12:03:23 <jki> #startmeeting CIP IRC weekly meeting
12:03:23 <collab-meetbot> Meeting started Thu Nov 10 12:03:23 2022 UTC and is due to finish in 60 minutes.  The chair is jki. Information about MeetBot at http://wiki.debian.org/MeetBot.
12:03:23 <collab-meetbot> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
12:03:23 <collab-meetbot> The meeting name has been set to 'cip_irc_weekly_meeting'
12:03:32 <jki> #topic AI review
12:03:48 <jki> 1. Add qemu-riscv to cip-kernel-config - patersonc
12:04:11 <patersonc[m]> No updates
12:04:37 <jki> 2. Create MR for disabling smc test in qemu - iwamatsu
12:05:09 <iwamatsu`> not touch yet
12:05:43 <jki> any other AIs?
12:06:14 <alicef> iwamatsu`: I will work on that
12:06:50 <jki> ok, AI transfered - thanks :)
12:06:54 <alicef> I talked today with kernelci about disabling smc test in qemu and I didn't got objection
12:07:03 <iwamatsu`> alicef: thanks
12:07:25 <alicef> today we had kernelci TSC and weekly meeting
12:07:54 <alicef> so I discussed a bit about topics that come out here
12:08:08 <alicef> iwamatsu`: np
12:10:38 <alicef> jki: thanks we can go on sorry for the interruption
12:10:43 <jki> np
12:10:47 <jki> #topic Kernel maintenance updates
12:11:05 <pave1> I did reviews on 5.10.153 and 5.10.154.
12:11:12 <uli> reviewing 5.10.154
12:11:41 <pave1> 4.4-cip & 4.4-cip-rt -- I guess I need to look at that. Sorry.
12:11:47 <pave1> 5.10-cip-rt -- released.
12:11:50 <iwamatsu`> I am reviewing 5.10.154.
12:12:10 <pave1> 4.19-cip-rt -- that will be fun.
12:13:07 <pave1> Upstream is preparing v4.19.255-rt114-rc2 update.
12:13:48 <pave1> But we are currently at 4.19.264-cip84. So ... upstream is behind by a lot.
12:14:06 <jki> do you know any reasons for that?
12:15:31 <pave1> I guess they hit some kind of problem with TIMER_IRQSAFE timers? Not sure.
12:15:43 <pave1> They should be releasing v4.19.255-rt114 today.
12:17:27 <pave1> I believe it makes sense to wait them for few more days to see what is going on there.
12:18:08 <pave1> Or I can ask when 4.19.255 announcement comes.
12:18:14 <jki_> would be good for us to be at least informed
12:18:30 <jki_> we may face questions from our users about why cip-rt is behind
12:18:45 <jki_> relates to my AOB topic...
12:19:54 <jki_> what's the 4.4 status now?
12:20:03 <pave1> 4.4-cip & 4.4-cip-rt -- I guess I need to look at that. Sorry.
12:20:30 <jki_> for 4.4-cip, we are still on a "once per month" promise
12:21:15 <pave1> I know. 4.4-cip could be released as is, if we wanted to keep the promise :-).
12:21:51 <pave1> But I am in the middle of move to newer 4.4-st, and there's that annoying warning from IP code that would be worth solving...
12:22:46 <jki_> if you need support, raise a hand, but we should at least clarify reasons for delays
12:24:03 <jki_> if there are no relevant fixes pending, not releasing is obviously not a problem
12:24:35 <pave1> We have
12:25:36 <pave1> Current 4.4-cip is up-to-date with 4.9.325. That is not yet released.
12:26:21 <pave1> 4.4-st-rc is mostly up-to-date with 4.9.328.
12:26:33 <pave1> Plus there's the ugly warning.
12:27:20 <pave1> If we want to do a release, we can... But we should be able to do release with 4.9.328 next week.
12:27:24 <jki_> is this basically the state for the last 3 months?
12:27:40 <jki_> unless I missed something, last release was in August
12:28:30 <pave1> Major changes are happening with lockstep with mainline kernels.
12:28:47 <pave1> I guess reducing release cadence for 4.4 would make sense.
12:29:29 <jki_> then we need reasoning for that
12:29:58 <jki_> but we would have missed even the cadence for 4.4-rt
12:32:37 <jki_> so, how to proceed now best for the current 4.4 situation?
12:33:10 <pave1> I believe best way forward is to do release with 4.9.328 changes next week
12:34:05 <jki_> ok
12:34:59 <pave1> Plus I can graph changes over time to 4.9. My feeling is that releasing 4.4 monthly does not make much sense any more because significant changes basically arrive every two months.
12:35:35 <pave1> I can gather data to support that after we do the release, I guess.
12:35:47 <jki_> sounds good to me
12:35:52 <pave1> ok :-)/
12:37:38 <jki_> anything else?
12:37:51 <jki_> 3
12:37:53 <jki_> 2
12:37:56 <jki_> 1
12:38:00 <jki_> #topic Kernel testing
12:38:52 <alicef> I have nothing other to add
12:39:05 <patersonc[m]> The gitlab runners have been upgraded to use the latest s/w. We've also upgraded the underlying EC2 instances
12:39:18 <patersonc[m]> So far I haven't seen any issues. Please keep an eye out though
12:39:37 <patersonc[m]> * gitlab runners used for test jobs have been
12:40:28 <patersonc[m]> lab-cip-denx is currently offline. I've notified Denx
12:42:02 <patersonc[m]> I don't think I have anything else to report
12:43:25 <jki_> anyone else with testing topics?
12:44:13 <jki_> 3
12:44:15 <jki_> 2
12:44:17 <jki_> 1
12:44:19 <jki_> #topic AOB
12:44:35 <jki_> - actively communicating deviations from the kernel release schedule
12:45:04 <jki_> I was thinking about this these days, to help our users follow the CIP kernel work
12:45:22 <jki_> what do you think?
12:47:24 <patersonc[m]> Makes sense to me
12:48:03 <alicef> also to me
12:48:09 <patersonc[m]> How did you plan to do it? Emails to cip-dev?  Or some sort of status page?
12:48:14 <pave1> I think we may try to use this as an opportunity to see who uses which -cip kernels and what release schedules they require.
12:48:30 <pave1> You are quite good at anouncing deviations at CIP TSC meetings :-).
12:48:32 <jki_> cip-dev would be my first idea
12:48:41 <jki_> same channel as for release announcements
12:49:10 <jki_> well, the TSC report template I inherited contains a report section on releases
12:49:16 <pave1> Early next year will have 4th kernel to maintain, and reducing the schedules for the old ones might help.
12:49:36 <jki_> and if I have to put all entries to "none", I scratch my head ;)
12:50:06 <jki_> sure, this is not about releasing needlessly frequent
12:50:15 <iwamatsu`> jki: don't understand the sentence correctly. Does Devisions mean a delay in release?
12:50:31 <pave1> :-). Yes, but noone at the TSC meeting seemed to care.
12:50:49 <jki_> I mean when we promised to release a kernel once per month, and we are delayed by a week already
12:51:21 <jki_> not when we are 2 days off for a biweekly release schedule
12:51:44 <jki_> we had the cases for RT in the past with missing upstream updates
12:52:15 <iwamatsu`> I understood.
12:52:16 <pave1> And it is quite apparent that we'll have more such issues in future, at least for RT :-(.
12:52:34 <iwamatsu`> s/understood/understand/
12:52:40 <jki_> sure, and we won't be able to resolve them all in time
12:52:47 <jki_> transparency would help here
12:53:09 <jki_> and may also help triggering awareness for our work - and maybe even more contributions
12:53:14 <patersonc[m]> Perhaps some sort of regular report on kernel work to cip-dev would be a good thing
12:53:42 <patersonc[m]> Help show the work we're doing with 4.4-st for example
12:54:26 <jki_> we could start with gathering a status more formally during this meeting
12:54:59 <jki_> I would volunteer sharing this via cip-dev, regularly or on demand - tbd
12:56:03 <pave1> If the goal would be to do more regular releases, that would work.
12:56:24 <pave1> But ... at least for rt ... maybe it would be more fair to clearly state that we depend on upstream for this.
12:56:45 <jki_> regular transparency would be my primary goal
12:57:12 <pave1> ...so there's basically not a "release schedule".
12:57:19 <pave1> ...but we release when upstream does.
12:58:19 <pave1> ...and ask users and our security team to tell us if it is not frequent enough.
12:58:51 <jki_> we have had no complaints about 4.4 delays yet, that is good news
12:59:12 <jki_> but we don't get the silent impressions of observers this way
12:59:41 <pave1> So my impression is that noone uses 4.4-rt. We probably still have some 4.4-cip users :-).
13:00:25 <iwamatsu`> Perhaps the user wants to know the reason for the delay, so first suggests the reason for the delay.
13:00:39 <jki_> again, we are not just late with 4.4-rt, but also with 4.4 itself
13:01:12 <jki_> I will add some topic section "kernel release status" for the next irc meeting to give this a start
13:01:36 <jki_> then we can think about how to make use of it best
13:02:14 <iwamatsu`> +1
13:04:03 <pave1> Yep. I should have some data about patches going to 4.4 next week.
13:04:17 <jki_> perfect
13:05:28 <jki_> anyone anything else for today?
13:06:12 <jki_> 3
13:06:14 <jki_> 2
13:06:15 <jki_> 1
13:06:18 <jki_> #endmeeting