Add migration code for QMutex/QRecursiveMutex/QMutexLocker

[uklotzde]

A quick PoC for portable QMutex/QRecursiveMutex/QMutexLocker code.

Not tested with Qt6 yet.

[coveralls]

Pull Request Test Coverage Report for Build 1018058343

• 87 of 147 (59.18%) changed or added relevant lines in 2 files are covered.
• 2 unchanged lines in 1 file lost coverage.
• Overall coverage decreased (-0.0008%) to 28.627%

---

Changes Missing Coverage	Covered Lines	Changed/Added Lines	%
src/track/track.cpp	85	145	58.62%

Files with Coverage Reduction	New Missed Lines	%
src/engine/readaheadmanager.cpp	2	88.79%

Totals
Change from base Build 1017932377:	-0.0008%
Covered Lines:	20082
Relevant Lines:	70150

---

💛 - Coveralls

[uklotzde]

I have chosen to migrate Track because it probably covers all use cases.

[uklotzde]

This is also an example why using auto instead of literally repeating the actual type declaration everywhere is useful.

[daschuer]

This is also an example why using auto instead of literally repeating the actual type declarations everywhere is useful.

Right, this overload and also template use case is the perfect fit for auto. I it also still fits to our rule to use auto to avoid repeatiton.

I don't think that it rectifies the switch to always use outo. Apart from the readability issues of diffs and code outside the editor it also makes the recent type refactoring less effective because some type checks are bypassed when using auto.

[daschuer]

LGTM. Thank you for wrapping you head around this hard nut.

[Be-ing]

Waiting for CI

[daschuer]

This is still draft state. Do we need some manual tests?

[uklotzde]

The compiler is the test.

The PR is marked as Draft to ask for feedback. Maybe someone has other/better ideas? It was just a quick test. I don't care how it finally looks like. It should just work.

[uklotzde]

This is also an example why using auto instead of literally repeating the actual type declarations everywhere is useful.

Right, this overload and also template use case is the perfect fit for auto. I it also still fits to our rule to use auto to avoid repeatiton.

I don't think that it rectifies the switch to always use outo. Apart from the readability issues of diffs and code outside the editor it also makes the recent type refactoring less effective because some type checks are bypassed when using auto.

If using auto introduces any weakness you should first and foremost rethink your type system, i.e. by using strict types and preventing implicit conversions. Implicit conversions between various native types are a weakness of C/C++.

[Be-ing]

You may want to share this on the Qt forum. I am sure others will find it useful.

[uklotzde]

I am not particularly proud of this simplistic solution 😅

Moreover we haven't tested it with Qt6 yet. It's an ugly workaround and Qt should have provided a better migration path.