path: root/wiki/Chat_log/20150819-core-chatlog

11:03 #periperi: < geertu> Two more missing?
11:03 #periperi: < geertu> I mean inactive
11:03 #periperi: < dammsan> hello
11:03 #periperi: < geertu> Simon and Laurent are in Seattle, so they won't join
11:04 #periperi: < morimoto> hello
11:04 #periperi: < geertu> Good, looks lijke we are complete
11:05 -!- Irssi: Pasting 5 lines to #periperi. Press Ctrl-K if you wish to do this or Ctrl-C to cancel.
11:05 #periperi: < geertu> Agenda:
11:05 #periperi: < geertu>   1. Upcoming Gen3 development for the Core group
11:05 #periperi: < geertu>        a. PFC
11:05 #periperi: < geertu>        b. DMAC (e.g. "dmas" for (H)SCIF)
11:05 #periperi: < geertu>   2. SoC Bus topology in DT
11:05 #periperi: < geertu> 1. Upcoming Gen3 development for the Core group
11:06 #periperi: < geertu> Sorry, I missed   3. Status check for tasks from last meeting - what is remaining?
11:06 #periperi: < geertu> Any update from the Tokyo side on Gen3?
11:07 #periperi: < dammsan> geertu: i've got a board now, currently installing it
11:07 #periperi: < geertu> great!
11:08 #periperi: < morimoto> I guess we need to have IRQC as core
11:08 #periperi: < morimoto> Maybe magnus will do it (?)
11:08 #periperi: < geertu> We have a task for that, right?
11:08 #periperi: < morimoto> I guess so
11:09 #periperi: < dammsan> my plan is to develop that portion yes
11:09 #periperi: < geertu> And magnus is the creator of irq-renesas-irqc.c ;-) 
11:09 #periperi: < dammsan> yay!
11:09 #periperi: < dammsan> where is the latest TODO list for the core group? on a wiki?
11:09 #periperi: < geertu> Is the "new" board identical to the "old" one?
11:10 #periperi: < geertu> https://osdr.renesas.com/projects/linux-kernel-development/wiki/Core-todo-list
11:10 #periperi: < dammsan> sweet
11:10 #periperi: < geertu> Going over that list is Topic 3. Or do you think it should be topic 0?
11:10 #periperi: < dammsan> nah, just wasn't sure where it was
11:11 #periperi: < dammsan> sorry, please continue
11:11 #periperi: < geertu> Is the "new" board identical to the "old" one?
11:11 #periperi: < geertu> Also U-Boot wise?
11:11 #periperi: < dammsan> the u-boot has been improved i've been told
11:11 #periperi: < dammsan> hardware is supposed to be the same
11:12 #periperi: < geertu> "improved" means Ether-AVB support?
11:12 #periperi: < dammsan> we'll see
11:13 #periperi: < dammsan> I got code for u-boot Ether-AVB one week ago but nothing worked at that point
11:13 #periperi: < dammsan> Maybe things are better this time
11:13 #periperi: < geertu> Does it have smsc Ethernet, too?
11:13 #periperi: < dammsan> nope
11:13 #periperi: < dammsan> that would be good
11:14 #periperi: < geertu> Is there any documentation about the board (schematics?)?
11:15 #periperi: < dammsan> morimoto: any comment?
11:16 #periperi: < morimoto> Now, we are under paper work for export it
11:16 #periperi: < morimoto> please wait
11:16 #periperi: < geertu> Perhaps Magnus can send a higher-resolution picture of the board to periperi?
11:18 #periperi: < dammsan> perhaps magnus can send a high resolution picture of his screen if someone asks nicely =)
11:19 #periperi: < geertu> hehe
11:19 #periperi: < geertu> a. PFC
11:19 #periperi: < geertu> I understand there's a preliminary driver?
11:20 #periperi: < morimoto> Yes
11:20 #periperi: < morimoto> I think we should (?) reuse it ?
11:20 #periperi: < morimoto> No one want to re-create over 5000line driver
11:20 #periperi: < geertu> IIRC, Laurent wanted to move more to DT?
11:21 #periperi: < geertu> (i.e. write script to convert 5000 lines?)
11:21 #periperi: < morimoto> I don't know, but BSP team is using this driver
11:21 #periperi: < geertu> But without something concrete, that's more like a wild idea...
11:22 #periperi: < geertu> So I'd say let's post the driver
11:22 #periperi: < geertu> Any objections to that?
11:22 #periperi: < dammsan> sounds like a perfect plan to me
11:22 #periperi: < morimoto> me too
11:23 #periperi: < dammsan> is it possible to post it early tomorrow?
11:23 #periperi: < morimoto> I can do it today
11:23 #periperi: < dammsan> i want to use it with IRQC
11:23 #periperi: < geertu> Great!
11:23 #periperi: < dammsan> thanks!
11:23 #periperi: < shimoda> Nice!
11:24 #periperi: < geertu>        b. DMAC (e.g. "dmas" for (H)SCIF)
11:24 #periperi: < geertu> Actually this is more about having core infrastructure pointed to by all device nodes in DT
11:25 #periperi: < geertu> - interrupts
11:25 #periperi: < geertu> - clocks
11:25 #periperi: < geertu> - power-domains = <&cpg_clocks>;
11:25 #periperi: < geertu> - dmas
11:25 #periperi: < geertu> To avoid lots of churn, I think we should try to get the core blocks in ASAP.
11:25 #periperi: < dammsan> sounds good to me
11:26 #periperi: < dammsan> would it be possible for someone to make a first example?
11:26 #periperi: < dammsan> (IRQC has no DMA =)
11:27 #periperi: < geertu> I was thinking about adding all (H)SCIF nodes if I find a spare moment
11:27 #periperi: < geertu> interrupts we already have (GIC node is there, IRQC will be soon)
11:27 #periperi: < geertu> (does anything use IRQC on the board?)
11:27 #periperi: < geertu> clocks we have some parts (exact location in the tree still TBD)
11:28 #periperi: < geertu> For power-domains I already posted the CPG/MSTP Clock Domain patches
11:28 #periperi: < geertu> dmas is lacking, but according to Laurent it's compatible with rcar-dmac, and I think a dummy node can do for now.
11:29 #periperi: < geertu> (have to test that on Koelsch, to see if scif still works with a dummy dmac node)
11:30 #periperi: < dammsan> sounds good. PMIC is hooked up to IRQ0. No quirk is needed.
11:30 #periperi: < geertu> Only PMIC?
11:30 #periperi: < dammsan> yes, a single one
11:30 #periperi: < dammsan> that's it
11:31 #periperi: < geertu> Nice
11:31 #periperi: < morimoto>  and I have its document. I can send it with schematics
11:31 #periperi: < geertu> Cool
11:31 #periperi: < dammsan> the rest is availabile on expansion connectors
11:31 #periperi: < geertu> Are these the same Samtec connectors like on Gen2?
11:32 #periperi: < dammsan> some of them are the same
11:32 #periperi: < dammsan> some are wider i think
11:32 #periperi: < dammsan> you'll see once you get the schematics
11:32 #periperi: < geertu> It's mostly the pitch that matters.
11:32 #periperi: < geertu> OK
11:33 #periperi: < geertu> I'll add a task for adding all (H)SCIF nodes
11:33 #periperi: < dammsan> thanks
11:33 #periperi: < dammsan> will you work with Simon to get the pm domains stuff merged too?
11:34 #periperi: < geertu> Yes
11:34 #periperi: < dammsan> sweet
11:34 #periperi: < geertu> (in the mean time it's in arm-soc for pre-Gen3)
11:34 #periperi: < geertu> Morimoto-san had a problem with it related to sound?
11:36 #periperi: < shimoda> morimoto: "its document" means PMIC's one? if so, we have to do paper work :)
11:36 #periperi: < geertu> PDF work?
11:37 #periperi: < morimoto> Yes it is PMIC document
11:37 #periperi: < dammsan> we need at least 3 different hankos
11:37 #periperi: < dammsan> to stamp both virtual and real papers
11:38 #periperi: < geertu> http://factsanddetails.com/media/2/20090801-customizedtitle6jun.gif
11:39 #periperi: < dammsan> i've started to greatly appreciate the popular software that makes gif images for stamps for inclusion in excel sheet
11:40 #periperi: < dammsan> now when i see an excel sheet without a stamp i somehow feel empty
11:40 #periperi: < geertu> Like the "please send me your digital signature" requests, where they actually want you to send a GIF instead of running gpg?
11:41 #periperi: < geertu> Any more "Upcoming Gen3 development for the Core group"?
11:42 #periperi: < dammsan> lets see
11:42 #periperi: < dammsan> IRQC SCIF PFC
11:43 #periperi: < dammsan> how about GPIO?
11:43 #periperi: < dammsan> do we have a task for that?
11:43 #periperi: < geertu> Linus W: "Patch applied with the ACKs.
11:44 #periperi: < dammsan> right, but does that include the integration stuff?
11:44 #periperi: < geertu> No, dtsi update seems missing
11:44 #periperi: < uli___> should i do that then?
11:45 #periperi: < dammsan> that would be great if so
11:45 #periperi: < uli___> ok
11:45 #periperi: < dammsan> btw: the new Gen3 board gives me random exceptions
11:46 #periperi: < geertu> r8a7795,v4.4,Add gpio nodes to DT
11:46 #periperi: < geertu> oops
11:46 #periperi: < geertu> It's a different board?
11:46 #periperi: < geertu> Different firmware?
11:48 #periperi: < shimoda> dammsan: it's different board than last week
11:48 #periperi: < shimoda> oops, also different firmware
11:49 #periperi: < shimoda> especially "bl2.srec". so please write old one if possible?
11:49 #periperi: < dammsan> shimoda: can i update early software using the SCIF mask rom loader
11:49 #periperi: < dammsan> ok, will try to use the old "bl2.srec"
11:49 #periperi: < dammsan> thanks
11:50 #periperi: < shimoda> thanks
11:50 #periperi: < dammsan> good news is that ether-avb seems to half-work - until i get exceptions
11:50 #periperi: < dammsan> i also get exception when i boot the kernel so...
11:50 #periperi: < dammsan> will downgrade bl2.srec
11:50 #periperi: < dammsan> end-of-gen3-board-stuff
11:51 #periperi: < geertu> Topic   2. SoC Bus topology in DT
11:52 #periperi: < geertu> I guess nothing has changed, so just a single bus
11:52 #periperi: < shimoda> I got one other Gen3 board now. I will try to boot kernel after i wrote some firmware
11:52 #periperi: < dammsan> shimoda: thanks, please do and let me know how it goes
11:53 #periperi: < dammsan> geertu: i think nothing has changed yes
11:53 -!- Irssi: Closing query with morimoto
11:53 -!- Irssi: Closing query with dammsan
11:53 #periperi: < geertu> Topic 3. Status check for tasks from last meeting - what is remaining?
11:53 #periperi: < geertu> https://osdr.renesas.com/projects/linux-kernel-development/wiki/Core-todo-list
11:53 #periperi: < geertu> Device: IPMMU
11:53 #periperi: < geertu> Finally get feedback from the hardware developers regarding the IPMMU + DMAC channel 0 issue
11:54 #periperi: < geertu> Any news?
11:54 #periperi: < dammsan> I don't think so
11:54 #periperi: < geertu> Device: r8a777x
11:54 #periperi: < geertu> Vote which missing support is valuable
11:54 #periperi: < geertu> I guess this can be marked completed?
11:55 #periperi: < dammsan> i guess so. if someone has a board locally and want to hack then that's great
11:55 #periperi: < dammsan> but i think adding USB support over remote access is too difficult to be worth dealing with
11:55 #periperi: < dammsan> especially for Gen1
11:55 #periperi: < dammsan> so yes, mark as completed
11:55 #periperi: < geertu> Device: IPMMU+DMAC
11:55 #periperi: < geertu> Discuss API changes on the dma-engine mailing list
11:55 #periperi: < geertu> Implement proper solution
11:56 #periperi: < geertu> Our DMAC expert is absent and busy?
11:56 #periperi: < dammsan> yes and yes =)
11:56 #periperi: < geertu> Device: SMP
11:56 #periperi: < geertu> Discuss SMP DT bindings with ARM SoC maintainers
11:56 #periperi: < geertu> No progress
11:56 #periperi: < dammsan> Nothing doe
11:56 #periperi: < geertu> Device: r8a7778
11:56 #periperi: < geertu> Add support for valuable devices to multi-platform r8a7778
11:56 #periperi: < geertu> Drop legacy (-legacy and -reference) r8a7778/bockw
11:56 #periperi: < shimoda> Oops, about IPMMU, I didn't ask hardware guys yet because I didn't explain the detail to the guys
11:56 #periperi: < dammsan> s/doe/done
11:56 #periperi: < geertu> I think this can be marked completed, too?
11:57 #periperi: < dammsan> yeah, thanks to hard work from ulrich we are in a much better state with bock-w and r8a7778 now!
11:57 #periperi: < geertu> Device: r8a7779
11:57 #periperi: < geertu> Add support for valuable devices to multi-platform r8a7779
11:57 #periperi: < geertu> And that's the USB one, which is still open
11:57 #periperi: < dammsan> shimoda-san, can we discuss IPMMU face-to-face next week?
11:58 #periperi: < dammsan> geertu: about USB, that's for r8a7779 right?
11:58 #periperi: < geertu> I think so
11:58 #periperi: < dammsan> i think we are limited by physcial access to hardware
11:58 #periperi: < shimoda> magnus-san, thank you!
11:59 #periperi: < dammsan> but if someone wants to hack then we should perhaps rearrange things
11:59 #periperi: < geertu> Sorry, that seems to be valid for both '78 and '79
11:59 #periperi: < geertu> So the valuable devices for '78 is still open, too
11:59 #periperi: < dammsan> geertu: i don't know if the IP is the same
11:59 #periperi: < dammsan> Gen1 seems rather different within the same family
12:00 #periperi: < dammsan> but anyway, keeping open seems good too
12:00 #periperi: < geertu> The remaining tasks are all open
12:00 #periperi: < geertu> Sorry, except for
12:00 #periperi: < geertu> Device: r8a7795
12:00 #periperi: < geertu> Group on-SoC devices under a simple-bus node on r8a7795
12:00 #periperi: < dammsan> shimoda: meeting on friday the 28th about IPMMU clarification?
12:01 #periperi: < dammsan> geertu: i thought the latest version from morimoto-san did this already?
12:01 #periperi: < geertu> Yes, cfr. the "Sorry" above
12:01 #periperi: < shimoda> dammsan: sure! thank you!
12:03 #periperi: < dammsan> ok!
12:04 #periperi: < morimoto> http://thread.gmane.org/gmane.linux.ports.sh.devel/47755
12:07 #periperi: < dammsan> this version is included in the latest renesas-drivers snapshot, right?
12:07 #periperi: < morimoto> I think so
12:08 #periperi: < geertu> morimoto: Thanks for PFC email!
12:10 #periperi: < morimoto> It seems SH-ARM ML didn't accept it. I guess it is too large ?
12:10 #periperi: < geertu> Possibly
12:10 #periperi: < morimoto> (This is always happen on SH-ARM ML...)
12:10 #periperi: < geertu> BTW, gmail started thinking Uli's emails are spam
12:11 #periperi: < uli___> gmail thinks everything is spam... :(
12:12 #periperi: < morimoto> This is prepare for alphabet.com ?
12:13 #periperi: < geertu> morimoto: You didn't CC LinusW?
12:13 #periperi: < morimoto> Oops, I didn't Cc to him
12:15 #periperi: < morimoto> Should I resend it ? or wait Laurent's opinion ?
12:15 #periperi: < geertu> Perhaps better wait for Laurent
12:16 #periperi: < geertu> Let's hope his mail server didn't crash due to the large patch while he's in Seattle ;-)
12:16 #periperi: < dammsan> geertu: if it crashes then perhaps he at least can have a silent vacation =)
12:17 #periperi: < morimoto> Hehe :)
12:18 #periperi: < dammsan> btw, the u-boot says "Salvator-X" for board name but we use "salvatorx", what's up with that?
12:19 #periperi: < geertu> Is that actually a beer?
12:19 #periperi: < dammsan> seems more like a spelling mistake to me =)
12:20 #periperi: < morimoto> According to HW team (they named it), it is standard beer name
12:20 #periperi: < morimoto> s/standard/public/
12:20 #periperi: < dammsan> salva-torx
12:20 #periperi: < morimoto> (This is the reason why we didn't need paper work for it when I sen patch to ML)
12:22 #periperi: < dammsan> just virtual hanko needed? =)
12:22 #periperi: < morimoto> it need no paper work if it is public name
12:23 #periperi: < morimoto> otherwise, we need to check all (?) names
12:23 #periperi: < dammsan> so if "google i'm feeling lucky" gives you a web page then it is fine? =)
12:23 #periperi: < morimoto> to avoid using same name
12:24 #periperi: < dammsan> i understand some uniqueness is needed
12:24 #periperi: < geertu> So it's not about the secrets in the .dts file, just about the name of the board?
12:25 #periperi: < morimoto> uniqueness is one reason. we need to avoid "pirate the trademark"
12:25 #periperi: < morimoto> (? correct English ??)
12:25 #periperi: < dammsan> how about trade the pirate? =)
12:26 #periperi: < dammsan> i think you want to avoid trademark violation
12:26 #periperi: < dammsan> which makes sense
12:27 #periperi: < dammsan> https://en.wikipedia.org/wiki/Torx
12:27 #periperi: < dammsan> apparently torx is trademarked
12:27 -!- Irssi: Starting query in freenode with dammsan
12:27 <geertu> http://events.linuxfoundation.org/sites/events/files/slides/stable.pdf doesn't list 3.10 under LTS
12:27 #periperi: < dammsan> but i will now stop to make everyone worried
12:28 #periperi: < dammsan> geertu: but that's from oracle
12:29 #periperi: < geertu> I think we're ready anyway, right?
12:29 #periperi: < geertu> Anything we missed?
12:29 #periperi: < dammsan> nothing from my side
12:30 #periperi: < geertu> OK, thanks, and have a nice day/evening/night/...
12:31 #periperi: < morimoto> thanks
12:31 #periperi: < dammsan> thank you guys!
12:31 #periperi: < shimoda> thank you!