I guess this also resolves https://bugs.python.org/issue33926 . Thanks.

A Python core developer has requested some changes be made to your pull request before we can consider merging it. If you could please address their requests along with any other requests in other reviews from core developers that would be appreciated.

Once you have made the requested changes, please leave a comment on this pull request containing the phrase I have made the requested changes; please review again. I will then notify any core developers who have left a review that you're ready for them to take another look at this pull request.

@methane: we need to use -O0 or -Og as optimization flag. This is because gdb tests are not run when CPython is compiled with the default -O3 level (IMHO, this is in itself a bug, however fixing that is more controversial that I expected, see #14331).

I guess this also resolves https://bugs.python.org/issue33926 . Thanks.

In fact, it fixes exactly that. So I'm changing the bpo number here.

@methane: we need to use -O0 or -Og as optimization flag.

When configure --with-pydebug is used, it is set automatically. No need to set it explicitly.

@jdemeyer Can you please change the issue number in PR description from 32926 to 33926? I guess the bot keeps linking this PR to the issue 32926 for every comment made in this PR.

When configure --with-pydebug is used, it is set automatically.

OK, let's test that.

OK, this seems to work. I have made the requested changes; please review again

Thanks for making the requested changes!

@methane: please review the changes made to this pull request.

Thanks @jdemeyer for the PR, and @methane for merging it 🌮🎉.. I'm working now to backport this PR to: 3.8.
🐍🍒⛏🤖

GH-14420 is a backport of this pull request to the 3.8 branch.

Does this change have any noticeable impact on CI timing? The merge run happened for 20 mins which used to be 10-12 mins earlier. The backport PR also completes in around 20 minutes.

Does this change have any noticeable impact on CI timing? The merge run happened for 20 mins which used to be 10-12 mins earlier. The backport PR also completes in around 20 minutes.

Yes, it affects. I will compare the time of -Og and -O0.

Does this change have any noticeable impact on CI timing? The merge run happened for 20 mins which used to be 10-12 mins earlier. The backport PR also completes in around 20 minutes.

• On PR #14390: "CPython tests" job took 15 min 39 sec
• On this PR, it took 22 min 23 sec: +6 min 44 sec, it's quite significant

• PR #14390: "Total duration: 11 min 8 sec" just for tests (regrtest)
• On this PR: "Total duration: 20 min 30 sec": +9 min 22 sec

It's 1.8x slower.

Previously, we used -O3:

commit 8ff53564730f7ba503f5ad94418c309c48e8516d
Author: INADA Naoki <methane@users.noreply.github.com>
Date:   Sat Feb 10 20:35:17 2018 +0900

    travis: Use -O3 option (GH-5599)
    
    We don't use debugger on Travis.

Is it really useful to make the pre-commit CI almost 2x slower just because of test_gdb? We already have a large fleet of post-commit buildbots which run test_gdb.

https://travis-ci.org/python/cpython/builds/551298808
16min with -Og

-Og isn't the default when building using --with-pydebug?

16 min 52 sec total, "Total duration: 12 min 4 sec" just for the tests, looks close to what we had using -O3. If test_gdb still pass, we should definitively use -Og :-) So yeah for #14423 :-)