**** BEGIN LOGGING AT Tue Jul 27 02:59:56 2010 Jul 27 03:12:23 so how do I fix it? Jul 27 03:19:28 03Chris Larson  07org.openembedded.dev * r1f9ca3be2c 10openembedded.git/recipes/perl/ (perl_5.10.1.bb perl_5.8.8.bb): Jul 27 03:19:28 Fix siteinfo issue in perl 5.8.8 and 5.10.1. Jul 27 03:19:28 Signed-off-by: Chris Larson Jul 27 04:25:50 kergoth: so how do I increase filesystem? Jul 27 04:26:26 hi Jul 27 04:26:54 when i run bitbake fso-illume-image it return error on do_config ncurses5.4 Jul 27 04:27:11 the error is " C compiler not able to create executables" Jul 27 04:27:25 can anyone tell what to do to resolve this..? Jul 27 06:15:29 anybody? I am failing on the last step: creating rootfs Jul 27 06:22:36 IgorK i cant remember, either distro or machine config, you can set the root image size Jul 27 06:22:50 i think you need to comment out that option Jul 27 06:44:30 warflyr: what is the fastest way of finding it? Jul 27 06:53:55 warflyr: http://paste.pocoo.org/show/242119 and http://paste.pocoo.org/show/242120. Which option should I use to increase it? Jul 27 06:58:39 gmorning. anyone has an oveerlay with a working x86_64 machine? Jul 27 07:04:27 nik0n: x86_64 as target or as build Jul 27 07:10:57 jo Jul 27 07:11:09 damn the gitserver makes problems again Jul 27 07:11:43 he khem Jul 27 07:11:56 khem git has some problems again Jul 27 07:12:08 hmmm Jul 27 07:12:20 hi Jul 27 07:12:27 morning blindvt Jul 27 07:12:48 woglinde: whats the error Jul 27 07:13:01 blindvt`: hi Jul 27 07:13:06 service dont react Jul 27 07:13:11 seems it hangs somewhere Jul 27 07:13:20 cgit too Jul 27 07:13:30 denial of service Jul 27 07:13:43 hmm I guess you have to wait another 9 hours Jul 27 07:13:54 everyone here is asleep Jul 27 07:14:42 jupp Jul 27 07:15:09 khem: target Jul 27 07:15:28 nik0n: look thru email Jul 27 07:15:30 ml Jul 27 07:15:35 I had it working once Jul 27 07:15:47 nik0n: there is a git branch also Jul 27 07:15:56 I hope not the wacky controller again Jul 27 07:15:56 from roman Jul 27 07:16:58 khem, okay ... read the mailinglist yesterday but wasnt able to build. gcc-cross always failed .... ok ill try again, now Jul 27 07:17:25 oh it may not be that handy anymore Jul 27 07:17:36 it was about a year or so ago Jul 27 07:17:51 i've been looking into the problem of SRCREV vs. PR=svn. I can't seem to find the spot that is responsible for setting up this name: gcc-cross-initial-svn-r0 Jul 27 07:18:36 the problem (AFAIU) is that this "r0" should really be r${SRCREV} Jul 27 07:18:48 any clues? Jul 27 07:21:42 (the problem with r0 is that no svn package gets rebuilt if SRCREV is bumped) Jul 27 07:27:09 hmuh kontron with rednumbers Jul 27 07:28:58 morning pb Jul 27 07:31:57 morning Jul 27 07:32:31 hi hrw Jul 27 07:32:36 morning Jul 27 07:33:59 doh. -PKGV = "4.6+svnr${SRCPV}" +PR = "4.6+svnr${SRCPV}" and all is well Jul 27 07:34:06 good morning Jul 27 07:35:45 ranjith: same error here Jul 27 07:36:51 ranjith: ncurses_5.4.bb failed, NOTE: Multiple libraries (libnss_nis.so.2, libnss_nisplus.so.2, libnss_hesiod.so.2) found, but LEAD_SONAME 'libc.so' doesn't match any of them Jul 27 07:39:07 mckoan: that's not an error Jul 27 07:39:26 (and, fwiw, that "Multiple libraries" message is coming from libc, not ncurses) Jul 27 07:39:41 pb__, re http://www.mail-archive.com/openembedded-devel@lists.openembedded.org/msg07945.html Jul 27 07:39:43 hi woglinde_ Jul 27 07:41:13 pb__, if i use only PKGV then i end up with foo-svn-r0 as PR and the package isn't rebuilt if i bump the SRCREV. Would it be acceptable to just set PR to "4.6+svnr${SRCPV}" (not PR_append, just PR) ? Jul 27 07:55:01 git down (again) Jul 27 07:55:21 mckoan yes Jul 27 08:04:36 blindvt: how about adding SRCREV to ${PV}? Jul 27 08:10:31 morning Jul 27 08:12:05 pb__, that doesn't work with PREFERRED_GCC_VERSION="svn" Jul 27 08:15:08 pb__, for this to work PV has to be untouched (i.e. it has to be plain "svn", from gcc-foo_svn), the PR could(should?) be e.g. 4.6-svnr${SRCPV} to trigger rebuilds if SRCREV is bumped Jul 27 08:52:08 impressive... OE does not even use --sysroot Jul 27 08:52:51 we dont? Jul 27 08:52:55 I thought we used Jul 27 08:52:57 now Jul 27 08:53:01 after -isystem Jul 27 08:53:44 yeah Jul 27 08:53:49 but not --sysroot Jul 27 08:54:28 good morning Jul 27 08:55:04 gi florian Jul 27 08:56:58 re zecke Jul 27 08:57:07 hi zecke Jul 27 08:58:14 blindvt: surely setting PREFERRED_GCC_VERSION="svn" is a bit silly anyway, since it doesn't say anything about which svn rev you want. if you want the head, you can set it to "svnr${AUTOREV}" or some such Jul 27 08:58:19 zecke: hail zecke Jul 27 08:58:24 florian: good morning Jul 27 09:16:03 I think OE on a single server is not ideal solution...what about mirrors like Gentoo? We received several hosting props iirc Jul 27 09:16:15 for free Jul 27 09:17:24 morning Jul 27 09:18:41 hi xora Jul 27 09:23:51 Hi. Anyone know what's wrong with http://cgit.openembedded.org? Jul 27 09:24:12 vhallac: it is down Jul 27 09:24:32 use the mirror Jul 27 09:24:40 Ah, thanks. Wanted to make sure it was not just me. :) Jul 27 09:24:59 anyone by any chance any experience/opinion on using http://www.dealextreme.com/details.dx/sku.42070 with oe? Jul 27 09:29:00 Oh, nice looking device, and price looks right, too. :) Jul 27 09:29:32 I wonder if the company would make the hardware specs, or at least the kernel/bootloader code available. Jul 27 09:32:22 By the way, is there any work or documents for performing in-field upgrade of an OE system? Jul 27 09:33:03 eFfeM_work: check does via wm8505 landed in mainline first Jul 27 09:33:26 eFfeM_work: This looks like an Eken M001 Jul 27 09:33:59 and how to get kernel source etc as this is a big problem with chinese stuff Jul 27 09:34:24 florian: they also have a device they name M001: http://www.dealextreme.com/details.dx/sku.39169 Jul 27 09:34:55 has 128 M (the 42070 has 256 M) Jul 27 09:35:14 eFfeM_work: ah right... that's a little difference Jul 27 09:35:42 hrw was wondering whether it was possible to run oe on it Jul 27 09:35:57 not sure if it can be rooted Jul 27 09:36:23 we have one of these at the office... didn't find any sources for these yet and even the original firmware does not do any power management Jul 27 09:36:55 http://www.mobileread.com/forums/showpost.php?p=1025376&postcount=31 Jul 27 09:37:17 "DO NOT buy one of these!" Jul 27 09:37:46 "The current WM8505 distribution depends on a Linux kernel built with the Android bells and whistles, as no source code has been made available (despite GPL obligation)." Jul 27 09:37:48 hehe Jul 27 09:37:54 http://bento-linux.org/wiki/vt8505/wm8505/android/android-kernel Jul 27 09:38:18 There are some debian ports, etc, but it looks like they just use the available android kernel. Jul 27 09:38:20 for playing with oe the smartq v5 and v7 might be a better choice Jul 27 09:39:24 eFfeM_work: OE or oe based distro? :D Jul 27 09:40:40 hrw, you know what I mean :-) angstrom or minimal Jul 27 09:42:18 humm well... one of the big archos tablets and you have a nice environment for a native oe ;-) Jul 27 09:42:45 those smartq things look nice, but seem to be a lot more $$$ (and so is archos) Jul 27 09:43:45 google also gave me this: http://www.irobotandroid.com/irobot/index.php?act=viewProd&productId=15 Jul 27 09:44:13 true... actually the archos ones are not much more expensive and pretty well supported Jul 27 09:44:39 effem whats the mirror now? Jul 27 09:45:23 woglinde_: oe mirror? see channel topic Jul 27 09:48:55 something like 'mirrorselect' would be nice feature Jul 27 09:49:17 but I think with git is different Jul 27 09:50:08 archos 5 seems to start at eur 249 around here, archos 7 at eur 149 Jul 27 09:54:22 how do you generate Packages.gz files in OE ? Jul 27 09:54:42 metaindex.bb Jul 27 09:56:20 thx ! Jul 27 09:57:24 woglinde_: is it inside OE ? Jul 27 09:57:27 yes Jul 27 09:57:41 find recipes -name "*metaindex*" Jul 27 09:57:59 [tridev@tridev OE]$ find openembedded/recipes/ -name '*metaindex*' Jul 27 09:57:59 [tridev@tridev OE]$ Jul 27 09:58:16 package-index ? Jul 27 09:58:18 I'm on dev Jul 27 09:58:35 ant_work: yes thx Jul 27 09:58:50 hm yes Jul 27 09:58:58 usualy you don't need it Jul 27 09:59:05 recipes/meta/package-index.bb Jul 27 09:59:07 image takes care Jul 27 09:59:18 ant for quick Index rebuild? Jul 27 09:59:20 I don't have an image yet Jul 27 09:59:29 I dont type bitbake fooimage Jul 27 09:59:42 I'm using task-boot and then a few other packages Jul 27 09:59:43 if your bbcache is good Jul 27 09:59:48 yoz can even use -b Jul 27 09:59:56 which fasten it up Jul 27 10:00:00 and have a script to build my rootfs Jul 27 10:00:24 nice thx guys Jul 27 10:02:25 woglinde_: I remember it just because recently I updated wiki Jul 27 10:02:32 ^_^ Jul 27 10:02:49 dj-death: it can be of interest http://wiki.openembedded.net/index.php/Advanced_configuration Jul 27 10:03:17 bottom Jul 27 10:04:01 that's kergoth's wisdom, read carefully ;) Jul 27 10:05:29 pb__: woglinde_ hi Jul 27 10:05:41 yo zecke Jul 27 10:06:10 12:04 < woglinde_> koen a quickfix is add in siteconfig.bbclass list() around itertools.chain(d.getVar("BBPATH", True).split(":"), d.getVar("FILESPATH", True).split(":"))) Jul 27 10:06:25 if anybody has siteconfig problem too Jul 27 10:13:45 is the git server down? Jul 27 10:14:04 yes Jul 27 10:14:25 ok. not just me then Jul 27 11:55:05 Hello, anyone use the Ralink RT3070 wifi dongle driver? I have some error building the recipe (rt3070_2.1.1.0.bb) Jul 27 14:29:08 shit -- now I'm getting inexplicable hangs of bitbake builds, just stops responding, end up with bitbake processes just sitting there, not executing tasks.. I had one successful build last night out of a total of .. 20 or so it was supposed to run :( Jul 27 14:29:13 * kergoth_ mutters Jul 27 14:49:07 kergoth_: doh Jul 27 15:26:38 re Jul 27 15:27:41 git up again? Jul 27 15:29:58 hi khem Jul 27 15:30:16 is it normal that my feed changed from armv6-novfp to armv6 Jul 27 15:30:43 I think not Jul 27 15:30:58 because messing armv6 without vfp and with vpf doesn't sound a good idea Jul 27 15:31:15 which armv6 has vfp and which not? Jul 27 15:31:40 htcdream: armv6-novfp Jul 27 15:31:45 bug: armv6 Jul 27 15:31:50 ugh Jul 27 15:31:58 I tried khem's patches Jul 27 15:32:04 it built fine Jul 27 15:32:06 but then.... Jul 27 15:32:10 hehe Jul 27 15:32:10 I noticed only today Jul 27 15:32:17 illegal instructions Jul 27 15:32:18 that it was renamed to armv6 Jul 27 15:32:20 hihi Jul 27 15:32:27 ah, so someone is aware that git's down? Jul 27 15:32:28 no illegal instructions Jul 27 15:32:35 cool i won't flood this channel with ?'s then Jul 27 15:32:43 just that if I share builddir Jul 27 15:32:48 jconnolly its down since 9 a.m. cest Jul 27 15:32:48 I would have huge issues Jul 27 15:32:53 ok Jul 27 15:32:57 hi jconnolly Jul 27 15:33:08 hi GNUtoo|laptop Jul 27 15:33:17 I'll come in #buglabs Jul 27 15:46:10 when i run bitbake for openssh i get this error, what dose it mean? ERROR: Please set the 'PERSISTENT_DIR' or 'CACHE' variable. Jul 27 15:56:01 cmkinne, incorrect or unset BBPATH Jul 27 15:56:13 thanks ill check thato ut Jul 27 15:57:25 specifically, your OE directory isn't in your bbpath, or if it is, check for typoes or see if you forgot to export it. that error means it found bitbake's conf/bitbake.conf rather than OE's conf/bitbake.conf. Jul 27 15:58:44 k Jul 27 15:59:43 It means that you should set the CACHE variable. Jul 27 16:00:00 Isn't that documented in the sample local.conf? Jul 27 16:00:08 (If not, we should do that) Jul 27 16:00:36 * mwester doesn't feel like working, and goes to check local.conf instead. :) Jul 27 16:01:31 mwester: IIRC is documented Jul 27 16:05:02 mwester, uh, CACHE is set in bitbake.conf. Jul 27 16:05:12 so no, that's not what it means Jul 27 16:07:29 * kergoth_ mutters Jul 27 16:09:53 re Jul 27 16:09:55 hi kergoth Jul 27 16:10:05 * woglinde mutters about the git-server Jul 27 16:10:15 hey Jul 27 16:10:38 i set the BBPATH to export BBPATH=/stuff/build:/stuff/openembedded Jul 27 16:10:44 but it still didnt seem to work Jul 27 16:10:53 im doing my bitbake commands in /stuff/build Jul 27 16:11:17 kergoth I would like to see your solution for the siteconfig stuff Jul 27 16:12:35 cmkinne: try KOM ;-) http://kom.kaeilos.com/ Jul 27 16:13:47 thanks ill give that a try Jul 27 16:19:01 well. I have Ben NanoNote now :) Jul 27 16:19:22 one more machine to support in OE :) Jul 27 16:21:53 hi woglinde Jul 27 16:54:04 ka6sox, ping Jul 27 16:55:41 jo crofton Jul 27 16:59:13 etrunko: here :) Jul 27 16:59:39 etrunko: I find building cmake non-native with native cmake neat Jul 27 16:59:47 hi zub Jul 27 16:59:51 hi woglinde Jul 27 17:00:05 etrunko: to unify the recipes, this has to be dropped IMHO Jul 27 17:00:06 zub good you read my answers Jul 27 17:00:27 zub: indeed Jul 27 17:00:42 woglinde: don't think it is possible to use BBCLASSEXTENDED here Jul 27 17:00:44 woglinde: you == Henning? :) Jul 27 17:01:04 zub yes Jul 27 17:01:08 ok :) Jul 27 17:01:16 /whois woglinde Jul 27 17:01:34 etrunko: it can IMHO, but we have to bootstrap the non-native cmake too (dunno yet how) Jul 27 17:01:43 but it won't be so nice :-P Jul 27 17:01:53 yet it would decrease number of recipes Jul 27 17:02:50 zub: yup, how would we get cmake to use cmake-native? Jul 27 17:03:21 etrunko? currently it dies, via inherit cmake; but if the recipes are to be unified, we have to drop that Jul 27 17:03:28 s/dies/does/; Jul 27 17:03:58 zub: yeah, i was meaning in the case of unifying the recipes Jul 27 17:04:08 cant inherit cmake Jul 27 17:04:24 yup Jul 27 17:04:58 http://lists.linuxtogo.org/pipermail/openembedded-devel/2010-July/022129.html Jul 27 17:05:10 the two options I see... Jul 27 17:05:34 the first one - I didn't think much about it, but it probably could be done with some ugly copypasting, basicaly pustting most of cmake.bbclass into recipe or an inc file Jul 27 17:05:42 but that wouldn't be nice Jul 27 17:06:27 etrunko: just FYI one can distinguish between native/non native "instances" of the recipe via _virtclass-native Jul 27 17:06:44 so we can e.g. pass diff. parameters to configure/whatever Jul 27 17:07:33 zub: never heard of _virtclass-native... /me sooo noob with oe Jul 27 17:07:36 :/ Jul 27 17:07:43 so am I, never mind :) Jul 27 17:09:25 yeah but the inherit may not work for only specific Jul 27 17:09:36 we need to talk to kergoth about it Jul 27 17:10:23 yes, that's why I think we'd have to bootstrap cmake non-native too Jul 27 17:10:44 ping kergoth Jul 27 17:10:51 if it's even possible, but I think it should Jul 27 17:11:18 just I'm not that sure it's worth, it decreases number of recipes, but they will be a bit more ugly Jul 27 17:11:21 woglinde, pong Jul 27 17:11:36 Crofton, pong Jul 27 17:11:47 I wanted to remote debug using gdbserver-gdb pair.... however I do not see gdb in the cross-compiler toolchain arm-angstrom-linux-gnueabi- in the path : tmp/cross/armv7a/bin Jul 27 17:12:01 Does OE not build gdb ? Jul 27 17:12:06 ka6sox, do you mind sending something to the oe-dev lis about the server issue Jul 27 17:12:27 kergoth is possible with BBCLASSEXTENDED to do an inherit for the non-native package? Jul 27 17:12:33 only Jul 27 17:12:39 bitbake gdb Jul 27 17:12:44 gitbake gdb-cross Jul 27 17:13:13 nope, you can't conditionally inherit at all, at least not easily. Jul 27 17:13:29 * zub imagines someone was using gitbake on the OE git... that would explain the issues :-P Jul 27 17:14:27 lol Jul 27 17:15:14 why is the git repo not working? Jul 27 17:15:16 kergoth okay thanks Jul 27 17:15:23 woglinde: ahh ... I think gdb-cross is what I was looking for... Jul 27 17:15:31 woglinde, zub: kergoth_'s statement makes our lives a bit more complicated then Jul 27 17:15:31 woglinde: Thanks .. lemme give it a try Jul 27 17:15:32 cmkinne ka6sox will say it us shortly Jul 27 17:15:45 k Jul 27 17:15:56 etrunko in this case I am fine with 2 packages for native and non-native Jul 27 17:16:10 you could, of course, programmatically, from an anonymous python function, run bb.parse. methods to pull it in, but then you have lots of other painful issues Jul 27 17:16:25 kergoth_: we need to patch perl to cope with siteinfo changes, I have a small patch Jul 27 17:16:25 kergoth that isnt it worth Jul 27 17:17:02 khem, how so? i pushed a perl fix for it last night that made it use SITEINFO_* instead of the functions directly Jul 27 17:17:05 woglinde, agreed Jul 27 17:17:13 http://pastebin.com/9dbB4fas Jul 27 17:17:20 yes Jul 27 17:17:21 thats already in Jul 27 17:17:22 ah Jul 27 17:17:24 it went in last night Jul 27 17:17:27 :) Jul 27 17:17:31 there was a mail on the list about it Jul 27 17:17:34 hmm why dont I have it then Jul 27 17:17:45 khem because git is down Jul 27 17:17:53 woglinde: still ? Jul 27 17:18:04 jupp Jul 27 17:18:07 can someone please pull from the git server? Jul 27 17:18:13 uh Jul 27 17:18:22 did you inform ka6sox-work Jul 27 17:18:38 I need a message pleaes. Jul 27 17:18:39 khem *g* Jul 27 17:18:45 ie what is it telling you? Jul 27 17:18:56 Your branch is ahead of 'origin/org.openembedded.dev' by 19 commits Jul 27 17:18:57 ka6sox: it hangs Jul 27 17:19:00 no message Jul 27 17:19:11 seems there was some msytical reset again Jul 27 17:19:24 I hope mickeyl told you the repait steps Jul 27 17:19:28 ups repair Jul 27 17:19:56 okay...when was the last time we *know* that we have good data/ Jul 27 17:19:57 ? Jul 27 17:20:09 we have had NO resets in like 5 days Jul 27 17:20:29 but as I khem said Jul 27 17:20:42 there is no output when pulling Jul 27 17:20:49 okay jas Jul 27 17:20:49 this is when I swicht branches Jul 27 17:21:03 args typing is hard today Jul 27 17:21:15 ka6sox: it starting to happen like 10-11 hours ago Jul 27 17:21:26 etrunko: so now the question is what else to do with the patches Jul 27 17:21:36 okay let me look @ some logs...bbiaf Jul 27 17:22:13 etrunko: the one change from your version that might be useful would be the OECMAKE_SOURCEPATH thing Jul 27 17:22:13 zub: we need at least the cmake.bbclass in so we can settle on the webkit-efl Jul 27 17:22:38 zub: and your install-launcher too Jul 27 17:22:47 etrunko: although I haveto say I like having all the sources in subdir of ${WORKDIR} :-P Jul 27 17:22:54 why do they wouldn't install the test program Jul 27 17:22:55 but OECMAKE_SOURCEPATH makes the recipe simpler, that's trie Jul 27 17:23:03 etrunko? Jul 27 17:23:41 etrunko: you mean why they didn't like installing the launcher? Jul 27 17:23:51 zub: yup Jul 27 17:24:45 etrunko: antognolli in the end yielded I think :) he understandably doesn't like it ending up on the "production"/user system Jul 27 17:24:56 okay I restarted the git server Jul 27 17:24:57 try now Jul 27 17:25:07 etrunko: but I think adding it to separe pkg is good for testing - that way is testing easier Jul 27 17:25:31 etrunko: for non-cross build installing is completely useless, for OE I think it makesmore sense Jul 27 17:25:42 woglinde, try now Jul 27 17:25:53 seems better now Jul 27 17:26:24 better as in its correct and no corruption or better relatively? Jul 27 17:26:29 yes Jul 27 17:26:37 could pull and rebase Jul 27 17:26:48 khem, now you try please. Jul 27 17:26:48 but dont know offhand if something is missing Jul 27 17:27:06 now lets see at kergoth siteinfo patch Jul 27 17:30:41 hm Jul 27 17:31:06 k Jul 27 17:31:52 hi tartarus Jul 27 17:32:24 ka6sox: seems to work now Jul 27 17:32:51 khem, up to date and no corruption? Jul 27 17:33:12 yes Jul 27 17:33:14 hi wogline Jul 27 17:33:14 seems so Jul 27 17:34:14 ka6sox hm thanks Jul 27 17:34:30 ka6sox anything in the hardwarelogs? Jul 27 17:35:19 kasox oestas server has denial of service too Jul 27 17:35:23 oestats Jul 27 17:35:29 aeh tinderbox Jul 27 17:36:31 woglinde, I need output of what it is telling you... Jul 27 17:36:55 is it like git Jul 27 17:36:58 I think Jul 27 17:37:01 denial of service Jul 27 17:37:34 http://tinderbox.openembedded.org/ Jul 27 17:37:37 no output Jul 27 17:37:43 ah now Jul 27 17:37:44 okay Jul 27 17:38:48 okay its working again Jul 27 17:39:10 dinner now Jul 27 17:39:12 till later Jul 27 17:58:25 hi marex Jul 27 18:01:06 ERROR: IO Error: file 'conf/bitbake.conf' not found Jul 27 18:01:26 i have a dir called stuff with oe and build inside of it on my /home directory Jul 27 18:01:35 i did the export also for that Jul 27 18:02:00 khem, can I have an ack for the gcc patches (or do you still need time to review them) (the TARGET_INCLUDES ones) Jul 27 18:02:08 nvm Jul 27 18:02:10 fixed it Jul 27 18:13:17 why is openembedded ignoring libraries that are installed to /lib64? Jul 27 18:13:44 i was told that there would be a x86_64 target branch but wasn't able to find it today, any hints? Jul 27 18:13:45 hm? Jul 27 18:14:01 maybee it got lost Jul 27 18:14:12 as the gitserver crashed Jul 27 18:14:43 03Roger Monk  07org.openembedded.dev * rfeffb40630 10openembedded.git/recipes/ti/ti-dsplink_1.65.00.03.bb: Jul 27 18:14:43 ti-dsplink: Add Latest dsplink 1.65.00.03 EA release Jul 27 18:14:43 * Add latest DSPLINK version Jul 27 18:14:43 Signed-off-by: Roger Monk Jul 27 18:14:43 Signed-off-by: Koen Kooi Jul 27 18:14:46 would be a pain. because building glibc for x86_64 works. the only problem i saw was in staging.bbclass so the lib64 folder was not staged to sysroot Jul 27 18:15:33 aand if thers no libc.so.6 in staging_target_dir gcc-cross fails ... moving it manually worked. rootfs was created nicely, the only problem was the image was lacking the /lib64 dir ... Jul 27 18:18:54 eFfeM_work: i will see them today Jul 27 18:22:50 khem, reminds me on my ppc patch that is also pending, i think :) Jul 27 18:25:39 blindvt get rw access Jul 27 18:27:27 woglinde, how's the process? Does somebody have to suggest me or how does that work in oe? Jul 27 18:28:04 blindvt: the ppc patch has comments from pb Jul 27 18:28:29 I am ok with the gcc patch itself but that PV thing needs to be sorted out Jul 27 18:28:31 khem, the ppc patch itself not, just tweaking PV/PR has comments from pb Jul 27 18:29:06 blindvt some has to write an email to suggest that you get rw access Jul 27 18:29:10 http://patchwork.openembedded.org/patch/2367/ needs acks from micro maintainers there was a reason why they stayed at 2.19.x Jul 27 18:29:28 http://patchwork.openembedded.org/patch/2366/ is obvious Jul 27 18:29:31 and then 2 core memembers or a lot of others have to give her ack's Jul 27 18:29:41 I have this one in queue Jul 27 18:31:21 khem: Trying a quick poke at moving everything around, bindir wise, now Jul 27 18:31:46 Tartarus: ok. I feel the approach I took is right one Jul 27 18:31:51 khem, http://patchwork.openembedded.org/patch/2367/ is hindering progress (2.19.x doesn't grok some arm asm that we use in uclibc). If there is a regression somewhere in binutils > 2.19 then i'd like to hear about it Jul 27 18:32:17 but I am ok if you have another approach and people are ok with it Jul 27 18:32:38 blindvt: I know those things but I dont maintain micro Jul 27 18:33:39 Tartarus: using gcc's logic to have multiple cross compilers coexist is the right way Jul 27 18:33:41 khem, if it was ?= then i wouldn't complain i think (because then i could just use binutils trunk or the current, stable release and be done) Jul 27 18:33:55 khem: Yes, and we are/will to some degree Jul 27 18:33:57 you can munge the stuff the way you want and make it work but it will come biting one day Jul 27 18:34:06 It's just that messing with the vendor string like that feels very wrong Jul 27 18:34:23 Have you tried poking the external toolchain stuff to make sure that still works easily? Jul 27 18:34:48 Tartarus: its not changing anything that external toolchain will be worried about I guess Jul 27 18:35:20 khem: Are you sure? The external toolchain stuff pokes the vendor stuff Jul 27 18:35:31 Tartarus: many use vendor string to differentiate stuff Jul 27 18:35:36 khem: Oh, and like pb mentioned, using the program prefix transform might be easier too Jul 27 18:35:47 and I think debian might move to it as well going forward Jul 27 18:35:48 khem: I'm referring to the external csl stuff Jul 27 18:36:39 yeah transform is good one but it has bugs in many gcc versions Jul 27 18:36:51 but thats something I am interested in Jul 27 18:37:23 why does external toolchain look at vendor str Jul 27 18:37:32 because external toolchains have vendor strings Jul 27 18:37:51 yes there is DISTRO_VENDOR Jul 27 18:37:58 no Jul 27 18:37:58 khem/blindvt any Idea when uclibc fixes undefined reference to `le16toh' Jul 27 18:38:00 they can use that Jul 27 18:38:16 khem: I'm not talking about that Jul 27 18:38:38 I'm talking about conf/distro/include/toolchain-external.inc Jul 27 18:39:01 And the stuff to use say a codesourcery installed toolchain w/ OE Jul 27 18:39:49 TOOLCHAIN_BRAND Jul 27 18:40:03 khem: how recently is the prefix stuff broken? Jul 27 18:40:24 khem: And yes, TOOLCHAIN_BRAND Jul 27 18:40:29 and that in turns messes with vendor Jul 27 18:40:30 after moving the cross stuff into native staging Jul 27 18:41:35 Tartarus: I dont see TOOLCHAIN_BRAND doing anything its a variable one needs to set Jul 27 18:41:48 and is empty for internal toolchain Jul 27 18:42:32 if we still want to create its own file heirarchy then there is no point of moving cross into native staging Jul 27 18:43:01 as we are unplugging it there and pulling into staging Jul 27 18:43:02 woglinde, didn't i apply this earlier today as c6d6237819037168a6923ac080e348e54615422c? Which rev are you using? I suggest you use current master's tip, that should have it Jul 27 18:43:58 khem: Humor me and try an external toolchain from csl, with your patchset please. Jul 27 18:43:58 blindtvt I am using the tip khem is setting Jul 27 18:44:14 khem: And the whole point of killing cross is that now we can mangle cross recipes for $ORIGIN Jul 27 18:44:54 woglinde: that function is not a uclibc one Jul 27 18:45:27 khem it comes from? Jul 27 18:45:47 oegit has 37aec506502e14dcddd6ca190929bbc8d33d8e07 Jul 27 18:45:58 khem newer udev are failing Jul 27 18:47:31 woglinde: yes Jul 27 18:47:39 woglinde: try patch http://permalink.gmane.org/gmane.comp.lib.uclibc.general/19631 Jul 27 18:47:54 or you can define this to something else Jul 27 18:47:59 woglinde, khem's tip in oe is too old, you should update if you need the BSD endian handlers (and if udev is silly _again_ to go down the BSD land then it should be fixed with a chainsaw -- or just use mdev to avoid all that udev bloat ;) Jul 27 18:48:04 Tartarus: ok Jul 27 18:49:10 blindvt: nice you applied that patch Jul 27 18:49:25 woglinde: then move uclibc tip to latest Jul 27 18:49:56 4e38c284bc8b3bff95543f246f3e0d3ee5a045ab Jul 27 18:50:51 Tartarus: how about then moving cross into /cross rather than /usr/bin Jul 27 18:52:01 khem: As a temporary fix, sure Jul 27 18:52:14 khem: I think longer term the right fix is also going to involve a real fix for multilib Jul 27 18:52:19 will that solve the need of ORIGIN Jul 27 18:52:36 Er, wait, no Jul 27 18:52:38 No it won't Jul 27 18:52:46 thats what I though Jul 27 18:52:50 Needs to be in usr Jul 27 18:53:00 otherwise we can get into path length issues, i believe Jul 27 18:53:05 usr/cross will work ? Jul 27 18:53:10 yes Jul 27 18:53:14 or does the libs have to go into usr/lib and so on Jul 27 18:53:46 The problem is we need to know that the start of the hard-coded rpath will be long enough to patch $ORIGIN in Jul 27 18:54:00 and that it's also going to leave enough room to go from where the binary is to where libraries are Jul 27 18:58:26 Tartarus: if we use program-transform-name may be we can have then live together in usr/bin Jul 27 18:58:55 Tartarus: may be not because this option will only rename the programs not the directories Jul 27 19:00:43 woglinde: a piece of advise, would it be ok to decrease PR in cmake.inc to r0? or instead of that, keep it ther as is and add to the new recipes? Jul 27 19:04:05 etrunko let it there Jul 27 19:04:15 ah no Jul 27 19:04:21 use INC_PR Jul 27 19:04:29 grep other recipes for it Jul 27 19:04:34 ok Jul 27 19:05:26 you never decrement PR Jul 27 19:05:34 in an existing recipe Jul 27 19:06:03 khem: understood Jul 27 19:06:35 woglinde: so with INC_PR in the end it will be something like 'r2.0'? Jul 27 19:07:29 yes Jul 27 19:07:53 i'll set INC_PR in the .inc file and PR=$"{INC_PR}.0}" on the others right? Jul 27 19:08:07 of course without that last } Jul 27 19:08:49 aeh Jul 27 19:09:05 PR = "${INC_PR}.0" Jul 27 19:09:48 yeah, just lazy enough to cut the white spaces Jul 27 19:10:19 etrunko: your quotes and dollar sign is also in the wrong order Jul 27 19:10:30 Tartarus: I guess combination of --program-transform-name and --prefix based on PACKAGE_ARCH might do the trick Jul 27 19:10:42 hmm hmm Jul 27 19:10:56 akheron: argh, sorry Jul 27 19:11:24 but i intended to write that correctly like Jul 27 19:11:34 heh Jul 27 19:11:55 *g* Jul 27 19:13:59 03Bernhard Reutner-Fischer  07org.openembedded.dev * r4fe8975385 10openembedded.git/recipes/gmp/ (gmp-5.0.1/amd64.patch gmp-5.0.1/configure.patch gmp_5.0.1.bb): Jul 27 19:13:59 libgmp: bump version Jul 27 19:13:59 Signed-off-by: Bernhard Reutner-Fischer Jul 27 19:13:59 Signed-off-by: Khem Raj Jul 27 19:14:00 03Bernhard Reutner-Fischer  07org.openembedded.dev * rf0c2a1d68f 10openembedded.git/conf/distro/minimal-uclibc.conf: Jul 27 19:14:00 minimal: commentary typo fix Jul 27 19:14:00 Signed-off-by: Bernhard Reutner-Fischer Jul 27 19:14:00 Signed-off-by: Khem Raj Jul 27 19:14:03 03Bernhard Reutner-Fischer  07org.openembedded.dev * r3d576c8b03 10openembedded.git/recipes/mpfr/ (mpfr-3.0.0/p3.patch mpfr-canadian_3.0.0.bb mpfr_3.0.0.bb): Jul 27 19:14:03 libmpfr: bump version Jul 27 19:14:03 Signed-off-by: Bernhard Reutner-Fischer Jul 27 19:14:03 Signed-off-by: Khem Raj Jul 27 19:17:24 blindvt: http://patchwork.openembedded.org/patch/2200/ needs rework Jul 27 19:17:36 you dont need that bx part Jul 27 19:17:39 its already fixed Jul 27 19:18:49 03Bernhard Reutner-Fischer  07org.openembedded.dev * rb47a00c4df 10openembedded.git/recipes/libmpc/ (libmpc-native_0.8.1.bb libmpc_0.8.1.bb libmpc_0.8.2.bb): Jul 27 19:18:50 libmpc: bump version and cleanup staging Jul 27 19:18:50 Signed-off-by: Bernhard Reutner-Fischer Jul 27 19:18:50 Acked-by: Paul Menzel Jul 27 19:18:50 Signed-off-by: Khem Raj Jul 27 19:20:11 gogogo Jul 27 19:20:32 yeah they were in my tree I was holding them up Jul 27 19:20:46 when y'all have about 2 minutes I want to restart the git server...did an upgrade. Jul 27 19:20:53 so moved to push branch and there it goes Jul 27 19:21:05 ka6sox: go for it Jul 27 19:21:13 I can wait 2 mins :) Jul 27 19:21:45 ka6sox update it Jul 27 19:23:05 re zecke Jul 27 19:32:18 few more mins please...checking things. Jul 27 19:41:56 woglinde_: thanks Jul 27 19:42:00 cbrake: yet? Jul 27 19:44:55 khem, try now please? Jul 27 19:51:43 khem, woglinde_ please check that its up. Jul 27 19:52:48 ka6sox hm in few minutes Jul 27 19:57:26 is something broken again? :) Jul 27 19:57:48 Jay7, not yet...just give me some time. Jul 27 19:58:11 I have to wait for a build Jul 27 19:58:27 git pull? Jul 27 19:58:36 Jay7, good start. Jul 27 19:58:37 :D Jul 27 19:58:44 I'll try 50 Jul 27 19:58:50 s/50/:)/ Jul 27 19:59:18 receiving objects Jul 27 20:00:26 updated Jul 27 20:00:51 good Jul 27 20:11:24 re kergoth Jul 27 20:19:29 anyone understand how the gst-plugins-* packages are dealt with? looking for input-selector which is supposed to be part of gst-plugins-bad but can't even figure out where all the plugin .so's are packaged Jul 27 20:33:41 03Khem Raj  07org.openembedded.dev * r3f5cdf3928 10openembedded.git/recipes/openssh/openssh.inc: Jul 27 20:33:41 openssh.inc: Enable pam if present in DISTRO_FEATURES Jul 27 20:33:41 Signed-off-by: Sean Hudson Jul 27 20:33:41 Signed-off-by: Khem Raj Jul 27 20:34:46 ka6sox: yep I pushed just now all seem to work Jul 27 20:34:59 khem, thanks Jul 27 20:50:26 hi khem Jul 27 20:53:19 GNUtoo|laptop: hey Jul 27 20:54:26 khem, hi,at the end it built fine but I've 2 big issues Jul 27 20:55:47 first issue: I've oetmps/htcdream/deploy/ipk/armv6 instead of oetmps/htcdream/deploy/ipk/armv6-novfp Jul 27 20:57:02 second issue: oetmps/htcdream/deploy/ipk/armv6 has no Packages.gz etc... Jul 27 20:57:19 GNUtoo|laptop: I have v3 of patches Jul 27 20:57:23 that should be better Jul 27 20:57:25 try that Jul 27 20:57:26 ok Jul 27 20:57:49 the issue is the armv6-novfp -> armv6 thing that we changed Jul 27 20:57:56 at least I bet Jul 27 20:57:56 hello, i'd like to inform, that recipes/openssh/openssh.inc is broken and unparsable - it lacks \ at the end of line, so multiline variable assingment does not work Jul 27 20:58:08 poor khem Jul 27 20:58:09 but i believe you already know it ;] Jul 27 20:58:11 khem, I'll try the patchs tomorrow Jul 27 20:58:18 (line 35) Jul 27 20:58:22 GNUtoo|laptop: sure Jul 27 20:58:27 woglinde_: :) Jul 27 20:58:40 (3g => very limited GB) Jul 27 20:58:44 nazgee khem will fix it Jul 27 20:59:06 not a problem for me, but good to know you are few steps ahead ;] Jul 27 21:00:37 nazgee: ups Jul 27 21:02:34 nazgee: update again Jul 27 21:02:57 03Khem Raj  07org.openembedded.dev * r31a171ffc2 10openembedded.git/recipes/openssh/openssh.inc: Jul 27 21:02:57 openssh.inc: Add missing '\' to EXTRA_OECONF Jul 27 21:02:57 Signed-off-by: Khem Raj Jul 27 21:05:15 khem: works now Jul 27 21:05:35 btw, is something wrong with my machine, or gettext keeps failing for everyone? lastly i am trying to build something on dev branch, and it seems I can't get bring the pieces together Jul 27 21:06:10 ettext_6_prg-gettext-6-prg.o: In function `main':gettext-0.18/gettext-tools/tests/gettext-6-prg.c:175: undefined reference to `rpl_unsetenv' Jul 27 21:06:10 nazgee: such riddles dont work you have to be more specific on errors Jul 27 21:06:15 ah Jul 27 21:06:30 nazgee: are you seeing this recently Jul 27 21:06:40 it is there for few days Jul 27 21:06:47 nazgee: bt.w whats your MACHINE and DISTRO Jul 27 21:06:50 i haven't look at the code Jul 27 21:07:05 as it is not urgent (though it gets so) Jul 27 21:07:25 igep0020 (beagle clone) and angstrom Jul 27 21:07:50 about few weeks ago i was able to build x11-image, but e-image was failing Jul 27 21:08:03 nazgee: you did not answer what I asked Jul 27 21:08:10 now i can't get anything (even meta toolchain fails) Jul 27 21:08:27 MACHINE = igep0020 Jul 27 21:08:49 DISTRO ?= "angstrom-2008.1" Jul 27 21:09:04 LIBC=glibc ? Jul 27 21:09:05 isn't that what you asked for? Jul 27 21:09:08 yes Jul 27 21:09:09 yes Jul 27 21:09:58 re Jul 27 21:10:12 nazgee: I fixed it few months ago Jul 27 21:10:46 sounds like a strange explanation for it not working now ;] Jul 27 21:10:57 or i just don't get it Jul 27 21:11:00 nazgee: git show 9991720b Jul 27 21:11:40 ok Jul 27 21:11:53 but how come it is not working now? Jul 27 21:12:13 you can git bisect Jul 27 21:12:18 when did it work last Jul 27 21:12:33 because its building fine here Jul 27 21:12:54 i can try it some time, but the problem is that my GF is shouting on me to go to bed ;] Jul 27 21:13:07 hehe Jul 27 21:13:14 well thats NMI Jul 27 21:13:20 so good night Jul 27 21:13:25 and i am working fo 14hrs a day... so i do not now when i will find some time Jul 27 21:13:38 i'll see what i can do with it Jul 27 21:13:51 ieehks Jul 27 21:13:58 working 14h a day suckz Jul 27 21:14:27 i work for two companies ;] Jul 27 21:14:36 work for one Jul 27 21:14:41 and save your life Jul 27 21:14:47 it does not work for me Jul 27 21:14:58 engineers here are paid shitty Jul 27 21:15:09 china? Jul 27 21:15:11 which place are you Jul 27 21:15:12 ;] Jul 27 21:15:15 poland Jul 27 21:15:25 come to us Jul 27 21:15:30 (between germany and russia) Jul 27 21:15:44 wehre are you guys? Jul 27 21:15:47 in germany you will find good payed jobs Jul 27 21:15:59 yes, but i do not like the language Jul 27 21:16:12 nazgee: ++ ;) Jul 27 21:16:13 just an oddity of my hearing system Jul 27 21:16:23 pfff Jul 27 21:16:27 khem likes it Jul 27 21:16:36 blah ;] Jul 27 21:17:09 :) Jul 27 21:17:12 khem are you going to elc in cambridge? Jul 27 21:17:43 really- i mean it. personally, i find nothing wrong about german, but this language seems a bit... aggresive to me? or maybe i am jus weird Jul 27 21:18:07 you are just weird Jul 27 21:18:09 woglinde_: no Jul 27 21:18:11 ;] Jul 27 21:18:19 woglinde_: I didnt make it to one in SFO Jul 27 21:18:24 cambridge is too far Jul 27 21:18:30 :( Jul 27 21:18:45 as you know I will proably never come to the US Jul 27 21:18:58 woglinde_: why not Jul 27 21:20:43 dont like it Jul 27 21:20:53 and I have to sign shitty documents to get it Jul 27 21:22:47 woglinde_: oh those association with SS and stuff Jul 27 21:23:09 woglinde_: well I might travel a lot more in couple of years Jul 27 21:24:15 yeah I hope we will meet some day Jul 27 21:24:57 you're in the US, woglinde? Jul 27 21:25:08 no I am from berlin Jul 27 21:25:14 aaah Jul 27 21:25:17 ;] Jul 27 21:26:44 ups, i just swaped you 2 withe eech other. then khem is from states, right? Jul 27 21:28:00 yes Jul 27 21:30:40 woglinde_: Are you student at berlin uni Jul 27 21:30:59 woglinde_: Jul 27 21:31:00 Technische Universität Berlin Jul 27 21:31:48 freie universität Jul 27 21:54:33 03Henning Heinold  07org.openembedded.dev * rd8e0ee5155 10openembedded.git/recipes/uclibc/uclibc_git.bb: uclibc-git: bump to latest git master, which fixes building of udev >= 151 Jul 27 21:54:44 03Henning Heinold  07org.openembedded.dev * ra1358b5f8b 10openembedded.git/recipes/dbus/ (10 files in 3 dirs): dbus-glib: converted to BBCLASSEXTENDED and remove do_stage stuff Jul 27 21:55:39 khem: what's today to test? Jul 27 21:56:02 ant___: heh Jul 27 21:56:06 v3 of the patches Jul 27 21:56:08 I posted Jul 27 21:56:27 although they may not go in as I am thinking/working on an alternative approach Jul 27 21:56:32 fwiw I didn't have any strange issue on building Jul 27 21:56:43 with previous version Jul 27 22:03:34 khem: the serie of 8( ? Jul 27 22:05:31 http://pastebin.com/MqMjfx09 anyone know how I can sort this conflict out? Jul 27 22:06:30 hm ah yes Jul 27 22:06:37 the fix is on the mailinglist Jul 27 22:06:57 oh sorry didn't find it I"ll search again Jul 27 22:20:23 good nite Jul 27 22:20:42 'nite Jul 27 22:30:47 khem: I'm testing after ./pw-am.sh 2475 2471 2470 2472 2473 2476 2474 2469 Jul 27 22:35:45 ant___: yeah choose the latest series with 8 patches Jul 27 22:36:40 ok, thx. Already teeing the buildlog Jul 27 22:37:14 andyj: why is it installing ipkg Jul 27 22:37:19 something is wrong Jul 27 22:37:22 in your env Jul 27 22:42:44 not sure I wasn't at the most recent git revision so I updated building again we'll see if that is still the case Jul 27 22:43:06 or are you saying something in my local.conf is borked? Jul 27 22:43:15 dont know Jul 27 22:43:23 did you build from scratch Jul 27 22:44:00 I am now Jul 27 22:44:13 previously I had a prior image built Jul 27 22:57:47 cmake 2.8.2 is being innovative - it executes the built binaries with --hepl-full and similar to produce doc, sigh Jul 27 22:59:06 * Jin^eLD hates cmake very, very much Jul 27 22:59:52 I used to like it, but it's not too cross-friendly in general Jul 27 23:00:09 I hated it from the very first moment I had to deal with it Jul 27 23:00:14 and that was not even doing cross stuff Jul 27 23:00:34 for a while I thought it might be the salvation from autohell Jul 27 23:00:59 well, autotools are not as bad as the rumors tell ;> Jul 27 23:01:20 maybe if you grok it, I never got to that level Jul 27 23:01:49 it took me a while, but once you get how it works - it's not bad at all Jul 27 23:02:04 the "hell" usually comes when you think you know what you are doing, but you're not Jul 27 23:02:14 then its easy to end up in a mess Jul 27 23:02:20 happened to me in the beginning a couple of times Jul 27 23:12:29 khem: fwiw NOTE: package eggdbus-native-0.6-r0: task do_compile: Failed Jul 27 23:12:33 | /usr/bin/ld: cannot find -ldbus-glib-1 Jul 27 23:13:55 03Tom Rini  07org.openembedded.dev * r8085586dd2 10openembedded.git/classes/ (cross.bbclass packaged-staging.bbclass): Jul 27 23:13:55 packaged-staging: Place cross/native/etc into the normal path Jul 27 23:13:55 With the changes to move ${CROSS_DIR} into sysroots/ we now Jul 27 23:13:55 $ORIGIN-mangle cross packages (and already were native) so change Jul 27 23:13:55 the comment from cross.bbclass while we're at this and then put Jul 27 23:13:56 these recipes into the usual packaged-staging cache. Jul 27 23:13:57 Signed-off-by: Tom Rini Jul 27 23:17:04 khem: I suspect is because of dbus-glib: converted to BBCLASSEXTENDED and remove do_stage stuff Jul 27 23:19:01 reverting it locally Jul 27 23:23:23 ok, I'v added last Tartarus commit on the top of your 8 patches Jul 27 23:23:31 ant___: ok Jul 27 23:23:43 relaunching the build now Jul 27 23:24:02 will let you know in 7 hours Jul 27 23:24:19 * ant___ -> zZZz Jul 27 23:31:48 anybody ever used the non-native cmake recipe? Jul 27 23:32:41 I wonder how come they were usable when they don't seem to correctly package the various .cmake scripts Jul 27 23:33:24 zub: I luckily had to make only one cross cmake recipe and inherit cmake was enough, luckil the makefiles seemed to be OK and worked for cross out of th ebox Jul 27 23:34:16 I'm actually trying to get a working cmake_2.8.2, i.e. to cross-compile cmake itself Jul 27 23:35:48 heh, that sounds tough Jul 27 23:36:06 surprisingly it compiles for me Jul 27 23:36:14 if I'm lucky, then just some polishing is left Jul 27 23:36:31 I just had to nuke the cmake's tendency to run the binaries it just generated :) Jul 27 23:37:03 and make it use system libs, instead of what it drags with itself (as then cross-hell breaks out it it uses them) Jul 27 23:37:19 heh Jul 27 23:37:27 they should have used autotools as their build environment ;) Jul 27 23:37:36 lol :) tell them :-P Jul 28 02:00:26 03Tom Rini  07org.openembedded.dev * ref1eb7cdc9 10openembedded.git/recipes/bison/bison.inc: Jul 28 02:00:26 bison-native: Disable packaged-staging Jul 28 02:00:26 bison-native does some hard-coded path nastiness to the binary so Jul 28 02:00:26 we disable packaged-staging. Jul 28 02:00:26 Signed-off-by: Tom Rini Jul 28 02:15:21 03Tom Rini  07org.openembedded.dev * r9a0e95f376 10openembedded.git/classes/packaged-staging.bbclass: (log message trimmed) Jul 28 02:15:22 packaged-staging: Make setting PSTAGING_ACTIVE be a named python func Jul 28 02:15:22 When the function to set PSTAGING_ACTIVE is an anon python function Jul 28 02:15:22 overrides are not set so we cannot set things to disabled in the context Jul 28 02:15:22 of BBCLASSEXTEND recipes. Call this function before we check Jul 28 02:15:22 PSTAGING_ACTIVE. In the case of packagedstaging_fastpath it's cleaner Jul 28 02:15:23 to convert to a full python function. While we're in here drop a **** ENDING LOGGING AT Wed Jul 28 02:59:56 2010