**** BEGIN LOGGING AT Wed May 23 03:00:03 2018
**** BEGIN LOGGING AT Wed May 23 08:27:17 2018
May 23 09:33:28 <yocti> New news from stackoverflow: how to load kernel modules on boot? <https://stackoverflow.com/questions/50484695/how-to-load-kernel-modules-on-boot>
May 23 09:36:45 <LetoThe2nd> nathani__: sorry, but would you mind fixing your connection some time? we've had lots and lots of join/quit messages from you the last days and weeks.
May 23 09:40:01 <dman_> Hi, I am starting to write a new utility and would like to include it as part of my linux image when building with Yocto, what would be the recommended build system to use for my apoplication? cmake, automake,...?
May 23 09:42:10 <LetoThe2nd> dman_: whatever fits your needs, and is already supported. obvious choices would be autotools or cmake, and lately meson too.
May 23 09:44:48 <dman_> LetoThe2nd: Thanks for the heads up, I am glad cmake is supported.
May 23 09:45:55 <LetoThe2nd> dman_: be aware that if you want or need to support older releases, then there might be an older cmake version provided. but thats about it.
May 23 09:46:49 <dman_> LetoThe2nd: Ah sweet, I'll keep that in mind.
May 23 09:52:50 <open-nandra> Hello, I'm using core-image-minimal from rocko build for BBB with enabled systemd and I see that sometimes it too too long board to boor and reason is that systemd timeout on waiting for non-existing node like: [ TIME ] Timed out waiting for device dev-mmcblkp1.device.
May 23 09:52:50 <open-nandra> [DEPEND] Dependency failed for /boot.
May 23 09:52:50 <open-nandra> [DEPEND] Dependency failed for Local File Systems.
May 23 09:52:57 <open-nandra> any ideas what can cause this
May 23 09:53:18 <open-nandra> obviously  dev-mmcblkp1.device should be dev-mmcblk0p1.device
May 23 09:53:33 <open-nandra> when remove line with this mountpoint from fstab board boots fine
May 23 09:54:03 <mckoan> open-nandra: same question here, I wasn't able to find who is adding such line at the end of fstab
May 23 09:54:41 <open-nandra> mckoan: but it's added runtime and before systemd was disabled board boots fine
May 23 09:54:50 <open-nandra> so I suspect some systemd thing add this
May 23 09:57:05 <open-nandra> mckoan: maybe systemd-fstab-generator?
May 23 09:57:26 <open-nandra> but it shouldn't adjust fstab
May 23 09:57:28 <open-nandra> puzzled :(
May 23 12:07:45 <mckoan> open-nandra: no clues
May 23 12:08:10 <open-nandra> mckoan: I checked systemd and they said systemd is not writing to fstab
May 23 12:09:26 <open-nandra> mckoan: Found it
May 23 12:09:38 <open-nandra> I used *.wic output to burn directly to SD card
May 23 12:09:49 <open-nandra> when use preformatted and only copy rootfs it was working
May 23 12:10:07 <open-nandra> when mount rootfs from wic it contains this wrong line which prevent to boot
May 23 12:10:21 <open-nandra> so need to investigate further but it's definitely yocto problem
May 23 12:16:11 <mckoan> open-nandra: in the correspondent .wks there is no evicence of the creation of such mountpoint though
May 23 12:17:10 <open-nandra> mckoan: ok but if you llok to generated rootfs.tar.gz in fstab it's not there
May 23 12:17:20 <open-nandra> so maybe also something else
May 23 12:18:15 <open-nandra> mckoan: where do you find wks
May 23 12:20:01 <mckoan> open-nandra: find -name '*wks'
May 23 12:21:48 <stoke_> Hey, this is probably a pretty simple question, but I'm missing something. I'm trying to create a new recipe based on a private repo I hold. I run "devtool add recipe URI;branch=branch-name" and devtool happily creates a recipe in my workspace layer. However, devtool fetches the latest commit from the repository rather than the commit from the branch that i specified earlier. I've tried modifying SRC_REV to the hash of the commit I wa
May 23 12:21:56 <stoke_> when I run devtool build recipe
May 23 12:23:11 <stoke_> is there something else I'm supposed to do when trying to access a specific commit?
May 23 12:24:23 <open-nandra> Imckoan: look at meta-yocto-bsp/wic/beaglebone.wks
May 23 12:24:35 <open-nandra> it's imo caused by first line in this file
May 23 12:24:58 <open-nandra> just guess though
May 23 12:33:19 <mckoan> stoke_: SRCREV = "numberofyourcommit"
May 23 12:33:57 <stoke_> mckoan: That's what I've got set
May 23 12:34:00 <mckoan> open-nandra: not in my case
May 23 12:34:29 <stoke_> but when I run devtool build recipe the source doesn't change to match the commit
May 23 12:34:35 <mckoan> stoke_: probably not
May 23 12:38:05 <open-nandra> mckoan: what do you mean, you don't have same wic file?
May 23 12:38:28 <stoke_> mckoan: Well I dont know what to tell you. I've copied my commit sha directly from gitlab to my .bb and verified it. It's definitly set to the commit I want
May 23 12:38:33 <stoke_> still have old source though
May 23 12:39:31 <mckoan> open-nandra: I do not have any evidence of the creation of the last line in fstab
May 23 12:39:54 <mckoan> stoke_: pastebin the recipe if you can
May 23 12:40:11 <mckoan> open-nandra: I'm not using BBB
May 23 12:41:10 <mckoan> open-nandra: my last line is "/dev/mmcblkp1 /boot   vfat    defaults        0       0"
May 23 12:42:00 <stoke_> mckoan: take a look https://pastebin.com/72P5C2Pk
May 23 12:46:51 <mckoan> stoke_: it should work as is
May 23 12:49:47 <open-nandra> mckoan: I have the same
May 23 13:22:09 <open-nandra> mckoan: The --no-fstab-update option is used to doesn't change fstab file. When
May 23 13:22:09 <open-nandra>     using this option the final fstab file will be same that in rootfs and
May 23 13:22:10 <open-nandra>     wic doesn't update file, e.g adding a new mount point. User can control
May 23 13:22:10 <open-nandra>     the fstab file content in base-files recipe.
May 23 13:22:12 <open-nandra> from wic manual
May 23 14:32:22 <svenn> is there a way to perform a check on which user is building the .bb file? Like if user=myname then src_uri= else src_uri= endif
May 23 14:34:34 <yocti> New news from stackoverflow: Yocto rtsp stream <https://stackoverflow.com/questions/50490512/yocto-rtsp-stream>
May 23 14:35:52 <LetoThe2nd> svenn: sounds like a pretty bad idea, care to elaborate on the usecase?
May 23 14:45:24 <svenn> jenkins access is different than regular user access
May 23 14:45:50 <svenn> jenkins use svn+ssh while users use https
May 23 14:47:00 <svenn> so the source code is the same, it is just a different transport for one particular user which is a build server
May 23 15:17:31 <mckoan> open-nandra: ths is the guilty : scripts/lib/wic/canned-wks/sdimage-bootpart.wks
May 23 15:45:18 <stefans__> Hi
May 23 15:45:45 <stefans__> I have this problem: do_package_write_rpm: The recipe libsdl2 is trying to install files into a shared area when those files already exist. Those files and their manifest location are:
May 23 15:46:20 <stefans__> The thing is that I've had a recipe called libsdl2-something and i've renamed that to libsdl2 - now when I'm building it complains about identical files in tmp/deploy/rpm
May 23 15:46:57 <stefans__> I understand why, but I want to workaround it without wiping tmp, or cleaning (ie. I'm commiting the rename and I'll break the build for a lot of people)
May 23 15:48:22 <stefans__> Is adding the filenames to SSTATE_DUPWHITELIST a good idea? Can I force from the recipe a clean of the older recipe? Or maybe even keep a dummy recipe with the older name and do sometricks?
May 23 16:04:52 <yocti> New news from stackoverflow: Change boot image on Yocto using Toaster <https://stackoverflow.com/questions/50492391/change-boot-image-on-yocto-using-toaster>
May 23 16:34:57 <yocti> New news from stackoverflow: Accessing initramfs-tools features in Yocto generated initramfs image <https://stackoverflow.com/questions/50492990/accessing-initramfs-tools-features-in-yocto-generated-initramfs-image>
May 23 16:36:08 <armpit> RP, stable/rocko-next good to go
May 23 16:39:44 <RP> armpit: thanks. Will probably trigger a release build
May 23 16:39:59 <armpit> k
May 23 16:47:46 <RP> rburton: fired a 2.4.3 build
May 23 16:48:57 <rburton> cool
May 23 17:15:26 <kergoth> so i stopped trying to use docker for windows drive sharing in favor of trying to use a samba container to share a volume back with windows instead, flipping it around, but that doesnt' seem to work either, likely due to how they're doing the host networking with user proxying.. https://github.com/gdiepen/volume-sharer/issues/3
May 23 17:15:27 <kergoth> hrmph
May 23 17:15:50 <kergoth> obviously could do *everything* inside the container, but it's a pain to set up dotfiles and editor configs and ssh keys and whatnot..
May 23 17:15:56 <kergoth> guess i'll stick to a VM for now
May 23 17:18:15 * kergoth would really rather be able to just do the build in docker, ideally transparently, while editing and dthe like  outside the container..doable from linux,  less so from windows :)
May 23 17:35:43 <otavio[m]> bluelightning1: are you there?
May 23 18:50:43 <RP> otavio[m]: I suspect bluelightning is asleep atm
May 23 18:52:06 <otavio[m]> RP: :-) I need to drop one layer from layer index. I have right to approve new layers but not drop them
May 23 18:55:01 <RP> otavio[m]: you could email...
May 23 18:55:47 <otavio[m]> Ok; will di
**** BEGIN LOGGING AT Wed May 23 19:14:08 2018
May 23 19:31:51 <radsquirrel> how do people feel about foo_git.bb vs foo_2.3.4.bb?
May 23 19:32:05 <radsquirrel> it seems like the convention is to use _git when the recipe tracks head
May 23 19:32:23 <radsquirrel> but if the recipe doesn't track head some use _git others use _x.y.z
May 23 19:32:30 <radsquirrel> is there a right answer?
May 23 19:35:35 <yocti> New news from stackoverflow: Yocto - Add Custom Driver <https://stackoverflow.com/questions/48423446/yocto-add-custom-driver>
May 23 19:41:35 <RP> radsquirrel: version is probably more obvious IMO
May 23 19:42:35 <radsquirrel[m]> RP: thanks.  _git preferred in the tracking-head case though right?
May 23 19:42:52 <radsquirrel[m]> so _git for tracking head and _x.y.z for tracking x.y.z
May 23 20:35:46 <yocti> New news from stackoverflow: Process with a dbus signal filter no longer receives signal <https://stackoverflow.com/questions/50496755/process-with-a-dbus-signal-filter-no-longer-receives-signal>
May 23 21:10:47 <RP> radsquirrel[m]: yes
May 23 21:11:58 <radsquirrel[m]> thanks!
May 23 21:31:47 <nathani_> hello
May 23 21:32:20 <nathani_> Anyone encountered this error: /home/dev/valid-poky-agl_sdk/tmp/sysroots/nitrogen6x/usr/include/stdlib.h:183:17: error: missing binary operator before token "("
May 23 21:32:20 <nathani_>  #if __GLIBC_USE (IEC_60559_BFP_EXT)
May 23 21:35:57 <yocti> New news from stackoverflow: How to configure Yocto / Boost to support Python 2? <https://stackoverflow.com/questions/50497738/how-to-configure-yocto-boost-to-support-python-2>
**** ENDING LOGGING AT Thu May 24 03:00:03 2018