Jump to content

OrangePi Zero2 - Allwinner H616

Werner

By Werner
in Orange Pi Zero 2

 Share

Recommended Posts

Recommended

Posted by XFer012,

Ok, so let's see the steps I followed.   1. Installed Armbian Unstable for OrangePi Zero2: https://imola.armbian.com/dl/orangepizero2/nightly/Armbian_21.05.0-trunk.77_Orangepizero2_hirsute_edge_5.11.11.img.xz   2. Forced ethernet at 100 mbit/s, otherwise it did not work: added in /etc/rc.local ethtool -s eth0 speed 100 duplex full autoneg off   3. Downloaded hexdump's precompiled kernel (derived from jernej's 5.11.0rc1 kernel): https://github.com/hexd
Recommended by Werner
5

Reputation Points

Go to this post
dolphs

dolphs

Posted
Posted

Hello to you all,

 

Haven't read 2825 yet, will do after this but

did a quick checkout and built both for H6 (opioneplus) and H616( zero2 ), :

  

git pull

sed -i 's/orange-pi-5.11/orange-pi-5.12/g' config/sources/families/include/sunxi64_common.inc


./compile.sh EXPERT=yes EXTRAWIFI=no USEALLCORES=yes

 

compilation.log shows
--- 

grep error output/debug/*.log

output/debug/compilation.log:drivers/clk/sunxi-ng/ccu-sun50i-h6-r.c:152:27: error: redefinition of ‘sun50i_h616_r_ccu_clks’
output/debug/compilation.log:drivers/clk/sunxi-ng/ccu-sun50i-h6-r.c:178:9: error: ‘CLK_NUMBER_H616’ undeclared here (not in a function); did you mean ‘CLK_NUMBER’?
output/debug/compilation.log:drivers/clk/sunxi-ng/ccu-sun50i-h6-r.c:194:35: error: redefinition of ‘sun50i_h616_r_hw_clks’
output/debug/compilation.log:drivers/clk/sunxi-ng/ccu-sun50i-h6-r.c:226:29: error: redefinition of ‘sun50i_h616_r_ccu_resets’
output/debug/compilation.log:drivers/clk/sunxi-ng/ccu-sun50i-h6-r.c:252:36: error: redefinition of ‘sun50i_h616_r_ccu_desc’

 

 

Suppose once that packaging issue is solved edge will be bumped , so sed no longer needed.

 

cheers chaps!

 

compilation.logFetching info...

dolphs

dolphs

Posted
Posted
  On 5/16/2021 at 9:43 PM, Igor said:

Another afternoon and all kernels for 5.12.y are done at build stage.

 

- cleaning

- do we need to archive 5.11.y. I would say no since its not LTS

- testing

Expand  

 

5.11 to be buried, is EOL ( already ).

imho the non LTS kernels could be buried once mainline kernel hits .1 release. Thus eg. when 5.13.1 will be released bump edge kernel asap so 5.12 can be buried ( and forgotten )

( needless to say , but  "currents" should stay on LTS )

Igor

Igor

Posted
Posted
  On 5/21/2021 at 8:06 AM, dolphs said:

5.11 to be buried, is EOL ( already ).

Expand  


Yes, yes :D 

 

For 5.12.y -> we have to fix packaging problems that were introduced with 5.12.y ... actually code adjustment is already done, testing is undergoing.

lampra

lampra

Posted
Posted

The test build  (Hirsute 5.12.6 May 23 2021) gets stack at Starting kernel ...

 

  Reveal hidden contents

 

Werner

Werner

Posted
  • Author
Posted
  On 5/26/2021 at 4:09 PM, lampra said:

The test build  (Hirsute 5.12.6 May 23 2021) gets stack at Starting kernel ...

Expand  

Try increasing verbosity to get more info

lampra

lampra

Posted
Posted
  On 5/26/2021 at 4:30 PM, Werner said:

Try increasing verbosity to get more info

Expand  

As per the documentation I increased verbosity to 7 at armbianEnv.txt but no new output.

 

  Reveal hidden contents

 

wota

wota

Posted
Posted
  On 4/28/2021 at 6:22 PM, Gabor5745737 said:

"reboot" is not turning on the board, just off, requires manual power cycling.

Anyone else noticed it?

Expand  

you can use sandisk tf card,it's will reboot normally .

the sbc  not work well with many tf cards, such as Samsung TF,which can't reboot.

Joerg

Joerg

Posted
Posted (edited)

 

  On 5/26/2021 at 4:09 PM, lampra said:

The test build  (Hirsute 5.12.6 May 23 2021) gets stack at Starting kernel ...

Expand  

For me too (Armbian_21.05.2_Orangepizero2_hirsute_edge_5.12.6.img).

Edited by Joerg
lampra

lampra

Posted
Posted
  On 5/27/2021 at 2:00 PM, Werner said:

Bummer....

Expand  

I am wondering if with the newer 5.12 kernel we need mainline ATF here rather than the beta version in order to boot the kernel. Does anyone know?

jernej

jernej

Posted
Posted

H616 ATF stuff should be upstreamed already, but I'm not sure if it is in any stable version.

lampra

lampra

Posted
Posted

Please accept my apologies if I should have posted this in a different place.

 

So I have a working local build environment and I want to test the mainline ATF. Which is the correct way to change the source and the tree and define PLAT=sun50i_h616?

When I change it at the families folder the and running ./compile.sh, I see the following in htop: 

./compile.sh  BOARD=orangepizero2 BRANCH=edge RELEASE=hirsute BUILD_MINIMAL=yes BUILD_DESKTOP=no KERNEL_ONLY=no KERNEL_CONFIGURE=no COMPRESS_OUTPUTIMAGE=sha,gpg,img,imgATF_PLAT=sun50i_h6

 

and it keeps spiting this:

  Reveal hidden contents

 

 

Alternative can see an ATF folder in the userscripts folder /build/userpatches/atf/atf-sunxi64$ but I couldn't find any documentation how to use it.

Thank you

Werner

Werner

Posted
  • Author
Posted
  On 6/18/2021 at 11:21 PM, lampra said:

Successfully built & booted kernel 5.13.rc1 using mainline ATF, mainline uboot and this kernel source.

Ethernet & usb works, reboot does not. More tests in the next few days

Armbianmonitor

Expand  

Great. Mind sending PR with the necessary changes to the build repository?

lampra

lampra

Posted
Posted
  On 6/19/2021 at 4:21 AM, Werner said:

sending PR

Expand  

I probably could, but I am not sure how to handle this without braking everything else.

  

diff --git a/config/sources/families/include/sunxi64_common.inc b/config/sources/families/include/sunxi64_common.inc
index bc59438..1082479 100644
--- a/config/sources/families/include/sunxi64_common.inc
+++ b/config/sources/families/include/sunxi64_common.inc
@@ -27,9 +27,9 @@ case $BRANCH in
 
        edge)
 
-               KERNELSOURCE="https://github.com/megous/linux"
-               KERNELBRANCH="branch:orange-pi-5.12"
-               KERNELPATCHDIR='sunxi-'$BRANCH
+               KERNELSOURCE="https://github.com/apritzel/linux/"
+               KERNELBRANCH="branch:h616-v8-rc"
+               KERNELPATCHDIR='sunxi-orange-pi-5.12'
 
        ;;
 esac
diff --git a/config/sources/families/sun50iw6.conf b/config/sources/families/sun50iw6.conf
index 0862996..0a04279 100644
--- a/config/sources/families/sun50iw6.conf
+++ b/config/sources/families/sun50iw6.conf
@@ -8,14 +8,17 @@ ASOUND_STATE='asound.state.sun50iw2-dev'
 
 if [[ $BOARD == orangepizero2 ]]; then
 
-       ATFSOURCE='https://github.com/apritzel/arm-trusted-firmware'
-       ATFBRANCH='branch:h616-beta'
+       ATFSOURCE='https://github.com/ARM-software/arm-trusted-firmware'
+       ATFBRANCH='branch:master'
 
        ATF_PLAT="sun50i_h616";
        ATF_TARGET_MAP='PLAT=sun50i_h616 DEBUG=1 bl31;;build/sun50i_h616/debug/bl31.bin'
 
-       BOOTSOURCE='https://github.com/jernejsk/u-boot'
-       BOOTBRANCH='branch:h616-v2'
+       BOOTSOURCE='https://github.com/trini/u-boot'
+        BOOTBRANCH='branch:master'
 
        [[ -z $CPUMAX ]] && CPUMAX=1512000

 

1
terrya

terrya

Posted
Posted

overlay for UART5

 

/dts-v1/;

/ {
    compatible = "allwinner,sun50i-h6";

    fragment@0 {
        target-path = "/aliases";

        __overlay__ {
            serial5 = "/soc/serial@5001400";
        };
    };

    fragment@1 {
        target = < 0xffffffff >;

        __overlay__ {

            uart5-pins {
                pins = "PH2\0PH3";
                function = "uart5";
                phandle = < 0x01 >;
            };
        };
    };

    fragment@2 {
        target = < 0xffffffff >;

        __overlay__ {
            pinctrl-names = "default";
            pinctrl-0 = < 0x01 >;
            status = "okay";
        };
    };

    __symbols__ {
        uart5_pins = "/fragment@1/__overlay__/uart5-pins";
    };

    __fixups__ {
        pio = "/fragment@1:target:0";
        uart5 = "/fragment@2:target:0";
    };

    __local_fixups__ {

        fragment@2 {

            __overlay__ {
                pinctrl-0 = < 0x00 >;
            };
        };
    };
};
 

1
maigonis

maigonis

Posted
Posted

I use self build trunk version whit disabled kernel update for a while now. Besides bugs that are already known it is stable. I have installed BIND9 as a local slave authoritative DNS server and it is up 24/7.

 

If devs need any logs from me let me know.

Werner

Werner

Posted
  • Author
Posted

Yeah seems missing in dtb package. Dont know why.

 

Edit: Got it. Seems like 5.12 does not include the necessary h616 patches but there is a patch folder for 5.13 that can be use for manual inclusion. Trying that atm.

If you want to try too locally edit this line to 5.13 and build:

https://github.com/armbian/build/blob/6844232dd70a47a7dedc066c630c875c3b55e881/config/sources/families/include/sunxi64_common.inc#L31

Also change symlink from 5.12 to 5.13 in patch/kernel folder

lampra

lampra

Posted
Posted

As mentioned by @mantouboji 5.13 is working well for the moment with the relevant patches. Though we are missing, among others,  overlays and wifi.

Regarding overlays, I tried to compile w1-gpio-cl but I had issues with the headers on 5.13. Did anyone manage to compile anything with 5.13?

Regarding wifi, as this is new AW hw,  I expect that we need to port the driver from the ventor's legacy sdk. Is anyone familiar with the process? 

1
kert

kert

Posted
Posted

What's the correct way to prepare a tftp-bootable kernel from the build outputs ? I've tried various combinations of `mkimage ..` and `booti .. | bootm .. | bootz' variations using `vmlinux` or `Image` as inputs, but i'm obviously lost, getting either "Bad Linux ARM64 Image magic!" or "Wrong Image Format for bootm command"

Guest
This topic is now closed to further replies.
 Share
Go to topic listing
×
×
  • Create New...

Important Information

Terms of Use - Privacy Policy - Guidelines

  I accept