--- Day changed

--- Log opened Thu Jul 17 00:00:02 2003
00:31 -!- Lathiat [lathiat@seven.sixlabs.org] has quit [Ping timeout: 490 seconds]
00:52 < Tusker> hmmm... somehow _I386_PTRACE_H is being defined... ie /usr/src/asm/ptrace.h is being included during the make linux ARCH=um
00:55 < Tusker> /usr/include/asm/ptrace.h rather
01:28 -!- stratsi [~stratsi@80.72.68.187] has quit [Ping timeout: 488 seconds]
01:31 < Tusker> anyone know why /usr/include/asm/ptrace.h would be included, even though I am building from a different directory and specifying the -I
01:52 -!- Snow-Man [~sfrost@66.93.83.236] has quit [Ping timeout: 488 seconds]
02:03 -!- Lathiat [lathiat@130.95.13.25] has joined #uml
02:16 < Tusker> Lathiat: you there ?
02:17 < Lathiat> yep
02:17 < Lathiat> sup?
02:17 < Tusker> have a compile problem with the UML kernel (when the host kernel has skas3 patch)
02:17 < Tusker> it is somehow including /usr/include/asm/ptrace.h (_I386_PTRACE_H)
02:18 < Tusker> which causes two structs to be "already defined"
02:18 < Lathiat> erm
02:18 < Lathiat> what version?
02:18 < Tusker> 2.4.21 
02:19 < Lathiat> hrm
02:19 < Lathiat> are you passing ARCH=um to every step?
02:19 < Tusker> for sure
02:19 -!- green [green@217.76.32.60] has joined #uml
02:20 < Lathiat> hrm
02:20 < Lathiat> ask green :P
02:20  * Lathiat hides from green
02:21 < Tusker> is green a compiling guru? :)
02:22 < green> hm?
02:22 < Tusker> UML kernel build on 2.4.21... host kernel has skas3 patch
02:23 < Tusker> fails to build since it is including /usr/include/asm/ptrace.h (_I386_PTRACE_H)
02:23 < green> how does it fails?
02:23 < Tusker> which means that some things are defined multiple times
02:24 < Tusker> In file included from process.c:27:
02:24 < green> ah, the multiple times thing (sprintf and stuff?) is probably because you need to undef GCOV and profiling support
02:24 < Tusker>  /usr/src/user-mode-linux/linux-2.4.21/arch/um/include/skas_ptrace.h:10: error: redefinition of `struct ptrace_faultinfo'
02:24 < Tusker>  /usr/src/user-mode-linux/linux-2.4.21/arch/um/include/skas_ptrace.h:15: error: redefinition of `struct ptrace_ldt'
02:24 < caker> http://marc.theaimsgroup.com/?l=user-mode-linux-user&m=101089127215367&w=2
02:28 < green> Tusker: btw, I do not even see where /usr/include/asm/ptrace.h is included, I only see /usr/include/sys/ptrace.h being included from arch/um/kernel/skas/process.c
02:28 < Tusker> ok, i'll build it in from a different directory then, totally seperate from /usr/src/
02:28 < caker> it says you screwed up your /usr/src/linux tree
02:29 < Tusker> green: yeah, I know... but, I did a #ifndef _I386_PTRACE_H and it was included
02:29 < Tusker> caker: it has the skas3 patch... hence screwed up :)
02:29 < Tusker> ? :)
02:29 < green> btw usually /usr/src/linux tree should have nothing to do with uml build. glibc keeps its headers separate
02:30 < caker> you build UML in /usr/src/linux -- so that email says .. but who knows .. Fix your tree for your host kernel (/usr/src/linux) then build UML some place else
02:30 < green> Tusker: put an #error in teh asm/ptrace.h and see where it was included from? ;)
02:30 < Tusker> my host kernel is vanilla + skas3
02:30 < Tusker> green: ahhh... good idea
02:31 < Tusker> one more build in a directory that is not /usr/src
02:31 < Tusker> and then i'll try that asm/ptrace.h thing
02:32 < Tusker> for your information... before I put the skas3 patch, the uml kernel built fine
02:32 < caker> where did you build your host kernel?
02:33 < Tusker> in /usr/src ...
02:34 < green> caker: I still insist that in normal glibc system it is completely irrelevant where you build the host kernel and if you have the sources of that kernel on the system at all
02:38 < caker> green: perhaps the makefile does weird things when it detects building out of /usr/src ?? i'm guessing :-)
02:38 < caker> or, it goes to include stuff from /usr/src/linux -- and he polluted it by patching it with UML
02:39  * caker shrugs
02:39 < Tusker> i've re-installed my host kernel a few times too... just vanilla + skas3... :)
02:41 -!- Snow-Man [~sfrost@snowman.net] has joined #uml
02:44 < Tusker> In file included from /usr/include/linux/ptrace.h:24,
02:44 < Tusker>                  from sigcontext.c:9:
02:48 < Tusker> and surprise surprise, host-skas3.patch adds stuff into host/include/asm-i386/ptrace.h
03:08 < green> hm
03:09 < green> you mean /usr/include/asm is a symlink to your real kernel tree?
03:09 < Tusker> if I answer yes will you hit me? :)
03:09 < green> yes ;)
03:10 < Tusker> (this is how the distro has it setup... 
03:10 < Tusker> )
03:10 < green> this is wrong wgong way! /usr/include/{linux,asm} stiff should represent kernel headers at the time of glibc compile
03:10 < Tusker> oh ok
03:12 < Tusker> so I can prove that there is something wrong with the way the distro is working, is there an RFC I can point to ?
03:13 < caker> don't build in /usr/src/linux ?
03:14 < green> Tusker: I think this is documented somewhere in either kernel or glibc. Do not remember right link, though
03:14 < Tusker> then where should kernel-headers go ?
03:15 < green> kernel headers should live in include/linux and include/asm, these are pglibc-build-precessed headers of the kernel headers of the system where this glibc was built, they are not equal to the kernel includes
03:15 < green> and real kernel carries its own headers with it
03:16 < Tusker> *nods*
03:16 < Tusker> so, /usr/src shouldn't even exist then?
03:17 < green> yes, it may not exist and still everything will work
03:17 < green> (this is true for properly built systems of course)
03:17 < Tusker> but in terms of standards compliance, there should be no need for /usr/src ?
03:19 < green> "man hier" have the definition, but does not declare it as mandatory
03:22 < Tusker> ok cheers
03:23 < green> Tusker: what is the broken distro you are using?
03:24 < Lathiat> laters
03:24 < Tusker> green: sourcemage
03:24 < green> Tusker: Ah, never heard of that ;)
03:24 < Tusker> green: it's pre 1.0 :)
03:24 < green> hehe.
03:24  * green thought that redhat clones are popular in Singapur
03:25 < Tusker> i'm actually australian... just working here
03:25 < green> ah
03:25 < green> still it would be natural if you'd have the same environment as everybody around you :)
03:26 < Tusker> well, I have about 6 linux boxes... :)
03:26 < Tusker> redhat + sourcemage + asplinux
03:27 < Tusker> oh, and a debian one
03:27 < green> hehe
03:28  * green prefers to have unified boxes, easier to take care of them this way
03:28 < Tusker> well... the asplinux boxes were because I was desperate (because I had no bandwidth, so that was the only one I could buy on CD)
03:29 -!- stratsi [~stratsi@80.72.68.187] has joined #uml
03:30 < green> Tusker: yup, that was the rh clone I referred to as "popular in Singapur" ;)
03:30 < Tusker> redhat ... because I was trying out rh9...
03:30 < Tusker> debian... co-sysadmin, and the other guy prefers debian
03:30 < Tusker> sourcemage... because I like to have a bit of a challenge sometimes :)
03:34 -!- snide [~snide@81.50.5.34] has joined #uml
04:15 -!- stratsi [~stratsi@80.72.68.187] has quit [Quit: BitchX: causing all sorts of havok!]
04:18 < Tusker> green: thanks for your insight
04:18 < Tusker> have to head off
04:20 < Tusker> cya
04:20 -!- Tusker [~tusker@cm237.omega39.maxonline.com.sg] has quit [Quit: [A] it brings us closer than dying ... and cancer ... and crying ...]
07:01 -!- desaster [~desaster@dsl-jklgw1o22.dial.inet.fi] has joined #uml
07:34 -!- wahlau [~sl@80.136.155.92] has joined #uml
07:34 < wahlau> hi
07:34 < wahlau> i have found out
07:35 < wahlau> that my samba does not work from my UML
07:35 < wahlau> anyone has similar experience?
07:59 -!- wahlau [~sl@80.136.155.92] has quit [Quit: using sirc version 2.211+KSIRC/1.2.4]
09:29 -!- jdike [~jdike@dhcp49.ists.dartmouth.edu] has joined #uml
09:29 < jdike> hi guys
09:30 < Lathiat> Hey jeff
09:30 < Lathiat> i looked at 2.6.0-test1 today, having not messed with 2.5 much its impressive
09:31 -!- mdz [~mdz@216-15-124-77.c3-0.smr-ubr3.sbo-smr.ma.cable.rcn.com] has quit [Quit: Client exiting]
09:32 < snide> Lathiat: yeah, but it seems that block device writers will have a hard time reworking their drivers
09:33 < Lathiat> they mmmunged the api?
09:33 < snide> Lathiat: i they hadn't already ;-)
09:33 < Lathiat> fun
09:33 < snide> Lathiat: the API changed meanwhile the 2.5... [ i don't know if it changed much actually ]
09:34 < jdike> The block driver interface has changed a fair bit over 2.5.x
09:34 < Lathiat> also linux leaving transmeta is interesting hehe
09:34 < Lathiat> ARGH
09:34 < Lathiat> TYPO
09:34  * Lathiat SLAPS HIMSELF
09:34 < Lathiat> linus damnit
09:35 < Lathiat> i just see linus liek on irc slapping me to death haha
09:37 < green> Hi Jeff
09:37 < jdike> Hey Oleg
09:38 < jdike> green: When are you leaving?
09:38 < green> leaving what?
09:38 < jdike> green: .ru
09:38 < green> Ah, on Saturday (July 19th)
09:39 < green> I still have almost two days left ;)
09:39 < jdike> green: yup, I'm driving up on Sunday, haven't decided when I'll get there
09:39 < green> jdike: BTW, I plan to move to your hotel room on July 22nd if you won't mind
09:40 < Lathiat> u guys catchin up?>
09:40 < jdike> green: np, it's available starting on Sunday
09:41 < green> jdike: I arrive on Saturday , and since usenix pays for it anyway, I decided I will provide you with three days without me ;)
09:41 < jdike> green: hehe
09:41 < jdike> green: OK
09:41 < green> actually that should read "two nights" ;)
09:42 < green> jdike: btw do you know what is the canadian customs' rules about alcohol crossing the border?
09:43 < jdike> green: not sure, I don't think there should be problems
09:43 < green> good to hear ;)
09:43 < jdike> green: Harald Welte brought a whole keg from Germany last year
09:43 < green> well, beer is not quite an alcohol (beer is not an alcohol by russian laws ;) )
09:44 < jdike> green: Well, it is according to wimpy North American standards
09:44 < green> bad news is the weather will be mostly rainy during entire next week
09:46 < jdike> green: Strange, the National Weather Service doesn't seem to provide Canadian weather :-)
09:47 < green> Last tiem I checked state of the art weather control was only at 20% success rate, or do you refer to weather forecasts? ;)
09:47 < jdike> green: forecasting
09:48 < green> ah, I used data from www.weather.com ;)
09:48 < jdike> green: showers next week, no big deal
09:49 < green> hm, is shower means something different from rain?
09:49 < jdike> green: it just means brief periods of rain, as opposed to all day
09:50 < green> Ah. Still probably shiny days would be better... Or may be not...
09:51 < green> But at least there will be no snow, so probably we still consider the weather will be mostly good rather than bad ;)
09:51 < jdike> green: most likely there will be no snow :-)
09:53 < green> and this is great ;)
10:57 -!- jdike [~jdike@dhcp49.ists.dartmouth.edu] has quit [Ping timeout: 501 seconds]
11:05 -!- jdike [~jdike@dhcp49.ists.dartmouth.edu] has joined #uml
11:07 -!- Lathiat_ [~Lathiat@203.15.140.23] has joined #uml
11:09 -!- Lathiat_ [~Lathiat@203.15.140.23] has quit [Client Quit]
11:09 -!- Lathiat [lathiat@130.95.13.25] has quit [Remote host closed the connection]
11:09 -!- Lathiat [lathiat@seven.sixlabs.org] has joined #uml
11:27 -!- Lathiat_ [~Lathiat@thump.bur.st] has joined #uml
11:28 -!- Lathiat_ [~Lathiat@thump.bur.st] has quit [Client Quit]
11:41 -!- revenger [revenger@p508B39AF.dip.t-dialin.net] has joined #uml
11:44 -!- revenger4 [revenger@p508B2DFA.dip.t-dialin.net] has quit [Ping timeout: 490 seconds]
12:08 < snide> funny, 2.4.20-um6 doesn't want to be compiled by gcc-3.3 (error in netrcv*)
12:26 < green> known thing
12:26 < green> 2.4.20-vanilla have the same issue
12:26 < green> just replace 'extern inline' with 'static inline' in some file in net/
12:27 < snide> green: did it appear in uml-devel ? (i'll search there then)
12:29 < green> no, it was in lkml. just grep for 'extern inline' in net, there is only one occurence
14:02 -!- albino [~albino@129.123.227.104] has quit [Ping timeout: 488 seconds]
14:41 -!- revenger [revenger@p508B39AF.dip.t-dialin.net] has quit [Quit: Lost terminal]
14:43 -!- ticallion [~ticallion@213.175.160.23] has joined #uml
14:44 < jdike> UML doesn't boot so well when you forget CONFIG_BINFMT_ELF
14:58 < jdike> Linux usermode 2.5.70-1um #2 SMP Thu Jul 17 14:59:24 EDT 2003 i686 unknown
14:59  * jdike is slightly behind the times
15:07 -!- ticallion [~ticallion@213.175.160.23] has quit [Quit: Client exiting]
15:07 -!- desaster [~desaster@dsl-jklgw1o22.dial.inet.fi] has quit [Quit: Hmmm. EPIC4-1.1.12 (558) has another bug.  Go figure...]
15:09 -!- desaster [~desaster@dsl-jklgw1o22.dial.inet.fi] has joined #uml
15:09 -!- desaster [~desaster@dsl-jklgw1o22.dial.inet.fi] has quit [Client Quit]
15:10 -!- desaster [~desaster@dsl-jklgw1o22.dial.inet.fi] has joined #uml
15:11 -!- desaster [~desaster@dsl-jklgw1o22.dial.inet.fi] has quit [Client Quit]
15:12 -!- desaster [~desaster@80.222.254.34] has joined #uml
15:16 -!- glommer [~glauber@200-158-193-117.dsl.telesp.net.br] has left #uml [Client exiting]
16:06 -!- \malex\ [~alex@207.44.210.17] has joined #uml
16:25 -!- desaster [~desaster@80.222.254.34] has quit [Ping timeout: 488 seconds]
16:31 < snide> jdike: why ? u could build it to use CONFIG_BINFMT_MISC (such as Java ;-)) )
16:32 < jdike> snide: you have an init that's written in Java?
16:33 < snide> jdike: actually, the biggest prob is the JRE ;-)
16:33 < jdike> snide: not to mention everything else the system needs?  JLinux ... would be interesting
16:33 < snide> jdike: but actually a init in java wouldn't be the hard part
16:33 < snide> -IMHO-
16:35 < snide> jdike: actually, why bother with Java ? we'll have Linux now... [ hmmm... should have an os_posix layer ]
16:39 -!- snide [~snide@81.50.5.34] has quit [Quit: [BX] Abort Retry Fail]
16:49 -!- peace [shanti@157.182.197.16] has joined #uml
16:49 -!- peace is now known as shanti
16:51 < shanti> jdike: I managed to graduate with whatever work I had done with uml/Ultrasparc, and now I got job at Cornell univ,NY
16:53 < jdike> shanti: cool, what are you doing there?
16:54 < shanti> jdike: Title is "Unix research support specialist"..have to administer 30+ redhat systems and some software development
16:54 < shanti> in C, Tk, etc
16:57 < jdike> shanti: how do you like it?
16:58 < shanti> A professor is doing some research on ELCAP (Early Lung Cancer Action Project), so I have work with him to do system administration
16:58 < shanti> jdike: Haven't started on it yet, will be joining there on July 21st
16:59 < jdike> shanti: OK, next week you start
16:59 < shanti> jdike: yes
17:03 < jdike> Wow.
17:03 < jdike> The last 2.5 patch I released was 2.5.67
17:18 -!- desaster [~saigai@80.222.254.34] has joined #uml
17:27 -!- shanti [shanti@157.182.197.16] has quit [Quit: Client Exiting]
17:38 -!- jdike [~jdike@dhcp49.ists.dartmouth.edu] has quit [Quit: Leaving]
18:33 -!- mdz [~mdz@216-15-124-77.c3-0.smr-ubr3.sbo-smr.ma.cable.rcn.com] has joined #uml
19:03 -!- david [~david@207.166.203.138] has quit [Excess Flood]
20:19 -!- slingo [~slingo@adsl-17-185-14.bhm.bellsouth.net] has joined #uml
20:20 -!- \malex\ [~alex@207.44.210.17] has left #uml []
20:43 -!- caker [~null@pcp507591pcs.nash01.tn.comcast.net] has quit [Ping timeout: 488 seconds]
20:49 -!- david [~d@vidcoulson.com] has joined #uml
21:23 -!- cameron [~cameron@139.80.30.192] has joined #uml
21:24 < cameron> Do I need an extra patch to put skas3 in my 2.4.21 kernel?
21:26 < david> no
21:26 < david> host-skas3 should work fine with 2.4.21
21:26 < cameron> Thanks
21:44 -!- Lathiat_ [lathiat@seven.sixlabs.org] has joined #uml
21:44 -!- Lathiat_ [lathiat@seven.sixlabs.org] has quit [Client Quit]
21:45 -!- Lathiat_ [lathiat@seven.sixlabs.org] has joined #uml
21:50 -!- cameron [~cameron@139.80.30.192] has quit [Quit: BitchX: its wax ecstatic]
21:58 -!- cameron [~cameron@crosstalk.otago.ac.nz] has joined #uml
22:02 -!- Lathiat_ [lathiat@seven.sixlabs.org] has quit [Quit: Lost terminal]
22:35 -!- caker [~null@68.52.196.167] has joined #uml
22:46 < slingo> hmm
22:58 -!- ticallion [~ticallion@213.175.160.215] has joined #uml
23:04 -!- cameron [~cameron@crosstalk.otago.ac.nz] has quit [Quit: My damn controlling terminal disappeared!]
23:30 -!- rlo [~roberto@200.138.100.242] has joined #uml
23:30 < rlo> is uml broke in 2.6.0?
23:31  * BB wonders if "2.6.0 is broke" as a whole ;)
23:31 < rlo> nah, it isn't that bad
23:31 < BB> maybe not for you then ;)
23:32 < rlo> oh, of course :)
23:32 < rlo> but for you it seems so ;)
23:33 < ticallion> rlo: I failed to make it work, doesn't build, somebody mentionned that the uml code in 2.6.0-test1 is ancient
23:35 < david> 2.6.0-test1 does not have current UML code
23:35 < david> it probably doesn't compile, much less work
23:35 < rlo> ticallion: ok thanks. jdike posted a patch for uml against 2.5.70, so I was wondering what was the prob with uml 2.6.0
23:35 < rlo> david: it certainly doesn't compile. just tried
23:36  * ticallion waves to david
23:36 < ticallion> rlo: url pls
23:39 < rlo> http://www.ussg.iu.edu/hypermail/linux/kernel/0307.2/0787.html
23:39 < rlo> ticallion: sorry for the delay...
23:39 < ticallion> rlo: it's all good, thanks
23:40 < rlo> ticallion: you're welcome
23:44 < david> rlo: Jeff is just trying to catch up with the diffs
23:45  * ticallion trying the patch
23:46 < rlo> david: Yeah, I saw his post. But the uml in 2.6.0 was  intriguing me, so I gave a shot...
23:47 < david> rlo: I understand
23:47 < david> rlo: you'll have to wait for jeff to catch up to the current 2.5 tree
23:47 < david> then to 2.6
23:48 < rlo> k, in the meantime I'll play with 2.5.70
23:49 < ticallion> have 3 failures
23:51 < ticallion> doesn't bloody build
23:53 -!- ticallion [~ticallion@213.175.160.215] has quit [Quit: later]
23:55 < rlo> yeah, doesn't build... 
23:55 < rlo> well, bedtime anyway. thank you guys for the help
23:56 -!- rlo [~roberto@200.138.100.242] has quit [Quit: Finalizando Cliente]
--- Log closed Fri Jul 18 00:00:00 2003