1165017737 J * ms_ ~ms@arkansas.doc.ic.ac.uk 1165018132 M * Guy- Bertl: suspend2 crashes with vanilla too 1165018164 M * Bertl okay, so probably something which will get fixed in a newer version 1165018229 M * Guy- hopefully :) 1165018790 Q * dna_ Quit: Verlassend 1165019271 P * stefani I'm Parting (the water) 1165020348 M * Guy- wow, the in-kernel swsusp seems to work though 1165020366 M * Guy- even with vserver (although I didn't try it with and running virtual contexts yet) 1165020391 M * jpachec swsusp? 1165020408 M * Guy- as in, software suspend 1165020415 M * jpachec ah 1165020417 M * Guy- handy for notebooks 1165020445 M * Guy- the mainline suspend never really worked for me so far, which is why I tried suspend2 1165020462 M * Guy- but now suspend2 is the one that doesn't work :) 1165020488 M * Bertl let me know how it goes with a guest running :) 1165020506 M * Bertl daniel_hozac: we have a bunch of duplicate inclusions in the ipv4 area? 1165020562 M * Guy- Bertl: I will :) 1165020564 Q * gerrit Ping timeout: 480 seconds 1165020571 M * Bertl seems like route.h pulls vs_base.h in ... 1165020592 M * Guy- but now, it's time to sleep() 1165020605 M * Guy- good night folks 1165020693 M * Bertl have a good one! cya! 1165022237 J * Aiken ~james@tooax6-138.dialup.optusnet.com.au 1165022322 M * Aiken is this expected with 2.6.19? http://paste.linux-vserver.org/723 1165022424 M * Aiken hatch egg 1165022438 M * Aiken wrong window :( 1165022686 M * hardwire ? 1165022696 M * hardwire you playing with your little pocket monster? 1165022737 M * Aiken packaging system 1165022756 M * hardwire pocket packaging system? 1165022765 M * hardwire hatch egg.. what were you trying to do? 1165022785 M * Aiken pocket? 1165022792 M * hardwire egg? 1165022822 M * Aiken the cmd was for my packaging system, I had irc with focus instead of the terminal to my alpha 1165022834 M * hardwire hehe 1165022843 M * hardwire what uses "hatch egg" as part of its packaging? 1165022849 M * hardwire thats where I got lost 1165022857 M * Aiken the cmd was for my packaging system 1165022875 M * hardwire what uses "hatch egg" as part of its packaging? 1165022966 Q * Piet Ping timeout: 480 seconds 1165022970 M * Aiken how many times to I have to say "the cmd was for my packaging system"? 1165023022 M * hardwire how many times do I have to ask "for what packaging system would a command like that apply"? 1165023040 M * Aiken mine 1165023082 M * hardwire see.. my packaging system on this workstation is apt.. 1165023090 M * hardwire if you have your own.. cool beans.. 1165023434 M * hardwire so what kind of packaging system is it? 1165023675 M * Aiken alpha kernel 2.6.19 utils 0.30.212-rc2 http://paste.linux-vserver.org/724 tests 001 & 031 show fail 1165023704 M * Aiken hardwire source, I went the way of LFS late 2002 or every early 2003 1165023867 M * Aiken testfs ext2 ext3 & resierfs all have test 116 fail which seems famillar 1165023872 M * hardwire Aiken: groovy 1165023877 M * hardwire do you have a distro you distribute? 1165023958 M * Aiken only for all of my machines 1165023972 M * hardwire ok 1165024458 Q * Aiken Ping timeout: 480 seconds 1165026407 J * Johnnie ~jdlewis@jdlewis.org 1165026665 J * gerrit ~gerrit@c-67-160-146-170.hsd1.or.comcast.net 1165027687 J * bronson__ ~bronson@c-67-188-227-156.hsd1.ca.comcast.net 1165027797 J * Hugon Hugon@host-62-141-242-236.tomaszow.mm.pl 1165028110 M * jpachec ah lfs 1165028120 M * jpachec the best teacher i ever had 1165028141 M * Bertl welcome Hugon! hey jpachec! 1165028644 Q * Hugon Quit: Leaving 1165028801 Q * Schaka Ping timeout: 480 seconds 1165029253 M * jpachec hey Bertl! 1165029901 Q * bronson__ Ping timeout: 480 seconds 1165030630 J * bronson__ ~bronson@adsl-75-36-144-172.dsl.pltn13.sbcglobal.net 1165032391 M * Bertl okay, off to bed now ... have a good one everyone! cya! 1165032413 N * Bertl Bertl_zZ 1165033071 Q * ms_ Ping timeout: 480 seconds 1165034469 J * Aiken ~james@tooax6-121.dialup.optusnet.com.au 1165037315 J * Piet hiddenserv@tor.noreply.org 1165039485 Q * Piet Quit: Piet 1165039800 J * ms_ ~ms@arkansas.doc.ic.ac.uk 1165041826 Q * Aiken Quit: Leaving 1165041954 J * Aiken ~james@tooax6-121.dialup.optusnet.com.au 1165047796 J * mrrm_ ~urkel@149.9.0.56 1165048658 Q * Aiken Quit: Leaving 1165048789 J * Aiken ~james@tooax6-121.dialup.optusnet.com.au 1165051068 Q * Aiken Ping timeout: 480 seconds 1165051094 J * bonbons ~bonbons@83.222.39.117 1165051322 Q * DreamerC_ Quit: leaving 1165051339 J * DreamerC ~dreamerc@59-115-48-61.dynamic.hinet.net 1165052862 J * pmenier ~pmenier@ACaen-152-1-83-187.w86-205.abo.wanadoo.fr 1165052876 M * pmenier hello 1165052910 M * pmenier does anybody know where i can get util-vserver-0.30.212 ? 1165053120 J * eyck_ ~eyck@nat-old.nowanet.pl 1165053132 J * DavidS ~david@chello062178045213.16.11.tuwien.teleweb.at 1165053210 M * bonbons that one has not been released yet... better check svn (svn.linux-vserver.org) for it, or ask daniel_hozac where he has tarballs 1165053223 M * pmenier ok thanks 1165053511 N * eyck_ Eyck- 1165053732 J * dna ~naucki@68-219-dsl.kielnet.net 1165054271 M * daniel_hozac http://people.linux-vserver.org/~dhozac/p/uv/experimental/util-vserver-0.30.212-rc2.tar.bz2 1165054332 M * daniel_hozac Bertl_zZ: yeah 1165054353 M * pmenier ok i try it and tell you results 1165054722 M * Eyck- Bertl_zZ: ok, so what do I do next about 2.4 release? 1165054756 M * daniel_hozac a new release for 2.4? 1.2.11 will finally be released? :) 1165054807 M * Eyck- I hope so 1165054852 M * Eyck- I need another year or two of life in 2.4 branch. 1165054910 M * daniel_hozac hehe. 1165054933 M * Wonka 2.4 is _so_ dead... 1165054936 M * DavidS Eyck-: that sounds like an interesing story :) 1165054969 M * Eyck- Wonka: yeah, and you should also burry your parents alive, because they are SOOO last week 1165054970 M * daniel_hozac Eyck-: you have the patch ready, right? do you just want Bertl_zZ to look over the changes? 1165055026 M * Eyck- daniel_hozac: nope, I just want to talk :) 1165055034 M * daniel_hozac ah :) 1165055081 M * DavidS huh? vnamespace: vc_xidopt2xid("/etc/vservers/smtpscanner"): No such file or directory 1165055101 M * DavidS but the directory exists 1165055110 M * DavidS (Debian etch here) 1165055128 M * daniel_hozac hmm. what were you running to get that? 1165055136 M * daniel_hozac and what util-vserver version? 1165055161 Q * ms_ Ping timeout: 480 seconds 1165055172 M * DavidS in my $HOME: sudo vserver smtpscanner enter 1165055187 M * DavidS 0.30.211-4 1165055258 J * h01ger ~holger@socket.layer-acht.org 1165055270 M * h01ger hi folks! 1165055276 M * daniel_hozac hello 1165055285 M * DavidS uname -r: 2.6.18-3-xen-vserver-686 1165055331 M * DavidS funny, exec works 1165055350 M * daniel_hozac enter is handled specially. 1165055364 M * h01ger i have a vserver, on which i have a chroot, in which the PRNG is not seeded. outside the chroot it is. what am i missing? 1165055366 M * daniel_hozac does vnamespace -e /etc/vservers/smtpscanner ls -l work? 1165055390 M * daniel_hozac h01ger: is the PRNG-device nodes available? 1165055407 M * DavidS no, same error 1165055453 M * daniel_hozac is /etc, /etc/vservers or /etc/vservers/smtpscanner on a different filesystem? 1165055477 M * daniel_hozac what does strace say? 1165055489 M * h01ger ah, /dev/random and /dev/urandom are missing... damn new debootstrap (relying on udev :) 1165055643 M * DavidS stat64("/etc/vservers/smtpscanner", {st_mode=S_IFDIR|0755, st_size=4096, ...}) = 0 1165055643 M * DavidS stat64("/etc/vservers/smtpscanner.upstream", 0xbfa45e14) = -1 ENOENT (No such file or directory) 1165055643 M * DavidS open("/etc/vservers/smtpscanner/context", O_RDONLY|O_LARGEFILE) = -1 ENOENT (No such file or directory) 1165055671 M * DavidS that's it probably .. i had once a /etc/vservers/smtpscanner.upstream directory with an earlier version of this vserver 1165055764 M * daniel_hozac once as in when you started it? 1165055798 M * DavidS i can't say for sure, but it is possible 1165055835 M * daniel_hozac you really should use static xids as well :) 1165055914 M * DavidS like "echo $RANDOM > /etc/vservers/muh/context"? 1165055972 J * sebastian ~info@pD957F582.dip.t-dialin.net 1165055981 M * daniel_hozac if $RANDOM is between 2 and 49151, sure. 1165056011 M * DavidS so static xids and a reboot will fix that? 1165056041 M * daniel_hozac well, just a reboot of the guest should fix it. 1165056058 M * daniel_hozac but as dynamic xids are going away, you shouldn't use them. 1165056108 M * DavidS daniel_hozac: once again you saved my day :) thanks! 1165056129 Q * DavidS Quit: Leaving. 1165057204 P * meandtheshell 1165057291 Q * ruskie Quit: Caught sigterm, terminating... 1165057484 J * meandtheshell ~markus@85-125-230-212.dynamic.xdsl-line.inode.at 1165057542 J * ruskie ~ruskie@ruskie.user.oftc.net 1165057959 M * pmenier Re-hello 1165057983 M * pmenier with util-vserver-0.30.212-rc2 it seems to work fine with kernel 2.6.19 1165058004 M * pmenier Linux version 2.6.19-vs2.1.x-t1 (root@lindows) (version gcc 3.3.5 (Debian 1:3.3.5-13)) #1 SMP 1165058031 M * pmenier util-vserver: 0.30.212-rc2; Dec 2 2006, 10:52:53 1165058187 M * h01ger daniel_hozac, hmmm. i cannot create devices on the vserver (which is a feature, i know). but so the created chroot cannot contain (nor create) the random-device nodes and i'm stuck. 1165058313 M * daniel_hozac h01ger: so either create them from the host, or give the guest CAP_MKNOD. 1165058383 M * h01ger both are suboptimal. the chroot is created automatically (on the vserver), and i dont want to give the guest the capability, cause i'm not the only root in that guest... 1165058388 A * h01ger scratches head 1165058515 M * Eyck- why not created them automatically when the chroot is created? 1165058536 M * Eyck- you're creating chroot inside vserver? 1165058560 M * daniel_hozac h01ger: so add a new ccap which permits creating a few secure devices? 1165058586 M * Eyck- oh, this potential new ccap became a reality recently? 1165058606 M * daniel_hozac no. 1165058617 M * daniel_hozac thus the add in the above sentence ;) 1165058621 M * Eyck- hehe 1165058644 M * Eyck- what about this 'mount few select secure filesystems inside vserver' ? 1165058647 M * h01ger Eyck-, yes i create the chroot in the vserver 1165058658 M * daniel_hozac Eyck-: yes, that's what secure_mount is supposed to do. 1165058661 M * h01ger daniel_hozac, howto? 1165058680 M * h01ger or is that a fictional feature? ;) 1165058701 M * daniel_hozac it's one you'd have to create yourself ;) 1165058737 M * Eyck- hmm, secure_mount would fix that... wouldn't it? you mount-bind your own /dev into newly-created chroot... 1165058739 M * h01ger but i can create custom capabilities? (without patching the kernel?) 1165058745 M * daniel_hozac Eyck-: no. 1165058755 M * daniel_hozac MS_NODEV is implied for all secure_mounts. 1165058764 M * daniel_hozac h01ger: no, of course not. 1165058767 M * daniel_hozac you have to patch the kernel. 1165058778 M * h01ger humpf 1165058784 M * h01ger ETOOSTUPID :) 1165058803 M * Eyck- daniel_hozac: hmm, for mount-bind that doesn't make much sense, hmm, but making an exception for this wouldn't make much sense either 1165058811 M * Eyck- ok. 1165058838 M * daniel_hozac Eyck-: it's done for all mounts, so it's not duplicated to each of the separate mount functions. 1165059079 M * daniel_hozac h01ger: if you're willing to test it, i could give it a try. 1165059094 A * h01ger would 1165059112 M * h01ger i want a solution for that problem :) 1165059155 M * daniel_hozac what kernel are you using? 1165059683 M * h01ger 2.8.18 from debian-sid 1165059735 M * daniel_hozac hehe, Debian sure are ahead of the times for once! :) 1165059847 M * h01ger sid quite often is. its stable that is lagging :) 1165060074 M * daniel_hozac i was referring to the 2.8 ;) 1165060144 M * h01ger hehe 1165060263 Q * ruskie Quit: killed 1165060544 J * ruskie ~ruskie@ruskie.user.oftc.net 1165060722 M * daniel_hozac rebooting now. 1165060849 Q * ensc Killed (NickServ (GHOST command used by ensc_)) 1165060859 J * ensc ~irc-ensc@p54B4EEF4.dip.t-dialin.net 1165061049 M * daniel_hozac seems to work ok here, http://people.linux-vserver.org/~dhozac/p/k/delta-secdev-poc01.diff 1165061065 M * daniel_hozac obviously this is just a quick hack, the list of allowed devices should be modifiable from userspace. 1165061079 M * daniel_hozac (current list is the same as the devices created by util-vserver by default) 1165061134 M * h01ger so it includes (u)random.. 1165061195 M * h01ger ok, thanks a lot!, will try this patch later (might become tomorrow) and report back 1165061290 M * daniel_hozac actually, it allows too much. creating block devices with those majors/minors works. 1165061347 M * h01ger uh 1165061394 M * daniel_hozac S_IFCHR and S_IFBLK weren't bits, as i expected them to be :) 1165061474 M * daniel_hozac http://people.linux-vserver.org/~dhozac/p/k/delta-secdev-poc02.diff should fix that though. 1165061564 M * daniel_hozac yep, doesn't work anymore. 1165061624 M * h01ger :)) 1165061970 J * ms_ ~ms@arkansas.doc.ic.ac.uk 1165062408 M * bonbons daniel_hozac: where can I find a list of variables with guest-information available to the scripts in /etc/vservers//scripts ? 1165062459 M * daniel_hozac the name is in $2. 1165062574 M * bonbons isn't the xid/nid available somewhere, as well as config directory? 1165062721 M * daniel_hozac well, $VSERVER_DIR might be available... i don't know if it's exported. 1165062786 M * daniel_hozac but you can quite easily get the directory/xid from the name by using vserver-info. 1165062952 M * bonbons ok 1165063357 M * bonbons daniel_hozac, is there one script that runs after context has been setup, but before guest's init is started? at pre-start the context is not available yet... 1165063368 M * daniel_hozac nope. 1165063434 M * daniel_hozac you should be able to use one of the non-executable scripts to setup VSERVER_EXTRA_CMDS though. 1165063457 M * daniel_hozac that's semi-inside the context. 1165063474 M * daniel_hozac what are you trying to do? 1165063488 M * bonbons will try with post-start and hope it's not too late 1165063612 M * bonbons I'm playing around with network context (things that util-vserver does not support yet / in this special case ipv6, but maybe later with address groups) 1165063966 Q * bronson__ Ping timeout: 480 seconds 1165063981 M * pmenier daniel_hozac: just a few bug in util-vserver0.30.212 when i enter in vweb1 the prompt stays on vweb3 1165064365 J * damn- ~damn@85.120.136.41 1165064440 M * daniel_hozac pmenier: hmm? 1165064464 M * daniel_hozac bonbons: well, didn't i tell you IPv6 is supported from 0.30.212-rc1+? 1165064471 M * daniel_hozac in the form your patch implements it. 1165064563 M * bonbons daniel_hozac: you told me, but on that box 0.30.210 is running... until the box (host) gets a complete update 1165064692 M * daniel_hozac well, VSERVER_EXTRA_CMDS should work for setting up the network context. 1165064704 M * pmenier daniel_hozac: i was trying to configure kernel-2.6.19 and you told me to try util-vserver-0.30.212 1165064726 M * daniel_hozac pmenier: yes, but what do you mean? do you have a paste showing the problem, e.g.? 1165064775 M * pmenier yes i was saying that it didn't cor'rectly switch into contect with util-vsevers0.30.211 1165064804 M * daniel_hozac well, it does. 1165064825 M * daniel_hozac it's just that it doesn't disconnect the guest uts and IPC namespaces from the host's. 1165064844 M * pmenier now with 0.30.212-rc2 it switchs correctly but stays with incorrect prompt 1165064879 Q * damn- 1165064892 M * pmenier i.e. i type vserver vweb1 enter the prompt is #:/vweb3 1165064939 M * pmenier i post the operations on mailing-list 1165065094 M * daniel_hozac so if you change the hostname for one guest, it changes for all of them? 1165065113 M * pmenier i didn't try again 1165065228 M * pmenier i got this message in /var/log/messages : get_xid_list(1) = 3 get_xid_list(4) = 0 1165065413 M * pmenier i will lunch now. back in a few minutes 1165065524 M * daniel_hozac pmenier: you're using legacy configs? that won't work. 1165065547 M * daniel_hozac legacy config implies legacy utils, and AFAIK they don't even use namespaces yet... 1165065564 J * s0undt3ch_ ~s0undt3ch@81.193.56.241 1165065577 Q * dmax Read error: Connection reset by peer 1165065582 Q * s0undt3ch Read error: Connection reset by peer 1165065583 N * s0undt3ch_ s0undt3ch 1165065635 J * dmax ~semaj@81.193.56.241 1165067376 Q * pmenier Quit: KVIrc 3.2.0 'Realia' 1165067733 J * NewSense lisa@bowzzer.com 1165067967 J * pmenier ~pmenier@ACaen-152-1-83-187.w86-205.abo.wanadoo.fr 1165069428 J * Piet hiddenserv@tor.noreply.org 1165070709 Q * Piet Quit: Piet 1165071097 M * rob-84x^ hi guys, i've just written a simple kernel module for logging all execve(2) syscalls: http://pastebin.com/837529 but it doesn't catch any execve from vservers :( it's logging everything from the host (as it should), but only the first execve in the vserver (for e.g.: vserver arch exec /bin/cat is logged: execve: 16 /bin/cat [/bin/cat] (16 is the xid)) but nothing after that [in the shell spawned by `vserver arch exec /bin/bash`] 1165071156 M * rob-84x^ hijacking execve(2) is not enought for logging that syscalls in guests? 1165071254 M * daniel_hozac i can't see your paste. 1165071332 M * rob-84x^ daniel_hozac: try this: http://rafb.net/paste/results/j68Qco88.html 1165071431 M * daniel_hozac you sure that's the execve it's using? 1165071445 M * daniel_hozac i.e. you're not running a 32-bit guest? 1165071483 M * rob-84x^ daniel_hozac: yeah, i'm running 32-bit guest on 64-bit host. so there something else to hijack? 1165071491 M * daniel_hozac well, yes. 1165071498 M * daniel_hozac you're just modifying the 64-bit syscall table. 1165071520 M * daniel_hozac (and why aren't you using kprobes for this? this seems like the sort of thing it was designed for) 1165071649 M * rob-84x^ daniel_hozac: i just need logging of execve in vservers and i'm not a kernel hacker. neither i know what kprobes are for ;-) but i'll take a look 1165071688 N * Bertl_zZ Bertl 1165071692 M * Bertl morning folks! 1165071703 M * daniel_hozac systemtap with a kprobeable kernel will create that sort of module for you. 1165071703 M * rob-84x^ anyway, that kprobes look nice on the first sight 1165071715 M * daniel_hozac morning Bertl! 1165071762 M * daniel_hozac Bertl: i just booted 2.6.19-vs2.1.x-t1... what did i miss if none of the guest processes show up in /proc? 1165071768 M * daniel_hozac (inside the guest, that is) 1165071889 M * daniel_hozac hmm, seems to only be for the guest that runs an init. 1165071998 M * Bertl interesting 1165072100 M * daniel_hozac and is the pid_task warning really helpful? seems to just be flooding the logs :) 1165072164 M * Bertl it works fine here, with and without 'plain' style guests 1165072181 M * daniel_hozac humm. 1165072223 M * rob-84x^ daniel_hozac: btw, pointers in the 32-bit syscall table are 32-bits long or 64-bits long? 1165072244 M * Bertl daniel_hozac: I have a sarge and sarge64 running here, the sarge is with 'plain' now 1165072283 M * daniel_hozac mine's an etch guest. 1165072303 M * daniel_hozac but it's on my x86 host. 1165072331 M * daniel_hozac chcontext --xid 666 --secure --disconnect -- ps faux fails for me. 1165072356 M * daniel_hozac replacing ps faux with ls -l /proc doesn't show any process directories whatsoever. 1165072371 M * daniel_hozac not even the self symlink. 1165072388 M * Bertl # chcontext --xid 666 --secure --disconnect -- ps faux 1165072388 M * Bertl New security context is 666 1165072388 M * Bertl [root@pluster-20 ~]# USER PID %CPU %MEM VSZ RSS TTY STAT START TIME COMMAND 1165072391 M * Bertl root 15872 0.0 0.0 4336 784 ? Rs 10:12 0:00 ps faux 1165072418 J * dna_ ~naucki@68-219-dsl.kielnet.net 1165072444 M * Bertl daniel_hozac: give me a minute to retry somewhere else 1165072598 M * Bertl daniel_hozac: any ideas about the pmenier.net mail? 1165072617 M * daniel_hozac well, legacy configs. 1165072632 M * daniel_hozac -> legacy utils -> no vnamespace at all. 1165072655 M * Bertl ah, okay, something we have to detect in the future 1165072684 M * Bertl I think we should not port 2.0 to 2.6.19 at all or we figure some way to hack the uts/IPC in 1165072713 M * Bertl would it work for those tools if I'd unshare them in the legacy context creation? 1165072715 M * daniel_hozac not porting it at all sounds ok to me. 1165072720 Q * dna Ping timeout: 480 seconds 1165072739 M * daniel_hozac hmm, yeah, i suppose that should work. 1165072766 M * Bertl okay, will try that in t3 then, t2 should be up in a minute 1165072841 M * Bertl only contains fixes for the issues we knew of yesterday, the '89 and I removed some duplicate inclusions manually 1165072854 M * Bertl ah, and ext4 is now included but not tested 1165072956 M * daniel_hozac ok, sounds good. 1165073053 M * Bertl you said, you get the task lookup warnings very often? 1165073073 M * daniel_hozac yes, ps faux in a guest causes at least one for each pid it doesn't have access to. 1165073092 M * daniel_hozac which looks like it's expected, if i'm reading the procfs code correctly. 1165073119 M * Bertl is that a new chcontext? 1165073133 M * daniel_hozac "new"? 1165073138 M * Bertl I mean, one which is handling the namespaces correctly 1165073156 M * daniel_hozac no, chcontext still isn't fixed. 1165073165 M * Bertl so what do you expect then? 1165073183 M * daniel_hozac well... i'd still expect the process visibility to work quite fine? 1165073200 M * daniel_hozac how did you fix your chcontext (i.e. where's the patch? :))? 1165073224 M * Bertl hmm, good point ... double checking everything 1165073368 M * Bertl the pid task messages I get here look like readdir is giving the full set 1165073376 M * Bertl does that look similar for you? 1165073409 M * daniel_hozac right. 1165073416 M * daniel_hozac and that looks like it's expected. 1165073431 M * daniel_hozac it's running pid_task for every pid to see if it exists. 1165073445 M * Bertl yep, as there is no check there 1165073465 M * daniel_hozac well, we have the check in pid_task already, no? 1165073526 M * Bertl so, you suggest to just remove the warning there 1165073542 M * daniel_hozac right, i think it's a bit too verbose. 1165073558 M * daniel_hozac maybe make it a vxdprintk(VXD_CBIT... instead? 1165073575 M * Bertl but that was different on 2.6.18, somehow? 1165073590 M * daniel_hozac yeah, next_tgid looks completely different 1165073699 M * Bertl okay, I'm pretty sure I had some thoughts about that when porting the code 1165073735 M * Bertl the first thing which catches my eye is that we use PIDTYPE_PID and not PIDTYPE_REALPID in vx_get_proc_task() 1165073964 M * daniel_hozac hmm. 1165073978 M * daniel_hozac i think find_ge_pid is returning NULL. 1165074065 M * daniel_hozac that would explain why it's not generating any pid_task messages in the init-guest. 1165074075 M * Bertl yep, very likely it ends up with an empty dir 1165074091 M * daniel_hozac find_pid doesn't do any of the pid mapping. 1165074097 M * Bertl btw, just confirmed the machine I tested on was running 2.6.18.4 1165074103 M * daniel_hozac ah, hehe. 1165074140 M * Bertl okay, have to run now ... will look into it later ... feel free to fix it up before I do if you feel like :) 1165074154 N * Bertl Bertl_oO 1165074164 M * daniel_hozac hehe, if i manage to figure it out. 1165074204 M * Bertl_oO well, you already found it .. the ge_pid and 'nr' handling is causing that 1165074219 M * daniel_hozac but i don't think that explains why it's not finding any of the processes. 1165074221 M * Bertl_oO we should keep 'real' pids there and 'map' them on the way to inodes 1165074272 M * Bertl_oO bbl 1165074956 M * daniel_hozac Bertl_oO: btw, is there any particular reason we're not using the pidspaces? too complex to implement the spectator, or? 1165075167 M * daniel_hozac or is there just a partial implementation currently? 1165075456 Q * virtuoso Ping timeout: 480 seconds 1165075500 M * h01ger daniel_hozac, you're patch fails to apply against 2.6.18-5~bpo - which is not really the version in sid :-/ 1165075521 M * daniel_hozac h01ger: well, the patch is against 2.6.18.3-vs2.1.1.2.3 so that's not entirely unexpected. 1165075602 M * h01ger i guess i'll manage to clean that up myself, but i wont have any idea whether i do the right thing :-/ :) 1165075629 M * daniel_hozac well, it's just the reject in fs/namei.c, right? 1165075642 M * daniel_hozac btw, you're using http://people.linux-vserver.org/~dhozac/p/k/delta-secdev-poc03.diff, right? 1165075662 M * daniel_hozac (poc02 would require VXC_SECURE_MKNOD for fifos as well) 1165075672 M * h01ger also in context.h 1165075689 M * daniel_hozac ah, you get a reject there? well, same thing there though. 1165075690 M * h01ger oh, no, i was using 02 1165075706 M * daniel_hozac it's just adding a single line, so it'd be pretty hard to screw it up ;) 1165075715 M * h01ger sure :) 1165075725 M * h01ger i just wonder if the lines which are missing here might be needed: 1165075727 M * h01ger #define VXC_ADMIN_MAPPER 0x00200000 1165075727 M * h01ger #define VXC_ADMIN_CLOOP 0x00400000 1165075735 M * daniel_hozac nah, those were added in devel. 1165075748 M * daniel_hozac stable is just missing them, like many other things :) 1165075761 M * h01ger ah, ok, fine. then i'll make food and other stuff and try that later... 1165075810 J * blues blues@blysk.ds.pg.gda.pl 1165075816 Q * blues^ Remote host closed the connection 1165075830 M * daniel_hozac sounds like a plan, thanks. 1165076631 Q * sebastian 1165077291 Q * ms_ Ping timeout: 480 seconds 1165077791 J * virtuoso ~s0t0na@shisha.spb.ru 1165078729 J * comfrey ~comfrey@201.243.176.219 1165079549 Q * pmenier Quit: KVIrc 3.2.0 'Realia' 1165081966 M * rob-84x^ daniel_hozac: thanks for information on kprobes. my simple module now looks nice and works great ;) http://robert.nowotniak.com/pliki/velog.c 1165082049 M * daniel_hozac np. 1165082605 M * daniel_hozac Bertl_oO: looks like the problem is that proc_pid_readdir tries to lookup real pid 1. 1165084020 J * ms_ ~ms@arkansas.doc.ic.ac.uk 1165084067 J * dna___ ~naucki@150-215-dsl.kielnet.net 1165084495 Q * dna_ Ping timeout: 480 seconds 1165084795 Q * virtuoso Ping timeout: 480 seconds 1165085163 Q * michal` Ping timeout: 480 seconds 1165085441 J * michal` ~michal@www.rsbac.org 1165086422 P * NewSense 1165088189 J * virtuoso ~s0t0na@shisha.spb.ru 1165092213 Q * Johnnie Ping timeout: 480 seconds 1165092218 J * Aiken ~james@tooax6-174.dialup.optusnet.com.au 1165094797 Q * mire_ Remote host closed the connection 1165095077 J * FireEgl proteus@adsl-17-159-214.bhm.bellsouth.net 1165095200 Q * FireEgl Killed (services.oftc.net (Too many invalid passwords)) 1165095214 J * FireEgl proteus@adsl-17-159-214.bhm.bellsouth.net 1165097294 J * kugg_ ~kugg@81-163-35-108.event.dreamhack.se 1165098043 J * kugg__ ~kugg@81-163-35-108.event.dreamhack.se 1165098151 Q * kugg_ Ping timeout: 480 seconds 1165098799 J * hardwire` ~hardwire@rdbck-5182.wasilla.mtaonline.net 1165098904 Q * kugg__ Read error: Connection reset by peer 1165099216 Q * hardwire Ping timeout: 480 seconds 1165099925 J * loard_1 ~root@fwinternet.aui.ma 1165099939 M * loard_1 hello 1165099952 M * loard_1 who can help me i have fedora 5 and i need to configure it 1165099985 M * daniel_hozac in what way? 1165099999 M * loard_1 i need to install libraries 1165100003 M * loard_1 but i don't know the code 1165100010 M * loard_1 this is my first time to use linux 1165100031 M * daniel_hozac it sounds like you're looking for #fedora @ irc.freenode.net. 1165100053 M * loard_1 thats the server for fedora? 1165100076 M * daniel_hozac yes. 1165100081 M * loard_1 thank you so much 1165100271 J * mire ~mire@243-166-222-85.adsl.verat.net 1165100933 Q * loard_1 Quit: Leaving 1165103806 J * DavidS ~david@chello062178045213.16.11.tuwien.teleweb.at