path: root/wiki/Chat_log/20171005-mm-chatlog

Multimedia-chat-meeting-2017-10-05

09:04 < pinchartl> neg: you can start
09:05 < neg> A)
09:05 < neg>     - Broke my 8 channel camera board by trying to measure the GMSL link with a scope.
09:05 < neg>     - Repaired half my 8 channel camera board... I have one MAX9286
09:05 < neg>       working and have confirmed all 8 cameras working.
09:05 < neg>     - Discovered that 3 out of 8 cameras needs auto ack on i2c from
09:05 < neg>       MAX9286 to function. Reason for this unknown.
09:05 < neg>     - Found issue with Sakari's subnotifier patchset for rcar-vin. Fix
09:05 < neg>       is posted and seems like it is picked up.
09:05 < neg>     - Got Acks from Hans on most VIN Gen3 patches, most of them are now
09:05 < neg>       OK from Hans point of view.
09:05 < neg> B)
09:05 < neg>     - Rebase VIN Gen3 patches and repost with Hans Acked-by.
09:05 < neg>     - Try to rescue 8 camera board.
09:05 < neg>     - If Laurents Kingfisher price information is OK from Renesas side
09:05 < neg>       try to order one including camera boards.
09:05 < neg>     - Rebase multiplexed pads on Sakari's patch series.
09:06 < neg> C)
09:06 < neg>     - Lost a lot of working time on breaking and trying to restore the 8 camera board.
09:06 < neg> EOT
09:06 < pinchartl> thank you
09:06 < pinchartl> do you think it's worth it trying to rescue the second MAX9286 ?
09:06 < pinchartl> it's likely that the chip itself is dead
09:07 < neg> I think it's dead
09:07 < pinchartl> I wouldn't spend more time on that
09:07 < neg> So my focus now is to make sure I don't damage anything else while using "half" the board
09:08 < pinchartl> :-)
09:08 < neg> Yes, I have no plan to try and rescue the second MAX9286 :-)
09:08 < pinchartl> ok, thanks
09:08 < pinchartl> next, uli__
09:09 < pinchartl> who should actually have gone first as I don't see any report through e-mail
09:09 < uli__> then you're not looking hard enough :)
09:09  * kbingham notices Uli got his report in before me :)
09:09 < uli__> anyway
09:10 < uli__> so i managed to fix the display on the Chromebook in mainline
09:10 < pinchartl> sorry, I might be having issues with my mail client :-(
09:10 < uli__> np
09:10 < uli__> panel didn't have regulators because of the weirdo device tree that deleted the correct description and replaced it with a bad one
09:11 < uli__> a lot of bisecting and two small hacks later it worked
09:11 < uli__> oh, and framebuffer console support was also missing
09:12 < uli__> (didn't know that was actually required...)
09:12 < pinchartl> :-)
09:12 < uli__> so that's really it. the only major hack remaining is the mmsys device that needs to be both a clock controller and a drm device
09:13 < uli__> nobody commented on that so far
09:13 < pinchartl> ok
09:14 < pinchartl> plans for the next two weeks ?
09:14 < uli__> fix some non-essential devices so i can use this thing myself :)
09:15 < uli__> that's as far as plans go for now
09:15 < pinchartl> I assume you'll start testing the GPU on mainline too ?
09:15 < pinchartl> (I found your e-mail now)
09:15 < uli__> that would be part of an additional task, right?
09:16 < pinchartl> yes
09:16 < pinchartl> they will come soon
09:16 < uli__> ok
09:16 < pinchartl> thank you
09:16 < pinchartl> next, kbingham 
09:16 < wsa_> what WLAN device is used on that chromebook?
09:16 < uli__> moment, i'll take a look
09:17 < kbingham> Most of my time has been looking at 8camera probing and such.
09:18 < kbingham> I started by helping Jacopo investigate his FSYNC task - and also looking at the breakout camera.
09:18 < kbingham> Initially the breakout camera wasn't capturing anything ... and after trying various things we managed to capture from it.
09:18  * kbingham regrets not doing more testing on the camera at that stage as we then assumed it was fine.
09:19 < kbingham> So then we tried ... and I failed ... to capture anything from the I2C on the other side of the MAX9271 ...
09:20 < kbingham> Stashing that task to see if there were other issues - I investigated the 'stabilising probe' task - and that's when I discovered the alternate MAX9286 was auto-acking for devices on the other MAX.
09:21 < kbingham> Fixing that doesn't stop my cameras from probing (they already did) but does seem to hinder 3 of niklas'
09:21 < kbingham> So more investigation required there anyway.
09:21 < kbingham> I've also prepared a tool to capture the profile and status information from the RDACM20 cameras and parse it.
09:21 < kbingham> Other than a lot of sensor configuration it shows very little configuration to the MAX9271
09:22 < kbingham> So ... next up ... Well - hopefully after this meeting we'll get together to discuss all of the current state and hte best next moves forwards.
09:22 < pinchartl> have you compared the profiles of all 8 cameras ?
09:22 < kbingham> I've compared 3.
09:22 < kbingham> They were the same.
09:23 < kbingham> I can do the full set if desired.
09:23 < kbingham> In fact it won't take long to do - so I'll do it today.
09:23 < pinchartl> it would be nice just to make sure
09:23 < pinchartl> neg: could you do the same with your cameras ?
09:24 < neg> yes
09:24 < pinchartl> thank you
09:25 < kbingham> Quickly finishing my status ... . Work next will be - discussed in todays meeting, as well as potetnitally looking at the DU/VIN loopback task before the next time I have a status update too :)
09:25 < kbingham> And my issues :
09:25 < kbingham> I need to check if I can publish code that parses the PROFILE of the cameras in case it conflicts with the NDA.
09:25 < kbingham> (I'll do that separately)
09:25 < kbingham> and I think I have broken the breakout camera.
09:26 < kbingham> Except ... I might not have ...
09:26 < kbingham> It might just not be responding - in exactly the same way that Niklas' three cameras require auto-ack ... (which I have disabled)
09:26 < kbingham> EOT from me.
09:26 < pinchartl> it would be nice if you could confirm that
09:27 < pinchartl> we have too much broken hardware already :-)
09:27 < pinchartl> thank you
09:27 < pinchartl> next, jmondi 
09:27 < jmondi> here I go
09:27 < uli__> wsa_: marvell 8897
09:27 < jmondi> A) -- fsync measurement with Kieran and his scope
09:28 < jmondi> --discussions and reviews on i2c stability
09:28 < wsa_> uli__: nice! that has upstream support :D
09:28 < jmondi> -- base respository for gmsl work
09:28 < jmondi> ceu:
09:28 < jmondi> -- resumed CEU development on GR-Peach
09:29 < jmondi> -- add OF parsing to already submitted ceu driver
09:29 < jmondi> -- some soldering and testing to add support for ov7670 camera board on gr-peach audiocamera shield
09:29 < jmondi> B)
09:29 < jmondi> -- capture on gr-peach
09:30 < jmondi> -- resume fsync investigation once i2c issues are mitigated
09:30 < jmondi> C)
09:30 < jmondi> -- could we sync how hw which we have to buy or will be shipped to us?
09:30 < jmondi> -- EOT
09:31 < pinchartl> thank you
09:31 < pinchartl> regarding hardware
09:31 < pinchartl> we need to buy the kingfisher camera setup ourselves
09:31 < jmondi> s/how/which
09:31 < pinchartl> that's the ULCB, the KF base board, the KF camera expansion board, cameras and cables
09:32 < pinchartl> I've started the process, ULCB is in back order, asked for a quote for KF, got quote for RDACM21 cameras and will buy cables
09:32 < pinchartl> at this point you don't need to do anything
09:32 < pinchartl> let's wait for the KF quote, and then we'll decide what to do, and who should buy hardware
09:32 < jmondi> pinchartl:  thanks
09:33 < pinchartl> you're welcome
09:33 < pinchartl> next, Morimoto-san, who reported by e-mail
09:33 -!- Irssi: Pasting 7 lines to #periperi. Press Ctrl-K if you wish to do this or Ctrl-C to cancel.
09:33 < pinchartl> Since last meeting:
09:33 < pinchartl> - Customer requested Sound new feature (= MIX volume), I posted this support to ML, under review now.
09:33 < pinchartl> - Posted H3 ES2.0 / D3 audio PFC patch, under review now.
09:33 < pinchartl> - Re-posted ALSA SoC cleanup patch
09:33 < pinchartl> - V3M/D3 shipping paper work
09:33 < pinchartl> - Pinged MAXIM datasheet question, no response
09:33 < pinchartl> For the next two weeks:
09:33 < pinchartl> - Continue V3M/D3 paper work
09:33 < pinchartl> - Ping MAXIM datasheet question
09:33 < pinchartl> Issues and Blockers:
09:33 < pinchartl> - No response from ALSA maintainer, MAXIM...
09:34 < pinchartl> I assume s/MAXIM/Omnivision/
09:35 < pinchartl> next, me
09:35 -!- Irssi: Pasting 5 lines to #periperi. Press Ctrl-K if you wish to do this or Ctrl-C to cancel.
09:35 < pinchartl> Since last meeting:
09:35 < pinchartl> - Completed the FCP near-lossless compression prototype
09:35 < pinchartl> - Assisted with the GMSL camera development
09:35 < pinchartl> - Embedded Recipes and Kernel Recipes conferences (September 26-29 in Paris)
09:35 -!- Irssi: Pasting 6 lines to #periperi. Press Ctrl-K if you wish to do this or Ctrl-C to cancel.
09:35 < pinchartl> For the next two weeks:
09:35 < pinchartl> - Finish the additional tasks descriptions for Q4
09:35 < pinchartl> - More patch review
09:35 < pinchartl> - V4L2 lifetime management issues (for VIN Gen3 support)
09:35 < pinchartl> - Display color keying support for Gen3
09:35 < pinchartl> Issues and Blockers: None
09:36 < pinchartl> next, dammsan 
09:37 < pinchartl> silence...
09:37 < pinchartl> I assume I can copy & paste last week's report
09:38 < pinchartl> Since last meeting: None
09:38 < pinchartl> For the next two weeks: None
09:38 < pinchartl> Issues and Blockers: None
09:38 < pinchartl> :-)
09:38 < pinchartl> that's it for status updates
09:38 < pinchartl> as everybody is here, let's discuss the topic of the next meeting
09:38 < pinchartl> I propose not skipping the one two weeks from now, just before Prague
09:39 < pinchartl> there's no plan for a formal meeting in Prague at this time, even if most of us will be there
09:39 < pinchartl> so I'd rather have a status update meeting the week before, if only to avoid wasting time on it when we'll meet
09:39 < pinchartl> what do geertu and wsa_ think ?
09:39 < geertu> Fine for me
09:40 < geertu> wsa_: ?
09:41 < pinchartl> wsa_ seems to be AFK, I'll let him chime in if he thinks it's a bad idea
09:41 < wsa_> i m here
09:41 < pinchartl> welcome :-)
09:41 < wsa_> i agree with laurent
09:41 < pinchartl> thank you
09:41 < pinchartl> (uli__: I've found your e-mail, it was an issue on my side, sorry about that)
09:41 < uli__> again, np :)
09:42 < pinchartl> last topic for multimedia
09:42 < pinchartl> Niklas, Kieran and Jacopo would like to sync up on the 8 cameras work today
09:42 < pinchartl> what would be a good time for you all ?
09:42 < neg> I'm free all day
09:43 < pinchartl> I'll have to skip the core and I/O meeting I'm afraid, and will be back at 11:45 CEST. should we start then, or after lunch ?
09:43 < jmondi> I have to be away from ~1hr before lunch, then I'll be here
09:43 < jmondi> so I would prefer after lunch if that's fine with everyone
09:44 < pinchartl> kbingham: how about you ?
09:44 < kbingham> pinchartl: For me the sooner the better.
09:44 < kbingham> pinchartl: I'll make myself avaialble as soon as everyone else is.
09:44 < kbingham> (But I'm available all day)
09:45 < pinchartl> jmondi: when can you start ?
09:45 < jmondi> sorry for not being here in the morning, but this is the one time in years I have to show up in my bank branch to update the security device
09:45 < jmondi> after lunch, I plan to be back home not later than 1pm
09:46 < pinchartl> kbingham: 12:00 for you, is that OK ?
09:46 < kbingham> So 12pm GMT :)
09:46 < kbingham> pinchartl: Ack.
09:46 < kbingham> Wait - not GMT - BST :)
09:46 < pinchartl> :-)
09:46 < pinchartl> OK, let's do that then
09:46 < kbingham> Stupid timezones
09:47 < pinchartl> 12:00 BST / 13:00 CEST / 14:00 EEST
09:47 < pinchartl> that's it for the multimedia meeting
09:47 < pinchartl> thank you all for attending