Ok, tried to fix the sign-off issue with a new commit but it didn't take.
How do I fix this?

EDIT: never mind, I got it.

Firmware built and tested with a combination of a RPI PICO and W5500 breakout. Works with simple test for UDP, Ping, ftp, bulk transfer. Bulk transfer speed 100kb/s.
Test suite coverage:

./run-multitests.py -i micropython -i pyb:a0 multi_net/*.py

12 tests performed
9 tests passed
2 tests skipped: multi_net/asyncio_tcp_client_rst.py multi_net/tcp_client_rst.py
1 tests failed: multi_net/ssl_cert_rsa.py

./run-multitests.py -i pyb:a0 multi_net/*.py

12 tests performed
9 tests passed
3 tests failed: multi_net/asyncio_tcp_client_rst.py multi_net/ssl_cert_rsa.py multi_net/tcp_client_rst.py

python run-tests.py --target pyboard net_inet/*.py net_hosted/*.py

17 tests performed (109 individual testcases)
12 tests passed
5 tests failed: connect_nonblock_xfer ssl_cert ssl_errors ssl_getpeercert test_tls_nonblock

Hmm, is it normal for W5500 and/or RP2040 to fail those tests in your test suite?

I don't get it. Why does the CI still fail? I amended the commit with a sign-off.

Hmm, is it normal for W5500 and/or RP2040 to fail those tests in your test suite?

That's a Micropython test suite from micropython/tests. I cannot remember a case where all tests passed. I do not know whetehr it is a problem of the test or the board to be tested.

It's this commit about which CI complains, before your first commit. Not one of yours. I wonder why it is not a problem for other PRs. Did you make a local "git rebase master" before doing a force push?

commit 9fb56d15627e831f315e8afcbd5115c5506b10e0
Author: Mark Grosen <[email protected]>
Date:   Mon Jul 17 17:03:04 2023 -0700

It's this commit about which CI complains, before your first commit. Not one of yours. I wonder why it is not a problem for other PRs. Did you make a local "git rebase master" before doing a force push?

I didn't. Should I have? Should I try that now?

It is always a good approach to rebase before pushing a PR. Then you see if there are conflicts. On your case probably not, but if previous commits changed the same file it may happen, and then you can resolve it.

I'm afraid nobody's going to bother to pull this PR as long as that stupid commit message error is present. What do I do?

Just wait. PR's can take a long time. Especially those for new boards. The wrong error flag does not stop it.

I think that commit message CI checker is being thrown off because this PR is made from the "master" branch of your fork, rather than a new feature branch. I don't think you could fix it now other then by re-opening as a new pr on a new branch name. I don't think that should be necessary though.

Ok I'll just wait and see. With the long amount of time between releases I'm just eager to get my board added before the next release, wouldn't want to miss it due to a CI issue that isn't real.
The odd thing is that the CI passed with the first couple of pushes, then failed on the last one with that odd failure on a commit that isn't mine. All commits were done from my master branch, so it's very odd that at first it passed and then suddenly failed.

So, MicroPython 1.21 was released. Without this getting merged. :((
VERY disappointed. I guess I'll have to remove MicroPython compatibility from the docs? Since there's no easy way for a user to do this.

There's hundreds of other pr's waiting too, I know it can be frustrating but this is the reality of open source projects.

In the mean time you can always build and release your own binaries; I've been doing that for years in my own git repo CI's for work projects that need features not available here in master.
Open source projects are never going to be able to meet all company deadlines.

You could also try squashing your commits into one single clean commit, rebased on master, that would make it quicker to review and eventually get merged. Might even resolve that error message being shown.
If nothing else having just one commit makes it easier to maintain in future if you need.

Rebased and merged in 3270d85

Sorry this took a while to get to! At least it will be in the next release, v1.22.0, which is due out this week.