**** BEGIN LOGGING AT Mon Jul 11 02:59:56 2011 Jul 11 07:11:54 good morning Jul 11 07:33:31 How to change the version of linux-libc-headers-native? Jul 11 07:33:57 I changed the version in distro configuration to 2.6.39 Jul 11 07:34:21 Target version of linux-libc-headers 2.6.39 is used, but for native 2.6.30 Jul 11 07:37:30 03Joel A Fernandes  07org.openembedded.dev * rd808b945d9 10openembedded.git/recipes/linux/ (2 files in 2 dirs): Jul 11 07:37:30 linux-omap-psp: Fix the user button GPIO to the correct value for a BeagleBoard-xM Jul 11 07:37:30 Signed-off-by: Joel A Fernandes Jul 11 07:37:30 Signed-off-by: Koen Kooi Jul 11 07:37:41 03Joel A Fernandes  07org.openembedded.dev * r2801aee777 10openembedded.git/recipes/beagleboard/beagleboard-test-scripts_git.bb: Jul 11 07:37:41 beagleboard-test-scripts: Add a script flash-fs.sh for flashing NAND Jul 11 07:37:41 This script flashes the NAND of a BeagleBoard if: Jul 11 07:37:41 * There is a valid image on the SD Card to flash Jul 11 07:37:42 * The board has NAND Jul 11 07:37:42 Signed-off-by: Joel A Fernandes Jul 11 07:37:43 Signed-off-by: Koen Kooi Jul 11 07:44:27 Anyone has ideas how to force it jump to the right version? Jul 11 08:08:44 hey guyes did any one work on touchscreen module for mini6410 friendly arm board,I am using lcd a70,Is there any problem with s3c touchscreen driver I am using linux kernel 2.6.28 ..... Jul 11 09:19:36 hi mickey|office Jul 11 09:21:03 hi pb_, all Jul 11 09:26:31 good morning pb_ Jul 11 09:26:32 hi bluelightning Jul 11 09:26:42 hi mickey|office Jul 11 09:33:19 hello, can somebody please fix the libgisi_git.bb recipe on old oe dev ? see http://shr-project.org/trac/ticket/1459 version ee474b2bf7a10b244d52f5333b195609003762cf is needed. Jul 11 09:48:07 khem: ping Jul 11 10:04:02 nschle85: khem is probably asleep, it's something like 3am where he is. Jul 11 10:11:31 morning pb_: RE your mail. Your instincts were right, a nonexistent package does cause an error. Although I can't fathom why. The resulting run script *looks* right, and shell experiments suggest it should work. Jul 11 10:15:46 pb_: can you have a look at : http://shr-project.org/trac/ticket/1459 version ee474b2bf7a10b244d52f5333b195609003762cf is needed. can you change that for me on old oe dev ? Jul 11 10:34:13 I think you should get one of the SHR maintainers to do that. Jul 11 10:34:24 celston: ah, hm. what's the error? Jul 11 10:34:56 That's the confusing bit, there isn't one :) Jul 11 10:35:14 I'll work it out, just wanted to moan. Jul 11 10:35:22 oh, right. heh. Jul 11 10:35:23 pb_: this is not an shr bug Jul 11 10:35:42 if I had to guess, I would say that it's probably that "opkg info xx" is exiting with an error status but not printing any sensible diagnostic Jul 11 10:36:08 pb_: the libgisi version does not match this should happen to all distributions Jul 11 10:38:23 pb_: the shr maintainer is in vacation, he wants that someone else does it (see ticket) Jul 11 10:38:57 pb_: do you have commit right to oe dev ? Jul 11 10:39:25 theoretically yes, though I don't use the .dev tree myself anymore Jul 11 10:51:52 hi folks, I just started a new project on oe. I did some work on oe already a few yares ago. Jul 11 10:54:38 would it be possible to get a private branch for some package addition, and fixes? Jul 11 10:55:12 is it better to ask here on irc or should I ask on the mailing list? Jul 11 10:55:32 well, asked already :-) Jul 11 10:57:24 yes, you can create as many private branches as you like Jul 11 10:59:17 pb_: on the oe git repo? I though somebody with write acces needs to create the branch. I have to admit I am somewhat new to git... Jul 11 11:00:19 from the wiki: Aspiring developers are encouraged to request a personal git tree where they can store and share changes... Jul 11 11:00:42 maybe I should have said "personal git tree" not branch. Jul 11 11:00:59 oh, right, I see what you mean now Jul 11 11:01:10 :-) Jul 11 11:01:12 the easiest way for you to do that is just to create a cloned repo on github Jul 11 11:02:02 ok. will that be visible for everyone, so that I can send patches on the mailing list, that then can be merged? Jul 11 11:02:06 yes Jul 11 11:02:09 coo! Jul 11 11:02:29 just go to https://github.com/openembedded/openembedded and click the "fork" button Jul 11 11:03:11 thanks! Jul 11 11:03:44 pb_:really great, I see still some known faces around, I we have chatted too some years ago... Jul 11 11:04:13 I still know koen, kergoth... Jul 11 11:04:26 pb_: got it, it's the grep. grepping for something in the empty string means grep returns 1. Jul 11 11:04:29 many new faces too... Jul 11 11:04:42 celston: aha, yeah Jul 11 11:05:58 pb_: Would this be a valid case for a bb warning, btw? i.e.: inform the user they've requested an unknown package be excluded, or just log to the task's log? Jul 11 11:06:10 just to the task's log, I think. Jul 11 11:06:30 I don't think it merits a diagnostic since the package might be a perfectly valid one that just doesn't happen to have been built Jul 11 11:07:34 and, if it hasn't been built, it clearly isn't going to be installed so there is no problem there. Jul 11 11:08:08 ack Jul 11 12:33:55 hello , I m getting this message i .... can anyone help me... what are they? Jul 11 12:35:22 http://pastebin.mozilla.org/1269390 Jul 11 12:35:30 this are the error... Jul 11 12:37:02 you're passing invalid parameters to XShmPutImage(). Jul 11 12:52:31 Good morning all Jul 11 12:53:05 I am having trouble with autoconf213_2.13.bb Jul 11 12:54:29 when I run the following command : bitbake autoconf213_2.13, I get the following error message: ERROR: Nothing PROVIDES 'autoconf213_2.13' Jul 11 12:54:56 i need autoconf213 to build mozilla Jul 11 12:55:21 can anyone help me please? Jul 11 12:56:32 cboyd_, DEPENDS on autoconf213 in the bb file for mozilla Jul 11 12:57:06 cboyd_: why are you trying to run that command? Jul 11 12:57:23 "bitbake autoconf213", or "bitbake autoconf213-2.13", would probably work Jul 11 12:57:33 i keep getting an error when trying to build my image Jul 11 12:57:37 but the underscore is erroneous and will indeed not do anything useful Jul 11 12:57:52 no rule for target -lX11 Jul 11 12:58:12 so, er, why did that lead you to run the command you mentioned? Jul 11 12:58:25 that error message doesn't have any obvious connection with autoconf213_2.13 Jul 11 12:58:52 mozilla needs autoconf213 to build Jul 11 12:59:15 my machine list autoconf 2.67 which is imcompatible Jul 11 12:59:39 which mozilla recipe are you trying to build, specifically? Jul 11 13:00:01 firefox_4.0.1.bb Jul 11 13:00:43 my image builds with firefox_3.6.16 Jul 11 13:00:55 but not with firefox_4.0.1 Jul 11 13:01:27 where did you get firefox_4.0.1.bb? I don't see it in my copy of the oe tree. Jul 11 13:03:03 sources/extraoss/recipes/firefox/firefox_4.0.1.bb Jul 11 13:03:25 what repo is that? Jul 11 13:05:08 i got it from my company's repo ... but i believe it can also be found at the git repo Jul 11 13:05:52 ls Jul 11 13:19:56 Hi all. Having fun bitbaking meta-sdk for a pxa board (using oe-core etc) Jul 11 13:20:09 getting: Jul 11 13:20:09 | + opkg-cl -f /home/jic23/src/beagle/setup-scripts/build/tmp-angstrom_2010_x-eglibc/work/armv5te-angstrom-linux-gnueabi/meta-toolchain-1.0-r4/opkg-sdk.conf -o /home/jic23/src/beagle/setup-scripts/build/tmp-angstrom_2010_x-eglibc/work/armv5te-angstrom-linux-gnueabi/meta-toolchain-1.0-r4/sdk/image --force-overwrite install task-sdk-host-nativesdk task-cross-canadian-arm | Unknown package 'task-sdk-host-nativesdk'. | Unknown package 'task-cross-canadian-arm Jul 11 13:20:22 Are they really supposed to exist as packages? Jul 11 13:31:28 meta-toolchain (rather than meta-sdk) Jul 11 13:36:17 ah. More digging later, looks like it's a name of architecture thing. Jul 11 13:36:56 opkg-sdk.conf includes all sorts of variants, but not x86_64-nativesdk which appears to be where the files actually are. Jul 11 13:37:09 Anyone know which side is wrong? Jul 11 13:45:09 ah. Seems to be an issue with the tune-xscale.inc file under meta-smartphone. The oe-core one works fine. Jul 11 13:58:14 morning kergoth_ Jul 11 14:28:40 hm, melo hosed again it seems Jul 11 14:31:56 looks like mysql Jul 11 14:32:10 web-server is working ok and redirecting to wiki Jul 11 14:32:19 really? I didn't think mysql ran on melo Jul 11 14:32:35 iirc, the db is on a separate host Jul 11 14:32:38 slow mysql or storage proble Jul 11 14:33:00 yeah, it looks to me as though the disks have gone away again Jul 11 14:33:46 iirc, cgit does not using any db Jul 11 14:33:58 so this is more likely storage problem Jul 11 14:34:38 yeah. rather annoying that this keeps happening Jul 11 14:35:16 mickey|office: did the ev board have any luck getting funding for better hardware/hosting? Jul 11 14:36:36 pb_: we granted quite a bit for a new switch and there's a saying that some other folks are donating as well. We didn't succeed yet in raising something for OE as is though yet Jul 11 14:36:48 one yet too much Jul 11 14:37:19 ah right, very good Jul 11 14:38:10 are the yocto folks donating anything to the eV? seems as though they ought to be really. Jul 11 14:38:13 it seems though that the yocto thingy didn't make our situation better, i.e the reasoning why OE as is should be funded when funds already go to yocto through the company membership or similar things Jul 11 14:38:59 makes it a bit harder for us to promote ourselves Jul 11 14:39:01 IMO Jul 11 14:39:07 yeah Jul 11 14:41:18 it goes straight in line with the problem of our volunteer membership fees, either we didn't communicate that well enough or the strategy just doesn't work out Jul 11 14:42:31 I guess I was rather expecting/hoping that the yocto project would make a centralised contribution to oe. I guess I can understand that the member companies might feel it was a bit much to pay "twice", though on the other hand I can't help feeling that those OE users with a market cap in excess of a billion dollars ought perhaps to be able to afford it. Jul 11 14:43:26 *nod* i'm going to put that on the agenda for the next e.V. meeting. Jul 11 14:43:31 er Jul 11 14:43:33 board meeting, i mean Jul 11 14:43:37 which happens to be in 20 minutes ;) Jul 11 14:43:56 thanks :-) Jul 11 14:53:53 * mwester really doesn't understand the *why* of the entire yocto/oe-core/oe-classic thing in the first place, much less horrifying things like accounting. Jul 11 14:55:33 I think that's probably at least two different questions. Jul 11 14:56:34 "why oe-core" is mostly to do with the monolithic oe.dev tree, and its associated workflow, becoming something of a maintainability and release engineering headache. Jul 11 15:05:02 pb_, Agreed that OE-classic was becoming a serious maintainability problem, but I find I cannot keep up with the gigantic "patch-bombs" that drop into oe-core at such a high rate (did we fall behind in OE-classic and this is just "catching up", or is this the new norm?) Jul 11 15:06:01 mwester: well, I think part of it is just that you have more visibility now of what's going in. with OE-classic, folks tended just to push their changes and not post them on the list, whereas now (almost) everything that goes in is also posted to the list. Jul 11 15:06:21 also, yeah, oe-core was effectively forked from oe-classic about four or five years ago and there is a certain amount of catching up that needs to be done Jul 11 15:07:17 and, finally, part of it is a git workflow artifact: you tend to get a patch series consisting of 10 or 20 or 100 little patches, each of which gets sent in a separate email even though it might be only a couple of lines being changed. Jul 11 15:08:14 mickey|office: was there another call for donations lately? Jul 11 15:08:20 it might have helped if oe-core had been made much more "core" like, with only core metadata, and not a sh*tload of basic recipes also Jul 11 15:08:29 denix: not that i know of Jul 11 15:08:36 denix: Crofton florian: ##oe-ev-board Jul 11 15:08:56 15 utc is now iirc Jul 11 15:09:06 mickey|office: yes, it's 1500 utc now Jul 11 15:09:22 mickey|office: indeed - many thanks Jul 11 15:10:35 mickey|office: anyway, my point being w/o the actual call for donations, how can we expect anyone to donate? Jul 11 15:10:44 urg Jul 11 15:10:49 esbenh: well, that's just a definitional question of what you consider "core". I'm not sure that redrawing those boundaries would make any qualitative difference. Jul 11 15:11:50 denix: ah yes, but i was mainly referring to the "membership fees" Jul 11 15:11:55 denix: there has been a long-standing and open-ended invitation for eV members to donate. I guess what has been missing is an explicit invitation/request to corporate users who are not eV members Jul 11 15:12:18 pb_: of-course, everyone is free to define core as they like Jul 11 15:12:31 which I guess is all of them, since (iirc) the eV only admits natural persons as members Jul 11 15:12:57 there were "offers", specifically from new Intel members. did anyone follow up with them? Jul 11 15:13:49 pb_: it might not directly influence quality, but it would be easier to track important changes, ie. changes in core .bbclass files and such Jul 11 15:15:19 esbenh: the number of .bbclass files which are truly "core" is quite small, maybe 3 or 4 at a guess. the majority of the classes are probably less important than, say, recipes like gcc and eglibc, in terms of the impact of changes to them. Jul 11 15:16:21 I personally would think that it is more "core" to most OE devs if base.bbclass is changed than fx. recipes-multimedia/musicbrainz Jul 11 15:16:41 pb_: well, that is again a matter of definition of "core" Jul 11 15:17:08 all bbclass files used by more than a small handfull of recipe could be considered core. Jul 11 16:24:59 khem: have you been bitten by https://bugs.launchpad.net/gcc-linaro/+bug/667490 lately? is this something I can help run to ground? Jul 11 17:21:01 michaedw: havent tried for long since I have that patch punted out of OE gcc Jul 11 17:21:55 michaedw: and I rely on folks with real hardware for such testing and person who helps me is on vacation these days Jul 11 17:22:09 khem: I've got real hardware; can I help? Jul 11 17:22:43 michaedw: do u run oe on it ? Jul 11 17:22:55 khem: I'm trying to find out whether it's still an issue, so I don't trip over it later Jul 11 17:23:03 khem: I can ;-) panda Jul 11 17:23:40 michaedw: ok, I think it needs that we enable this patch in OE and rebuild cairo Jul 11 17:24:08 when Martin is back I can ask for his time Jul 11 17:25:10 khem: maybe I can try it with a Linaro toolchain? Jul 11 17:25:19 khem: I build cross Jul 11 17:26:06 khem: can you point me to a configure/make log? Jul 11 17:35:29 kergoth_: did you see RP__ comments about the environment variables? Jul 11 17:39:46 michaedw: default 4.5 toolchain in oe has all linaro patches so it should give the same trouble Jul 11 17:40:07 additionally if you want to reproduce it on oe-core and you have meta-linaro then thats fine too Jul 11 17:40:14 but it might be too much work Jul 11 17:40:20 oe.org is AWOL Jul 11 17:40:33 ka6sox: anything wrong with apache on melo Jul 11 17:41:16 oh boy melo is hosed too Jul 11 17:49:33 looking Jul 11 17:50:07 khem: we will give it a shot with both toolchains Jul 11 17:50:35 khem: but if you can point me to a log, that would be easiest; I could just build against the sysroot we've got Jul 11 18:08:33 khem ping Jul 11 18:15:57 khem, its up....what are you guys doing...all of a sudden the load goes thru the roof. Jul 11 18:19:35 evening Jul 11 18:20:06 ka6sox: melo strikes back? :) Jul 11 18:20:22 Jay7, its trying! Jul 11 19:09:39 ka6sox: I tried accessing wiki.openembedded.org from my browser Jul 11 19:10:05 ah now it seems ok Jul 11 19:10:43 did melo got rebooted Jul 11 19:10:49 nitink: yes Jul 11 19:11:25 khem: which is best way to override the site macro (like ac_cv_sizeof_long=${ac_cv_sizeof_long=8} Jul 11 19:11:26 ) for a machine Jul 11 19:12:03 nitink: for a machine ? why would you override long Jul 11 19:12:09 that seems wrong premise to me Jul 11 19:12:36 if you explain the problem I might have more inputs Jul 11 19:13:35 this is for x32 Jul 11 19:13:51 I also think it is hacky Jul 11 19:13:53 nitink: x32 is it mix of both Jul 11 19:14:10 but I wanted to get something to test quickly Jul 11 19:14:20 Now I am making a new site file for x32 Jul 11 19:14:42 khem yes Jul 11 19:17:35 ka6sox: ok whats happening to melo these days Jul 11 19:17:45 it gets grumpy often Jul 11 19:17:57 nitink: yes x32 is a new ABI I suppose Jul 11 19:18:03 Yes Jul 11 19:18:04 nitink: I havent delved too much into it Jul 11 19:18:05 khem, dunno...all of a sudden the load goes nutz Jul 11 19:18:17 but its an effort to make x86 access larger address space Jul 11 19:18:23 primarily Jul 11 19:18:24 nitink: Well, provide x32 siteinfo files Jul 11 19:18:30 is my understanding correct Jul 11 19:18:32 * Tartarus is updating oe-core w/ oe.dev's version now Jul 11 19:18:39 which might make multilib site fun easier Jul 11 19:19:28 thanks khem Tartarus, I am going with the siteinfo file already. Jul 11 19:19:56 Tartarus: heh you might introduce bugs Jul 11 19:20:02 subtle one's Jul 11 19:20:15 khem: I'm going to have to give the site files themselves a good read, yes Jul 11 19:20:19 Since oe-core has some fixes Jul 11 19:20:22 and oe.dev has some too Jul 11 19:20:35 Tartarus: yes it should be a merge Jul 11 19:20:40 rather than overwrite Jul 11 19:20:42 But step one is to get the infrastructure bits that allow for common-linux, etc Jul 11 19:20:55 Then I'm gonna try and fix some of the stupid duplication Jul 11 19:20:57 Tartarus: and if possible compare the config.cache Jul 11 19:21:06 before and after Jul 11 19:21:20 A good idea, thanks Jul 11 19:23:47 nitink: x32 abi is not committed to GNU toolchain upstream yet Jul 11 19:23:57 are you using the gcc from branch ? Jul 11 19:24:31 khem: I am using a custom branch Jul 11 19:25:26 ok Jul 11 19:25:40 I would be interested to know how it fares Jul 11 19:25:48 over traditional ABI Jul 11 19:51:02 So, in oe-core, if I mangle meta-toolchain or meta-environment, what're some good sanity test targets? RP__? sgw? Jul 11 19:53:40 Tartarus: meta-toolchain meta-toolchain-gmae Jul 11 19:53:59 Tartarus: and SDKs Jul 11 19:54:07 which sdks? :) Jul 11 19:55:32 Tartarus: task-core-sdk-gmae Jul 11 19:55:33 e.g. Jul 11 19:55:48 Tartarus: meta-toolchain-sdk Jul 11 19:56:22 k, thanks Jul 11 19:56:32 Tartarus: sorry, its now meta-toolchain-gmae Jul 11 19:56:36 * Tartarus waits for busybox to finish before trying core-image-minimal or so Jul 11 19:57:16 Tartarus: core-image-sato, core-image-sato-sdk Jul 11 19:57:37 k Jul 11 19:58:06 Tartarus: if this is for site config changes you probably need to build world Jul 11 19:58:49 oh that's right, world is supposed to be sane :) Jul 11 19:59:04 heh Jul 11 20:00:02 RP__: I am reading scripts/oe-setup-builddir Jul 11 20:00:37 The Yocto Project has extensive documentation about OE including a reference manual Jul 11 20:00:37 which can be found at: http://yoctoproject.org/documentation Jul 11 20:00:37 For more information about OpenEmbedded see their website: http://www.openembedded.org/ **** ENDING LOGGING AT Mon Jul 11 20:00:49 2011 **** BEGIN LOGGING AT Mon Jul 11 20:02:14 2011 Jul 11 20:02:25 Tartarus: world should work Jul 11 20:03:13 alright, so once I get some stuff done locally I'll xfer this over to a bigger box and see about world Jul 11 20:10:53 heh, so bash is a bad example and doesn't load site files Jul 11 20:11:09 but, whee: build-eglibc/work/i586-oe-linux/ncurses-5.9-r0.1/temp/log.do_configure.896:configure: loading site script /home/trini/work/tmp/oe-core-template/oe-core/meta/site/common-linux Jul 11 20:25:50 jo ant Jul 11 20:32:45 hi Jul 11 20:36:29 hi ant Jul 11 20:36:58 hiho pb Jul 11 20:37:28 hi woglinde Jul 11 20:56:00 hi Jul 11 20:56:20 hi gnutoo Jul 11 20:56:24 what last bitbake version works with oe.dev? Jul 11 20:56:50 I don't have navitD.png in the workdir Jul 11 20:57:09 | install: cannot stat `/home/gnutoo/embedded/oe/oetmps/shr/work/armv4t-oe-linux-gnueabi/navit-1_0.2.0+svnr4530-r8.12/navitD.png': No such file or directory Jul 11 20:58:01 distro:shr machine:om-gta02 Jul 11 21:05:13 woglinde, do you have any idea? Jul 11 21:07:24 navitD.png is only at ./image/usr/share/applications/navitD.desktop Jul 11 21:07:38 navitD is used by SHR Jul 11 21:07:48 it's an incon for a .desktop Jul 11 21:08:07 that is used to wrap navit in a freesmartphone command Jul 11 21:08:12 in order to keep display on Jul 11 21:08:15 when using navit Jul 11 21:09:58 no Jul 11 21:10:00 asked jama Jul 11 21:11:34 JaMa is on vacation Jul 11 21:35:53 ahhh space handling Jul 11 21:36:10 bitbake -e navit gives: Jul 11 21:36:22 svn://anonymous@navit.svn.sourceforge.net/svnroot/navit/trunk;module=navit;proto=httpsfile://navitD.png Jul 11 21:40:19 world ain't happy, btw, RP__ :) Jul 11 21:40:39 maybe it's meta-oe tho Jul 11 21:40:44 Tartarus: how unhappy? Jul 11 21:40:56 yeah, meta-oe problem Jul 11 21:41:42 pangomm wants mm-common which isn't in, but it's meta-oe, so, false alarm :) Jul 11 22:57:29 * Tartarus kicks a meta-oe+oe-core world for qemux86 Jul 11 22:58:03 (no changes, outside of some world related fun, will sort out real from silly and submit a few changes) Jul 11 23:19:12 blah, looks like libtelepathy unrelated problem Jul 11 23:27:52 sgw: killer round of patches Jul 12 00:09:12 * Tartarus drops meta-oe, restarts test builds Jul 12 00:16:10 Tartarus, gets lonely over there in the afternoon :) Jul 12 00:44:27 hiya Crofton Jul 12 00:45:10 hey Jul 12 00:45:41 back from the hinterlands? Jul 12 00:46:49 yeah Jul 12 00:46:58 off to DC for a couple of days though Jul 12 00:55:58 you get around :D **** ENDING LOGGING AT Tue Jul 12 02:59:57 2011