From 55e3b2f45880faaf06f3c660ca9e8a6d9aa14bce Mon Sep 17 00:00:00 2001 From: Kuninori Morimoto Date: Mon, 9 Dec 2019 15:29:52 +0900 Subject: wiki: Porting wiki: Porting Chat Log Signed-off-by: Kuninori Morimoto --- wiki/Chat_log/20190523-io-chatlog | 159 ++++++++++++++++++++++++++++++++++++++ 1 file changed, 159 insertions(+) create mode 100644 wiki/Chat_log/20190523-io-chatlog (limited to 'wiki/Chat_log/20190523-io-chatlog') diff --git a/wiki/Chat_log/20190523-io-chatlog b/wiki/Chat_log/20190523-io-chatlog new file mode 100644 index 0000000..77f60b7 --- /dev/null +++ b/wiki/Chat_log/20190523-io-chatlog @@ -0,0 +1,159 @@ +09:02 < wsa> let's start with the IO meeting then +09:02 < jmondi> Hello, good morning +09:02 < wsa> usual status updates +09:02 < wsa> A - what have I done since last time +09:02 < wsa> ------------------------------------ +09:02 < wsa> Kaneko-san +09:02 < wsa> : posted v3 of update calculation formula for Gen3 Thermal +09:02 < wsa> Niklas +09:02 < wsa> : investigated Geert's comments on SDHI PM issue and came to an agreement +09:02 < wsa> Shimoda-san +09:02 < wsa> : sent patches for USB and SDHI with IOMMU support, reviewed a lot of USB +09:02 < wsa> patches as well as SDHI and SCIF patches +09:02 < wsa> Simon +09:02 < wsa> : posted v3 of H3 IPA support and dynamic power coefficient patches, tested +09:02 < wsa> v3 of update calculation formula for Gen3 Thermal +09:02 < wsa> Ulrich +09:02 < wsa> : sent RAVB patch to implement MTU change while device is up and investigated +09:02 < wsa> review comments +09:02 < wsa> Wolfram +09:02 < wsa> : sent and merged better API for i2c_new_device and i2c_new_dummy, sent a +09:02 < wsa> watchdog cleanup series, resent DA9063 cleanup patches after rc1, tried to +09:02 < wsa> reproduce an OOPS, upported patch avoiding SDHI CRC error, reviewed SDHI and +09:02 < wsa> SCIF and I2C slave support patches, submitted talk proposal for plumbers +09:02 < wsa> B - what I want to do until next time +09:02 < wsa> ------------------------------------- +09:02 < wsa> Kaneko-san +09:02 < wsa> : wants to keep at update calculation formula for Gen3 Thermal +09:02 < wsa> Shimoda-san +09:02 < wsa> : wants to continue discussions about SDHI with IOMMU support, Gen2 ethernet +09:02 < wsa> driver, SCIF with DMAC issue, send patches for USB, remove some meanwhile +09:02 < wsa> obsolete USB stuff and revise USB2 phy nodes for Gen3 +09:02 < wsa> Simon +09:02 < wsa> : wants to send v4 of H3 IPA support and dynamic power coefficient patches, +09:02 < wsa> and follow up on RAVB on E3/D3 Errata +09:02 < wsa> Ulrich +09:02 < wsa> : wants to send v2 of RAVB patch to implement MTU change while device is up +09:02 < wsa> Wolfram +09:02 < wsa> : wants to upport new SDHI HS400 patch, start converting i2c_new_* users, +09:02 < wsa> handle 'arbitration lost' better with the IIC core, investigate SDR104 +09:02 < wsa> regression with SDIO +09:02 < wsa> C - problems I currently have +09:02 < wsa> ----------------------------- +09:02 < wsa> Niklas +09:02 < wsa> : needs review tags for his SDHI PM patch +09:02 < wsa> Wolfram +09:02 < wsa> : had no success in reproducing the OOPS so far +09:02 < wsa> let me know if it needs updates +09:03 < wsa> neg, geertu: can you give me a short summary of your agreement? Were the things Geert was seeing another issue? +09:04 < wsa> kbingham: you there already? +09:05 < neg> wsa: In short, geertu noticed more clock imbalances which turned out to be caused by other drivers (and non sdhi related clocks) +09:05 < geertu> wsa: my suspend script changed pm_qos_resume_latency_us, which caused some issues +09:05 < geertu> Changing pm_qos_resume_latency_us used to be needed, but apparently it now works without +09:05 < geertu> hence the issues caused by it are gon +09:05 < geertu> s/gon/gone/ +09:06 < geertu> So while SDHI is fine, there's another issue in MMCIF +09:06 < wsa> neg: maybe it is a good idea to resend the patch then, and geert gives a tag for that resend +09:06 < wsa> should make it more visible to Ulf +09:07 < neg> geertu: thanks again for your effort working with this patch +09:07 < wsa> geertu: only with APE6 or with Gen2 as well? +09:07 < neg> wsa: OK, sounds good I will resend the patch +09:07 < geertu> wsa: Only on older SH/R-Mobile +09:08 < wsa> Does it look complicated? +09:08 < geertu> MMC is always complicated ;-) +09:08 < wsa> :D +09:09 < neg> +1 ;-) +09:10 < neg> Clocks that show imbalance are mmcif and sys-dmac +09:10 < wsa> Well, if it is kinda simple, I'd like to ask neg to handle it somewhen in his base time. If it is complicated, then I think it is not worth it for that old platform. +09:10 < neg> I only looked into mmcif, and I'm not sure it's a issue and might even be expected behavior +09:11 < neg> Did not dig into sys-dmac after confirmed it was not related to the SDHI driver +09:12 < wsa> I think Geert should have an opinion if that is expected behaviour? ;) +09:12 < neg> I added it to my list of tasks to potentially work on as it's old boards. The SDHI fix seemed to be more important as it also effects Gen3 +09:12 < wsa> But I let you decide if it is worth the effort... +09:12 < wsa> I'm all happy that the SDHI patch works as is :) +09:13 < wsa> Thanks for working on that (with all the APE6 trouble), Niklas and Geert. +09:13 < wsa> ok +09:14 < geertu> the sys-dmac imbalance is on gen3 +09:14 < wsa> kbingham seems not here yet, I wanted to ask him about potential preferences for i2c_new_* conversion +09:14 < wsa> will do that later today +09:14 < wsa> any questions / comments from your side? +09:15 < Marex> yep +09:15 < geertu> I'll have a deeper look at the sys-dmac issue +09:15 < Marex> I just learnt this https://git.kernel.org/pub/scm/linux/kernel/git/horms/renesas-bsp.git/commit/?h=v4.14.75-ltsi/rcar-3.9.5.rc2&id=a0d396ede95a55a4dff6aa15ea314a3d35e2e842 patch might be limited to M3W only +09:15 < Marex> just FYI +09:16 < wsa> argh +09:16 < wsa> all of M3-W? +09:16 < horms> ouch! +09:16 < Marex> wsa: I mean, I hope I wasnt off topic and that was a general question/comment call +09:16 < wsa> "might" means we still need confirmation from the HW team? +09:17 < wsa> definately on topic +09:17 < Marex> wsa: I submitted similar patch to U-Boot, but Goda-san told me REL is working on confirming it on other SoCs and that it's confirmed on M3W only for now +09:17 < wsa> good to know +09:17 < Marex> wsa: that is all I know +09:17 < kbingham> wsa, Hola +09:17 < kbingham> Sorry - just saw your message :) +09:17 < Marex> wsa: I said I'll wait for further information +09:17 < kbingham> Oh - it wasn't even that long ago :) +09:18 < Marex> wsa: will keep you updated +09:18 < wsa> I'll postpone my todo-item for that patch, then +09:18 < wsa> Marex: thanks for the heads up! +09:18 < wsa> hi kbingham! +09:18 < kbingham> wsa, Heya! +09:18 < kbingham> Seems I forgot today was meeting day :) +09:18 < wsa> kbingham: dunno if you saw it but i2c_new_dummy and _device now return errnos +09:18 < Marex> wsa: feels like E3 might also be affected +09:18 < wsa> you once requested that ont he I2c list, too +09:18 < kbingham> wsa, I reviewed the patches didn't I :) +09:18 < Marex> wsa: but that's just my guess +09:18 < wsa> right +09:19 < wsa> Marex: ok, will wait for official confirmation from Renesas +09:19 < shimoda> Marek: all Gen3 SoCs has the HS400 correction issue. but we can look the issue on M3-W easily +09:19 < shimoda> for some reason +09:20 < wsa> kbingham: in that old mail, you said you were also interested in i2c_new_secondary_device() returning errno. Is that still needed? +09:20 < shimoda> anyway, the correction feature is corrupt on the SoCs, BSP team made a kernel patch that Marek-san shows +09:20 < kbingham> wsa, Hrm... new me doesn't remember what long ago me said - I'll go see. +09:21 < wsa> kbingham: the thread was "i2c_new_{secondary_device,dummy,device}() return type." +09:21 < Marex> shimoda: sooooo, how shall we proceed ? :) +09:21 < wsa> shimoda: what would be your suggestion then? Should I upport that patch for Linux or wait until we have more information? +09:22 < Marex> wsa: +1 +09:23 < kbingham> wsa, Aha, wow - old thread :) I do indeed make a lot of use of i2c_new_secondary_device() so Yes, i would also like to see that be converted. +09:23 < shimoda> Marex: wsa: sorry for lack conclution. we should upport the patch for Linux. +09:23 < kbingham> I think the number of users of i2c_new_secondary_device() are probably low enough that it could be converted directly ? +09:23 < wsa> kbingham: yes +09:23 < kbingham> (now that we have the return types on the other apis +09:23 < kbingham> ) +09:23 < kbingham> Is that something you would like me to tackle ? +09:24 < kbingham> Seems you have now done the blocking work that was necessary from my original request. +09:24 < wsa> kbingham: there a lot of ways to start, so I wanted to ask you which would be most benefitial to you :) +09:24 < wsa> kbingham: I'll do it +09:24 < Marex> shimoda: is that OK with Goda-san too ? +09:24 < wsa> shimoda: thanks for the conclusion +09:25 < wsa> shimoda: I will work on upporting the patch, then. It needs a bit of preparational refactoring first, anyhow +09:25 < kbingham> wsa, My usage is in "drivers/media/i2c/adv748x/adv748x-core.c:186"" +09:25 < shimoda> Marex: let me talk with goda-san now. please wait for a while +09:26 < kbingham> And it seems there are only 5 users of the call ... so it should be an easy conversion. +09:26 < wsa> yup +09:26 < Marex> shimoda: thank you :) +09:26 < wsa> "adv748x_initialise_clients" +09:27 < wsa> ? +09:27 < wsa> is that OK in English? +09:27 < kbingham> "Initialise the client" with an adv748x prefix ? +09:28 < Marex> kbingham: s/s/z/ probably ? +09:28 < wsa> i thought it must be "initialize" +09:28 < pinchartl> wsa: British English vs. American English :-) +09:29 < kbingham> https://simple.wikipedia.org/wiki/British_English +09:29 < geertu> wsa: That's US English, isn't it? +09:29 < kbingham> Some words in British English use "-ise" where "-ize" is used in American English. Spelling them with "-ize" is also done in Britain sometimes. However, this is very much frowned upon and would not be considered acceptable if writing for a British audience. +09:29 < kbingham> :) +09:29 < wsa> Good, so I learned something today +09:29 < pinchartl> kbingham: I'm afraid that function names in the kernel use american english... +09:29 < kbingham> (I had that page up already while writing a review to Neg :D heheh) +09:29 < Marex> kbingham: but kernel is written in simplified English, so -ize ? +09:29 * geertu realizes Trump wants a Brekzit +09:30 < geertu> simplified -> "init" +09:30 < pinchartl> you could make everybody happy with adv748x_init_clients() +09:30 < shimoda> Marex: wsa: I'm sorry. my understanding is wrong. As goda-san said, we should wait for hw team conclusion because hw team still investigate this root cause. So, our upstream team also should wait. +09:30 < wsa> shimoda: fine, too. will do +09:30 < Marex> shimoda: will do, thank you for checking :-) +09:31 < shimoda> Marex: wsa: thanks! +09:31 < wsa> ok, I think we are done with IO -- cgit v1.2.3