Multimedia-chat-meeting-2016-03-22 hello [16:55] Hi morning let's give Ulrich and Magnus a few more minutes to join [16:57] or maybe they're here already ? dammsan: hello uli___: hello I couldn't get any response from Magnus during today [16:58] Maybe he is super busy let me try to ping him [16:59] hi [17:00] sorry for being late hello ! no worries i have more details for the questions let's get started [17:01] topics for today - tasks status update - VSP questions from Magnus - next meeting / ELC anything else ? nope, not from me ok [17:02] I sent request mail morimoto: about vsp suspend/resume ? Yes, and I would like to 100% for v4.5 on peripelist :) good point [17:03] Ahh, and from BSP - request API (delayed) 4. How will the 50% contract affect development - plane-alpha, pixel-alp added to the agenda [17:05] let's get started then thanks let's try to be quick on the tasks status update I'll try to handle it per developer this time [17:06] ADV7482,v4.7,plan,niklas,Prototype on Gen3 ADV7482,v4.8,plan,niklas,Gen3 support upstream ADV7482,v4.8,plan,niklas,Interlace support upstream VIN+CSI2,v4.7,noplan,niklas,Prototype on Gen3 VIN+CSI2,v4.8,noplan,niklas,Gen3 support upstream VIN+CSI2,v4.8,noplan,niklas,Interlace support upstream VIN,v4.8,noplan,niklas,Gen3 support upstream (without CSI-2) neg: what's your status ? ? noplan ? [17:07] Bit of downtime since last meeting waiting for new contract. Working on VIN+adv7180 hope to be done with VIN for gen2 before ELC morimoto: I think we have a plan, that's a mistake [17:08] OK thanks neg: what's left to be done for VIN+ADV7180 ? minior shuffeling of code between files per review comments and extra check for NV16 alignment, nothing big [17:09] so you think the next version will be merged ? [17:10] yes ok I'll add a task for that in the list thanks VIN,v4.7,public,niklas,New VIN driver without soc-camera (tested on Gen2) [17:11] But I was wondering if we should try to get the false gen3 support out of soc_camera we can, it's just a simple patch [17:12] does the above schedule still holds for you ? [17:13] depending on if I can get the same ammount of time for multimedia as before, yes [17:14] neg: you will, but the time may come in boosts lets discuss more f2f at ELC [17:15] let's discuss that as part of topic 3 then we have - DU+HDMI,v4.7,plan,ulrich,HDMI output on Gen3 prototype - DU+HDMI,v4.8,plan,ulrich,HDMI output on Gen3 upstream - DU+HDMI,v4.7,prototype,ulrich,Test setup with HDMI output to HDMI input loopback (without EDID) - DU+HDMI,v4.7,public,ulrich,EDID generation support for the HDMI loopback test setup - VIN,v4.7,public,ulrich,Add DV timings support to rcar-vin - VIN,v4.7,public,ulrich,Upstream Lager HDMI input bug fixes Ulrich isn't here to update us on the status also I'll ask him to do so by e-mail in a reply to the meeting report my feeling is that we may want a HDMI video out prototype for Gen3 earlier [17:16] if possible earlier than v4.7 ? well, the earlier the better =) since it is a prototype it does not need to be aligned to upstream schedule perhaps something for ulrich to focus on sometime soon if possible [17:17] ok I'll write it down othanks!! then [17:18] HDMI+sound+output,2016-03-31,plan,morimoto,Prototype on Gen3 HDMI+sound+output,2016-03-31,plan,morimoto,Upstream support without hotplug on Gen2 HDMI+sound+output,2016-06-30,plan,morimoto,Hotplug support upstream RSND,2016-06-30,plan,morimoto,Hotplug support upstream on Gen3 morimoto: that's for you I have no update for these I think I need +3 month for these ? [17:19] should we try to align that with upstream kernel versions then ? instead of dates I don't care details :) [17:20] we should still set milestones [17:21] I means version vs date just +3 months for all four tasks then ? [17:22] Because these are based on HDMI+output [17:23] and not yet supported ? correct ok, I've recorded that [17:24] next - VIN,v4.7,plan,magnus,Investigate if/how parallel VIN can be used on Salvator-X i did figure that out it is possible but we decided to go with it as backup in case CSI gets too hairy I'll remove the task then as investigation is done [17:25] good thanks DU+VSPD,2016-03-31,plan,laurent,Z-order support prototype [17:26] DU+VSPD,v4.7,plan,laurent,Z-order support upstream VSP,v4.8,public,laurent,V4L2 request API upstream VSP,v4.7,public,laurent,CLU/LUT support submitted upstream on Gen3 VSP,v4.7,prototype,laurent,HGO support upstream on Gen3 the plan for z-order support still holds, I'll try to complete it this week [17:27] regarding the request API, I'll try to post the next version before ELC [17:28] as I'll present it in my ELC talk :-) thanks! do you have Display-List plan ? [17:29] display list support is upstream [17:30] already do you mean ? yes thanks ! it has been merged in Linus' tree there are still enhancements that will be merged in v4.7 I couldn't get everything upstreamed in one go there shouldn't be any delay, the code is ready how about request API status? [17:31] is it ready to go, or shall we expect more updates? as you probably know I've had a bit of a conflictual relationship lately with the V4L2 maintainer [17:32] it resulted in delays sure, that's fine Mauro said he wouldn't merge (and didn't even review) any media controller patch before his rework could be merged internal renesas guys needed to make some workaround to use the request API in your latest code and we're still handling the fallout so there will be a new version of the request API I hope to post it before ELC [17:33] can you make sure we can consume it in renesas-drivers? I will thanks. pinchartl: can you add "Display-List" with "complete" status ? s/complete/merged/ [17:34] we would like to know when it was merged yes thanks done next, [17:35] - DU,?,noplan,?,HDMI output support on Alt - DU,?,noplan,?,LVDS output support on Gose - DU,?,noplan,?,LVDS output support on Alt - FDP,2016,noplan,?,Upstream driver we still have no plan although, for FDP, I have someone who could start working on it right after ELC but we'll need to sort out the budget the Gen2 DU bits can be slow going IMO, not very important ones pinchartl: can you send his cost ? I send this mail before to you [17:36] s/send/sent/ morimoto: sure. I'll reply to your e-mail thanks ok, that's it for the tasks status update [17:38] topic 2 - VSP development Magnus mentioned that 1) request API is not working. 2) multiple input is not working. 3) UDS (both scale up/down) is not correct result. and Morimoto-san, that suspend/resume crashes is there anything else ? [17:39] no, those are the ones i know about it's enough for me :-) i bet =) hehe :) do you have more information about 2 and 3 ? can we begin by 1) ? [17:40] we covered it already perhaps but there is some patch on some internal mail thread to fix something I think we did. there will be a new version for ELC now you're being very precise :-D could you send them my way ? morimoto: can you make sure pinchartl: gets the patch related to request API? OK, will try [17:41] thanks [17:42] morimoto: please hook up to me via google chat if you need details pinchartl: while morimoto digs through his email, can we move to 2? yes do you have more information about 2 and 3 ? [17:43] yes [17:44] apparently RPF -> BRU -> WPF is working [17:45] however in case RPF.1 is connected to BRU.1 there is trouble RPF.0 -> BRU.0 + RPF.1 -> BRU.1 ? sink->source [17:46] is filled somehow it used to work, but might have got broken in recent patches. I'll investigate that pinchartl: I send 0001-media-vsp1-workaroud-request-API.patch to you morimoto: thanks dammsan: I'd like to create automated test scripts for the vsp driver [17:48] but haven't had time to work on that yes pinchartl: thanks i've asked morimoto-san to send a patch related to 2) as well ok :-) how about 3 ? [17:49] right result of UDS is not correct we don't know much detail about it but perhaps you can show us your test case? [17:50] that is known to work? =) it could also be a regression I've tested it before I'll retest it thanks the multimedia guys say they want to add more feature requests pinchartl: sending done [17:51] because we don't have enough work ? :-) but prefer to get the basic features finished first morimoto: thanks i guess they may not be aware of the delay invovled when do the multimedia guys want all that to be fixed ? [17:52] then theralready err already they said they requested to have it in february, but got it semi-working in march [17:53] they seem to be concerned about plane-alpha and pixel-alpha ass well the request API, or the multiple inputs + UDS ? [17:54] multiple inputs + UDS is definitely a regression I've successfully tested it when the features were merged I'll add plane-alpha + pixel-alpha to the todo list target is v4.7 thanks how about we confirm priority of the current list when we meet these guys f2f tomorrow [17:55] that's a good idea [17:56] * uli___ sneaks in hi Ulrich [17:57] last item is suspend/resume crashes I haven't tested that on Gen3 as we don't have propery suspend/resume support however I expect the crash to happen on Gen2 as well fix gen2 first? of course [17:59] I'll add it to the todo list too which gets quite full for v4.7 I'll mark it for v4.8, but I'll let you confirm priorities in your meeting tomorrow thanks [18:01] anything else for the vsp ? we will bring information from that meeting to f2f meet up at ELC not that i am aware of ok, thank you Does datasheet for VSP was OK for you ? [18:02] topic 3 morimoto: I think so good how will the 50% contract affect multimedia development ? [18:03] we have no shortage of work and everything is urgent yet, the contracts will be reduced to 50% for April how do we deal with that ? i doubt it will be very visible from the outside [18:04] if one month drops and two other months increases but that is not your concern perhaps? =) does that mean May and June could be covered by contracts at 125% then ? I think Base 50% + Additional 25% + Additional 25% [18:05] the total amount of cash remains unchanged if we have 50 in April, 100 in May and 100 in June, that's 250 instead of 300 (I'm talking about development time, no cash) but we never have that amount of time anyway with other customers and paperwork due dates [18:06] let's put it another way so first month is always short before invoice? sure, i'll be silent =) when I have to send a report by the 17th of the first month, I still work from the 17th to the end of the month [18:07] uli___: please chat with me using google chat about tasks ok pinchartl: sure are we expected to work less during April compared to what we've done so far ? you may include the second half of your patches in the second report? less = 50% of a normal month it is up to you really [18:08] dammsan: I might, but I still work during the second half of the last month of the contract since you only have the base contract you will not be sure how much you can adjust your work later on my point is that, if we reduce the contract scope to 50% for April and leave it to 100% in May and June, that's 250 work time points instead of the usual 300 meaning development will be slowed down [18:09] so may and june may get more busy so if you want to reduce your burden then work a bit more for base in april to make room for may and june? so, to put it another way, is Renesas asking us to work for free half of April because we might get more work later ? no? renesas is asking you to do what is written in the contract. [18:10] period. but at the same time they're asking us to deliver features faster than what we've done so far hm..? "please deliver faster, but work only 50% to do so" doesn't make too much sense to me :-) that portion that i don't understand who asks you to deliver sooner? you mentioned earlier that HDMI output for Gen3 should be done before v4.7 if possible [18:11] and that the VSP1 issues are expected to be fixed already that means we're considered to be too slow i think you are right [18:12] sorry about that feeling but it is related to too much work and too little resources about that HDMI Gen3 drop it is just from my side that we should aim for something sooner IMO but we need to look at the big picture [18:13] so, what are we expected to do ? with reduced amount of time in April it sure looks like April will move slower we do what we can do with the amount of time we have? following priority? I can certainly leave the option to all developers in the team to work 100% in April while being paid 50% but I won't request anyone to do so of course spreading out work evenly over 3 months makes sense [18:14] but we need time to finalize the additional tasks so there will unfortunately be a delay the additional tasks for May and June will cover 50%, right ? [18:15] so that's 250 instead of 300 in total for Q2, correct ? in my mind they may be more fixed-price than time based but we need to discuss more 250 has never been the goal [18:16] it has always been 300 so we need to find a good balance for the amount of work in the last two months of the quarter could it be that the additional tasks will cover more than 50%, resulting in ~300 for Q2 in total ? [18:17] it depends on how much time each individual has but yes, if the total should be 300 then it must become like that of course, but do you think it could be so ? it could be how? [18:18] because in that case it could make sense to work more than what the contract covers in April and deliver the result in May and June (but again it would be a personal decision, I won't request anyone to do so) yes, working more earlier would be excellent but no papers yet i'm afraid [18:19] I'm sure Renesas won't complain if we work more than we're paid :-) so it is difficult to say what to work on =) the question is whether there's a reasonable chance that that "overtime" could be covered by additional tasks in May and June define overtime =) working 100% in April instead of 50% [18:20] obviously guessing what would be requested in May and June but we can have a reasonable guess I'm pretty sure HDMI output will be requested for instance if you tell me now that, if I can prepare already in April, there's a 90% chance Renesas will want to request more from me in May and June, there's an incentive [18:22] if the chance is 5%, less so :-) I'm trying to understand what will happen ifok i see [18:23] (sorry, got kicked out from my VM) and with the multiple hops between our team and the decision makers in Renesas, plus the language barrier and the business culture differences, it's hard for me to be confident about my understanding of the situation if people are willing to work more w/o papers then i propose that we simply list one major task for each member that can be focused on in april and we think will be included as additional task [18:24] ok I believe we can keep our current focus in April [18:25] on HDMI, VIN and VSP each will require more than 10 days (50%) to complete they might be covered by the base contract or additional tasks in May and June but in both cases I believe they will be covered [18:26] yes, good plan we need to make sure we only include 50% of the deliverables in the April report though no need to play too many paper games IMO but it is up to you how you want to roll if code is queued up and working then report contents are not important [18:27] just a formality I personally won't change my work pace for April (well, except that I won't code much during ELC) same here ok [18:28] should I record this in the meeting report or should I leave it out ? same here if it's ok for me to post patches based on datasheets while I do not have a contract [18:29] I don't think your NDA expires at the end of the SoW it is up to you so that shouldn't be a problem I'll leave it out of the meeting minutes, I'm not sure how I should phrase it [18:30] last topic my only additional comment about this is yes ? if you are planning vacation, feel free to do so in april instead of may/june =) [18:31] good point :-) I'll write that down thanks ok thanks, it was the NDA part that gave me the most grief :) neg: please read your NDA in details though [18:32] I'm not a lawyer topic 4 [18:33] next meeting the next meeting would be in 2 weeks which is ELC time who will attend ELC ? [18:34] Magnus, you said you will be there yes Niklas as well ? yes Ulrich, you don't plan to attend if I remember correctly ? how about Morimoto-san ? correct is Morimoto-san stuck in an interrupt storm again ? [18:36] maybe [18:37] he is coming i know it =) :-) [18:38] I was ked-napped by Magnus, so yes I will got to ELC :-D very nice we'll have the next meeting there F2F meeting :) I think that's it for today then thank you everybody for joining [18:39] thanks pinchartl: do you have any more information about the mini multimedia meeting? it will be held on Thursday pinchartl: is this meeting including members outside this group? [18:40] that's all I know for now or is it for our group? that's the linux media mini-summit ok cool we'll have another meeting for our group during the week any ideas when to meet? My departure time is 8th April for me meeting with you guys is top prio [18:41] it may make sense to decide right away if possible ok I got a mail from Hans telling me to let Maruo know if I wanted to attend, I drop him a mail and see if I can find out more let me see if the conference schedule has been published Do I need something for this mini-summit ? [18:42] I have nothing else planed during ELC so anytime works for periperi meetup ok, my talk is on Tuesday morning [18:43] morimoto: if you want to join the mini-summit you just have to be there and listen. and possibly talk too :-) I've sent a mail and CC'ed the three of you to ask if we have seats [18:44] I think we will I'd prefer the F2F meeting to be on Tuesday afternoon or on Wednesday pinchartl: OK, no ticket, no register are needed Thanks [18:45] i'm fine either Tuesday or Wednesday let's go for Tuesday right after lunch ? or maybe starting at lunch time ? [18:46] pinchartl: every time is OK for me during 4th - 8th (morning) [18:47] start at lunch time must be good I'll see if I can find a nice place to have a team lunch away from the crowd [18:48] on Tuesday works for me ok, I think we're done and I have to go good timing :-) thank you all for attending and see you in San Diego ! thank you!! thanks see ya in CA Thanks [18:49]