20:00:32 #startmeeting 20:00:32 Meeting started Wed Sep 26 20:00:32 2012 UTC. The chair is pwhalen. Information about MeetBot at http://wiki.debian.org/MeetBot. 20:00:32 Useful Commands: #action #agreed #halp #info #idea #link #topic. 20:00:32 #chair pwhalen jonmasters bconoboy ctyler pbrobinson dgilmore 20:00:32 Current chairs: bconoboy ctyler dgilmore jonmasters pbrobinson pwhalen 20:00:37 .fas jcapik 20:00:38 jcapik: jcapik 'Jaromír Cápík' 20:00:42 good afternoon guys 20:00:43 .fas djdelorie 20:00:43 djdelorie: djdelorie 'DJ Delorie' 20:00:44 .fas chris@tylers.info 20:00:46 ctyler: ctyler 'Chris Tyler' 20:00:48 * masta here 20:00:49 .fas pwhalen 20:00:49 pwhalen: pwhalen 'Paul Whalen' 20:00:57 .fas parasense 20:00:58 masta: parasense 'Jon Disnard' 20:01:01 .fas frojoe 20:01:02 Frojoe: jacwang 'Jordan Cwang' - burningfool 'Jordan Cwang' 20:01:10 .fas darthjava 20:01:10 DarthJava: darthjava 'Dmitry Kozunov' 20:01:42 .fas dmarlin 20:01:43 dmarlin: dmarlin 'David A. Marlin' 20:02:16 .fas ahs3 20:02:16 ahs3: 'ahs3' Not Found! 20:02:16 #topic 1) F18/19 Build status - problem packages 20:02:23 :) 20:03:00 we got glibc sorted out 20:03:07 is anyone aware of our current blockers for f18/19? 20:03:44 hola 20:04:20 nothing, excellent! 20:04:27 might be some blowback later for ARM atomics 20:04:35 pwhalen: mainly kernel for f18, as far as stopping progress to alpha 20:04:44 * masta unsure if that was glibc or gcc 20:05:06 dgilmore, right, we'll come to that one specifically.. 20:05:22 #topic 2) 3.6 kernel mmc driver - status update 20:05:47 dgilmore, how goes it? 20:06:18 right so have identified the resource in omap world for mmc + dma, no reply so far in regards to =m 20:06:30 works when =y 20:07:40 does the dma engine thing impact all the kernels? 20:08:01 not sure if all of them converted mmc to dma engine 20:08:26 masta: trimslice didn't seem to need it (yet) 20:09:05 ok great, was unsure if that was a global mmc change or specific to one or another 20:09:40 pwhalen: slowly 20:09:45 we need to decided when we go with =y if thing do not move upstream? 20:09:56 for omap at least 20:10:06 masta: and tegra 20:10:55 it would suck to diverge from the common config for mmc, but we need to start havign working kernel boots 20:11:33 whats our course of action here? Last week I believe there was some consensus, where do we stand now? 20:11:48 since until recently these were built-in, and have not worked since the change, I'd call it a regression, and suggest we revert the change until a bootable kernel using modules is available 20:11:59 well I think we decided to wait for this meeting to decide 20:12:11 a week to get it working as modules 20:12:23 else we risk fallign behind 20:12:29 falling 20:12:55 but we didn't engage upstream until a few days ago... 20:13:06 for omap... and no reply so far 20:13:11 dmarlin, I would agree. 20:13:34 +1 20:13:38 +1 20:14:10 we can still work in getting the kernel to boot using modules via scratch builds, and once they are working push out the config change 20:14:25 any -1 on this? dgilmore ? 20:15:40 welp, I agree we can work on the module thing later 20:16:16 who is able to fix the kconfig in the kernel srpm? 20:17:50 masta: I have working configs... I can send a diff to anyone who will apply it 20:18:08 well that implies we need more folks willing to edit the kernel, or able... 20:18:12 masta: not config, just config changes 20:18:17 not kconfig 20:18:24 sry my bad 20:18:48 before we say agreed on this, lets take it to one last round on the list? 20:19:06 somebody who has commit rights? 20:19:07 +1 20:19:16 pwhalen: why? 20:19:20 jcapik: yes 20:19:21 the folks we need to commit this aren't here 20:19:39 dgilmore and pbrobinson 20:19:41 it was a suggestion, if you guys disagree, please let me know 20:20:02 ^^ and/or ^^ 20:20:04 I just dont want to say agreed, then turn to the folks we disagreed last time and ask them to make the change 20:20:13 s/we/who/ 20:20:13 * dgilmore is tsill working on some options 20:20:17 pwhalen: I'll send the patches and see if they get committed, or if there are objections 20:20:27 (to the list) 20:20:52 dmarlin, that sounds good 20:21:02 dmarlin: you can create a regular bug 20:21:09 jcapik: sure 20:21:36 jcapik: I'll file a BZ, attach the patches, and send a link to the list 20:21:46 dmarlin: +1 20:22:01 sounds good +1 20:22:25 #action dmarlin to send kernel config patches to the list, open BZ for a final round of discussion 20:22:42 #topic 3) VFAD summary 20:22:59 masta, was there any documentation of the vfad I missed? 20:23:05 oh yea... last week 20:23:39 sry, I totally failed at this task, we mostly identified the aforementioed mmc thing 20:23:57 so we had problems booting rc5/6 20:24:39 * masta would love to host another vfad, having learned from this... would improve 20:24:49 VFAD at masta' 20:24:52 s house! 20:25:04 perhaps when we have new test kernels 20:25:42 I hear 3.6-rc7 is out 20:26:06 but we would block on the email to the list thing, about mmc 20:26:44 #topic 4) Raspberry Pi status update 20:27:32 Updates to Kernel and WiringPi and then we ship. 20:27:53 ctyler, fantastic, is there an eta? 20:27:58 Updates will be out this week, aim to ship next week. 20:28:10 * djdelorie notes that Fedora "It’s no longer our recommended distro" by the RPi foundation: http://www.raspberrypi.org/archives/805 near the end :-( 20:28:24 what is the kernel being updated to? 20:28:29 3.2.27 20:29:13 cool 20:29:33 Would be cooler if they upstreamed to master, but we'll take what we can get. 20:29:53 #topic 5) your topic here 20:30:24 open floor for anything else 20:30:46 ok so the f18 alpha images thing... ppc just posted theirs... kinda jelous 20:31:17 just for future reference, when we make changes to the kernel config (or other critical packages), I would suggest we test on all supported platforms before pushing upstream. 20:31:28 what is the short list of things we need to achieve alpha images? 20:32:28 * working kernels 20:32:38 * lorax? 20:32:42 working f18 kernels, and f18 image creation tools 20:33:32 f18 image creation tools may take a while, but are being worked by the anaconda team 20:33:53 we can used patched f17 tools to create f18 images, at least for testing 20:34:42 I think that should be acceptable for an alpha, lest we get further behind 20:34:51 missing from anaconda/lorax are: no-virt support, text mode install, image-only creation 20:34:59 (in F18) 20:35:25 so if we have some python/anaconda hackers willing to help. :-) 20:35:53 I think text-mode installs is getting close 20:36:37 (didn't we conclude that vnc was a viable alternative to text-mode installs?) 20:36:58 ok so no image-only and text mode... does that mean... it doesn't work right now? 20:37:02 at all? 20:37:03 ctyler: it was suggested, but I don't know if anyone has tried it 20:37:27 masta: anaconda has been largely rewritten in F18... 20:37:49 masta: and those features have not yet been implemented 20:38:13 masta: text-mode is getting close... no-virt and image-only are not yet 20:38:27 just want to know if right now we have a tool-chain to build images... if no what is the complexity to make that happen? 20:38:55 * masta is willing to learn python to help here 20:39:07 using only F18 tools... pretty comples at this time 20:39:12 complex 20:39:44 dmarlin: where do you see --image-only missing? 20:39:47 the anaconda team is working on these issues, but may have higher priority issues, so no ETA 20:40:05 dgilmore: bcl said it was missing 20:40:12 dgilmore: I have not looked 20:40:18 dmarlin: it doesnt appear to be so 20:40:29 dmarlin: it very much looks present 20:40:33 dgilmore: cool. maybe he added it 20:40:50 dmarlin: ultimately someone on the arm team needs to do the work if the anaconda folks dont 20:40:59 dmarlin: i doubt it was ever removed 20:41:22 if its something high level i can try :) 20:41:24 dgilmore: ok, that was my understanding, but I could be wrong 20:41:36 i rem getting anaconda to run in text mode on a guru which was kind of useless tho 20:41:36 ok so we need to determine if the tools work 20:42:02 just had to tell it to skip some steps and point to the right partition and so on 20:42:24 masta: we have a query in to the anaconda team... I may know more before the next meeting 20:42:33 fossjon_room, sounds like your our man :) 20:42:50 no im not sure, just saying 20:43:14 http://fossjon.wordpress.com/2012/04/09/anaconda-gui-running-on-a-guru-armv5/ 20:43:18 haha self plug 20:43:25 sorry 20:45:19 the only device that would work the best with anaconda is the trimslice or something with a hard disk 20:45:29 something useful that can be installed to? 20:46:06 what we have been doing is creating 'disk images' that can be copied to an SDCard (using livemedia-creator) 20:46:06 fossjon_room, lets perhaps talk after the meeting if you want to give it a shot.. 20:46:28 ok cool 20:46:33 we have created images for vexpress (qemu), panda, and trimslice 20:46:53 is this process documented? 20:47:08 yes, but not up to date 20:47:08 * masta would love to reproduce 20:47:21 * dmarlin still has that on his todo list 20:47:50 https://fedoraproject.org/wiki/Architectures/ARM/Installer 20:48:40 I need to update that for F18 and other platforms (vexpress and panda) 20:49:34 shall we wrap, or anything else? 20:50:14 dmarlin: thx, I'll give it a try to make an image 20:50:50 going... 20:51:10 gone! 20:51:13 #endmeeting