path: root/wiki/Chat_log/20180322-mm-chatlog

Multimedia-chat-meeting-2018-03-22

10:31 < pinchartl> so welcome to the multimedia meeting
10:31 < neg> I'm ready
10:31 < morimoto> me too
10:32 < pinchartl> I would have started with Jacopo as he hasn't sent his report
10:32 < pinchartl> but I assume that could be related to his e-mail problems
10:32 < pinchartl> so I'll let him report here
10:32 < pinchartl> * Kieran
10:32 < pinchartl> Since last meeting:
10:32 < pinchartl> - Submitted a talk proposal for ALS
10:32 < pinchartl> - 2 versions of DU/Interlaced submitted (and it works now)
10:32 < pinchartl> - vsp1/tlb-optimise/v7 posted. (RB tags welcome ...)
10:32 < pinchartl> - Some back breaking holiday time.
10:32 < pinchartl> - Tried and failed to boot the wheat remotely.
10:32 < pinchartl> Until next meeting:
10:32 < pinchartl> - Test Wheat (D3 I2C conflict work) at last.
10:32 < pinchartl> - Other work tasks to be discussed in Multimedia meeting.
10:32 < pinchartl> Issues and Blockers: None
10:32 < pinchartl> kbingham: any comment ?
10:33  * jmondi back
10:33 < kbingham> I realised I could also add " - Helped Jacopo with Eagle-v3m display testing" to what I have done
10:33 < pinchartl> OK I'll add that
10:33 < pinchartl> * Laurent
10:33 < pinchartl> Since last meeting:
10:33 < pinchartl> - Submitted a talk proposal for ALS
10:33 < pinchartl> - Got DU LVDS rework merged
10:33 < pinchartl> - Skiing holidays
10:33 < pinchartl> - Patch review
10:33 < pinchartl> Until next meeting:
10:33 < pinchartl> - Get the GMSL patches posted to public mailing lists
10:33 < pinchartl> - Harvest BSP for multimedia patches
10:33 < pinchartl> - Upstream pending VSP patches
10:34 < pinchartl> Issues and blockers: None
10:34 < pinchartl> * Magnus:
10:34 < pinchartl> Since last meeting: None
10:34 < pinchartl> Until next meeting: None
10:34 < pinchartl> Issues and blockers: None
10:34 < pinchartl> * Morimoto-san
10:34 < pinchartl> Since last meeting: None
10:34 < pinchartl> Until next meeting: None
10:34 < pinchartl> Issues and Blockers: None
10:34 < pinchartl> dammsan, morimoto: any comment ? :-)
10:34 < morimoto> no comment
10:35 < pinchartl> * Niklas
10:35 < pinchartl> Since last meeting:
10:35 < pinchartl> - [PATCH v11 00/32] rcar-vin: Add Gen3 with media controller
10:35 < pinchartl> - [PATCH v2 0/2] rcar-vin: always run in continues mode
10:35 < pinchartl> - [PATCH v2] i2c: adv748x: afe: fix sparse warning
10:35 < pinchartl> - [PATCH v12 00/33] rcar-vin: Add Gen3 with media controller
10:35 < pinchartl> - [PATCH v2 0/2] rcar-vin: always run in continues mode
10:35 < pinchartl> - Attended ELC, noteworthy discussions I had
10:35 < pinchartl>   - Talked to Gustavo, vb2 fences series seems to reach its final form 
10:35 < pinchartl>     and will likely hit upstream in a not so distant future.
10:35 < pinchartl>   - Talked to Hans, he likes the VIN continues mode and sent a pull 
10:35 < pinchartl>     request during ELC for that work :-) He also liked the current v12 
10:35 < pinchartl>     of VIN Gen3 patches and he have reviewed all patches in the series.
10:35 < pinchartl> - Fixed left-over DT compatibility string for CSI-2 DT which are 
10:35 < pinchartl>   submitted for renesas-drivers as detected by Geert.
10:35 < pinchartl> Until next meeting:
10:35 < pinchartl> - Await Laurents comments on VIN Gen3 v12 and repost hopefully the last  version :-) Laurent do you want me to post v13 before you review to  address Hans small nit-pick?
10:35 < pinchartl> - Try to scope out some MM related task(s) for my base time for Q2.
10:35 < pinchartl> - If there is time or if I have down time fixup CEC register mismatch  for adv on Gen2 (low prio).
10:35 < pinchartl> Issues and blockers:
10:35 < pinchartl> - Kieran reviewed the CSI-2 driver, do anyone else in the MM team review  it before I try to push upstream to review or hopefully pick up the  patch? I think the best way forward is to ask Sakari to have a lock at  it Laurent is over loaded.
10:35 < pinchartl> neg: I won't have time to review the VIN patches before the middle of next week, so if it's not much work you can post a v13
10:35 < pinchartl> or wait for me, as you wish
10:36 < neg> pinchartl: OK then I will hold off for hans pull-requst to hit the media-tree and rebase ontop of the continues mode and repost as soon as thay happens
10:36 < pinchartl> ok
10:36 < pinchartl> regarding the CSI-2 driver, I reviewed an old version, I'd like to review the last one too, but that's not mandatory
10:36 < pinchartl> if you can get a review from Sakari in the meantime it would be nice
10:38 < neg> OK what do you suggest is the best way to get motion for the CSI-2 driver? I don't want to bug Sakari too much as he also seem to have a lot to do these days
10:38 < neg> If you have time to look at in before next meeting that is fine, but it's been sitting on the ML since Feb so I kind of want it to pick up speed :-)
10:38 < pinchartl> maybe you can ask Maxime for a review too, he has posted CSI-2 drivers recently
10:38 < neg> good idea
10:38 < pinchartl> February was yesterday :-)
10:39 < neg> :-)
10:39 < pinchartl> I'll try to review it around the end of next week
10:39  * kbingham would like to borrow pinchartl's time machine :D
10:39 < pinchartl> kbingham: it only works in one direction, all it does it shrink time
10:39 < pinchartl> I don't recommend it
10:39 < pinchartl> * Ulrich
10:39 < pinchartl> Since last meeting:
10:39 < pinchartl> - Sent fixes for Intel IGT test suite for false negatives on R-Car DU
10:39 < pinchartl> Until next meeting:
10:39 < pinchartl> - Check out what seems like legitimate IGT test fails, of which there are a handful.
10:39 < pinchartl> Issues and Blockers: None
10:39 < pinchartl> uli___: any comment ?
10:40 < neg> Thanks, and if you run of time please let me know and I will try to push forward with Sakari :-)
10:40 < pinchartl> OK
10:40 < pinchartl> feel free to ping me in a week
10:40 < pinchartl> I will be travelling again tomorrow and back home next Wednesday
10:40 < uli___> i'll send a summary of test fails to periperi later, for reference
10:41 < pinchartl> uli___: thank you, that will be nice
10:41 < pinchartl> jmondi: your turn
10:41 < jmondi> pinchartl: thanks
10:41 < jmondi> sorry I have not sent report, email's not working
10:41 < neg> pinchartl: ok thanks, will ping you late next week before I ask Sakari for review thanks for the information
10:42 < jmondi> Multimedia) 10 A) Since last time 11 - THC631024LVD 12 -- [PATCH v6 0/3]  drm: Add Thine THC63LVD1024 LVDS decoder bridge 13 -- [PATCH v2 0/4] Renesas CEU: SH7724 ECOVEC + Aptina mt9t112 14  15 B) Next week 16 - Propose and API extensions to propagate LVDS mode through bridges 17  18 C) NULL 19
10:42 < jmondi> ups
10:42 < jmondi> A) Since last time
10:42 < jmondi> - THC631024LVD
10:42 < jmondi> -- [PATCH v6 0/3]  drm: Add Thine THC63LVD1024 LVDS decoder bridge
10:42 < jmondi> - soc camera
10:42 < jmondi>  -- [PATCH v2 0/4] Renesas CEU: SH7724 ECOVEC + Aptina mt9t112
10:42 < jmondi> B) next week
10:43 < jmondi> - Propose and API extensions to propagate LVDS mode through bridges
10:43 < jmondi> - more soc camera removal
10:43 < jmondi> C) Problems
10:43 < jmondi> None
10:43 < jmondi> Hans will send me the Ecovec camera module he found
10:43 < jmondi> --eot
10:43 < jmondi> brb, door ring
10:44 < pinchartl> nice
10:44 < pinchartl> I think that's all for the status updates. who have I forgotten this time ?
10:44 < neg> jmondi:  is the Ecovec camare module listed under C) ? :-)
10:47 < pinchartl> as I don't seem to have forgotten anyone,
10:47 < pinchartl> Topic 2. BSP Team Requests
10:48 < pinchartl> morimoto: I think the VIN question was answered on the mailing list already
10:48 < pinchartl> with continuous mode support being merged upstream for Gen2
10:48 < morimoto> yes, thanks
10:48 < pinchartl> and included in renesas-drivers for Gen3
10:48 < morimoto> sounds nice !
10:48 < pinchartl> - Display DPLL support on ULCB board
10:49 < pinchartl> I've seen your e-mail
10:49 < jmondi> sorry, I may be away for 10 minutes :(
10:49 < pinchartl> I'll reply after the meeting
10:49 < pinchartl> to my knowledge DPLL should work
10:49 < pinchartl> but it might not
10:49 < pinchartl> in which case I'll have to fix it
10:49 < jmondi> people fixing the doorbell in the building...
10:50 < morimoto> ULCB team want to use HDMI, and it seems it needs DPLL
10:50 < morimoto> they solved their issue, but wrong way
10:50 < morimoto> s/wrong way/using wrong way/
10:50 < pinchartl> the DPLL is internal to the SoC, and I'm pretty sure it works on H3 as without it HDMI output wasn't working
10:50 < pinchartl> which ULCB board is the BSP team testing ?
10:50 < morimoto> H3 ULCB
10:51 < morimoto> It was just question mail. I can post more detail if you have time to forcus it
10:51 < pinchartl> I'll have a look and reply to the e-mail
10:52 < morimoto> OK, but again, it is generic question. no stress
10:52 < morimoto> Ohh, they want to use ULCB + HDML + 4K
10:52 < morimoto> s/HDML/HDMI/
10:52 < pinchartl> + 4k might make a difference :-)
10:53  * jmondi back (again, and hopefully untile the end of the meeting)
10:53 < pinchartl> anyway I'll have a look and reply to the e-mail ASAP
10:53 < pinchartl> I will fly tomorrow though, and come back home on Wednesday
10:53 < pinchartl> - DRM plane types
10:53 < pinchartl> The rcar-du driver uses two plane types, with one DRM_PLANE_TYPE_PRIMARY plane per CRTC and all other planes set as DRM_PLANE_TYPE_OVERLAY. The BSP team wants to know if it would be possible to set DRM_PLANE_TYPE_OVERLAY for all planes.
10:53 < pinchartl> morimoto: you also mentioned
10:53 < pinchartl> "The background is that our customer want to clip some part from plane A, and want to indicate it on plane B."
10:53 < pinchartl> I'm not sure to understand that
10:54 < pinchartl> what do you mean by "indicate it on plane B" ?
10:54 < morimoto> they want to print clipped image (= from plane A) on plane B
10:55 < morimoto> print ? indicate ? put ? I don't know which English is OK
10:56 < pinchartl> could you get a detailed description ? what is the order of planes A and B ? what part of the screen do they span ?
10:56 < pinchartl> I'm afraid I don't understand the problem
10:56 < morimoto> OK, will ask to them
10:56 < pinchartl> thank you
10:57 < pinchartl> that should be all for the BSP team requests
10:57 < pinchartl> Topic 3. Additional Tasks for 2018 Q2/1
10:57 < pinchartl> we haven't started discussing those yet
10:58 < pinchartl> and I won't have time before the end of next week
10:59 < pinchartl> dammsan: any comment ?
10:59 < kbingham> Are there priorities from renesas as to the direction we should be working on for ATs ? I.e. - upporting, virtualisation etc ?
10:59 < dammsan> lets chat about that in a little while
11:00 < dammsan> up-porting seems like a hot topic
11:01 < dammsan> and a mixed burger lacking the answer to why in the commit message
11:01 < pinchartl> please all feel free to propose tasks you think would be useful, and we can then discuss them at the end of next week
11:01 < pinchartl> that's all I have for today
11:02 < pinchartl> any comment or question ?
11:02 < jmondi> small one for morimoto-san and damsan
11:02 < neg> For upport work shall we split the work per driver or other method to not step on eachothers toes?
11:03 < jmondi> neg: go on, my question is very simple
11:03 < dammsan> thanks morimoto: what do you think?
11:04 < pinchartl> neg: I think we should coordinate before starting the work
11:05 < pinchartl> I planned to go through the BSP patches next week to see what can be up-ported
11:05 < neg> I plan to start before next meeting to mine BSP for VIN and CSI-2 upport work if no one else is really keen on geting into that ?
11:05 < pinchartl> but feel free to beat me to that
11:05 < pinchartl> I think that, for non-trivial patches, the work should be assigned to the person working on the related driver(s)
11:06 < pinchartl> but I'm sure there will be overlaps here and there, so we need to sync up
11:07 < neg> OK shall we schdule this sync as part of periperi meeting or sync some other way? I'm looking for a way so I don't duplicate work but have no real preference on how the sync is done
11:08 < pinchartl> we can sync up as soon as the BSP analysis is done
11:08 < pinchartl> I won't work on it before next Thursday
11:08 < pinchartl> so we can sync up next Thursday on IRC if you want to have a look already
11:09 < kbingham> What is required to do the BSP analysis ? Just go through each patch and identify /classify ? or is this where Geert's magic script comes in ?
11:10 < pinchartl> kbingham: just going through the patches, yes
11:10 < neg> Sounds like a plan, I will try to get a overview before next week. Next Friday is a public holiday here so I will be offline but sync on Thursday or Monday works for me
11:11 < pinchartl> kbingham: if you have time for that before next Thursday you can coordinate with Niklas
11:11 < kbingham> Ah yes - I almost forgot - we're coming up to easter holidays here too.
11:11 < pinchartl> please avoid classifying the patches twice independently :)
11:11 < neg> pinchartl: :-)
11:11 < kbingham> Ack :) - although perhaps that's a useful verification :D
11:13 < neg> kbingham: I will for a start only look at the VIN and CSI-2 drivers for MM upport and will sync with you if I jump and look at other drivers
11:13 < kbingham> neg: Ack ... perhaps we can use an etherpad or something so we can see eachothers progress in realtime and avoid duplication.
11:14 < pinchartl> kbingham: or just coordinate on IRC ? you might not need to do the work at the same time
11:14 < neg> kbingham: or a clone of periupport so the result is more easly integrated by our mighty group leader
11:15 < kbingham> yup ... however it works ... Haven't seen what's required yet :D
11:15 < kbingham> As long as we sync when we start.
11:17 < pinchartl> jmondi: you also had a question ?
11:17 < jmondi> yup
11:17 < jmondi> very quick and simple
11:17 < jmondi> morimoto: I asked mt9t112 sensor manual, but it's covered by NDA
11:17 < morimoto> yes
11:17 < jmondi> does this NDA extends to Magnus and from Magnus to me?
11:18 < morimoto> Uhm... dificult question. it is too old for us
11:18 < morimoto> and actually
11:18 < morimoto> I don't have this datasheet anymore...
11:18 < jmondi> I don't want to create more paper work for you, or fights with legals for something like this
11:18 < jmondi> ok, problem solved
11:19 < jmondi> morimoto: thanks for looking into it
11:19 < morimoto> sorry I couldn't help you
11:19 < jmondi> no worries!
11:20 < pinchartl> jmondi: is it similar to the MT9T111 ?
11:20 < morimoto> I think these 2 have no big difference
11:20 < dammsan> it's like r8a7795 and r8a7796
11:21 < pinchartl> http://www.uctronics.com/download/MT9T111_MT9T112_DS_A.pdf
11:21 < dammsan> no big difference
11:21 < jmondi> pinchartl: yes, it's the same driver
11:21 < dammsan> =)
11:21 < pinchartl> does that help ?
11:21 < morimoto> Wow ! open datasheet !
11:21 < jmondi> pinchartl: thanks, I have that
11:21 < jmondi> there are no registers description there
11:23 < neg> morimoto: at ELC we talked about my renesas-tests repo for automated testing of som IPs that the BSP team might be interested in and I told you I publish that. I looked at the repo and I need to do some cleaning before that can happen due to some borrowed none GPL code :-) And this might take some time, is this OK for you?
11:23 < pinchartl> jmondi: http://www.ideasonboard.org/jmondi.tar.bz2.gpg
11:23 < pinchartl> does that help ?
11:23 < morimoto> neg: thanks. it is OK
11:24 < jmondi> pinchartl: let me look into that
11:24 < pinchartl> last topic, next meeting
11:24 < pinchartl> two weeks from now ?
11:24 < pinchartl> geertu: ^^
11:24 < pinchartl> wsa_: ^^
11:25 < geertu> pinchartl: ack
11:25 < geertu> Time (DST)?
11:25 < neg> morimoto: the VIN continues mode was merged in media-tree just now :-)
11:26 < jmondi> pinchartl: thanks! it will help for sure!
11:26 < morimoto> neg: sounds good !
11:26 < pinchartl> geertu: 09:00 BST / 10:00 CEST / 11:00 EEST / 17:00 JST
11:26 < pinchartl> (if I'm not mistaken)
11:26 < pinchartl> jmondi: you're welcome
11:27 < jmondi> morimoto: as now I have hardware and documentation, I cannot refuse anymore to be listed as the new v4l2 driver maintainer I guess. As the original driver author is this ok for you?
11:27 < jmondi> morimoto: I should have asked for ov772x as well probably
11:28 < morimoto> do you mean I'm auther for new ov772x/mt9t112 ?
11:28 < neg> pinchartl: Meeting in 2 weeks from now works for me
11:29 < pinchartl> so this is all for the multimedia meeting
11:29 < pinchartl> thank you all for attending
11:29 < jmondi> morimoto: you're author of the original soc_camera based driver, I have moved those driver and made a proper v4l2 subdev driver out of them.