**** BEGIN LOGGING AT Wed Jun 29 02:59:56 2011 Jun 29 08:11:20 morning all Jun 29 08:11:45 ant_work: sorry I wasn't there last night when you were trying to talk to me Jun 29 08:14:30 np Jun 29 08:15:16 I had those strange failures with oe-core (angstrom-scripts) Jun 29 08:19:57 re meta-handheld, things are moving along, at very slow pace but things are happening Jun 29 08:20:40 one hitch is that we need to sort out gnu-tar.gz... Jay7 said late last year we could have it fetched from ltg rather than in-tree, I think that would be a good idea Jun 29 08:21:05 ah, the tarball... Jun 29 08:21:25 in BSP? Jun 29 08:22:31 yeah Jun 29 08:22:43 well, it's not a tarball, just a gzipped binary Jun 29 08:22:57 but we shouldn't have binaries in the tree Jun 29 08:23:09 not executable ones at any rate Jun 29 08:23:14 actuallty all that flashingstuff can be done by shell script Jun 29 08:23:31 you have to have some way of decompressing though right? Jun 29 08:23:42 unpacking on spitz hdd will fail without resizing Jun 29 08:23:50 so you'll be in console Jun 29 08:24:01 for resizing the partitions Jun 29 08:24:12 ok Jun 29 08:24:22 I never noticed any issues that I recall Jun 29 08:24:35 really? Jun 29 08:24:42 not that I remember, no... Jun 29 08:24:48 iirc it defaults to overwriting a 9MB partition :) Jun 29 08:24:54 with images > 20MB Jun 29 08:24:59 :p Jun 29 08:24:59 maybe I did something the first time, but that was quite a while ago Jun 29 08:25:23 is there something in any instructions we have that talks about that issue? Jun 29 08:25:43 I guess repartitioning is documented as pre-requisite Jun 29 08:27:38 right, ok Jun 29 08:27:58 I have a script for flashing Jun 29 08:28:05 (using nandlogical binary) Jun 29 08:28:28 so this is a replacement for updater.sh? Jun 29 08:28:32 yes Jun 29 08:28:46 bfor all but collie Jun 29 08:28:58 right, collie has its own inbuilt flash stuff Jun 29 08:29:11 others have too Jun 29 08:29:18 so forgive my ignorance but what's the benefit over updater.sh? Jun 29 08:29:50 you don't need to extract the battery in the shadow grasping in the dark Jun 29 08:29:55 ;) Jun 29 08:31:26 (if you mean the script) Jun 29 08:33:01 tghe other C+D flasher is in the PROM, was useful when testing u-boot and zapping the emergency kernel Jun 29 08:35:11 all Zaurus share the same Angelboot bootloader Jun 29 08:36:38 NOTE: package zlib-native-1.2.5-r0: task do_populate_sysroot: Failed Jun 29 08:36:49 ^^ does it ring a bell? Jun 29 08:49:23 hmm, not specifically, let me have a look around Jun 29 08:50:10 I did like described here: http://sakrah.homelinux.org/ Jun 29 08:50:19 vanilla, for qemu Jun 29 08:51:13 is there any other useful error for the zlib-native issue? Jun 29 08:51:20 fwiw at the beginning bitbake complained pseudo was missing and would have been rebuilt Jun 29 08:52:09 yeah, that's expected - it's not so much as complaining, it's just telling you it's what it has to do first Jun 29 08:52:28 then a couple of shell errors /mismatches Jun 29 08:52:51 do you have a complete log? Jun 29 08:53:00 yes, @home ... Jun 29 08:53:02 or at least partial Jun 29 08:53:13 ok, please send it to me when you can Jun 29 08:53:18 thx Jun 29 08:53:40 which distro are you building on btw Jun 29 08:53:41 ? Jun 29 08:54:05 angstrom Jun 29 08:54:14 like khem example Jun 29 08:55:16 bitbake complained about sanity and finally broke with AttributeError: 'NoneType' object has no attribute 'find' Jun 29 08:55:31 1.13.2 Jun 29 08:58:12 er, I meant the host machine distro Jun 29 08:58:21 ubuntu/fedora/etc. Jun 29 08:59:00 Gentoo, so dash Jun 29 08:59:04 is excluded Jun 29 08:59:19 ah right Jun 29 08:59:44 o-dev builds just fine there Jun 29 09:00:30 (I had to ASSUME_PROVIDED linux-libc-headers to avoid an intermitten build race) Jun 29 09:00:41 but that was for oe-dev Jun 29 09:01:43 hmm ok, well we shouldn't have races in oe-core, if there are any they should be reported as bugs Jun 29 09:03:06 ofc I meant # ASSUME_PROVIDED += "linux-libc-headers-native" Jun 29 09:04:32 I don't think it was a race, I retried bitbake console-image and was still failing there Jun 29 09:04:52 then I removed TMPDIR and restarted from scratch: same soup Jun 29 09:08:38 I'm rstarting a quick rebuild here in office. I have a Gentoo server Jun 29 09:14:00 ok, thanks Jun 29 09:19:08 (hope I'll not kill the raid by build-heat :) Jun 29 09:19:49 heh... are you having hot weather where you live at the moment also? Jun 29 09:20:02 well, where you work Jun 29 09:20:22 yea, 34C° those days in Northern Italy Jun 29 09:20:33 that's pretty warm Jun 29 09:20:59 break for swim at 16 o' clock ;) Jun 29 09:21:43 ouch oe-core is fetching slowly Jun 29 09:22:02 yes the oe git server is a bit stretched atm Jun 29 09:22:28 60-80Kb right now Jun 29 09:26:06 ok, some oddities... Jun 29 09:26:08 http://paste.debian.net/121362/ Jun 29 09:31:52 now needing pysqlite on host...building it Jun 29 09:32:48 Pseudo is not present but is required, building this first before the main build Jun 29 09:32:50 Parsing recipes...sh: line 0: cd: None: No such file or directory Jun 29 09:32:51 sh: line 0: cd: None: No such file or directory Jun 29 09:32:53 sh: line 0: cd: None: No such file or directory Jun 29 09:32:55 sh: line 0: cd: None: No such file or directory Jun 29 09:33:25 I'm logging right now Jun 29 09:34:38 argh../usr/bin/env: cvs: No such file or directory Jun 29 09:34:41 installing it Jun 29 09:37:08 fwiw this error-tail seems the same http://paste.debian.net/121363/ Jun 29 09:37:14 so it's bitbake noise Jun 29 09:53:11 ant_work: can you do a quick check in the bitbake sources it has fetched, is the working dir clean? Jun 29 09:53:27 http://paste.debian.net/121365/ Jun 29 09:53:45 I noticed some weird things on my system with the angstrom setup scripts and git where I keep getting crap changes on top after updating Jun 29 09:53:46 not yet seen Jun 29 09:54:21 this looks really weird :/ Jun 29 09:54:56 I suppose I have same python at home Jun 29 09:55:58 bitbake seems ok.. Update version to 1.13.2 Jun 29 09:58:35 * Starting Python Updater... Jun 29 09:58:36 * Main active version of Python: 2.6 Jun 29 09:58:38 * Active version of Python 2: 2.6 Jun 29 09:58:39 * Active version of Python 3: 3.1 Jun 29 09:59:48 I'm rebuilding python just in case Jun 29 10:00:11 in fact, the migration 2.6 -> 3.1 is painfull Jun 29 10:01:26 hmm... Jun 29 10:01:28 [oe-commits] Tom Rini : file: Fix relocation patch Jun 29 10:06:54 bitbake isn't compatible with v3 Jun 29 10:07:13 so I hope it's not trying to use any part of it... Jun 29 10:10:21 ah, I lack quilt and chrpath on this host Jun 29 10:11:29 heh Jun 29 10:11:35 chrpath ! Jun 29 10:12:05 build is now going further Jun 29 10:12:31 did the sanity check not flag those up? Jun 29 10:12:49 no it seems Jun 29 10:13:00 hmm, it should have Jun 29 10:13:12 it caught cvs Jun 29 10:14:12 and pysqlite Jun 29 10:14:56 next time you get a chance, try removing chrpath and then do "which chrpath" and see what it returns Jun 29 10:14:58 now, I've added chrpath and quilt (which added ed and diffstat) Jun 29 10:15:11 ok, buil failed :/ Jun 29 10:15:20 same zlib Jun 29 10:15:24 -native Jun 29 10:15:28 moment Jun 29 10:15:44 you have all of these right? patch help2man diffstat texi2html makeinfo cvs svn bzip2 tar gzip gawk hg chrpath wget cpio Jun 29 10:16:31 here in office I have to check but at home I had all what was needed for oe-dev Jun 29 10:17:19 http://paste.debian.net/121372/ Jun 29 10:20:33 bluelightning: I'll retry later at home. I have an Ubuntu lucid installed as well Jun 29 10:21:04 well I was using lucid, using maverick now on my main system Jun 29 10:21:06 let's hope is Gentoo-specific Jun 29 10:21:28 it may well be, esp. if nobody else reported that exact issue (I can't see anyone talking about it via google) Jun 29 12:06:22 it looks like gen-site-config is Poky stuff Jun 29 20:18:05 ping Jay7 Jun 29 20:18:33 just wanted to let you know that I got kexeboot working, with some modifications: I bypass the menu entirely and just use the highest priority one Jun 29 20:45:10 Now I just have to fix kexec :p **** ENDING LOGGING AT Thu Jun 30 02:59:57 2011