cambrian_invader has quit [Ping timeout: 260 seconds]
thomas_ has joined #yocto
adrian__ has joined #yocto
nemik has quit [Ping timeout: 252 seconds]
cambrian_invader has joined #yocto
nemik has joined #yocto
nemik has quit [Ping timeout: 256 seconds]
nemik has joined #yocto
goliath has joined #yocto
alessioigor has joined #yocto
alessioigor has quit [Client Quit]
kevinrowland has quit [Ping timeout: 252 seconds]
mvlad has joined #yocto
Wouter0100 has quit [Read error: Connection reset by peer]
Wouter0100 has joined #yocto
kranzo has joined #yocto
<kranzo>
hey all
davidinux has joined #yocto
GuestNew has joined #yocto
<GuestNew>
During some tests on kirkstone branch I was not able to give variables to bitbake from cmd line (MACHINE) anymore. I must have to set it inside the local.conf. There a way to bybass this "limitation" ? In order to not change all my CI... Thx
<kroon>
GuestNew, i think there is an "allowlist" for environment variables
<kroon>
previosly "whitelist"
<GuestNew>
kroon thx for the hint I will take a look :)
<kroon>
might have renamed that variable in kirkstone
<landgraf>
GuestNew: MACHINE should work. We're using it all the time
<GuestNew>
MACHINE=foo bitbake bar-image is call, MACHINE ?=fake, results MACHINE=ted-fake is invalid
<landgraf>
this is kirkstone 7f78bb7a7b
<GuestNew>
MACHINE=foo bitbake bar-image is call, MACHINE ?=fake (in conf/local.conf), results MACHINE=ted-fake is invalid
<landgraf>
GuestNew: MACHINE ??= "qemux86-64"
<landgraf>
I have this in local.conf
<landgraf>
more weak assignment :)
<kroon>
GuestNew, use -e to see how it got its value
<landgraf>
+1
<GuestNew>
Ho super nice thanks for all this morning tips :)
<kroon>
or bitbake-getvar
davidinux has quit [Ping timeout: 240 seconds]
goliath has quit [Quit: SIGSEGV]
Tyaku has joined #yocto
<Tyaku>
Hi, I have a question about "bests practices". Currently we have a project A and B (Based on CMake) that require files (.c) from a repository made by SiliconLabs (gecko_sdk). To include these projects in Yocto recipe, what is the best:
<Tyaku>
1. Put in the project A and B a submodule of C, adapt the CMakeLists.txt to keep the .c files from the submodule. (good for yocto, but gecko_sdk is huge, for development we have to download it two times (each project)).
nemik has quit [Ping timeout: 240 seconds]
nemik has joined #yocto
<Tyaku>
2. Keep our CMakeLists.txt as is, (currently we specify an environment variable with the path to gecko_sdk. So in the bitbake recipe: download project A and gecho_sdk for example (don't know yet if it's possible but I think), and then pass the correct path to the build with EXTRA_OECMAKE or something like this.
leon-anavi has joined #yocto
<GuestNew>
kroon -e doesn't work because of the previous error. but bitbake-getvar doesn't see the cmd line : https://pastebin.pl/view/ea335c6a
nemik has quit [Ping timeout: 240 seconds]
nemik has joined #yocto
<kroon>
GuestNew, i cant read the pastebin since its in polish
<kroon>
but you can check with -e without giving a recipe as argument
<GuestNew>
the global pastebin is under maintenance ... sorry
rfuentess has joined #yocto
dev1990 has joined #yocto
amahnui1 has joined #yocto
tnovotny has joined #yocto
olani has joined #yocto
mihai has quit [Ping timeout: 248 seconds]
goliath has joined #yocto
bps has joined #yocto
bps has joined #yocto
guysoft42 has joined #yocto
<kroon>
GuestNew, bitbake-getvar does seem to respect my passing of MACHINE via the env.
<GuestNew>
kroon you are right, after bumping to the HEAD i got the expected output : from env data.py:112 [inheritFromOS]:) sorry for troubles and thanks again for the support
Guest1179 has joined #yocto
ykrons has quit [Quit: Ex-Chat]
<Guest1179>
Hello, how can I disable the sanity checker? I added my own MACHINE but bitbake says "Either fix the the cause of this error or at your own risk disable the checker (see sanity.conf)"
<Guest1179>
It says DISTRO '.....' not found
<Guest1179>
and "OE-core's config sanity checker detected a potential misconfiguration"
Guest1179 has quit [Quit: Client closed]
<qschulz>
GuestNew: DISTRO not found sounds like an error you don't want to silence actually
<landgraf>
qschulz: you're replying to wrong Guest :) Guest1179 has left already
<GuestNew>
qschulz still okay to get some tips :P
<qschulz>
landgraf: the joy of nickname tab completion :)
<qschulz>
GuestNew: Tip: you can share sstate-cache and downloads directories between all developer machines and projects for every distro, machine and image you want to build
<GuestNew>
To be honnest you have already point me this very usefull tips months ago, my available disk space thanks you every day ^^
nemik has quit [Ping timeout: 252 seconds]
nemik has joined #yocto
<qschulz>
GuestNew: you can even save more with INHERIT += "rm_work" in your local.conf if you're really tight on space
<qschulz>
and if you have CI/CD and want to keep storage space under control, a find -delete based on the latest access time is usually a good thing too (I know a few people doing this here :)
<GuestNew>
qschulz Ho I have to put inside my cheat sheet:) :)
<qschulz>
s/latest/last/
<GuestNew>
you make my day
florian has joined #yocto
nemik has quit [Ping timeout: 250 seconds]
nemik has joined #yocto
florian_kc has joined #yocto
sstiller has joined #yocto
leon-anavi has quit [Quit: Leaving]
leon-anavi has joined #yocto
kroon has quit [Quit: Leaving]
nemik has quit [Ping timeout: 240 seconds]
nemik has joined #yocto
nemik has quit [Ping timeout: 240 seconds]
nemik has joined #yocto
<GuestNew>
qschulz are you mainteneur of Yocto Project ? After bumping to th HEAD of kirkstone branch I got python3-wheel build failed http://pastebin.fr/105533. there is a note inside Migration notes for 4.0 (kirkstone)
<qschulz>
GuestNew: nope, not a maintainer, just an occasional contributor
<GuestNew>
ok thx :)
<qschulz>
but you don't have to reach the maintainer to have a bug reported or fixed
<GuestNew>
i'm curious
amahnui1 has quit [Quit: Connection closed for inactivity]
<qschulz>
to which note in the migration notes of 4.0 are you referring to?
<qschulz>
which host distribution are you using (Ubuntu something, Debian, fedora etc..). Do you have all necessary packages installed (c.f. system requirements in the docs)
<qschulz>
I know moto-timo worked a lot on this python stuff lately so maybe he'll be able to help once he's awake.
<qschulz>
Are you able to reproduce on kirkstone 4.0 (and not HEAD of kirkstone branch) ?
<GuestNew>
I'm trying to migrate (under WSL2) from gatestgarth branch to kirkstone. All works like a charm from months on gatesgarth. Now I have checkout HEAD of kirkstone and I don't understand clearlt the issue
<thomas_>
When I rebuild it, after I changed the SRCREV the second time, I got no errors.
<thomas_>
I would expect, that the AUTOINC variable would increment from 0 to 1, after I changed the SRCREV. But it seems not to happen. What could be the reason for that?
<rburton>
almost all variables *can* be used in local.conf
kranzo has quit [Quit: Client closed]
<wkawka>
Thank you
nemik has quit [Ping timeout: 240 seconds]
nemik has joined #yocto
<wkawka>
So, if I would like to enable u-boot to my device, I have to use UBOOT_CONFIG or something else? I already have an u-boot bin copied into sd card and saved as kernel8.img
<wkawka>
I'm running this on rpi0-2w
Tyaku has quit [Quit: Lost terminal]
<qschulz>
wkawka: UBOOT_CONFIG will configure u-boot but not build it by itself
<qschulz>
virtual/bootloader virtual recipe should be set to build u-boot instead
<qschulz>
c.f. PREFERRED_PROVIDER
nemik has quit [Ping timeout: 260 seconds]
nemik has joined #yocto
<jclsn[m]>
Has anyone come up with an advanced .clangd config file on the project level to automatically search for the right toolchain and sysroot in /opt/...
<jclsn[m]>
?
<jclsn[m]>
Something you could put into version control of you projects and never worry about it again?
<jclsn[m]>
Guess that is utopian, but asking is for free
<rburton>
using a SDK, i presume
<jaskij[m]>
For CMake based projects, there's a setup-less CMakeToolchain generation which I think made it to Kirkstone, if that is of any
ykrons has joined #yocto
<wkawka>
Then I have to set it to sth like this: PREFERRED_PROVIDER_virtual/bootloader = "kernel8.img" or how?
<qschulz>
wkawka: no, the name of the u-boot recipe you want to use
<qschulz>
might be u-boot, might be u-boot-ti, u-boot-fsl, whatever
<qschulz>
also, I assume that kernel8.img is actually the name o the Linux kernel iamge and not U-Boot but I may be wrong, I've seen some weird naming choices in the past :)
tnovotny has quit [Remote host closed the connection]
<rburton>
would be nice if the rpi people cared about meta-raspberrypi :(
<GuestNew>
rburton same for meta-xilinx :(
GNUmoon has joined #yocto
<qschulz>
GuestNew: didn't AMD say they are going to put some effort in Yocto support since they bought Xilinx?
<rburton>
meta-raspberrypi is maintained by people unrelated to the rpi foundation, whereas at least meta-xilinx is worked on by people at xilinx/amd
<seninha>
Hi, I'm new to yocto. I generated a linux image signed with pregenerated keys for secure boot. How can I run the image on qemu? Do I have to import the keys to qemu bios or something like that?
prabhakarlad has quit [Quit: Client closed]
<qschulz>
My assumption is that if you don't check the signature of the image during your boot process, there should be no issue at all?
<seninha>
When I run `runqemu`, qemu freezes on the "Booting from Hard Disk..." screen.
<jaskij[m]>
AMD's hiring people to work on Xilinx Yocto support
<GuestNew>
qschulz <3
<rburton>
they already have people working on it of course, but more people sounds good
Guest11 has joined #yocto
<Guest11>
Hi, I have a strange bitbake error: core-image-minimal-1.0-r0 do_rootfs: core-image-minimal: usermod command did not succeed. ... Error executing a python function in exec_func_python() autogenerated: .....run.set_user_group.16238 failed with exit code 1 Do you have any idea?
<seninha>
Oh, I had to pass runqemu the path to the ovmf file.
nemik has quit [Ping timeout: 260 seconds]
nemik has joined #yocto
PaowZ has quit [Ping timeout: 240 seconds]
bps has joined #yocto
bps has joined #yocto
xmn has joined #yocto
<Guest11>
I see, seninha is guilty for the failed run.set_user_group because he has not pass runqemu the path to the ovmf file :)
nemik has quit [Ping timeout: 246 seconds]
nemik has joined #yocto
xmn has quit [Ping timeout: 240 seconds]
GuestNew has quit [Quit: Client closed]
PaowZ has joined #yocto
xmn has joined #yocto
amahnui1 has quit [Quit: Connection closed for inactivity]
davidinux has joined #yocto
manuel1985 has joined #yocto
<ernstp>
looks like the git safe.directory thing is causing a lot of issues!
davidinux has quit [Ping timeout: 252 seconds]
<RP>
ernstp: sadly, yes
<RP>
hopefully master is fixed now
<ernstp>
ok good!
<ernstp>
looks like sakoman is working on getting dunfell in order
davidinux has joined #yocto
<ernstp>
the current dunfell head is probably a bit broken but I see some stuff in progress on dunfell-nut
wkawka has quit [Quit: Client closed]
GuestNew has joined #yocto
raghavgururajan has quit [Remote host closed the connection]
raghavgururajan has joined #yocto
Tyaku has joined #yocto
<Tyaku>
Hi, I have this error: "QA Issue: -dev package mypackage-dev contains non-symlink .so" when I build a cmake project that generate a library .so
<Tyaku>
Did someone understand the problem ?
Guest11 has quit [Quit: Guest11]
<qschulz>
Tyaku: yes
<qschulz>
you're not supposed to have .so files which aren't symlinks
<qschulz>
best practices dictate that shared libraries should be versioned in their filename too
<qschulz>
e.g. .so.X.Y.Z
<qschulz>
then symlinks exist from .so to .so.X to .so.X.Y to .so.X.Y.Z
<qschulz>
so fix your cmake to create a .so.X.Y.Z
<Tyaku>
oh :D I didn't excpect this kind of problem :D
<Tyaku>
Thanks
nemik has quit [Ping timeout: 276 seconds]
nemik has joined #yocto
nemik has quit [Ping timeout: 256 seconds]
nemik has joined #yocto
GNUmoon has quit [Remote host closed the connection]
PaowZ has quit [Remote host closed the connection]
GNUmoon has joined #yocto
PaowZ has joined #yocto
sakoman has joined #yocto
GuestNew has quit [Quit: Client closed]
<Tyaku>
qschulz: Problem solved, it was it.
thomas_ has quit [Ping timeout: 252 seconds]
manuel1985 has quit [Ping timeout: 260 seconds]
kroon has quit [Quit: Leaving]
manuel1985 has joined #yocto
Schiller has joined #yocto
<moto-timo>
why would a pr_info() in a out of tree module that is auto loaded only print out when you "rmmod" it? (kernel-lab update to 5.17.5 from 5.15.2)
manuel1985 has quit [Ping timeout: 240 seconds]
<Tyaku>
It's strange, the build works, I correctly have mylib.so > mylib.so.4 > mylib.so.4.4.1 in sysroot-destdir/usr/lib but the image build with yocto only contains the .so.4 and .so.4.4.1 files
<rburton>
that's not strange, that's correct
<qschulz>
Tyaku: because the -dbg package has the .so file and not your normal packaghe
<rburton>
.so is in pn-dev
<qschulz>
-dev sorry
<qschulz>
thx rburton :)
<qschulz>
Tyaku: why do you need the .so symlink?
<Tyaku>
I think you are right :)
<qschulz>
Tyaku: oe-pkgdata-util find-path "mylib.so" will tel you that
<Tyaku>
I need it for a JAVA application. The .so contains the "JNI" of the JAVA application
<rburton>
the .so is a symlink
<Tyaku>
Yes but when my JAVA software starts, it looks for mylib.so not mylib.so.4
<Tyaku>
*But I think it's what you said before (with -dev package) because yeah i have this file in the -dev. I'm rebuilding the image to test.
<rburton>
you'll have to manually put it in PN, set FILES:PN-dev to "", and turn off the sanity check
davidinux has quit [Ping timeout: 260 seconds]
* moto-timo
bites his tongue about JAVA
davidinux has joined #yocto
<rburton>
dotnet does the same thing and needs .so files
<rburton>
idiocy
<Tyaku>
QA Issue: non dev package XXXXX contains symlink .so '/usr/lib/libXXXX.so' [dev-so]
<Tyaku>
Is this the sanity check to remove ?
<rburton>
yes
<Tyaku>
How ?
<rburton>
though if the library is actually a module that JNI is loading and it has no other purpose, then don't version it
<rburton>
build it as libfoo.so, and put it in PN. the sanity check is only for symlinks, modules are fine.
<rburton>
INSANE_SKIP:${PN} += "dev-so"
<Tyaku>
INSANE_SKIP_${PN} += "dev-so"
<Tyaku>
Oh ok, so I don't need the .so.x.y.z thing in that case ?, i'm going to remove it.
<rburton>
if and only if you never link to the library from C code
<rburton>
(i've forgotten almost everything i knew about java)
<Tyaku>
Yes never
<Tyaku>
This library is only for ONE JAVA application
nemik has quit [Ping timeout: 252 seconds]
nemik has joined #yocto
nemik has quit [Ping timeout: 240 seconds]
nemik has joined #yocto
goliath has quit [Quit: SIGSEGV]
ThomasRoos has joined #yocto
MiguelHorta has joined #yocto
Schiller has quit [Ping timeout: 252 seconds]
<MiguelHorta>
Hey all, given a new recipe how can I add a file to /dev/polkit-1/rules.d/? I tried to simply copy it in the do_install() step, but then it fails during the do_rootfs() "file /etc/polkit-1/rules.d conflicts between attempted installs of polkit-0.116-r0.cortexa9t2hf_neon and ..."
nemik has quit [Ping timeout: 240 seconds]
nemik has joined #yocto
vvn has joined #yocto
Tyaku has quit [Quit: Lost terminal]
Minvera has joined #yocto
vvn has quit [Ping timeout: 240 seconds]
nemik has quit [Ping timeout: 252 seconds]
nemik has joined #yocto
vvn has joined #yocto
wkawka has joined #yocto
<wkawka>
Hi, I need to explore u-boot process. I have to change something in u-boot scripts to make it work with different image. Where can I find good sources about that process?
ThomasRoos has quit [Quit: Leaving]
nemik has quit [Ping timeout: 252 seconds]
nemik has joined #yocto
Tyaku has joined #yocto
nemik has quit [Ping timeout: 252 seconds]
nemik has joined #yocto
davidinux has quit [Ping timeout: 246 seconds]
<kayterina[m]>
how does devtool uses $S in do_fetch? In a recipe with hacked S = "${WORKDIR}/$1" it does not fetch
<rburton>
fetch doesn't use S
<rburton>
what is $1?
<kayterina[m]>
it is the first argument, it hacks the distutils,not my recipe
<rburton>
that's unbelievably horrid
<kayterina[m]>
it builds ok when not used from devtool
<rburton>
I literally don't know how
<kayterina[m]>
DISTUTILS_SETUP_PATH ?= "${S}"
<kayterina[m]>
I changed this variable and it works
<rburton>
stop what you're doing and explain what the problem you're trying to solve with $1 is
<kayterina[m]>
DISTUTILS_SETUP_PATH ?= "${WORKDIR}/$1" instead of changing S
<kayterina[m]>
the recipe builds an executable with pyinstaller
<kayterina[m]>
the hack with $1 is for calling distutils3_do_compile pyinstaller-5.0.1 inside do_compile
<rburton>
you're still not explaining what you want to do
<rburton>
you want to call distutils3_do_compile in a directory that isn't S?
Tyaku has quit [Quit: Lost terminal]
<kayterina[m]>
build with pyinstaller with yocto. Until now I just installed a tar, now it has to build it automated.
<rburton>
isn't that just a matter of pyinstaller foo.py?
UmaKadam[m] has quit [Quit: You have been kicked for being idle]
deee101[m] has quit [Quit: You have been kicked for being idle]
qrsBRWNanyall[m] has quit [Quit: You have been kicked for being idle]
Alban[m] has quit [Quit: You have been kicked for being idle]
<kayterina[m]>
I suppose not, something about the oe recipe is not ok with the project we are on?
<kayterina[m]>
or something about not creating a pyinstaller recipe?
<kayterina[m]>
The correct way would be to create seperate recipes for altgraph, pyinstaller-hooks-contrib and pyinstaller-hooks-contrib?
adrian_ has joined #yocto
florian_kc has quit [Ping timeout: 276 seconds]
MiguelHorta has quit [Quit: Connection closed]
adrian__ has quit [Ping timeout: 256 seconds]
<rburton>
i'd create a native recipe for pyinstaller, so you can just invoke it in recipes where you want to run it
<rburton>
it's hard to comment when i can't see what you've done
adrian__ has joined #yocto
adrian_ has quit [Ping timeout: 240 seconds]
goliath has joined #yocto
manuel1985 has joined #yocto
rfuentess has quit [Remote host closed the connection]
florian_kc has joined #yocto
<kergoth>
Oi, now that it's 10:30 after 3 meetings I can finally go eat breakfast and wash and stuff. /eyeroll
florian_kc has quit [Ping timeout: 276 seconds]
manuel1985 has quit [Ping timeout: 248 seconds]
seninha_ has joined #yocto
seninha has quit [Ping timeout: 240 seconds]
leon-anavi has quit [Remote host closed the connection]
seninha_ has quit [Quit: Leaving]
nemik has quit [Ping timeout: 240 seconds]
nemik has joined #yocto
nemik has quit [Ping timeout: 240 seconds]
nemik has joined #yocto
ivand has joined #yocto
Tokamak has quit [Ping timeout: 250 seconds]
<ivand>
I'm struggling to start cog browser on i.MX6 platform, built with meta-webkit and meta-freescale. If someone had similar setup, I'd be grateful for advice. Details described here: https://github.com/Igalia/meta-webkit/issues/367
Tokamak has joined #yocto
Tokamak_ has joined #yocto
guysoft42 has quit [Remote host closed the connection]
guysoft42 has joined #yocto
Tokamak has quit [Ping timeout: 246 seconds]
Wouter0100 has quit [Remote host closed the connection]
frieder has quit [Remote host closed the connection]
<derRichard>
after upgrading to kirkstone i see base hash mismatches for literally every single recipe in all layers, even on a fresh build. build dir is empty except for conf/
Wouter0100 has quit [Ping timeout: 240 seconds]
<derRichard>
what could be wrong? :-/
<derRichard>
ah. in one layer's layer.conf the BB_BASEHASH_IGNORE_VARS seems to have no effect anymore.
seninha has joined #yocto
Minvera has quit [Remote host closed the connection]
<JaMa>
derRichard: are you sure you have the right bitbake version? (the one which reads BB_BASEHASH_IGNORE_VARS instead of BB_HASHBASE_WHITELIST)?
<derRichard>
JaMa: yes. in this layer is more to clean up, i fear ;-\
Wouter0100 has joined #yocto
florian_kc has quit [Ping timeout: 240 seconds]
mvlad has quit [Remote host closed the connection]
bps has quit [Ping timeout: 240 seconds]
alimon has quit [Quit: Leaving.]
kevinrowland has joined #yocto
florian_kc has joined #yocto
alimon has joined #yocto
Nate20 has joined #yocto
<Nate20>
Phew. Has anybody had any success adding a "/data" drive to the qemuarm64 machine before? It doesn't look to use a WKS so I can't just try throwing in an additional line in a WIC file somewhere like..
<Nate20>
part /data --source empty --ondisk sda --fstype=ext4 --label data --size="4097M" --align 1024 --use-uuid
<Nate20>
Actually it seems like there is a meta-arm repository that seems to be a bit better maintained than the machine definitions in core OE. They seem to use WKS for defining partitions so I could probably just switch to those 🤞