<hauke>
it fails with "warning: Not a git repository. Use --no-index to compare two paths outside a working tree"
bluew has quit [Ping timeout: 480 seconds]
mrkiko has quit [Quit: leaving]
<f00b4r0>
ynezz: got mail, also here: https://clbin.com/m3wrn - i believe this will fix the kmodupload issue, which should hopefully allow for a complete run now
tidalf has joined #openwrt-devel
nick[m]1234 has joined #openwrt-devel
zatwai_ has joined #openwrt-devel
zatwai has quit [Read error: Connection reset by peer]
tidalf_ has joined #openwrt-devel
tidalf has quit [Ping timeout: 480 seconds]
tidalf has joined #openwrt-devel
tidalf_ has quit [Ping timeout: 480 seconds]
<ynezz>
f00b4r0: thanks, cooking the update
<f00b4r0>
ynezz: cool. Not sure why the kmodupload doStepIf didn't like the lambda construct, but it's my only explanation for the behaviour. So I just replaced the lambda with a proper function, that should do it
<ynezz>
on gitlab it was possible to push the container from the merge request by the actor, so it was quite easy to test drive them before merging the contribution
<f00b4r0>
i'm not surprised we're seeing bugs with the network operations: it's the only ones I couldn't test, for obvious reasons :)
<ynezz>
I can't find a proper way to doing that on github, looks like the permissions doesnt allow that
<ynezz>
so perhaps we need to hook in some external container registry just for this purpose
whatevs111[m] has joined #openwrt-devel
tidalf has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<ynezz>
error while parsing config file: can only concatenate str (not "Interpolate") to str (traceback in logfile), really need to find a way how to better test this containers
djfe has quit [Ping timeout: 480 seconds]
<f00b4r0>
ok
<ynezz>
should probably do a full config on the the CI
<ynezz>
thats for my change, now will trying with your fix
<f00b4r0>
I know how to fix that. Sorry I don't have my buildbot setup at hand so I'm not configcheck'ing the changes
<[Pokey]>
I would really like to see that merged as I want to add another SoC
<[Pokey]>
would be good to have something to base it upon
<f00b4r0>
ynezz: I'm guessing no news is good news? :)
barhom has joined #openwrt-devel
<ynezz>
f00b4r0: builds started 2 mins ago, so check back in an hour :P
<f00b4r0>
ok. If build is started I'm confident it will go through this time.
<ynezz>
but it runs, thanks for fixes
<f00b4r0>
I think ;)
hexagonwin[m] has joined #openwrt-devel
tidalf has joined #openwrt-devel
Slimey has joined #openwrt-devel
tidalf_ has quit [Ping timeout: 480 seconds]
tidalf has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
tidalf has joined #openwrt-devel
hauke[m] has joined #openwrt-devel
<wigyori>
ynezz: couldn't work out yet what's the issue with the 5.15 kernel boots, so i'm now moving forward to 6.1
<wigyori>
even with the latest bootloader something is screwed up with every second or third kernel update
<wigyori>
looks like that binutils downgrade was a red herring
PtitGNU_ has quit [Read error: Connection reset by peer]
<wigyori>
there were some other reports out in the wild on issues with 5.15 kernels, so...
PtitGNU has joined #openwrt-devel
ctdvqgg445[m] has joined #openwrt-devel
jakllsch_ has joined #openwrt-devel
jakllsch has quit [Ping timeout: 480 seconds]
<ynezz>
wigyori: producing silicon looks dead simple, software is hard :P
<ynezz>
I'm happy, that I didn't bought anything yet, was tempted to add some device into daily runtime testing pipeline
<ynezz>
wigyori: what about simply including it as source-only for the start, it might bring some attention/help eventually
<ynezz>
but if you don't want to bother with 5.15 anymore, I understand :)
<ynezz>
[Pokey]: BTW what hardware do you've?
jakllsch_ is now known as jakllsch
minimal has joined #openwrt-devel
tohojo has joined #openwrt-devel
<f00b4r0>
ynezz: yay, first success!
<wigyori>
ynezz: i've spent long enough time on 5.15 to say i did bother :) but at this point i'd rather go on 6.1 with these. the starfive boards would be on 6.1 anyway, as well as the d1-based ones
gladiac has joined #openwrt-devel
<Slimey>
=)
george_p56 has joined #openwrt-devel
<rmilecki>
hauke: could you send your approvals for the two ARM: dts: BCM5301X patches I sent?
<rmilecki>
hauke: maybe Acked-by
tidalf_ has joined #openwrt-devel
<ynezz>
f00b4r0: indeed, nooice, thanks!
tidalf has quit [Ping timeout: 480 seconds]
<f00b4r0>
ynezz: 👍
<f00b4r0>
i think we can do the same for phase2, and ultimately have both phases under the same master, which will enable us to use DependentScheduler for phase2, which AIUI will be convenient
<f00b4r0>
but that'll have to wait a bit because I have a busy end of May ahead
<Slimey>
heh someone to hold my hand while i submit devices i've ported to openwrt?
<f00b4r0>
ynezz: you'll also want to enable other branches to make sure switchover works, eventually. But I'm not overly worried about that, I did test it
valku has joined #openwrt-devel
<ynezz>
f00b4r0: the plan is to add 23.05 and 22.03 after
<f00b4r0>
👍
schwicht has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<f00b4r0>
ynezz: when all is going and depending on the number of available workers, it will probably be a good idea to adjust the treeStableTimer.
<ynezz>
you mean, increase it?
<f00b4r0>
I think there's a way to cancel ongoing builds that are superseded by newer commits, I'd have to check that again
<f00b4r0>
yes
<f00b4r0>
alternatively we can switch to nightlys for master. With the amount of CI done everywhere, there's probably less need for continuous building
<f00b4r0>
anyway, there's still room for fine tuning, but at least we're off to something more flexible than what we have :)
goliath has quit [Quit: SIGSEGV]
<ynezz>
BTW the scheduling seems a bit off, master_armvirt/64 was built already 3 times
<f00b4r0>
no, it was failed 2 and is now building once
<f00b4r0>
I think that's correct
schwicht has joined #openwrt-devel
<f00b4r0>
I don't remember if I touched the priority logic
<f00b4r0>
git blame would tell :)
schwicht has quit []
schwicht has joined #openwrt-devel
<ynezz>
well, do you want a failing targets consume all the resources?
<f00b4r0>
they won't
<f00b4r0>
it may be an artifact though
<f00b4r0>
look at the other failed targets, they haven't been retried yet
<f00b4r0>
looking at the code I don't think I broke the logic you had implemented, I only added a layer of branch priority
<f00b4r0>
see d86ad174
<ynezz>
well, not blaming you, we've jumped buildbot from 3.5.0 to 3.8.0 so I would expect something over there
<ynezz>
lets see how it fares in long term
<f00b4r0>
i think it's still a bit early to spot statistical patterns :)
<f00b4r0>
yeah
decke[m] has joined #openwrt-devel
evils[m] has joined #openwrt-devel
schwicht has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
schwicht has joined #openwrt-devel
<f00b4r0>
great. I see a bug in libradcli on openwrt/ath79 which I cannot reproduce on linux/x86
<f00b4r0>
that's going to be a joy to debug
george_p56 has quit [Ping timeout: 480 seconds]
goliath has joined #openwrt-devel
<mirko_>
is any of openwrt's "u"-utils capable of asymmetrically encrypting arbitrary data (string / small file) ? can be fully asymmetrically (without deriving a shared symmetric key) or not - data is only a few bytes
Lynx- has quit [Quit: Going offline, see ya! (www.adiirc.com)]
gladiac has quit [Quit: k thx bye]
gladiac has joined #openwrt-devel
<f00b4r0>
and I don't understand CMake. /RAGE
gladiac has quit []
gladiac has joined #openwrt-devel
owrt-images-builds has joined #openwrt-devel
tidalf_ has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<ynezz>
the bot owrt-images-builds is here, probably needs some channel flag to actually output something here?
MaxSoniX has quit [Quit: Konversation terminated!]
<philipp64>
I saw that a couple of people messaged me last week but Colloquy crashed when I tried to read the channel... need to find a better IRC app for MacOS. Wasn't ignoring anyone...
znullptr[m] has joined #openwrt-devel
<f00b4r0>
jow_: you wouldn't happen to have kept a copy of the email that is referred in master.cfg regarding the gitverify step?
<f00b4r0>
it no longer is archived in the wayback machine :(
owrt-images-builds has quit [Quit: buildmaster reconfigured: bot disconnecting]
owrt-images-builds has joined #openwrt-devel
<ynezz>
it is
<f00b4r0>
oh weird. It first gave me a not found
<f00b4r0>
so we have a problem
<f00b4r0>
because checking out the branch HEAD commit puts the repo in detached head state
<f00b4r0>
does not have the git once more and yet it works
<f00b4r0>
and I can't spot the difference
<ynezz>
it happens only on a fresh box, once
<ynezz>
if the repository doesnt exist
<ynezz>
any consecutive Git() call works fine, just the initial one ends up in the detached HEAD state
* f00b4r0
reads the Git() doc again
tidalf has joined #openwrt-devel
<f00b4r0>
ynezz: you know what, I think your fix is as good as any ;P
oliv3r[m] has joined #openwrt-devel
<f00b4r0>
it's harmless, zero performance impact and Just Works™
<ynezz>
maybe git clone has some switch to not leave the clone in that detached HEAD state and we need to propagate that into Git() somehow?
<f00b4r0>
the problem is that the detached HEAD *is* the branch tip
<f00b4r0>
so it's not a "problem". It only becomes one because of gitverify
<f00b4r0>
the fact that Git() performs different operations depending on whether it's starting fresh or updating *is* a problem, and one that isn't under our control :(
<f00b4r0>
what puzzles me is that the rev-parse for the detached-HEAD-yet-branch-tip doesn't say what we expect
patrykk has quit [Read error: Connection reset by peer]
patrykk has joined #openwrt-devel
<f00b4r0>
the doc is unhelpful. I'll check the buildbot code later
<f00b4r0>
meanwhile I think we should go with your workaround.
Gramdalf has joined #openwrt-devel
<f00b4r0>
or accept initial failures for new workers
<ynezz>
if we want to add workers dynamically, its not an option, specifically for releases
patrykk has quit [Read error: Connection reset by peer]
<ynezz>
nick[m]1234: hi, and welcome to the team! :) can't you push it yourself? :P
<nick[m]1234>
ynezz: I said to aparcar that I will wait the additional hours until voting really ends, so after 16th of may. Not sure if that was smart. ^^ So I still miss the onboarding. ;)
<robimarko>
No point really as you already have enough votes
<nick[m]1234>
ok :)
<ynezz>
did he mentioned timezone? I assume its 16th somewhere already :P
<ynezz>
err 17th
<robimarko>
Its 17th in half of Asia already :)
<f00b4r0>
ynezz: agreed. Your workaround seems harmless and works. Let's go with it for now. I'll revisit when I have more time but I believe this is an incompatibility between our getver.sh and buildbot's way of doing things. The latter being most likely not easy to change.
<Gramdalf>
Is there anyone with write access to the repository that could look over this PR, and hopefully merge it (thus supporting a new platform)? https://github.com/openwrt/openwrt/pull/3802 I'm working with the creality wifi box on a development project and it would be nice to have it officially supported.
<Gramdalf>
More information is available on the most recent comment
owrt-images-builds has quit [Quit: buildmaster reconfigured: bot disconnecting]
owrt-images-builds has joined #openwrt-devel
skorpy[m] has joined #openwrt-devel
bluse-blue[m] has joined #openwrt-devel
ThiloteE_ has joined #openwrt-devel
ThiloteE_ has quit []
tidalf has quit [Remote host closed the connection]
cbeznea1 has quit [Ping timeout: 480 seconds]
<f00b4r0>
i'm taking a quick look at the pending patches in patchwork and I see this: http://patchwork.ozlabs.org/project/openwrt/list/?series=349989 I thought this had been shot down already but maybe it was my imagination. The rationale is: it would add a new target to 22.03, and that's a no-no.
<robimarko>
I share that opinion, dont think adding a new subtarget to 22.03 is something that should be done
<hauke>
The "Build Packages with external toolchain (x86/64) / Build x86/64" step is building all packages, but it uses the precompield toolchain
<hauke>
we should do this step with the internal toolchain when we change the toolchain
fpsusername[m] has joined #openwrt-devel
djfe_ is now known as djfe
domon has joined #openwrt-devel
<nick[m]1234>
weird, I can not merge in opewrt/packages via github web interface. anyone else seeing this bug?
dgcampea is now known as Guest378
dgcampea has joined #openwrt-devel
<ynezz>
nick[m]1234: I've just made you an owner, so you can check it yourself
<ynezz>
the convention is, that all core members have same privileges
<hauke>
nick[m]1234: I can not push to my github repo
<hauke>
looks like github has problems
<hauke>
now it works again
<ynezz>
yes, i needed to use OTP, mobile auth is not working
<nick[m]1234>
ynezz: thanks :)
Guest378 has quit [Ping timeout: 480 seconds]
<ynezz>
and it seems like azure builder has connectivity issues all day
<Snuupy>
hauke: you meant to reply to someone else, right?
<hauke>
Snuupy: sorry yes
<Snuupy>
:)
<hauke>
I wanted to reply to ynezz
<nick[m]1234>
aparcar: with the musl bump a lot of packages do not build anymore. Just hunting for those packages and fixing them. Maybe waiting with the branch would be nice.
<apritcha>
working on adding device support for the first time, and this is probably a stupid question, but my pipeline failed due to kernel patches requiring refresh, the error said to run "make target/linux/refresh" and then force push, but git says everything's already up to date when trying to push, anyone have any ideas? not 100% up on things beyond the basics with git sorry