1226275589 N * Guest2451 Genghis
1226275630 N * Genghis Guest2459
1226276630 J * trippeh atomt@uff.ugh.no
1226276706 Q * pisco_ Ping timeout: 480 seconds
1226276728 J * pisco ~pisco@86.59.118.153
1226276755 Q * nou Ping timeout: 480 seconds
1226278181 Q * hparker Quit: Read error: 104 (Peer reset by connection)
1226278317 J * hparker ~hparker@linux.homershut.net
1226279280 N * Guest2459 Genghis
1226279320 N * Genghis Guest2463
1226279703 J * nou Chaton@causse.larzac.fr.eu.org
1226281845 Q * nou Quit: Lost terminal
1226282970 N * Guest2463 Genghis
1226283010 N * Genghis Guest2473
1226283584 Q * hparker Quit: Read error: 104 (Peer reset by connection)
1226285160 J * hparker ~hparker@linux.homershut.net
1226285380 Q * hparker Remote host closed the connection
1226286660 N * Guest2473 Genghis
1226286700 N * Genghis Guest2483
1226288347 J * hparker ~hparker@linux.homershut.net
1226289645 Q * rgl Ping timeout: 480 seconds
1226289656 Q * trippeh Ping timeout: 480 seconds
1226289915 J * trippeh atomt@uff.ugh.no
1226290350 N * Guest2483 Genghis
1226290390 N * Genghis Guest2495
1226291876 Q * aj__ Ping timeout: 480 seconds
1226294043 N * Guest2495 Genghis
1226294080 N * Genghis Guest2501
1226297627 J * ktwilight__ ~ktwilight@102.93-66-87.adsl-dyn.isp.belgacom.be
1226297655 J * derjohn_mob ~aj@e180206112.adsl.alicedsl.de
1226297734 N * Guest2501 Genghis
1226297770 N * Genghis Guest2508
1226297906 Q * ktwilight_ Ping timeout: 480 seconds
1226298175 J * mtg ~mtg@dialbs-088-079-143-204.static.arcor-ip.net
1226299284 J * ntrs__ ~ntrs@77.29.10.142
1226299995 J * yarihm ~yarihm@whitehead2.nine.ch
1226301425 N * Guest2508 Genghis
1226301460 N * Genghis Guest2515
1226301701 Q * derjohn_mob Ping timeout: 480 seconds
1226301881 Q * larsivi Ping timeout: 480 seconds
1226303074 J * davidkarban ~david@193.85.217.71
1226303709 M * fb daniel_hozac: some (latest?) from http://people.linux-vserver.org/~dhozac/t/uv-testing/ ?
1226303773 M * fb daniel_hozac: this will work with older (vs2.3.0.35) kernel patch?
1226303851 Q * hparker Remote host closed the connection
1226305115 N * Guest2515 Genghis
1226305150 N * Genghis Guest2524
1226305877 Q * fb Quit: leaving
1226307211 J * fb fback@red.fback.net
1226307218 M * fb hello again :)
1226308809 N * Guest2524 Genghis
1226308840 N * Genghis Guest2529
1226310272 J * ntrs_ ~ntrs@77.29.10.239
1226310700 Q * ntrs__ Ping timeout: 480 seconds
1226312343 J * rgl ~rgl@bl8-131-68.dsl.telepac.pt
1226312345 J * Pazzo ~ugelt@reserved-225136.rol.raiffeisen.net
1226312499 N * Guest2529 Genghis
1226312530 N * Genghis Guest2535
1226313636 Q * Tuxbubling Remote host closed the connection
1226314026 Q * esa` Ping timeout: 480 seconds
1226314036 J * esa ~esa@ip-87-238-2-45.static.adsl.cheapnet.it
1226314355 J * mrfree ~mrfree@host1-89-static.40-88-b.business.telecomitalia.it
1226315848 J * gnuk ~F404ror@pla93-3-82-240-11-251.fbx.proxad.net
1226316190 N * Guest2535 Genghis
1226316230 N * Genghis Guest2541
1226316726 Q * kiorky Remote host closed the connection
1226317268 N * Bertl_zZ Bertl
1226317272 M * Bertl morning folks!
1226318822 J * kiorky ~kiorky@cryptelium.net
1226319880 N * Guest2541 Genghis
1226319920 N * Genghis Guest2547
1226320468 M * hijacker_ afternoon Bertl 
1226320906 Q * Aiken Quit: Leaving
1226322517 M * fb Bertl :-)
1226322537 M * fb Bertl: daniel_hozac was right, it works with latest util-vserver 
1226322563 M * Bertl good :)
1226322605 M * fb Bertl: afair you wanted to test something on xfs
1226322629 M * fb Bertl: all my guests are on xfs, so maybe I can help a bit
1226322655 M * Bertl well, xfs needs testing .. I personally do not want to test xfs a lot, as I do not use xfs myself :)
1226322679 M * Bertl testing here means to test it in a real world setup with hashification and such
1226322703 M * fb unfortunately, i'm not using hashification :(
1226322798 M * fb maybe in the future, when grsec is ported to now-2.3
1226322816 M * Bertl IIRC, the future is here :)
1226322847 M * Bertl means: harry provided a grsec patch for recent Linux-VServer kernels
1226322970 M * Bertl daniel_hozac: pre2824 is recent?
1226323034 M * fb Bertl: grsex+pax is important on this machine, it's exposed to an Internet, and it has shell accounts I don't trust, so security is a priority there
1226323052 M * Bertl so why not use that patch?
1226323083 M * fb Bertl: it's in harry's home on people.linux-vserver?
1226323091 M * Bertl yup, I moved it there
1226323096 M * Hawq hello
1226323101 M * Bertl hey Hawq!
1226323127 M * fb Bertl: okay, I'll take a look at this
1226323169 M * fb Bertl: we hawe holliday tomorrow, maybe it's a good time to play with this machine
1226323217 M * Hawq is it normal and expected behavior while running X inside guest that I can't switch back to text console using ctrl+alt+f1? actually, I do switch but display remains at X session and I'm not able to switch back.
1226323237 M * fb Bertl: it runs few guests that are nearly identical, but a few configuration files
1226323261 M * Bertl Hawq: kind of, you are lacking the proper permissions for arbitrary terminal switching
1226323264 M * fb so hashification may make sense there :)
1226323299 M * Bertl fb: go crazy there ... let us know how it goes (but make backups)
1226323300 M * Hawq Bertl: anything can be done to get normal expected behavior? :)
1226323332 M * Bertl Hawq: well, normal and expected behaviour differ here .. why are you running X inside a guest btw?
1226323489 M * Hawq Bertl: its my home router as well as test machine and desktop for rest of my family. and I want those things separated
1226323548 M * Bertl but you are aware, that you are already giving that guest (the one with X running) a lot more permissions than can be considered secure?
1226323571 N * Guest2547 Genghis
1226323584 M * Hawq Bertl: yeah, but I'm fine with that.
1226323585 M * Bertl (and you are probably going to give it even more to make the switching work :)
1226323610 N * Genghis Guest2552
1226323611 M * Bertl if you really want to keep that separate, run X on the host, and connect to the guest for the desktop
1226323700 Q * mrfree Quit: Leaving
1226323842 M * Bertl makes it secure and completely separated
1226323887 M * Hawq I was thinking about that too, but I'd like to try get switching working anyway :) just to know how :)
1226323921 M * Bertl okay, get strace -fF up and running, check where it fails atm (you probably get an EPERM somewhere)
1226323949 M * Bertl upload the relevant trace part (e.g. to paste.linux-vserver.org) and we'll see what is missing
1226324021 M * Hawq stupid question: to which process should I connect strace?
1226324077 M * Bertl most likely to X
1226324928 M * Hawq weird thing. when I try to switch console with strace attached my screen goes black and keyboard stops working. after killing X via ssh screen goes messy and I have to reboot machine to get it back.
1226324946 M * Hawq w/o strace it doesn't behave like that
1226324983 M * Bertl did you get the trace?
1226325009 M * Bertl but messy screen sounds like a X11 problem
1226325028 M * Hawq yeah. just going to paste it. grep doesn't find EPERM there though.
1226325627 J * mrfree ~mrfree@host1-89-static.40-88-b.business.telecomitalia.it
1226325738 M * Hawq problem. strace is 1.5meg and while it uploaded ok to pld.pastebin.com it doesn't work and my pastebinit script doesn't recognize paste.linux-vserver.org. just a moment, I'll upload this trace somewhere
1226325776 M * Hawq hm, not 1.5 meg, 800 kb. sorry. 1.5 meg was from kdm :)
1226325794 M * Bertl can you put it somewhere where I could download it?
1226325823 M * Hawq yes. uploading now.
1226325900 M * Hawq http://hawk.furud.net/files/
1226326073 M * Hawq I attached to both kdm and x while on login prompt and tried to switch to tty1.
1226326104 A * ard thinks that X in a vserver is a good idea
1226326114 M * ard just to keep the base system clean from X stuff :-)
1226326152 M * Hawq ard: true :)
1226326170 M * Bertl you can do that with a chroot too
1226326219 M * Hawq Bertl: sure, but since machine is vserver enabled why to bother with chroots? :)
1226326248 M * ard vservers is way better than a chroot
1226326266 M * Hawq easier to maintain for me
1226326277 M * Hawq than bunch of chroots
1226326281 M * ard but I haven't tried it with fglrx though
1226326283 M * Bertl that is true, but there is no real advantage over a chroot, if you give dangerous capabilities :)
1226326288 M * ard at least: i did once, and it worked
1226326304 M * ard there is! no weird daemons on the host system :-)
1226326319 M * ard there is no real security advantage, that's true
1226326330 M * ard but if you are running debian/sid :-)
1226326362 M * Bertl Hawq: was that done with -fF ?
1226326370 M * Hawq Bertl: yes
1226326389 M * Bertl because I do not see any thred ids
1226326422 M * Hawq I used 'strace -fF -p 22584 1>x.log 2>&1' to generate it
1226326460 M * Bertl maybe it didn't fork ... could be
1226326501 M * Bertl /dev/tty* exists inside the guest, yes?
1226326507 M * Hawq yes
1226326511 M * Hawq all of them
1226326519 M * Bertl and they are the proper device nodes too, right?
1226326528 M * Hawq yup, copied from host
1226326616 M * Bertl what happens if you switch (from X) with chvt or switchto?
1226326680 M * Hawq just a moment, I'll check
1226326693 M * Bertl maybe have them straced too
1226326730 M * Bertl /dev/ttyv* does exist too, inside the guest? (just making sure here)
1226327114 M * Hawq yes, all /dev/tty* are in guest
1226327209 M * Hawq after executing 'chvt 1' it looked like enter key was pressed and X was restarted.
1226327214 M * Hawq http://paste.linux-vserver.org/12561
1226327222 M * Hawq here is strace from chvt
1226327260 N * Guest2552 Genghis
1226327278 M * Bertl okay, that looks fine and IMHO should have switched correctly
1226327300 N * Genghis Guest2558
1226327304 M * Bertl at least I see no error on the last three ioctls
1226327347 M * Hawq but didn't :( I got back to X login screen. maybe it did switched correctly but immediately switched back because X got restared
1226327373 M * Bertl that could be .. leaves the question why X restarted
1226327397 M * Bertl mayber try to do the chvt via ssh to the guest?
1226327495 Q * rgl Read error: Connection reset by peer
1226327515 J * rgl ~rgl@bl8-131-68.dsl.telepac.pt
1226327552 M * Hawq it switched but same way as with ctrl+alt+f1, X display was left. but at least X wasn't restarted
1226327581 M * Bertl so you still see the X screen, but input goes to the terminal?
1226327611 M * Hawq yeah
1226327635 M * Bertl what kernel version/patches?
1226327693 M * Hawq its PLD kernel, 2.6.25.19. let me check what version of patch this release was using...
1226327796 M * Hawq it was built with patch-2.6.25.15-vs2.3.0.35.diff
1226327833 M * Hawq I may try to boot this machine with 2.6.27.4 using patch-2.6.27.4-vs2.3.0.35.9.diff
1226327848 M * Hawq I mean, vanilla 2.6.27.4 with just vserver patch
1226327890 M * Bertl yep, maybe do that, then we have a common code base
1226328074 M * Bertl CAP_SYS_TTY_CONFIG seems to be the only capability responsible for console switching
1226328159 M * Bertl but it seems to schedule a callback to do the switch
1226328518 M * Bertl I presume the callback does some signaling between kernel and (int this case X) which doesn't get delivered properly, because the X pid is not visible at this point
1226328538 M * Bertl I think we need to enable Linux-VServer debuging to look at the signals there
1226328616 M * Hawq ok. I'll rebuild kernel with debugging enabled
1226328644 M * Bertl keep the kernel tree around, we'll have to test/change a few things, I guess
1226328934 M * Hawq ok. its building now. it will take a while
1226328964 M * Hawq in the meantime I'll try to boot 2.6.27.4 without debugging
1226328965 M * Bertl np, notify me when you're ready
1226330000 J * xdr_ ~xdr@gote2.109.cust.blixtvik.net
1226330111 Q * xdr Ping timeout: 480 seconds
1226330465 Q * mtg Quit: Verlassend
1226330577 Q * mrfree Quit: Leaving
1226330950 N * Guest2558 Genghis
1226330990 N * Genghis Guest2563
1226331221 Q * xdr_ Ping timeout: 480 seconds
1226331895 J * ntrs__ ~ntrs@77.29.22.233
1226332093 J * mib_m03ztr 597b160e@webchat.mibbit.com
1226332113 M * Bertl welcome mib_m03ztr!
1226332128 M * mib_m03ztr :)
1226332141 M * mib_m03ztr hi
1226332171 Q * mib_m03ztr 
1226332198 J * nickk 597b160e@webchat.mibbit.com
1226332213 M * Bertl wb
1226332325 Q * ntrs_ Ping timeout: 480 seconds
1226332536 Q * nickk Quit: http://www.mibbit.com ajax IRC Client
1226333322 M * Hawq Bertl: I booted 2.6.27.5 with newest patch however there is problem. everything works until vprocunhide is executed. after that any access to /proc is denied
1226333359 M * Bertl inside the guest or what?
1226333364 M * Hawq on host
1226333376 M * Bertl that sounds strange, what util-vserver version?
1226333386 M * Hawq inside guest too, just checked
1226333433 M * Hawq util-vserver 0.30.215, but as usual with bunch of patches we use in PLD
1226333445 M * Bertl update to a more recent version
1226333458 M * Bertl http://people.linux-vserver.org/~dhozac/t/uv-testing/
1226333479 M * Hawq latest one will do fine?
1226333483 J * dowdle ~dowdle@scott.coe.montana.edu
1226333484 M * Bertl yep
1226333502 M * Bertl hey dowdle! LTNS!
1226333538 M * dowdle Bertl: Usually Mon-Fri.
1226333918 J * nou Chaton@2001:6f8:328:bbc:6666:6667::
1226334644 N * Guest2563 Genghis
1226334680 N * Genghis Guest2567
1226334724 Q * davidkarban Quit: Ex-Chat
1226334792 J * doener ~doener@i577ADEF2.versanet.de
1226334899 Q * doener_ Ping timeout: 480 seconds
1226335003 J * mtg ~mtg@dialbs-088-079-143-204.static.arcor-ip.net
1226336626 J * hparker ~hparker@2001:470:1f0f:32c:215:f2ff:fe60:79d4
1226336743 Q * hparker 
1226336748 J * hparker ~hparker@2001:470:1f0f:32c:215:f2ff:fe60:79d4
1226337279 Q * mtg Quit: Verlassend
1226337298 M * Hawq Bertl: no change with newest util-vserver
1226337315 M * Hawq # ps ax
1226337317 M * Hawq Error: can not access /proc.
1226337327 M * daniel_hozac Hawq: sounds like you've got a broken vprocunhide configuration.
1226337346 J * bonbons ~bonbons@2001:960:7ab:0:2c0:9fff:fe2d:39d
1226337400 M * Hawq daniel_hozac: but it works on 2.6.25. error appeared after using 2.6.27
1226337476 M * daniel_hozac maybe you had proc security disabled on 2.6.25.
1226337664 M * Hawq daniel_hozac: CONFIG_VSERVER_PROC_SECURE=y on both
1226337687 M * Hawq daniel_hozac: yet 2.6.25 and even 2.6.26 works (just checked 2.6.26).
1226337807 M * Hawq daniel_hozac: any changes are required to vprocunhide setup to get it working with 2.6.27?
1226337811 M * daniel_hozac no.
1226337989 M * Hawq weird then
1226338147 M * Bertl nothing changed there, maybe you have a broken vprocunhide config as daniel_hozac suggested?
1226338199 M * Hawq Bertl: any particular files I should check?
1226338218 M * Hawq Bertl: as current config doesn't work only on 2.6.27
1226338272 M * daniel_hozac so you're running a completely vanilla 2.6.27.5 with the 2.6.27.4-vs2.3.0.35.9 patch?
1226338333 M * daniel_hozac with vanilla util-vserver 0.30.216-pre2724?
1226338334 N * Guest2567 Genghis
1226338345 M * daniel_hozac uh, 2824.
1226338370 N * Genghis Guest2575
1226338433 J * duckx ~Duck@81.57.39.234
1226338647 M * micah i've got one vserver that just refuses to stop
1226338656 M * micah and the stop operation doesn't timeout, it just hangs there
1226338679 M * micah until I issue a control-c, and then it continues... this makes a reboot of the box painful
1226338755 M * daniel_hozac what's it doing?
1226338804 M * Bertl Hawq: check vprocunhide-files
1226338911 J * derjohn_mob ~aj@e180206112.adsl.alicedsl.de
1226338946 M * Hawq daniel_hozac: true except util-vserver is not vanilla, I'll try vanilla one in a moment
1226338993 Q * SpComb Ping timeout: 480 seconds
1226339236 M * micah daniel_hozac: well, good question, at first it was trying to stop apache2, I could control-c that after it was taking too long, then it got hung up on 'Asking all remaining processes to terminate...' which wouldn't respond to control-c, but could suspnd it
1226339267 M * micah trying to do a vps auxw |grep 60 doesn't return
1226339290 M * micah sorry, even without the grep
1226339323 M * daniel_hozac what kernel is that on? it sounds broken.
1226339357 M * micah yeah, this is one of my remaining older 2.6.18 debian kernels
1226339443 M * Hawq Bertl: my vprocunhide-files http://paste.linux-vserver.org/12562 
1226339496 M * Bertl looks good to me ..
1226339506 M * Hawq Bertl: don't look at it with non-working console switching in mind, file got overwritten during util-vserver update
1226339507 M * micah interesting, I can do a vps and see some of the context's processes
1226339525 M * fb micah: i've built most recent util-vserver package available in daniel_hozac home using your debian's .diff; it built without any issues :)
1226339548 M * micah http://paste.linux-vserver.org/12563
1226339558 M * micah fb: yea, there are very few differences
1226339607 M * fb micah: only changes i made were adding changelog entry and removed two rev 2772 patches
1226339625 M * daniel_hozac micah: looks like your ps processes aren't going away like they should.
1226339643 M * daniel_hozac micah: or anything using /proc to list processes, for that matter.
1226339865 M * micah daniel_hozac: the odd thing is, its just this particular guest
1226339908 M * daniel_hozac micah: and vps. so something's wrong with that guest's processes.
1226339922 M * micah i was able to vkill a few of the remaining proceses, but two of them are immune to even a kill -9 (vkill -s 9 -- 32357)
1226339959 M * Bertl try to kill pid 1, 0 and -1
1226339977 M * daniel_hozac sounds like it's stuck in kernel space.
1226339985 M * Bertl and make sure that those processes are not connected to 'outside' processes (like helpers)
1226340084 M * micah Bertl: pid 1, 0 and -1 on the host?
1226340092 M * Bertl with vkill
1226340283 M * micah those pids aren't found
1226340316 M * micah ah, wait, I need to specify the context
1226340322 M * Bertl yep
1226340349 M * micah doesn't seem to kill them (1 is not found)
1226340368 M * micah but 0, -1 receive a kill signal, but stay
1226340705 J * Piet ~piet@86.59.118.153
1226340858 A * micah updates kernel
1226341309 Q * kir Quit: Leaving.
1226342024 N * Guest2575 Genghis
1226342060 N * Genghis Guest2586
1226342206 Q * rgl Ping timeout: 480 seconds
1226342333 M * Hawq just tried with vanilla util-vserver, still no access to proc
1226342361 Q * weasel Ping timeout: 600 seconds
1226342422 M * Bertl Hawq: so, you don't see any files in proc on the host?
1226342433 M * Bertl (and procfs is mounted there ?)
1226342528 Q * gnuk Quit: NoFeature
1226342682 J * weasel ~weasel@anguilla.debian.or.at
1226342691 M * Hawq Bertl: yes, no files, any access to /proc returns permission denied, and yes proc is mounted
1226342714 M * daniel_hozac what do you get in dmesg?
1226342722 Q * weasel 
1226342755 J * weasel ~weasel@anguilla.debian.or.at
1226342756 M * Bertl what does showattr /proc give you?
1226342769 M * Hawq daniel_hozac: good point :) seems problem lies there.
1226342771 M * Hawq daniel_hozac: w[4094]: segfault at c ip b7f36a1c sp bfe72ae0 error 6 in libproc-3.2.7.so[b7f31000+b000]
1226342813 M * Hawq # showattr /proc
1226342816 M * Hawq chdir(): Permission denied
1226342896 M * Hawq daniel_hozac: http://paste.linux-vserver.org/12564
1226342903 M * Hawq daniel_hozac: only this is in dmesg
1226342928 M * Hawq daniel_hozac: I mean, no other vserver related stuff appears there
1226343095 M * Bertl 2.6.27.4-vs2.3.0.35.9 works fine here with 0.30.216-pre2824
1226343107 M * Bertl nothing unusual with proc can be observed
1226343861 M * micah hrm, my new kernel doesn't start up my md devices on boot, leaving me in an initramfs shell, i wonder why not
1226343881 Q * Hawq Ping timeout: 480 seconds
1226343898 M * Bertl micah: check mdadm.conf
1226343922 M * micah Bertl: its identical between the two initrds
1226343935 M * daniel_hozac does the initrd have the requisite modules?
1226343945 M * Bertl dmesg?
1226344001 M * micah the two initrd's seem to contain identical stuff, a diff -r produces no differences except for the module directory name
1226344027 M * Bertl that sounds a little weird, the modules _should_ differ, no?
1226344057 M * Bertl what kernel versions are we talking about, btw, and what's in dmesg?
1226344124 M * micah well, I can't see dmesg, because I dont get a rootfs on boot
1226344133 M * micah i'm trying to go from 2.6.18 to a 2.6.26
1226344137 M * daniel_hozac busybox doesn't have dmesg?
1226344172 M * Bertl 2.6.18 -> 2.6.26 is a huge leap, are you sure your devices didn't change?
1226344176 M * micah ah, it may... I'll reboot to find out
1226344180 M * Bertl e.g. from hda to sda or so?
1226344211 M * micah Bertl: its sda in 2.6.18 and in the busybox shell I saw in /dev sda etc. 
1226344223 M * micah unless in 2.6.26 evreything moved to hda?
1226344245 M * Bertl don't trust dev, it is manually populated in initrd, check with dmesg or /proc/partitions
1226344284 M * micah yeah
1226344317 M * micah i'll reboot to drop into the busybox shell, I went back to 2.6.18 so I could inspect the initrd contents
1226344390 M * micah is there a way to interrupt the timeout that it has waiting for devices and get to the shell faster?
1226344397 M * micah it takes a long time before it drops there
1226344429 M * Bertl if you rewrite the init script, probably yes, but I don't think that is a good idea
1226344430 M * fb Bertl: switching to uuid seems to solve all issues with changing /dev/device names with me :-)
1226344593 M * Bertl fb: hmm? please elaborate ...
1226344696 M * fb Bertl: root=UUID=16502caa-1288-433a-a383-3d7ded5c630b option passed by grub to kernel
1226344724 M * daniel_hozac s/kernel/initrd/
1226344727 M * Bertl ah, well, yeah, works nicely ... unless you clone the partitions :)
1226344754 Q * yarihm Quit: Leaving
1226344758 M * fb daniel_hozac: you don't need initrd
1226344787 M * Bertl fb: unfortunately, nowadays you need initramfs to boot linux
1226344797 M * fb Bertl: i think you can generate new uuid for a partition, at least with xfs :)
1226344844 M * fb Bertl: why you say so?
1226344879 M * Bertl fb: because device nodes need to be created and devfs was abandoned some time ago
1226344916 M * fb Bertl: but you still can use static /dev
1226344942 M * Bertl yeah, just that it is not static anymore, so it wont work in many cases :)
1226344953 M * daniel_hozac e.g. luks, lvm, etc.
1226344961 M * fb but yes, for most flexible use, initrd is best and easiest to maintain
1226345016 M * micah arg, this IRA clears the screen, so its impossible to log output
1226345031 M * Bertl get a serial cable :)
1226345050 M * micah yeah, but its 3253 miles away :P
1226345086 M * micah so /proc/partitions, /dev, /proc/cmdline, /proc/modules all look like they contain the right stuff
1226345109 M * Bertl what about /proc/mdstat ?
1226345123 M * micah only thing missing is /dev/md0
1226345136 M * micah which is required to assemble the device
1226345156 M * micah Bertl: yeah, /proc/mdstat shows nothing assembled, and if I attempt to do a mdadm --assemble --scan, it complains about the missing device file
1226345180 M * Bertl well, looks like your initramfs isn't able to create the device node
1226345199 M * Bertl i.e. update mkinitrd to a more recent version and remake the initramfs
1226345210 M * micah good idea
1226345366 M * micah it looks like those devices get created by mdadm automatically, they are not in the initrd
1226345377 M * micah and I need to pass --auto=yes to mdadm
1226345393 M * micah which is probably not set in my mdadm.conf when I generated the initramfs
1226345698 J * Hawq ~hawk@limanowa.net
1226345698 J * cga ~weechat@94.36.130.212
1226345708 M * Hawq re
1226345713 M * Bertl wb
1226345715 N * Guest2586 Genghis
1226345750 N * Genghis Guest2595
1226345765 M * Hawq my server died :/ isp problems. is there real time log of this channel to see what I lost?
1226345791 M * Bertl yep, http://irc.13thfloor.at/LOG/
1226345899 M * Hawq nice. but I didn't lost much :)
1226345930 M * Hawq now I'm trying manually what vprocunhide startup script does in PLD
1226345986 Q * Pazzo Quit: Ex-Chat
1226346075 M * Hawq setattr -Rx --hide /proc; vprocunhide; is that correct?
1226346167 M * Hawq so, if I'm getting it right, setattr removes access to /proc and vprocunhide should fix it? should setattr remove /proc access for host too?
1226346225 M * daniel_hozac vprocunhide only processes the configuration file entries.
1226346282 J * ntrs_ ~ntrs@77.29.22.233
1226346716 M * Hawq then setattr either works bad at my system (ie. shouldn't affect host) or its working ok on 2.6.27 but wasn't working ok on older kernels
1226346718 Q * ntrs__ Ping timeout: 480 seconds
1226346736 M * Hawq because after setattr I'm loosing /proc access on host
1226346746 M * Bertl could you upload the output of 'vserver-info - SYSINFO'?
1226346796 M * Hawq http://paste.linux-vserver.org/12565
1226346821 J * Hawq_ hawk@furud.net
1226346845 M * Bertl quite a number of APIs you compiled in, daniel_hozac?
1226346853 M * Hawq looks like my server is accessible again :)
1226346896 M * daniel_hozac shouldn't matter...
1226347025 M * Bertl so, next step, Linux-VServer debugging?
1226347052 M * daniel_hozac Hawq: what does showattr -d /proc show?
1226347059 M * daniel_hozac before and after.
1226347293 Q * Hawq Quit: leaving
1226347303 N * Hawq_ Hawq
1226347307 M * Hawq awH-bui- /proc
1226347309 M * Hawq lstat(): No such file or directory
1226347316 M * Hawq thats after
1226347497 M * Hawq awh-bui- /proc
1226347505 M * Hawq thats before
1226348468 Q * weasel Quit: brb
1226348772 J * weasel ~weasel@anguilla.debian.or.at
1226349146 M * Hawq so... anything more I could do/test/check/whatever?
1226349182 M * daniel_hozac your kernel's whack.
1226349186 M * Bertl well, you can upload your kernel config
1226349405 N * Guest2595 Genghis
1226349440 N * Genghis Guest2600
1226349506 M * Hawq http://pld.pastebin.com/f322f097e
1226350032 M * Bertl well, I can build a kernel with your config ... and see if that works
1226350075 M * Bertl (but I guess it will), OTOH, you have Linux-VServer debugging enabled in your kernel, so maybe we should see what calls make your /proc inaccessible
1226350120 M * Hawq it would be weird if it will work. its just vanilla, vserver patch and nothing more.
1226350138 M * Bertl well, no problem with trying it :)
1226350197 M * Hawq if it will work then it means that my util-vserver build is guilty one here...
1226350221 M * Bertl not necessarily, it could be your toolchain as well
1226350237 M * Bertl (even for the kernel compile, but that's unlikely)
1226350303 M * daniel_hozac well, something isn't setting IATTR_PROC_DEFAULT correctly, and vx_hide_check doesn't notice until HIDE is set too.
1226350507 M * Hawq I'll try rebuilding util-vserver on 2.6.27 headers. when 2.6.25 was out such thing + recent util-vserver helped for my problems :)
1226350549 M * daniel_hozac util-vserver doesn't rely on the system's kernel headers
1226350561 M * daniel_hozac for anything of importance, anyway.
1226351089 M * Hawq nothing interesting in dmesg after setattr -Rx --hide (with debugging enabled). just same stuff I pasted before.
1226352056 J * rgl ~rgl@lx2-84-90-232-86.netvisao.pt
1226352979 Q * bonbons Quit: Leaving
1226353095 N * Guest2600 Genghis
1226353117 M * Hawq I just booted 2.6.27.5 built on different version of PLD. it works. and it has dozen more patches. now I'll try to built exactly same kernel on my system. if it won't work, its +1 for toolchain
1226353130 N * Genghis Guest2612
1226353146 M * Bertl build finished here, test booting shortly
1226353323 M * Hawq its not util-vserver fault for sure. 2.6.27.5 that I booted works with old 0.30.215
1226353506 J * ntrs__ ~ntrs@77.29.17.50
1226353932 Q * ntrs_ Ping timeout: 480 seconds
1226354213 Q * cga Quit: WeeChat 0.2.6
1226354412 M * Bertl yep, after some initial trouble with the missing IDE/ATA support :) your config is working fine here
1226354608 M * Hawq then I guess the one that I'm compiling won't work for me.
1226354637 M * Hawq I'll update binutils and see if it helps, if not I'll go for newer gcc
1226354660 J * larsivi ~larsivi@9.80-202-30.nextgentel.com
1226354939 M * Bertl hmm, I remember we had some PLD toolchain issues some time ago (i.e. somebody reported them)
1226354947 M * Bertl maybe google knows more ...
1226355315 M * Hawq when I'll got it working I'll go back to debugging X inside guest + console switching :)
1226355328 M * Bertl good idea :)
1226356012 M * Hawq hm. it didn't even built. gave memory fault when creating bzImage. not nice.
1226356054 M * Hawq anyway, bed time here. I'll upgrade gcc & binutils tomorrow and see what happens
1226356067 M * Hawq thanks and good nite.
1226356293 M * Bertl have a good one!
1226356356 J * marl_ ~marl@78.149.78.173
1226356537 Q * FireEgl Read error: Connection reset by peer
1226356762 J * Aiken ~Aiken@ppp118-208-15-107.lns1.bne1.internode.on.net
1226356785 N * Guest2612 Genghis
1226356820 N * Genghis Guest2619
1226357438 Q * ntrs__ Ping timeout: 480 seconds
1226359565 J * FireEgl Proteus@2001:470:e056:1:4::
1226360475 N * Guest2619 Genghis
1226360510 N * Genghis Guest2624
1226361096 Q * larsivi Ping timeout: 480 seconds