[12:10] lamont: please enable CONFIG_DISCONTIGMEM in all hppa configs (to support more than 4GB RAM). kylem says it's ok on all configs, debian has it [12:11] (more than 3.75GB to be precise) [12:13] Total Memory: 3840 Mb [12:13] grukble === lamont needs more RAM, dammit [12:15] T-Bone: can 32-bit machines have > 4GB RAM? [12:15] lamont: sure, if you run a 32bit kernel [12:16] doh [12:16] should ask in #parisc, they're discussin that issue [12:16] actually get 2^44 bits of address space for physical [12:16] but we *really* want it for -34 for the builders ;) [01:22] fabbione: next ABI event, hppa should have CONFIG_DISCONTIGMEM=y in all 4 configurations. (causes an ABI event) [01:44] fabbione: and is the abi checker smart enough that one can just change the abi version, and it just generates the new files and ignores the old? that'd be cool.:-) === mainer [~mainer3@bb-205-209-66-252.gwi.net] has joined #ubuntu-kernel [03:53] hi: new to ubuntu: are there any people or sites that provide .deb pkgd kernel-images? [03:54] anyone home? === mainer [~mainer3@bb-205-209-66-252.gwi.net] has left #ubuntu-kernel ["Leaving"] [03:59] goodness [03:59] hah === kylem groans. [04:09] i found out the problem i was having yesterday with mdz's help [04:10] the initrd doesn't load the chipset specific ide driver so it ends up getting loaded after the generic one [04:50] morning [04:50] morning fabbione [04:51] evening [04:51] fabbione: we planning to keep the 2.6.10 tree alive for breezy, or roll to 2.6.11? [04:51] i thought it was 2.6.12 [04:51] 2.6.13!! [04:52] or will they skip that one out of superstition? [04:52] lamont: if version X has an abi change, it ignores X-1, but we still need to generate the abi for X+1 [04:52] right [04:52] i was thinking to skip .11 completely [04:52] according to dilinger is a crappy kernel [04:53] and it would be heaps load of work to port to .11 [04:53] although I must admit that I'm tempted to let -34 be an hppa-abi event and bastardize things to keep it at 5. It's not even in the archive and all that... :-) [04:53] specially the security fixes that are not in .11 [04:53] right [04:53] 12 sounds good to me [04:53] lamont: i have no isseus if you want to modify hppa [04:53] and that can go in as soon as breezy opens, yes? [04:53] nobody other than you and T-bone are using it right now [04:53] fabbione: is ABI event... I have to debate whether I care enough to worry about it. [04:53] and it has never been published [04:53] right [04:54] lamont: heh its giong to be a party once breezy opens ;) [04:54] zul: zact;y [04:54] zactly, even [04:54] lamont: we will go .12 as soon as .12 is released [04:54] doh [04:54] lamont: that will take us more or less 2/3 days to prepare [04:54] no 2.6.11.really.2.6.12? [04:55] is there any reason why you need .12 so badly? [04:55] when does breezy open? [04:55] I rather expect after UDU, truthfully [04:55] dilinger: a few days after hoary is released [04:55] lamont: i think james will be faster this time :) [04:55] prior to really opening breezy, there are plans to rebuild everything with gcc-4.0... [04:56] not sure if that's going to lockstep or not. [04:56] ah right [04:56] gcc-4 transition [04:56] iirc there are some kernel build issues with gcc-4 [04:56] so it's really a question of "what are the plans between hoary-release and breezy-start... [04:56] otherwise, breezy could start _now_ [04:56] zul: yes [04:56] zul: there are build issues with lots of stuff.. [04:56] zul: we can't stay with 2.6.10 [04:57] the full build wouldn't go into the actual archive [04:57] it will probably FTBFS [04:57] heh one of should try building with gcc-4 :) [04:57] i can do that right away :) [04:57] ahhh...i was going to do that... [04:57] :) [04:57] but iirc the miscompilations were on arch != i386 [04:58] given that the current gcc-3.4, gcc-4.0 are FTBFS on hppa..... :( [04:58] zul: go ahead and do it [04:58] goody [04:58] hppa sucks :) [04:58] fabbione: feh [04:58] they build on sparc [04:58] java assertions, doko thinks he knows what the fix is, won't go in before release [04:58] does ubuntu actually work on sparc now/ [04:58] if I get bored this week, I'll test his fix === lamont has 3 ubuntu/hppa machines in his house [04:59] (2 buildd's and a router) [04:59] anyway, bed time for me., [05:00] zul: yes it does === dilinger slaves over a hot 2.6.11 [05:01] lamont: don't rush.. i woke one hour earlier than usual :) [05:01] cool...i have an ultra1 lying around [05:01] zul: unfortunatly it won't install [05:01] because elmo stopped rsyncing sparc.u.c [05:01] blah [05:01] due to some server load problems [05:02] he promised me to move it somewhere else [05:02] but afaik he didn't yet [05:03] ok...well i can wait...somtimes patiently [05:03] perhaps i can convince elmo to do a pulse todya [05:03] let see [05:04] lamont: sometimes soon i will need some help to add breezy/hoary-security/hoary-updates to the buildd. [05:04] because tbh i never had the need to build more than one release [05:05] and i can't remember crap on how to do it :) [05:10] i think it might be time for me to go to bed...later folks [05:11] lamont: i was thinking that it could actually be an option to open a 2.6.12 branch [05:12] creating an orig from bk [05:12] but that means that the orig must be used internally and not published [05:12] becuase we will replace it with the real one once it's out === cc [~cc@c210-49-121-44.eburwd3.vic.optusnet.com.au] has joined #ubuntu-kernel === Seveas [~seveas@ksl403-uva-131.wireless.uva.nl] has joined #ubuntu-kernel === zul [~chuck@198.62.158.205] has joined #ubuntu-kernel [02:43] morning [02:45] fabbione: other option is to open a 2.6.12 branch with a 2.6.11pre12.orig.tar.gz === makx [~max@baikonur.stro.at] has joined #ubuntu-kernel [02:47] lamont around? [02:47] yo [02:47] i wanted to ask question how you'd think daily builds should be properly set up. [02:48] doing it d-k wise but it should be same, i guess. [02:48] fabbione hinted me that you may have usefull pointers on last u-k meeting. [02:49] i've played around a bit and have a skript that svn co d-k and builds images atm. [02:51] if the plan is to have these in the archive, then it's simply a matter of uploading fresh source every day. Ideally, it'd have an orig.tar.gz that had nearly everything, and then just a diff.gz for the day, until such time as the diff got large enough and we cut a new orig.tar.gz [02:51] less bits to up/down load that wya [02:52] fabbione: cutting even an internal-only orig.tar.gz that's incorrect is a bad plan... [02:52] lamont: aah you are building u-k directly out of source. [02:53] forgot that bit. [02:54] experimental still sounds like a nice target for those. [02:55] ok, so i'll have to work on the unifying of the d-k packaging. [02:58] lamont: ok thanks for the hint. [03:00] kids->school [03:01] figgin xp === makx [~max@baikonur.stro.at] has left #ubuntu-kernel ["bella] [03:27] oh this sucks..ply a matter of uploading fresh source every day. Ideally, it'd have an orig.tar.gz that had nearly everything, and then just a diff.gz for the day, until such time as the diff got large enough and we cut a new orig.tar.gz [03:27] lamont less bits to up/down load that wya [03:28] frig...stupid gnome...sorry [03:28] what i really wanted to say dpkg-architecture: warning: Couldn't determine gcc system type, falling back to default (native compilation) [03:34] uh...nevermind [03:39] ftbs with gcc-4.0 === lamont returns [04:09] hey lamont [04:27] fabbione: you around? === lamont changes the abicheck code [04:34] heh that will get his attention [04:36] grumble. worse than I wanted it to be [04:49] changing the ${arch}.ignore case to still do diffs [04:49] if it can, that is. [04:53] lamont: i am now [04:54] i completely crashed [04:57] heh [04:57] added some code to the .ignore case to still do the diff if there's both an abi file and the .ignore file [04:58] lamont: i make no objections if it has been tested :) [04:58] test build is running now [04:59] cool [04:59] well, actually, the .dsc is packaging now. [04:59] if everybody is happy and the situation is normal i am off again to enjoy a bit of sunshine [04:59] and i will be back in an hour or so [05:00] later [05:00] later :) [05:00] toodles [05:04] fabbione: of course, the new code is so that hppa can have a well documented non-abi event. :-) We're testing to see if there are more than 2 users, you see... :) [05:06] heh...its not like you would break something on i386 like i usually do and people notice these things [05:13] zul: this is intentionally ignoring the abi change that really should create 2.6.10-6, but won't. [05:13] ah i c [05:17] yeah. both t-bone and I agree that we should, so that's 100% of the known user community of our non-existant port. [05:17] well, s/non-existant port/non-advertised archive/ [05:18] which means that we have abi files, we're breaking abi compatibility on that archtecture only, and specifically ignoring that fact. [05:18] lalalalalala [05:18] heh [05:19] crap..i have a dentist appointment today [05:23] zul: that'd be protcologist. :-) [05:24] speaking of protcologist...nah...i dont think you wanna know :) [05:25] really crappy work [05:25] hehe [05:25] gotta wonder what makes a med student _decide_ to specialize that direction... [05:26] fetish maybe === lamont bets money [05:27] could be a little from column a, a little from column b === lamont grumbles [05:42] gotta figure out where the hppa.ignore file gets removed [05:46] ok. empty files and debuild -S don't seem to mix. [05:58] lamont: it's the kernel make clean target that kills empty files [05:58] Unpacking libmpfr-dev (from .../libmpfr-dev_2.1.0-2ubuntu1_sparc.deb) ... [05:58] dpkg: error processing /opt/sparcbuildd/chroots/chroot-hoary/var/cache/apt/archives/libmpfr-dev_2.1.0-2ubuntu1_sparc.deb (--unpack): [05:58] trying to overwrite `/usr/share/info/dir.old.gz', which is also in package texinfo [05:58] dpkg-deb: subprocess paste killed by signal (Broken pipe) [05:58] hmmm [06:00] i guess i have some crap in the buildd cache [06:00] was old texinfo bug did that. [06:01] yeah i remember that [06:01] but gcc-4.0 has been built a while ago [06:02] i mean that i did build gcc-4 other times [06:02] without problems [07:06] fabbione: ccache me thinks [07:07] zul: nah. i am talking about building gcc-4.0 [07:07] ccache can cope with a different compilare [07:07] compiler [07:07] it did build before today [07:07] fabbione: i got the same with export CC="ccache gcc-4.0" [07:13] zul: so 2.6.10 doesn't build with gcc-4.0 [07:13] nope [07:13] how suprising :P === fabbione isn't [07:13] neither am i, just being sarcastic ;} [07:14] sure sure [07:14] T-Bone: i could understand that... you are french :P [07:14] fabbione: and? Aren't you italian? :) [07:14] yes i am [07:14] what so different about us then? :) [07:15] that you are french and i am not? [07:15] the way we cook spaghettis? :) [07:15] no.. you don't cook spaghettis [07:15] LOL [07:15] here we go =) [07:15] :) === T-Bone must schedule yet another Italian trip for next summer. That country is probably the one I'd want to live in, should I have to leave .fr ;] [07:16] fabbione: dies in drivers/acpi [07:17] fabbione: that has to be a compliment! =] [07:17] T-Bone: why would you leave .fr because you are too french? [07:17] hehe [07:17] T-Bone: i would never go back to live in italy.. why would you do that? [07:17] zul: no. I was hypothetising :) [07:17] zul: ok [07:17] ah [07:17] fabbione: why wouldn't you? [07:18] T-Bone: because Italy is only a very nice place to have holidays [07:18] LOL [07:18] that's about it [07:18] working there.. sucks [07:18] ah [07:18] internet there .. sucks [07:18] can't tell [07:18] government.. questionable [07:18] lol [07:19] bersecloni? [07:19] all the rest != some resturant = the sucks [07:19] berlusconi [07:19] yet, food is cool, weather is cool, ragazza cool, monuments cool, landscape cool... ;) [07:19] lazio cool ;) [07:19] T-Bone: aren't you married? [07:19] zul: ahahha [07:19] fabbione: hell no! [07:19] ;) [07:20] T-Bone: if you really want to get laid to death you should go and visit siciliy [07:20] just remember to get a fake id before going there [07:20] i'm 24 dude. I'm not gonna emprison myself while I'm at the peak of my sex appeal ;) [07:20] not going to touch it.. [07:20] i went to sicily ;) [07:20] otherwise one of these hot chicks might show up at your door in france with a mini T-Bone [07:20] LOL [07:21] dude... 24? [07:21] hell you are kid [07:21] fabbione: that's what they say in movies, is that so true? [07:21] you are a young'un t-bone [07:21] T-Bone: i lived in sicily for 4 years... [07:21] fabbione: sure. I'm trying to grow chest hair. Ask lamont-away =) [07:21] well it is true in certain parts of siciliy [07:21] ahahahaha [07:22] i guess this might be true in certain parts of Corsica as well :) === ..[topic/#ubuntu-kernel:fabbione] : Ubuntu kernel development discussion | http://www.ubuntulinux.org/wiki/KernelTeam | http://people.u.c/~lamont/Archives/kernel-team@ubuntu.com--2005/ stable: kernel-debian--pre34--2.6.10 playground: kernel-debian--experimental--2.6.10 | There are no kernel bugs.. only broken hardware | T-Bone: I'm trying to grow chest hair. [07:22] hah [07:22] well corsica is france territory and i have never been there [07:23] fabbione: you're late on topic. Lamont has put that one in the very early days of #u-kernel ;] [07:23] fabbione: never been to france? [07:23] yeah but i wasn't here [07:23] ah ok :) [07:23] T-Bone: oh yeah.. i have been to france several times [07:23] ah! [07:23] specially in Paris [07:24] that's just because the best airshow in EU is/was there [07:24] said to be the most beautiful city in the world... [07:24] i been to paris once [07:24] T-Bone: bullshit.. Rome > Paris [07:24] airshow? [07:24] fabbione: lol, no way :) [07:24] T-Bone: yeah.. Le Bourget [07:24] fabbione: you don't have the Eiffel tower in Rome ;) [07:24] (sorry for the spelling) [07:24] fabbione: ah! I see [07:24] spelling is correct btw :) [07:24] T-Bone: you don't have the colosseum in Paris [07:25] fabbione: right. we don't collect ruins ;) === T-Bone ducks! [07:25] T-Bone: that's why it is still up and rocking after 2000 years.... [07:25] ehhe [07:25] anyway let's cut the speech france <-> italy here [07:26] otherwise i will get T-Bone to commit suicide for being born there [07:26] LOL [07:26] in your widest dreams... =) [07:26] +l [07:26] T-Bone: ok.. listen to this story as a starter [07:27] you are probably too young to remember [07:27] ? [07:27] you know that Italy and France always fight for who has the best wine? [07:27] it's kinda of a friendly war ... [07:27] but you can smell it [07:27] fabbione: I don't. The answer is obvious. I wasn't aware there was a debate 8) === T-Bone laughs like evil =] [07:28] well we always accused France to steal Italian wine, mix it with some local flavour and sell it with a different name [07:28] France of course always negated that [07:28] berk [07:28] should some people do that in France, that's not wine they're selling :P [07:28] local flavour would be anti-freeze [07:29] all of a sudden, in Italy, they found a big amount of poison in italian wine [07:29] due to some kinda of infection... [07:29] probably generated by some chemical shit [07:29] fabbione: it is obvious some producers sell something that should barely be called cat piss. What's even more unfortunate is that there are people buying it as "French wine" :P [07:29] well.. of course we closed down the selling of wine [07:29] fabbione: no kidding? [07:30] now.. how come some of the major France wine producers did the same? :) [07:30] T-Bone: no i am really serious [07:30] i am not kidding [07:30] wow [07:30] oh as a side note [07:30] looks pretty much unbelievable to me [07:30] the part of italy that was damaged [07:30] at least for the "Grands Crus" [07:30] was the north-west.. close to the french border line [07:31] Piemont? [07:31] now.. really *CASUALLY* [07:31] all the southern-east french wine were not distributed [07:31] yeah Piemonte [07:31] ;) [07:31] good wine in Piemonte, that's for sure :) [07:31] oh.. of course.. France did NEVER import wine from italy.... [07:31] heh [07:32] anyway [07:32] i think it's time to go and help my wife [07:32] and get some food [07:32] cya tomorrow guys [07:32] fabbione: well i don't know much about that story, but the best wines in France aren't southern-east ones. Bordeaux and Bourgognes are respectively southern west and middle-east, so to speak [07:32] fabbione: heh, cya. I'm not done with you yet ;) [07:33] T-Bone: well still pretty close to france :) [07:33] T-Bone: next one will be italian kitchen vs french cousine [07:33] :P [07:33] LOL [07:37] french football vs italian football? [07:38] lol [07:38] i don't like football, i'd be a bad advocate :) [07:39] OTOH, french rugby vs italian rugby would prove interesting... 8) [07:41] i dont really pay attention to rugby anymore [07:45] fabbione: there is alot of warnings like this include/linux/skbuff.h:1017: warning: pointer targets in passing argument 2 of 'csum_and_copy_from_user' differ in signedness [07:45] that should probably be fixed as well === lamont_r [~lamont@m010f36d0.tmodns.net] has joined #ubuntu-kernel [09:22] WARNING! ABI check override for hppa has been detected, and there were changes! [09:22] woot! [09:26] lol [09:27] T-Bone: that means that my check changes worked.... [09:27] indeed [09:28] fabbione: actually, I have another quesion about the abi check.... [09:28] normal use case is going to be that the user installs the new kernel and then doesn't reboot... [09:28] hence adding symbols should mean that the new modules won't load in the old kernel, no? [09:29] :q === lamont_r needs focus_mode: eyes [09:40] wohoo...helium balloons [09:41] zul: get a real job, slacker [09:42] i tell you i work for the government.. [09:42] zactly! :0-) [09:42] lol [09:42] however i wouldnt mind being paid to work on ubuntu full time :) [09:43] CONFIG_DISCONTIGMEM committed. I feel so dirty [09:43] lamont_r: i like you very much, you know? Especially when you're dirty ;) [09:43] ok...take it to #ubuntuhppaorgy [09:43] what better way to find the 3rd user, eh? [09:44] lol [09:44] lamont_r: i've finished reviewing my howto, and will advertise it soon [09:45] mjg59: why does the kernel think ACPI says I have no batteries? [09:45] T-Bone: and you fixed your /boot vs / issue? [09:45] lamont_r: I can't tell. I'll tell you in next upload [09:45] s/upload/upgrade/ [09:46] lamont_r: all i can say is that it doesn't work within the chroot, that's for sure [09:46] which brings us the question of "how is it supposed to work in the (yet hypothetical) installer" [09:47] grep link_in /etc/kernel-img.conf [09:47] link_in_boot = yes [09:47] have that? [09:47] yes [09:47] link_in_boot = Yes === jbailey [~jbailey@CPE000ded9d787c-CM014260028338.cpe.net.cable.rogers.com] has joined #ubuntu-kernel [09:52] hey jbailey [09:52] Heya Chuck [09:52] how is it going? [09:53] lamont_r: is there any way to educate popcon about an http proxy? [09:53] http_proxy=? [09:53] zul: Good! I got a place. =) [09:53] lamont_r: in popularity-contest.conf? [09:53] dunno [09:53] ah [09:53] jbailey: oh yeah wehre? [09:53] because i can't get it to work behind a proxy, you see... [09:54] zul: In the plateau/ [09:55] cool [09:55] jbailey: dude, i've started reading "WYFL", it's fuckin amusing! ;) [09:55] jbailey: i'm learning a handful of useful expressions ;) [10:03] later [10:06] WYFL? === lamont_r saw a t-shirt this past weekend that simply said 'WTFWJD?' [10:07] Watch Your Fucking Language [10:07] subtitled "How to swear effectively, explained in explicit detail and enhanced by numerous examples taken from everyday life" [10:08] URL? [10:08] lamont_r: btw, seen what I said in #parisc? I wonder how bad it would be to run a debian 2.6.8 kernel on Ubuntu for the time being? [10:08] lamont_r: that's a fuckin book! ;) [10:08] lamont_r: amazon.com ;) [10:08] heh [10:09] T-Bone: probably no worse than any other. [10:09] lamont_r: ok. I guess that i'll do that then, because the j6k don't have GSP and i don't feel like going to school every 4 days or so to power cycle them [10:09] T-Bone: alternatively, figure out what we broke in 2.6.10SMP [10:09] lamont_r: btw, we're down to 2 builders until friday, since one of them died :P [10:09] xorg builds just fine on a UP kernel. [10:10] lamont_r: everything built fine [10:10] lamont_r: i'm almost certain that's the cache_grow panic that stroke [10:10] lamont_r: unfortunately that box is headless [10:10] xorg build died twice on 2.6.10-33.1-hppa64-smp [10:10] T-Bone: crossover serial cables, dude. [10:11] lamont_r: that's what's there. Except minicom isn't always running on the other end of the cable :P [10:11] lamont_r: i've been building xorg on 2.6.12 64bit smp [10:11] well, yeah === T-Bone does a test install of bastardized kernel on ubuntu to check how it behaves [10:14] lamont_r: unfortunately, as i kept bashing on #parisc, our kernel is in a very awful state since 2.6.8.1 :( [10:15] T-Bone: modulo the "expect bug" [10:16] lamont_r: alas === lamont_r thinks maybe -34 with discontigmem will make it die less often... [10:19] lamont_r: all in all i can keep the buildd setup on the A500 to build sensitive packages, unless you manage to get the A500 in DC, and that one to build these [10:20] lamont_r: install procedure advertised on #parisc and linked on the website ;) [10:21] just in time for the abi event. cool. :=( [10:21] who cares, they all know :) [10:21] yeah [10:21] i haven't m-l'd it yet :) [10:22] well, if you do, include the fact that -34 will have abi breakage, so they'll need to reboot after they install [10:22] (hppa, for our listeners here) [10:22] i won't m-l anything before we get the archive in the right place and -34 in it === lamont_r likes that plan [10:22]