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
107
108
109
110
111
112
113
|
09:03 < wsa_> welcome :)
09:03 < wsa_> here is the condensed summary, please double check:
09:03 < wsa_> A - what have I done since last time
09:03 < wsa_> ------------------------------------
09:03 < wsa_> Geert
09:03 < wsa_> : upported MSIOF DMA completion improvement and fixed SCIF race conditions
09:03 < wsa_> during shutdown.
09:03 < wsa_> Kaneko-san
09:03 < wsa_> : got PCIE support for M3-W and M3-N merged.
09:03 < wsa_> Niklas
09:03 < wsa_> : resent a patch series to improve SDHI tuning, started reworking the BSP
09:03 < wsa_> SDHI reset patches, and started looking into DMA max seg size for SDHI.
09:03 < wsa_> Shimoda-san
09:03 < wsa_> : added USB3 role switch support, sent a bugfix for the SDHI DMA RX case,
09:03 < wsa_> implemented USB support for r8a77990, and investigated a SCIF overrun
09:03 < wsa_> issue.
09:03 < wsa_> Simon
09:03 < wsa_> : got SDHI HS400 support merged.
09:03 < wsa_> Ulrich
09:03 < wsa_> : sent first version of D3 MSIOF support and researched origins of SCIF
09:03 < wsa_> REIE handling
09:03 < wsa_> Wolfram
09:03 < wsa_> : fixed a problem in and improved R-Car support for I2C bus recovery and
09:03 < wsa_> improved wiki documentation for it, sent a new version of the I2C RXDMA
09:03 < wsa_> workaround patch, guided the SCCB redesign including needed preparations,
09:03 < wsa_> fixed a regression introduced with the I2C fault injector and resent a
09:03 < wsa_> new fault injector, and looked into a SDHI DMA mapping problem.
09:03 < wsa_> Before that, he attended ALS2018 and held a talk about DMA safe buffers.
09:03 < wsa_> B - what I want to do until next time
09:03 < wsa_> -------------------------------------
09:03 < wsa_> Kaneko-san
09:03 < wsa_> : wants to upport M3-N SATA and D3 MSIOF.
09:03 < wsa_> Niklas
09:03 < wsa_> : wants to post patch for the now started work and continue with
09:03 < wsa_> SDHI tuning improvements.
09:03 < wsa_> Shimoda-san
09:03 < wsa_> : wants to make a plan for USB role switch on E3 despite the connector
09:03 < wsa_> issue, continue with the SCIF issue if needed, add PWM for E3, and
09:03 < wsa_> continue to learn QEMU with USB.
09:03 < wsa_> Simon
09:03 < wsa_> : wants to work CPG and DTS changes for HS400.
09:03 < wsa_> Wolfram
09:03 < wsa_> : wants to merge the I2C recovery and RXDMA work, keep guiding the SCCB
09:03 < wsa_> refactoring, review Niklas SDHI patch series, and check if the SDHI
09:03 < wsa_> NON_REMOVABLE workaround is still needed.
09:03 < wsa_> C - problems I currently have
09:03 < wsa_> -----------------------------
09:03 < wsa_> Niklas
09:03 < wsa_> : mentions that figuring SDHI testing slows him down.
09:03 < wsa_> Shimoda-san
09:03 < wsa_> : needs more info from the HW team to make renesas_usbhs work again.
09:03 < wsa_> Wolfram
09:03 < wsa_> : is so busy with all the other tasks that QEMU development is lagging.
09:06 < wsa_> I have a question for shimoda about the SCIF issue and if I got the current status right
09:06 < wsa_> We all did some research about the REIE and TOIE bit and you reported back our findings to the BSP team
09:06 < wsa_> and we are waiting now if this is enough for them
09:07 < wsa_> Is this correct or am I forgetting something?
09:07 < wsa_> Do we need to work more on this?
09:08 < wsa_> shimoda: ? (let's try the IRC notification thingie)
09:09 < wsa_> Other than that, IO group is promised an E3 board
09:09 < neg> Nice
09:10 < wsa_> Since E3 is mainly DTS enablement which we agreed to be worked on by Kaneko-san, I voted to have it sent to Simon
09:10 < wsa_> neg: core and M/M will get one, too
09:10 < wsa_> so he can test the patches
09:11 < wsa_> horms: is hat okay with you?
09:11 < horms> sure!
09:12 < wsa_> nice
09:12 < shimoda> wsa_: sorry for the delayed. about REIE and TOIE, this is enough for now.
09:12 < wsa_> shimoda: thanks!
09:14 < shimoda> by the way, i'd like to get some opinions about SCIF console overrun issue
09:14 < shimoda> i sent an email at 2018-07-09 pm 2:48 JST
09:15 < geertu> shimoda: This is input overrun, right?
09:15 < shimoda> geertu: yes
09:16 < geertu> I'm afraid that with flow control disable, there's not much we can do.
09:16 < geertu> s/disable/disabled/
09:18 < shimoda> geertu: thanks for your opinion! does this mean this is not a bug on the driver?
09:19 < geertu> shimoda: Well, printing kernel messages on the serial console is known to interfere with normal serial operation.
09:19 -!- ulih [~Mutter@x2f7fd43.dyn.telefonica.de] has joined #periperi
09:20 < geertu> To make it perfectly safe without any interference would mean loosing most of its value (kernel messages may not come through)
09:20 * geertu wonders if ulih is the SCIF expert?
09:20 < ulih> morning. network outage, using my wife’s iphone
09:21 < ulih> what is the matter?
09:21 < geertu> ulih: We were just talking about the SCIF console overrun issue
09:23 < wsa_> shimoda-san wrote an email about it on 2018-07-09 pm 2:48 JST
09:23 < wsa_> but geert already mentioned that "kernel messages on the serial console is known to interfere with normal serial operation"
09:23 < wsa_> and "I'm afraid that with flow control disable, there's not much we can do."
09:24 < ulih> ok. thanks for filling me in
09:25 < wsa_> while ulih thinks about it, are there any other topics from your side?
09:26 < wsa_> I am done with the "bigger" topics
09:26 < wsa_> details about upporting will be done 1:1 privately
09:26 < wsa_> vaishali: I really hope to have the SDHI check done this week
09:27 < geertu> Has anyone tried QEMU+KVM and I/O pass-through (vfio) with a DMA-capable device directly connected to the IPMMU (i.e. not via SYS-DMAC)?
09:27 < vaishali> wsa_: Thanks. :)
09:27 < wsa_> but I'd think the other topics do not depend on it, or?
09:27 < wsa_> geertu: nope
09:28 < Marex> wsa_: two more PCIe fixes got in
09:28 < Marex> wsa_: you wanted to ask about the 32bit limitation of the PCIe at some point
09:28 < dammsan> geertu: i don't see how the guest dma map/unmap operations would affect the host that has the IPMMU driver?
09:28 < wsa_> Marex: nice
09:29 < wsa_> Marex: yup, we missed to talk about this in Japan :( Seems we need to do it by email
09:29 < dammsan> geertu: i have not tried it, but i dismissed it before doig so =)
09:29 < wsa_> I will do this later
09:30 < Marex> wsa_: the thread (Re: [PATCH][RFC] PCI: rcar: Add bus notifier so we can limit the DMA range) died on the 6th of June, sadly
09:30 < Marex> wsa_: I can poke the thread if you think it makes sense, although I don't have anything to contribute without really digging in
09:31 < geertu> dammsan: I hoped it would just work out-of-the-box, as IOMMU and pass-through should work on other systems
09:31 < geertu> Marex: I can assign "ARM64,?,noplan,?,Handle 32-bit DMA limitation (https://patchwork.kernel.org/patch/10416861/)
09:31 < geertu> " to you with a single vim edit ;-)
09:32 < geertu> If you're interested
09:32 < wsa_> geertu: how about I give you the mic and we talk about it in the core meeting? :)
09:32 < geertu> wsa_: Sounds like a good idea!
09:32 < wsa_> good!
09:32 < dammsan> geertu: maybe reproducing a x86-based environment with vfio might bbbe good?
09:32 < wsa_> thank you all for this meeting!
|
