--- Day changed

--- Log opened Tue Mar 29 23:59:01 2005
00:06 -!- monrad [~monrad@213083190130.sonofon.dk] has quit [Quit: Leaving]
00:08 -!- Bluefox [~bluefox@pcp484971pcs.whtmrs01.md.comcast.net] has quit [Server closed connection]
00:08 -!- Bluefox [~bluefox@pcp484971pcs.whtmrs01.md.comcast.net] has joined #xen
04:24 -!- surriel [~riel@imladris.surriel.com] has quit [Server closed connection]
04:24 -!- surriel [~riel@imladris.surriel.com] has joined #xen
06:45 -!- mael [~mael@nat.inha.fr] has joined #xen
06:45 < mael> hi
06:48 < mael> I had a hard time trying to find this chan
06:48 < mael> as I went there trhu freenode the first time and now it's closed
06:49 < mael> maybe it would be nice if the informations about this chan were published somewhere (in xen's FAQ?)
07:15 < mael> Tv: hehe my colleague (tamtam) is talking with you on another chan :)
07:29 < surriel> I guess I could add a pointer to this channel on the Xen wiki
07:41 < mael> you mean the one on terrabox?
07:42 < surriel> no, I mean http://wiki.xensource.com/
07:42 < muli_> please do
07:42 < surriel> I will, once I get to the office
07:42 < surriel> my browser there remembers the password ;)
07:42 < surriel> (and I don't, this early in the morning)
07:43 < mael> :)
07:48 < mael> mmmh btw there's no link to the wiki from xensource nor from the cl.cam.ac.uk site
07:52 < mael> is there anything on the wiki surriel ?
07:52 < mael> I'm stuck with french and when I want to search the site I'm redirected to 192.168.0.34
08:20 -!- DEac- [~deac@xdsl-84-44-150-6.netcologne.de] has quit [Ping timeout: 480 seconds]
08:33 -!- DEac- [~deac@xdsl-213-196-201-126.netcologne.de] has joined #xen
08:37 -!- mael [~mael@nat.inha.fr] has quit [Read error: Connection reset by peer]
08:43 -!- mael [~mael@nat.inha.fr] has joined #xen
08:47 -!- MadMethod [Method@pcp0010742058pcs.howard01.md.comcast.net] has quit [Ping timeout: 480 seconds]
08:53 < Tv> mael: This is an unofficial channel, I guess. None of the original xen authors are ever here, I think.
08:54 < mael> yeah true
08:54 < mael> but it is an efficient addition to xen user list I think
08:54 < mael> Tv:  so did you decided a name with tamtam? :)
08:55 < surriel> Tv: they might show up once they learn of the existance of this channel ;)
08:56 < Tv> mael: yes, dhiccup
08:56 < Tv> surriel: I sort of get the feel they are too scientisty to IRC :)
08:57 < mael> hehe fun
08:57 < mael> Tv:  it was pretty strange to hear about you the way I did
08:58 < mael> irc is a small world finally :)
09:00 -!- Method [~Method@stanford.columbia.tresys.com] has joined #xen
09:25 -!- hollis [~hollis@user-0vvde2g.cable.mindspring.com] has joined #xen
09:48 -!- unriel is now known as riel
09:49 < riel> Tv: I've met them in person, they're geeky just like us ;)
09:50 < mael> :)
09:50 -!- rharper [~rharper@pixpat.austin.ibm.com] has joined #xen
09:52 < riel> mmm ok, so the front page can't be edited
09:53 < mael> locked I guess
09:53 < mael> is there any content on the wiki?
09:54 < mael> I was not able to do a search on it and I stuck to french because this is my prefered langage choice
09:54  * riel edits http://trac.xensource.com/xenwiki/DeveloperDiscussion instead
09:54 < riel> mael: you can configure your language choice
09:54 < riel> and yes, there is content
09:55 < riel> go to FrontPage, then select "Show Local Site Map" from the "More Actions" dropdown
09:56 < mael> then?
09:56 < riel> then you get a list of all the pages
09:56 < mael> I only have french pages there
09:56 < mael> :\
09:56 < riel> then you should configure your account to see the english language pages
09:57 < riel> did you create an account ?
09:57 < mael> nope
09:57 < mael> but I can :)
09:57 < riel> after that you can simply set the preferred language to english
09:57 < riel> and you'll get all the english language pages
10:09 < mael> riel:  thx it worked
10:10 < riel> np
10:11 < mael> btw it could be nice to fix this i18n problem
10:11 < mael> and at least modify the frontPages (the one you can't modify :))
10:11 < riel> *nod* ;)
10:11 < mael> I think I should send a mail on the ML
10:13 < riel> sounds like a good idea
10:26 < mael> done
10:28 < mael> I just hope the mail is written in something close enough to english so it'll be clear
10:38 -!- hbaum [~hbaum@bi01p1.co.us.ibm.com] has joined #xen
10:47 -!- hollis [~hollis@user-0vvde2g.cable.mindspring.com] has quit [Quit: leaving]
11:09 < riel> there, a little bit closer to having the xen kernels compile again in rawhide
11:10  * riel just fixed up page table handling, pci stuff and /dev/mem handling
11:10 < riel> and the execshield bits may still work
12:29 -!- hollis [~hollis@pixpat.austin.ibm.com] has joined #xen
12:38 -!- hollis [~hollis@pixpat.austin.ibm.com] has quit [Quit: leaving]
12:48 -!- hollis [~hollis@pixpat.austin.ibm.com] has joined #xen
14:08 -!- anticw is now known as cw
14:08 < cw> i should execshield was working with Xen as-is?
14:08 < cw> i don't see any reason why it shouldn't
14:10 < riel> the upstream kernel doesn't have segment based execshield, I think
14:10 < riel> and Xen doesn't have PAE
14:10 < riel> but with the patches in the Fedora RPM, execshield works
14:10 < cw> ok, i think im confusing things here
14:10 < cw> what is arjan's thing called?
14:10 < cw> the randomization thing
14:11 < riel> address space randomisation
14:11 < riel> also, I think the upstream kernel has NX based execshield
14:11 < riel> but that only works in PAE mode
14:12 < cw> ok, im getting my nouns confused ... ignore me :)
15:04 -!- Lampy [~jon@dynamic-62-56-43-157.park-s46b.dslaccess.co.uk] has joined #xen
15:09 -!- Lampy [~jon@dynamic-62-56-43-157.park-s46b.dslaccess.co.uk] has left #xen []
15:22 -!- matta [~matta@69.93.28.254] has joined #xen
15:24 -!- knewt [~jmb@zeus.pimb.org] has joined #xen
15:28 -!- buggs [~noidentd@n80-237-228-135.cnet.hosteurope.de] has joined #xen
15:29 < buggs> hoi
15:30 < buggs> riel, what is your involvement in xen ?
15:30 < matta> buggs: he is the fedora/rawhide maintainer of it
15:30 < matta> and holds redhat's interest in it at heart
15:30 < buggs> ah
15:35 < riel> I'm working on two main areas:
15:36 < riel> 1) make sure Xen gets integrated into Fedora - ie. it should work out of the box, without the user having to do anything special
15:36 < riel> 2) test the latest features in Xen, and fix any bugs I run into
15:36 < riel> the idea is to work together with upstream development (ie. the xen-devel mailing list) as closely as possible
15:37 < riel> part of (2) is making it easier for users to try out Xen
15:37 < riel> after all, the more users there are, the faster we'll find bugs ;)
15:37 < buggs> they made it your number one job ?
15:37 -!- greenrd [~greenrd@66-188-75-198.cpe.ga.charter.com] has joined #xen
15:37 < riel> buggs: yes
15:38 < riel> though it's unfair to blame "they" - I was involved with it too ;))
15:38 < buggs> have you had contact to matt dill on lately ?
15:38 < buggs> hehe
15:38 < buggs> *dillon
15:39 < riel> not in the last year
15:40 -!- matta-lt [~matta@69.93.28.254] has joined #xen
15:49 < mikegrb> riel: <3
15:51 < riel> looks like the port from 2.6.11 to 2.6.12-rc1 is almost done
15:51 < riel> that was a LOT of work ...
15:51 < riel> well - vmlinuz compiles
15:51 < riel> it might still blow up in the modules
15:51 < mikegrb> riel doesn't love me :<
15:52 < riel> I'm doing whatever I can to get kernel-xen[0U] back into rawhide this week
15:52 < riel> if only because FC4 test 2 freezes next week ;)
15:59 -!- aliguori [~anthony@cpe-70-112-81-91.austin.res.rr.com] has joined #xen
16:06 -!- jeroney [~jeroney@pixpat.austin.ibm.com] has joined #xen
16:33  * riel wonders if he should put a reference to this #xen in his email signature
16:35 -!- matta-lt [~matta@69.93.28.254] has quit [Quit: Sphinx may be image of a failed attempt to cross humans with lions.]
16:47 < aliguori> so what makes oftc better than freenode?  i thought the whole point of freenode was to provide an irc network for open source projects
16:48 < aliguori> i think a lot of people just used to freenode.. not that it's better.
16:49 < schweeb> I think there were some differences in opinion... some people didn't like some of the political aspects of freenode or something and split off
16:49 < aliguori> ahh
16:49 < aliguori> so oftc is freenode with less politics?  that's nice
16:50 < riel> I moved my server to OFTC a few years ago
16:50 < aliguori> :-)  freenode does do a lot of strange motds and broadcasts
16:50 < knewt> i'm already on rhizo, freenode and efnet, so oftc is just another for me
16:50 < aliguori> gees, i didn't realize oftc was that old
16:50 < schweeb> yea, I think a lot of people had probs with lilo
16:50 < schweeb> and other stuff
16:51 < riel> around the time when lilo was arguing that we should all pay him, because he removed the oper rights from the volunteers who used to run the network
16:51 < riel> I think that was in the (northern) summer of 2002
16:52 < aliguori> interesting.  had no idea
16:53 < jeroney> ok lets get some conversation...so it looks like in xen/include/xen/mm.h  page_scrub_lock is declared as an extern, then in xen/common/page_alloc it declared static...gcc 3.4 has no problem with this...gcc 4 does (and it should)..anyone see any reason why this should be static and declared in the beginning of the file..I don't see one!?
16:53 < aliguori> xen has killed my irc habits.. i end up rebooting my systems to frequently.
16:53 < aliguori> jeroney: is this in unstable?
16:53 < jeroney> aliguori: yeah
16:54 < riel> jeroney: if it isn't called from any other file, I think it should become static
16:54 < riel> if it is called from other files, we should remove the static
16:54 < riel> cleanup patches like that are useful
16:54  * knewt found one very useful side effect of xen a little while back. the ability to trigger a reboot over the serial console even when dom0 has completely died on you :)
16:54 < jeroney> what about the extren decleartion in mm.h ...one or the other has to go
16:55 < jeroney> riel: let me look quick like
16:55 < riel> if the function is not called from any other file, the extern declaration can disappear from mm.h
16:55 < aliguori> jeroney: looks like it's a little more complicated than you describe
16:56 < aliguori> page_scrub_lock is a spinlock for page_scrub_list
16:56 < aliguori> both are declared as extern in mm.h
16:56 < aliguori> page_alloc.c and domain.c both use page_scrub_list
16:57 < aliguori> domain.c doesn't actually modify the list so i'm guessing it's safe to not hold that lock
16:57 < jeroney> yeap I see that now
16:57 < aliguori> i'd say that lock shouldn't be static just instead of removing the extern declaration
16:57 < aliguori> b/c if someone else uses page_scrub_list, they should have access to the lock in case they modify it
16:58 < aliguori> my guess is it was originally static and someone needed to access it so they removed the static keyword and forgot to remove static from page_scrub_lock
16:58 < jeroney> that could be and gcc 3 never complained
16:59 < aliguori> let me pull up an annotated version and see how it happened
17:01 < aliguori> hmm, keir just added this on the 22nd
17:01 < aliguori> i'd ask him
17:01 < jeroney> of this month
17:01 < aliguori> yup
17:01 < riel> aliguori: even if you don't modify the list, you still need to hold the lock when you read the list
17:01 < riel> otherwise somebody could destroy the list item you're looking at, and your pointer would point to lala-land
17:02 < aliguori> eh, it depends.. i don't know the code well enough
17:02 < aliguori> it's possible there's no way the list could be modified.. i think page scrubbing happens at well defined time periods
17:05 -!- Method [~Method@stanford.columbia.tresys.com] has quit [Ping timeout: 480 seconds]
17:08 < aliguori> so that's a check to see if the page scrubbing list is empty or not.  if it's empty, we return ENOMEM durnig domain creation.  if it's not, we return EAGAIN (which means we have to wait until the page scrubber is scheduled)
17:09 < aliguori> i think the worst case that could happen is that we return EAGAIN at the exact wrong moment when there is really a free page.  that seems harmless
17:09 < riel> oh, it may not be buggy in this case, but defensive programming never hurts
17:10 -!- niv [~niv@bi01p1.co.us.ibm.com] has joined #xen
17:10 < aliguori> i agree
17:11 -!- riel is now known as unriel
17:12 < aliguori> jeroney: have you gotten xen to boot into multiuser mode on x86_64 yet?
17:15 < jeroney> aliguori: sadly no..havn't gotten any updated patches either ... let me make contact today and see if I can get the hook up 
17:43 -!- surriel is now known as riel
17:51 -!- Method [Method@pcp0010742058pcs.howard01.md.comcast.net] has joined #xen
18:01 -!- hollis [~hollis@pixpat.austin.ibm.com] has quit [Quit: leaving]
18:10 -!- hbaum [~hbaum@bi01p1.co.us.ibm.com] has quit [Quit: Client exiting]
18:48 -!- niv [~niv@bi01p1.co.us.ibm.com] has quit [Quit: ChatZilla 0.9.61 [Mozilla rv:1.7.5/20041217]]
18:53 -!- Mark [~Mark@maw48.kings.cam.ac.uk] has joined #xen
18:55 < Mark> Hey ho
18:55 < Mark> Anyone about?
18:55 < knewt> 'lo
18:56 < Mark> ah hello
18:56 < Mark> just saw your mail about this new fangled IRC business ;-)  so i thought it check it out
18:56 < riel> hi Mark, good to see you online
18:57 < riel> looks like I have xen compiling again in rawhide
18:57 < riel> the xen kernel, that is - the hypervisor was easy ;)
18:57 < knewt> i must have been on another channel on oftc at some point for a while, as i had already registered my nick in the past :)
18:57 < Mark> Hi Rik
18:58 < rharper> hey Mark, glad you could make it
18:59 < Mark> knewt: scary when one happens across evidence of having done things you can't remember - i do it all the time!
18:59 < Mark> rharper: you're Ryan, right?
19:01 < rharper> yeah
19:01 < Mark> nice to meet you
19:01 < rharper> nice to meet you as well
19:01 < Mark> although i guess i might have seen you in Cambridge...?
19:02 < rharper> heh, no I havent made it out there yet
19:02 < riel> will you be there next week ?
19:02 < rharper> riel: nope
19:02 < riel> ;(
19:02 < riel> how about OLS ?
19:02 < rharper> riel: no invite for me =(
19:03 < riel> OLS isn't an invitation-only event
19:03 < rharper> riel: I hope so.  We have a proposal accepted.  working on that 
19:03 < knewt> unfortunately i can't make it to the Xen Summit :(
19:03 < riel> I'm looking forward to drinking a pint with you guys in Ottawa
19:04  * rharper is all for drinking pints
19:04 < riel> US pints are kind of sucky, though ;)
19:05 < rharper> so I've heard
19:08 < riel> 470 or so ml
19:08 < riel> compared to 560 or so ml for UK pints
19:08 < knewt> which is disappointing. but things requiring money are off the agenda until i manage to get a job
19:08 < riel> knewt: can't blame you
19:10 < knewt> and when you're talking a good 140 pounds just on travel costs to get over to the uk and back...
19:14 < riel> that's a cheap ticket
19:14 < riel> it's around $500 from Boston to London and back
19:14 < knewt> ryanair
19:14 < knewt> this is from germany
19:14 < riel> ahh ok
19:14 < knewt> i'd love to be back in the uk
19:15 < knewt> interviewed for a job in oxford a couple of weeks back, but not heard yet
19:23 < riel> yay, the xen0 kernel + all modules finished compiling
19:35 < Mark> riel: with gcc4?
19:35 < riel> yes
19:41 < riel> gcc4 so far seems to be picky, not buggy
19:41 < riel> though optimisation differences are triggering some xen bugs that gcc-3.4 happens to skip over
19:41 < Mark> sounds promising.  does the resulting kernel behave correctly yet?
19:41 < riel> it seems to work here
19:42 < riel> it's frequently recompiling itself, on an SMP domain 0
19:42 < Mark> nice!
19:43 < riel> call me reckless, but I see nothing wrong in following xen-unstable, with gcc-4 and SMP guests ;)
19:45 < riel> oh, and closely following the 2.6 upstream kernel
19:53 < Mark> you like to live on the edge :-)
19:54 < riel> well, if something breaks I'll just fix it
19:54 < riel> somebody needs to clear the path ;)
19:59 < Mark> indeed
20:00 < Mark> when is FC4 intended for release?
20:00 < riel> besides, rawhide seems to work fairly well despite the "follow the newest version everywhere" strategy ;)
20:00 < riel> may or june
20:01 < Mark> what xen are you intending to release with?  still unstable?
20:01 < riel> if 3.0 isn't out yet, I won't have much choice ;)
20:01 < riel> chances are I'll keep updating xen in FC4 until 3.0 is out
20:02 < riel> then I'll follow 3.0 updates in FC4, while tracking -unstable again for FC5
20:02 < Mark> true :-)  there might be a 2.1 somewhere along the way
20:02 < Mark> but i think we're still waiting for the platform init code to arrive for that...
20:02 < riel> yeah, but 2.1 doesn't have SMP, x86-64 or VT
20:02 < Mark> no, *must have acronyms* :-)
20:04 < riel> it would be nice to have a Linux distro support VT, just before the hardware hits the market
20:04 < riel> I want Linux to be first with this ;)
20:07 < Mark> yup, that'd be well cool.  it'll defy the hardware support argument.
20:08 < Mark> used to have hardware support issues, now it supports hardware before you can buy it :-)  much better
20:09 < rharper> heh
20:16 < riel> "if you want to use that, you'll want Xen"
20:17 < jeroney> hey that whould be a good line for Xen
20:21  * riel wonders if "extreme compile waiting" counts as an extreme sport
20:22  * rharper broke down and set up distcc pools for the two kernel compiles 
20:22 < jeroney> rharper: that super machine under your desk wasn't enough
20:23 < rharper> heh, well, I didnt have that until a few days ago
20:23 < rharper> it helps the pool out nicely though
20:23 < rharper> have it running distcc in a 32-bit chroot 
20:23 < jeroney> hehe
20:23 < rharper> accepting up to 32 cons since its rather beefy
20:24 < rharper> mmm, full xen builds (minus docs) in 3m39s  
20:24 -!- jeroney [~jeroney@pixpat.austin.ibm.com] has quit [Quit: Leaving]
20:47 -!- DEac- [~deac@xdsl-213-196-201-126.netcologne.de] has quit [Ping timeout: 480 seconds]
20:54  * Mark yawns
20:54 < Mark> gnite all
20:55 -!- Mark [~Mark@maw48.kings.cam.ac.uk] has left #xen [Kopete 0.10 : http://kopete.kde.org]
20:58 -!- rharper [~rharper@pixpat.austin.ibm.com] has quit [Quit: Leaving]
20:58 -!- DEac- [~deac@xdsl-213-196-206-117.netcologne.de] has joined #xen
21:10 < riel> ok, it's assembling the RPMs ;)
21:10 -!- muli_ [~muli@nesher3.haifa.il.ibm.com] has quit [Ping timeout: 480 seconds]
21:22 -!- rusty [~rusty@bh02i525f01.au.ibm.com] has joined #xen
21:23 -!- niv [~niv@bi01p1.co.us.ibm.com] has joined #xen
21:49 -!- niv [~niv@bi01p1.co.us.ibm.com] has quit [Quit: ChatZilla 0.9.61 [Mozilla rv:1.7.5/20041217]]
21:56 -!- greenrd [~greenrd@66-188-75-198.cpe.ga.charter.com] has quit [Ping timeout: 480 seconds]
22:06 -!- greenrd [~greenrd@66-188-75-198.cpe.ga.charter.com] has joined #xen
22:21 -!- crisen [~crisen@spank.terdmonk.com] has joined #xen
22:43 -!- Bluefox [~bluefox@pcp484971pcs.whtmrs01.md.comcast.net] has quit [Ping timeout: 480 seconds]
22:54 -!- Gollum [~gollum@pcp01411681pcs.phnixv01.pa.comcast.net] has joined #xen
22:55 -!- Bluefox [~bluefox@pcp484971pcs.whtmrs01.md.comcast.net] has joined #xen
22:55 < Gollum> is there anybody here?
22:55 < rusty> No.
22:57 -!- Gollum [~gollum@pcp01411681pcs.phnixv01.pa.comcast.net] has left #xen []
23:00 -!- Gollum [~gollum@pcp01411681pcs.phnixv01.pa.comcast.net] has joined #xen
23:04 -!- cw [cw@adsl-67-124-117-212.dsl.snfc21.pacbell.net] has left #xen []
23:04 -!- cw [cw@adsl-67-124-117-212.dsl.snfc21.pacbell.net] has joined #xen
23:04 -!- mode/#xen [+o cw] by ChanServ
23:05 -!- Gollum [~gollum@pcp01411681pcs.phnixv01.pa.comcast.net] has left #xen []
23:13 -!- Bluefox [~bluefox@pcp484971pcs.whtmrs01.md.comcast.net] has quit [Quit: Linux sucks, it doesn't run all my viruses.]
23:18 -!- mode/#xen [+oo rusty riel] by cw
23:29 -!- Bluefox [~bluefox@pcp484971pcs.whtmrs01.md.comcast.net] has joined #xen
--- Log closed Wed Mar 30 23:59:00 2005