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
|
09:04 < wsa_> so, let's start the IO meeting
09:05 < wsa_> welcome everyone
09:05 < wsa_> let's start with the summary:
09:05 < wsa_> A)
09:05 < wsa_> Simon posted a new version of the HS400 patches addressing review comments and
09:05 < wsa_> started the virt+network task. Niklas sent out the first patch for variable
09:05 < wsa_> MTU size for RAVB. Shimoda-san got various USB-related bugfixes merged in
09:05 < wsa_> upstream. Ulrich fixed a SCIF lockup and converted the SCIF driver to use HR
09:05 < wsa_> timers. Wolfram reviewed a ton of patches all around, discussed some watchdog
09:05 < wsa_> isses, made preparations for OSS+AGLJ, got the IP core switcher series merged, and
09:05 < wsa_> worked on virt tasks, generic one and I2C specific one.
09:05 < wsa_> B)
09:05 < wsa_> Simon wants to look at the CLK changes needed for HS400 and will continue with
09:05 < wsa_> the virt task. Niklas wants to send out V2 of RAVB-MTU size patch and, if time
09:05 < wsa_> permits, look into his previous SDHI hack. Shimoda-san wants to work on USB3.0
09:05 < wsa_> role swap, USBPHY GPIO support, M3-N USB enablement and the virt task. Ulricht
09:05 < wsa_> will send out patches to the solutions of the above SCIF tasks. Wolfram will
09:05 < wsa_> finish the IO virt task today and fully move to the virt+I2C task, continue
09:05 < wsa_> reviewing stuff, submit the talk proposal for AGL, keep at the watchdog issues,
09:05 < wsa_> and plans to talk with Kieran about reprogrammed I2C addresses.
09:05 < wsa_> C)
09:05 < wsa_> Geert noticed eMMC problems on resume. Simon needs a SDR104 card in the Porter
09:05 < wsa_> board of Magnus' farm.
09:05 < wsa_> Please check if I misunderstood something. And fire away questions if you have.
09:06 < wsa_> I have one
09:06 < wsa_> geertu: which branch showed the eMMC problem? linus or ren-drivers?
09:06 < geertu> wsa_: That was my dev branch, based on renesas-drivers.
09:07 < geertu> Haven't tried it on anything else.
09:07 < geertu> I recently started using the eMMC for native compilation of QEMU, as it's much faster than nfsroot
09:07 < wsa_> oh, and anothso, with HS400?
09:07 < geertu> Hence I don't know if it happened before.
09:07 < wsa_> with HS400?
09:08 < geertu> I guess you have experience with mounted SD cards and s2idle/s2ram?
09:08 < geertu> No HS400, AFAIK
09:08 < wsa_> Simon once reported suspend&resume problems with SDHI, too
09:08 < wsa_> don't recall the details anymore
09:09 < wsa_> there was never time to look into it :(
09:09 < wsa_> i'll boost priority on this one a bit
09:10 < wsa_> neg: what about the hack to increase the MTU size beyond 2k? Not tried or didn't work?
09:10 < geertu> I can do some more tests, after reporting
09:10 < wsa_> geertu: Thanks, much appreciated!
09:11 < wsa_> geertu: and doing native compiles, you would probably also like the HS400 patches :)
09:11 < horms> wsa_: those s2ram/sdhi problems were a long time ago
09:11 < neg> wsa_: it did not work out as I planed :-( Not sure if it can be done, not given up entierly on it, but won't presue it in the near future. Shimoda-san told me of this HW limitation at the beginning of the task but I tought I could break it :-)
09:11 < horms> I think they have been resolved though I don't recall how
09:12 < horms> On the porter/sdr104 front, there is a card present, but I think its in the wrong slot (not all slots support that speed)
09:13 < wsa_> neg: yeah, that's understandable. Well, maybe somewhen...
09:14 < wsa_> horms: that should be easily fixable ;)
09:14 < horms> yes, its not a large problem
09:14 < horms> nor an urgent one
09:14 < horms> my larger problem is the clk patch for HS400. I think you and I need to talk about that some time.
09:15 < wsa_> yes, we can do that
09:15 < horms> I suspect you have a different eMMC chip on hour H3 ES1.0 board
09:15 < wsa_> somewhen next week
09:15 < horms> sure, lets schedule a time (some time)
09:15 < wsa_> I can't check right now, but I recall something like this, yes
09:15 < horms> ok, i guess you need to go home first
09:15 < horms> let me know when the timing is good for you
09:16 < wsa_> monday should be good
09:16 < geertu> wsa_: How long before the eMMC will wear out?
09:17 < horms> ok, i should be around on monday. but if you want to schedule a time let me know
09:17 < horms> geertu: just before we confirm the bug fix
09:17 < wsa_> horms: let's try the ad-hoc method, I'd say
09:17 < horms> ok, sur
09:18 < horms> in general mornings are better for me, say from 9 to noon
09:18 < wsa_> geertu: so, tried to wear out a 64MB SD card once because I wanted to check if SDHI catches such errors well.
09:18 < horms> but mondays are pretty good in general
09:18 < wsa_> I wrote continously for days while hacking, no luck
09:19 < wsa_> but maybe that card was not MLC
09:20 < wsa_> so, do you guys have questions?
09:20 < wsa_> I am done, I think
09:21 < wsa_> JapaPERI is happy, too?
09:22 < shimoda> yes, but i cannot catch up the emmc issue though :)
09:23 < horms> shimoda: issue is that hs400 is working fine for me. but it involves patches that wsa_ has had trouble with before. next step if for wsa_ to verify the current patchset in his environment, then discuss
09:24 < horms> It seems likely his H3 1.0 and mine have different eMMC chips. but we should test before speculating more
09:25 < shimoda> horms: thanks! and geertu has other issue (doesn't work after resume)
09:27 < kbingham> wsa_: I'm happy to talk I2C addreses whenever you're free
09:27 < shimoda> about H3 1.0, i think it is very bad chip about clock thing
09:27 < kbingham> :)
09:27 < wsa_> kbingham: nice, we will see when. hopefully next week, but no promise yet
09:27 < kbingham> wsa_: Ack :)
09:28 < horms> shimoda: ok. so some boards have a chip that works badly with the clk fix? that would explain things
09:28 < horms> do you know which chip?
09:29 < wsa_> or maybe it is because some clks are running only at 50%?
09:29 < wsa_> we need to check, but keep in mind that ES1.0 is special regarding to clocks
09:30 < geertu> Yes, it has completely different handling for SD clocks, IIRC
09:31 < geertu> You may need to measure the actual clock signals...
09:31 < wsa_> heh
09:31 < horms> I suggest we take this offline and move onto Core
09:31 < wsa_> we will come back to this once I am back home
09:31 < wsa_> agreed
09:32 < wsa_> thanks for the meeting!
09:32 < wsa_> geertu: the stage is yours
|
