12:01:19 <jki> #startmeeting CIP IRC weekly meeting
12:01:19 <collab-meetbot> Meeting started Thu Nov 20 12:01:19 2025 UTC and is due to finish in 60 minutes.  The chair is jki. Information about MeetBot at http://wiki.debian.org/MeetBot.
12:01:19 <collab-meetbot> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
12:01:19 <collab-meetbot> The meeting name has been set to 'cip_irc_weekly_meeting'
12:01:20 <iwamatsu__> hello
12:01:30 <jki> #topic AI review
12:01:40 <jki> i don't have any
12:01:47 <jki> 5
12:01:49 <jki> 4
12:01:51 <jki> 3
12:01:52 <jki> 2
12:01:54 <jki> 1
12:01:58 <jki> #topic Kernel maintenance updates
12:02:13 <masami> This week reported 31 new CVEs and 0 updated CVEs.
12:02:14 <uli_> i'm working on 4.19 and 4.4
12:02:45 <masami> While triaging CVEs, I found a bug in the fix patch for CVE-2025-39808 and sent a patch.
12:02:51 <pave1> I am reviewing 6.12.58.
12:02:52 <iwamatsu__> I reviewed 6.12.57
12:02:53 <masami> it was merged into the mainline. https://github.com/torvalds/linux/commit/53f731f5bba0cf03b751ccceb98b82fadc9ccd1e
12:03:50 <jki> very good!
12:05:00 <jki> more topics?
12:05:24 <jki> 5
12:05:26 <jki> 4
12:05:27 <jki> 3
12:05:29 <jki> 2
12:05:31 <jki> 1
12:05:32 <jki> #topic Kernel release status
12:05:48 <jki> 4.19 is late but in the making
12:05:59 <jki> anything in addition?
12:06:34 <jki> 5
12:06:36 <jki> 4
12:06:38 <jki> 3
12:06:40 <jki> 2
12:06:43 <jki> 1
12:06:45 <jki> #topic Kernel testing
12:06:56 <patersonc> Hello all
12:07:22 <patersonc> I've been looking into the kernelCI pipelines in more detail and been adding some more support. Goal is to add some RT tests
12:07:36 <patersonc> Pavel, are there any specific tests you'd like?
12:07:59 <jki> doesn't kernelCI have tests already?
12:08:05 <arisut> patersonc, please rebase the PR
12:08:22 <patersonc> arisut: Yes I'll do that after this meeting
12:08:33 <pave1> There's cyclictest, but I've not really looked into other test.
12:08:38 <patersonc> jki: Yes. We just need to enable them for CIP
12:08:39 <pave1> Testing realtime is not easy.
12:08:59 <patersonc> RT tests already supported by KernelCI can be seen here: https://github.com/kernelci/kernelci-pipeline/blob/main/config/jobs.yaml#L2190
12:09:06 <patersonc> Look for the rt-tests-*
12:09:15 <jki> yes, KernelCI lacks I/O testing, but there should be cyclictest + functional rt-tests
12:09:39 <jki> all that under stress-ng, most likely
12:09:42 <pav31> Lets do that :-).
12:10:07 <pave1> It would be interesting to know how members use/test realtime.
12:10:26 <pave1> Hopefully people having real use for realtime also have some real tests ;-).
12:10:56 <jki> yes, we have, but those are system tests with concrete products and often infrastructure around
12:11:27 <jki> I'm remininding folks all the time that some tests also need to be externalized
12:11:28 <pave1> jki: I know, we had tests that involved oscilloscopes and hardware event sources.
12:12:01 <jki> pavel: it's different today, more RT Ethernets with complex container stacks
12:12:22 <jki> but also some tests that involve other I/O
12:12:31 <pave1> jki: Would it be possible to send a report 'we are testing these and these kernels with these results' maybe once in a year?
12:13:10 <pave1> jki: I assume it differs by industry, really. That elevator company is going to be interested in "are we toggling gpios when we should".
12:13:37 <pave1> jki: I assume some train control is on different level, perhaps doing gpio toggling by non-linux systems connected over ethernet.
12:13:54 <jki> we are working to towards that, just reminded everyone at my event today of sharing kernel configs...
12:14:09 <pave1> Thank yuo!
12:14:13 <pave1> Thank you!
12:15:17 <jki> so, what about the state of kernelci testing with de0-nano-soc?
12:15:22 <jki> https://lore.kernel.org/cip-dev/f2c73b72-cbb3-49c1-8555-9341b3cd3535@siemens.com/T/#u
12:16:36 <jki> also the Renesas board setups looked improveable to me, firmware-wise
12:16:59 <jki> just like we standardized the worker images, we should likely standardized the firmware versions
12:16:59 <patersonc> Did you manage to update the bootloaders in the end on your Renesas boards?
12:17:18 <jki> yes, but that update was from 2013.01 to 2013.01 U-Boot
12:17:27 <jki> that's... not helpful
12:17:30 <patersonc> ha
12:17:30 <iwamatsu__> about de0-nano-soc: I apologize for the delay in reporting this.I am continuing investigation, but I still do not know the cause.
12:18:12 <jki> iwamatsu__: thanks - you can cross-check with the isar-cip-core integration I shared, that worked for me
12:18:48 <iwamatsu__> jki: ok
12:18:57 <jki> we can also update what we have in our lab on the de0 in Munich, unifying things ideally
12:19:26 <jki> I just gave one nano-soc to pasquale for deployment at Siemens Prague lab
12:19:37 <jki> will take some time, though
12:20:52 <jki> regarding other board shippments: I'm awaiting for support from our export control folks, they are very busy unfortunately
12:21:36 <patersonc> I'm assuming that not all devices have upstream bootloader support. Maybe it would be a good side-project for CIP to get this sorted for their reference platforms?
12:21:46 <patersonc> Thank you jki
12:22:31 <jki> while kernelci may be happy with tftp booting, full cip images need recent upstream U-Boot with EFI support
12:22:47 <jki> so we may need to be more picky regarding firmware in the future
12:23:23 <jki> the TI AM62px looks good so far, and the upcoming BeagleV Fire will likely do so as well
12:23:59 <arisut> my de0-nano-soc patch got merged today
12:24:00 <jki> the Nano SoC was... left behind by the SoC vendor in upstream
12:24:12 <jki> oh! didn't get a notice yet
12:24:48 <arisut> is adding the de0-nano-soc context on boot
12:25:09 <jki> arisut: which patch(es) are you exactly referring to?
12:25:17 <arisut> and add nfs
12:25:35 <arisut> https://github.com/kernelci/kernelci-pipeline/pull/1325
12:25:55 <jki> ah, kernelci - didn't track that
12:26:42 <jki> interesting - do they help us already?
12:26:58 <arisut> what do you mean by help us?
12:27:39 <jki> help with making de0-nano-soc boot in kernelci
12:27:49 <jki> maybe even without firmware updates?
12:28:18 <arisut> it should boot with the correct context but is one month old pull request
12:28:30 <arisut> and add nfs
12:28:35 <arisut> so should help
12:29:08 <jki> already ran pipelines with them that failed before?
12:29:18 <jki> like with 6.12-cip?
12:29:33 <arisut> I should check
12:30:10 <arisut> is still a patch I sent one month ago
12:30:48 <jki> I cannot track which of the pipelines you shared last times where run with our without those changes
12:30:48 <arisut> for trying to workout de0 problems
12:31:01 <arisut> without
12:31:15 <jki> but even if those changes should resolve issues, we should still consolidate our firmware versions on our test devices
12:31:22 <arisut> last time pipeline was on production
12:31:35 <jki> ...then without those changes
12:31:35 <arisut> there is no such patch on production yet
12:31:40 <arisut> as is merged today
12:32:28 <patersonc> I guess we'll come back to it next week after the next production roll-out and subsequent CIP push
12:32:34 <arisut> yes
12:33:48 <jki> further testing topics?
12:34:29 <jki> 5
12:34:31 <jki> 4
12:34:33 <jki> 3
12:34:35 <jki> 2
12:34:37 <jki> 1
12:34:39 <jki> #topic AOB
12:35:08 <jki> BeagleV Fire - TSC confirmed our proposal
12:35:22 <jki> how many to request now, how to distribute them?
12:36:12 <pave1> I would not mind having a toy to play with, but I'm probably fine with doing testing remotely.
12:36:42 <patersonc> If TSC approves, I'd like 4 for the LAVA labs. 2 for lab-cip-iwamatsu and 2 for lab-cip-cariad
12:37:12 <jki> uli_: will you need one as well?
12:37:29 <uli_> i would not mind to have one
12:37:31 <jki> for isar-cip-core enabling, I'd like to have one at least initially
12:38:24 <iwamatsu__> Contribution to isar-cip-core, Toshiba can do it.
12:38:59 <jki> iwamatsu__: thanks in advance!
12:39:08 <jki> then we should be fine with 6
12:39:36 <jki> maybe we at Siemens will get some independently, we are also enabling Xenomai on RISC-V right now
12:39:58 <jki> and my colleague doing that expressed interest to follow the CIP board selection :)
12:40:36 <patersonc> \o/
12:40:43 <jki> patersonc: then you have all data to request the boards, likely one-by-one, right?
12:40:53 <jki> direct shippings recommended ;)
12:41:31 <jki> other topics for today?
12:42:03 <patersonc> jki: Do you want me to order all of them?
12:42:28 <jki> I would try to make the requests on one bunch, linking them
12:42:52 <patersonc> sure
12:42:54 <jki> but not one request for all boards, because then you would end up having to redistribute them
12:43:09 <jki> if you can handle that, would be great, yes
12:43:09 <patersonc> I'll figure it out :)
12:43:22 <jki> otherwise raise your hand if I can help
12:43:42 <jki> okay...
12:43:57 <jki> closing remarks, topics?
12:44:20 <jki> 5
12:44:22 <jki> 4
12:44:24 <jki> 3
12:44:26 <jki> 2
12:44:28 <jki> 1
12:44:30 <jki> #endmeeting