I finally completed the porting of ipq806x to kernel 5.10.
Main difference is dsa for the common device...
I would really appreciate any comment or test. Everything should work... but i can test this only on my r7800.
Also as i posted in the pr... I would really like to report any io error about mtd since we now have a way to fix this.
Reported working
Dubious working
Anyway uptime of 5 hours as the main router with 8-9 devices... all fine
Almost 4 hours here but I only have 3 devices.
what router?
R7800 
I can't seem to compile with your latest commit.
arch/arm/boot/dts/qcom-ipq8064.dtsi:1276.28-1297.5: ERROR (phandle_references): /soc/ethernet@37000000: Reference to non-existent node or label "phy4"
also defined at arch/arm/boot/dts/qcom-ipq8064-db149.dts:172.8-180.3
arch/arm/boot/dts/qcom-ipq8064.dtsi:1322.28-1343.5: ERROR (phandle_references): /soc/ethernet@37400000: Reference to non-existent node or label "phy6"
also defined at arch/arm/boot/dts/qcom-ipq8064-db149.dts:193.8-198.3
arch/arm/boot/dts/qcom-ipq8064.dtsi:1345.28-1366.5: ERROR (phandle_references): /soc/ethernet@37600000: Reference to non-existent node or label "phy7"
also defined at arch/arm/boot/dts/qcom-ipq8064-db149.dts:200.8-205.3
will fix it asap
Just to let you know I've had your DSA work running on my C2600 for the past week. Granted it's seeing low load, I have not had any stability issues in it... Good work 
@castiel652 can you test r7800 with latest commits? make sure to create a backup since i'm not certain but there is a problem with the switch caused by a patch. (so it will probably brick, and have to flash the previous firmware with the tftp method)
I was planning to upgrade my R7800 this weekend. What changed that can now brick the router?
Led Blink config in dsa driver
Doesn't work.
Router with no connection?
Yep. No connection at all
Pushed new version and should work.
just putting this one here incase
your branch isn't booting for me at the moment with an ea8500
I can't check exactly why right now, but I think there's some kind of problem with the bootloader passing along what the root filesystem is
@Ansuel
I applied your PR #3954 patchset (bba0cf59 - 3ca482f7) on top of the current master (91 commit ahead of your PR).
I left all my local customizations intact, and even edited the network config by by hand for DSA changes ("lan1 lan2 lan3 lan4" and "wan"). So I sysupgraded with -F and keeping settings 
R7800 boots ok, and at the first glance everything pretty much works, with the exception of the WAN LED that stays dark. That might be due to me editing network config by hand, but I think that there is no updated logic at https://github.com/openwrt/openwrt/blob/bba0cf599da20be7c6fd4c36acc80289b7dc87f8/target/linux/ipq806x/base-files/etc/board.d/01_leds#L39 yet.
In case you are interested, the kernel and system boot logs can be found at
https://gist.github.com/hnyman/95e9d3d07e949b5c3517f0e540ea3254
Ps.
I got the wan LED to work by looking example from mvebu WRT3200ACM,
and this works in /etc/config/system of R7800:
config led 'led_wan'
option name 'WAN'
option sysfs 'white:wan'
option trigger 'netdev'
option mode 'link tx rx'
option dev 'wan'
Corresponding 01_leds line in wrt3200acm is (but I did not yet test the in R7800):
ucidef_set_led_netdev "wan" "WAN" "pca963x:rango:white:wan" "wan"
With no booting you mean bootloop or no connection?
Ok i only read now about bootargs... Do you have a serial or anything that can give me a kernel log?
(failsafe should work? and can be used to get a kernel log?)
R7800 has been up since this morning. Nothing special to report
Except that kernel log shows that lan1 has strangely disconnected twice.
[ 43.292995] qca8k 37000000.mdio-mii:10 lan1: Link is Up - 1Gbps/Full - flow control rx/tx
[ 43.294975] br-lan: port 1(lan1) entered blocking state
[ 43.302776] br-lan: port 1(lan1) entered forwarding state
...
[ 61.104697] br-lan: port 6(wlan0) entered forwarding state
[ 3481.708786] ath10k_pci 0001:01:00.0: Invalid VHT mcs 15 peer stats
[16597.967728] qca8k 37000000.mdio-mii:10 lan1: Link is Down
[16597.967840] br-lan: port 1(lan1) entered disabled state
[16601.087405] qca8k 37000000.mdio-mii:10 lan1: Link is Up - 1Gbps/Full - flow control rx/tx
[16601.087468] br-lan: port 1(lan1) entered blocking state
[16601.094637] br-lan: port 1(lan1) entered forwarding state
[33194.228227] qca8k 37000000.mdio-mii:10 lan1: Link is Down
[33194.228343] br-lan: port 1(lan1) entered disabled state
[33196.310170] qca8k 37000000.mdio-mii:10 lan1: Link is Up - 1Gbps/Full - flow control rx/tx
[33196.310236] br-lan: port 1(lan1) entered blocking state
[33196.317418] br-lan: port 1(lan1) entered forwarding state
The strange part is uptime similarity:
lan1 up at 43, down at 16600, --> uptime period 16560
lan1 next down at 33194 --> uptime period 16590