--- Day changed

--- Log opened Tue Jul 15 00:00:02 2003
02:17 -!- amnesia is now known as amn_reads_kernel_bible
02:45 -!- amn_reads_kernel_bible is now known as amn_away
04:23 -!- anklt [~mangled@pD952ABD8.dip.t-dialin.net] has joined #uml
04:23 < anklt> Salut...
04:23 < anklt> Anyone listening?
04:32 < green> yup
04:32 < anklt> Oh, nice.
04:32 < anklt> I ran into building trouble..
04:32 < anklt> The problem seems that the linker finds several symbols redefined, even with different sizes.
04:33 < green> what are the symbols? what kernel version what uml patch?
04:33 < anklt> (From objects which export them in fact, but it's not redefined in the source code.)
04:33 < anklt> Ok, Linux 2.4.20-6uml and 2.4.20-5uml under Linux 2.4.22-pre2, 2.5.72 and 2.6.0-test1.
04:34 < anklt> Symbols are the stuff you get out of object files using nm
04:35 < anklt> However, the result looks like this:
04:35 < anklt> /home/thunder/src/kernel/linux-2.4.20/lib/string.c:280: multiple definition of `strpbrk'
04:35 < anklt> arch/um/kernel/tt/unmap_fin.o(.text+0x36560): first defined here
04:35 < anklt> and stuff.
04:35 < anklt> /usr/local/bin/ld: Warning: size of symbol `strpbrk' changed from 179 to 157 in vmlinux.o
04:36 < green> ah, so glibc and kernel symbols collide.
04:36 < anklt> Hm, looks like that..
04:36 -!- dfri [~dfri@194.204.113.2] has joined #uml
04:36 < dfri> hi there
04:36 < anklt> Salut.
04:37 < dfri> i am trying to set up my netowrk in UML but there seams to be a progb
04:37 < green> anklt: in fact it is strange that you do not have __HAVE_ARCH_STRPBRK defined
04:38 < dfri> in the  howto it says that enable Netweork device support. Cannot find this :S
04:38 < anklt> green: Where do you think it ought to be defined?
04:38 < dfri> any idea where i can find this option?
04:38 < anklt> dfri: Type slowly. Leave you some time.
04:38 < green> anklt: somewhere in include/asm
04:38 < dfri> k
04:40 < anklt> I only see it undefined for s390 and m68k..
04:40 < anklt> I grepped for it, but I couldn't find any #define. Juust #undef.
04:40 < anklt> s/uu/u/
04:42 < green> hm.
04:43 < anklt> dfri: Network device support is under Network device support in the main menu...
04:43 < green> looks like so. Though it builds for me which is strange
04:43 < anklt> dfri: (If you enabled "Networking support" under General setup.)
04:43 < anklt> green: Might that be a compiler problem?
04:44 < green> should not be. What compiler do you use?
04:45 < anklt> I tried gcc 3.3.1 yet..
04:47 < anklt> I first thought of the binutils, but neither 2.13.2 nor 2.14.90.0.4 were able to build it. (That was before you made me think of glibc.)
04:47 < green> ah, hm. never tried that gcc 3.3, 3.2.3 and 2.95 worked fine for me. May be this is some funny optimisation logic that repleced one functions with other
04:48 < dfri> in the main menu there is no Network device support but it is Network device i suppose it is that Right?
04:48 < anklt> Yes, I think.
04:48 < dfri> ok then
04:50 < dfri> the prob is that when i execute eth0=ethertap,tap0,,192.168.0.254 nothing is happening
04:50 < dfri> ifconfig only shows the loopback
04:52 < dfri> any idea what am i missing or doing wrong?
04:52 < anklt> Hm, ifconfig tap0 up maybe?
04:53 < dfri> no such dev
04:53 < anklt> Then it might be that it's not compiled into your UM?
04:54 < anklt> L
04:54 < dfri> could it be the rrot_fs prb?
04:56 < anklt> Ça veut dire quoi?
04:56 < anklt> (What does that mean?)
04:57 < anklt> green: I'm just compiling the whole thing using GCC 2.95.4...
04:59 < green> that should work, I guess ;)
04:59 < anklt> The thing is that everything else I built using gcc 3.3.1 works just perfectly...
04:59 < anklt> (That is, my whole system)
05:02 < anklt> BTW, some days ago I successfully booted Linux 1.2.16 on a modern machine. :)
05:03 < green> lkml have some reports about gcc 3.3.1 miscompiling 2.6.0-test1
05:04 < anklt> My Linux 2.6.0-test1 was compiled using gcc 3.3.1...
05:04 < anklt> And it's running. :)
05:05 < anklt> The only kernel I just can't get to compile using gcc 3.3.1 is 2.0.40-rc6.
05:06 < anklt> (I use gcc 2.7.2.1 for that. Maybe that one will compile UML too? *g*)
05:07 < anklt> dfri: Can you specify a little more how you think it's a "root fs" problem?
05:08 < dfri> It was a question, as the config seams to be right
05:08 < anklt> So you think it's a problem with your root fs?
05:09 < dfri> nope, as i tryed a diff root fs
05:09 < anklt> Then I don't understand your suggestion.
05:09 < dfri> how can i be sure that the seeting in the config are taking effect?
05:09 < anklt> Hmm, did you enable "Ethertap support"?
05:10 < dfri> yes
05:10 < anklt> dfri: By remaking the dependencies... If you want to ensure that *everything* is recompiled in correct order, make mrproper.
05:10 < dfri> Ethertap, tun/tap are both on
05:11 < dfri> hmm good idea let me do it
05:11 < anklt> Ahm BTW, is this server IPv6 accessible?
05:11 < dfri> it's off
05:12 < anklt> I meant this IRC server.
05:12 < anklt> IPv6 shouldn't matter for your UML.
05:12 < dfri> sorry no idea
05:12 < dfri> :)
05:12 < anklt> (As long as the host supports IPv4, though.)
05:22 < dfri> ARG.. recompiled but stil did not work
05:22 < dfri> i am using kernel 2.4.20 any idea what can i try?
05:23 < anklt> Some different networking method?
05:24 < dfri> need to set up an eth
05:30 < anklt> Damn, it doesn't link either....
05:30 < anklt> green: didn't work.
05:33 < dfri> I am running uml not as a root does that create a prob in setting the eth ?
05:44 -!- dfri [~dfri@194.204.113.2] has quit [Quit: Client Exiting]
05:45 < green> anklt: same problem?
05:45 < anklt> green: yes..
05:46 < green> anklt: have you done "make ARCH=um distclean" prior to retry?
05:46 < anklt> I did mrproper.
05:47 < anklt> Should suffice, shouldn't it? :)
05:47 < green> with ARCH=um ?
05:47 < anklt> Yes.
05:47 < green> yes, this is the same thing
05:47 < anklt> Just a bit more destruction. :)
05:48 < green> well, quick qorkaround seems to be to disable offending functions in lib/string.c
05:49 < anklt> Hm, shouldn't the rest of the kernel miss them?
05:49 < green> try it and you will see. After all kernel will link with glibc's functions with the same prototypes. 
05:49 < anklt> (Sorry I'm so slow in responding, but I'm just setting up Linux 2.6.0 on an UltraSPARC.)
05:50 < anklt> Hm, isn't it compiled using -nostdinc -nostdlib ?
05:51 < green> certainly not -nostdlib, uml is linked against glibc
05:51 < green> s/glibc/libc on your system which is most probably glibc/ ;)
05:51 < anklt> Hm, but only parts of it?
05:51 < anklt> Well, most of my systems are using libc5, actually. :)
05:51 < green> no. it links against whoel thing. That's why you see symbols clash
05:51 < anklt> (Though this one is not.)
05:53 < anklt> I don't find sprintf in there...
05:53 < anklt> I guess I'll have to disable half of the kernel.
05:54 < green> so you have a clash with sprintf too?
05:54 < anklt> Yes.
05:54 < green> hm, weird.
05:55 < anklt> It's a lot of stuff...
05:56 < green>  something is terribly broken as it builds for lots of people
05:56 < anklt> Hmm.
05:56 < anklt> gcc can't be the problem, I suppose, as 2.95.4 fails as well.
05:57 < anklt> The source itself is vanilla + patch. Should be OK, shouldn't it?
06:00 < green> yes it should
06:01 < anklt> libc is glibc 2.3.1-17 ...
06:02 < anklt> binutils are 2.14.90.0.4, but 2.13.2 didn't work either...
06:02 < anklt> I'm not sure if they can be the problem... I guess not?
06:06 < green> I use: binutils 2.14.90.0.4, gcc 3.3 and glibc 3.2.1 and it works, worked with earlier versions too
06:07 < anklt> Hum.
06:31 < anklt> Current drops: sscanf, sprintf, __mon_yday....
06:32 < anklt> What if I inline these functions?
06:35 < anklt> Only  __mon_yday left..
06:36 < anklt> With a little problem..
06:36 < anklt> How to find it?!
06:44 -!- amn_away [eugeneteo@tacheon.cc.uic.edu] has left #uml []
07:41 -!- ticallion [~ticallion@213.175.160.198] has joined #uml
07:41 < ticallion> hi guys
07:42 < ticallion> anybody tried applying the skas patch to 2.4.20-18.8 (red hat errata kernel)?!
07:46 -!- ticallion [~ticallion@213.175.160.198] has quit [Quit: Client exiting]
08:04 -!- dfri [~dfri@194.204.113.2] has joined #uml
08:05 < dfri> hi guys, i managed to set the network from the host to the UML. I need to ping other computers on the host's network any idea how i can do it?
08:05 < anklt> You can ping the host?
08:05 < dfri> yes
08:06 < anklt> Then you just have to set it as the default gateway, and set up IP forwarding on that host.
08:06 < anklt> (And define a route to the client through the host on all other hosts who need to access it -- or use NAT...)
08:07 < dfri> ok. that is the main problem.. the default gateway, as ipfowarding is on
08:07 < anklt> Now just get the packets a way back!
08:07 < anklt> The other hosts are looking for your host on the local net, where it isn't.
08:08 < anklt> You have to route it back *through* the host to your UML.
08:08 < dfri> can you give me a hand on the default gateway pls
08:08 < anklt> I think the default gateway is just fine!
08:09 < anklt> It's just the routing into the other direction which doesn't work.
08:09 < anklt> You know, the UML is so to say in another subnet.à
08:09 < anklt> s/à//
08:09 < anklt> But I must leave now... Sorry...
08:10 < dfri> thanks just the same.,
08:10 < anklt> Set a host route on all other hosts.
08:10 < anklt> # route add -host $(UMLIP) gw $(HOSTIP)
08:10  * anklt <- gone.
08:38 -!- dfri [~dfri@194.204.113.2] has left #uml [Client Exiting]
09:05 < anklt> I did a #define __mon_yday __k_mon_yday in kernel.h... Works...
09:05 < green> heh
09:06 < anklt> (In addition to inlining sprintf and sscanf.)
09:06 < anklt> Do you want a patch? :)
09:08 < green> no, I want to understand why I do not see this problem locally first ;)
09:08 < anklt> You can look at the patch in order to find out. :)
09:09 < anklt> Hm, "Profiling timer expired."...
09:10 < green> hm...
09:10 < green> why haev you enabled gcov support after all? ;)
09:10 < anklt> Where did I enable that beast? *g*
09:11 < anklt>        SIGPROF     27,27,29     A      Profiling timer expired
09:11 < anklt> That one.
09:12 < anklt> Where do I find the GCOV support optin?
09:12 < anklt> option...
09:13 < green> in gernel debugging section
09:13 < green> "GCOV support"
09:13 < anklt> Whoops, I see it.
09:13 < anklt> Thanks!
09:13 < anklt> I usually enable everything under "Kernel hacking" :)
09:13 < anklt> Anything else to switch off there?
09:14 < anklt> ptrace? gprof?
09:16 < green> gprof is not needed
09:16 < green> ptrace is if you plan to use gdb on tt mode
09:16 < anklt> Sometimes. :)
09:16 < green> then let it on ;)
09:17 < anklt> Merci :)
09:51 -!- Jax [~computer_@148.100.215.25] has joined #uml
09:57 -!- mdz_ [~mdz@h00034795fb2d.ne.client2.attbi.com] has joined #uml
10:33 < anklt> The UML is booting well.
11:36 -!- AquaJo [~torbofh@pD95592F0.dip.t-dialin.net] has joined #uml
11:44 -!- revenger3 [revenger@p508B2A62.dip.t-dialin.net] has quit [Ping timeout: 501 seconds]
11:46 -!- revenger [revenger@80.139.36.155] has joined #uml
12:12 -!- anklt [~mangled@pD952ABD8.dip.t-dialin.net] has quit [Quit: Je retournerai bientôt..]
12:58 -!- ticallion [~ticallion@213.175.160.169] has joined #uml
13:00 < ticallion> anybody tried out 2.6-test1?!
13:04 < green> yup
13:04 < ticallion> green: fails when I'm building the damn thing
13:04  * ticallion thinks it's got to do with gcc4
13:04 < ticallion> I mean
13:04 < ticallion> 3
13:04 < ticallion> which gcc u got rigged?!
13:05 < green> gcc 3.3
13:06 < ticallion> grrr, I'm running 3.2, lemme check what the error is maybe C gurus abound around
13:08 < green> gcc 3.2.2-5 from RH 9 by any chance?
13:08 < ticallion> nope... RHL 8.0
13:09 < ticallion> but i think they're the same cept for threads
13:09 < ticallion> cos mine's the errata dist
13:09 < green> they are not same
13:10 < ticallion> yeah well, my C mojo barely exists
13:11 < ticallion> it fails in arch/um/util
13:11 < ticallion> lots of errors
13:11 < ticallion> undeclared crapolas and missing semicolons and misc. mayhem of the sort
13:11 < ticallion> unfortunately kernel.org's bugzilla only covers 2.5
13:12 < green> you need to patch in the uml patch before it will compile. there is outdated uml in there
13:12 < ticallion> I'm about to check the mailing list
13:12 < green> unfortunatelly no such patch ixists in the wild.
13:12 < ticallion> green: oh... that's cool... thanks for the tip will try
13:12 < ticallion> grrrrr
13:12 < green> I routinuely maintain my own uml tree that is up to date with Linus' kernel for example
13:13 < ticallion> green: lin is all about sharing :P
13:14 < green> sure. ;)
13:14 < green> I store it in bk://namesys.com/bk/reiser4-linux-2.5 only I have some unrelated reiser4 bits there too
13:19 -!- AquaJo [~torbofh@pD95592F0.dip.t-dialin.net] has quit [Quit: Ich komme wieder, sicher?!]
13:22 -!- ticallion [~ticallion@213.175.160.169] has quit [Ping timeout: 488 seconds]
13:28 -!- stratsi [~stratsi@80.72.68.187] has quit [Ping timeout: 490 seconds]
14:39 -!- glommer [~glauber@200-158-193-117.dsl.telesp.net.br] has joined #uml
14:40 < glommer> Hi Pals.... I'm trying to run UML, but I got a problem. I would appreciate if someone here could help me
14:40 < glommer> I did exactly as the instructions in website
14:40 < glommer> and got:
14:40 < glommer> VFS: Cannot open root device "ubd0" or 62:00
14:40 < glommer> Please append a correct "root=" boot option
14:40 < glommer> Kernel panic: VFS: Unable to mount root fs on 62:00
14:40 < caker> glommer: what's your startup command line?
14:41 < glommer> udb0=root_fs.md-8.2-server.pristine.20020324
14:41 < glommer> this root_fs was downloaded from the uml site too
14:41 < caker> what are it's permissions?
14:41 < glommer> I am root
14:41 < caker> try adding "devfs=nomount"
14:42 < glommer> got the same thing
14:42 < glommer> But... what bothers me most, is this line: Please append a correct "root=" boot option
14:43 < caker> a few lines up from that, is there anything else in the output?
14:43 < glommer> sure
14:43 < glommer> let me paste it here
14:43 < glommer> Initializing software serial port version 1
14:43 < glommer> mconsole (version 2) initialized on /root/.uml/pKV0dY/mconsole
14:43 < glommer> unable to open root_fs for validation
14:43 < glommer> UML Audio Relay
14:43 < glommer> Initializing stdio console driver
14:43 < glommer> VFS: Cannot open root device "ubd0" or 62:00
14:43 < glommer> Please append a correct "root=" boot option
14:43 < glommer> Kernel panic: VFS: Unable to mount root fs on 62:00
14:44 < caker> udb should be ubd
14:44  * BB does that lots
14:44 < glommer> shit
14:44 < caker> yes
14:44 < BB> that too
14:44 < glommer> such a silly thing
14:44 < glommer> heheh
14:44 < glommer> :)
14:45 < glommer> but thanks anyway.. :)
14:46 < caker> no prob
15:00 -!- ticallion [~ticallion@213.175.160.79] has joined #uml
15:01 < ticallion> green: 404s galore on that url u gave me
15:01 < ticallion> sorry I disappeared, power outage
15:23 -!- stratsi [~stratsi@80.72.68.187] has joined #uml
15:30 -!- ticallion [~ticallion@213.175.160.79] has quit [Quit: Client exiting]
16:02 < green> ticallion: well, you should use bitkeeper to pull from there ;)
16:02 < green> oops
16:29 -!- Jax [~computer_@148.100.215.25] has quit [Quit: gone home]
17:00 < QSECOFR> will root hostfs work with a cow?
17:02 < BB> mooo
17:02 < caker> ummm huh?
17:02 < caker> cow as another ubd device?
17:18 < QSECOFR> can i overlay a cow on a hostfs mount?
17:18 < caker> no
17:19 < QSECOFR> k
17:19 < QSECOFR> thanks.
17:51 -!- Dorian [~dorian@host81-132-236-196.in-addr.btopenworld.com] has joined #uml
18:16 -!- iStormy [~stormy@209.205.169.19] has joined #uml
18:35 -!- Dorian [~dorian@host81-132-236-196.in-addr.btopenworld.com] has quit [Quit: Download Gaim [http://gaim.sourceforge.net/]]
21:07 -!- jdike [~jdike@h69-21-248-50.69-21.unk.tds.net] has joined #uml
21:07 < jdike> hi guys
--- Log closed Tue Jul 15 21:40:12 2003

--- Log opened Tue Jul 15 21:40:26 2003
--- Log closed Tue Jul 15 21:46:26 2003

--- Log opened Tue Jul 15 21:46:38 2003
21:49 -!- jdike [~jdike@h69-21-248-50.69-21.unk.tds.net] has quit [Quit: Leaving]
22:14 -!- glommer [~glauber@200-158-193-117.dsl.telesp.net.br] has left #uml [Client exiting]
23:38 -!- ticallion [~ticallion@213.175.160.16] has joined #uml
23:45 -!- iStormy [~stormy@209.205.169.19] has quit [Quit: Nope, not here anymore]
23:54 -!- coryb [~cory@24.124.40.190] has joined #uml
23:54 -!- stratsi [~stratsi@80.72.68.187] has quit [Ping timeout: 488 seconds]
--- Log closed Wed Jul 16 00:00:00 2003