**** BEGIN LOGGING AT Mon Jun 18 02:59:58 2012 Jun 18 07:20:19 infinity: re: SGX for omap3/armhf. i don't know ;-( Jun 18 16:08:48 janimo: *poke* Jun 18 16:09:17 janimo: Any chance you'd have some Ubuntu time this week to look at why mono on armel is broken? It's starting to hold up the world. Jun 18 16:10:45 hello Jun 18 16:14:08 i need some help, if a kind soul would spare a few minutes of their time :| Jun 18 16:21:16 hey Jun 18 16:25:16 oh well. have fun Jun 18 20:37:04 Hi...is it possible to build a kernel for a PNA? I've seen some Windows CE devices wich ran on linux...i would like to use ubuntu, because i use ubuntu since version 5, but to be honest, i have no idea how to build a kernel remotly, and get it booting on a win CE device...although i know how to get into bios and how to flash the rom on that device Jun 18 20:39:19 hello folks.. Jun 18 20:39:26 I have a quck question Jun 18 20:39:35 after editing /boot/boot.script Jun 18 20:40:14 do I have to run anything for the changes to be applied? like the equivalent of "grub-update" ? Jun 18 20:44:03 alexmoldovan1: flash-kernel Jun 18 20:44:03 alexmoldovan1: flash-kernel I believe based off of: http://www.omappedia.com/wiki/Ubuntu_on_OMAP_FAQ#I_want_to_install_Ubuntu_on_external_USB_hard_disk_instead_of_sluggish_SD_card Jun 18 20:45:58 infinity, TypoNAM thx.. Jun 18 20:46:00 ..rebooting Jun 18 20:48:13 nice...now I can see boot messages and luks password prompt on my monitor instead of the serial console Jun 18 20:48:14 :) Jun 18 21:40:03 I am trying to do the netboot, following this, http://testcases.qa.ubuntu.com/Install/ARM/NetBoot Jun 18 21:40:23 to get 12.04 armel Jun 18 21:40:58 for an Overo, and it cannot find the ethernet devices Jun 18 21:45:54 cvanvliet: Which means, probably, that our kernel doesn't support your network device... Jun 18 21:46:00 ok Jun 18 21:46:18 I was fairly sure, but wanted to see if anything I was missing Jun 18 21:46:27 thanks Jun 18 22:02:10 rsalveti: it would be nice to have robclark's xf86-video-omap in 12.10. we just noticed today that someone packaged it for debian/experimental. who can help on the ubuntu side? Jun 18 22:02:37 ndec: If it's in experimental, all we need to do is sync it. Jun 18 22:02:47 yup Jun 18 22:02:53 sync should be enough already Jun 18 22:03:15 ndec: Does it, like, work? Jun 18 22:03:21 ndec: I'm happy to sync it right now. Jun 18 22:04:01 Oh, I see a nice RC bug on it. :P Jun 18 22:04:12 http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=671897 Jun 18 22:04:13 Debian bug 671897 in src:xf86-video-omap "xf86-video-omap: current binary segfaults, please update to fixed 0.3.0 release" [Grave,Open] Jun 18 22:05:14 Filed by the maintainer... Jun 18 22:05:18 * infinity scratches his head. Jun 18 22:06:07 Ahh, it's blocked by http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=667572 Jun 18 22:06:08 Debian bug 667572 in src:libdrm "libdrm: please provide OMAP API on armel and armhf" [Wishlist,Open] Jun 18 22:06:54 And that may be blocked by our kernel supporting the KMS bits correctly. Jun 18 22:06:56 yeah, you'll need to add --enable-expermental-omap-api (or something roughly like that) when building libdrm Jun 18 22:07:13 robclark: Plus making sure our kernel has the omap kms bits required. Is that true in quantal? Jun 18 22:07:26 possibly you'll need the upstream omapdrm kernel stuff.. Jun 18 22:07:35 so should be true in 3.4+ Jun 18 22:07:42 any upstream since 3.3 would work. Jun 18 22:07:46 ah 3.4... Jun 18 22:07:49 I think an older version was backported for 3.2 Jun 18 22:07:57 which is why 12.04 doesn't have it Jun 18 22:07:59 Right, don't care about 3.2... Jun 18 22:08:00 quantal will be 3.5... Jun 18 22:08:17 robclark: Can you verify that our 3.4 kernels in quantal actualy have the right CONFIG bits turned on? Jun 18 22:08:37 robclark: If not, filing kernel bugs would be helpful. If so, then we can do nasty things to libdrm and such. Jun 18 22:08:40 perhaps.. can someone point me at the kernel for quantal? Jun 18 22:08:42 infinity: on our side we will work with the debian maintainer since we started our own package. Jun 18 22:09:04 robclark: it's the linaro kernel, so it does. Jun 18 22:09:08 ndec: The Debian maintainer seems to know what he's doing here, he's just waiting on the Debian kernel and libdrm to support what he needs. Jun 18 22:09:15 We can certainly move ahead of his packaging for now. Jun 18 22:09:20 ahh, ok.. then you are good Jun 18 22:09:25 ndec: eh? Jun 18 22:09:26 that's the same kernel we use for our releases. Jun 18 22:09:34 ndec: We don't use the Linaro kernel. Jun 18 22:09:43 infinity: yes, that's what I meant, we will get rid of our packages, and move to the 'upstream' one Jun 18 22:09:56 infinity: yes, you do. what else would you use? Jun 18 22:10:05 Our omap3 kernel is pure mainline, and our omap4 is the linaro LT sources, but with our configs. Jun 18 22:10:21 infinity: sure... that's what i call 'linaro kernel' Jun 18 22:10:24 Hence why I asked about the confics. Jun 18 22:10:27 configs* Jun 18 22:10:40 i see. Jun 18 22:10:44 you need DRM_OMAP. Jun 18 22:11:04 Is xf86-video-omap usable on all omap platforms (or, rather, the two we ship), or is it poorly named? Jun 18 22:11:30 that would be really good for us if you get that as early as possible on 12.10... it's a much better non accelerated X driver at least. Jun 18 22:11:48 infinity: robclark has tested it on beagle, iirc Jun 18 22:12:07 And omap4 as well? Or just omap3? Jun 18 22:12:11 (This isn't clear to me) Jun 18 22:12:16 the only trick indeed is the experimental flag in libdrm, and the new package that results out of it. Jun 18 22:12:33 it is supposed to work on on beagle (although perhaps sans sgx stuff).. Jun 18 22:12:35 omap4 and omap5 is what we care about these days, so that work on them for sure. Jun 18 22:12:47 Mmkay. Jun 18 22:13:04 I probably test it more on o4/o5.. but it is intended to support o2+ (but I have no o2 hw, so essentially o3+) Jun 18 22:13:18 Yeah, we don't support any omap2 hardware anyway. Jun 18 22:13:38 right, I didn't think you'd mind too much if I don't test it on o2 :-P Jun 18 22:15:37 Right, so. I don't have the time this week to own this. The two best options would be to either file a bug on libdrm and assign it to me (and I'll deal with it later), or better yet, work with the Debian maintainer on getting the upstream stuff in good shape, and either sync when it is, or push it to me because Debian's blocking on something that doesn't affect us. Jun 18 22:15:45 well, if we got the 2d working fine on beagle it's good already Jun 18 22:16:00 the current beagle sgx driver is not supported anymore, because of the wrong X abi Jun 18 22:16:02 infinity: it's already upstream in libdrm. Jun 18 22:16:28 ndec: Yes, I know. I meant "upstream" as in Debian here, not upstream upstream. Jun 18 22:16:39 argh. Jun 18 22:16:51 (But I'm happy to just do it directly in Ubuntu too) Jun 18 22:16:53 let me file a bug in LP for now... Jun 18 22:16:59 Since Debian looks like they'll do it eventually. Jun 18 22:17:06 And we can re-sync later. Jun 18 22:18:32 ndec: The bug helps, yes. And assign it to me so it doesn't fall off my radar. (~adconrad) Jun 18 22:18:41 Though, I expect one of you to bug me next week anyway. ;) Jun 18 22:19:05 infinity: oops. ubuntu only has 2.4.33 libdrm. Jun 18 22:19:16 robclark: do you know is 2.6.33 is 'enough'? Jun 18 22:19:41 2.4.33 is what was in Debian when the maintainer filed the bug claiming that adding the configure flag would fix it. Jun 18 22:19:44 So... Jun 18 22:19:46 We'll see. Jun 18 22:20:13 Anyhow, I might get around to this "after hours", but if it's on work hours, it'll wait a little bit. Busy week ahead. Jun 18 22:20:18 2.4.33 is min required version Jun 18 22:20:26 robclark: Willing to trade after hours Ubuntu development for a Panda5. Just sayin'. Jun 18 22:20:34 :-P Jun 18 22:21:13 you might want a slightly newer version if possible.. or at least there were a few things fixed.. Jun 18 22:21:18 * robclark just looking at commit log Jun 18 22:21:52 will 12.10 use libdrm 2.4.36 (which doesn't exist yet, but I guess will soon)? Jun 18 22:21:54 I'm sure revving from .33 to .35 isn't rocket science. Jun 18 22:21:56 We can do that. Jun 18 22:22:15 And we can move to .36 when it releases, I'm sure. Jun 18 22:22:21 I'll talk to the desktop team about it. Jun 18 22:22:24 Put it all in the bug. :P Jun 18 22:22:36 in the nearish future, I was planning to push rotation support.. but that will introduce a dependency on some patches that went in after 2.6.35 Jun 18 22:22:45 But I see no reason we wouldn't want the latest and shiniest, unless it breaks other drivers. Jun 18 22:23:12 libdrm is usually safe-ish to take a new version.. Jun 18 22:23:20 Given the stable ABI, I sure hope it's safe. :P Jun 18 22:24:13 well, every now and then you get something like 'nouveau: pull in major libdrm rewrite' ;-) Jun 18 22:24:21 (but I didn't hear anyone complain so I guess it worked) Jun 18 22:24:21 * infinity smirks. Jun 18 22:25:53 infinity: i opened bug 1014879 Jun 18 22:25:55 Launchpad bug 1014879 in libdrm "Enable omap experimental support" [Undecided,New] https://launchpad.net/bugs/1014879 Jun 18 22:26:09 ndec: Danke. Jun 18 22:26:10 i am not sure i can assign it to you, but i added you explicitely. Jun 18 22:26:15 thanks in advance, btw. Jun 18 22:26:23 I assigned it to myself, no worries. Jun 18 22:30:18 infinity: ok. thx. time out for me , bye ;-) **** ENDING LOGGING AT Tue Jun 19 02:59:58 2012