1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
|
09:22 < geertu> Welcome to today's Core Group Chat Meeting!
09:22 < geertu> Agenda:
09:22 < geertu> 1. Status Updates
09:22 < geertu> 2. Discussion Topics
09:23 < geertu> Topic 1. Status updates
09:23 < geertu> A) What have we done since last time:
09:23 < geertu> Marek worked on OpenOCD (patch resync), U-Boot (SDHI performance,
09:23 < geertu> Blanche and SMC911x DM conversion), and Linux (PCIe suspend).
09:23 < geertu> Shimoda-san submmited rcar-usb2-clock-sel patches, and discussed the
09:23 < geertu> plan for the R-Car V3U bootloader.
09:23 < geertu> Ulrich sent v2 of the ignore-unused clock series, and reviewed patches.
09:23 < geertu> Geert converted the CPG/MSSR DT bindings to json-schema, upported
09:23 < geertu> CPUIdle for R-Car M3-N and E3, sent a second batch of pull requests for
09:23 < geertu> v5.7, and reworked the GPIO Aggregator.
09:24 < geertu> B) What we plan to do till next time:
09:24 < geertu> Marek plans to continue SMC911x work in U-Boot for Blanche.
09:24 < geertu> Shimoda-san plans to convert DT binding docs to json-schema, and will
09:24 < geertu> continue internal discussions about BL31 code for R-Car V3U.
09:24 < geertu> Geert will post v6 of the GPIO Aggregator, and hopes to resume QEMU GPIO
09:24 < geertu> virtualization.
09:25 < geertu> C) Problems we have currently:
09:25 < geertu> Shimoda-san regrets not being able to share the R-Car V3U Hardware
09:25 < geertu> User's manual.
09:25 < geertu> Covid-19 has reached EuroPeri space.
09:25 < geertu> ---EOL---
09:25 < geertu> Anything I missed?
09:26 < geertu> shimoda22: What is ICUMXA?
09:27 < shimoda22> This is a CPU core which R-Car V3H has (section 68A)
09:27 < shimoda22> IICU, this is RH850 CPU core so that we don't have any free compiler..
09:29 < geertu> Thx, forgot to search in the datasheet ;-) Google didn't really know
09:29 < geertu> RH850 is different from V850?
09:31 < shimoda22> perhaps same. V850 is NEC and RH850 is Renesas
09:32 < uli_> there's a surprisingly long wikipedia article on that: https://en.wikipedia.org/wiki/V850, including a section on "ancient prom writers" :)
09:32 < geertu> yep
09:32 < geertu> and gcc has v850 support
09:33 < Marex> Variant(s) V850 Family,
09:33 < Marex> RH850 Family
09:33 < Marex> well ...
09:34 < shimoda22> Oh, I didn't know gcc supports RH850.
09:35 < geertu> perhaps not the latest variant
09:35 < geertu> And it might have been removed in recent gcc versions
09:35 < wsa> Marex: I don't see this patch in next "[RFC] PCI: rcar: Fix incorrect programming of OB windows" although you reviewed it. Could you give a ping?
09:36 < geertu> At least gcc-9.1.0 still has v850-specific options
09:36 < geertu> Now, is that of any help? ;-)
09:39 < shimoda22> I don't know :) especially, the hardware manual doesn't say ICUMXA is related to RH850.
09:40 < Marex> wsa: sure, although I suspect that if Lorenzo took forever to review patches before this current mess, he would just ignore patches by now
09:41 < shimoda22> about pci, i reviewed other patches about pcie endpoint support.
09:41 < wsa> Marex: I see. Well, might be worth pointing out that the patch is not RFC anymore
09:42 < wsa> shimoda22: yes, great! I just updated this task in periject :)
09:42 < Marex> wsa: what surprises me more is that the original author is literally sitting with Lorenzo in the same office
09:42 < Marex> oh well
09:43 < Marex> wsa: I guess that unit of ARM is a mess
09:43 < wsa> well, then pinging Andrew might make more sense? ;)
09:44 < Marex> wsa: um, that's the author of the patch, you already asked me to ping him again
09:44 < kbingham> wsa, Is that Andrew Murray?
09:44 < kbingham> wsa, Andrew Murray left ARM.
09:44 < wsa> Marex: Initially, I was aiming for the maintainer, but if the author sits next to him, the latter might make more sense...
09:45 < wsa> ... that is, unless he left the company
09:45 < wsa> kbingham: where to?
09:45 < Marex> kbingham: didn't he join like a few months ago ?
09:46 < Marex> morimoto: btw thank you for the hint to buy instant ramen, I bought a large-ish crate
09:46 < morimoto> Hehe :) no worry
09:46 < kbingham> wsa, He took some external contract or something.
09:47 < kbingham> wsa, I can hire him if you (or rather Renesas) want me to get him to work directly on Renesas :-)
09:48 < wsa> kbingham: does he have any experience in I2C dynamic address assignment? :D
09:48 < kbingham> Marex, Well techincally he 're-joined' ... I don't know why he didn't want to stay there. Maybe he preferred being independent.
09:49 < kbingham> wsa, I think he's more experienced in PCI topics :) But I'm sure he's done I2C work too :)
09:49 < geertu> Topic 2. Discussion Topics
09:50 < geertu> Any other topics to discuss?
09:50 < wsa> kbingham: it is basically the same anyhow
09:50 < kbingham> wsa, :-D
09:50 < wsa> uli_: any idea why my testing of ignore-unused clocks fail?
09:51 < geertu> wsa: Because the clock was not enabled in the first place at all?
09:51 < wsa> geertu: I started it in U-Boot
09:51 < wsa> If I don't hack the code to bail out and forbid the kernel to ping the watchdog, it will reboot
09:52 < geertu> wsa: OK.
09:53 < geertu> Looking at drivers/clk/clk.c, CLK_IGNORE_UNUSED just prevents disabling the clock.
09:53 < wsa> that's what we want
09:53 < wsa> IIUC
09:53 < geertu> But only in the clk_disable_unused late_initcall
09:54 < geertu> it can still be disabled elsewhere
09:54 < geertu> I.e. manually, or through Runtime PM?
09:54 < geertu> WDT probe fails -> detached from PM Domain -> clock disabled?
09:54 < uli_> what i have checked is whether it will be temporarily disabled when the probing is deferred
09:54 < uli_> and that works
09:54 < wsa> yup
09:54 < wsa> uli_: ok
09:55 < geertu> works == not temporarily disabled?
09:55 < uli_> yes
09:57 < geertu> Looks like cpg_mstp_clock_endisable() needs some debug code to track what happens to the WDT clock
09:57 < patersonc> wsa: geertu: Fabrizio left us in December. But in a twist of fate he's actually going to be re-joining us b/April. Huzzah!
09:58 < wsa> uli_: can you do that?
09:58 < uli_> sure
09:58 < patersonc> But this time he'll have a @renesas.com email address instead of a @bp.renesas.com address :)
09:58 < wsa> uli_: thanks!
09:58 < wsa> patersonc: leaving and rejoining is a thing these days? :)
09:59 < patersonc> wsa: It's a long complicated story!
09:59 < geertu> patersonc: Great! Looking forward to seeing a patch for .mailmap
09:59 < geertu> wsa: I'll rejoin public life in 5 weeks or so (or not)
10:00 < wsa> patersonc: anyway, glad he will be back!
10:00 < wsa> geertu: isn't it more like public life will rejoin us?
10:00 < patersonc> Same here
10:01 < geertu> Let's finish the Core meetinng with this great news. Unless someone has something to add?
10:02 < patersonc> Sorry to interrupt!
10:02 < geertu> patersonc: np, we like positive news
10:03 < geertu> Thanks for joining, and have a nice continued day!
|
