you don’t need _nodt and dtb files for r2 because normal uImage has dtb appended like the existing uImage…
have you changed kernel var in your uEnv.txt to the new filename?
kernel=uImage_4.19.88-main
if there is no such line, add it 
do you have only hdmi connected or can you look on debug-uart?
i guess “root=/dev/sda1” is wrong or do you have your OS on Sata-HDD?
i have added rtl815[02] to my 4.19-main and updated to .92 you can downloaded package from github releases
the kernel line had no effect. So I build the kernel from scratch
fetched original configs with
./build.sh importconfig
Add my desired driver and build and installed the kernel again.
And now…machine is booting with the new kernel…puhhhhh…hard work!
Also the kernel module for my device works. A bit ugly here, is the behavior of renaming the new interface from eth2 to enxd037453fe21b.
[ 10.973153] r8152 4-1.4:1.0 eth2: v1.09.9
[ 11.525462] r8152 4-1.4:1.0 enxd037453fe21b: renamed from eth2
Thanks a lot for your support! 
And yes, I use a ssd as root device.
A single qustion is left.
Is it possible to suppress all these kernel messages on the console?
Or let them only appear on tty1?
Best regards
Jochen
which driver was it?
https://www.freedesktop.org/wiki/Software/systemd/PredictableNetworkInterfaceNames/
=> “I don’t like this, how do I disable this?”
for console:
dmesg --console-off
should do what you want…else you can disable debug and drm.debug in your cmdline (bootargs)
driver is r8152. It’s a 3 port usb 3.0 hub with a gigabit ethernet port
an other solution for the naming problem is this
create a file:
/etc/udev/rules.d/70-persistent-net.rules
SUBSYSTEM==“net”, ACTION==“add”, DRIVERS=="?", ATTR{address}==“b8:ac:6f:65:31:e5”, ATTR{dev_id}==“0x0”, ATTR{type}==“1”, KERNEL=="eth", NAME=“eth0”
changing the mac address to your device and chance the NAME to your desired interface name. In my case for the R2 it is eth2
I removed the “drm.debug=0x1” entry from the uEnv line, but it is still logging 
now I’ve added the “dmesg --console-off” to my rc.local and the spooke has an end.
Thank you!
this one does not work in my case (pcie wifi card)…
dropped dev_id,type,kernel and changed name and mac
checked values with “udevadm info --attribute-walk /sys/class/net/wlp1s0”, have now added net.ifnames=0 to my bootopts via uEnv.txt
Hello Frank,
it seems I have a hw related issue since a few days ago!
[ 528.629555] mt7530 mdio-bus:00 lan2: Link is Down [ 530.704485] mt7530 mdio-bus:00 lan2: Link is Up - 100Mbps/Full - flow control off [ 530.711975] br0: port 3(lan2) entered blocking state [ 530.716954] br0: port 3(lan2) entered forwarding state [ 542.144605] br0: port 3(lan2) entered disabled state [ 542.149764] mt7530 mdio-bus:00 lan2: Link is Down
Before that, I don’t change anything! I’ve tried reinstall OS, everything(other cabel, other stuff), but bpi-r2 doesn’t want to communicate on this lan port!
Has somebody faced with this?
do other ports work/is this the only port? only 1 port connected to a switch (not 2 ports to same switch!)? tried release port out of bridge and working directly over it (other subnet)?
i don’t think it is OS-related…maybe kernel, but i don’t remember such…but i use no br0-config
have you anything hardware-related in dmesg?
hai frank… can i burn with sd card with etcher? cause i’m not using linux
i don’t know etcher…basicly image needs to be written 1:1 to sd-card…if etcher does this, then it should work
Etcher is fine to write images to SD-card. I used it many times.
Nothing to hardware related in dmesg!
I suspected the power supply - but it should’ve worked! (12V 2A)
I’ve got away both sata hdd and pci-e wifi card, but after a few hours symptoms are the same! Sometimes port 1, after that port 3…etc not able to give ip address! So, only one can’t get working! I’ll try remove bridge-utils, and take a switch put on lan 3 just for instance! What else should I test?
I’m really going white soon … ! 
After a reconfig, I can tell you, I’m sure it was a faulty power supply! Would you advise a 12V/3A for replacing?
I have only 12v2A
I think I miss some kernel configuration, and should recompile it! (CONFIG_NFT_SET_RBTREE=m CONFIG_NFT_SET_HASH=m CONFIG_NFT_SET_BITMAP=m) I need it for geoip nftables filtering! What is the most simple way to do this? Could you advise me some hints? The power supply issue is resolved temporarily with a separate switch hanging on lan port till I will have ordered a new one!
Which kernel-version do you use? Then i can add them to defconfig and you can use binary from github releases after build…
edit: i did not found them in 5.4/4.19…
https://cateee.net/lkddb/web-lkddb/NFT_SET_RBTREE.html says “found in Linux kernels: 4.9 - 4.17”
based on this patch it seems they are replaced by CONFIG_NF_TABLES_SET (set in 5.4, but not on 4.19)
I do not understand why a switch on lanport solves your power-issue…have you a very long distance to other side (between r2 and other client/router)? Or is there a cable problem (e.g. shortcut)
I use this :
"Linux bpi-r2 4.19.92-bpi-r2-main #1 SMP Sat Jan 4 12:32:40 UTC 2020 armv7l GNU/Linux "
Does it mean, that we can’t add it to the kernel? It’s strange!
I used R2 with two attached 2.5 HDD’s, and a wireless card in pci-e, and all lan ports. After a few days power supply got hot, rebooted and didn’t want to work perfectly. That’s the cause I made a decision dropping hdd out and using only 1 lan port behind a switch - seems to me working flawlessly!
We can’t add the old options,but maybe functionality is available with new option (have added it to 4.19-main too, should be build after ~30 minutes)
looks like it is this Patchset: https://www.spinics.net/lists/netfilter-devel/msg54692.html
Thank you Frank! I’ve just added your deb package, but couldn’t test it yet! A short supplement to your table on github : by kernel 4.19 OpenVPN, and LXC are functioning! I could apply both of them, furthermore OVPN runs from LXC with ip address configured from bridge, and opened a port to it on nftables firewall.
PS : It’s working! 
Hello Frank!
Finally the ordered MT7612e Wifi card arrived! Unfortunately I can"t compile a firmware! It’s all the same wether I try it on R2 or on VM(Debian 10)! I see followings (it comes in step build.sh =>1, point 4 success) :
" /home/bpi_r2_mt76/u-boot-mt/include/linux/compiler-gcc.h:95:1: fatal error: linux/compiler-gcc8.h: No such file or directory #include gcc_header(GNUC) ^~~~ compilation terminated."
I get a compiler-gcc8.h file from your github repo (4.14) and inserted but it looks out more awful and throws multiple errors!
Do I something wrong? Maybe should I use another gcc?