1138320084 M * cohan btw, is that useful what i do here in any way?
1138320176 M * Bertl_oO sure, will investigate it from the kernel point when I'm done here
1138320267 Q * vrwttnmtu Quit: Leaving
1138320419 M * jkl Bertl_oO: did hollow say anything to me?
1138320568 M * jkl seems like he went to sleep again
1138320576 M * Bertl_oO 19:24 < jkl> on ALL my vservers! it's hectic
1138320576 M * Bertl_oO 19:25 < Hollow> well, you can either workaround it by adding udev to  package.provided, or find the ebuild depending on udev
1138320578 M * jkl *sigh* 
1138320580 M * Bertl_oO 19:27 < daniel_hozac> vmware? heh.
1138320582 M * Bertl_oO 19:28 < daniel_hozac> addr2line?
1138320590 M * jkl ahhh
1138320592 M * jkl thank you
1138320607 A * jkl wonders where package.provided may be
1138320625 M * Bertl_oO jkl: but in general if your client doesn't log, use the real-time irc channel logs
1138320970 M * cohan so, have got a "postit" shell function ready to directy push files on that webspace
1138320998 M * cohan https://servy.dynip.yawsp.de/~install/kern.log <--- this is the output of the latest, extended log messages
1138321098 M * Bertl_oO very interesting
1138321116 M * cohan of this patched source: https://servy.dynip.yawsp.de/~install/cvirt.c
1138321217 M * cohan is there a _simple_ way to see when there are writes to that memory address?
1138321247 M * Bertl_oO not without attaching a debugger
1138321288 M * Bertl_oO currently I just have the irc here, but I'll look into it in an hour or so
1138321355 M * cohan i hope anybody can reproduce that at all - i'll also upload my kernel config
1138321391 M * Bertl_oO btw, what gcc/binutils do you use?
1138321437 M * cohan gcc 3.4.4, binutils 2.16.1
1138321816 M * cohan should be asleep soon - but this kinda hypnotizes me now
1138322100 M * cohan where are you currently located, Bertl_oO?
1138322172 M * Bertl_oO austria
1138322187 M * cohan hm, so you swapped day and night?
1138322210 M * Bertl_oO somewhat
1138322254 M * cohan okay, compiled kernel without COW, no change...
1138322423 M * Bertl_oO didn't expect it to change anything ...
1138322454 M * Bertl_oO guess the character was just coincidence
1138322525 M * jkl uhhh!
1138322533 M * jkl this is incredibly frustrating
1138322590 M * cohan using different names, shorter domains (vserv1.de) 
1138322634 M * Bertl_oO well, as you saw in your print, the domain isn't even written there
1138322656 M * cohan not at the get, but at the set
1138322673 M * cohan just playing around
1138322685 M * Bertl_oO okay, np .. maybe you find something
1138322754 M * cohan next kernel with legacy again? ;)
1138322880 P * undefined 
1138322930 M * cohan okay, it continues mutating
1138323005 M * cohan you see, two successive get-names produce different trailing garbage
1138323018 M * cohan https://servy.dynip.yawsp.de/~install/kern.log
1138323856 M * jkl ah fixed it
1138323863 M * jkl removed udev from the world file
1138323909 A * jkl shakes his fists
1138323916 M * jkl I showed you!! silly vserver
1138323944 M * Bertl_oO hrm, those are gentoo issues .. not vserver issues
1138323952 M * jkl yes, i know
1138323963 M * jkl *silly gentoo !
1138325651 M * cohan i'll get some sleep no
1138325651 M * cohan w
1138325668 M * Bertl_oO okay, good night, I guess we should know tomorrow
1138325712 J * gerrit ~gerrit@c-24-22-19-162.hsd1.or.comcast.net
1138325771 J * cohan_zzz ~martin@p54AD2271.dip0.t-ipconnect.de
1138325787 Q * bragon Remote host closed the connection
1138325796 J * bragon ~bragon@sd866.sivit.org
1138325816 M * cohan_zzz if you need any more information, just let me know
1138325831 M * Bertl_oO okay, great!
1138325872 M * cohan_zzz in 6 hours i will be back again, you can also query me
1138326134 Q * cohan Ping timeout: 480 seconds
1138327806 J * lilo_ ~lilo@lilo.usercloak.oftc.net
1138327909 Q * lilo Ping timeout: 480 seconds
1138328360 M * blackfire hello all :) I'm back for a while
1138328456 M * blackfire does anybody know how compatible is util-vserver .210 compared to .208?
1138328518 M * blackfire I'm mostly interested in /usr/sbin/vserver <args> compatibility (basically start, stop, enter)
1138329475 J * ag- ag@caladan.roxor.cx
1138331398 J * click click@ti511110a080-4706.bb.online.no
1138332105 N * Bertl_oO Bertl
1138332111 M * Bertl back now ...
1138332120 M * Bertl welcome click! ag-!
1138332130 M * click re re
1138332164 M * click another powersurge, another day of fixing dnszones due to an ISP refusing to set up a static ip... *sigh*
1138332901 M * mnemoc CNAME to yi.org or dyndns is your friend :)
1138333524 Q * Johnnie Quit: G'bye!
1138334164 J * Johnnie ~jdlewis@dynamic-acs-24-154-53-16.zoominternet.net
1138334802 Q * Loki|muh Remote host closed the connection
1138334812 J * Loki|muh loki@satanix.de
1138335749 J * Smutje_ ~Smutje@xdsl-84-44-246-249.netcologne.de
1138335854 Q * Smutje Ping timeout: 480 seconds
1138336579 Q * mkhl Ping timeout: 480 seconds
1138337753 Q * mnemoc Ping timeout: 480 seconds
1138337883 J * mnemoc ~amery@200.75.27.75
1138338450 J * balbir ~balbir@59.145.136.1
1138338458 M * Bertl welcome balbir!
1138338527 M * Bertl anybody still awake?
1138339706 M * daniel_hozac good morning!
1138339718 M * Bertl good morning daniel_hozac!
1138339769 M * Bertl daniel_hozac: could you give the testme.sh-0.15 a spin (on as many vserver hosts you can reach :)
1138339788 M * daniel_hozac sure.
1138339893 M * daniel_hozac works fine on all (3, heh) of them.
1138339906 M * Bertl great!
1138339906 M * daniel_hozac one of which is stuck on an ancient kernel and 0.30.209.
1138339947 M * Bertl my oldest was: Linux 2.4.28-p4vs1.29 with jacks 0.30
1138339970 M * daniel_hozac heh, so it wasn't that ancient then. (2.6.11 based)
1138340032 M * Bertl mnemoc: ping!
1138340150 J * stefani ~stefani@c-24-19-46-211.hsd1.wa.comcast.net
1138340156 M * Bertl welcome stefani!
1138340172 M * stefani good evening.
1138340189 M * Bertl do you have an older linux-vserver system (or several) in reach?
1138340268 M * Aiken what do you call old?
1138340278 M * Bertl daniel_hozac: what do you think, should we add some kind of auto-mail option, so you can do something like testme.sh -m and it will mail the stuff to the mailing list?
1138340287 M * Bertl Aiken: anything you got is appreciated :)
1138340293 M * Bertl http://vserver.13thfloor.at/Stuff/SCRIPT/testme.sh-0.15
1138340299 M * Aiken 2.4.30-vs1.2.10 + 0.30?
1138340305 M * Bertl excellent!
1138340306 M * Aiken testme-0.15 blows up on that
1138340314 M * Bertl really?
1138340317 M * Bertl how?
1138340324 M * Bertl could you upload it?
1138340331 M * daniel_hozac Bertl: for the architecture reports? sure, makes sense.
1138340371 M * Bertl can mailx be assumed on most systems?
1138340375 M * Aiken never mind
1138340381 M * Aiken that system defaults to bash1
1138340386 M * Aiken it works fine with bash2
1138340397 M * Bertl hmm, good point, maybe we should check that too
1138340403 M * daniel_hozac hmm, well, none of my systems have mailx.
1138340420 M * Bertl hmm, what do they have? mail?
1138340435 M * daniel_hozac yeah.
1138340447 M * Bertl I mean, in theory we could use telnet :)
1138340451 M * Aiken the blow up with bash 1.14.7 http://pastebin.com/525184
1138340504 M * Aiken probably not worth worrying about, T think I was/am the only person running vserver on a system that old
1138340568 M * Aiken stil getting one error http://pastebin.com/525186
1138340606 M * daniel_hozac does mktemp -d /tmp/test.XXXXXX give it too?
1138340629 M * Aiken it is an older mktemp that does not support -d
1138340638 M * Bertl ah, okay ...
1138340690 M * Aiken fixed, updated mktemp
1138341575 J * rene- rene@201.144.61.222
1138341584 M * Bertl welcome rene-!
1138341598 A * rene- feels warm and fuzzy
1138341601 M * rene- hi
1138341656 M * rene- thanks, ive just found out about this, i plan to virtualize my box, i want to run 4-5 instances of debian in a P3 1Ghz / 256 of ram, does it sound reasonable?
1138341673 M * Bertl piece of cake :)
1138341722 M * Bertl lycos pushed a dual PIII to more than 150 guests ... (when testing)
1138341755 M * rene- really? i ve read that i can apt-get my way into a virtualized environment, is that true?
1138341805 M * Bertl well, yes, there are many debian packages out there, just make sure to get recent ones
1138341907 M * rene- regarding networking, can a nic be shared among different installations? like each one has a unique IP so they can beaccessed from the outside? what about communicating instances with each others via tcp/ip.. can it be done?
1138341932 M * Bertl yes, works fine, networking is isolated but not virtualized
1138341952 M * Bertl i.e. you share a common networking on the host, each guest gets a subset of ips
1138342087 M * rene- my plan is to have a debian box running five or so instances of linux each one running ssh and asterisk, i will be connecting ip phones to the different instances and they should all be able to communicate themselves, i intend to give training, and i dont think im getting a real asterisk box for each participant
1138342156 M * Bertl IIRC, some folks are already doing something like that
1138342181 M * rene- so i think vserver will do the job, yes my inspration comes fom here: http://www.telephreak.org/papers/vpa/
1138342208 M * rene- they dont address the actual vserver config just asterisk in the virtual environmen
1138342211 M * rene- environment
1138342241 M * rene- how involved is it? i have yet to compile my own linux kernels
1138342256 M * Aiken Bertl line 11 http://pastebin.com/525209
1138342294 M * Bertl ah, funny, tx!
1138342319 M * Bertl rene-: did you have a look at linux-vserver.org yet?
1138342376 M * rene- i have take a glance i need to read more i guess
1138342386 M * rene- took a glance at
1138342389 M * rene- brb
1138342913 M * rene- yay lots of good docs
1138343126 M * rene- it doesnt look that difficult at all
1138343162 M * Bertl well, it's really quite simple
1138343177 A * rene- thinks, my brother is sleeping, maybe he wont mind me  turning his winbox into a vserver
1138343236 M * Bertl well, problem is, win* doesn't use linux kernels :)
1138343319 M * rene- i was thinking about formatting it, but i will save him the grief
1138343389 P * stefani parting (is such sweet sorrow)
1138345192 M * Bertl Aiken: what does BASH_VERSINFO contain on your 1.x bash?
1138345262 M * Aiken the above was on bash2
1138345270 M * Aiken on a normal system, not the old system
1138345296 M * Bertl yup, but what does $BASH_VERSINFO contain on the 1.x bash?
1138345318 M * Aiken BASH_VERSION=1.14.7(1)
1138345337 M * Bertl okay, so no VERSINFO, I'd asume?
1138345349 M * Aiken (root@barney) set | grep BASH
1138345350 M * Aiken BASH=/bin/bash
1138345350 M * Aiken BASH_VERSION=1.14.7(1)
1138345358 M * Bertl okay, great!
1138345835 M * Bertl Aiken: do you have the old mktemp still around somewhere?
1138346091 M * Aiken yes
1138346191 M * Aiken (root@barney) mktemp -d /tmp/test.XXXXXX
1138346191 M * Aiken error: -d option is not supported on Linux
1138346203 M * Bertl http://vserver.13thfloor.at/Stuff/SCRIPT/test-0.01.sh
1138346215 M * Bertl could you give that a try? it should work around this issue
1138346232 M * Bertl (does just create a temp dir and check for a few commands)
1138346267 M * Bertl should also complain on your 1.x bash that it is too old
1138346272 M * Aiken (root@barney) bash /home/ftp/incoming/test-0.01.sh 
1138346272 M * Aiken bash seems too old.
1138346281 M * Bertl :)
1138346297 M * Aiken (root@barney) bash2 /home/ftp/incoming/test-0.01.sh 
1138346297 M * Aiken error: -d option is not supported on Linux
1138346297 M * Aiken drwxr-xr-x    2 root     root         1024 Jan 27 17:18 /tmp/test-0.01.sh.zWSNpQ
1138346314 M * Bertl okay, except for the error message, this is fine now
1138346385 Q * gerrit Ping timeout: 480 seconds
1138346498 M * Bertl okay, updated it in place, should now be fine
1138346515 M * Bertl erm, no,sec
1138346541 M * Bertl now it should be fine
1138346577 M * Bertl hmm, close, but no banana
1138346740 N * Smutje_ Smutje
1138347098 M * Bertl okay, two new functions, but that should do the trick (updated in place)
1138347238 M * Aiken what do you make of /tmp/testme.sh-0.15: 16#VCISyscall: value too great for base (error token is "16#VCISyscall")
1138347265 M * Bertl seems like one of the scanners is picking the wrong line/data
1138347293 Q * Aiken Quit: Leaving
1138347317 J * Aiken ~james@tooax6-123.dialup.optusnet.com.au
1138347322 M * Bertl there should be a number there, but as the script is very hacky (because grown)
1138347332 M * Aiken net connection went splat so missed any answer
1138347350 M * Bertl I decided to rewrite it, and to start with a solid framework
1138347371 M * Bertl you should not have missed anything (I waited for your return ;)
1138347374 M * Aiken so that message is nothing to worry about at the moment?
1138347395 M * Bertl no, I hope it will be fixed tomorrow
1138347542 J * _are_ ~are@62.112.159.81
1138348525 M * Hollow morning
1138348533 M * Bertl good morning!
1138348594 M * Hollow http://vwiki.croup.de/Documentation ;)
1138348670 M * Bertl looks nice!
1138348695 M * Hollow do you think the hierarchy fits?
1138348744 M * Bertl well, I'm not sure the 'devel' stuff should be there
1138348755 M * Hollow yeah, we could also put it in the devel section ;)
1138348830 M * Hollow do you have some spare time this evening?
1138348862 M * Bertl very likely ... 
1138348872 M * Hollow great
1138349169 M * cohan_zzz moin
1138349175 N * cohan_zzz cohan
1138349182 M * Bertl morn cohan!
1138349328 M * Bertl cohan: could you do a few simple tests for me?
1138349359 M * cohan of course, i am looking forward to... machine just booting
1138349489 M * Bertl chcontext --xid 42 -- sleep 1000 &
1138349504 M * Bertl vuname -s --xid 42 -t context="123456789"
1138349530 M * Bertl vuname -g --xid 42 context
1138349710 M * cohan without "false" the last line?
1138349736 M * Bertl yup, just as is :)
1138349818 M * cohan seems that it worked, hm
1138349842 M * Bertl so that gives you the right value, yes?
1138349845 M * cohan no truncates visible, yes
1138349864 M * Bertl okay, now please let's try with a 64 char long string
1138349879 M * Bertl i.e. continue with 012345...
1138349901 M * cohan same xid?
1138349907 M * Hollow oh lovely persistant contexts... ;)
1138349915 M * Bertl yes, just skip the first command
1138349957 M * cohan works... ah, okay, i used a length of 42 now ;) *fixing*
1138349984 M * cohan 64 works... 65 now?
1138349988 M * Bertl yep
1138350008 M * cohan vuname: vc_set_vhi_name(): Arg list too long
1138350020 M * Bertl okay, so that is correct too
1138350032 M * Bertl as the 64 char string is actually 65 chars long
1138350039 M * Bertl (because of the terminating zero)
1138350054 M * cohan yes, right - i'll just give the phoenix another "g'morning"-try
1138350057 M * Bertl what did your kernel log capture of those tests?
1138350098 M * cohan what one expects, i'll upload it anyway
1138350183 M * cohan https://servy.dynip.yawsp.de/~install/ kern2.log
1138350218 M * Bertl hmm, connection refused
1138350251 M * cohan https?
1138350280 M * Bertl yep, even copied the funny space :)
1138350304 M * Bertl servy.dynip.yawsp.de (84.173.34.113)
1138350317 M * cohan that was not meant real, though yesterday it worked
1138350330 M * cohan correct IP
1138350339 M * cohan try without space
1138350368 M * cohan god knows what a browser does with that space ;)
1138350404 M * cohan just seen an external connect, that was you, right?
1138350409 M * Bertl seems my browser cached the name ...
1138350426 M * Bertl ip worked fine ...
1138350467 M * cohan from yesterday? ah, ic - well, it's a dynip service i run myself, but usually update takes about 5 mins, but perhaps some caching policy validation/propagation issue
1138350472 M * Hollow Bertl: btw, the suggested dd line for bzImage extraction does not work.. gzip always complains about invalid format
1138350498 M * Bertl Hollow: might require slight adjustments ... best is to search for the gzip header
1138350541 M * Bertl cohan: np, so I guess _something_ changes the name shortly after the set
1138350588 M * Bertl cohan: do you have vserver debugging enabled?
1138350594 M * cohan in the kernel? yes
1138350602 M * Bertl excellent, please do:
1138350627 M * cohan this is, btw, a kernel with legacy enabled again, just "hoped" that solves it, but no change again
1138350636 M * Bertl sysctl -w vserver.debug_switch=255
1138350652 M * cohan done
1138350666 M * Bertl then start/stop the guest, and look in the kernel logs
1138350773 M * cohan i'll never get used to vkill using -c OR --xid, i always type --ctx
1138350924 M * cohan kern3.log
1138351080 M * Bertl where was that diff again (of your changes)?
1138351189 M * cohan cvirt.c
1138351212 M * cohan sorry, no diff right now, because i lost original source due to vi not breaking hardlinks
1138351223 M * cohan (i'll regenrate a virgin sourcetree now...)
1138351226 Q * shedi Quit: Leaving
1138351232 M * Bertl cohan: ah, no
1138351254 M * cohan those changes where _only_ about log messages
1138351256 M * cohan https://servy.dynip.yawsp.de/~install/cvirt.c
1138351264 M * Bertl that's perfect
1138351483 M * Bertl we are on x86 there, right?
1138351502 M * cohan right
1138351520 M * cohan # uname -a
1138351520 M * cohan Linux servytest 2.6.15-vs2.1.0.5 #9 SMP Fri Jan 27 02:19:44 CET 2006 i686 pentium4 i386 GNU/Linux
1138351655 M * Bertl could you add the following to include/linux/vserver/context.h and recompile the kernel?
1138351669 M * cohan do i have any choice? ;) tell me what to add
1138351678 M * Bertl right after the one and onle #include
1138351682 M * Bertl *only
1138351688 M * Bertl #ifndef CONFIG_VSERVER
1138351688 M * Bertl #warning config options missing
1138351688 M * Bertl #endif
1138351735 M * Bertl somehow I suspect the context struct changes size :)
1138351839 M * cohan make in progress...
1138351849 M * Bertl check for that specific warning
1138351921 M * cohan should really give that devel node a faster disc... it's a 2GHz machine, but an ancient old 4GB harddisc, the first one i get to grab when i started ;)
1138352027 M * cohan did not see a warning yet, but perhaps i missed one
1138352039 M * Bertl it was just an idea ...
1138352103 M * cohan shall i re-touch the header, re-make and redirecting make's output?
1138352117 M * cohan but i wonder... this time it builds the whole kernel, damn
1138352136 M * Bertl because the context.h is very popular :)
1138352162 M * cohan yes :/ should make a slimmer kernel config with only the required modules
1138352184 M * Bertl so no warning there, yes?
1138352210 M * cohan would have come early?
1138352220 M * Bertl can happen any time
1138352245 M * cohan i'll break this build and redirect output, ctrl+c should not blew the whole show?
1138352280 M * Bertl no, but you will miss the comples till now
1138352294 M * Bertl best would be to rebuild the entire kernel
1138352340 M * cohan oookay - make clean enough?
1138352346 M * Bertl yep
1138352468 M * cohan this will need some time now... switching to a faster hard disk could also save time... ;)
1138353437 M * Bertl okay, I will be off to bed shortly, any estimate?
1138353575 M * cohan started right now
1138353692 M * cohan every 10 seconds https://servy.dynip.yawsp.de/~install/make.out is updated
1138353732 M * cohan well, every 30 seconds should be enough
1138354171 M * Bertl cohan: you get something like this, yes?
1138354172 M * Bertl can not determine xid of vserver 'test101'; returned value was ''
1138354322 M * cohan yes, all the time
1138354336 M * cohan this is what util-vserver-tools say on startup, shutdown etc.
1138354385 M * Bertl okay, seems I can at least reproduce it here now
1138354419 M * cohan wow!
1138354538 M * cohan so any idea why no one else stumbled across that yet?
1138354552 M * cohan what is so special about my/your now generated conditions?
1138354653 M * Bertl no idea yet
1138354761 M * Bertl well, you have legacy enabled right now?
1138354840 M * cohan yes
1138354850 M * cohan the kernel i am building now has no legacy again
1138354853 M * Bertl could you try with the 0.30.209 tools for me?
1138354868 M * cohan i'll give it a try... any special ./configures, 
1138354869 M * Bertl (they should work with legacy enabled)
1138354884 M * Bertl I used: ./configure --prefix=/usr --sysconfdir=/etc --localstatedir=/var --with-vrootdir=/vservers
1138354896 M * Bertl but take your 'known working' configs
1138354912 M * cohan i'd just use this:
1138354912 M * cohan https://servy.dynip.yawsp.de/~install/vports/util-vserver/Pkgfile
1138354919 M * Bertl (or better, known-not-working)
1138354926 M * cohan with the version number decreased
1138354936 M * Bertl okay, good
1138354998 M * cohan and without the "debugpatch" of course
1138355554 M * cohan kernel is built now - 0.30.209 tools are building, shall i give the old tools a try first,
1138355576 M * cohan or - well, rebooting makes no sense, see the log, no warning related to vserver
1138355589 M * Bertl yes, no reboot
1138355597 M * cohan https://servy.dynip.yawsp.de/~install/make.config produced https://servy.dynip.yawsp.de/~install/make.out
1138355598 M * Bertl without the legacy, the old tools will not work
1138355834 M * cohan okay, i replaced the tools...
1138355842 M * cohan ...just trying to launch "as usual" ?
1138355852 M * Bertl yes please
1138355879 M * cohan same error... do you want the kernel log output?
1138355897 M * Bertl no, tx
1138355932 M * cohan do you want login to that box? ;)
1138355943 M * Bertl I do one more test now, then I'm off to bed ... but, if you find the time you could do the following:
1138355996 Q * balbir Quit: Leaving
1138356013 M * Bertl get some older kernel + patch e.g. 2.6.14 and try with that, work your way up the releases and see which one introduced this issue
1138356034 M * Bertl (usually some kind of bisection is good)
1138356046 J * shedi ~siggi@tolvudeild-205.lhi.is
1138356056 M * cohan bisection? saying "introduced with kernel 2.6.x" ?
1138356084 M * Bertl cohan: it seems to be a kernel issue, but we do not know since when or what exactly happens
1138356106 M * Bertl it could be some mainline issue as well, scribbling over the memory
1138356107 M * cohan i am "glad" that at least you where able to reproduce the issue 
1138356133 M * Bertl but wait, I have another idea ...
1138356134 M * cohan yes, okay - not sure whether my time can effort it, but i'll give it a try
1138356159 M * Bertl what character do you get scribbled over the correct name?
1138356173 M * cohan depends, changes
1138356184 M * Bertl when? when you change the context id?
1138356196 M * cohan see kern.log
1138356206 M * cohan https://servy.dynip.yawsp.de/~install/kern.log
1138356212 M * cohan is a quite good example
1138356236 M * cohan but the phoenix X seems to be overridden with } mostly
1138356243 M * Bertl what xid has phoenix?
1138356246 M * cohan 235
1138356259 M * cohan vserv2 had 231
1138356272 M * cohan oh, 232
1138356305 M * cohan ah, it was vserv2, that makes sense in my internal context naming scheme ;)
1138356439 M * cohan okay, i'll update utils again, and use the no-legacy-kernel, and will try to use older kernels
1138356472 M * cohan btw, i never got a network interface alias up - but i think thats related, right?
1138356494 M * Bertl yes, very likely ... what kernel do you use right now?
1138356507 M * cohan (e.g. the vserver ... start never brought up or down any network alias etc)
1138356516 M * cohan # uname -a
1138356516 M * cohan Linux servytest 2.6.15-vs2.1.0.5 #9 SMP Fri Jan 27 02:19:44 CET 2006 i686 pentium4 i386 GNU/Linux
1138356520 M * cohan this was the one with legacy
1138356530 M * cohan where i just tried the old tools
1138356544 M * cohan another one without legacy is ready now
1138356567 M * Bertl so it must have been in 2.6.15 already
1138356582 M * Bertl I tested with 2.6.16-rc1-vs2.1.0.7
1138356677 M * cohan yes, only used 2.6.15 the last two days
1138356753 M * Bertl the values change gradually
1138356765 M * Bertl so something is using that memory for something else
1138356803 M * cohan yes, some vserver-related function which is called in between - but i can't read the kernel debug output properly ;)
1138356939 M * cohan shall i try it with 2.6.14 and 2.1.0 ?
1138357103 M * Bertl http://vserver.13thfloor.at/Stuff/BUGHUNT/cohan-0001/klog.txt
1138357115 M * Bertl let me explain the log
1138357128 M * cohan okay
1138357152 M * Bertl I've done a small modification to __lookup_vx_info() logging the context utsname too
1138357175 M * cohan ic
1138357195 M * Bertl the switch log outputs VCMD_CC_N[V]
1138357203 M * cohan test101 becomes tevt101, later tee}101
1138357213 M * Bertl where CC is the category, N the command and V the version
1138357222 M * cohan *trying to get that*
1138357239 M * Bertl include/linux/vserver/switch.h
1138357254 M * Bertl for example the VCMD_52_2[0]
1138357274 M * Bertl is category 52, command 2 version 0
1138357286 M * Bertl now 52 is flags, so we check there
1138357307 M * cohan yes, got it
1138357313 M * Bertl #define VCMD_set_cflags         VC_CMD(FLAGS, 2, 0)
1138357375 M * Bertl 46_1[0] is VCMD_task_xid
1138357395 M * cohan in include/linux/vserver/context_cmd.h
1138357421 M * Bertl commands are logged twice, once before the call
1138357430 M * Bertl once afterwards ... with the return code :)
1138357495 M * Bertl the problem is, we cannot pinpoint the change yet
1138357495 M * Bertl the __lookup() is called at the beginning of most commands
1138357556 M * Bertl so IMHO the VCMD_52_2 or the VCMD_46_1 and VCMD_46_5 commands could change it
1138357575 M * Bertl but it looks like the VCMD_52_2 is doing evil stuff
1138357700 M * Bertl now, I can enable more logging regarding the get/put, which I will no now to check refcounting
1138357717 M * cohan no now / do now?
1138357723 M * Bertl yep
1138357740 M * cohan if i can be of any help... just provide me a patch or sth, i'll apply & try - but i fear you are more effective
1138357770 M * Bertl you can try compiling a 2.6.14 kernel with devel
1138357891 M * cohan okay, thats 2.1.0 then, right?
1138357953 M * cohan 2.6.15.1 will be of no change, right? ;) 
1138358035 M * Bertl yep
1138358052 M * Bertl http://vserver.13thfloor.at/Stuff/BUGHUNT/cohan-0001/klog2.txt
1138358060 M * Bertl here with full ref counting
1138358060 M * cohan okay, building 2.6.14.6 in progress...
1138358088 M * cohan *looking*
1138358150 M * Bertl nothing unusual ... looks quite normal, except for the change
1138358195 M * cohan but some line in between has to be responsible, right
1138358195 M * cohan ?
1138358205 M * cohan one of that forks?
1138358213 M * Bertl not necessarily ..
1138358242 M * Bertl they are just points in time ...
1138358245 J * marl_mobile ~matt@84.92.193.226
1138358368 M * Bertl but we can make the grid smaller ... by for example adding the output to the put
1138358510 M * marl_mobile anyone around help with a kernel compile problem? im getting the following error when trying to compile 2.6.15-1 with the 2.6.15-vs2.0.1.2.diff patch ive searched the site and google, but the only thing i can see if a referance to the vx_file_rsspages_dec in the diff file itself, anyone point me in the right direction? the error is : mm/built-in.o: In function `__xip_unmap':filemap_xip.c:(.text+0x218f4): undefined reference to `vx_file_rsspages_dec'
1138358527 M * marl_mobile btw, running debian testing, on a minimal install
1138358624 M * Bertl 2.6.15-1 ?
1138358632 M * marl_mobile yup, fresh from kernel.org
1138358650 M * Bertl url?
1138358658 M * marl_mobile 2.6.15.1 sorry
1138358662 M * Bertl okay :)
1138358681 M * marl_mobile lol, jumping back and forward between console and irc :)
1138358682 M * Bertl the vx_file_rsspages_dec is a synthetic function
1138358708 M * Bertl it is not present in the patch it is generated by combining elements
1138358770 A * cohan wants a whip to make the kernel compiling faster... ;)
1138358775 M * marl_mobile so does that meen its not to do with the vserver patch, or is it a by-product of the patch?
1138358797 M * Bertl it is a by product .. I just wonder why you get that, what arch do you use?
1138358805 M * marl_mobile cohan, cobble together a couple of old machines and run parrel compiling :)
1138358825 M * marl_mobile the kerenl config was set to amd athalon/duron
1138358835 M * Bertl so 64bit, yes?
1138358838 M * marl_mobile nope
1138358857 M * Bertl hmm .. give me a few seconds
1138358862 M * marl_mobile thanks :)
1138358911 M * Bertl http://vserver.13thfloor.at/Devel/PAT-2.6.15/delta-xip-fix01.diff
1138358917 M * marl_mobile cohan, im currently designing a computer 'room' to run 20 500Mhz-1Ghz mahcines on a network to allow parrel compilation of software
1138358928 M * Bertl apply that and run 'make' again
1138358934 M * Bertl (no need to do make clean)
1138358962 M * marl_mobile thanks was just about to ask that :)
1138359020 M * cohan ;) nice "room" then - i just hope i don't have to stay _inside_ that room
1138359075 M * marl_mobile lol, im waiting on hearing about some old car A/C units to cool it down, absolute nitght mare in the summer, but saves a fortune in the winter in heating bills :)
1138359459 M * SiD3WiNDR mm
1138359470 M * SiD3WiNDR my "serverroom" with 4-5 machines and some switches already makes noise and is "warm" :p
1138359608 N * ebiedermOo ebiederm
1138359613 M * marl_mobile cant shut the door on myn now, wuithin 15 seconds the temp alarms start going :(
1138359619 M * cohan hm, so marl_mobile is expected to get the same problems we are just tracking, Bertl?
1138359636 M * marl_mobile thanks Bertl btw, that compiled ok this thime :)
1138359662 M * marl_mobile cohan, wat kinda problems? problems caused by parrel compilation?
1138359667 A * ebiederm laughs We are running the AC at work and there is snow on the ground outside.
1138359749 M * cohan nope, problems with 2.6.15/16 and vs2.1.0.5
1138359774 M * marl_mobile :( great, just finished compiling it :(
1138359789 M * cohan then try it, we need a wider test base ;)
1138359819 M * marl_mobile ok, be back in a bit, going to try and reboot to it now
1138359837 M * Bertl marl_mobile: you are using stable, I doubt this is an issue there
1138359985 M * marl_mobile brb
1138359988 Q * marl_mobile Quit: Leaving
1138360009 M * cohan rebooting to 2.6.14.6-vs2.1.0 now
1138360283 M * cohan Bertl, how are 2.1.0 and 2.1.0.5 related ? 
1138360306 M * cohan (e.g. only 2.6.15-transition-changes, or also newer stuff?)
1138360424 M * Bertl both
1138360522 M * cohan 2.6.14.6 with 2.1.0 seems to work perfectly
1138360537 M * cohan network interface up etc... seamless
1138360575 M * Bertl thought so, we would have heard about it otherwise :)
1138360609 M * Bertl okay, I'm off to bed now (finally) if you like you can try 2.1.0.1-2.1.0.4
1138360636 M * Bertl will look into it this evening ...
1138360637 M * cohan applying to 2.6.15 or 2.6.16 ?
1138360645 M * ebiederm Sleep well...
1138360647 M * cohan aeh, 2.6.14 or 15
1138360657 M * Bertl if you look at the patch, it should show you the kernel version
1138360665 M * cohan perhaps adding a note somewhere that 2.6.15 with 2.1.0.5 has major issues?
1138360707 M * Bertl feel free to do that, or send an email
1138360722 M * cohan okay, so gn8 and sleep well
1138360733 M * Bertl tx, cya all!
1138360741 N * Bertl Bertl_zZ
1138360997 N * Bertl_zZ Bertl
1138361006 M * Bertl back again, just a short question
1138361020 M * Bertl cohan: what gcc do you use for the kernel?
1138361062 M * cohan dejavu ;) 3.4.4
1138361105 M * Bertl ah, okay, good wrong idea then ... I'm at 4.0.3
1138361111 M * Bertl night then :)
1138361117 N * Bertl Bertl_zZ
1138361634 Q * Aiken Ping timeout: 480 seconds
1138361836 M * cohan okay, dropped a note on the mailing list
1138362043 J * Milf ~Miranda@ipsio107.ipsi.fraunhofer.de
1138362815 J * meebey meebey@booster.qnetp.net
1138364321 M * daniel_hozac hmm, is it really dependent on static contexts? i'd expect it for dynamic contexts too.
1138364398 N * ebiederm ebiederm_zZ
1138364426 M * cohan maybe - i am not sure, i thought i have no dynamic contexts in kernels with no legacy, 
1138364453 M * cohan furhtermore, i thought "dynamic contexts" would not use that kernel stuff for name resolution, but i believe i mixed things up here
1138364476 M * marl ok, i know this isnt quite a vserver problem, but ive just compiled 2.6.15.1 with vserver useing a chroot and basic debian install (debootstrap testing) i can boot up to the new system, and apear to be able to type commands etc. but i have no more output apearing on my monitor after the kernel inits the ACPI stuff, if i try cat file >/dev/tty0 i still dont get anything on screen, but i can type reboot and the system WILL reboot, i dont have networking enabled 
1138364476 M * marl yet, so i cant ssh in from outside to check any more, anyone got any pointers?
1138364685 M * daniel_hozac does ctrl+alt+f[123456] work?
1138364798 M * marl screen dosnt apear to change, but i need to log in to be able to run commands, its almost as if its sending all output to /dev/null but accepting input from keyboard
1138364840 M * daniel_hozac did you try with acpi=off or noacpi or whatever that argument is?
1138364900 M * marl 2 secs
1138365117 M * cohan i'll now give http://vserver.13thfloor.at/Experimental/patch-2.6.15-vs2.1.0.2.diff a try
1138365169 M * marl daniel_hozac, can u remember how to turn off framebuffer support from kernel boot option?
1138365331 J * Viper0482 ~Viper0482@p54977A08.dip.t-dialin.net
1138365421 J * ghoost ~MaNoLyA@85.103.0.39
1138365434 M * ghoost hi all
1138365462 M * marl hi ghoost 
1138365497 M * ghoost i havesome problms about of linux
1138365525 M * ghoost who can help me ?
1138365527 M * schellh just shoot
1138365579 M * ghoost how ?
1138365602 M * schellh tell us of your problems
1138365626 M * ghoost i can't mount sound card
1138365631 M * ghoost its on board
1138365675 M * ghoost i need commands
1138365681 M * ghoost but i dont know
1138365707 M * schellh well i dont know also because i dont use linux for my workstation only for servers 
1138365711 M * schellh sorry
1138365723 M * ghoost ok not problem
1138365735 M * ghoost i just ask
1138365744 M * ghoost maybe anyone know
1138365751 M * ghoost in server
1138365831 Q * ghoost Quit: -( MaNoLyA Script v7.1 )-  ßy TyRaNT  DownLoad : -( Http://MaNoLyA.GeN.TR )-
1138365923 M * marl in what way are u trying to 'mount' the sound card?
1138365930 M * cohan i've seen a ghoost ;) asking things from netherworld i cannot even understand ;)
1138365970 M * schellh lol
1138367889 J * prae ~prae@ezoffice.mandriva.com
1138369166 M * cohan ah, very nice results: 2.6.15-vs2.1.0.2 does _not_ show the context issues
1138369189 M * cohan daniel_hozac: giving 2.1.0.3 or 2.1.0.4 a try now? ;) suggestions?
1138369211 M * daniel_hozac cohan: which base kernel? 2.6.15?
1138369219 M * daniel_hozac oh, sorry. my reading is off....
1138369226 M * daniel_hozac that's great though.
1138369381 M * cohan i'll try it with .4 now
1138370425 Q * shedi Quit: Leaving
1138370976 J * lilo ~lilo@cpe-24-167-94-255.houston.res.rr.com
1138371067 J * mkhl ~mkhl@200-148-41-148.dsl.telesp.net.br
1138371084 Q * lilo_ Ping timeout: 480 seconds
1138371664 M * cohan 2.6.15-vs2.1.0.4 booting...
1138371728 M * blizz ;)
1138371734 M * blizz im cookin, will catch up later if u
1138371737 M * blizz s/if/with/
1138371774 M * cohan also working - ! so the error has to be introcuded with the latest 2.1.0.5 !
1138371782 M * cohan *diffing the diffs*
1138371899 M * daniel_hozac interdiff is your friend ;)
1138371913 M * daniel_hozac if you recompile 2.1.0.5 now, does it still fail?
1138372042 M * cohan ahem - why do you expect it should not?
1138372062 M * cohan nothing changed in the environmet - and bartl also reproduced the error on 2.6.16rc1
1138372087 M * daniel_hozac i expect it should, just to rule out some bizarre compilation mishap or whatever.
1138372091 M * blizz because of the mysterious dark kernel magic :)
1138372125 Q * schellh Quit: 
1138372294 M * cohan well, i made hundreds of kernels the last 48 hours, so why not just recompiling 2.4.15-vs2.1.0.5...
1138372314 M * cohan but i really, really doubt thats it - the code is simply broken, thats it
1138372397 M * cohan okay, interdiff is an extra piece of software?
1138372430 M * daniel_hozac part of patchutils.
1138372489 M * daniel_hozac but that should only be necessary if your diff of the diffs gets annoying.
1138372518 M * cohan yes, it does, because they where made with different diff-parameters
1138372549 M * daniel_hozac ah.
1138372634 M * cohan okay, with some sedding its 600 lines now with classical diff approach
1138372654 M * cohan but i'll give interdiff a try - furthermore, i think i screw this now, i am not into that code at all
1138372670 M * cohan i think others can much, much easier track down what changed from 2.1.0.4 -> 2.1.0.5
1138372840 M * cohan but for your convenience, i'll give 2.6.15.1 a try, a version not listed yet in the hall-of-failures ;)
1138373644 J * shedi ~siggi@inferno.lhi.is
1138373740 J * gerrit ~gerrit@c-24-22-19-162.hsd1.or.comcast.net
1138373748 M * cohan okay, while that new 2.6.15.1 builds, i made an interdiff - interested, daniel_hozac?
1138373783 M * daniel_hozac sure.
1138373805 M * cohan https://servy.dynip.yawsp.de/~install/vs2.1.0.4-5.diff
1138373997 M * daniel_hozac so, to recap, 2.6.15-rc7-vs2.1.0.1 and 2.6.15-vs2.1.0.5 fails, but the ones in between work fine?
1138374027 M * daniel_hozac oh, no, i read that wrong again.
1138374034 M * daniel_hozac wow, i guess i really should sleep more.
1138374066 M * cohan to recap, as far as i can tell, the breakage is introducet with the switch from 2.1.0.4 to 2.1.0.5
1138374172 Q * Viper0482 Remote host closed the connection
1138374191 M * cohan summay: introduced __vx_page_fault, a change in struct _vx_cacct, some changes in limit.h, changes in cvirt_init.h, vx_info_proc_cacct(),  .... i am lost what could be responsible
1138374788 M * cohan rebooting with 2.6.15.1-vs2.1.0.5
1138374828 M * cohan added the interdiff to pastebin for reference in the mailinglist: http://pastebin.com/525691
1138374878 M * cohan 2.6.15.1-vs2.1.0.5 erroring out as expected
1138374898 M * cohan completely fresh build, from virgin vanilla .bz2 extracted sources
1138374902 M * daniel_hozac good.
1138375242 M * cohan okay, ml update is out - i think i cannot be of any use anymore
1138378133 P * rene- 
1138379235 J * Smutje_ ~Smutje@xdsl-84-44-187-168.netcologne.de
1138379379 Q * Smutje Ping timeout: 480 seconds
1138380730 J * meandtheshell ~markus@85-124-10-86.dynamic.xdsl-line.inode.at
1138380968 Q * jkl Quit: BitchX-1.1-final -- just do it.
1138381276 Q * Milf Ping timeout: 480 seconds
1138381600 J * stefani ~stefani@superquan.apl.washington.edu
1138382763 M * meandtheshell under debian you just have to "apt-get install gaim-themes" - that installs you 50 different ones on the fly 
1138382779 M * meandtheshell under windoof you have to search until you die 
1138382780 M * meandtheshell LOL
1138382794 M * meandtheshell FUCK closed source ;-
1138382850 M * meandtheshell oops - sorry - wrong window *blushing*
1138383694 Q * prae Quit: Execute Order 69 !
1138384085 J * anarcat ~anarcat@H144.C72.B0.tor.eicat.ca
1138384091 M * anarcat hello
1138384099 M * anarcat micah, are you there?
1138384111 M * anarcat you want a amd64 build for bind9?
1138384125 M * anarcat because i'm makeing one anyways
1138384180 J * dearaujo ~dan@cpe-66-25-189-193.austin.res.rr.com
1138384333 J * menomc ~amery@200.75.27.19
1138384442 Q * mnemoc Ping timeout: 481 seconds
1138384442 N * menomc mnemoc
1138384618 J * Viper0482 ~Viper0482@p549771CA.dip.t-dialin.net
1138384630 M * anarcat micah, so i guess you won't mind if i just upload it...
1138384639 M * micah anarcat: hey
1138384642 M * anarcat :)
1138384645 M * anarcat gotcha there :)
1138384656 M * anarcat btw, you were at gplv3 right?
1138384663 M * micah anarcat: oh, heh... didn't respond because of that, just happened to have a context switch available
1138384675 M * micah anarcat: I was, how did you know?
1138384687 M * anarcat omar saw you on a photograph
1138384696 M * anarcat we were meters apart :)
1138384704 M * anarcat i'm trying to find the picture
1138384705 M * micah anarcat: re: bind9, an amd64 version would be great, but the repo doesn't work yet with anything other than i386, I need to fix that
1138384710 M * micah wait, you were there?
1138384714 M * micah ah, crap!
1138384722 M * anarcat yep
1138384725 M * anarcat that sucks eh? :)
1138384728 M * micah haha
1138384732 M * micah ah internet
1138384753 M * anarcat yeah, a bit ridiculous
1138384770 M * micah thats one thing that sucks about the net... you could be playing an online game with your brother and you dont even know, or be bidding against your sister in an ebay auction and have no clue
1138384783 M * anarcat hehehe
1138384788 M * anarcat yeah.. it's pretty bad
1138384789 M * micah or be sitting next to someone who you would like to meet in person :)
1138384801 M * anarcat yep
1138384806 M * anarcat omar is looking for the picture
1138384857 M * anarcat so... are you using debarchiver for that repo of yours?
1138384871 M * anarcat i had bunch of fun trying to figure out how to fix debian.alternc.org for amd64
1138384947 M * anarcat http://gallery.hipatia.info/gplV3-Boston/fsfmit73
1138384954 M * anarcat i'm the dreaded guy :)
1138384960 M * anarcat which one are you?
1138384974 M * anarcat I: Successfully built in /usr/src/bind9-9.2.4
1138385082 M * micah I'm using something else, what is it called
1138385085 M * micah debpool
1138385132 M * micah hehe, i saw you and thought you must be an interesting person that I should talk to
1138385139 M * micah i've got the massive blue coat
1138385157 M * anarcat interesting
1138385166 M * anarcat i was incredibly tired the whole 2 days
1138385176 M * anarcat we made the trip from montreal by car after a day of hiking
1138385181 M * anarcat we got to the hotel at 3am
1138385185 M * anarcat it was ridiculous :)
1138385189 M * anarcat but i was glad to be there
1138385211 M * micah I was pretty tired too actually, had a hard time focusing on things
1138385216 M * micah are you on one of the committees?
1138385230 M * anarcat hmmm... 
1138385235 M * anarcat not *really*
1138385238 M * anarcat no
1138385261 M * micah haha there are some funny fotos there
1138385265 M * anarcat yep :)
1138385267 A * anarcat sleeping
1138385273 M * anarcat so glad :)
1138385279 M * anarcat named: -u not supported on Linux kernels older than 2.3.99-pre3 or 2.2.18 when using threads
1138385280 M * micah http://gallery.hipatia.info/gplV3-Boston/fsf44
1138385282 M * anarcat whee
1138385283 M * anarcat getting there
1138385292 M * micah anarcat: ah, comment that line out from /etc/default/bind9
1138385357 M * anarcat which line?
1138385391 M * micah the one with the -u part
1138385430 M * micah do you know those hipatia guys?
1138385433 M * anarcat yse
1138385434 M * anarcat yes
1138385438 M * anarcat juan carlos, mostly
1138385454 M * micah they are really cool folks
1138385457 M * anarcat you know him?
1138385458 M * anarcat yes!
1138385503 M * micah i've met them in other contexts, but I actually hung out and talked with them at mako's for some time during gplv3
1138385511 M * anarcat interesting
1138385521 M * anarcat i'm not very good at meeting people, just like that'
1138385530 M * anarcat conferences are new for me
1138385539 M * anarcat (you run bind as root??)
1138385546 M * micah yeah, I'm not good at it either
1138385553 M * micah well I have a vserver that runs nothing but bind
1138385558 M * micah so I'm not so concerned :)
1138385580 M * anarcat aaah.. i see
1138385583 M * anarcat hmmm... 
1138385595 M * anarcat otherwise, i don't get it, the kernel is much older than 2.3.99...
1138385615 M * anarcat oooh... i need --disable-threads
1138385620 M * anarcat http://linux-vserver.org/ProblematicPrograms
1138385870 M * anarcat ok well, micah, thanks for everything
1138385894 M * anarcat see you at a conference somewhere :)
1138385918 M * micah right, yeah
1138385927 M * micah next time we'll have a chance of knowing who is who :)
1138385931 M * anarcat yep :)
1138385936 M * anarcat i'm kinda hard to miss anyways :)
1138386071 Q * Hunger arion.oftc.net neutron.oftc.net
1138386077 J * Hunger Hunger.hu@Hunger.hu
1138386620 Q * Viper0482 Quit: bin raus, 
1138387289 Q * mire Ping timeout: 480 seconds
1138388074 N * Bertl_zZ Bertl
1138388079 M * Bertl morning folks!
1138388088 M * cohan wback, Bertl
1138388134 Q * _are_ Ping timeout: 480 seconds
1138388294 J * mire ~mire@150-167-222-85.COOL.ADSL.VLine.verat.net
1138388574 M * Bertl welcome mire!
1138388581 M * Bertl hey cohan! good work!
1138388597 M * cohan is there any solution around already? 
1138388785 M * Bertl hey, I just got up .. will have to get reasonable caffein levels first :)
1138388959 M * cohan hehe, okay - i still think you are kinda vampire, regarding to your day/night rythm
1138388989 M * Bertl nocturnal is the technical term :)
1138389042 P * anarcat thanks all
1138389129 J * Blissex ~Blissex@82-69-39-138.dsl.in-addr.zen.co.uk
1138389139 M * Bertl welcome Blissex!
1138389152 M * Blissex thanks!
1138389177 M * derjohn cohan, this would explain why he can spend so much time for us in coding vserver .. well if you are immortal you may have some time left :)
1138389336 M * Bertl derjohn: yes, mere mortal :)
1138389479 M * cohan i am just mirroring my devel sys onto a faster harddisk, so i can hopefully compile faster, just in case...
1138389503 M * Bertl can't hurt, first, I'll do some basic compile time tests ...
1138390150 J * _are_ ~are@62.112.159.81
1138390173 M * Hollow morning Bertl 
1138390241 M * Bertl hey Hollow! everything fine?
1138390247 M * Bertl wb _are_!
1138390253 M * Hollow yeah, i started working on libvconfig
1138390308 M * _are_ hi
1138391087 J * manjak ~manjak@212.110.81.81
1138391228 Q * manjak Quit: 
1138392340 M * Bertl cohan: are you fine if I 'abuse' you to test a few changes?
1138392806 J * BaZzOKa ~BaZzOKa@82.211.155.50
1138392860 M * Bertl welcome BaZzOKa!
1138392878 M * BaZzOKa Bertl :)
1138393156 Q * BaZzOKa Quit: I am Using Ebless ScripT v8 Special For Ops, If you Want Best Channel Protection, DownLoad your Copy From - http://i.am/Ebless - ScripT in mIRC v6.12 With Register - Have Fun & GooD Luck
1138393157 J * anarcat ~anarcat@H144.C72.B0.tor.eicat.ca
1138393161 M * anarcat hello
1138393173 M * Bertl wb anarcat!
1138393614 M * cohan Bertl: i am fine...
1138393620 M * anarcat hey micah 
1138393624 M * cohan it's just we are playing some cardgame here,
1138393629 M * anarcat care to share why debpool more than debarchiver?
1138393631 M * cohan so i'll be absent ocassionally
1138393638 M * anarcat esp. since it's not yet even in unstable :)
1138393642 M * cohan and mirroring took longer then expected
1138393760 M * micah anarcat: it was a long time ago that this decision was made, so I can't remember
1138393777 M * micah anarcat: but I tried a couple others before someone mentioned debpool and I think it was easy to setup quickly so I went with it
1138393809 M * anarcat hmmm
1138393825 M * anarcat debarchiver seems to work pretty well but it's giving me headaches on weird cases
1138393834 M * anarcat so you run unstable on the archive?
1138393954 M * Bertl cohan: okay, this one is supposed to change nothing in the behaviour, i.e. the bug should remain unchanged: http://vserver.13thfloor.at/Experimental/delta-config-clean01.diff
1138393964 M * Bertl cohan: could you verify that this is true?
1138395080 J * Aiken ~james@tooax6-064.dialup.optusnet.com.au
1138395090 M * Bertl morning Aiken!
1138395103 M * Aiken hi
1138395259 M * cohan Bertl: i'll queue that on the next occasion
1138395274 M * Bertl ok, great!
1138395374 M * Aiken was there a problem with 2.1.0.5? or is cohan just having problems with 2.1.0.5?
1138395390 M * Bertl there is a problem with 2.1.0.x
1138395436 M * Aiken oh .x? um, I have 2 machines running with 2.1.0.4 and seem ok, the alpha and this athlon
1138395463 M * Bertl 16:01 < cohan> to recap, as far as i can tell, the breakage is introducet with  the switch from 2.1.0.4 to 2.1.0.5
1138395501 M * Aiken ok
1138395654 M * cohan so the problem is in 2.1.0.5, not in 2.1.0.4, see mailinglist, just to avoid confusions ;)
1138395701 M * Bertl seems to be, but I agree it's most likely the page_fault code
1138395956 M * Aiken noticed in the mailing list 2.6.16-rc1-vs2.1.0.6.1 works for them
1138395961 M * Aiken any others have it working?
1138395970 M * Aiken it is still no go for me
1138396081 M * Bertl in what way?
1138396097 M * Bertl (I remember something, but no details)
1138396285 M * Aiken anther set of attribute statements seemed to deal with the oops
1138396331 M * Aiken am now getting   eth0 DMAing conflict in ne2k_pci_get_8390 and net2k_pci_block_something
1138396341 M * Aiken actually was getting them all along
1138396451 M * Bertl hmm, okay, so probably the alignment is still not perfect?
1138396474 M * Aiken possible I surpose
1138396479 P * dearaujo 
1138396483 M * Aiken the ne2k code is the same as it was in 2.6.15
1138396572 M * Bertl okay, will look into that shortly
1138396585 M * Aiken instead of chasing alignment issues my next step is to try a vanilla kernel
1138396596 M * Bertl hmm, good idea!
1138396612 M * Bertl cohan: okay, another patch which should not change anything :)
1138396614 M * Bertl http://vserver.13thfloor.at/Experimental/delta-config-clean02.diff
1138396744 Q * mkhl Quit: 
1138397405 M * Bertl cohan: and here is the fix for the bug :)
1138397407 M * Bertl http://vserver.13thfloor.at/Devel/PAT-2.1.1/detla-pfault-fix01.diff
1138397522 M * cohan shall i apply both at once, or first verify that delta-config does not solve the bug?
1138397567 M * Bertl I'm pretty sure that is the bug we were seeing
1138397578 M * Bertl so your choices are:
1138397595 M * Bertl - compile a few test kernels one with the other without
1138397601 M * Bertl - just use the fix for now
1138397616 M * Bertl - wait for the next release (will happen shortly)
1138398189 Q * meandtheshell Quit: a brave heart left ... 
1138399650 M * cohan # wget http://vserver.13thfloor.at/Experimental/delta-config-clean02.diff
1138399655 M * cohan HTTP request sent, awaiting response... 404 Not Found
1138399662 M * cohan ah, i see the typo
1138399975 M * Aiken Bertl vanilla is refusing to crash
1138399995 M * Bertl cohan: you see more than I do ...
1138400015 M * Bertl Aiken: okay, good ... crash means hang now?
1138400035 M * Aiken no hang
1138400051 M * Bertl yes, but 'latest' kernel+patch hangs?
1138400055 M * Aiken no
1138400062 M * Bertl instead it does what?
1138400063 M * Aiken not vanilla
1138400066 M * Aiken it works
1138400073 M * Bertl yeah, vanilla, but vserver
1138400074 M * Aiken whoops
1138400092 M * Aiken vanilla seems ok
1138400116 M * Aiken afetr a  2nd lot of attribute lines + vserver patch I am getting the eth0 DMAing errors
1138400135 M * Aiken the 2nd lot of attribute lines stopped the oopses I was getting
1138400135 M * Bertl which you do not get on vanilla, right?
1138400136 M * cohan s/detla/delta/ on your other URL, sorry
1138400146 M * Aiken the eth DMAing error was there all a long with the vserver patch
1138400159 M * Aiken I don't get that error with vanilla
1138400190 M * Bertl okay, could you show me that error once again please?
1138400232 M * Aiken eth0 DMAing conflict in ne2k_pci_get_8390 (also in net2k_pci_block_something)
1138400238 M * Bertl cohan: had the wrong name, just changed it, see ML
1138400303 M * cohan np, got it
1138400327 M * cohan but i just realize... i have to use 2.6.16-rc1, one patch failes vs. 2.6.15.1
1138400343 M * cohan can't find file to patch at input line 108
1138400344 M * Bertl np either, the cleanups are harmless
1138400350 M * cohan |diff -NurpP --minimal linux-2.6.16-rc1-vs2.1.0.7.1/include/linux/vs_tag.h linux-2.6.16-rc1-vs2.1.0.7.2/include/linux/vs_tag.h
1138400361 M * cohan ah, okay
1138400362 M * Bertl does not exist in older patches
1138400365 M * Aiken grumble, looks like vanilla finally went splat
1138400930 M * Aiken vanilla gives different errors again
1138401222 T * Bertl http://linux-vserver.org/ | latest stable 2.01, 1.2.10, 1.2.11-rc1, devel 2.1.0, exp 2.1.0.8, 2.0.1.2 | util-vserver-0.30.210 | libvserver-1.0.2 & vserver-utils-1.0.3 | He who asks a question is a fool for a minute; he who doesn't ask is a fool for a lifetime -- share the gained knowledge on the wiki, and we'll forget about the minute ;)
1138402011 M * Bertl Aiken: okay, started a kernel compile for alpha (will take some while) will look into fixing all the alignment issues but it will take a while ...
1138402027 M * Bertl for now I'm off to bed again, kind of nap attack ... back later ...
1138402033 N * Bertl Bertl_zZ
1138402430 J * cyfer ~cyfer@tor-irc.dnsbl.oftc.net
1138402654 Q * Aiken Ping timeout: 480 seconds
1138402837 P * cyfer 
1138402843 P * stefani I'm Parting (the water)
1138403759 M * cohan Bertl_zZ, you patches seem to work... so i'm looking forward for the next release
1138404011 M * SiD3WiNDR :)
1138405541 Q * michal_ Ping timeout: 481 seconds
1138405547 Q * Blissex Read error: Connection reset by peer