1186790404 J * buzztracker ~buzztrack@pelikan.garga.net
1186790919 Q * meandtheshell Quit: Leaving.
1186791030 Q * onox Quit: zzzz
1186792500 Q * Aiken Quit: Leaving
1186793989 Q * Guest322 Ping timeout: 480 seconds
1186793994 Q * buzztracker Quit: Leaving
1186794003 J * buzztracker ~buzztrack@pelikan.garga.net
1186794332 J * Aiken ~james@ppp121-45-255-55.lns2.bne4.internode.on.net
1186794718 J * gerrit_ ~gerrit@c-67-169-199-103.hsd1.or.comcast.net
1186795491 Q * blizz Ping timeout: 480 seconds
1186795539 J * friendly12345 ~friendly@ppp121-44-237-56.lns2.mel4.internode.on.net
1186795650 Q * ntrs_ Quit: Leaving
1186795660 J * blizz ~stephan@evilhackerdu.de
1186797594 Q * buzztracker Quit: Leaving
1186797605 J * buzztracker ~buzztrack@pelikan.garga.net
1186798064 Q * Aiken Remote host closed the connection
1186798357 J * Satriani ~linus@bl7-133-239.dsl.telepac.pt
1186798393 Q * bzed Ping timeout: 480 seconds
1186798783 Q * Freax Ping timeout: 480 seconds
1186799745 Q * FireEgl Read error: Connection reset by peer
1186800254 Q * Johnnie Read error: Connection reset by peer
1186800441 J * Johnnie ~jdlewis@c-67-163-142-234.hsd1.pa.comcast.net
1186800683 Q * gerrit_ Ping timeout: 480 seconds
1186800690 J * FireEgl FireEgl@Sebastian.Atlantica.CJB.Net
1186800738 Q * Ashsong Remote host closed the connection
1186801195 Q * buzztracker Quit: Leaving
1186801203 J * buzztracker ~buzztrack@pelikan.garga.net
1186801403 J * gerrit_ ~gerrit@c-67-169-199-103.hsd1.or.comcast.net
1186804451 J * Aiken ~james@ppp121-45-255-55.lns2.bne4.internode.on.net
1186804567 N * Bertl_oO Bertl
1186804574 M * Bertl evening folks!
1186804793 Q * buzztracker Quit: Leaving
1186804803 J * buzztracker ~buzztrack@pelikan.garga.net
1186806933 M * daniel_hozac evening Bertl!
1186807343 M * Bertl how's going?
1186807439 M * daniel_hozac good, you?
1186807605 M * Bertl yeah, fine too .. we actually got some work done today :)
1186807650 M * daniel_hozac oh?
1186807664 M * daniel_hozac what kind of work?
1186807692 M * Bertl mostly infrastructure setup for testing and such
1186807705 M * daniel_hozac ah, nice.
1186807718 M * Bertl but it looks like there will be some intensive ipv6 testing very soon
1186807759 M * daniel_hozac sounds good. did you get a chance to look at all my patches yet?
1186807833 J * ntrs ntrs@68-188-55-120.dhcp.stls.mo.charter.com
1186807894 M * Bertl nope, but that is scheduled for tomorrow
1186807911 M * Bertl including an update of 2.3.x + olpc
1186807933 M * daniel_hozac cool.
1186808393 Q * buzztracker Quit: Leaving
1186808406 J * buzztracker ~buzztrack@pelikan.garga.net
1186810878 M * Bertl okay, I'm off to bed now ... have a good one everyone! cya!
1186810884 N * Bertl Bertl_zZ
1186811995 Q * buzztracker Quit: Leaving
1186812003 J * buzztracker ~buzztrack@pelikan.garga.net
1186814166 J * dna ~dna@168-235-dsl.kielnet.net
1186815593 Q * buzztracker Quit: Leaving
1186815604 J * buzztracker ~buzztrack@pelikan.garga.net
1186816690 J * gerrit ~gerrit@dslb-084-060-193-255.pools.arcor-ip.net
1186817047 J * AstralSt ~astralsto@tor-irc.dnsbl.oftc.net
1186817275 Q * AStorm Ping timeout: 480 seconds
1186818255 J * Courtney ~Courtney@222-155-108-46.jetstream.xtra.co.nz
1186818733 Q * dna Quit: Verlassend
1186819193 Q * buzztracker Quit: Leaving
1186819215 J * buzztracker ~buzztrack@pelikan.garga.net
1186819940 Q * gerrit Read error: Connection reset by peer
1186820283 N * DoberMann[ZZZzzz] DoberMann
1186820712 M * arachnist http://www.linux.com/articles/114286
1186820713 M * arachnist lol
1186822793 Q * buzztracker Quit: Leaving
1186822803 J * buzztracker ~buzztrack@pelikan.garga.net
1186825593 J * meandtheshel1 ~markus@85.127.114.133
1186826055 J * ema ~ema@fw.galliera.it
1186826393 Q * buzztracker Quit: Leaving
1186826403 J * buzztracker ~buzztrack@pelikan.garga.net
1186826498 Q * Courtney Quit: pling pimeout
1186829546 J * dreamind apwdsl@p548A88E4.dip0.t-ipconnect.de
1186829574 M * dreamind Hi folks :D
1186829584 N * dreamind Guest376
1186829587 Q * Satriani Remote host closed the connection
1186829955 J * bzed ~bzed@10-205-116-85.dsl.manitu.net
1186829993 Q * buzztracker Quit: Leaving
1186830002 J * buzztracker ~buzztrack@pelikan.garga.net
1186830305 J * rgl ~Rui@84.90.10.107
1186830309 M * rgl hello
1186830710 M * rgl the timer interrupt (as seen on /proc/interrupts) shouldn't generated to both CPU cores?   
1186831138 Q * AstralSt Ping timeout: 480 seconds
1186831750 J * AstralSt ~astralsto@tor-irc.dnsbl.oftc.net
1186831933 J * meebey meebey@booster.qnetp.net
1186831972 M * cohan i moved some old vserver guests to a debian etch system, and created a "config tree" instead of the config file, including: echo "plain" >/etc/vservers/yawsp/apps/init/style
1186831986 P * meebey 
1186831993 M * derjohn cohan, that is nice to hear !
1186832017 M * cohan since my guests are bsd-style init
1186832026 M * cohan ah, hi derjohn ! :)
1186832050 M * derjohn cohan, I mean ... well ... that's not the qemu project we talked 'bout ;) sad :)
1186832055 M * cohan but now my recently migrated vserver cycles all the time, but i believe the trouble does not lie within the vserver itself
1186832078 M * derjohn well, you might try to vserver start --debug
1186832094 M * cohan derjohn: no, not yet - that project is next on the agenda this weekend, although i consider using debian etch and xen for that particular task.
1186832110 M * derjohn (i am not sure if that gives you much info on plain initsytle ... but worth a try )
1186832138 Q * gerrit_ Ping timeout: 480 seconds
1186832140 M * derjohn xen ?! with non-VT hardware? Then you can stay on qemu ... no difference
1186832172 M * cohan but in speed. at least to my little experience
1186832189 M * derjohn really? even with kqemu module installed and activated ?
1186832252 M * derjohn bzw: on VT hardwrae you could also load the kvm module (newer kernels... which is uses exactly like qemu but really really really fast. I use it on my X60 notebook for runnign a WinXP ....
1186832253 M * cohan did not run benchmarks, but i think so - well, test box right beside me will show this later, if i do not have to assist my family due to a sudden case of severe illness :(
1186832330 M * derjohn with VT Hardware XEN and KVM are surely pretty fast. Sad to hear about the illness. *thumbs*
1186832373 M * cohan the --debug seems useless, since the vserer start command terminates correctly. i think i miss something in the new util-vserver-stuff that indicates that the guest "did not start correctly, and needs cycling" - something like that
1186832394 M * derjohn for what kind of guest doy ou need "plain" ?
1186832424 M * derjohn yes, --debug helps only in case vserver does the init.
1186832458 M * derjohn on etch you should consider upgrading to .213 utils. (from backports.org or my vserver repo)
1186832458 Q * vasko Read error: Connection reset by peer
1186832459 J * vasko ~vasko@unreal.rainside.sk
1186832459 A * vasko is gone. Gone since Tue Jul 31 15:17:00 2007
1186832480 M * cohan old crux-guests, bsd-style init, and fakeinit - it works perfect with the old tools. and i don't like the idea vserver-tools are doing that much of the work - they should just provide a context and do the low-level-stuff - i do not consider running init-scripts very low-level
1186832480 Q * vasko Read error: Connection reset by peer
1186832493 M * derjohn with plain you can enter the guest while it's starting and see what goes on
1186832519 J * vasko ~vasko@unreal.rainside.sk
1186832519 A * vasko is gone. Gone since Tue Jul 31 15:17:00 2007
1186832534 M * derjohn just vserver start & vserver enter  or such
1186832557 M * cohan i can also look into it with vps xafuw  -  inside the geust, everything is okay, even apache is started, but something from the outside (visible due to the different init[6] of the context) lines in vps output forces shutdown again
1186832600 M * cohan vserver enter usually fails or is killed after some seconds, due to fact that outside stuff terminates the vserver without any obvious reason :( 
1186832646 M * derjohn that fast ??
1186832677 M * cohan i really consider placing my old, stupid-simple, low-level scripts in place, using simple, single file config files and where the host os and vserver-utils now just nothing about the distro :( </upsetting>
1186832678 M * derjohn well, I think if vserver receives an exit 1, it will not continue anymore.
1186832709 M * derjohn try step by step disable you initscripts ... or write a dmesg / bootlog within the guest ?
1186832757 M * cohan even vserver stop only succeeds once in a while, since most of the time, the vserver is "in cycle" and thus not considered running by - whatever stupid function the current vservers use to determine the state of a guest
1186832762 M * derjohn what is the last thing you see in syslog of the guest ?
1186832769 M * cohan the initscripts (well, it's bsd-style, just one :) completes sucessfull
1186832792 M * cohan Aug 11 13:46:15 yawsp exiting on signal 15
1186832803 M * cohan always the same
1186832811 M * cohan after
1186832812 M * cohan Aug 11 13:46:15 yawsp syslogd 1.4.1: restart.
1186832812 M * cohan Aug 11 13:46:15 yawsp init: Switching to runlevel: 6
1186832829 M * cohan the apache and other services produce some output
1186832852 J * gerrit_ ~gerrit@c-67-169-199-103.hsd1.or.comcast.net
1186832857 M * cohan so i'd EXPECT the vserver to stay running, since the restart seems not to be triggered from the inside
1186832868 M * derjohn hm , sounds like vserver utils dont see that vserver is running and terminate it after $timeout ?
1186832873 M * cohan even if networking or whatever else should not be set up as expected
1186832902 M * derjohn I remember having similar provs with a suse 7 guest 1-2 years ago 
1186832904 M * cohan yes, this is what it sounds like. but HOW does this vserver version look for a running guest?
1186832945 M * derjohn cohan, daniel_hozac is the de-facto maintainer currently. ask him. but before that you should try .213 utils. etch ships .212.
1186832956 M * derjohn (deb available, see above)
1186832961 M * cohan in the old days, it just looked for the context existing - at least my scripts did :(
1186832994 M * cohan derjohn: do you have plans on saturday, the 18th?
1186833037 M * derjohn yep, I am invited for a wedding (not mine ;)) ...
1186833045 M * derjohn but over the day i should be vail.
1186833048 M * derjohn *avil
1186833051 M * derjohn *avail
1186833059 N * fede eSa|
1186833094 M * cohan ah, ic - i am celebrating my birthday (and some kind of farewell party) somewhere around heidelberg, starting in the early afternoon
1186833130 M * derjohn i heard you'll leave for the emperor bush's country for a time ?
1186833232 M * cohan yes, leaving on 4th of september - destination ann arbor. but i hope i do not see much more of the emperor than his voting out :)
1186833281 M * derjohn so, you dont have plans to return (at least not short-term) ?
1186833300 M * derjohn sad, that I cant joim the celebration, as there is a collision with the wedding
1186833472 J * Pazzo ~ugelt@195.254.225.136
1186833519 M * cohan the PhD thesis will take 4-5 years, I expect to return only around christmas - and for some weddings i am supposed to attend :)
1186833558 M * derjohn rgl, you should see both cpus in /proc/interrupts. You need 'apt-get install irqbalance' to make the irq distributed among both cpus.
1186833589 M * cohan if the wedding takes place somewhere in the south of frankfurt, you could just show up for an hour :)
1186833592 Q * buzztracker Quit: Leaving
1186833593 M * derjohn cohan, PhD ? Wow !
1186833595 M * rgl derjohn, thx.  I'm reading that I can also set that has an kernel compile time option (CONFIG_IRQBALANCE).
1186833603 J * buzztracker ~buzztrack@pelikan.garga.net
1186833646 M * derjohn rgl, yes, I think so. but as I want to stay as close as possible to the debian kernel I use the usespace daemon.
1186833661 M * derjohn cohan, :) yes
1186833705 M * rgl derjohn, I'm using ubuntu.  and it doesn0t have irqbalance installed by default.
1186833756 M * derjohn rgl, yep, right. but why ?
1186833774 M * rgl derjohn, why I'm using ubuntu?
1186833814 M * derjohn no, why they dont install it by default? (at least if the installed detects a dualcore cpu ....) .. maybe file bug on launchpad ?
1186833907 M * friendly12345 derjohn: why have irqbalance enabled in the kernel? Are there really any benefit to make using it worthwhile?
1186833915 M * friendly12345 derjohn: *benefits
1186834006 M * derjohn friendly12345, good question, do you have an answer? My stomach tells me, it's a good idea to share the icq load amoung all avail cpus. I cannot prove that scientifically, though
1186834029 M * cohan damn, diagnosis of my uncle's disease, it's a cerebral hemorrhage 
1186834053 M * cohan the icq load *lol*
1186834070 M * cohan is icq multi-threaded? :)
1186834107 M * derjohn hmmm, yes, the icq load causes much interrupts on my work ;)
1186834133 M * friendly12345 derjohn: from what I've read any real benefits just aren't there. What I do know is that it causes a moderate amount of cpu wakeups, which isn't good for power consumption
1186834139 M * derjohn cohan, so you uncle will be under a brain-surgery very soon ? (to stop bleeding)
1186834196 M * cohan at least drainaged, to remove the pressuse - that is happening right now. don't know about the next steps, depends on size and nature of the bleeding, i believe
1186834226 M * derjohn friendly12345, so nothing good for a notebook .... but to be honest: on my dual-core amd server, I saw load avg dropping after installing irqbalance. the server was syncing it's large md0 device.
1186834280 M * friendly12345 derjohn: hmm well that is interesting
1186834285 M * derjohn cohan,so that a hemorrhagic stroke (schlaganfall) ?
1186834302 M * derjohn friendly12345, as I sad : I have to "scientific" prove ....
1186834328 M * cohan i prepare to help out at the farm of my uncle, but until it is decided that i can be of any help, i still try to get my old vservers up
1186834445 M * cohan i think it is not a stroke. it is "just" a bleeding, he started to feel a bad headache in the morning which did not vanish and lost orientation - that are symptoms for pressure in the brain, not for reduced blood supply (that is the condition for a stroke, at least afaik)
1186834487 M * derjohn well, at least close to it. good that yu ve been pretty fast in diagnosis ...
1186834670 M * cohan well, i was not involved in the diagnosis, i just get updated and help organizing - 30 cows need to be milked and fed this day, regardless of the state of my uncle
1186834704 M * friendly12345 cohan: better get a moove on then
1186834722 M * derjohn oh well, besides a server-farm you have a real-farm to care for ? 
1186834727 M * friendly12345 cue the bovine jokes in 5..4..3.. 2
1186834753 M * derjohn then hurry up, before you trigger the "xploding-cow" bug ;)
1186834758 M * cohan i am picked up if required - no, not me :) just the family of my uncle... but the family now has other troubles. but i spent much time on the farm when i was a child, nice times ;)
1186834798 M * cohan no, not me relates to "my farm" ;)
1186834805 J * yarihm ~yarihm@84-75-109-39.dclient.hispeed.ch
1186834948 M * cohan that is another thing about farmers: absolute pragmatism. he and his family will not expect the world to stop turning because of such events.
1186835001 M * derjohn what is "** glibc detected *** free(): invalid pointer: 0xb7e50d38 ***" ? A broken glibc ?
1186835045 M * derjohn cohan, well, yes, what other choise do they have ...
1186835374 M * cohan hm - interesting glibc message. when does it occur?
1186835459 M * AstralSt cohan, ah, rural lifestyle :-)
1186836330 J * Piet ~piet@tor.noreply.org
1186836523 M * Guest376 bye
1186836529 Q * Guest376 Quit: Guest376
1186836561 Q * Aiken Quit: Leaving
1186836652 Q * ema Quit: leaving
1186836723 J * DoberMann_ ~james@AToulouse-156-1-171-70.w90-38.abo.wanadoo.fr
1186836832 Q * DoberMann Ping timeout: 480 seconds
1186837193 Q * buzztracker Quit: Leaving
1186837203 J * buzztracker ~buzztrack@pelikan.garga.net
1186837332 M * cohan weird - i just started the vserver again - this time it keeps up running - but i did not change a thing, only did some more phonecalls...
1186837359 M * derjohn I wish i could solve problems that way ...
1186837440 M * cohan i don't like that way, i miss the answers!
1186837448 M * Loki|muh 42!
1186837461 M * cohan in that case i miss the question :)
1186837476 M * derjohn Loki|muh, nowadays 90-60-90 
1186837485 M * cohan i am picked up in an hour, derjohn - i fear we have to delay the other project :/ again :(
1186837504 M * derjohn when did you plan to start it ?
1186837566 M * derjohn cohan, could it be something in your guest was unhappy and did a shutdown ? 
1186837896 M * cohan don't think so - but you never know. now some network config problems reside, but those i can at least track down :)
1186837901 P * friendly12345 
1186837911 M * cohan derjohn: i planned it today and tomorrow, but this plans are void now
1186837970 J * friendly12345 ~friendly@ppp121-44-237-56.lns2.mel4.internode.on.net
1186838032 M * derjohn cohan, ok, sure, there are things that have really higher prio than IT....
1186838060 M * cohan not many, but yes, there are :)
1186838069 M * cohan better :( in that case
1186838077 Q * friendly12345 
1186839537 Q * rgl Quit: Leaving
1186840792 Q * buzztracker Quit: Leaving
1186840803 J * buzztracker ~buzztrack@pelikan.garga.net
1186842336 Q * FireEgl Read error: Connection reset by peer
1186843279 J * FireEgl FireEgl@Sebastian.Tcldrop.US
1186843347 J * friendly12345 ~friendly@ppp121-44-237-56.lns2.mel4.internode.on.net
1186843366 P * friendly12345 
1186844393 Q * buzztracker Quit: Leaving
1186844403 J * buzztracker ~buzztrack@pelikan.garga.net
1186844788 J * BobR_afk_ odie@IRC.13thfloor.at
1186844802 J * Bertl_zZ_ herbert@IRC.13thfloor.at
1186844890 Q * the-me Ping timeout: 480 seconds
1186844905 Q * BobR_afk Ping timeout: 480 seconds
1186844920 Q * derjohn Ping timeout: 480 seconds
1186844920 Q * Bertl_zZ Ping timeout: 480 seconds
1186844934 J * derjohn ~derjohn@80.69.41.3
1186846580 J * lilalinux ~plasma@dslb-084-058-198-043.pools.arcor-ip.net
1186847993 Q * buzztracker Quit: Leaving
1186848004 J * buzztracker ~buzztrack@pelikan.garga.net
1186848126 J * djbclark dclark@opensysadmin.com
1186851241 J * rgl ~rgl@84.90.10.107
1186851243 M * rgl hi
1186851594 Q * buzztracker Quit: Leaving
1186851603 J * buzztracker ~buzztrack@pelikan.garga.net
1186852100 Q * rgl Quit: Zzzzz
1186853852 J * rgl ~rgl@84.90.10.107
1186853854 M * rgl hello
1186854304 J * ema ~ema@fw.galliera.it
1186855192 Q * buzztracker Quit: Leaving
1186855207 J * buzztracker ~buzztrack@pelikan.garga.net
1186855682 N * DoberMann_ DoberMann[PullA]
1186856024 Q * yarihm Quit: Leaving
1186856958 J * fatgoose ~samuel@76-10-147-136.dsl.teksavvy.com
1186856980 Q * Pazzo Quit: ...
1186857093 J * fatgoose_ ~samuel@76-10-147-136.dsl.teksavvy.com
1186857094 Q * fatgoose Read error: Connection reset by peer
1186857209 J * marcfiu ~marcfiu@64.129.10.212
1186857218 P * marcfiu 
1186857220 J * oliwel ~mail-at-o@ppp-82-135-65-11.dynamic.mnet-online.de
1186857220 Q * fatgoose_ 
1186857231 A * oliwel waves hello to the crowd
1186857438 M * oliwel Anybody here ?
1186857529 M * sid3windr nope
1186857846 M * eyck nope
1186858145 M * oliwel looking for a good idea....updated my gentoo guest and now it wont boot....
1186858152 M * oliwel hangs when executing sysinit
1186858159 M * oliwel and have no clue how to dbug ;(
1186858797 Q * buzztracker Quit: Leaving
1186858806 J * buzztracker ~buzztrack@pelikan.garga.net
1186859582 J * Patrick Patrick@Linux-Dev.org
1186860130 Q * oliwel Quit: ChatZilla 0.9.78.1 [Firefox 2.0.0.4/2007060520]
1186860417 N * Bertl_zZ_ Bertl
1186860424 M * Bertl morning folks!
1186860788 M * bXi yo Bertl 
1186860815 J * Blissex ~Blissex@82-69-39-138.dsl.in-addr.zen.co.uk
1186860850 M * rgl oh, octopus Bertl is here!
1186860859 M * sid3windr Bertl's an octopus?!
1186860919 M * rgl and big one :)
1186860932 M * sid3windr o_O
1186860938 M * Bertl lol
1186860958 M * sid3windr \ /
1186860959 M * sid3windr =o=
1186860961 M * sid3windr / \
1186860981 M * sid3windr Bertl: I didn't know you were an octopus :p
1186861005 M * Bertl sid3windr: neither did I :)
1186861176 M * bXi Bertl: all well and stable here
1186861210 M * Bertl bXi: excellent, so you are running 2.3.x + daniel's patches now, yes?
1186861228 M * bXi yeah
1186861293 M * Bertl okay, I will update the kernel tree today (next few hours), so if you feel like, you can upgrade too (will include all the fixes and use 2.6.22.2+)
1186861364 M * bXi cool
1186861367 M * bXi will do obviously
1186861546 M * rgl the 2.6.22.x are bug fix releases only?
1186861614 M * bXi venice ~ # uname -a
1186861614 M * bXi Linux venice 2.6.22.1-vs2.3.0.15 #1 Thu Aug 9 22:35:52 CEST 2007 x86_64 AMD Athlon(tm) 64 Processor 3000+ AuthenticAMD GNU/Linux
1186861637 M * Bertl rgl: yep
1186861684 M * rgl woah. the .2 seems to have a lot of fixes *G*
1186861805 M * rgl you guys enable CONFIG_IRQBALANCE ?
1186861869 M * eyck hmm, sometimes it's better to leave it to userspace irqbalance daemon 
1186861873 M * Bertl it is new (again) so we'll see :) but in general it is a good choice
1186861880 M * eyck  ( or even all-manual managing )
1186861907 M * eyck not always, people from irqbalance claim their solution is smarter, 
1186861931 M * eyck  and manual irq managing works best if you know what your server is doing 
1186861950 M * rgl humm, what can I read to known what is best?
1186861977 M * eyck  umm. 
1186862033 M * Bertl first, the kernel option's help, of course
1186862185 M * rgl only if I could find the LOCAL_APIC thingy *G*
1186862202 M * rgl I think it needs to be enable before using the IRQBALANCE
1186862249 M * rgl humm, that an i386 thingy.  and I have x86_84
1186862280 M * rgl so, how can I even enable it?
1186862292 M * rgl ./arch/x86_64/kernel/io_apic.c:#if defined(CONFIG_GENERIC_PENDING_IRQ) || defined(CONFIG_IRQBALANCE)
1186862305 M * rgl but, the IRQBALANCE thing is inside i386 :|
1186862316 M * rgl I mean: CONFIG_IRQBALANCE
1186862319 M * Bertl do you have a multi processor x86_64 system?
1186862323 M * rgl err I mean: ./arch/i386/Kconfig:config IRQBALANCE
1186862342 M * rgl I have a Core2Duo with two cores.
1186862396 Q * buzztracker Quit: Leaving
1186862404 J * buzztracker ~buzztrack@pelikan.garga.net
1186862603 M * bXi rgl: are the core2duo with single cores then?
1186862725 M * rgl bXi, sorry I mislead you.  all C2D have at least 2 cores.
1186862742 M * bXi rgl: i know :p was just messing aroud
1186862774 M * rgl *G*
1186862792 M * rgl well, the irqbalance.org page explains this stuff :)
1186863284 M * rgl ok.  I've installed the irqbalance daemon.
1186863467 Q * ema Quit: leaving
1186863647 J * jmcaricand ~jm@dyn-91-168-10-69.ppp.tiscali.fr
1186863684 M * bXi Bertl: if you want to have something tested you can ask me :)
1186863713 M * Bertl bXi: excellent! as I said, should have something in a few hours
1186863812 M * bXi just throw m at me :D
1186863879 M * Bertl k, will do :)
1186863905 M * bXi unless it takes 4 hours :p i'll be asleep then
1186864014 M * Bertl shouldn't take too long
1186864071 M * bXi i'll be playing some civilization in the meantime
1186864097 Q * jmcaricand Quit: Parti
1186864162 M * rgl Bertl, are you hacking on the 2.6.22.2 patch?
1186864341 M * Bertl yup, why?
1186865343 Q * mnemoc Ping timeout: 480 seconds
1186865614 M * rgl Bertl, I can try it in this desktop too :D
1186865643 M * rgl I'm currently running 2.6.22.1
1186865735 M * Bertl cool
1186865810 J * mnemoc ~amery@kilo105.server4you.de
1186865977 M * rgl what HZ setting you guys use?
1186865993 Q * buzztracker Quit: Leaving
1186866005 J * buzztracker ~buzztrack@pelikan.garga.net
1186866051 M * rgl my other server (2.6.19) is using 100Hz, but this new one is using 250 Hz, and I don't remember setting this.
1186866140 M * doener 250 is the default since .20 or so
1186866174 M * rgl ah, then it was that.
1186866191 M * rgl what you think about this tickeless kernel?
1186866269 M * doener nice for powersaving :-) but until cpuidle finally makes it into Linus' tree, it can also make things worse, e.g. with the ipw* wireless stuff, because the current acpi idle handler sucks
1186866284 M * doener s/sucks/is not quite compatible with tickless/
1186866300 M * doener ;-)
1186866377 A * rgl googles cpuidle *G*
1186866400 M * doener add "venki" or so to your search terms, otherwise you'll end up with some windows crap
1186866452 M * doener should be in Len's acpi branch by now, too (just IIRC)
1186866578 J * yarihm ~yarihm@84-74-20-183.dclient.hispeed.ch
1186867063 Q * yarihm Quit: Leaving
1186867089 M * rgl I see.  I'll wait until its in mainline :D
1186867114 M * rgl I never really understood why we need a tick at all.
1186867308 M * rgl what is the LOC interrupt?
1186867666 Q * Piet Quit: Piet
1186868554 J * Aiken ~james@ppp121-45-255-55.lns2.bne4.internode.on.net
1186868745 J * virtuoso ~s0t0na@ppp91-122-94-50.pppoe.avangard-dsl.ru
1186869054 Q * Johnnie Ping timeout: 480 seconds
1186869165 Q * Blissex Remote host closed the connection
1186869256 J * yarihm ~yarihm@84-74-20-183.dclient.hispeed.ch
1186869318 J * virtuoso_ ~s0t0na@pppoe-14.2.110.89-adsl.spbnit.ru
1186869595 Q * buzztracker Quit: Leaving
1186869603 J * buzztracker ~buzztrack@pelikan.garga.net
1186869720 Q * virtuoso Ping timeout: 480 seconds
1186870036 N * virtuoso_ virtuoso
1186871091 Q * FireEgl Read error: No route to host
1186871280 N * AstralSt AStorm
1186871290 M * AStorm doener, cpuidle is in Linus' tree
1186871298 M * AStorm since 2.6.21 or so
1186871300 M * AStorm for x86
1186871305 M * AStorm x86_64 is in -mm
1186871350 M * AStorm most other archs are tickless from start
1186871730 M * Bertl daniel_hozac: ping?
1186871853 M * bXi Bertl: how goes the coding?
1186871980 M * Bertl should be finished in half an hour
1186872005 M * bXi cool
1186872036 M * bXi i got 25 turns left on my civ4 game :p
1186872658 Q * AStorm Remote host closed the connection
1186872692 J * AStorm ~astralsto@ns.km10530.keymachine.de
1186873193 Q * buzztracker Quit: Leaving
1186873207 J * buzztracker ~buzztrack@pelikan.garga.net
1186873224 J * lilalinux_ ~plasma@dslb-084-058-214-036.pools.arcor-ip.net
1186873685 Q * lilalinux Ping timeout: 480 seconds
1186873879 M * daniel_hozac Bertl: pong
1186874064 M * Bertl ah, I have a few questions/comments, got a few minutes?
1186874120 M * daniel_hozac sure.
1186874245 M * Bertl delta-lback-fix01.diff
1186874274 M * Bertl why do we want to allow the check for disabled ipv4/6 here?
1186874292 M * daniel_hozac primarily we want it for IPv4.
1186874309 M * daniel_hozac otherwise the loopback interface isn't shown in e.g. ip a, even though an lback address is setup and configured.
1186874343 M * Bertl okay, I guess we should handle the lback explicit there, any problems with that?
1186874353 M * daniel_hozac no, sounds fine.
1186874364 M * bXi Bertl: i'm going to bed now
1186874372 M * bXi could you pm me the places where i can get the patch?
1186874375 M * bXi when its done
1186874379 M * Bertl bXi: okay, will do so, tx!
1186874383 M * Bertl daniel_hozac: delta-nxc-feat01.diff
1186874387 M * bXi cool
1186874396 M * Bertl I think, that one is changing the semantics :)
1186874418 M * daniel_hozac how so?
1186874467 M * daniel_hozac i mean, it is, but i think it's more correct this way. or do we want users in guests to have root-equivalent privileges there?
1186874516 M * Bertl good argument
1186874557 M * Bertl okay, would have been better to have two patches there
1186874568 Q * bzed Quit: Leaving
1186874572 M * Bertl one which adjust the semantics and the other which does the moving around
1186874601 M * daniel_hozac yeah, a few of the patches aren't really split up in a sane way.
1186874624 M * Bertl I will include the semantical change (as it sounds plausible) but I prefer to keep the 'explicit' override check there, because I consider that better readable
1186874649 M * daniel_hozac okay, makes sense.
1186874689 M * Bertl so we make that:
1186874722 M * bXi good night!
1186874728 M * Bertl         if ((protocol == IPPROTO_ICMP) &&
1186874728 M * Bertl                 nx_capable(answer->capability, NXC_RAW_ICMP))
1186874728 M * Bertl                 goto override;
1186874746 M * daniel_hozac yeah, looks good.
1186874746 M * Bertl bXi: have a good one!
1186874814 M * Bertl delta-connect-fix01.diff
1186874831 M * daniel_hozac yeah, ugly hack.
1186874835 M * Bertl can we assume that sk->sk_nx_info->v6.ip is valid/fine?
1186874863 M * daniel_hozac i thought about that too... i guess it would be best to add an && current_nx_has_v6()
1186874887 M * daniel_hozac the thing is, creating an IPv6 socket should be disallowed if there are no addresses assigned.
1186874904 M * Bertl ah, btw, I renamed that nx_current_has_v6() ;)
1186874914 M * daniel_hozac s/should be/is/, i guess we haven't really defined the semantics there yet.
1186874928 M * Bertl yeah, but they could be removed lateron (the ipv6 ips)
1186874929 M * daniel_hozac hehe, that's fine. i just went with current_nx_info and such.
1186874933 M * daniel_hozac exactly.
1186875044 M * Bertl okay, this part is doing the connect, and we want the loopback/first ip, right?
1186875075 M * daniel_hozac yeah.
1186875097 M * Bertl but we do not have a v6 loopback mapping yet
1186875126 M * daniel_hozac right.
1186875163 M * Bertl but that shouldn't be too hard to add ... 
1186875175 M * Bertl is that case very common?
1186875182 M * daniel_hozac connect to ::?
1186875196 M * daniel_hozac the comment above that piece of code describes it as a weird BSD-ism
1186875291 M * daniel_hozac i guess if we do the loopback thing in a more centralized place, that should cover this case too.
1186875309 M * Bertl okay, I'll postpone that then and just check for v6
1186875382 M * Bertl delta-udp-fix01.diff
1186875403 M * Bertl this one is changing the final score ... which probably is not a good idea, no?
1186875503 M * daniel_hozac yeah, that didn't seem quite right to me either... i figured bonbons had thought about it more than me.
1186875528 M * Bertl I think we should make that a continue case only
1186875541 M * daniel_hozac sounds good to me.
1186875625 M * Bertl and I'm going to move the check to the end of the list
1186875633 M * Bertl as it might be more expensive than the other checks
1186875690 M * daniel_hozac but then we have to duplicate the ipv6_addr_any(&np->rcv_saddr) check.
1186875736 Q * rgl Quit: Zzzzz
1186875737 M * Bertl yes
1186875783 M * Bertl hmm, well, let me think about that again ...
1186876008 M * Bertl http://paste.linux-vserver.org/5355
1186876045 M * Bertl forget it, that is nonsense
1186876071 M * daniel_hozac hmm? looks fine to me.
1186876100 M * Bertl what is the purpose of this check?
1186876146 M * daniel_hozac making sure that an :: bound socket only gets packets destined for an address in that context.
1186876203 M * Bertl okay, yes, that should be fine then, got confused by the daddr
1186876311 M * Bertl delta-lback-fix02.diff
1186876326 M * Bertl I'm completely unhappy with that one :)
1186876371 M * daniel_hozac me too, two completely unrelated hunks in one patch :)
1186876379 N * DoberMann[PullA] DoberMann[ZZZzzz]
1186876381 M * Bertl but I guess I managed to rewrite this in a way I like, sec
1186876389 M * daniel_hozac the first hunk i wanted to talk to you about.
1186876448 M * Bertl http://paste.linux-vserver.org/5356
1186876459 M * Bertl (for the second)
1186876527 M * daniel_hozac couldn't the last branch just be else?
1186876544 M * daniel_hozac as v4_addr_in_nx_info checks for the bcast.
1186876564 M * Bertl okay, yes
1186876575 M * daniel_hozac but yes, that looks much better.
1186876592 M * daniel_hozac btw, speaking of the bcast, i think we should initialize that to 255.255.255.255.
1186876609 M * Bertl yep, that is on my todo :)
1186876617 M * daniel_hozac okay :)
1186876625 M * Bertl 0 is a bad default for that
1186876629 M * daniel_hozac yeah.
1186876633 M * Bertl although it should not hurt immediately
1186876649 M * daniel_hozac well... it should give the guest access to all IPs, no?
1186876673 M * Bertl no, I don't think we have any checks which would do that
1186876685 M * Bertl we only check 'known' ips against bcast
1186876740 M * Bertl i.e. 0 should only have the meaning of a specific ip using 0.0.0.0 not any
1186876762 M * Bertl (which probably isn't even a possibility in the checks)
1186876793 M * Bertl but anyways, ~0 is a much better default :)
1186876793 Q * buzztracker Quit: Leaving
1186876798 M * daniel_hozac heh, yeah.