--- Day changed

--- Log opened Fri May 30 00:00:03 2003
00:22 -!- pirlouit [~peter@64.162.195.202] has quit [Remote host closed the connection]
00:22 -!- pirlouit [~peter@64.162.195.202] has joined #uml
00:33 -!- pirlouit [~peter@64.162.195.202] has quit [Remote host closed the connection]
00:35 -!- pirlouit [~peter@64.162.195.202] has joined #uml
03:02 -!- lemon [~lemon@137.132.81.172] has joined #uml
03:04 -!- lemon_ [~lemon@137.132.81.172] has joined #uml
03:11 -!- lemon_ [~lemon@137.132.81.172] has quit [Quit: Client exiting]
03:11 < lemon> quit
03:12 -!- lemon [~lemon@137.132.81.172] has quit [Quit: Client exiting]
03:12 -!- lemon [~lemon@beijing.ddns.comp.nus.edu.sg] has joined #uml
03:23 < lemon> Hi, I'd like to ask a question regarding "redirecting uml boot message to a file".
03:23 < lemon> i use 'fork' to create the uml.
03:24 < lemon> before 'system(linux, ubd0=root_fs...)' is called. i remaped stdin/stdout to file.
03:25 < lemon> but i only saw part of the boot mesg appearing in the file, say 'boot_mesg'.
03:26 < lemon> the boot message stops at ''NET4: Unix domain sockets 1.0/SMP for Linux NET4.0.
03:27 < green> everything else gets into con0 and so it depends on where do you redirect that
03:27 < lemon> the reason why i need a boot mesg mapped to a file is that i want to check the boot mesg in my perl script.
03:28 < lemon> i start the uml with option 'con0=fd:0,fd:1 con=pts'
03:29 < green> do you have FD_CHAN compiled in?
03:30 < lemon> yes.
03:31 < lemon> i reopen stdin to /dev/null and stdout to 'boot_mesg' in my perl script before call 'system(linux ...)'
03:31 < green> then you should see the rest of boot messages, I think
03:32 < lemon> i guess so. but it surprised me that i can't see the second half of the boot mesg.
03:33 < green> well, if you run uml manually, do you see it?
03:33 < lemon> when i run it manually, i can see the full mesg on my terminal
03:34 -!- snide [~snide@AMontsouris-108-1-8-3.w217-128.abo.wanadoo.fr] has joined #uml
03:37 < green> lemon: hm. does it looks like buffering issue then?
03:37 < green> (i.e. try to disable buffering on the fd before running uml)
03:37 < lemon> good idea. i am tring it now. 
03:48 < lemon> be back soon
03:51 -!- lemon [~lemon@beijing.ddns.comp.nus.edu.sg] has quit [Quit: Client exiting]
04:31 -!- ElectricElf [david@elf.noc.oftc.net] has quit [Quit: Reboot]
04:46 -!- ElectricElf [david@elf.noc.oftc.net] has joined #uml
05:23 -!- lemon [~lemon@beijing.ddns.comp.nus.edu.sg] has joined #uml
05:25 -!- Lathiat [lathiat@2001:388:7094:4080:2e0:29ff:fe07:1eb2] has joined #uml
05:27 -!- mace [mace@bub.ipv6.darksilence.net] has joined #uml
05:27 -!- lemon [~lemon@beijing.ddns.comp.nus.edu.sg] has quit [Quit: Leaving]
05:28 -!- lemon [~lemon@137.132.81.172] has joined #uml
06:03 < lemon> green: hi, did this in my script: $|=1; open(STDIN, "< /dev/null"); open(STDOUT, "> boot_mesg"); system("linux", "con0=fd:0,fd:1", con=pts);...
06:03 < lemon> still, it did not work. : (
06:04 < lemon> of cause those statement is within the 'forked' process.
06:15 < green> I am not sure this will work, perhaps you need to do this with the file descriptor itself
06:18 < lemon> thanks. checking it now. 
06:36 < lemon> green: hi, i've checked the file descriptor of STDIN, STDOUT and STDERR. they are 0, 1 and 2 respectively.
06:37 < green> sure ;)
06:37 < green> but you need to disable buffering somehow. Otherwise when it see that fd is just some file and not char device, it usually buffers in 4k chunks
06:39 < lemon> besides 'fileno', is there any other subroutine i can use to control the buffering of file descriptor in perl?
06:39 < lemon> is it possible that there is some buffering control inside uml which i should indeed disable?
06:40 < green> well may be so, too.
06:42 < lemon> inside uml, con0 == /dev/console, right? 
06:43 < green> yes
06:43 < lemon> so if do "echo 'hello' > /dev/console" as root in uml, it 'should' goes to the console.
06:43 < green> yes
06:44 < green> this is only if you have correc node for /dev/console of course
06:44 < green> (there was a practice when /dev/console is a symlink to /dev/tty0)
06:45 < lemon> i see.
06:50 < snide> green: it could be a symlink to /dev/vc/0, no ?
06:51 < lemon> just checked. /dev/console is a char device with majoy 5, minor 1, belongs to group tty.
06:52 < snide> green: hmm... actually, i'm wrong, iirc /dev/vc/0 is the currently displayed vc [ vc/1, vc/2, etc ]
06:52 < lemon> inside uml
06:52 < green> snide: /dev/ttys/0 i think
06:52 < green> lemon: so try to output lots of stuff there and see if something comes out to your script/file
06:53 < lemon> good idea. so i am forcing the buffer to flush. ; )
06:53 < green> kind of
06:58 < lemon> no luck : (
06:58 < lemon> perhaps the data is not large enough.
06:59 < lemon> green: thanks for your help. i think i will continue trying tomorrow. see you.
06:59 < green> ok
06:59 < lemon> time to leave the library. : )
06:59 < lemon> bye
07:00 < green> bye
07:00 -!- lemon [~lemon@137.132.81.172] has quit [Quit: Leaving]
09:24 -!- ichilton [~ian@pc3-stoc3-4-cust203.midd.cable.ntl.com] has quit [Read error: Connection reset by peer]
09:24 -!- ichilton [~ian@80.6.255.203] has joined #uml
09:38 < Lathiat> im being most evil
09:38 < Lathiat> i have a uml setup
09:38 < Lathiat> and im runnign virtual servers (chroots plus extra ock downs) inside it
09:39 < Lathiat> how evil ;)
09:39 < green> hehe.
09:40 < Lathiat> its quite cool this virtual server stuff
09:41 < Lathiat> it adds liike IP locking and stuff
09:41 < david> green: here /dev/vc/0 and /dev/console have different major/minors
09:41 < Lathiat> like freebsd jails
09:41 < david> Lathiat: did you patch your kernels so people can root you via them?
09:43 < BB2> hmm
09:43 < BB2> do they have an BFUB modeling section ?
09:43 < green> david: sure
09:43 < BB2> (re uml mailing list incase i've totaly confused everytone)
09:46 < Lathiat> david; well im hoping that
09:46 < Lathiat> they cant get outside the uml
09:46 < Lathiat> right?
09:46 < Lathiat> heh
09:46 < Lathiat> so even if they get out of the magic chroot
09:47 < green> they only need to poke in /dev/mem to get out ;) or to load some nasty kernel module ;)
09:47 < Lathiat> dev/ram doesnt exist ;)
09:47 < Lathiat> norcan they mknod
09:48 < green> even if they are root? ;)
09:48 < Lathiat> yep
09:49 < green> even when they are out of magic chroot?
09:49 < Lathiat> uh once theyre out of the chroot then oh well
09:49 < Lathiat> atm the users have access to the box anyway
09:49 < Lathiat> im locking them away as a security measure
09:50  * green does similar thing
10:02 -!- ichilton [~ian@80.6.255.203] has quit [Ping timeout: 488 seconds]
10:04 -!- ichilton [~ian@pc3-stoc3-4-cust203.midd.cable.ntl.com] has joined #uml
10:19 < Lathiat> hey
10:19 < Lathiat> if i turn on tty logging
10:19 < Lathiat> how do watch them?
10:26 < Lathiat> hrm
10:27 < Lathiat> getting this virtual serve patch to work with uml
10:27 < Lathiat> is gona require hacking
10:27 < Lathiat> all the stuff is in arch/i386
10:27 < Lathiat> mm fun
10:29 < caker> morning
10:30 < Lathiat> moaning
10:30 < Lathiat> afternon here
10:30 < Lathiat> well night actually
10:31  * Lathiat is gonna learn to hack the kernel 
10:31 < caker> Lathiat: does the vserver-in-uml look doable?
10:31 < Lathiat> dear god, what im i getting myself into
10:31 < Lathiat> caker: well the problem is, msot of the extra syscalls etc arre in arch/i386
10:31 < Lathiat> so i guess stuff neds to be in arch/um
10:32 < Lathiat> so umm it shouldnt be _that_ hard i dont think
10:34 < Lathiat> but itd probably be good to get a local copy
10:34 < Lathiat> hacking in vim over an ssh from a dialup is laggy and all bad
10:41 -!- Maniac [~daManiac@209.5.247.105] has joined #uml
10:53 < Lathiat> Microsoft is a long way from its ultimate goal where users can take security for granted in its products...the majority of viruses written attack Microsoft products.'"
10:53 < Lathiat> lol like every virus targets microsoft products except like 1
10:58 -!- amrut [~amrut@203.124.158.218] has quit [Quit: Client Exiting]
11:30 -!- dilinger [irc@207.99.115.73] has quit [Quit: Terminated with extreme prejudice - dircproxy 1.0.5]
11:30 -!- dilinger [irc@207.99.115.73] has joined #uml
11:33 -!- dilinger [irc@207.99.115.73] has quit [Client Quit]
11:33 -!- dilinger [irc@sloth.voxel.net] has joined #uml
11:43 -!- Ryback_ [~ra993482@143.106.24.143] has joined #uml
11:51 < snide> btw, is someone aware of a 'hanging' bug in uml_switch + 2.5.69-1um ?
12:18 -!- snide [~snide@AMontsouris-108-1-8-3.w217-128.abo.wanadoo.fr] has quit [Quit: going back home...]
13:18 -!- thy_ [koubt9am@u-pl7.ms.mff.cuni.cz] has joined #uml
13:27 -!- Newbee [~user01@141.165.40.68] has joined #uml
13:29 < Newbee> Just built 5.69 but UML exits wo starting console, what are correct cmd ln opts?
13:30 < caker> con=null con0=fd:0,fd:1 ?
13:33 < Newbee> same output, uml exits after Initializing stdio console driver
13:33 < caker> exits, or hangs?
13:33 < Newbee> exits
13:33 < caker> whats your command line?
13:34 < Newbee> ./linux con=null con0=fd:0,fd:1
13:34 < caker> heh
13:36 < caker> not sure, .. it isn't supposed to do that<tm> .. I'm not very familar with 2.5, but I don't think it would be much different.
13:36 < caker> exiting without giving you any indication of what went wrong is very strange
13:36 < caker> hanging is a different story
13:37 < Newbee> hang on, i had the wrong sym link path for root_fs, i'll fix and try again
13:38 < caker> ln -s brain Newbee
13:38 < caker> there
13:38 < caker> :-p
13:40 -!- |thy| [koubt9am@u-pl7.ms.mff.cuni.cz] has joined #uml
13:40 -!- |thy| [koubt9am@u-pl7.ms.mff.cuni.cz] has left #uml []
13:52 < Newbee> no same problem, tried booting 4.20 on same root_fs, it works fine
13:53 < caker> did you compile in console support into your uml kernel?
13:55 < Newbee> tt=y skas=y stdio_console=y con_zero_chan=fd:0,fd:1 con_chan=xterm fd/null/port/pty/tty/xterm not set
13:55 < caker> there's the problem
13:55 < caker> turn all that stuff on (fd/null,etc..)
13:56 < Newbee> what?  I hit ENTER on all default options in make config. will edit .config directly and try again.
14:00 < Lathiat> hey if i have tty logging on - how do i watch the tty?
14:01 < caker> screen?
14:01 < caker> tail?
14:01 < caker> where does it log to?
14:01 < Lathiat> u can just do that?
14:02 < Lathiat> i dunno
14:02 < Lathiat> thats why im asking ;)
14:02 < Lathiat> its that magic option in UML mostly usefull for honeypotting etc
14:02 < caker> don't you have to tell it where to log?
14:06 < Lathiat> i dunno i havent seen any docs on it
14:17 < Newbee> 5.69 boot gets past Init console driver to VFS: Cannot open root device "6200" or unknown block 98,0
14:17 < Newbee> am using root_fs_toms1.7.205 which works with 4.20
14:17 < Newbee> ls
15:08 -!- Lathiat [lathiat@2001:388:7094:4080:2e0:29ff:fe07:1eb2] has quit [Quit: Lost terminal]
15:09 -!- Lathiat [lathiat@seven.sixlabs.org] has joined #uml
15:18 -!- Newbee [~user01@141.165.40.68] has quit [Quit: Client Exiting]
15:18 -!- thy_ [koubt9am@u-pl7.ms.mff.cuni.cz] has quit [Read error: Connection reset by peer]
15:36 -!- karm [karm@conurb.net] has quit [Ping timeout: 492 seconds]
17:29 -!- Maniac [~daManiac@209.5.247.105] has left #uml [Client exiting]
18:10 -!- Ryback_ [~ra993482@143.106.24.143] has quit [Quit: Client Exiting]
20:53 < david> hello
20:59 < caker> hello
22:34 -!- ElectricElf [david@elf.noc.oftc.net] has quit [Quit: Reboot]
22:44 -!- ElectricElf [david@elf.noc.oftc.net] has joined #uml
--- Log closed Sat May 31 00:00:00 2003