<risca>
Today I learned that git suppresses output if it can't detect a tty, e.g. during "docker build". However, "expect" can happily create a pseudo tty for me :)
sakoman has quit [Quit: Leaving.]
geoffhp has quit [Ping timeout: 260 seconds]
florian_kc has quit [Ping timeout: 260 seconds]
qschulz has quit [Remote host closed the connection]
qschulz has joined #yocto
camus has quit [Remote host closed the connection]
camus has joined #yocto
camus has quit [Remote host closed the connection]
camus has joined #yocto
sudip has quit [Ping timeout: 268 seconds]
davidinux has quit [Ping timeout: 268 seconds]
camus has quit [Ping timeout: 252 seconds]
camus has joined #yocto
davidinux has joined #yocto
sudip has joined #yocto
prabhakarlad has quit [Ping timeout: 260 seconds]
starblue has quit [Ping timeout: 260 seconds]
starblue has joined #yocto
Estrella__ has joined #yocto
dmoseley has quit [Excess Flood]
Estrella has quit [Ping timeout: 264 seconds]
Estrella_ has quit [Ping timeout: 268 seconds]
Estrella has joined #yocto
dmoseley has joined #yocto
<PhoenixMage>
hmmmm, having problems a go program (k3s), segfaults when run. Decided to see what happens if I compile go from source on the SBC using the yocto version to bootstrap it following the official go instructions misc/cgo/test fails :/
Estrella_ has joined #yocto
money has joined #yocto
money is now known as Guest2037
Guest2037 has quit [Changing host]
Guest2037 has joined #yocto
Guest2037 is now known as money
Estrella has quit [Ping timeout: 246 seconds]
Estrella__ has quit [Ping timeout: 272 seconds]
Estrella has joined #yocto
xmn has quit [Ping timeout: 265 seconds]
jclsn has quit [Ping timeout: 265 seconds]
jclsn has joined #yocto
demirok has quit [Quit: Leaving.]
money has quit [Quit: late]
<PhoenixMage>
Is there an environment variable I am missing that allows you to run the go tests during the build? At least for the native compilation?
amitk has joined #yocto
camus has quit [Remote host closed the connection]
camus has joined #yocto
thomasd13 has joined #yocto
camus has quit [Remote host closed the connection]
camus has joined #yocto
camus has quit [Remote host closed the connection]
camus has joined #yocto
schtobia has quit [Ping timeout: 260 seconds]
camus1 has joined #yocto
camus has quit [Ping timeout: 252 seconds]
camus1 is now known as camus
schtobia has joined #yocto
demirok has joined #yocto
schtobia_ has joined #yocto
bps has joined #yocto
nemik has quit [Ping timeout: 260 seconds]
nemik has joined #yocto
alessioigor has joined #yocto
nemik has quit [Ping timeout: 272 seconds]
nemik has joined #yocto
bps has quit [Ping timeout: 252 seconds]
hmw[m] has joined #yocto
alessioigor has quit [Quit: alessioigor]
alessioigor has joined #yocto
zpfvo has joined #yocto
invalidopcode has quit [Remote host closed the connection]
invalidopcode has joined #yocto
Payam has quit [Quit: Leaving]
goliath has joined #yocto
zpfvo has quit [Quit: Leaving.]
zpfvo has joined #yocto
schtobia_ has quit [Ping timeout: 246 seconds]
leon-anavi has joined #yocto
<LetoThe2nd>
yo dudX
* alessioigor
waves all!
<tomzy_0>
morning
mouser has quit [Quit: leaving]
lamm1 has joined #yocto
mouser has joined #yocto
gsalazar has joined #yocto
bps has joined #yocto
bps has quit [Changing host]
bps has joined #yocto
prabhakarlad has joined #yocto
<TRO[m]>
Payam, rburton: agree, S3 is probably not a good solution for that. I vote for the solution we made - works for many EC2 instances in parallel: https://www.youtube.com/watch?v=CNCjtQj74VI
<TRO[m]>
Min: 10:39 ++
invalidopcode has quit [Read error: Connection reset by peer]
<termi24>
Hi! Does anybody know where the initial system start timestamp comes from? I would like to modify this timestamp to be the same as the build timestamp or last commit timestamp. For example last year I was working on a honister build a lot and no matter when and where I built the image, no matter which host machine, no matter which kernel
<termi24>
commit...it is always like this -> "Journal begins at Fri 2021-08-06 16:41:00 CEST". I saw that some directories like /sbin /lib /boot /etc /bin /usr also have a modification close to this time -> "drwxr-xr-x 2 root root 4096 2021-08-06 21:14:37.388000000 +0200 bin", but I have no clue where the ~5.5 hour offset might come from, so maybe I'm
<termi24>
still on the wrong track?
<kanavin>
termi24, what usually helps is a bit of code archeology. Specifically, something does print that line, so you could grep your build tree (or sub-parts of it), and see where it comes from.
<kanavin>
then unroll from there.
* mcfrisk
remembers sending close to 100 patches to lkml, and several iterations :/
<RP>
mcfrisk: there are some limits that groups.io has
<rburton>
termi24: systemd has a time epoch as configured in the recipe, which is the SOURCE_DATE_EPOCH of the build.
<mcfrisk>
with lkml, it was quite challening to find proper maintainers to Cc, and of course that resulted in 10's of dead email addresses and lots of gone-missing maintainers..
<termi24>
Thanks rburton, this was the missing piece I was looking for. Don't know why I missed that.
<qschulz>
mcfrisk: scripts/get_maintainer.pl ?
<mcfrisk>
qschulz: yep, I used that automatically with git format-patch, but naturally input data is unreliable etc
florian_kc has joined #yocto
xmn has joined #yocto
<rburton>
RP: in ab-helper, why do some layers have no-layer-add set in the config.json? if layer-config is passed just the needed repos, why couldn't they be added automatically?
<rburton>
RP: i thought it might be because some repos have sublayers so you can't just add the top-level, but meta-virt doesn't. then again that might be copypasta.
<rburton>
same for meta-intel
seninha has quit [Quit: Leaving]
ldts_ is now known as ldts
<RP>
rburton: I think things have just got confused over time and I suspect copy/paste
<rburton>
cool
<rburton>
glad i'm not missing something
<rburton>
but i did miss a bit from that ab2 patch yesterday :(
<RP>
rburton: it is possible I don't remember something
<RP>
rburton: ah :(. I can probably work out how to restart the controller if needed
<rburton>
can't fire a test run of my helper branch if there's no entry in the ab2 config!
invalidopcode has quit [Remote host closed the connection]
<RP>
rburton: I remembered how to get in now
invalidopcode has joined #yocto
<rburton>
oh whilst i'm here i'll rip out oe-selftest-arm, as armhost runs for qemuarm64 now
<rburton>
RP: you should write that down :)
<RP>
rburton: I did!
<RP>
rburton: I was trying to solve the wrong problem yesterday :/
<qschulz>
RP: what happens if you release a lock you didn't hold in lock_timeout?
<qschulz>
RP: challenging the use of finally instead of else in the try-catch basically
<RP>
qschulz: that code is only reachable if the lock is held?
<qschulz>
RP: os._exit() is enough to not reach the finally?
<RP>
qschulz: that is what os._exit() is designed for (ish)
<RP>
meant to be used with fork() but I think this is a valid use
florian_kc has quit [Ping timeout: 252 seconds]
nemik has quit [Ping timeout: 260 seconds]
nemik has joined #yocto
nemik has quit [Ping timeout: 260 seconds]
nemik has joined #yocto
<RP>
rburton: I reconfig'd the controller
<RP>
rburton: well, I tried. We have a worker out of space and it all went wrong :(
xmn has quit [Ping timeout: 272 seconds]
xmn has joined #yocto
sakoman has joined #yocto
invalidopcode has quit [Remote host closed the connection]
invalidopcode has joined #yocto
demirok has quit [Quit: Leaving.]
xmn has quit [Ping timeout: 256 seconds]
<rburton>
oh nuts
termi24 has quit [Quit: Client closed]
paulg has joined #yocto
demirok has joined #yocto
zpfvo has quit [Ping timeout: 264 seconds]
<rburton>
RP: is the ab ready to fire a test build or should i hold back?
<RP>
rburton: you can try
<rburton>
fire in the hole
<RP>
rburton: either the update applied or it didn't, I can't tell
<rburton>
well i managed to start a job, so that's a good sign
<rburton>
armpit: is it you i beg at to get the mickledore layercompat patch merged into meta-oe?
zpfvo has joined #yocto
florian_kc has joined #yocto
<RP>
rburton: isn't it khem?
manuel1985 has joined #yocto
meego has joined #yocto
<JaMa>
it's already in master-next for meta-oe, but as you pinged armpit meta-security needs his or mine version to be merged as well (not sure why one of the meta-security layers wasn't updated in armpit's version)
<rburton>
clang shouldn't take that long to build!
prabhakarlad has quit [Quit: Client closed]
<meego>
Can someone kindly help me understand the rationale behind this : in recipes, PACKAGECONFIG often checks the DISTRO/IMAGE/TUNE_FEATURES vars. However, it seems there is no way to conditionally set configure flags based on wether some specific packages are set to be installed or not. Why is that? It seems like something a lot of recipe developers would want (e.g. ffmepg) so I'm guessing there's something I'm not seeing
thomasd13 has quit [Ping timeout: 268 seconds]
<RP>
meego: a recipe can't know which packages an image might install, they're separate things and vary by image. The recipe doesn't rebuild per image
<RP>
rburton: that does sound somewhat worrying :/
<meego>
RP: makes sense, thanks!
manuel1985 has quit [Ping timeout: 265 seconds]
<risca>
thanks RP and rburton for the help yesterday. It took some time, but I think most stuff from the SDK is actually building for me now =) Have to go mirror hunting from time to time, but that's okay
<risca>
I'm amazed that so much code is still online after 10 years. Ancient packages and whatnot
<rburton>
risca: a lot of it might be coming from our mirror, for exactly that reason :)
<risca>
Yeah, I suspect so too =) thanks a lot
<RP>
risca: I do get asked what data we can delete now and again so I'm glad we've managed to keep the sources mirror working!
<risca>
The only reason I'm doing this is to have a "known good" system, so I can have a reference. Next is porting the stuff to mainline
<risca>
Don't you just love propriertary vaporware?
* risca
does a little happy dance
goliath has quit [Quit: SIGSEGV]
bps has quit [Ping timeout: 248 seconds]
<RP>
risca: I started working on OE in the first place as I wanted to try and improve the reproducibility of software for devices!
<rburton>
i wonder if poky 1.0 still builds with a sufficiently old container
<risca>
Why don't you investigate? It's a nice little rabbit hole
<risca>
Hot tip: recompile host git to link with openssl instead of gnutls or gitlab.gnome.org won't work
zpfvo has quit [Ping timeout: 246 seconds]
<qschulz>
risca: GIT_SSL_NO_VERIFY environment variable maybe would be enough?
<qschulz>
I mean... security issue then :D
<risca>
Nope, unfortunately
<qschulz>
risca: too new I guess
<risca>
It's some kind of bug in gnutls which makes encryption not work
<mark_>
After upgrading dunfell to kirkstone, bitbake always run into this python warning: /usr/lib64/python3.6/importlib/_bootstrap.py:219: ImportWarning: can't resolve package from __spec__ or __package__, falling back on __name__ and __path__ Have anyone run into this and what's your solution?
<whuang0389>
if a package provides multiple alternatives, is it possible to give each alternative an their own priority?
invalidopcode has quit [Remote host closed the connection]
sudip_ is now known as sudip
invalidopcode has joined #yocto
alessioigor has joined #yocto
alessioigor has quit [Client Quit]
<ferlzc>
Hi i'm using Yocto 4.0 on Debian Buster and for some reason i'm having this error related to QEMU when I try to build an image:
<ferlzc>
../qemu-6.2.0/meson.build:2264:2: ERROR: Problem encountered: fdt not available but required by targets arm-softmmu, aarch64-softmmu, i386-softmmu, mips64el-softmmu, ppc-softmmu, ppc64-softmmu, riscv32-softmmu, riscv64-softmmu, x86_64-softmmu
<ferlzc>
however when I try to build core-image-minimal the image builds without any problem
florian_kc has quit [Ping timeout: 260 seconds]
<ferlzc>
removed wic.vmdk from my IMAGE_FSTYPES and the error stopped
Haxxa has quit [Quit: Haxxa flies away.]
whuang0389 has quit [Ping timeout: 260 seconds]
Haxxa has joined #yocto
xmn has quit [Ping timeout: 252 seconds]
xmn has joined #yocto
sakoman has joined #yocto
amitk has quit [Ping timeout: 268 seconds]
Payam has joined #yocto
florian_kc has joined #yocto
<ferlzc>
Hi, i'm having error building the qemu-system-native on Kirkstone: | ../qemu-6.2.0/meson.build:2264:2: ERROR: Problem encountered: fdt not available but required by targets arm-softmmu, aarch64-softmmu, i386-softmmu, mips64el-softmmu, ppc-softmmu, ppc64-softmmu, riscv32-softmmu, riscv64-softmmu, x86_64-softmmu
<ferlzc>
any hints on what could be missing ?
<Payam>
is it possible to use s3 for sstate?
<mark_>
I thought about that but never tried it, I use EFS instead
<Payam>
mark_, can you tell me how you created it?
<Payam>
I am a noob on aws
<Payam>
it is Elastic Beanstalk right?
<Payam>
it says "create a web app"
<JaMa>
elastic file storage
<JaMa>
s3 should IMHO work as well if you enable static web endpoint
<mark_>
Our AWS build nodes setup a few years ago. I forgot most of the details tbh. In general, me the dev-ops guy asked our IT-cloud team to get this EFS storage set up and showed me how to mount it. And I just mount it on the EC2 node userdata. And I point the sstate directory under that EFS mount point. Works pretty well
<rburton>
our CI is the same: runner nodes spun up as needed, all sharing a EFS mount for sstate/dldir
<rburton>
trivial
<Payam>
so rburton should I use s3 or efs for sstate and dl?
<rburton>
as i said yesterday and you've been told by several people, EFS
<JaMa>
EFS is better, because you can use it as rw SSTATE_DIR while s3 static web can be used only as ro SSTATE_MIRROR (and you need to s3 sync the sstate there after the builds)
<rburton>
EFS is basically an amazon-branded NFS mount
<rburton>
the extra work to sync s3 just isn't worth it
<Payam>
yes but yoctor has support for EFS?
<rburton>
its just a directory you mount
<Payam>
can I do sOURCE_MIRROR_URL ?= \"efs..?
<JaMa>
but if you want to expose that sstate to other people to use in builds outside AWS you need s3
<rburton>
Payam: read up on what EFS is and how it would work. you just mount it and use it as a normal directory, it's nothing special
<Payam>
ah really?
<Payam>
so bitbake will not use any fetching?
<JaMa>
yes, it will create all necessary sources from pixie dust provided by EFS :)
<Payam>
the aws guy wil be back on monday
manuel1985 has joined #yocto
<Payam>
so I let him do that. But until then I think I will go with s3 on sstate
Payam has quit [Quit: Leaving]
Starfoxxes has quit [Ping timeout: 264 seconds]
Starfoxxes has joined #yocto
xmn has quit [Ping timeout: 272 seconds]
seninha has quit [Ping timeout: 268 seconds]
Algotech75 has joined #yocto
xmn has joined #yocto
mark_ has quit [Quit: mark_]
meego has quit [Read error: Connection reset by peer]
seninha has joined #yocto
<risca>
Do anyone in here recognize "arago-glsdk-multimedia-image"? I've tried searching for it online, but all I find is other people not finding it and customer support asking them to contact sales :/
<risca>
The SDK I'm trying to build has this hard coded in their build script "MACHINE=$1 bitbake arago-glsdk-multimedia-image", then fails on "Nothing provides..."
<risca>
I've followed their steps and checked out all the layers they have in their config file
Minvera has quit [Remote host closed the connection]
azcraft has quit [Remote host closed the connection]
manuel1985 has quit [Ping timeout: 265 seconds]
ferlzc has quit [Quit: Leaving]
warthog9 has quit [Quit: Leaving]
mvlad has quit [Remote host closed the connection]