**** BEGIN LOGGING AT Wed Sep 01 02:59:57 2010 Sep 01 07:18:18 good morning Sep 01 07:27:14 angstrom+x11-image fails : http://tinderbox.openembedded.net/builds/85412/ Sep 01 07:34:09 Hi all Sep 01 07:34:21 i was converting sip-native recipe to new style staging Sep 01 07:34:53 in this recipe STAGING_PYDIR variable is used Sep 01 07:35:39 which is initialized by ${STAGING_DIR}/lib/python2.4 Sep 01 07:35:59 previously STAGING_BIN_DIR was converted to bindir Sep 01 07:36:38 now with with varaible should I change STAGING_PYDIR or keep it like ${D}${STAGING_PYDIR} Sep 01 07:50:31 eFfeM_work: u there? Sep 01 07:50:53 morning Sep 01 07:51:08 hello hrw Sep 01 07:51:25 morning Sep 01 07:54:12 hrw: hi Sep 01 07:56:05 please correct me for new style staging only those things will go to ipk file for native recipe which are present in tamp/work/i686-linux/package/image/..sysroot/i686-linux Sep 01 07:57:02 if any thing is present in other than i686 directory it will not go into ipk file Sep 01 08:20:13 anybody is it correct to STAGING_PYDIR to ${D}${libdir}/python2.4? Sep 01 08:21:14 coz in new style staging with STAGING_PYDIR it will not go in ipk file as it is not in i686-linux dir for native recipe Sep 01 08:34:00 mckoan: update to latest snapshot Sep 01 08:34:36 mckoan: there was a patch to base.bbclass which could have caused this which was reverted yesterday Sep 01 08:34:52 good morning Sep 01 08:35:01 hi florian Sep 01 08:35:04 good morning Sep 01 08:35:15 hi ericben Sep 01 08:40:33 hmmm now its doing 5880 tasks for native-sdk-image a week back it was ~4200 I wonder someone checked in some long dep chain Sep 01 08:41:24 I guess git bisect time Sep 01 08:41:55 khem: thx, build is in progress Sep 01 08:51:08 03Khem Raj  07org.openembedded.dev * ra64073a585 10openembedded.git/recipes/make/make_3.82.bb: Sep 01 08:51:08 make_3.82.bb: Add new recipe. Sep 01 08:51:08 Signed-off-by: Khem Raj Sep 01 08:51:08 Acked-by: Leon Woestenberg Sep 01 08:51:12 03Khem Raj  07org.openembedded.dev * rf3f80ed068 10openembedded.git/recipes/unionfs/ (unionfs-utils_1.0.11.bb unionfs-utils_1.0.13.bb unionfs.inc): Sep 01 08:51:12 unionfs-utils: Fix the QA error about hashing style. Sep 01 08:51:12 * Use INC_PR Sep 01 08:51:12 Signed-off-by: Khem Raj Sep 01 09:00:11 For a new project I need an ARM system supported by OE, with LCD,eth,usb,case (no open frame) Any suggestion ? Sep 01 09:00:30 mckoan: mini2440? Sep 01 09:00:44 but no.. it does not have case Sep 01 09:01:43 hrw: yes it has Mini2440 Enclosure with 3.5" Display Frame Sep 01 09:01:52 may be a choice Sep 01 09:02:01 right, this would be a cheap one... Sep 01 09:02:13 I'd preper something more handy, PDA-like Sep 01 09:02:48 ethernet is complicated for this gormfactor Sep 01 09:02:55 s/g/f Sep 01 09:04:03 pdalike will be rather ethernet less Sep 01 09:04:16 I'll evaluate Mini2440, thanks Sep 01 09:04:21 mckoan: TI blaze maybe? but expensive like hell Sep 01 09:04:21 any of you are using it? Sep 01 09:04:23 the first zoom had ethernet iirc Sep 01 09:04:39 mckoan: yes Sep 01 09:05:00 florian: which image do you mount on it? Sep 01 09:05:38 mckoan: built all sorts of different stuff for it in the past... Sep 01 09:06:41 mckoan: it is a rather common scenario on this device - upstream compatible kernel and u-boot is available which mekes it easy to handle. Sep 01 09:07:02 its not really fast but lots of connectivity Sep 01 09:07:18 the mini6410 would be nice too but there is no case for it iirc Sep 01 09:07:56 hmm... a gumstix+baseboard+display might fit in some industrial handheld case. Sep 01 09:08:52 hrw: hello, i would like build some graphic library like gtk+,qte........ do you have any idea how can I do that Sep 01 09:09:42 hrw: for powerpc Sep 01 09:09:57 kamel: "bitbake gtk+" Sep 01 09:10:04 another question, do you know any USB-eth adatper running without pain with a kernel 2.6.9 ? Sep 01 09:11:05 maybe is an oxymoron :-D Sep 01 09:11:31 hrw: I did that but I find many problems like : files .so not found Sep 01 09:11:33 2.6.9??? Sep 01 09:11:59 mckoan: stop digging in graveyard... Sep 01 09:12:22 hrw: LOL Sep 01 09:19:09 if device needs 2.6.9 kernel then it is several years old and forgotten by anyone Sep 01 09:21:31 2.6.9 is nearly 6 years old - 19.10.2004 was day of release Sep 01 10:03:08 hi ericben I had several ideas, I'll PM you Sep 01 10:03:35 mckoan: why does it have to be ARM? We are using BluePro by TQ Components which works fine with OE. QT embedded Sep 01 10:14:51 vps: link please Sep 01 10:15:06 Here: http://www.tq-components.com/en/products/industrial-pcs/industrial-pcs-with-power-architecture/ Sep 01 10:15:48 vps: too big and too expensive I suppose Sep 01 10:15:50 I'm building images from the stable/2009 tree, with qte 4.6.2 backported. Patches have been posted but I have newer around Sep 01 10:16:08 OK, I have no idea about prices Sep 01 10:16:32 vps: industrial stuff is always expensive Sep 01 10:16:44 vps: thx anyway Sep 01 10:17:07 My boxes here cost 1M EUR, so I don't care :-D Sep 01 10:18:04 * mckoan -> food Sep 01 10:23:20 can somebody confirm that sqlite3-native package is not building correctly Sep 01 10:24:45 this recipe is inherited from autotools so and sqlite3.inc file has changed Sep 01 10:25:15 its inheritance has been moved to autotools from autotools_stage Sep 01 10:25:53 so it follows the new legacy staging but in the recipe do_install in defined as : Sep 01 10:26:19 i cheked yum-metadata-parser-native package Sep 01 10:26:41 this package is dependent on sqlite3-native Sep 01 10:27:07 it is failing saying cannot find -lsqlite3 Sep 01 10:27:29 so I think sqlite3-native recipe is broken at the moment Sep 01 11:23:45 anybody can confirm my observation? Sep 01 11:24:17 khem: can you confirm that sqlite-native is being built correctly? Sep 01 12:12:36 Noor: it "build" here but I'm not sure it builds anything ! Sep 01 12:18:25 did u bild yum-metadata-parser-native Sep 01 12:18:43 I did bitbake sqlite-native Sep 01 12:18:45 sqlite3-native builds without any problem Sep 01 12:19:02 do bitbake yum-metadata-parser-native Sep 01 12:20:17 what i think is sqlit3-native is not staging all the stuff Sep 01 12:21:01 sqlite-native builds but I see nothing in staging-pkg nor image nor sysroot-destdir Sep 01 12:26:09 I am note sure about sqlite-native Sep 01 12:27:33 same for sqlite3-native Sep 01 12:29:30 yeah Sep 01 12:30:00 I think that is due to update on sqlite3.inc file Sep 01 12:30:33 when inheritance changed to autotools from autotools_staging Sep 01 12:31:52 this is recipe is in between new staging and old staging Sep 01 12:33:34 coz AUTOTOOLS_NATIVE_STAGE_INSTALL = '1" and do_stage is do_stage_native which allows to do new style staging Sep 01 12:33:54 but in recipe do_install is defines as ":" Sep 01 12:34:39 yeah thats the problem **** BEGIN LOGGING AT Wed Sep 01 12:36:30 2010 Sep 01 12:45:21 hmmm Sep 01 12:45:23 git: error while loading shared libraries: libcrypto.so.1.0.0: cannot open shared object file: No such file or directory Sep 01 12:45:25 just to tell Sep 01 12:45:31 I'll rebuild git-native Sep 01 12:45:42 if someone could bump PR Sep 01 12:45:51 I have built an image for an avr32 machine (atngw100) - bitbake without errors. But if I try to boot the kernel, I get this error: http://nopaste.info/d2968cd774.html -- Has anyone an idea what is wrong? Sep 01 12:47:00 anyway I use autorev Sep 01 12:47:12 so git-native can't be built without -b Sep 01 12:47:21 because it uses git in the parsing Sep 01 12:47:32 and parsing fails because of git Sep 01 12:47:42 git: error while loading shared libraries: libcrypto.so.1.0.0: Sep 01 12:47:50 and so it won't be able to upgrade git Sep 01 12:48:57 with PR bump it seem to work Sep 01 12:49:08 if someone could bump it, else it will go in my next push Sep 01 12:49:21 that is to say not now Sep 01 12:49:28 maybe in some days Sep 01 13:19:06 Radiao: it seems there is a problem in a driver of this kernel, can't you try a more recent kernel ? Sep 01 13:20:42 ericben: I haven't set a specific kernel version anywhere. Just Machine=atngw100 and Distro=angstrom-2010.x --- Where would I set to use a more recent kernel? Sep 01 13:21:23 preffered kernel version Sep 01 13:21:55 Radioa: I don't have experience on this platform but I think aavr are supported by mainline kernel so maybe you can try 2.6.35 or git Sep 01 13:24:24 Hello people Sep 01 13:24:24 I'm trying to assess whether a device can run Linux instead of Windows CE (currently) Sep 01 13:24:27 I have the device, I tried the HaRET (it runs, and console connects). Sep 01 13:24:29 Now what? Please advise. Sep 01 13:24:39 Can i set this in local.conf? I was told not to tamper with files in OE/conf/distro/* Sep 01 13:25:04 Device is: Mobile Compia (M3 Mobile) MC6500, ARM920T PXA27x, Windows CE 5.00 Sep 01 13:27:22 good morning Sep 01 13:28:02 any oestats "experts" here? Sep 01 13:28:23 I'm having some trouble setting up a server Sep 01 13:28:41 Lachezar, now it depends on what you want to do next Sep 01 13:29:10 I would like to boot a kernel (using HaRET) as first step. Sep 01 13:29:27 ahhhh Sep 01 13:29:29 And just to know, that it is POSSIBLE. Sep 01 13:29:32 I have the device, I tried the HaRET (it runs, and console connects). means haret console Sep 01 13:29:51 Yep Sep 01 13:29:54 I thought it meant console in angstrom Sep 01 13:30:07 Lachezar: you have to customize a kernel for your device, a hard work Sep 01 13:30:14 Nope... HaRET console on my laptop connects to the device, that runs HaRET Sep 01 13:30:54 Yeah, I know, but before I throw myself in the deep, I thought a minimalistic generic ARM920T kernel would make a nice first step. Sep 01 13:31:04 Lachezar, get a kernel,compile it , setup haret to boot the kernel,then watch the kernel panic,then put a rootfs on sd, put root=/dev/mmcblk0p1 or something like this and boot a console image Sep 01 13:31:22 Anything half-brain generic? Sep 01 13:31:45 Lachezar: do you have any datasheet or block schematics of your device? Sep 01 13:31:52 Cross-compiling the kernel is a distant 3-rd or 4-th step (after understaning what hardware drivers I need to put) Sep 01 13:32:36 Ahmmm... mckoan: nope... I could try finding one. Sep 01 13:32:47 Lachezar: good luck! Sep 01 13:33:29 Lachezar, before compiling the kernel Sep 01 13:33:39 Lachezar,you need to find the right source tree Sep 01 13:34:08 the one with support for your hardware Sep 01 13:34:09 GNUtoo: You mean the correclty patched tree? Sep 01 13:34:13 indeed Sep 01 13:34:28 GNUToo: Yeah, but I FIRST need to find what hardware it is :( Sep 01 13:34:29 else you would have to do the patches yourself Sep 01 13:34:35 ah? Sep 01 13:34:44 it's not already supported somewhere? Sep 01 13:35:01 Nada. WindowsCE from the source and nothing further. Sep 01 13:35:14 source -> provider Sep 01 13:35:19 ok Sep 01 13:35:33 did you look on internet for infos on your device? Sep 01 13:35:41 specially for some community port Sep 01 13:35:48 of the linux kernel Sep 01 13:35:51 It;s an EDA with limited market, not for the general public. Sep 01 13:35:58 ok Sep 01 13:36:04 then look at : Sep 01 13:36:06 *the specs Sep 01 13:36:10 So there is NO information apart from the manufacturer's site. Sep 01 13:36:13 *similar hardware that is supported Sep 01 13:36:21 maybe the specs can be found with: Sep 01 13:36:23 *windows ce Sep 01 13:36:26 *haret Sep 01 13:36:31 I'm not a haret specialist Sep 01 13:36:38 people in #htc-linux are tough Sep 01 13:36:47 Ahha... So I need to find a HaRET guru? :) Sep 01 13:37:02 HaRET ran, and I got a spark of hope... Sep 01 13:37:20 basically most people there(but me and some other people) are doing wince->android ports Sep 01 13:37:25 But then it said: put kernel in, and I paniced Sep 01 13:37:32 personally I do android->SHR Sep 01 13:37:34 Radiao: you will need to create conf file for a recent kernel and then set DEFAULT_PREFERENCE_atngw100 = "1" in the recipe of this kernel Sep 01 13:37:43 Oh. Android is my target OS eventualy! Sep 01 13:38:25 Lachezar, we don't do andorid here, only plain openembedded based OS, altough there is an android NDK that can be built with oe Sep 01 13:38:59 Well... I thought to take it one step at a time... Win-Ce -> Linux -> Android. Sep 01 13:39:05 ok Sep 01 13:39:11 that's wiser Sep 01 13:39:16 ?? shr Sep 01 13:39:27 an angstrom derived distro for phones Sep 01 13:39:29 anyway Sep 01 13:39:35 Ahha... Sep 01 13:39:54 go in #htc-linux and ask questions about haret+wince on how to get your machine specs Sep 01 13:40:24 10x. Going there now. Sep 01 13:46:33 hi! I need help in understanding some kernel segmentation fault. How do I read the following line (or is there a better channel for this?): Sep 01 13:46:37 [ 225.514801] Backtrace: [ 225.517272] [] (wait_for_common+0x0/0x1a4) from [] (wait_for_completion+0x18/0x1c) Sep 01 13:46:58 what do +0x0/0x1a4 and +0x18/0x1c mean? Sep 01 14:09:25 hi Ken Sep 01 14:33:29 hmmm. Sep 01 14:58:42 damnit Sep 01 14:58:58 i hate when i send a patch to the list and *then* notice the typo in the commit message Sep 01 15:09:34 uh, google, this email from zecke is not Haitian, it's english Sep 01 15:09:37 * kergoth_ smacks gmail Sep 01 15:16:08 anyone building oe for palmtt3? git://git.hackndev.com has been down for some time i guess? Sep 01 15:16:17 or perhaps all has been in vanilla 2.6 kernel and the config will just work? Sep 01 15:25:22 Hi Sep 01 16:58:31 sakoman: are you still working on your oestat server ? Sep 01 16:58:54 ericben: I would like to, but am kind of stuck :-) Sep 01 16:59:13 sakoman: I've setup one and it seems to work fine Sep 01 16:59:33 did you do the "development" or "production" setup? Sep 01 16:59:37 just following the instructions and launching python manage.py runserver IP:PORT Sep 01 16:59:42 dev at the moment Sep 01 16:59:58 for production I have to check as I'm not using apache but lighttpd Sep 01 17:00:06 yeah, I've used that in the past with no issues. I was trying to do the "production" method Sep 01 17:00:26 I'll keep you informed, I plan to do this this evening Sep 01 17:00:40 it "sort of" works: http://build.sakoman.com/ Sep 01 17:01:13 what is not working ? Sep 01 17:01:57 at first glance you might think that the media directory isn't working, but I can access stuff there directly: Sep 01 17:02:00 http://build.sakoman.com/media/css/dashboard.css Sep 01 17:02:28 ericben: page layout is wrong and build machines won't connect Sep 01 17:02:33 on the dev platform, I have the same layout Sep 01 17:02:41 byt the build machines do connect Sep 01 17:02:50 and I get the waterfall OK Sep 01 17:03:26 On my browser I just get an all text rendering Sep 01 17:03:41 like the css is not being honored Sep 01 17:04:23 ok, I'll keep you infromed if that works Sep 01 17:23:45 gm Sep 01 17:34:38 hi all Sep 01 17:44:16 are there any specific python packages that need to be installed for bitbake / oe to work? Sep 01 17:54:42 ron, not that I am aware of (you could install psyco though), there is a page on the wiki stating what you need Sep 01 17:55:02 you need the python sqlite bindings at a minimum, its used for the persist_data module in bitbake Sep 01 17:59:40 ah ok, guess that was already installed for me Sep 01 18:04:48 hi likewise Sep 01 18:08:39 kergoth_: is there an op in bitbake to say stop after preparing runqueue ? Sep 01 18:10:46 03Graham Gower  07org.openembedded.dev * r2f421560ed 10openembedded.git/recipes/gstreamer/gst-plugins-bad_0.10.19.bb: Sep 01 18:10:46 gst-plugins-bad_0.10.19.bb: add libmodplug to DEPENDS. Sep 01 18:10:46 Several images want gst-plugin-modplug (e.g. illume-image and the Sep 01 18:10:46 fso images). They fail to find it without this. Sep 01 18:10:46 Signed-off-by: Graham Gower Sep 01 18:10:46 Signed-off-by: Koen Kooi Sep 01 18:18:21 <[Rui]> Hi all, I'm wondering if someone could bump omnewrotate in oe tree, please? Sep 01 18:18:39 khem, not exactly. you could hack it in with a class -- do a bb.fatal() at BuildStarted event - i *think* thats after the runqueue but before it runs any tasks Sep 01 18:19:54 <[Rui]> Heinervdm: thanks! the diff is here: http://pastebin.com/cmyDecwc (valid for one hour) Sep 01 18:23:26 kergoth_: ah good idea Sep 01 18:23:40 kergoth_: I am doing a git bisect so need it temporarily Sep 01 18:26:34 kergoth_: when bisecting if the bisected tree turns out to be bad then I should do git bisect bad right to move to next step Sep 01 18:26:41 I always confuse it Sep 01 18:36:30 would be a nice enhancement: bitbake -dump_tasks or so Sep 01 18:40:22 hm, anyone an idea why I get this ? Sep 01 18:40:23 ERROR: '[]' RDEPENDS/RRECOMMENDS or otherwise requires the runtime entity 'virtual/arm-oe-linux-gnueabi-gcc-2.95' but it wasn't found in any PACKAGE or RPROVIDES variables Sep 01 18:40:34 and some of this Sep 01 18:40:36 Unknown Event: Sep 01 18:40:46 don't think I changed anything that adds this Sep 01 18:42:47 pastebin: http://pastebin.ca/1930684 machine sheevaplug distro angstrom Sep 01 18:42:52 eFfeM: bitbake -g ? Sep 01 18:45:15 khem, yeah actually as can be seen in the pastebin it already says that virtual/arm-oe-linux-gnueabi-depmod- is not buildable Sep 01 18:45:35 ASSUME_PROVIDED += "virtual/arm-linux-gcc-2.95" Sep 01 18:46:05 virtual/arm-oe-linux-gnueabi-gcc-2.95 Sep 01 18:46:08 actually Sep 01 18:46:17 yeah, but why would anyone want to depend on 2.95 ?????? And this used to work a month ago or so Sep 01 18:47:53 thats what bitbake -g would tell Sep 01 18:50:36 I'm checking it atm, also working on something else Sep 01 18:50:49 btw any idea when you are going to push your patches ? Sep 01 18:54:02 khem, oh, pretty sure -g dumps based on the runqueue, could try that as a test Sep 01 19:06:31 03Frans Meulenbroeks  07org.openembedded.dev * rb1f7b5ac41 10openembedded.git/recipes/iscsi-target/ (9 files in 3 dirs): Sep 01 19:06:31 iscsi-target: moved to 1.4.20.2 Sep 01 19:06:31 also renamed the files dir to the more appropriate iscsi-target Sep 01 19:06:31 added a DEFAULT_PREFERENCE = "-1" for beagleboard as Sep 01 19:06:31 angstrom and minimal use a modified 2.6.32 kernel which breaks this recipe Sep 01 19:06:32 Signed-off-by: Frans Meulenbroeks Sep 01 19:06:39 03Frans Meulenbroeks  07org.openembedded.dev * r18cbd67a6b 10openembedded.git/MAINTAINERS: Sep 01 19:06:39 MAINTAINERS: updated my entry Sep 01 19:06:39 Signed-off-by: Frans Meulenbroeks Sep 01 19:13:08 khem, bitbake -g does not help me, the odd thing is that it says ERROR '[]' RDEPENDS Sep 01 19:13:37 no mentioning of 2.95 in bitbake -g, will try -D -D -v -v or so Sep 01 19:15:58 03Frans Meulenbroeks  07org.openembedded.dev * r4bbb807bf5 10openembedded.git/recipes/tgt/tgt_1.0.7.bb: Sep 01 19:15:58 tgt: moved to version 1.0.8 Sep 01 19:15:58 Signed-off-by: Frans Meulenbroeks Sep 01 19:31:50 hm, problem is still there (btw was machine sheevaplug, distro minimal, not angstrom), bitbake 1.10 Sep 01 19:32:13 bitbake -D -D -v -v gives lines like: Sep 01 19:32:14 DEBUG: Added dependency virtual/arm-oe-linux-gnueabi-gcc-2.95 for /home/frans/oe/openembedded/recipes/linux/poodle-kernel-2.4-embedix.bb Sep 01 19:33:00 no idea yet why I get that 2.4 stuff, the build continues after the notice, and nicely builds the preferred linux-kirkwood Sep 01 20:18:14 eFfeM: KERNEL_CCSUFFIX = "-2.95" Sep 01 20:18:18 thats the one I guess Sep 01 20:19:49 yep see classes/kernel.bbclass Sep 01 20:19:52 DEPENDS += "virtual/${TARGET_PREFIX}gcc virtual/${TARGET_PREFIX}depmod-${@get_kernelmajorversion('${PV}')} virtual/${TARGET_PREFIX}gcc${KERNEL_CCSUFFIX} update-modules" Sep 01 20:22:45 khem, well the issue is that if I build minimal for beagleboard I do not get this problem Sep 01 20:23:12 and the beagleboard bitbake -v -v -D -D log says DEBUG: Skipping /home/frans/oe/openembedded/recipes/linux/shepherd-kernel-2.4-embedix.bb Sep 01 20:23:19 which is not in the sheeva log Sep 01 20:24:33 sheeva includes kirkwood.inc which says; PREFERRED_PROVIDER_virtual/kernel = "linux-kirkwood" Sep 01 20:26:38 hm, mihgt have found something' Sep 01 20:29:35 eFfeM: prolly its in ANGSTROM_BLACKLIST Sep 01 20:30:14 khem, no the beagle build was also using minimal (and i checked blacklisting only distro that blacklists these is shr) Sep 01 20:30:37 thought I had to add ( ) around the list of compatible machines, but that did not help Sep 01 20:44:49 re Sep 01 20:46:38 hi florian Sep 01 20:48:56 03Khem Raj  07org.openembedded.dev * r13b2674e39 10openembedded.git/recipes/vlc/vlc-davinci_0.8.6h.bb: Sep 01 20:48:56 vlc-davinci_0.8.6h.bb: Rename codec-engine to ti-codec-engine Sep 01 20:48:56 Signed-off-by: Khem Raj Sep 01 20:48:56 Acked-by: Frans Meulenbroeks Sep 01 20:48:58 03Khem Raj  07org.openembedded.dev * rebbb55cb3d 10openembedded.git/recipes/clutter/ (clutter-gtk-0.6_git.bb clutter_0.6svn.bb): Sep 01 20:48:58 clutter_0.6svn.bb,clutter-gtk-0.6_git.bb: Unused and unbuildable for long hence removed. Sep 01 20:48:58 Signed-off-by: Khem Raj Sep 01 20:48:58 Acked-by: Frans Meulenbroeks Sep 01 20:48:59 03Khem Raj  07org.openembedded.dev * r257041edcb 10openembedded.git/recipes/minisip/libmsip_svn.bb: Sep 01 20:48:59 libmsip_svn.bb: Rename libmnetutil0->libmnetutil in DEPENDS Sep 01 20:49:00 Signed-off-by: Khem Raj Sep 01 20:49:00 Acked-by: Frans Meulenbroeks 03Khem Raj  07org.openembedded.dev * r235a6f5fcc 10openembedded.git/recipes/gpephone/libmsgenabler2_svn.bb: Sep 01 20:49:01 libmsgenabler2_svn.bb: Replace gsmd-lips with gsmd in DEPENDS Sep 01 20:49:02 Signed-off-by: Khem Raj Sep 01 20:49:02 Acked-by: Frans Meulenbroeks Sep 01 20:49:52 bt950-cs_0.1.bb: Dont override PACKAGES but append to it. Sep 01 20:49:53 Signed-off-by: Khem Raj Sep 01 20:49:53 Acked-by: Frans Meulenbroeks Sep 01 20:49:53 03Khem Raj  07org.openembedded.dev * r87350d079f 10openembedded.git/recipes/ (27 files in 12 dirs): Sep 01 20:49:53 recipes: Move some recipes to nonworking for getting bitbake world. Sep 01 20:49:53 * If a recipes has DEPENDS on another recipe which is nonworking Sep 01 20:49:54 then move the recipe in question to nonworking too. Sep 01 20:49:54 Signed-off-by: Khem Raj Sep 01 20:49:55 Acked-by: Frans Meulenbroeks Sep 01 20:53:54 khem, thanks, curious what the next build gives Sep 01 20:54:23 btw it is also strange that we still have those 2.4 recipes that require 2.95 but we have no 2.95 Sep 01 20:54:34 guess that should put them to nonworking either Sep 01 20:59:15 found my problem, SOC_FAMILY patch is broken I'd say Sep 01 21:01:48 eFfeM: well some people might have external toolchains providing 2.95 Sep 01 21:01:50 so Sep 01 21:02:04 its safe to assume it provided Sep 01 21:02:07 and get on Sep 01 21:02:39 03Khem Raj  07org.openembedded.dev * r251f9f2362 10openembedded.git/recipes/openmoko-base/openmoko-session_svn.bb: Sep 01 21:02:39 openmoko-session_svn.bb: Replace openmoko-today with openmoko-today2 Sep 01 21:02:39 Signed-off-by: Khem Raj Sep 01 21:02:39 Acked-by: Frans Meulenbroeks Sep 01 21:02:41 03Khem Raj  07org.openembedded.dev * r726082f805 10openembedded.git/recipes/networkmanager/cnetworkmanager_git.bb: Sep 01 21:02:41 cnetworkmanager_git.bb: python-uuid is not provided remove it from RDEPENDS Sep 01 21:02:41 Signed-off-by: Khem Raj Sep 01 21:02:41 Acked-by: Frans Meulenbroeks Sep 01 21:04:28 khem found the problem, actually most machines will have this error for non-angstrom distro Sep 01 21:04:34 testing patch Sep 01 21:05:26 ti made a mess of it Sep 01 21:05:47 ok good Sep 01 21:09:59 heh, I suspect they are focused on having it work for a limited set of distro/machine pairs :) Sep 01 21:10:05 03Khem Raj  07org.openembedded.dev * rcd62d504db 10openembedded.git/recipes/sugar/sugar-fructose.bb: Sep 01 21:10:05 sugar-fructose.bb: sugar-web-activity is not provided by any recipe dont DEPEND on it. Sep 01 21:10:05 Signed-off-by: Khem Raj Sep 01 21:10:05 Acked-by: Frans Meulenbroeks Sep 01 21:10:07 03Khem Raj  07org.openembedded.dev * r460eb9a6ca 10openembedded.git/recipes/images/ (2 files): Sep 01 21:10:07 initramfs-bootmenu-debug-image.bb,initramfs-bootmenu-image.bb: Rename klibc-utils-static-modprobe->klibc-static-utils-modprobe Sep 01 21:10:07 Signed-off-by: Khem Raj Sep 01 21:10:07 Acked-by: Frans Meulenbroeks Sep 01 21:10:07 03Khem Raj  07org.openembedded.dev * rb3e68674ef 10openembedded.git/recipes/initrdscripts/ (2 files): Sep 01 21:10:08 initramfs-module-bootmenu_1.0.bb,initramfs-module-kexecboot_1.0.bb: Rename klibc-utils-static-fstype -> klibc-static-utils-fstype Sep 01 21:10:08 Signed-off-by: Khem Raj Sep 01 21:10:08 Acked-by: Frans Meulenbroeks Sep 01 21:18:54 Crofton yeah; really good work, introducing a patch that breaks virtually all distro/machine pairs expect angstrom and ti soc's. And that person whined for weeks that I broke his build with openssl Sep 01 21:18:55 bah Sep 01 21:19:11 anyway, mail sent to list with 2 possible fixes Sep 01 21:20:36 03Khem Raj  07org.openembedded.dev * r8b82ba3b25 10openembedded.git/recipes/angstrom/angstrom-bootmanager.bb: Sep 01 21:20:36 angstrom/angstrom-bootmanager.bb: Remove Sep 01 21:20:36 Signed-off-by: Khem Raj Sep 01 21:24:31 03Khem Raj  07org.openembedded.dev * rc31313b175 10openembedded.git/recipes/tasks/task-arago-toolchain-target.bb: Sep 01 21:24:31 task-arago-toolchain-target.bb: remove libpng12-dev which is not provided and e2fsprogs-libs-dev is now e2fsprogs-dev Sep 01 21:24:31 Signed-off-by: Khem Raj Sep 01 21:24:31 Acked-by: Frans Meulenbroeks Sep 01 21:24:31 Acked-by: Denys Dmytriyenko Sep 01 21:25:44 khem, if lets say in 2hrs I wanted to start reworking garnet would that work for you? Sep 01 21:26:27 calling it a day, cya Sep 01 21:27:07 03Khem Raj  07org.openembedded.dev * r500a032003 10openembedded.git/recipes/tasks/task-demo-x11.bb: Sep 01 21:27:08 task-demo-x11.bb: Replace non existing libgles-omap3-demos with libgles-omap3-x11demos Sep 01 21:27:08 Signed-off-by: Khem Raj Sep 01 21:27:08 Acked-by: Frans Meulenbroeks Sep 01 21:28:10 ka6sox-work: hmmm lemme see Sep 01 21:28:48 before I sent an email to the ML I wanted to check first Sep 01 21:29:07 ka6sox-work: sure that should be fine Sep 01 21:29:23 I have a job running on that one which should be done in 2 hrs Sep 01 21:29:37 okay just ping me here when its done... Sep 01 21:29:43 I can wait...np Sep 01 21:30:47 03Khem Raj  07org.openembedded.dev * rf7e048c9c7 10openembedded.git/recipes/tasks/task-gmae.bb: Sep 01 21:30:47 task-gmae.bb: it uses XSERVER but doesnt really define it so define it. Sep 01 21:30:47 Signed-off-by: Khem Raj Sep 01 21:30:49 03Khem Raj  07org.openembedded.dev * r932d107f7c 10openembedded.git/recipes/meta/ (meta-angstrom-2007.bb meta-opie.bb): Sep 01 21:30:49 meta-angstrom-2007.bb,meta-opie.bb: task-opie-bluetooth is added only if it appeas in COMBINED_FEATURES Sep 01 21:30:49 Signed-off-by: Khem Raj Sep 01 21:30:49 Acked-by: Frans Meulenbroeks Sep 01 21:30:52 03Khem Raj  07org.openembedded.dev * re56e07be95 10openembedded.git/recipes/tasks/task-gpe.bb: Sep 01 21:30:52 task-gpe.bb: virtual/xserver can not be RDEPEND'ed on it can be DEPEND'ed upon Sep 01 21:30:52 Signed-off-by: Khem Raj Sep 01 21:30:52 Acked-by: Frans Meulenbroeks Sep 01 21:35:41 03Khem Raj  07org.openembedded.dev * r3a182c0f32 10openembedded.git/recipes/meta/meta-toolchain-arago-dvsdk.bb: Sep 01 21:35:42 meta-toolchain-arago-dvsdk.bb: Remove Sep 01 21:35:42 * As suggested by Denis on mailing list. Sep 01 21:35:42 Signed-off-by: Khem Raj Sep 01 21:58:45 03Khem Raj  07org.openembedded.dev * rb74fea69ca 10openembedded.git/recipes/meta/meta-toolchain-arago-gst.bb: Sep 01 21:58:45 meta-toolchain-arago-gst.bb: Remove Sep 01 21:58:45 * Per Denys's suggestion. Sep 01 21:58:45 Signed-off-by: Khem Raj Sep 01 22:00:28 good night all Sep 01 22:23:14 ka6sox-work: my work on garnet finished so its all yours Sep 01 22:23:28 khem, thanks Sep 01 22:23:29 ka6sox-work: send me a note when its redone and is up again Sep 01 22:23:38 I'll send a message to the group Sep 01 22:23:44 ok Sep 01 23:09:46 i wish people would stop announcing new android tablets, and instead someone actually shipped one Sep 01 23:12:10 kinda wish I didn't miss the one that was in K-Mart, heh Sep 01 23:12:22 and i should see if that "rugged" one ever shipped Sep 01 23:13:04 i want one, and the ipad is getting tempting again just due to the fact that it actually exists Sep 01 23:13:06 :) Sep 01 23:14:08 03Tom Rini  07org.openembedded.dev * r50320c1452 10openembedded.git/contrib/oe-stylize.py: Sep 01 23:14:08 oe-stylize.py: Move exports up to before normal functions Sep 01 23:14:08 As Mickey pointed out, these modify all functions so it helps Sep 01 23:14:08 reading comprehension if they are listed before them. Sep 01 23:14:08 Signed-off-by: Tom Rini Sep 01 23:16:19 kergoth, there are some chinese ipad knock offs being sold cheaply on ebay Sep 01 23:16:31 look for iped, epad, etc. Sep 01 23:16:39 android based i believe Sep 01 23:35:49 probably doesn't want one made out of lead and cadmium tho :) Sep 01 23:55:15 kergoth: ipad is nice but it cost around 1000 usd delivered Sep 01 23:56:14 that price given its lack of expandability is a bit much, indeed Sep 01 23:56:23 kergoth: I thinks I like the fact that it has roll over 3G Sep 01 23:56:29 so I pay for what I use Sep 01 23:56:46 and not like iphone where I trade my balls to get one for next 2 years Sep 02 00:00:20 kergoth: are you familiar with debina renaming area in OE Sep 02 00:00:28 kergoth: we have a problem with that Sep 02 00:00:38 the deps are not marked correctly in resulting ipks Sep 02 00:01:04 and also it does not rename -dev -doc -dbg packages which is confusing Sep 02 00:01:36 it renames individual packages, the -doc/-dev are recipe wide, not per binary package, so it doesn't make any sense to rename them to match Sep 02 00:02:24 not sure what you mean about the deps, it adjusts them to match, see package_runtime_rename or whatever Sep 02 00:03:00 hrm Sep 02 00:08:46 * kergoth re-reads debian.bbclass Sep 02 00:29:19 e-image now builds for me. w00t! Sep 02 00:31:09 is there a clear way to know if an image is supposed to be distro specific or distro agnostic? (other than a distro prefix, obviously) Sep 02 00:31:15 or board specific Sep 02 00:55:56 well, in general they should be distro/machine agnostic :) Sep 02 00:56:14 but certainly some are more specific than others Sep 02 00:56:39 for example the beagle demo images may depend on stuff only supported by that hw, such as the dsp's Sep 02 00:57:49 btw, this signal is generated by gnuradio running on Angstrom built with OE Sep 02 00:57:50 http://twitpic.com/2kbxth Sep 02 00:58:34 kergoth, khem is this the -dev for libstdc++ not pulling in libstdc++ issue? Sep 02 01:01:20 probably, heh Sep 02 01:07:56 khem, about 30 minutes longer...you needing it right away? Sep 02 01:12:02 grg: people have been adding image recipes which are specific to distros sometimes boards Sep 02 01:13:02 grg: in general it would be desirable to have all images build for all machines distro combos but that may not be practical Sep 02 01:13:23 we might have to have few generic image targets which build for all distros Sep 02 01:13:39 and then few which are distro specific and machine specific Sep 02 01:14:30 {{-}{-}} would be something good Sep 02 01:15:30 yeah, it just needs to be made more explicit that it is only expected to build with a particular distro and/or machine combo Sep 02 01:15:50 heh Sep 02 01:15:58 compatible MACHINE/DISTRO Sep 02 01:16:13 grg, at the end of the day, people need to do work Sep 02 01:16:13 I've been doing some building yesterday and today and have gotten the following images built for qemumipsel (with some minor fixes along the way): x11-image, native-sdk-image, nas-server-image, imote2-image, illume-image, e-image-core, e-image Sep 02 01:16:23 Crofton: yes thats the issue Sep 02 01:16:35 Crofton: and same issue for libgcc too Sep 02 01:16:39 cool Sep 02 01:16:46 people are seeing missing provider for libgcc_s.so Sep 02 01:16:49 its same problem Sep 02 01:17:30 bitbake world has 83000 tasks :) Sep 02 01:17:37 :) Sep 02 01:17:37 well we are making progress :) Sep 02 01:17:49 anything we understand can be resolved Sep 02 01:17:54 down to almost 7000 recipes though. Sep 02 01:18:09 and I figured out why my mothers cell phone died Sep 02 01:18:19 the charging cable is nicke Sep 02 01:18:29 nicked, well, almost cut in half Sep 02 01:19:31 I am building bitbake world its been churning bits for past 4 hours Sep 02 01:19:40 lets see when and where it dies Sep 02 01:26:37 Crofton: nice to gnuradio is kicking Sep 02 01:26:46 nice to know I meant Sep 02 01:29:17 grg: cool that you got so many of them building Sep 02 01:30:21 khem, well they *should* build. They're in the tree. Sep 02 01:30:36 opie-image builds too Sep 02 01:31:00 of these that i've tried to boot on real hardware, none of them start x successfully Sep 02 01:31:01 grg: cool Sep 02 01:31:34 grg: buildablily and runnability are two different beasts Sep 02 01:32:57 grg: getting more dead meat out of OE will make it better Sep 02 01:33:08 yeah absolutely Sep 02 01:33:19 and my dream is that one day we will be able to move to time based release Sep 02 01:33:31 like once a year or twice a year predictably Sep 02 01:34:08 that wouldn't be too difficult, just have the TSC say that only fixes can be committed to the tree for e.g. a two week window before the release Sep 02 01:34:34 yeah we need some release policy Sep 02 01:34:52 that will raise the OE brand Sep 02 01:36:50 i think our current focus in regard to buildability is a good start Sep 02 01:37:12 yes I believe so Sep 02 01:37:20 many new users come along and have failing builds for the first thing they try. Which is not a good first impression Sep 02 01:37:33 * khem nods Sep 02 01:38:15 if its in the tree and it doesn't build, there better be a damn good reason why not or a fix pending on the mailing list Sep 02 01:38:29 people have been paying attention to parts they are interested and sometimes its poisons the other parts and we dont know Sep 02 01:38:31 just be careful, because one mans dead meat is anothers steak Sep 02 01:38:53 Crofton_|work: that steak has to be cookable Sep 02 01:38:56 I have a couple of long term projects in the tree Sep 02 01:38:57 and eatable Sep 02 01:39:14 * ka6sox-work thanks eFeneM for all his work in resurecting freenx Sep 02 01:39:16 Crofton_|work, there are plenty of festering carcasses Sep 02 01:39:26 we don't want to drive everyone into their own repos Sep 02 01:39:26 Crofton_|work: the thing is if we need to chug along then they need to evolve with OE Sep 02 01:39:45 so if you maintain them healthy thats good Sep 02 01:40:03 Crofton_|work, if something is broken because its a work in progress, thats ok, as long as its marked as such Sep 02 01:40:35 Crofton_|work: thats true, thats why we are making sure that none of used stuff is not deleted Sep 02 01:40:59 you've noticed a few of the .de regulars haven't been seen in a few weeks? Sep 02 01:41:32 it will interesting to see what happens after people get back from summer holidays Sep 02 01:42:35 summer? what a crazy hemisphere! Sep 02 01:42:42 rofl Sep 02 01:42:53 grg, you in the southern hemisphere? Sep 02 01:42:58 australia Sep 02 01:43:01 cool Sep 02 01:43:16 my mother was showing me some candles my cousin had melt outside Sep 02 01:43:20 he is in Adelaide Sep 02 01:43:35 i am in adelaide also Sep 02 01:43:43 cool Sep 02 01:43:49 My gransmother was from there Sep 02 01:44:40 a couple of years ago we had 7 days over 40C in a row, 46C one day. Candles will melt. Sep 02 01:49:54 Crofton_|work: I think .eu regulars will find things better thats my hope :) Sep 02 01:51:25 I've got some patches in patchwork that appear to have been forgotten (some even with acks)... any chance someone could take a look? Sep 02 01:51:34 grg: sure Sep 02 01:51:36 my succeeding builds probably wont succeed without them Sep 02 01:51:39 well having to add libstdc++ to all images using c++ will be annoying :) Sep 02 01:51:41 khem, thanks :) Sep 02 01:51:43 give me patchwork ids Sep 02 01:52:06 Crofton_|work: if I am lucky I will figure out a fix in classes Sep 02 01:52:14 yeah, thanks for looking Sep 02 01:52:40 I got two tones to come out of my board and am going out of town until Tuesday :) Sep 02 01:53:17 khem, 2636, 2303, 2364 Sep 02 01:53:28 there are some others, but they don't unbreak things Sep 02 01:54:26 hmm... gpsd has had new recipes added, so that patch may need updating Sep 02 01:54:59 2636 had questions using inherit gettext Sep 02 01:55:43 khem, yes. it still fails with inherit gettext however Sep 02 01:56:15 the gettext m4 macro is required to even autoconf Sep 02 01:56:32 grg: ok then the patch in there is fine I guess Sep 02 01:58:59 let me build pulseaudio now Sep 02 01:59:00 khem, 2615 and 2616 will make it possible to build a system without opkg installed. Sep 02 01:59:50 khem, and 2349 is the gcc -ibad patch... not sure how much interest there is in that Sep 02 01:59:53 grg: you need to redo 2364 Sep 02 02:00:19 khem, sure. Sep 02 02:03:20 03Graham Gower  07org.openembedded.dev * rf43e388d81 10openembedded.git/recipes/gnome/gnome-doc-utils.inc: (log message trimmed) Sep 02 02:03:20 gnome-doc-utils: Add dependency upon gettext-native. Sep 02 02:03:20 Failing log.do_configure: Sep 02 02:03:20 automake (GNU automake) 1.11.1 Sep 02 02:03:20 Copyright (C) 2009 Free Software Foundation, Inc. Sep 02 02:03:21 License GPLv2+: GNU GPL version 2 or later Sep 02 02:03:22 This is free software: you are free to change and redistribute it. Sep 02 02:03:25 03Khem Raj  07org.openembedded.dev * rbfb1161244 10openembedded.git/conf/distro/include/ (2 files): Sep 02 02:03:25 preferred-opie-versions-1.2.3.inc,preferred-opie-versions-1.2.4.inc: There is only one version of opie-securityplugin-dummy Sep 02 02:03:25 Signed-off-by: Khem Raj Sep 02 02:03:34 03Graham Gower  07org.openembedded.dev * r11ed546d25 10openembedded.git/recipes/pulseaudio/ (29 files in 5 dirs): (log message trimmed) Sep 02 02:03:34 pulseaudio: Fix most recent recipe and remove broken recipes. Sep 02 02:03:34 Three months ago, commit 9cf826839c3cffd18f037d2ca8271affb85bab9d broke Sep 02 02:03:34 6 of the 7 pulseaudio recipes. No one noticed. Clearly the broken recipes Sep 02 02:03:35 are unused. Sep 02 02:03:35 0.9.15 appears to be in use as it still builds. 0.9.21 is the Sep 02 02:03:36 most recent recipe and has been retained. Sep 02 02:04:30 grg: I have staged 2615 and 2616 will test them and consider them applied Sep 02 02:04:56 grg: I am not too enthused with ibad thing I might have a deeper look to understand its value Sep 02 02:07:07 khem, thanks. The ibad thing was just so i could add a path or two to the cross compile badness paths. Also, the current cross compile badness patch calls abort() in gcc, which gives a message saying to report the bug to gcc devs - and looking at gcc archives this has happened. Sep 02 02:19:09 btw, I noticed pulse has been borked for a while Sep 02 02:19:24 but it was on a low priority project Sep 02 02:23:41 03Graham Gower  07org.openembedded.dev * r47f52315e1 10openembedded.git/recipes/tasks/task-boot.bb: (log message trimmed) Sep 02 02:23:41 task-boot.bb: Don't pull in u-a if ONLINE_PACKAGE_MANAGEMENT="none". Sep 02 02:23:41 This makes it possible to build an opkgless system, without having to set Sep 02 02:23:41 DISTRO_UPDATE_ALTERNATIVES. In the unlikely event that someone doesn't want Sep 02 02:23:41 opkg, but does want update-alternatives, one can set DISTRO_UPDATE_ALTERNATIVES Sep 02 02:23:41 in the local.conf. Sep 02 02:23:42 Signed-off-by: Graham Gower Sep 02 02:24:02 03Graham Gower  07org.openembedded.dev * rc5e0035ea3 10openembedded.git/classes/package_ipk.bbclass: Sep 02 02:24:02 package_ipk.bbclass: remove redundant dependencies upon opkg/opkg-collateral. Sep 02 02:24:02 Signed-off-by: Graham Gower Sep 02 02:24:02 Signed-off-by: Khem Raj Sep 02 02:26:36 03Graham Gower  07org.openembedded.dev * r0f08f4a52c 10openembedded.git/recipes/gpe-package/ (4 files in 2 dirs): Sep 02 02:26:37 gpe-package: remove some ancient recipes. Sep 02 02:26:37 Signed-off-by: Graham Gower Sep 02 02:26:37 Signed-off-by: Khem Raj Sep 02 02:26:38 03Graham Gower  07org.openembedded.dev * ra970a3ab04 10openembedded.git/recipes/gpe-package/gpe-package_0.4.bb: Sep 02 02:26:38 gpe-package_0.4.bb: fix build problems introduced in previous commit. Sep 02 02:26:38 gpe-package really depends on libipkg, not IPKG_VARIANT. It will not build Sep 02 02:26:38 against a recent opkg. Sep 02 02:26:39 Also, SRC_URI+=foo must be after inherit gpe or gpe.bbclass overwrites it. Sep 02 02:26:39 Signed-off-by: Graham Gower Sep 02 02:26:40 Signed-off-by: Khem Raj Sep 02 02:31:27 thanks khem! Sep 02 02:34:26 hmm gnuradio is arm only due to an arm only python-numpy recipe Sep 02 02:43:56 sorry Sep 02 02:44:10 and it is ancient version Sep 02 02:44:33 right now I am using a srctree recipe Sep 02 02:44:40 but some people still try that one out **** ENDING LOGGING AT Thu Sep 02 02:59:58 2010