path: root/wiki/Chat_log/20160411-io-chatlog

--- Log opened Mon Apr 11 09:55:04 2016
09:55 -!- wsa_ [~wsa@p4FE259C0.dip0.t-ipconnect.de] has joined #periperi-io
09:55 -!- Irssi: #periperi-io: Total of 6 nicks [1 ops, 0 halfops, 0 voices, 5 normal]
09:55 -!- Irssi: Join to #periperi-io was synced in 0 secs
09:55 < wsa_> hiya
09:55 < wsa_> wow
09:55 < horms> Hi wsa_ 
09:56 -!- shimoda [~shimoda@relprex1.renesas.com] has joined #periperi-io
09:56 < wsa_> hi guys
09:56 < shimoda> hi
09:56 < morimoto> hi
09:56 < neg> morning
09:56 < uli___> hi
09:56 < wsa_> nice to see you again!
09:57 < wsa_> so, ELC was good?
09:57 < neg> yes it was nice
09:58 < neg> good for me who are new to meet people in the flesh
09:58 < morimoto> yes I think so
09:58 < wsa_> neg: yes, hallway track is usually my fav, too :)
09:59 < wsa_> okay, then, let's get started
10:00 < wsa_> I didn't notice updates for the todo file, please tell me if I missed something
10:00 < wsa_> from my side, SDHI support for UHS-I/SDR50 was merged
10:00 < wsa_> and WDT support should show up in linux-next any day
10:01 < wsa_> SDHI probably needs DMA updates to achieve full speed
10:01 <@geertu> wsa: Have you tried Günter's restart series?
10:01 < shimoda> oh, it's nice! I (or Jinso) will test the UHS-I/SDR50 :)
10:01 < wsa_> I guess this is why we have this in the todo file:
10:01 < wsa_> SDHI,?,noplan,?,Confirm & check + tune up DMA transport speed on Gen2
10:02 < wsa_> geertu: not yet, but will surely do today or tomorrow. It looks great to me.
10:03 < wsa_> shimoda: I get transfer rates of about 35MB/s while the card should do 90MB/s. It does 75MB/s on a Tegra-board.
10:03 <@geertu> wsa: Then the remaining question is when (if ever) to make the RWDT restart priority lower than the PSCI priority ;-)
10:04 < wsa_> geertu: yes
10:04 < wsa_> and how to activate the SWDT, because I still don't know how/where/when to change a register in secure-mode
10:05 < wsa_> oh, how impolite of me. I missed something!
10:05 < shimoda> wsa_: i see. I guess PIO is impossible to be 90MB/s on Gen3.
10:05 < wsa_> let's welcome Niklas to our group who is contracted to do IO work in Q2.
10:05 < wsa_> I didn't know until today :)
10:05 <@geertu> neg: Welcome to I/O!
10:06 < neg> thank you
10:06 < wsa_> uli___: are you also contracted for IO work or are you just interested? :)
10:06 < uli___> i'm in
10:06 < uli___> contractually, that is
10:07 < wsa_> shimoda: I agree. I first want to add eMMC support and then port the Gen3 DMA driver.
10:07 < wsa_> uli___: then welcome, too!
10:07  * geertu is not contracted for I/O work
10:07  * horms neither
10:07 < wsa_> geertu: then, bye ;)))))
10:08 < wsa_> okay, we need to talk about this contracting soon
10:08 < wsa_> in a second
10:08 <@geertu> My I/O status: Not much happened on the SCIF front due to SYSC and Easter
10:08 <@geertu> Will resume SCIF this week
10:09 < horms> geertu: quick question. am I right in thinking there is no QSPI present in Gen 3?
10:09 < shimoda> wsa_: add eMMC support first and then DMA support are good to me
10:09 < wsa_> the not surprising thing about SDHI: it sometimes really needs refactoring before adding something new to it.
10:09 < wsa_> shimoda: good
10:09 <@geertu> horms: There is. But It's for secure mode only.
10:09 < wsa_> then we will see how Gen3 DMA performs and where the bottlenecks are
10:10 < horms> geertu: ok. is it called something else in the data sheet? or absent?
10:11 < horms> wsa_: the sdhi driver has a long and cheqquered history
10:11 <@geertu> horms: It's called RPC
10:11 < horms> geertu: ok, i saw that one. thanks.
10:11 < wsa_> horms: i noticed :D
10:12 < wsa_> but i found an old TMIO datasheet which was not detailed but helped in understanding the "history"
10:12 < wsa_> and the different names for bits
10:13 < wsa_> okay, if there are no more updates to the todo, then let's talk about the situation with the contracts
10:14 < wsa_> so, i learned a lot of things today who is contracted and who is not
10:15 < wsa_> and, to be honest, i am a bit surprised to find ulrich and niklas here because they did great work for the multimedia group which is not short of tasks, is it?
10:16 < neg> my situation is that I got 5 days each for i/o, core, and multimedia (15 in total) during Q2 and ontop of that there will be additional contracts mainly for multimedia during Q2 as it looks now
10:16 < uli___> same here
10:17 < wsa_> ah, okay
10:17 < wsa_> geertu: and you are officially not contracted for IO?
10:18 < wsa_> horms: and you?
10:18 <@geertu> wsa: Indeed. Just Core Group Lead and Development
10:18 < wsa_> geertu: similar here, just with IO group, by the way
10:19 < horms> presently i am primarily working on maintainance and secondarily with Kaneko-san on upporting 
10:20 < wsa_> horms: doing ethernet as a side-project hobby ;)
10:20 < horms> that has been an ongoing side project but of late its mainly been handled by Kaneko-san via the upporting work.
10:20 < wsa_> okay
10:21 < wsa_> so we need to find tasks for Uli and Niklas
10:21 < horms> iirc the main outstanding issue there is reducing dma descriptor usage for the aligned case
10:21 < horms> also Sergei sort of stepped up at some point, which imho is better than bad
10:22 < wsa_> technically, i think it would be nuts to move SCIF tasks away from Geert
10:22 < wsa_> horms: nice to hear that
10:23 < wsa_> at the the baud generator stuff is public and WIP anyhow
10:23 < wsa_> geertu: is this FIFO issue something which can be done by someone else
10:23 < wsa_> i mean reasonably
10:24 <@geertu> wsa: I assume there will be an "additional contract" in the future
10:24 <@geertu> so I don't expect to retire from I/O ;-)
10:24 < wsa_> \o/
10:24 <@geertu> BTW, what do you mean by "WIP"?
10:24 < wsa_> ah, not baud generator
10:24 < wsa_> flow control pins
10:25 <@geertu> because BRG is in
10:25 <@geertu> There are some flushing patches in the BSP. and in Hamza's git tree
10:26 < wsa_> so, tasks I see for uli___ and neg I see from our todo-list:
10:26 < wsa_> Thermal,2016-06-30,plan,morimoto,Upstreaming For H3
10:26 < wsa_> I2C,2016-06-30,plan,wolfram,Gen3 I2C DMA support
10:26 < wsa_> SPI,2016-05-31,plan,geert,Implement initial SPI slave prototype support for R-Car Gen2
10:26 < wsa_> which all need reassignment
10:27 < wsa_> looks OK? did i miss something?
10:28 < morimoto> wsa_: RVC Khiem want to do Thermal for Gen3
10:28 < morimoto> s/Khiem/Khiem-san/
10:28 < wsa_> I see
10:28 < neg> I can start by looking at I2C DMA since I poke around with DMA anyhow
10:28 < horms> Not wanting to go around in circles, but is there any application of the SPI task to Gen 3? If not, I'm curious to know its motivation.
10:29 < wsa_> IIRC think it was a customer request? shimoda-san, do you recall more?
10:29 <@geertu> The other SPI on Gen3 is MSIOF, which has hardware issues that are supposed to be fixed in ES2, or in M3-W
10:30 < horms> Ok, so it might be used in the context of MSIOF on Gen 3?
10:30 <@geertu> And DRIF, which is sort-of half-duplex SPI slave
10:30 < wsa_> could we use Gen2 for the prototype?
10:30 <@geertu> That's what it says: "... for R-Car Gen2"
10:30 < wsa_> morimoto: can i add khiem-san to the todo file as "khiem"?
10:30 < horms> wsa_: no argument about prototyping on Gen 2
10:30 < horms> I was just wondering what the end game was
10:30 < wsa_> geertu: got me there :D
10:31 < horms> neg: my notes tell me SYS-DMAC Supports I2C 0 to 6 and IIC-DVFS. v0.51e 17.1.1
10:31 < morimoto> wsa_: that is good idea for me :)
10:31 < horms> neg: in case that is of any help
10:32 < neg> horms: thanks
10:33 < wsa_> uli___: how does implementing SPI slave support sound to you? that's a new thing for linux...
10:33 < horms> neg: i think the implication is that you can do dma to/from 40 bit addresses using the SYS-DMAC. I don't have notes of any other dma capabilities for I2C
10:33 < wsa_> geertu: would you be ok to reassign the task?
10:34 < uli___> i have no clue about spi, so i might as well as well have a look at it...
10:34 <@geertu> wsa: SPI? I think it may be a bit premature
10:35 < wsa_> geertu: what do you mean?
10:36 <@geertu> wsa: There will be additional contracts, right?
10:36 <@geertu> I already spent lots of investigation about SPI slave
10:36 < wsa_> shimoda: how is USB going? is there a place where you could need help?
10:38 < neg> horms: nice, the manual even have flow-charts for DMA operation :)
10:38 < horms> :)
10:39 < wsa_> geertu: i see
10:39 < shimoda> i am doing USB3-host, USB-PHY tasks. But, I forgot the "USB2-Func,?,noplan,shimoda,IPMMU issues" :)
10:42 < shimoda> and i have a problem about usb2 host with IPMMU (that is not listed the todo list yet)
10:42 < wsa_> shimoda: so, could the IPMMU tasks be done by someone else?
10:42 < wsa_> uli___: how does USB & IPMMU sound?
10:42 < shimoda> and I sent a question about it to Magnus-san
10:42 < shimoda> wsa_: it is nice idea
10:43 < uli___> i can look at that
10:43 < wsa_> \o/
10:44 < shimoda> uli___: thanks!
10:45 < shimoda> Oh, i have a note about IPMMU of Gen3.
10:45 < wsa_> so, I'll assign usb2-func & IPMMU task to uli
10:45 < wsa_> and wait for shimoda-san if i should add a task about usb2-host & IPMMU
10:46 < shimoda> we need a new firmware which will be released on this month if we use Gen3
10:46 < horms> v2.8.0?
10:47 < shimoda> horms: yes
10:48 < shimoda> the default setting of IPMMU is not suitable for linux and the register can be modified on secure-mode :)
10:48 < wsa_> I want a u-boot command to change registers in secure-mode :)
10:49 < wsa_> shimoda: can i set the usb3-host task to planned for v4.8?
10:49 < shimoda> wsa_: about usb2-host & IPMMU, I would like to get Magnus-san's reply
10:50 < wsa_> yes, i agree
10:51 < wsa_> i meant this one:
10:51 < wsa_> USB3-Host,?,noplan,shimoda,suspend problems
10:51 < wsa_> since you said you currently work on usb3-host tasks
10:51 < wsa_> or is it something else?
10:51 < shimoda> I see, i'm working the task now because Gen3 also has such an issue :)
10:52 < shimoda> I would like to ask HW team why this issue happen
10:53 < shimoda> so now i'm trying to get register dump and then pass it to hw team
10:54 < wsa_> i see
10:54 < shimoda> about workabout is simple, we just add XHCI_SLOW_SUSPEND quirk on Gen3, but we need a reason for upstreaming
10:55 < shimoda> s/workabout/workaround/
10:57 < wsa_> ok
10:57 < horms> shimoda: we need to know which hw is effected and why?
10:58 < wsa_> uli___: I set the the milestone to v4.8. let me know if this is realistic after you get an overview of the problem
10:58 < shimoda> horms: yes, i will try to get such information
10:59 < horms> shimoda: thanks. fwiw that makes sense to me
10:59 < uli___> will do
10:59 < wsa_> shimoda: can you mail uli a description of the problem?
10:59 < wsa_> neg: I will give some introduction to the I2C DMA topic as well
11:00 < wsa_> OK, I think we are done then?
11:00 < shimoda> wsa_: sure
11:00 < neg> wsa_: thanks
11:01 < uli___> shimoda: thank you
11:01 < wsa_> we included three new members today!
11:01 < wsa_> the new contracts really shuffle things around ;)
11:02 < wsa_> last chance for news from your side
11:02 < horms> nothing from my side
11:03 < wsa_> Then, thank you all! It seems we found suitable tasks for everyone in the end. I hope everyone is happy.
11:03 < shimoda> nothing from me
11:04 < wsa_> then, see you next time :)
11:05 < uli___> see you
11:05 < neg> thanks all, see you all next time
11:05 < horms> likewise
11:05 < morimoto> wsa_: can you add this meeting log to Redmine ?
11:05 <@geertu> Thx, CU
11:06 < wsa_> morimoto: sure thing
11:06 < morimoto> Thanks. and bye
11:06 -!- horms [~horms@reginn.isobedori.kobe.vergenet.net] has quit Quit: Leaving
11:06 < shimoda> thank you, bye!
11:06 -!- morimoto [~user@relprex2.renesas.com] has left #periperi-io ["ERC Version 5.3 (IRC client for Emacs)"]
11:06 -!- shimoda [~shimoda@relprex1.renesas.com] has quit Quit: WeeChat 0.4.2
11:12 -!- neg [~neg@unaffiliated/neg] has left #periperi-io []
11:15 -!- geertu [~geert@d54C36A7B.access.telenet.be] has left #periperi-io []
11:20 -!- Irssi: #periperi-io: Total of 2 nicks [0 ops, 0 halfops, 0 voices, 2 normal]
--- Log closed Mon Apr 11 11:36:19 2016