--- Day changed

--- Log opened Thu Jun 19 00:00:02 2003
00:00 -!- tarcan [~tarcan@200.97.184.70] has left #uml [Finalizando Cliente]
03:23 -!- desaster [~desaster@80.222.254.34] has quit [Quit: There are 3 kinds of people: those who can count & those who can't.]
03:25 -!- mistral [mistral@212.159.71.212] has quit [Ping timeout: 488 seconds]
03:32 -!- desaster [~desaster@dsl-jklgw1o22.dial.inet.fi] has joined #uml
03:44 -!- mistral [mistral@212.159.71.212] has joined #uml
04:29 -!- pflanze [~chris@cc-linux4.ethz.ch] has joined #uml
04:44 -!- ido [~Ido@81.218.140.160] has joined #uml
04:44 < ido> hey people ;p
04:46 < ido> can anyone help me getting started with uml ?
05:21 -!- desaster [~desaster@dsl-jklgw1o22.dial.inet.fi] has quit [Remote host closed the connection]
05:29 -!- desaster [~desaster@80.222.254.34] has joined #uml
06:29 -!- ido [~Ido@81.218.140.160] has quit [Ping timeout: 488 seconds]
06:40 -!- yeti [~yeti@pD950B002.dip.t-dialin.net] has joined #uml
06:45 -!- _slh__ [~chatzilla@68.17.240.64] has joined #uml
06:48 < yeti> i think i dont understand the bridge module good enough up to now... my /var/log grows incredibly... how can i reduce the amount the bridge logs?
06:54 < green> redirect the logs to /dev/null ;)
06:55 < yeti> i only want to reduce the bridge's verbosity...
06:55  * yeti thinks...
06:55  * yeti . o O ( use the source, luke )
06:55 < desaster> if you use something like logcheck then you can filter those off
06:55 < desaster> and syslog-ng can filter them before they even hit the log files
06:56 < yeti> but that still would eat cpu cycles for logging and filtering... first i try to find a way to tell the bridge to be more quiet
06:57 < yeti> i have too many tasks at once... bridge is new to my... i try to polish my uml environment... i try to install uos in uml...
06:57 < yeti> hmmm...
06:57 < yeti> clone me!
07:00 < yeti> i am rsyncing a cdimage from one k6-2/500 to an other system...
07:01 < yeti> syslogd and klogd eat 33% uf the cpu power!!!
07:01 < yeti> i dont think filtering the logs is the answer...
07:01 < yeti> there must be a way to log less
07:01 < yeti> or bridge is useless to me
07:02 < yeti> net performance drops to the speeds of a 10mbps lan
07:02 < yeti> :-(
08:24 < yeti> thing i found it: netfilter debugging :-)   disabled it and it seems that does the trich... now testing on 2 more systems...
08:24 < yeti> this will need a reboot of this box too... cya later...
08:25 -!- yeti [~yeti@pD950B002.dip.t-dialin.net] has quit [Quit: Client exiting]
09:04 -!- litost [~mondain@phynp6.phy-astr.gsu.edu] has joined #uml
09:21 -!- jdike [~jdike@129.170.249.105] has joined #uml
09:22 < jdike> hi guys
09:22 < desaster> jeff jeff
09:22 < jdike> desaster disaster
09:22  * jdike looks for green
09:22 < desaster> he was last alive 13:53
09:22 < desaster> erm
09:22 < desaster> now the time is 16:22
09:24 < jdike> hmmmph
09:25 < ichilton> Hi Jeff
09:26 < jdike> Hi Ian
09:33 < Lathiat> Heyas
09:33 < Lathiat> jdike: soo, coming to LCA2k4?
09:33 < Lathiat> cfp just came out a few days ago
09:34 < jdike> Lathiat: Jeez, haven't even thought about it
09:34 < jdike> Lathiat: Where is it?  Melbourne?
09:34 < Lathiat> jdike: adelaide
09:34 < jdike> Lathiat: Ah
09:35 < Lathiat> im running the ipv6 mini-conf, you can come and i can enslave more peopel into v6 *hehe*
09:35 < jdike> Lathiat: LCA is going to have to start repeating cities, .au is running short :-)
09:35 < Lathiat> jdike: quite a few to go yet :)
09:35 < Lathiat> jdike: i figure itl be back in perth in 5-10 years
09:35 < jdike> Lathiat: Melbourne and Canberra
09:35 < jdike> Lathiat: and people might dispute Canberra
09:36 < Lathiat> and hobart in theory, but dont think there is much of a userbase there
09:36 < jdike> Hmmm, Hobart would be interesting
09:36 < Lathiat> but who knows, met a guy in my year 11 class who runs mandrake on his laptop the other day, never would have guessed.
09:38 < Lathiat> ive got me a sexy new server, not as sexy as everyoen else tho
09:38 < Lathiat>  Hostname: nine.sixlabs.org - OS: FreeBSD 4.8-RELEASE/i386 - CPU: 2 x Pentium Pro (199.43-MHz 686-class CPU) - Processes: 27 - Uptime: 56 Minutes - Load Average: 0.00 - Memory Usage: 6.89mb/317.10mb (2.17%) - Disk Usage: 0.59gb/0.88gb (67.05%) - Network Traffic (tl0): In/Out 516.0mb/521.1mb
09:38 < Lathiat> ~
09:38 < Lathiat> soon to have debian on it of course :P
09:38 < ljlane> Why are you switching?
09:40 < Lathiat> i just put freebsd on here for fun to try it out
09:40 < ljlane> ah
09:40 < Lathiat> but debian is my preferred server environment
09:40 < Lathiat> also, i can run uml :)
09:41 < jdike> Lathiat: port UML to it then :-)
09:42 < Lathiat> wonder how much effort itd take heh
09:42 < Lathiat> more ffort than i have time to do :P
09:53 < Lathiat> jdike: so, gonna think about it? :)
09:54  * jdike will think about it
09:55 < jdike> Lathiat: when is it, exactly?
09:56 < Lathiat> january 2k4
09:56 < Lathiat> exact day, see http://conf.sixlabs.org
09:56 < Lathiat> ugh
09:56 < Lathiat> i mean http://conf.linux.org.au
09:56 < Lathiat> the former is for the ipv6 conf hehe
10:08 < jdike> Has anyone got the 2.5.70 xconfig to work?
10:11 < Lathiat> havent tried
10:11 < Lathiat> might try oit later
10:11 < Lathiat> jdike: whats the problem, doesnt start or crashes in the middle of it?
10:12 < jdike> scripts/kconfig/qconf.o(.text+0x35): undefined reference to `QApplication::translate(char const *, char const *, char const *, QApplication::Encoding) const'
10:12 < jdike> A zillion of those when linking qconf
10:13 < jdike> Ah, figured it out
10:15 < Lathiat> xconfig uses Qt?
10:15 < jdike> Lathiat: yup
10:16 < Lathiat> wicked, Qt owns :P
10:16 < Lathiat> puity it has a crap license
10:16 < jdike> Lathiat: nicer than the old tk interface actually
10:16 < jdike> Lathiat: just way too big a window
10:16 < Lathiat> heh
10:16 < Lathiat> so they changed in 2.5?
10:16 < Lathiat> ill 'ave to try it
10:17 < jdike> anyway the problem was I was accidentally using 2.95.3 gcc to compile and 3.3 gcc libs to link
10:18 < Lathiat> haha
10:18 < jdike> Surprisingly, that doesn't work well
10:21  * jdike builds a 2.5.70 UML for the first time
10:23 < jdike> Then comes .71 and .72
10:24 < jdike> and then Linus will ignore me some more
10:26 < Lathiat> you get that eh?>
10:28 < jdike> Lathiat: I'll check with the gf and see how much she fancies a trip to .au :-)
10:30 < Lathiat> jdike: okies :)
10:30 < Lathiat> jdike: gonna do a paper, or just drag alogn with the rest of the corwd?
10:31 < jdike> Lathiat: Well, a paper might mean a free trip :-)
10:35 < Lathiat> indeed
10:35 < Lathiat> does for me :P
10:35 < jdike> Lathiat: where are you, BTW?
10:37 < Lathiat> perth, australia
10:37 < Lathiat> was an org for last years lca
10:53 -!- bshyu [~bshyu@113.c210-58-173.ethome.net.tw] has joined #uml
10:58 < green> jdike: push the stuff to Andrew Morton! ;)
10:59 < jdike> green: hehe
10:59 < jdike> green: I have a question for you
11:00 < jdike> green: what's the proper way to send IO to a device?
11:00 < green> hmm, what do you mean?
11:00  * green actuall had a question for Jeff, too. Only I had it two days ago and forgot about it already
11:00 < jdike> green: I have a bunch of data in kernel memory (COW bitmap) and I want to have a device write it out
11:01  * jdike is making a COW driver
11:01 < jdike> which layers above two arbitrary other block devices
11:02 < jdike> green: I guess I need to get some buffer_headers and send them to ll_rw_blk
11:02 < green> hm, you mean ll_rw_blk is not suitable for you?
11:02 < jdike> green: I think it is, what's the usual way of getting the buffer_heads?
11:03 < green> I think that drivers/block/loop.c have all the questions solved already.
11:03 < jdike> green: ah, didn't think to look there
11:03 < jdike> green: Also, I have ordering constraints which loop doesn't have
11:03 < green> buffers can be created with e.g. create_empty_buffers
11:04 < green> create_empty_buffers( page, blocksize, 0) (I do not remember what 0 is for, right now though
11:04 < jdike> green: the backing data has to be written before the bitmap on the COW device
11:04 < green> so you send backing data first, then wait on it, then send bitmap stuff
11:05 < jdike> green: yeah, I was wondering if there was a way to tell the block layer that, to avoid actually waiting
11:06 < green> no, elevator can reorder stuff at will
11:06 < jdike> green: except I don't have a struct page handy
11:06 < green> at least until this new "elevator barriers" code is merged in mainline
11:06 < jdike> green: and this stuff isn't going through the page cache
11:06 < green> what do you have then?
11:06 < jdike> green: a vmalloced buffer
11:07 < green> hm. But why? I thought that the best way would be to hack on loop.c to understand cow files, why do you think this is bad idea?
11:08 < jdike> green: just that's the way it's already done
11:08 < green> and you want to do it differently?
11:09 < jdike> green: the bitmaps can be large, so I'd prefer not to try to put them in kmalloc/gfp memory
11:09 < jdike> green: bitmap size == device size / 4K (with 512 byte sectors)
11:10 < green> jdike: Ah. Well, I think you cannot write vmalloked stuff directly, you'd need to do getblk()/memcpy()/ll_rw_blk it seems
11:11 < jdike> green: YUCK
11:25 -!- pflanze [~chris@cc-linux4.ethz.ch] has quit [Ping timeout: 496 seconds]
11:26 -!- smcavoy [~smcavoy@grimlock.drive-megawheels.net] has joined #uml
11:28 < green> jdike: btw, have you figured out how to mage gdb happy wrt stacktraces in uml by any chance?
11:31 -!- _slh__ [~chatzilla@68.17.240.64] has quit [Ping timeout: 488 seconds]
11:36 < jdike> green: no, haven't looked at that yet
11:38 < jdike> 2.5.70 builds
11:46 < green> check if it halts correctly, too
11:47 < jdike> green: first things first
11:47 < jdike> green: it hangs on boot
11:47 < green> hehe. strange
11:47 < green> works for me like a charm ;)
11:48 < jdike> green: looks like the ubd rounding bug
11:48 < jdike> green: now I have an excuse to fix it :-)
11:49 < green> ;)
12:08 -!- ido [~Ido@bzq-218-140-160.cablep.bezeqint.net] has joined #uml
12:10 < ido> anyone here ?
12:11 < alt> hi
12:11 < alt> sorta
12:11 < jdike> hmmm, hangs in the same place
12:16 < alt> ubd rounding bug?
12:16 < jdike> no, it's something else
12:18 < alt> ah
12:27 < ido> hrmf
12:27 < ido> could anyone please help me getting started with uml ?
12:31 < jdike> hmmph
12:31 < jdike> rm is segfaulting and the fault isn't being fixed
12:33 < ido> hmmpf
12:34 < ido> hey jdike
12:37 < alt> ido: what do you need to know?
12:38 < ido> welp
12:38 < ido> first thing, was that i had trouble patching the 2.4.20 kernel
12:38 < ido> actually patching worked ok, but it gave me an error while compiling it
12:38 < ido> so i moved to using a prebuilt 2.4.20 kernel
12:39 < ido> anyhow, as for my first question, is if i could run gentoo over uml ?
12:39 < alt> is this the host kernel or the UML kernel?
12:39 < ido> i've already installed gentoo on one of the boxes here, and wondered if it could be moved onto a uml host..
12:39 < alt> oh, I'm sure you can
12:40 < alt> I've just never done that :\
12:40 < alt> I've done everything except install my own distro
12:40 < ido> would would be the steps in doing that ?
12:41 < alt> hmm... I know there's a howto on that
12:41 < alt> somewhere
12:41 < ido> eg, creating a new virtual 'filesystem', booting a prebuilt kernel, mk'ing the virtual partitions, and then moving the data from the current box to the new one over the network ?
12:41 < ido> got any idea where ?
12:41 < alt> http://user-mode-linux.sourceforge.net/UserModeLinux-HOWTO.html
12:41 < alt> that's a great one to start with
12:41 < alt> but doesn't address installing your own distro
12:41 < jdike> ido: you can use the existing install, except you need to change device names in fstab and inittab
12:41 < alt> http://user-mode-linux.sourceforge.net/slack_readme.html <-- that's for slackware
12:42 < alt> might help 
12:42 < ido> and how would i import the existing install partition onto the virtual uml disk ?
12:42 < alt> cp -a
12:43 < ido> hmm. hrmf.
12:43 < jdike> alt: dd
12:43 < ido> lemme try installing it
12:43 < alt> jdike: that too
12:44 < alt> jdike: I keep a base install of Debian in a reference directory, then I make a new "drive", mount it and do "cp -a" from the reference to the new drive
12:44 < alt> works great
12:45 < jdike> alt: yeah, although I generally prefer dd for that sort of thing
12:46 < alt> personal preference I guess 
12:46 < alt> :)
12:46  * alt loves linux 
12:56 < ido> damnit
12:56 < ido> does anyone have cloop ?
12:56 < ido> cant' seem to find it anywhere (searching like 20 minutes on google for it)
13:00 < jdike> ido: mount -o loop isn't what you want?
13:05 < ido> truth is i'm working with a howto
13:12 -!- jdike [~jdike@129.170.249.105] has quit [Ping timeout: 488 seconds]
13:13 < ido> whats 'mount none /mnt/loop -t hostfs -o /path/to/gentoo-working-dir' ? -> mount none -t hostfs ?!
13:14 < ido> how to i make hostfs ?
13:21 < ido> whats the config filename thats stores the 'make config' configuration in the kernel sources ?
13:21 -!- jdike [~jdike@dhcp05.ISTS.dartmouth.edu] has joined #uml
13:21 < jdike> aha
13:21 < jdike> it's not really hung
13:22 < jdike> just taking forever clearing /tmp because I dropped a huge directory structure there earlier
13:25 < jdike> Linux usermode 2.5.70-1um #5 SMP Thu Jun 19 13:01:04 EDT 2003 i686 unknown
13:31  * jdike does the bk dance
13:32 < ido> hrmf
13:32 < ido> how do i make hostfs.o ?
13:32 < ljlane> ido: .config
13:32 < ido> ... ?
13:32 < ido> cant' seem to find it
13:32 < jdike> enable CONFIG_HOSTFS
13:32 < ido> in kernel 2.4.20 
13:32 < jdike> ido: what's the real problem?
13:32 < ido> just add CONFIG_HOSTFS=y ?
13:33 < jdike> ido: no
13:33 < ido> then.. ?
13:33 < ido> going through a gentoo over UML howto
13:33 < ido> http://www.omote.com/gentoo/install-in-uml.html
13:35 < BB> ah now i know why you were looking for cloop
13:36 < ido> aha
13:36 < BB> those instructions are a tad old and weird if you ask me
13:36 < ido> hrmf
13:36 < ido> they seems weird, and quite outdated
13:37 < ido> actually most of the stuff i've seen on UML seems quite outdated
13:37 < ido> thats why i'm having such trouble installing it
13:38 < ido> could you point me to anything newer ?
13:38 < BB> i might have a smallish gentoo rootfs somewhere if you're able to download it
13:38 < ido> sure, that'd be helpfull
13:39 < BB> I'l see if i can find which server i've put it on
13:40 < ido> great.
13:40 < ido> got a howto/manual that comes with it?
13:40 < ido> or any other type of up-to-date instructions ?
13:40 < BB> nope, i installed it into a partition and copied it off
13:41 < ido> oh, thats fine by me
13:42 < BB> i've found it, i'l just tidy it up alittle first ;)
13:43 < ido> :)
13:44 < ido> tidy up what exactly ?
13:44 < jdike> ido: remove the pr0n
13:44 < BB> lol
13:44 < BB> its running uk2.php.net atm, so i need to basicaly remove the whole php mirror :P
13:45 < ido> eh
13:45 < ido> cool.
13:45 < ido> you've got each website running on a different uml ?
13:46 < BB> well i have a raq550 running on one doing most vhost stuff, then i put the php one on its own as it gets a fair few hits
13:46 < ido> eh.
13:51 < ido> what exeactly does the cloop extractor do ?
13:56 < ido> cloop-0.66/extract_compressed_fs livecd.cloop > livecd
13:56 < ido> what would this do ?
13:56 < BB> convert the compressed image into an uncompressed one
13:57 < ido> could it be done otherwise ?
13:57 < BB> no idea
13:57 < ido> since i dont have cloop.
13:57 < ido> hrmf.
13:57 < BB> i gave in as i couldnt find cloop
13:57 < ido> diddo.
13:57 < ido> is the gentoo root_fs ready ?
13:57 < jdike> what does 'file livecd.cloop' say?
13:57 < BB> just testing it
13:58 < ido> oh, ok.
13:58 < alt> http://lazyeyez.homelinux.net/cola/movies/mac.wmv
14:00 < BB> lol boat anchor
14:01 < alt> hee hee
14:01 < alt> my cable modem is busy :)
14:01 < BB> I think this needs to be sent to every Mac user ;)
14:01 < alt> oh yes :)
14:02 < alt> hmm.. I'm not sure if that's a neat feature
14:02 < alt> but if you put the cursor over an mp3 in Konqueror, it plays the file
14:03 < ido> fuck its late already
14:09 < ido> arg.
14:27 < BB> hmm ok its larger that i expected it to be
14:27 < BB> http://rain.max-hosting.net/~chris/gentoo.test.bz2 its 532MB
14:28 < Pahan> Heh.
14:28 < Pahan> cloop is a stupid compressed filesystem created specifically for Knoppix, AFAIK.
14:28 < Pahan> It only supports one image mount per system.
14:28 < Pahan> And it can't be created on the fly, AFAIK.
14:29 < Pahan> That is, you need to form a full image to create a compressed version of it.
14:31 < ido> welp, wow
14:31 < ido> huge image
14:31 < ido> what does it include ?
14:32 < BB> erm basicaly an install of gentoo, with php and a few other things
14:32 < ido> hrmf
14:32 < ido> basic install ? or is it partly configured already ?
14:33 < BB> i've uninstalled things i've altered the config files for
14:33 < BB> so it has no apache/htdig
14:34 < ido> no problem
14:35 < ido> if i run a couple of UML's on a host, and they're nost running any pograms (except default background deamons), would it consume a lot of my host's resources
14:36 < Pahan> In skas mode, no.
14:36 < Pahan> That is "not at all"
14:37 < Pahan> Not more than a minute of CPU time per day.
14:38 -!- pflanze [~chris@217.162.113.252] has joined #uml
14:45 -!- Ndiin [~eandres@bosporos.mars.asu.edu] has joined #uml
14:52 -!- Ndiin [~eandres@bosporos.mars.asu.edu] has quit [Quit: raven[0m[36m:[0m .]
14:56 < ido> how do i set skas mode ?
14:57 < alt> patch your host kernel
14:57 < ido> ah
14:57 < alt> the UML will detect the patch
14:57 < ido> against what ?
14:57 < alt> and use it 
14:57 < ido> where do i get it? 
14:57 < alt> the patch or the kernel?
14:57 < Pahan> Use google or something.
14:57 < ido> patch ofcourse ;p
14:58 < alt> the patch is at user-mode-linux.sourceforge.net I believe
14:58 < alt> the kernel can be had from www.kernel.org
14:58 < Pahan> You patch kernel with skas, turn "/proc/mm" in config and boom.
14:58 < ido> anyhow, to copy an existing partition to with with a prebuilt kernel, i could dd if=/dev/hda1 of=ptn1
15:00 < ido> afterwards, while booted with a prebuilt kernel with a basic temporary partition, i dd if=/mnt/host/ptn1 of=/dev/ubd1
15:00 < ido> right ?
15:01 < ido> i need the partitions to be the exact same size though, right?
15:27  * BB would go for the mount+tar option
15:27 < Pahan> ido: All relevant filesystems can be resized.
15:28 < ido> which mount+tar option ?
15:28 < ido> ah, eh, can't quite do that
15:28 < ido> since we're talking on taking the original distro from another computer..
15:29 < BB> well why do you need to do it within the uml, why not mkfs the new one, mount them both and copy over the data is what i mean
15:30 < ido> hrmf
15:30 < ido> i could do that
15:30 < ido> but it'd have to install networking first
15:30 < ido> besides, this sounds faster
15:30 < ido> and would be an exact copy
15:34 < BB> why do you need to copy ptn1 to the ubd1 device anyway tho? just use the ptn1 file or a copy/COW ofit
15:34 -!- Ndiin [~eandres@bosporos.mars.asu.edu] has joined #uml
15:35 < Ndiin> Hey, I don't suppose there are any devs hanging around? I'm tried to get AFS's kernel module to compile for UML.. =)
15:35 < Pahan> AFS sucks, though :P
15:37 < Ndiin> Perhaps, but it's needed ;)
15:37 < jdike> Ndiin: Where is the __verify_write coming from?
15:38 < Ndiin> I found that one, thats from the asm-i386/uaccess.h, in the access_ok define, so I simply used the first defined version of it
15:38 < Ndiin> All thats left now is the boot_cpu_data
15:39 < jdike> Ndiin: Something was very wrong if you were getting stuff from asm-i386
15:39 < jdike> Ndiin: boot_cpu_data just might not be exported
15:40 < Ndiin> AFS is using the i386.. it can't tell the difference between the host, mostly because it doesn't know about arch=um. I'm simply trying to get it working with the least fuss.. although it likely will have problems, still
15:40 < Ndiin> [root@bosporos bin]# nm linux |grep boot_cpu_
15:40 < Ndiin> a01ec450 D boot_cpu_data
15:40  * Ndiin shrugs
15:42 < Ndiin> Doesn't appear to be anyone else that's gotten AFS working under UML, from what I've found on google though.. all I've found is people asking questions about the first few problems I came accross, and one suggesting the proper solution (making afs properly support uml), but no followups or fixes
15:43 < Ndiin> I don't know enough about arch specific code other than enough to be dangerous, heh.
15:44  * jdike ponders the 2.5.71 and .72 diffs
15:45 < ido> hrmf
15:45 < ido> after i've dd'd /dev/hda1 to a file
15:45 < ido> isn't mount -o loop -t reiserfs filename /mnt   supposed to mount it ?
15:46 < BB> should do
15:46 < Ndiin> jdike: btw, were you able to reproduce those IO problems using mem=2048M at all?
15:46 < ido> getting an error saying failed: Invalid argument
15:47 < jdike> Ndiin: didn't try
15:47 < jdike> Ndiin: I don't doubt that I can, just need to decide to sit down and fix it
15:48 < Ndiin> Just curious.. that little bug drove me crazy for a few days ;)
15:51 < ido> hrmf
15:51 < ido> could anyone help me with the mount thingy ?
15:52 < BB> i take it, it was reiser
15:53 < ido> aha
15:53 < ido> another question
15:54 < jdike> ido: you can leave off the -t anyway
15:55 < ido> how come when ./linux and the root_fs reside on the root partition, they work fine, and then they're run from a partition mounted to /whatever, they dont, and i get a Kernel panic: VFS: Unable to mount root fs on 62:00 error message (although "root=root_fs" has been set)
15:56 < jdike> ido: what did it say before that?
15:56 < ido> ah, seems it works when runnning as 'linux root_fs', although when i run it from anydir which is on the root partition, it works without that
15:57 < ido> seems like a bug if you ask me
15:58 < BB> depends where and what you ran i guess
16:07 < Ndiin> hmm, odd.. /proc/ksyms in the UML doesnt have boot_cpu_data, but the system map does
16:08 < jdike> Ndiin: You did read what I said about exporting it?
16:09 < Ndiin> er, no, must have missed that
16:09 < Ndiin> ah, hmm
16:14  * Ndiin recompiles
16:29 < Ndiin> haha, that was fun.. I did expect it, though.
16:30 < Ndiin> Thats what the problem was for the unresolved symbol, thanks.
16:30 < Ndiin> What I expected, did happen though: *boom* ;)
16:30 < Ndiin> So much for the cheap way out.
16:31 -!- sssssssh_____ [~chatzilla@adsl-19-240-82.jax.bellsouth.net] has joined #uml
17:09 -!- yeti [~yeti@pD950B002.dip.t-dialin.net] has joined #uml
17:22 -!- smcavoy [~smcavoy@grimlock.drive-megawheels.net] has quit [Remote host closed the connection]
17:23 -!- litost [~mondain@phynp6.phy-astr.gsu.edu] has quit [Quit: Run Away!! Run Awaayyyy...]
17:47 -!- jdike [~jdike@dhcp05.ISTS.dartmouth.edu] has quit [Quit: Leaving]
18:12 -!- ASY- [~someone@66.36.133.68] has joined #uml
18:12 -!- ASY [~someone@66.36.133.68] has quit [Read error: Connection reset by peer]
18:12 -!- ASY- is now known as ASY
18:42 -!- war- [war@lucidpixels.com] has joined #uml
19:08 -!- Dorian [~Dorian@host81-132-149-57.in-addr.btopenworld.com] has joined #uml
19:08 -!- Dorian [~Dorian@host81-132-149-57.in-addr.btopenworld.com] has left #uml []
19:17 < Ndiin> Has anyone ever tried doing some really heavy processing / IO in a UML? I'm trying some reads/converts/writes of 44mb, 120mb, and 740mb files..
19:18 < Ndiin> the 44 and 120 went through fine, the 44 being very close to the host in processing time, the 120 about half as fast
19:18 < Ndiin> but the 740 seems to be hanging, although the UML is still responsive
19:19 < yeti> i just jad a hang in an "apt-get upgrade"
19:19 < Ndiin> its using like 80% cpu on the host, but it thinks it's using 12.5% cpu in the UML
19:19 < yeti> i already killed that process :-()
19:19 < Pahan> Ndiin: skas mode?
19:20 < Ndiin> no, tt
19:20 < Pahan> Heh heh.
19:20 < ido> trying to complie a 2.4.21 kernal, patched against http://kernels.usermodelinux.org/patches/uml-patch-2.4.21.bz2
19:20 < Ndiin> That significant a difference, Pahan?
19:20 < ido> and when make linux ARCH=um
19:21 < ido> it dies near the end
19:21 < ido> with a : /usr/lib/libc.a: could not read symbols: Malformed archive
19:21 < ido> collect2: ld returned 1 exit status
19:21 < ido> make: *** [linux] Error 1
19:22  * yeti switches to the vt... will be kinda away again
19:22 < ido> anyone encountered that before ?
19:31 -!- Ndiin [~eandres@bosporos.mars.asu.edu] has quit [Quit: raven[0m[36m:[0m Going home.]
19:51 -!- bshyu [~bshyu@113.c210-58-173.ethome.net.tw] has quit [Read error: Connection reset by peer]
20:38 -!- yeti [~yeti@pD950B002.dip.t-dialin.net] has quit [Quit: Client exiting]
20:43 -!- sssssssh_____ [~chatzilla@adsl-19-240-82.jax.bellsouth.net] has quit [Ping timeout: 490 seconds]
21:13 -!- Synchro [~dark@earth.noc.teilam.gr] has joined #uml
21:13 -!- Synchro [~dark@earth.noc.teilam.gr] has left #uml []
23:06 -!- war- [war@lucidpixels.com] has quit []
23:17 -!- alt is now known as aLt
23:17 -!- aLt is now known as AlT
23:17 -!- AlT is now known as aLT
23:18 -!- aLT is now known as alt
23:49 -!- mistral [mistral@212.159.71.212] has quit [Ping timeout: 488 seconds]
23:50 -!- mistral [mistral@212.159.71.212] has joined #uml
--- Log closed Fri Jun 20 00:00:00 2003