12:02:32 <jki> #startmeeting CIP IRC weekly meeting
12:02:32 <collab-meetbot`> Meeting started Thu Apr 20 12:02:32 2023 UTC and is due to finish in 60 minutes.  The chair is jki. Information about MeetBot at http://wiki.debian.org/MeetBot.
12:02:32 <collab-meetbot`> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
12:02:32 <collab-meetbot`> The meeting name has been set to 'cip_irc_weekly_meeting'
12:02:36 <jki> #topic AI review
12:02:48 <jki> 1. create kernelci pipeline for buster images (arisut)
12:03:46 <jki> arisut: ?
12:03:56 <arisut> just talked with kernelci about the project of using debian style for testing CIP kernel on KernelCI
12:04:11 <arisut> just recently
12:04:27 <jki> ok, so "ongoing"
12:04:34 <arisut> yes
12:04:42 <jki> good
12:04:47 <jki> other AIs?
12:05:33 <arisut> not from me
12:05:48 <jki> then let's move on in
12:05:50 <jki> 1
12:05:51 <jki> 3
12:05:53 <jki> 2
12:05:55 <jki> 1
12:05:58 <jki> #topic Kernel maintenance updates
12:06:13 <masami> This week reported 8 new CVEs and 2 updated CVEs.
12:06:28 <uli> reviewing 5.10.178
12:06:45 <iwamatsu> I reviewed 5.10.178.
12:08:47 <jki> anything else?
12:09:09 <jki> 3
12:09:11 <jki> 2
12:09:14 <jki> 1
12:09:17 <jki> #topic Kernel release status
12:09:21 <jki> -4.4
12:09:26 <uli> on track
12:09:35 <jki> -4.19
12:09:45 <iwamatsu> RT has not followed LTS...
12:10:12 <jki> but we would have to release now?
12:11:01 <jki> strictly spoken, no
12:11:17 <jki> 0.5 per month, last 4.19-rt was end of March
12:12:15 <jki> ok, then
12:12:17 <jki> -5.10
12:12:28 <iwamatsu> same 4.19.
12:12:45 <iwamatsu> s/same/same as /
12:12:58 <iwamatsu> If RT is not released this week or next week, it will be difficult to release cip-rt.
12:12:59 <jki> here we should do once per month, thus here we are affected then
12:13:07 <jki> understood
12:13:52 <jki> is the same one doing 4.19-RT and 5.10-RT upstream?
12:14:12 <jki> just checked - no
12:14:23 <jki> 4.19 is Daniel, 5.10 is Luis
12:15:33 <iwamatsu> yes, right
12:17:07 <jki> sorry, SIGREALLIFE
12:17:08 <iwamatsu> and Daniel has been release 4.19.280-rt123-rc1. he is updating.
12:17:33 <iwamatsu> hi, pavel
12:17:44 <pave1> Hi!
12:18:06 <pave1> Sorry for the delay. I was reviewing 5.10.178.
12:18:44 <jki> back
12:19:16 <jki> ok, then we are done on this
12:19:22 <jki> #topic Kernel testing
12:21:03 <patersonc[m]> I've nothing special to report this week. Just the usual maintenance & stable-rc result reporting
12:21:33 <arisut> the isar-cip-core work for kernelci is ongoin
12:22:08 <jki> what's the state now for 6.1-cip?
12:23:31 <arisut> without branch there is no testing
12:23:35 <jki> pavel: IIRC, you wanted to follow-up with next steps needed, right?
12:23:54 <pave1> Yes.
12:24:11 <arisut> if we have a branch we can start testing it
12:24:22 <pave1> I guess we should announce that we now expect patches for 6.1
12:24:32 <pave1> for any issue fixed in 5.10.
12:25:05 <pave1> Then we agree on starting point, branch off, and check for any patches in
12:25:10 <patersonc[m]> Should we also ask the members if they have any particular configs they want built/tested/supported for 6.1?
12:25:40 <jki> sure, configs will also be needed
12:25:52 <pave1> 5.10 but not in 6.1.
12:25:56 <jki> at least forward ports of the qemu configs, to have a starter
12:26:00 <pave1> Yes, and configs.
12:26:50 <pave1> We can start from those used for testing  6.1-stable.
12:27:14 <jki> when did we branch off 5.10-cip back then?
12:27:23 <jki> after the first backport arrived or earlier?
12:27:25 <patersonc[m]> We already have configs copied from 5.10 (which are used for stable testing). But maybe members what to review/refresh/add to those configs
12:28:00 <jki> pavel: can you do the call-out on the list for the configs?
12:28:01 <patersonc[m]> https://gitlab.com/cip-project/cip-kernel/cip-kernel-config/-/merge_requests/74/diffs?commit_id=570fa322ab77a7c6b5779a6a89bccbdb77573374
12:28:15 <arisut> configs will be added to cip-kernel-config repo?
12:28:25 <jki> yep
12:28:46 <arisut> ok
12:29:08 <jki> we already have a 6.1-cip branch there
12:29:24 <jki> in fact, we have many branches there, not only for cip kernel it seems
12:29:40 <jki> "6.2.y-cip"?
12:30:57 <patersonc[m]> Could be better named 6.2.y tbh
12:31:20 <jki> would be less confusing, yes
12:31:50 <jki> not sure if anything elsewhere (kernelci?) is relying on the current path names, though
12:31:51 <patersonc[m]> It was agreed in the MR to keep it like that
12:32:53 <patersonc[m]> We don't have any -rt configs for 6.1 yet btw, so we'd also need to get those added
12:33:04 <patersonc[m]> Assuming we're doing a linux-6.1.y-cip-rt
12:33:22 <iwamatsu> Yes, I've heard that 6.2 is removed from cip-kernel-config when it is removed from stable.
12:33:27 <jki> I see no reason for not doing 6.1-rt
12:35:10 <jki> ok - so, who is doing what next?
12:35:51 <patersonc[m]> I'll add the pipeline for 6.1 cip testing
12:35:51 <pave1> I have sending that mail on my todo.
12:36:30 <pave1> Plus I have a script for for checking if we have all the patches from old kernel.
12:36:30 <alicefm> I will follow/do the work on kernelci for 6.1 testing
12:37:05 <jki> sounds like a good plan, thanks
12:37:16 <jki> anything else on testing?
12:37:46 <arisut> same
12:38:13 <patersonc[m]> Not from me
12:38:23 <jki> then let's move on...
12:38:25 <jki> 3
12:38:27 <jki> 2
12:38:29 <jki> 1
12:38:31 <jki> #topic AOB
12:39:34 <jki> anyone anything?
12:40:02 <jki> btw, no reply from Greg so far after my last, longer answer
12:40:23 <patersonc[m]> MR for the 6.1-cip CI pipeline is up: https://gitlab.com/cip-project/cip-testing/linux-cip-pipelines/-/merge_requests/38
12:40:43 <pave1> jki -- at some point it will make sense to make our bosses to talk to his boss...
12:41:06 <jki> does he have a boss? :)
12:41:34 <patersonc[m]> ha
12:41:35 <pave1> Well, no, but someone pays him ...
12:41:50 <jki> bosses need input on what we need and how much effort we estimate for that
12:41:58 <jki> and I think we are still at that level
12:42:14 <patersonc[m]> It's going to take more than CIP to change the LTS length policy. It'll need lots of companies to commit to supporting the efforts.
12:42:20 <jki> sure
12:42:52 <jki> that's why I was proposing to something like an LTS project - if that makes sense - under LF, with real, paying members
12:43:08 <patersonc[m]> jki: Cunning
12:43:10 <jki> or CIP would need a different budget and partially a different mission
12:43:19 <pave1> Taking over -lts from greg?
12:43:46 <jki> CIP can do that as LTS is way broader in scope and we are still too small
12:44:20 <jki> also, CIP would need members from - say - consumer domains if consumer .configs should come into scope
12:44:29 <jki> or server .configs
12:45:01 <jki> but Debian LTS was also asking a bit into that direction as they have members from domains not present in CIP
12:46:44 <Dr_Who> we've been exploring doing LTSes that pickup where Greg leaves off. (We == Linaro)
12:47:18 <Dr_Who> We already do quite a bit of kernel regression testing that like other projects report into Greg right now
12:47:27 <pave1> jki -- 'can not'?
12:47:54 <jki> Dr_Who: not sure who you are ;), but Grant was indicating this as well to me
12:48:25 <jki> pavel: I was thinking "cannot" and then wrinting "yes, we can" ;)
12:48:31 <jki> but I meant "cannot"
12:48:43 <Dr_Who> so Linaro is an open source engineering company that a number of companies came together and founded back in 2010
12:48:44 <pave1> Thanks, makes sense.
12:48:56 <Dr_Who> Arm, Google, Samsung, ST just to name a few
12:49:05 <jki> all well known
12:49:43 <jki> and Linaro did LTS work before, that's where KernelCI and LAVA had its roots to my understanding
12:49:50 <Dr_Who> yup
12:50:03 <pave1> Well, if Linaro is willing to take over lts mainainance. that would certainly help us...
12:50:39 <jki> and if that needs extra funding, we should make that formal, but it should be scoped on LTS
12:51:10 <jki> I can imagine that this can attract more - CIP would surely consider supporting such an effort
12:51:23 <jki> in any case, someone needs to take a lead here IMHO
12:51:38 <Dr_Who> We're at the exploring phase, there seem to be a number of companies in the same situation, wanting an LTS - as you rightly point out, there is a cost associated with it, engineers like to get paid
12:53:11 <pave1> Perhaps Linaro should be included in that mailthread with Greg?
12:53:15 <Dr_Who> certainly up for a chat
12:53:30 <Dr_Who> interestingly I have talked to Greg about this a few weeks back
12:53:57 <jki> Dr_Who: who should be on the table for such a chat?
12:54:19 <Dr_Who> that'd be me tom.gall@linaro.org is my address
12:55:30 <jki> Dr_Who: great, will follow-up with you
12:55:35 <Dr_Who> thank you!
12:56:21 <jki> good
12:56:30 <jki> anything else for today?
12:57:06 <jki> 3
12:57:07 <jki> 2
12:57:09 <jki> 1
12:57:12 <jki> #endmeeting