**** BEGIN LOGGING AT Fri Jul 20 02:59:59 2012 Jul 20 05:26:43 Hi, I have a question, Hope some good soul could help me Jul 20 05:27:06 I am newbie and using Tiny Os QUick start guide Jul 20 05:27:24 THe thing is I have TinyOs installed with me Jul 20 05:27:41 I am using Imote2 by Crossbow and want to burn C code on it Jul 20 05:27:48 when I try to do motelist Jul 20 05:27:54 I get No devices found Jul 20 05:33:01 any one :P Jul 20 06:49:03 Hi. I need to downgrade autotools after the update to automake 1.12, which packages are affected? I have downgraded automake to the previous verison in OE, but now I get this undefined macro AC_DEFINE when running autotools. Jul 20 07:45:32 Hello Jul 20 07:48:52 I built the following files on oe-classic (using muffinman-image.bb): Jul 20 07:48:53 $ ls -l tmp/deploy/images/ Jul 20 07:48:53 total 86968 Jul 20 07:48:53 Jlime-muffinman-image-2010.1-ben-nanonote.rootfs.jffs2 Jul 20 07:48:56 Jlime-muffinman-image-2010.1-ben-nanonote.rootfs.tar.bz2 Jul 20 07:48:59 Jlime-muffinman-image-2010.1-ben-nanonote.rootfs.tar.gz Jul 20 07:49:02 modules-2.6.36-r0-ben-nanonote.tgz Jul 20 07:49:04 muffinman-image-ben-nanonote.jffs2 -> Jlime-muffinman-image-2010.1-ben-nanonote.rootfs.jffs2 Jul 20 07:49:08 muffinman-image-ben-nanonote.tar.bz2 -> Jlime-muffinman-image-2010.1-ben-nanonote.rootfs.tar.bz2 Jul 20 07:49:11 muffinman-image-ben-nanonote.tar.gz -> Jlime-muffinman-image-2010.1-ben-nanonote.rootfs.tar.gz Jul 20 07:49:14 vmlinux.bin-2.6.36-r0-ben-nanonote.bin Jul 20 07:49:17 vmlinux.bin-ben-nanonote.bin -> vmlinux.bin-2.6.36-r0-ben-nanonote.bin Jul 20 07:49:20 Jul 20 07:50:25 I flashed the target device with Jlime-muffinman-image-2010.1-ben-nanonote.rootfs.jffs2 and the uImage-2.6.36 file (found inside Jlime-muffinman-image-2010.1-ben-nanonote.rootfs.tar.bz2) Jul 20 07:51:16 I get the following kernel error message at boot: Jul 20 07:51:16 "kernel panic - not syncing: VFS: Unable to mount root fs on unknown-block(0,0)" Jul 20 07:53:03 does anyone know what's wrong with my rootfs ? (I suppose the problem lies with the rootfs, since the kernel seems to be booting) Jul 20 07:53:30 and how can I debug that issue ? Jul 20 07:54:51 I have absolutely no clue here, any help would be appreciated... Jul 20 07:56:12 apelete: is jffs2 compiled in kernel and available at boot? Are the mtd partitions correctly declared? The issue here is the kernel cannot mount the rootfs Jul 20 07:56:18 you miss some driver Jul 20 07:57:10 btw, what is your kernel commandline? root= rootfstype= ? Jul 20 07:57:53 note, it's a very common error :) Jul 20 07:59:20 Hi Jul 20 07:59:33 I am using TinyOs quick start guide Jul 20 07:59:39 when I do motelist cmd Jul 20 07:59:49 i get "No devices found" Jul 20 07:59:59 can any body help ? Jul 20 08:00:21 I am using Imote2 by crossbow Jul 20 08:00:54 ant_work: what if jffs2 isn't compiled in the kernel ? how can I add the driver in the kernel configuration in oe ? Jul 20 08:10:41 btw, what is your kernel commandline? root= rootfstype= ? Jul 20 08:11:38 I don't know about those. I used a recipe that was already available to build rootfs + kernel Jul 20 08:13:05 I don't know th edevice nor the bootloader. Does it boot from nand as default? Jul 20 08:13:36 it boots from nand, and I used a ready-to-use uboot bootloader Jul 20 08:14:10 didn't built the bootloader myself in oe, it was a separate file. Jul 20 08:15:01 hi Jul 20 08:15:18 ant_work: hmm. I remember know that the bootloader was initially used with a uImage + ubi rootfs Jul 20 08:15:47 and I'm trying to reuse it with uImage kernel + jffs2 rootfs Jul 20 08:15:55 is that a problem ? Jul 20 08:16:30 I should definetly look at the kernel commandline I think. Jul 20 08:17:48 don't know where I can find it, but it has to be somewhere in the oe repository. Jul 20 08:19:55 s/remember know/remember now/ Jul 20 08:20:15 I see here it's UBI for NAND http://www.jlime.com/wiki/downloads/nanonote/extra Jul 20 08:20:55 maybe the u-boot you're using lacks jffs2 for size reasons, strange but possible Jul 20 08:21:21 I think th eproblems is the kernel/cmdline though, not the bootloader Jul 20 08:24:18 thanks, will look into the kernel first then. Jul 20 08:25:24 ant_work: is the bootloader involved with the loading of the rootfs ? I thought it was in charge of loading the kernel (which in turn loads the fs) Jul 20 08:26:06 right, there are second-stage bootloaders, though, scanning the devices Jul 20 08:26:08 building gcc* fails here in do_unpack/do_patch stage with recent oe-core; but worked fine with an 12-18h older version. Looks like a race between multiple BB_NUMBER_THREADS accessing the shared source dir. Does anybody else see this? Jul 20 08:26:28 apelete: imagine you have an initrd Jul 20 08:28:51 ant_work: how are they related ? (talking about second-stage bootloader and initrd) Jul 20 08:31:51 apelete: you're going to be confused.we're going OT. Try the jffs2 rootfs+kernel+combo provided by that distro then try to build yours. Jul 20 08:32:23 I suspect you may have ubi-formatted your mtd thus no jffs2 boot is possible Jul 20 08:33:12 check the latest status of ben-nanonote, th ecurrent botloader, th erootfstype people are using now in 2012 Jul 20 08:34:19 * ant_work has lot of crumbs under 'h' key, very sticky :) Jul 20 08:35:11 finally, abandon oe-classic asap :) Jul 20 08:36:47 yeah, porting to oe-core is planned. but I cannot do so as long as the recipe is actually broken in oe-classic. Jul 20 08:58:05 hi all, I try to compile oe for xilinx mach but it fail at kernel fetch. It should be fetched by git but http appear to be used. http://pastebin.com/g3xx1Vqv suggestion? Jul 20 09:02:48 I am using TinyOs quick start guide, when I do motelist cmd I get "No devices found" can any body help ? I am using Imote2 by crossbow Jul 20 09:35:13 mmh... bitbake.conf sets 'CCACHE_DISABLE = 0' but ccache checks for existence of this variable only, not for its value. So this disables ccache :( Jul 20 09:52:35 is there a good reason why pkg.m4 is put in FILES_${PN} instead of FILES_${PN}-dev? this moves pkg.m4 to $toolchain-sysroot/armv* instead of toolchain-sysroot/x86_64* in the toolchain, where I feel it should be in the x86 directory Jul 20 09:52:44 am I way of, or is there some good reason for this? Jul 20 12:57:15 hello Jul 20 12:59:17 i have a problem with the priority of my overlay. do you can tell me why bitbake builds only the default curl package from oe-core but didnt build my custom bb file? Jul 20 13:05:12 foobla: try using bitbake-layers show-overlayed Jul 20 13:05:27 foobla: so you are providing your own curl recipe? Jul 20 13:05:57 yes just to compile curl with ssl instead of gnutls Jul 20 13:06:13 foobla: you could probably do that more effectively with a bbappend Jul 20 13:07:38 so i have to write a bbappend in my meta directory? Jul 20 13:08:48 foobla: you don't have to; the method you are using should work; but using a bbappend means you only change exactly the variables you need to and it's easier to handle when the core recipe gets updated Jul 20 13:09:32 bluelightning: thats ok :D i'll try it with an append file thanks for helping me Jul 20 13:20:08 np Jul 20 13:51:36 quick question: if we have two recipes of same package in the same layer, but different versions, which one will be picked if PREFERRED_VERSION is not defined for it? my understanding is that it should be the one with higher version number Jul 20 13:52:18 it works great with the bbappend :D Jul 20 13:54:19 fusman: that's correct, also DEFAULT_PREFERENCE will have an effect if set in either of the recipes Jul 20 13:54:27 foobla: great :) Jul 20 14:04:09 mmh.. is it expected that the meta-systemd layer requires the meta-gnome + meta-efl one? Else, bitbake fails due to .bbappend for gdm + elsa without a matching .bb Jul 20 14:07:24 ensc|w: the README suggests using BBMASK to mask out the directories for those layers Jul 20 14:07:42 bluelightning: DEFAULT_PREFERENCE is not set in either recipes Jul 20 14:08:07 fusman: so you are finding that the older recipe is being built? Jul 20 14:08:15 yeah Jul 20 14:09:18 hi Jul 20 14:09:42 i'd like to override the defconfig of the core busybox recipe in my own layer Jul 20 14:09:43 fusman: and bitbake -e | grep PREFERRED_VERSION_packagename doesn't list anything? Jul 20 14:10:22 i tried it with a bbappend recipe and SRC_URI += "file://defconfig" Jul 20 14:10:37 while putting my new defconfig unter files/ Jul 20 14:10:38 MaxPowers: did you extend FILESEXTRAPATHS ? Jul 20 14:10:48 unfortunatly it took the old one Jul 20 14:10:52 no i didn't Jul 20 14:11:20 MaxPowers: you need FILESEXTRAPATHS_prepend := "${THISDIR}/files:" Jul 20 14:11:21 do i have to extend it so the folder where the .bbappend recipe lies is parsed? Jul 20 14:11:36 ok i c Jul 20 14:11:37 MaxPowers: not so it is parsed, but so it is searched for files during fetching Jul 20 14:11:51 the prepend ensures it is searched first Jul 20 14:12:03 ok, i'll try it thx Jul 20 14:17:32 bluelightning: allright, it works :) Jul 20 14:50:39 good bye :) Jul 20 15:03:31 woglinde? Jul 20 15:36:15 cbrake: did ka6sox ever contact you regarding openembedded.org backups? There's some broken pages out there. Jul 20 15:59:57 wmat: yes, I took a quick look, the files are present, but in a different location than WP is expecting. I suspect WP got moved or something Jul 20 16:26:53 anyone know how i get access to: http://cgit.openembedded.org/meta-openembedded-contrib/? Jul 20 16:27:30 msm, what kind of access? Jul 20 16:28:25 push access Jul 20 16:28:35 shoragan pong Jul 20 16:28:47 i was going to start a for_denzil branch for review Jul 20 16:29:02 and even for_master as well Jul 20 16:29:54 I would use the openembedded-devel list and propose what you want to do there. Jul 20 16:31:21 ka6sox: yes, I do that already Jul 20 16:32:16 ah, okay..thats the method for getting these things done... Jul 20 16:32:37 if the infra is broken I can fix that...but granting access to things is not me :D Jul 20 16:37:39 ka6sox: well i was just trying to maintain a transitory branch for meta-oe Jul 20 16:37:48 i do it for poky-contrib (not oe-contrib) Jul 20 16:37:56 it seemed like a good idea, although meta-oe volume is low Jul 20 17:02:10 cbrake: thx Jul 20 17:03:42 hi woglinde :) Jul 20 17:03:56 it seems that meta-java/master is broken for gcc 4.7 Jul 20 17:05:08 which would you recommend openjdk-6 or 7? Jul 20 17:11:15 msm: i expect cbrake can still give access to repositories there. i might be able to, i haven't tried in a while. remind me later Jul 20 17:22:42 shoragan hm right Jul 20 17:22:53 I fiex in openjdk-7 branch only Jul 20 17:22:56 ups fixed Jul 20 18:04:31 How does bitbake virtual/kernel and bitbake my-image play together? For instance if I clean, with my kernel config, and recompile creating some new kernel modules are they automatically going to get picked up next time I issue bitbake my-image? Jul 20 18:05:15 images are always reconstructed from scratch using the current binary packages Jul 20 18:05:29 woglinde, i've found some other stuff in the -7 branch; will send you updates till monday Jul 20 18:05:30 cya Jul 20 18:06:33 Ok, I thought was seeing an issue where I needed to bitbake my-image -c clean, to get it to pick up the new .ko Jul 20 18:14:09 kergoth: ok, thanks Jul 20 18:17:58 guys Jul 20 18:18:18 does anyone know an example of a recipe that depends on gsoap? Jul 20 19:28:43 is there a way to revert (unexport) the effect of 'export CCACHE_DISABLE ??= "0"'? E.g. so, that CCACHE_DISABLE is not set Jul 20 19:29:10 ccache checks for existence of 'CCACHE_DISABLE' so that current bitbake.conf disables ccache unconditionally Jul 20 19:40:48 unexport CCACHE_DISABLE Jul 20 19:48:37 kergoth: really? afais, bitbake supports 'export' only but not 'unexport' Jul 20 19:52:12 CCACHE_DISABLE[unexport] = "1" Jul 20 19:52:14 i should say Jul 20 19:52:20 grep the metadata, there are multiple uses of it Jul 20 20:14:03 is not really usable in a conditional context because only existence of flag is checked, not the value :( Jul 20 20:30:52 huh? Jul 20 20:31:06 oh, i see. just use anonymous python to delete hte flag if necessary Jul 20 21:50:44 Hi guys, what is it about the do_rootfs stage that 1) takes a long time and 2) doesn't seem to parallelize? I ask because I'm kinda tired of waiting >15 minutes every time I make my basic image Jul 20 21:53:57 correction: >20min Jul 20 21:59:05 do_rootfs is basically copying everything built and downloaded in previous stages into the requested image format(s), right? Jul 20 22:15:25 spacecolonyone: I don't think tar can be parallelized, can it? Jul 20 22:17:32 hollisb: yea, but taring 140 MB into a 75MB tarball on a brand new machine shouldn't take 30 min Jul 20 22:19:32 I wonder if it would pay to comment out the ubi image... Jul 20 22:43:06 I'm trying to set some kernel options with menuconfig and the options aren't showing up, anyone up for a go? Jul 20 22:44:05 I've talked about it a bit over in v4l, where Laurent confirms that the options (OMAP_IOMMU & OMAP_IOVMM) should be there, so I'm thinking it is an oe/bitbake issue Jul 20 22:44:20 rather an issue with my usage of OE :) Jul 20 22:46:35 anyway, when menuconfig failed to present the options I manually went through and made the changes to the defconfig for my kernel (following Koen's guidance @ https://groups.google.com/forum/?fromgroups#!topic/beagleboard/UlL4e0_O_UE) Jul 20 22:46:57 but the changes don't seem to be taking Jul 20 23:15:53 maybe you did not copy out the modified config to the recipe dir? Jul 20 23:16:57 see Koen's note about step 3) Jul 20 23:20:05 ant__: sadly, no :( Jul 20 23:20:10 that isn't the case Jul 20 23:20:50 it seems to be a bigger issue in that menuconfig won't evel let me select OMAP_IOMMU (even after pushing z) Jul 20 23:29:42 ok, so some guys just suggested "make menuconfig ARCH=arm", which I've never seen mentioned in the context of kernel configuration with bitbake, is there a reason for this or was it just assumed no one would be that silly? Jul 20 23:40:14 later guys, thanks for the help Jul 21 01:23:09 is there any way to cheek if an option got compiled into the kernel, WITHOUT booting and scatting /proc config.gz Jul 21 01:23:18 zcatting* **** ENDING LOGGING AT Sat Jul 21 02:59:58 2012