CI workflows, run with standard mod_http2

[icing]

We used to include a special mod_h2 in our CI that supports the directive H2MaxDataFrameLen for test_02_20. Since then, ubuntu-lastest includes a more recent apache httpd. Let's see if we can live without the special.

[icing]

@bagder, @vszakats and @dfandrich : test_02_20 gets SKIPPED, due to the server not supporting the directive. I think we'll leave the special mod_h2 in there for now. While it is unlikely that we stumble onto the 7.88 data corruption again - the code is basically rewritten - I think having some tests with other HTTP/2 frame lengths is worth it.

But, if you prefer even more speed in CI, we can merge this PR. The test will continue to run on a local, fresher httpd install or once GH updates the image again.

[bagder]

I think we'll leave the special mod_h2 in there for now.

I agree. Good test coverage is more important than speedy CI.

[dfandrich]

Analysis of PR #15353 at 4ed426fc:

Test 987 failed, which has NOT been flaky recently, so there could be a real issue in this PR. Note that this test has failed in 3 different CI jobs (the link just goes to one of them).

Generated by Testclutch

[vszakats]

Isn't test_02_20 disabled for all CI runs? (regardless of apache support for this attribute)

master test run:

http/test_02_download.py::TestDownload::test_02_20_h2_small_frames[0] SKIPPED [  7%]

https://github.com/curl/curl/actions/runs/11438522033/job/31820289894#step:48:136

pytest source:

curl/tests/http/test_02_download.py

Lines 297 to 298 in e89491e

	@pytest.mark.skipif(condition=Env().ci_run, reason="not suitable for CI runs")
	def test_02_20_h2_small_frames(self, env: Env, httpd, repeat):

[icing]

Isn't test_02_20 disabled for all CI runs? (regardless of apache support for this attribute)

You are correct. Need to enable this test in CI to get the real result. Doh!

[icing]

Thanks for the sharp eyes, @vszakats ! Now test_02_20 passes in CI and we can merge this PR without losing coverage.