**** BEGIN LOGGING AT Wed Mar 25 02:59:57 2009 Mar 25 03:00:16 argh, i need to get around to redoing the whole bb.fetch module Mar 25 03:00:25 terrible hacks Mar 25 07:37:57 morning Mar 25 07:39:25 morning mckoan ! Mar 25 07:45:26 mckoan: morning Mar 25 07:46:00 mckoan: believe or not... i'm here since we talked yesterday Mar 25 07:46:16 working... Mar 25 07:46:52 mckoan: probably i cant work with you morning late in qt-build Mar 25 07:47:05 could we postpone some hours? Mar 25 07:47:11 i need to sleep a little... hheh Mar 25 07:53:15 mckoan, news from your Marvell distributor ? Mar 25 07:56:55 aloisiojr: yes of course Mar 25 07:57:20 mckoan: tks Mar 25 07:58:35 gremlin[it]: yes after 2 days he told me they don't have any SheevaPlug in Italy Mar 25 07:58:36 hi everybody Mar 25 07:58:53 hahah ok ... Mar 25 07:59:08 ~curse Avnet Mar 25 07:59:10 May the fleas of a thousand camels infest your most sensitive regions, Avnet ! Mar 25 07:59:46 hi recalcati_oe hi Mar 25 08:00:53 hi mckoan Mar 25 08:01:01 gremlin[it]: BTW I can get a Kirkwood processor eval board Mar 25 08:01:46 mhhh can be interesting :) ... Mar 25 08:02:00 I've seen that packages directory has been renamed in recipes. now bitbake beagleboard-demo-image doesn't work. I need to write bitbake -b openembedded/recipes/images/beagleboard-demo-image.bb . but I'm not sure that it build all changes correctly. Mar 25 08:02:01 http://www.marvell.com/products/embedded_processors/kirkwood/index.jsp Mar 25 08:02:52 recalcati_oe, take a look at your conf/local.conf file, you rhave to change there all *.bb files are ... Mar 25 08:04:45 done, now I restart wtih -c rebuild Mar 25 08:06:02 its working Mar 25 08:06:11 thx Mar 25 08:06:45 mckoan, what interest me more is the "point of service terminal" ... with some kind of "standard" PCI-Ex graphic card, such as Volari or other low cost, low power device could be a nice device Mar 25 08:21:31 good morning Mar 25 08:31:05 is there any chance that a solution for http://patchwork.openembedded.org/patch/181/ gets commited? Mar 25 08:32:09 it doesn't matter how it is done, I'd just like to have a way to put my DEPLOY_DIR outside of TMPDIR with Angstrom Mar 25 09:06:56 good morning Mar 25 09:16:38 03Jeremy Lainé  07org.openembedded.dev * rcfb474d939 10openembedded.git/recipes/qt4/qt4-native_4.4.3.bb: qt4-native: add package to replace qmake2-native and uicmoc4-native Mar 25 09:29:13 mpoullet|work: changing only angstrom.conf seems not acceptable Mar 25 09:29:24 well, angstrom.inc Mar 25 09:30:35 unaware users would get in trouble...the value would be taken from bitbake.conf if there are no overides in local.conf Mar 25 09:31:13 fwiw I adapted Tom's suggestion and created ANGSTROM_DEPLOY_DIR...semms to work nice Mar 25 09:32:47 I found an error in libgsm_1.0.10.bb : I have commented the line 24 because it tries to create a link of gsm.h , but the file is already present and equal to file ${STAGING_INCDIR}/gsm.h Mar 25 09:32:56 now it works Mar 25 09:37:13 ant_work: hi, it's not a problem for me, I'd just want to know if something will get committed, so that I don't have to change anything, simply edit my local.conf Mar 25 09:37:34 * mpoullet|work will try Tom's suggestion as well Mar 25 09:38:37 mpoullet|work: we have yet to read what Koen thinks about this Mar 25 09:43:01 ant_work: ok, fine. Mar 25 09:44:53 mpoullet|work: imho the point is in bitbake.conf the value is not suffixed by *libc, just DEPLOY_DIR ?= "${TMPDIR}/deploy" Mar 25 09:45:06 Hi all Mar 25 09:46:08 I add my own package - libx264 in my oe and built it successfully .. but when I try to build vlc with x264 support, it says that pkg cannot find libx264 package .... how should I add it in pkg ? Mar 25 09:47:56 03Jeremy Lainé  07org.openembedded.dev * r1476b97656 10openembedded.git/recipes/qt4/ (qt4-native-4.4.3/qt-config.patch qt4-native_4.4.3.bb): qt4-native: add qt-config.patch from qmake2-native Mar 25 09:56:32 aloisiojr: ping? Mar 25 09:57:28 jeremy_laine: pong Mar 25 09:58:13 aloisiojr: ok, I am currently trying to progressively simplify OE's build of Qt Mar 25 09:58:38 good morning Mar 25 09:59:03 aloisiojr: I am a bit concerned that suddenly replacing everything by rules written from scratch is going to have some side effects Mar 25 09:59:43 aloisiojr: one thing I have done is to introduce a qt4-native package, to avoid building qmake2-native + uicmoc4-native separately Mar 25 10:00:27 aloisiojr: I have thought about the approach you have taken (build the native tools as part of the target Qt build) and I think it has some issues Mar 25 10:00:57 1/ I don't think we should be touching ${STAGING_DIR_NATIVE} when building a non-native package Mar 25 10:01:39 2/ if both qt4-x11-free and qt-embedded provide the native tools, it becomes impossible to have both packages in staging Mar 25 10:03:25 aloisiojr: on a related topic, what packages do you suggest for regression testing? Mar 25 10:04:05 aloisiojr: I have looked around and a bunch of Qt packages don't build for reasons other than Qt (for instance I can't build qsvn because subversion won't build for my platform) Mar 25 10:14:17 jeremy_laine: the 2 issues you've pointed are good ones Mar 25 10:14:29 03Robert Schuster  07org.openembedded.dev * r5575ee4e7a 10openembedded.git/recipes/llvm/ (llvm-2.5/fix-build.patch llvm_2.4.bb llvm_2.5.bb): llvm 2.5: Updated patch to include preprocessor definitions. Mar 25 10:17:06 jeremy_laine: i was concerning about the regression tests / side effects too... I was thinking in make all 4.x qt recipes build with the new bbclass Mar 25 10:21:16 my 'package' depends on 'libxxx-dev' (for building) and 'libxxx' (for runtime). I specified DEPENDS = "libxxx" and RDEPENDS = "${DEPENDS}" for 'package'. But I can't get 'libxxx-dev' contents in my staging anyway. Am I doing smth wrong? Mar 25 10:24:46 jeremy_laine: your aproach of create a qt4-native package is good.. but you must just build qmake and src/tools parts, to prevent useless compilation Mar 25 10:25:30 aloisiojr: I don't build everything, only the host tools Mar 25 10:25:46 aloisiojr: I basically adapted the uicmoc4-native package Mar 25 10:26:19 jeremy_laine: it is a nice approach Mar 25 10:26:59 morning Mar 25 10:27:30 hi hrw Mar 25 10:28:25 jeremy_laine: another nice change is package the (qmake/moc/uic/rcc) from qt4 build Mar 25 10:28:44 they are arm binaries in your approach, right? Mar 25 10:28:55 now is the webkit-gtk compilation the fails: http://pastebin.com/d741d5906 Mar 25 10:29:22 qt4-tools would be very useful to SDKs Mar 25 10:31:06 aloisiojr: do you think I should rename "qt4-native" to "qt4-tools-native" ? Mar 25 10:31:31 aloisiojr: it would probably be a good idea, as it's not a full Qt Mar 25 10:32:05 anyone can help with my little problem I specified above? Mar 25 10:32:12 aloisiojr: by the way, what is the meaning of "SDK" in qmake2-sdk, as opposed to qmake2-native Mar 25 10:33:37 jeremy_laine: anyway, there is a bug regarding qreal + qvariant in moc, we already talked to qt guys and we're studying a provisional workarround to solve this Mar 25 10:33:51 booxter_: -dev deps are created automatically Mar 25 10:33:59 booxter_: share recipe Mar 25 10:37:34 hrw, here is 'package': http://pastebin.com/m7fb2873 and here is 'libzip': http://pastebin.com/m28cb7ff4 Mar 25 10:38:46 ('package' recipe is bad I know) Mar 25 10:41:04 booxter_: problem is with libzip or with 'package? Mar 25 10:41:22 03Jeremy Lainé  07org.openembedded.dev * r8e64571067 10openembedded.git/recipes/qt4/ (qt4-native-4.4.3/qt-config.patch qt4-native_4.4.3.bb): qt4-native: rename to qt4-tools-native Mar 25 10:41:24 florian: in order to give a better idea of OE, hide the fact there is Bugzilla ;-) Mar 25 10:41:29 Anyone here used wpa-supplicant with dbus system activation? Mar 25 10:41:42 I don't know. libzip recipe generates libzip-dev with "staging" files (includes and libzip.a) Mar 25 10:41:53 I'd like to see these files when building 'package' Mar 25 10:42:08 ant_work: heh Mar 25 10:42:31 don't confess..even under torture Mar 25 10:42:35 I'd like to see these files IN MY STAGING when building 'package' Mar 25 10:42:35 :-) Mar 25 10:42:40 booxter_: but 'package' installs only one binary nothing more Mar 25 10:43:09 but it links with lots of libraries Mar 25 10:43:21 booxter_: s/^RDEPENDS/DEPENDS, drop currents DEPENDS line which is wrong Mar 25 10:43:34 booxter_: OE should add all needed deps to resulting package Mar 25 10:44:09 booxter_: 3rd line should be DEPENDS not RDEPENDS (same content as it is), 4th line dropped Mar 25 10:44:22 RDEPENDS = "directfb glib-2.0 gstreamer neon libzip" ? Mar 25 10:45:05 no, DEPENDS = "directfb glib-2.0 gstreamer neon libzip" and no RDEPENDS Mar 25 10:45:49 OE will generate RDEPENDS from objdump output Mar 25 10:45:58 florian: my limited skills (and free time...) won't allow me to fix so many bugs as I wish...some seem however trivial. Most of them have been probably solved in the meanwhile. Mar 25 10:46:19 (all compile bugs I suspect) Mar 25 10:47:07 ant_work: yes we need to get rid of all this automatically created crap Mar 25 10:48:43 hrw, doesn't work Mar 25 10:49:30 there is no 'libzip-dev' contents in 'mipsel-angstrom-linux' staging Mar 25 10:49:46 let me build libzip Mar 25 10:50:03 booxter_: libzip is not in OE - right? Mar 25 10:50:43 hi pb_ Mar 25 10:51:10 booxter_: and which ver it is? Mar 25 10:51:30 hrw, right. It's my own recipe at the moment Mar 25 10:51:41 booxter_: 0.9? Mar 25 10:51:41 0.9 version Mar 25 10:51:52 hi florian Mar 25 10:51:57 do you need checksums? Mar 25 10:52:04 booxter_: no Mar 25 10:53:09 booxter_: add do_stage() { autotools_stage_all } into libzip recipe Mar 25 10:55:36 03Jan Lübbe  07fso/milestone5.5 * r42b5fdd5fa 10openembedded.git/packages/bluez/bluez4_4.33.bb: bluez4: use update-rc.d Mar 25 10:58:32 hrw, works ok, tnx. Where can I find the info regarding the magic?:) Mar 25 10:58:50 its not magic Mar 25 10:59:09 booxter_: to put sth into staging/ you need stage task Mar 25 11:00:08 understood. Tnx. Mar 25 11:00:30 Crofton|work: updating the ProjectDevices page which seems to have gone backwards, who did you send that poodle to last year? Mar 25 11:00:48 hrw: I'm getting ready to make a ubifs image and realized there is no endian option like there was with mkfs.jffs2, is this correct? Mar 25 11:01:05 does anyone know who maintains the neuros* packages? Mar 25 11:01:26 slapin: looks like Mar 25 11:01:57 " Mar 25 11:02:12 svolpe: looks like Mar 25 11:04:55 hey mickey|aalborg Mar 25 11:06:10 hi florian Mar 25 11:06:33 03Jeremy Lainé  07org.openembedded.dev * r2fa972014d 10openembedded.git/ (7 files in 2 dirs): qt4: make use of qt4-tools-native instead of qmake2-native and uicmoc4-native Mar 25 11:29:09 XorA, I'll need to look Mar 25 11:29:15 hvontres? Mar 25 11:32:23 Crofton|work: yeah, any idea his actual name :-D Mar 25 11:32:49 Henry von trasckow? Mar 25 11:32:52 03Jeremy Lainé  07org.openembedded.dev * r88c677023e 10openembedded.git/ (15 files in 3 dirs): qt-embedded: rename to qt4-embedded for consistency Mar 25 11:32:52 something like that Mar 25 11:33:07 Crofton|work: good enough for me :-) Mar 25 11:33:28 we have 2 poodle, 3 tosa? Mar 25 11:34:32 yes Mar 25 11:36:48 to early, need coffee Mar 25 11:42:51 tresckow Mar 25 11:47:35 Crofton|work: cheers, ProjectDevices should be upto date with all devices that passed through me now anyway Mar 25 11:49:35 XorA: at fosdem you also got wifi cards from me ;D Mar 25 12:01:23 hrw: are you opening #oe-stable? Mar 25 12:01:31 yes Mar 25 12:04:26 hrw: ah yes, I put them in different part of desk Mar 25 12:04:39 hrw: dlink + libertas ones? Mar 25 12:04:43 yes Mar 25 12:04:52 * XorA goes to update Mar 25 12:04:59 both were donated for me (not OE) in past Mar 25 12:06:47 hrw: done Mar 25 12:07:05 * XorA begins to feel like a zaurus library Mar 25 12:07:25 XorA: that libertas is probably Ambicom but unbranded sounds better Mar 25 12:07:42 schurig got Ambicom one from same source Mar 25 12:07:49 cool Mar 25 12:07:51 and wrote driver with it Mar 25 12:08:06 need to find out why tosa backlight is not working then get libertas driver knocked about abit on tosa Mar 25 12:08:15 XorA: Libertas is the chip name family Mar 25 12:08:17 need pop out and post begging letters Mar 25 12:08:23 XorA: both SparkLan and Ambicon produce them Mar 25 12:08:51 schurig: card hrw gave me has no manufacturer marking at all Mar 25 12:09:30 XorA: I got one too, also without marking and without FCC number. But the card worked anyway :-) Mar 25 12:09:49 XorA: however, we now bought so around 300 of those cards from Sparklan in the meantime Mar 25 12:10:07 03Jeremy Lainé  07org.openembedded.dev * r8809e3ba81 10openembedded.git/recipes/qt4/qt_staging.inc: qt_staging.inc: fix symlink to qmake Mar 25 12:11:23 hrw: was discussion about stable postponed? Mar 25 12:11:38 it is on #oe-stable Mar 25 12:12:21 I don't have that chan Mar 25 12:12:49 now ys... Mar 25 12:22:02 03Jeremy Lainé  07org.openembedded.dev * rcfabec5a23 10openembedded.git/recipes/neuros-public/ (7 files): neuros-*: don't force DEPENDS, let qt4e.bbclass handle it Mar 25 13:16:01 03Jeremy Lainé  07org.openembedded.dev * ra1f6020e21 10openembedded.git/recipes/qt4/ (14 files in 3 dirs): qt4-x11-free: drop obsolete versions as per oe-devel Mar 25 13:23:18 How can I try to solve the webkit-gtk compilation problem: http://pastebin.com/d741d5906 . Where can I find the mantainer ? Mar 25 13:26:15 I don't know how is the openembedded mantainers structure. I found a webpage containing the mantainers, but now I don't find it. Mar 25 13:29:38 recalcati_oe: use git log --follow webkit-gtk.bb Mar 25 13:32:16 thank you very much: git log --follow recipes/webkit/webkit-gtk_svn.bb Mar 25 13:33:08 it is not maintainer list but who changed it Mar 25 13:34:43 hrw: yes, I understand. Can it be a bitbake version problem ? I have 1.8.13 Mar 25 13:35:06 recalcati_oe: I lack time to check sorry Mar 25 13:35:38 ok, I'll try this evening to ude a newer one. Mar 25 13:35:41 ybye Mar 25 13:35:43 bye Mar 25 14:13:02 florian: you here? Mar 25 14:13:33 Esben: yes Mar 25 14:22:41 what is the current hosting situation, is there anything that could be improved? At dotsrc.org we have some server resources which maybe could be offered to OE. We have some Linux servers sitting in an A/C room with UPS and backup on the danish research network (10Gbit). Mar 25 14:24:51 Esben: This might be interesting. You should contact mickey|zzZZzz who is in charge for the main OE server. Mar 25 14:30:21 hehe, he sounded interested last night. I met him after FOSS-Aalborg :-) Mar 25 14:30:50 I'll discuss with him if there is anything we can do to help. Mar 25 14:35:30 i am unable to compile libxine. i get configure errors such as "error: m4_require: circular dependency of AC_LANG_COMPILER(C++)" Mar 25 14:35:35 are my autoconf utils old? Mar 25 14:36:48 i use default fedora core 8 system Mar 25 14:38:53 03Jeremy Lainé  07org.openembedded.dev * r8ec53c8f0b 10openembedded.git/recipes/qt4/ (30 files in 4 dirs): Mar 25 14:38:53 qt4: unify patches, drop obsolete files, fix mkspecs Mar 25 14:38:53 * unify qt4-x11-free / qt4-embedded patches into files/ Mar 25 14:38:53 * remove unused files and patches Mar 25 14:38:53 * fix mkspecs to define RCC and UIC3 locations Mar 25 14:59:14 hi Mar 25 14:59:50 is it just my (4Mb/s) connection or is cloning from git.openembedded.net incredibly slow? Mar 25 15:00:14 it's taking hours.. Mar 25 15:00:31 under heavy load maybe? Mar 25 15:01:58 mickey|zzZZzz: ping? Mar 25 15:02:56 delire it takes a fair amount of time but as far as i recall not several hours Mar 25 15:03:02 unless the server has a hiccup again Mar 25 15:03:14 eFfeM: ok, cheers Mar 25 15:03:39 zecke: ping? Mar 25 15:03:48 eFfeM: it averages around 2kb/s here.. Mar 25 15:03:54 server seems fine from what I can tell -- no processes taking lots of memory or CPU time Mar 25 15:04:07 cbrake: right. tis odd. Mar 25 15:04:29 i try to compile gxine, i get configure errors such as "error: m4_require: circular dependency of AC_LANG_COMPILER(C++)" Mar 25 15:04:30 * delire is dialing from Spain Mar 25 15:05:33 morning Mar 25 15:06:35 delire: I'm receiving data at: 7.27 MiB | 609 KiB/s (git clone) Mar 25 15:06:47 cbrake: where are you cloning from? Mar 25 15:06:55 git clone git://git.openembedded.net/openembedded Mar 25 15:06:58 * delire doesn't whois out of politeness Mar 25 15:07:08 delire: US Mar 25 15:07:48 anybody knows any link to OE presentations/slides? I want to prepare a speech for a conference. Mar 25 15:11:49 cbrake: hehe odd.. when i git clone git://git.openembedded.net/openembedded into another directory, the first git clone, the one taking a long time, jumps up 100kb/s.. Mar 25 15:12:29 ugh.. then after a few seconds both drop to 1k. ohh well, i'll come back tomorrow and see if it's done. Mar 25 15:35:38 03Jeremy Lainé  07org.openembedded.dev * r82b9cae9c7 10openembedded.git/recipes/qt4/qt4-tools-native_4.4.3.bb: qt4-tools-native: stage mkspecs for the host platform Mar 25 15:38:18 bye all Mar 25 15:39:34 cya Mar 25 15:40:26 delire seems something local, the reason you get fast speed initially could be because some of the data is still somewhere in a cache Mar 25 15:42:11 perhaps, yep.. Mar 25 15:48:28 why does the error come when we compile libxine: "error: m4_require: circular dependency of AC_LANG_COMPILER(C++)" Mar 25 16:01:39 03Jeremy Lainé  07org.openembedded.dev * rac41ae95b1 10openembedded.git/recipes/qt4/ (9 files): qt4: consolidate include files into qt4.inc Mar 25 16:25:09 vsr: there were reports http://bugs.openembedded.net/show_bug.cgi?id=4488 Mar 25 16:42:51 morning :) Mar 25 16:47:58 hey Jay7 Mar 25 16:50:46 ant_work: hi :) Mar 25 16:51:22 what is short resume about stable branch? :) Mar 25 17:00:04 03Stefan Schmidt  07fso/milestone5.5 * r638dc47abc 10openembedded.git/packages/wpa-supplicant/ (4 files in 2 dirs): Mar 25 17:00:04 wpa-supplicant: Prepare >=0.6 for dbus system activation Mar 25 17:00:04 * Fix binary path in service file Mar 25 17:00:04 * Install service under the right name taht it gets detected and used correctly Mar 25 17:00:04 * Bump PR's Mar 25 17:00:14 Jay7, hrw has the notes, but he had another meeting right after Mar 25 17:00:34 Jay7: I don't find the link in hystory... Mar 25 17:00:35 well, waiting for report in ML :) Mar 25 17:00:36 03Stefan Schmidt  07fso/milestone5.5 * r3753273cc3 10openembedded.git/packages/connman/connman_0.13.bb: Mar 25 17:00:36 connman: Remove version 0.13 as it is b0rked and segfaults on startup. Mar 25 17:00:36 Wait for 0.14 or use 0.12 Mar 25 17:01:12 I've forget run irc-client to create own log :) Mar 25 17:02:08 Jay7: juice is in 20-30 lines of text on hrw server Mar 25 17:02:41 lost the link... Mar 25 17:04:23 Topic for #oe-stable: Discussion about stable branch for OpenEmbedded. channel is logged, log will be published. WIP summary at http://marcin.juszkiewicz.com.pl/tmp/oe-stable.summary.txt Mar 25 17:05:27 Jay7: remember to send your ssh keys! Mar 25 17:05:35 ah.. Mar 25 17:05:54 * Jay7 have forget about ssh keys :) Mar 25 17:06:21 ant_work: but I have nothing to commit now :) Mar 25 17:06:56 I guess you'll have a lot of free-time now... Mar 25 17:07:01 oking ;-) Mar 25 17:08:20 the first year will be ok, then your daughter will reach the keyboard :d Mar 25 17:09:23 ant_work: and you will buy a laptop for her (like I did for mine ;-) ) Mar 25 17:09:55 ant_work: yeah :)) Mar 25 17:12:16 then you'll have to fix that 'sugar' image Mar 25 17:12:31 :p Mar 25 17:13:57 stefan_schmidt: read the wpa-supplicant-0.6.inc file, we beat you to dbus starting Mar 25 17:22:23 XorA: Ah, I see the right name is already used in dev. Still you have the wrong path for the binary that get's executed? Mar 25 17:22:43 stefan_schmidt: that is fixed with sed in the same file Mar 25 17:22:47 XorA: The service file contains /sbin and OE has it in /usr/sbin Mar 25 17:22:58 hmm Mar 25 17:23:01 stefan_schmidt: we need to use sed because ${sbindir} is the correct value Mar 25 17:23:16 which may not be /usr/sbin/ Mar 25 17:23:31 stefan_schmidt: Ive been using connman on tosa for months Mar 25 17:24:27 good Mar 25 17:24:50 much saner than network manager Mar 25 17:26:29 indeed Mar 25 17:26:51 So if you already use it that long what do you use for dhcp. dhclient or udhcpc? Mar 25 17:33:24 udhcpc Mar 25 17:33:39 although that was failing at FOSDEM Mar 25 17:33:49 it hasnt failed on all the networks I tested within UK though Mar 25 17:33:58 even dodgy pub ones Mar 25 17:35:06 hmm, it working here sometimes and sometimes not. Strange. Mar 25 17:35:30 Ah, your 'months' is actually since february. :) Mar 25 17:35:31 odd, only place I had it fail was FOSDEM Mar 25 17:35:44 stefan_schmidt: thats 2 months :-) Mar 25 17:35:45 I wondered why I missed it, but that was the time for my studies Mar 25 17:36:00 XorA: In two weeks ;) Mar 25 17:36:04 We used to ship udhcp in products at $WORK-1, never gave us any bother. Mar 25 17:36:18 stefan_schmidt: well I was running it long before we worked out the kinks and committed Mar 25 17:36:25 broonie: udhcpc not. Only in combination with connman here. Mar 25 17:36:29 stefan_schmidt: had to start wpa supplicant by hand though :-) Mar 25 17:36:46 XorA: Yeah, was doing the same before. Mar 25 17:37:12 * XorA eventually looked at debian and discovered the path issues :-) Mar 25 17:37:12 Anyway, need to dig into why connman does not like to play with udhcpc friendly Mar 25 17:37:22 heh, same here Mar 25 17:37:27 Also for the file name Mar 25 17:37:44 No idea why they do not give it the right name in the upstream tarball tho Mar 25 17:38:03 I think dbus changed how it worked Mar 25 17:38:38 there is no point in Name=fi.epitest.hostap.WPASupplicant if the file has to have that name Mar 25 17:38:51 just 2 oppurtunities to fuckup Mar 25 17:39:05 yup Mar 25 17:39:30 anyway go forth and connect freely :-) Mar 25 17:40:30 wxwidgets does not build for Beagle, but does build for OSK Mar 25 17:40:38 bother Mar 25 17:41:20 XorA: I'm a bit behind...blocked by bootloader hacks...how's bluetooth with connman? Mar 25 17:41:32 Crofton|work: Hey, latest default kernel does boot itself but did not bring us the userland correctly on osk5912 Mar 25 17:41:43 heh Mar 25 17:41:45 Crofton|work: Have you tried one recently Mar 25 17:41:52 ant_work: havent tried it, no reliable BT device Mar 25 17:41:53 the one before did mount usbfs :( Mar 25 17:42:08 I tried enabling tmpfs, but seems to break everything Mar 25 17:42:11 same userspace boots with a "known to work" kernel Mar 25 17:42:29 hmmm Mar 25 17:42:30 I see Mar 25 17:42:51 I ahve powered it up though Mar 25 17:42:54 hang on Mar 25 17:42:56 udev is trying to use tmpfs and I guess it dies doing so Mar 25 17:43:40 I'm not sure what the problem is Mar 25 17:43:48 Crofton|work: I was trying to boot 2.6.29 with the .config from OE kernel and I have similar problem Mar 25 17:43:49 I'm concerned OSK support is bitrotted Mar 25 17:43:59 so I think it could be something in udev needs love Mar 25 17:44:00 we need tmpfs, right? Mar 25 17:44:22 Crofton|work: huh you mean in kernel Mar 25 17:44:32 yes Mar 25 17:44:32 XorA: is udev working fine on Tosa btw? Mar 25 17:44:43 ant_work: yes Mar 25 17:44:48 Crofton|work: I hope not Mar 25 17:44:52 on other Z there is a prolem with devices tarball Mar 25 17:44:54 well, the angstrom fstab wants it Mar 25 17:45:27 03Stefan Schmidt  07fso/milestone5.5 * rdd37403ed0 10openembedded.git/packages/wpa-supplicant/ (4 files in 2 dirs): Mar 25 17:45:27 Revert "wpa-supplicant: Prepare >=0.6 for dbus system activation" Mar 25 17:45:27 This reverts commit 638dc47abc7157500630cccefe10521378771364. Mar 25 17:45:32 03Stefan Schmidt  07fso/milestone5.5 * rb0af66603d 10openembedded.git/packages/wpa-supplicant/wpa-supplicant-0.6.inc: Mar 25 17:45:32 wpa-supplicant 0.6.7: fix dbus service file Mar 25 17:45:32 Conflicts: Mar 25 17:45:33 i do not know much about tmpfs Mar 25 17:45:33 ant_work: but still need to work out why backlight is not working Mar 25 17:45:34 packages/wpa-supplicant/wpa-supplicant_0.6.7.bb Mar 25 17:45:38 03Stefan Schmidt  07fso/milestone5.5 * r9bac4d763b 10openembedded.git/packages/wpa-supplicant/wpa-supplicant-0.6.inc: Mar 25 17:45:41 wpa-supplicant-0.6.inc : fix supplicant dbus service file name Mar 25 17:45:43 wpa-supplicant_0.6.7.bb : PR bump for above change Mar 25 17:45:44 XorA: offhand I don't remember exactly...possibly tarball is read-only Mar 25 17:45:45 Conflicts: Mar 25 17:45:47 packages/wpa-supplicant/wpa-supplicant_0.6.7.bb Mar 25 17:45:49 03Stefan Schmidt  07org.openembedded.dev * r9c889bdeb8 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into org.openembedded.dev Mar 25 17:45:52 03Stefan Schmidt  07org.openembedded.dev * r3bc21dfdeb 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into org.openembedded.dev Mar 25 17:45:55 03Stefan Schmidt  07org.openembedded.dev * r158be4c48f 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into org.openembedded.dev Mar 25 17:45:59 WTF? Why did git changed the authors for a cherry-pick? Mar 25 17:46:02 03Stefan Schmidt  07org.openembedded.dev * r14438b191f 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into org.openembedded.dev Mar 25 17:46:05 03Stefan Schmidt  07org.openembedded.dev * r6767ca5043 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into org.openembedded.dev Mar 25 17:46:10 03Stefan Schmidt  07org.openembedded.dev * r451b1c6871 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into org.openembedded.dev Mar 25 17:46:13 03Stefan Schmidt  07org.openembedded.dev * r95e0882e2d 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into org.openembedded.dev Mar 25 17:46:15 stefan_schmidt: You become the committer. Mar 25 17:46:16 03Stefan Schmidt  07org.openembedded.dev * rc26d2cb4e1 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into org.openembedded.dev Mar 25 17:46:30 broonie: But why? It was a cherry-pick Mar 25 17:46:30 tick tick tick Mar 25 17:46:34 stefan_schmidt: The author will (if you did a cherry pick) still be recorded probably but you did the actual commit. Mar 25 17:46:49 stefan_schmidt: git records both bits of information. Mar 25 17:46:54 broonie: ok, that makes sense Mar 25 17:46:55 khem, I would really like to debreak the OSK Mar 25 17:47:09 any thoughts on how to proceed? Mar 25 17:47:20 We can go back kernels and see what works Mar 25 17:47:33 Crofton|work: I have been using a custom kernel for long time Mar 25 17:47:47 based on 2.6.28 release Mar 25 17:48:31 Crofton|work: I think I need to try few things Mar 25 17:48:44 different compiler/binutils to compile the kernel itself Mar 25 17:49:06 this shouldn't be that hard :) Mar 25 17:49:29 yep Mar 25 17:49:51 I'm worried something bitrotted in linux-omap Mar 25 17:49:57 then also try out different rootfs one with pure arm and one with mixed thumb-arm like we have these days in OE Mar 25 17:50:38 pure arm should just boot Mar 25 17:50:54 there is a guy trying to get the OSK working atm Mar 25 17:51:03 his interest is USB serial dongles Mar 25 17:51:12 hmm Mar 25 17:51:42 I dont have much of hardware to play around other than qemu Mar 25 17:51:47 :) Mar 25 17:53:13 I ahve an osk Mar 25 17:53:23 building .27 with LL DEBUG Mar 25 17:53:26 and will flash Mar 25 17:54:09 Crofton|work: I use tftpboot generally is faster Mar 25 17:56:29 bb Mar 25 18:30:39 awesome I have backlight :-) Mar 25 18:34:03 i'm newbie Mar 25 18:34:14 need startup guide or book or tutorial to learn embedded programming Mar 25 18:57:21 hotbabvy1: embedded programming need not be that different from normal coding these days and if your desire is to write apps then OpenEmbedded is not the place to look. Mar 25 19:52:05 exit Mar 25 20:03:52 Crofton|work: I would like to move preferred kernel to 2.6.28 for osk5912 Mar 25 20:04:05 ok :) Mar 25 20:04:09 that kernel boots fine both Mar 25 20:04:17 uclibc/thumb and eglibc/arm userspaces Mar 25 20:04:22 can you push Mar 25 20:04:32 I assume it will work on glibc? Mar 25 20:04:55 I'm caught up on some stuff I ahve to do atm Mar 25 20:05:17 unpleasant stuff, like ordering a laptop to tun windows on :( Mar 25 20:05:34 put it in a VM instead Mar 25 20:06:21 almost practical Mar 25 20:16:42 what would you review before you will be interviewed for embedded application developer position? Mar 25 20:19:19 Re-read the K&R book Mar 25 20:19:36 why ?CosmicPenguin? Mar 25 20:19:37 your facebook profile.... Mar 25 20:19:38 :) Mar 25 20:19:50 Crofton|work: yes it should. OE guys should use eglibc Mar 25 20:20:25 why OcsmicPenguin? Mar 25 20:20:32 why CosmicPenguin? Mar 25 20:20:47 hey khem Mar 25 20:20:50 don't read anything Mar 25 20:20:56 I've got a patch to gcc-4.3.3 for cortex a9 support Mar 25 20:20:59 just answer the questions as best you can Mar 25 20:21:01 wanna see it quick before I push sometime soon? Mar 25 20:21:08 and do not try to BS the interviewers Mar 25 20:21:12 Tartarus: hi Mar 25 20:21:34 Tartarus: ok cool Mar 25 20:21:57 is it a backport ? Mar 25 20:22:43 Yeah Mar 25 20:22:48 just sent w/ a pointer to the original Mar 25 20:22:57 I think it's an easy backport :) Mar 25 20:23:13 BS the interviewers? Mar 25 20:23:25 Bull shit them Mar 25 20:24:51 did we get trolled? Mar 25 20:25:15 03Khem Raj  07org.openembedded.dev * r5f7f81339d 10openembedded.git/recipes/linux/linux-omap_2.6.28.bb: Mar 25 20:25:15 linux-omap_2.6.28.bb: omap5912osk's default preferred kernel. Mar 25 20:25:15 Signed-off-by: Khem Raj Mar 25 20:25:29 Tartarus: OK I will have a a look Mar 25 20:26:07 k, it's apparently not quite right yet :( Mar 25 20:35:56 khem, bah, I think some other stuff needs backporting too Mar 25 20:36:46 reading the log on http://gcc.gnu.org/viewcvs/trunk/gcc/config/arm/arm.c?view=log I think r139860 / 65 / 74 / 81 need pulling in too Mar 25 20:39:55 I'm trying to add a 'kernel-headers' package to kernel.bbclass without success. When I add 'kernel-headers' on the end of PACKAGES defined there bitbake doesn't seem to know that its provided by kernel recipes - any ideas? Mar 25 20:42:24 my patch and explanation/questions are here http://www.pastebin.ca/1372060 Mar 25 21:05:56 tharvey: We have linux-libc-headers recipe for same Mar 25 21:06:31 although I always thought it would be better to use headers from the same kernel sources that we use to build the kernel Mar 25 21:06:32 khem, ya I saw that... for includes in libc - is that really the same thing as kernel headers needed for building out of tree kernel modules with meta-toolchain though? Mar 25 21:06:46 ah, yes you just answered the question then... they are not the same headers? Mar 25 21:07:12 tharvey: yes they are more of a glibc needed kernel headers Mar 25 21:07:54 tharvey: although it would be good to have kernel headers as you are doing and then probably we can use same for glibc Mar 25 21:08:13 true... do you see what I might be doing wrong then? Mar 25 21:08:20 * khem looks Mar 25 21:08:23 perhaps I need DYNAMIC_PACKAGES? Mar 25 21:09:21 tharvey: kernel headers do not install into / Mar 25 21:09:31 I'm a bit mystified why bitbake would say kernel-headers is an unresolved dep when clearly an ipk was created by rebuilding virtual/kernel Mar 25 21:09:40 they should be installed into /usr/include Mar 25 21:10:22 true, thats easy to fix though Mar 25 21:10:42 you mean INSTALL_HDR_PATH="${D}/usr" ? Mar 25 21:10:50 whats error you get Mar 25 21:11:19 will take a min to recreate the error from populate-sdk Mar 25 21:11:46 but in general... why can't I just 'bitbake kernel-headers' if kernel.bbclass has declared that in PACKAGES? Mar 25 21:19:53 should work Mar 25 21:21:39 khem: aren't linux-headers a 'sanitized version' ? Mar 25 21:21:57 of kernel headers? Mar 25 21:27:43 i'm closely following the instructions on http://elinux.org/BeagleBoardAndOpenEmbeddedGit but am having problems seeing why i get an error: ERROR: opening $OE_HOME/openembedded/recipes/*/*.bb: file $OE_HOME/openembedded/recipes/*/*.bb not found" while bitbaking despite ls seeing that actual path. Mar 25 21:27:52 here is the total output: http://rafb.net/p/0sOwAb77.html Mar 25 21:27:57 any ideas where i'm going wrong? Mar 25 21:28:39 delire: the instrcutions there for the local.conf are wrong. Mar 25 21:28:46 it should be ${OE_HOME} not $OE_HOME in the .conf file Mar 25 21:28:47 'ls $OE_HOME/openembedded/recipes/*/*.bb | wc -l' lists thousands of files. Mar 25 21:28:53 kergoth: ahah.. Mar 25 21:28:59 * delire tries this. Mar 25 21:30:38 kergoth: awesome, that's definitely doing something. cheers. Mar 25 21:30:47 :) Mar 25 21:31:02 * delire notes a distinct feeling of relief coupled with general gladness. Mar 25 21:44:23 kergoth: can you pls explain ALTERNATIVE_PRIORITY = "x" ? Mar 25 21:44:37 (psplash) Mar 25 21:44:55 I've never used psplash in my life Mar 25 21:45:00 kergoth: ERROR: Multiple .bb files are due to be built which each provide virtual/psplash (/oe/org.openembedded.dev/recipes/psplash/psplash-angstrom_svn.bb /oe/org.openembedded.dev/recipes/psplash/psplash_svn.bb). Mar 25 21:45:00 This usually means one provides something the other doesn't and should. Mar 25 21:45:37 to take a gues, id' say alternative_prioritiy is for update-alternatives, and has nothing to do with bitbake's provider selection Mar 25 21:45:41 but i haven't looked at it Mar 25 21:45:52 ok, thx Mar 25 21:46:47 I'll see how other virtuals are implemented then Mar 25 21:50:12 khem, sorry got distracted - the kernel-headers added to kernel.bbclass (http://www.pastebin.ca/1372060) does not work when I add 'kernel-headers' to RDEPENDS of a meta-toolchain based recipe Mar 25 21:50:17 fails with RDEPENDS/RRECOMMENDS or otherwise requires the runtime entity 'kernel-headers' but it wasn't found in any PACKAGE or RPROVIDES variables Mar 25 21:50:50 I'm thinking this has something to do with the fact that kernel recipes are virtual providers? Mar 25 21:51:32 note that rebuilding my virtual/kernel with the patched kernel.bbclass 'does' build a kernel-headers_*.ipk with the correct kernel headers Mar 25 22:10:30 kergoth: seems fixed setting a distro PREFERRED_PROVIDER for the new virtual/psplash. I'll send the patch to the angstrom ML. Possibly 'minimal' is also affected... Mar 25 22:10:46 ah Mar 25 22:11:23 otavio added the new virtual recently Mar 25 22:27:26 XorA: ping Mar 25 22:34:48 hello xjqian Mar 25 22:35:03 hi ant Mar 25 22:36:10 kergoth: your help would be appreciated again..sorry I'm exaggerating... Mar 25 22:36:15 base.bbclass Mar 25 22:36:23 package_stagefile_shell() { Mar 25 22:36:39 if you're askinga bout the sed, i have no idea what it's trying to accomplish, or how to fix it, offhand Mar 25 22:36:44 haven't investigated it Mar 25 22:36:55 I see a recipe breaking if DEPLOY_DIR is moved Mar 25 22:37:05 and yes, I edited that sed Mar 25 22:37:21 this way: destfile=`echo $srcfile | sed s#${DEPLOY_DIR}#${PSTAGE_TMPDIR_STAGE}#` Mar 25 22:37:42 all fine but zaurus-updater...sob Mar 25 22:37:55 in do_deploy Mar 25 22:37:57 package_stagefile_shell ${DEPLOY_DIR_IMAGE}/updater.sh.${MACHINE} Mar 25 22:38:35 result: | cp: cannot stat `updater.sh.c7x0': No such file or directory Mar 25 22:42:18 kergoth: I'd fix the recipe doing cp ...instead of calling package_stagefile_shell Mar 25 22:42:33 atm it's the only one broken Mar 25 22:47:49 03Graeme Gregory  07org.openembedded.dev * re1433571e5 10openembedded.git/recipes/linux/ (4 files in 2 dirs): linux_2.6.29.bb : new version, tosa support added. Mar 25 22:48:05 :-) Mar 25 22:49:39 ant__: tosa pulls ahead in the race :-) Mar 25 22:49:51 and how! Mar 25 22:50:32 tose = longer legs? Mar 25 22:52:40 * XorA wonders if it suspends Mar 25 22:54:07 XorA: I finally have somehow updater.sh in packaged staging... Mar 25 22:54:41 wow it recovers from suspend Mar 25 22:54:55 so..you're ahead spitz/akita !!! Mar 25 22:55:13 Dmitry did a good job obviously Mar 25 22:55:25 now please back-fix 2.6.28 Mar 25 22:55:27 I had to talk to him about one problem which we came up with a patch for Mar 25 22:56:03 power-man? Mar 25 22:56:18 does battery recharge? Mar 25 22:56:26 ant__: seems to Mar 25 22:56:40 so..full-working...great Mar 25 22:58:28 bugger Ive lost my libertas firmware Mar 25 23:00:35 XorA: what do you think about http://rafb.net/p/Ib7JMm11.html Mar 25 23:00:50 would you try a rebuild? Mar 25 23:01:01 just zaurus-updater :-) Mar 25 23:01:22 gimme a few mins, working on libertas Mar 25 23:01:26 np Mar 25 23:03:42 XorA: the four patches http://rafb.net/p/QqXirk94.html (so you understand) Mar 25 23:05:05 ant__: its too far in the pm for me to work out what you are doing :-) Mar 25 23:07:12 he he...the problem is DEPLOY_DIR get replaced Mar 25 23:07:17 by sed Mar 25 23:24:37 * ant__ doesn't receive mail anymore from openembedded-commits Mar 25 23:26:59 tharvey: what happens when you use DEPENDS Mar 25 23:30:37 khem, same thing Mar 25 23:31:34 shouldn't I simply be able to do a 'bitbake kernel-headers'? that fails as well with 'ERROR: Nothing PROVIDES 'kernel-headers'' Mar 25 23:33:49 tharvey: let me check couple of things Mar 25 23:34:51 03Stefan Schmidt  07kergoth/master * r9c889bdeb8 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into org.openembedded.dev Mar 25 23:34:52 03Stefan Schmidt  07kergoth/master * r14438b191f 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into org.openembedded.dev Mar 25 23:34:52 03Stefan Schmidt  07kergoth/master * rc26d2cb4e1 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into org.openembedded.dev Mar 25 23:34:53 03Stefan Schmidt  07kergoth/master * r3bc21dfdeb 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into org.openembedded.dev Mar 25 23:34:56 03Stefan Schmidt  07kergoth/master * r158be4c48f 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into org.openembedded.dev Mar 25 23:34:59 03Stefan Schmidt  07kergoth/master * r6767ca5043 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into org.openembedded.dev Mar 25 23:35:00 erm, thats odd Mar 25 23:35:04 03Stefan Schmidt  07kergoth/master * r451b1c6871 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into org.openembedded.dev Mar 25 23:35:06 stupid CIA Mar 25 23:35:07 03Jeremy Lainé  07kergoth/master * rcfb474d939 10openembedded.git/recipes/qt4/qt4-native_4.4.3.bb: qt4-native: add package to replace qmake2-native and uicmoc4-native Mar 25 23:35:10 03Robert Schuster  07kergoth/master * r5575ee4e7a 10openembedded.git/recipes/llvm/ (llvm-2.5/fix-build.patch llvm_2.4.bb llvm_2.5.bb): llvm 2.5: Updated patch to include preprocessor definitions. Mar 25 23:35:13 03Jeremy Lainé  07kergoth/master * r8e64571067 10openembedded.git/recipes/qt4/ (qt4-native-4.4.3/qt-config.patch qt4-native_4.4.3.bb): qt4-native: rename to qt4-tools-native Mar 25 23:35:16 03Jeremy Lainé  07kergoth/master * r1476b97656 10openembedded.git/recipes/qt4/ (qt4-native-4.4.3/qt-config.patch qt4-native_4.4.3.bb): qt4-native: add qt-config.patch from qmake2-native Mar 25 23:35:17 * kergoth sighs Mar 25 23:35:25 03Jeremy Lainé  07kergoth/master * rcfabec5a23 10openembedded.git/recipes/neuros-public/ (7 files): neuros-*: don't force DEPENDS, let qt4e.bbclass handle it Mar 25 23:35:28 03Jeremy Lainé  07kergoth/master * ra1f6020e21 10openembedded.git/recipes/qt4/ (14 files in 3 dirs): qt4-x11-free: drop obsolete versions as per oe-devel Mar 25 23:35:31 03Jeremy Lainé  07kergoth/master * rac41ae95b1 10openembedded.git/recipes/qt4/ (9 files): qt4: consolidate include files into qt4.inc Mar 25 23:35:34 03Jeremy Lainé  07kergoth/master * r2fa972014d 10openembedded.git/ (7 files in 2 dirs): qt4: make use of qt4-tools-native instead of qmake2-native and uicmoc4-native Mar 25 23:35:34 those hooks need work. Mar 25 23:35:37 03Jeremy Lainé  07kergoth/master * r88c677023e 10openembedded.git/ (15 files in 3 dirs): qt-embedded: rename to qt4-embedded for consistency Mar 25 23:35:42 03Chris Larson  07kergoth/master * rf74b74c526 10openembedded.git/classes/ (4 files): Mar 25 23:35:47 First pass of cleanup of messages outputted to the user. Mar 25 23:35:47 they shouldn't blindly dump commits that's been previously submitted like hat Mar 25 23:35:49 OpenEmbedded outputs a lot of messages that the user is likely to never Mar 25 23:35:51 care about. We should only output something when it reflects upon their Mar 25 23:35:53 recipe (i.e. unpacking their sources, applying their patches), or is quite Mar 25 23:35:57 significant or unusual. Mar 25 23:35:59 Signed-off-by: Chris Larson Mar 25 23:36:01 03Jeremy Lainé  07kergoth/master * r8809e3ba81 10openembedded.git/recipes/qt4/qt_staging.inc: qt_staging.inc: fix symlink to qmake Mar 25 23:36:04 03Stefan Schmidt  07kergoth/master * r95e0882e2d 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into org.openembedded.dev Mar 25 23:36:11 03Chris Larson  07kergoth/master * rf01404828b 10openembedded.git/classes/cross.bbclass: Mar 25 23:36:14 Revert "cross.bbclass: set autoconf's program prefix to keep TARGET_SYS and TARGET_PREFIX decoupled" Mar 25 23:36:16 This reverts commit 114155243344c15847e73598e82d02926443cbf1. Mar 25 23:36:18 03Jeremy Lainé  07kergoth/master * r82b9cae9c7 10openembedded.git/recipes/qt4/qt4-tools-native_4.4.3.bb: qt4-tools-native: stage mkspecs for the host platform Mar 25 23:36:22 * kergoth kicks CIA-2 Mar 25 23:36:23 03Khem Raj  07kergoth/master * r5f7f81339d 10openembedded.git/recipes/linux/linux-omap_2.6.28.bb: Mar 25 23:36:26 linux-omap_2.6.28.bb: omap5912osk's default preferred kernel. Mar 25 23:36:26 khem, actually makes sense that I can't bitbake kernel-headers as its not a recipe but a package... the fact that bitbake things its not provided though stumps me Mar 25 23:36:28 Signed-off-by: Khem Raj Mar 25 23:36:30 (6 lines omitted) Mar 25 23:36:36 ow Mar 25 23:37:20 tharvey: you can add it to PROVIDES Mar 25 23:37:24 then it will work Mar 25 23:38:08 add this PROVIDES += "kernel-headers" Mar 25 23:38:13 to the bbclass Mar 25 23:40:09 tharvey: I will try to leaverage your patch to use in toolchain builds too and then we wont need linux-libc-headers package Mar 25 23:41:01 that would be great... esp if we can figure out why it doesn't work :) Mar 25 23:41:13 khem: I'm pretty sure other people discourages compiling glibc against kernel headers (Gentoo) Mar 25 23:41:29 Gentoo provides sanitized headers Mar 25 23:41:41 hmm... I missed this: NOTE: Runtime target 'kernel-headers' is unbuildable, removing... Mar 25 23:41:41 does this makes sense? Mar 25 23:41:52 doesn't tell me why its unbuildable though Mar 25 23:42:47 ant__: santized headers is not integrated into linux kernel Mar 25 23:43:04 we do not need extra bits as we did in the past Mar 25 23:43:21 ofcourse I am talking about 2.6.18+ kernels Mar 25 23:43:27 interesting Mar 25 23:44:05 its more integrated if you use same kernel sources for generating the headers for glibc and rest of things Mar 25 23:44:18 right now we do not do that Mar 25 23:44:48 exactly as others are doing Mar 25 23:45:31 tharvey: unbuildable ? Mar 25 23:46:09 usually these sanitized headers are some release older (I remember nptl thing years ago...) Mar 25 23:47:51 03Graeme Gregory  07org.openembedded.dev * reda924db4b 10openembedded.git/conf/machine/include/zaurus-2.6.inc: zaurus-2.6.inc : use linux as the tosa kernel instead of linux-rp Mar 25 23:47:51 03Graeme Gregory  07org.openembedded.dev * re2e4fe0cfb 10openembedded.git/recipes/linux/ (linux_2.6.28.bb linux_2.6.29-rc6.bb): linux_2.6.28, linux_2.6.29-rc6 : demote tosa for these recipes Mar 25 23:48:56 ant__: those days are gone Mar 25 23:49:16 :-) Mar 25 23:49:39 ant__: now you can take any upstream release and use the headers make INSTALL_HDR_PATH=/some/where/ headers_install Mar 25 23:49:45 ant__: I have endured that pain Mar 25 23:49:53 khem, let me patch my change into meta-toolchain and pastebin the whole diff/error Mar 25 23:50:12 tharvey: what happens if you make it virtual Mar 25 23:54:00 khem, all I do for that is add kernel-headers to PROVIDES in kernel.bbclass right? that didn't help either Mar 25 23:54:40 03Chris Larson  07kergoth/master * r534eae1fbc 10openembedded.git/classes/ (base.bbclass packaged-staging.bbclass patch.bbclass): Mar 25 23:54:40 Shorten some full paths printed to the user. Mar 25 23:54:40 Adds a base_path_out convenience function, which prepares a full path for Mar 25 23:54:40 display to the user. The initial implementation just makes it relative to Mar 25 23:54:40 ${TOPDIR}. This function is then used for some messages outputted to the Mar 25 23:54:44 user (packaged-staging, patch application, clean, unpack tasks). Mar 25 23:54:46 Signed-off-by: Chris Larson Mar 26 00:03:23 tharvey: Mar 26 00:03:24 NOTE: multiple providers are available for kernel-headers (linux, linux-omap1, linux-bast, linux-omap-pm, linux-rt, linux-omap, linux-omap-2.6, gumstix-kernel); Mar 26 00:03:28 NOTE: consider defining PREFERRED_PROVIDER_kernel-headers Mar 26 00:03:30 this is what I get Mar 26 00:03:53 so now if I add PREFERRED_PROVIDER_kernel-headers="linux-omap" Mar 26 00:07:12 khem, grep -r kernel-headers org.openembedded.dev/recipes/linux/* returns nothing? Mar 26 00:07:45 I applied your patch Mar 26 00:13:01 khem, where? don't see it on http://cgit.openembedded.net/cgit.cgi?url=openembedded/log/ Mar 26 00:15:16 i mean locally Mar 26 00:15:51 ah... still trying to get a meta-toolchain built which includes it Mar 26 00:17:07 khem, looks like your the one that can answer the question I posted to #beagle a bit ago - can you explain the purpose/reason for all the linux-omap kernel recipes? trying to understand what linux-omap1 and linux-omap2 is when beagleboard (omap3x) uses linux-omap- Mar 26 00:17:45 and I notice that gumstix tree uses a linux-omap3- for overo which makes sense to me if omap1, omap2, omap3 refer to family generation Mar 26 00:18:30 tharvey: yes linux-omap recipes are using tony's tree Mar 26 00:18:46 we have omap to indicate family Mar 26 00:18:59 but now we are trying to consolidate into linux-omap Mar 26 00:19:02 only Mar 26 00:19:07 ahhh... ok Mar 26 00:19:25 osk5912 already used linux-omap Mar 26 00:19:29 now Mar 26 00:19:39 earlier it was using linux-omap1 Mar 26 00:20:08 yes, I noticed your recent patch there Mar 26 00:20:18 ok, so consider everything but linux-omap-* out of date? Mar 26 00:29:09 khem, I'm not seeing where kernel-headers is declared for linux-omap for example Mar 26 00:29:19 or do you mean after my patch to kernel.bbclass? Mar 26 02:04:38 after your patch **** ENDING LOGGING AT Thu Mar 26 02:59:57 2009