CI failures in pylons, profile without code changes

[emmettbutler]

Between ff9e040 and 7c5783b, the 1.x branch's CI started failing. The changes in 7c5783b do not appear to have caused this.

#6154 was applied to unblock CI by adjusting certain test expectations to match CI's behavior. It didn't find the root cause and make the code resilient against it. Let's find the root cause of the original failures.

The pylons test failures appear to be caused by an exception here leading to _DDWAF_LOADED being False. The code is already written in such a way as to indicate that this is an expected state, so adapting the tests to handle it was straightforward.

The profile test failures have no known common denominator so far. The closest I was able to discern is that they may have something to do with code provenance not being enabled, but I'm not confident in that assessment. The test expectations here and here should be reevaluated to make sure they represent the expected behavior of the profiler.

[emmettbutler]

@christophe-papazian @P403n1x87 I'm tagging you on this because you may have an easier time finding the root cause in your respective areas of expertise.