diff options
Diffstat (limited to 'wiki/Chat_log/20180509-mm-chatlog')
| -rw-r--r-- | wiki/Chat_log/20180509-mm-chatlog | 169 |
1 files changed, 169 insertions, 0 deletions
diff --git a/wiki/Chat_log/20180509-mm-chatlog b/wiki/Chat_log/20180509-mm-chatlog new file mode 100644 index 0000000..a74ce94 --- /dev/null +++ b/wiki/Chat_log/20180509-mm-chatlog @@ -0,0 +1,169 @@ +Multimedia-chat-meeting-2018-05-09 + +<geertu> kbingham: Wake up! +<uli___> kbingham[m]: ^ +<geertu> MultiMedia is in (cryogenic?) stasis... [17:20] +<neg> Maybe he is on some odd none EU timezone ;-) In the meantime shall we do + status reports individually in name order (jmondi, kbingham, morimoto, + neg, uli___ )? +<jmondi> neg: fine with me [17:21] +<jmondi> even if I guess, the important part is collecting it, more than + re-writing here what we reported by email +<jmondi> I can start though +<jmondi> - DRM bridge format support [PATCH 0/8] drm: bridge: Add support for + static image formats HAve to really look into comments, unfortunately + seems like we opened pandora's box of bridges lifetime management. + [17:22] +<jmondi> - MT9T111 media: i2c: mt9t112: Add OF tree support[3~ +<jmondi> - Gen-3 VIN parallel support on D3 [17:23] +<jmondi> capture now sort-of-wroking on v4.16 -> v4.17 has severe issues with + memory corruption while capturing +<jmondi> Until next time) +<jmondi> - Complete VIN parallel support (due date 5/M) [17:24] +<jmondi> - Re-tackle the bridges issues +<jmondi> - Have MT9T111 capture working on peach, so I can integrate it on + Armadillo +<jmondi> - soc_camera removal +<jmondi> Problems I have now) +<jmondi> - Capturing on v4.17 is very fragile +<jmondi> --eot-- +<neg> soc_camera removal \o/ [17:25] +<jmondi> neg: slow down, it's the third time in a row that it gets post-poned + to "until next time" [17:26] +<jmondi> is very low priority +<neg> morimoto: do you have status update for MM? [17:28] +<morimoto> OK +<morimoto> A) What have I done since last time +<marex-cloud> geertu: not bootargs to your scripts, but mtdparts to your + bootargs, which you are clearly passing already ... btw we can + continue the discussion in #perimtd if you want ( dammsan too ) +<morimoto> IRQ +<morimoto> OK, continue [17:29] +<morimoto> - Enjoyed GW ! +<morimoto> - DMAC driver error handling fixup +<morimoto> # Current DMAC driver is sharing error IRQ for all channels, +<morimoto> # but, it will be issue if we uses Virtual Machine, Multi OS, etc. +<morimoto> # We can handle it on each channel's IRQ handler +<morimoto> - Considering new PeriPeri tool (= periupport + peripelist = + "periject") +<morimoto> - Posted ALSA SoC cleanup patches final step +<morimoto> B) What I plan to do till next time +<morimoto> - Consider periject design +<morimoto> C) Problems I have currently +<morimoto> - I want to know each member's opinion about periject. +<morimoto> but on-email is very OK +<morimoto> -- EOF -- +<shimoda> sorry, i have to go home. bye! [17:30] +*** shimoda (~shimoda@relprex3.renesas.com) has quit: Quit: WeeChat 0.4.2 +<morimoto> any question ? +<morimoto> if no, next is neg +<neg> So far I like the name periject :-) +<neg> thanks +<neg> rcar_csi2_write(priv, PHYCNT_REG, PHYCNT_RSTZ); +<neg> rcar_csi2_write(priv, PHTC_REG, PHTC_TESTCLR); [17:31] +<neg> wops +<neg> A) +<neg> - Posted and got merged second batch of VIN enhancement patches. +<neg> - VIN M3-N done. +<neg> - Rework CSI-2 based on datasheet v1.0, new patches pending. +<neg> - Investigated ADV748x behavior of AFE pixel clock. +<neg> B) +<neg> - More CSI-2 work +<neg> - Post next batch of VIN patches that piled up prior to Gen3 merger. + [17:32] +<neg> - Post fix for ADV748x +<neg> - Post prototype for v4l2-subdev video standard control +<neg> C) +* geertu joined #perimtd +<neg> - CSI-2 rework will maybe make it too late for v4.18 :-( +<neg> -- eot -- +<neg> if no questions please go a head uli___ [17:33] +<uli___> A) +<uli___> MM: wikified the instructions for running the IGT test suite +<uli___> (https://elinux.org/R-Car/Tests:igt) +<uli___> MM: rebasing IGT DU fixes, WIP +<uli___> MM: adding DU support to D3, WIP +<uli___> B) [17:34] +<uli___> MM: finish DU on D3 +<uli___> MM: send a new IGT DU fix series +<uli___> C) +<uli___> none +<uli___> correction: C) copy-and-paste doesn't work very well in xfce terminal +<uli___> that's it [17:35] +<neg> thanks, out of curiosity is there any DU work for M3-N in the pipeline? +<uli___> not from me [17:36] +<jmondi> neg: I think Kieran was about to work on that +<neg> jmondi: ok thanks I'm just curious [17:37] +<neg> anything else we need to discuss for MM, I had counted on kbingham to + handle this so I have nothng prepared but have not read all status mails + in detail if there where any open questions +<morimoto> I have nothing [17:39] +<jmondi> neg: I only have the issues with v4.17 capture not working +<uli___> none here +<geertu> neg: [PATCH v2 00/11] r8a77965: M3-N DU Enablement +<jmondi> but actually v4.16 behaves the same, I'm finding out as we speak +<neg> geertu: thanks how could I miss that, I need a beter interface to gmane + archive I keep missing patches [17:40] +<uli___> jmondi: how exactly does it not work? [17:41] +<jmondi> uli___: by 'randomly' hangs the CPU while capture operations is + starting [17:42] +<jmondi> uli___: yesterday Niklas suggested that power cyclcing the board + would help as it may be some v4.17 memory corruption issue + [17:43] +<kbingham> ?? +<kbingham> Have I just walked into a meeting that I didn't know was happening + ? +<jmondi> uli___: and it actually did, this morning I re-based on v4.16, first + boot was ok, but now I still need to power cycle to have capture + working +<neg> jmondi: have you nailed it down to know if s_stream() have been called + or if it's the set_fmt() that is the last v4l2 callback? [17:44] +<morimoto> kbingham: meeting is almost ending stage now :) +<kbingham> morimoto: Oh my ... I'm sorry ! I thought meetings were thursdays! +<jmondi> neg: with strace I see g_fmt being called and that's it +<neg> kbingham: :-) +<jmondi> but I would not trust strace too much to debug these kind of issues +<morimoto> kbingham: :) [17:45] +<neg> jmondi: i think some trace printouts in rcar-vin/rcar-dma.c would be + usefull to know if it's s_stream() or something else cousing the lockup + [17:46] +<jmondi> neg: you see, it is freaking 'random'.. I've been capturing images + for a good 2 hours, rebooted the board, no kernel updates, now it + hangs +<kbingham> neg: M3-N DU should be in renesas-drivers already. [17:47] +<jmondi> neg: I was not correct, I changed the pixel clock polarity between 2 + reboots, I hardly think that can hang the CPU [17:48] +<neg> jmondi: ouch that is bad, I had the impression it always failed on the + second capture attempt by bad +<jmondi> btw, let's tak it offline, I have nothing else for the meeting +<jmondi> neg: no, I made several captures in a row, until I have rebooted the + board [17:49] +<neg> Yes lets keep talking about that later to not hold everyone up :-) +<kbingham> So Shall I post my ABC ... and then that covers the status' :D +<neg> kbingham: please do so if you have it handy :-) [17:50] +<kbingham> Thursday 10th May +<kbingham> +<kbingham> A) What I have done since last time +<kbingham> - M3-N DU Enablement +<kbingham> - Got JTAG working with OpenOCD and Salvator-XS +<kbingham> - TLB-Opimise v7,8,9 +<kbingham> - DU/Interlaced v3, v4 +<kbingham> B) What I will do next +<kbingham> - DMA Virtualisation +<kbingham> C) Any issues I have currently +<kbingham> - None +<kbingham> Ahem ... ignore the first line ... [17:51] +<kbingham> I'll post that to the ML as a late addition +<neg> thanks, I think the last item on the agenda is to pick who will record + the MM log into the format pinchartl usually do and make sure it's sent + out and recorded in the wiki and what else normaly happens with it + [17:52] +<kbingham> neg: I think that was supposed to be me - but I'm not certain I + have all the scrollback ... [17:53] +*** horms (~horms@a80-127-179-77.adsl.xs4all.nl) has quit: Ping timeout: 264 + seconds +<morimoto> neg: I can put it to wiki [17:54] +<morimoto> kbingham: can you create it by using wiki ? +<kbingham> morimoto: Sure :) +<neg> kbingham: I can email you the full scrollback if you need it [17:55] +<kbingham> neg: morimoto: - An e-mail is probably easier thanks |