**** BEGIN LOGGING AT Thu Jul 28 02:59:57 2011 **** ENDING LOGGING AT Thu Jul 28 04:47:30 2011 **** BEGIN LOGGING AT Thu Jul 28 04:42:15 2011 Jul 28 05:10:24 Circuitsoft: why don't you throw it into the OE bugtracker Jul 28 05:10:26 ? Jul 28 06:57:41 Marex: OE bugtracker is dead afaik. Nobody cares about it. Jul 28 06:57:54 better to send bugreports to ML Jul 28 06:59:04 Jay7: I'm old school Jul 28 06:59:10 morning all Jul 28 06:59:13 yeah.. Jul 28 06:59:19 morning, bluelightning Jul 28 06:59:24 bluelightning: good Jul 28 06:59:35 bluelightning: good morning Jul 28 06:59:37 Jay7: but don't worry ... I don't care either ;D Jul 28 06:59:38 Marex: sleep well Jul 28 06:59:53 hi sgw, Jay7 Jul 28 07:00:09 bluelightning: who said a thing about sleeping, I just drank a coffee to avoid that ;D Jul 28 07:00:29 Marex: ah ok :) Jul 28 07:00:42 bluelightning: been trying to build atom-pc and n450 with master and failing badly! Jul 28 07:01:02 sgw: where's it falling over? Jul 28 07:01:03 bluelightning: not sure if you have looked at any of RP's code around tune Jul 28 07:01:17 sgw: I haven't built it yet no Jul 28 07:01:20 PACKAGE_EXTRA_ARCHS seem to get clobbered Jul 28 07:01:26 sgw: but I will fire off some builds today Jul 28 07:01:58 really easy test is to MACHINE=atom-pc bitbake core-image-minimal -e | grep PACKAGE_EXTRA_ARCH Jul 28 07:04:55 bluelightning: it should be: all any noarch x86 i386 i486 i586 i686 core2 atom_pc Jul 28 07:05:06 but its: PACKAGE_EXTRA_ARCHS="i586" Jul 28 07:05:18 rootfs then fails because the core2 packages are not found. Jul 28 07:05:51 something is mangling it, I have not found it yet. Jul 28 07:05:54 sgw: koen reported it yesterday Jul 28 07:06:13 [OE-core] do_rootfs broken, was: Re: [PATCH 1/3] Add ARM tune file overhaul based largely on work from Mark Hatle Jul 28 07:06:39 JaMa: this is x86 also, not just arm clearly. Jul 28 07:08:07 yup from what he said in 2nd mail it looks like generic problem for all archs Jul 28 07:08:38 Got buried under all the mail and trying to get M3 ready to see that, any hints of solutions? Jul 28 07:09:43 thumb is broken too :/ Jul 28 07:10:04 sgw: right, I see the same thing now... Jul 28 07:11:23 bluelightning: OK, I am not an overrides expert, I know RP is gone for the next few days, I will create an M3 branch in my morning, can you dig into this, I am not sure what else is on your plate, but we need to get this fixed ASAP. Jul 28 07:11:46 sgw: sure, I have nothing this important on my plate at least Jul 28 07:11:58 sgw: will see what I can find Jul 28 07:12:15 ao2: hey, morning Jul 28 07:12:18 JaMa: ciao Jul 28 07:12:56 hi Jul 28 07:13:04 bluelightning: ao2: got g+ already ? Jul 28 07:13:10 bluelightning: thanks. I am still but an egg when it comes to the core of bitbake Jul 28 07:13:36 Marex: not yet, no, had the invites but haven't been bothered to sign up Jul 28 07:13:49 Marex, I don't need it for now Jul 28 07:14:02 sgw: I'm still learning every day :) Jul 28 07:14:07 bluelightning: I just got in ... I can track a few interesting people so I use it as a news feed Jul 28 07:14:31 though I don't produce any "news" myself ... >:-) Jul 28 07:14:46 * Marex is badly ditching the concept of social networking by being antisocial Jul 28 07:16:53 bluelightning: pull in fray when he wakes up if you can. Jul 28 07:17:05 sgw: ok, good idea Jul 28 07:18:23 Marex: cau Jul 28 07:19:35 bluelightning: aren't you comming for ELCE2011 ? Jul 28 07:19:40 bluelightning: let's have a drinking party Jul 28 07:19:55 ao2: and you ... ? Jul 28 07:19:55 Marex: I hope so :) Jul 28 07:20:03 bluelightning: you are comming ? cool Jul 28 07:20:21 may be dependent on whether my talk is accepted Jul 28 07:20:39 bluelightning: what talk ? :) Jul 28 07:20:52 Marex: will let you know later :) Jul 28 07:20:52 on what topic ? or is that a secret ? Jul 28 07:20:58 bluelightning: PM me :b Jul 28 07:21:00 no, not really secret... Jul 28 07:22:09 Marex, ah in Prague, I forgot about it. Jul 28 07:25:09 good morning Jul 28 07:30:23 ao2: so ... what's your status about that ? Jul 28 07:32:10 Marex, I don't know yet Jul 28 07:34:41 ack :) Jul 28 09:23:25 03Fabrice Aeschbacher  07master * r1aeea48674 10openembedded.git/classes/packaged-staging.bbclass: (log message trimmed) Jul 28 09:23:25 packaged-staging: Fix fetching from PSTAGE_MIRROR with file:// URI Jul 28 09:23:25 * This patch fixes fetching pre-built staging packages from PSTAGE_MIRROR having file:// uri Jul 28 09:23:25 see: http://lists.linuxtogo.org/pipermail/openembedded-devel/2011-July/033938.html Jul 28 09:23:25 (thanks to Tom Rini for helping me fixing this) Jul 28 09:23:26 Signed-off-by: Fabrice Aeschbacher Jul 28 09:23:26 Signed-off-by: Tom Rini Jul 28 09:56:43 morning all Jul 28 10:05:41 hi pb_ Jul 28 10:15:29 hi florian, pb_ Jul 28 10:15:40 hi bluelightning Jul 28 10:30:42 hi all Jul 28 10:37:36 hi lumag Jul 28 10:37:57 hi florian Jul 28 12:04:20 pb_, I have a shell thing that makes git do tab completion, I wonder if we should figure out how that works and see about a similar thing for oe-core? Jul 28 12:05:31 although I fear it might slow things down Jul 28 12:24:08 JaMa: stop debating about a temporary layer. Do as you think is correct, remembering all will go to meta-handhelds and you'll have to let bluelightning decide for the styling Jul 28 12:24:15 thanks again for your time Jul 28 12:27:05 ok Jul 28 12:27:45 I was trying to show what's wrong in new tar.gz series.. Jul 28 12:28:42 np, this week I have strange free-time windows (1-3 AM) :/ Jul 28 12:29:21 I'm not surprised by the outcomes :) Jul 28 12:29:36 luckily you are the filter Jul 28 12:30:19 and I have to add this week is rather 'beefy' in term of oe-core change Jul 28 12:30:28 bad luck ;) Jul 28 12:30:52 Crofton: I did see more than completion: yesterday doing -c patch on a WIP recipe another term opened, well, more than one I suppose. Jul 28 12:31:05 I closed 3 or 4 Jul 28 12:31:15 must be a recent change Jul 28 13:06:45 what is the semantic of the new TUNE_CCARGS? Are these really only tuning parameters (-mtune=), or are options like '-mcpu=' allowed which generate code which might be executed only on this cpu? Jul 28 13:38:24 03Mario Schuknecht  07org.openembedded.dev * r7712977b6e 10openembedded.git/recipes/linux/ (linux-2.6.24/hipox/defconfig linux_2.6.24.bb): Jul 28 13:38:24 linux-2.6.24: enable tun device support (hipox machine only) Jul 28 13:38:24 Signed-off-by: Mario Schuknecht Jul 28 13:38:24 Acked-by: Steffen Sledz Jul 28 13:38:32 03Mario Schuknecht  07org.openembedded.dev * r6786333d52 10openembedded.git/recipes/linux/ (linux-2.6.24/hipox/hipox-gmac-deadlock.patch linux_2.6.24.bb): Jul 28 13:38:32 linux-2.6.24: fix deadlock situation in gmac driver (hipox machine only) Jul 28 13:38:32 Signed-off-by: Mario Schuknecht Jul 28 13:38:32 Acked-by: Steffen Sledz Jul 28 13:43:22 Crofton: worth a try, though my experience with custom bash completions has been that generally they are so slow as to be more annoying than helpful Jul 28 13:48:58 yeah Jul 28 13:49:08 the git one I have can feel that way at times Jul 28 13:51:15 the annoying thing about the git one is it gets in the way of the standard file/path completion Jul 28 13:51:34 or rather, it doesn't integrate it Jul 28 13:51:40 ah, that sucks Jul 28 13:52:07 the dpkg one is like that too, if you go to deploy/ipkg and do "dpkg-deb -c libc6" it doesn't do what you want. Jul 28 13:52:42 I should try to teach it the difference between dpkg and dpkg-deb some day, I guess Jul 28 13:59:57 pb_: do you foresee any longer fallout wrt tune files overhaul? There is some curiosity about destiny of armv4 say for one collie... Jul 28 14:12:21 I don't think armv4 support is going to go away. It does seem to be broken at the moment but I think that is just a bug rather than a deliberate act. Jul 28 14:13:38 sure, looks like that Jul 28 14:14:03 new tune files spell 'cortex' as 'coretex' at some places Jul 28 14:14:40 heh, that is no good Jul 28 14:14:51 just in comments, or in places where it actually matters? Jul 28 14:14:58 where it matters Jul 28 14:15:10 I was a bit startled by that "darwin" vs "darvin" thing as well :-} Jul 28 14:15:11 meta/conf/machine/include/tune-cortexr4.inc:TUNE_FEATURES_tune-cortexr4 = "armv7r vfp coretexr4" Jul 28 14:15:19 sigh Jul 28 14:15:26 ditto for m3 + m1 Jul 28 14:15:39 I am a bit annoyed that richard merged that patch, it does seem to have all kinds of things wrong with it Jul 28 14:15:51 nostalgy of oe-dev? Jul 28 14:16:37 seems we are back in the golden days :p Jul 28 14:17:03 obviously it's his tree and he is perfectly entitled to merge whatever he wants, but I'm not quite sure I understand why there was such a rush to get this landed before he went away for the long weekend. Jul 28 14:17:46 we'll forgive him for this time, isn't Jul 28 14:18:55 oh, sure Jul 28 14:20:18 feel free to submit patches, if they're straightforward then Saul should be able to merge them immediately Jul 28 14:20:33 let's hope he'll not break arm or thumb....in motorcycle ;) Jul 28 14:20:52 yes that would not be good Jul 28 14:21:00 he'll has much to type behind Jul 28 14:21:02 I'm not sure that the armv4 thing is amenable to a straightforward fix. It seems to just be misdesigned. Jul 28 14:21:30 the coretex stuff and the big-endian things probably are fixable fairly trivially, though auditing the code to make sure there aren't more of the same is probably a bit more of a challenge Jul 28 14:22:48 hi, has git.openembedded.org been moved recently? my last pull was on jul 19, and when I want to pullnow I get a dns spooing detection warning Jul 28 14:22:54 yes Jul 28 14:23:01 tom moved it from melo to some other vm Jul 28 14:23:11 you should be pulling from github though :-} Jul 28 14:26:53 ah ok, missed that this has changed, Jul 28 14:28:48 guess I missed the announcement (if any was made) Jul 28 14:30:14 eFfeM_work: there was one... probably not as loud as it could have been Jul 28 14:30:49 yeah, and life has been quite hectic recently so guess I just overlooked it Jul 28 14:31:27 anyway, what need to be done if I still want to be able to push a bugfix patch to openembedded ? Jul 28 14:31:46 I am somewhat a git n00b Jul 28 14:32:42 eFfeM_work: just edit your oe/.git/config file and change the repo push URL Jul 28 14:34:36 bluelightning: I fear it is not that trivial, currently it has remote = origin and there is only one url in the config file Jul 28 14:35:36 argh sorry Jul 28 14:35:42 see git-remote(1) Jul 28 14:35:42 (actually perhaps the wiki could use some info, then again the old openembedded tree seems to become less and less used) Jul 28 14:36:02 eFfeM_work: assuming your remote URL is not openembedded.net it should be OK Jul 28 14:36:29 then all you need to do is remove the old host key for openembedded.org in your ~/.ssh/known_hosts Jul 28 14:36:35 nope it is url = git@git.openembedded.org:openembedded Jul 28 14:36:41 yeah, that seems the simplest Jul 28 14:36:54 just wanted to know if there was an issue or so Jul 28 14:36:59 sure Jul 28 14:39:12 morning kergoth Jul 28 15:00:57 bbl, bye Jul 28 15:02:01 hey Jul 28 15:06:33 fray: ping Jul 28 15:07:08 I've sent a patch to the list for the "coretex" thing Jul 28 15:25:09 eFfeM_work: I guess we should link to the new oe-core page somewhere... I was hoping for a bit more feedback first though Jul 28 15:25:48 03Tom Rini  07master * reaf7e8181d 10openembedded.git/recipes/bluez/bluez4/fix-build-races.patch: Jul 28 15:25:48 bluez4: Add Upstream-Status to fix-build-races.patch Jul 28 15:25:49 This has been applied upstream now. Jul 28 15:25:49 Signed-off-by: Tom Rini Jul 28 16:53:32 So, I'm having a build issue related to libtool. https://gist.github.com/a1dceb578a6e73fc7153 Is any more detail necessary? Jul 28 19:23:31 are there any classes or layers that let you create users on an image for the 2011.03-maintenance branch? Jul 28 19:50:01 what do you mean? Jul 28 19:51:40 er, ah. Just the normal hacky post inst script Jul 28 19:51:47 cleaner way is in oe-core Jul 28 20:06:30 Tartarus, is the oe-core method too different to be backported to 2011.03-maintenance branch? Jul 28 20:07:03 I think so, yes Jul 28 20:08:20 oh well, i can do hacky post inst stuff or even hacky ROOTFS_POSTPROCESS_COMMAND stuff Jul 28 20:10:30 i just never like maintaining my hacks after more than 3 weeks go by Jul 28 20:56:04 JaMa: ping Jul 28 20:57:04 pong Jul 28 20:57:18 may I pm you? Jul 28 20:57:47 yes, but I'm going to bed soon Jul 28 21:10:06 hi ant__ Jul 28 21:13:01 hello pb_ Jul 28 21:14:44 hi bluelightning Jul 28 21:15:19 pb_: bluelightning: I've noticed oe-core lacks the 'ability' to unpack the files in subdirs Jul 28 21:15:19 hi ant__ Jul 28 21:15:53 ant__: really? if so we should add that capability Jul 28 21:15:55 i.e. if [ -e ${WORKDIR}/${LOGO_SIZE}/logo_linux_clut224.ppm ]; then Jul 28 21:15:58 fails Jul 28 21:16:14 when LOGO_SIZE is declared Jul 28 21:16:39 iirc Tartarus added that stuff Jul 28 21:17:23 I did? :) Jul 28 21:17:27 (btw the file resides in /files/$LOGO_SIZE) Jul 28 21:17:49 hi Tom Jul 28 21:22:33 ant__: it was Robert Schuster, revision b25264651415 Jul 28 21:23:06 my change to linux.inc was 99c8541 Jul 28 21:23:14 although it was revised by kergoth a bit Jul 28 21:24:15 should I OR it ? Jul 28 21:24:50 or remove the LOGO_SIZE stuff ? Jul 28 21:25:52 in fact this deprecated inc file is neither in meta-oe nor in oe-core Jul 28 21:27:28 I'm not sure; the suggestion for oe-core is to use linux-yocto Jul 28 21:28:06 for handheld targets it doesn't bother me too much either way as long as a working kernel gets built Jul 28 21:28:39 ^_^ Jul 28 21:28:41 meta-handheld will have linux.inc for quite some time anyway if it's nowhere else, in order to build the older kernels Jul 28 21:29:53 I don't remember which recipe was needing the subdir feature Jul 28 21:30:04 presumably for this case you could just have LOGO_SIZE in SRC_URI instead? Jul 28 21:30:08 we have just adapted to the changes Jul 28 21:32:19 I'd just revert http://cgit.openembedded.org/cgit.cgi/openembedded/commit/?h=org.openembedded.dev&id=99c8541b7184df8912b6cd07e3aaae0eaf31bf93 Jul 28 21:32:37 in oor layer Jul 28 21:46:11 bluelightning: we still have ARM_KEEP_OABI ?= "1" Jul 28 22:01:38 heh ${KERNELORG_MIRROR} Jul 28 22:02:03 seems http://kernel.org/pub/ in oe-core Jul 28 22:03:12 what has changed then? Jul 28 23:11:50 ant__: yes I noticed the latter Jul 28 23:12:01 easy to work around **** ENDING LOGGING AT Fri Jul 29 02:59:56 2011