Get CI testing for Qt for WebAssembly up and running. There are several subtasks:

• CI builds: QTBUG-67588
• QTestLib and tests: Fix event loop issues.
  • Browsers expect that web page code will return control to the browser after processing events.
  • QTestLib is designed to keep control and call processEvents() to process more events.
  • This will make browsers flag the web page code as unresponsive; which may interfere with test execution.
  • This is a special instance of the general problem of the Qt event loop not returning control.

• Develop script for running tests in a browser and collecting the results.

• Establish a test pass/fail baseline. Many of the tests are going to fail, and we don’t want to manually audit all of them. Develop a script that records a baseline, and make the CI system use it. We will then use the CI system to enforce that there are no additional failures.

• Dynamic (shared libraries) builds. Not strictly needed, but building all the tests statically is going to be time and space consuming.