path: root/wiki/Chat_log/20160811-io-chatlog

--- Log opened Thu Aug 11 10:00:02 2016
10:00 -!- wsa_ [~wsa@dslb-178-008-085-090.178.008.pools.vodafone-ip.de] has joined #periperi-io
10:00 -!- Irssi: #periperi-io: Total of 3 nicks [1 ops, 0 halfops, 0 voices, 2 normal]
10:00 -!- Irssi: Join to #periperi-io was synced in 1 secs
10:00 < wsa_> hi guys!
10:00 < morimoto> Hi
10:00 -!- shimoda [~shimoda@relprex1.renesas.com] has joined #periperi-io
10:00 -!- neg [~neg@unaffiliated/neg] has joined #periperi-io
10:00 < neg> morning
10:00 < shimoda> hello
10:00 < wsa_> good morning
10:00 -!- geertu [~geert@d54c189fd.access.telenet.be] has joined #periperi-io
10:00 < morimoto> Kon-ni-chi-wa
10:01 -!- khiemnguyen [d2a0fca8@gateway/web/cgi-irc/kiwiirc.com/ip.210.160.252.168] has joined #periperi-io
10:01 < geertu> Mornin'
10:01 < wsa_> Guten Morgen!
10:01 -!- horms [~horms@217.111.208.18] has joined #periperi-io
10:01 -!- Irssi: #periperi-io: Total of 8 nicks [1 ops, 0 halfops, 0 voices, 7 normal]
10:02 < geertu> Goeiemorgen Simon
10:02 < horms> Goed Morgen
10:03 < wsa_> okay, let's start
10:03 < wsa_> sort -R did some magic again
10:04 < wsa_> shimoda: you are first today
10:04 < shimoda> yes
10:04 < shimoda> should i talk about a) b) c)?
10:04 < wsa_> please do
10:05 < shimoda> A) What have I done since last time
10:05 < shimoda> I sent some bugfix patches of usbhs driver
10:05 < shimoda> B) What I plan to do till next time
10:05 < shimoda> I will send performance improvement patch for usbhs + g_ncm
10:05 < shimoda> I will send OTG support for gen3 usbhs driver
10:05 < shimoda> C) Problems I have currently
10:05 < shimoda> investigate an issue that M3-W USB3 doesn't work if high speed device is connected with HW guys
10:05 < shimoda> done
10:05 < wsa_> thanks
10:06 < wsa_> from last time: can you send the email about the suspend problem? khiem wanted to have a look at it, I think
10:06 < wsa_> for usb3-host
10:07 < shimoda> oops, maybe i don't do that
10:09 < khiemnguyen> shimoda: please send the email.
10:09 -!- neg [~neg@unaffiliated/neg] has quit Read error: Connection reset by peer
10:09 < wsa_> ah, so you want to work on the problem yourself?
10:09 < wsa_> or do you say you just forgot it?
10:09  * geertu is called bu the door bell
10:10 -!- neg [~neg@unaffiliated/neg] has joined #periperi-io
10:10 < shimoda> sorry, just forgot it...
10:10 < wsa_> okay
10:10 < shimoda> khiemnguyen: yes, i will send it
10:10 < wsa_> great
10:10 < wsa_> horms: you are next
10:10 < horms> sure
10:11 < horms>   a) what have I worked on since last time
10:11 < horms>      * Resolved timeouts during tuning which is related to UHS-I/SDR104
10:11 < horms>        support; reposted sdr104 patchset; awaiting review
10:11 < horms>      * Posted patches to enable SDHI0, 1 and MMCIF on r8a7794/Alt,
10:11 < horms>        Modeled on r8a7794/Silk; lightly tested; awaiting review
10:11 < horms>   b) what will I work on until next time
10:11 < horms>      * Work on merge of above
10:11 < horms>      * Extend SDR104 coverage to other boards
10:11 < horms>   c) I have the following problems (or not)
10:11 < horms>      * SD cards in Magnus's board farm, maybe
10:11 < horms> end
10:11  * geertu back (sort of)
10:11 < wsa_> so, no timeouts with SDR104 anymore?
10:13 < horms> I have not seen them with the latest patchset
10:13 < horms> They may still be lurking
10:13 < wsa_> awesome
10:13 < horms> Possibly something will show up when I extend support to other Boards
10:13 < horms> Or try other SD cards
10:13 < horms> but I did try the usual-suspect SD cards
10:13 < horms> in particlar one that liked to timeout a lot was my main test case
10:13 < wsa_> as i said in my review, i think we can work incrementally from where we are and merge the patches
10:14 < horms> I can give details on the cards tested etc...
10:14 < wsa_> let's hope ulf agrees :)
10:14 < horms> actually its on the e-linux wiki
10:14 < horms> ok
10:14 < wsa_> uli___: you are next
10:14 < horms> do you think you could respond to the latest patchset with an ack or something?
10:15 <@uli___> a) nothing i/o-related; b) check sdr104 performance (card's in the mail); c) nothing i'm aware of
10:15 <@uli___> horms: could you send a link to that e-linux page?
10:15 < horms> one moment
10:15 < geertu> uli___: Did my MSIOF patch help for mmc-over-msiof?
10:15 < wsa_> horms: I send a Tested-by just yesterday
10:16 < wsa_> sent
10:16 < horms> wsa_: thans!
10:16 <@uli___> geertu: didn't have time to check that yet
10:16 < horms> uli___: http://elinux.org/Tests:SD-SDHI-SDR104
10:16 <@uli___> thanks
10:17 < horms> uli___: Samsung Card 2 is the one that liked to timeout the most. But to be fair there were bugs in the code.
10:17 <@uli___> are these sdr50 or sdr104?
10:18 < wsa_> uli___: you think you can work on MSIOF this month?
10:18 <@uli___> i think i can fit that in
10:18 < wsa_> good
10:19 < wsa_> wsa_: you are next
10:19 < wsa_> with pleasure
10:19 < wsa_> a) i reviewed and tested simons patchset
10:19 < wsa_> i tried to find the i2c-demux regression but no luck so far
10:20 < wsa_> b) I will find this i2c-demux regression
10:20 < wsa_> a) worked on additional tasks, I hope you got the contracts today?
10:20 < geertu> wsa_: Got the draft SoW today
10:21 < wsa_> c) the problem with the i2c-demux issue is that a pointer within the devicetree seems to change although it should be constant
10:21 < wsa_> (but we are talking OF_DYNAMIC here, so a bit more tricky)
10:22 < horms> uli___: the test numbers are sdr104
10:22 < wsa_> and Magnus being totally gone
10:22 < wsa_> like for some negotiations or changes to his lab
10:22 < horms> uli___: which is only available on one slot on the r8a7790/Lager
10:23 <@uli___> horms: ok, thanks
10:23 < wsa_> that's it
10:23 < horms> wsa_: thanks, my SoW arrived
10:23 <@uli___> [off-topic] geertu: i got my core draft today as well
10:23 < wsa_> has someone any news from Magnus?
10:23 < horms> wsa_: your pointer findings match my earlier inestigation
10:24 < morimoto> wsa_: nothing to Renesas
10:24 < neg> wsa_: I tried to contact him a few days back to get a copy of my SoW still no response
10:24 < horms> wsa_: you may want to look at the presentation by Frank Rowland at LCJ. The tool he discussed there for dumping and diffing kernel DT may be useful somehow
10:24 < horms> wsa_: no, I have not heard from M
10:24 < horms> not for about a month now
10:24 < horms> hopefully he is still alive
10:24 < geertu> I talked to him before I went on holidays. He planned to be back in Tokyo August/M
10:25 < wsa_> horms: oh, that tool could be interesting, thanks
10:25 < wsa_> so, maybe just a few more days
10:25 < horms> wsa_: i specifially asked if it was possible to diff between kernel DT states over time. He said yes.
10:25 < wsa_> nice!
10:25 < horms> Its probably in the recording of the presentation near the end if such a thing exists
10:26 < wsa_> If all fails, I will mail him ;)
10:26 < wsa_> geertu: your turn now
10:26 < horms> Yes, he seemed quite happy to share his knowledge
10:27 < geertu> A) What have I done since last time
10:27 < geertu> Fixed invalid clock generator parameters with MSIOF
10:27 < geertu> B) What I plan to do till next time
10:27 < geertu> r8a7795 MSIOF parent clock control prototype (core)
10:27 < geertu> Implement initial SPI slave prototype support for R-Car Gen2
10:27 < geertu> C) Problems I have currently
10:27 < geertu> None
10:28 < geertu> From my current state, I think static configuration in DT of the MSIOF parent clock frequency is the way to go, though
10:29 < wsa_> ok, let's see how it goes
10:29 < wsa_> neg: your turn
10:30 < neg> a) EtherAVB suspend/resume patches posted and picked up in net-next, still trying to reproduce 'OHCI-PCI doesn't seem to work with CONFIG_DMA_CMA=y' problem.
10:30 < neg> b) Update U-Boot on my Koelsch and see if I can reproduce the CONFIG_DMA_CMA=y problem then
10:30 < neg> c) My toolchain complains when I try to build new u-boot image, need to get older version and procedure to upgrade u-boot seems tricky, hope I wont brick by Koelsch in the process :-)
10:31 < geertu> neg: Have you tried Magnus' Lager instead?
10:32 < wsa_> can't you just use geert's binary?
10:32 < neg> geertu: no, I will try it parallel to trying to update my u-boot
10:33 <@uli___> you could compile u-boot to a different address and chain-load it from the old version
10:33 -!- khiemnguyen [d2a0fca8@gateway/web/cgi-irc/kiwiirc.com/ip.210.160.252.168] has quit Quit: http://www.kiwiirc.com/ - A hand crafted IRC client
10:33 -!- khiemnguyen [d2a0fca8@gateway/web/cgi-irc/kiwiirc.com/ip.210.160.252.168] has joined #periperi-io
10:34 < geertu> uli___: If you do that, you'll have side-effects of hardware initialization of the first one, right?
10:34 < neg> uli___: thanks I will try that
10:34 <@uli___> possible. but it might just work. :)
10:34 <@uli___> i.e., break, in this case
10:35 < neg> flash update, I managed to build u-boot using the toolchain from kbuild so I think I should be fine :)
10:35 < wsa_> good luck then!
10:35 < wsa_> :)
10:35 < neg> so thats about it from me
10:35 < wsa_> khiemnguyen: your turn
10:37 < wsa_> knock knock
10:37 < khiemnguyen> yes.
10:37 < khiemnguyen> A) What have I done since last time
10:38 < khiemnguyen> No update
10:38 < khiemnguyen> It meant that I have not sent any updates into ML yet.
10:38 < khiemnguyen> I have tracked R-Car Gen3 Thermal driver in in-house BSP (v3.3.2), if there's any hints to update my patchset.
10:38 < khiemnguyen> B) What I plan to do till next time
10:38 < khiemnguyen> Complete Gen3 Thermal driver
10:38 < khiemnguyen> C) Problems I have currenty
10:38 < khiemnguyen> (same as for Core group)
10:38 < khiemnguyen> - Time flies but nothing complete.
10:38 < khiemnguyen> - Will need to check my to-do list again.
10:39 < khiemnguyen> As Laurent information, v4.9 might be next LTS.
10:39 < khiemnguyen> I think I will try to complete R-Car Gen3 Thermal for v4.9.
10:39 < wsa_> and does the BSP indicate updates to the thermal driver?
10:40 < khiemnguyen> yes, we did have some updates in BSP v3.3.2
10:41 < khiemnguyen> one of the consideration points is difference of the thermal initialization processing in H3 WS1.0/WS1.1 and 2.0
10:41 -!- khiemnguyen [d2a0fca8@gateway/web/cgi-irc/kiwiirc.com/ip.210.160.252.168] has quit Read error: Connection timed out
10:41 < wsa_> i see
10:42 < wsa_> morimoto: your turn now
10:42 < morimoto> A) I asked about my posted thermal driver patch to maintainer. thermal maintenance system seems confusable. 1 of them was accepted.
10:42 < morimoto> B) I will keep ask and wait for it, and will post next bug fix patch.
10:42 < morimoto> C) no headache
10:42 < morimoto> Here, my thermal is for Gen2, khiemnguyen's one is for Gen3
10:42 -!- khiemnguyen [d2a0fca8@gateway/web/cgi-irc/kiwiirc.com/ip.210.160.252.168] has joined #periperi-io
10:42 < morimoto> EOF
10:43 < khiemnguyen> that's all of my status.
10:44 < morimoto> wsa_: that's all of my status, too :)
10:44 < khiemnguyen> geert: how about the status of SoC revision checking ? Dirk patchset
10:46 < geertu> khiemnguyen: No updates. That's core work, for after August 15
10:46 < wsa_> any other questions from your side?
10:46 < wsa_> i have none
10:47 < horms> no io questions from me
10:47 < wsa_> only the inofficial question what the "H3ULCB" board is? Some custom board? Some (semi-)official Renesas board?
10:47 < shimoda> i have no questions, but I got some SCIF driver's requests from our front team
10:48 < wsa_> shimoda: what are they?
10:48 < geertu> wsa_: Have you read the Core Group Meeting Minutes?
10:48 < shimoda> they needs to improve timeout interval for highspeed (1Mbps or more)
10:49 < shimoda> the current driver uses normal timer API
10:49 < horms> wsa_: yes, i would like to follow up with morimoto on H3ULCB. I can talk to you about it i you like but as Geert mentions its detailed in the core group minutes. afik there are no io issues relating to it at this time.
10:49 < shimoda> they suggest highres timer
10:50 < wsa_> geertu: I have; it doesn't answer my question, though
10:50 < wsa_> Is it a Renesas board?
10:50 < horms> what is your question?
10:50 < geertu> wsa_: I don't know more. "low-cost" means it's the silk/porter equivalent for H3
10:50 < horms> I think so. Morimoto-san, Shimoda-san?
10:50 < geertu> wsa_: According to Sergei, it's Renesas
10:51 < wsa_> thanks
10:51 < morimoto> LCB = Low Cost Board
10:51 < wsa_> shimoda: so i add a task like:
10:52 < shimoda> horms: yes, it is like silk/porter, but I heard our customers is possible to use many the ULCB instead of salvator
10:52 < shimoda> wsa_: thanks for adding
10:53 < horms> morimoto: shimoda: are we ok to proceed with upstreaming. Sergei has been pinging Geert and I on #renesas-soc
10:53 < wsa_> SCIF,?,noplan,?,use highres timers for timeouts with speeds >1MBps
10:53 < shimoda> so, Munakata-san is thinking upstream team should support it as well, not like silk/porter
10:53 < wsa_> geertu: does this task make sense to you?
10:54 < geertu> wsa_: I think so (which highres timer do we have, that we can use? ARM arch timer?)
10:54 < morimoto> H3ULCB board will to e-linux page on Oct. I know horms/geertu want to have schematic soon. So, our other team is now doing export paper work for it. So, hopefuly, you can get schematic 8/E
10:55 < horms> Thanks. In the mean time is it ok from a Renesas pov for us to take patches from Cogent into mainline? From my pov it is ok.
10:55 < geertu> morimoto: Thanks, looking forward to it.
10:56 < shimoda> horms: Goda-san hope the dts is merged into v4.9 for LTS(I)
10:56 < horms> got it
10:56 < horms> I will give Cogent the green light
10:57 < shimoda> horms: thanks for it!
10:57 < wsa_> ok
10:58 < wsa_> i think we are done now with the official part
10:58 < wsa_> H3ULCB is not really IO topic anymore ;)
10:58 < wsa_> thanks guys!
10:58 < shimoda> thank you!
10:58 < neg> thanks all
10:59 < horms> thanks!
10:59 < morimoto> Thanks. Renesas Japan will have Summer Vacation from 13th to 21th.
10:59 < khiemnguyen> thanks.
10:59 < geertu> thx, bye, enjoy the holidays!
10:59 < wsa_> yes, have great holidays
11:00 < morimoto> geertu/wsa_: Thanks !
11:00 < wsa_> i'll be mostly away 15th-21st as well
11:01 < morimoto> I want to ship M3 board to you, but because of time-limit, it is 50%
11:03 < wsa_> for me, no hurry
11:04 < wsa_> IIRC the only one from IO group having a M3 related task is Simon
11:04 < wsa_> which is optional task, even
11:06 < morimoto> OK, but unfortunately, I could confirm that the chance of M3 board before summer vacation is now 0%
11:06 < morimoto> today was shipping time-limit
11:07 < morimoto> So, I will ship it on 22th or later
11:07 < wsa_> ok
11:07 < morimoto> thanks, and bye
11:07 -!- morimoto [~user@relprex1.renesas.com] has left #periperi-io ["ERC Version 5.3 (IRC client for Emacs)"]
11:11 < wsa_> cya!
--- Log closed Thu Aug 11 11:11:03 2016