CPU:0, Error: cbb-fabric@0x13a00000, irq=34

When I exec “devmem2 0x2310000”, it shows:

[690515.740921] CPU:0, Error: cbb-fabric@0x13a00000, irq=34
[690515.746418] **************************************
[690515.751442] CPU:0, Error:cbb-fabric, Errmon:2
[690515.756021] 	  Error Code		: TIMEOUT_ERR

[690515.761775] 	  Error Code		: TIMEOUT_ERR
[690515.765902] 	  MASTER_ID		: CCPLEX
[690515.769498] 	  Address		: 0x2310000
[690515.773178] 	  Cache			: 0x0 -- Device Non-Bufferable
[690515.778461] 	  Protection		: 0x2 -- Unprivileged, Non-Secure, Data Access
[690515.785544] 	  Access_Type		: Read
[690515.789136] 	  Access_ID		: 0x12
[690515.789138] 	  Fabric		: cbb-fabric
[690515.796232] 	  Slave_Id		: 0x35
[690515.799564] 	  Burst_length		: 0x0
[690515.803156] 	  Burst_type		: 0x1
[690515.806570] 	  Beat_size		: 0x3
[690515.809900] 	  VQC			: 0x0
[690515.812781] 	  GRPSEC		: 0x7e
[690515.815926] 	  FALCONSEC		: 0x0
[690515.819252] 	**************************************
[690515.824374] ------------[ cut here ]------------

But “devmem2 0x6910000” is ok. Both “0x2310000.ethernet” and “0x6910000.ethernet” are ethernet nodes. And I didn’t add them to dts. I would like to know how to fix it so that I can successfully access “0x2310000.ethernet” .

Hi,

I don’t see any meaningful purpose of doing this.

Please let official kernel driver do the work instead of doing this.

Actually, I want to read/write PHY registers with mdio-tools. It’s fine when I access 0x6810000/0x6910000, but it’s not fine when I access 0x2310000. So I tested those addresses with devmem2. They have the same result. I also found that “0x31e0000.i2c” can be accessible, but “0x3210000.spi” is not. I don’t understand why some addresses are accessible and others aren’t. 0x2310000 is the address I need to access in my project.

please use mdio tool + c22 for 0x2310000.

It also reports a cbb-fabric error.

What kind of command you are using there?

mdio 2* phy 1 raw 2

Did you already enable device tree node for RGMII?

Yes, I added it. If I run mdio I can see that.

# mdio
2310000.ethernet
fixed-0

So the RGMII interface is working but you just cannot use mdio tool for it?

Yes, I think some code restricts access to 0x2310000 and other addresses. We can confirm this using devmem2. Some addresses in the /sys/devices/platform/ directory can be accessed, but others can’t.

# ls /sys/devices/platform/
# devmem2 <address>

please share the full dmesg first.

dmesg.txt (165.0 KB)

Thank you, I fixed it. There was something wrong with mac_rst.

1 Like

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.