path: root/wiki/Chat_log/20191010-mm-chatlog

Multimedia-chat-meeting-2019-10-10

10:18 < pinchartl> welcome to the multimedia meeting
10:18 < pinchartl> Topic 1. Status Check for the Multimedia Tasks
10:18 < pinchartl> * Jacopo
10:18 < pinchartl> Since last meeting:
10:18 < pinchartl> - Restarted work on CMM
10:18 < pinchartl> Ezequiel (Collabora) is developing a very similar feature for Rockchip,
10:18 < pinchartl> and Sean's review on his patches will be taken into account for CMM.
10:18 < pinchartl> Until next meeting: N/A
10:18 < pinchartl> Issues and blockers: None
10:18 < pinchartl> jmondi: any comment ?
10:19 < jmondi> pinchartl: yes, quick one
10:19 < jmondi> Ezequiel has moved all their gamma LUT handing to atomic_enable/begin
10:20 < jmondi> Sean suggested that, and that removes the need to restore the color_management_changed flag after suspend as we do now
10:20 < jmondi> I plan to do the same for CMM, but before going down that path I would like your opinion there
10:20 < pinchartl> that seems wrong, the gamma LUT should be modifyable at runtime
10:20 < jmondi> [PATCH v4 0/3] RK3288 Gamma LUT
10:20 < pinchartl> why did they do that ?
10:21 < jmondi> I'm not sure I did completely understood Sean's reply to v3
10:21 < jmondi> I asked eze too yesterday, but then I had to run
10:21 < pinchartl> could you check it with them ? please raise the issue of modifying the gamma table at runtime
10:22 < jmondi> also, they moved everything to crct_enable/begin, while we had that in the commit_tail helper
10:22 < jmondi> so the CMM would be set per-CRTC ?
10:22 < pinchartl> to determine if it has to be supported (and if not, why, and where that is documented)
10:22 < jmondi> I will keep pressing Ezequiel on that
10:22 < jmondi> it's fun how many corner cases we're hitting with CMM
10:22 < jmondi> well, fun..
10:23 < pinchartl> it doesn't matter if it's set in the atomic tail handler or the CRTC operations, that's not relevant
10:23 < pinchartl> what is relevant is if modifying the gamma LUT without a modeset needs to be supported
10:23 < jmondi> ah well, no
10:23 < jmondi> we do
10:23 < jmondi> for_each_old_crtc_in_state
10:24 < jmondi> so we go through each crtc anyhow
10:24 < pinchartl> sure, we have to as gamma tables are per CRTC
10:24 < jmondi> indeed, my bad, I got confused
10:25 < jmondi> I'll try to know more
10:25 < pinchartl> so please check with them before changing anything, and please let me know
10:25 < pinchartl> thank you
10:25 < jmondi> you'll be cc-ed
10:26 < pinchartl> ok. I don't plan to participate in the discussion though, please make sure to drive it
10:26 < pinchartl> next, Kieran, who is AWOL due to having a new baby
10:27 < pinchartl> (what wouldn't people do to have a few weeks of off ? :-))
10:27 < pinchartl> * Laurent
10:27 < pinchartl> Since last meeting:
10:27 < pinchartl> - CMM & other patches review
10:27 < pinchartl> - Attended XDC (X.Org Developer's Conference)
10:27 < pinchartl> Discussed, among other topics, a new kernel API for cameras,
10:27 < pinchartl> interoperability between displays and cameras, fences for V4L2, buffer
10:27 < pinchartl> allocator API with constraint resolution, ...
10:27 < pinchartl> Until next meeting:
10:27 < pinchartl> - CMM review to finish the upstreaming
10:27 < pinchartl> - DU group refactoring
10:27 < pinchartl> - Attending ELC-E (with V4L2 workshop)
10:27 < pinchartl> Issues and Blockers: None
10:27 < pinchartl> any question ?
10:27 < pinchartl> * Morimoto-san
10:27 < pinchartl> Since last meeting:
10:27 < pinchartl> - ALSA SoC framework cleanup
10:27 < pinchartl> Posted patches have been accepted, posted next set.
10:27 < pinchartl> - ALSA SoC small bug fixes
10:27 < pinchartl> Until next meeting:
10:27 < pinchartl> - Continue progressing cleanup patches
10:27 < pinchartl> Issues and Blockers:
10:27 < pinchartl> - Can't test Intel-related changes
10:27 < pinchartl> Some of the cleanup patches are related to an Intel-specific features.
10:27 < pinchartl> The code is compile-tested only due to lack of a test platform.
10:27 < pinchartl> - Large number of cleanup patches to upstream
10:27 < pinchartl> ALSA SoC is too complex today, hence the need for a large cleanup, with
10:27 < pinchartl> more than 500 local patches still pending. The next step will focus on
10:27 < pinchartl> soc-pcm that supports flexible CPU/codec connections, and this area is
10:27 < pinchartl> too complex.
10:27 < pinchartl> morimoto: any comment ?
10:28 < morimoto> Intel said that they are using Github to test hardware
10:28 < morimoto> He can test it
10:29 < pinchartl> ? how does that work ?
10:29 < morimoto> I'm not sure detail, but if he push patches to Github,
10:29 < morimoto> test script (?) works automatically
10:30 < morimoto> Some magical machine is working
10:30 < pinchartl> I would have assumed them to use gitlab for self-hosted CI integration
10:30 < morimoto> And it seems my patches were OK
10:30 < pinchartl> that's good news :-)
10:30 < pinchartl> so it's not an issue anymore ?
10:30 < morimoto> yes :)
10:30 < morimoto> Yeah, thanks
10:31 < pinchartl> thank you
10:32  * pinchartl wonders if ALSA SoC cleanup is Morimoto-san's secret plan to boost his kernel commits statistics ;-)
10:32 < pinchartl> * Niklas
10:32 < pinchartl> Since last meeting:
10:32 < pinchartl> - [PATCH 0/2] rcar-vin: Cleanup how subdevice format is handled
10:32 < pinchartl> - [PATCH v2 0/2] rcar-vin: Support V4L2_FIELD_SEQ_{TB,BT}
10:32 < pinchartl> - Attended LPC
10:32 < pinchartl> Until next meeting:
10:32 < pinchartl> - Attend ELC-E and join the V4L2 workshop
10:32 < pinchartl> - Keep pushing VIN patches unlocked by recent acceptance of cleanup patches
10:32 < pinchartl> Issues and blockers: None
10:32 < pinchartl> neg: any comment ?
10:32 < neg> No comment, thanks
10:33 < morimoto> pinchartl: not *secret* but *official* plan :)
10:33 < pinchartl> :-)
10:33 < pinchartl> * Ulrich
10:33 < pinchartl> Since last meeting:
10:33 < pinchartl> - Patch review
10:33 < pinchartl> Until next meeting: N/A
10:33 < pinchartl> Issues and Blockers: None
10:33 < pinchartl> uli_: any comment ?
10:33 < uli_> nope
10:34 < pinchartl> thank you
10:34 < pinchartl> any discussion topic ?
10:35 < neg> Anything to plan/bring to ELC-E ?
10:35 < wsa> I could bring my Gen2 Alt board for Kieran if he still needs that?
10:36 < neg> I'm thinking if we need some face-to-face talks or exchange of hardware. I take it we will not have a periperi meeting day ?
10:36 < pinchartl> wsa: could you check with him ?
10:36 < pinchartl> kbingham[m]: ^^
10:36 < geertu> wsa: kbhingham won't be there
10:36 < pinchartl> geertu: of course. stupid me
10:36 < wsa> right
10:36 < pinchartl> Kieran can't get the hardware if doesn't attend
10:36 < geertu> and the goods may receive increased scrutiny at border control
10:36 < pinchartl> neg: that was my next question. when will the next meeting be ?
10:37 < pinchartl> 3 weeks from now is October the 31st
10:37 < pinchartl> it will collide with the gstreamer conference for Jacopo and me
10:37 < wsa> that's elce time
10:37 < pinchartl> and I believe Niklas won't be back home yet either
10:38 < neg> I will go back home on the 3rd of Nov
10:38 < pinchartl> should we push it back by one week ?
10:38 < wsa> yup
10:38 < pinchartl> so November 7th, same time as usual ?
10:38 < wsa> works for me
10:38 < pinchartl> we will have switched to winter time
10:39 < geertu> pinchartl: Will we? ;)
10:39 < pinchartl> won't we ?
10:39 < morimoto> not in Japan :)
10:39 < geertu> Seems like the killing of DST has been moved to the background noise...
10:40 < pinchartl> geertu: I think we would have been notified if there would be no chance this autumn
10:41 < geertu> pinchartl: Right. So ELC-E will be held in Winter Time
10:42 < geertu> I hope Thalys knows ;-)
10:42 < pinchartl> so the meeting will start at 17:00 JST, right ?
10:42 < geertu> I think so
10:43 < pinchartl> ok
10:43 < pinchartl> that's all for the multimedia meeting
10:43 < pinchartl> I propose adjourning, does anyone second ?
10:43 < neg> second
10:44 < pinchartl> meeting adjourned. thank you all for attending