Device-tree improvements

[TheSaint]

I did a primary test. This is the report.
The impressions are no much positive. I haven't yet wrote the setup permanently, as you may see I tested the starting by manually setting the commands at the u-boot shell.
The USB, still unable to detect the plug in. Also the SD card detection still joggling if non card inserted.

I'm tinkering why the TinkerOS works with rk3288-miniarm and if I try to use it will give so much hassle that will stop to boot. I suppose that they patched with a non standard manner.
The bad thing is that none is willing to come around to give a help. The only support is just within debian based tools.

[summers]

Thanks for that. Actually it does what I hoped:
1) Discovers 3 sdio devices
2) brings up mmc on two of them
3) starts the wi-fi
Now it only talked about partitions on one of the mmc locations, could you use "fdisk" followed by "p" to see the partition table on both mmc devices (eMMC and the sdcard)? (oh yes exit "fdisk" with "q" to avoid writing the partition table).

My patch to kernel botched again, needs to be done in a specific way using git - and spent over an hour on it last night - and it barfed (mangled patches). I'll try again either later this week, or next weekend. At least I've the format set up in a better way - it more maintainable with both TB and TBS on a common framework.

Oh yes, when you did the device tree modifications in uboot, did you note $this->bbcode_second_pass_quote('', 'l')ibfdt fdt_setprop(): FDT_ERR_NOSPACE

[TheSaint]

Undeniably a great job, Summers !
From my side I still shooting in the dark to understand which dft option may work.
Here the further reports you asked for.
Well last error was probably cause when I was trying to add$this->bbcode_second_pass_code('', ' fdt addr ${fdt_addr_r}
fdt set /dwmmc@ff0c0000 broken-cd
fdt set /usb@ff500000 no-relinquish-port')
then i repeated once more. But this time I used without these lines because I forgotten to copy the boot.txt content. BTW is there a way to print a file on the u-boot shell ? Similar to cat ?

Regarding the reports, you and whom may be interested, have permission of peruse them.
I hope the reports last a bit longer. Usually the host will remove them after few months or even less than one.

[summers]

yes try changing $this->bbcode_second_pass_code('', ' fdt addr ${fdt_addr_r}
fdt set /dwmmc@ff0c0000 broken-cd
fdt set /usb@ff500000 no-relinquish-port') for $this->bbcode_second_pass_code('', ' fdt addr ${fdt_addr_r}
fdt resize 1024
fdt set /dwmmc@ff0c0000 broken-cd
fdt set /usb@ff500000 no-relinquish-port
fdt resize')
That should get rid of the errors ...

And thanks for the fdisk output - yes confirms the eMMC is working as expected.

[TheSaint]

I just forgot to mention that I booted manually and copying the command to shell.

EDIT
I got this crash while attempting to use the WiFi as AP$this->bbcode_second_pass_code('', '[ 259.822523] ------------[ cut here ]------------
[ 259.827683] kernel BUG at mm/slub.c:3903!
[ 259.832157] Internal error: Oops - BUG: 0 [#1] PREEMPT SMP ARM
[ 259.838669] Modules linked in: xt_REDIRECT xt_tcpudp nf_nat_pptp nf_nat_proto_gre nf_conntrack_pptp nf_conntrack_proto_gre iptable_filter iptable_nat ipt_MASQUERADE nf_conntrack_ipv4 nf_defrag_ipv4 nf_nat_ipv4 nf_nat nf_conntrack bpfilter 8021q garp mrp stp llc snd_soc_hdmi_codec rc_cec dw_hdmi_cec dw_hdmi_i2s_audio snd_soc_simple_card snd_soc_simple_card_utils snd_soc_rockchip_i2s r8723bs(C) snd_soc_core ac97_bus snd_pcm_dmaengine rk_crypto dw_wdt rockchipdrm joydev evdev realtek snd_usb_audio pwm_rockchip analogix_dp snd_usbmidi_lib cfg80211 dw_hdmi snd_hwdep cec snd_rawmidi snd_seq_device rc_core rfkill snd_pcm dwmac_rk snd_timer stmmac_platform stmmac uio_pdrv_genirq uio rockchip_saradc sch_fq_codel crypto_user ip_tables x_tables
[ 259.911422] CPU: 3 PID: 299 Comm: RTW_CMD_THREAD Tainted: G C 4.18.8-1-ARCH #1
[ 259.920943] Hardware name: Rockchip (Device Tree)
[ 259.926200] PC is at kfree+0x190/0x1b4
[ 259.930418] LR is at nl80211_send_station+0xa6c/0xc70 [cfg80211]
[ 259.937126] pc : [<c04a80ec>] lr : [<bf1a6244>] psr: 40010013
[ 259.944124] sp : ed78fda8 ip : 00000008 fp : ed101014
[ 259.949955] r10: 00000000 r9 : ed101030 r8 : 00000580
[ 259.955787] r7 : ed78fe58 r6 : ef4e9504 r5 : ed512600 r4 : c1605d88
[ 259.963077] r3 : c1601ebc r2 : 00000024 r1 : e9c8f355 r0 : f1579000
[ 259.970366] Flags: nZcv IRQs on FIQs on Mode SVC_32 ISA ARM Segment none
[ 259.978335] Control: 10c5387d Table: 2baf806a DAC: 00000051
[ 259.984751] Process RTW_CMD_THREAD (pid: 299, stack limit = 0x295df052)
[ 259.992129] Stack: (0xed78fda8 to 0xed790000)
[ 259.996992] fda0: 00000011 ed78fe58 c1605d88 ed512600 00000011 ed78fe58
[ 260.006119] fdc0: 00000580 bf1a6244 00000005 00000013 ed101014 00000000 fffff04b c1605d88
[ 260.015255] fde0: ed101034 ed101030 00000000 00000000 00000000 78660e19 00000001 ed5c2000
[ 260.024392] fe00: 00480020 e9c8f30a ed78fe58 ed5c19a0 00000000 ed512600 fffff04b bf1a65e4
[ 260.033528] fe20: 00000000 ed5c1800 ed5c2000 e9c8f30a ed78fe58 f1579000 c1605d88 0000001c
[ 260.042664] fe40: 00000000 00000000 eabc8600 f157a450 f1579000 bf2f3660 00000000 00000000
[ 260.051800] fe60: 00000065 f1579000 00000008 c1605d88 00000001 bf2ab3f4 00000065 bf2d75ac
[ 260.060928] fe80: 00000000 ef3fb76c 8020001a eabc3c00 ffffe000 c04a7a34 00000000 00000001
[ 260.070064] fea0: 80200019 00000001 00000000 eabc3c00 00000000 f1579000 e9c8f31c 00000039
[ 260.079200] fec0: 006c6048 00060019 80200019 78660e19 f1579000 bf2d803c 00000000 ee001d00
[ 260.088336] fee0: eabc3c00 ef3fb76c ffffe000 00005e9b f157a450 f1579000 001fff00 ffffe000
[ 260.097463] ff00: 00000055 78660e19 fffff04b f1579000 f168d3b4 00000055 e9c8f300 bf2af3ec
[ 260.106599] ff20: f1579000 00000000 f157a470 c1604d00 eabc8600 bf2b99a8 f157a000 f157c000
[ 260.115726] ff40: f157a470 bf2a55f8 ee3cb9f4 60000013 ed78e000 e8888200 bf30b000 f157a460
[ 260.124854] ff60: ed590c9c ed590c80 e8ab1740 00000000 ed78e000 f1579000 bf2a5474 ea843cbc
[ 260.133990] ff80: ed590c9c c0369358 ed78e000 e8ab1740 c0369210 00000000 00000000 00000000
[ 260.143126] ffa0: 00000000 00000000 00000000 c03010e8 00000000 00000000 00000000 00000000
[ 260.152253] ffc0: 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000
[ 260.161390] ffe0: 00000000 00000000 00000000 00000000 00000013 00000000 00000000 00000000
[ 260.170555] [<c04a80ec>] (kfree) from [<bf1a6244>] (nl80211_send_station+0xa6c/0xc70 [cfg80211])
[ 260.180445] [<bf1a6244>] (nl80211_send_station [cfg80211]) from [<bf1a65e4>] (cfg80211_new_sta+0x6c/0xfc [cfg80211])
[ 260.192284] [<bf1a65e4>] (cfg80211_new_sta [cfg80211]) from [<bf2f3660>] (rtw_cfg80211_indicate_sta_assoc+0x74/0x90 [r8723bs])
[ 260.205106] [<bf2f3660>] (rtw_cfg80211_indicate_sta_assoc [r8723bs]) from [<bf2af3ec>] (rtw_stassoc_event_callback+0x1ac/0x1c4 [r8723bs])
[ 260.218989] [<bf2af3ec>] (rtw_stassoc_event_callback [r8723bs]) from [<bf2b99a8>] (mlme_evt_hdl+0x74/0xa8 [r8723bs])
[ 260.230831] [<bf2b99a8>] (mlme_evt_hdl [r8723bs]) from [<bf2a55f8>] (rtw_cmd_thread+0x184/0x3c4 [r8723bs])
[ 260.241661] [<bf2a55f8>] (rtw_cmd_thread [r8723bs]) from [<c0369358>] (kthread+0x148/0x150)
[ 260.250987] [<c0369358>] (kthread) from [<c03010e8>] (ret_from_fork+0x14/0x2c)
[ 260.259053] Exception stack(0xed78ffb0 to 0xed78fff8)
[ 260.264692] ffa0: 00000000 00000000 00000000 00000000
[ 260.273827] ffc0: 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000
[ 260.282962] ffe0: 00000000 00000000 00000000 00000000 00000013 00000000
[ 260.290351] Code: 1a000003 e5963004 e3130001 1a000000 (e7f001f2)
[ 260.297157] ---[ end trace 6234368a3866fcbe ]---')
I presume it's a wrong device-tree addressing.

EDIT2
It might be of some interest what the linaro dmesg has to say

[summers]

btw - making some progress on mainlining the tinker board s device tree. Got my first sign off from the rock chip maintainer.

May only get the eMMC in first time round, but at least thats enough to get the tinker board s to come up ...

[TheSaint]

You just gave me a great new, Summers !
I'm happy that there may be a mainlined kernel, as far as Asus pushed this board as RPi competitor, I think they are interested to cope as much as Raspberry does. Therefore, Archlinux, or other distro, should do.

[summers]

$this->bbcode_second_pass_quote('summers', 'b')tw - making some progress on mainlining the tinker board s device tree. Got my first sign off from the rock chip maintainer.

May only get the eMMC in first time round, but at least thats enough to get the tinker board s to come up ...

[TheSaint]

$this->bbcode_second_pass_quote('Summers', 'I')'ll be submitting it to the various people, in next half hour

[summers]

Interesting that it saw the card ejected at @131 - and only then went into switching speed. Says that it did at least partially understand that card was going. So have to wonder where the speed changes are coming from. Anyway eMMC and wifi comming up, which is all we can hope for.

After the TBS device tree goes mainline, I make look some more into getting wifi out there in the world. Anyway first need to understand what the maintainers will take for the wifi code ...