[Solved] Asus TB S

This is for any ARMv7 device that we do not officially support.

Re: Asus TB S

Postby TheSaint » Tue Aug 28, 2018 9:53 am

The u-boot is the previous post configuration file and all armian patches for u-boot. So, I'd propose to rearrange the autoboot delay to 2 seconds, instead of the current 5.
The /boot is like this:
Code: Select all
# ls /boot/*
/boot/boot.scr  /boot/initramfs-linux.img  /boot/zImage
/boot/boot.txt  /boot/mkscr

/boot/dtbs:
rk3288-tinker.dtb  rk3288-tinkerS.dtb

Perhaps the first fdt file is less important, but it may be needed for the non S model.

I'm trying to arrange some personalization....
TheSaint
 
Posts: 345
Joined: Mon Jul 23, 2018 7:57 am

Re: Asus TB S

Postby summers » Tue Aug 28, 2018 10:07 am

If you look at the begining of "dmesg" it will say if you have loaded the tinker or tinkerS device tree blob. (Or indeed miniarm)

If the tinkerS (which i expect it will be) - then I'll do that patch first ...
summers
 
Posts: 868
Joined: Sat Sep 06, 2014 12:56 pm

Re: Asus TB S

Postby TheSaint » Tue Aug 28, 2018 11:15 am

It seems that there's no explicit mentions
Code: Select all
# dmesg |grep -i tinker
[    0.000000] OF: fdt: Machine model: Rockchip RK3288 Tinker Board

But I'm almost sure that it will fail without it.

Note to correct the command line as follow:
Code: Select all
console=ttyS2,115200n8 root=PARTUUID=xxxxxxxxx quiet rw rootwait

Of course the PARTUUID is automatically guessed.

P.S. how to build a ftd file ?
TheSaint
 
Posts: 345
Joined: Mon Jul 23, 2018 7:57 am

Re: Asus TB S

Postby summers » Tue Aug 28, 2018 8:30 pm

OOps, just chect, my dtb has
Code: Select all
model = "Rockchip RK3288 Tinker Board";
I'll have to change it.

As you just had the two files in /bot/dtbs it can't have been rk3288-tinker.dtb - as that wouldn't have been up what you found. So it must have been my changes. Pity it comes up with error when no sd card, I'll think about that. I'll do another dtb, that gets the model right - but no other changes, so no need to change.

Oh yes, to change dts to dtb and back again install "dtc" and use e.g:

Code: Select all
dtc -I dtb -O dts < rk3288-tinkerS.dtb > rk3288-tinkerS.dts
dtc -I dts -O dtb < rk3288-tinkerS.dts > rk3288-tinkerS.dtb

Me though, becuase the main dts in the kernel includes thing around, I set up dts in
Code: Select all
linux/arch/arm/boot/dts/

Then make with
Code: Select all
make dtbs

IIRC "dtc" is also in the kernel source, IIRC down in tools ...

To my mind it easier to work from
Code: Select all
linux/arch/arm/boot/dts/
as it has comments in ...
Last edited by summers on Wed Aug 29, 2018 9:33 am, edited 2 times in total.
summers
 
Posts: 868
Joined: Sat Sep 06, 2014 12:56 pm

Re: Asus TB S

Postby TheSaint » Tue Aug 28, 2018 9:24 pm

summers wrote:Bity it comes up with error when no sd card, I'll think about that

I just thought that it could stay inactive to check the SD card presence. Then only to be triggered by a voltage level on insertion, such as hotplug would do.
Perhaps is what it's doing, but it uses a big verbosity.

I think I'll try a brand new u-boot, to suite my needs and perhaps I'll see what the patches doing in dtbs. So we may consider to have the fix in the tinkerS.dtb , Which is probably has been done.
TheSaint
 
Posts: 345
Joined: Mon Jul 23, 2018 7:57 am

Re: Asus TB S

Postby TheSaint » Wed Aug 29, 2018 6:49 am

Today I redone a brand new u-boot, without the network support. I never had need for such booting mode.
It's working fine, but I honestly confess I don't know what it makes things going. So perhaps we should try one of your recipes and see whether it goes up to the login.

One more new, the messages of joggling are gone, now it's calm even in the UART2 login, without the SD card :)
So by my position, I might declare it solved, but you'll have the power to finalize one for the mainline.

I'm going to try the last update, with the kernel version 4.18 :D
TheSaint
 
Posts: 345
Joined: Mon Jul 23, 2018 7:57 am

Re: Asus TB S

Postby summers » Wed Aug 29, 2018 11:15 am

Oh yes - on u-boot did you try my version that enabled scanning of both the eMMC and the sd card? Would be useful to know how far it goes. Alas I suspect mainlining all of armian patches looks a challenge to me. Instead focus of the minimum changes needed so that u-boot can boot from either the eMMC or the sd card.

On the sd card errors when there is no sd card installed. Yes this should be handled with the CD = Card Detection. Now its not clear if this is enabled on the TB S - the schematic http://dlcdnet.asus.com/pub/ASUS/mb/Linux/Tinker_Board_2GB/Tinker_Board_Schematics.zip doesn't seem to show the connection to the sd card! Pity as then we could see what is wired. There is a standard way to describe CD given here https://github.com/torvalds/linux/blob/master/Documentation/devicetree/bindings/mmc/mmc.txt - we could force it to do polling, but that may be what is doing right now. On the other hand it may be the driver adjusting the phasing: https://github.com/torvalds/linux/blob/master/Documentation/devicetree/bindings/mmc/rockchip-dw-mshc.txt - which would explain why it quietens down after a bit. Le me know if you want me to try out the device tree.

On the device tree, the maining kernel and debian are slightly different. The mainline kernel just has:
Code: Select all
&sdmmc {
   bus-width = <4>;
   cap-mmc-highspeed;
   cap-sd-highspeed;
   card-detect-delay = <200>;
   disable-wp;         /* wp not hooked up */
   pinctrl-names = "default";
   pinctrl-0 = <&sdmmc_clk &sdmmc_cmd &sdmmc_cd &sdmmc_bus4>;
   status = "okay";
   vmmc-supply = <&vcc33_sd>;
   vqmmc-supply = <&vccio_sd>;
};

Whereas Debian instead has
Code: Select all
&sdmmc {
   bus-width = <4>;
   cap-mmc-highspeed;
   cap-sd-highspeed;
   sd-uhs-sdr12;
   sd-uhs-sdr25;
   sd-uhs-sdr50;
   sd-uhs-sdr104;
   card-detect-delay = <200>;
   disable-wp;         /* wp not hooked up */
   num-slots = <1>;
   pinctrl-names = "default";
   pinctrl-0 = <&sdmmc_clk &sdmmc_cmd &sdmmc_cd &sdmmc_bus4>;
   status = "okay";
   supports-sd;
   vmmc-supply = <&vcc33_sd>;
   vqmmc-supply = <&vccio_sd>;
   sd-uhs-sdr104;
};

So its just a lot more verbose about the modes that the sd card slot can handle. Anyway aside from disabling write protection, neither does anything unusual with card detect ...
summers
 
Posts: 868
Joined: Sat Sep 06, 2014 12:56 pm

Re: Asus TB S

Postby TheSaint » Wed Aug 29, 2018 11:47 am

The message is exceeding the permitted length.

Anyway, summarizing, the test fails to boot from eMMC. Also it's trying to boot the kernel even when plugged to PC, which it should just behave as a SCSI interface, as it's doing with my u-boot.
TheSaint
 
Posts: 345
Joined: Mon Jul 23, 2018 7:57 am

Re: Asus TB S

Postby summers » Wed Aug 29, 2018 12:48 pm

Yes - the u-boot just totally failing to see the eMMC. I'll dig some more when time allows. Probably though first I'll try doing the kernel device tree patches.
summers
 
Posts: 868
Joined: Sat Sep 06, 2014 12:56 pm

Re: Asus TB S

Postby TheSaint » Thu Aug 30, 2018 6:40 am

I have a setup that works with my Tinker Board S. There's just a little personalization that exclude to boot via PXE.
So I think we have all the ingredient and tested results to whip up a good release candidate :)
I wish that volunteers will come to give a test and support.

Regarding the patches, I think is a bit of study what they are tweaking.

EDIT

Bad news. The messages are just silenced when I put command to the kernel line. But still happen, but it's non continuous.
The kernel doesn't detect USB plugging.
Last edited by TheSaint on Sat Sep 01, 2018 3:26 am, edited 1 time in total.
TheSaint
 
Posts: 345
Joined: Mon Jul 23, 2018 7:57 am

PreviousNext

Return to Community Supported

Who is online

Users browsing this forum: No registered users and 2 guests