I didn't find the first crash and there is nothing shows up on the second crash (probably because the first crash make the device enter a crash cycle...)

Could it be possible the same issue as Issue 770791?

This is the line from the first crash
https://chromium.googlesource.com/chromium/src/+/e8a498472788c1c6834c/chrome/browser/ui/webui/settings/chromeos/cups_printers_handler.cc#311

The crash indicates SIGSEGV but I'm not sure how that's possible in the code.

I wonder if there's a configuration we're not testing.  Is Sync enabled?

This seems to happen when I get a printer added with a particular invalid config. I can't tell what that config is because now as soon as I go into printer config it immediately crashes.

Is there some info I can collect?

It looks like that it's possible that GetPrinterInfo() returns a null pointer. In this case, when |printers_list| tries to append the value to the list, it will try to use * to get the value the pointer points to (see https://cs.chromium.org/chromium/src/base/values.cc?rcl=c9d234ccd01690deaeba611d6e13d9105b68c025&l=1314), then it will crash if it's a null pointer.

It seems like we only need to make sure not to append a nullptr to the list should fix the issue.

I see it now.  If the uri is invalid, we return a nullptr which we probably shouldn't do.  I think it's possible that an invalid url can be entered into settings.