[12:13] It's replacing an existing patch already that has the same name. [12:13] So the load order should be already determined. [12:14] (And I didn't play with it outside of that environment... I'd be hard pressed to remember how to build a kernel other than 'debuild' *g*) [12:17] ok :) [12:20] jbailey: is there a bug that this patch fixes? [12:22] If bugzilla ever comes up, I'll let you know.../ [12:22] *sigh* [12:38] BenC: 13535 [12:38] It occured to me that I could just search my email. =) === TheMuso [n=luke@dsl-202-173-132-131.nsw.westnet.com.au] has joined #ubuntu-kernel === doko [n=doko___@dsl-084-059-064-154.arcor-ip.net] has joined #ubuntu-kernel === BenC [n=bcollins@ip68-107-227-12.hr.hr.cox.net] has joined #ubuntu-kernel === doko [n=doko___@dsl-084-059-064-154.arcor-ip.net] has joined #ubuntu-kernel === smurfix [n=smurf@debian/developer/smurf] has joined #ubuntu-kernel === jbailey [n=jbailey@modemcable139.249-203-24.mc.videotron.ca] has joined #ubuntu-kernel === infinity [n=adconrad@loki.0c3.net] has joined #ubuntu-kernel === dilinger [i=dilinge@mouth.voxel.net] has joined #ubuntu-kernel === lamont [n=lamont@mix.mmjgroup.com] has joined #ubuntu-kernel [06:09] morning [06:17] BenC: did you commit the patch? [06:17] or do i need to? [06:18] argh [06:18] he is not online [06:18] jbailey: where is your patch? [06:19] fabbione: My homedir on concordia [06:19] ok [06:19] thanks [06:19] BenC: Whenever you're back, the initramfs DSDT patch is in my homedir on concordia and is called drivers-acpi-osl_attach-dsdt-to-initrd.dpatch [06:20] drivers-acpi-osl_attach-dsdt-to-initrd.dpatch ? [06:20] ehhe [06:20] Right. Replaces the current version. [06:20] After 3 attempts to figure out how to call it something else, I gave up and gave it the same name. =) [06:20] ahaha [06:20] (I think you generate your patch lists somehow, but the kernel packaging makes gcc's look simple) [06:21] yes we do.. [06:21] from 00list-$version to 00list [06:21] because we support patching per arch and per flavour [06:21] so we need to generate it dynamically [06:22] dpatch isn't exactly working [06:22] what do i need to add in the changelog? [06:26] jb?? [06:26] what do i need to add in the changelog? <- [06:30] ok nevermind.. i wrote something you lazy canadian :P [06:42] Eh, I don't have a nick highlight on 'jb' [06:42] =) [06:42] *sleepy* Canadian. [06:42] yeah right :) [06:42] lazy lazy... [06:42] :P [06:42] Right. [06:42] If I were less lazy, I'd have gotten up and gone to bed already. [06:43] ahaha [06:44] Sleep now [06:48] night jeff! [06:54] lamont: ping? [06:54] eep! [06:55] lamont: i am on f11.. do we have enough pkgs to build the kernel in the breezy chroot? [06:55] or do i still need to use hoary? [06:56] I think you can use breezy [06:56] ouch. [06:56] heck i can't remember my passwd on f11 [06:56] grrr... [06:56] need to have gcc-3.4-hppa64 (still universe) installed in the chroot. [06:56] lamont: can you reset it for me please? [06:56] hppa-hacks/expect-tcl8.3-dev_5.43.0-2_hppa.deb [06:56] hppa-hacks/expect-tcl8.3_5.43.0-2_hppa.deb [06:56] hppa-hacks/expectk-tk8.3_5.43.0-2_hppa.deb [06:56] hppa-hacks/gcc-3.4-hppa64_3.4.4-5ubuntu1_hppa.deb [06:56] hppa-hacks/palo_1.9_hppa.deb [06:57] that's all the universe-should-be-main packages that I have hacked in right now. [06:57] otherwise i will just use hoary... [06:57] fabbione: sure [06:57] just put a temp one that i can change on the fly [06:57] i am already logged in [06:58] yeah, ISTR that sshkeys are the only way to actually login... [06:59] yeah but i need the passwd for sudo :) [06:59] so i can install the B-D [06:59] or you can install them for me.. [06:59] i really don't care tbh 100% [07:03] lamont: is the chroot reasonably updated? [07:03] or do i need to upgrade it first? [07:04] hasn't been touched by me in forever [07:04] ok [07:04] I'll let you fly it [07:04] deb http://people.ubuntu.com/~lamont/ubuntu-hppa/tree hoary main restricted universe [07:04] i guess this is not needed anymore.. [07:04] lamont: that's fine :) [07:05] and should not be present if breezy is [07:05] (it has .debs with diff md5sums, but the same name) [07:05] i think you did create that chroot during the rebuild of breezy === fabbione kills === lamont isn't sure if breezy is actually debootstrappable yet... [07:06] so you may have to clone the hoary chroot, nuke /var/cache/apt/archives/*, fix /etc/apt/sources.list, and dist-upgrade [07:06] for what i need, a dist-upgrade will do [07:06] maybe [07:07] it's old enough that it's upgrading almost everything :) [07:07] if you get any bitching about md5sums on debs being wrong, remove the afflicted .deb :-) [07:07] eheh yeah.. have been there before [07:11] chroot breezy/ apt-get build-dep linux-source-2.6.12 [07:11] Reading package lists... Done [07:11] Building dependency tree... Done [07:11] 0 upgraded, 0 newly installed, 0 to remove and 0 not upgraded. [07:11] i guess that's it :) [07:17] ok. sparc and hppa building [07:20] nice.. the latest OCFS2 will compile on sparc and hppa as well.. [07:22] and the others too now... === BenC [n=bcollins@ip68-107-227-12.hr.hr.cox.net] has joined #ubuntu-kernel [07:55] hey Ben === Seveas [n=seveas@seveas.demon.nl] has joined #ubuntu-kernel === michele [n=michele@host147-90.pool8540.interbusiness.it] has joined #ubuntu-kernel [09:47] fabbione: what's the hold-up with 9157? [09:48] Mithrandir: that the fix we did try, didn't work? [09:49] I can't remember, it's been a whlie. [09:49] yeah i do remember... [09:49] that's why.. the patch doesn't help at all... [09:49] but i don't know why [09:50] we can ask BenC when he is awake... === chmj [n=chmj@196.36.161.235] has joined #ubuntu-kernel === calc [n=ccheney@ip70-185-4-246.ma.dl.cox.net] has joined #ubuntu-kernel === BenC [n=bcollins@ip68-107-227-12.hr.hr.cox.net] has joined #ubuntu-kernel [10:18] apparently, it should load i2o_core, but not i2o_block. [10:18] what's responsible for loading sd_mod for instance? I think the same mechanism should be used to load i2o_block [10:18] hotplug in theory [10:19] but the point is that it loads correctly the driver for the card, but not the possible associated device manages [10:19] managers [10:19] if you load a scsi driver for card foo [10:19] you don't know what's connected to it [10:19] it could be a disk or a cdrom [10:19] yeah [10:20] ok, so hotplug should then traverse the device tree and look at what's connected and then load modules as appropriate? [10:20] that's why initramfs parses /sys after to identify what needs to be loaded [10:20] yup [10:24] hm, I guess I should see if I can test that machine sometime in the future, then [10:27] Mithrandir: initramfs atm doesn't check for i2o [10:27] but it's trivial to implement [10:27] that would solve your problem i think [10:28] check in initramfs scripts/functions [10:28] there is a call to scsi_init_something [10:29] that's a little bash function that parses /sys in seek of scsi devices and modprobe [10:29] the same can be done for i2o without any problem... [10:43] ok ladies [10:43] 7.11 is ready === calc [n=ccheney@ip70-185-4-246.ma.dl.cox.net] has joined #ubuntu-kernel [10:53] I think I have a fix for i2o, now I just need to get access to the box scheduled.. [12:12] fabbione: any idea where Xu's test kernel is? [12:13] Mithrandir: he did upload already.. it's building mostlikely, waiting for pitti to bless the USN [12:13] ok [12:13] but i don't have the changelog [12:13] so i don't know if he added the patch or not [12:13] it's mostlikely in the archive already [12:14] because sparc picked it up for building [12:14] * Fixed sync_page_io bio leak in drivers/md/md.c (Neil Brown). [12:14] yup [12:14] cool [12:14] food time === Mithrandir [n=tfheen@c5100BC63.inet.catch.no] has joined #ubuntu-kernel === BenC [n=bcollins@ip68-107-227-12.hr.hr.cox.net] has joined #ubuntu-kernel [12:52] hey BenC [12:52] already awake? === Seveas [n=seveas@seveas.demon.nl] has joined #ubuntu-kernel [01:27] jbailey: does my i2o patch for initramfs look remotely sane? [01:27] Mithrandir: I just got here, lemme look. =) [01:28] ok [01:29] Mithrandir: Where did you send it? [01:31] jbailey: new bug [01:32] 13806 [01:34] Yes, this looks right. [01:34] Assuming that's where things are in /sys and such. I don't have i2o systems here. [01:36] well, it's the only thing I saw which resembled a block device. [01:36] Cool. Will you have a chance to test it, or should I just apply it? [01:38] I can test it, but not for a while, that box is responsible for all the ads we're sending out here (which means it'll cost about 3000 EUR/day just in lost revenue if we take it down unplanned) [01:43] Yup, fair enough. [01:44] I'll just apply it for now then. It can't be worse than it is now. [01:44] true [01:44] And I don't see anything harmful in there. [01:45] I won't upload just for this, though. I'll probably pull in the LVM fix first. [02:08] what time is it in BenC land? === zul [n=chuck@CPE0006258ec6c2-CM000a73655d0e.cpe.net.cable.rogers.com] has joined #ubuntu-kernel [02:10] heylo [02:10] hey zul [02:11] ok time to start uploading the kernel [02:11] i can't wait forever [02:12] you can do it! [02:12] Woo kernel [02:15] eheh [02:17] accepted [02:17] this time sparc will be the fastest arch to upload :) [02:19] because there are only a couple of users? ;) [02:22] zul: He' [02:22] s probably primed the ccache. =) [02:22] im evil [02:22] doko: Quick! Upload gcc! [02:26] fabbione: That's -7.11? [02:26] yes [02:27] jbailey: no.. even better.. i did a full sbuild while waiting.. the kernel was ready for upload this morning [02:27] Ahahahah [02:28] but i was waiting Ben for the baz dance and stuff [02:28] so in the meanwhile i can start [02:28] it always takes a few hours before it's builded on all 4 arches === BenC [n=bcollins@ip68-107-227-12.hr.hr.cox.net] has joined #ubuntu-kernel [02:42] c === TheMuso curses speakup. [02:43] hey ben === TheMuso doesn't remember being called ben unless zul is referring to someone else. [02:46] hey BenC [02:46] TheMuso: you are not the only one in this chan :) [02:46] Just realised that. :S [02:49] weird [02:49] not sure why I keep getting disconnected [02:49] fabbione: ready to do this kernel? [02:49] BenC: yes.. [02:50] i did upload the kernel in the meanwhile [02:50] all the rest needs to be done [02:50] as it stands now, preX,11 branch is finished [02:50] so it's like you completed your devel cycle [02:51] i usually wait for katie to get the ACCEPT message and only after tag/prepare the new devel branch [02:51] so when you are ready we can start the baz dance [02:51] let's do it [02:52] ok.. === BenC hears the bas line to Funky Cold Madina as he says that [02:52] the first thing to do is to get mainline... [02:52] bass line [02:52] ahah [02:52] kernel-team@ubuntu.com--2005/kernel-debian--mainline--2.6.12 [02:52] since it's the first time: [02:53] baz get kernel-team@ubuntu.com--2005/kernel-debian--mainline--2.6.12 debian-mainline [02:53] store it whenever you want.. [02:53] it doesn't need to be in the kernel source [02:55] ok [02:55] tell me when you have done that.. [02:55] it's done [02:55] ok [02:55] now we merge from the pre branch [02:56] baz merge kernel-team@ubuntu.com--2005/kernel-debian--preX,11--2.6.12 [02:56] you do that inside the debian-mainline dir you just got [02:57] this will take some time... [02:57] it's a quite slow operation [02:57] ok... [02:57] in the meantime i suggest you check your enviroment on people.u.c to be sure 100% that your umask is set to 002 [02:57] for sftp operation [02:58] before or after I do this? [02:58] during... [02:58] ok, did that [02:58] it's important the umask is set before you commit [02:59] and the merge is done [02:59] for reading it's no problem [02:59] ok [02:59] did you get any conflict? [02:59] nope [02:59] ok.. now you can commit... [02:59] baz commit -s'2.6.12-7.11' [02:59] -s is the comment... [02:59] or commit log or whatever [03:00] baz doesn't spawn a nice editor to make commit logs [03:00] it's sort of retarded that way [03:00] ok, that finished, but the last line was [03:00] arch: no arch user id set [03:00] that a warning? [03:00] hmmmm [03:00] or did it fail [03:00] what's the output of baz my-id ? [03:01] arch: no arch user id set [03:01] ok [03:01] baz my-id Fabio M. Di Nitto [03:01] some env I need to set? [03:01] or whatever fits better to you [03:01] the commit did fail... [03:01] or it didn't go trough [03:01] also remmeber your gpg private key [03:01] you need to sign the commit [03:02] otherwise it fails even worst [03:04] * committed kernel-team@ubuntu.com--2005/kernel-debian--mainline--2.6.12--patch-10 [03:04] what's your keyid? [03:04] "Ben Collins " [03:05] 60E80B5B [03:05] sorry... i mean the 0x.. [03:05] yeah === fabbione updates locally... [03:06] so the "--patch-10" is something that baz adds? is that like a revision level? [03:07] it's like the amount of commits on that branch [03:07] sort of -rXXXX in svn [03:07] right [03:07] it's added automatically normaly [03:07] but you can use it to retrive specific revisions [03:07] and stuff like that [03:07] ok.. now [03:08] we need to tag and prepare the next devel branch [03:08] do a baz abrowse [03:08] very useful command to see all the branches in that repo [03:09] now.. baz as tla sucks hard on how you can name branches.. [03:09] specially when it goes to . - and stuff like that [03:09] lots of them [03:09] yup [03:11] ok, ready to create a branch [03:11] baz branch kernel-team@ubuntu.com--2005/kernel-debian--mainline--2.6.12 kernel-team@ubuntu.com--2005/kernel-debian--mainline-2,6,12-7,11-00 [03:11] yes sorry.. connection to uk is slow from here.. [03:11] it was still abrowsing [03:11] this will create the tag [03:11] no! [03:11] there is an weeoe [03:11] error [03:11] ok [03:11] baz branch kernel-team@ubuntu.com--2005/kernel-debian--mainline--2.6.12 kernel-team@ubuntu.com--2005/kernel-debian--mainline-2,6,12-7,11--0 [03:11] there [03:12] note how . - and , change [03:12] that's the tag for the last release, right? [03:13] yes, that's correct [03:13] done [03:13] ok [03:13] now let's create the branch for the new devel [03:13] baz branch kernel-team@ubuntu.com--2005/kernel-debian--mainline--2.6.12 kernel-team@ubuntu.com--2005/kernel-debian--preX,12--2.6.12 [03:14] done [03:14] perfect [03:14] now.. go in you kernel source debian/ [03:15] (that i guess you manage with baz)... [03:15] baz switch kernel-team@ubuntu.com--2005/kernel-debian--preX,12--2.6.12 [03:15] brb.. wife is calling... [03:19] re [03:19] tell me when you are done... [03:19] done [03:20] ok just one second that i will fix one small thing in the rules file [03:20] it will make your life simpler :) [03:21] if you notice at the end of the rules file there are some "I am lazy BOFH" targets to help you do some boring tasks [03:21] one of them is startnewrelease [03:21] (that i am fixing now) [03:22] that's the first command you run on a new devl branch... [03:22] arch_commit: unable to acquire revision lock (internal error in archive-pfs.c(pfs_lock_revision)) [03:22] GRRRr === fabbione bets on wrong umask === BenC set his umask correctly [03:23] unless .bash_profile isn't getting loaded [03:23] BenC did you set it at the top or the bottom? [03:23] top [03:23] I just changed what was already there [03:23] 4 drwxr-xr-x 3 bcollins warthogs 4096 Aug 19 14:14 kernel-debian--preX,12--2.6.12 [03:24] yeah.. it didn't load.. [03:24] let me check where i do it... [03:24] i do it in .bashrc [03:24] that's because sftp doesn't load .bash_profile [03:25] so i need you to do some chmod on people :) [03:25] cd /home/lamont/public_html/Archives/kernel-team@ubuntu.com--2005 [03:25] yeah, doing it now [03:25] yeah [03:25] we need it in all the branches/commit you do [03:25] done === fabbione commits [03:26] $ find . -user bcollins | xargs chmod g+w [03:26] should get everything [03:26] it did [03:26] ok get the change i did to rules and run that makenewrelease as user [03:27] no need of fakeroot [03:27] added umask to my bashrc too [03:27] perfect :) [03:28] it should basically: create the new changelog entry, add debian/patches/00list-* stuff [03:28] you might want to correct the info in the changelog for the name [03:28] they depend on your passwd file.. [03:29] but clearly, modify them as you wish [03:29] these were ok for me.. they might be useless or not enough for you [03:30] startnewrelease? [03:30] yeah [03:30] make -f debian/rules startnewrelease [03:31] done [03:31] ok.. now.. [03:31] baz lint [03:31] it will tell you if there are files in the archive that are not in rcs [03:31] no output [03:31] ok [03:31] baz diff [03:31] so you can see what has been done [03:32] changes related startnewrelease [03:32] perfect [03:32] now.. we did bump the abi [03:32] already been using that command :) [03:32] so there is a manual step to do [03:32] ehe ok.. since i don't know how much you know about baz.. i go step by step.. [03:33] go in debian/abi [03:33] ok [03:33] baz mv 2.6.12-6.10 2.6.12-7.11 [03:33] once you have done that, commit [03:33] the whole tree? [03:33] ? [03:33] not just the abi change, right? [03:33] no no.. the whole tree [03:34] this is to start a new release after an abi change.. [03:34] the most complex case basically [03:34] so once you get this one right.. the others will be a breeze.. [03:34] tell me when you are done with the commits... [03:35] so i can add the new abi files [03:35] done [03:35] ok gimme a couple of minutes now... [03:35] in future, when we have abi changes, you will have to take care of i386/amd64/ppc/ia64.. [03:36] that's what you are mandate to :) [03:36] i will take care of hppa/sparc [03:36] since it's non-paied porting job [03:36] but clearly.. [03:36] if you take care of them too.. you will become a Kernel Hero :) [03:38] well, I have both systems here, so I can probably squeeze them out once I settle in to the "have to do" stuff :) [03:38] well it's simple [03:38] we officially support only 3 arches [03:38] i386/amd64/ppc [03:39] so the rest is work that should come from the community [03:39] given that i will get pobably ia64 and hppa, and that i love sparc... [03:39] i took care of them as well [03:39] it's not high cost to run a build in parallel [03:39] ia64 is at the DC [03:39] hppa ping me or lamont [03:39] sparc we can both do i guess [03:40] amd64 is the only thing I am missing locally [03:40] BenC: tough.. i only have i386 and sparc :) [03:40] you build it at the DC [03:41] somebody will test for you [03:41] and for what i suggest.. build and work as much as you can on the DC machines [03:41] they are fast :) === fabbione waits for baz to do a diff.... [03:46] ok done [03:46] BenC: update.. and you are good to go with the next kernel [03:47] now we need to look at all the surronding of an abi change... [03:47] BenC: ssh to chinstrap and go to fabbione@chinstrap:~/benc [03:48] i did prepare all the pkgs for you already signed [03:48] since i didn't know if you already have upload privileges [03:48] when changing the abi the sequence to do is: [03:48] kernel [03:49] linux-restricted-modules [03:49] and in no preferred order linux-meta and debian-installer [03:49] ok [03:49] now.. as soon as the kernel is built on all arches [03:49] what exactly is changed in each of those? [03:49] http://people.ubuntu.com/~lamont/buildLogs/l/linux-source-2.6.12/ <- [03:49]