Monday, 2019-04-29

*** HyP3r <HyP3r!~HyP3r@andreas-fendt.de> has quit IRC00:00
*** HyP3r <HyP3r!~HyP3r@andreas-fendt.de> has joined #yocto00:00
*** bluelightning_ <bluelightning_!~paul@pdpc/supporter/professional/bluelightning> has joined #yocto01:00
*** bluelightning <bluelightning!~paul@pdpc/supporter/professional/bluelightning> has quit IRC01:04
*** fatalhalt <fatalhalt!~fatalhalt@c-67-163-60-93.hsd1.il.comcast.net> has quit IRC05:02
*** agust <agust!~agust@p54833389.dip0.t-ipconnect.de> has joined #yocto05:45
yoctiNew news from stackoverflow: How to execute a script in swupdate image recipe in Yocto? <https://stackoverflow.com/questions/55897498/how-to-execute-a-script-in-swupdate-image-recipe-in-yocto>06:05
*** lfa_ <lfa_!~lfa@213-47-163-50.cable.dynamic.surfer.at> has joined #yocto06:05
*** lfa <lfa!~lfa@213-47-163-50.cable.dynamic.surfer.at> has quit IRC06:08
*** lfa_ <lfa_!~lfa@213-47-163-50.cable.dynamic.surfer.at> has quit IRC06:12
*** xtron <xtron!~xtron@110.93.212.98> has joined #yocto06:13
yoctiNew news from stackoverflow: Where can I find Yocto "sub"-packages not listed in OpenEmbedded layer index like "kbd-keymaps"? <https://stackoverflow.com/questions/55668617/where-can-i-find-yocto-sub-packages-not-listed-in-openembedded-layer-index-lik>06:35
*** lfa <lfa!~lfa@217.19.35.51> has joined #yocto06:35
*** jmiehe <jmiehe!~Thunderbi@p578c106e.dip0.t-ipconnect.de> has joined #yocto06:46
*** sk_tandt <sk_tandt!~sk_tandt@net-5-88-141-17.cust.vodafonedsl.it> has joined #yocto06:48
*** nayfe <nayfe!uid259604@gateway/web/irccloud.com/x-tmbvbaylmdypbwci> has joined #yocto06:55
*** varjag <varjag!~user@122.62-97-226.bkkb.no> has joined #yocto06:58
*** lfa <lfa!~lfa@217.19.35.51> has quit IRC07:08
*** lfa <lfa!~lfa@217.19.35.51> has joined #yocto07:08
*** cvasilak <cvasilak!~cvasilak@2a02:587:811e:db00:b8ba:4618:a687:a37b> has joined #yocto07:11
*** AndersD <AndersD!~AndersD@h-98-128-162-82.NA.cust.bahnhof.se> has joined #yocto07:12
*** vineela <vineela!~vtummala@134.134.139.76> has joined #yocto07:13
*** lusus <lusus!~lusus@62.91.23.180> has joined #yocto07:13
*** lucaceresoli <lucaceresoli!~lucaceres@45.11.168.109.cust.ip.kpnqwest.it> has joined #yocto07:25
*** tprrt <tprrt!~tprrt@217.114.201.133> has joined #yocto07:29
*** yann <yann!~yann@lfbn-1-3372-5.w90-127.abo.wanadoo.fr> has quit IRC07:35
*** florian <florian!~florian_k@Maemo/community/contributor/florian> has joined #yocto07:42
*** angelo_ts is now known as ad__07:44
*** AndersD <AndersD!~AndersD@h-98-128-162-82.NA.cust.bahnhof.se> has quit IRC07:44
*** AndersD <AndersD!~AndersD@h-98-128-162-82.NA.cust.bahnhof.se> has joined #yocto07:45
*** Bunio_FH <Bunio_FH!~bunio@81-18-201-214.static.chello.pl> has joined #yocto07:55
*** prabhakarlad <prabhakarlad!~prabhakar@194.75.40.178> has joined #yocto07:56
*** malanecora <malanecora!b23cc82c@gateway/web/freenode/ip.178.60.200.44> has joined #yocto07:58
*** vineela <vineela!~vtummala@134.134.139.76> has quit IRC08:06
*** BCMM <BCMM!~BCMM@unaffiliated/bcmm> has joined #yocto08:10
*** yann <yann!~yann@85.118.38.73> has joined #yocto08:11
*** raem <raem!5bc64f1c@gateway/web/freenode/ip.91.198.79.28> has joined #yocto08:25
*** jeanba <jeanba!~jbl@213.83.168.178> has joined #yocto08:25
*** jeanba <jeanba!~jbl@213.83.168.178> has left #yocto08:26
raemI'm trying to install p11tool to my system. I was hoping that it is provided in the gnutls package (tried to include gnutls as well as gnutls-bin in my image)08:28
raemam i missing something?08:28
LetoThe2ndraem: maybe this? http://cgit.openembedded.org/openembedded-core/tree/meta/recipes-support/p11-kit/p11-kit_0.23.15.bb?h=master08:29
raemthat's p11-kit not p11tool (which is also requried and i already have it in my image)08:30
LetoThe2ndraem: was just an idea :)=08:32
*** bluelightning <bluelightning!~paul@pdpc/supporter/professional/bluelightning> has joined #yocto08:32
*** bluelightning_ <bluelightning_!~paul@pdpc/supporter/professional/bluelightning> has quit IRC08:37
*** malanecora <malanecora!b23cc82c@gateway/web/freenode/ip.178.60.200.44> has quit IRC08:39
bluelightningraem: try oe-pkgdata-util find-path '*/p11tool'08:44
miwaI'm getting a QA error building nativesdk-gnu-config, Files/Directories were installed but not shipped in any package, is this dependent on my host distribution and that's why the error is occuring? But shouldn't do_install work the same on all hosts regardless of distribution?08:55
raembluelightning: thanks for the tip, that's a really nice util tool i didn't know. However doesn't find anything. I now tried to included it via a .bbappend with `FILES_${PN}-bin += "${bindir}/p11tool"`08:57
varjaglong shot, but does anyone has a layer that makes sumo build on ubuntu bionic08:58
varjag?08:58
bluelightningraem: I don't think that's going to help on its own, if the file isn't in another package within that recipe already08:58
varjagright now a bunch of native recipes fail to build (glibc version conflicts and such) so i have to shadow them manually08:59
raembluelightning: yeah didn't work :). it's just weird that the binary is provided in the gnutls package but somehow not everything packed in the recipe09:00
raemis there a way to see what's going on during the build process of the gnutls recipe. to figure out if files are missing etc?09:01
varjagrelated, what's the easiest way to drop qemu?09:01
bluelightningraem: you could have a look at the logs for gnutls - find the workdir using: bitbake -e gnutls | grep ^WORKDIR=09:02
bluelightningraem: then look under temp/ under that directory09:02
raembluelightning: thanks, will try this out09:03
bluelightningvarjag: not sure you can easily... a good start would be to build for a non-qemu MACHINE, but that's only part of the dependency09:03
varjagi do build for non-qemu machine yes09:05
varjageverything was fine while the build box was 16.04 ununtu09:06
varjagnow it's retired and i have to move my toys to my workstation09:06
varjagwhich is 18.0409:06
varjagqemu then gets fussy about glibc09:06
varjagnow one solution would be to build it in a vm like it's common with legacy toolchains09:07
varjagbut i haven't even released the product yet, this would be humiliating :)09:07
varjagso what i did now, made a transition layers with recipes-devtools09:09
*** jeanba <jeanba!~jbl@213.83.168.178> has joined #yocto09:09
varjagand populate it with newer versions of conflicting packages09:09
paulbarkervarjag: I'd recommend spinning up an Ubuntu 16.04 container. The glibc update from 16.04 -> 18.04 is a bit of a jump09:09
*** jeanba <jeanba!~jbl@213.83.168.178> has left #yocto09:10
varjagi see09:10
varjagwarrior, is it expected to work on bionic?09:10
*** tprrt <tprrt!~tprrt@217.114.201.133> has quit IRC09:11
paulbarkerI bailed out after 3 full days of work trying to get an older branch (around Yocto Project 2.0) working on Ubuntu 18.04. Would have taken at least an extra week or two of full time effort which just wasn't worth it09:11
paulbarkerthud should work under Ubuntu 18.0409:11
varjaghm.. tried moving qemu from thud into my recipes-devtools, but there's hard dependency to quemu 2.10.1 someplace09:12
varjagmaybe i should just move my project to thud..09:13
paulbarkerYea, you're going to have a long string of dependencies. By the time you've got it compatible you'll have changed a lot of core recipes and have a Frankenstein's monster that no-one else has tested09:13
varjagpoint09:13
paulbarkerI use LXD containers for a pretty lightweight way to run older Ubuntu versions within an Ubuntu 18.04 base image. You can also use Docker if that's your thing09:14
paulbarkerBuilds are slower in VMs, I think due to increased I/O latency, but builds in containers seem to be just as fast as outside09:15
varjagi didn't use docker for anything yet09:15
varjagdid have build environments in plain chroots tho09:15
*** tprrt <tprrt!~tprrt@217.114.201.133> has joined #yocto09:23
*** nayfe <nayfe!uid259604@gateway/web/irccloud.com/x-tmbvbaylmdypbwci> has quit IRC09:35
*** rburton <rburton!~rburton@35.106.2.81.in-addr.arpa> has joined #yocto09:43
bluelightningraem: looks like we configure gnutls with --without-p11-kit by default so that's probably why it's not there09:59
bluelightningraem: there's a PACKAGECONFIG for it though so you'd only need to add PACKAGECONFIG_append_pn-gnutls = " p11-kit" to your configuration10:01
*** raem <raem!5bc64f1c@gateway/web/freenode/ip.91.198.79.28> has quit IRC10:13
*** conso <conso!~conso@p57B951F7.dip0.t-ipconnect.de> has joined #yocto10:21
erboI use the crops/poky docker containers for building. It's so nice to not have to worry about which build host environment I'm using. I can really recommend that, and I cannot notice any performance penalty.10:27
*** AndersD_ <AndersD_!~AndersD@h-98-128-162-82.NA.cust.bahnhof.se> has joined #yocto10:30
*** raem <raem!5bc64f1c@gateway/web/freenode/ip.91.198.79.28> has joined #yocto10:31
*** AndersD <AndersD!~AndersD@h-98-128-162-82.NA.cust.bahnhof.se> has quit IRC10:32
raembluelightning: thanks for the tip, just found out also that p11kit is missing. Now I also know how PACKAGECONFIG is working :)10:44
*** jeanba <jeanba!~jbl@213.83.168.178> has joined #yocto10:57
*** jeanba <jeanba!~jbl@213.83.168.178> has left #yocto10:57
*** rburton <rburton!~rburton@35.106.2.81.in-addr.arpa> has quit IRC11:07
*** rburton_ <rburton_!~rburton@35.106.2.81.in-addr.arpa> has joined #yocto11:07
*** AndersD_ <AndersD_!~AndersD@h-98-128-162-82.NA.cust.bahnhof.se> has quit IRC11:07
*** AndersD <AndersD!~AndersD@h-98-128-162-82.NA.cust.bahnhof.se> has joined #yocto11:09
*** nayfe <nayfe!uid259604@gateway/web/irccloud.com/x-ynbugumjhfeazvlh> has joined #yocto11:09
*** saraf <saraf!~a_saraf@123.252.238.18> has joined #yocto11:34
*** berton <berton!~berton@177.194.203.109> has joined #yocto11:41
rburton_kanavin: please send a patch changing perl owner to you. i'll give you actual money.11:43
LetoThe2ndrburton_: in a valid currency? ;-)11:45
rburton_sure11:46
* LetoThe2nd had expected Italian Lira or such.11:46
*** berton <berton!~berton@177.194.203.109> has quit IRC11:49
*** saraf <saraf!~a_saraf@123.252.238.18> has quit IRC11:50
RPrburton_: haha :)11:52
*** berton <berton!~berton@177.194.203.109> has joined #yocto11:53
*** jeanba <jeanba!~jbl@213.83.168.178> has joined #yocto11:56
yoctiNew news from stackoverflow: How can I auto load kernel module on boot in Yocto? (like "insmod /xxx/oo.ko") <https://stackoverflow.com/questions/55902608/how-can-i-auto-load-kernel-module-on-boot-in-yocto-like-insmod-xxx-oo-ko>12:06
*** jeanba <jeanba!~jbl@213.83.168.178> has quit IRC12:06
*** jeanba <jeanba!~jbl@213.83.168.178> has joined #yocto12:07
*** jeanba <jeanba!~jbl@213.83.168.178> has left #yocto12:11
*** tprrt <tprrt!~tprrt@217.114.201.133> has quit IRC12:29
*** tprrt <tprrt!~tprrt@217.114.201.133> has joined #yocto12:32
paulbarkerFinally tracked down some recent build issues. The golang compiler sets 0500 permissions on directories for cached dependencies12:33
paulbarkerThat's professionally helpful12:33
*** jeanba <jeanba!~jbl@87-49-146-140-mobile.dk.customer.tdc.net> has joined #yocto12:35
*** jeanba <jeanba!~jbl@87-49-146-140-mobile.dk.customer.tdc.net> has left #yocto12:35
paulbarkerhttps://github.com/golang/go/issues/27161#issuecomment-433102849 *facepalm*12:42
zeddiiwhee12:43
rburton_fun12:46
paulbarkerI'm just going to have to add a recursive chmod to the end of do_compile for the affected package (netns)12:48
zeddiiyup. sometimes the best thing to do is just work around it.12:48
paulbarkerIt doesn't seem to be an issue for most go recipes and only affects netns after updating to >= v0.5.012:48
paulbarkerNot sure it's something we can fix universally as the module cache ends up in different places for different recipes12:49
paulbarkerI think it's just one more Feature (TM) of the go compiler12:50
zeddiiheh. I was about to say that.12:50
zeddiia never ending source of fun.12:50
paulbarkerLike using a tmpdir for intermediate object files and unconditionally removing them when the `go` tool exits so you can't debug any linker errors12:50
zeddiiwhat is this debug you speak of ? "full stack" go developers never have such issues.12:51
*** AndersD <AndersD!~AndersD@h-98-128-162-82.NA.cust.bahnhof.se> has quit IRC12:52
paulbarkerThings do seem to work if you don't use cgo12:52
paulbarkerAll the interesting software needs cgo to do some magic though12:52
zeddiiyah, when I've had to patch multiple makefiles to even allow CGO to be enabled, you know you are off the beaten path :D12:53
* zeddii wonders why on his brand new build server that icu-native just went boom.12:54
paulbarkerI think I'm permanently scarred after trying to debug the C code burried in runc12:54
zeddiiheh. no doubt.12:54
*** erakis <erakis!~erakis@199.58.239.58> has joined #yocto12:56
paulbarkerMy new advice when looking at an issue is, if `git grep longjmp` shows anything just run away immediately and pretend you never saw the issue12:57
LetoThe2ndpaulbarker: and 'git grep shortjmp' ;-)13:03
LetoThe2ndpaulbarker: i'm just preparing sstate and stuff for the show tomorrow ;-)13:03
paulbarkerLetoThe2nd: Excellent, will be around for that13:03
LetoThe2ndpaulbarker: \m/13:05
* zeddii reverts an april 24th icu version bump to see if that helps13:07
RPzeddii: new servers are such fun13:11
LetoThe2ndRP: heh, we literally ordered new gear for ~15k 5 minutes ago13:12
*** xtron <xtron!~xtron@110.93.212.98> has quit IRC13:14
zeddiiindeed. I'm seeing an icu-native build failure that clearly no one else is, and one that I'm not seeing on my "broken in" servers. Reverting the april version bumps has me building again. I'll take a closer look shortly.13:14
*** bluelightning <bluelightning!~paul@pdpc/supporter/professional/bluelightning> has quit IRC13:25
*** leitao <leitao!~leitao@2620:10d:c092:200::1:9682> has joined #yocto13:32
leitaohow do I know which package is adding a usergroup in my final image?13:33
RPleitao: grep?13:36
leitaoRP: well, I tried that, grepping for ADDGROUP, but no clue.  It seems someone is creating a 'video' usergroup. I am interested who.  I am wondering if it is being created using a different way13:37
JPEWleitao: base-passwd maybe?13:39
RPleitao: start be looking at base-passwd which has the basics13:39
leitaook, let me take a look there. Thanks13:39
*** AndersD <AndersD!~AndersD@h83-209-191-235.cust.a3fiber.se> has joined #yocto13:56
*** AndersD_ <AndersD_!~AndersD@h-98-128-162-82.NA.cust.bahnhof.se> has joined #yocto13:58
*** AndersD <AndersD!~AndersD@h83-209-191-235.cust.a3fiber.se> has quit IRC14:00
*** JPEW <JPEW!cc4da337@gateway/web/freenode/ip.204.77.163.55> has quit IRC14:06
*** JPEW <JPEW!cc4da337@gateway/web/freenode/ip.204.77.163.55> has joined #yocto14:07
*** florian <florian!~florian_k@Maemo/community/contributor/florian> has quit IRC14:07
*** andycooper <andycooper!uid246432@gateway/web/irccloud.com/x-fppdaarbqlonxybn> has joined #yocto14:09
*** varjag <varjag!~user@122.62-97-226.bkkb.no> has quit IRC14:36
yatesif there are multiple patch files for a recipe, with an associated .bbappend, do the patches get applied in the order in whcih they are listed in the SRC_URI?14:51
rburton_yes14:52
rburton_the order that the src_uri is evaluated to14:52
rburton_its literally for patch in src_uri14:52
yatesok14:53
yatesso this becomes important if the same file is patched in more than one .patch file?14:53
yates(probably not a good situation)14:56
erakisHi, I'm running bitbake on fido branch and I try adding the `vim` package to my image. But, I'm getting the error `No checksum specified for... ../downloads/vim_vim.googlecode.com_.hg._v7-4-│meta-intelia14:56
erakis481.tar.gz, please add at least one to the recipe:`. I can't upgrade from fido. Does anyone have a solution?14:56
LetoThe2nderakis: um, do what the error message tell you to do?15:00
LetoThe2nderakis: -> add a checksum for the download15:00
erakis@LetoThe2nd: Is this means that I need to modify the vim recipe itself ? Or I can add a patch somewhere else in my meta branch ?15:01
LetoThe2nderakis: you can either copy it over into your own layer and modify it there, or use a bbappend.15:02
*** berton <berton!~berton@177.194.203.109> has quit IRC15:05
*** AndersD_ <AndersD_!~AndersD@h-98-128-162-82.NA.cust.bahnhof.se> has quit IRC15:07
*** berton <berton!~berton@177.194.219.152> has joined #yocto15:07
*** berton_ <berton_!~berton@177.194.219.152> has joined #yocto15:09
*** berton <berton!~berton@177.194.219.152> has quit IRC15:12
erakis@LetoThe2nd: So I can add the file `vim_7.4.481.bbappend` into `meta-mymeta/recipes-core/image-myimage/files/vim_7.4.481.bbappend` and add the `FILESEXTRAPATHS_prepend := "${THISDIR}/${PN}:"`, `SRC_URI[md5sum]=... SRC_URI[sha256sum]=...` into ? But what do I have to put into my recipe so that the bbappend be applied ?15:15
*** sk_tandt_ <sk_tandt_!~sk_tandt@net-5-88-141-17.cust.vodafonedsl.it> has joined #yocto15:15
*** sk_tandt <sk_tandt!~sk_tandt@net-5-88-141-17.cust.vodafonedsl.it> has quit IRC15:18
*** Bunio_FH <Bunio_FH!~bunio@81-18-201-214.static.chello.pl> has quit IRC15:27
erakis@LetoThe2nd: Ok I got it, I have to put it in `meta-mymeta/recipes-support/vim/vim_7.4%.bbappend`.15:29
*** cvasilak <cvasilak!~cvasilak@2a02:587:811e:db00:b8ba:4618:a687:a37b> has quit IRC15:31
*** sk_tandt_ <sk_tandt_!~sk_tandt@net-5-88-141-17.cust.vodafonedsl.it> has quit IRC15:36
yoctiNew news from stackoverflow: Yocto, Fixing Computing transaction error in do_rootfs? <https://stackoverflow.com/questions/55906525/yocto-fixing-computing-transaction-error-in-do-rootfs> || What is the new IMAGE_DEPENDS_wic_append_suffix notation? <https://stackoverflow.com/questions/55906202/what-is-the-new-image-depends-wic-append-suffix-notation>15:36
*** leitao <leitao!~leitao@2620:10d:c092:200::1:9682> has quit IRC15:56
JPEWHas anyone given thought to making qemux86 be i686 instead of i586?15:57
*** jmiehe <jmiehe!~Thunderbi@p578c106e.dip0.t-ipconnect.de> has quit IRC15:59
yoctiNew news from stackoverflow: meta-marvell update do_image_ [depends] with MACHINE override <https://stackoverflow.com/questions/55358358/meta-marvell-update-do-image-typedepends-with-machine-override>16:06
*** raem <raem!5bc64f1c@gateway/web/freenode/ip.91.198.79.28> has quit IRC16:08
*** lusus <lusus!~lusus@62.91.23.180> has quit IRC16:17
*** yann <yann!~yann@85.118.38.73> has quit IRC16:24
*** conso <conso!~conso@p57B951F7.dip0.t-ipconnect.de> has quit IRC16:59
*** moosnat__ <moosnat__!~moosnat@unaffiliated/moosnat> has joined #yocto17:15
*** moosnat_ <moosnat_!~moosnat@unaffiliated/moosnat> has joined #yocto17:17
*** moosnat__ <moosnat__!~moosnat@unaffiliated/moosnat> has quit IRC17:20
*** comptroller <comptroller!~comptroll@47-213-222-253.paolcmtc01.res.dyn.suddenlink.net> has quit IRC17:30
yoctiNew news from stackoverflow: Config parsing error on xl create for Xen Guest DomU <https://stackoverflow.com/questions/55907917/config-parsing-error-on-xl-create-for-xen-guest-domu>17:37
*** WillMiles <WillMiles!~Will@static-209-87-231-80.storm.ca> has joined #yocto17:39
*** comptroller <comptroller!~comptroll@47-213-222-253.paolcmtc01.res.dyn.suddenlink.net> has joined #yocto17:46
erakisHi, I would like to add alias for all users. Ex: alias ls='ll='ll -la'. What is the proper way ? Adding them from IMAGE_POSTPROCESS_COMMAND is a good choice ? Or overriding base-files with a bbappend ?18:04
*** likewise <likewise!uid442@beta.alwyzon.com> has quit IRC18:17
*** tprrt <tprrt!~tprrt@217.114.201.133> has quit IRC18:18
JPEWerakis: We have a bbappend for base-files that adds them to .profile19:02
JPEWor .bashrc19:03
*** BCMM <BCMM!~BCMM@unaffiliated/bcmm> has quit IRC19:03
erakis@JPEW: I find this https://git.yoctoproject.org/cgit.cgi/poky/plain/meta/recipes-core/base-files/base-files/share/dot.bashrc. But I'm on sh. So If I well understand, I should override https://git.yoctoproject.org/cgit.cgi/poky/plain/meta/recipes-core/base-files/base-files/share/dot.profile with a bbappend file ?19:06
*** bl4ckb0ne <bl4ckb0ne!znc@2604:bf00:710:0:5054:ff:fe94:ef02> has joined #yocto19:07
bl4ckb0nehi, im looking for a recipe for the intel compiler, is there one available ?19:07
JPEWerakis: Ya, thats sounds like the right track19:08
erakis@JPEW: Thank you19:10
erakis@JPEW: But, what will looks like the .bbappend ? meta-mylayer/recipes-core/base-files/base-files/share/dot.profile.bbappend ?19:11
JPEWerakis: Just one line: FILESEXTRAPATHS_prepend := "${THISDIR}/${PN}:"19:12
JPEWoh, sorry19:13
erakis@JPEW: I put "FILESEXTRAPATHS_prepend := "${THISDIR}/${PN}:" as first line and on the second line "alias l='ls -CF'" in this file and this will work ?19:13
JPEWYou need 2 files: meta-mylayer/recipes-core/base-files/base-files.bbappend which has the line I gave you19:13
JPEWand meta-mylayer/recipes-core/base-files/base-files/dot.profile which has the complete .profile you want (it will completely overwrite the core dot.profile)19:14
JPEW.bbappend file only append to .bb files19:15
JPEWSo, you write a .bbappend that prepends to the path that bitbake uses when it looks for "fiel://dot.profile" so it finds yours before the core dot.profile file19:16
erakisThank you so much, that make sense. I need to go, but I will try it later in the evening.19:16
*** rcw <rcw!~rcw@partnervpn-ala.windriver.com> has joined #yocto19:17
rburton_bl4ckb0ne: ISS comes with a meta-iss layer19:19
bl4ckb0nei found a lot of things about compiling a project with an icc toolchain, but nothing about packaging it19:20
rburton_there's totally a layer installed as part of the installer because i was poking at it last month19:20
rburton_https://software.intel.com/en-us/articles/using-intel-system-studio-with-the-yocto-project19:21
bl4ckb0neneat, i think thats exactly what I need19:23
bl4ckb0nethanks!19:23
*** wowwow <wowwow!~w@2600:1700:ce01:1100:dd2c:1275:6373:8ab6> has joined #yocto19:26
bl4ckb0neI also experience trouble with a clone from a git lfs repo, does the recipe uses it own git binary ?19:27
bl4ckb0nethat iss seems to be only for runtime compiler19:45
*** rcw <rcw!~rcw@partnervpn-ala.windriver.com> has quit IRC20:09
*** tprrt <tprrt!~tprrt@upc31-1-78-208-110-13.fbx.proxad.net> has joined #yocto20:13
*** bluelightning <bluelightning!~paul@pdpc/supporter/professional/bluelightning> has joined #yocto20:22
*** nasamuffin <nasamuffin!~nasamuffi@192.157.233.35.bc.googleusercontent.com> has joined #yocto20:35
radsquirrelhi everyone I'm Brad from OpenBMC20:35
radsquirrelhey we are exploring ways to improve developer workflow.20:36
*** EdTanous <EdTanous!~Ed_Tanous@134.134.139.73> has joined #yocto20:36
radsquirrelone idea that came up was some kind of way to include the source code for our applications right in the metadata layer20:36
radsquirrelnot sure how that would look, still exploring.20:36
radsquirrelI vaguely remember reading somewhere that source-code-in-meta isn't recommended but I can't find that anymore.20:37
yoctiNew news from stackoverflow: unable to execute Yocto Extensible sdk (eSDK) script <https://stackoverflow.com/questions/55910465/unable-to-execute-yocto-extensible-sdk-esdk-script>20:37
radsquirrelDoes the yocto community have any opinions to help guide us on this?20:37
radsquirrelthanks!20:37
JPEWradsquirrel: Are you looking for a "temporary" solution, as in the developers want to point a recipe to their local code, do some development with it, then swicth it back to the "offical" recipe when finished?20:39
radsquirrelwell I think there is devtool for that right?20:39
JPEWradsquirrel: Yes, that and externalsrc :)20:40
RParmpit: any idea what I missed in the ltp config of https://autobuilder.yocto.io/pub/non-release/20190429-6/testresults/testresult-report.txt ?20:41
nasamuffinHiya, I'm working on OBMC too - I think the root here is that right now we have ~1 layer per Git repo, and one possible solution is to instead shoot for multiple layers in one git repo20:41
* armpit looks20:41
RParmpit: all the first ltp tests show 0 tests20:41
nasamuffinI'm kind of hoping for something more like 1 layer per git submodule (although, as brad said, still under investigation)20:41
radsquirrelJPEW: One issue we have is that as we implement features and review them, the wip patches are distributed across 3-4 different repositories.20:42
radsquirrelJPEW: maybe a couple application repositories, and couple bitbake meta-layers.20:42
nasamuffinI'm not sure that necessarily means sticking the source right into the distro layer. The manual seems to indicate that it's not necessary to put source into additional SCMs at all20:42
radsquirrelJPEW: with all those repos it can be hard to try out a feature before things have been merged.20:43
bl4ckb0nebut /b 420:43
armpitRP, that is odd because I ran resulttool on the AB and it looked fine20:43
bl4ckb0nefail, sorry20:43
JPEWradsquirrel: I share your pain :) We have the same problem, changes have to trickle through a lot of layers to into a final build20:43
radsquirrelJPEW: well if nothing else it helps to know it is a real pain point.20:44
kergothit should certainly be possible to put a recipe inside a project source directory, and have that source idr as a submodule of a layer with bbfiles set to match those recipes20:44
radsquirreland not just something we are doing wrong.20:44
JPEWradsquirrel: Is it necessary to have the new features immediately? We more or less mitigated the problem with some daily build shenangians20:44
kergothor something along similar lines, but there are downsides to it not being out of band20:44
*** wak-work <wak-work!wak-workma@gateway/shell/matrix.org/x-vtmevctlijfutvvq> has joined #yocto20:45
paulbarkerWe're using git submodules, for our own meta layers I've considered scripting an automatic submodule update commit in the top level repository when there's a merge to master in the layer repository20:45
nasamuffinpaulbarker: do you think you could link your repo? i'd like to see how you've done it20:45
kergothcan look at alternatives to submodules (i.e. repo), or assisting tools to submodules (i.e. git-meta) too20:45
kergothgit-meta is interesting since it lets you do a 'git meta commit' which commits to every modified submodule and commits to the master updating the submodules accordingly with one command20:46
paulbarkernasamuffin: I've considered it, not implemented it yet, our churn isn't fast enough to push me to do it yet20:46
* kergoth needs to try that one20:46
JPEWpaulbarker, radsquirrel: We do effectively the same thing with a nightly build job; it loops through the recipes we care about, checks out the head of their branch, does a few spot builds, then pushes up a change for review (we use gerrit)20:46
paulbarkerBut it should be fairly straightforward with a git hook on each layer repository20:46
nasamuffini mean specifically the git-sm setup, not necessarily the script20:47
radsquirrelJPEW: that is an interesting though.  some kind of tag in the pull request perhaps gets it into a build.20:47
radsquirrels/though/thought20:47
paulbarkernasamuffin: https://gitlab.com/oryx/oryx20:47
nasamuffinty!20:47
paulbarkerWe used to use repo tool but it had too many failure modes20:48
nasamuffinactually if i can put on my git contributor hat quickly i'll state that we're hoping people will stop using repo and start using submodules soon :)20:48
JPEWradsquirrel: Well.... thats where the shenangins come in :) We have a *separate* job that builds those same recipes with ${AUTOREV}.... so if you pull a "daily" build you get those20:48
JPEWAnd the auto-upgrade jobs ensures that the recipe is more-or-less up to date with the daily build20:48
paulbarkernasamuffin: You may find our nightly trigger script interesting. Does a tentative update of all layers and tags the result for a nightly build20:49
nasamuffinpaulbarker: tyvm, i'll be sure to take a look.20:49
paulbarkerhttps://gitlab.com/oryx/nightly-trigger/blob/master/nightly-trigger.sh20:49
paulbarkerStraightforward but fairly effective20:49
radsquirrelJPEW: I follow.20:50
radsquirrelI see an option for us to pursue in that, so thanks.20:51
paulbarkerThe only thing I struggle with in git submodules is mirroring20:51
nasamuffincan you elaborate?20:52
*** jwessel <jwessel!~jwessel@unknown-3-103.windriver.com> has joined #yocto20:52
paulbarkerWe point directly at third party layers in our .gitmodules file. If they move or disappear our old commits of the top-level repository no longer build20:52
radsquirrelkergoth what were the downsides to not being out-of-band?20:53
radsquirrelthats what I thought I had read at one point but could not find again.20:53
kergothmainly overhead, changes which aren't bound to the branch end up having to be cherry picked to every affected source branch. of course the alternatives have their own overhead to deal with, and that could be alleviated with tooling. i don't know of anyone doing much work with metadata within source trees at the moment20:54
radsquirrelkergoth: actually I was still parsing your first...20:54
radsquirrelmeta-data within source?20:55
radsquirrelI think we were looking at it from the other way - source in metadata?20:55
radsquirrelkergoth: will keep trying to wrap my head around that :-)20:58
radsquirrelkergoth paulbarker nasamuffin JPEW thanks for the ideas!20:58
*** WillMiles <WillMiles!~Will@static-209-87-231-80.storm.ca> has quit IRC21:06
*** shauno <shauno!~soneil@pdpc/supporter/professional/shauno> has quit IRC21:08
*** nasamuffin <nasamuffin!~nasamuffi@192.157.233.35.bc.googleusercontent.com> has quit IRC21:12
*** shauno <shauno!~soneil@pdpc/supporter/professional/shauno> has joined #yocto21:14
*** berton_ <berton_!~berton@177.194.219.152> has quit IRC21:21
*** rewitt <rewitt!~rewitt@134.134.139.72> has joined #yocto21:24
RParmpit: https://autobuilder.yoctoproject.org/typhoon/#/builders/95/builds/1/steps/7/logs/step1c - ltp_groups took 30s?!21:37
RParmpit: compared to 8000s in https://autobuilder.yoctoproject.org/typhoon/#/builders/81/builds/161/steps/7/logs/step1c21:38
RParmpit: has to be a missing dependency...21:38
yoctiNew news from stackoverflow: How to tell bitbake to variable's calculate basehash value after specific task? <https://stackoverflow.com/questions/55911423/how-to-tell-bitbake-to-variables-calculate-basehash-value-after-specific-task>22:07
*** wowwow <wowwow!~w@2600:1700:ce01:1100:dd2c:1275:6373:8ab6> has quit IRC22:29
*** tprrt <tprrt!~tprrt@upc31-1-78-208-110-13.fbx.proxad.net> has quit IRC22:35
*** JaMa <JaMa!~martin@ip-217-030-068-212.aim-net.cz> has quit IRC23:03
*** georgem_home <georgem_home!uid210681@gateway/web/irccloud.com/x-pgsytftayfmqiayh> has joined #yocto23:20
*** agust <agust!~agust@p54833389.dip0.t-ipconnect.de> has quit IRC23:22

Generated by irclog2html.py 2.11.0 by Marius Gedminas - find it at mg.pov.lt!