WebHosting Paid by #1Payday.Loans
[00:12] <BlueLaguna> daja77: hmm, that doesn't seem to work, "show the currennt package list" is exactly the same [00:43] BlueLaguna (~BlueLagun@c-67-191-131-28.hsd1.ga.comcast.net) left #rocklinux. [01:36] kasc_ (kasc@dsl-082-083-128-039.arcor-ip.net) joined #rocklinux. [01:47] kasc (kasc@dsl-082-083-058-120.arcor-ip.net) left irc: Read error: 110 (Connection timed out) [01:47] Nick change: kasc_ -> kasc [03:24] <madtux> https://www.jwz.org/images/love_your_job.gif [04:21] mnemoc_ (~amery@200.75.27.111) joined #rocklinux. [04:24] SteffenP (steffen@p549948B2.dip.t-dialin.net) left irc: "https://www.bomberclone.de" [04:25] mnemoc (~amery@200.75.27.12) left irc: Read error: 60 (Operation timed out) [04:49] netrunne1 (~andreas@p54980706.dip0.t-ipconnect.de) joined #rocklinux. [05:06] netrunner (~andreas@p5498078E.dip0.t-ipconnect.de) left irc: Read error: 110 (Connection timed out) [05:06] Nick change: netrunne1 -> netrunner [08:04] <blindcoder> moin [09:14] <madtux> hi blindy [09:14] <blindcoder> moni madtux [09:14] <blindcoder> madtux: got some sleep? [09:15] <madtux> yeah [09:15] <madtux> slept from like 9am till 4pm [09:15] <madtux> so thats something [09:16] <blindcoder> nice :) [09:16] <blindcoder> I'm currently trying to set up udev [09:17] <netrunner> brr, oral exam in 30 minutes :/ [09:22] <madtux> ready for it? [09:23] <netrunner> still reading :) [09:23] <blindcoder> hmm [09:23] <blindcoder> okay [09:23] <blindcoder> there is exactly NIL support for udev in the ROCK startup script [09:23] <netrunner> blindcoder: yes there is. it works out of the box. [09:23] <blindcoder> netrunner: so why can't I find udevd nowhere to be started? [09:23] <netrunner> blindcoder: I use it on 3 machines now :) [09:24] <netrunner> blindcoder: it is started by hotplug [09:24] <blindcoder> ah, okay [09:24] <blindcoder> so how is it put on the initrd? [09:24] <netrunner> can we discuss this in ~1h? :) [09:25] <netrunner> have to examine some petrinets now [09:25] <blindcoder> k [09:28] <blindcoder> grmpf [09:28] <blindcoder> devfs still being mounted [09:29] <blindcoder> udevd not running [09:29] <blindcoder> so much for working udev [09:29] <netrunner> blindcoder: udev would move the mount of devfs to /dev/devfs [09:29] <blindcoder> netrunner: that means it is not started [09:29] <netrunner> blindcoder: have you selected udev? have you selected hotplug? [09:29] <blindcoder> yes ye- [09:29] <blindcoder> 2 <*> Use udev /dev fliesystem. [09:30] <netrunner> hm. you have udev installed? [09:30] <blindcoder> 5 <*> Use hotplug to configure hardware [09:30] <blindcoder> udev 063 0 [09:30] <blindcoder> hotplug 2004_09_23 1 [09:31] <netrunner> well, I did nothing more. (besides the patches to system.init, but that was nothing that kept udev from starting) [09:31] <blindcoder> what patches to system.init? [09:31] <blindcoder> hmm [09:31] <blindcoder> is udevd running on your machine? [09:31] <netrunner> is in submaster, it ran hwscan to load all modules if you selected hotplug. [09:31] <netrunner> yes [09:31] <blindcoder> well, it is not running here :( [09:31] <netrunner> and as I said, it was hotplug who started it for me. [09:33] <netrunner> at least I think so. [09:33] <blindcoder> can you check that? [09:33] <blindcoder> pstree or something? [09:33] <netrunner> not now. [09:33] <netrunner> seems to be from init. [09:35] <blindcoder> starting udevd manually doesn't move devfs anywhere either [09:35] <netrunner> started in /etc/initscript [09:35] <netrunner> when in /etc/conf/devtype:devtype=udev [09:35] <blindcoder> I have no occurance of udev in there :( [09:35] <blindcoder> hmm [09:35] <blindcoder> now this is strange [09:36] <netrunner> case "$devtype" in [...] udev) ... [09:36] <blindcoder> I installed the current sysfiles package an hour ago [09:36] <blindcoder> but it didn't replace /etc/initscript [09:36] <netrunner> find / -name "*.GEMnew" :) [09:37] <blindcoder> I used Emerge-Pkg [09:39] <blindcoder> would a duplicate file error prevent Emerge-Pkg from installing anything at all? [09:39] <netrunner> nah, duplicate files can only be evaluated after installing. [09:40] <blindcoder> hmm [09:40] <blindcoder> even after re-running Emerge-Pkg sysfiles [09:40] <blindcoder> I still have the old /etc/initscript [09:40] <blindcoder> hmm [09:40] <blindcoder> this sucks :( [09:41] <blindcoder> ARGH [09:41] <netrunner> -noupdate don't backup/restore modified package files [09:41] <netrunner> maybe? [09:41] <blindcoder> sysfiles.conf line 32 [09:41] <blindcoder> I'm going to kill someone [09:41] <blindcoder> VERY PAINFULLY [09:42] <netrunner> hehehe [09:42] <netrunner> blindcoder: use .gem's, that's much better than fiddling around with emerge-pkg [09:42] <blindcoder> netrunner: where can I find gems? [09:42] <netrunner> blindcoder: on your build machine :) [09:43] <blindcoder> :P [09:43] <blindcoder> guess I'll just reinstall my laptop [09:43] <netrunner> I can borrow you mine, the sysfiles one should be system-independent [09:43] <blindcoder> nah, I'll just copy the file manually [09:43] <netrunner> 13 minutes to exam [09:43] <th> there is a manual to that file? [09:43] <th> hmm [09:43] <th> ;> [09:48] Action: blindcoder trying again [09:49] <blindcoder> Mounted /dev (udev) from /etc/initscript. [09:50] <blindcoder> w00t [09:56] <blindcoder> now I just need to get my initrd stuff and encryption going for that [10:04] <netrunner> damn, my schedule had an off-by-one. [10:05] <blindcoder> one hour late? [10:18] Action: blindcoder adding a sanity check to his filesystem encryption [10:40] <blindcoder> so, now getting familiar with device-mapper [10:40] <blindcoder> and finally ditching loop-aes [10:43] <th> echo 0 `blockdev --getsize $DEV` crypt aes-plain XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX 0 $DEV 0 | dmsetup create crypted_volume [10:45] <blindcoder> yeah, got that, but dmsetup fails [10:45] <blindcoder> device-mapper ioctl cmd 9 failed: Invalid argument [10:45] <blindcoder> both aes_i586 and dmcrypt modules are loaded [10:47] <th> what's your $DEV [10:47] <th> ? [10:47] <blindcoder> tried with /dev/acer_stick and /dev/discs/disc0/part2 [10:47] <th> hmm [10:47] <blindcoder> hmm [10:48] <blindcoder> Jul 20 10:47:35 localhost kernel: device-mapper: crypt: Error setting key [10:48] <th> perhaps the key needs to be generated somehow? [10:49] <th> propably not [10:49] <blindcoder> good question [10:49] <blindcoder> for testing purposes I tried a key: 12345678901234567890 [10:49] <th> wrong length [10:49] <th> match the length of what i pasted [10:49] <th> XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX [10:49] <blindcoder> min keysize: 16 [10:49] <blindcoder> max keysize: 32 [10:50] <blindcoder> according to /proc/crypto [10:50] <blindcoder> ah [10:50] <blindcoder> forget it [10:50] Action: blindcoder bangs head on table [10:50] <blindcoder> blocksize: 16 [10:50] <th> so it is 16 or 32 [10:50] <blindcoder> so _either_ a 16 byte key _or_ a 32 byte key [10:51] <blindcoder> and suddenly it works :) [10:53] <blindcoder> so I need hashalot, too [10:54] <blindcoder> okay, on-the-fly de/encrypting also works fine :) [10:54] <blindcoder> so that means I can totally dump loop-aes :) [10:57] Action: blindcoder sees the initrd growing beyond the 8MB limit [12:24] SteffenP (~steffen@p5499548A.dip.t-dialin.net) joined #rocklinux. [13:25] <blindcoder> uwah [13:25] <blindcoder> with udev I need 1 MB more space on the initrd [13:39] <blindcoder> okay, first test with dmcrypt [13:47] Action: blindcoder looks at his screen in bewilderment... [13:47] <blindcoder> it worked [13:47] <blindcoder> it freaking worked [13:47] <blindcoder> I now have a dmcrypt encrypted partition on first try [13:48] <blindcoder> but it needs 5.8 MB on the initrd [13:57] <blindcoder> wow [13:57] <blindcoder> it... just worked [13:58] <blindcoder> encrypting [13:58] <blindcoder> booting with encrypted [13:58] <blindcoder> decrypted [13:58] Action: blindcoder is astonished over his own achievement [14:56] <th> oh [14:56] <th> hmmm [14:56] <th> i've machine doing that rescue-lvm-cycle thing which does the same [15:01] <blindcoder> somehow, I fail to parse that sentence [15:03] <th> there is a "a" missing [15:03] <th> s/ / a / [15:04] <th> or in other words: i have a machine which boots via my lvm-cycle thing through the rescue target [15:04] <th> and that machine does dmcrypting in stage1 too [15:04] <blindcoder> ah, now I get it [15:04] <blindcoder> well, I'd like to have a working rootfs encryption directly in rock [15:05] <blindcoder> as well as finally a usable initrd framework ;) [15:05] <blindcoder> and loop-aes will completely vanish with the next LVP update [15:05] <blindcoder> which will dump loop-aes and use dmcrypt, too [15:06] <th> yea that's good [15:07] <blindcoder> I'm now going to write a few words about that into RockDoc so that people can look at how it works [15:19] <blindcoder> what I like most about this is that it's easy to encrypt an existing filesystem :) [15:19] Action: blindcoder is proud of himself :) [15:22] <th> how would you encrypt the existing fs? [15:24] <blindcoder> setup the dmcrypt [15:24] <blindcoder> dd if=/dev/discs/disc0/part3 of=/dev/mapper/disc0_part3_encrypted [15:24] <blindcoder> and done :) [15:24] <blindcoder> hm [15:25] <blindcoder> I just sense an eeeevil bug in the current devicemapper plugin [15:25] <blindcoder> need to address that later [15:26] <blindcoder> I only use the "partx" part of the device to create the device mapper [15:27] <blindcoder> so part3 gets mapped to /dev/mapper/part3_encrypted [15:27] <blindcoder> this will get ugly once you have two disks [15:33] <th> well that method (dd) is not that nice. [15:33] <th> cause you need enough free space [15:33] <blindcoder> ehm [15:33] <blindcoder> if I copy from a harddisk to itself, I have enough free space [15:33] <blindcoder> like this: [15:34] <blindcoder> echo "blablabla /dev/discs/disc0/part3 blabla" | dmsetup crypt part3_encrypted [15:34] <blindcoder> dd if=/dev/discs/disc0/part3 of=/dev/mapper/part3_encrypted bs=1k [15:35] <th> ahh [15:35] <th> on itself [15:35] <blindcoder> yes :) [15:36] <th> you should better not issue the dd twice [15:36] <blindcoder> that way you can encrypt/decrypt existing filesystem [15:36] <blindcoder> hehe, no :) [15:36] <th> well [15:36] <blindcoder> or you would have to issue it twice to decrypt it again ;) [15:36] <th> that would be double encryption [15:36] <th> could propably undo be swapping if and of [15:36] <blindcoder> yeah, but the script doesn't do so [15:37] <blindcoder> at least, if you do it right :) [15:37] <blindcoder> if you instruct it to encrypt the filesystem although it's already encrypted you're screwed [15:37] <blindcoder> but what happens if I decrypt an unencrypted filesystem? [15:37] <blindcoder> would that still count as encryption? [15:37] <blindcoder> </philosophy> [15:37] <th> you could not mount the fs cause it's broken [15:38] <th> the point is: [15:38] <blindcoder> yeah, that much is obvious [15:38] <th> you dont crypt on the FS [15:38] <th> but on the block [15:38] <blindcoder> but would it still be considered ENcryption by law? shouldn't it be DEcryption? [15:38] <blindcoder> yes [15:38] <th> ahhhhhh [15:38] <th> gotcha [15:38] <th> well [15:38] <th> i think it's all the same [15:39] <th> is there any law in germany that would be of interest for us wrt crypting? [15:39] <blindcoder> to ROCK? no [15:39] <blindcoder> to me as a user of encryption? yes [15:39] <th> what laws are that? [15:39] <th> i mean: we're not in france here [15:39] <blindcoder> because I'll be travelling a bit to countries with evil import restrictions [15:40] <th> ahh k [15:42] <blindcoder> okay, information added to rockdoc [15:42] <blindcoder> still need some more [15:43] <blindcoder> but I need the laptop for that [15:43] <blindcoder> and it's still busy encrypting /home :) [16:08] <blindcoder> done [16:17] <blindcoder> now [16:17] <blindcoder> this sucks [16:20] <blindcoder> devfs still got mounted at boot and udevstart doesn't do anything [17:42] <blindcoder> okay, here is my official statement [17:42] <blindcoder> UDEV SUCKS [17:42] <daja77> hm you have sysfs mounted? [17:43] <blindcoder> yes [17:43] <blindcoder> it is up and running [17:43] <blindcoder> but don't ask... [17:43] <blindcoder> udevstart follows different rules than udevd [17:43] Action: daja77 asking blindcoder [17:43] <blindcoder> my soundcard isn't recognized anymore [17:43] <blindcoder> and all this userspace crap is taking up a hell of a lot of space in the initrd [17:44] <blindcoder> then /dev/random and /dev/urandom are created with root.root and 660 [17:44] <_Ragnar_> post it to the kernel devel list [17:44] <blindcoder> making it impossible to use ssh from a user account [17:44] <blindcoder> _Ragnar_: when I calm down I will [17:45] <daja77> sounds like stupid defaults which can be changed ... [17:45] <blindcoder> can? MUST! [17:45] <blindcoder> oh, and udevstart always return 0 with no messages [17:45] <daja77> I meant we are able to do that ... [17:45] <blindcoder> even if it fails at everything it tries to do [17:46] <blindcoder> but I'd really like to have my soundcard back now [17:47] <daja77> modprobe [17:47] Action: daja77 hides [17:47] <blindcoder> the modules are loaded [17:47] <blindcoder> snd_intel8x0 33568 0 [17:49] <daja77> hm which app can't access? [17:49] <blindcoder> any [17:49] <blindcoder> aplay [17:49] <blindcoder> xmms [17:49] <blindcoder> mplayer [17:49] <blindcoder> /dev/sound non-existant even with oss compatibility layer [17:49] <blindcoder> /dev/dsp also nonexistant [17:50] <blindcoder> alsactl also can't do shit [17:50] <_Ragnar_> sounds like udev is a really big step back from devfs [17:51] <blindcoder> HELL, YEAH! [17:51] <blindcoder> devfs worked [17:51] <blindcoder> udev doesn't [17:51] <blindcoder> BIG step backward [17:51] <daja77> sounds like the devices are not created for whatever reason [17:51] <esden> does anyone know how far xmms2 is? [17:51] <blindcoder> Jul 20 17:49:14 localhost kernel: ACPI: PCI interrupt 0000:00:1f.5[B] -> GSI 10 (level, low) -> IRQ 10 [17:51] <blindcoder> looks fine to me [17:52] <blindcoder> Jul 20 17:49:14 localhost kernel: intel8x0_measure_ac97_clock: measured 49475 usecs [17:52] <blindcoder> Jul 20 17:49:14 localhost kernel: intel8x0: clocking to 48000 [17:52] <blindcoder> just the fucking piece of crap that is udev doesn't get them [17:54] <blindcoder> argh [17:54] <blindcoder> udevd isn't started [17:54] <blindcoder> what is this piece of crap? [17:55] <daja77> sounds like the reason for your problem ... [17:56] <blindcoder> yeah, but why? [17:56] <blindcoder> shouldn't it be started by hotplug [17:57] <daja77> what happens if you start it manually [17:58] <blindcoder> then it works [17:58] <blindcoder> according to andreasj from #linux.de udevd is NOT started by hotplug [17:58] <blindcoder> *sigh* another thing to add to the initrd [17:59] <th> well hotplug wouldnt start devfsd for you, would it? [17:59] <clifford> blindcoder: in the real system udevstart is executed by /etc/initscript (fyi) [17:59] <blindcoder> clifford: yes, but that neither helps me in initrd nor does it start udevd [17:59] <blindcoder> clifford: udevstart doesn't care about /etc/udev/rules.d/ [18:02] <blindcoder> that's why I only get /dev/hda[1234] and no /dev/discs/... [18:04] <blindcoder> reboot again... [18:10] <blindcoder> argh [18:10] <blindcoder> I was missing the /bin/sh -> bash symlink >_< [18:11] <daja77> you should calm down and relax ... [18:11] <blindcoder> yeah, really... [18:12] <blindcoder> sooooo [18:12] <blindcoder> how can I tell hotplug to always load certain modules? [18:12] <blindcoder> namely snd-pcm-oss snd-seq-oss and snd-mixer-oss? [18:13] <daja77> .oO (that's a huge sm sync) [18:28] <blindcoder> okay, that works [18:28] <blindcoder> now for udev permissions [18:29] <blindcoder> hmm [18:29] <blindcoder> permissions.d [18:29] <blindcoder> is _totally_ ignored by udev [18:33] <blindcoder> HAH [18:33] <blindcoder> udev 063 handles permissions differently from 050 [18:53] <netrunner> blindcoder: use my patch from submaster and write it to /etc/conf/kernel [18:56] <blindcoder> netrunner: id? [18:57] <netrunner> 2005071120083612746 [19:03] <blindcoder> ah [19:03] <blindcoder> well, I've configured udev correcly now, that solves the permissions problem [19:03] <blindcoder> see 2005072018472004222 [19:03] <blindcoder> and the modules have been solved with modules.conf though I probably won't remember this after the next reinstallation [20:08] steffen_ (~steffen@fortuna.bingo-ev.de) joined #rocklinux. [20:47] <daja77> 43 confilcts ^^ [20:48] <blindcoder> ouch [20:49] blindcod1r (~blindcode@dsl-082-082-102-183.arcor-ip.net) joined #rocklinux. [20:49] blindcoder (~blindcode@dsl-082-082-097-087.arcor-ip.net) left irc: Nick collision from services. [20:52] Nick change: blindcod1r -> blindcoder [20:53] blindcoder (~blindcode@dsl-082-082-102-183.arcor-ip.net) left irc: Nick collision from services. [20:53] blindcoder (~blindcode@dsl-082-082-102-183.arcor-ip.net) joined #rocklinux. [21:05] <daja77> everything resolved ;) [21:23] h-k (~horrorkid@p549A7327.dip.t-dialin.net) joined #rocklinux. [21:41] h-k_ (~horrorkid@p549A47FC.dip.t-dialin.net) left irc: Read error: 110 (Connection timed out) [22:29] steffen_ (~steffen@fortuna.bingo-ev.de) left irc: Remote closed the connection [23:18] SteffenP (~steffen@p5499548A.dip.t-dialin.net) left irc: "https://www.bomberclone.de" [23:18] SteffenP (steffen@p5499548A.dip.t-dialin.net) joined #rocklinux. [23:45] <daja77> clifford: jiba just told me, that he'll release a new version of pyopenal soon, with my patch included ;) [00:00] --- Thu Jul 21 2005