[hist] prevent nullptr access with copy from default constructor

[ferdymercury]

Fixes #16771

[github-actions]

Test Results

    18 files      18 suites   3d 14h 24m 27s ⏱️
 2 699 tests  2 698 ✅ 0 💤 1 ❌
46 103 runs  46 102 ✅ 0 💤 1 ❌

For more details on these failures, see this check.

Results for commit ae2a9ab.

♻️ This comment has been updated with latest results.

[pcanal]

Is the failure in roottest-root-hist-h2root at https://github.com/root-project/root/pull/16773/checks?check_run_id=32223807661 understood?

[ferdymercury]

Is the failure in roottest-root-hist-h2root at https://github.com/root-project/root/pull/16773/checks?check_run_id=32223807661 understood?

Is it related to #15915 ?
pinging @couet and @hahnjo

[dpiparo]

@ferdymercury à propos roottest-root-hist-h2root: I think this was clarified by the last round of tests. We see a failure on the GPU platform, which is known and unrelated to your changes. If you agree, I would proceed with the merge of this PR

[ferdymercury]

Agreed thks!

[ywkao]

Hi @ferdymercury, thank you for the swift fixes!
Would it be possible to backport these changes to versions 6.30 and 6.32?

[guitargeek]

We could, but there will probably be no new patch release anyway, since ROOT 6.34.00 is only a few weeks away. And we're not going to make a new patch release just for this bugfix.

Given this information, could you live without a backport? I guess the only way in which it could help you is if you build your own ROOT from source from the tip of the 6.30 and 6.32 branches, without needing a patch release. But I guess this is not the case?

[guitargeek]

Hi @ywkao, following up on my last comment, what would be motivation for the backports? Are you using these older ROOT version in production and this is a crucial fix? Or you're using ROOT on your laptop and if would be easy to update to 6.34 risk-free? How we proceed depends also on that

[ywkao]

Hi @guitargeek, thank you for the information. I found that the fix has been included in the latest CMSSW release so I can continue with the unit test in cms-sw/cmssw#41932 (comment). I will let you know if there is still a need of the backport after confirming with the CMS DQM experts.

[ywkao]

Hi @guitargeek, after confirming with CMS people, we would like to have TH2Poly functional in the default CMSSW. A backport to v6.30 and v6.32 will be very helpful.

[guitargeek]

I understand. The backports will be merged tonight or tomorrow morning.

[ywkao]

Thank you for the help with the root backports!