path: root/wiki/Chat_log/20180301-io-chatlog

09:03 < wsa_> so, simon is excused i think we can start the meeting
09:03 < pinchartl> hi Vaishali
09:04 < vthakkar> Hi Laurent
09:04 < morimoto> Japan becaming hot :)
09:04 < wsa_> and Vaishali also rejoined successfully, welcome!
09:04 < vthakkar> Yes, my network went off for a minute. Thanks
09:06 < wsa_> so, I'd like to start with the status updates, then officially welcome Vaishali (maybe with some words from her about her), and then have a few open questions, mostly about SDHI
09:06 < wsa_> let's see how much we can talk about with Simon not being here
09:07 < wsa_> Any complaints about that?
09:07 < vthakkar> sounds good.
09:07 < wsa_> nice :)
09:08 < wsa_> so, here are the collected and slightly reworded status updates
09:08 < wsa_> A)
09:08 < wsa_> Simon took part in the HS400 patch discussion and did performance testing of
09:08 < wsa_> host/guest networking with x86 QEMU. Kaneko-san prototyped support for D3
09:08 < wsa_> thermal. Geert fixed alias usage for serial drivers and enabled IIC on M3N.
09:08 < wsa_> Niklas got his resizable MTU patch merged (what about David's remark?).
09:08 < wsa_> Shimoda-san added USB support for M3N. Wolfram finished his generic
09:08 < wsa_> virtualization task and is now evaluating I2C passthrough, reviewed HS400
09:08 < wsa_> patches and assorted SDHI clock dividing, and worked on watchdog requests.
09:08 < wsa_> Uli got his SCIF patches merged.
09:08 < wsa_> B)
09:08 < wsa_> Simon wants to test HS400 on M3M and continue with his QEMU task. Kaneko-san
09:08 < wsa_> wants to send out support for D3 thermal. Geert wants to remove hard dependency
09:08 < wsa_> of serial aliases (needed for I2C as well?). If time allows, Niklas wants to
09:08 < wsa_> recheck the SDHI issue workaround patch he once created. Shimoda-san has
09:08 < wsa_> multiple USB issues to work on (role swap for SS, GPIO support for USB PHY,
09:08 < wsa_> resume issue) and also learns about USB virtualization. Wolfram wants to finish
09:08 < wsa_> his virtualization task, give HW team feedback about I2C IP core, pick up more
09:08 < wsa_> SDHI work, write a talk proposal for AGL, and still wants to talk with Kieran
09:08 < wsa_> about I2C addresses.
09:08 < wsa_> C)
09:08 < wsa_> Simon checked that SDR104 on Porter still fails (test Niklas patch maybe?).
09:08 < wsa_> Wolfram so far depends on gmane NNTP for reporting, but it has become
09:08 < wsa_> unreliable in the last days.
09:08 < wsa_> so, one question to neg 
09:08 < wsa_> neg: can something be done about David's comment about the MTU patch?
09:09 < wsa_> he said most users would want to have interfaces enabled when they resize MTU?
09:09 < wsa_> and a question to geertu 
09:10 < neg> wsa_: I'm sure it can be done, but if we choose to do so I also think we should apply a simlar work to sh_eth
09:10 < wsa_> geertu: I don't think we need to remove hard dependencies on I2C busses, because they match the schematics. Am I correct?
09:10 < wsa_> neg: for sure. I just wondered if it can be done at all and if it was feasible to do so
09:11 < geertu> wsa_: The issue is not related to matching buses with schematics.
09:11 < wsa_> neg: I don't know the HW in enough detail to unterstand if it can be done at all with reasonable effort
09:11 < geertu> The sh-sci driver has a hard dependency on th presence of the serialN aliases. Without them, the driver won't probe.
09:12 < wsa_> geertu: Ah!
09:12 < geertu> This is mostly a relic from the serial core working with a fixed array of ports.
09:12 < neg> wsa_: Sure it's feasible but I think it will take some work as we would have to resize the rings while the nic is up. I have no plan to address this at this time, but if you think it's a good idea I can look at it as part of base
09:12 < geertu> It makes DT overlays more difficult (need to update /aliases)
09:12 < wsa_> geertu: i see. that definately should be removed.
09:12 < neg> wsa_: I think in it self it would be a usefull change for both our network drivers
09:13 < geertu> Note that on SPI the aliases are already optional, as the SPI subsystem handles dynamic bus numbers.
09:13 < geertu> I think I2C is similar to SPI w.r.t. this
09:13 < wsa_> geertu: yes
09:14 < wsa_> neg: well, if David says "most users would want this" and you also say it would be useful, I think we should target that
09:14 < geertu> wsa_: (checked) yes, see i2c_add_adapter()
09:14 < geertu> I think I can do something similar in sh-sci. I won't touch core serial, though ;-)
09:15 < wsa_> geertu: dragons live there
09:15 < wsa_> neg: I will add a todo item for it
09:15 < wsa_> neg: and you want to do it as part of a base task?
09:16 < neg> wsa_: I would be fine to do that in base but are open to other arangements
09:17 < wsa_> okay
09:17 < wsa_>  we will see about that when the next round of add. tasks is due
09:18 < wsa_> if you happen to have free base time (yeah, right) until then, please have a go
09:18 < wsa_> so, simon is not here, we will handle my remark then by mail
09:19 < wsa_> status updates done, if there are no remarks from your side?
09:19 < wsa_> Then
09:20 < wsa_> Vaishali, nice to have you here. Really looking forward to working with you. Do you want to introduce yourself to the group?
09:21 < vthakkar> Wolfram, Sure. thanks.
09:22 < vthakkar> Hi, everyone. I'm Vaishali Thakkar and I started my journey with Linux Kernel via Outreachy. And later worked on security and mm related stuff at Oracle.
09:22 < kbingham> Hello Vaishali :-) (I'm somewhat amused by the coincidence of finding you on twitter as kernel_girl - then the next day being told you were joining the team!)
09:23 < vthakkar> kbingham: Ah, I assumed that you knew that before following me. :)
09:23 < vthakkar> Nice to meet you here.
09:24 < jmondi> vthakkar: Nice to meet you (we have been introduced in Prague if I'm not wrong). Welcome!
09:24 < jmondi> (and good morning to everyone else)
09:25 < vthakkar> jmondi: Yes, you're right. Nice to meet you too.
09:25 < kbingham> vthakkar: Nope :-) - You came up in some post from either google+ or twitter
09:25 < vthakkar> kbingham: Cool. :)
09:25 < kbingham> vthakkar: Sounds like I missed you in Prague - (I couldn't attend as we were having our first baby here)
09:26 < pinchartl> kbingham: that's just an excuse. lots of people have babies in Prague :-)
09:26 < wsa_> vthakkar: out of interest, did you do some device driver work so far?
09:26 < kbingham> pinchartl: I'm not sure Keri would agree :)
09:27 < vthakkar> wsa_: I've experience working on microcontrollers [Arduino, Tessel etc] for small projects but not for kernel drivers.
09:27 < wsa_> vthakkar: nice! well, that connection can be easily made ;)
09:28 < vthakkar> And this is exactly why I wanted join because so far I had worked on mm and security related stuff only. :)
09:28 < vthakkar> specifically in mainline.
09:28 < vthakkar> wsa_: Sure
09:29 < geertu> vthakkar: We met in Prague, but I believe you were at ELCE in Dublin, too, right?
09:29 < wsa_> I think you found a good spot to do this :) Again, welcome to our team!
09:29 < neg> vthakkar: welcome
09:29 < vthakkar> geertu: I wasn't in Dublin. My visa was rejected ;) Julia presented my work.
09:30 < vthakkar> neg: thanks
09:30 < vthakkar> geertu: But yes, we met in Prague.
09:30 < vthakkar> wsa_: Thanks :)
09:30 < uli___> vthakkar: hi and welcome! may i ask how you got roped into this job specifically?
09:31 < geertu> vthakkar: Oops.  But I did attend Julia's session, so that must be how I remembered.
09:32 < wsa_> okay, so for the remaining issues
09:32 < vthakkar> uli__: Magnus contacted me when Julia presented my work at Dublin one. Then for some reason, I had to take up a job at Oracle and then I met Magnus at 2-3 other conferences in last 2 years and we talked about the work your groups are doing.
09:33 < wsa_> morimoto: shimoda: what is your take on maybe not enabling eMMC HS400 mode on H3 ES1.0?
09:33 < wsa_> as far as I can see, we have no "recommended HW setting" for CPG for ES1.0
09:34 < wsa_> and the one I tried on my own didn't work for writing data
09:34 < shimoda> wsa_: I don't remember about eMMC HS400 mode on H3 ES1.0
09:34 < wsa_> or we could ask the HW guys if there is some recommended HW setting for ES1.0 maybe?
09:34 < shimoda> but, H3 ES1.0 is not good, especially clock thing
09:36 < shimoda> wsa_: hmm, I guess hw guys said you should not use ES1.0, but I'll ask anyway :)
09:36 < wsa_> shimoda: thank you. I agree that makes most sense, if they say "don't do it", then we won't do it
09:37 < wsa_> it = HS400
09:37 < shimoda> wsa_: i got it
09:37 < wsa_> dammsan: are you here?
09:37 < wsa_> Marex: and you?
09:38 < dammsan> yep im here
09:38 < geertu> wsa_: Marex is excused (@EmbeddedWorld)
09:39 < wsa_> I see
09:39 < wsa_> dammsan: did you read the mail from Shimoda-san about listing the revisions in the SDHI driver?
09:41 < dammsan> yeah, I discussed this with him in the office today, and i only care about disabling disallowed combos from use, ie early ES versions.
09:41 < dammsan> one way to solve this could be a black list of those ES versions
09:41 < dammsan> but i have no strong feeling really
09:42 < wsa_> like blacklisting HS400 on ES1.0? ;)
09:42 < wsa_> I see
09:42 < wsa_> thanks for the heads up!
09:43 < dammsan> s/on ES1.0//g =)
09:43 < dammsan> thanks
09:43 < wsa_> okay, this was it from my side
09:43 < kbingham> wsa_: What topics would you like to discuss regarding I2C addresses, and when would you like to do so - I'm assuming it's related to us talking when we were in Brussels - but my memory has now faded. We do have an issue in core where we might need to be able to mark an address as 'unusable' without actually using it ourselves... (unless we use it in which case it's - "in use" ...)
09:44 < wsa_> kbingham: I wanted to talk about how to encode "reprogammable addresses"
09:44 -!- horms [~horms@penelope-musen.horms.nl] has joined #periperi
09:44 < wsa_> and how to handle them (including power-on default addresses)
09:44 < kbingham> Ah yes - handle it better at the core
09:44 < wsa_> That seemed to me like the first step
09:45 < kbingham> We do have a lot of reprogrammable addresses now.
09:45 < kbingham> Ok - now I understand the context - we can schedule outside of this meeting :)
09:45 < wsa_> all the other stuff we talked in Brussels seem to either depend on it or at least would be considerable easier when we have that
09:46 < wsa_> we should have a seperate meeting/thread about that
09:46 < wsa_> can't really predict when, though :(
09:46 < kbingham> wsa_: Ok - understood :)
09:46 < wsa_> hi simon
09:46 < pinchartl> I'd like to participate to that discussion too
09:47 < wsa_> pinchartl: sure, you are welcome
09:47 < horms> hi, sorry I am late
09:47 < wsa_> hope you enjoyed the performance at the school :)
09:47 < horms> I did, thanks
09:47 < wsa_> okay, any more issues?
09:49 < wsa_> okay, then I'll hand over to Geert (sorry for the delay!)
09:49 < geertu> np