<zeddii>
moto-timo: I managed to make ceph get past the python3 detection issues, now it blows up building c programs after about 12% of the build. definitely giving up for now. It can't find libgcc, etc, which should be all available via the cmake bbclass
qschulz has quit [Remote host closed the connection]
qschulz has joined #yocto
lexano has quit [Ping timeout: 264 seconds]
goliath has quit [Quit: SIGSEGV]
nerdboy has quit [Remote host closed the connection]
rfs613 has quit [Ping timeout: 260 seconds]
nerdboy has joined #yocto
davidinux has quit [Ping timeout: 264 seconds]
davidinux has joined #yocto
lexano has joined #yocto
rfs613 has joined #yocto
vlrk has quit [Quit: Client closed]
starblue has quit [Ping timeout: 246 seconds]
starblue has joined #yocto
<paulg>
translation: "you get what you get." :-P
pedrowiski has joined #yocto
jclsn has quit [Ping timeout: 268 seconds]
jclsn has joined #yocto
mulk has quit [Ping timeout: 252 seconds]
vlrk has joined #yocto
vlrk94 has joined #yocto
vlrk has quit [Ping timeout: 250 seconds]
mulk has joined #yocto
vlrk94 has quit [Quit: Client closed]
Xagen has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
vlrk94 has joined #yocto
vlrk26 has joined #yocto
vlrk94 has quit [Ping timeout: 250 seconds]
vlrk26 has quit [Quit: Client closed]
ablu has quit [Read error: Connection reset by peer]
ablu has joined #yocto
paulg has quit [Ping timeout: 246 seconds]
vlrk26 has joined #yocto
johnmcnuggets has quit [Ping timeout: 255 seconds]
Chaser has joined #yocto
vlrk26 has quit [Quit: Client closed]
vlrk26 has joined #yocto
vlrk26 has quit [Client Quit]
jmd has joined #yocto
alessioigor has joined #yocto
guestkati has joined #yocto
alessioigor has quit [Remote host closed the connection]
alessioigor has joined #yocto
mckoan|away is now known as mckoan
<mckoan>
good morning
rob_w has joined #yocto
xmn has quit [Ping timeout: 264 seconds]
mkazantsev has joined #yocto
Tyaku has joined #yocto
frieder has joined #yocto
<alessioigor>
good morning to you
Michael_Guest has joined #yocto
goliath has joined #yocto
rfuentess has joined #yocto
rfuentess has quit [Read error: Connection reset by peer]
rfuentess has joined #yocto
zpfvo has joined #yocto
sa7mfo has joined #yocto
zpfvo has quit [Ping timeout: 268 seconds]
<sa7mfo>
Hello! I do not have a shell (/bin/sh is missing) unless dev-pkgs is specified as an image feature. My image file inherit from core-image and require core-image-minimal.bb. Is this expected?
jmd has quit [Read error: Connection reset by peer]
zpfvo has joined #yocto
leon-anavi has joined #yocto
mkazantsev has quit [Quit: Leaving]
belsirk has joined #yocto
rfuentess has quit [Ping timeout: 256 seconds]
Tyaku has quit [Quit: Lost terminal]
<LetoThe2nd>
sa7mfo: unless your specific setup does something weird, core-image-minimal definitely comes with a shell. I'd start by checking your BSP/distro layers.
<sa7mfo>
Thank you, I will check those layers
prabhakalad has quit [Quit: Konversation terminated!]
prabhakalad has joined #yocto
guestkati has quit [Ping timeout: 250 seconds]
kpo has quit [Ping timeout: 246 seconds]
<jdiez>
is it possible to keep the sstate of a package build with a slightly different recipe? i find myself changing a dependency's recipe back and forth while troubleshooting, and this causes bitbake to rebuild the dependency every time I change it - even though I had already built it with that config before
amitk has quit [Remote host closed the connection]
Guest18 has joined #yocto
belsirk has quit [Remote host closed the connection]
mvlad has joined #yocto
luc4 has joined #yocto
<luc4>
Hello! I have a doubt about the difference between DEPENDS and RDEPENDS. If I state that my recipe DEPENDS on, let’s say, libconfig, is libconfig also added to my image? And is it also added to the SDK?
<mckoan>
luc4: quick answer, yes
<luc4>
mckoan: thanks! So should I expect libconfig headers to be present in the SDK?
<mckoan>
luc4: DEPENDS means something required by recipe A to build, RDEPENDS something required by the final artifact generated by recipe A to run on the target
<luc4>
mckoan: from your last sentence, I’d say I should add libconfig to both DEPENDS and RDEPENDS. But I guess DEPENDS means “required to build” and also “required to run”. Is this correct?
vlrk26 has quit [Quit: Client closed]
<mckoan>
luc4: yes, sorry. DEPENDS means something required by recipe A to build and then to run
<luc4>
mckoan: then it means I probably did something else wrong cause I can’t find the headers in the sdk… thanks!
<mckoan>
luc4: are you using bitbake -c populate_sdk ?
<luc4>
mckoan: actually I added inherit populate_sdk populate_sdk_qt6 in my recipe. Is this different?
<mckoan>
luc4: never used in this way
amitk has joined #yocto
<luc4>
mckoan: the command you provided did not include the Qt tools that I need
<luc4>
mckoan: that is why I’m using this other technique
starblue has quit [Ping timeout: 252 seconds]
starblue has joined #yocto
mrpelotazo has quit [Read error: Connection reset by peer]
mrpelotazo has joined #yocto
kpo has joined #yocto
Saur80 has quit [Quit: Client closed]
Saur80 has joined #yocto
guestkati has quit [Quit: Client closed]
vlrk26 has joined #yocto
vlrk has joined #yocto
guestkati has joined #yocto
sa7mfo has quit [Quit: Client closed]
vlrk26 has quit [Ping timeout: 250 seconds]
<Ad0>
I have a defconfig I want to convert to fragment instead. how can I trick the system into applying my defconfig and then output a fragment instead? it will be huge for starters for sure, but I want to reduce it till it brakes making the fragment smaller
<Ad0>
breaks*
kpo has quit [Ping timeout: 264 seconds]
florian has quit [Ping timeout: 255 seconds]
vmeson has quit [Ping timeout: 268 seconds]
florian_kc has joined #yocto
vlrk has quit [Quit: Client closed]
vlrk has joined #yocto
warthog9 has quit [Quit: Leaving]
warthog9 has joined #yocto
Saur11 has joined #yocto
Saur80 has quit [Ping timeout: 250 seconds]
guestkati has quit [Quit: Client closed]
ptsneves has joined #yocto
vmeson has joined #yocto
ptsneves has quit [Ping timeout: 260 seconds]
vlrk has quit [Quit: Client closed]
paulg has joined #yocto
luc4 has quit [Ping timeout: 264 seconds]
luc4_ has joined #yocto
Guest18 has quit [Quit: Client closed]
ben62 has joined #yocto
<yocton>
Ad0: "make savedefconfig" will output the minimal config generating the current .config (removing all the default configs). This should provide a good starting point to create fragments.
<kanavin>
RP: I have a version update patchbomb that should be ready and a-full tested in a few minutes. Just wondering if I should send it or hold it, as it does close the gap with upstream significantly. https://git.yoctoproject.org/poky-contrib/log/?h=akanavin/package-version-updates
luc4 has joined #yocto
luc4_ has quit [Ping timeout: 246 seconds]
<RP>
kanavin: send I guess but this means I'll have just more rounds of testing before I can build M3 :/
<rburton>
i'd say an upgrade bomb now is too late for M3
* RP
is torn on it
<rburton>
i'm not :)
<kanavin>
I'll send, if only to avoid other people spending cycles on those updates
vlrk has joined #yocto
vlrk62 has joined #yocto
<kanavin>
but first, a bicycle ride - still need to read through the final versions to ensure no test leftovers, correct patch metadata, decent descriptions etc.
vlrk has quit [Ping timeout: 250 seconds]
luc4_ has joined #yocto
luc4 has quit [Ping timeout: 256 seconds]
luc4_ is now known as luc4
rfuentess has joined #yocto
ak77 has quit [Ping timeout: 252 seconds]
ak77 has joined #yocto
Saur11 has quit [Quit: Client closed]
vlrk62 has quit [Quit: Client closed]
Saur11 has joined #yocto
johnmcnuggets has joined #yocto
goliath has quit [Quit: SIGSEGV]
vlrk62 has joined #yocto
vlrk62 has quit [Client Quit]
vlrk62 has joined #yocto
ben62 has quit [Quit: Client closed]
<sakoman>
kanavin: perhaps as a middle ground break it into to patch bombs: one with important major version updates to squeeze in now, and another with point releases that I can take as part of stable branch update.
<sakoman>
s/to/two/
alperak has joined #yocto
Michael_Guest has joined #yocto
florian_kc has quit [Ping timeout: 268 seconds]
sev99 has joined #yocto
johnmcnuggets has quit [Changing host]
johnmcnuggets has joined #yocto
<sakoman>
if you decide to do this, tag the second set with scarthgap so it ends up in my patchworks queue
<kanavin>
sakoman, unfortunately I don't look into changelogs. My job is covering for inactive recipe maintainers who don't do their job on time.
<kanavin>
I can't possibly triage 50+ updates this way, sorry.
justache has joined #yocto
<sakoman>
kanavin: Ok, just a thought
<sakoman>
in that case I vote for taking them now so *I* don't have to wade through 50 update changelogs to see if it is safe to take them for stable ;-)
Saur11 has joined #yocto
goliath has joined #yocto
Saur6 has joined #yocto
* sakoman
curses companies who copy me on every internal CI failure. Like I need more useless clutter in my inbox!
Martin0815 has joined #yocto
johnmcnuggets has quit [Ping timeout: 272 seconds]
Martin0815 has quit [Client Quit]
justache has quit [Remote host closed the connection]
Martin12345 has joined #yocto
justache has joined #yocto
<rburton>
sakoman: replying and pointing out their leaking internal infrastructure details usually makes them stop :)
<sakoman>
rburton: yes, I do that. But another company soon steps up to fill the void!
<kanavin>
on the other hand, that way we can find out stealth yocto freeloaders :)
<kanavin>
and approach them about project membership :D
Saur11 has quit [Ping timeout: 250 seconds]
<sakoman>
kanavin: there is that :-)
<Martin12345>
Hello community, I need to integrate a closed source software package written in rust. Besides cross compiling inside yocto, I would also need support for cross compiling rust with the yocto sdk. This is where I stumble. Which way should I go: using the oe-core rust integration or meta-rust? I am on yocto langdale btw. Thanks :]
<kanavin>
Martin12345, meta-rust is generally obsolete
<rburton>
Martin12345: use the integrated rust
<rburton>
also langdale is end-of-life so I recommend you move off that
Saur75 has joined #yocto
<Martin12345>
Thanks kanavin and rburton (y) can you recommand some specific rust in yocto documentation? In the *.rst files of my yocto checkout I only found basic infos, somewhat missing a "full documentation".
<rburton>
if you want rust in the sdk then i expecting add nativesdk-rust to the SDK will be sufficient
Saur6 has quit [Ping timeout: 250 seconds]
<kanavin>
Martin12345, I'm not sure what there is to be honest. It should work like any other sdk toolchain.
rob_w has quit [Remote host closed the connection]
<mckoan>
Ad0: is the keyboard working?
<Martin12345>
Great, thanks rburton kanavin :]
<Ad0>
not yet. I compared the localectl and both are C
<Ad0>
so it has to be something else
<Ad0>
all the keyboard mappings are already there
<Ad0>
I tried to set it manually, no diff
<Ad0>
System Locale: LANG=C
<Ad0>
it should just detect mye keyboard and just ... work
<Ad0>
I made no changes to this in kirkstone vs dunfell
justache has quit [Remote host closed the connection]
kpo has joined #yocto
justache has joined #yocto
Saur75 has quit [Quit: Client closed]
Vonter has quit [Quit: WeeChat 4.2.1]
Saur75 has joined #yocto
<fullstop>
Hi all, I have a third party layer which has do_deploy:append:machine(), and do_deploy:append(). The machine specific one generates some files, and the do_deploy:append one uses them. It looks like it is executing the do_deploy:append before the machine specific task (mickledore). Is there a preferred method of controlling order?
<kanavin>
fullstop, probably the easiest is to convert the part that needs to run last into a task
<kanavin>
there is no preferred method for dealing with badly written third party layers though
<moto-timo>
zeddii: so now I think it needs some education on AR but I am not a compiler expert lol
<moto-timo>
zeddii: also, there is no way that all four of those variables should all be the same LDFLAGS ? I'm assuming...
<fullstop>
thanks, kanavin, I will update the layer.
<zeddii>
it took me about an hour to even convince myself that the toolchain.cmake was being consulted, and then I did determine it was. I went from knowing almost nothing about cmake, to knowing slightly more yesterday
<moto-timo>
same here
<moto-timo>
it's just that I've done this blind guessing game before
<moto-timo>
lol
<zeddii>
moto-timo: I saw so many different, very similar, superceeded variables in my travels that I just kept putting them all in until the behaviour changes.
<moto-timo>
a long time ago someone even advised me to use the CMake gui to set up the equivalent of the defconfig. I did it that time. It felt dirty.
<LetoThe2nd>
tgamblin: sound advice!
<zeddii>
like do you, or don't you need the <VARIABLE>:PATH=<value> or not, it doesn't error if you don't have the PATH= .. is it silently ignoring it ? when you are flailing at variables you don't understand, that would be nice to know :)
<moto-timo>
LetoThe2nd: tgamblin and I lost several years of our lives getting k8s to work and then realizing it wouldn't be sustainable (or even work in some cases for yocto-tekton)
<moto-timo>
zeddii: I don't know what we need for AR... there is CMAKE_AR (which seem like it is already passing in our *-ar)
<moto-timo>
"We know better than you how to compile our software, so we will just silently fail and leave you guessing"
<moto-timo>
No. No you do not.
<LetoThe2nd>
moto-timo: actually neither getting it to work nor understanding it is actually required, just a general "feeling" for it (beyond the obvious icky one, of course)
<moto-timo>
LetoThe2nd: Buy more beer immediately.
<LetoThe2nd>
moto-timo: nah not a drinking day.
<moto-timo>
yet
<LetoThe2nd>
generally.
<moto-timo>
we call it h8s now for a very solid reason (or 1000 reasons)
<moto-timo>
zeddii: maybe I created this problem with the wrong CMAKE_STATIC_LINKER_FLAGS
Xagen has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<LetoThe2nd>
fun story from my YP life: 1) I get a fancy board 2) I look at bsp docs, says latest "use mickledore", or "legacy kirkstone". 3) me builds kirkstone, doesn't boot. 4) me builds mickledore, boots. 5) me tells vendor 6) vendor answers "yeah we know, you got silicon v2.0, its not working with anything else than mickledore" -EWONTFIX
<moto-timo>
-EWONTBUY
<moto-timo>
fun story, variscite finally added DART-6UL support so the old meta-mender-community variscite recipes could work again.
<moto-timo>
(unrelated)
<LetoThe2nd>
nice
<moto-timo>
while a bit late for what I cared about, at least they did it :)
<zeddii>
my kingdom for a clear error message on badly formatted options.
<moto-timo>
all the settings are in `../build/CMakeCache.txt` if you do `bitbake -c devshell ceph`
<moto-timo>
they really like their ALL_CAPS_YELLING_AT_YOU_VARIABLES_WITH_EXTRA_LONG_NAMES_THAT_ARE_NOT_OBVIOUS
<zeddii>
yesterday I attempted to learn the syntax to put a print statement into the template files for python.
<zeddii>
I couldn't stop thinking ... why, oh ... why!?
vlrk62 has quit [Quit: Client closed]
joekale has quit [Ping timeout: 256 seconds]
joekale has joined #yocto
<moto-timo>
so it seems like CMAKE_STATIC_LINKER_FLAGS should be something like 'rcs' (whatever we should be passing to ${AR})
vlrk62 has joined #yocto
luc4 has quit [Quit: luc4]
<moto-timo>
zeddii: clearly also not correct `x86_64-poky-linux/13.2.0/ar: rcs: No such file or directory`
rfuentess has quit [Remote host closed the connection]
sakoman has quit [Ping timeout: 255 seconds]
<RP>
moto-timo: I saw "rcs" and wondered if you were ok at first :)
<moto-timo>
RP: lol
* moto-timo
has forgotten so much about compilers
<moto-timo>
use it or lose it
<moto-timo>
also, never had to write one in Comp Sci 101, because I was not a CS major
<RP>
moto-timo: what is CS? :)
vlrk22 has joined #yocto
<fray>
lol I taught myself lex and yacc.. but have pretty much forgotten it all at this point
<moto-timo>
the author of "The Lion, the Witch and the Wardrobe"?
<moto-timo>
number of core YP/OE devs that are physics or hard sciences background 99.9%
<moto-timo>
lol
<RP>
"It isn't rocket science" - no, I was taught that :)
<moto-timo>
lol
<rburton>
looking at my side hussle whilst a build runs. anyone have an explanation for why linux-firmware wants to make its own -license packages?
<moto-timo>
"It isn't nuclear fusion"
<RP>
rburton: I think some of the firmware packages require the license be installed or something
* vmeson
sees a keyword...
<RP>
moto-timo: was taught that too :)
vlrk62 has quit [Ping timeout: 250 seconds]
<moto-timo>
"It isn't quantum mechanics"
<moto-timo>
lol
<vmeson>
moto-timo: stop with the keywords! ;-)
<RP>
moto-timo: relativistic quantum mechanics is more interesting ;-)
<moto-timo>
ha. mission accomplished "beetlejuice, beetlejuice, vmeson"
* RP
has a section in a file on that too...
justache has joined #yocto
<rburton>
RP: but we have license.bbclass and the -lic packages for that
<rburton>
like the firmware isn't so special that it has to ship its own copy of GPL.txt
starblue has quit [Ping timeout: 252 seconds]
mckoan is now known as mckoan|away
sakoman has joined #yocto
sakoman has quit [Read error: Connection reset by peer]
<moto-timo>
Dear lazyweb, what the heck are the arguments supposed to be that are passed to ${AR} (or better yet, where do they come from in the bbclasses)
<moto-timo>
grepping for two characters is not likely to give me anything useful
<rburton>
depends on the use, so there's no ARFLAGS or anything defined centrally
<rburton>
ie ${AR} -rcs would make sense
<rburton>
(create, add, update index)
yocti has quit [Ping timeout: 255 seconds]
<moto-timo>
rburton: thank you. I see that when I grep for ${AR} in openembedded-core/meta
<moto-timo>
well, without the - apparently...
yocti has joined #yocto
alperak has quit [Quit: Client closed]
frieder has quit [Remote host closed the connection]
zpfvo has quit [Remote host closed the connection]
<smokey>
This was working, with an older kirkstone commit of https://git.yoctoproject.org/poky. I checked out yocto-4.0.16 and now see this issue.
<khem>
zeddii: does thrift export .cmake files ? if so check that. I think thats what its trying to find from
<rburton>
would it be after thirft-native too?
sakoman has quit [Read error: Connection reset by peer]
<smokey>
rburton: my auto.conf has: INHERIT += "sign_rpm_ext"
<khem>
Smokey: add meta-integrity before meta-oe perhaps its being inherited in global INHERIT maybe and it becomes chicken-egg problem for add-layers
<rburton>
that class is horrible as a global class
leon-anavi has quit [Quit: Leaving]
flom84 has joined #yocto
flom84 has quit [Remote host closed the connection]
joekale has quit [Ping timeout: 255 seconds]
<smokey>
khem: afraid that didn't help. Does this parse error imply that the inherited bbclass is malformed?
florian_kc has joined #yocto
<khem>
can you comment it out in auto.conf during bitbake-layers add
<smokey>
khem: yes, that gets me to where I can build. As a work-around I could just write bblayers.conf directly instead of add-layer'ing them.
Saur75 has quit [Quit: Client closed]
Saur75 has joined #yocto
<smokey>
thanks for your help khem
florian_kc has quit [Ping timeout: 255 seconds]
vlrk22 has quit [Quit: Client closed]
starblue has joined #yocto
smokey has quit [Quit: Client closed]
vlrk22 has joined #yocto
vlrk22 has quit [Client Quit]
alperak has joined #yocto
<RP>
rburton: some of the firmware licenses may be :/
<khem>
smokey, its better to write your auto.conf at end of setting up your workspace
<khem>
you are doing fine by using add-layers tools
smokey has joined #yocto
dgriego has quit [Ping timeout: 264 seconds]
florian_kc has joined #yocto
khimaros has joined #yocto
<khimaros>
hi all, i'm trying to build meta-raspberrypi using the provided kas config and it is failing with a missing dependency. i'm new to yocto and would appreciate any help troubleshooting or building up my mental model.
dgriego has joined #yocto
<khimaros>
the specific error is: `Nothing RPROVIDES 'linux-firmware-rpidistro-bcm43456' (but layers/poky/meta/recipes-core/packagegroups/packagegroup-base.bb RDEPENDS on or otherwise requires it)`
jmd` has quit [Remote host closed the connection]
<khimaros>
okay, more progress made with the non-kas method. maybe the configuration there is out of date or something?
<moto-timo>
zeddii: I'm realizing the cmake/modules in ceph source tree is not necessarily what we are resolving...
Haxxa has quit [Quit: Haxxa flies away.]
<zeddii>
oh ?
Haxxa has joined #yocto
<moto-timo>
Findthrift.cmake vs FindThrift.cmake
<moto-timo>
or those are vendored deeper
<moto-timo>
so thrift_HOME in the former and Thrift_HOME in the later
<khimaros>
documentation seems to be really broken for meta-raspberrypi. none of the documented MACHINE= types are working.
<moto-timo>
several other libs build with 'thrift' maybe that is the oddball or just the first to fail
ptsneves has joined #yocto
Chaser has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
prabhakalad has quit [Ping timeout: 252 seconds]
prabhakalad has joined #yocto
prabhakalad has quit [Ping timeout: 268 seconds]
Vonter has quit [Ping timeout: 252 seconds]
Vonter has joined #yocto
ptsneves has quit [Quit: ptsneves]
vladest has quit [Remote host closed the connection]
alperak has quit [Quit: Client closed]
sev99 has quit [Quit: Client closed]
starblue has quit [Ping timeout: 264 seconds]
vladest has joined #yocto
JaMa has quit [Ping timeout: 272 seconds]
JaMa has joined #yocto
alessioigor has quit [Ping timeout: 264 seconds]
starblue has joined #yocto
<khimaros>
found a bug with poky (libxcrypt) on debian-trixie which is resolved by inheriting `perlnative` in the build recipe
Xagen has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
Vonter_ has joined #yocto
Vonter has quit [Ping timeout: 252 seconds]
<khem>
khimaros: if you can send a patch to relevant release branch and see if we need it on master as well. Or atleast file a ticket in bugzilla might be helpful too
kpo has joined #yocto
<khimaros>
definitely needed on hardknott (trying to reproduce an old rpi4 image). i'm guessing master has moved on from using the deprecated switch (given/when) semantics and probably works fine with perl-5.38 -- i was just looking at the contribution process and will try to get a patch submitted at least for hardknott
roussinm has joined #yocto
Saur75 has quit [Quit: Client closed]
<Xogium>
huh, isn't hardknott OLD and no longer supported ?
<Xogium>
maybe I mix it up with another release
Saur75 has joined #yocto
<rburton>
hardknott is ancient yes
<khimaros>
Xogium: it is probably quite old. i think maybe 3 years?
<khimaros>
maybe it's not worth bothering with a patch then. i'll just patch locally and carry on. maybe better to focus my effort on updating the top level recipes to a newer branch. i'm picking up some unmaintained code for a personal project.
<Xogium>
I mean since the version is no longer maintained, I don't think your patch would be applied
Saur75 has quit [Quit: Client closed]
mvlad has quit [Remote host closed the connection]
Saur75 has joined #yocto
<roussinm>
Anyone using efibootguard on nanbield? We having issue running it on QEMU. It's basically the same error as: https://github.com/systemd/systemd/issues/11541 but all the makefiles are correct in efibootguard.
<roussinm>
It works on Mickledore. So it has to do with toolchain somehow.
<roussinm>
No, not yet. I think it's probably time... at this point.
<vmeson>
K. I had never heard of efibootguard or the meta-efibootguard layer. Perhaps if you think it's generally useful ,you can ask Andreas to get it added to the layers.openembedded.org index.
* vmeson
heads off for the day.
cambrian_invader has quit [Ping timeout: 268 seconds]