07:10:35 * ** rajm has joined #cip 07:11:08 * ** patersonc has joined #cip 07:20:54 * rajm runs the daily health on the BBB with an email to the list. 07:33:40 * rajm runs the daily health check on the iwg20m with an email to the list 07:39:14 <patersonc> :) 07:40:09 <patersonc> I haven't received the iwg20m email though, just the bbb one (#192). 07:45:02 * ** toscalix has joined #cip 07:46:17 <rajm> hmm - it definitely came via cip-testing-results did you get the one yesterday patersonc ? 07:46:56 <patersonc> Ah, it's just arrived! 07:47:07 <patersonc> And yes I did get the one yesterday 07:47:10 <patersonc> Sorry for the fuss 07:54:16 <rajm> ok no problem! 07:58:02 * ** sangorrin has joined #cip 07:59:16 * ** mungaip has joined #cip 07:59:19 * ** wagi has joined #cip 07:59:58 <szlin> wagi: wave 08:00:17 <szlin> #startmeeting 08:00:20 <szlin> #roll call 08:00:23 <wagi> szlin: hi :) 08:00:27 <szlin> please say hi if you're around 08:00:34 <sangorrin> 你好 08:00:35 <szlin> hi ._./ 08:00:35 <patersonc> hi 08:00:36 <rajm> hi 08:00:38 <szlin> sangorrin: ! 08:01:11 <bwh> hi 08:01:11 <szlin> #topic kernel maintenance updates 08:01:23 <szlin> The following patch was merged into 4.4-stable. 08:01:23 <szlin> d733f7542ad4 ("drivers: net: cpsw: fix segfault in case of bad phy-handle") 08:01:24 <szlin> https://www.spinics.net/lists/stable/msg258336.html 08:01:52 <szlin> === 4.4.155 reviewing === 08:01:53 <szlin> I have reviewed patches below without any issue 08:01:53 <szlin> f7f501c753f3 ("getxattr: use correct xattr length") 08:01:53 <szlin> c2b736ff27b4 ("spi: davinci: fix a NULL pointer dereference") 08:02:00 <szlin> Whilst reviewing the patch c2b736ff27b4 ("spi: davinci: fix a NULL pointer dereference"), I think below patches should be also backported. 08:02:03 <szlin> 7f3ac71ac3b0 ("spi: davinci: fix spurious i/o error") 08:02:06 <szlin> 35fc3b9ff66b ("spi: davinci: Handle return value of clk_prepare_enable") 08:02:56 * ** hungtran has joined #cip 08:03:52 <sangorrin> szlin, thank you. 08:03:55 <wagi> Can't really say anything usefull about those parts of the kernel 08:04:43 <szlin> The reivew progress is quite slow, but keep going. 08:05:13 <szlin> I am thinking about to create a wiki page for listing reviewing progress? 08:06:25 <sangorrin> it would be nice to distribute the review among the kernel team members 08:07:49 <sangorrin> do you think this patches could be accepted in cip? https://lore.kernel.org/patchwork/cover/622352/ 08:08:04 <sangorrin> they touch some critical parts 08:08:20 <sangorrin> such as drivers/base/property.c 08:08:32 <szlin> if no one objects, I will try to create a wiki page for it 08:08:34 <sangorrin> but they are necessary to support non-ACPI platforms such as Intel Skylake 08:08:47 <patersonc> szlin: Good idea 08:08:52 * ** iwamatsu__ has joined #cip 08:09:42 <szlin> IMHO, stable won't accept this kind of patch but only bug-fixed 08:10:08 <sangorrin> szlin: you can also send us homework. I can probably review a few patches per week 08:10:21 <sangorrin> the important thing is that we dont review the same patches 08:10:27 * ** iwamatsu__ has quit IRC 08:10:38 <bwh> szlin: Right, it's definitely not OK for stable. Might be OK for CIP (I didn't look very far). 08:10:43 <sangorrin> I said CIP not stable 08:11:01 * ** iwamatsu__ has joined #cip 08:11:10 <szlin> sangorrin: it depends on maintainer ? 08:11:56 <sangorrin> yes, I'm afraid of causing regressions.. 08:13:14 <szlin> sangorrin: yes, this patch contains lots of files and changes... 08:14:08 <szlin> Furthermore, I will send the v2 dts patch to upstream for our board this week 08:14:50 <szlin> any other point? 08:15:32 <patersonc> o/ 08:15:35 * ** iwamatsu__ has joined #cip 08:15:40 <szlin> wagi: Im so happy to see you, do you have any updates in rt-stable? 08:15:59 <wagi> yes, there is a problem with spinlocks 08:16:21 * ** iwamatsu__ has quit IRC 08:16:25 <wagi> This was I was haunting now for 1 year. As it turns out the qspinlock code is not working correctly 08:16:43 <wagi> just a sec I fetch an url 08:16:44 <szlin> wagi: do you have related link? 08:17:06 <szlin> patersonc: sorry, please go ahead 08:17:11 <patersonc> Speaking of patchwork - is there any value in setting this up (or something else) for CIP? Especially if we're hoping to start getting more maintainers. 08:17:19 <wagi> https://lore.kernel.org/lkml/20180921120226.6xjgr4oiho22ex75@linutronix.de/ 08:17:43 <wagi> so for 4.4. I'll drop back to the ticket spinlock implementation 08:18:20 <wagi> I am testing this right now on the hardware I was observing the problem 08:18:27 <wagi> so far 12h without an issue 08:19:27 <bwh> patersonc: I haven't used it before as a maintainer, but I think it could be useful 08:19:48 <szlin> wagi: thanks, please keep us posted 08:20:20 <patersonc> bwh: The renesas-soc mailing list uses it and it seems to work. Helps to organise reviewers and I assume it helps the maintainers keep an eye on everything. 08:20:31 <szlin> patersonc: I know some subsystem use patchwork but some don't use it. 08:20:32 <wagi> szlin: sure 08:21:17 <szlin> patersonc: I think it could be useful, too 08:21:53 <patersonc> Should we ask TSC? Or can it just be the maintainers call? I don't know if it would cost for LF to set it up for us 08:22:02 <szlin> #agreed create a wiki page for stable patches review 08:22:13 <patersonc> (unless there are other suggestions) 08:22:32 <bwh> I would hope that we could use an existing instance (patchwork.kernel.org or patchwork.ozlabs.org) 08:22:57 <szlin> #action SZ Lin create a wiki page for collaboration 08:23:28 <patersonc> bwh: good point 08:24:06 <szlin> bwh: do you know the differences between patchwork.kernel.org and patchwork.ozlabs.org? 08:25:15 <bwh> szlin: AFAIK they are running the same software but different people administer them (and might have different rules for which lists they include) 08:25:54 <szlin> bwh: do you have idea that which one is more suitable for CIP? 08:26:23 <bwh> I don't 08:27:27 <szlin> I see, so we need to do some survey if we want to use patchwork 08:28:32 <szlin> patersonc: I think we can discuss this in TSC meeting 08:29:26 <patersonc> Okay 08:30:45 * ** hungtran has quit IRC 08:30:50 <szlin> #action Discuss patchwork item in TSC meeting 08:30:55 <szlin> any other point? 08:31:01 * ** hungtran has joined #cip 08:31:08 <szlin> #topic kernel testing 08:31:19 <rajm> I tried Nguyen's suggestion on #186 (the iwg20m) and so far it works reliably, looking how to alter the config so that a uImage doesn't need building manually 08:31:36 <rajm> but I'm hoping to resolve this today 08:31:44 <szlin> rajm: thanks 08:32:05 <rajm> I'm now using lavacli within b@d for non health check testing and have updated the wiki 08:32:30 <szlin> I saw the slide from Nguyen yesterday 08:32:33 <szlin> https://events.linuxfoundation.org/wp-content/uploads/2017/11/A-Flexible-Test-Automation-System-for-Various-Embedded-Linux-Usecases-Khiem-Nguyen-Renesas-Design-Vietnam.pdf 08:33:05 <rajm> More testing of lava-docker but I'm still getting uboot timeouts - looking into the settings within lava-docker on this 08:33:20 <rajm> thanks szlin I'll take a look 08:33:47 <patersonc> FYI, that is a different Nguyen 08:34:02 <sangorrin> That's a very common name in Vietnam ;) 08:34:09 <rajm> ok 08:34:16 <szlin> oops, I see 08:34:17 <patersonc> But is one who is very experienced with testing - he works heavily on AGL's setup 08:34:47 <patersonc> Renesas update: We have a lava lab instance set up, and now have access to the CIP lava master. We're just waiting on LF to exchange keys and then we can start getting the communication working between master and lab. 08:35:10 <szlin> this slide mentioned B@D and CIP 08:35:56 <patersonc> So it does! 08:36:01 <szlin> patersonc: thanks for update 08:36:07 <sangorrin> Khiem Nguyen? 08:36:18 <szlin> sangorrin: yes 08:36:30 <sangorrin> he is coming to ELCE/ATS 08:37:17 <sangorrin> patersonc: great, I already have admin permissions on https://lava.ciplatform.org/ 08:37:26 <szlin> he mentioned it at page 22 in the slide 08:38:12 <szlin> any other points? 08:38:19 <sangorrin> patersonc: are you using the "polling" setup for the lava slave? 08:38:20 <szlin> 3 08:38:50 <patersonc> sangorrin: I guess so - assuming that means the slave will be polling the master for jobs 08:38:57 <sangorrin> in the past, Lava server connected with a direct ssh connection, which is almost impossible if you have a firewall 08:39:20 <patersonc> As far as I understand it, it's the other way around now. 08:39:24 <sangorrin> yes, that's correct 08:39:27 <patersonc> Good. 08:39:28 <sangorrin> ok, good. 08:40:11 <szlin> 2 08:40:14 <szlin> 1 08:40:17 <szlin> #topic AOB 08:41:34 <szlin> 3 08:41:35 <szlin> 2 08:41:38 <szlin> 1 08:41:40 <szlin> #endmeeting