I was approached about this topic on November 2025. I have nothing new to
add but to quote my private response to George-Alexandru Popa, minus the
rants. So please stop copying me on this thread. Ask the netdev mailing
list instead; there are qualified people such as Vladimir, Andrew, and
Linus that can provide guidance.

Chester A.

Ok then tell the chap working on the ethernet driver to use the 
"mediatek,mt7621" compatible string for the switch as that's the correct one 
for MCM MT7530. It's called that because MCM MT7530 only existed on MT7621 SoC until this 
EcoNet SoC, apparently. It's still unclear if MCM for EcoNet SoC needs different setup 
operations though. I can help with that if it doesn't work out of the box, once the 
Ethernet driver is finalised.

Chester A.

On 18 November 2025 00:15:41 GMT+03:00, George-Alexandru Popa 
<[email protected]> wrote:
I suspect is multichip module because the OEM dump says "MCM MT7530":

insmod ETH_LAN drivermacType = 0qdma_lan: physical_size is 0x00200000ether.c:v1.00-NAPI 29.Mar.2015eth0: FE MAC Ethernet address: 50:E0:39:00:09:90Use External switch ,type 10EN7512 TRGMII V1.0 GE Rxet cal (1)(0)(0)(0x21) GE Rext AnaCal Done! (2)(0x22) GE Rext AnaCal Done! (2)(0x22) GE R50 AnaCal Done! (2) (0x27)(0xa7) GE R50 AnaCal Done! (2) (0x27)(0xa7) GE R50 AnaCal Done! (2) (0x27)(0xa7) GE R50 AnaCal Done! (2) (0x27)(0xa7) GE Tx offset AnaCal Done! (5)(0x25) GE Tx offset AnaCal Done! (4)(0x24) GE Tx offset AnaCal Done! (2)(0x22) GE Tx offset AnaCal Done! (3)(0x23) GE Tx amp AnaCal Done! (pair-0)(0x12)(0x1020) GE Tx amp AnaCal Done! (30) (0x2) GE Tx amp AnaCal Done! (pair-1)(0x17)(0x420) GE Tx amp AnaCal Done! (30) (0x2) GE Tx amp AnaCal Done! (pair-2)(0x19)(0x920) GE Tx amp AnaCal Done! (25) (0x7) GE Tx amp AnaCal Done! (pair-3)(0x21)(0x220) GE Tx amp AnaCal Done! (32) (0x0)tcPhyVerLookUp 3887 0tcPhyVerLookUp 3887 0tcPhyVerLookUp 3887 0tcPhyVerLookUp 3887 9412 7530 Ge,
0x1MCM MT7530tcPhyVerLookUp 3887 9412 7530 Ge, 0x1MCM MT7530 [tcPhyInit] in  
[tcPhyInit] out , r31 = (1, 0x0)

On Monday 17 November 2025 at 18:58:54 GMT, Chester A. Unal <[email protected]> wrote: You're not writing an ethernet driver for MT7530. MT7530 is an ethernet
switch. You'd write a DSA driver, which you don't need. You're writing an
ethernet driver for EcoNet EN751221 but your chaps are struggling to make
the MDIO bus of it work. So I have nothing to help with because I don't do
ethernet drivers so once whomever is working on the Ethernet driver makes
the MDIO bus work, the MT7530 switch that is connected to the bus will
work fine using the MT7530 DSA subdriver. That is, of course, if this
MT7530 switch is a standalone IC on the Zyxel device you mention.

Please let me know if the MT7530 switch on the Zyxel is standalone or if
it's in the EcoNet SoC as part of a multi-chip module, or if it's like the
switch component in MT7988 SoC meaning it's MMIO-accessed.

Chester A.

On 17 November 2025 17:14:42 GMT+03:00, George-Alexandru Popa 
<[email protected]> wrote:
P.S. Also, feel free to highlight any misinformation as that will help us speed 
things up.
On Monday 17 November 2025 at 14:12:06 GMT, George-Alexandru Popa <[email protected]> wrote:
  The link that i've sent is where the relevant discussion actually starts. 
Basically, we want to know if we need to write an entire MT7530 ethernet driver 
or it's already implemented and we're just wasting time on this part. CJD 
already worked on it on here: https://github.com/cjdelisle/econet_eth



On Monday 17 November 2025 at 12:32:25 GMT, Chester A. Unal <[email protected]> wrote:
Hey Alex.

I'm not going to read all of that thread which is riddled with misinformation. 
You're going to have to summarise what do you want me to weigh on.

Chester A.

On 17 November 2025 14:57:09 GMT+03:00, George-Alexandru Popa 
<[email protected]> wrote:
Hi there.
I've seen that you've worked on the MT7530 DSA 
driver(https://lists.infradead.org/pipermail/linux-mediatek/2024-March/076381.html)
 and would greatly appreciate if you could give us some pointers on how could 
we tackle our issue. Does the driver needs rewriting entirely or is just a 
per-device issue with the MIDO bus? Discussion here: 
https://forum.openwrt.org/t/openwrt-support-for-zyxel-pmg5617ga-and-econet-socs-first-gpon-support/144231/235


Thanks,
Alex




_______________________________________________
openwrt-devel mailing list
[email protected]
https://lists.openwrt.org/mailman/listinfo/openwrt-devel

Reply via email to