**** BEGIN LOGGING AT Wed May 06 02:59:58 2015
May 06 05:26:30 <drix> sup
May 06 07:22:00 <miandonmenmian> does anyone know where is the "interfaces" file on yocto?
May 06 07:22:36 <nrossi> miandonmenmian: as in the /etc/network/interfaces file?
May 06 07:23:11 <miandonmenmian> ok.. i got no network folder
May 06 07:23:28 <miandonmenmian> but i got a interface that keeps loading up at boot and i am not sure how to disable it
May 06 07:23:52 <nrossi> is it loading during kernel boot or during init?
May 06 07:25:31 <nrossi> aka is getting setup before or after the "INIT:..." line during boot
May 06 07:26:43 <miandonmenmian> INIT line is after uboot?
May 06 07:27:49 <miandonmenmian> i believe this is happening on systemd
May 06 07:27:50 <nrossi> yes, it should appear when the kernel has finished booting and hands off to userspace init
May 06 07:27:55 <miandonmenmian> if that answers the question...
May 06 07:28:01 <nrossi> oh in that case completely different outcomes :P
May 06 07:28:36 <miandonmenmian> i might be wrong.. but what do you mean?
May 06 07:29:00 <nrossi> are you using systemd+sysvinit mixed mode, or purely systemd
May 06 07:29:21 <nrossi> if full on systemd you will need to look at systemd-networkd for network interfaces
May 06 07:29:31 <nrossi> /etc/systemd/netowkr/
May 06 07:29:38 <nrossi> /etc/systemd/network/*
May 06 07:29:38 <miandonmenmian> ah! ok
May 06 07:30:13 <miandonmenmian> i'm still quite confused with systemd stuff. what is sysvinit mixed mode? the old init.d stuff?
May 06 07:30:50 <nrossi> yep, mixed in with systemd
May 06 07:31:07 <nrossi> because systemd does support starting init.d scripts
May 06 07:31:50 <miandonmenmian> oh, i read it was compatible. good to know it isnt..
May 06 07:32:10 <nrossi> Depends on your definition of 'compatible'  :)
May 06 07:32:21 <miandonmenmian> interesting, the interface i want to get rid is there, but the one i want to use is not. but seems like im on the right track
May 06 07:32:36 <miandonmenmian> i suppose most confs of systemd should be on that folder, right?
May 06 07:32:43 <miandonmenmian> /etc/systemd/...
May 06 07:33:02 <nrossi> i believe so, i am not completely familar with all of systemd yet :)
May 06 08:12:15 <bluelightning> morning all
May 06 09:13:53 <drix> Is it possible to have something variable in an overriden file ? For instance, having a /etc/network/interfaces that contains ${IP_ADDRESS} or something like that ?
May 06 09:15:09 <LetoThe2nd> drix: you can have variables in the recipes. in the file itself, you'd have to replace it, for example with sed
May 06 09:15:22 <LetoThe2nd> otherwise the file content goes onto the target verbatim
May 06 09:17:20 <drix> alright, so I can call sed in the context of the recipe ?
May 06 09:17:34 <LetoThe2nd> sure
May 06 09:18:31 <LetoThe2nd> i think sed-native is even implicitly there, but i'm not sure
May 06 09:20:38 <drix> yes, i'm just wondering how i can apply the sed replacement
May 06 09:21:05 <LetoThe2nd> um... sed -i filename 's/old/new/g' ;)
May 06 09:22:12 <drix> yes sure, but what's the variable that stores the conf/machine/interfaces file !
May 06 09:23:16 <LetoThe2nd> i personally would just skim/grep existing layers for post_install examples, i'm pretty sure there should be something comparable to take a look at
May 06 09:25:48 <drix> pkg_postinst_PACKAGENAME () { #!/bin/sh -e # Commands to carry out }
May 06 09:25:50 <drix> sounds good
May 06 09:27:43 <LetoThe2nd> hm, i was thinking more along the lines of appending to the install command in a bbappend, if you modify the behaviour of an existing recipe.
May 06 09:28:42 <LetoThe2nd> but of course there are probably multiple ways of archieving this
May 06 09:31:43 <drix> actually, the main purpose for me is to avoid creating, lets say 10 machines, for 10 different network configurations, because it will be the only difference between the machines
May 06 09:32:33 <LetoThe2nd> i know what you mean but this is pretty much unrelated to what i said earlier ;)
May 06 09:33:04 <LetoThe2nd> its "reason to do something" versus "musing on various ways to do something"
May 06 10:06:05 <mcfrisk> Hi! How is bitbake calling shell script functions like buildhistory_commit() in meta/classes/buildhistory.bbclass? With set -e or just plain /bin/sh?
May 06 11:04:33 <shakalaka> hi
May 06 11:05:20 <shakalaka> im looking for some information about creating own recipes
May 06 11:06:01 <shakalaka> anybody here?
May 06 11:06:06 <bluelightning> shakalaka: hi
May 06 11:06:20 <bluelightning> shakalaka: have you seen http://www.yoctoproject.org/docs/current/dev-manual/dev-manual.html#new-recipe-writing-a-new-recipe ?
May 06 11:07:14 <shakalaka> i did it with https://wiki.yoctoproject.org/wiki/How_do_I#Q:_How_do_I_put_my_recipe_into_Yocto.3F
May 06 11:08:08 <shakalaka> but when im trying switch gnu helloworld with my application in dropbox, i get Nothing RPROVIDES error
May 06 11:08:35 <bluelightning> shakalaka: can you use pastebin to show me your recipe please?
May 06 11:08:48 <shakalaka> yeah, sure, 1 minute
May 06 11:09:43 <shakalaka> http://pastebin.com/kENh8dNm
May 06 11:10:13 <shakalaka> md5 is ok
May 06 11:10:25 <bluelightning> shakalaka: so what is this recipe called now and what is the exact error you receive?
May 06 11:11:17 <shakalaka> i also add this string in local.conf in build directory IMAGE_INSTALL_append = " cool"
May 06 11:12:13 <shakalaka> this is layer.conf http://pastebin.com/EAyjAh5t
May 06 11:12:42 <shakalaka> my folder with recipes "meta-vc"
May 06 11:13:39 <shakalaka> then "source oe-init-build-env", then "bitbake core-image-minimal", and i get error ERROR: Nothing RPROVIDES 'cool' (but /home/valentin/Desktop/Projects/Yocto/poky/meta/recipes-core/images/core-image-minimal.bb RDEPENDS on or otherwise requires it)
May 06 11:14:15 <neverpanic> Has anybody ever tried to use a keyring for passwords to be used by Subversion & Git inside a bitbake build?
May 06 11:14:51 <shakalaka> and this is my bblayers.conf http://pastebin.com/zWDAE3hE
May 06 11:22:34 <jku> shakalaka: what's the recipe filename?
May 06 11:25:03 <bluelightning> if the recipe file name isn't "cool" then that will be the source of your problem, since by default the main package produced by the recipe is the same name as the recipe
May 06 11:25:39 <bluelightning> otherwise, it's possible the system is not seeing your recipe at all
May 06 11:26:20 <bluelightning> which would be caused by it not being in a path that bitbake is looking (as defined by bblayers.conf and layer.conf in each layer)
May 06 11:26:38 <bluelightning> bitbake-layers show-layers and bitbake-layers show-recipes should give you a way to check those
May 06 11:30:07 <drix> What's the best way to install a file into a /etc/ subdirectory ? I put it in the do_install_append part of my recipe but it's not working
May 06 11:31:07 <drix> cp: cannot create regular file '/home/cedric/yocto/poky-dizzy-12.0.1/build/tmp/work/core2-32-poky-linux/toast-001/1.0-r0/image/etc/network/interfaces': No such file or directory
May 06 11:31:27 <drix> looks like the /etc/network tree doesn't exist. I'm not sure I'm doing it the correct way.
May 06 11:33:58 <rburton> drix: you need to create the directories before you can put stuff in them
May 06 11:44:15 <shakalaka> my-program_1.0.bb
May 06 11:44:25 <shakalaka> recipe name
May 06 11:45:06 <jku> shakalaka: should be cool_1.0.bb to get packagename "cool"
May 06 11:45:36 <drix> rburton: thanks, that's what i thaught, but i wasn't sure it was a good practice!
May 06 11:46:25 <shakalaka> omg, it's working! thanks :)
May 06 11:47:10 <shakalaka> i get another error Function failed: cool: LIC_FILES_CHKSUM points to an invalid file, but i think i can handle this
May 06 13:38:35 <TobSnyder1> can I override ERROR_QA of poky/meta-yocto/conf/distro/poky.conf in my layer?
May 06 13:39:50 <rburton> yes
May 06 13:39:58 <rburton> or create your own distro and override it there
May 06 13:43:56 <TobSnyder1> how to do so in my layer?
May 06 13:44:08 <TobSnyder1> actually I do have the problem of already-stripped binary
May 06 13:44:11 <TobSnyder1> QA Issue: File was already stripped, this will prevent future debugging!
May 06 13:48:19 <rburton> right, don't do that in the layer
May 06 13:48:38 <rburton> ideally, fix the build so that ti doesnt strip
May 06 13:48:50 <rburton> unless of course its a binary you don't build yourself so you can't
May 06 13:48:57 <rburton> in which case you want to look up INSANE_SKIP in the manual
May 06 14:00:12 <TobSnyder1> e.g INSANE_SKIP_${PN} += "already-stripped"
May 06 14:02:07 <rburton> yes
May 06 14:18:27 <limotec-dev> does anyone have experience with cups for arm?
May 06 14:23:06 <LetoThe2nd> limotec-dev: care to elaborate in which ways it might differ from cups on any other platform?
May 06 14:23:36 <limotec-dev> we are trying to install cups on a arm board with cortexA9 processor, we were able to build a yocto image with the cups package but we arent able to print anything
May 06 14:24:01 <limotec-dev> the error cups give is : Unsupported format "application/vnd.cups-banner".
May 06 14:24:14 <lpapp> bluelightning: since I could not get more help with the kernel boot issue on the openembedded core mailing list, what shall I try next, the linux kernel and gcc mailing lists?
May 06 14:24:31 <limotec-dev> we already googled this error but we dont't find a fix for this
May 06 14:24:51 <bluelightning> lpapp: you did see Gary's reply, right?
May 06 14:26:29 <LetoThe2nd> limotec-dev: concerning yocto, i'd check http://permalink.gmane.org/gmane.comp.printing.cups.general/28106
May 06 14:26:48 <bluelightning> lpapp: http://lists.openembedded.org/pipermail/openembedded-core/2015-April/104352.html
May 06 14:26:53 <LetoThe2nd> limotec-dev: e.g., get further diagnostics first
May 06 14:27:34 <imrehg> Hi, I have someone reporting a compile error with the yocto BSP package our engineers made, but I can't reproduce it. For me compiles all good, for the person reports errors
May 06 14:28:04 <imrehg> here's the log apparently: http://pastebin.com/eR7SaP2s  any ideas what might be the root cause, any similar things people seen?
May 06 14:28:44 <imrehg> "ERROR: Error in compiling python function in <SNIP>/sources/poky/meta/classes/base.bbclass"
May 06 14:29:09 <LetoThe2nd> imrehg: um, are you sure that the person reporting is really using the same layers/revisions/etc. as you do?
May 06 14:29:36 <LetoThe2nd> imrehg: like, bsp is for dizzy, person is using dora or dylan
May 06 14:29:46 <bluelightning> imrehg: at a guess, it does not like the brackets in the path
May 06 14:30:10 <bluelightning> parentheses, that is ()
May 06 14:30:32 <imrehg> he says we are working from the same zip bundle, same OS (Ubuntu 12.04.5), same packages installs, dora
May 06 14:30:34 <LetoThe2nd> bluelightning: ++
May 06 14:30:41 <imrehg> yeah.... I can see that.
May 06 14:30:59 <imrehg> I was wondering why he extracted it like that, that's a difference for me.
May 06 14:31:06 <imrehg> *ouch*
May 06 14:31:07 <bluelightning> logically it should not care, but it is incorporating the path into the function name
May 06 14:31:25 <LetoThe2nd> we should all revert to 8.3 naming. would avoid so many problems. only ansi 7bit, only lowercase.
May 06 14:31:48 <bluelightning> I wouldn't be surprised if brackets in the path caused other issues later in the build though
May 06 14:32:55 <imrehg> yup, just reporocuded it
May 06 14:33:04 <imrehg> *reproduced
May 06 14:33:21 <imrehg> thanks a lot!
May 06 14:34:30 <imrehg> guess I have more ammunition now to tell the engineers to leave the darn brackets out of the package filename too
May 06 14:47:08 <limotec-dev> LetoThe2nd: thankyou, we will try this tomorrow!
May 06 15:32:28 <lpapp> bluelightning: actually, it did not come up in my inbox, strangely enough :(
May 06 15:32:50 <lpapp> bluelightning: thanks, will try!
**** ENDING LOGGING AT Thu May 07 02:59:58 2015