**** BEGIN LOGGING AT Fri Oct 14 02:59:57 2011 Oct 14 05:48:31 I want to build a simple filesystem(should be less than 40 Mb) which can support browsing ,RDP,and X11 sessions only ,my target board is mini2440 ,if any one can just give a list of minimum packages that i would be needed to up the system Oct 14 05:48:52 advices are welcomed ....... Oct 14 06:59:32 good morning Oct 14 08:41:44 gm Oct 14 08:58:10 there are now two pieces of low hanging fruit one for German speaker and one for English speaker to be done before GA Oct 14 09:04:24 Hi there, First off, thanks for OE, I really enjoy working with it. I've been a keen user of Gentoo and Linux from Scratch for years, so OE fits in very well! I've got a question regarding the DISTRO_FEATURES variable. Oct 14 09:07:47 I've been creating my own distro based on the Angstrom configurations. My distro is supposed to be a very much stripped console-image. From the original configurations, I stripped most of the DISTRO_FEATURES that Angstrom applies, however, when I try to build the 'console-image' (that I expect to be quite bare), still Xorg and all its dependencies are sucked in the build process. Oct 14 09:08:42 The question is, should I be looking at the DISTRO_FEATURES variable, or do I have to dig a bit deeper? Is there a list of supported DISTRO_FEATURES anyway? Oct 14 09:08:56 Oh, and thanks in advance :-) Oct 14 09:14:36 hi jonathan_ Oct 14 09:15:14 jonathan_: do not care too much about the build process... its really scary what gets pulled in building a very simple image. Oct 14 09:16:02 packaging is pretty good so that you will end up an a sane image - without all these X11 and similar bits used in the build process Oct 14 09:19:24 Hi florian, thanks for the apt response. OK, I see. I was running 'bitbake -n' all the time, and was indeed scared of by all the dependencies. So as I understand now all these dependencies (like Xorg) are pulled in, build and packaged, but not included in the end-result? (console-image) Oct 14 09:20:37 Up until now, I've been working on helloworld-image to strip down the Linux kernel configuration to a bare minimum. This was quite successful, and the build process is very lean as well (no Xorg et cetera). Oct 14 09:22:10 jonathan_: yes right... Oct 14 09:22:37 its most likely bluez pulling in stuff for build you dont want, if you remove bluetooth as a feature Oct 14 09:23:19 bluez depends on gstreamer which depends on everything including 3 kitchen sinks Oct 14 09:23:55 bluetooth was already gone. My DISTRO_FEATURES now is: DISTRO_FEATURES = "ext2 ipv4 ipv6 largefile pam xattr" Oct 14 09:33:42 use bitbake -g to see which recipes are actually depending on x11. Oct 14 09:33:59 though, if you're on contemporary oe-core, x11 shouldn't even be buildable if it isn't in DISTRO_FEATURES. what branch are you using? Oct 14 09:36:09 I installed OE using 'git clone http://github.com/openembedded/openembedded.git' Oct 14 09:51:03 jonathan_, you should also consider looking at openembedded-core, because openembedded.dev is kind of old Oct 14 09:52:43 Ah OK...so I guess the manual is wrong on this? Oct 14 10:04:34 Ok, so I did: 'git clone http://github.com/openembedded/oe-core.git'. But now it seems I should revise my way of working a bit... Oct 14 10:06:19 Aha. I see I've been starting out wrong already. *now reading oe-core manual* Oct 14 10:07:17 jonathan_, yes it's now modular, you might want to choose a distro Oct 14 10:07:19 like angstrom Oct 14 10:07:27 and use their setup scripts Oct 14 10:07:30 it's a lot easier that way Oct 14 10:11:52 OK, thanks. I'll look into it. Oct 14 10:14:09 jonathan_, http://sakrah.homelinux.org/blog/2011/03/using-openembedded-core-to-build-angstrom-for-qemu/ Oct 14 10:16:57 Thanks! That's very helpful of you. Oct 14 10:23:17 As I understand, there's no need for the environment variables BBPATH and BB_ENV_EXTRAWHITE anymore? And I guess I can better use the bitbake that is in GIT over the one installed by Gentoo (1.10.2)? Oct 14 10:24:04 I see bitbake in GIT is a bit newer. Oct 14 10:24:53 simply use the angstrom scripts for now Oct 14 10:25:02 it'll setup everything for you Oct 14 10:25:54 and since bitbake is still bitbake there is still need of BBPATH, BB_ENV_EXTRAWHITE depend on your setup Oct 14 10:26:13 and you should not use your distro's bitbake Oct 14 10:26:19 it's rare to need to set those variables yourself however Oct 14 10:27:10 In openembedded-classic I needed to do this, according to the manual.. Oct 14 10:27:38 where's your manual located? Oct 14 10:28:48 Here it is: http://docs.openembedded.org/usermanual/usermanual.html Oct 14 10:29:00 The section about setting up is here: http://docs.openembedded.org/usermanual/usermanual.html#gettingoe_getting_oe Oct 14 10:29:44 But, this seems all outdated now. So I will 'un-setup' my environment first, before starting out with oe-core. Oct 14 10:31:18 ok Oct 14 10:41:26 I will use the Yocto project documentation, is that OK? (I'm kinda paranoid now.. :-P) Oct 14 10:46:13 what device+distro are you targetting? Oct 14 11:01:47 For now, I'm targetting a BeagleBoard but mostly just for the ARM core. In the near future we'll build a prototype board with a similar (if not the same) core. Peripherals will greatly differ from the BeagleBoard, and speed is important, so my idea was to create a minimal distro based on preferred versions of an existing distro, like Angstrom. But then stripped to the bare minimum. Oct 14 11:02:47 doh, why is it always such a pain to get rid of X11 in the console image? I got rid of bluetooth in distro features and still something is pulling in all the X crap Oct 14 11:03:22 I already stripped the Linux kernel (using the helloworld-image from OE-classic) to a minimum (about 1.1MiB, booting in under 1,1secs at 1GHz). Oct 14 11:03:33 Oh, the BeagleBoard is a -xM btw. Oct 14 11:04:12 s/a/an Oct 14 11:40:39 Ok. Now where to look for beagleboard machine configs? I found the BSP with binaries (http://www.yoctoproject.org/download/bsp/texas-instruments-arm-cortex-a8-development-board-beagleboard). But I'd like to do a custom build. The only machine configs I see now are qemu types. Oct 14 12:30:18 Geez this new building method is quite something else from what it used to be. Oct 14 12:31:49 Unfortunately I did not succeed building anything yet. I tried GNUtoo's blog example, but used './oebb.sh config beagleboard' instead. Oct 14 12:32:40 will read the backlog soon, I must try to call my om-gta02 to see if a bug was fixed by a patch Oct 14 12:32:50 Then: 'bitbake -n helloworld-image' which stops at .bb file indexing: ERROR: Could not include required file recipes-graphics/xorg-xserver/xserver-xf86-dri-lite.inc####################################### | ETA: 00:00:11 ERROR: Command execution failed: Exited with 1 Oct 14 12:34:48 jonathan_: remove meta-intel layer if you don't need it Oct 14 12:34:59 jonathan_: or you have to apply few patches from oe-core ML to it Oct 14 12:36:23 Iam using overo board and kernel linux-omap-psp-2.6.32-r102g, When i set CONFIG_FB_OMAP2_NUM_FBS=1 my kernel crashes Oct 14 12:36:45 [] (omap_vout_probe+0x85c/0x9b0) from [] (platform_drv_probe+0x1c/0x24) Oct 14 12:36:50 [] (platform_drv_probe+0x1c/0x24) from [] (driver_probe_device+0xa4/0x174) Oct 14 12:36:55 [] (driver_probe_device+0xa4/0x174) from [] (__driver_attach+0x60/0x84) Oct 14 12:37:30 what patch need to be added to overcome this crash ? Oct 14 12:41:36 Thanks JaMa|Wrk. I edited conf/bblayers.conf and removed all meta-intel layers. Now the build stalls somewhere else... Oct 14 12:43:19 ERROR: No recipes available for: tiff, libgcrypt and atk Oct 14 12:44:27 jonathan_: you have to update Oct 14 12:46:21 Tssss..that worked! Thanks. Did I just need to update because a fix was committed or did I need to update because of the bblayers.conf change? Oct 14 12:47:20 because it was fixed few mins ago Oct 14 12:47:35 welcome to rapid development :) Oct 14 12:47:41 Ok. Lucky me :-) Oct 14 12:49:23 khem: can we do something about libGL in qemu? I mean introduce mesa-native or something? Oct 14 12:49:42 khem: I really don't want to install libgl to my minimal chroot just to build qemu-native Oct 14 12:54:17 hi i just build the kernel for my board Oct 14 12:54:19 it load Oct 14 12:54:23 everything seems ok Oct 14 12:54:36 but i can load the rfs Oct 14 12:54:39 cant* Oct 14 12:55:01 Please append a correct "root=" boot option Oct 14 12:55:11 where do i find this boot option? Oct 14 12:57:05 this is in the kernel commandline Oct 14 12:57:39 where to find the command line? Oct 14 12:57:56 bootloader or kernel sources Oct 14 12:58:45 can be hardcoded during kernel build Oct 14 12:58:49 so which file inside the kernel i have to look into? Oct 14 12:58:56 .config Oct 14 12:59:11 can i specify the memory address in it? Oct 14 12:59:15 don't you use u-boot or such? Oct 14 12:59:23 ya i use uboot Oct 14 12:59:32 ok, then set it there :) Oct 14 12:59:37 u mean i can set inside uboot? Oct 14 12:59:42 sure Oct 14 13:00:01 is it this command? Oct 14 13:00:02 setenv bootargs console=ttyO2,115200 root=/dev/nfs Oct 14 13:00:10 yes Oct 14 13:00:13 whereby my root = 0x???????? ? Oct 14 13:01:25 http://www.kernel.org/doc/Documentation/kernel-parameters.txt Oct 14 13:01:44 nfsroot= [NFS] nfs root filesystem for disk-less boxes. Oct 14 13:01:45 See Documentation/filesystems/nfs/nfsroot.txt. Oct 14 13:02:41 yea that is an example Oct 14 13:02:54 just wondering if i can just root= memory address? Oct 14 13:03:52 an initramfs? Oct 14 13:04:20 as i wanna load my rfs into a memory address Oct 14 13:04:51 so my kernel will know which address to find my rfs Oct 14 13:07:46 isn't easier to load it as initramfs? Oct 14 13:08:57 not really sure and how also Oct 14 13:09:09 still figuring out how to load my rfs Oct 14 13:09:15 really need help on it Oct 14 13:09:23 i have my kernel and my hello world image Oct 14 13:09:29 my kernel is working fine i presume Oct 14 13:09:46 but helloworld how to act as rfs i am still finding out how Oct 14 13:10:43 hi, all! Oct 14 13:11:08 trying to make new OE-core setup, and get this error constantly: http://pastebin.com/jByFqt1J Oct 14 13:11:15 any ideas? Oct 14 13:11:49 kaylessa, I suppose the easy way is 1) create a cpio.gz with your rootfs 2) embed it in the kerenl or pass it as initrd= argument in the bootloader Oct 14 13:11:54 http://www.kernel.org/doc/Documentation/filesystems/ramfs-rootfs-initramfs.txt Oct 14 13:11:58 slapin: latest bitbake? Oct 14 13:12:10 JaMa|Wrk: everything latest Oct 14 13:12:20 ok can thanks ant_work Oct 14 13:12:27 just used Koen's scripts Oct 14 13:15:17 my script: http://pastebin.com/tjFPRwMy Oct 14 13:16:38 full output: http://pastebin.com/gS4UQSPh Oct 14 13:28:24 slapin: check which sanity check is failing.. maybe there is bug Oct 14 13:29:20 i have micro-helloworld-image-uclibc-ipk-20111004-sarge-at91.rootfs.tar.gz Oct 14 13:29:28 how to i convert to a ramdisk? Oct 14 13:31:41 JaMa|Wrk: this is from execution of line 257: if check_conf_exists("conf/site.conf", e.data): Oct 14 13:31:58 conf/site.conf is in perfect condition though Oct 14 13:32:30 (as I compare with different setup, from two weeks ago) Oct 14 13:39:18 slapin: check what's inside check_conf_exists Oct 14 13:40:27 JaMa|Wrk: i found it, it was wrong assumption originally. Oct 14 13:40:59 JaMa|Wrk: error cause is line 257 of check_sanity function (not very helpful error message issue), if oeroot.find ('+') != -1: Oct 14 13:41:31 JaMa|Wrk: shomhow oeroot at tis place is None, which is unexpected by code. Oct 14 13:42:09 JaMa|Wrk: somehow COREBASE variable is not set Oct 14 13:44:32 JaMa|Wrk: where it should be set? Oct 14 13:45:49 slapin: openembedded-core $ cat meta/conf/layer.conf Oct 14 13:52:00 ah, I see, it is due to ordering in bblayer.conf Oct 14 13:52:15 or not Oct 14 13:54:03 yes, that fixed things up Oct 14 13:54:10 JaMa|Wrk: thanks a lot! Oct 14 13:54:16 yw Oct 14 14:05:07 is there some locking protection in bitbake to prevent multiple builds in the same TMP directory? Oct 14 14:06:51 isn't Oct 14 14:11:16 has anyone tried to compile openssl for powerpc recently? I get http://pastebin.mozilla.org/1355052 Oct 14 14:44:18 regarding openssl... had to remove this line from the recipe to make it work http://pastebin.mozilla.org/1355085 Oct 14 14:47:13 I want to run periodic builds, so I want to know if previous build completed before running new one. Is there some way to do that? Oct 14 14:48:35 slapin: you could setup buildbot Oct 14 14:49:56 Jin^eLD: any links? Oct 14 14:51:47 http://trac.buildbot.net/ Oct 14 14:52:05 but I do not have any links to a ready to use OE setup/configuration, I just remember I used it a couple of years ago Oct 14 14:52:12 for periodic OE builds Oct 14 14:52:31 It might be one in contrib Oct 14 15:11:23 03Marco Cavallini  07master * r1831408cf1 10openembedded.git/recipes/python/python-numpy/ (config.h numpyconfig.h): Oct 14 15:11:23 python-numpy: copied missing files from arm directory Oct 14 15:11:23 * this is a temporary patch that allow build to complete Oct 14 15:11:23 * python-numpy mantainer should verify and reorganize files in this recipe Oct 14 15:23:06 hi all i got a question on initramfs Oct 14 15:23:36 i am still confuse what should be the argument for CONFIG_INITRAMFS_SOURCE Oct 14 15:24:55 can be a file or a directory or a cpio Oct 14 15:25:00 http://www.kernel.org/doc/Documentation/filesystems/ramfs-rootfs-initramfs.txt Oct 14 15:25:06 yup i read that Oct 14 15:25:19 so what happens after i did that Oct 14 15:25:23 i make it? Oct 14 15:25:43 is there an oe style of making it? Oct 14 15:25:53 if you want to embed it in kernel, yes Oct 14 15:25:59 does bitbake virtual/kernel remake my kernel? Oct 14 15:26:08 cos i already did virtual/kernel once Oct 14 15:26:38 bitbake -c clean and bitbake -c cleansstate Oct 14 15:26:59 JaMa|Wrk: there are two faction. One who dont want to build so much host software and then other like you who want minimal chroot Oct 14 15:27:09 keine ahnung :-) Oct 14 15:27:19 oops... wrong window Oct 14 15:27:25 i will try :D Oct 14 15:27:28 ok thanks alot ant_works Oct 14 15:29:29 khem: with gentoo I have to build them in both cases :) Oct 14 15:29:49 khem: but ok.. Oct 14 15:30:42 khem: I've removed EXTRA_IMAGEDEPENDS += "qemu-native qemu-helper-native" from qemu.inc as I don't plan to test those qemu images in my chroot anyway Oct 14 15:35:14 ant_works just wanna double comfirm Oct 14 15:35:41 bitbake -c clean virtual/kernel create a new kernel file after i edit the .config ? Oct 14 15:38:06 no, clean the previous build Oct 14 15:38:52 looks like Angstrom build libav_git and mplayer_svn with a missing yasm Oct 14 15:38:58 any hint? Oct 14 15:39:57 the problem is that build fails on both packages Oct 14 15:41:29 then how u rebuild the new kernel using opemembedded style rather than "MAKE" Oct 14 15:44:11 according: well, you need a different defconfig Oct 14 15:44:43 so you can 1) start with bitbake -c menuconfig and then bitbake virtual/kernel Oct 14 15:44:57 but this .config will be lost Oct 14 15:45:11 or 2) provide a proper defconfig Oct 14 15:48:25 whats the different between the defconfig and .config? Oct 14 15:48:38 as for the image, you can declare INITRAMFS_IMAGE = "initramfs-xx-image" and this will be built Oct 14 15:48:47 and embedded in one single pass Oct 14 15:48:52 ah Oct 14 15:48:52 magic of OE ;) Oct 14 15:48:59 yea thats it! Oct 14 15:49:05 it is under local.conf right? Oct 14 15:49:21 well, or in the kernel recipe Oct 14 15:49:29 cool! Oct 14 15:49:35 i beginning to feel the power Oct 14 15:49:36 of OE Oct 14 15:49:38 thanks Oct 14 15:49:44 i go play around Oct 14 15:49:51 look at this: http://cgit.openembedded.org/meta-handheld/tree/recipes-bsp/images/initramfs-kexecboot-image.bb Oct 14 15:50:23 ok i go take a look Oct 14 15:51:17 the image.bb must specify IMAGE_FSTYPES = "cpio.gz" at least Oct 14 15:52:10 bbl, bye Oct 14 16:07:22 hmm, are there any news on this issue? http://comments.gmane.org/gmane.linux.distributions.angstrom.devel/4615 as a temporary solution it points to a workaround at http://patchwork.openembedded.org/patch/2069/ but I somehow do not see a relation between the two Oct 14 16:07:26 does anyone have an idea? Oct 14 16:19:29 Jin^eLD: patchwork.openembedded.org was old pw installation but somehow now its resolving to new patches.openembedded.org Oct 14 16:19:38 which should not happen I think Oct 14 16:20:28 khem: I tried to bump uclibc in my build to .3, let's see what happens Oct 14 16:20:44 as it seems that the issue with e2fsprogs is still around Oct 14 16:22:21 Jin^eLD: are you with 2011.03 release ? Oct 14 16:22:31 khem: btw, have you seen any caching issues with nginx? sometimes cgit returns different page on every ctrl+f5 Oct 14 16:22:45 khem: and it didn't do it before.. so I guess it's not http proxy here Oct 14 16:22:55 JaMa|Wrk: hmm Oct 14 16:23:10 JaMa|Wrk: there has been some work going on with url rewrite Oct 14 16:23:17 so it could be because of that Oct 14 16:23:34 but if you keep seeing it Oct 14 16:23:39 say for a week Oct 14 16:23:46 khem: oh hmm seems I am on angstrom-2010.x Oct 14 16:23:51 ie directly on http://git.openembedded.org/ is sometimes concrete patch, then tree then list of repos, then message "no such repository" and ie after fifth refresh there is expected list of repos Oct 14 16:23:51 my bad Oct 14 16:24:14 JaMa|Wrk: yes that seems to be due to url rewrite Oct 14 16:24:22 khem: it was in last 2-3 days.. but not easily reproducible Oct 14 16:24:28 JaMa|Wrk: once we figure out correct redirects Oct 14 16:24:34 ok.. I'll let you know if I see it again Oct 14 16:24:37 khem: uhm.. I do not see an angstrom 2011 conf? Oct 14 16:24:51 Jin^eLD: no there is 2010.x Oct 14 16:24:57 and angstrom-bleeding Oct 14 16:24:59 thanks Oct 14 16:25:31 khem: I must be blind as I do not see any -bleeding config Oct 14 16:26:22 Jin^eLD: I am talking about new world order Oct 14 16:26:27 Jin^eLD: using layers Oct 14 16:27:55 khem: sorry I've been "away" for over a year, was working with angstrom-2008.1 (started back then when it was the current stable release) Oct 14 16:28:10 what nwo and what layers? do you have a link? Oct 14 16:28:35 so the last 2 years I've been working with the old release and have no clue about recent changes Oct 14 16:28:51 just now got my iomega storcenter out of the box, off work, and am trying to build an image for it :) Oct 14 16:29:33 bumping uclibc to uclibc-0.9.30.3 did not solve the e2fsprogs compile problem :( Oct 14 16:30:06 Jin^eLD: lot has happened since then Oct 14 16:30:15 what problem do u see Oct 14 16:31:29 Jin^eLD: For the new structure you can read http://sakrah.homelinux.org/blog/2011/03/using-openembedded-core-to-build-angstrom-for-qemu Oct 14 16:32:16 khem: I am working on oe.dev master and angstrom-2010.x configuration, building for storcenter/popwerpc, so far after a little battle to get rid of X11 everything seemed fine until I hit the problem where e2fsprogs fails to compile (with the message in the link I posted before) Oct 14 16:33:46 Jin^eLD: and also here http://www.angstrom-distribution.org/building-angstrom Oct 14 16:33:50 khem: that link you posted.. what exactly is the difference to the old proceedings? Oct 14 16:34:16 let me check out the oebb.sh script though Oct 14 16:34:19 that seems new Oct 14 16:34:27 Jin^eLD: its using the layers Oct 14 16:34:59 layers as in collections or what exactly is that? Oct 14 16:35:08 what's the difference to the regular local.conf concept? Oct 14 16:36:12 reading oebb.sh now.. Oct 14 16:37:17 khem: Did you add the rewrite rules I gave you? Oct 14 16:37:35 Tartarus: they did not work Oct 14 16:37:40 The complex one? Oct 14 16:37:43 yes Oct 14 16:37:48 neither of them Oct 14 16:37:56 Is it enabled now? Oct 14 16:38:41 Tartarus: no Oct 14 16:38:53 Can you enable it or tell me what was wrong? Oct 14 16:39:02 rewrite ^/.*p=(.*)\.git&a=(.*)\;h=([0-9a-f]+)$ http://git.openembedded.org/$2/$3/?id=$4 last; Oct 14 16:39:07 is what I have been working on Oct 14 16:39:11 but it does not work either Oct 14 16:39:50 Tartarus: it just says pattern Does not match Oct 14 16:39:54 khem: this layer stuff looks like regular usage of collections... unless I am missing something Oct 14 16:40:01 Can you paste back to me the complex one? Oct 14 16:40:11 My friend put that up on a server and it looked like it was redirecting right Oct 14 16:45:58 Tartarus: this is what I have http://pastebin.com/FLpwqzXy Oct 14 16:46:24 Jin^eLD: yes collections but now core meta-data is lot less Oct 14 16:47:06 khem: I'd have to see it through and understand how it works first Oct 14 16:47:39 but that's a different issue :P trying to compile e2fsprogs for uclibc first Oct 14 16:48:56 Tartarus: hmmm I change .net to .org and now I get this This webpage has a redirect loop Oct 14 16:49:17 Jin^eLD: paste your error Oct 14 16:49:20 somewhere Oct 14 16:49:36 Do you have other redirects on? Oct 14 16:49:38 The webpage at http://git.openembedded.org///?id= has resulted in too many redirects Oct 14 16:49:41 nope Oct 14 16:50:45 khem: http://pastebin.mozilla.org/1355118 basically the same as described in one of the comments here: http://comments.gmane.org/gmane.linux.distributions.angstrom.devel/4615 Oct 14 16:52:38 Jin^eLD: Find which .o is asking for __tls_get_addr Oct 14 16:52:49 readelf -s *.o | grep __tls_get_addr Oct 14 16:57:32 khem: its in ./lib/et/elfshared/error_message.o Oct 14 16:58:08 ok can --disable-tls option to e2fsprogs Oct 14 16:58:11 configure Oct 14 16:58:24 will try that, thank you Oct 14 17:02:12 khem: that worked indeed, thanks! Oct 14 17:04:53 Jin^eLD: you should try the new angsrom I strongly reccommend Oct 14 17:05:13 khem: I do fear too much automagic with this oebb.sh and so on Oct 14 17:05:27 do I really have to use that in order to use the new angstrom? Oct 14 17:21:53 Crofton, Crofton|Work you about? Oct 14 17:23:16 a little Oct 14 17:49:41 khem: ping, try http://pastebin.com/FtBsgrez Oct 14 17:58:53 Tartarus: nginx does not parse that Oct 14 17:59:19 It does not like if followed immediately with if Oct 14 18:08:56 khem: appologies, http://pastebin.com/wD3p9qPY Oct 14 18:09:59 Tartarus: yes I tweaked it to something like that Oct 14 18:10:01 already Oct 14 18:10:12 k Oct 14 18:10:15 Is it working finally? Oct 14 18:10:30 yes it seems Oct 14 18:10:54 I wonder why we can not have all three regexp in onw Oct 14 18:13:02 hi Oct 14 18:13:22 I asked on the list, but no reply -- are the meta-openembedded-contrib and oebuildstats repos supposed to be broken? I cannot clone either of them. Oct 14 18:36:04 uwe_: meta-openembedded-contrib is for users Oct 14 18:36:28 let me see whats wrong Oct 14 18:39:02 if I add add the following to a recipe what is the difference Oct 14 18:39:10 VARFOO = blah Oct 14 18:39:12 vs. Oct 14 18:39:29 VARFOO_pn-recipename = blah Oct 14 18:39:49 seems like if this is in the recipe file itself it makes no difference Oct 14 18:40:52 uwe_: meta-openembedded-contrib should be fixed Oct 14 18:46:50 about my question about - i guess the VARFOO would not be defined for a native build Oct 14 18:49:39 msm: if the recipe does BBCLASSEXTEND then this does not apply Oct 14 18:49:48 and also if you have FILE_xxx Oct 14 18:49:54 where xxx is not recipename Oct 14 18:51:19 right Oct 14 18:51:39 ok so this probably from before BBCLASSEXTEND was used for -native Oct 14 19:03:35 is the CDC-Ethernet gadget not set to dhcp by default ? Oct 14 19:07:12 khem: yup, works now, thanks! oebuildstats is still broken though. Oct 14 19:47:46 OT: does ubi(block) support snapshots or can it be stacked with md? Oct 14 20:04:08 Anyone know why mirrors.openembedded.org is down? Oct 14 20:07:43 How can I debug distribute_sources execution? Oct 14 20:41:01 otavio: how does it fail? you always have the usual -DDD Oct 14 20:53:34 zecke: doesn't work Oct 14 20:53:36 zecke: http://paste.debian.net/136542/ Oct 14 20:53:42 this is my WIP fix Oct 14 20:53:52 but it doesn't copy the patches over there Oct 14 20:56:00 AFAIK we need to distribute the sources and the patches Oct 14 20:56:08 so seems it is missing the patches now Oct 14 20:56:53 RP_: please look at http://paste.debian.net/136542/; it works now but fails to put the patches on source deploy dir Oct 14 20:57:07 * otavio is out now Oct 14 20:57:10 bbl **** ENDING LOGGING AT Sat Oct 15 02:59:56 2011