=== MTeck is now known as MT- === MT- is now known as MTeck === jMyles is now known as ausimage42721 === ausimage42721 is now known as jMyles [14:19] rtg: will these changes that you threaten to push to master break a rebase? [14:20] amitk, rebase to a topic branch? not if you _completely_ remove the topic branch debian directory first. [14:21] * amitk will try it out once the change is made... [14:21] amitk, you're gonna have to change the rebase script in the fsl topic branch [14:21] I will [14:22] and then push it to master for the next poor soul who has to start a new branch [14:22] amitk, agreed [14:22] we'll get it figured out eventually [14:30] anyone else seeing us.archive.u.c being awfully slow today? [14:31] amitk, i recon the rebase script should by default pull in the contents of the scripts/misc directory from master by default [14:31] pgraner, where are you timezone wise? its always slow from here [14:32] apw: back in EST (NC) [14:32] apw: I'm getting a steady 41.8kB/s [14:32] thats utterly shite :) [14:33] --help [14:33] apw: yea now you see why I'm asking about the apt-cacher-ng thing [14:33] does it belong to us? [14:33] rtg, for the upgrade issue from linux-imx51 you simply need to add: Conflicts: linux-imx51 and Replaces: linux-imx51 to the linux-babbage debian/control file [14:33] apw: my rsync script stopped running while I was gone so I'm a week out of date :-( [14:34] yeeks life is over [14:34] ogra, send a patch to the mailing list so I don't forget [14:34] apw: so I'm reworking my setup .... [14:34] apw: the only thing I'll need to rsync are the daily ISO bits [14:35] yeah same here, iso's get syned apt-cacher for the rest [14:36] Ng, ? whats up with '--help' in all the channels? [14:36] us.archive.u.c has dailies ? [14:36] * ogra thought they would not be mirrored from cdimage.u.c [14:36] ogra: no, just talking about how we do it [14:36] rtg: shockingly poor use of irssi's /foreach :( [14:36] Ng, ah, I have that problem some days too :) [14:37] pgraner, ah, i thought it was related to your question about us.a.u.c [14:37] Ng: and here I thought you had so deep insight to the slow us.archive discussion :-( [14:46] rtg: I got an email from someone complaining that we dropped the DSDT patch [14:47] pgraner, we didn't drop it, I just never forward ported it. [14:47] rtg: ok, at this point its the same difference [14:47] rtg: IIRC we decided at the sprint that we would not fwd port it, correct? [14:48] mjg59, do you think the DSDT patch is still worth it? I consider it a bit dangerous since you can brick your BIOS with it. [14:49] rtg: I'd be surprised if you can brick the BIOS [14:49] It's certainly possible to damage the hardware by breaking thermal control, though [14:49] * cking concurs with mjg59 [14:49] I don't think it's necessary any more [14:49] We don't ship it in Fedora and there's a miniscule number of bugs from people wanting it [14:49] mjg59, I have not checked, but are you carrying it in Fedora? [14:50] mjg59, cool, thanks. [14:50] pgraner, I stand by our decision to _not_ carry it. [14:50] rtg: ack, I'll respond, I'll document on the wiki on the Kernel FAQ [15:08] amitk, what was the syncy thing you use called? [15:09] * rtg thinks apw speaks in pidgin english [15:11] i have a way with words don't you think [15:11] apw: unison [15:11] thx ... stupid name :) [15:11] apw: wait till you find out what its written in :) [15:17] Ocaml? [15:17] amitk, you used this thing with more than one replica? [15:21] cking: correct [15:21] and there was much gnashing of teeth... [15:21] apw: simultaneously? no [15:21] cking, one of your favourites then? [15:22] * cking notes it's another scripting language to learn and then forget about [15:22] apw: you can do several 1-to-1 syncs to achieve the effect [15:22] amitk, any idea which end keeps state for the sync? src/dst/both? [15:23] both keep their end of the state [15:24] so doing a sync from A to B from either end is equivalent and interchangeable [15:24] absolutely [15:24] http://www.cis.upenn.edu/~bcpierce/unison/ , incase you haven't found it already [15:27] amitk, is this thing interactive? ie is it asking me a question when it says: [15:28] <---- file bin/kpkg-check [f] [15:30] apw: use -auto, it will then only bother you with the ones with conflicts [15:30] handy thanks [15:33] apw: I use this script: http://pastebin.ubuntu.com/250854/ . I can then call it with unison-sync.sh [15:34] yeah that looks handy [15:35] i assume that makes the sync unit the whole account but then limits us to one subdir [15:36] if by whole account, you mean your home dir, then yes [16:55] jjohansen, will we achieve a uec (aka ec2) flavour for this A4 release? [17:05] rtg: I hope so [17:06] rtg: I have run into a few problems with the suse patches though [17:06] jjohansen, its pretty much gotta happen today. otherwise I won't have time to integrate the patch set. I may not have time anyway. [17:07] rtg: right now I am running a paravirt ops build while I work on the suse patches, to see if that will work [17:07] perhaps you just oughta take your time and get it right. An A4 release isn't critical. [17:55] hi, can someone help me with linux-image-debug- package? [17:56] i understand that its not available for all kernel versions. how can i make one myself? [17:57] sumanc, look in debian/rules.d/0-common-vars.mk for CurrentlyBuilding [17:58] ogra, re: fsl imax51 build times. what do you suggest? [17:58] rtg, faster buildds :) [17:59] we'll get them soon [17:59] ogra, I've been working to speed developer build times by skipping some packages. [17:59] rtg, that wasnt a serious comment in the other channel [17:59] ogra, couldn't tell if you were just grumbling [18:00] nah [18:00] i'll wait patiently [18:00] have my nightshifts for the week planned anyway so it doesnt matter when the buildd spits it out [18:01] it just feels longer if you watch it all the time [18:01] ogra, go have a beer and come back in awhile. [18:02] yeah, i probably should do that [18:12] rtg, thanks. i read that file. can you please dumb it down for me a little. [18:12] * apw wonders if it is entirly reasonable to be checking floppies for partition tables? [18:13] if i do AUTOBUILD=1 fakeroot debian/rules binary-debs, will it build that package? [18:13] i assume they are supported in theory? [18:13] sumanc, i would not have expected you to need AUTOBUILD=1, otherwise i would expect that to make you some binaries to install [18:17] sumanc, 'sudo mkdir /CurrentlyBuilding', then rerun your build. You can shortcut some of the build by running 'fakeroot debian/rules binary-debs flavours=generic' [18:20] apw and rtg, thanks. command 'fakeroot debian/rules binary-debs flavours=generic' is running right now [18:20] you might want to make it skip abi and modules checking in the end of the build [18:20] if you made any config changes [18:20] but CurrentlyBuilding is created under / right? what does this do? i am running fakeroot as non-root user. will it create any permission issues? [18:20] * ogra always runs into that [18:21] and sedly the check happens only at the very end of the build [18:21] *sadly [18:21] i c. [18:22] sumanc, its a convention used by the buildd's. if '/CurrentlyBuilding' exists (which it will on a buildd), then the debug packages (and some other stuff) are created. [18:22] i have not made any config changes, i want the linux-debug- package for the systemtap to work. [18:22] sumanc, you should be fine. note the debug package is quite large [18:23] rtg, thanks for the explanation. large is fine. i got a lot of space :) [18:23] sumanc, and hopefully a fast disk [18:24] 5400 rpm [18:24] build is crawling the laptop though :( [18:26] you have an hour of grinding i'd say [18:27] wow [18:29] i am just getting into kernel hacking. is systemtap something you guys recommend? or are there other tools and practices? i want to be able to debug/look inside the hood as much as possible [18:31] i am used to printk, but it gets in the way and not always appropriate, i think. [18:33] cking, and every time it finds a floppy, it checks it for a partition table [18:34] which leads it to read at least 2 blocks, leading to a 50s ish delay [18:43] apw, i doubt we would drop devicekit within the next two weeks to switch to udev only design [18:43] and after FF such a change is inappropriate [18:43] so fixing devkit is surely the way to go for floppies [18:44] it'll be a udev config change i think, but yes [18:44] geez ! [18:44] seems linux-fsl is in its finishing stages [18:44] na you are hulucinating from holding your breath [18:44] lol [18:46] sumanc, i tend to use printk myself for most of my debug [18:47] ogra, armel FTBS [18:48] bohooo [18:48] hmm, dpkg-gencontrol: error: package linux-image-debug-2.6.31-0-babbage not in control info [18:49] thought I fixed that [18:50] glad to see: dpkg-deb: building package `linux-image-debug-2.6.28-14-generic' [18:50] rtg, well, your debian/rules seems to still have lines with -plinux-image-debug-2.6.31-0-babbage [18:51] ogra, indeed. Is there a requirement for a debug package? [18:51] not now, no [18:51] the kernel wont even work properly, we have no working USB and no NIC driver atm [18:52] its all about booting and making sure the image layout in karmic works right [18:52] drop debug [18:53] gurgle, arm, ... pop [19:12] thanks guys. i have the debug package installed now. i need to reboot to test it. thanks a lot for your help. === zul_ is now known as zul [23:36] hi everyone [23:36] I need help [23:37] anyone know if this would be supported by the kernel? [23:37] http://www.ppa-usa.com/products/usb/usb_pci/1325.htm [23:39] mattwj2002: hard to tell without knowing what's the chipset used for each function. [23:39] oh okay [23:39] yeah I tried googling it [23:39] no luck :(