EthernetC33 debugging

[JAndrassy]

the missing volatile on bool dhcp_acquired in CNetIf.h caused problems for WiFi DHCP too. The waiting for DHCP ended only on timeout, but status() check then was OK because the IP address was set. So WiFi.begin just took long time.

I use Uno R4 to test Renesas Core LwIP networking. For WiFi I have an esp32 with the Hosted firmware on SPI and for Ethernet I replaced the EthernetDriver implementation with one over the W5500 driver (Ethernet 2 shied) from the esp8266 core. So I use Ethernet.begin methods with the MAC address parameter where some of this bugs appeared. The missing restartAsyncRequest(); in setMacAddress stopped the periodic task so DHCP then timed out. After that I had in EthernetDriver a buffer smaller than the DHCP packet and the code just hang. It was the the assert in CEth::output).

tested with LegacyEthernetTest

EthernetDriverW5500.cpp should work on Portenta C33 with MKR ETH shield too

[aentinger]

LGTM 👍

[aentinger]

Looks like this branch is kept from merging by the license/cla check not bring signed. Can you sign the CLA agreement? Likely you've already signed it, but maybe you made this commit with a different email address? (This is a known cause for CLA errors).

[aentinger]

Update: wasn't a CLA issue but a merge conflict with main. All is resolved now.

[aentinger]

Why the force-push? Now there's another merge-conflict 😉 . You'd have needed to pull --rebase, then add a new commit on top (uncommenting the include). Do you want to clean this up or shall I?

[JAndrassy]

sorry, the force push was just to start the check for the CLA.
yes please resolve the conflict.

[aentinger]

Hi @JAndrassy ☕ 👋 can you check that your merges are in? (They should be).

[JAndrassy]

looks OK. thank you