path: root/wiki/Chat_log/20160223-io-chatlog

--- Log opened Tue Feb 23 08:56:22 2016
08:56 -!- wsa_ [~wsa@79.226.92.116] has joined #periperi-io
08:56 -!- Irssi: #periperi-io: Total of 2 nicks [1 ops, 0 halfops, 0 voices, 1 normal]
08:56 -!- Irssi: Join to #periperi-io was synced in 0 secs
09:00 -!- geertu [~geert@84.195.106.123] has joined #periperi-io
09:01 < geertu> Hi
09:01 <@horms> hi geertu 
09:02 < wsa_> hi geert
09:02 < wsa_> hi simon
09:02 <@horms> hi wsa_ 
09:03 -!- morimoto [~user@relprex1.renesas.com] has joined #periperi-io
09:03 < morimoto> Hi
09:03 < wsa_> hi morimoto-san
09:03 < morimoto> shimoda-san will be here
09:03 < morimoto> soon
09:04 < wsa_> good
09:04 -!- shimoda [~shimoda@relprex1.renesas.com] has joined #periperi-io
09:04 < shimoda> hi
09:04 < wsa_> hi shimoda-san
09:04 < shimoda> I'm sorry for late
09:05 < wsa_> no worries
09:05 < wsa_> but let's get started
09:05 < wsa_> any questions to the updates I did to the todo file? any further updates?
09:06 <@horms> none from my side
09:06  * geertu checks git
09:07 < wsa_> horms: I will apply your ARCH_RENESAS patch for I2C today or tomorrow
09:08 <@horms> thanks. I hope to roll out the remaining dozen or so patches for ARCH_RENESAS (not I2C!) soonish
09:08 < geertu> I think the SCIF tasks should be postponed to v4.7.
09:09 < geertu> For MSIOF, I tested the fix in the BSP, but it doesn't work for me.
09:09 < geertu> Perhaps it only works for other SPI modes than mode 0?
09:09 < geertu> The fix modifies some timing parameters. I have a deeper look with the logic analyzer to see if I can tune the parameters to make it work.
09:09 < wsa_> geertu: ok, moved SCIF tasks to v4.7
09:10 < geertu> For SPI slave, I noticed something interesting on the mailing list.
09:11 < geertu> Apparently some devices have an SPI_READY pin, so the slave can indicate it's ready to provide data.
09:11 < wsa_> geertu: so, it is mode0 that doesn't work? does it work on gen2?
09:11 < geertu> msiof works on Koelsch with my test board (2 HC595 shift registers and LEDs)
09:12 < geertu> Cfr. the pictures in thread "r8a7795 SYS-DMAC issues (was: Re: [PATCH] [RFC] arm64: renesas: r8a7795: Complete SYS-DMAC nodes)" on periperi-ml
09:13 < geertu> I emailed the SPI_READY guy to ask for SPI slave hardware supporting this. He couldn't say more than it's "for automotive", and CCed our friend Dirk Behme.
09:13 < geertu> So perhaps I have found the requestor of the feature ;-)
09:14 < wsa_> :)
09:15 < wsa_> Nice detective work ;)
09:15 < geertu> I'll check with Dirk
09:15 < wsa_> so, i'll leave the MSIOF issues for 4.6 and wish you good luck
09:15 < geertu> Thx! ;-)
09:16 < wsa_> shimoda-san: any news on the USB side?
09:17 < shimoda> wsa_: today usb phy patch was merged to allow cpg driver of gen3
09:17 < shimoda> so
09:17 < shimoda> so, i would like to send some integration patches.
09:18 < shimoda> however, for ehci ch 0 patch is "RFC", I will not send such patch.
09:18 < shimoda> horms: is it acceptable for v4.6?
09:18 <@horms> shimoda: if you are very quick
09:19 <@horms> I'd like to finalise my branches tomorrow
09:19 < shimoda> horms: i got it, i will try it today
09:20 < wsa_> great
09:20 < shimoda> and I have other topic from BSP team
09:20 < shimoda> they would like to support DRIF module for gen3
09:21 <@horms> shimoda: thanks
09:21 < shimoda> and i got a datasheet of DRIF
09:21 < geertu> DRIF used to be in older versions of the datasheet
09:22 < shimoda> I heard REE/Phil is developing it
09:22 < geertu> It's "just" 2x MSIOF slave, right?
09:22 < geertu> Older R-CarH3_01_U12_DRIF_0010_e.doc says:
09:22 < geertu> "This LSI includes two sets of DRIF (Digital Radio Interface). Each of DRIF is composed of two MSIOF modules, which supports Slave Receive Mode only."
09:22 < geertu> But latest 0.5E datasheet has only some DRIF remainings (pfc/mstp/dmac)
09:23 < shimoda> on gen3 manual of DRIF, "This LSI includes four sets of DRIF"
09:25 < wsa_> Hmm, so is this basically the same as this task?
09:25 < wsa_> "SPI,2016-05-31,plan,geert,Implement initial SPI slave prototype support for R-Car Gen2"
09:25 < morimoto> geert: latest DRIF says
09:25 < wsa_> we haven't specified which SPI-core to use for that, have we?
09:25 < morimoto>  
09:25 < morimoto> This LSI includes four sets of DRIF (Digital Radio Interface). Each of DRIF is composed by two sub modules. In this
09:25 < morimoto> module, S3D2φ is used as module clock, clock input from LSI is used as serial clock for reception.
09:25 < geertu> wsa_: MSIOF slave
09:26 < geertu> Yeah, besides working MSIOF slave, another difficulty may be in tying the two instances together
09:28 < geertu> Do we know what (dual SPI master) hardware is connected to DRIF?
09:28 < geertu> Can we get test hardware?
09:29 < wsa_> shimoda: so, the task looks to me to watch for Phil's DRIF patches and test them when they come along?
09:29 < shimoda> geertu: i will ask someone
09:30 < shimoda> wsa_: i think so if possible
09:30 < wsa_> shimoda: was there a deadline mentioned?
09:32 < shimoda> wsa_: i don't know, so I will ask bsp team
09:32 < geertu> FWIW, SPI (MSIOF) slave has 2016-05-31
09:32 < wsa_> geertu: I'm inclined to assign this task to you? ;)
09:33 < geertu> Let's wait and see...
09:34 < wsa_> OK, so I added this task:
09:34 < wsa_> "DRIF,?,plan,?,Assist Phil in developing the driver"
09:34 < shimoda> wsa_: this task is good to me
09:35 < wsa_> morimoto: any news on the Gen3-Thermal driver?
09:35 < wsa_> shimoda: good
09:35 < morimoto> We will have meeting about that with RVC next month
09:35 < morimoto> (I think I can create this driver before that :)
09:36 < wsa_> :)
09:36 < wsa_> so, the deadline "2016-03-31" is still valid?
09:36 < morimoto> I think postpone is good for me
09:38 < wsa_> 2016-06-31?
09:38 < morimoto> good !
09:38 < morimoto> And I think it will be RVC task
09:39 < wsa_> khim-san?
09:39 < morimoto> I think so
09:39 < wsa_> I was about to ask about him anyway
09:40 < wsa_> I assume you would ask on the periperi-list if khim-san needs additional tasks to be assigend?
09:40 < wsa_> Or how is it handled?
09:41 < morimoto> Umm
09:41 < morimoto> I don't know. Do we need track his task in periperi-list ?
09:43 < wsa_> Either him, or the person petting him :)
09:43 < wsa_> we will see
09:43 < wsa_> for now, I leave the thermal task to you, morimoto-san. OK?
09:43 < morimoto> OK.
09:43 < morimoto> I will report the result
09:44 < wsa_> thanks
09:46 < wsa_> shimoda, morimoto: can one of you ask about this strange bit in SDHI?
09:47 < wsa_> this thread on periperi: "[periperi] Q: ACC_REG in Gen2 SDHI Docs?"
09:47 < morimoto> OK, I can ask it to HW guys
09:48 < wsa_> and is my knowledge accurate that the next Gen3-BSP release is end of February?
09:48 < wsa_> morimoto: Thanks!
09:48 <@horms> wsa_: i expect to get an update on that on Friday
09:49 < wsa_> horms: urgent BSP release?
09:50 <@horms> wsa_: there is a BSP meeting on friday. I expect the timing of the forthcoming releases to be discussed there.
09:51 < wsa_> ah, okay
09:51 < wsa_> horms: can you post the result for the next release on periperi?
09:52 <@horms> sure
09:52 < wsa_> thanks
09:52 <@horms> be aware that there are internal and external releases
09:53 <@horms> its the external ones that go in kernel.org, to customers, etc...
09:53 < wsa_> i am interested in the next update I can have access to
09:53 <@horms> understood
09:53 <@horms> me too!
09:53 < wsa_> :D
09:54 < shimoda> wsa_: next Gen3 BSP release is end of March
09:54 <@horms> I can see the internal ones, but its not so interesting afik
09:54 < wsa_> end of march, okay thanks
09:54 < wsa_> I ask because I couldn't get 8-bit eMMC to work
09:55 < wsa_> it is planned for the next BSP, so I wanted to check if there is some gory detail i am missing
09:55 < wsa_> shimoda-san kindly sent me some patches but the gory detail was not in there
09:55 < wsa_> so I was interested in building the BSP and check if/how it works
09:56 < wsa_> that's the full story
09:56 < shimoda> wsa_: thanks for the full story
09:57 < shimoda> I have internal patch set of the current internal bsp
09:57 < wsa_> one more info from my side: i have now access to the Coverity-reports for the upstream kernel. I tried some patterns finding renesas-specific files and there were some hits.
09:58 < wsa_> i'll investigate if they are not false positives and fix them along if there are not too many
09:58 < wsa_> if there is a bunch, it may be worth to make a task out of it
09:59 < wsa_> but we can discuss that later
09:59 < wsa_> that's all from my side
09:59 < wsa_> are we done?
09:59 < geertu> I have another question
10:00 < geertu> Anyone who knows what's the rationale behind the private review of the CANFD driver?
10:00 <@horms> not i
10:01 < morimoto> I don't know background, but Munakata-san and REE had meeting about it
10:01 < morimoto> and it seems it was REE's challenge
10:01 < morimoto> REE want to send it to ML
10:01 < geertu> Good ;-)
10:01 < morimoto> but they want private review as 1st step
10:02 < morimoto> So, we need to tell him to send it to ML :)
10:02 < wsa_> wearing my maintainer's hat, I appreciate if companies do the first rounds of review internally :)
10:02 < morimoto> :)
10:03 < wsa_> looks like we are done now
10:03 < geertu> wsa_: And wearing your company-internal hat? ;-)
10:04 < wsa_> geertu: that hat tells me it is benefitial to be nice to maintainers :D
10:06 < wsa_> and, it looks more professional to have proper patches sent out
10:06 < wsa_> and the first version really had some way to go
10:06 < wsa_> OK, let's call it a day
10:06 < wsa_> thank you guys!
10:06 < morimoto> thanks
10:07 < shimoda> thank you!
10:07 < morimoto> wsa_: I would like to confirm
10:07 < geertu> thx
10:07 < geertu> bye
10:07 < morimoto> bye
10:07 < morimoto> wsa_: about your question about SDHI
10:07 < wsa_> yes
10:08 < morimoto> Are you checking sh_mobile_sdhi_sdbuf_width() ??
10:08 < morimoto> I'm missing point
10:09 < wsa_> yes, i mean this function
10:09 < wsa_> for version 0x490c, it always uses 0x0001 for 32-bit
10:10 < morimoto> yes
10:10 < wsa_> now, this document for channels 0+1
10:10 < wsa_>  R8A7790(R-CarH2) SD Host Interface CH0,CH1 Hardware Manual(V00R01E).pdf
10:10 < wsa_> chapter 1.2.23
10:11 < morimoto> verion = 0x490c = CH2,3
10:11 < wsa_> 0: 32-bit access
10:11 < wsa_> and in 1.2.22
10:11 < morimoto> for CH0,1 ?
10:11 < wsa_> 16’h490C
10:12 < morimoto> ?
10:12 < wsa_> the filename of the document says channel 0+1, yes
10:12 < wsa_> shall i send you the document I am referring to?
10:13 < wsa_> (i got this from the renesas-ftp once)
10:13 < morimoto> please wait
10:14 < morimoto> ver = 0x490c = CH2,3 = 0: 16bit, 1:32bit
10:14 < morimoto> ver = 0xCB0D = CH0,1 = 0: 32bit, 1:16bit
10:15 < morimoto> Is this correct for you?
10:16 < wsa_> that would be nice
10:16 < wsa_> so, the document i have has an error?
10:17 < morimoto> does you document is not same as above ?
10:17 < wsa_> because it says in 1.2.22 initial value is "16’h490C"
10:17 < morimoto> if so, your document was wrong version 
10:18 < morimoto> I ping-pong:ed this with SDHI guys before
10:18 < morimoto> (I forgot detail...)
10:18 < wsa_> can i have access to the latest version?
10:18 < morimoto> What is your document version ?
10:18 < morimoto> I have is
10:19 < morimoto> CH2,3 = V00R03E
10:19 < wsa_> i have this
10:19 < wsa_> R8A7790(R-CarH2) SD Host Interface CH0,CH1 Hardware Manual(V00R01E).pdf
10:19 < morimoto> CH0,1 = V00R05E
10:19 < morimoto> oops, too old :)
10:19 < wsa_> and R8A7790(R-CarH2) SD Host Interface CH2,CH3 Hardware Manual(V00R01E).pdf
10:20 < wsa_> yes, looks like it :)
10:20 < morimoto> OK, we now find the bug :)
10:20 < morimoto> I wounder how did you get it ?
10:20 < morimoto> Ah, from FTP
10:20 < wsa_> i was SO hoping for a documentation error :)
10:21 <@horms> I think I'll depart now unless there is anything more (non SDHI) to discuss
10:21 < morimoto> Oops, sorry Simon
10:22 < morimoto> Yes, IO meeting was finished I think
10:22 <@horms> No problem
10:22 -!- horms [~horms@reginn.isobedori.kobe.vergenet.net] has left #periperi-io ["Leaving"]
10:22 < morimoto> OK, I will ask to Jinso to send latest SDHI version to you
10:22 < wsa_> thanks!
10:22 < morimoto> was_: do you think your issue will be solved by it ?
10:22 < morimoto> s/was_/wsa_/
10:23 < wsa_> yes, I will add comments to the upstream driver to make things clear
10:23 < wsa_> luckily, no code changes are needed
10:23 < morimoto> I hope so :)
10:23 < morimoto> Thank you for your hard work
10:24 < morimoto> I will ask it tomorrow. (because of time-zone issue :)
10:24 < wsa_> very welcome
10:24 < morimoto> OK, thanks. bye
10:24 < wsa_> thanks for all your assistance
10:24 < wsa_> tommorow is fine, cya!
10:25 < shimoda> bye!
10:26 -!- shimoda [~shimoda@relprex1.renesas.com] has quit Quit: WeeChat 0.4.2
10:27 -!- Irssi: #periperi-io: Total of 3 nicks [0 ops, 0 halfops, 0 voices, 3 normal]
--- Log closed Tue Feb 23 10:27:25 2016