=== lamont spams jbailey privately with a list of i386 failures [12:12] I need a faster computer. [12:13] listen(5, 5) = -1 EOPNOTSUPP (Operation not supported) [12:13] this is weird.... [12:15] not socket? [12:15] it is a socket [12:15] socket(blabla); [12:15] bind(foobar) [12:15] listen [12:15] and i get that error [12:20] i guess i am too tired to spot the error [12:20] better i get some sleep [12:20] cya tomorrow [12:20] night fabbione [12:32] hrm.. need a name for 2.6.12-1 sometime === cc [~cc@byte.fedora] has joined #ubuntu-kernel === zul [~chuck@CPE0006258ec6c2-CM000a73655d0e.cpe.net.cable.rogers.com] has joined #ubuntu-kernel [01:56] hey [01:57] howdy [02:02] how is it going [02:25] not bad [02:25] just about to grill catfish [02:25] and got to watch a roommate slice open his finger [02:25] a fully complete night, i'd say [02:25] multisync works now, so I can sync my phone and evo [02:26] the hoary kernel should really have had this small change, since the x40's irda is broken as-is [02:26] but it's fixed in .12rc1 [02:27] anyone doing boot process speedup stuff around? [02:29] i have the house to myself tonight [02:31] fabbione: so our linux-source-2.6.10 package is useless, fwiw... (bye bye debian/ in the clean target...) [02:32] heh i think i missed something [02:32] zul: that was your doing??? [02:32] wha? [02:32] that you missed something? [02:32] oh [02:32] heh [02:34] bleah...im going to go back to what i was diong [02:41] how do you change the concurrency level again? [02:44] duh...never mind [07:07] morning [07:08] lamont: that's the same problem zul had [07:08] you need change the Makefile that does the cleaning [07:15] actually.. it should contain all the changes.. including the Makefile one.. === Seveas [~seveas@seveas.demon.nl] has joined #ubuntu-kernel === cc [~cc@byte.fedora] has joined #ubuntu-kernel === chuck_ [~chuck@CPE0006258ec6c2-CM000a73655d0e.cpe.net.cable.rogers.com] has joined #ubuntu-kernel [01:50] heylo [01:51] fabbione: it builds now you can do a merge [01:52] ahi [01:52] you managed.. [01:53] i am fighting with cman atm [01:53] devfs? [01:53] cluster manager [01:53] kernel side is ok [01:53] cool... [01:53] i had to create a monster to make it working [01:53] hehe... [01:53] but something in userland doesn't work properly yet [01:54] i am going to reboot and test out the new kernel [01:54] write(2, "Cannot bind multicast address: I"..., 48Cannot bind multicast address: Invalid argument [01:54] interesting [01:54] brb === zul [~chuck@CPE0006258ec6c2-CM000a73655d0e.cpe.net.cable.rogers.com] has joined #ubuntu-kernel [01:58] Linux ubuntu 2.6.12-1-686 #1 Fri Apr 15 21:22:21 EDT 2005 i686 GNU/Linux [01:58] no problems so far [01:58] same here [01:58] just don't play with cman until i commit the fixes [01:58] otherwise it's oopsorama [01:58] although dmesg output has changed a bit [01:59] done and done === Seveas [~seveas@seveas.demon.nl] has joined #ubuntu-kernel [02:07] yes i added printk time stamps [02:10] thats cool [02:17] huh? === T-Gone is now known as T-Bone [02:18] printk timestamps in the to-be-shipped kernel? [02:18] hey T-Bone [02:19] hey zul [02:19] T-Bone: only for now... [02:19] ah ok [02:19] we will not put it for final [02:19] *ouf*. Been a bit worried :P [02:19] it will be useful to see order of crashes on loaded systems [02:19] order of crashes? [02:20] if there is an OOPS on a machine, it might happen before another event, but reported afterwards [02:22] cat /proc/cluster/status [02:22] Protocol version: 5.0.1 [02:22] Config version: 1 [02:22] Cluster name: fabbione [02:22] Cluster ID: 25573 [02:22] Cluster Member: Yes [02:22] DIE DIE DIE MY DARLING! [02:22] T-Bone: ph34r my d3bugg1ng sk1llz: [02:23] fabbione: I'm not quite sure about what you mean but heh. My understanding was that the primary use for printk timestamps was to track down long delays between two kernel ops [02:23] https://www.redhat.com/archives/linux-cluster/2005-April/msg00051.html [02:23] and i need to update all our GFS patch set :/ [02:24] this morning it was an OOPS [02:24] fabbione: i'd rather ph34r your w1c|<3d m1|\|d if anything ;} [02:24] now it's a working cluster [02:24] certainly.. using my laptop and a secondary workstation in failover is pointless :) [02:29] gnbd_export -d /dev/hda3 -e testhd [02:29] gnbd_export: created GNBD testhd serving file /dev/hda3 [02:29] YEAHHHH [02:29] ok [02:29] ok === zul [~chuck@CPE0006258ec6c2-CM000a73655d0e.cpe.net.cable.rogers.com] has joined #ubuntu-kernel === jbailey [~jbailey@CPE000ded9d787c-CM014260028338.cpe.net.cable.rogers.com] has joined #ubuntu-kernel [02:38] hey jeff [02:39] hey jb [02:39] i am off for a nap :) [02:39] i will merge later... [02:39] ok.. === fabbione is way too tired to get it right now [02:39] liar! [02:41] Heya Chuck, Fabio! [02:41] fabbione: You get up for work when I go to sleep and when I show up in the morning you take a nap ;) [02:41] jbailey: eheheh [02:41] fabbione: Before you go - merge orgy proceeding okay for you? [02:42] im off to tim hortons later though [02:42] jbailey: build orgy and yes :) [02:42] Right, sorry. =) [02:42] Good. Glad to hear it. =) [02:42] i am going to play kernel merge later [02:42] there are a few FTBFS due to toolchain not properly alligned [02:42] but we will get to that [02:43] Makes sense. I need to finish some stuff for silbs, then I will do either cdbs2 or initramfs bits. Dunno which. [02:43] But for now, one bug report on metacity and then video games. =) [02:44] eheheh [02:44] later === T-Bone is now known as T-Gone === dilinger crosses his fingers for cdbs2 [03:28] dilinger: Me too. [03:29] dilinger: I'm getting to the point where I have trouble looking at the cdbs1 code base. [03:29] But first! [03:29] Rise of Nations! [03:53] hehe [03:53] i really want to start working on cdbs2, but i *need* to get silo fixed first [03:54] and that means reading through ieee std 1275-1994 and the sparcv9 manual [03:54] which i've been putting off :) [03:59] slacker :) === cc [~cc@byte.fedora] has joined #ubuntu-kernel [05:04] fabbione: apt-get install linux-source-2.6.10; fakeroot debian/rules clean -> we missed that, eh? [05:53] bbl [06:02] re [06:02] lamont: not that i know off... that one is created copying the patched source to another directory [06:16] zul: is your archive down again??? [06:19] oh hell it is [06:51] lamont: any news on the hppa archive/autobuilders? If not i'll wander off again, so that's really a yes/no question :) [06:52] T-Gone: without gcc-4 there is no archive :) [06:52] oh btw... [06:52] sparc has almost finished breezy/main [06:52] where is hppa again? [06:52] in your ass [06:52] straight up to it [06:52] i hope it feels good [06:52] ;P [06:52] ah that's why i couldn't feel it.. there is none [06:53] MUHA MUHA === T-Gone wonders who's the kid on that chan, sometimes :P [06:54] " [06:54] ops === T-Gone -> off. lamont : ping me when something happens (mail is best). [07:16] http://www.askmen.com/jokes/current/index.html [07:16] hahaha [07:34] fabbione: not its up [07:35] grr.. [07:36] stupid dyndns [07:37] zul: can you fix that up please? [07:37] yeah doing so right now [07:37] i really need to get a more stable connection towards your repo... [07:37] otherwise i can give you some kind of sftp access to my server where to store the repo [07:37] yeah i need to find one [07:37] and you could mirror there [07:37] i think that would be better === LycoN [~lycon@70-32-9-78.frdrmd.adelphia.net] has joined #ubuntu-kernel [07:38] ok.. i will need to look how and what to setup === jbailey [~jbailey@CPE000ded9d787c-CM014260028338.cpe.net.cable.rogers.com] has joined #ubuntu-kernel [07:38] sure [07:38] but we can do it on monday or so [07:38] no probs [07:39] i am spending time updating the GFS patch set in the meanwhile [07:39] *** glibc detected *** double free or corruption (!prev): 0x00022050 *** [07:39] jbailey: that doesn't look nice [07:40] sparc building iproute [07:40] fabbione: Eh? I didn't think that glibc had any checking like that in it. [07:41] jbailey: weird is that it builded on all other arches [07:41] Is there any context to that bug that you can paste? [07:41] jbailey: ~/logs/ ? [07:42] Or even just 2 or 3 lines before that for me to guess what's up? [07:42] But truly, I don't think glibc has any detection code in it for double frees or whatever. [07:42] Sounds like it's coming from something else. [07:42] /usr/include/linux/socket.h:2:2: warning: #warning "You should include . This time I will do it for you." [07:42] gcc -D_GNU_SOURCE -O2 -Wstrict-prototypes -Wall -g -I../include -DRESOLVE_HOSTNAMES -DCONFIG_GACT -DCONFIG_GACT_PROB -o normal normal.c -lm [07:42] ./normal >normal.dist [07:42] *** glibc detected *** double free or corruption (!prev): 0x00022050 *** [07:42] /bin/sh: line 1: 16195 Aborted ./normal >normal.dist [07:42] DOH! [07:42] #warning "You should include . This time I will do it for you." [07:43] gotta love this [07:43] Yeah, it's the new l-k-h. =) [07:43] jbailey: well it's correct [07:43] no user land should include kernel headers [07:44] That and IIRC, linux/socket.h doesn't contain anything that sys/socket.h doesn't. [07:44] But that still shouldn't fail your build, it's only a @warning. [07:44] #warning, rather. [07:45] Actually, hm. [07:45] oh i need to create a clean chroot for you [07:45] It's not the build. It's ./normal >normal.dist that's failing. [07:45] What's that script do? [07:45] can i wipe out the home or do you want me to backup? [07:45] fabbione: Wipe away. [07:45] jbailey: it failed 3 minutes ago.. exactly when you joined the chan :) [07:46] Wow. At this rate, I might develop a complex ;) [07:46] eheheh [07:47] jbailey: if you want i can give you access to the imap inbox where all the logs land :) [07:47] i trash the success & buildd admin mails [07:47] but i keep the failures === LycoN [~lycon@70-32-9-78.frdrmd.adelphia.net] has left #ubuntu-kernel [] === zul [~chuck@CPE0006258ec6c2-CM000a73655d0e.cpe.net.cable.rogers.com] has joined #ubuntu-kernel [07:58] heh...that was weird [08:02] what? [08:03] the connection timing out [08:03] fabbione: do you have a list of drivers that users have emailed you about? [08:05] nobody (THANKS GOD!) mails me directly [08:05] cool [08:05] makes my job kind of easier...kind of [08:07] 'nother grub2 boot test, bbias === jbailey [~jbailey@CPE000ded9d787c-CM014260028338.cpe.net.cable.rogers.com] has joined #ubuntu-kernel [08:13] how did the test go? [08:14] Good. I now have a usable update-grub. [08:14] I've also discovered that I wasn't using 2.6.10-5 to boot my machine, and that it doesn't. =) [08:14] uh? [08:15] -4 is still fine, though. I should hunt that down. I suspect it's just initrd fuckage though, in which case I really don't care. === fabbione builds new GFS [08:15] jbailey: -5 should boot fine.. and it does basically everywhere [08:15] we only have one bug report where it doesn't [08:15] Is it someone on a pegasos box? =) [08:16] jbailey: svenl :) [08:16] Right - so probably affects me too then. =) [08:16] jbailey: the report is about i386 [08:16] nothing to do with ppc [08:17] Ah, okay. =) [08:17] My -4 works fine, though. I'm not fussed. [08:17] fabbione: could you look at 8817 someone wants an option enabled in the kernel [08:17] jbailey: i would be more glad to know why it doesn't [08:17] I'll move my box over to an initramfs config soon for testing anyway, so no point in putting any thought into it. [08:17] fabbione: Bah. [08:17] fabbione: Well, since you ask. [08:18] zul: yeah i saw that... we just need to test a build of it. i remember some horror stories about Zoran and 2.6 [08:18] jbailey: i might hand you directly 2.6.12rc2 [08:18] sure. [08:18] there is nothing we can fix in 2.6.10 anyway [08:18] so don't bother with that [08:18] Sure, that works. [08:19] jbailey: i am planning to unleash 2.6.12 power during UDU [08:19] not before [08:19] there is still too much work to do [08:19] Even better. [08:19] the core is all there and rocking [08:19] it's all the small crappy details that are not [08:20] zul: is your archive fixed now? [08:20] And you say I have to get a patch into Debian's kernel-package before you'll accept it? =( [08:20] fabbione: should be accessible now === jbailey contemplates a dpkg-divert on /usr/bin/mkinitrd [08:20] jbailey: not at all.. we work very close with the Debian kernel team [08:20] i see no point in doing that [08:20] we can test and push [08:21] Ah, that'll work too. [08:21] jbailey: since the kernel is a completely different package from debian, we are not trapped in the sync merge-o-matic shit [08:21] Yeah. I have to be careful with what I do with glibc. [08:22] yes :) [08:22] I don't want to go too far from Debian, but I think the likelyhood of us ever having the same version at the same time is pretty slim. [08:22] jbailey: so do i need to backup your home in the sparc chroot or can i just trash it and give you a clean one? [08:22] fabbione: trash away. [08:22] jbailey: ok [08:23] fabbione: All the lkh bits are now in the archive. =) [08:23] jbailey: rocking [08:23] jbailey: i will give you a chroot that is basically hoary + the new toolchain [08:23] no more than that [08:23] no dist-upgrade yet [08:23] fabbione: Cool. What am I troubleshooting this time? =) [08:23] so you can play with it as much as you like [08:24] jbailey: you said you wanted to test the new NTPL stuff on sparc.. didn't you? [08:24] Oh, right! =) [08:24] so i am giving you the sandbox for it :) [08:25] Do I need to nice my build to keep it out of the way of the buildd? [08:26] jbailey: no, don't worry. all the chroot has ccache so once you build the first, it will be there [08:26] jbailey: just keep an eye on the disk space [08:26] that machine doesn't really have a lot [08:27] Cool. [08:27] if you won't manage to get your sparc up and running, i will explain to you how the stuff works there in details [08:27] so you can do everything you need all alone :) [08:28] i really don't mind that it is used for good stuff [08:28] Cool. [08:28] new chroot in place [08:28] have fun :) [08:28] Yeah, it's better if I can wait to setup my other boxes like the sparc and hppa box until I'm in Montral [08:28] I'm a little short on outlets and stuff here, so if I have access to other machines where I can control the chroot it's nice. [08:28] What I desperately need is amd64. [08:29] i need one of those too :) [08:29] send one to ottawa [08:29] I've been eying a lovely amd64 laptop, though. [08:29] care of lazy ass chuck [08:29] jbailey: you already have full control on that machine.. that's what you haven't realized yet :) [08:29] It was under $1000, 15" screen. [08:29] fabbione: Right, that's why it's really nice. [08:29] amd64.. use concordia :) [08:30] jbailey: you can also create extra chroots if you need [08:30] Takes too long to get chroots updated. [08:30] jbailey: chroots on all the porting machines are updated daily afaik [08:30] jbailey: anyway.. check in ~/bin [08:30] Right, which wasn't useful when previewing toolchain updates. [08:30] there is a simple script called create-chroot :) [08:30] Ah, cool. =) [08:31] You're on a local mirror, right (or at least a proxy) [08:31] local mirror [08:31] it is updated almost every hour [08:31] it depends on how busy is archive [08:31] Cool, so I won't worry about blowing things away and pulling again if I need. [08:31] (15 rsync connection limit) [08:31] nope.. no worries at all [08:32] just remember NOT to leave .deb + .changes in ~/chroots/ [08:32] Yeah. I've had to be careful about that. =) [08:34] ehhe [08:34] well usually katie is pretty careful, but binnmu are a pain [08:40] zul: do you have any objections if i drop the gcc-4 fix for now? [08:40] nope ifs still in my archive for safe keeping [08:41] bah it's a line! [08:41] i will just add it :) [08:41] i was expecting a full rewrite of it :) [08:41] also.. you need to learn to write the changelog :) [08:41] i will fix that for you [08:41] yeah yeah :) === fabbione is an anal retentive changelog bitch [08:42] nah...you are just anal [08:42] i know [08:42] and trust me.. on packages like this is GOOD [08:42] true...hehheh...i said anal [08:43] heh everyone in my house is a sleep except for me [08:44] i am alone :) [08:44] where is your wifey? [08:44] boy scouts [08:45] ah.. [08:46] well only until tomorrow morning [08:46] i have around 12 hours left :) [08:46] cool...i have to get ready for church soon.. [08:46] actually.. i should consider to cook some food [08:46] have fun :) [08:46] see you later probably [08:46] yeah depends on what time i get back from dinner with my parents in law as well [08:47] uhuhu [08:47] sure [08:47] i think i am going to stay up long anyway [08:47] :P [08:47] i slept too much this afternoon [08:48] now to search the forums [08:50] HELL [08:50] i did build the smp version instead of the UP one [08:50] die [08:51] right im off again..if any one wants to lart me can foad ;) [08:52] kthxbye [08:54] later dude [09:35] Setting up linux-image-2.6.12-1-686 (2.6.11.90-1) ... [09:35] cpio: /lib/ld-linux.so.2 (0xb7feb000): No such file or directory [09:35] cpio: (0xffffe000): No such file or directory [09:35] wtf === jbailey remembers to run 'sparc32' [09:41] fabbione: Eh, what'cha doing? Is that when setting up the initramfs stuff built into the kernel? [09:55] no, that's mkinitrd [09:55] sorry i was having dinner :) [09:55] i am debugging it now [09:56] Don't appologise for eating. =) [09:56] I'm just starting the build, btw. [09:56] NPTL for the 3 sparc32 passes and LT for the sparc64 pass. [09:57] nice :) [10:01] ok.. it's not an initrd-tools problem === jbailey cheers! [10:02] But I gues sI knew that. [10:02] We don't use cpio. [10:02] hmm yes you do [10:03] grep cpio /usr/sbin/mkinitrd [10:03] < tmp3 sort -u | cpio -pLd --quiet initrd [10:03] Eh, really? === jbailey looks. [10:03] yes we do :) [10:03] it's the only instance [10:04] but it started to fail today [10:04] i have never seen it before [10:04] cpio: /lib/ld-linux.so.2 (0xb7feb000): No such file or directory [10:04] it tries to copy a file called: [10:04] /lib/ld-linux.so.2 (0xb7feb000) [10:04] something screwed with bash? [10:05] You're on ia32, right? [10:05] That file should exist. [10:05] the symlink is there [10:05] otherwise nothing would work [10:05] Right =) [10:05] and yes.. ia32 [10:05] If it's doing cpio -L, it'll dereference the symlink. [10:06] you know.. in the attemp to mv ld-2.3.5.so to that symlink i trashed the machine :) [10:06] -p [10:06] copy-pass mode [10:06] fabbione: You can recover from that, just call the linker directly. [10:06] So if you mv /lib/ld-2.3.5.so / [10:07] Just do /ld-2.3.5.so mv /ld-2.3.5.so /lib [10:07] nope.. everything is linked against it [10:07] you rm the symlink and you are doomed [10:07] Not at all. [10:07] i had to netboot to recreate the symlink [10:07] PT_INTERP is ignored if you call the linker directly. [10:07] there was not a single command that was working [10:07] As long as you still have a running shell you can do it. [10:08] i did have the shell [10:08] If you close your terminal, you're fucked though. =) [10:08] but mv/cp/ln/ls didn't work [10:08] Right, you missed what I said, though. [10:08] 'mv' won't work [10:08] '/path/to/ld-2.3.5.so mv' will work. [10:08] ahh [10:08] ok [10:09] did the output from ldd changed? [10:10] Show me ldd /bin/ls ? [10:10] I have a Hoary i386 box I can compare against. [10:10] (just booting it) [10:12] yeah i can see that there are some differences [10:13] and mkinitrd uses ldd to determine the libs to copy [10:13]