 Q * SNy jupiter.oftc.net quasar.oftc.net 1133352861 N * _nokoya nokoya 1133352946 J * MostlyHarmless ~mh@melbourne.mostly-harmless.ca 1133352948 J * harry ~harry@d515321D1.access.telenet.be 1133352965 J * mugwump ~samv@202.78.240.73 1133352967 J * AndrewLee ~andrew@tnlug.linux.org.tw 1133352974 J * sizo janek@openbug.org 1133352975 J * FaUl j9QQBAFJax@verbrennung.org 1133352978 J * SNy a08c0001b5@bmx-chemnitz.de 1133352980 J * bragon ~bragon@god.geeknode.org 1133352994 J * nox ~nox@noxlux.de 1133352998 J * Hunger Hunger.hu@Hunger.hu 1133353160 J * alexx ~alexx@proxy.ikse.net 1133353168 J * mountie ~mountie@CPEdeaddeaddead-CM000a739acaa4.cpe.net.cable.rogers.com 1133353244 J * derjohn ~derjohn@80.69.35.186 1133353266 J * daniel_hozac ~daniel@c-6f1472d5.010-230-73746f22.cust.bredbandsbolaget.se 1133353482 M * cemil damn fuck 1133353489 M * cemil after sshÄ'in into a new vserver i get: 1133353509 M * cemil stdin: is not a tty 1133353901 Q * Hollow Remote host closed the connection 1133354414 A * [MUPPETS]Gonzo notiert: Cemil, derjohn... alles auf einem Haufen... 1133354886 J * Hollow ~hollow@home.xnull.de 1133355672 M * harry //dev ok? 1133355718 M * harry crw-rw-rw- 1 root root 5, 0 2005-11-26 14:01 tty 1133355734 M * harry etc 1133356280 N * Bertl_zZ Bertl 1133356286 M * Bertl morning folks! 1133356320 Q * Milf Ping timeout: 480 seconds 1133356511 M * [MUPPETS]Gonzo hi bertl 1133356624 J * click click@ti511110a080-4353.bb.online.no 1133356947 M * Bertl okay off for now .. back later ... 1133357414 M * cemil harry: defiant:~# ls -la /vservers/vls08/dev/tty 1133357414 M * cemil crw-rw-rw- 1 root tty 5, 0 2005-11-30 13:32 /vservers/vls08/dev/tty 1133357880 Q * mcp Read error: Connection reset by peer 1133357941 J * mcp ~hightower@wolk-project.de 1133358756 Q * infowolfe Remote host closed the connection 1133358759 Q * xzu Quit: leaving 1133359136 J * blackfire blackfire@nat1.radiostrada.pl 1133359168 M * blackfire hi, anybody alive? 1133359254 M * Bertl barely, what's up? 1133359448 M * Bertl okay, off then ... back later 1133359455 N * Bertl Bertl_oO 1133359535 M * blackfire ehh missed ya :/ 1133359545 M * blackfire crashes crashes crashes :/ 1133359700 Q * blackfire Quit: brb 1133359761 J * DaCa ~danny@mail.limehouse.org 1133359819 J * jkl eric@c-67-173-248-142.hsd1.co.comcast.net 1133359827 J * blackfire blackfire@dp70.internetdsl.tpnet.pl 1133359998 J * infowolfe ~infowolfe@209-112-215-110-cdsl-rb1.nwc.acsalaska.net 1133360651 Q * Loki|muh Ping timeout: 480 seconds 1133360666 Q * nox Ping timeout: 480 seconds 1133360987 J * shedi ~siggi@inferno.lhi.is 1133361039 J * nox ~nox@noxlux.de 1133361160 J * mef ~mef@targe.CS.Princeton.EDU 1133361236 J * Loki|muh loki@satanix.de 1133361830 M * blackfire any kernel hacker around here by chance? :) 1133362201 Q * tchan Remote host closed the connection 1133362236 J * tchan ~tchan@c-67-174-18-204.hsd1.il.comcast.net 1133362244 M * FaUl seems not to 1133362251 M * blackfire yep ;/ 1133363143 Q * infowolfe Remote host closed the connection 1133363370 Q * monrad Quit: Leaving 1133364335 M * harry blackfire: depends 1133364354 M * blackfire harry: are you on the linux-vserver mailing list? 1133364392 M * harry nope 1133364397 M * harry i should really do that... 1133364464 J * monrad ~monrad@213083190130.sonofon.dk 1133364502 M * blackfire ah, so you haven't followed my crusade on chasing oops? ;) 1133364515 M * blackfire i've been spamming the list for a few days now 1133364552 M * harry hehe 1133364555 M * harry nope 1133364579 M * harry i'll look @ the archives... 1133364593 M * harry but... if Bertl_oO can't solve it, i doubt if i can... 1133364633 M * blackfire it's an obvious null pointer dereference but if i insert a check just before, it still happens :( 1133364645 M * BWare heh 1133364654 M * BWare I checked netconsole last night 1133364674 M * blackfire hi BWare :) 1133364678 M * blackfire and? 1133364689 M * BWare and I don't get an oops but a dma_timeout of some sort, so I'm guessing I have a faulty disk taking down the box 1133364708 M * BWare but I don't get a lot of output from netconsole 1133364726 M * BWare perhaps I forgot to add some debugging options in the kernel 1133364731 M * blackfire does it look like the one i posted today? (nmi watchdog detecting a hang) 1133364745 A * harry now subscribed to ml 1133364750 M * BWare Not really 1133364757 M * BWare I didn't get an oops 1133364774 M * blackfire well, my second report wasn't an oops either :) 1133364782 M * BWare just a message about the dma on the /vservers disk and p00f box gone ;) 1133364802 M * FaUl BWare: sounds like a kernel-bug 1133364832 M * BWare FaUl: A vanilla kernel did not show it ;) 1133364852 M * FaUl woha, that sucks 1133364855 M * BWare But then again the vanilla kernel was older than the vserver kernel I'm using now 1133364872 M * blackfire could you check my mail from 14:20 and compare the reports? or mail your log to the list? 1133364872 M * BWare So I definately have to dig in a bit deeper 1133364896 M * blackfire yeah, well I'm knee deep in the kernel at the moment ;) 1133364926 M * BWare hehehe 1133364965 M * BWare My log did not show any dump, hence my guess I forgot to enable some debug options in the kernel 1133365000 M * BWare Anyway tonight I'll get a shovel and start digging 1133365002 M * blackfire you should get a stack trace if it was a kernel bug 1133365028 M * harry blackfire: what's your real name? (on the ml?) 1133365081 J * Viper0482 ~Viper0482@p54976974.dip.t-dialin.net 1133365110 M * blackfire harry: G. Nosek 1133365226 M * BWare blackfire: This was the log entry: Nov 29 21:17:29 src@jukebox hdg: dma_timer_expiry: dma status == 0x21 1133365233 M * BWare But that was all 1133365242 M * blackfire harry: though it may not be in the archives yet 1133365283 J * infowolfe ~infowolfe@209-112-215-110-cdsl-rb1.nwc.acsalaska.net 1133365295 M * BWare and looking at the logs it happens every crash 1133365308 M * BWare hmmm.... time for a diskchange I think 1133365316 M * blackfire hmm... /me thinks... as a wild guess, is it possible that some vserver function (vx_lookup_info maybe) is holding a spinlock for so long that it invokes varius race conditions? 1133365409 M * blackfire s/varius/various/ 1133365632 Q * Viper0482 Read error: Connection reset by peer 1133365901 M * blackfire v2144:/srv/linux-2.6# while true ; do make clean ; make -j24 ; done <---- i've just started this, i'll see how long the box lasts (v2144 is a vserver) 1133365914 J * Viper0482 ~Viper0482@p54976974.dip.t-dialin.net 1133366000 M * blackfire gone 1133366596 M * blackfire Nov 30 16:51:54 40 [42949920.940000] Unable to handle kernel NULL pointer dereference 1133366599 M * blackfire Nov 30 16:51:54 40 at virtual address 00000000 1133366601 M * blackfire Nov 30 16:51:54 40 [42949920.940000] printing eip: 1133366604 M * blackfire Nov 30 16:51:54 40 [42949920.940000] Unable to handle kernel NULL pointer dereference at virtual address 00000000 1133366607 M * blackfire Nov 30 16:51:54 40 [42949920.940000] printing eip: 1133366610 M * blackfire Nov 30 16:51:54 40 [42949920.940000] a01ae9eb 1133366612 M * blackfire Nov 30 16:51:54 40 [42949920.940000] *pde = 00000000 1133366615 M * blackfire Nov 30 16:51:54 40 [42949920.940000] Oops: 0000 [#1] 1133366617 M * blackfire Nov 30 16:51:54 40 [42949920.940000] SMP 1133366620 M * blackfire Nov 30 16:51:54 40 [42949920.940000] Modules linked in: ipt_owner ipt_state iptable_filter netconsole e100 ip_conntrack_ftp ip_conntrack 1133366623 M * blackfire Nov 30 16:51:54 40 [42949920.940000] CPU: 0 1133366625 M * blackfire Nov 30 16:51:54 40 [42949920.940000] EIP: 0060:[] Not tainted VLI 1133366628 M * blackfire Nov 30 16:51:54 40 [42949920.940000] EFLAGS: 00010247 (2.6.14.2-test) 1133366631 M * blackfire Nov 30 16:51:54 40 [42949920.940000] EIP is at do_task_stat+0x93/0x8a9 1133366633 M * blackfire Nov 30 16:51:54 40 [42949920.940000] eax: 00000000 ebx: 00000000 ecx: a05f4740 edx: a90469a0 1133366636 M * blackfire Nov 30 16:51:54 40 [42949920.940000] esi: bf9f8000 edi: b5efa550 ebp: ae3d3f20 esp: ae3d3d9c 1133366639 M * blackfire Nov 30 16:51:54 40 [42949920.940000] ds: 007b es: 007b ss: 0068 1133366642 M * blackfire Nov 30 16:51:54 40 [42949920.940000] Process pidof (pid: 14236, threadinfo=ae3d2000 task=b997a550) 1133366946 M * blackfire over and over again, in the same place... I have placed _two_ checks before dereferencing the task variable... 1133367994 J * robig ~Miranda@212.80.250.154 1133368008 M * robig good evening 1133368013 M * blackfire hi 1133368035 M * robig ive a problem with my vserver 1133368058 M * blackfire yeah, me too ;) what's the problem? 1133368061 M * robig what does " host kernel: VSW: xid=49183 messing with the devpts." in host syslog mean? 1133368216 M * robig no idea? 1133368231 M * blackfire no big deal 1133368260 N * Bertl_oO Bertl 1133368265 M * Bertl back now ... 1133368273 M * robig the problem is, that my apache in the vserver is quiet slow! 1133368286 M * robig from one minute to another 1133368307 M * blackfire check /proc/virtual//sched, it may be running out of tokens 1133368329 M * Bertl robig: the message means that a guest context tried to access a 'hidden' pty of a different guest 1133368329 M * blackfire hi Bertl :) any insight on my problems...? 1133368331 M * daniel_hozac blackfire: are you sure you're checking the right variable? 1133368365 M * robig where do i get the xid? 1133368365 M * blackfire yup, the variable is task_struct *task 1133368383 M * blackfire cat /etc/vservers//context in new layout 1133368383 M * Bertl blackfire: the dump looks like a kernel issue, which version do you use? 1133368400 M * blackfire 2.6.14.2, vs2.1.0rc8 1133368419 M * Bertl robig: the xid also shows that you are _not_ using static context ids (which are advised) 1133368445 M * Bertl robig: so you have to check all contexts for the proper xid 1133368456 M * blackfire i have a whole collection of oopses, all in the same place 1133368468 M * Bertl blackfire: there is a 2.6.14.3 out there, maybe check that too? 1133368482 M * Bertl blackfire: do you have an easy way to 'crash' it? 1133368499 M * blackfire yup i will but it's been so since 2.6.13.4 at least 1133368513 M * blackfire well, make -j24 in a kernel dir usually does it 1133368525 M * robig how do i use them? 1133368536 M * Bertl probably a special module or setup you use, or maybe the machine is broken? 1133368549 M * blackfire several machines? 1133368576 M * blackfire dual xeon ht, dual core amd64 - these are the most affected 1133368578 M * Bertl robig: just create a file called 'context' in your config tree, and put a number between 2 and 49151 into it 1133368625 M * Bertl blackfire: okay, good, that very much eliminates the hardware part 1133368640 M * Bertl blackfire: what compiler do you use? 1133368651 M * blackfire gcc-3.4 1133368657 M * blackfire gcc-3.3 was the same 1133368680 M * Bertl okay, good, binutils? 1133368772 M * blackfire dpkg says 2.15-6 1133368785 M * Bertl okay fine .. that should do ... 1133368801 M * Bertl now the kernel config itself, could you upload that somewhere please? 1133368814 M * blackfire yup sure, wait a second 1133368910 M * blackfire http://dev.tld.pl/config/ 1133368932 M * blackfire these are all configs that i'm using 1133368978 M * blackfire (erup is a 3rd party patch but it's been in production since forever at our data center) 1133369006 Q * robig Ping timeout: 480 seconds 1133369032 M * Bertl what is erup doing? 1133369132 M * blackfire elevates specific privileges for normal users (set*id in a limited range) 1133369158 M * Bertl hmm, url? 1133369160 M * blackfire so you can setuid e.g. 2000 -> 2001 but not 2000 -> 0 or something 1133369230 M * blackfire http://dev.tld.pl/patches/ 1133369238 M * blackfire here are the patches that i'm using 1133369247 M * blackfire - vserver 1133369252 M * blackfire - erup 1133369267 M * blackfire - an old kerneldoc tweak (useless actually now) 1133369275 M * blackfire - the semi-famed sendfile fix 1133369289 M * blackfire - a version change in makefile 1133369308 M * blackfire - an extra device id for an usb-serial converter (for one machine) 1133369317 M * Bertl hmm, the makefile change is to confuse folks, right? :) 1133369330 M * blackfire yup ;) 1133369369 M * blackfire and it'd be a bit unwieldy to have a kernel name of 2.6.14.2-vs2.1.0-rc8-p4smp.20 :) 1133369497 M * Bertl k, well, do you see a chance to test without erup? 1133369525 M * Bertl it's a quite intrusive patch and it might clash with recent kernels 1133369554 M * blackfire i'll have a try... it'll fsck up our apache setup but who cares ;) i wanted to replace it anyway 1133369582 M * Bertl let's also enable ksyms and debug info in the kernel config 1133369584 M * blackfire it's been ported from 2.6.13 so it isn't very outdated 1133369588 M * blackfire sure 1133369595 M * Bertl and keep the build tree around 1133369650 M * blackfire no problem 1133369724 M * blackfire hmm what's exactly the config option to ksyms? do you mean kallsyms? i thought that's always on in 2.6? 1133369733 M * blackfire debug info is checked now :) 1133369751 M * Bertl yes that and the VERBOSE_DEBUG + DEBUG_INFO IIRC 1133369867 M * blackfire CONFIG_DEBUG_BUGVERBOSE=y 1133369868 M * blackfire CONFIG_DEBUG_INFO=y 1133369874 M * blackfire ok? 1133369937 M * Bertl yep, should be fine 1133369980 M * blackfire ok, doing a make clean just to be sure 1133370091 M * blackfire there's a static string array: 1133370091 M * blackfire static const char *task_state_array[] = { 1133370092 M * blackfire "R (running)", /* 0 */ 1133370094 M * blackfire ... etc 1133370125 M * blackfire in the compiled object file is it going to have its real address or is it replaced after linking? 1133370130 J * kermit ~kvirc@i577B46C5.versanet.de 1133370147 M * Bertl welcome kermit! 1133370158 M * kermit hi everybody ! 1133370169 Q * Viper0482 Remote host closed the connection 1133370180 M * Bertl blackfire: depends, but it will get its final address in the vmlinux 1133370190 M * kermit anybody here with experience in vserver2.0 + 2.6.12.4 and lvm2 problems ? 1133370205 M * Bertl kermit: problems no, the rest, yes :) 1133370261 M * kermit i am running a setup with a logical volume for each server. the fs is mounted via /etc/fstab in the "base"-system at boot-time. 1133370274 M * kermit problem is.. 1133370278 M * kermit 1. stop vserver 1133370284 M * kermit 2. umount lvm-partition 1133370298 M * kermit lvdisplay still shows that the lv is "open" 1133370306 M * kermit so i am unable to remove the lv. 1133370331 M * kermit distri is debian sarge, util-vserver from sarge, kernel compiled myself. 1133370348 M * blackfire Bertl: there's a movl $0, %ecx and a mov (%ecx), %eax so that makes me a bit concerned but it's probably filled out by the linker ;) 1133370371 M * Bertl kermit: well, first, make sure to replace the debian tools, they are broken ... 1133370401 M * Bertl kermit: then, you might consider that if you 'mount' the lvm on startup, and unmount it on shutdown 1133370426 M * Bertl the second started guest will have _both_ lvm mounts in his namespace 1133370472 M * Bertl kermit: so you might consider unmounting it after the startup (in the host namespace) 1133370545 M * Bertl blackfire: yes, that looks like relocation addresses :) 1133370611 M * kermit Bertl: i am mounting the lvm-paritions via the base-fstab because i wanted to keep the setup simple as i am just helping the admin and he has no experience with vservers... but what do you mean by " _both_ lvm mounts in his namespace" ? 1133370642 M * Bertl look, it's simple ... when the guest is created ... 1133370657 M * Bertl a 'copy' of the current namespace is made 1133370672 M * Bertl this includes _all_ currently mounted devices 1133370706 M * Bertl so, for example, if you mount /mnt/cdrom, and then start a guest, the guest will 'copy' that mount 1133370733 M * Bertl now even if you unmount the cdrom on the host, you will not be able to eject it 1133370745 M * Bertl because the guest is still keeping an 'open' mount 1133370819 M * kermit ok, i think i got it, but whats the point about it ? Why is the whole namespace copied ? I mean i can't access /mnt/cdrom within the guest. so why copy it ? 1133370850 M * Bertl because this is how CLONE_NAMESPACE works and the tools do not cleanup the namespaces yet 1133371016 M * Bertl and you can access it, if you enter the namespace 1133371016 M * kermit i ?can? access it ? 1133371016 M * Bertl yes, just not from within the guest 1133371016 M * Bertl but if you use vnamespace to enter the namespace, you will see the mounts 1133371016 M * aba but that's just because the guest is chrooted away 1133371036 M * Bertl --rbind ed to be precise 1133371088 M * aba ("chroot" just for the common understanding - "secure chrooted aka --rbind ed" is technically more precise :) 1133371117 M * Bertl k, I let you win for today :) 1133371119 M * kermit ok, so a "vnamespace --enter XX; umount " should free the lvm-parition ? 1133371138 M * Bertl kermit: precisely 1133371168 M * Bertl kermit: of course, you ahve to do that for every guest which got a copy 1133371171 M * aba kermit: eh, not a "vnamespace --enter XX umount " ? 1133371195 M * Bertl ah, yes, indeed 1133371208 M * Bertl aba: now you really earned it :) 1133371238 M * kermit thanks a lot ! i'll try this... any more things i should be aware of with namespaces ? 1133371244 M * Bertl aba: btw, how are you doing? 1133371266 M * aba Bertl: good, thanks. Just a bit busy currently 1133371278 M * kermit i am still used to vserver+2.4. never got such "problem" there... 1133371302 M * kermit . 1133371331 M * aba Bertl: btw, what I really would like is to make a small and reviewable package of "how to escape from vservers", and publish it (as some sort of "quality assurance") 1133371407 M * aba (I know that there exists some code of you, but I'd like to see it "easier useable" for normal users) 1133371429 A * blackfire crosses fingers and reboots 1133371614 M * blackfire i had to tune2fs the disks to -i0 -c0 because they'd fsck several times a week ;) 1133371655 M * kermit Bertl: thanks a lot, vnamespace did it ! 1133371885 M * blackfire dead :( 1133371901 M * blackfire Nov 30 18:31:09 40 [42949670.500000] CPU: 1 1133371901 M * blackfire Nov 30 18:31:09 40 [42949670.500000] EIP: 0060:[] Not tainted VLI 1133371904 M * blackfire Nov 30 18:31:09 40 [42949670.500000] EFLAGS: 00010247 (2.6.14.2-test) 1133371906 M * blackfire Nov 30 18:31:09 40 [42949670.500000] EIP is at do_task_stat+0x93/0x8a9 1133371909 M * blackfire Nov 30 18:31:09 40 [42949670.500000] eax: 00000000 ebx: 00000000 ecx: a05f36c0 edx: bae2e5e0 1133371912 M * blackfire Nov 30 18:31:09 40 [42949670.500000] esi: a7b95000 edi: b8e9d550 ebp: b1603f20 esp: b1603d9c 1133371915 M * blackfire Nov 30 18:31:09 40 [42949670.500000] ds: 007b es: 007b ss: 0068 1133371917 M * blackfire Nov 30 18:31:09 40 [42949670.500000] Process pidof (pid: 15390, threadinfo=b1602000 task=b9c23030) 1133371920 M * blackfire all the same story 1133371953 M * blackfire a single make -j24 in a vserver killed it 1133371993 M * blackfire i don't know if it was a coincidence but i have issued a vsched --xid 2144 (the vserver that compiled the kernel) --fill-rate 8 1133371994 M * Bertl blackfire: okay, now please do the following: 1133372012 M * Bertl inside the kernel build tree: 1133372022 M * micah if I have /etc/vservers//interfaces/1/dev with "eth1" in it because I want the IP specified there to appear on that device, does that mean when that vserver is stopped it will attempt to deconfigure that device as well? 1133372054 M * Bertl blackfire: 'addr2line -e vmlinux a01add9f 1133372055 M * blackfire Bertl: just a sec, it is still booting ;) 1133372075 M * Bertl micah: yes 1133372183 M * micah Bertl: is there a way to specify what device an IP gets put on without allowing the vserver to tear the interface down on restart? 1133372201 M * blackfire ??:0 1133372204 M * blackfire ehh 1133372224 M * Bertl blackfire: missing ksyms 1133372251 M * blackfire well the vmlinux has ~50mb, i'd figure the debug info is there ;) 1133372273 M * blackfire nm vmlinux shows the symbols 1133372286 M * Bertl micah: well, you have to decide _what_ you want .. either the tools should create the ip for you (in this case they will tear it down again) or you create it on your own ... there is no open for create but do not destroy ... 1133372339 M * blackfire can i objdump a single function from vmlinux? dumping it all will take ages probably 1133372357 M * micah Bertl: I want the tools to create the IP for me and tear down that IP, but I dont want the entire interface to be taken down when it does so 1133372410 M * Bertl micah: it will not do so if you have a primary interface ip set or the new magic interface option configured (no idea what it is called right now, but 2.6.14 should have it) 1133372458 M * Bertl just to clarify, the tools are never taking down the interface, jsut removing the ip 1133372490 M * micah hmmm ok 1133372516 M * micah i think i am experiencing a different problem, where bringing down a vserver is removing the IP from the interface of a different vserver 1133372659 M * micah for example, I have vs/interfaces/1/ip set to 192.168.0.10, I started that vserver, and that IP existed, but after I have restarted other vservers (that do not have 192.168.0.10 configured) that IP is no longer listed in ip addr list 1133372731 M * Bertl yes, that's the 'primary' issue 1133372755 M * Bertl actually it's very old and I still consider it a 'buggy behaviour' 1133372764 M * micah what should I do to stop it? 1133372769 M * blackfire Bertl: ok, I got a full vmlinux objdump 1133372789 M * Bertl but the network folks told me several times that this is intentional 1133372826 M * Bertl micah: if you start your guests, and have a look with 'ip addr ls' you will see, that for each subnet, the first one is called 'primary' 1133372829 M * micah is my only option to put each IP on its own virtual interface (eth1:x etc) 1133372850 M * Bertl eth1:x will not help you a bit, as it is an alias 1133372899 M * Bertl check the setup, look at the IPs (with ip addr ls) and you will find, that taking down the one IP inside a subnet labeled 'primary' will tear down all others too 1133372989 M * micah in my ip addr list it doesn't say primary, it just does not say secondary: 1133372989 M * micah inet 192.168.0.1/24 brd 192.168.0.255 scope global eth1 1133372989 M * micah inet 192.168.0.7/24 brd 192.168.0.255 scope global secondary eth1 1133373008 M * micah but that is fine... so maybe I should create that primary IP off of the host? 1133373376 M * Bertl yes, that's the usual workaround 1133373401 M * Bertl either assign a 'dummy' ip on the host, or alternatively use /32 networks 1133373525 Q * sladen Ping timeout: 480 seconds 1133373695 J * sladen paul@starsky.19inch.net 1133373996 M * Bertl wb sladen! 1133374064 M * micah Bertl: I met sladen a week or so ago 1133374086 M * micah we hung out in NYC and Boston... small world 1133374246 Q * Duckx Ping timeout: 480 seconds 1133374246 Q * DuckMaster Ping timeout: 480 seconds 1133374513 J * stefani ~stefani@superquan.apl.washington.edu 1133374545 M * Bertl micah: yeah, funny ... where are you now? 1133374785 M * micah Bertl: i'm near NYC now, but I think sladen is back in northern europe 1133375020 M * Bertl micah: I'm in Princeton, NJ 1133375101 J * DuckMaster ~duckx@195.75.27.155 1133375114 J * Duckx ~duckx@195.75.27.155 1133375115 M * micah Bertl: you are right now? what are you doing there? thats about 45 minutes from here! 1133375360 Q * infowolfe Ping timeout: 480 seconds 1133375609 M * Bertl micah: yeah, that's why I asked/mentioned it ... 1133375665 J * JonB ~NoSuchUse@kg49.kollegiegaarden.dk 1133375815 M * micah Bertl: how long are you there for? 1133375837 M * Bertl till 16th of december at least 1133375912 M * micah wow, long time. we should have a $beverage some time 1133376003 M * Bertl definitely! 1133376056 M * Bertl micah: are you in NYC permanently or just visiting? 1133376084 M * Bertl (actually mef, sitting beside me wants to know that :) 1133376355 M * micah i'm actually in highland park, mef might know where it is.. just moved here for the next half year 1133376407 M * Bertl so, mef asks: "if you are in highland park, why 45minutes?" :) 1133376450 J * infowolfe ~infowolfe@209-112-219-176-cdsl-rb1.nwc.acsalaska.net 1133376468 J * Legend ~chatzilla@cp676338-a.landg1.lb.home.nl 1133376474 M * Legend hello all!! 1133376485 M * Bertl welcome Legend! 1133376519 M * Legend short question... im using debian.. i apt-getted de vserver patch and used make-kpkg to create an new kernel with default debian setting + vserver 1133376537 M * Bertl which kernel version? 1133376547 M * Legend i reboot and with the testme.sh i get chcontext ok but chbind fail 1133376550 M * Legend 2.6.12 1133376573 M * Legend with vc_set_ipv4() Function blbal 1133376615 M * micah if using 2.6.12 then is using patch-2.6.12.4-vs2.0.diff 1133376618 M * Bertl hmm, you should go for 2.6.14 or later 1133376631 P * JonB Leaving 1133376647 M * micah Legend: do you have the util-vserver packages installed? 1133376656 M * micah Bertl: 45 minutes because it takes 30 minutes to walk to the train :) 1133376671 M * Legend wait chainging computer brb 1133376705 Q * Legend Quit: 1133376739 M * micah Bertl: I'm trying to track down the kernel patches later than 2.6.12, the only thing I can find are refs to http://www.13thfloor.at/vserver/d_rel26/overview/ which only have 1.9.5 patches 1133376757 M * Bertl micah: try google 1133376761 M * blackfire micah: http://vserver.13thfloor.at/Experimental/ ? 1133376770 M * micah ah google! :p 1133376799 M * micah I was meaning the path that one would take from the wiki or from 13th floor 1133376836 M * blackfire micah: ah.. i don't know then ;) 1133376937 M * micah I'm thinking of changing the [Development 2.6 Releases] link under Downloads to point to the Experimental list, but maybe its too experimental? 1133376947 J * Legend ~chatzilla@cp676338-a.landg1.lb.home.nl 1133376955 M * Legend im back 1133376961 M * Bertl micah: we'll restructure this pretty soon 1133376987 M * micah ok, no rush, someone pointed it out to me 1133376998 M * Legend im using debian testing and yes ive install the utils-vserver package aswel as the debiantools-vserver 1133377313 M * micah Legend: I would remove the debiantools-vserver, it is not useful 1133377527 Q * kermit Quit: KVIrc 3.2.0 'Realia' 1133377750 M * Legend ok but it doesnt conflict or so and it doesnt explain the fault i get 1133378143 M * mef micah: what are you doing in highland park? and why are you there for just six months? where will you return to when the six months are over? 1133378171 M * Bertl Legend: did you accidentially disable the legacy networking? 1133378238 M * Legend how can i see this? 1133378607 M * Legend Bertl: my config looks fine 1133378621 M * Legend chcontext is working. 1133378632 M * Legend chbind: vc_set_ipv4root(): Function not implemented 1133378874 M * Legend config_vserver_legacy=y 1133378986 M * micah Legend: what are all your vserver kernel config options set to? 1133379021 M * Legend CONFIG_VSERVER_LEGACY=y 1133379023 M * Legend CONFIG_VSERVER_LEGACY_VERSION=y 1133379025 M * Legend CONFIG_VSERVER_NGNET=y 1133379026 M * Legend CONFIG_VSERVER_PROC_SECURE=y 1133379028 M * Legend CONFIG_VSERVER_HARDCPU=y 1133379029 M * Legend CONFIG_VSERVER_HARDCPU_IDLE=y 1133379031 M * Legend # CONFIG_INOXID_NONE is not set 1133379032 M * Legend # CONFIG_INOXID_UID16 is not set 1133379034 M * Legend # CONFIG_INOXID_GID16 is not set 1133379035 M * Legend CONFIG_INOXID_UGID24=y 1133379037 M * Legend # CONFIG_INOXID_INTERN is not set 1133379038 M * Legend # CONFIG_INOXID_RUNTIME is not set 1133379040 M * Legend CONFIG_XID_TAG_NFSD=y 1133379042 M * Legend CONFIG_VSERVER_DEBUG=y 1133379043 M * Legend CONFIG_VSERVER_HISTORY=y 1133379045 M * Legend CONFIG_VSERVER_HISTORY_SIZE=64 1133379170 A * michal_ has been killed by the flood ;] 1133379486 M * Bertl Legend: remove the NGNET=y line 1133379501 M * Legend what does that mean? 1133379502 M * Bertl execute 'make oldconfig' and take the default 1133379518 M * Bertl Legend: and next time use pastebin.com or similar :) 1133379576 M * Legend hehe sorry for that 1133379940 M * Legend AH it says disable legacy networking hehe read read read 1133380132 M * Bertl Legend: should work fine once you have that changed 1133380173 M * Legend compiling.... 1133380428 J * Aiken ~james@tooax7-084.dialup.optusnet.com.au 1133380774 Q * monrad Quit: Leaving 1133380945 J * dhansen ~dave@sprucegoose.sr71.net 1133380953 J * shuri ~shuri@64.235.209.226 1133381324 M * Bertl welcome Aiken! dhansen! shuri! 1133381387 M * Aiken good 1133381399 M * michal_ GUG ! 1133381403 M * michal_ Generic UTC Greeting 1133381524 M * shuri Hello Bertl ! 1133381534 M * shuri LTNS 1133381575 A * mnemoc stole 'moin' fom germans, good wildcard 1133381693 Q * mountie Remote host closed the connection 1133381708 J * mountie ~mountie@CPEdeaddeaddead-CM000a739acaa4.cpe.net.cable.rogers.com 1133381820 M * Bertl hey michal_! mnemoc! 1133381827 M * Bertl shuri: yeah, indeed! 1133381985 M * micah Bertl: I have a config help re-word suggestion for the COW piece 1133382006 M * micah I would suggest changing it to: 1133382007 M * micah This enables the COW (Copy-On-Write) link break code. This allows 1133382007 M * micah you to treat unified files like normal files when writing 1133382007 M * micah to them (this will implicitely break the link and create 1133382007 M * micah a copy of the unified file) 1133382134 M * Bertl micah: k, please provide a patch ... I'll look into it shortly 1133382142 M * micah ok 1133382887 M * micah sent to ML 1133382978 M * Bertl great, tx! 1133383286 J * lilo ~lilo@lilo.usercloak.oftc.net 1133383665 Q * lilo_ Ping timeout: 480 seconds 1133383966 J * monrad ~monrad@213083190130.sonofon.dk 1133384026 Q * FireEgl Remote host closed the connection 1133384429 Q * ntrs Read error: Connection reset by peer 1133384437 J * ntrs ~ntrs@68-188-50-87.dhcp.stls.mo.charter.com 1133384589 M * blackfire ok, i'm calling it a day 1133384623 M * blackfire thanks everybody 1133384625 M * blackfire cya 1133384665 M * bragon hello :) 1133385419 J * yarihm ~yarihm@84-73-119-83.dclient.hispeed.ch 1133385441 Q * shuri Remote host closed the connection 1133386070 Q * neofutur_ Ping timeout: 480 seconds 1133386762 J * neofutur ~neofutur@neofutur.net 1133386953 M * [MUPPETS]Gonzo is there a way to add an ip to a vserver while it is running? If so, how? 1133387165 Q * roeb[] Ping timeout: 480 seconds 1133387287 M * derjohn [MUPPETS]Gonzo, i assume with chbind, but is it possible on a "running" vserver? 1133387340 Q * neofutur Ping timeout: 480 seconds 1133387786 J * miller7 none@213.239.180.101 1133387807 M * daniel_hozac [MUPPETS]Gonzo: you'll need to restart the services anyhow, so might as well just restart it. 1133387828 M * [MUPPETS]Gonzo ok 1133387840 M * miller7 hello guys 1133388508 Q * Legend Quit: Chatzilla 0.9.68.5.1 [Firefox 1.5/undefined] 1133391263 Q * yarihm Quit: Leaving 1133392747 P * stefani I'm Parting (the water) 1133393013 M * Bertl okay, off for now ... back later 1133393034 N * Bertl Bertl_oO