LetoThe2nd changed the topic of #yocto to: Welcome to the Yocto Project | Learn more: https://www.yoctoproject.org | Community: https://www.yoctoproject.org/community | IRC logs: http://irc.yoctoproject.org/irc/ | Having difficulty on the list, with someone on the list or on IRC, contact Yocto Project Community Manager Letothe2nd | CoC: https://www.yoctoproject.org/community/code-of-conduct
florian has quit [Ping timeout: 268 seconds]
jmiehe has quit [Quit: jmiehe]
goliath has quit [Quit: SIGSEGV]
johnmcnuggets has joined #yocto
johnmcnuggets has quit [Read error: Connection reset by peer]
johnmcnuggets has joined #yocto
davidinux has quit [Ping timeout: 246 seconds]
davidinux has joined #yocto
starblue has quit [Ping timeout: 255 seconds]
starblue has joined #yocto
jclsn has quit [Ping timeout: 272 seconds]
jclsn has joined #yocto
Saur75 has quit [Quit: Client closed]
Saur75 has joined #yocto
mulk has quit [Ping timeout: 260 seconds]
mulk has joined #yocto
sotaoverride has quit [Ping timeout: 252 seconds]
sotaoverride has joined #yocto
joekale has joined #yocto
sakoman has quit [Ping timeout: 272 seconds]
ablu has quit [Read error: Connection reset by peer]
ablu has joined #yocto
xmn has quit [Ping timeout: 264 seconds]
sakoman has joined #yocto
Saur75 has quit [Quit: Client closed]
Saur75 has joined #yocto
Chaser has joined #yocto
sakoman has quit [Ping timeout: 240 seconds]
lexano has quit [Ping timeout: 272 seconds]
Dracos-Carazza has quit [Quit: ZNC 1.8.2 - https://znc.in]
Dracos-Carazza has joined #yocto
johnmcnuggets has quit [Ping timeout: 246 seconds]
sakoman has joined #yocto
sakoman has quit [Ping timeout: 260 seconds]
schtobia has quit [Quit: Bye!]
schtobia has joined #yocto
sakoman has joined #yocto
benkard has joined #yocto
mulk has quit [Ping timeout: 256 seconds]
benkard is now known as mulk
mbulut has joined #yocto
mbulut has quit [Client Quit]
leon-anavi has joined #yocto
davidinux has quit [Ping timeout: 268 seconds]
davidinux has joined #yocto
roussinm has quit [Ping timeout: 255 seconds]
Noor has joined #yocto
nerdboy_ has joined #yocto
nerdboy has quit [Ping timeout: 255 seconds]
nerdboy_ is now known as nerdboy
nerdboy has quit [Client Quit]
nerdboy has joined #yocto
vladest has joined #yocto
vladest has quit [Ping timeout: 272 seconds]
mvlad has joined #yocto
deribaucourt has joined #yocto
ederibaucourt has quit [Ping timeout: 256 seconds]
Noor has quit [Read error: Connection reset by peer]
rfuentess has joined #yocto
vladest has joined #yocto
florian has joined #yocto
sakoman has quit [Ping timeout: 272 seconds]
sakoman has joined #yocto
florian has quit [Ping timeout: 252 seconds]
rsalveti has quit [Quit: Connection closed for inactivity]
Saur75 has quit [Quit: Client closed]
Saur75 has joined #yocto
Noor has joined #yocto
merit has joined #yocto
locutusofborg has quit [Read error: Connection reset by peer]
florian has joined #yocto
locutusofborg has joined #yocto
luc4 has joined #yocto
locutusofborg has quit [Read error: Connection reset by peer]
alessioigor has joined #yocto
locutusofborg has joined #yocto
ptsneves has joined #yocto
<merit> oh weird - the matrix and irc aren't linked
KanjiMonster has quit [Quit: ZNC 1.8.2 - https://znc.in]
Guest52 has joined #yocto
KanjiMonster has joined #yocto
KanjiMonster has quit [Client Quit]
<merit> hi - sorry to crosspost from matrix
<merit> how would I go about using source from inside a tarbal inside a zip inside a zip that's hosted on a remote server in a .bb recipe?
<merit> adding it in "SRC_URI" seems straightforward enough, but I'm stumped on how I should unpack each successive layer - a custom "do_unpack"? "do_unpack_append"?
KanjiMonster has joined #yocto
KanjiMonster has quit [Client Quit]
KanjiMonster has joined #yocto
<rburton> i'd add a do_unpack postfunc that did the rest of the unpackging
<rburton> bitbake will fetch and undo one layer of compression, you'll have to do the rest
KanjiMonster has quit [Client Quit]
KanjiMonster has joined #yocto
Saur75 has quit [Quit: Client closed]
Saur75 has joined #yocto
vladest has quit [Remote host closed the connection]
vladest has joined #yocto
<merit> tried this, but it crashes claiming "Variable do_unpack_append contains an operation using the old override syntax."
<alperak> PIP_INSTALL_PACKAGE is no longer used, but it is still present in some recipes. is there something i don't know or does it need to be cleaned?
<merit> do_unpack:append() ?
jmiehe has joined #yocto
Guest52 has quit [Quit: Client closed]
<rburton> do_unpack is python, you can't just throw shell on it
<rburton> which is why i said use a postfunc
<rburton> grep for postfunc, you'll find examples in meta/
<rburton> alperak: obsolete, remove
davidinux has quit [Ping timeout: 252 seconds]
davidinux has joined #yocto
johnmcnuggets has joined #yocto
<merit> oh, this is kinda funny
<merit> the url is making the uri parser blow up in fetch because it's got a query that's not "k=v" but just a param
<merit> so I guess I have to implement my own "do_fetch"?
<rburton> escape the ?
<rburton> you _should_ be able to escape it enough
<merit> like, with backslashes?
<rburton> its a url so %
<merit> you're talking urlencodings?
<rburton> yes
<rburton> ? is %3F
<rburton> its probably the ; that is breaking, escape that
Noor has quit [Quit: Leaving]
ray-san has joined #yocto
<merit> escaping the ? made new stuff break instead
florian_kc has joined #yocto
<merit> it never made it past looking for a = on the query parm
<rburton> try just escaping the ;
<merit> not sure why
<merit> but, changing it back to
<merit> (escaping only the semicolon)
<merit> huh
<merit> that's fun
<merit> it's the same error either way - invalid syntax
<rburton> thats not where i expected it to fail, can you pastebin the entire recipe
<rburton> as i said above you can't append shell to do_unpack
<rburton> invalid syntax because you're putting shell on the end of a python function
<rburton> :append is literally string concatenation. do_unpack is python code.
<merit> ohhh, dang
<merit> okay
<rburton> use a postfunc instead
<rburton> (see my messages at 11:02)
<merit> so that's why you end up with a chimera of functions
<merit> thanks for the patience
Saur75 has quit [Quit: Client closed]
<merit> I am _very_ new at this whole ecosystem
<merit> did not realize that an _append'd function is literally string appended into the function
<rburton> somehow expressing a url with query strings in a format that takes urls and then having to triple-unpack a compressed file isn't the easiest thing to do
<merit> thought it was merely some sort of abstract hook
Saur75 has joined #yocto
<merit> starting to think that the answer, today, is to unpack it by hand
<merit> do the stupid build
<merit> send firmware to germany
<merit> and go to sleep
<KanjiMonster> downloadfilename is a parameter for the fetcher, not the webserver URI, so I don't think you need or should escape the semicolon. At least for me the url only works without it when I throw it at wget
<rburton> ah true yeah
<rburton> i didn't pay complete attention and thought it was a parameter for the server
<rburton> download urls that are all magic parameters for the server are the worst
<KanjiMonster> also nice matroshka archive you get there, a tar.gz in a zip in a zip
<merit> the main part is that there's a server param that's making the uri parsing barf in the fetcher code in python
Vonter has quit [Ping timeout: 264 seconds]
<merit> or at least, that's the problem with the uri
Vonter has joined #yocto
<rburton> if I were going to rewrite yocto i'd not use URIs like this, but hey that's where we are
goliath has joined #yocto
<rburton> what if you escape the ? with %3F
<merit> same error
<rburton> cursed thing
<rburton> i hate URLs
<merit> URI wakes up screaming
<merit> well, same error, slightly different
<merit> let me confirm
<KanjiMonster> purely guessing from the output, it seems the code trips up about no seconds (or more) parameter via & if it's a query
<merit> yes
<merit> that's the primary issue
<KanjiMonster> and no amount of escaping will fix that
<merit> I guess the uri doesn't cause crashing after that
<merit> if I escape, it just makes everything wrong for varying reasons
<merit> ie - escape the semicolon, and it can't get the params
<rburton> right you need the ; to be literal as that's for bitbake
<merit> escape the ? and it's no longer sending a query
<KanjiMonster> you could try adding a & at the end, at least that does not break the url when throwing it at the webserver (but may trip up the fetcher at a later stage)
<merit> ya - assumed bitbake stripped the ; stuff at the very top of the *init*
<merit> oo
<merit> or
<merit> I could add nonsense value and see if the server cares
<KanjiMonster> it does
<merit> wellllll
<merit> it cares if the value has a value
<merit> but bitbake survives when I add the trailing =
<merit> blursed
<KanjiMonster> might trip up their statistics, but who cares ;D
<merit> now, to actually add some unpacking
<rburton> that feels like a bitbake fetcher bug trying to come out
<rburton> looks good
<merit> still not seeing any .ko nuggets in my rootfs
<rburton> oe-pkgdata-util list-pkg-files -r [recipe name]
<rburton> verify it actually build stuff and put it in a package
<merit> just blew everything away in case bb doesn't know to do that
johnmcnuggets has quit [Quit: Leaving]
<rburton> so yeah the url parser assumes that a query string is a key-value mapping
<rburton> but that's not actually required and typically URLs you fetch don't have them
<rburton> so congrats you found a bug, i have a test case
<merit> lol yay
<rburton> hm interestingly enough the last change to this code was me, three years ago, fixing what sounds like the same issue
<rburton> ah no different issue
<merit> hm,
<merit> thought it was working, then ran bb against the recipe standalone
<merit> now I'm running into stuff like license requirements
<merit> whew
Noor has joined #yocto
<KanjiMonster> there is one at the end of the README.txt
<KanjiMonster> I'm positively surprised, the driver files have SPDX tags
<KanjiMonster> merit: btw, the code here https://github.com/Aquantia/AQtion is identical to what's contained in the zip
<KanjiMonster> I checked out the repo, copied over all files, and git only showed an untracked file "release_notes.txt", but no file changes
<KanjiMonster> might be a bit less hassle
Guest9743 has quit [Quit: Connection closed]
jmiehe has quit [Quit: jmiehe]
johnmcnuggets has joined #yocto
<merit> wellll
<merit> not quite _everything_
<merit> the readme on the zip file has the license replaced with "uhh, we totally do GPL2"
<merit> vs "this is not free software but it doesn't cost money" on the official
<merit> lol
Saur75 has quit [Quit: Client closed]
chep has quit [Quit: ZNC 1.8.2 - https://znc.in]
Saur75 has joined #yocto
chep has joined #yocto
Saur75 has quit [Client Quit]
Saur75 has joined #yocto
<merit> lbeh
<merit> "/bin/sh: 1: unzip: not found"
Saur75 has quit [Client Quit]
Saur75 has joined #yocto
Saur75 has quit [Client Quit]
Saur75 has joined #yocto
Guest9743 has joined #yocto
alessioigor has quit [Quit: alessioigor]
alessioigor has joined #yocto
<rburton> merit: do_fetch[depends] += "unzip-native:do_populate_sysroot"
<rburton> but i'd definitely be grabbing the git repo!
<KanjiMonster> merit: huh? both README.txt files in the zip file are identical to the one in the github repo
<KanjiMonster> and all (build relevant) files in the zip and repo have a GPL-2.0-only SPDX tag, so claiming GPL 2.0 seems to be correct
<KanjiMonster> merit: though I guess you mean this: https://github.com/Aquantia/AQtion/issues/7
creich has quit [Ping timeout: 268 seconds]
unknown1 has joined #yocto
mckoan|away has quit [Ping timeout: 246 seconds]
<merit> ok... finally got it running
<merit> not sure where this has my actual OS build, but the recipe cooks
xmn has joined #yocto
lexano has joined #yocto
Saur75 has quit [Quit: Client closed]
Saur75 has joined #yocto
<merit> thanks for everything
<merit> it's so far... building
<merit> so at least it's not broken yet
<merit> lol
<merit> and I've still got like... a few hours I might even be able to sneak in for sleep
unknown1 has quit [Ping timeout: 252 seconds]
Saur75 has quit [Client Quit]
Saur75 has joined #yocto
unknown1 has joined #yocto
luc4 has quit [Ping timeout: 272 seconds]
sev99 has joined #yocto
<rburton> merit: what's your email? I'd like to cc you on a patch if that's okay.
mckoan|away has joined #yocto
alessioigor has quit [Ping timeout: 255 seconds]
alessioigor has joined #yocto
pedrowiski has joined #yocto
vladest1 has joined #yocto
vladest1 has quit [Client Quit]
<rburton> merit: never mind. would be useful if you could test that https://lore.kernel.org/bitbake-devel/20240308141439.1119130-1-ross.burton@arm.com/T/#u solves your problem if you remove the url bits you added
vladest has quit [Ping timeout: 264 seconds]
<merit> crap, just decided to give up and sleep for the "night"
<merit> system image finished building with no sign of my ko
<rburton> merit: did you add it to your image?
<rburton> building a recipe != putting it into image
<merit> I dropped it into the spot where the vendor said to drop the recipe
<rburton> you add a package to an image by using IMAGE_INSTALL in the image recipe
<rburton> having a recipe available just means you can build that recipe. having built a recipe just means you _can_ put the packages into an image.
<vvn> is there a best practice for DISTRO_VERSION? single word vs semantic numbers, etc.
<rburton> your choice, it just gets written out in various places
<vvn> is it ok to follow the yocto project version? e.g. mickledore/4.2 in order to give a clue on which yocto version you're based on witout needing to providing a compatibility table?
<rburton> i guess nobody can stop you :)
<vvn> just talking about conventions and best practices here rburton :)
* vvn is wondering what the industry does regarding distro versioning
alessioigor has quit [Quit: alessioigor]
<rburton> "it depends" i guess. if you've got an actual product you might want to bump major version when you want to, not when yocto does.
alessioigor has joined #yocto
alessioigor has quit [Client Quit]
<vvn> good point
<vvn> one should refer to VERSION_ID= from os-release(5) then
Chaser_ has joined #yocto
Chaser has quit [Ping timeout: 264 seconds]
paw has joined #yocto
vladest has joined #yocto
vladest has quit [Remote host closed the connection]
johnro1987 has joined #yocto
davidinux has quit [Ping timeout: 255 seconds]
davidinux has joined #yocto
<johnro1987> Hello guys, I have yocto enviroment(dunfell) with  external toolchains.For development I am building an SDK (to run in docker and do smaller projects compilation).In this generated SDK I am missing some files compared to that external toolchains install. I am mostly interested in ASAN suport.My compilation against the generated SDK  fail because
<johnro1987> is missing libasanpre_init.o . Is there are any fix for this?
vladest has joined #yocto
mckoan|away has quit [Ping timeout: 260 seconds]
mckoan|away has joined #yocto
alessioigor has joined #yocto
roussinm has joined #yocto
unknown1 has quit [Ping timeout: 264 seconds]
luc4 has joined #yocto
mckoan|away has quit [Ping timeout: 272 seconds]
mckoan|away has joined #yocto
Chaser_ has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
<paulg> "WARNING: Host distribution "ubuntu-18.04" has not been validated with this version of the build system"
<paulg> splitting hairs, but that is factually not true.
<paulg> it may not be officially supported, but I guarantee you there are clowns like me still "validating" it works.
leon-anavi has quit [Quit: Leaving]
<roussinm> Is it possible to have 2 target sdk from one shar? Or do we have to deploy 2 different sdk?
Guest9743 has quit [Quit: Connection closed]
johnmcnuggets has quit [Remote host closed the connection]
johnmcnuggets has joined #yocto
<RP> paulg: it isn't in our test matrix so it isn't officially tested
<paulg> I know - I get it, and I'm largely being a jerk (imagine that?) - just sayin our message doesn't say what you just said - that release/distro XYZ isn't officially supported.
luc4 has quit [Ping timeout: 255 seconds]
Vonter has quit [Ping timeout: 256 seconds]
Vonter has joined #yocto
Noor has quit [Quit: Leaving]
pedrowiski has quit [Ping timeout: 272 seconds]
<JaMa> paulg: you don't test it often or deep enough as I haven't seen you reporting or fixing issues when patchelf got broken for 18.04 or when qemu-native stopped building
<JaMa> being able to finish some small build from time to time isn't validation (and that's me being a jerk :))
<RP> paulg: I'm open to patches to change it to something different
<paulg> Apparently I'm just a sponge/leech and I don't send patches. :-P
florian has quit [Quit: Ex-Chat]
<RP> paulg: sucker? :)
florian_kc has quit [Ping timeout: 260 seconds]
rfuentess has quit [Remote host closed the connection]
ptsneves has quit [Ping timeout: 272 seconds]
amitk has joined #yocto
<nerdboy> moin
florian_kc has joined #yocto
simonew has joined #yocto
johnmcnuggets has quit [Remote host closed the connection]
florian_kc has quit [Ping timeout: 272 seconds]
paw has quit [Ping timeout: 268 seconds]
johnmcnuggets has joined #yocto
goliath has quit [Quit: SIGSEGV]
alessioigor has quit [Quit: alessioigor]
alessioigor has joined #yocto
johnmcnuggets has quit [Remote host closed the connection]
amitk has quit [Ping timeout: 256 seconds]
johnmcnuggets has joined #yocto
dmoseley has quit [Quit: ZNC 1.9.0 - https://znc.in]
florian_kc has joined #yocto
dmoseley has joined #yocto
florian_kc has quit [Ping timeout: 272 seconds]
amitk has joined #yocto
<mischief> is there a nice way to make perf use my unpacked source tree from `devtool modify` for the kernel? it seems STAGING_KERNEL_DIR isnt overridden..
<mischief> also, why doesn't yocto build systemd-native for things like systemd-sysusers, systemctl?
kpo has joined #yocto
florian_kc has joined #yocto
<rburton> mischief: because for the limited systemctl we need its easier to reimplement a small script than manage to build a minimal systemd. last time i tried it wasn't possible to build _just_ the tools.
<mischief> q
<mischief> rburton: well, i think the 'sysusers' code is somewhat broken, in kirkstone at least
<rburton> file a bug then :)
<rburton> presumably that's the logic in meta/classes-recipe/rootfs-postcommands.bbclass
<rburton> though the point of sysusers is that it doesn't actually do much at build time surely
<mischief> ah, i guess that got fixed in 2023 for the main branch.
<mischief> tho it still appears to be some python goo, instead of shell goo now :-)
amitk has quit [Ping timeout: 252 seconds]
<vvn> Is mediamtx packaged somewhere?
johnro1987 has quit [Quit: Connection closed]
<merit> ok, rburton, I've tracked down where I think the image is being put together and appended onto the IMAGE_INSTALL
jmd has quit [Remote host closed the connection]
<vvn> I guess I'll have to write a recipe for mediamtx. Does someone recommend a go-based recipe I could use as a reference?
<mischief> prepare for sadness - with go modules, tracking down the dependencies is rather difficult, and you require network access at compile time unless you are careful
simonew has quit [Remote host closed the connection]
mvlad has quit [Remote host closed the connection]
<vvn> mischief: luckily there are standalone precompiled releases to download
Saur75 has quit [Quit: Client closed]
Saur75 has joined #yocto
<mischief> if that is acceptable to you, that might be easier.
alessioigor has quit [Quit: alessioigor]
goliath has joined #yocto
<merit> well,s hoot
<merit> I think I'm getting closer
<merit> now I've got the module building and failng because it's grabbing kernel headers from my system instead of the device root
<rburton> if your following the guide for building an out of tree module that should just work, so you might have makefiles which hardcode /usr or something in them
<merit> tried this: "arager@churchill:~/src/fsl-auto-yocto-bsp/build_s32r45evb$ bitbake -e fsl-image-auto | grep -e '^KERNEL_.*='"
<merit> it's giving me "bitbake -e fsl-image-auto | grep -e '^KERNEL_.*='
<merit> whoops
<merit> might have overshared a little
<rburton> follow that if you're not already
<rburton> if you are then congratulation you've found broken makefiles, and you need to fix them
* rburton -> bed
<merit> I'm following that, but I get this in my build log as the makefile line called:
<merit> make -j4 -C /lib/modules/5.15.0-97-generic/build M="/home/arager/src/fsl-auto-yocto-bsp/build_s32r45evb/tmp/work/s32r45evb-fsl-linux/egpl-t101-mod/0.1-r0/git" modules
<merit> vOv
<merit> I'll figure it out
<mischief> /lib/modules/5.15.0-97-generic is your host kernel?
<merit> thanks for all the pointers
<merit> it's been invaluable
<merit> yeah
<merit> "Linux LOCALHOST 5.15.0-97-generic #107~20.04.1-Ubuntu SMP Fri Feb 9 14:20:11 UTC 2024 x86_64 x86_64 x86_64 GNU/Linux"
jmiehe has joined #yocto
<mischief> i *think* its supposed to be pointing into .. build/tmp/work-shared/$MACHINE/kernel-source
<mischief> perhaps your module makefile hardcodes /lib/modules/$(uname -r), or you forgot to pass it in in your recipe..
<merit> not much to the module recipe
<merit> nobody hardcoding against any headers in particular
<merit> unless
<merit> KDIR isn't being set
<merit> ofc
<mischief> perhaps you can stuff it in EXTRA_OEMAKE
<mischief> we have a class in our layer that just replaces do_compile for out of tree modules, since most of ours have broken makefiles