TheGrebs.com:IRC Logs:#uml:2006:07:31

---	Log	opened Mon Jul 31 00:00:52 2006
00:10	\|-\|	Cowboy_ [~Cowboy@netblock-68-183-193-69.dslextreme.com] has joined #uml
00:11	\|-\|	Cowboy_ [~Cowboy@netblock-68-183-193-69.dslextreme.com] has left #uml []
00:11	\|-\|	Supaplex [~supaplex@shell.aros.net] has quit [Ping timeout: 480 seconds]
00:15	\|-\|	Supaplex [~supaplex@shell.aros.net] has joined #uml
00:51	\|-\|	tanuki [~TOTEHM00N@dialup-4.247.134.53.Dial1.Tampa1.Level3.net] has quit [Ping timeout: 480 seconds]
00:58	\|-\|	tanuki [~TOTEHM00N@dialup-4.247.134.170.Dial1.Tampa1.Level3.net] has joined #uml
01:19	\|-\|	Ancalagon [~PtitKosmi@215.208-78-194.adsl-fix.skynet.be] has joined #uml
01:27	\|-\|	kokoko1 [~Slacker@203.148.65.8] has quit [Ping timeout: 480 seconds]
01:32	<Ancalagon>	hi all
01:40	\|-\|	pingu [~seraph@40-163.127-70.tampabay.res.rr.com] has quit [Remote host closed the connection]
02:18	\|-\|	MrX [~urk@219.95.238.72] has quit [Quit: urk IRC v0.-1.4 - http://urk.sf.net/]
02:28	\|-\|	tanuki_ [~TOTEHM00N@dialup-4.247.134.225.Dial1.Tampa1.Level3.net] has joined #uml
02:30	\|-\|	tanuki [~TOTEHM00N@dialup-4.247.134.170.Dial1.Tampa1.Level3.net] has quit [Ping timeout: 480 seconds]
03:28	\|-\|	Ancalagon [~PtitKosmi@215.208-78-194.adsl-fix.skynet.be] has quit [Ping timeout: 480 seconds]
03:37	\|-\|	tanuki_ [~TOTEHM00N@dialup-4.247.134.225.Dial1.Tampa1.Level3.net] has quit [Quit: Rising Above Mediocrity: http://bgeiger.net/weblog/ ]
03:41	\|-\|	Ancalagon [~PtitKosmi@215.208-78-194.adsl-fix.skynet.be] has joined #uml
04:28	\|-\|	Ancalagon [~PtitKosmi@215.208-78-194.adsl-fix.skynet.be] has quit [Ping timeout: 480 seconds]
04:57	\|-\|	Ancalagon [~PtitKosmi@215.208-78-194.adsl-fix.skynet.be] has joined #uml
05:00	\|-\|	HuK0B [~NIKOV@62.213.165.194] has joined #uml
05:29	\|-\|	Ancalagon [~PtitKosmi@215.208-78-194.adsl-fix.skynet.be] has quit [Ping timeout: 480 seconds]
05:39	\|-\|	rgl [Rui@217.129.151.190] has joined #uml
05:39	<rgl>	hi
05:40	<rgl>	known a backport package of uml to debian stable?
05:41	<HuK0B>	? what you mean
05:42	<HuK0B>	I use uml with debian sarge host and work perfect
05:42	<rgl>	how did you install it HuK0B ?
05:42	<rgl>	I mean, there is a debian package for it, or did you compile it?
05:45	<HuK0B>	debian package
05:46	<HuK0B>	if you want to make you fast step to install
05:46	<rgl>	HuK0B, which one?
05:46	<HuK0B>	apt-get install uml-utilities
05:46	<rgl>	I can only find user-mode-linux package on the testing dist of debian
05:46	<HuK0B>	in sarge there are package too
05:46	<rgl>	where?
05:47	<rgl>	http://packages.debian.org/cgi-bin/search_packages.pl?keywords=user-mode-linux&searchon=names&subword=1&version=stable&release=all
05:47	<rgl>	nothing returned
05:47	<HuK0B>	http://ftp.debian.org/pool/main/u/uml-utilities/uml-utilities_20040406-1_i386.deb
05:47	<HuK0B>	this is stable realise in debian sarge
05:48	\|-\|	tyler [~tyler@ARennes-356-1-156-27.w86-220.abo.wanadoo.fr] has joined #uml
05:48	<rgl>	humm, that package actually contains the uml kernel? AFIKT it only contains the utilities to manage uml instances.
05:48	<HuK0B>	everything is ok you don't need to compile anything except host kernel with skas3 and ghost kernel(if you prefer to compile it) or you can just download ghost kernel
05:49	<HuK0B>	uml kernel you mean ghost kernel?
05:49	<HuK0B>	no this package is only utillities
05:49	<HuK0B>	uml_mconsole... etc.
05:49	<rgl>	yes, the guest (ghost?) kernel
05:49	<HuK0B>	you can download compiled ghost kernel
05:49	<HuK0B>	or to compile one
05:50	<rgl>	how are you compiling the host kernel with skas patch? are you using the kernel-source package?
05:51	<HuK0B>	w8 I will give you link for howto for debian sarge
05:51	<HuK0B>	I useing host kernel with skas3 patch and normal ghost kernel 2.6.17 without any patches
05:51	<HuK0B>	just download one kernel 2.6.17 vanilla from kernel.org
05:52	<rgl>	ok :)
05:54	<rgl>	HuK0B, the howto you are talking about is the one at http://uml.ligatura.org/ ?
05:55	<HuK0B>	http://uml.openconsultancy.com/
05:55	<HuK0B>	here good howto
05:55	<HuK0B>	yes this is good too
05:57	<HuK0B>	here http://www.user-mode-linux.org/~blaisorblade/patches/skas3-2.6/skas-2.6.17-rc5-v9-pre9/skas-2.6.17-rc5-v9-pre9.patch.bz2
05:57	<HuK0B>	skas patch for host kernel 2.6.17
05:57	<rgl>	how are you building the host kernel?
05:57	<rgl>	you also build a stock kernel?
05:57	<rgl>	that is, you don't use the debian kernels?
05:58	<HuK0B>	no
05:58	<HuK0B>	I don't use debian kernels
05:58	<HuK0B>	if you want skas3 I think you need to compile it
05:59	<HuK0B>	but you can start uml with debian kernel too but without skas3 mode
05:59	<rgl>	I see, ok thx for the tips :)
06:10	\|-\|	Ancalagon [~PtitKosmi@215.208-78-194.adsl-fix.skynet.be] has joined #uml
06:24	\|-\|	debugger_ [~Rui@217.129.151.190] has joined #uml
06:25	\|-\|	debugger_ [~Rui@217.129.151.190] has quit []
06:32	\|-\|	rgl [Rui@217.129.151.190] has quit [Ping timeout: 480 seconds]
07:13	\|-\|	Ancalagon [~PtitKosmi@215.208-78-194.adsl-fix.skynet.be] has quit [Ping timeout: 480 seconds]
07:13	\|-\|	Ancalagon [~PtitKosmi@215.208-78-194.adsl-fix.skynet.be] has joined #uml
07:22	\|-\|	Ancalagon [~PtitKosmi@215.208-78-194.adsl-fix.skynet.be] has quit [Ping timeout: 480 seconds]
07:22	\|-\|	Ancalagon [~PtitKosmi@215.208-78-194.adsl-fix.skynet.be] has joined #uml
07:45	\|-\|	HuK0B [~NIKOV@62.213.165.194] has quit []
08:08	\|-\|	Segora [~gefla@p4523e3d4.np.schlund.de] has joined #uml
08:09	\|-\|	Ancalagon [~PtitKosmi@215.208-78-194.adsl-fix.skynet.be] has quit [Ping timeout: 480 seconds]
08:35	\|-\|	Ancalagon [~PtitKosmi@215.208-78-194.adsl-fix.skynet.be] has joined #uml
08:37	\|-\|	kokoko1 [~Slacker@203.148.65.8] has joined #uml
08:48	\|-\|	laptop [~david@dsl-146-220-190.telkomadsl.co.za] has quit []
08:49	\|-\|	pjb [~pjb@thalassa.informatimago.com] has quit [Quit: ERC Version 5.0.4 $Revision: 1.726.2.19 $ (IRC client for Emacs)]
09:02	\|-\|	kybe [~kyrre@legolas.iu.hio.no] has quit [Ping timeout: 480 seconds]
09:32	<Ancalagon>	bye
09:32	\|-\|	Ancalagon [~PtitKosmi@215.208-78-194.adsl-fix.skynet.be] has quit [Quit: Chatzilla 0.9.74 [Firefox 1.0.4/20060629]]
09:45	\|-\|	HuK0B [~NIKOV@62.213.165.194] has joined #uml
09:46	\|-\|	HuK0B [~NIKOV@62.213.165.194] has quit []
09:46	\|-\|	MrX [~urk@219.95.238.72] has joined #uml
09:58	\|-\|	tyler [~tyler@ARennes-356-1-156-27.w86-220.abo.wanadoo.fr] has quit [Ping timeout: 480 seconds]
10:13	\|-\|	kybe [~kyrre@legolas.iu.hio.no] has joined #uml
10:15	\|-\|	tyler [~tyler@ARennes-356-1-149-23.w86-220.abo.wanadoo.fr] has joined #uml
10:39	\|-\|	jdike [~jdike@pool-71-248-190-113.bstnma.fios.verizon.net] has joined #uml
10:39	\|-\|	MrX [~urk@219.95.238.72] has quit [Quit: urk IRC v0.-1.4 - http://urk.sf.net/]
10:39	<jdike>	Hi guys
10:47	\|-\|	the_hydra [~a_mulyadi@202.59.168.5] has joined #uml
10:59	\|-\|	hfb [~hfb@pool-71-106-220-165.lsanca.dsl-w.verizon.net] has joined #uml
10:59	\|-\|	wimble [~c06ac502@mail.thegrebs.com] has joined #uml
10:59	<wimble>	hello there
10:59	<the_hydra>	hello wimble
11:00	<wimble>	is there any way to implement a virtual time scheduler as in umlsim without the performance drop of breakpoints?
11:00	<wimble>	hi the_hydra :)
11:01	<the_hydra>	virtual time scheduler
11:01	<wimble>	yep
11:01	<the_hydra>	hm, mind to explain briefly?
11:01	<wimble>	like in a simulator
11:01	<the_hydra>	hmm
11:01	<wimble>	you know ns-2 for example?
11:01	<wimble>	there's a scheduler
11:01	<wimble>	that fires events in chronological order
11:02	<wimble>	suppose your pc is not fast enoug
11:02	<wimble>	enough
11:02	<the_hydra>	ns2-> network simulator?
11:02	<wimble>	yes, you said it
11:02	<wimble>	=)
11:02	<the_hydra>	:))
11:02	<wimble>	if it gets too long to simulate what's going on
11:02	<the_hydra>	ok please continue
11:02	<wimble>	then your simulation might take longer than real time
11:02	<the_hydra>	i c...
11:03	<the_hydra>	i think I got your point
11:03	<wimble>	e.g. simulation time of 1 hour in order to reproduce 10 minutes of simulated time
11:03	<the_hydra>	so, you want to find a way to get rid that "delay"?
11:03	<wimble>	now, if i need to emulate a network of hundreds nodes with a single machine
11:04	<wimble>	there are little chances that i am going to do it realistically in real time
11:04	<wimble>	the only alternative i know is to patch UML with the patches provided by UMLSIM
11:04	<wimble>	but those involve the use of breakpoints
11:05	<the_hydra>	I think, with respect to UMLSim, i think you will always experience delays
11:05	<wimble>	and the author say that that technique is extremely slow because each breakpoint involves hundreds of system calls
11:05	<the_hydra>	no matter how small or big it is
11:05	<wimble>	say it again?
11:05	<the_hydra>	I think, with respect to UMLSim, i think you will always experience delays
11:05	<wimble>	(please)
11:05	<the_hydra>	no matter how small or big it is
11:05	<wimble>	yes, but in other words, i mean :)
11:05	<the_hydra>	especially in virtualized/emulated environment
11:05	<wimble>	which delays are you talking about?
11:05	<the_hydra>	hm ok
11:06	<wimble>	sure,
11:06	<the_hydra>	timer shot delay
11:06	<wimble>	i don't mind _too much_ about speed
11:06	<wimble>	i just care about realism
11:06	<the_hydra>	i.e you want it to expire in 100 ms, it might be fired up after 110 ms
11:06	<wimble>	but i would like not to have such a big slowdown
11:07	<wimble>	but the uml clock should advance only when umlsim wants it to, isn't it?
11:07	<the_hydra>	I think so
11:07	<wimble>	i mean, what matters is not that things happen timely
11:07	<wimble>	but that it looks like they were
11:08	<wimble>	and that if some machine is running ahead of the other ones
11:08	<wimble>	it stops and waits them
11:08	<the_hydra>	you mean, to make the clock synchronized between virtualized nodes?
11:08	<wimble>	yes
11:08	<wimble>	exactly
11:08	<the_hydra>	ok
11:09	<the_hydra>	I think, that is what jdike currently work on
11:09	<wimble>	and to make those that are "going faster" sleep and wait for the other ones
11:09	<wimble>	oh, that would be really cool
11:09	<the_hydra>	i.e: making UML fetch "time" from virtual source
11:09	<the_hydra>	i don't remember the exact explanation
11:09	<the_hydra>	try to dig LWN
11:10	<the_hydra>	using this approach, UML won't rely on gettimeofday or other such slow call
11:10	<wimble>	!google LWN virtual time source uml
11:10	<linbot>	wimble: Search took 0.40 seconds: LWN : Virtual time: <https://lwn.net/Articles/180375/>; LWN : OLS: Xen and UML: <http://lwn.net/Articles/144765/>;; LWN : A look at Xen: <http://lwn.net/Articles/139964/>;; LXer: An Interview with Jeff Dike - The creator of User Mode Linux: <http://lxer.com/module/newswire/view/62755/index.html>;; LXer: User Mode Linux: <http://lxer.com/module/newswire/view/60766/index.html>;; (1 more message)
11:10	<the_hydra>	jdike: jeff, are you there?
11:10	<jdike>	yup
11:10	<the_hydra>	please CMIIW
11:10	<the_hydra>	wimble:the 1st one
11:11	<wimble>	the_hydra: the first three say 404 not found :(
11:11	<the_hydra>	what? hmm
11:11	<wimble>	how do i get the next message from linbot?
11:12	<the_hydra>	linbot: help
11:12	<linbot>	the_hydra: (help [<plugin>] [<command>]) -- This command gives a useful description of what <command> does. <plugin> is only necessary if the command is in more than one plugin.
11:13	<wimble>	linbot: help google
11:13	<linbot>	wimble: (google <search> [--{language,restrict} <value>] [--{notsafe,similar}]) -- Searches google.com for the given string. As many results as can fit are included. --language accepts a language abbreviation; --restrict restricts the results to certain classes of things; --similar tells Google not to filter similar results. --notsafe allows possibly work-unsafe results.
11:13	[~]	wimble mumbles...
11:14	<wimble>	i'll get the cache copy from the real google
11:15	[~]	the_hydra kicks linbot
11:15	<wimble>	why that?
11:16	<the_hydra>	to make it shows the next result set :)
11:19	<wimble>	?
11:19	<wimble>	jdike: haven't yet grabbed the patch
11:19	<wimble>	it looks like it's an host patch, if i understand what it does
11:20	<jdike>	all it does is make UML gettimeofday faster
11:20	<the_hydra>	jdike: no more syscall latency, jeff?
11:20	<jdike>	no more ptrace latency
11:20	<the_hydra>	oh ok
11:21	<wimble>	but it would allow also other processes than uml linuces to have their own time space?
11:22	<jdike>	yeah
11:23	<wimble>	ok, so when i'll find the patches i should apply them to the host, i guess
11:23	<wimble>	(when i'll find 'em =)
11:23	<the_hydra>	or just dig uml-devel archieves
11:23	<the_hydra>	i am sure that is in there too
11:24	<wimble>	thank you
11:24	<wimble>	why don't they leave attachments on the archives? is it because the vger policy is to paste patches?
11:26	<wimble>	i have to go
11:26	<wimble>	will look at it tomorrow
11:26	<the_hydra>	cu wimble
11:26	<wimble>	thank you!
11:26	<wimble>	cheers
11:26	<the_hydra>	cheers!
11:26	\|-\|	wimble [~c06ac502@mail.thegrebs.com] has left #uml []
11:36	\|-\|	Cowboy_ [~Cowboy@129.42.184.35] has joined #uml
11:39	\|-\|	the_hydra [~a_mulyadi@202.59.168.5] has quit []
11:42	\|-\|	Cowboy_ [~Cowboy@129.42.184.35] has left #uml [Client exiting]
12:00	\|-\|	tanuki [~TOTEHM00N@40-163.127-70.tampabay.res.rr.com] has joined #uml
12:23	\|-\|	da-x [~Dan@cpe-68-173-44-101.nyc.res.rr.com] has joined #uml
12:30	\|-\|	dan [~Dan@cpe-68-173-42-9.nyc.res.rr.com] has quit [Ping timeout: 480 seconds]
12:38	\|-\|	da-x [~Dan@cpe-68-173-44-101.nyc.res.rr.com] has quit [Ping timeout: 480 seconds]
12:39	\|-\|	nikkne_ [~nikkne@nslpc3.epfl.ch] has joined #uml
12:39	\|-\|	nikkne [~nikkne@nslpc3.epfl.ch] has quit [Read error: Connection reset by peer]
12:50	\|-\|	da-x [~Dan@cpe-68-173-44-101.nyc.res.rr.com] has joined #uml
13:14	\|-\|	da-x [~Dan@cpe-68-173-44-101.nyc.res.rr.com] has quit [Ping timeout: 480 seconds]
13:14	\|-\|	da-x [~Dan@cpe-68-173-44-101.nyc.res.rr.com] has joined #uml
13:21	\|-\|	pingu [~seraph@70.127.163.40] has joined #uml
13:27	\|-\|	da-x [~Dan@cpe-68-173-44-101.nyc.res.rr.com] has quit [Ping timeout: 480 seconds]
13:30	\|-\|	pingu [~seraph@70.127.163.40] has quit [Quit: using sirc version 2.211+KSIRC/1.3.12]
13:37	\|-\|	da-x [~Dan@cpe-68-173-44-101.nyc.res.rr.com] has joined #uml
13:39	\|-\|	tyler [~tyler@ARennes-356-1-149-23.w86-220.abo.wanadoo.fr] has quit [Ping timeout: 480 seconds]
13:48	\|-\|	nikkne_ [~nikkne@nslpc3.epfl.ch] has quit [Quit: Konversation terminated!]
13:51	\|-\|	tyler [~tyler@ARennes-356-1-131-129.w86-220.abo.wanadoo.fr] has joined #uml
14:00	\|-\|	da-x [~Dan@cpe-68-173-44-101.nyc.res.rr.com] has quit [Ping timeout: 480 seconds]
14:03	\|-\|	da-x [~Dan@cpe-68-173-44-101.nyc.res.rr.com] has joined #uml
14:12	\|-\|	kos_tom [~thomas@humanoidz.org] has joined #uml
14:14	\|-\|	tyler [~tyler@ARennes-356-1-131-129.w86-220.abo.wanadoo.fr] has quit [Ping timeout: 480 seconds]
14:22	\|-\|	tyler [~tyler@ARennes-356-1-149-166.w86-220.abo.wanadoo.fr] has joined #uml
14:35	\|-\|	da-x [~Dan@cpe-68-173-44-101.nyc.res.rr.com] has quit [Ping timeout: 480 seconds]
14:36	\|-\|	da-x [~Dan@cpe-68-173-44-101.nyc.res.rr.com] has joined #uml
14:38	\|-\|	tyler [~tyler@ARennes-356-1-149-166.w86-220.abo.wanadoo.fr] has quit [Read error: Connection reset by peer]
14:40	\|-\|	Segora [~gefla@p4523e3d4.np.schlund.de] has quit [Ping timeout: 480 seconds]
14:42	\|-\|	ram_away changed nick to ram
14:47	\|-\|	da-x [~Dan@cpe-68-173-44-101.nyc.res.rr.com] has quit [Ping timeout: 480 seconds]
14:52	\|-\|	da-x [~Dan@cpe-68-173-44-101.nyc.res.rr.com] has joined #uml
14:56	[~]	tanuki wonders idly whether it's possible for a kernel to muck with the memory manager's page-replacement algorithm
14:56	<tanuki>	erk
14:56	<tanuki>	a kernel module, even.
15:14	\|-\|	da-x [~Dan@cpe-68-173-44-101.nyc.res.rr.com] has quit [Ping timeout: 480 seconds]
15:16	<jdike>	no
15:16	<tanuki>	:-(
15:16	<jdike>	you can fiddle statistics maybe, if they're exported
15:17	<tanuki>	this is for Yet Another SOFTICE Lab
15:17	<tanuki>	(before anyone asks, pingu is my employer)
15:35	\|-\|	tyler [~tyler@ARennes-356-1-149-166.w86-220.abo.wanadoo.fr] has joined #uml
15:36	\|-\|	da-x [~Dan@cpe-68-173-42-9.nyc.res.rr.com] has joined #uml
15:41	\|-\|	pingu [~seraph@40-163.127-70.tampabay.res.rr.com] has joined #uml
15:42	<pingu>	hola people
15:43	<pingu>	if someone has time (or is curious) the latest of the problem w/ compiling the guest scheduler inside a uml guest kernel is at http://pastebin.ca/100336. Didn't touch it in last couple of days, resuming today :)
15:46	\|-\|	tyler [~tyler@ARennes-356-1-149-166.w86-220.abo.wanadoo.fr] has quit [Ping timeout: 480 seconds]
17:17	\|-\|	kos_tom [~thomas@humanoidz.org] has quit [Quit: I like core dumps]
18:13	\|-\|	ram changed nick to ram_away
18:58	\|-\|	pingu [~seraph@40-163.127-70.tampabay.res.rr.com] has quit [Quit: using sirc version 2.211+KSIRC/1.3.12]
19:09	\|-\|	tanuki [~TOTEHM00N@40-163.127-70.tampabay.res.rr.com] has quit [Quit: Rising Above Mediocrity: http://bgeiger.net/weblog/ ]
20:16	\|-\|	jdike [~jdike@pool-71-248-190-113.bstnma.fios.verizon.net] has quit [Quit: Leaving]
20:33	\|-\|	hfb [~hfb@pool-71-106-220-165.lsanca.dsl-w.verizon.net] has quit [Quit: Leaving]
22:31	\|-\|	Nem^1 [~Nem@dslb-084-056-238-196.pools.arcor-ip.net] has joined #uml
22:38	\|-\|	Nem^ [~Nem@dslb-084-056-232-022.pools.arcor-ip.net] has quit [Ping timeout: 480 seconds]
22:39	\|-\|	Nem^1 changed nick to Nem^
---	Log	closed Tue Aug 01 00:00:57 2006