02:08:51 * ** mturquette_ has joined #cip 02:12:52 * ** mturquette has quit IRC 02:12:56 * ** mturquette_ is now known as mturquette 06:08:11 * ** patersonc has joined #cip 06:33:54 * ** weshuang has quit IRC 07:14:48 * ** rajm has joined #cip 07:31:22 * ** moxavict1r has joined #cip 07:31:59 * rajm runs a BBB healthcheck with email to the list 07:33:15 * ** moxavict1r has left #cip 07:35:45 <rajm> and a successful renesas board HC with another list email 07:39:29 * ** weshuang has joined #cip 08:09:22 * ** toscalix has joined #cip 08:11:07 * ** moxavict1r has joined #cip 08:11:28 * ** moxavict1r has left #cip 08:12:54 * ** moxavict2r has joined #cip 08:15:40 * ** moxavict2r has left #cip 08:29:30 * ** moxavict1r has joined #cip 08:31:19 * ** moxavict1r has left #cip 08:42:04 <toscalix> szlin: any point on the agenda for today's meeting? 08:43:09 * ** moxavict1r has joined #cip 08:48:51 <szlin> toscalix: which version of kernel should we start to review? Based on ben's email yesterday, it should be 4.4.129 08:50:37 <toscalix> yep 08:56:19 * ** moxavict1r has left #cip 08:59:04 <szlin> I don't have question before jumping into review work. 08:59:51 <szlin> BTW, I just sent kernel patch via stable process - https://patchwork.kernel.org/patch/10364735/ 09:00:59 <patersonc> Mornin all 09:01:10 <szlin> patersonc: hi 09:01:47 <patersonc> szlin: Potential agenda item: Kernel maintenance testing requirements 09:01:58 <szlin> I see. 09:02:21 <szlin> let's start the meeting 09:02:24 <szlin> #startmeeting 09:02:31 <szlin> #topic roll call 09:02:41 <szlin> please say hi if you are in the meeting 09:02:44 <szlin> ._./ 09:02:50 <rajm> hi 09:02:51 <bwh> hi 09:02:52 <toscalix> o/ 09:02:55 <toscalix> hi 09:03:20 <weshuang> hi 09:03:25 <patersonc> hi 09:03:33 <szlin> #topic 1. Kernel maintenance testing requirements 09:04:00 <szlin> patersonc: Any thoughts? 09:04:22 <patersonc> I'm asking wrt infrastructure. KernelCI/Jenkins etc. Is any required? Or will everyone be doing everything locally? 09:05:06 <patersonc> At the moment CIP are setting up KernelCI and LAVA Labs for remote testing, but no Jenkins instance. I want to know if you see a need for a centralised build system. 09:05:17 <patersonc> Would this help with maintenance? 09:05:43 <patersonc> New commits can trigger builds/tests automatically etc. 09:06:11 <bwh> I thought KernelCI had some provision for that already 09:06:24 <toscalix> it does 09:06:25 <bwh> Maybe not 09:06:39 <toscalix> but simple 09:06:57 <toscalix> since it was designed for testing the kernel only initially 09:07:55 <szlin> Daniel Sangorrin has mentioned that he is going to use Fuego for testing 09:08:07 <toscalix> linaro is already using kernelci to test beyond kernel, like agl 09:08:32 <toscalix> and agl has plans to make fuego and lava compatible 09:08:33 <szlin> I am not sure that if Fuego is overkill 09:09:12 <toscalix> there is a big gap between where we are and needing anything beyond kernelci 09:09:39 <toscalix> I mean, we need to set up our labs, get familiar with the lava set up, create/import tests 09:10:02 <toscalix> probably adapt the kernelci interface to our use case a little... 09:10:07 <patersonc> Okay 09:10:19 <patersonc> I wasn't sure what KernelCI could do 09:10:41 <szlin> toscalix: that's what Moxa doing...we are building the testlab with CI/ CD/ LT 09:11:06 <toscalix> szlin: based on kernelci? 09:11:15 <toscalix> using openQA? 09:11:59 <szlin> based on jenkins and LAVA. But yes, openQA could be a option 09:12:08 <toscalix> ok 09:12:39 <toscalix> so you will be able to send results to the infra that renesas is setting up, cool 09:14:40 <szlin> toscalix: Have you ever used openQA? 09:15:08 <toscalix> szlin: yes. I was the manager of the team who developed initially, a few years back 09:15:31 <szlin> cool. 09:15:54 <toscalix> but the tool has come a long way since then 09:16:03 <szlin> how come? 09:16:44 <toscalix> now it has become a product almost 09:16:53 <toscalix> Lots of effort on it 09:17:18 * rajm has used it but very much as an initial evaluation, was beginning to scratch surface 09:18:35 <szlin> I see. 09:18:37 <rajm> very dep on using particular distros but that may not be a problem 09:19:09 <toscalix> rajm: fedora, opensuse.... more and more distros using it 09:19:37 <rajm> yes those 2 but maybe things are opeing out more 09:20:12 <rajm> *opening 09:20:37 <szlin> AFAIK, Debian is not using openqa. 09:20:49 <szlin> bwh: please correct me if I am wrong. 09:21:22 <bwh> Well there seems to be some unofficial attempt at using it: http://openqa.debian.net/ 09:21:33 <szlin> https://debconf17.debconf.org/talks/69/ 09:21:43 * ** vidda_ has joined #cip 09:21:48 <szlin> Adam Majer had a session with openqa in debconf17 09:22:19 <bwh> Right, I wasn't there 09:22:58 <szlin> Back to the topic, any other discussion? 09:23:02 <toscalix> In my opinion, there will be the default test framework for any complez linux system. The cost/benefit is insane. But I am bias since I feel part of the baby so... 09:23:11 <toscalix> yes, offtopic 09:23:22 <patersonc> Thanks for the info. It sounds like there is no big push for a CIP Jenkins server at this time. I'll look more into what KernelCI can do. 09:23:40 <toscalix> patersonc: if there is need, let's discuss it, yes 09:23:51 <toscalix> I have a question, 09:24:25 <patersonc> Sure 09:24:26 <toscalix> bwh: sent the patches for review. Is there anybody who is interested in review patches that cannot access to them or has problems with them? 09:24:39 <szlin> patersonc: Since Daniel Sangorrin is using Fuego for testing, Fuero is a Jenkins-based test framework 09:26:15 <patersonc> szlin: Okay. I guess it's a question of whether CIP wants to start centalizing Fuego etc. or not. 09:26:32 <patersonc> toscalix: AGL uses Fuego right? 09:26:39 <patersonc> Or just LTSI? 09:26:40 <toscalix> not yet 09:27:00 <toscalix> fuego compatibility with lava is not there yet 09:27:15 <szlin> toscalix: I am reviewing the patch via linux-stable.git 09:27:22 <patersonc> Ah okay 09:27:39 <bwh> szlin: OK, but you won't be able to send replies without the mail. 09:27:40 <toscalix> szlin: cool 09:28:29 * ** HarryYJ_Jhou has joined #cip 09:28:44 <szlin> bwh: how about 1. cherry-pick 2. sign-off the patch 3. send the mail via git send-email if there is no issue 09:29:18 <bwh> Sorry, why would you be sending a patch? 09:30:12 <szlin> bwh: shouldn't we add "Signed-off-by" after reviewing the patch? 09:30:46 <bwh> No, since the patches are already getting applied to a git branch they can be merged into CIP's branch 09:30:48 <szlin> or just send out the mail and mention that this patch has be done 09:31:14 <szlin> I see 09:31:32 <patersonc> bwh: So there's no addition of reviewed-by tags either? 09:32:19 <bwh> patersonc: Not usually but Greg would probably be willing to add them if that is sent before the end of the review cycle 09:32:32 <patersonc> bwh: okay 09:32:35 <szlin> I need to go since the meeting is overtime 09:32:48 <bwh> szlin: So I suggest you make a note of the patches you reviewed and didn't see any issue with, and then send a list of those to cip-dev 09:33:05 <bwh> plus separate mails about any problematic patches 09:33:22 <szlin> bwh: ok, thanks. 09:33:30 <szlin> any questions? 09:33:44 <szlin> #endmeeting