1113609766 M * Doener basically the same here, i really don't consider myself a kernel hacker... Linux-VServer was (and still is) a great project for me to learn about the kernel while doing something useful and actually getting involved into something. That's to a great deal like that because of Bertl, who has helped me to get where I stand now in several ways. (And because of me, being crazy enough to start really using Linux and reading/hacking kernel code nearly at the same 1113609767 M * Doener time ;). 1113609792 M * Doener IOW: i don't know for sure either, what's The Right Thing to do there ;) 1113610407 Q * ndim Ping timeout: 480 seconds 1113610697 J * ndim ~hun@helena.bawue.de 1113611009 N * Bertl_oO Bertl 1113611024 M * Bertl evening folks! 1113611034 M * Doener evening Bertl 1113611036 M * DaCa hi Bertl 1113611109 M * Bertl hey Doener! DaCa! 1113611118 M * Bertl do you have a short version for me? 1113611153 M * Bertl kevinp: still around? 1113611177 M * Doener http://pub.limehouse.org/ - take a look at the mlock_{vs,grsec,merge} files... we discussed how to do the merge the right way 1113611232 M * Bertl what's the orig files? 1113611234 M * Doener i assumed that replacing both changes to locked_vm would be right, and that it would need some closer look to fix that everywhere (i.e. also in those places where no conflicts occur) 1113611249 M * Doener vanilla mm/mlock.c ;) 1113611257 M * Doener ah, sorry, misread your question... 1113611308 M * DaCa Bertl: .orig are the diffs for the whole mlock.c 1113611313 M * Doener i'd say those contain all hunks for that file, while the others just contain the conflicting hunk 1113611349 M * Bertl okay, I guess I got it, thanks folks! 1113611367 M * Bertl what does the VM_MIRROR do? 1113611434 M * DaCa Bertl: PaX mirrors the VMA 1113611465 M * Bertl to get some overhead? 1113611502 M * DaCa Bertl: http://pax.grsecurity.net/docs/vmmirror.txt 1113611679 M * Bertl I don't see any point in doing that? besides it very likely breaks if you do not take special care with different splits ;9 1113611747 M * Bertl it's basically polluting the address space with various copies, what should that buy you? 1113612995 M * Bertl okay, now that we discussed that ;) how is 2.0-pre1 ? 1113613005 M * Doener *lol* 1113613071 M * Bertl well, forgive me, but I considered 20 minutes time enough to come up with some obscure justification ... *G* 1113613119 M * Doener i know, i just liked your way of expressing that ;) 1113613135 M * Doener (that was a compliment) 1113613162 M * Bertl ah, well, I just don't want that DaCa or anybody else doing the grsec stuff get's it wrong ... 1113613203 M * Bertl I'm just trying to figure out useful parts which might make sense to integrate in a hardened (hehe what a trem ;) vserver release 1113613208 M * Bertl s/trem/term/ 1113613320 M * Doener btw, did you see SNy complaining about the 127.0.0.1? he has a point there... we basically turn stuff listening on 127.0.0.1 from 'reachable only locally' to 'reachability depends on the vservers first assigned ip address' 1113613334 M * Doener s/listening/trying to listen/ 1113613490 A * Bertl just read up on that ... 1113613508 M * Bertl well, it is correct that we 'remap' 127.0.0.1 to the first ip 1113613535 M * Bertl this is part of how the 'original' network restriction makes localhost work ... 1113613580 M * Bertl I don't think that we can avoid that without at least the first half of ngnet ... 1113613708 M * Doener yep, i agree on that. and i thought that a minimalistic approach might be worth a try, i.e. only lo, only 127.0.0.1, hardcoded, maybe only used on demand 1113613758 M * Doener but i didn't do any research yet, what changes would actually be needed 1113613791 M * Bertl basically you would need to 'tag' local trafic (i.e. virtualized the lo) 1113613796 M * Bertl -d 1113613843 M * Bertl the problem is not the real local traffic, it is more the 'not strictly local' traffic between guests 1113613874 M * Bertl of course you could detect that via the ADDR_LOCAL() check 1113614245 M * Bertl okay, regarding 2.0-pre1, did you get around testing the reboot/helper stuff? 1113614340 Q * itamarjp Quit: 1113614385 M * Doener no, will do that now 1113614538 M * Bertl thanks! 1113614945 M * Doener hm, i should re-think my way of keeping my source trees... everytime i get a new patch from you, i basically end up with a complete rebuild... 1113614987 M * Doener (as i do: copy clean vanilla, patch, copy to build tree, build) 1113615033 M * Bertl doing copy vanilla, patch, diff to old copy, apply to working tree 1113615053 M * Bertl would probably simplify that ... 1113615160 M * Doener yeah, but that would break my way of working on new patches... while checking for build errors, i turn of backup in vim (if i know that that is fine) and thus modify the clean source tree and the build tree at once... 1113615184 M * Doener patching both, source and build tree, i'd break hardlinks 1113615206 M * Doener (or not create them in the first place) 1113615374 M * Bertl hmm, how do you turn of backup for 'some' vim sessions? 1113615587 M * Doener i just do :set bkc=yes when i need it 1113615617 M * Bertl hmm, okay, I would always forget that ;9 1113615735 M * Doener i got used to it... but probably some vimscript could do that automatically, based on the directory name for example... all my build trees are in a build/ directory 1113615918 M * Doener hum, hum... the vserver is just gone after reboot, does not come up again... 1113615958 M * Doener and i got: /usr/local/lib/util-vserver/vshelper: ( 10) returned with 256 1113615971 M * Doener is that correct? 1113615977 M * Doener (that was on startup) 1113616025 M * Bertl interesting ... 1113616047 M * Bertl please do 'echo 2 >/proc/sys/vserver/debug_switch' 1113616078 M * Doener hm, the second start gave: 1113616082 M * Doener /usr/local/lib/util-vserver/vshelper: (startup 10) returned with 512 1113616082 M * Doener /usr/local/lib/util-vserver/vshelper: ( 10) returned with 256 1113616088 M * Doener turning on the debugging now... 1113616106 M * Bertl the null is definitely wrong ... 1113616126 M * Bertl interesting is the fact that your helper bails out for some reason ... 1113616141 M * Doener because it doesn't know about startup (yet) 1113616145 M * Bertl IIRC this is somehow related to enricos strange exec script ... 1113616159 M * Bertl 512 and 256 should not be returned 1113616183 M * Doener hm, i actually got a delegate for startup... 1113616205 M * Bertl what does 'head -2 /usr/local/lib/util-vserver/vshelper' say? 1113616230 M * Doener #! /usr/local/lib/util-vserver/sigexec /bin/bash 1113616240 M * Bertl see, that's the problem 1113616248 M * Bertl remove the sigexec stuff, and it work 1113616251 M * Bertl +s 1113616272 M * Doener hmm... i used that setup for ngnet stuff... and IIRC it worked back then... 1113616295 M * Bertl sure? 1113616321 M * Bertl because I had those issues from the beginning ... 1113616321 M * Doener 99%... and after removing the sigexec, the output is still the same 1113616353 M * Bertl what about using testme.sh? 1113616414 M * Doener all suceeded, no message, startup fails with 512, shutdown with 256... 1113616453 M * Bertl hmm, testme works quite fine 1113616463 M * Bertl here, with no error at all ... 1113616496 M * Bertl but I have the sigexec stuff enabled too, strange ... 1113616512 M * Doener the 512 error seems to have been caused by the delegate script i had (ngnet stuff) with an empty script i get 256 1113616523 M * Bertl VCI: 0001:0025 273 03110076 1113616523 M * Bertl --- 1113616523 M * Bertl vxD: /sbin/vshelper: (startup 100) returned with 0 1113616523 M * Bertl vxD: /sbin/vshelper: (shutdown 100) returned with 0 1113616525 M * Bertl [001]# succeeded. 1113616539 M * Bertl (for example) 1113616589 M * Doener ok, i now put "exit 0" into those scripts, no more errors on startup/shutdown 1113616598 M * Doener but still the message if i start the vserver 1113616628 M * Bertl okay, I'll check that one ... maybe it is caused by the reboot ... 1113616676 M * Doener kevinp: that's also one for you ;) add "exit 0" to those files i told you to create, then you'll actually get rid of the error message, creating them isn't enough 1113617314 M * Bertl hmm, btw, legacy servers don't get a startup, but a shutdown .... 1113617472 M * Bertl argh! ;) 1113617489 M * Bertl okay, I'll do a fixup for the null ;) 1113617532 M * Doener hm, that sounds like something obvious :) 1113617587 M * Bertl it is .. it is ... 1113617828 M * Bertl http://vserver.13thfloor.at/Experimental/FOR-2.0/delta-vshelper-fix01.diff 1113617832 M * Bertl last hunk 1113617856 M * Doener ah, right, makes sense :) 1113618511 M * Doener ok, no more vshelper messages now 1113618543 M * Bertl good ;) 1113618549 M * Doener even with the sigexec stuff... 1113618584 M * Bertl yeah, sorry, it is active here too .. no idea why I believed that it has to be removed ... 1113618808 M * Doener still no success on reboot though... 1113618818 M * Doener the vshelper is called, but nothing happesn 1113618819 M * Doener vxD: /usr/local/lib/util-vserver/vshelper: (restart 10) returned sync with 0 1113618820 M * Doener vxD: /usr/local/lib/util-vserver/vshelper: (shutdown 10) returned sync with 0 1113618841 M * Bertl hum, the shutdown is executed ... 1113618854 M * Bertl that means that the restart managed to kill the vserver, right? 1113618873 M * Doener i guess so... 1113618905 M * Bertl so the problem is more the restart ... which of course can not any longer happen in the 'restart' handler 1113618986 M * Bertl the actual server restart must happen _after_ the shutdown 1113618991 M * Doener hm, right... the context will never be removed while the restart handler is still running, as we're doing that call in sync, right? 1113619044 M * Doener hm... did we have a race there before? that could explain the strange random succesful reboots i experienced 1113619151 M * Bertl I guess so 1113619163 M * Bertl that's why I wanted to make it sync ;) 1113619175 M * Doener heh :9 1113619276 M * Doener i'm pretty tired... since university started again, my day/night rhythm is even more messed up than before ;) getting up at 8am monday to wednesday and whenever i like to the rest of the week... somehow that's not a good thing to do i guess ;) 1113619331 M * Bertl well, if it doesn't kill you, it makes you stronger ... ;) 1113619341 M * Bertl anyway, I take this as a good night ;) 1113619407 M * Doener right :) have a good night Bertl, #vserver ! 1113619412 N * Doener Doener_zZz 1113625320 J * itamarjp lualele@200-225-242-013-dynamic.idial.com.br 1113625348 M * Bertl morning itamarjp! 1113625368 M * itamarjp hi 1113625368 M * mikegrb Bertl! it's evening you crazy man 1113625393 M * itamarjp Bert1: I am see your picture. 1113625401 M * Bertl mikegrb: hmm, I'm almost off to bed, so it must be morning ... 1113625408 M * mikegrb heh 1113625414 M * mikegrb I was about to do the same as well 1113625441 M * itamarjp I have a question about irc, Can anyone help-me ? 1113625466 M * Bertl itamarjp: s/I am/may I/ or s/am see/am seeing/ ? 1113625509 M * itamarjp Bert1: I am sorry for my bad english (I am from brazil, I speak only portuguese.) 1113625528 M * Bertl completely wrong, I would not understand you ;) 1113625539 M * itamarjp I am was see your picture. 1113625552 M * itamarjp I donīt know how to speak in english 1113625556 M * Bertl so you did see my picture *G* ? 1113625566 M * itamarjp Bert1: yes 1113625593 M * itamarjp Bert1: you have a long hair.... 1113625611 M * Bertl excellent, now you know who will hunt you down if I find you flooding another channel ;) 1113625658 M * itamarjp me ? 1113625672 M * itamarjp I am not flooding any channels. 1113625705 M * itamarjp Bert1: Why you say this ? 1113625737 M * Bertl because I remember you flooding this and #kernelnewbies ;) 1113625769 M * itamarjp no, are you wrong..... 1113625784 M * Bertl http://irc.13thfloor.at/LOG/2005-04/LOG_2005-04-13.txt 1113625830 M * Bertl or do you have somebody else using your accounts? ... 1113625836 M * Bertl anyway, it was more a joke ;) 1113625839 M * itamarjp I am only asked about how to resize a disk image. 1113625875 M * Bertl well, 28 lines in a row are considered flooding ... 1113625907 M * itamarjp Bert1: I am sorry, I am new in irc. 1113625929 M * Bertl fine, let's just forget about it ... okay? 1113625934 M * itamarjp I am new in vserver to... 1113625994 M * itamarjp Bert1: ok, I will send only single lines..... 1113626127 M * Bertl the rule is usually, if it fits in 3-4 lines, paste it here, if not use pastebin.com or simialr 1113626176 M * itamarjp Bert1: Can you help-me about irc ? 1113626203 M * Bertl there are a bunch of IRC newcomer howtos out there 1113626226 M * Bertl just use google to search for IRC, HOWTO and maybe LINUX 1113626266 M * itamarjp Bert1: ok, I will search in google. 1113626272 M * Bertl a nice client IMHO (in my humble opinion, see acronym finder ;) is irssi 1113626919 Q * itamarjp Quit: 1113628277 M * Bertl okay, off to bed now ... back later ... 1113628283 N * Bertl Bertl_zZ 1113628967 J * alex234 new@p548C97CE.dip.t-dialin.net 1113632712 Q * sebd Read error: Operation timed out 1113633302 Q * alex234 Read error: Connection reset by peer 1113634497 J * sebd ~sebd@lesdeveloppementsdurables.org 1113635485 Q * sizo Quit: leaving 1113637730 J * erwan_ho ~erwan@lns-vlq-39f-81-56-133-136.adsl.proxad.net 1113639442 J * kalou ~kalou@80.13.75.19 1113639446 M * kalou 'lo 1113639584 P * erwan_ho Leaving 1113640161 Q * sebd Ping timeout: 480 seconds 1113641886 Q * kalou Ping timeout: 480 seconds 1113643460 J * kalou ~kalou@AToulon-201-1-24-151.w80-13.abo.wanadoo.fr 1113644921 Q * kalou Ping timeout: 480 seconds 1113644930 Q * Beirdo Remote host closed the connection 1113644994 J * Beirdo ~gjhurlbu@beirdo.usercloak.oftc.net 1113647104 Q * Jdogg Read error: Connection reset by peer 1113647105 J * alex234 new@pD955EBB5.dip.t-dialin.net 1113647119 J * Jdogg Jdogg@c-24-1-174-133.hsd1.tx.comcast.net 1113647181 Q * alex234 Read error: Connection reset by peer 1113647528 J * alex234 new@pD955EBB5.dip.t-dialin.net 1113647664 P * alex234 1113649801 J * erwan_ho ~erwan@lns-vlq-39f-81-56-133-136.adsl.proxad.net 1113650216 J * alex234 new@pD955EBB5.dip.t-dialin.net 1113650364 P * alex234 1113650942 J * alex234 new@pD955EBB5.dip.t-dialin.net 1113650965 P * alex234 1113653192 Q * erwan_ho Remote host closed the connection 1113653365 J * erwan_ho ~erwan@lns-vlq-39f-81-56-133-136.adsl.proxad.net 1113653518 Q * erwan_ho Remote host closed the connection 1113654838 J * erwan_ho ~erwan@lns-vlq-39f-81-56-133-136.adsl.proxad.net 1113658475 N * Doener_zZz Doener 1113658488 M * Doener morning 1113659011 M * DaCa morning Doener 1113659028 M * Doener hey DaCa, what's up? 1113659057 M * DaCa me, i'm just up :) 1113660533 Q * maharaja Ping timeout: 480 seconds 1113660923 J * DuckMaster ~Duck@dyn-83-157-176-183.ppp.tiscali.fr 1113661045 Q * duckx Read error: Operation timed out 1113661296 J * mcp ~hightower@217.171.201.37 1113661616 M * mcp re 1113662190 M * mikegrb er 1113663578 M * mst spoing 1113663905 N * Bertl_zZ Bertl 1113663915 M * Bertl morning folks! 1113663923 M * Bertl hey mcp! is it really you? 1113663965 M * TheSeer hey bertl ;) 1113663968 M * TheSeer hey everyone *g* 1113663973 M * Bertl hey TheSeer! 1113663998 M * TheSeer i just learned Longhorn is going to have a brand new revolutionary feature 1113664016 M * TheSeer they'll add a firewall that can scan OUTGOING traffic 1113664022 M * TheSeer *woooow* 1113664028 M * TheSeer amazing, isn't it? 1113664038 M * TheSeer i call that "freedom to innovate" (tm) 1113664040 M * TheSeer ;-P 1113664050 M * TheSeer tell me again, why do people buy windows??? 1113664100 M * Bertl because of the marketing ;) 1113664203 M * TheSeer ;) 1113664320 M * mst because it does the job for them 1113664322 M * mst sadly 1113664353 M * TheSeer that reminds me on the marketing blabla for windows 95 1113664360 M * mst and remember very, VERY few people actually buy windows as such 1113664368 M * TheSeer Thanks to microsoft, filenames now can be longer then 8.3 chars 1113664388 M * TheSeer now if that wasn't something the world hadn't seen before... 1113664390 M * TheSeer erm... 1113664392 M * TheSeer or did it? 1113664420 M * TheSeer i mean.. it's not like even Windows NT 3.5 was able to do that 1113664432 M * TheSeer but hey, windows 95 was a very innovative product 1113664439 M * TheSeer sometimes i just hate marketing ;) 1113664450 M * TheSeer (and i do work in an advertising agency ;-> ) 1113664473 M * Bertl and don't forget, microsoft brought us the fancy tools (Microsoft Services for Unix) so now we can use sed and awk ;) 1113664531 M * TheSeer :) 1113664545 M * TheSeer to bad those are bsd licensed 1113664559 M * TheSeer it would be kewl to actually sue m$ for license abuse ;> 1113664679 M * Bertl hey after all they 'invented' the gnu de-buggler ;) 1113664698 M * Bertl or no it was gnu be-duggler ;) 1113664758 M * Bertl http://www.google.com/url?sa=U&start=1&q=http://www.microsoft.com/windows/sfu/productinfo/features/default.asp&e=8092 1113664809 M * Bertl including make, rcs, yacc, lex, cc, c89, nm, strip, gbd ... 1113665124 M * Bertl @all so how is 2.0-pre1 for you? 1113666117 M * Doener hi Bertl! 1113666190 M * Doener seems the hints in the 2.6 guide do their magic :) the related problem-report on the ml was pretty nice IMHO 1113666290 M * Bertl yup, and morning! 1113666401 M * Bertl btw, it might be good to have 'lables' there to reference from other pages (folks did that for some parts of the paper) 1113666409 M * Bertl s/lables/labels/ 1113666448 M * Bertl that would allow the -install-howtos to be reduced to a minimum ... 1113666468 M * Bertl (referencing the relevant sections from there) 1113666481 M * Doener yep, wanted to add some, once it's 'finished'... don't want it to mess up other pages' links if i choose to restructure anything 1113666667 M * Bertl okay, off again for now ... back later ... 1113666674 N * Bertl Bertl_oO 1113667509 Q * mst Quit: leaving 1113668257 N * Bertl_oO Bertl 1113668270 M * Bertl back now ;) 1113668319 M * Bertl Doener: any ideas regarding the reboot; how to simplyfy it or such? 1113668425 M * Doener we could have a mark on the context. the sys_reboot call is the last call anyways. we could then mark the context and issue a "vshelper reboot" instead of a "vshelper shutdown" when the context goes away 1113668481 M * Bertl well, we could do that in userspace too (e.g. mark the vserver for reboot, and act accordingly on shutdown 1113668518 M * Bertl I think the issue will be that the shutdown vshelper call is no good place to restart the vserver ;) 1113668611 M * Bertl personally I'd prefer something in userspace similar to the inetd (which checks if a vserver should be started, and if the vserver was stopped, simply restarts it) 1113668712 M * Bertl an alternative would be to add real 'init' restart support from the kernel side ... (i.e. when init exits, it is restarted by some kernel thread) 1113668737 M * Bertl but basically I don't think it's worth the overhead 1113669096 M * Doener hum... 1113669149 M * Doener as we can't change all settings on the fly, a "init 0; init 3" isn't enough either... 1113669224 M * Bertl well, the following should work in any case: 1113669266 M * Bertl in the startup do: rm -f /var/run//reboot 1113669278 M * Bertl in the reboot do: touch /var/run//reboot 1113669301 M * Bertl in the shutdown do: (/path/to/reboot-script &) 1113669314 M * Bertl hmm, sorry: 1113669326 M * Bertl [ -f /var/run//reboot ] && (/path/to/reboot-script &) 1113669342 M * Bertl and the reboot script would do: 1113669365 M * Bertl while [ -d /proc/virtual/ ]; do sleep 0.5; done 1113669376 M * Bertl vserver start 1113669567 M * Bertl of course, I could also do a second shutdown call, once the context is gone ... 1113669728 M * Bertl so, in which direction should we go with 2.0 (but keep future ngnet in mind ;) 1113669747 M * Doener hmm... 1113669776 M * Bertl busy? no problem ;) 1113669790 M * Doener brain's working slowly 1113669847 J * ripman ~leukejong@d594e3d7.dsl.concepts.nl 1113669848 M * Bertl of course we have yet another alternative ... 1113669852 M * Bertl welcome ripman! 1113669887 M * Bertl Doener: we could add an userspace daemon, connected via netlink which receives all the various events (enrico and I talked about something like that some time ago) 1113669908 M * ripman hello 1113669919 M * ripman chbind: vc_set_ipv4root(): Function not implemented what does this mean when i start the vserver ? 1113669957 M * Doener either you're not using a vserver kernel or you didn't choose network legacy support while configuring the kernel 1113669963 M * Bertl there are some resons for that; the most likely one is that the support is missing in your kernel 1113669987 M * Bertl ripman: give the testme.sh script a try ... 1113669998 M * ripman ohh i see i am new into vserver stuff.. 1113670005 M * ripman i downloaded a kernel with was patched 1113670038 M * Bertl hmm, sure? 1113670038 M * Doener http://vserver.13thfloor.at/Stuff/SCRIPT/testme.sh 1113670071 M * ripman Linux version 2.4.18ctx (root@biochip) (gcc version 2.95.4 20011002 (Debian prerelease)) #1 Tue Mar 2 20:05:50 CET 2004 1113670086 M * ripman i downloaded it from a vserver site 1113670098 M * Bertl hmm, quite an ancient one ... 1113670105 M * Bertl what tools do you use? 1113670147 M * ripman debian-vservers tools 1113670216 M * ripman yes just testing it with the old kernel 1113670219 M * Bertl well, I'd suggest to build your own kernel, but of course running the testme.sh can't hurt ... 1113670224 M * ripman okay 1113670227 M * ripman let me test the .sh 1113670792 Q * grecea Remote host closed the connection 1113670817 M * Doener dinner time.. back in a few... (minutes ;) 1113670821 N * Doener Doener|gone 1113670866 J * Tbery ~tb@pha-84-242-95-4.nat.karneval.cz 1113670874 M * Bertl welcome Tbery! 1113670887 M * Tbery Hi Bertl 1113670897 M * Tbery How are you?? 1113670906 M * Bertl fine thanks! and you? 1113670910 M * Tbery to. 1113670916 M * Tbery new contrakts.. 1113670924 M * Tbery after exibithion.. 1113670957 M * Tbery Me and friend we would like to make some vservers aplication.. 1113670965 M * Tbery for monitoring vservers.. 1113670996 M * Tbery You can help us with it..if you want,... 1113671075 M * Bertl well, sure I help you, it's just a question of how much time I have for that ;) 1113671084 M * Tbery ok] 1113671162 M * Tbery how can I set up for users vserver withch aplication he cant install?? 1113671210 M * Bertl hmm, you want to 'restrict' the guest root user so that he can not install some applications? 1113671389 M * Tbery yes 1113671400 M * Tbery no in virtual server 1113671417 M * Tbery deny install some aplication.. 1113671418 M * Bertl virtual server == guest, no? 1113671423 M * Tbery nfs portmap 1113671437 M * Tbery yes 1113671444 M * Tbery in guest 1113671462 M * Bertl I don't see a good way to deny them some apps, unless you do not give them root in the first place 1113671501 Q * monrad Ping timeout: 480 seconds 1113671530 M * Tbery yes..but if I will give one virtual server somebody else.. 1113671536 M * Tbery as his own.. 1113672246 Q * ripman Ping timeout: 480 seconds 1113672339 J * ripman ~leukejong@d594e3d7.dsl.concepts.nl 1113672710 Q * Tbery Quit: Ukončuji 1113672782 N * Doener|gone Doener 1113672797 M * Bertl and, did you enjoy your dinner? 1113672851 M * Doener sure i did! ;) 1113672869 M * Bertl excellent! 1113672882 A * Doener always enjoys food 1113672890 M * Bertl yeah, me too ... 1113672907 M * Bertl had a good self made pizza in the meantime ;) 1113673067 M * Doener i just had some bread today, we're invited for lunch at a chinese restaurant tomorrow, so i mustn't eat too much today ;) 1113673305 M * Bertl i.c. 1113673331 Q * ripman Ping timeout: 480 seconds 1113675801 J * yarihm ~yarihm@217-162-204-144.dclient.hispeed.ch 1113676064 M * Bertl welcome yarihm! 1113677206 Q * DaemonDazz Ping timeout: 480 seconds 1113677836 M * yarihm yo Bertl 1113678872 M * Hollow heya 1113678882 M * Doener hi Hollow! 1113678885 M * Bertl hey Hollow! 1113678900 M * Hollow Bertl: http://www.underemotion.com/preview/Unbenannt-2.png 1113679049 M * Doener wiki design suggestion? 1113679056 M * Hollow yup 1113679088 M * Doener the name is Linux-VServer ;) but that's not too important for a suggestion i guess... 1113679122 M * Doener how does the design perform in lynx/elinks(2)? 1113679128 M * Bertl looks good, can I see/test it in action somewhere? 1113679179 M * Bertl especially with some
-formatted text and a little more in the menu?
1113679258 M * Hollow no, just a graphic atm, but i can implement it if you like
1113679261 M * Doener IMHO the hard borders of the sub-menu and the headline would fit better if it gets smoothed a little as the header's bottom border is made to appear rounded, it's looks a little not fitting together
1113679298 M * Bertl Hollow: well, I can say nothing without testing the functionality, can I?
1113679336 M * Hollow you can, it's just step-by-step ;)
1113679352 M * Bertl okay, if it works as expected, it looks fine, better?
1113679385 M * Hollow yeah, jsut wanted to know if it looks ok, and you said it does, so implemenation would be next then ;)
1113679847 M * Bertl but don't misinterpret me here, I do not really care about the 'look' so much, others care, and I leave the judgement to them ... I  _do_ care about the functionality and that it looks similar on all browsers ...
1113679880 M * Doener Bertl: did you have a look at git yet?
1113680121 J * ripman ~leukejong@d594e3d7.dsl.concepts.nl
1113680152 M * Bertl Doener: no, no time yet ...
1113680186 M * Doener ok... i'll have a look at it now... let's see if we could make use of that
1113680344 M * DaPhreak well i tried it, on monday or so .. but i must tell you that i'm obviously to stupid for this git thingy :)
1113680448 M * Doener well, i just 'make'd it... and now? *g*
1113680503 M * DaPhreak that's what i also acomplished ... cd git-0.04; make :P
1113680536 M * Hollow hm, anything known about java causing problems in a vserver?
1113680539 M * Doener the README says what it does (theoretically), but not how you make it do that
1113680563 M * DaPhreak yeah that was also my problems (no docs)
1113680573 M * DaPhreak s/problems/problem/
1113680606 Q * ripman Ping timeout: 480 seconds
1113680619 M * DaPhreak hmm .. Hollow, nothing heard so far .. as far as i know a buddy of me got jakarta and friends running inside a vserver
1113680651 M * Hollow hm, cannot compile with javac, or even get the help displayed... gets a SIGKILL
1113680673 M * Hollow http://phpfi.com/58141
1113680675 M * Bertl well, use strace on it, and if that doesn't work, then gdb?
1113680683 M * Hollow this is the strace
1113680729 M * Doener maybe you hit some limit?
1113680763 M * Hollow hm..
1113680827 M * Hollow it's a standard installation.. not hard-cpu sched or so...
1113680844 M * Bertl I'd go for compiler bug ;)
1113680887 M * Doener schedhard would kill the process ;) i was thinking of ulimits...
1113680908 M * Hollow how can i find out the limits?
1113680913 M * Doener ulimit -a
1113680924 M * Bertl indeed, could be possible ...
1113680932 M * Hollow stoned root # ulimit -a
1113680936 M * Hollow core file size        (blocks, -c) 0
1113680940 M * Hollow data seg size         (kbytes, -d) unlimited
1113680944 M * Hollow file size             (blocks, -f) unlimited
1113680948 M * Hollow max locked memory     (kbytes, -l) 32
1113680952 M * Hollow max memory size       (kbytes, -m) unlimited
1113680955 M * Doener s/schedhard would/schedhard wouldn't/ sorry ;)
1113680956 M * Hollow open files                    (-n) 1024
1113680960 M * Hollow pipe size          (512 bytes, -p) 8
1113680964 M * Hollow stack size            (kbytes, -s) 8192
1113680964 M * Hollow cpu time             (seconds, -t) unlimited
1113680964 M * Hollow max user processes            (-u) 1983
1113680964 M * Hollow virtual memory        (kbytes, -v) unlimited
1113680998 M * Doener interesting process limit ;) 
1113681013 A * Hollow shrugs
1113681029 M * Doener reminds me that my gf has a birthday soon... (it's her year of birth)
1113681037 M * Hollow lol
1113681178 M * DaPhreak haha .. same here ;P and my birthday is also not so far away ..
1113681191 J * monrad ~monrad@213083190130.sonofon.dk
1113681269 M * DaPhreak damn am i old :P on wednesday, or was it thuesday ?
1113681301 M * Hollow will util-vserver go stable along with the kernel patches?
1113681341 M * Doener what shall others say? i'm about a year older than you... and i guess we also have some folks here being even older^Wwiser
1113681351 M * Hollow haha
1113681378 M * Doener ;)
1113681412 M * DaPhreak heh .. na, don't be ashamed Hollow :P you'll also get older^H wiser some day *runs away*
1113681425 M * DaPhreak s/ashamed/blue/
1113681437 M * Hollow sure
1113681439 M * Hollow :P
1113681467 M * DaPhreak *shurgs* methods of activity analysis
1113682011 J * Nik ~Nik@cable-153-130.online.bg
1113682016 M * Nik hi all
1113682019 M * Doener welcome Nik 
1113682031 M * Bertl Hollow: well tools and kernel will get stable by testing and reporting bugs ;)
1113682035 M * Bertl hey Nik!
1113682259 M * Nik w0w! 2.0-pre1! respect! :-)
1113682318 M * DaPhreak hmm Bertl, is there any diff/patch for 2.0_pre1 and ngnet ?
1113682392 M * Bertl nope
1113682603 J * kjo ~krischan@pD9E2DE72.dip.t-dialin.net
1113682604 Q * kevinp Read error: Connection reset by peer
1113682619 M * Bertl welcome kjo!
1113682643 M * kjo hi bertl
1113682802 M * Doener ok, now i got git-pasky... that seems to be more usable for one not knowing what this is all about ;)
1113682845 M * Bertl lol
1113683826 Q * Nik Ping timeout: 480 seconds
1113683835 J * Nik ~Nik@cable-153-130.online.bg
1113684321 Q * Nik Ping timeout: 480 seconds
1113684800 Q * monrad Quit: Leaving
1113684877 M * DaPhreak DaCa: you should definitly look into your grsecurity/gracl_shm.c since it fails during compilation (simply add #include  to the list of includes)
1113684945 M * Bertl careful, IIRC there are some weird include dependancies with grsec
1113684954 M * DaCa DaPhreak: yes, it has been corrected in the current patch
1113684957 M * Bertl so it might be that this simply depends on your grsec config
1113685144 M * DaPhreak yeah .. if you actually doesn't enable the grsec stuff you wouldn't trigger this :)
1113685205 M * DaCa DaPhreak: thats why I didnt notice it on my first compile as I just fed it a .config with everything as default
1113685248 M * DaPhreak yeah .. same here at the first time .. :)
1113685480 M * Bertl who is doing the grsec stuff atm?
1113685551 M * DaPhreak well DaCa is doing the current
1113685566 M * DaPhreak since i wiped out all my grsec stuff
1113685583 M * Bertl DaCa: okay, you ahve to 'track' the dependancies of the include files and how they change with grsec options
1113685609 M * Bertl and to make sure that the vs_* files are not included twice ...
1113685881 M * DaCa Bertl: as far as I understood in this case, grsecurity/gracl_shm.c is using a some macro that got overridden by vserver, calling a vserver macro that it doesnt know about and as such fails with an undefined reference when linking vmlinuz
1113685939 M * Bertl could happen, don't know the error in detail ;)
1113685954 M * Bertl but I'd suspect it to be simpler ;)
1113685982 M * DaPhreak i'll show you in prox. 2min's
1113686028 M * Doener hmm... i still didn't suceed in creating a git repository (or whatever that storage thingy is called) and commiting files...
1113686067 M * DaPhreak heh, seems like i'm not to stupid to handle git ;)
1113686095 M * Doener well, then tell me how to use it...
1113686123 M * DaCa what is git?
1113686136 M * DaPhreak well same as you Doener, i couldn't create a repo :) that was the point were i get stuck
1113686157 M * Doener "GIT - the stupid content tracker" - Linus' temporary(?) bk replacement
1113686181 M * Doener DaPhreak: ah, so you meant you're no the only one being too stupid, right?
1113686194 M * DaPhreak yeah :)
1113686259 M * Doener ok, that's fine then ;)
1113686262 M * DaPhreak well if there were some --help or so .. would be nice :) or even a better README
1113686342 M * Doener i'm using git-pasky now... that has a central script called (who would have expected) "git" and it even has a "git help"... but that doesn't tell me anything either...
1113687021 M * Doener hm, now that I actually _read_ the error message it makes sense... xargs can't find a program, although its in my path... guess i'll just modify the script to use `which $PROGRAM` instead...
1113687306 J * monrad ~monrad@213083190130.sonofon.dk
1113687400 J * ripman ~leukejong@d594e3d7.dsl.concepts.nl
1113687420 M * Doener welcome monrad, ripman 
1113687607 M * DaPhreak Bertl: here's the compile error message #1
1113687608 M * DaPhreak grsecurity/gracl_shm.c: In function `gr_handle_shmat':
1113687608 M * DaPhreak grsecurity/gracl_shm.c:19: warning: implicit declaration of function `vx_rmap_pid'
1113687669 M * Doener what is line 19 in gracl_shm.c?
1113687676 M * Bertl yep, thought so, means that the include file was _not_ included in gracl_shm.c
1113687712 Q * kjo Quit: Verlassend
1113687808 M * DaPhreak Doener: 19: task = find_task_by_pid(shm_cprid);
1113687810 M * monrad evening
1113687821 M * Bertl evening monrad!
1113687838 M * Bertl find_task_by_pid is a macro, calling find_task_by_real_pid with the rmapped pid
1113687859 M * Bertl (so the macro is there, but the vx_rmap_pid declaration is not)
1113687881 Q * ripman Ping timeout: 480 seconds
1113687891 M * Doener DaPhreak: yeah, i had a similar problem with 2.6.12-rc2
1113687919 M * Doener Bertl: shouldn't the file defining find_task_by_pid then include the file defining vx_rmap_pid?
1113687930 M * DaCa it has been corrected in patch-2.4.30-grsec2.1.5-vs1.2.10-dm-3.diff.bz2 which I uploaded yesterday night
1113687983 M * DaCa duh patch-2.6.11.7-grsec2.1.5-vs2.0pre1-2.diff.bz2 I ment :)
1113688214 M * Bertl Doener: no, because of the dependancies
1113688224 M * Doener ok
1113688232 M * Bertl simple exmaple:
1113688250 M * Bertl file a.h:  #define A_1(x) X(x)
1113688267 M * Bertl        #define A_2(x) Y(x)
1113688281 M * Bertl file b.h: #define X(x) (x+1)
1113688290 M * Bertl file c.h: #define Y(x) (x+2)
1113688311 M * Bertl now if file test.c uses A_1 but not A_2, c.h is not required
1113688345 M * Bertl (i.e. would add dependancies where not needed)
1113688389 M * DaPhreak so it is better to include the cvirt.h in gracl_shm.c, right ?
1113688416 M * DaPhreak s/cvirt.h/vs_cvirt.h/
1113688429 M * Bertl yep, it's always 'better' IMHO to include .h files in .c files than in other .h files ;)
1113688432 M * DaCa thats what I did, isn't it the right solution?
1113688445 M * DaPhreak it is :)
1113688460 M * Bertl probably, I just wanted to point out that you have to test different configs ;)
1113688503 M * DaPhreak heh, or use make allyesconfig but that probably might break other stuff
1113688522 M * Bertl well, allyes is only one conefig to test ;)
1113688750 M * DaPhreak another question 'bout ngnet Bertl :)
1113689168 M * Doener ok, no git for me, i simply don't get it...
1113689291 M * Bertl or better "I simply don't git it ;)"
1113689449 M * DaPhreak heh .. yeah i also didn't git it ;)
1113690471 J * ripman ~leukejong@d594e3d7.dsl.concepts.nl
1113690477 M * ripman hello
1113690485 M * ripman i have execute the tesme.sh
1113690490 M * ripman and i got this error
1113690490 M * ripman chcontext: vc_new_s_context(): Function not implemented
1113690490 M * ripman chcontext failed!
1113690494 M * ripman ??
1113690600 Q * erwan_ho Read error: Connection reset by peer
1113690643 M * Bertl ripman: could you upload the complete output somewhere? maybe pastebin.com?
1113690646 J * erwan_ho ~erwan@lns-vlq-39f-81-56-133-136.adsl.proxad.net
1113690668 M * ripman sure
1113690675 M * ripman bert
1113690736 M * ripman okay its there
1113690741 M * ripman on pastebin
1113690756 M * Bertl url?
1113690775 M * ripman http://pastebin.com/272195
1113690807 M * ripman what do you think ?
1113690817 M * Doener hm, pretty recent tools vs. a really ancient kernel...
1113690827 M * Doener ripman: you're on debian sarge, right?
1113690842 M * ripman so that might be the problem right ?
1113690846 M * ripman yes debian sarge
1113690855 M * ripman with 2.4 kernel
1113690865 M * Doener any objections against using a 2.6 kernel?
1113690896 M * Bertl the tools should be backwards compatible ...
1113690898 M * Doener 2.4 + vs1.2.10 should also be fine, but we got a nice howto for 2.6 vserver kernels ;)
1113690912 M * ripman yes really ?
1113690914 M * ripman for 2.6
1113690923 M * ripman didnt find it with google..
1113690927 M * Bertl btw, is this the second one today coming here with a 2.4.18ctx kernel, whatever that might be?
1113690946 M * ripman i am the second one probaly hahah
1113690949 M * Doener Bertl: that's what i was going to say... we don't even know what version that is ;)
1113690963 M * ripman and the first one ..
1113690965 M * ripman ??
1113690985 M * Bertl ah, no, it was both times you, right? ;)
1113690995 M * Doener ripman: it was just written two(?) days ago... so maybe google doesn't know about it yet
1113690995 M * ripman yep
1113691000 M * ripman http://lena.franken.de/linux/debian_and_vserver/vserver.html
1113691004 M * Doener http://linux-vserver.org/Step-by-Step+Guide+2.6
1113691006 M * ripman this i sthe link i got the kernel from
1113691041 J * Nik ~Nik@cable-153-130.online.bg
1113691078 M * Bertl lol @ 137GB ;)
1113691111 M * Bertl ripman: okay, could you upload (pastebin) the output of 
1113691124 M * Bertl 'vserver-info - SYSINFO' please?
1113691142 M * DaPhreak haha Bertl :) i know what went wrong :)
1113691170 M * DaPhreak pwd (compiling) ~/oldwork/kernel/trees/vserver-2.0_pre1.orig
1113691174 M * ripman http://pastebin.com/272197\
1113691180 M * DaPhreak pwd (editing) ~/oldwork/kernel/trees/vserver-2.0_pre1
1113691335 M * Bertl ripman: hmm, really interesting ...
1113691349 M * ripman there is no step by step for debian 2.6 right ?
1113691377 M * Bertl (see the doc Doener referred to)
1113691403 M * Doener building a kernel shouldn't be too much debian specific (except that you may use make-kpkg for some of the steps).
1113691439 M * Doener you can skip the installation of the tools, and just use those you got from the debian repository
1113691473 M * Doener Bertl: hm, maybe his 2.4.18ctx doesn't use the vserver syscall yet?
1113691490 M * ripman ohh i see.
1113691511 M * Bertl Doener: that's probably the most likely explanation
1113691526 M * Bertl IIRC we had the syscall number coded in rpoc somewhere
1113691532 M * Bertl s/rpoc/proc/
1113691568 M * Doener with 2.6 it's /proc/virtual/info ...
1113691599 M * Doener but I don't even know much about the 2.4 series, not to think of the time when it was called ctx
1113691600 M * Doener ;)
1113691641 M * ripman so what should i do  ?
1113691681 M * Bertl ripman: either use the ancient tools matching that kernel
1113691686 M * Doener get a recent kernel, either 2.4 + vs1.2.10 or 2.6 + vs1.9.5.x
1113691701 M * Bertl (or that, which is much better IMHO)
1113691771 M * ripman okay than i will try 2.6 
1113691854 M * Hollow how can i use vshelper to reboot vservers from inside?
1113691938 M * Bertl Hollow: depends on the aptch version, but you're welcome to help to test/fix that with 2.0-pre1 ;)
1113691959 M * Hollow i have 1.9.5 atm
1113691980 M * Hollow how should it work? ;)
1113692007 M * Bertl well, it calls the helper, which in turn brings down the guest, and restarts it
1113692066 M * Hollow who calls the helper?
1113692091 M * Bertl the kernel
1113692101 J * jsambrook ~jsambrook@aelfric.plus.com
1113692108 P * jsambrook 
1113692129 M * Hollow and how does the kernel know when to send the reboot cmd to the helper?
1113692145 M * Bertl when the guest did send the reboot syscall to the kernel
1113692182 M * Hollow using shutdown?
1113692200 M * Bertl well, whatever calls the syscall ...
1113692211 M * Bertl can be reboot, or halt or whatever
1113692226 M * Hollow ok.. using 'shutdown -r now` i get "init: /dev/initctl: No such file or directory"
1113692243 M * Doener try: reboot -f
1113692243 M * DaPhreak well try reboot -f ;)
1113692246 M * Hollow using the gentoo init style there never is a init process
1113692255 M * Hollow hm
1113692289 M * Doener reboot/poweroff/shutdown work through init if not forced to do without it
1113692358 M * Hollow works, thx
1113692405 M * DaPhreak naaah .. i need something to eat ... brb
1113692767 M * Bertl okay, need a break too .. back later ...
1113692778 N * Bertl Bertl_oO
1113692926 Q * Nik Ping timeout: 480 seconds
1113692990 Q * ripman Ping timeout: 480 seconds