<LetoThe2nd>
gmorell: hiya, did I understand correctly that you're seeing clashes of Mender with other software? In which specifically? I definitely have meta-mender and meta-virtualization building together.
<mckoan>
good morning
Vonter has quit [Ping timeout: 260 seconds]
Vonter has joined #yocto
<Guest98>
how can i override kernel Image file.
rfuentess has joined #yocto
xmn has quit [Ping timeout: 260 seconds]
<LetoThe2nd>
Guest98: please elaborate? The kernel image type, like zImage, uImage, etc? Or choosing a different kernel altogether?
<Guest98>
I want to overwrite the Image created by Yocto with my own Image file. In the documentation of the custom board it says that i have to override the Image.
Schlumpf has quit [Quit: Client closed]
Kubu_work has joined #yocto
brrm has joined #yocto
Guest59 has joined #yocto
Guest98 has quit [Ping timeout: 245 seconds]
<Guest59>
LetoThe2nd if you wrote anything. i didnt see, disconnected from chat.
Daanct12 has joined #yocto
slimak has joined #yocto
Guest59 has quit [Client Quit]
brrm has quit [Ping timeout: 255 seconds]
<LetoThe2nd>
Guest59: what you are referring to is the output format file of the kernel build process. You can effectively set it through 'KERNEL_IMAGETYPE = "zImage"' for example in your local.conf for testing, and optimally in the machine configuration. But "I want to override the image" sounds like you want to inject a kernel (which to a linux distribution build process is a bit more than just a file) with something built outside of Yocto,
<LetoThe2nd>
and thats a whole different story.
sakman has quit [Ping timeout: 255 seconds]
Guest98 has joined #yocto
<LetoThe2nd>
Guest98: what you are referring to is the output format file of the kernel build process. You can effectively set it through 'KERNEL_IMAGETYPE = "zImage"' for example in your local.conf for testing, and optimally in the machine configuration. But "I want to override the image" sounds like you want to inject a kernel (which to a linux distribution build process is a bit more than just a file) with something built outside of Yocto,
<LetoThe2nd>
and thats a whole different story. Can you please share the documentation which claims you have to do that?
<LetoThe2nd>
Guest98: this randomly injects binary stuff, completly breaking any kind of source/license handling. Tell them to provide a proper BSP layer.
Daanct12 has quit [Quit: WeeChat 4.0.5]
Daanct12 has joined #yocto
<mckoan>
LetoThe2nd: that's a common scenario nowadays, more tha you can imagine :-(
<LetoThe2nd>
Guest98: its a prime example of yet another HW vendor putting Frankensoftware right on top of NVidias Frankenbuntu. Have them provide a proper BSP, or get source and create the BSP yourself. Or pay someone to do.
<LetoThe2nd>
mckoan: I can imagine a whole lot, but I can also speak up and tell people that its shit.
<mckoan>
LetoThe2nd: in those cases we usually re-engineer the whole build system, if final customer understands the reason
<LetoThe2nd>
mckoan: hence my remark "Or pay someone to do".
<mckoan>
LetoThe2nd: LOL
zpfvo has quit [Ping timeout: 264 seconds]
<Guest98>
i'm a rookie. they asked me to stand up the board on the yocto side and i'm trying to do it, at workplace. there's no one to help me or give me advice. :/
<Guest98>
thanks for knowledge.
<LetoThe2nd>
mckoan: I am actually just sitting on a train, writing a keynote for $COMPANY who want to get more involved
<LetoThe2nd>
Guest98: I feel for you, really. We've all been there. But this is the time to step up to your supervisor and honestly tell them: "the Yocto support for this board does not exist, and getting it into production ready shape will cost $$$"
<LetoThe2nd>
Guest98: you can go searching for a board with better support, or you can pay people to do the integration, and/or can get training to build up knowledge in your company. We will happily help with all of that. But "what are the handful of lines to make this work like the botched-up NVidia Ubuntu", that just is not going to happen because it doesn't work that way.
rfuentess has quit [Read error: Connection reset by peer]
<Guest98>
LetoThe2nd i wrote to company and they shared their custom kernel source with me.
Guillaume has quit [Quit: Client closed]
Guest4 has joined #yocto
<Guest4>
Hello !
leonanavi has joined #yocto
alessioigor has quit [Quit: alessioigor]
alessioigor has joined #yocto
goliath has joined #yocto
slimak has joined #yocto
Daanct12 has quit [Quit: WeeChat 4.0.5]
Daanct12 has joined #yocto
029AAA4MD has joined #yocto
<029AAA4MD>
Hello, I have a dummy question about a file populated by systemd recipe. So: I need to change the configuration of '/etc/systemd/timesyncd.conf' I used "oe-pkgdata-util find-path /etc/systemd/timesyncd.conf" to know which recipe provide this file. it returns "systemd".
<029AAA4MD>
Wow wait a minute
<029AAA4MD>
exit
<029AAA4MD>
I come back
029AAA4MD has quit [Client Quit]
Tyaku_ has joined #yocto
<Tyaku_>
.
<Tyaku_>
I'm, back, the kaspersky thing (intrusive tool) was maybe changing my nickname.
<rburton>
Tyaku_: write a bbappend and either sed the file, or ideally install your own in a do_install:append
<Tyaku_>
Hello, I have a dummy question about a file populated by systemd recipe. So: I need to change the configuration of '/etc/systemd/timesyncd.conf' I used "oe-pkgdata-util find-path /etc/systemd/timesyncd.conf" to know which recipe provide this file. it returns "systemd". In systemd the file only appear in 'systemd_247.6.bb' CONFFILES_${PN} = ""
<rburton>
systemd installs the file in its do_install stage
<Tyaku_>
CONFFILES_${PN} = "....here...." , but I don't know where the file is
<rburton>
its in the systemd source tree
<rburton>
its just one of many files systemd installs
<Tyaku_>
But strangely I don't find the file in "build/tmp/work" folder/subfolders
<Tyaku_>
But I also don't see the sources of systemd in the systemd folder so yeah.
<rburton>
builds from sstate won't have the sources unpacked, as there's no point
<rburton>
use devtool and you can edit the sources, generate a patch, and write a bbappend without having to dig around the tmp/work folders
<Tyaku_>
Ok! Thanks I'm going to do it.
Guest98 has quit [Quit: Client closed]
Guest1119 has quit [Quit: Quit]
Guest98 has joined #yocto
Guest98 has quit [Client Quit]
Guest98 has joined #yocto
Net147 has joined #yocto
Net147 has quit [Changing host]
Net147 has joined #yocto
Marmottus has joined #yocto
Guest98 has quit [Quit: Client closed]
Guest98 has joined #yocto
<Guest4>
Hello, I have a question about something quite basic on the paper but apparently not :(. I am working with a RPi, trying to manually switch root partition from U-Boot. In the boot partition, there is a cmdline.txt file, that specifies "root=/dev/mmcblk0p2". When I change it, it works well. But I want to specify the root partition from U-Boot. So I
<Guest4>
removed this file. And I am trying to use bootargs. Unfortunately, whatever I put in bootargs, it keeps booting well and mounting /dev/mmcblk0p2. How can I transfer bootargs from U-Boot to the kernel ?
<Guest4>
Thanks !
<qschulz>
Guest4: I would suggets to look into how U-Boot is creating its bootargs before booting the kernel
<qschulz>
Guest4: you can start by printing the environment, with printenv
slimak has quit [Ping timeout: 255 seconds]
<qschulz>
and check the bootcmd environment variable and follow the breadcrumbs until you get to the thing that reads the config.txt or sets defaults
<Guest4>
With printenv, the bootargs is well set, but it is never used in any other env variable. Then I have also the uboot.scr script, that specifies this:
<Guest4>
fdt addr ${fdt_addr} && fdt get value bootargs /chosen bootargs
<Guest4>
So is it picking the bootargs value from the dtb ?
Guest98 has quit [Quit: Client closed]
<qschulz>
Guest4: could very well do that indeed
<qschulz>
Guest4: if you have the dtb in hand, you can run dtc -I dtb -O dts my.dtb and check what's in /chosen/bootargs
Guest98 has joined #yocto
starblue has quit [Ping timeout: 255 seconds]
JulienP has joined #yocto
starblue has joined #yocto
rfuentess has quit [Remote host closed the connection]
<JulienP>
Hi again I changed my name ^^ So here is the content:
<DvorkinDmitry>
I have a problem setting pasword for my image. Using: EXTRA_USERS_PARAMS += "usermod -p \"$(openssl passwd -salt 'Xkm4BMK42Po.njiE' -6 '123' | sed -e 's/\$/\\\$/g')\" root;"
<DvorkinDmitry>
log.do_rootfs says: NOTE: img-tps-free: Performing usermod with [-R /workdir/build.tppg2/tmp/work/tp
Danct12 has quit [Read error: Connection reset by peer]
prabhakarlad has quit [Quit: Client closed]
amitk_ has joined #yocto
Daanct12 has quit [Ping timeout: 255 seconds]
luc4 has joined #yocto
<luc4>
Hello! I met a recipe that includes COMPATIBLE_MACHINE variables. Apparently, my build is not included, but I know it works (I'm using it). Is there a way to somehow force ignore that constraint and build anyway?
<rburton>
use a bbappend in your bsp layer to enable the recipe
<rburton>
if its something generic like "this recipe claims to be x86 specific but it works fine on arm" then upstream a patch too
<rburton>
(I had to do that for some UEFI pieces which claimed to be x86 specific)
<luc4>
rburton: the recipe does not include armv8 in 32bit mode apparently, probably because yocto itself does not seem to include it. So I do not think they would merge my patch. I'll read about bbappend though, thanks!
<rburton>
luc4: with my arm hat on: why are you using 32-bit code on an arm8?
<luc4>
rburton: because I have closed source drivers that are 32bit only
<rburton>
scream and shout at the vendor
<luc4>
rburton: they moved to more standard API, cannot complain about it
<rburton>
absolutely can if they're only releasing 32-bit binaries
<rburton>
only 15 years until Y2038!
amitk has quit [Remote host closed the connection]
amitk has joined #yocto
<luc4>
rburton: porting openmax to 64bit? Nah, I won't complain about it.
<luc4>
rburton: I removed the COMPATIBLE_MACHINE and I moved it to a bbappend. Will it rebuild the entire recipe? Cause it includes chromium... rebuilding means I can go on holiday for the next 10 hours :-)
amitk has quit [Remote host closed the connection]
amitk has joined #yocto
sakman has joined #yocto
amitk_ has quit [Quit: Lost terminal]
amitk_ has joined #yocto
Vonter has quit [Ping timeout: 255 seconds]
Xagen has joined #yocto
Vonter has joined #yocto
Xagen has quit [Client Quit]
<tlwoerner>
qschulz: what was the branch you wanted to rock-5b stuff on?
<qschulz>
tlwoerner: kirkstone but I guess it makes sense to backport to every branch between master and that one? don't know :)
zpfvo has quit [Ping timeout: 240 seconds]
zpfvo has joined #yocto
<tlwoerner>
qschulz: langdale is eol, so we can probably skip that one
amitk_ has quit [Ping timeout: 255 seconds]
<tlwoerner>
being a new-ish board, the older kernel and u-boot won't support it. we're already using linux-yocto-dev on master for kernel support
zpfvo has quit [Ping timeout: 240 seconds]
<qschulz>
tlwoerner: mmmm, would have been nice to split rock5b support from rk3588 support then :)
<qschulz>
so we could have backported only some part
zpfvo has joined #yocto
<qschulz>
tlwoerner: I guess I can send a backport once I work on this again (i'll have to do it in October)
<qschulz>
and i'll just send rk3588 support
<qschulz>
i mean it depends...
<qschulz>
maybe you don't want something you cannot test in your layers
<tlwoerner>
let me take a look and see what happens, maybe we'll get lucky
<qschulz>
though you already have px30 support and have no board to test it
<tlwoerner>
we already have px20 support for nothing that we have
<tlwoerner>
(in the layer)
<tlwoerner>
oops, px30
<qschulz>
yup, which I use in our public layer (thanks!)
Danct12 has joined #yocto
alessioigor has quit [Quit: alessioigor]
alessioigor has joined #yocto
Guest99 has quit [Quit: Client closed]
xmn has joined #yocto
Guest12 has joined #yocto
<Guest12>
Yocto index doesn’t list explicitly whether there is a lawyer for openmpi
<Guest12>
Found mpich but not exactly the same thing
<Guest12>
Any clue?
Guest12 has quit [Client Quit]
Kubu_work has quit [Quit: Leaving.]
<JaMa>
hehe like the "UNEXPECTED" INFO: UNEXPECTED SUCCESS: test_patches_upstream_status (common.CommonCheckLayer.test_patches_upstream_status)
prabhakarlad has joined #yocto
<gmorell>
LetoThe2nd: adding podman building against go 1.18 to an existing project is causing mender client to fail to build, we've added patches to backport the (single) go 1.18 feature being used to 1.17 and patched the go.mod for podman to use 1.17 but it's not.. great.
<LetoThe2nd>
gmorell: hmm. not great, agreed. and surprising: I have tested meta-mender against mickledore, which is go 1.20 and not run into any problems.
Guest72 has joined #yocto
<Guest72>
Storey was disconnected
<Guest72>
Was asking about openmpi
<Guest72>
Any clue?
<Guest72>
Meant sorry was disconnected
Xagen has joined #yocto
JulienP has joined #yocto
|Xagen has joined #yocto
Xagen has quit [Ping timeout: 255 seconds]
zpfvo has quit [Ping timeout: 264 seconds]
zpfvo has joined #yocto
Guest72 has quit [Ping timeout: 245 seconds]
zpfvo has quit [Ping timeout: 255 seconds]
zpfvo has joined #yocto
Xagen has joined #yocto
SuperGuest has joined #yocto
|Xagen has quit [Ping timeout: 255 seconds]
Danct12 has quit [Quit: What if we rewrite the code?]
<SuperGuest>
Hello world! I know nothing about rust. I am tring to integrate a python package with rsut dependencies. it is labourious. I am almost done, but I obtain an ld error at the end: /home/yoctouser/build/tmp/hosttools/ld:
<SuperGuest>
/home/yoctouser/build/tmp/work/armv8a-enchanted_tools-linux/python3-snips-nlu-parsers/0.4.3-r0/git/python/ffi/target/release/deps/libbacktrace_sys-31c993e55041b594.rlib: error adding symbols: file in wrong format
<SuperGuest>
| collect2: error: ld returned 1 exit status
<SuperGuest>
any idea ?
<qschulz>
SuperGuest: you're probably not crosscompiling properly
<qschulz>
probably missing to pass the target compilation flags to the build system of your application
<qschulz>
or you do pass it, but it's ignored by it
<rburton>
yeah, the hint is you can't use hosttools/ld in a cross build
linfax has quit [Ping timeout: 240 seconds]
Danct12 has joined #yocto
<SuperGuest>
I just specify the github SRC_URI/SRCREV link and 91 crate.io dependencies, and
<SuperGuest>
inherit python_setuptools3_rust, what should I do more? (I work with langdale)
<JaMa>
hmm is there a way to use something like THISDIR but pointing to the bbclass which is being parsed? (e.g. for referencing file relative to the .bbclass not the recipe which inheritted the bbclass)
<rburton>
SuperGuest: its entirely possible that the project is broken, not your recipe
leonanavi has quit [Remote host closed the connection]
leon-anavi has quit [Read error: Connection reset by peer]
alessioigor has quit [Quit: alessioigor]
luc4 has quit [Ping timeout: 240 seconds]
alessioigor has joined #yocto
kscherer has joined #yocto
<LetoThe2nd>
rburton: an upstream project broken for cross compilation? impossible, thats completely unheard of!
radanter has joined #yocto
jbo has quit [Ping timeout: 245 seconds]
justache has quit [Ping timeout: 258 seconds]
tct has joined #yocto
justache has joined #yocto
SuperGuest has quit [Quit: Client closed]
xantoz has quit [Ping timeout: 245 seconds]
xantoz has joined #yocto
ak77 has quit [Ping timeout: 252 seconds]
ak77 has joined #yocto
LocutusOfBorg has quit [Ping timeout: 245 seconds]
LocutusOfBorg has joined #yocto
iokill has quit [Ping timeout: 258 seconds]
iokill has joined #yocto
zpfvo has quit [Remote host closed the connection]
ecdhe has quit [Read error: Connection reset by peer]
ecdhe has joined #yocto
alessioigor has quit [Quit: alessioigor]
alessioigor has joined #yocto
demirok has quit [Ping timeout: 258 seconds]
demirok has joined #yocto
NishanthMenon has quit [Ping timeout: 252 seconds]
NishanthMenon has joined #yocto
<gmorell>
/w 2
florian_kc has quit [Ping timeout: 255 seconds]
radanter has quit [Remote host closed the connection]
florian has quit [Quit: Ex-Chat]
slimak has joined #yocto
JulienP has quit [Quit: Client closed]
goliath has quit [Quit: SIGSEGV]
<khem>
RP: I would like to upgrade test_patches_upstream_status to an error on meta-oe layercheck
<khem>
we have cleaned up all layers and its time !
<RP>
JPEW: interesting. Is that for a single connection or for streaming over them?
<JPEW>
That's for the stream API, which is the critical one
<JPEW>
ON the others, the performance doesn't matter as much
<RP>
JPEW: right, makes sense
brazuca has quit [Quit: Client closed]
<RP>
rburton: I think I worked out the numpy thing
<RP>
rburton: compare a build in a "reproducible" directory with a build in a directory with a different name like "qemux86-world". One will use O3, the other will not
<RP>
it is all a little bit sad
brazuca has joined #yocto
leon-anavi has joined #yocto
sotaoverride has quit [Ping timeout: 240 seconds]
amitk has quit [Ping timeout: 248 seconds]
flom84 has quit [Ping timeout: 255 seconds]
<abelloni>
RP: wth!
<RP>
abelloni: I'm nearly tempted to make this a challenge of finding the bug now I can explain how to trigger it :)
mvlad has quit [Remote host closed the connection]