<rsalveti>
which seems to be the standard for modules, to avoid TUNE_PKGARCH depending on MACHINE_ARCH
nemik has joined #yocto
<rsalveti>
JPEW: so in order to be aligned with what oe-core expects, removing from the runtime dependency list might be the easiest fix, but won't be 100% ideal, as you said
<rsalveti>
at least now we know what is going on :-)
<rsalveti>
and I guess a similar issue could show with the other cases covered by sstate_rundepfilter (e.g. SIGGEN_EXCLUDERECIPES_ABISAFE / SIGGEN_EXCLUDE_SAFE_RECIPE_DEPS)?
pgowda_ has joined #yocto
nemik has quit [Ping timeout: 244 seconds]
nemik has joined #yocto
nemik has quit [Ping timeout: 276 seconds]
nemik has joined #yocto
jmiehe1 has joined #yocto
jmiehe has quit [Ping timeout: 276 seconds]
jmiehe1 is now known as jmiehe
nemik has quit [Ping timeout: 244 seconds]
nemik has joined #yocto
nemik has quit [Ping timeout: 276 seconds]
nemik has joined #yocto
amitk has joined #yocto
nemik has quit [Ping timeout: 244 seconds]
nemik has joined #yocto
nemik has quit [Ping timeout: 244 seconds]
nemik has joined #yocto
thomasd13 has joined #yocto
kroon has joined #yocto
manuel1985 has joined #yocto
nemik has quit [Ping timeout: 256 seconds]
nemik has joined #yocto
manuel1985 has quit [Ping timeout: 276 seconds]
goliath has joined #yocto
frieder has joined #yocto
gpanders has quit [Write error: Connection reset by peer]
raghavgururajan has quit [Read error: Connection reset by peer]
gpanders has joined #yocto
raghavgururajan has joined #yocto
mckoan|away is now known as mckoan
mvlad has joined #yocto
thomasd13 has quit [Ping timeout: 256 seconds]
Schlumpf has joined #yocto
thomasd13 has joined #yocto
manuel1985 has joined #yocto
Guest5 has joined #yocto
bradfa has quit [Read error: Connection reset by peer]
diamondman has quit [Ping timeout: 244 seconds]
denix has quit [Ping timeout: 244 seconds]
bradfa has joined #yocto
diamondman has joined #yocto
dagmcr has quit [Read error: Connection reset by peer]
denix has joined #yocto
dagmcr has joined #yocto
jpuhlman has quit [Remote host closed the connection]
wkawka has joined #yocto
nemik has quit [Ping timeout: 240 seconds]
nemik has joined #yocto
florian has joined #yocto
leon-anavi has joined #yocto
nemik has quit [Ping timeout: 276 seconds]
nemik has joined #yocto
jpuhlman has joined #yocto
<wkawka>
Hi, how can I force bitbake to do do_image_tar instead of do_image_wic?
<wkawka>
I have already do_image_wic[noexec] = "1"
<qschulz>
wkawka: use the proper IMAGE_FSTYPES
<wkawka>
That was it, thanks
<wkawka>
Is it possible my docker has missing configuration? I am making a recipe for container which contains some lib in it, but for now when i import it and then run it, with for example echo "hello world" i dont see any error and any output, it is possible that the container created by recipe works correctly but docker doesn't? I have incluede busybox
<wkawka>
so commands like echo should work. I didn't include it earlier and there was an error
xmn has quit [Quit: ZZZzzz…]
nemik has quit [Ping timeout: 272 seconds]
nemik has joined #yocto
<wkawka>
what is the proper name of qemu basic machine?
<wkawka>
qemux86-64 will do?
<rburton>
yes
janvermaete[m] has quit [Quit: You have been kicked for being idle]
Alban[m] has quit [Quit: You have been kicked for being idle]
<wkawka>
Recipe which creates container must be based on quemu like machine or can it be based on a target machine?
<wkawka>
qemu*
<rburton>
target machine makes sense if its a container to run on a real target
<rburton>
the qemu machine is designed and tailored for the qemu emulated hardware
<wkawka>
yes, it will be run on target
<wkawka>
when i built it on target machine I checked the manifest, and there were also packages specific to that machine
<wkawka>
So I thought that is not ok and I have changed the machine to qemu
thomasd13 has quit [Remote host closed the connection]
<rburton>
well, you don't need a kernel, so don't install a kernel.
thomasd13 has joined #yocto
<rburton>
there's an example image in oe-core
<wkawka>
If i built a container on specific machine, will I be able to run it with docker on PC?
<rburton>
if that specific machine is compatible with a PC, yes
<wkawka>
I tried multiple examples, sadly none of it worked
<rburton>
what exactly did you try?
<rburton>
there's a worked example in oe-core for building a generic container image (containerimage.py selftest) and meta-virtualisation most likely has tested examples that include the metadata etc that docker would want
<wkawka>
when i Import them and run wth docker there is no error, but no output also
<wkawka>
they just end
thomasd13 has quit [Remote host closed the connection]
thomasd13 has joined #yocto
<rburton>
what machine?
<wkawka>
this is PCengines apu3
<wkawka>
I am using kas, yocto version is dunfell
<rburton>
assuming the MACHINE you're using is compatible with the hardware youre running on, you'll need to look at what docker is doing. Does it enter the container okay? What command does it run in the container? etc
<wkawka>
As I said before, there are no errors, but also no output and it just exits
<wkawka>
Is there a way to debug such things?
<wkawka>
docker debug mode etc
<rburton>
no idea, maybe ask a docker channel
<rburton>
zeddii might be able to help, but he'll not be awake yet
jmiehe has quit [Ping timeout: 276 seconds]
starblue has quit [Ping timeout: 240 seconds]
<Guest82>
Hi, I would like to add a complete git repository in my image as if it was cloned locally. AFAIK what I want sits in the downloads directory in /git2. When I set the SRC_URI in my recipe to my repo, the content of the .git folder in the ${WORKDIR}/git does not fully reflect that. Would any of you know what the best approach would be? Use
<Guest82>
externalsrc?
<rburton>
Guest82: just do a git clone in the do_install task?
<rburton>
add it to SRC_URI and then you can just clone directly from the DL_DIR
starblue has joined #yocto
<Guest82>
rburton, OK, so something like:
<Guest82>
install -d ${D}/mydir
<Guest82>
cd ${D}/mydir
<Guest82>
git clone ${DL_DIR}/git2/ ... Not out of lazyness, but is there any automatically correct reference the exact repo folder for my recipe? Instead of typing it mannualy :)
nemik has quit [Ping timeout: 256 seconds]
nemik has joined #yocto
nemik has quit [Ping timeout: 244 seconds]
nemik has joined #yocto
<rburton>
Guest82: ${WORKDIR}/git is where bitbake unpacked to, so you should be able to clone from there.
<rburton>
Guest82: if you set S=WORKDIR/git as is conventional with git repos, then you can just do git clone ${S} ${D}/mydir
<rburton>
i *think* git hardlinks where possible in local clones so it will work correctly when packaged up
xmn has joined #yocto
ardo has quit [Read error: Connection reset by peer]
ardo has joined #yocto
Schlumpf has quit [Ping timeout: 252 seconds]
<Guest82>
@rburton: Thanks, I've tried that. The result in my image rootfs folder is identical to doing a cp -rT ${S} ${D}/mydir.
<Guest82>
There is no content in the objects folder other than an info and pack folder, and the info folder contains an "alternates" file with:
<Guest82>
Where _git_user and _repo_name are my repository names. Maybe I can make use of that somehow?
<Guest82>
whoops that went wrong trying to type in a directory here..
<rburton>
man git-clone would be useful, you want to create an entirely independent clone.
<rburton>
the local checkout is as lightweight as possible, you likely want to force git to unassociate the links
<rburton>
maybe --dissociate
<Guest82>
OK thanks I'll look at the manual and try your suggestion
<rburton>
the problem with copying ${S} is that the repo is a mostly-linked clone to the canonical repo in DL_DIR, for speed. git can just point to the central repo instead of copying the files, which is a huge win when you have a repo like the kernel (9GB here) or gcc (2GB)
<rburton>
but obviously that's no good if you want a standalone copy
nemik has quit [Ping timeout: 240 seconds]
nemik has joined #yocto
nemik has quit [Ping timeout: 244 seconds]
nemik has joined #yocto
Wouter01001 has quit [Remote host closed the connection]
Wouter01001 has joined #yocto
Wouter01001 has quit [Read error: Connection reset by peer]
Wouter01001 has joined #yocto
mihai has joined #yocto
ruslan has joined #yocto
nemik has quit [Ping timeout: 240 seconds]
nemik has joined #yocto
<jclsn[m]>
Can the u-boot-imx binary not be flashed via uuu? I tried flashing u-boot.bin, but uuu complains that it can't find a valid IVT header
<jclsn[m]>
We were still building our bootloader with the Android toolchain and I have the task to move to Yocto's bootloader now
nemik has quit [Ping timeout: 276 seconds]
nemik has joined #yocto
<Guest82>
rburton: Tried and although there is now a repacked file in the objects/pack dir, the repo on my target still points to the repo dir in tmp/work/etc instead of my github repo. I'll dig some more and try to find a solution.
<Guest82>
Maybe I just have to put the URL in the do_install funtion seperatly and clone like that.
<rburton>
you could do git clone http://whatever --reference ${S}, and it will copy stuff locally really fast but still point at the parent repo
florian has quit [Ping timeout: 240 seconds]
<rburton>
yeah of course, it will still think the repo you cloned from is 'upstream'
<ruslan>
Hello!
<ruslan>
I want to know how to add script that will be executed during build within image? E.g., I want to execute systemctl enable service@mod, but I don't want to do it in each new image manually.
nemik has quit [Ping timeout: 240 seconds]
nemik has joined #yocto
ptsneves has joined #yocto
<rburton>
Ideally, the recipe that provides that service file can do it, using the systemd class
<rburton>
worst case you just need a pkg_postinst script which does the systemctl call and install that into the image
<rburton>
look up package maintainer scripts
nemik has quit [Ping timeout: 240 seconds]
nemik has joined #yocto
seninha has joined #yocto
<Guest82>
Hmm somehow I'm not allowed to do git clone "git://git@github.com:whatever" as if I have no network access from within do_install. fatal: unable to look up ... (port 9418) (Name or service not known)
<Guest82>
Trying to solve this first now.
<qschulz>
Guest82: do_install[network] = "1" in your recipe
wCPO has quit [Quit: Ping timeout (120 seconds)]
mario-goulart has quit [Remote host closed the connection]
mario-goulart has joined #yocto
wCPO has joined #yocto
<Guest82>
qschulz: Thanks I've put that in above the do_install function but it has the same result.
<Guest82>
I'm trying to do this so I can create a development image to easily develop and debug my application on the target, and use version control on it as well without having to set up the repo everytime I flash my board.
<Guest82>
The production image doesn't need this so maybe I'm taking a wrong approach here anyway.
<qschulz>
Guest82: i'm not sure spending time on this to save a simple git clone on your target is wise tbh
<LetoThe2nd>
.... and why one would favor that over, say, devtool build && deploy-target
Guest54 has joined #yocto
<Guest54>
Hi! I use yocto dunfell. My problem is iptables which gives me this error message: "iptables v1.8.4 (legacy): Couldn't load match `limit':No such file or directory". Now I added libnftnl to my yocto but the error message remains
<Guest54>
Any idea what I must do to get it work?
ardo has quit [Read error: Connection reset by peer]
<Guest54>
The exact call of iptables was: iptables -A syn_flood -m limit --limit 3/s --limit-burst 5 -j RETURN
<Guest54>
I guess I still need to add any package or lib to my yocto distro for getting it work
ardo has joined #yocto
<Guest54>
Does this error also happen on your Yocto Linux?
<qschulz>
Guest54: did you add libnftnl to your image or did you add the correct PACKAGECONFIG option for iptables?
<Guest54>
qschulz: yes, it's installed (as you suggested yesterday)
<qschulz>
why does it still say legacy then?
<Guest54>
meta-mylayer\recipes-extended\iptables contains iptables_%.bbappend now and its content is: PACKAGECONFIG_append = " libnftnl"
<qschulz>
i thought this should have been enough to use nftables (and provided "legacy" in the iptables output actually means it';s not using nftables)
<qschulz>
Guest54: which version of yocto are you using?
<Guest54>
dunfell
<qschulz>
also, check with bitbake-getvar -r iptables PACKAGECONFIG it's set correctly
<qschulz>
ah, no bitbake-getvar in dunfell yet I think
<qschulz>
so, bitbake -e iptables | grep -e "^PACKAGECONFIG="
<RP>
qschulz: I think it was backported
<qschulz>
RP: I've seen patches from sakoman recently but I think it's still too recent to expect people to have it available?
<qschulz>
RP: two weeks ago the patch was sent to the ML for bitbake
<qschulz>
but yes, thanks for the reminder, soon we'll be able to more or less expect bitbake-getvar to always be there :)
<Guest54>
qschulz: Your bitbake command outputs: PACKAGECONFIG="ipv6 libnftnl"
<qschulz>
so I think you should be able to use iptables with nftable support by calling xtables-nft-multi directly?
<qschulz>
(and/or add the change manually in a do_install_append in your bbappend for iptables)
<Guest82>
qschulz: Yes I'm probably overcomplicating this trying to streamline and document the process.
<Guest82>
It's a NodeJs application and I hoped to use npmsw in the SRC_URI to also fetch my node modules at the same time as they are not in my application repository. I wanted to do that for the production image anyway.
<Guest82>
LetoThe2nd: I'm using vscode over ssh at the moment for remote debugging of the project on my target. I'm not sure how I could use devtool for this kind of work instead?
<Guest82>
Maybe I can trick vscode into pushing the project to the target prior to debugging.
<Guest82>
Thanks for your feedback guys!
<Guest54>
qschulz: I'm not sure how to use xtables-nft-multi. Simply exchanging the string "iptables" with "xtables-nft-multi" in my script doesn't work.
<Guest54>
I will try the do_install_append way
<qschulz>
Guest54: or try if the symlink is enough?
<ruslan>
I wrote postinstallation script, which is executed on target, and the system goes into the infinite boot loop.
<ruslan>
last message during the boot: "A start job is running for Run pending postinst"
kroon has quit [Quit: Leaving]
<Guest54>
qschulz: OK, I did the symlink directly in my current yocto distro and after that I again tried: xtables-nft-multi -A syn_flood -m limit --limit 3/s --limit-burst 5 -j RETURN
<Guest54>
The result is now: iptables/1.8.4 Failed to initialize nft: Protocol not supported
<qschulz>
Guest54: might need to tweak your kernel config if nftables is not enabled there?
Guest5 has quit [Quit: Client closed]
<qschulz>
zgrep /proc/config.gz NF_TABLES ?
<Guest54>
qschulz: OK, I'll check this
<JPEW>
RP: rsalveti and I found an interesting thing yesterday: create-spdx.bbclass trawls through BB_TASKDEPDATA to find the runtime dependencies for packages, but BB_TASKDEPDATA isn't filtered by siggen, so includes "stale" references to things that aren't dependencies on the task from the signature perspective (meaning, the task doesn't rerun when they change)
sotaoverride has joined #yocto
<JPEW>
I'm not sure if BB_TASKDEPDATA *should* be filtered, or if we need a new BB_TASKSIGDEPDATA that is filtered? The filtering has to be done by the cooker because it can have multiconfig dependencies (otherwise I'd just filter it in create-spdx.bbclass)
<Guest54>
zgrep not found :)
<qschulz>
Guest54: not a surprise :D
<qschulz>
Guest54: just copy the file somewhere on your pc and extract it or something
<Guest54>
hmm not set :-) I'll rebuild with that one on. Thanx. I hope it works then :)
olani has quit [Ping timeout: 244 seconds]
olani has joined #yocto
<ruslan>
qschulz: I've tried to add config filename to SYSTEMD_SERVICE, but bitbake says that he can't find this file. How to properly point to file, which is placed in files://conf/wg0.conf?
<qschulz>
ruslan: remove pkg_postinst_ontarget and what's coming after, not needed
<qschulz>
PACKAGE_WRITE_DEPS:append don't understand why this is needed, so likely can be removed
<qschulz>
ruslan: as for SYSTEMD_SERVICE:${PN} set it to the name of the systemd unit filename
<qschulz>
not the path
<qschulz>
your systemd unit file then shoudl be installed in ${D}${systemd_unitdir}/system by your recipe
<qschulz>
aaaaaaah
<qschulz>
wait a sec, I'm starting to understand your things now
<qschulz>
sorry, didn't read the whole thing
<qschulz>
you're just providing a configruation file for wg-quick and want to start wg-quick at boot for this configuration
<ruslan>
qschulz: Yes
<qschulz>
ruslan: so the short answer is: i don't know how to do it properly
<qschulz>
I **think**, it should be possible to have a bbappend for wireguard-tools
<RP>
JPEW: I think nothing else has needed the filtered data. This all worries me a bit though, not least that the dependency may be valid for license reasons even if not for signature
<qschulz>
where you have SYSTEMD_SERVICE:${PN} += "wg-quick@wg0.service" in there
<qschulz>
that's what I'd try next, but those are wild guesses
<JPEW>
RP: Right. The specific problem is that SPDX documents are all linked together with hashes of the document themselves (which is good for tracing). This means that you have to regenerate the SPDX if anything upstream of you changes. Using the filtered list is not 100% ideal, but it at least prevents builds from breaking. The other option would be to ignore the siggen filter for the task so that it gets all the signatures and rebuilds
<JPEW>
appropriately
jmiehe has quit [Ping timeout: 244 seconds]
sakoman has joined #yocto
Guest54 is now known as fbre
Guest82 has quit [Quit: Client closed]
gpanders has left #yocto [WeeChat 3.4.1]
fitzsim has quit [Remote host closed the connection]
<ptsneves>
@RP On the topic 'Re: [OE-core] [PATCH 2/3] oeqa/sysroot.py: Check bitbake return status', the first patch was correct. If i do not ignore status the test will fail like was the case for v2. The test actually asserts it fails. The only way to have the test passing without ignoring status is to catch the exception but that is even uglier. TLDR: The v1 patch is correct, the v2 is not.
thomasd13 has quit [Ping timeout: 240 seconds]
wkawka has quit [Quit: Client closed]
<ruslan>
qschulz: What I need is to run these 2 rows after installation or during first boot.
<ruslan>
I've tried to add a postinst, but with postinst script system boots infinitely
<RP>
ptsneves: Ah, right, I think I misunderstood things then, sorry
<RP>
JPEW: I'm a bit torn on that, I think I would need to think about it
mckoan is now known as mckoan|away
<qschulz>
ruslan: the variables should do it for you automagically
<ruslan>
qschulz: But it is custom recipe witch custom conf files.
<qschulz>
ruslan: install the custom conf file from your recipe if you want, and you can add a bbappend for wireguard-tools to enable wg-quick@wg-0 service at boot using the correct variables
<qschulz>
that could be hack-ish indeed
<qschulz>
i dont know how you're supposed to provide a configuration file to a systemd unit from another recipe, sorry
amitk_ has joined #yocto
amitk has quit [Ping timeout: 244 seconds]
ruslan has quit [Quit: Client closed]
<fbre>
ruslan: don't do exit or return at the end of postinst
ruslan has joined #yocto
<jclsn[m]>
What does this mean?
<jclsn[m]>
```
<jclsn[m]>
```
<jclsn[m]>
Manifest /home/user/Workspace/Yocto/vti2-doom/build/tmp/sstate-control/manifest-x86_64_x86_64-nativesdk-freedoom.package_write_rpm not found in vti2 cortexa53-crypto-mx8mm cortexa53-crypto cortexa53 armv8a-crc-crypto armv8a-crypto armv8a-crc armv8a aarch64 allarch x86_64_x86_64-nativesdk (variant '')?
jpuhlman has quit [Remote host closed the connection]
<ruslan>
fbre: ok, I'll try.
<ruslan>
Where can I find example with enabling systemd services during build/first boot stage?
jpuhlman has joined #yocto
<ruslan>
jclsn[m]: try to disable sstate cache
ptsneves has quit [Ping timeout: 240 seconds]
Wouter01001 has quit [Read error: Connection reset by peer]
Wouter01001 has joined #yocto
Guest82 has joined #yocto
peoliye has joined #yocto
gsalazar has quit [Remote host closed the connection]
pabigot has quit [Quit: Leaving.]
pabigot has joined #yocto
atomic_null[m] has quit [Quit: You have been kicked for being idle]
ptsneves has joined #yocto
pabigot has quit [Quit: Leaving.]
jmiehe has joined #yocto
ptsneves has quit [Ping timeout: 256 seconds]
pabigot has joined #yocto
F_Adrian has joined #yocto
ruslan has quit [Quit: Client closed]
Guest54 has joined #yocto
Guest54 is now known as fbre
JudgeMaygarden[m has joined #yocto
fbre has quit [Quit: fbre]
leon-anavi has quit [Read error: Connection reset by peer]
nemik has quit [Ping timeout: 240 seconds]
nemik has joined #yocto
nemik has quit [Ping timeout: 256 seconds]
nemik has joined #yocto
frieder has quit [Remote host closed the connection]
nemik has quit [Ping timeout: 244 seconds]
nemik has joined #yocto
nemik has quit [Ping timeout: 244 seconds]
nemik has joined #yocto
manuel1985 has quit [Ping timeout: 240 seconds]
nemik has quit [Ping timeout: 244 seconds]
nemik has joined #yocto
nemik has quit [Ping timeout: 276 seconds]
nemik has joined #yocto
<otavio>
RP: I *think* we fixed the Rust SDK
<otavio>
can you put it on autobuilder so we confirm this?
pgowda_ has quit [Quit: Connection closed for inactivity]
contre has joined #yocto
<vmeson>
JaMa: or anyone, Is there a recipe/layer for qtcreator for qt6? I don't see it in meta-qt6 but the source code is available at least: https://www.qt.io/offline-installers?hsLang=en
<RP>
otavio: interesting. I'm sure abelloni will include in the next run so we'll see
<RP>
otavio: they don't "feel" quite right but I need to do some thinking about them, it is hard to tell
otavio has quit [Ping timeout: 244 seconds]
<JaMa>
vmeson: I'm not aware of any, did you check the layer index?
<vmeson>
JaMa: yes, the Qt company hasn't even indexed the basic qt6 layer. :-(
otavio has joined #yocto
jpuhlman has quit [Remote host closed the connection]
<vmeson>
I'll check with the folks who worked on meta-qt6, thanks.
jpuhlman has joined #yocto
pabigot has quit [Remote host closed the connection]
pabigot has joined #yocto
fitzsim has joined #yocto
pabigot has quit [Remote host closed the connection]
pabigot has joined #yocto
pabigot has quit [Remote host closed the connection]
pabigot has joined #yocto
pabigot has quit [Remote host closed the connection]
pabigot has joined #yocto
otavio has quit [Quit: leaving]
otavio has joined #yocto
nemik has quit [Ping timeout: 244 seconds]
nemik has joined #yocto
nemik has quit [Ping timeout: 240 seconds]
nemik has joined #yocto
ptsneves has joined #yocto
adrian_ has joined #yocto
F_Adrian has quit [Ping timeout: 272 seconds]
Guest82 has quit [Quit: Client closed]
adrian__ has joined #yocto
ptsneves has quit [Read error: Connection reset by peer]
adrian_ has quit [Ping timeout: 244 seconds]
ptsneves has joined #yocto
peoliye has quit [Quit: Client closed]
florian has joined #yocto
v0n has quit [Quit: WeeChat 3.5]
vvn has quit [Quit: WeeChat 3.5]
vvn has joined #yocto
v0n has joined #yocto
florian has quit [Ping timeout: 256 seconds]
v0n has quit [Quit: WeeChat 3.5]
vvn has quit [Quit: WeeChat 3.5]
vvn has joined #yocto
mvlad has quit [Remote host closed the connection]