Concept ACK, though the ever-complicating knot of shell script is not something I feel confident to review.

I guess the alternative would be to replace those tests with BOOST_CHECK(true);. But anything seems like a hack. Why is the file compiled at all when none of the tests are compiled?

My guess is to avoid checking for the boost define in the build system.

This doesn't work with Boost 1.47 and makes gitian hang because its boost doesn't support the --list_content option.

In search for a better solution, I ended up falling back to basically #16564

Theoretically, we could have configure check the define for us and remove the file entirely, but I'm not sure it's worth the trouble (and doesn't give the user any indication the test was skipped!).

This doesn't work with Boost 1.47 and makes gitian hang because its boost doesn't support the --list_content option.

Now we have minimum Boost 1.64.0.

Also this PR makes make check compatible with https://www.boost.org/doc/libs/1_64_0/libs/test/doc/html/boost_test/tests_organization/enabling.html.

Maybe reopen?

What is the status of this? Needs conflicts fixed. Needs the CI to run. Also needs the PR description updated.

Rebased

🐙 This pull request conflicts with the target branch and needs rebase.

_{Want to unsubscribe from rebase notifications on this pull request? Just convert this pull request to a "draft".}

Rebased again, and added a fix for a regression introduced by #19385

This would previously cause make check to fail.

I tried to test this, but failed. make check passes. See also:

./src/test/test_bitcoin -t raii_event_tests -l message && echo $?
Running 1 test case...
Skipping raii_event_tess: libevent doesn't support event_set_mem_functions
Test case raii_event_tests/raii_event_tests_SKIPPED did not check any assertions

*** No errors detected
0

This was already fixed in 091cc4b

Good point.