1167350449 J * mrrm_ ~urkel@tor-irc.dnsbl.oftc.net 1167350893 Q * mrrm Ping timeout: 480 seconds 1167351635 Q * benj0r Quit: Pwet 1167352007 Q * dna_ Quit: Verlassend 1167354494 M * Bertl nap attack .. off for now .. 1167354499 N * Bertl Bertl_zZ 1167357559 J * Aiken_ ~james@tooax6-050.dialup.optusnet.com.au 1167357888 Q * Aiken Ping timeout: 480 seconds 1167359122 J * bronson ~bronson@pool-71-116-237-101.lsanca.dsl-w.verizon.net 1167359929 J * marcfiu ~mef@c-68-39-177-97.hsd1.nj.comcast.net 1167359934 P * marcfiu 1167360897 J * shedi ~siggi@inferno.lhi.is 1167361017 N * Bertl_zZ Bertl 1167361021 M * Bertl back now ... 1167364024 Q * duckx Remote host closed the connection 1167366649 Q * FireEgl Quit: ... 1167366678 J * FireEgl ~FireEgl@2001:5c0:84dc:1:211:9ff:feca:b042 1167372223 J * jmcaricand ~jmcarican@d90-144-91-194.cust.tele2.fr 1167372231 M * Bertl welcome jmcaricand! 1167372264 M * jmcaricand thank 1167372418 J * marcfiu ~mef@c-68-39-177-97.hsd1.nj.comcast.net 1167372426 M * Bertl wb marcfiu! 1167372431 M * marcfiu hey Bertl 1167372452 M * Bertl how's going? 1167372458 M * marcfiu well 1167372482 M * marcfiu Eric Biederman will be stopping by Princeton in, oh, 9 hours or so. 1167372502 M * Bertl great! send greetings from me ... 1167372511 M * marcfiu Will be picking his brains for a few hours. :) 1167372528 M * marcfiu Will send greetings along! 1167372538 M * marcfiu Do you have anything in specific that I should be asking him about? 1167372547 M * Bertl make sure that he doesn't convince you that the full network virtualization is the only way to go 1167372560 M * marcfiu right 1167372590 M * Bertl as it is not one of his requirements, the layer 3 isolation is something he likes to forget about, so please remind him :) 1167372601 M * marcfiu I will remind him. 1167372621 M * Bertl probably same goes for lightweigth init-less contexts 1167372635 M * Bertl (with a fake init) 1167372654 M * marcfiu From what I gather, it is an important element for sure in non-server environments (e.g., little laptops). 1167372679 M * Bertl yes, and I'm pretty sure we will make that possible in the future too 1167372680 M * marcfiu I don't think we will be chatting about anything else besides network virtualization. 1167372719 M * Bertl but it would be nice to have the functionality in mainline, instread of hacking it into an otherwise unused pid virtualization 1167372738 M * marcfiu You had mentioned that there might be a way to have network isolation operate side-by-side with full network virtualization vs. isolation being done within a virtualized network stack. 1167372758 M * Bertl yep, IMHO those are candidates for two namespaces 1167372768 M * marcfiu expand 1167372773 M * Bertl i.e. one layer 2 virtualzation, and one layer 3 isolation space 1167372792 M * Bertl the layer 3 isolation can also sit ontop of a layer 2 virtualization 1167372800 M * marcfiu how does one operate those to simultaneously yet side-by-side? 1167372834 M * Bertl simple, if you want 100% network virtualization, you 'clone' or 'unshare' a layer 2 space 1167372853 M * marcfiu or are you suggesting that one either choose just pure isolated stack, virtualized stasck, or isolation in virtualized stack? 1167372859 M * Bertl if you are fine with the much more lightweight isolation, you create a layer 3 space 1167372866 M * Bertl yep 1167372878 M * marcfiu is this something one can do dynamically or at kernel config time? 1167372886 M * Bertl dynamically 1167372896 M * Bertl that's why we want a space for that 1167372911 M * Bertl i.e. you decide at guest startup what type you wnat 1167372913 M * marcfiu isn't a space==virtualized stack? 1167372913 M * Bertl *want 1167372954 M * Bertl yes, but with the network virtualization at layer 2, the host has to be in some layer2 space already 1167372971 M * Bertl so by unsharing a layer 3 isolation, no overhead is occured 1167372986 M * marcfiu got it... 1167373021 M * Bertl OTOH, if you want a separate space (layer2) for the guest 1167373031 M * Bertl you simply unshare the layer2 space 1167373045 M * Bertl (you can then still use layer 3 isolation inside the guest) 1167373060 M * marcfiu I suppose the decision of incorporating isolation will be based purely on its performance & scalability. 1167373085 M * Bertl the isolation has, as we already showed, no measureable overhead 1167373099 M * Bertl the virtualization (layer 2) still needs to show that 1167373131 M * marcfiu right... so the good news is that one just needs to measure the performance in both time/space of layer 2 virtualization for 1,10,100,... guests. 1167373141 M * Bertl i.e. I think there is measureable overhead, but you folks should know better by now (testing *hint* ) 1167373175 M * marcfiu Sapan is working on that, but he got distracted to work on a paper for a workshop deadline next Friday Jan 4th. 1167373191 M * marcfiu Have you peeked at Eric's network virtualization code? 1167373204 M * marcfiu If so, how is it different from openvz? 1167373225 M * Bertl I got a glance at the early version, which wasn't even near working 1167373237 M * marcfiu My guess is that it hits a different point in the overall design spectrum. 1167373242 M * marcfiu oh 1167373242 M * Bertl there is also some 'proposed' implementation flying around 1167373244 P * marcfiu 1167373248 J * marcfiu ~mef@c-68-39-177-97.hsd1.nj.comcast.net 1167373255 M * marcfiu whoops 1167373262 M * Bertl boss button? :) 1167373264 M * marcfiu you'll have to repeat the last thing you typed. 1167373274 M * marcfiu accident. 1167373281 M * Bertl Bertl> there is also some 'proposed' implementation flying around 1167373286 M * Bertl lol @ accident 1167373296 M * marcfiu well, time for me to head to bed... 1167373298 M * marcfiu cheers 1167373306 M * Bertl okay, have a good one! cya! 1167373319 P * marcfiu 1167373740 J * justinwick ~justinwic@dynamic-acs-24-154-34-43.zoominternet.net 1167373952 Q * justinwick 1167374981 J * m`m`h ~simba@deb30.mgts.by 1167376053 Q * borgfish Ping timeout: 480 seconds 1167376075 Q * chaotika_ Ping timeout: 480 seconds 1167376320 J * chaotika chaotika@ccc2.rbg.informatik.tu-darmstadt.de 1167376636 Q * bronson Read error: Operation timed out 1167379368 Q * Aiken_ Quit: Leaving 1167379604 J * Aiken ~james@tooax6-050.dialup.optusnet.com.au 1167381498 J * borgfish ~bla@141.12.9.118 1167381502 M * borgfish moin 1167381509 M * Bertl good morning! 1167381531 M * borgfish :) 1167381880 M * Bertl hmm, frauenhofer, Integrierte Publikations- und Informationssysteme ... interesting ... 1167382552 Q * Aiken Quit: Leaving 1167382703 J * Aiken ~james@tooax6-050.dialup.optusnet.com.au 1167383021 M * meandtheshell Bertl: URL regarding the frauenhofer stuff is? 1167383040 M * Bertl ah, was just looking up borgfish's ip :) 1167383055 M * meandtheshell Bertl: lol - I see 1167383075 A * meandtheshell already checked heise and slashdot 1167383085 M * Bertl lol ... 1167383128 M * Bertl meandtheshell: but if you are interested in important stuff ... 1167383148 M * Bertl the Linux kernel now has Galois Field support for 2^128 *G* 1167383180 A * meandtheshell needed this since ages - now he's happy :) 1167383313 Q * jmcaricand Quit: Quitte 1167383660 M * meandtheshell Bertl: except the fields for scientific computing and cryptography, where (application areas) do you think resp. why do you consider this important? (not to say it is not ... I am just curious to hear your point of view ...) 1167383754 M * Bertl sorry, was a joke ... of course only cryptography will need that in the kernel 1167383776 M * Bertl nevertheless, it made me look up the short life of Evariste Galois ... 1167383882 M * meandtheshell I see - oh my god - 20 years 1167383928 M * Bertl yeah, who knows what he could have solved when he had made it to 35 ... 1167384050 J * prae ~benjamin@foxhound.sherpadown.net 1167384068 M * Bertl wb prae! 1167384080 M * prae mornin' 1167384389 Q * prae Quit: Pwet 1167384768 J * prae ~benjamin@foxhound.sherpadown.net 1167385144 M * borgfish IPSI is closed since 1.11 1167385152 M * borgfish thats why milf is LBF now 1167385249 M * sid3windr ah, with the CPP on the KBC I'm MIA on the FTP. :) 1167386787 M * Bertl okay, just uploaded a 2.3.x for 2.6.20-rc2 .. in case somebody wants to test it ... (usual place) 1167386803 M * Bertl and now I'm off to bed ... so have a good one everyone ... 1167386811 N * Bertl Bertl_zZ 1167388645 J * dna ~naucki@p54BCFD19.dip.t-dialin.net 1167389732 Q * dna Quit: Verlassend 1167390131 J * Tyan ~tyan@F2671.f.strato-dslnet.de 1167390199 M * Tyan hi everybody 1167390355 M * Tyan I'm using linux-vserver version 2.6.17.13-vs2.0.2.1 and although I've set CAP_SYS_RAWIO I have no /proc/sys/usb inside the vserver to access USB-devices. Is there anything else I need to do? 1167390466 M * borgfish well i really dont know and am not very skilled but i would try hardlinking the real proc device into the vservers proc 1167390486 M * borgfish maybe copy is enough 1167390532 M * Tyan hmm... but wouldn't i have to do this on every vserver restart then? 1167390545 M * borgfish dont think so 1167390573 M * Tyan ok, I'll give it a try 1167390737 J * click_ click@ti511110a080-3060.bb.online.no 1167390775 M * Tyan I copied it from the root system but the vserver cant see it though 1167390851 Q * click Ping timeout: 480 seconds 1167390870 M * borgfish sorry maybe you better wait for someone skilled :) 1167390949 M * Tyan well, got no other choice :) 1167390969 M * Tyan couldnt find anything in the FAQs about this issue 1167391013 Q * Aiken Ping timeout: 480 seconds 1167391808 Q * shedi Quit: Leaving 1167392857 J * yarihm ~yarihm@84-74-17-6.dclient.hispeed.ch 1167393735 Q * ensc Killed (NickServ (GHOST command used by ensc_)) 1167393745 J * ensc ~irc-ensc@p54B4FA31.dip.t-dialin.net 1167394714 Q * prae Quit: Pwet 1167395510 J * prae prae@foxhound.sherpadown.net 1167395591 J * lilalinux ~plasma@dslb-084-058-218-130.pools.arcor-ip.net 1167396013 Q * prae Quit: Pwet 1167396248 M * brcc w 1167396563 Q * mrrm_ Remote host closed the connection 1167396567 J * mrrm ~urkel@tor-irc.dnsbl.oftc.net 1167398172 M * daniel_hozac Tyan: it's been discussed on the mailing list, as well as here. 1167398209 M * daniel_hozac http://www.paul.sladen.org/vserver/archives/200610/0053.html 1167398245 Q * micah Remote host closed the connection 1167399109 J * micah ~micah@micah.riseup.net 1167399370 Q * yarihm Ping timeout: 480 seconds 1167400001 J * marcfiu ~mef@c-68-39-177-97.hsd1.nj.comcast.net 1167400006 P * marcfiu 1167400352 Q * sebastian|aw Quit: ( www.nnscript.de :: NoNameScript 4.02 :: www.XLhost.de ) 1167401059 J * Sebastian ~sebastian@office.star-hosting.de 1167401159 Q * Sebastian 1167401698 J * dna ~naucki@p54BCF307.dip.t-dialin.net 1167403783 Q * tso Ping timeout: 480 seconds 1167404307 J * Johnnie ~jdlewis@jdlewis.org 1167404821 Q * Johnnie Quit: G'bye! 1167405147 J * harvey ~maxwiezha@218.246.211.53 1167405198 P * harvey 1167405428 J * Johnnie ~jdlewis@jdlewis.org 1167405547 J * _dmax ~semaj@81.193.62.228 1167405813 J * s0undt3ch_ ~s0undt3ch@81.193.62.228 1167405896 Q * dmax Ping timeout: 480 seconds 1167405903 N * _dmax dmax 1167405973 Q * s0undt3ch Ping timeout: 480 seconds 1167405973 N * s0undt3ch_ s0undt3ch 1167406900 Q * dna Quit: Verlassend 1167408623 J * duckx ~Duck@tox.dyndns.org 1167409574 J * stefani ~stefani@tsipoor.banerian.org 1167412026 J * dna ~naucki@p54BCEB6A.dip.t-dialin.net 1167412773 M * matti Anyone want some coffee? 1167412925 N * Bertl_zZ Bertl 1167412928 M * Bertl morning folks! 1167413349 M * matti Bertl: Coffee? 1167413487 M * Bertl yes, please 1167413577 M * matti Aye. 1167413584 M * bXi morning Bertl 1167413626 A * matti hands a mug of fresh, hot and delicious coffee to Bertl! Enjoy! 1167413634 M * matti Anyone else? 1167413699 M * nebuchadnezzar I don't drink coffee 1167413724 M * matti nebuchadnezzar: Nobody's perfect ;p 1167413783 M * nebuchadnezzar I just have beer, belgium beer ;-) 1167413798 M * matti nebuchadnezzar: Bannana one? ;p 1167413861 M * matti nebuchadnezzar: I can make you some green tea if you like. But, this will take some time, since green tea must be prepare in very special way i order to keep it excellent and teasty. 1167413887 M * nebuchadnezzar I don't like tee :-) 1167413896 M * nebuchadnezzar I'll stay with my beers 1167414047 M * matti nebuchadnezzar: Aye. :) 1167416975 J * bonbons ~bonbons@83.222.37.103 1167416982 M * Bertl welcome bonbons! 1167416996 M * bonbons Hey Bertl! 1167417165 M * daniel_hozac Bertl: any new ideas on how to solve the sparc syscall problems? 1167417209 M * daniel_hozac Bertl: would using bcs do it? 1167417337 M * daniel_hozac and shouldn't l1 be in the clobber list? 1167417383 M * Bertl we can add that, yes, sec phone 1167418035 M * daniel_hozac nebuchadnezzar: could you test my ugly hack? 1167418061 M * nebuchadnezzar daniel_hozac: which one ? 1167418097 M * hardwire oh.. slam! 1167418107 A * hardwire snickers 1167418127 M * daniel_hozac http://people.linux-vserver.org/~dhozac/p/uv/experimental/delta-sparc-hack01.diff i don't even know if it'll compile as is... 1167418188 M * daniel_hozac (i guess it's time to dust off the sparcs) 1167418266 M * hardwire well I got an ISP to start using vserver 1167418272 M * hardwire woot.. now if only they would hire me 1167418273 M * hardwire heh 1167418437 M * Bertl daniel_hozac: okay, back again 1167418453 M * Bertl daniel_hozac: no, I don't think so 1167418467 M * Bertl the problem IMHO is again with the gcc optimizations/assumptions 1167418484 M * daniel_hozac well, one of the problems, yes. 1167418485 M * Bertl I had a deep look at the code, and the issue is with %o 1167418491 M * daniel_hozac oh? 1167418506 M * Bertl not being recognized as clobbered/changed/used 1167418508 M * nebuchadnezzar daniel_hozac: to the libc or diet ? 1167418518 M * daniel_hozac nebuchadnezzar: either. 1167418528 M * daniel_hozac nebuchadnezzar: i guess diet is the most relevant one though. 1167418535 M * nebuchadnezzar ok 1167418539 M * daniel_hozac Bertl: %o0, or all of them? 1167418562 M * Bertl sec, let me upload the IMHO relevant part 1167418599 Q * Roey Read error: Operation timed out 1167418628 M * nebuchadnezzar daniel_hozac: just make clean && make is sufficient ? 1167418635 M * daniel_hozac nebuchadnezzar: make clean shouldn't be required. 1167418644 M * daniel_hozac make should figure out which files need rebuilding on its own. 1167418671 J * Roey ~katz@h-69-3-4-130.mclnva23.covad.net 1167418695 M * nebuchadnezzar ok, but I have an error during compilation of lib/lib_libvserver_a-syscall.o 1167418702 M * nebuchadnezzar /tmp/ccGVOfHe.s:163: Error: symbol `success' is already defined 1167418721 M * daniel_hozac humm. 1167418760 M * daniel_hozac i guess the name needs to vary somehow. 1167418774 M * nebuchadnezzar it's during assembly 1167418783 M * nebuchadnezzar it's an assembler message 1167418788 M * daniel_hozac yeah. 1167418812 M * Bertl http://paste.linux-vserver.org/830 1167418839 M * daniel_hozac right, gcc is doing the wrong thing there. 1167418864 M * daniel_hozac but even so, that shouldn't be executed, should it? 1167418866 M * Bertl yes, the important part is, sec looking for the C code 1167418888 M * daniel_hozac __sysc_retv 1167418944 M * Bertl nebuchadnezzar: did you remove the vxid.E/S ? 1167418954 M * nebuchadnezzar arf no 1167418969 M * Bertl I'm getting 404 1167418977 M * Bertl http://www.asgardr.info/~nebu/vserver/vxid.E 1167418978 Q * Hollow Ping timeout: 480 seconds 1167418979 M * daniel_hozac i don't see them in the dir listing either. 1167419005 M * Bertl nebuchadnezzar: do you still have them somewhere? 1167419087 M * nebuchadnezzar Bertl: arf, you ask me for the vxid.o :-) I never do the .E/S 1167419112 M * Bertl hum, but it seems I downloaded it .. strange 1167419185 M * nebuchadnezzar Bertl: ok, now they are on my http 1167419207 M * Bertl tx 1167419273 M * Bertl http://paste.linux-vserver.org/831 1167419301 M * Bertl so we do here __err = __ret 1167419313 M * Bertl (local variable __err !!) 1167419327 M * Bertl then we call the __errno_location() 1167419337 M * Bertl and assign it the local variable 1167419374 M * Bertl unfortunately gcc decides to leave the __err = __ret out and 'remembers' that __ret is in %o0 1167419394 M * daniel_hozac yeah. 1167419394 M * Bertl and the procedure call also returns the address in %o0 1167419416 M * Bertl that's it, it basically writes the _errno location as error code 1167419423 M * daniel_hozac yep. 1167419438 M * daniel_hozac however, that branch shouldn't be getting executed at all. 1167419449 M * daniel_hozac i mean, the version syscall should succeed. 1167419451 J * Hollow ~hollow@styx.xnull.de 1167419467 M * Bertl good point 1167419503 M * Bertl and yes, l1 should be marked as clobbered 1167419516 M * daniel_hozac is it possible to do bcc [__HERE__ + 8] in some way? 1167419522 M * daniel_hozac i.e. avoid the use of labels? 1167419530 M * Bertl eheh ... no 1167419542 M * Bertl but we can use local labels 1167419560 M * Bertl I'll modify that, nebuchadnezzar can you test it? 1167419568 M * nebuchadnezzar sure 1167419601 M * Bertl daniel_hozac: shiny10 was the last version, yes? 1167419604 M * daniel_hozac yep. 1167419657 M * hardwire hmm.. 1167419672 M * hardwire so if I wanted to unify my instances.. 1167419678 M * daniel_hozac don't, hashify. 1167419682 M * hardwire I would shut them down and do some sort of magic? 1167419693 M * daniel_hozac hashify works even if they're running. 1167419702 M * daniel_hozac (though you'll have to restart them in order to see the benefits) 1167419728 M * hardwire I don't mind killing them 1167419738 M * hardwire how do the cow/vfs parts work? 1167419749 M * daniel_hozac what do you mea? 1167419751 M * daniel_hozac +n 1167419759 M * hardwire so I just hard linked tons of data 1167419779 M * hardwire I want that data to be able to change of course. when written to via a vserver 1167419787 M * hardwire without effecting other linked instances 1167419791 M * hardwire just curious how that works 1167419806 M * daniel_hozac COW will only trigger if the files are IUNLINK|IMMUTABLE. 1167419823 M * hardwire ok 1167419828 M * daniel_hozac (hashification would set that) 1167419839 M * hardwire so if I made a program to find hard linkable files.. then set them all to IMMUTABLE.. all is well? 1167419855 M * daniel_hozac uh, you realize that's what vhashify is, right? 1167419861 M * hardwire no actually.. i don't 1167419891 M * daniel_hozac but IMMUTABLE isn't sufficient. IUNLINK is required too. 1167419904 M * daniel_hozac and obviously, for COW, you'll need a 2.1+ kernel with CONFIG_VSERVER_COWBL=y. 1167419919 M * hardwire IUNLINK&IMMUTABLE then 1167419923 M * daniel_hozac |. 1167419927 M * daniel_hozac & would result in 0:) 1167419936 M * hardwire && 1167419944 M * hardwire + 1167419951 M * hardwire notor 1167419953 M * hardwire heh 1167421215 M * daniel_hozac Bertl: will the ,a there really do what we want? 1167421295 M * daniel_hozac ah, yes it will. 1167421356 M * Bertl okay, uploaded a new version, please reload :) 1167421372 M * Bertl nebuchadnezzar: please get the following file: 1167421377 M * Bertl http://vserver.13thfloor.at/Experimental/SYSCALL/syscall_shiny11.h 1167421392 M * nebuchadnezzar ok 1167421393 M * Bertl and put it into the vxid build dir, replacing the previous one 1167421419 M * Bertl i.e. name it szscall.h 1167421420 Q * dna Quit: Verlassend 1167421436 M * daniel_hozac s/z/y/ :) 1167421450 M * nebuchadnezzar ok, builded 1167421464 M * Bertl yeah, tx, I have three different keyboard layouts here, and x2x is confused 1167421489 M * nebuchadnezzar do I need to have a util-vserver installed ? I have none for the moment 1167421497 M * Bertl nope 1167421508 M * daniel_hozac ./vxid -V should work now. 1167421516 M * nebuchadnezzar ok so ./vxid -V give me vc_get_version: Unknown error 4160581024 1167421535 M * Bertl so that doesn't change anything, right? 1167421572 M * Bertl nebuchadnezzar: try make clean && make 1167421574 M * nebuchadnezzar I don't remember the error code before 1167421585 M * daniel_hozac Bertl: is 1 a special local label that's replaced by gcc, or is it the 1f in the bcc,a that makes it local? 1167421587 M * nebuchadnezzar Bertl: already done 1167421609 M * Bertl 1f means the next local label 1 (forward) 1167421615 M * Bertl 1b would be backwards :) 1167421617 M * hardwire http://pastebin.ca/295828 :( 1167421624 M * daniel_hozac ok, cool, thanks. 1167421706 M * Bertl nebuchadnezzar: 1167421718 M * Bertl please change the makefile to not do the strip 1167421731 M * Bertl then recompile and run the vxid with gdb 1167421760 M * nebuchadnezzar ok 1167421809 M * Bertl run it with a break @ main 1167421815 M * Bertl i.e. do 'break main' 1167421818 M * Bertl then r -V 1167421838 M * nebuchadnezzar breaked 1167421839 M * Bertl then step till you hit the 'ta as we did last time 1167422026 M * nebuchadnezzar with 'si' IIRC 1167422100 M * nebuchadnezzar that strange because I do not have the assembler instruction like last time 1167422118 M * Bertl ah, you need the display/i $EIP or so 1167422185 M * nebuchadnezzar display/i $pc 1167422191 M * Bertl zep 1167422193 M * Bertl yep 1167422201 J * WorkRoey ~katz@h-69-3-4-130.mclnva23.covad.net 1167422284 M * nebuchadnezzar well, can I break after the argument managment ? 1167422296 M * nebuchadnezzar arf, it's ok now 1167422299 M * Bertl good 1167422301 M * nebuchadnezzar I'm at the right place 1167422305 M * Bertl yeah, gdb is a bitch 1167422357 M * nebuchadnezzar well, I think it's not to be used directly 1167422362 M * nebuchadnezzar what do you want ? 1167422402 M * nebuchadnezzar for the moment I'm here "2: x/i $pc 0x10d60 : ta 0x10" 1167422448 M * Bertl excellent, now please do the 'info registers 1167422458 M * daniel_hozac does gdb show the condition code register? 1167422463 M * Bertl it should 1167422469 M * daniel_hozac ok. 1167422473 M * nebuchadnezzar I paste the ouput 1167422477 M * Bertl yes please 1167422488 M * Bertl ah, make that: 1167422494 M * nebuchadnezzar http://paste.linux-vserver.org/832 1167422494 M * Bertl info all-registers 1167422516 M * nebuchadnezzar http://paste.linux-vserver.org/833 1167422542 M * Bertl now step over the syscall 1167422555 Q * Roey Ping timeout: 480 seconds 1167422562 M * Bertl then do the info again (normal info is enough) 1167422594 M * nebuchadnezzar http://paste.linux-vserver.org/834 1167422624 M * Bertl what does the instruction display show now? 1167422637 M * nebuchadnezzar 2: x/i $pc 0x10e10 : sethi %hi(0x11000), %o0 1167422649 M * nebuchadnezzar before the info registers 1167422671 M * Bertl hmm, okay, let's redo that and use si instead of s 1167422679 M * nebuchadnezzar ok 1167422680 M * Bertl (when you are on td 1167422683 M * Bertl *ta 1167422745 M * nebuchadnezzar before ta or when ta is displayed ? 1167422755 M * daniel_hozac both, i guess. 1167422812 M * nebuchadnezzar well, when an instruction is displayed, it's executed ? or it will be executed at the next step ? 1167422820 M * Bertl next step 1167422822 M * nebuchadnezzar ok 1167422835 M * nebuchadnezzar so now I'm at the ta 1167422878 M * daniel_hozac info registers, si, info registers 1167422882 M * Bertl yep 1167422889 M * nebuchadnezzar not all registers ? 1167422894 M * Bertl nope 1167422949 M * nebuchadnezzar ok, http://paste.linux-vserver.org/835 1167423055 M * Bertl okay, let's do a few si and see what it does 1167423082 M * nebuchadnezzar http://paste.linux-vserver.org/836 1167423085 M * nebuchadnezzar before a call 1167423105 M * daniel_hozac Bertl: are %o0, %o1, %o2 really supposed to be 0? 1167423136 M * nebuchadnezzar they are %clr before the ta :-) 1167423147 M * nebuchadnezzar I supposed %clr is for clear ? 1167423167 M * daniel_hozac oh, 0 is the version command, nevermind me... 1167423197 M * Bertl everything looks fine so far 1167423248 M * nebuchadnezzar the cmp fail so it calls __errno_loca.... ? 1167423269 M * daniel_hozac nebuchadnezzar: are you running it as root? 1167423273 M * nebuchadnezzar no 1167423299 M * Bertl let's try that then :) 1167423305 M * nebuchadnezzar erf 1167423329 M * nebuchadnezzar sure now vxid -V give version: 0002:0102 1167423331 M * daniel_hozac that would explain why the version check errors, i guess :) 1167423352 M * daniel_hozac nebuchadnezzar: what happens if you use shiny10 and run it as root? 1167423362 M * Bertl yep, that would be interesting 1167423382 M * nebuchadnezzar can you give me the URI please ? 1167423390 M * daniel_hozac http://vserver.13thfloor.at/Experimental/SYSCALL/syscall_shiny10.h 1167423428 M * nebuchadnezzar it works version: 0002:0102 1167423487 M * daniel_hozac hehe 1167423589 J * deepsoul ~sandro@87.5.221.235 1167423684 M * Bertl welcome deepsoul! 1167423699 M * Bertl daniel_hozac: so that leaves us with the gcc confusions 1167423699 M * deepsoul hi Bertl 1167423750 M * Bertl daniel_hozac: I'll try something there, but we first have to fake a failing syscall as in the vhashify 1167423773 M * Bertl daniel_hozac: btw, why does the vhashify syscall fail? 1167423798 M * daniel_hozac i'm not entirely certain that it does now... 1167423956 M * daniel_hozac nebuchadnezzar: could you run vhashify through gdb and set a breakpoint at vc_get_iattr and display/i $pc, si until you get the ta? 1167423978 M * nebuchadnezzar the diet enabled tools or the libc only ? 1167423997 M * daniel_hozac diet, i guess. 1167424012 M * daniel_hozac shouldn't matter for vhashify though. 1167424077 M * nebuchadnezzar ok, I'll use a fresh util-vserver with none of your patches :-) 1167424503 J * eimann ~eimann@mrtg.cablesurf.de 1167424505 M * eimann hi 1167424532 M * hardwire hi 1167424538 M * hardwire what can I do for you this fine day 1167424539 M * eimann i get ncontext: vc_net_create(): Invalid argument with util-vserver-0.30.212 on 2.6.19.1-vs2.2.0-rc6, any idea how to fix this? 1167424547 M * hardwire did you reboot? 1167424569 M * eimann hm, nope, not yet ,but i can try that 1167424572 M * hardwire ok 1167424577 M * daniel_hozac eimann: you disabled dynamic contexts in your kernel configuration, yet you forgot to set a static xid for your guest. 1167424590 M * hardwire your no fun 1167424591 M * daniel_hozac eimann: echo > /etc/vservers//context should fix it. 1167424619 M * hardwire its a shame you can't just echo `uuidgen` > ... 1167424620 M * hardwire heh 1167424676 M * eimann daniel_hozac: i've set a static context id 1167424685 M * daniel_hozac eimann: and what is it? 1167424735 M * eimann it's between 49000 and 49005, i've 6 vserver 1167424736 M * eimann s 1167424753 M * hardwire so /etc/vservers/.../context = 49000 ? 1167424793 M * daniel_hozac eimann: so dmesg doesn't say anything about dynamic? 1167424798 M * eimann jep, hardwire 1167424804 M * hardwire ok 1167424808 M * eimann daniel_hozac: dynamic disabled in dmesg 1167424815 M * eimann one moment, telephone 1167424817 M * daniel_hozac eimann: there you go. 1167424837 M * hardwire he meant uno momento telefono 1167424842 M * hardwire *sigh* 1167424845 M * hardwire why am I so bored. 1167424851 M * hardwire and why do I have to bug you guys when I am 1167424865 M * daniel_hozac i have a rather long TODO list. 1167424871 M * nebuchadnezzar daniel_hozac: ok, I'm at the ta 1167424872 M * daniel_hozac if you're interested, i mean. 1167424880 M * daniel_hozac nebuchadnezzar: ok, could you paste the output thus far? 1167424880 M * hardwire daniel_hozac: I can't progrizzam 1167424884 M * Bertl hardwire: maybe try doing something useful? 1167424895 M * hardwire Bertl: working on a plone site atm 1167424901 M * hardwire I just hate css 1167424905 M * Bertl hardwire: like testing the 2.6.20-rc2 vserver kernel 1167424913 M * hardwire Bertl: eww.. why 1167424914 M * nebuchadnezzar http://paste.linux-vserver.org/837 1167424927 M * Bertl hardwire: because it would benefit from testing? 1167424929 M * daniel_hozac Bertl: is there anything interesting in 2.6.20? 1167424941 M * Bertl daniel_hozac: minor scheduler change 1167424990 M * daniel_hozac ok. 1167424997 M * eimann ah, cool. works. i've rebooted the box. now it works :) 1167425036 J * marcfiu ~mef@aegis.CS.Princeton.EDU 1167425036 M * daniel_hozac nebuchadnezzar: ok, please continue with the si until you get to the ta in vc_get_iattr, but try to keep the entire output. 1167425041 M * marcfiu hey 1167425046 M * daniel_hozac hello 1167425057 M * marcfiu had a good discussion with Eric Biederman today about network virtualization. 1167425076 M * daniel_hozac oh? 1167425088 M * marcfiu Bertl knows the details behind this. 1167425098 M * marcfiu Overall I think the discussion went well. 1167425123 M * marcfiu He showed us some new stuff that he's currently working on and will talk about on the container mailing list. 1167425128 M * daniel_hozac already? you have some sort of telepathic link? :) 1167425140 P * deepsoul 1167425146 M * Bertl daniel_hozac: yeah :) 1167425159 M * bonbons marcfiu: are the details somewhere, or could you make us a small summary? 1167425160 M * daniel_hozac awesome! that's even better than a crystal ball. 1167425166 M * marcfiu oh... met with him face-2-face and he we walked some code. 1167425219 M * marcfiu and described ideas... next best thing lacking a vulcan mind meld. :) 1167425232 M * bonbons so go for option 2 as option one is pretty difficult unless you had a recorder :) 1167425252 M * nebuchadnezzar daniel_hozac: do you want me to just do 's' and not 'si' over call ? 1167425271 M * daniel_hozac over which call? 1167425279 M * marcfiu bonbons: its primarily about L2 virtualization. 1167425302 M * marcfiu The basic idea is similar to Herbert's dual-headed loop back devices. 1167425338 M * bonbons dual-headed? don't remember something by that name... 1167425339 M * marcfiu The implementation probably will be more explicit compared to the approach taken by openvz, which should make it easier for others to grok the code and move forward. 1167425339 M * nebuchadnezzar I'm in utilvserver_checkCompatConfig and there is a call to vc_get_vci 1167425387 M * marcfiu bonbons: I'm not sure what Herbert called those p2p loopback devices, but the term dual-headed or similar came up about 6-10 months ago. 1167425404 M * daniel_hozac nebuchadnezzar: i guess you could just s over that one. it shouldn't be relevant for vc_get_iattr. 1167425410 M * daniel_hozac n, i mean. 1167425415 M * marcfiu Bertl: what did you call 'em again? 1167425505 M * bonbons there are two loop-back approaches I know about, contact-tagging loopback skbs or simulation of loopback addresses... 1167425517 M * Bertl marcfiu: hmm, no idea what you are talking about atm 1167425540 M * hardwire mwa ha ha 1167425544 M * hardwire thats terrible! 1167425567 M * Bertl marcfiu: you are referring to 'pipe' like interfaces, or so? 1167425572 M * marcfiu yes 1167425590 M * Bertl those are the source of all overhead there 1167425618 M * marcfiu yes... the overheady kind. 1167425641 M * Bertl marcfiu: did he give some date when 'his' network virtualization should be there? 1167425667 M * marcfiu I think he was motivated to get something done sooner, as we are willing to do performance testing etc. 1167425684 M * marcfiu within the next 2-3 weeks, but who knows for sure. 1167425800 M * Bertl okay, sounds good 1167425807 M * marcfiu it does. 1167425808 M * nebuchadnezzar daniel_hozac: Sorry but there is something I do not understand, I re-run the debuging session, I si since a call to "call 0xf7f2ace4 " and 's' made me enter in it, inside I get a ta... I'm quite loss 1167425836 M * marcfiu Bertl: I'm hopeful something good will come of this. 1167425838 M * Bertl nebuchadnezzar: looks good so far 1167425840 M * daniel_hozac what is it that you don't understand? 1167425852 M * nebuchadnezzar which ta do you want ? 1167425883 M * nebuchadnezzar Bertl: well, when I get out utilvserver_checkCompatVersion I do not find any ta... I'm running it again to be sure 1167425885 M * daniel_hozac the one in vc_get_iattr is the most interesting. 1167425899 M * daniel_hozac (there should be one. 1167425927 M * daniel_hozac nebuchadnezzar: you are running this as root too, right? :) 1167425933 M * nebuchadnezzar yes 1167425968 M * Bertl marcfiu: what was his overall position towards Linux-VServer? 1167426012 M * daniel_hozac nebuchadnezzar: you could just use n when you get to call utilvserver_checkCompatVersion, and run info register (o0 is the interesting one). 1167426025 M * daniel_hozac (run that after it) 1167426083 M * daniel_hozac ensc: ping? could you fix the download link at savannah? 1167426104 P * eimann going to frankfurt, afk 1167426124 M * daniel_hozac (and point it to http://ftp.linux-vserver.org/pub/utils/util-vserver/ instead) 1167426222 M * nebuchadnezzar daniel_hozac: When I get the call to utilvserver_checkCompatVersion, instead of s or si I do n... and then info registers 1167426238 M * daniel_hozac nebuchadnezzar: right. 1167426257 M * nebuchadnezzar well, I get o0 0xffffffff -1 1167426263 M * daniel_hozac or p ver might be better. 1167426296 M * daniel_hozac so p ver returns -1 too? 1167426338 M * ensc daniel_hozac: ok; should be updated 1167426353 M * daniel_hozac ensc: thanks 1167426360 M * nebuchadnezzar I got an error 1167426375 M * nebuchadnezzar http://paste.linux-vserver.org/838 1167426437 M * nebuchadnezzar ok, need to eat now 1167426496 M * daniel_hozac ok, thanks. 1167426589 M * nebuchadnezzar When I get out of utilvserver_checkCompatVersion, I quite quickly return from vc_get_iattr and never se the ta instruction 1167426601 M * nebuchadnezzar I'll made an output of the complete run after diner 1167426616 M * daniel_hozac yeah, it looks like the version syscall is failing. 1167426647 M * daniel_hozac you can set a breakpoint at utilvserver_checkCompatVersion and we'll focus on the ta there. 1167426655 M * nebuchadnezzar ok 1167426932 J * DreamerC_ ~dreamerc@125-225-96-36.dynamic.hinet.net 1167426975 J * Aiken ~james@tooax8-250.dialup.optusnet.com.au 1167427339 Q * DreamerC Ping timeout: 480 seconds 1167428593 J * ntrs ~ntrs@68-188-55-120.dhcp.stls.mo.charter.com 1167428656 M * nebuchadnezzar daniel_hozac: but that's strange, the problem was not here :-/ 1167428664 M * daniel_hozac hmm? 1167428682 M * nebuchadnezzar in utilvserver_checkCompatVersion, IIRC 1167428689 J * dreamind apwdsl@p548AB183.dip0.t-ipconnect.de 1167428695 M * daniel_hozac what do you mean? 1167428708 M * daniel_hozac it seems to be the cause of the problems, IMHO. 1167428727 M * dreamind Hi Folks ;) 1167428783 Q * dreamind 1167428790 J * dreamind apwdsl@p548AB183.dip0.t-ipconnect.de 1167428796 M * daniel_hozac hello 1167428811 M * nebuchadnezzar daniel_hozac: ok, I'm on the ta in utilvserver_checkCompatVersion 1167428829 M * daniel_hozac nebuchadnezzar: ok, info register, si, info register 1167428877 M * nebuchadnezzar http://paste.linux-vserver.org/839 1167428905 Q * dreamind 1167428912 J * dreamind apwdsl@p548AB183.dip0.t-ipconnect.de 1167428917 Q * dreamind 1167429058 M * daniel_hozac looks like it worked. 1167429062 M * daniel_hozac si again 1167429082 M * daniel_hozac then get the value of l1 from info registers. 1167429183 M * nebuchadnezzar 0x0 1167429226 M * daniel_hozac hmm, ok. do si a few more times. 1167429258 M * nebuchadnezzar I have a call to __errno__location@plt 1167429357 M * daniel_hozac si until you get to the ret from utilvserver_checkCompatVersion 1167429397 M * daniel_hozac then p res, p v_errno. 1167429422 M * nebuchadnezzar res is -135312416 1167429427 M * nebuchadnezzar v_errno is 0 1167429429 M * daniel_hozac hmm. 1167429436 M * daniel_hozac could you paste the output thus far? 1167429538 M * nebuchadnezzar ok, http://paste.linux-vserver.org/840 at the end 1167429836 M * Bertl hmm, cmp l1,0, bne? isn't that the wrong logic? 1167429907 M * daniel_hozac gcc optimizations again? 1167429969 M * daniel_hozac nebuchadnezzar: could you upload the output of objdump -d /usr/lib*/libvserver.so.0 | awk '/:/ { do_print=1 } /^$/ { do_print=0 } do_print==1 { print }'? 1167429975 M * Bertl well, it would make more sense to branch on eq, but mazbe that's just me 1167430061 M * daniel_hozac well, the test is if ((__err)), no? 1167430101 M * nebuchadnezzar http://www.asgardr.info/~nebu/vserver/utilvserver_checkCompatVersion.asm 1167430312 M * daniel_hozac are you sure that's the entire function? 1167430340 M * nebuchadnezzar that's the complete output of your command 1167430347 M * Guy- what's the best way to add a new network interface/IP to an existing vserver? edit the files in /etc/vservers/servername by hand? 1167430442 M * daniel_hozac if by edit you mean add, then yes. 1167430467 M * daniel_hozac nebuchadnezzar: that's strange, but the entire command looks wrong. 1167430475 M * daniel_hozac s/command/function/ 1167430516 M * daniel_hozac nebuchadnezzar: could you perform the "make-.S" routine for lib/checkversion.c? 1167430527 M * nebuchadnezzar sure 1167430613 M * nebuchadnezzar there is 2 checkversion... a lib_libvserver_a-chechversion.o and lib_libvserver_la-chechversion.lo 1167430617 M * nebuchadnezzar there are 1167430628 M * daniel_hozac ah right, the lib_libvserver_la_* one. 1167430639 M * nebuchadnezzar ok 1167430661 M * Guy- daniel_hozac: yes, add is what I meant 1167430766 M * nebuchadnezzar daniel_hozac: http://www.asgardr.info/~nebu/vserver/lib_libvserver_la-checkversion.S 1167431115 M * daniel_hozac okay, Bertl's right as always :) the bne is wrong. 1167431208 M * daniel_hozac or.. hum. 1167431216 P * marcfiu 1167431950 Q * bonbons Quit: Leaving 1167432026 M * daniel_hozac ok, it's correct. 1167432051 M * Bertl still the %o0 issue remains 1167432084 M * daniel_hozac indeed. 1167432145 M * Bertl but I still don't understand the check login 1167432161 M * Bertl if everything else is fine, the syscall should succeed, no? 1167432172 M * Bertl so the error code should not be stored, right? 1167432217 M * Bertl now iif something fails, then the return code will be bogus 1167432218 M * daniel_hozac well, errno is set to v_errno. 1167432227 M * Guy- I have a vserver that uses the host as its nameserver, and DNS replies appear to be coming from the isolated address of the vserver, not the one in /etc/resolv.conf 1167432241 M * Guy- I can mask this with iptables 1167432241 M * daniel_hozac res and v_errno are static variables. 1167432257 M * Guy- but is there an intrinsic vserver way to make it work like I think it should? 1167432270 M * Bertl Guy-: I assume your host's dns is bound to _all_ ips 1167432271 M * daniel_hozac run the nameserver in a guest? 1167432284 M * Bertl otherwise it would reply with the specific ip it is bound to 1167432298 M * Guy- Bertl: it is 1167432304 M * Guy- Bertl: so that's a no-no then? 1167432336 M * Guy- Bertl: for other reasons, I'd like it bound to more than one IP, but since it's DJB's dnscache, it can only bind to 0.0.0.0 or one particular IP 1167432356 M * daniel_hozac so, use chbind. 1167432362 M * Guy- but in that case, I'll bind it to one IP only and set up NAT to make it reachable on the other IPs 1167432378 M * Bertl Guy-: lol @ djb universe 1167432415 M * sid3windr hmm 1167432415 M * Bertl Guy-: you can use chbind to limit the 0.0.0.0 to a subset of host ips 1167432435 M * sid3windr although it's called "initialize" on the flower page, the script should be called "pre-start" 1167432441 M * sid3windr is initialize for .212 perhaps? 1167432447 M * sid3windr if so, shouldn't it be noted there? :) 1167432455 M * Guy- Bertl: even for processes running on the host? cool 1167432461 M * daniel_hozac pre-start isn't initialize. 1167432465 M * sid3windr hmm 1167432467 M * Bertl sid3windr: it is started before prepre-start :) 1167432470 M * sid3windr ah no 1167432473 M * sid3windr I see 1167432479 M * sid3windr but I had a script called initialize 1167432482 M * sid3windr and it wasn't run 1167432485 M * sid3windr so I did --debug 1167432494 M * daniel_hozac IIRC it was added in 0.30.211. 1167432495 M * sid3windr and it seemed to be looking for pre-start 1167432505 M * sid3windr pre-start is fine for me, I didn't really need initialize ;) 1167432508 M * daniel_hozac yep, 0.30.211. 1167432516 M * sid3windr I'm at 204 it seems 1167432520 M * sid3windr silly debian sarge ;) 1167432539 M * Bertl Guy-: strictly speaking, they will not be running 'on the host' anymore 1167432539 M * sid3windr I haven't added backports package on every host yet 1167432543 M * daniel_hozac use backports? 1167432550 M * sid3windr thanks ;) 1167432552 M * sid3windr yes, see above ;) 1167432552 M * Bertl Guy-: they will run in a separate network namespace 1167432564 M * Bertl Guy-: but in the same context space as the host uses 1167432591 M * Guy- Bertl: OK, that's good enough for me, thanks 1167432621 M * Bertl daniel_hozac: I should have a 4.1.1 sparc gcc soon, will try to mend the miscompilations regarding error value 1167432636 M * daniel_hozac okay. 1167432663 M * daniel_hozac i'm still curious as to why it's interpreting the return value as -1 though. 1167432788 M * daniel_hozac IMHO it should be the address of errno. 1167433000 M * Bertl that is a bug, the return code is -1 1167433010 M * Bertl the error code is in errno 1167433087 M * Bertl what gcc option causes the errno to become a function? 1167433169 M * Bertl i.e. -fpic or -fPIC? 1167433327 M * daniel_hozac neither, glibc does that, i think. 1167433343 M * Bertl hmm, not here :( 1167433364 M * Bertl but I guess I can fake that 1167433399 M * daniel_hozac maybe it's -D_REENTRANT 1167433415 M * Bertl I think it is more a define issued before the shiny include 1167433484 M * Bertl __sysc_setret 1167433496 M * Bertl and __sysc_seterr 1167433515 M * daniel_hozac echo 'errno' | gcc -E -include errno.h - returns (*__errno_location ()) here. 1167433538 M * Bertl ah, so that is defined there ... let's see 1167433613 M * Bertl is that an int or long ptr? 1167433658 M * daniel_hozac it's an int here. 1167433672 M * Bertl ah, I have something which is reasonably close to the issue 1167433730 M * Bertl http://paste.linux-vserver.org/841 1167433801 M * daniel_hozac that looks a lot better indeed. 1167433807 M * Bertl nope 1167433814 M * Bertl it trashes %o0 too 1167433821 M * daniel_hozac ? 1167433823 M * Bertl but that is what we actually want to see 1167433859 M * Bertl ah, no, that is correct :( 1167433878 M * Bertl okay, so it doesn't happen with gcc-4.1.1 :( 1167433900 M * daniel_hozac nebuchadnezzar: what command line did you use to compile it? 1167433919 M * nebuchadnezzar to compile what ? 1167433935 M * daniel_hozac lib_libvserver_la_checkversion... 1167433950 M * nebuchadnezzar http://paste.linux-vserver.org/842 1167434170 M * Bertl ah, I compiled the gcc 3.3.6 toolchain .. sec 1167434194 M * nebuchadnezzar you want me to try with gcc3.3 ? 1167434197 M * nebuchadnezzar 3.4 1167434202 M * daniel_hozac 3.4 was good, no? 1167434210 M * nebuchadnezzar yes 1167434214 M * nebuchadnezzar it works with 3.4 1167434227 Q * stefani Quit: I Quit! 1167434228 M * Tyan thanks for the replay, daniel_hozac ... has anything about this been worked out yet? 1167434265 M * daniel_hozac Tyan: see the first reply in that thread. 1167434269 M * daniel_hozac that worked for me. 1167434316 M * daniel_hozac Bertl: i still don't see how -1 ends up in %o0 in the success case. 1167434335 M * Tyan what usb-device were you trying to use? 1167434354 M * daniel_hozac Tyan: none, i just checked that the files were visible from the guest. 1167434397 M * Tyan hm... I'll check that out tomorrow and report back then 1167434413 M * daniel_hozac sounds good. 1167435493 Q * ruskie Quit: killed 1167435529 Q * DreamerC_ Quit: leaving 1167435571 M * Guy- I'm trying to find a clean way to have vservers be supervised by runit 1167435586 M * Guy- the current method of using vserver foo start doesn't fit this well 1167435601 J * DreamerC ~dreamerc@125-225-96-36.dynamic.hinet.net 1167435628 M * Guy- what I'd like to do is run a 'runit' or 'runsvdir' process in the vserver, without going through the sysv init stuff that 'vserver foo start' does 1167435648 J * ruskie ruskie@goatse.co.uk 1167435669 M * Guy- vserver exec would be nice, but it only works if I do vserver start first 1167435697 M * Guy- what else does vserver start do, other than run the initscripts of the vserver? 1167435727 M * Guy- should I perhaps just add runit as an 'initscript' in the vserver? 1167435728 M * daniel_hozac create the contexts, setup the IP addresses, set disk limits, mount things, etc. etc. 1167435748 M * Guy- daniel_hozac: and is there a way to do that without starting any processes in the vserver? 1167435761 M * daniel_hozac sure. 1167435769 M * daniel_hozac make the context persistent, and the initscript a noop. 1167435783 M * Guy- OK 1167435812 M * Guy- there is no initstyle to forgo starting anything, is there? 1167435827 M * daniel_hozac no. 1167435834 M * Guy- pity 1167435839 M * daniel_hozac just set apps/init/cmd.start to /bin/true. 1167435910 M * Guy- what's the default? 1167435924 M * daniel_hozac default what? 1167435934 M * daniel_hozac start command? it depends on the initstyle. 1167436002 M * Guy- for sysv? 1167436020 M * daniel_hozac /etc/{rc,init}.d/rc 1167436028 M * nebuchadnezzar go to bed now 1167436060 M * daniel_hozac ok, good night nebuchadnezzar! 1167436119 M * Guy- daniel_hozac: OK, I assume there is a cmd.stop for shutdown? 1167436127 M * daniel_hozac yes. 1167436157 M * daniel_hozac as http://www.nongnu.org/util-vserver/doc/conf/configuration.html says. 1167436205 M * Guy- indeed :) 1167436250 M * Guy- I'm sorry, I'm not really expecting many of the things I wonder about to be documented 1167436262 M * Guy- but I'll check before asking next time 1167436645 M * Guy- are vservers 'destroyed' when there are no processes left in them?