<RP>
kanavin: I've merged the other fixes but not the enable patch until we get that one fixed. All looking good though thanks!
dtometzki has quit [Ping timeout: 268 seconds]
<JosefHolzmayr[m]>
yo dudX
mckoan|away is now known as mckoan
<mckoan>
good morning
rfuentess has joined #yocto
gsalazar has joined #yocto
leon-anavi has joined #yocto
<ThomasD13>
Hi everyone. Where is the appropriate place to specify the used linux kernel (+config) and the devicetree for a linux build? Image configuration? Machine configuration?
<ThomasD13>
thanks JosefHolzmayr[m]. Am I allowed to ask a git related question here too? It might be obvious for some git cracks :)
<fbre>
yes, allowed
<ThomasD13>
When I rebase like "git rebase --onto targetBranch feature~3 feature" it will "move" the last 3 commits of feature ontop on targetBranch
<ThomasD13>
When I checkout feature and do "git rebase --onto targetBranch HEAD~3 HEAD" it will COPY the last 3 commits of feature ontop of targetBranch
<ThomasD13>
Why the difference? I thought HEAD is a reference as well as <branchname> is?
<ThomasD13>
The first rebase does a MOVE, the second a COPY of these commits
zyga has quit [Quit: Leaving]
zyga has joined #yocto
Belsirk has joined #yocto
rfuentess has quit [Read error: Connection reset by peer]
Belsirk is now known as rfuentess
chrfle has quit [Ping timeout: 252 seconds]
chrfle has joined #yocto
florian has joined #yocto
mtudan has joined #yocto
<ad__>
hi, on gatesgarth, i am getting the follwoing imx-sc-firmware was skipped: incompatible with machine "xxx-imx8mp" (not in COMPATIBLE_MACHINE), what's should the proper way to solve this ?
<ad__>
well i added proper machine in OVERRIDES, hope is ok
<qschulz>
ad__: usually it's in MACHINEOVERRIDES you should put it
vd has quit [Ping timeout: 256 seconds]
<mtudan>
@khem
<mtudan>
is it possible to compile the kernel recipe only with clang? I've set the following variables in the kernel recipe but cat /proc/version still reports gcc was used to compile kernel:
<mtudan>
TOOLCHAIN = "clang"
<mtudan>
RUNTIME = "llvm"
Guest33 has joined #yocto
<JosefHolzmayr[m]>
mtudan: i don't think the kernel can be built by clang without being patched, and so it probably is in the UNCLANGABLE variable of meta-clang, therefore skipped.
<mtudan>
I've searched if it is in the config file, but couldn't find possible entry regarding the kernel recipe:
<JosefHolzmayr[m]>
bitbake -e always gives the authorative answer.
<ad__>
qschulz, yes, done, thanks
Guest65 has joined #yocto
Guest65 has quit [Client Quit]
fray has quit [Ping timeout: 268 seconds]
mtudan has quit [Quit: Client closed]
mtudan has joined #yocto
fray has joined #yocto
gsalazar_ has joined #yocto
gsalazar has quit [Ping timeout: 240 seconds]
mckoan is now known as mckoan|away
Tyaku has joined #yocto
<Tyaku>
Hi, If we have spaces in folders names and we want to "install" all headers files from these folders, how to do it ? ex: install ${S}/src/darwin/CHIPTool/CHIPTool/View Controllers/*.h ${D}${includedir}/matter/src/darwin/CHIPTool/CHIPTool/View Controllers
<qschulz>
Tyaku: it's run in a shell, so quote the whole string
<qschulz>
and or escape the space
xmn has joined #yocto
<fbre>
Hi! Where is the .bb file which defines the partition sizes of the .wic file?
<fbre>
(I mean, which meta layer and which recipe-... dir)
<qschulz>
fbre: find where the WKSFILE variable points to
<qschulz>
bitbake -e should help
<qschulz>
or look directly at WKS_FULL_PATH
<qschulz>
with bitbake -e
<qschulz>
on the image recipe
<qschulz>
the logic of wic is handled in meta/classes/image_types_wic.bbclass
ThomasD13 has quit [Ping timeout: 265 seconds]
<fbre>
(y)
<fbre>
the .wic file contains boot and root partition, and my problem is the root partition is of different size from bitbake build to bitbake build. I actually wanted a fixed size for the root partition.
<qschulz>
if the root partition is the last partition on the flash, it makes more sense to resize at first boot I'd say
<fbre>
I have more partitions behind the rootfs partition (my partitions 3 and 4).
<qschulz>
otherwise, yeah, you need to reserve the space
<fbre>
I currently exchange boot and rootfs partition but if the new .wic file is bigger, it exceeds the size of device :)
<fbre>
I currently keep partitions 3 and 4 because they contain my user data
Tyaku has quit [Ping timeout: 265 seconds]
<user_123>
Hi , Please Support to setup ECLIPSE IDE for Yocto Tool chains( arm-poky-linux-gnueabi)
<JosefHolzmayr[m]>
user_123: in case you are looking for the yocto eclipse plugin or support on it: it is explicitly deprecated and not supported for years.
<user_123>
i am trying cross compilation with my yocto tool chain
<user_123>
i am facing issues with stdint.h not found
<JosefHolzmayr[m]>
user_123: then i'd suggest to either stick to non-integration (e.g., not using the eclipse magic) or trying vscode as shown here: https://youtu.be/PY4godidHx4
<JosefHolzmayr[m]>
user_123: sorry, but eclipse is and oten has been a constant source of headaches.
<user_123>
using VScode can i able to communicate ioctl and driver layer communication using SDK , toolchain ??
<user_123>
I am developing Linux System Programming in C , is it support? if support Please provide doc...Thank you
<JosefHolzmayr[m]>
user_123: using the setup that the video shows, you can write c code and compile it using a yocto based toolchain.
<JosefHolzmayr[m]>
and using a standard toolchain on linux you can usually acess files, and therefore drivers.
camus has quit [Ping timeout: 252 seconds]
<JosefHolzmayr[m]>
but if you are just getting started, its probably way easier to learn the basics without yocto and all, just try on your desktop.
xmn has quit [Ping timeout: 240 seconds]
goliath has joined #yocto
<user_123>
OK Thank you
gsalazar_ is now known as gsalazar
wwilly has joined #yocto
tgamblin has quit [Remote host closed the connection]
tgamblin has joined #yocto
camus has joined #yocto
camus has quit [Ping timeout: 260 seconds]
camus has joined #yocto
zyga has quit [Ping timeout: 260 seconds]
Guest33 has quit [Quit: Client closed]
jwillikers has joined #yocto
fbre has quit [Quit: fbre]
jwillikers has quit [Remote host closed the connection]
jwillikers has joined #yocto
<jonmason>
looking at stale email addresses, there appear to be many in the patches
sakoman has joined #yocto
<jonmason>
A lazy grep also shows "scripts/combo-layer:# Richard Purdie <richard.purdie@intel.com>"
<jonmason>
I'm sure that bounces
stwcx has joined #yocto
<stwcx>
Hello. I managed to backport the override syntax support in bitbake to Zeus and Rocko (LOL). I know those are EOL but if I were to send patches what is the probability they'd be applied to those branches?
chrfle has quit [Ping timeout: 265 seconds]
wwilly has quit [Quit: Leaving]
stwcx has quit [Quit: Ping timeout (120 seconds)]
m4ho has quit [Read error: Connection reset by peer]
stwcx has joined #yocto
m4ho has joined #yocto
arlen has joined #yocto
Tokamak has quit [Read error: Connection reset by peer]
Tokamak has joined #yocto
<berton[m]>
Hi! I was looking the PACKAGE_BEFORE_PN variable and I don't understand what the difference is between:
<berton[m]>
PACKAGE_BEFORE_PN = "${PN}-foo"
<berton[m]>
PACKAGES =+ "${PN}-foo"
<kergoth>
berton[m]: PACKAGE_BEFORE_PN is exactly that, right before the PN package, but still after src, dbg, dev, etc. If you don't want to risk overriding something belonging to those, PACKAGE_BEFORE_PN is best.
<kergoth>
for example, lets say you wanted to package ${bindir}, if you do that with PACKAGES =+ you could accidentally grab .debug files, depending on your configuration for dbg file paths
<berton[m]>
kergoth: OK. This example with debug files clarifies usage for me. Thanks!
vd has joined #yocto
<kergoth>
np
<kergoth>
i expect that unless you explicitly want to override those special packages, it'd be best to use PACKAGE_BEFORE_PN to clarify your intention with the line
dev1990 has joined #yocto
* kergoth
shrugs
paulg has joined #yocto
Tokamak has quit [Ping timeout: 240 seconds]
Tokamak has joined #yocto
<RP>
jonmason: the challenge is some are signed-off-bys which you can't really update
Tokamak has quit [Read error: Connection reset by peer]
<paulg>
more e-mail update sweeps?
florian_kc has joined #yocto
<jonmason>
RP: yes, it can be a little bit of a nightmare. Just pointing out that there are some out there that might be able to be changed. And probably, if they are changed to be applied to a newer version of the package, they need to be modified anyway.
gsalazar_ has joined #yocto
Tokamak has joined #yocto
gsalazar has quit [Ping timeout: 252 seconds]
gsalazar_ is now known as gsalazar
<RP>
jonmason: right, we could do with cleaning some of this up
florian has quit [Quit: Ex-Chat]
rfuentess has quit [Quit: chelas en viernes!]
florian_kc has quit [Ping timeout: 252 seconds]
leon-anavi has quit [Quit: Leaving]
ecdhe has joined #yocto
ecdhe has quit [Changing host]
<tgamblin>
does meta-browser have its own chat?
<tgamblin>
Trying to uprev chromium to fix a pile of CVEs, unsurprisingly running into do_compile issues
<jonmason>
Why would ssh fail when just the GPU changed?
<jonmason>
Anyway, I'll run locally and see if I can replicate and fix
<RP>
jonmason: I'm not saying it was due to that. Just hard to untangle
<jonmason>
It's fair, and I was overeager to get it out last night before doing a full regression myself
<RP>
jonmason: the later build worked so it probably isn't virtio but I don't know what to make of it
<RP>
I guess I'm asking if you have any thoughts on it
* RP
's thoughts are to go and find food :)
<jonmason>
Fresh from McD here, with lots of sugary soda in my bloodstream
<RP>
jonmason: I could use some sugar after being out on the mountain bike
mattsm has quit [Read error: Connection reset by peer]
mattsm has joined #yocto
jwillikers has quit [Ping timeout: 265 seconds]
jwillikers_ has joined #yocto
jwillikers_ is now known as jwillikers
<rburton>
RP: i recommend chimichurri steak on focaccia. in related news, the cafe at phoenix bike park really does exceed expectations
vmeson has quit [Ping timeout: 265 seconds]
vmeson has joined #yocto
florian_kc has joined #yocto
mtudan has quit [Quit: Client closed]
ant__ has quit [Ping timeout: 268 seconds]
gsalazar has quit [Ping timeout: 265 seconds]
arlen_ has joined #yocto
arlen has quit [Ping timeout: 252 seconds]
zyga-mbp has joined #yocto
sakoman has quit [Quit: Leaving.]
amitk has quit [Ping timeout: 268 seconds]
<RP>
rburton: I loved the chimichurri steak I had in portland as Brian would attest
<kanavin>
RP: I am back :)
<kanavin>
RP: was another visit to linutronix office
<RP>
kanavin: thanks for the patches, I saw the cpan one :)
<kanavin>
RP: ah right then it's sorted
<RP>
kanavin: I'll run a few more tests over the weekend but I hope so
<kanavin>
RP: there's a bunch more warnings, but those are not directly visible in AB logs, so will be fixed gradually later
<RP>
kanavin: sounds good
<kanavin>
RP: or some only start showing up when bitbake itself runs on py 3.10
<kanavin>
(I fixed the most serious one though)
<RP>
kanavin: we no doubt have that to look forward to but this at least puts us in a better position
<RP>
it is nice to try and get ahead of things for a change!
<kanavin>
RP: one serious item is that distutils is deprecated and will be removed in 3.12. There's a fair number of spots where it is used in oe-core/poky (and not inside components but in yocto itself).
<kanavin>
we have two years I guess :)
<RP>
kanavin: we should probably put some plans in place to resolve that
<RP>
at least someone has realised which is a good first step
* RP
is still trying to decide which invasive issues are worth tackling in 3.5
<kanavin>
RP: on my favorite subject :) I tend to think that sato->weston transition should be deferred to after the next LTS
<kanavin>
so that LTS is looking just like people are used to, and then we can do it :)
<kanavin>
maintaining the patches is not hard
<RP>
kanavin: yes, I think that makes sense
<kanavin>
by then it should also be more obvious to the community that standalone X server is effectively EOL
<kanavin>
there's some hobbyist volunteer from poland who has proposed to do a new x server release, but that by itself doesn't make it alive and kicking
<RP>
kanavin: right, I think it will go one a little yet but it is ramping down
<kanavin>
RP: I do wish we had wayland eye candy, but it just isn't showing up
<JPEW>
kanavin: Now that rust is in place, it might be more reasonable to get phosh running
<JPEW>
kanavin: I *think* many of the issues I had with it were due to old librsvg.... need to try again
<kanavin>
JPEW, we need to finally resolve the old librsvg too
<moto-timo>
JPEW: it would be nice to get phosh or similar working
* kanavin
hasn't heard of phosh until now
<moto-timo>
I was at least able to build librsvg with vmeson contrib. But I don’t think I runtime tested it.
<kanavin>
a smartphone shell as a reference UI is not a bad idea!
<kanavin>
(for yocto)
<kanavin>
because we can never otherwise agree what a reference ui is supposed to be
Vonter_ has quit [Ping timeout: 268 seconds]
<RP>
kanavin: feel free to give librsvg a go! :)
<moto-timo>
kanavin: you should come to a meeting once in a while we’re we tend to discuss things like this ;)
<kanavin>
RP: I've heard others are working on it? and working and working and working, and no patches are showing up :-/
<kanavin>
that discourages me from trying ;)
<moto-timo>
s/we’re/where/
<RP>
kanavin: vmeson is trying but doesn't have a lot of time
<RP>
he is also on vacation atm so we can talk about him :)
<moto-timo>
So far my issues with rust are mostly setuptools_rust not being cross friendly
* RP
fixed a few build issues but knows little about it
<kanavin>
moto-timo, I'm generally content with reading the meeting notes :)
<moto-timo>
kanavin: you would have heard about phosh a dozen times already. But point taken ;)
<kanavin>
RP: I finally got a green build for py 3.10 and openssl 3 as well yesterday. So this whole patchset is ready, except I need to replace wip commit messages with real ones http://git.yoctoproject.org/cgit.cgi/poky-contrib/log/?h=akanavin/package-version-updates
<RP>
kanavin: that is rather nice, good work! :)
zyga-mbp has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<RP>
kanavin: Should let us make a good start in 3.5
ant__ has joined #yocto
<kanavin>
RP: we might beat all other distros to having py 3.10 *and* openssl 3 ;)
<moto-timo>
kanavin: I’ll be sending you upgrades fir my core recipes soon for staging. python3-hypothesis and -importlib-metadata update frequently, so not sure if we want sequential patches or rebase
<kanavin>
when RP says 'slightly', one should get *very* concerned ;)
<moto-timo>
“soon” meaning a few days
<RP>
kanavin: would be nice to "win" that race
<RP>
kanavin: small but quite important small detail wrong
<kanavin>
moto-timo, I generally don't stash other people's updates, you can keep them in your branch unless there's a reason I should have them?
<kanavin>
RP: ouch!
<moto-timo>
kanavin: fair enough
<kanavin>
RP: one is often blind to one's own changes that way
<kanavin>
RP: should be squashed withe the commit introducing ZSTD_THREADS anyway
jaskij[m] has joined #yocto
<jaskij[m]>
are multiconfigs with only a string of numbers in the name supposed to work?
<RP>
kanavin: it is easily done, I just mention it since I saw it
<RP>
moto-timo: we are kind of pulling together a kirkstone-next (between abelloni and I)
<RP>
jaskij[m]: from the question I guess they don't? There are probably reasons it might break but its not something I've ever tried
<jaskij[m]>
jaskij[m]: or rather, did they break in hardknott? I had a multiconfig named 504, worked fine in dunfell, stopped working after switching to hardknott
<RP>
jaskij[m]: I don't think we intentionally broke that
<jaskij[m]>
wanted to confirm on chat before I fire anything on the mailing list
<jaskij[m]>
ERROR: Unable to register event handler '504sstate_eventhandler_stalesstate':
<jaskij[m]>
File "/home/jaskij/yocto/layers/poky/meta/classes/sstate.bbclass", line 1
<moto-timo>
RP: ok, at least the 3 recipes that don’t change every three days might be worth staging in kirkstone-next. I’ll let you know
<RP>
jaskij[m]: hmm, right, you can't start a python function with a number
<RP>
jaskij[m]: I am tempted to say "don't do that" and we should just warn against it
<jaskij[m]>
curiously enough, this did work in dunfell
<kanavin>
moto-timo, right, the problem with all those python3-thingy recipe updates is that they have a very short shelf life
<RP>
jaskij[m]: that eventhandler is new
<kanavin>
not all, but quite many
<RP>
jaskij[m]: we also fixed multiconfig event handlers iirc so it could be that
<kanavin>
so one is better off updating them on an independent cadence
<jaskij[m]>
RP: fair enough, just that it threw a *massive* amount of errors at me while switching, manually wiping the cache cleared the output enough for me to notice what was going on
<jaskij[m]>
renaming the mc to something starting with a letter fixed the issue
<RP>
jaskij[m]: a bug (or patch) to give a nice error instead would be good
<moto-timo>
kanavin: absolutely. AUH run on meta-python usually results in about 50 upgrades (with a couple handfuls addressed by patches to the ML)
<moto-timo>
kanavin: I just need to asses the ones in core that might have more impact to other core usage
<kanavin>
moto-timo, sometimes I wonder if all those python module maintainers really need to tag versions several times a week, and what prompts them doing that. Is it a tag on every commit principle?
<jaskij[m]>
RP: fair enough, I'll try to mail something in
<jaskij[m]>
tbh, mailing lists are super intimidating to me
florian_kc is now known as florian
<RP>
jaskij[m]: please don't be intimidated, we're all a friendly lot
<RP>
jaskij[m]: feedback should be constructive
<jaskij[m]>
it's just something about mailing lists, not the people in the project
<jaskij[m]>
like, if it was a forum or a GH repo or something, I'd have no issue with it
<RP>
jaskij[m]: bugzilla is another option FWIW
<RP>
jaskij[m]: there isn't a huge difference between a forum and a mailing list really. I have been around them for too many years though so I probably wouldn't see it :/
<jaskij[m]>
it's in my mind, and I gotta deal with it anyway sooner or later
<RP>
jaskij[m]: if we can help somehow let us know
r0bin has joined #yocto
<jaskij[m]>
sent
warthog9 has quit [Ping timeout: 268 seconds]
warthog9 has joined #yocto
<JPEW>
kanavin: I like phosh as an option because it's just gnome under the hood, so you get a huge library immediately
<JPEW>
And phone ui makes a lot of sense from a input perspective (supports touchscreen and keyboard)
florian has quit [Ping timeout: 260 seconds]
xmn has joined #yocto
prabhakarlad has quit [Quit: Client closed]
prabhakarlad has joined #yocto
<jaskij[m]>
have there been any serious attempts to mock `pg_config`?
<jaskij[m]>
Probably wrong chat
<jaskij[m]>
My bad, tired mind didn't notice chat isn't as split as mailing lists
florian has joined #yocto
sgw has quit [Quit: Leaving.]
sgw has joined #yocto
r0bin has quit [Remote host closed the connection]
r0bin has joined #yocto
JaMa has quit [Quit: leaving]
jwillikers has quit [Remote host closed the connection]
otavio has quit [Remote host closed the connection]
<moto-timo>
kanavin: python3-hypothesis definitely tags and releases on every commit :/