13:02:32 <jki> #startmeeting CIP IRC weekly meeting
13:02:32 <collab-meetbot> Meeting started Thu Jun  5 13:02:32 2025 UTC and is due to finish in 60 minutes.  The chair is jki. Information about MeetBot at http://wiki.debian.org/MeetBot.
13:02:32 <collab-meetbot> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
13:02:32 <collab-meetbot> The meeting name has been set to 'cip_irc_weekly_meeting'
13:02:43 <jki> #topic AI review
13:02:53 <jki> I have none recorded
13:03:02 <jki> anything to add?
13:03:14 <jki> 5
13:03:16 <jki> 4
13:03:18 <jki> 3
13:03:18 <pave1> Nothing from last time.
13:03:21 <jki> 2
13:03:23 <jki> 1
13:03:26 <jki> #topic Kernel maintenance updates
13:03:40 <uli_> i'm working on 4.19
13:03:41 <masami> This week reported 7 new CVEs and 31 updated CVEs.
13:04:09 <pave1> I was reviewing 6.12.31 and .32
13:04:18 <iwamatsu__> I reviewed 6.12.31 and 32. And I released 6.12.y-cip.
13:04:57 <pave1> There's series of unix domain socket changes queued for 6.1.X.
13:05:08 <pave1> Apparently, garbage collection is hard.
13:05:24 <pave1> If someone uses them heavily, it may be worth testing extra hard.
13:07:51 <jki> ok - anything to add?
13:08:08 <jki> 5
13:08:10 <jki> 4
13:08:11 <jki> 3
13:08:14 <jki> 2
13:08:15 <jki> 1
13:08:18 <jki> #topic Kernel release status
13:08:29 <jki> all kernels are on track right now
13:08:39 <jki> any problems ahead?
13:09:15 <jki> 5
13:09:17 <jki> 4
13:09:19 <jki> 3
13:09:21 <jki> 2
13:09:23 <jki> 1
13:09:25 <jki> #topic Kernel testing
13:09:48 <patersonc> Nothing from me
13:09:53 <arisut> sent kci-dev token
13:10:00 <arisut> by email
13:10:11 <patersonc> Thanks. I tested it and it works for me
13:10:18 <arisut> nice
13:10:49 <iwamatsu__> alisut: I revieved. but I don't test  yet.
13:11:42 <pave1> Attempted to test it. Did not get kci-dev to install on work machine.
13:12:23 <arisut> pave1, did you get any error ?
13:12:35 <pave1> Lets talk after the meeting?
13:12:49 <arisut> write the issue on the kci-dev github if you enconter any problem
13:12:51 <arisut> sure
13:13:56 <arisut> also there was a proposal from kernelci to testing this in the future https://github.com/kernelci/kernelci-pipeline/issues/1126
13:14:09 <arisut> .kernelci.yml file
13:14:38 <arisut> should allow to manage kernelci pipeline from cip kernel repository
13:14:48 <patersonc> Could be an easy way for us to control our pipelines
13:15:51 <arisut> the issue talks about adding it in the kernel tree but after today discussion looks like that is possible to do it in different ways
13:16:04 <arisut> like adding the file to a ad-hoc branch
13:16:30 <arisut> today kernelci meeting discussion
13:17:07 <jki> yeah, we should not carry (more) special CI files in-tree unless really unavoidable
13:17:25 <pave1> it really does not work well while bisecting, for example.
13:17:27 <arisut> yes that was today discussion at KernelCI meeting
13:18:51 <arisut> I agree on not carry special CI files in-tree unless unavoidable
13:20:28 <jki> are there already discussions regarding onboarding of the new referencen board of TI for us?
13:20:46 <arisut> ?
13:21:09 <jki> the AM62x... we discussed at the TSC this week
13:21:30 <jki> it was said it is already part of the kernelci farm
13:22:02 <arisut> I'm not aware of such discussion
13:22:14 <arisut> afair
13:22:20 <jki> the ball is with TI, just curious
13:22:58 <jki> but I assume we also want it then in the CIP farm, in addition to kernelci, or is that no longer needed?
13:23:06 <arisut> there was no such discussion at today kernelci meeting
13:23:26 <jki> i rather meant discussions in cip context
13:24:35 <patersonc> I guess if they already have labs connected to KernelCI, once we've finished adding our confius etc. to KernelCI then we wouldn't need TI's boards specifically in CIP labs
13:24:46 <patersonc> But that's only relevant for kernel testing
13:24:56 <patersonc> For CIP core testing etc. we'd need a board in a CIP lab
13:25:05 <jki> right, we could still benefit from it for core testing
13:27:56 <jki> ok, anything else on testing?
13:28:46 <jki> 5
13:28:48 <jki> 4
13:28:50 <jki> 3
13:28:52 <jki> 2
13:28:54 <jki> 1
13:28:57 <jki> #topic AOB
13:29:28 <jki> ppc32 - i just wanted to bring that topic up here as well, after only talking about it at the TSC meeting
13:29:48 <jki> everyone followed that? or should I summarize first?
13:30:04 <pave1> Probably summarize.
13:31:00 <jki> sure: we (Siemens) have a some products with PowerPC32 SoCs still under maintenance, and that long into the next decade, likely
13:31:23 <jki> question was brought to me if the CIP kernel could help here as well
13:31:57 <uli_> what kernels are they running now?
13:32:01 <jki> if there are others intested in putting this arch to rest in a similar way
13:32:28 <jki> I don't have stats yet, but they were trying to stay very close to mainline so far
13:32:38 <jki> and mainline did not discontinue ppc32 - yet
13:32:50 <jki> yocto arlready did (that's what they use)
13:33:14 <jki> but may accept the arch as tier 2 (no official support, no release blocker) still
13:33:55 <jki> I mean, we are even getting m68k patches for CIP, don't we? :)
13:33:57 <iwamatsu__> Which series do we use? I understand there are many types of PPC32.
13:34:35 <pave1> jki: I guess adding ppc32 target to test farm would help a lot.
13:34:39 <jki> also no data yet, I already asked the same
13:34:52 <pave1> jki: If it is broken, Siemens gets to fix it. After that, we can keep eye on that.
13:35:06 <jki> that would already be a great starting point
13:35:08 <iwamatsu__> And what about rootfs and cross-tools? Debian is no longer supported.
13:35:17 <pave1> jki: I don't expect that to be too much work.
13:35:31 <jki> given that it hard to impossible to get any EVMs for those SoCs anymore, duty would be on us anyway
13:35:37 <jki> right
13:36:00 <jki> my personal interest was just to find out if there aren't others with similar problems now
13:36:34 <jki> and then the question is if we could already learn something from it thinking of the inevitible arm32 retirement one day...
13:36:49 <pave1> I guess I have two ppc32 machines here. One of them got indirect lightning strike tonight.
13:37:04 <jki> ouch
13:37:12 <pave1> But they are called ADSL routers and I don't normally hack on them, so... :-)
13:37:23 <jki> yes, this kind of losses will make maintenance also harder and harder over time :)
13:37:59 <pave1> (And I may have the architectures confused. You are talking ppc32, and this is likely mips.)
13:39:26 <iwamatsu__> I also have a ppc (mac mini).
13:39:36 <iwamatsu__> I noticed that ppc32 can be used for unstable. It is also built in debian-ports.;-)
13:40:26 <jki> let me clarify what hardware would actually be helpful at all - and if we cannot help here
13:40:42 <jki> but already qemu would be nice as basic test
13:41:28 <pave1> Yeah. If -cip had regression from -stable that would be visible on qemu, that would be kind of bad.
13:41:54 <jki> it is already great news that you do not see general blockers
13:43:09 <jki> anything else for today?
13:44:44 <jki> 5
13:44:46 <jki> 4
13:44:48 <jki> 3
13:44:50 <jki> 2
13:44:52 <jki> 1
13:44:55 <jki> #endmeeting