<RP>
khem: I was trying to help Saur debug something so yes, don't worry about it
<RP>
landgraf: I'd have hoped it would pull into the repo if present on a mirror. If it is exists it should delete it as we lose data that way potentially
wojci has joined #yocto
ray-san has quit [Ping timeout: 276 seconds]
ray-san has joined #yocto
beneth has quit [Ping timeout: 244 seconds]
beneth has joined #yocto
<landgraf>
RP: if clonedir (git2/github.com.boto.botocore.git) exists then try_premirror false and bitbake doesn't try to fetch premirror at all and fails because the revision is not in the repo
<RP>
landgraf: we might want to rethink that I guess :/
<landgraf>
once the clonedir deleted it works and pulls premirror ( github.com.landgraf.botocore.git) with the revision which is needed for the build
<RP>
landgraf: did we fix the mirroring so they don't destroy the existing clone?
<landgraf>
RP: I think so, we don't delete clone anymore but fetch into it from the temporary location
<RP>
landgraf: the original problem was that trying the mirror destroyed the existing clone. If we've fixed that, the mirror logic could probably be tweaked
<RP>
landgraf: right, I thought I remember we'd changed that (for the better)
<landgraf>
RP: right, there's comment explaining this in the try_premirror function
<landgraf>
RP: I will try to create test cases for the problem (once I have time :-( )
leon-anavi has joined #yocto
rfuentess has joined #yocto
goliath has joined #yocto
gsalazar has joined #yocto
rob_w has joined #yocto
Kubu_work has joined #yocto
mckoan|away is now known as mckoan
florian has joined #yocto
Jones42_ has joined #yocto
Jones42 has quit [Ping timeout: 246 seconds]
mrtatillon has joined #yocto
Jones42__ has joined #yocto
Jones42_ has quit [Read error: Connection reset by peer]
florian has quit [Ping timeout: 252 seconds]
mrtatillon has quit [Quit: Client closed]
mrtatillon has joined #yocto
florian has joined #yocto
mrtatillon has quit [Ping timeout: 256 seconds]
xmn has quit [Ping timeout: 248 seconds]
mvlad has joined #yocto
wojci has quit [Ping timeout: 252 seconds]
wojci has joined #yocto
<JaMa>
would it make sense to extend buildpaths QA check to also show the line where tmpdir was found (after checking that the file is text file and not binary and that the line isn't way too long or way too many of such lines to show in the message?)
olani has quit [Remote host closed the connection]
<JaMa>
mc on some build hosts fail with "ERROR: mc-4.8.31-r0 do_package_qa: QA Issue: File /usr/libexec/mc/extfs.d/uzip in package mc-helpers-perl contains reference to TMPDIR [buildpaths]" and if I don't have shell access to such host it's more complicated to guess what went wrong there (I assume some full path to zip in WORKDIR instead of /usr/bin/zip or something like that)
<mcfrisk_>
JaMa: would be nice. the specific build config would be nice to know too. AB build configs are a bit hard to figure out, at least for me.
vthor has quit [Excess Flood]
olani has joined #yocto
vthor has joined #yocto
vthor has quit [Changing host]
vthor has joined #yocto
florian_kc has joined #yocto
Piraty has quit [Quit: -]
Piraty has joined #yocto
prabhakalad has quit [Quit: Konversation terminated!]
prabhakalad has joined #yocto
mbulut has joined #yocto
sakoman has quit [Ping timeout: 255 seconds]
sakoman has joined #yocto
rob_w has quit [Remote host closed the connection]
vthor has quit [Quit: kill -9 $pid]
vthor has joined #yocto
vthor has quit [Changing host]
vthor has joined #yocto
Saur_Home has quit [Quit: Client closed]
Saur_Home has joined #yocto
Jones42__ is now known as Jones42
ablu has quit [Read error: Connection reset by peer]
<mcfrisk_>
"rm -rf tmp && bitbake core-image-minimal" fails with: FileNotFoundError: [Errno 2] No such file or directory: '/home/builder/src/build/tmp/log/eventlog/20240830112650.json
<mcfrisk_>
how should the build directory by cleaned now?
ablu has joined #yocto
<mcfrisk_>
reverting 4eedb58a285bdb6d3620f6ca15396dee14fad354 seems to help, or cures the situation due to changed hashes
<JaMa>
don't you use memres bitbake and it was still running while you wiped tmp?
BhsTalel has joined #yocto
<mcfrisk_>
I don't, I just call bitbake from command line. Maybe there are leaked processes but at least I did not start them on purpose. will check..
BhsTalel has quit [Ping timeout: 256 seconds]
alessioigor has quit [Remote host closed the connection]
alessioigor has joined #yocto
xmn has joined #yocto
wojci has quit [Ping timeout: 245 seconds]
mrtatillon has joined #yocto
<mrtatillon>
(sorry if I spam, I'm disconnected all the time, and I don't know if my messages are sent)
<mrtatillon>
Oh sorry, I wrote too fast, I meant that I added `IMAGE_INSTALL += "kernel-module-bcm2835-gpiomem"` of course
<mrtatillon>
Thanks for your help
<mrtatillon>
For more context, I'm trying to move my Yocto project, built from the Krogoth version, to a Dunfell version.
<mrtatillon>
Actually in my base project I have both packages `linux-firmware-brcm43430` and `kernel-module-bcm2835-gpiomem`, and with both I first got an error with `linux-firmware-brcm43430` (`ERROR: Nothing RPROVIDES 'linux-firmware-brcm43430'`) after many unsuccessful searches I decided to remove it and see later. After removing it and some errors later I
<mrtatillon>
got the error with `kernel-module-bcm2835-gpiomem`. So I removed this module as well.
<mrtatillon>
Now the build completed without errors. Of course my image is not bootable after removing so many packages, so now I restore all the modules I removed, including the latest `kernel-module-bcm2835-gpiomem`
<JaMa>
not sure if you're using 6.6 kernel, it's not in dunfell by default, but at least it will point you to config options you should check in the defconfig
<JaMa>
it's renamed in 6.1 already (at least I have deploy/ipk/raspberrypi4_64/kernel-module-raspberrypi-gpiomem-6.1.77-v8_6.1.77+git0+43d1723dbe_77fc1fbcb5-r0.0webosrpi38_raspberrypi4_64.ipk already)
<mrtatillon>
Yes, I saw that yesterday that Dunfell was EOL, but I'm pretty sure that wasn't the case when I started my migration :/
sanbeam has joined #yocto
leon-anavi has quit [Read error: Connection reset by peer]
leonanavi has joined #yocto
amitk has joined #yocto
florian has joined #yocto
sanbeam has quit [Remote host closed the connection]
sanbeam has joined #yocto
sanbeam has quit [Ping timeout: 248 seconds]
<gmorell>
wired: shitty ai slop couches for your storefront
<gmorell>
my b, wrong channel
alessioigor has quit [Remote host closed the connection]
alessioigor has joined #yocto
sanbeam has joined #yocto
sanbeam has quit [Ping timeout: 248 seconds]
alessioigor has quit [Ping timeout: 256 seconds]
mrtatillon has quit [Quit: Client closed]
<khem>
JaMa: your comments on clang19 pr are valid
<khem>
JaMa: for now I had to demote buildpaths to warning because not all layers in yoe are buildpaths clean
<khem>
secondly. about bcc and bpftrace I am fine to move it under one dynamic-layer
<JaMa>
khem: ok, I just wanted to check if you're seeing the same (or if it's specific to my setup)
<JaMa>
I can send the PR to move bpftrace, but it can wait after clang19 is merged to avoid conflicts
rfuentess has quit [Remote host closed the connection]
<JaMa>
I've sent the fixes for what buildpaths I was still seeing in meta-oe layers from my world builds
<JaMa>
qt5 and qt6 is the biggest source of buildpaths issues for my builds, so I'm selectively demoting it to WARN_QA in recipes which need it until https://bugreports.qt.io/browse/QTBUG-105913 is fixed
reatmon_ has quit [Remote host closed the connection]
reatmon_ has joined #yocto
<khem>
thanks for fixes, I have fixed AB issues thus far but AB does not test all the layers or even all the combos
<khem>
I can not apply gcab patch for whatever reason
<JaMa>
thanks for doing great job of maintaining meta-oe (so that I don't have to :))
<JaMa>
and it depends on the previous one (which demotes ERROR_QA to WARN_QA)
<JaMa>
because I wasn't sure if both will be applied (was waiting for Mikko to explain why it was on purpose - which seems to be fine after his reply on ML)
<khem>
I see that makes sense now
<khem>
so perhaps it is better to squash both of them into one
<JaMa>
do you think armin will take both squashed to scarthgap or just revert the one from Mikko which was backported there?
<khem>
he wont
<khem>
lets keep it as such
<JaMa>
I would prefer to keep them separate as to teach other people not to use INSANE_SKIP in such cases
<khem>
re. meta-oe fixing, I think I am occupied too much in chasing breakages from everyone fmt/libxml2/ffmpeg/automake have caused a ton of stuff to break
mckoan is now known as mckoan|away
<khem>
its untenable at one point I might just disappear :)
<khem>
or lets say choose to disappear
<JaMa>
I'm aware how painful it is and I am grateful that you lasted (probably almost as long as I did) and I won't be surprised if you choose to disappear
leonanavi has quit [Quit: Leaving]
<JaMa>
I'm trying to help as time allows, but leaving the issues in "world builds" only after fixing the build failures in our images (which sometimes take longer then you fixing meta-oe recipes)
<khem>
yeah, cycles are too long, before you can run one something in oe-core has come that will shoot next
<JaMa>
e.g. every ffmpeg is breaking quite a lot (many multimedia related recipes as you might expect in TVs) and component owners are useless (I'm still waiting for them to review my changes for compatibility with ffmpeg-5)
<khem>
freakin hamster wheel
<JaMa>
indeed
<khem>
yeah then I question why do we keep pushing for new versions
<JaMa>
and it's not just about keeping up with oe-core changes, our internal developers are breaking kirkstone based builds often enough that it takes me half of my time just to unbreak stuff on daily basis
<khem>
self-inflicted harm
<khem>
I thought of disabling at a whim but that does not work effectively, because dependencies run wild
<khem>
and suddenly one cant build a reasonable system by disabling
<JaMa>
yeah, I take latest versions as nice-to-have but if there aren't enough people to keep meta-oe and other layers buildable, then always having latest-greates might cause more harm than good
<khem>
so thats not an answer either
<JaMa>
and if you start using SKIP_RECIPE more aggressively then people still won't help, just complain later when they upgrade to the next LTS and discover that the recipe they were building in dunfell is now blacklisted for 3 years
<JaMa>
I'm still getting occasional e-mail why I've deleted something when the PNBLACKLIST message was quite clear what was wrong with the recipe 8 years ago
<khem>
yeah I think I will stop triaging world builds and do yoe images and fix whats needed that might restore sanity
<JaMa>
you need to get some OE member drunk on some conference and ask him/her to maintain meta-oe, I've asked Armin long time ago in Prague and was really surprised he said yes :)
<JaMa>
and he wasn't even drunk I think :)
<khem>
anyone is welcome
<JaMa>
and I feel better since then, but my sanity still suffers from insane internal developers
<JaMa>
and for the record, you're doing great job and I thank you for that
<JaMa>
(the only "thanks" I got for my years maintaining meta-oe was from darknighte after I've passed the maintenance to armin
sanbeam has joined #yocto
sanbeam has quit [Changing host]
sanbeam has joined #yocto
<moto-timo>
JaMa: I think I have also thanked you a few times ;)
<moto-timo>
I wish I had answer for you khem but I know it's not me to take up the role.
<JaMa>
moto-timo: sorry, so it was you and Sean :)
<moto-timo>
JaMa: but you make a fair point and what khem does is largely not appreciated either.
<moto-timo>
The Maintainer's Dilemma
jmd has joined #yocto
<JaMa>
right, there are more people complaining that something isn't fixed or merged yet, than people showing appreciation for maintainer's work
Kubu_work has quit [Quit: Leaving.]
sanbeam9 has joined #yocto
sanbeam has quit [Ping timeout: 260 seconds]
sanbeam9 has quit [Ping timeout: 260 seconds]
sanbeam9 has joined #yocto
<moto-timo>
complain all you want, I can ignore you :)
<moto-timo>
mute on
sanbeam9 has quit [Ping timeout: 276 seconds]
<moto-timo>
JaMa: I think about all the PN BLACKLIST that you had for meta-ros all those years and nobody fixed a danged thing. So yeah, SKIP_RECIPE or even demote error to warning is just a quick fix. Folks need to step up and contribute to what they consume.
<moto-timo>
khem: I am always amazed at how much you do. I am convinced you do not sleep more than 4 hours a day. Thank you for everything.
<JaMa>
meta-ros was (and still is) special and long-term SKIP_RECIPE makes more sense there (with so many automatically generated recipes where some of them might never be used in any OE based product)
mbulut_ has quit [Ping timeout: 272 seconds]
<moto-timo>
indeed. things are in a pretty good place lately though, thanks to RobW
<JaMa>
another task where I'm glad I'm no longer responsible for :) indeed thaks to RobW
<moto-timo>
but pretty much anything good we have is the result of often one single person doing most of the work, like madisongh for meta-tegra.
<moto-timo>
there are many things I would love to work on for FOSS full-time, but nobody is standing in line to write the checks to cover my salary for that.
<JaMa>
I was never paid for meta-oe work :)
<moto-timo>
meta-java will not be even slightly functional in a year unless someone steps up
<moto-timo>
I was never paid for any meta-python or meta-perl work. I was always told "you can do that on your own time."
<moto-timo>
so if you wonder why I don't do more? that is why. My own time is precious. As is yours and khem's and all the rest of our tireless maintainers.
* tgamblin
wonders what would happen if meta-openembedded went kaboom tomorrow
jmd has quit [Remote host closed the connection]
<moto-timo>
can you ship product without meta-openembedded? if the answer is no, then you should allocate budget to keep it alive.
<moto-timo>
ad nauseum.
<JaMa>
I understand, when I was younger I didn't value my own time enough, now I feel how precious it is, so I'm trying to spend it more visely (and failing most of the time)
<moto-timo>
having kids also changes your perspective on what is important
<moto-timo>
:)
gsalazar has quit [Ping timeout: 260 seconds]
<JaMa>
true, but if you're workaholic then you might realize it only after they leave home (either to school or to have their own life as you were busy with work)
<JaMa>
and that's why I'm still here, chatting with you while waiting for another build to finish (100 more in build queue) while kids went to say hello to grand parents who arrived 20 minutes ago
<JaMa>
and the build queue is indeed never-ending like that hamster wheel
steelswords has quit [Ping timeout: 255 seconds]
jmd has joined #yocto
* moto-timo
is a recovering workaholic
<moto-timo>
I am learning to step away
<moto-timo>
Sorry. Not sorry.
<khem>
I know people appreciate the work and I am happy about that, I am sure others are too but I wish people showed up more on reviewing patches, testing them out, assuming responsibility, it looks like repeat of Tragedy of Commons parable :)
smooge has quit [Quit: I have resigned. I will not be pushed, filed, stamped, indexed, briefed, debriefed, or numbered! My life is my own.]