21:00:37 #startmeeting Fedora ARM weekly status meeting 21:00:37 Meeting started Wed Jan 30 21:00:37 2013 UTC. The chair is pwhalen. Information about MeetBot at http://wiki.debian.org/MeetBot. 21:00:37 Useful Commands: #action #agreed #halp #info #idea #link #topic. 21:00:37 #chair pwhalen jonmasters bconoboy ctyler pbrobinson dgilmore 21:00:37 Current chairs: bconoboy ctyler dgilmore jonmasters pbrobinson pwhalen 21:00:44 .fas pwhalen 21:00:46 pwhalen: pwhalen 'Paul Whalen' 21:00:50 .fas jcapik 21:00:52 jcapik: jcapik 'Jaromír Cápík' 21:01:15 .fas dmarlin 21:01:16 dmarlin: dmarlin 'David A. Marlin' 21:01:17 * masta is here 21:01:18 here 21:01:19 finally 21:01:22 .fas vicodan 21:01:22 dan408: vicodan 'Dan Mashal' 21:01:22 .fas parasense 21:01:25 masta: parasense 'Jon Disnard' 21:01:27 1st meeting evar made 21:01:32 .fas blc@ 21:01:33 bconoboy: blc '' 21:01:34 hola 21:02:05 #topic 0) Status of ACTION items from previous meeting 21:02:18 #info Last weeks minutes and action items: 21:02:26 #link http://meetbot.fedoraproject.org/fedora-meeting-1/2013-01-23/fedora-meeting-1.2013-01-23-21.01.html 21:02:32 .fas jonmasters 21:02:33 jonmasters: jcm 'Jon Masters' 21:03:15 bconoboy, finished the vexpress script, works great and reduces it to a single script 21:03:47 sweet! 21:04:01 jonmasters, has a couple of the items on the go, completed the most important for our release 21:04:14 I've added notes to the growing release notes 21:04:17 I am now debugging omap 21:04:24 ...then working on curing a rainy day 21:04:42 summer's eve helps 21:04:42 masta is working on ltrace now 21:04:58 masta: any update? 21:05:19 nope... I kinda forgot about it until today... I'm going to build it in mock and findout why it fails 21:05:23 pwhalen: photos link mailed out just now 21:05:34 * masta pulls down the ltrace SRPM 21:05:34 oatley's got the tarballs on the radar and will have them out when RC images are available 21:05:54 pwhalen: following up with Linaro on LLVM when I visit next week. Also mongodb has taken a backseat but will poke when done with F18 21:06:08 jonmasters: did mongodb need love on armv7? 21:06:33 yes 21:06:38 (I already explained that) 21:08:17 did I miss anything there? 21:08:42 I think jonmasters said he fixed it.. ? 21:08:49 there were a whole lot of actions... 21:09:10 jonmasters: rescheduling fudcon wrapup meeting? 21:09:38 right, I will be taking that one over 21:09:43 ah, cool 21:10:04 Let's record these... 21:10:12 #action jonmasters to continue on mongodb next week 21:10:12 bconoboy: yes, already spoke with pwhalen about it 21:10:22 #action pwhalen to organize fudcon wrapup meeting 21:10:34 #action masta to continue ltrace investigation 21:10:51 #action jonmasters will raise llvm issue with linaro 21:11:03 Oh, kernel dtb subpackage? 21:11:26 it's not going to be a subpackage 21:11:30 dgilmore has the ball 21:11:46 okay, we'll come back to that 21:11:58 next? 21:12:01 jonmasters: did fudcon whiteboard iamges go out? 21:12:14 jonmasters: right 21:12:26 bconoboy: yes 21:12:35 bconoboy, got them 21:12:41 super 21:12:42 next! 21:12:45 #topic 1) Current Problem packages 21:13:03 Peter is afk (away message) 21:13:13 There's two packages, in process 21:13:44 glibc: There's a bug here that is the cause of weird problems where shared libraries can't be found. https://bugzilla.redhat.com/show_bug.cgi?id=905184 21:13:50 Carlos is working on it now 21:14:09 at some point I would be interested to know why we went from 3.6.11.3 to 3.6.10 in the RC but we're not changing it now 21:14:24 #info glibc: There's a bug here that is the cause of weird problems where shared libraries can't be found. 21:14:27 webkitgtk3: Another issue where non-JIT enabled architectures can't build the latest. Have a bz, one sec... 21:14:32 #link https://bugzilla.redhat.com/show_bug.cgi?id=905184 21:14:37 #link bz for glibc bug is https://bugzilla.redhat.com/show_bug.cgi?id=905184 21:14:41 #undo 21:14:41 Removing item from minutes: 21:14:42 sorry 21:14:58 https://bugzilla.redhat.com/show_bug.cgi?format=multiple&id=905593 21:15:04 that's the webkitgtk3 bz 21:15:17 #info webkitgtk3: Another issue where non-JIT enabled architectures can't build the latest 21:15:20 I'm doing a test build with JIT re-enabled. It's been so long since we turned it off I thought we better see if it has magically started working 21:15:25 that bug looks really weird 21:15:28 #link https://bugzilla.redhat.com/show_bug.cgi?format=multiple&id=905593 21:15:40 nevermind 21:15:51 that's it AFAIK 21:16:03 #action bconoboy doing a test build with JIT re-enabled 21:16:28 jonmasters: what went from 3.6.11.3 to 3.6.10? 21:17:04 dgilmore, the kernel 21:17:13 pwhalen: it never did 21:17:24 3.6.11.3 was only ever in updates-testing 21:18:39 I could be wrong but i thought 3.6.11 shipped on x86 with 18.. 21:19:10 seano: your wrong 21:19:18 dgilmore: ok, thanks, my Panda upgraded to it but it must have had updates-testing enabled 21:19:23 dgilmore, right, I guess i found it confusing because we have updates-testing enabled 21:19:35 koji latest-pkg f18 kernel 21:19:35 Build Tag Built by 21:19:38 ---------------------------------------- -------------------- ---------------- 21:19:41 kernel-3.6.10-4.fc18 f18 jforbes 21:19:42 dgilmore ill accept that, i couldnt remember if i updated first or not. :) 21:19:59 ok, move on? 21:20:07 dgilmore: we don't disagree, just wondered, thanks for the clarification :) 21:20:08 :) 21:20:14 #topic 2) F18 final RC planning 21:20:29 i need to get the last packages signed 21:20:34 and run pungi 21:20:39 and make images 21:20:50 that is planned to be today 21:21:02 we have an omap issue. 21:21:08 pwhalen: describe? 21:21:31 * ctyler pops in late 21:22:05 have a $dayjob meeting to go to, have fun all 21:22:08 bconoboy: what issue? 21:22:19 we spun up a pre-rc for testing, when booting the 3.6.10-8 kernel, once the system boots up there is an hdmi power error displayed : 21:22:24 [ 76.405090] Failed to set PLL_PWR_STATUS 21:22:34 [ 76.405181] omapdss HDMI error: failed to power on device 21:22:40 no, we don't have an omap issue 21:22:45 I just tested it succesfully 21:22:50 jonmasters? 21:22:54 ahhh .... the HDMI/DVI one 21:22:57 pwhalen: there are known to be some issues with a few monitors 21:22:57 with display? 21:23:02 pwhalen: I would chalk this up to that 21:23:04 pwhalen: indeed 21:23:17 ok, great, we needed a second panda 21:23:23 Fedora release 18 (Spherical Cow) 21:23:24 Kernel 3.6.10-8.fc18.armv7hl.omap on an armv7l (ttyO2) 21:23:31 and the oops? 21:23:36 pwhalen: did beta work with your monitor without that issue? 21:23:39 jonmasters: to be clear, you took the new kernel and it booted to xfce successfully? 21:23:57 dmarlin, it did.. 21:23:59 bconoboy: or the new image 21:24:02 ? 21:24:11 bconoboy: yes, not meaning to be offensive, but when I say I tested it, I mean I am currently logged in on a graphical console using Xfce 21:24:23 pwhalen: so it should not be a monitor issue 21:24:25 okay, we don't have an omap issue. we just need to get a new one for pwhalen. 21:24:48 now to be fair, I am booting a Panda ES rev EB3 21:24:49 dgilmore: how will you be making images? 21:25:04 and it's possible pwhalen has an older one. There are a bunch of known issues with the DSS output 21:25:17 I'm inclined to say this new kernel has unlikely made it any worsr 21:25:29 bconoboy: right now i plan to make them using livemedia-creator 21:25:36 bconoboy: in a chroot 21:25:36 pwhalen: have you power cycled the monitor, and ensured it was turned on during boot from the moment you applied power to the board? 21:25:52 pwhalen: bconoboy: jonmasters: I have the same issue here 21:25:55 jonmasters: if the beta image worked, and the new one does not (on the same hardware) 21:25:56 jonmasters, I did.. 21:26:10 i can hook up a monitor to my panda 21:26:15 jonmasters: then is it hardware? 21:26:15 Sounds like pwhalen needs a new 60" HDMI to test. 21:26:25 ctyler: +1 ") 21:26:28 dgilmore: Why not use the images dmarlin is creating? It's what we used for f18-beta, and means the least amount of churn. it's also done with livemedia-creator. 21:26:35 ill take his old 60 " test monitor. :) 21:26:37 when using the beta image (3.6.10-6) and upgrading to 3.6.10-8, there is no problem 21:26:56 guys .... the issue appears randomly 21:27:07 dgilmore: I have not had much success with livemedia-creator in a chroot 21:27:17 ctyler, I will drop by and grab 1042's :) 21:27:22 dgilmore: device discovery sometimes fails 21:27:38 If it's randomly occuring I'd rather not block on it- instead let's get people onto 3.7. IF it presents itself there we'll fix it post-release. 21:27:41 dmarlin: shouldnt happen when you bind mount /dev 21:27:53 bconoboy: +1 21:27:53 dgilmore: I did 21:27:56 bconoboy, +1 21:28:09 dgilmore: but maybe I missed somethign 21:28:09 dmarlin: thats what we have to do for applances and livecds for x86 21:28:11 something 21:28:12 bconoboy: it disappeared with built-in PANEL_TFP410 in my case 21:28:31 bconoboy: im going to try do it right 21:28:44 bconoboy: and i will seek feedback from dmarlin 21:28:45 dgilmore: bcl said my results were to be expected when running novirt 21:28:55 dgilmore: If you do not succeed by say... tomorrow afternoon. Would you agree to use dmarlin's images then? 21:29:12 * jonmasters will arrange for pwhalen to get a nice shiny new display 21:29:30 bconoboy: i wont commit 21:29:34 I believe he does urgently require something shiny from Samsung, justifiable on the grounds that those displays work. I'll send email. 21:29:35 I have no desk space, I have 5 right now on my desk :) 21:29:41 :) 21:29:55 dgilmore: we need time to test whatever images are made in order to populate mirrors and release tuesday. 21:30:55 bconoboy: well we need to put on the mirrors Friday 21:31:04 so we need to test and sign off by then 21:31:12 dgilmore: exactly 21:31:25 can we get the images made today or early tomm, then have a vfad test tomm afternoon? 21:31:37 #info PandaBoard successfully running Xfce on 3.6.10-8: http://www.flickr.com/photos/jonmasters/8430085259/ 21:31:42 dgilmore: I'm cool with pungi for the installer stuff- we did that for the beta. Changing image generation at this point strikes me as an f19 thing. 21:32:03 dmarlin: Correct me if I'm wrong, but we have images right now, yes? 21:32:27 bconoboy: we cant we dont have a final tree 21:32:33 bconoboy: we have 3, but I want to respin them... had a couple of typos in the kickstart (now corrected)... 21:32:38 we may have some testing images 21:32:49 bconoboy: and we'll need to use the official repo for the final 21:32:56 dgilmore: Indeed, we used a bleed repo 21:32:57 (signed packages) 21:33:41 right, I requested we use a bleed repo for generating a testing image to get this moved along 21:33:51 So sometime in the next 24 hours the presumed final signed packages will be out there. Let's make RC images and test them out at this time tomorrow. 21:33:53 after the release we will migrate the kickstarts to the spin sig 21:33:57 we all know we need signed packages, and once those are available, official images will be created 21:34:09 bconoboy: ive not signed all the packages yet 21:34:16 dgilmore: That's happening today? 21:34:25 so the bleed repo would have to have unsigned ones in it 21:34:29 bconoboy: yes 21:34:44 #action dgilmore to sign all packages for f18arm final today 21:34:57 dgilmore: when will that be done and out in the wild? 21:35:25 bconoboy: if the kernel and js are signed, will we still need a bleed repo? 21:35:25 dgilmore: yea, the bleed repo was intentionally created with unsigned packages, only for testing 21:35:35 bconoboy: assumimg all are tagged right they will be on mirrors tomorrow signed 21:35:43 excllent 21:35:48 jonmasters: we use a bleed repo on primary 21:35:55 but we make sure the rpms are signed 21:35:55 I'm aware of that also 21:36:02 and that :) 21:36:04 dmarlin: presumably not- if it's all on the mirrors 21:36:13 bconoboy: cool 21:36:20 dgilmore: So, final image generation one way or another can happen tomorrow? 21:36:39 jonmasters: just stating that there is no way we have images that could potentially be considered final, contrary to bconoboy's statement that he has some 21:36:51 bconoboy: it should happen tonight 21:36:56 I'm sure dmarlin and dgilmore will figure out whether a bleed is still needed, but it sounds like not. And it sounds like the final image can be made tomorrow and made available for a VFAD that pwhalen will immediately organize 21:37:20 we'll just give it a quick poke as an RC then bless as release on Tue 21:37:20 dgilmore: They've the final bits, maybe not signed, but no need to be pedantic... they're testable 21:37:43 #info signed rpms should be available tonight 21:37:45 bconoboy: I specifically asked that they not be released outside of RH so as to not upset dgilmore 21:37:51 jonmasters: sure 21:37:59 we'll have testable images tomorrow, anyway, we're going in circles here 21:38:06 =( 21:38:19 #action dgilmore to sign packages, dmarlin to co-ordinate with dgilmore and generate final images tomorrow (Thu) 21:38:20 #action dgilmore &/| dmarlin to create images once the packages are signed 21:38:27 #undo 21:38:27 Removing item from minutes: 21:38:30 yeah, that's better 21:38:42 are we going to have a vfad for testing tomorrow? 21:38:45 #action pwhalen to organize a VFAD immediately upon the release of those images for the same day (even an hour later) for a quick soak test 21:38:46 we are! 21:39:00 =) 21:39:01 excellent! 21:39:02 tomorrow +1 21:39:26 sounds good.. 21:39:33 #action images will be announced as RC, only acceptable reason to delay is a major regression, otherwise these *are* going out as GA on Tuesday (2013/2/5) 21:40:00 #topic 3) Mass Rebuild - Does anything need to be done? 21:40:03 The earlier the better. In case there is are items we might be able to respoin.. 21:40:33 #action dgilmore, astone, and msalter are co-ordinating on getting a small aarch64 related change into mass rebuild on PA 21:40:43 #undo 21:40:43 Removing item from minutes: 21:40:50 #info dgilmore, astone, and msalter are co-ordinating on getting a small aarch64 related change into mass rebuild on PA 21:40:55 The current missing package count is 1200- I'm concerned this is going to impair our ability to mass-rebuild 21:41:08 I hope the mass rebuild happens after the PHX gear is racked 21:41:11 dgilmore: any comment? 21:41:18 pwhalen: please co-ordinate a hitlist with pbrobinson et al of things we need to play whackapackage on 21:41:35 we'll then prioritize that activitity as F18 will be done 21:42:02 To confirm: mass rebuild is happening after PHX cut-over, right? 21:42:21 bconoboy: mass rebuild has shifted from Feb 1 to Feb 8 21:42:32 gives us enough time to be moved to phx 21:42:33 oh thats good 21:42:40 dgilmore: great! 21:42:43 jonmasters, will do 21:42:48 =) 21:43:02 #action pwhalen to co-ordinate a hitlist with pbrobinson et al of things we need to play whackapackage on 21:43:13 we should be helped by not rebuilding v5tel 21:43:14 pwhalen: I will see Peter this weekend (and buy him free beer) and we will discuss packages 21:43:32 jonmasters, excellent, and the shirt1 21:43:36 jonmasters: s/beer/whiskey/ 21:43:38 pwhalen: and that too 21:43:41 cool 21:43:50 * jonmasters will have some vegan non-fat non-fun fruit juice 21:44:10 aka wine? 21:44:25 too much fun in wine 21:44:27 * dgilmore suggests jonmasters has a good belgium beer 21:44:32 too much C2H5OOH 21:44:37 anything else before the rebuild? 21:44:37 dgilmore: +1 21:44:38 stella 21:44:48 pwhalen: we just need the hit list 21:44:54 fossjon: funny story, I did some consulting for a company that was next to the Stella factory 21:45:03 heh 21:45:09 fossjon: that whole town stinks, and it's not a good smell :) 21:45:14 ewww 21:45:27 :) 21:45:34 ok, I'm repacking stuff, need to get on a plane. Will be talking about Fedora ARM at FOSDEM this weekend 21:45:43 #topic 4) Your topic here! 21:45:46 sounds like OMAP is all set and you guys have this 21:45:49 see ya! 21:45:55 #action jonmasters to do us all proud as FOSDEM 21:46:02 jonmasters, safe flight 21:46:20 have fun 21:46:24 pwhalen: note, I'm packing an HDMI cable just in case :) Will be at my folks' place tomorrow 21:46:25 [gone] 21:46:31 pwhalen: armv5tel is shut off for f19 21:46:42 #info armv5tel is now shut down on f19 21:46:47 excellent! 21:46:52 ... and there is much rejoicing 21:47:04 i like my armv5tel ;) 21:47:09 me too 21:47:11 * ctyler does soft-float happy dance 21:47:15 todays rawhide removed the arm tree 21:47:24 * masta marks the day 21:47:25 and we have disabled it in koji 21:47:29 ... and there was much disk space savings 21:47:46 life will become nice without armv5 atomics 21:47:56 true that 21:48:10 masta: true, but still can't afford an upgrade. lol 21:48:19 dgilmore: any comments on migration plan? will we see much downtime? 21:49:20 bconoboy: really depends 21:49:31 bconoboy: hopefully there will not be much 21:49:39 but if the last rsync is slow 21:49:42 its slow 21:50:04 with the faster builders, we should be able to make up any lost time in short order 21:50:08 im trying to keep /mnt/koji as close as possible 21:50:28 dmarlin: thats the idea 21:51:16 wow.... lot of things happening this week, new builders, removing v5, a release.... 21:51:16 dgilmore: will the builders all be running f18 from our final (soon to be released) ? 21:51:31 dmarlin: ideally yes 21:51:37 dgilmore: very cool 21:51:40 ya it is a bit of stuff to do masta 21:52:21 anything else folks? 21:52:36 i have nothing 21:52:51 did we document that the arm server hardware has landed? 21:53:02 probably not 21:53:16 #info ARM servers have arrived in PHX. Hope to have them installed soon. 21:54:06 ill give you until midnight to have them up and running, and a full rebuild done. :) 21:54:09 ok so I'm going to try to get a chromebook image out by ~ monday next week... based on the final release 21:54:18 ^^ a remix ^^ 21:54:24 masta: what kernel? 21:54:33 masta, cool, looking forward to trying it 21:54:37 ctyler: the masta kernel 21:54:49 ctyler: with all the fixes 21:54:49 We're looking to hook up a fairly standard usb-eth adapter, will it support that? 21:55:05 Lots of module goodness? 21:55:18 ctyler: I'll let you beta test before it is announced so we can fix that if there is an issue 21:55:29 lots of modules 21:55:42 dmarlin: are you going to add the f18-highbank.ks file to the wiki? 21:55:45 well, mostly just fixes for AUDIT_SYSCALL 21:55:52 masta: cool. --> fossjon has our chromebook 21:55:52 and various other things, like the audio 21:55:53 mlangsdorf: I can 21:56:10 audio works now? 21:56:16 masta ctyler cool 21:56:24 #info Masta working on a chromebook remix, hopefully out by monday next week 21:56:36 mlangsdorf: once we have the final repo to point to 21:56:42 dmarlin: I think it's better/more stable than the f17 version. 21:56:54 mlangsdorf: very good 21:57:07 mlangsdorf: we like it. :) 21:57:14 seano: supposedly... the fix is in my kernel, but I've too chicken to test it 21:57:20 seano: the audio 21:57:51 thanks all! 21:57:54 #endmeeting