Ryzen 7000 Testing

[mariettosun]

I hope I am not wrong here

 

[mariettosun]

now is OpenCore Configurator ( you meant this I think) @CaseySJ

 

[CaseySJ]

now is OpenCore Configurator ( you meant this I think) @CaseySJ
View attachment 8831

My mistake -- I thought you were using Clover. If you're using OpenCore then the screenshot is correct.

Let's also add pci=0x02 to boot arguments, so this means we will have this:

• ixgbe=0 debug=0x100 pci=0x03 apple-axge-debug=0x1 ... and any others you normally use ...

Update:

• Changed apple-axgbe-debug to apple-axge-debug

 

[CaseySJ]

** UPDATE: We need to add 2 more patches to Aquantia Action driver **

After we are done with testing we can reduce the number of patches​

Please see updated post:

Ryzen 7000 Testing

Where does one find these kernel patches? … Huh?? :) Just take another look at the post you quoted from.

forum.amd-osx.com

Just tested all six patches together in Ventura 13.1, and they work fine. The Aquantia is also able to resume after wake. There may be some issue getting the system to sleep when Ethernet cable is connected to Aquantia -- this problem happens on Intel platforms as well.

 

[mariettosun]

@CaseySJ pci=0x02 ths hangs my system in early stage

added two new patches and I have the same result

 

[CaseySJ]

@CaseySJ pci=0x02 ths hangs my system in early stage

added two new patches and I have the same result

Let's try once more, but with pci=0x03 that will enable both _IOLog (dmesg) and kprintf (system log).

As you can see in my previous post, I just tested the 6 patches successfully in Ventura 13.1. Although the system does not sleep when cable is connected, it does work after waking from sleep under both of these conditions:

1. Unplug cable --> sleep works properly --> then wake computer --> connect cable to Aquantia --> works
2. Keep cable plugged --> sleep process begins, computer goes into standby for 1-2 seconds, then does a dark wake --> press key to do full-wake --> Aquantia works

 

[mariettosun]

Let's try once more, but with pci=0x03 that will enable both _IOLog (dmesg) and kprintf (system log).

bootlog.txt always the same here
dmesg output could be useful?

 

[CaseySJ]

bootlog.txt always the same here
dmesg output could be useful?

Apologies @mariettosun I made a mistake while sipping coffee by the fireplace...

The boot argument is actually apple-axge-debug=0x1

 

[mariettosun]

1184 instances now
upload the txt? or I have to search something here?

 

[DarkSilentSC]

Hohoho looks like I just got a whole bunch of test to do! I’ll get back to you guys when I get my access to rig once again. Truly amazing work to all of you!!!

 

[mariettosun]

@CaseySJ maybe is it for this?

 

[CaseySJ]

@CaseySJ maybe is it for this?
View attachment 8836

Yes -- that is a problem

If you like, you're welcome to post your config.plist in private message so I can review the patches.

By the way, your system log is showing very promising results so if we can fix the patch, the port should work...

# Packet 2
<== AppleEthernetAquantiaAqtion::interrupt_msi_handler
==> AppleEthernetAquantiaAqtion::interrupt_msi_handler index=0
==> AppleEthernetAquantiaAqtion::rx_pkt_queue
==> AppleEthernetAquantiaAqtion::allocate_mbuf
<== AppleEthernetAquantiaAqtion::allocate_mbuf mbuf=<private>
reg_put32(<private>, 0x00000002)
Flushing input queue.
<== AppleEthernetAquantiaAqtion::rx_pkt_queue
reg_put32(<private>, 0x00000001)
<== AppleEthernetAquantiaAqtion::interrupt_msi_handler


# Packet 3
==> AppleEthernetAquantiaAqtion::interrupt_msi_handler index=0
==> AppleEthernetAquantiaAqtion::rx_pkt_queue
==> AppleEthernetAquantiaAqtion::allocate_mbuf
<== AppleEthernetAquantiaAqtion::allocate_mbuf mbuf=<private>
reg_put32(<private>, 0x00000003)
Flushing input queue.
<== AppleEthernetAquantiaAqtion::rx_pkt_queue
reg_put32(<private>, 0x00000001)
<== AppleEthernetAquantiaAqtion::interrupt_msi_handler


# Packet 4
==> AppleEthernetAquantiaAqtion::interrupt_msi_handler index=0
==> AppleEthernetAquantiaAqtion::rx_pkt_queue
==> AppleEthernetAquantiaAqtion::allocate_mbuf
<== AppleEthernetAquantiaAqtion::allocate_mbuf mbuf=<private>
reg_put32(<private>, 0x00000004)
Flushing input queue.
<== AppleEthernetAquantiaAqtion::rx_pkt_queue
reg_put32(<private>, 0x00000001)
<== AppleEthernetAquantiaAqtion::interrupt_msi_handler

 

[CaseySJ]

@CaseySJ maybe is it for this?

Problem Found:



Patches 3 and 4 are actually for tx ring (not rx ring). I have re-listed them here:

Patch 3: Transmit Ring Patch 1

• Arch: x86_64
• Identifier: com.apple.driver.AppleEthernetAquantiaAqtion
• Base: __ZN27AppleEthernetAquantiaAqtion13alloc_tx_ringEj
• Comment: CaseySJ - Fix alloc_tx_ring #1
• Find: 4C89E741 B8030000 00
• Replace: 4C89E741 B8130000 00
• MinKernel: 21.0.0
• MaxKernel: 22.99.99
• Count: 1
• Enabled: True

Patch 4: Transmit Ring Patch 2

• Arch: x86_64
• Identifier: com.apple.driver.AppleEthernetAquantiaAqtion
• Base: __ZN27AppleEthernetAquantiaAqtion13alloc_tx_ringEj
• Comment: CaseySJ - Fix alloc_tx_ring #2
• Find: BA001000 00BF0200 0000
• Replace: BA001000 00BF1200 0000
• MinKernel: 21.0.0
• MaxKernel: 22.99.99
• Count: 1
• Enabled: True

 

[mariettosun]

@CaseySJ a correct ip is now there but system does not surf internet

first find patch is not found in my log as before

 

[CaseySJ]

@CaseySJ a correct ip is now there but system does not surf internet

first find patch is not found in my log as before

Let me send you modified config.plist in Private Message...

 

[CaseySJ]

Hohoho looks like I just got a whole bunch of test to do! I’ll get back to you guys when I get my access to rig once again. Truly amazing work to all of you!!!

Awesome. We need as many testers as we can get!

 

[mariettosun]

@CaseySJ your config did the job
first boot it reboots after few verbose lines
then it starts correctly and Aquantia is working well
Tested a cycle of Sleep/wake
It works perfectly (as before in my system)

So now, in MSI trx40 Pro work two internal i211 lan with new kext of @mbarbierato dev and Aquantia PCIE card with your patches

 

[CaseySJ]

@CaseySJ your config did the job
first boot it reboots after few verbose lines
then it starts correctly and Aquantia is working well
Tested a cycle of Sleep/wake
It works perfectly (as before in my system)

So now, in MSI trx40 Pro work two internal i211 lan with new kext of @mbarbierato dev and Aquantia PCIE card with your patches

Thank you — this is the first independent confirmation!!

Can you post a link to the i211 driver?

 

[ExtremeXT]

Thank you — this is the first independent confirmation!!

Can you post a link to the i211 driver?

Here, but unfortunately it's not open source.

 

[mariettosun]

to avoid errors (as I did in transcription) , and if it is not useful I will delete it, these are the patches in a plist

Add in yours config and enable also the Aquantia quirk

Thanks @CaseySJ job!