mrnuke has quit [Read error: Connection reset by peer]
beneth has quit [Remote host closed the connection]
nemik has quit [Ping timeout: 252 seconds]
nemik has joined #yocto
nemik has quit [Ping timeout: 248 seconds]
nemik has joined #yocto
beneth has joined #yocto
GNUmoon2 has quit [Ping timeout: 268 seconds]
GNUmoon2 has joined #yocto
rob_w has joined #yocto
stuom has joined #yocto
<stuom>
Hi, I'm little lost how I add sources to src package. I install the sources in ${D}/usr/src/ and then add FILES_${PN}-src = "/usr/src/*" but I get installed-vs-shipped error?
mrpelotazo has quit [Ping timeout: 244 seconds]
mrpelotazo has joined #yocto
MWelchUK2 has joined #yocto
dkl_ has joined #yocto
wCPO has quit [Ping timeout: 244 seconds]
MWelchUK has quit [Ping timeout: 256 seconds]
MWelchUK2 is now known as MWelchUK
dkl has quit [Read error: Connection reset by peer]
wCPO has joined #yocto
nemik has quit [Ping timeout: 248 seconds]
nemik has joined #yocto
nemik has quit [Ping timeout: 256 seconds]
nemik has joined #yocto
mihai has quit [Quit: Leaving]
rob_w has quit [Ping timeout: 256 seconds]
rob_w has joined #yocto
frieder has joined #yocto
<wyre>
hi guys, I've got `IMAGE_FSTYPES += "tar.xz ubifs wic"` in my machine conf file ... but can't see any tar.xz file in my tmp/deploy/image/<my machine>/ folder
<wyre>
am I missing something?
lexano has quit [Ping timeout: 268 seconds]
kranzo has joined #yocto
Schlumpf has joined #yocto
zpfvo has joined #yocto
lexano has joined #yocto
denisoft81 has joined #yocto
zpfvo has quit [Quit: Leaving.]
zpfvo has joined #yocto
adrian_ has joined #yocto
mvlad has joined #yocto
adrian__ has quit [Ping timeout: 248 seconds]
zpfvo has quit [Ping timeout: 248 seconds]
ptsneves has joined #yocto
zeddii has quit [Ping timeout: 240 seconds]
zeddii has joined #yocto
goliath has quit [Quit: SIGSEGV]
vladest has joined #yocto
vladest has quit [Client Quit]
vladest has joined #yocto
zpfvo has joined #yocto
leon-anavi has joined #yocto
stuom has quit [Ping timeout: 252 seconds]
<qschulz>
wyre: what does bitbake-getvar -r <your-image> IMAGE_FSTYPES return?
zpfvo has quit [Ping timeout: 252 seconds]
zpfvo has joined #yocto
denisoft81 has quit [Quit: Leaving]
davidinuxsea has joined #yocto
camus1 has joined #yocto
camus has quit [Ping timeout: 248 seconds]
camus1 is now known as camus
camus has quit [Remote host closed the connection]
camus has joined #yocto
<wyre>
qschulz, it was because I was overriding IMAGE_FSTYPES in my image recipe 😞
<wyre>
but thank you very much, you are always very kind answering :)
<qschulz>
wyre: you could use :append in your machine.conf
<qschulz>
wyre: my pleasure :)
florian has joined #yocto
denisoft81 has joined #yocto
<wyre>
qschulz, yes, I think I'll use the :append in the image recipe, because I'd say this might depend on the image I'm actually building 😊
<qschulz>
wyre: mmmm why an :append in the recipe?
<qschulz>
can't you just += to it from you image recipe instead?
<wyre>
hmmm, not sure what's the criteria I should follow to use += or :append 🤔
nemik has quit [Ping timeout: 248 seconds]
nemik has joined #yocto
GNUmoon2 has quit [Remote host closed the connection]
GNUmoon2 has joined #yocto
<qschulz>
wyre: using an :append in a recipe would be weird to me (in a bbappend, different story)
<qschulz>
except if you want to do a machine specific append
<qschulz>
e.g., "I want this to happen to all ARM64 machines"
nemik has quit [Ping timeout: 256 seconds]
nemik has joined #yocto
camus has quit [Ping timeout: 248 seconds]
camus has joined #yocto
goliath has joined #yocto
<qschulz>
wyre: I would actually put IMAGE_FSTYPES:append = " ubi wic" in your machine configuration file
<qschulz>
because ubi and wic layouts are machine specific very often
camus has quit [Remote host closed the connection]
camus has joined #yocto
camus1 has joined #yocto
camus has quit [Ping timeout: 248 seconds]
camus1 is now known as camus
td12 has joined #yocto
<td12>
Hi,
<td12>
is there any way to apply the patches to kernel with patch tool not git?
<td12>
it doesn't work for kernel, since patching is done in kernel-yocto.bbclass using kgit-s2q
<wyre>
qschulz, ubi or ubifs? 🤔
<td12>
would it be accepted if we prepare the patch for kernel to also support PATCHTOOL variable?
<td12>
or there is some specific requirement so that kernel support only git?
mihai has joined #yocto
<landgraf>
td12: I don't know about kernel in particular. default PATCHTOOL value is guilt. quilt doesn't support binary patchers for example so if you want to use one you have to switch to different patchtool
ptsneves has quit [Ping timeout: 248 seconds]
<td12>
landgraf: changing a PATCHTOOL doesn't do anything because kernel recipe overrides do_patch task and doesn't check the value of PATCHTOOL and always uses git
hcg has joined #yocto
<landgraf>
td12: there was bug in patchtool implementation which prevented it from being used with nongit repositories (like tarballs). maybe because of than
<landgraf>
td12: which kernel recipe/file are you talking about?
<td12>
meta/classes/kernel-yocto.bbclass
nemik has quit [Ping timeout: 256 seconds]
nemik has joined #yocto
<rburton>
LetoThe2nd: hands on with yocto is full up already! annoying, but good. I get monday off I guess! :)
<landgraf>
yup. git is hardcoded everywhere.
<landgraf>
td12: maybe do_patch::append or bbappend with you custom patches ?
<landgraf>
or better postfunc
<td12>
hmm, that could be the way
<rburton>
td12: why do you need to change the patch tool?
camus has quit [Remote host closed the connection]
camus has joined #yocto
zpfvo has quit [Ping timeout: 252 seconds]
<td12>
rburton: We have a specific coverage test job that uses a target with prebuild image and rebuilds the kernel module with --coverage flag. But for the coverage we need 1 patch
<td12>
to be applied on kernel source code. Then we install the compiled (with --coverage) module and try to load it. The problem is that after we updated the poky version it start
<td12>
ed to fail and it's because now git is used for patch applying in kernel and it creates the commit thus changes the hash of the repository and the module load fails with "Exe
<td12>
c format error"
<rburton>
not sure how the version changing would result in exec format error
<rburton>
are you sure the module is actually the right executable format?
<td12>
yes, the only difference is the patch applied as a new commit. when I remove the patch it loads correctly
<rburton>
so why would the hash changing break the module?
<td12>
I think that's just the way kernel works, it breaks when the module is not compiled against the same version
kranzo has quit [Quit: Client closed]
<td12>
journal gives this output: `Aug 18 09:53:10 fct-0a kernel: <MODULE_NAME>: disagrees about version of symbol module_layout
<rburton>
rebuild the entire kernel then
starblue has quit [Ping timeout: 252 seconds]
<rburton>
you're changing the kernel ABI and then trying to load a module that doesn't match
<rburton>
if you change the kernel abi (enabling coverage) then the module needs to load in a kernel that has a matching abi
<td12>
yes, but that's only because the hash changed. the same would work if the patch would be applied without new hash
starblue has joined #yocto
<td12>
I know it's a bit nasty what we're doing but we'd like to not rebuild the whole image in the test that just tries to check only one module with --coverage
<qschulz>
td12: IIRC there's a kernel config/option to disable this check
<td12>
I think we don't want to disable that check totally, just workaround it in this single case
<td12>
and the problem is that applying the patch in the kernel source for the image would increase the size of the kernel, so we don't want to do that
<td12>
I mean, if there is no other way
florian_kc has joined #yocto
zpfvo has joined #yocto
<td12>
would it be accepted in upstream if we suggest the change to add the option for kernel to also use PATCHTOOL variable?
zpfvo has quit [Ping timeout: 252 seconds]
zpfvo has joined #yocto
<RP>
td12: some of our kernels have a hard requirement on using git for patch management, that can't be optional for them
<td12>
ok that's what I wanted to know, thanks
<td12>
I'll try to merge the patch to the repository and hope the size will not increase in a way it will break anything
<qschulz>
What happens if I have a recipe pointing to a git commit that gets removed upstream?
<qschulz>
Specifically, what happens to the git repo in DL_DIR once I update this git commit to point to one that exists upstream?
nemik has quit [Ping timeout: 256 seconds]
nemik has joined #yocto
<qschulz>
Is the git commit that existed locally only disappear from the local git repo in DL_DIR once I update the recipe to point to a new git commit?
<agherzan>
Any hint why we raise exception in oecore_update_localconf (NotImplementedError) as opposed to bb.fatal (and have a cleaner user-ending message)?
zpfvo has quit [Ping timeout: 252 seconds]
BenW_ has joined #yocto
BenW__ has quit [Ping timeout: 268 seconds]
td12 has quit [Quit: Client closed]
rob_w has quit [Remote host closed the connection]
<wyre>
hmmm, just can see image_types.bbclass and image_types_wic.bbclass 🤔
<wyre>
I can see this in the image_types.bbclass https://bpa.st/7I6A but not sure if this does mean that ubi is just ubifs ...
<wyre>
it seems more like ubi depends on ubifs ...
td95 has quit [Quit: Client closed]
adrian__ has joined #yocto
adrian_ has quit [Ping timeout: 268 seconds]
Schlumpf has quit [Quit: Client closed]
<landgraf>
qschulz: DL_DIR should be updated. once fether doesn't see commit in local mirror it will try upstream and fetch it from there if exists
<landgraf>
*fetcher
<landgraf>
it's a bit more complicated because of premirrors and mirrors but in general should work this way
<agherzan>
Re NotImplementedError. There seems to be a bug where sanity_check_conffiles doesn't catch NotImplementedError exceptions anymore. I see this on both kirkstone and master.
<qschulz>
landgraf: so the fetcher is only doing like a git fetch on usptream remotes, and not completely getting rid of the git repo and then cloning againb (that would be extremely inefficient :) )
<qschulz>
wyre: devtool reset <recipe>
<landgraf>
qschulz: it does git clone only if clonedir doesn't exist
<wyre>
qschulz, hmmm, still the same https://bpa.st/PL6A but devtool reset says there is not any recipe named virtual/kernel in the workspace 🤔
wkawka has joined #yocto
xmn has joined #yocto
<wyre>
may be this because I have a bbappend recipe which deploys some patches for this recipe? 🤔
camus has quit [Read error: Connection reset by peer]
camus has joined #yocto
prabhakarlad has quit [Quit: Client closed]
<qschulz>
wyre: I'm not sure virtual recipes work with devtool, might have to pass the actual name of the recipe
Schlumpf has joined #yocto
wkawka has quit [Quit: Client closed]
ptsneves has joined #yocto
zpfvo has quit [Ping timeout: 248 seconds]
zpfvo has joined #yocto
Minvera has joined #yocto
sakoman has joined #yocto
hcg has quit [Quit: Client closed]
<qschulz>
landgraf: thx
Estrella_ has joined #yocto
thomasd13 has quit [Ping timeout: 248 seconds]
wkawka has joined #yocto
jmiehe has quit [Quit: jmiehe]
seninha has quit [Remote host closed the connection]
<ernstp>
do you know if there was something special needed to support ubuntu-22.04 on master?
<ernstp>
with ubuntu-22.04 + dunfell I'm getting some KeyError: 'getpwuid()' that i'm not getting with older ubuntu + dunfell...
<vmeson>
ernstp: I don't remember any special requirements, it was more a matter of people having higher priority tasks and then getting that distro added to the Yocto test cluster. sakoman might know more especialy about kirkstone and ubu-22.04
<vmeson>
ernstp: earlier this week, sakoman mentioned that once the 4.0.x is done, he'll add ubu-22.04 to kirkstone, I'm not sure if he intends to supprot dunfell.
rber|res has quit [Ping timeout: 256 seconds]
<sakoman>
vmeson: ubuntu 22.04 is fine for kirkstone
<sakoman>
vmeson: I do regular local builds of dunfell on ubuntu 22.04 (i.e. bitbake -k world for smoke test)
<ernstp>
was there a good way to inspect file properties under pseudo?
goliath has quit [Quit: SIGSEGV]
<ernstp>
I'm getting this error with 22.04 but not with 18.04... strange!
<ernstp>
Path ./package/etc/apparmor.d is owned by uid 1000, gid 1000, which doesn't match any user/group on target. This may be due to host contamination.
<Xagen_>
as a part of one of my recipes, i'm doing an `install -d` to ${D}/usr/lib and subsequently an `install -Dm 644` of files to that directory that i create in the first step
rber|res has joined #yocto
<Xagen_>
they're also in FILES_${PN}, minus the ${D}
Xagen_ is now known as Xagen
<Xagen>
but, they don't end up in the packaged sdk
<Xagen>
any thoughts on what i've missed?
<Xagen>
i imagine it's something simple that has to do with packaging
<rburton>
Xagen: first, oe-pkgdata-util list-pkg-files -r [recipename], to verify that the packaging did what you expect
prabhakarlad has joined #yocto
<Xagen>
rburton: the non-nativesdk version of it looks right, but the nativesdk (which is the problem) gives a `ERROR: Unable to find any built runtime package` error
<rburton>
yeah you need to faff to use the nativesdk pkgdata
<rburton>
you need to tell pkgdata-util where the pkgdata directory is, which is tedious
<rburton>
but if the normal package is good then the nativesdk one should be
<rburton>
presumably you've added the package to the image you're building a sdk from
<Xagen>
interesting, i'll have to add that word to my repertoire
<Xagen>
i have
<Xagen>
the interesting thing is that the normal version of it, for the target system, packages fine but it's only referenced as a dependency for something else
<Xagen>
the nativesdk one for the host is directly referenced by the image recipe
<Xagen>
but doesn't package
<rburton>
how is it directly referenced in the image recipe?
* zeddii
grumbles. why do people continue to sing up to mailing lists with yocto@<foo> and then don't even sign their email with their name ?
* zeddii
writes a kill file.
zpfvo has quit [Ping timeout: 268 seconds]
zpfvo has joined #yocto
Estrella_ has quit [Remote host closed the connection]
hcg has joined #yocto
mihai has quit [Quit: leaving]
<Xagen>
rburton: with a `TOOLCHAIN_HOST_TASK_append`
rber|res has quit [Ping timeout: 268 seconds]
rber|res has joined #yocto
<Xagen>
rburton: i got the pkgdata util to work for the nativesdk version and the file list looks correct
elfenix[m] has quit [Quit: You have been kicked for being idle]
mux[m] has quit [Quit: You have been kicked for being idle]
RichardGolledge[ has quit [Quit: You have been kicked for being idle]
elfenix[m] has joined #yocto
<Xagen>
yet it didn't make it into the packaged sdk at the end
mux[m] has joined #yocto
RichardGolledge[ has joined #yocto
RichardGolledge[ has left #yocto [#yocto]
mux[m] has left #yocto [#yocto]
seninha has joined #yocto
<halstead>
Thanks kanavin. I have it cleaned up and I'll get it back to work. It appears the ab-janitor was still running but stuck on a git clone and had stopped removing files.
Andrew20 has joined #yocto
<kanavin>
halstead, thanks :)
<kanavin>
halstead, RP I still wonder about those extraordinarily long running world-alt jobs, which you can see e.g. here at the bottom of the list https://autobuilder.yoctoproject.org/typhoon/#/
<kanavin>
is there something special about world-alt that it gets deprioritized?
<RP>
kanavin: the docs builds don't build anywhere else so no sstate quite often there
<RP>
kanavin: just shows how helpful sstate is overall
<RP>
abelloni: "$ORIGIN/../../../mapi/shared-glapi:/home/pokybuild/yocto-worker/qa-extras2/build/build/tmp/work/x86_64-linux/llvm-native/14.0.6-r0/recipe-sysroot-native/usr/lib:/home/pokybuild/yocto-worker/qa-extras2/build/build/tmp/work/x86_64-linux/llvm-native/14.0.6-r0/recipe-sysroot-native/lib" is the mesa problem string
<kanavin>
22 hours, and still 15000/28000 done
<RP>
kanavin: that does seem slow :(
<RP>
kanavin: it is still going. The AB has been under a lot of load over the last day
adrian_ has joined #yocto
<Andrew20>
Does the archiver class only place source of certain packages in build/tmp/deploy/sources or does it do all packages for a given image?
adrian__ has quit [Ping timeout: 268 seconds]
nemik has quit [Ping timeout: 248 seconds]
nemik has joined #yocto
nemik has quit [Ping timeout: 268 seconds]
adams[1] has joined #yocto
nemik has joined #yocto
florian_kc has quit [Ping timeout: 256 seconds]
leon-anavi has quit [Quit: Leaving]
frieder has quit [Remote host closed the connection]
<JPEW>
Andrew20: I _think_ it should be all the recipes (not packages) that go into an image
<JPEW>
Or perhaps more precisely, all the recipes that generated packages that went into an image
<Xagen>
when packaging is done for a given image, does it package all the paths in FILES_${PN} from ${D} as the base directory for each?
<Andrew20>
JPEW: Thanks that is sort of what I assumed. However, if I image_install_append, for example vim, and rebuild it doesn't seem to place the vim source in the deploy dir. I'm wondering if it has something to do with the COPYLEFT_LICENSE_INCLUDE/EXCLUDE. I also notice that my archiver-sources in the work dir is empty for some recipes.
seninha has quit [Remote host closed the connection]
<ferlzc>
Hi guys! i'm trying to use the recipe libssh2 as an DEPEND for to build another recipe. I'm using yocto Kirkstone version. However during ./configure i get the following errors: checking for ssh2 support... yes, shared
<ferlzc>
| checking for ssh2 files in default path... not found