<agherzan>
How would one handle machine features that are based on peripherals? For example adding to a non-rtc machine, an rtc module.
<agherzan>
Is it expected for such a setup to always come with a separated machine?
florian_kc has quit [Ping timeout: 272 seconds]
pasherring has quit [Read error: Connection reset by peer]
pasherring has joined #yocto
jclsn7 has quit [Ping timeout: 240 seconds]
jclsn7 has joined #yocto
dev1990 has quit [Quit: Konversation terminated!]
jclsn7 has quit [Ping timeout: 240 seconds]
jclsn7 has joined #yocto
chep has quit [Read error: Connection reset by peer]
chep` has joined #yocto
chep` is now known as chep
jclsn7 has quit [Ping timeout: 272 seconds]
jclsn7 has joined #yocto
LocutusOfBorg has quit [Remote host closed the connection]
jclsn7 has quit [Ping timeout: 272 seconds]
LocutusOfBorg has joined #yocto
jclsn7 has joined #yocto
Bardon_ has quit [Ping timeout: 252 seconds]
jclsn7 has quit [Ping timeout: 272 seconds]
jmiehe has quit [Quit: jmiehe]
Wouter0100 has quit [Remote host closed the connection]
Wouter0100 has joined #yocto
jclsn7 has joined #yocto
sakoman has quit [Quit: Leaving.]
rber|res has joined #yocto
RobertBerger has quit [Ping timeout: 272 seconds]
jclsn7 has quit [Quit: Ping timeout (120 seconds)]
jclsn7 has joined #yocto
Bardon has joined #yocto
jclsn7 has quit [Ping timeout: 256 seconds]
jclsn7 has joined #yocto
jclsn7 has quit [Ping timeout: 240 seconds]
jclsn7 has joined #yocto
pasherring has quit [Remote host closed the connection]
pasherring has joined #yocto
jclsn7 has quit [Ping timeout: 256 seconds]
jclsn7 has joined #yocto
amitk has joined #yocto
jclsn71 has joined #yocto
jclsn7 has quit [Ping timeout: 272 seconds]
jclsn71 is now known as jclsn7
pasherring has quit [Remote host closed the connection]
pasherring has joined #yocto
AKN has joined #yocto
Thorn has quit [Ping timeout: 272 seconds]
Thorn has joined #yocto
<AKN>
Hi, Everyone Trying to brinup the pruss core in yocto image, linux-ti-kernel 5.10.100, modified the am5729-beaglebone-ai.DTS according to the one in the Beaglebone git page build was successful but couldn't get the pruss core Shows error cannot lookup hwmod pruss1
<AKN>
Include pru-icss , pruss-lld recipes still couldn't brinup the pru cores
Starfoxxes has quit [Ping timeout: 260 seconds]
Starfoxxes has joined #yocto
<moto-timo>
RP: looks like there must be something in the wheels (guessing timestamps) that is breaking reproducible build :/
<mcfrisk>
thanks Saur[m] RP, I'll have a look and possibly cherry-pick to my dunfell tree
<mcfrisk>
oops, need coffee
davidinux has joined #yocto
davidinux has quit [Ping timeout: 256 seconds]
davidinux has joined #yocto
osama1 has joined #yocto
osama2 has joined #yocto
osama1 has quit [Ping timeout: 256 seconds]
GNUmoon has quit [Ping timeout: 240 seconds]
osama2 has quit [Ping timeout: 240 seconds]
rob_w has joined #yocto
alessioigor has joined #yocto
alessioigor has quit [Client Quit]
Wouter0100 has quit [Ping timeout: 272 seconds]
Wouter0100 has joined #yocto
Etheryon has joined #yocto
mvlad has joined #yocto
mvlad_ has joined #yocto
mvlad_ has quit [Client Quit]
GNUmoon has joined #yocto
GNUmoon has quit [Ping timeout: 240 seconds]
GNUmoon has joined #yocto
Guest61 has joined #yocto
Guest61 has left #yocto [#yocto]
goliath has joined #yocto
lucaceresoli has joined #yocto
mckoan|away is now known as mckoan
kevinrowland has quit [Ping timeout: 256 seconds]
alessioigor has joined #yocto
alessioigor has quit [Client Quit]
rfuentess has joined #yocto
<Etheryon>
Morning, is there any way I can investigate why rfkill is soft-blocking my device?
kanavin_ has quit [Remote host closed the connection]
creich_ has joined #yocto
<LetoThe2nd>
yo dudX
creich has quit [Ping timeout: 250 seconds]
tre has joined #yocto
<mckoan>
good morning, probably not
gsalazar has joined #yocto
gsalazar_ has joined #yocto
gsalazar_ has quit [Remote host closed the connection]
gsalazar_ has joined #yocto
gsalazar_ has quit [Remote host closed the connection]
gsalazar has quit [Ping timeout: 240 seconds]
Guest5255 has joined #yocto
leon-anavi has joined #yocto
<zyga[m]>
indeed
<Guest5255>
Hi, how can I test if a certain out-of-tree kernel module actually works (the module is visible with lsmod)?
<qschulz>
michaelo: they build because the SPHINXOPTS patch was merged c.f. 931d409b255a85f2217ca093d8391a678ce00ddb
<LetoThe2nd>
Guest5255: by testing it? That totally depends on what the module is supposed to do.
<LetoThe2nd>
Guest5255: 1) apply temperature 2) read value :)
<Guest5255>
I've checked the /sys/bus devices directory
<Guest5255>
`I've read that there are specific testing tools for kernel modules, I have the module connected and I want to read out the temperature using the /sys/bus directory, that is reading directly from the kernel module virtual filesystem.
<LetoThe2nd>
Guest5255: i mean, what you are talking about is generic driver development. this is a) absolutely hardware dependent, and b) not related to the YP at all. would you mind looking at literature that covers this topic, like LDD3 first?
<Guest5255>
LetoThe2nd You are absolutely right, I already did but I will check again, thanks!
<RP>
So we have a likely kernel crash that needs investigating, a hang in ltp in the controllers test, a problem with libsecret and meson autodownloading when docs are enabled and a reproducibility issue with the python changes
<LetoThe2nd>
RP: perfect time to start day drinking.
<LetoThe2nd>
or a mosh pit.
<LetoThe2nd>
or both.
<Guest5255>
or green tea :)
<landgraf>
is there a way to specify task dependecies in form "depend on task if exists" ? do_mytask[depends] = $PN:do_install if $PN:do_install else ""
booboo1212[m] has quit [Quit: You have been kicked for being idle]
<qschulz>
landgraf: I'm not sure you can know from a recipe if another recipe has a task?
<qschulz>
landgraf: what's the actual usecase?
booboo1212[m] has joined #yocto
<landgraf>
qschulz: It's class wide function. I'm trying to run abi checker function for each recipe
<landgraf>
but some of them like buildtools-tarball.bb don't have do_install and bitbake is not happy about it :)
booboo1212[m] has left #yocto [#yocto]
<Guest5255>
What is the difference (advantages/disadvantages) between generating root filesystem with busybox and with yocto?
<RP>
qschulz, michaelo: mhalstead added some patches to fix the version switch issues
<RP>
(on the docs)
<landgraf>
qschulz:postfuncs work but it requires do_install to be rerun for abi_check or stamp disabled which I try to avoid and add do_abicheck with nostamp
<qschulz>
RP: yeah, I hate those :) We need to do something about it so it's either not forgotten or not required
<qschulz>
hate is a strong word
<rber|res>
Guest5255, There are different possibilities to write a driver for a temperature sensor. What kind of driver is yours? industrial IO?
<qschulz>
I'm disappointed we're still forgetting about those changes before tagging
<qschulz>
so it means we need to do something about it :)
<RP>
qschulz: agreed
<RP>
qschulz: trouble is it isn't tagging, it needs to be done before the release is built :/
<RP>
qschulz: but if we do it then, the website can update too soon
<Guest5255>
rber|res yes industrial IO, I am searching for two days (books and internet) but they all discuss the basics but not actually how to see if it works (if I write an almost empty file then lsmod will still show the module!) :)
<RP>
(needs to be before the release is built as that is the docs revision that is used
<RP>
qschulz: perhaps we can have an addition tag which is used in preference for the site if present?
Dracos-Carazza_ is now known as Dracos-Carazza
kanavin has joined #yocto
<rber|res>
Guest5255, Do you see something related to your driver in /sys/bus/iio/devices/ ?
goliath has quit [Quit: SIGSEGV]
<Guest5255>
rber|res Initially the /sys/bus/iio directory was not visible, after installing libiio it was, now it isn't again. But I did check /sys/bus/iio/devices and it didn't show anything
<RP>
qschulz: true, but I'm not sure I like the idea of the docs output for a give tag being variable either
<RP>
qschulz: I guess it depends on the implementation
<qschulz>
RP: In the future, I guess it'll make sense to have the content of scripts/run-docs-build or whatever its name is in a git repo where we can also put all the common files that don't actually belong in the yoooooooocto-docs (e.g. migration guides, javascript switcher script, release notes, etc..)
<qschulz>
RP: if you're using the git tag as version number, it shouldn't variate? (though, yes, I had initially thought of using some magic around git tags for YOCTO_DOC_VERSION_MINUS_ONE in poky.yaml
<qschulz>
in which case it wouldn't be reproducible over time
<qschulz>
except if we go for major releases only in this variable
<RP>
qschulz: when cloning repos, not everyone fetches tags :/
AKN has quit [Read error: Connection reset by peer]
AKN has joined #yocto
<pasherring>
Hey there! I am trying to make use of ads7846 (an SPI touch controller). Whenever I enable the dtoverlay, I get an error on uboot, with "ERROR : xrealloc failed". Any thoughts on this? Not sure if there this could be some fdt memory size related thingy, or defective dtbo
dtometzki has quit [Ping timeout: 256 seconds]
rfuentess has quit [Remote host closed the connection]
rob_w has quit [Remote host closed the connection]
jmiehe has joined #yocto
_wmills has quit [Quit: Leaving]
GNUmoon has quit [Ping timeout: 240 seconds]
<dwagenk>
<manuel1985> "Can anyone tell me why https://..." <- yes it's common. do_install is already populated by the lower layers (don't know of it's bitbake itself that provides the default or if its in oe-core). It may also be appended or overridden by the classes that the recipe inherits.
<dwagenk>
usually you want to keep the default do_install task but also copy some custom files into the image, thus do_install:append() is what you usally see/use.
wmills has joined #yocto
<rburton>
manuel1985: in that case, the autotools class provides a do_install
<manuel1985>
dwagenk, rburton: Thanks both
<lukma>
If I may have a question regarding kernel-meta repository and *.scc files
<lukma>
I would like to reuse bsp/beaglebone.scc
<lukma>
but remove some of "includes" (which I don't need)
<lukma>
Shall I:
<lukma>
1. Prepare local *.cfg with # CONFIG .... not defined
<lukma>
2. Provide local beaglebone.scc and copy it to the kernel-meta/bsp/beaglebone directory at do_kernel_configure:prepend()
<lukma>
3. Use KBUILD_DEFCONFIG with some local set of *.cfg files?
<manuel1985>
Can I have a do_install:append() in both a recipe and an append recipe? Do they get merged or does one overwrite the other?
<manuel1985>
With variables they kinda acumulate but not sure if it's the same for functions.
<qschulz>
manuel1985: :append always get applied
<qschulz>
in the order in which they were parsed
<manuel1985>
qschulz: kthx :)
<qschulz>
manuel1985: also, bitbake -e prints the content of the task after parsing IIRC
pasherring has quit [Read error: Connection reset by peer]
AKN has quit [Read error: Connection reset by peer]
pasherring has joined #yocto
tre has quit [Remote host closed the connection]
ar__ has joined #yocto
sakoman has joined #yocto
codavi has joined #yocto
ar__ has quit [Ping timeout: 245 seconds]
florian_kc has joined #yocto
<RP>
moto-timo: I have patches to help the reproducible issues
pasherring has quit [Remote host closed the connection]
<moto-timo>
RP: thank you
pasherring has joined #yocto
rber|res has quit [Ping timeout: 256 seconds]
rber|res has joined #yocto
paulg has joined #yocto
dvorkindmitry has joined #yocto
<RP>
moto-timo: patch sent out
rber|res has quit [Ping timeout: 268 seconds]
marc1 has quit [Ping timeout: 240 seconds]
rber|res has joined #yocto
marc1 has joined #yocto
osama has quit [Ping timeout: 240 seconds]
dtometzki has joined #yocto
alicef has quit [Quit: install gentoo]
alicef has joined #yocto
alicef_ has joined #yocto
alicef has quit [Ping timeout: 256 seconds]
rfuentess has joined #yocto
alicef has joined #yocto
alicef_ has quit [Read error: Connection reset by peer]
alicef has quit [Client Quit]
alicef has joined #yocto
pasherring has quit [Ping timeout: 250 seconds]
pasherring has joined #yocto
<moto-timo>
Saur[m]: thank you for the review of my patches... it is helpful to get a second set of eyes on things
mrpelotaz0 has quit [Ping timeout: 240 seconds]
AKN_R has joined #yocto
AKN has joined #yocto
florian_kc has quit [Ping timeout: 256 seconds]
osama has joined #yocto
mrpelotazo has joined #yocto
pasherring has quit [Remote host closed the connection]
osama1 has joined #yocto
pasherring has joined #yocto
osama has quit [Ping timeout: 256 seconds]
goliath has quit [Quit: SIGSEGV]
Etheryon has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<landgraf>
RP: it's pidgin/libgnt_2.14.1.b from meta-oe
<RP>
landgraf: scarily, pidgin is still using mercurial so we probably shouldn't drop it
<kergoth>
Huh, I'm surprised any projects are still using it
<RP>
kergoth: so am I, we were just talking about dropping it
florian_kc has joined #yocto
<landgraf>
RP: ok. will fight with it then then :)
rfuentess has quit [Remote host closed the connection]
<RP>
landgraf: I think we might just tell people to use mecurial from the host though
<RP>
landgraf: I'm not sure if is worth too much more time on it :/
mauro_anjo has quit [Read error: Connection reset by peer]
* JaMa
was cleaning some older workdirs and found many svn, cvs, bzr, hg checkouts as well, but only a few of them still had working upstream repo
<JaMa>
and damn, they are difficult to use after so many years using mostly git everywhere
<JaMa>
even with "git svn", git-remote-hg, git-remote-bzr (sadly the last 2 require normal mercurial and bzr to be installed as well and also python2 :)
<RP>
JaMa: we removed the bzr fetcher since it was so unused
AKN has quit [Read error: Connection reset by peer]
<JaMa>
yes, that's fine with me, even gentoo removed ebuild for bzr, because nobody was interested to update it for python3
<JaMa>
ubuntu still has a package as they still use it on launchpad
lucaceresoli has quit [Quit: Leaving]
<RP>
JaMa: I thought it was totally gone!
<landgraf>
RP: well. mailman2 is still in use here and there
<landgraf>
and it's bazaar based iirc
mckoan is now known as mckoan|away
<JaMa>
RP: I thought so as well until my recent digging, but I'm not aware of any OE recipe still activelly using it, so I'm fine with fetcher removal
dev1990 has joined #yocto
<RP>
JaMa: it isn't coming back :)
<JaMa>
zombies are real.. :)
* kanavin
feels like crap today
<kanavin>
my mom is safe
<fray>
I'm glad to hear she's safe..
<RP>
kanavin: it is pretty horrible and sad, glad she is ok :(
<kanavin>
if you didn't know I am 25% ukrainian, 25% belorussian, 50% russian
<khem>
yes agree, glad to hear that
<kanavin>
my mom lives in st.petersburg, far away from any hostilities, but I did transfer her money just now via western union, as economy will plummet there
<fray>
Ya, good chance all of that (money transfer) is going to get shut down really soon
<fray>
going to hurt a lot of innocent people
<kanavin>
fray, not exactly innocent. they did vote for putin. my mom did not, but she's 76, and going to protests at that age - when the police will take you into custody - isn't exactly a good idea.
<moto-timo>
RP: I'm not finding BB_REMOVED_CLASS_MSGS[image-prelink] = "fix USER_CLASSES in your local.conf" anywhere... did you perhaps have a WIP branch for that?
<fray>
I was just given a new RHEL install, and I'm getitng an error from bitbake/lib/bb/pysh/pyshyacc.py "import pyshtables, no module named pyshtables'
<fray>
any idea what I need to get installed?
<kergoth>
fray: pyshtables is generated. is the bitbake dir writable?
<moto-timo>
RP: I'm guessing it would need to have been some tooling in bitbake/lib/bb/parse
<fray>
yes, it's writable
<moto-timo>
kanavin: hoping for the best... certainly distressing times :/
<fray>
pretty clear I'm missing something from my install.. it won't generate it, and thus can't find it
huseyinkozan has joined #yocto
<fray>
found it.. problem is 'from hashlib import md5' there is no md5 since this is a FIPS install
<fray>
that is in bitbake/lib/ply/yacc.py
jmiehe has quit [Ping timeout: 240 seconds]
goliath has joined #yocto
<fray>
ply doesn't support the fips stuff the way it's implemented.. there is a simple fix, but I've no way to verify it outside this environment.. Once I get a little further I'll send something to the list
<kergoth>
I wonder if we should look at alternatives eventually. I think there are some new parsing options in python nowadays. i.e. guido's series about peg parsers, lark, etc
<fray>
more or less the fix for this (and another issue) is replacing 'hashlib.md5()' with 'hashlib.new('MD5', usedforsecurity=False)'
<fray>
it's possible that usedforsecurity can be an argument to the md5, but I'm not sure.. the references I found all said you need to use 'new'
<kergoth>
simple enough
<fray>
what I find funny.. the reports of failures I had for using on a FIPS install, the failures were during RPM packaging.. NOT in bitbake or the classes.. so far I've gotten a raft of failures in bitbake/classes.. (easy ones, but failures)
<fray>
this should give us a list of places we may need to consider moving to something stronger then md5.. (or if it's not for 'security' we can use the 'new' approach)
<fray>
things like the LIC_FILE_CHKSUM are likely suspect and probably shouldn't use md5 anymore. Things like hashes for variables in teh cache, it'll be fine to use md5s..
prabhakarlad has quit [Quit: Client closed]
pasherring has quit [Read error: Connection reset by peer]
pasherring has joined #yocto
pasherring has quit [Read error: Connection reset by peer]
pasherring has joined #yocto
<rburton>
RP: when are you merging the python wheel stuff into oe-core?
<rburton>
(if its this evening, i'll leave it, otherwise I'll start pinning out CI to specific SHAs)
Maxxed has joined #yocto
<moto-timo>
rburton: we need to see if the reproducible jobs are healthier on AB
<rburton>
i'll send a patch to drop meta-oe's layer priority tomorrow as this is madness
<rburton>
i forgot it was not equal to core
florian_kc has quit [Ping timeout: 245 seconds]
<khem>
hmmm vnc to runqemu is not working any ideas ?
<khem>
it used to work fine in same setup. This is master
<kergoth>
rburton: good idea. I have an override in meta-mentor's setup scripts that forces the meta-oe priority to 1 due to previous issues.
<kergoth>
horrible hack, but..
<kergoth>
I still wonder if bits to sort bblayers by layer priority would be worthwhile upstream. I have a little bitbake-layers command to do it on demand at least, maybe i should submit that
amitk has quit [Ping timeout: 245 seconds]
Minvera has joined #yocto
GNUmoon has joined #yocto
<moto-timo>
Seems like oeqa/files/buildhistory_filelist1.txt and oeqa/files/buildhistory_filelist2.txt pointing at python3.8 should be moved to python3.10?
<JaMa>
kergoth: if it supports selecting own priorities as well, then it would be interesting - now we're generating bblayers.conf with own mcf tool which decideds the order as well as sets priorities of all included layers
<kergoth>
Huh, it doesn't currently but certainly could. That's interesting
<kergoth>
I've been considering submitting a variety of new bitbake-layers commands to try to upstream at least some of what we're doing with our scripts, but haven't gotten it to completion
<RP>
rburton: given there are still a couple of issues, won't be today
<smurray>
khem: I successfully used publicvnc with runqemu earlier today with a qemux86-64 build using master from sometime yesterday, that was with remmina as client
<rburton>
moto-timo: print a message if it needs to use that class and the meta-python layer isn't enabled?
huseyinkozan has quit [Quit: Konversation terminated!]
<moto-timo>
rburton: that's what I was thinking... just need to figure out proper syntax
<RP>
moto-timo: python3-flit-core uses pip install with no --root
<RP>
moto-timo: can you remember why --target instead of what the class does?
* RP
hasn't looked anything up yet
<moto-timo>
ah the do_install is different
<RP>
moto-timo: Looking at the docs, can we replace that with --prefix and --root ?
<moto-timo>
RP: --target was just what worked at first... and then I was resisting using pip_install_wheel class because it wasn't straight forward.
<RP>
moto-timo: ok, I think if we swap options around it might work
<moto-timo>
yes, exactly... do the same thing as the pip_install_wheel class. that was just an oversight when we fixed that
<armpit>
RP, does 14745 help ?
<RP>
moto-timo: local test says that fixes flit
<RP>
armpit: yes, thanks. But I still plan to talk about on Monday
<sgw>
RP: Is INCOMPATIBLE_LICENSE = "<some lic>" valid? All the selftests seem to focus on the pn-<image> override style
<RP>
sgw: I'd think it should be
<moto-timo>
RP: I think I will inherit pip_install_wheel and use the knobs to set the wheel path properly... so we have the flags in one place
<RP>
moto-timo: ok, I guess I should stare at numpy then
<khem>
smurray: vnc worked ok for me it was pilot error, due to wrong port. but I still have problem where it does not launch with gtk or sdl, I dont really use it in that capacity so its ok. vnc is what I needed
<smurray>
khem: okay. I've not tried anything other than vnc for a long time, as it's way more convenient for me with my setup
<RP>
moto-timo: and bottom line is the hash is different
<moto-timo>
RP: salt problem?
<moto-timo>
RP: at least for python3-docutils/0.18.1-r0/image/usr/lib/python3.10/site-packages/docutils/nodes.py it has the path to the file in "filename" so that would be different paths for build A and build B correct?
<moto-timo>
or am I reading the output wrong from the tool :/
<RP>
moto-timo: is there a problem with nodes.py? It looks like other pyc files in docutils in the bin dir?
<moto-timo>
RP: right... I'm being dense
codavi has quit [Ping timeout: 256 seconds]
<moto-timo>
RP: we could just not copy the /usr/bin/__pycache__
<RP>
moto-timo: it is odd it is just that single bindir
<RP>
moto-timo: the numpy issue looks quite different
<RP>
although pyelftools is another issue :/
<RP>
moto-timo: I think I know. I'd bet it hashes the file before we update the shebang
<RP>
moto-timo: so we need to recompile after changing that
<moto-timo>
RP: aha!
<RP>
moto-timo: I'm going to guess you can figure out the thing to do after the sed to do that (I think there is a line in the main python recipe we can crib from)
* RP
can take a quick look at numpy as I have a feeling I recognise it
<RP>
moto-timo: python3_3.10.2.bb: # Unfortunately the following pyc files are non-deterministc due to 'frozenset'
<RP>
moto-timo: some of the code in numpy is using frozenset so that is the issue there
<RP>
moto-timo: fix is to delete the pyc for that bit
kriive has quit [Read error: Connection reset by peer]
kriive has joined #yocto
<moto-timo>
RP: docutils also has '#!/usr/bin/env python' in those scripts
<RP>
moto-timo: we're probably changing it to python3?
<moto-timo>
RP: that's my thought
<RP>
moto-timo: I have a numpy fix ready FWIW
goliath has quit [Quit: SIGSEGV]
<RP>
moto-timo: want me to write the recompile patch?
<RP>
I'm keen to try and get something onto the AB before I collapse/sleep
<moto-timo>
RP: I think you are closer to getting it done than I am
<moto-timo>
RP: I was going with 'devtool modify' on python3-docutils approach
* RP
would definitely not have started there
<moto-timo>
:/
<RP>
moto-timo: although my idea of fixing this results in a pseudo abort()
<moto-timo>
RP: there's always something else to slow progress