allocator | experimental high performance , low fragmentation memory | Performance Testing library

std::vector and other containers (except std::array) are specified to have a copy constructor. This is not specified to be conditional on whether or not the element type is copyable. Only instantiation of the copy constructor's definition is forbidden if the element type is not copyable.

As a result std::is_copy_constructible_v on the container will always be true. There is no way to test whether an instantiation of a definition would be well-formed with a type trait.

It would be possible to specify that the copy constructor is not declared or excluded from overload resolution if the element type is not copyable. However, that would come with a trade-off which is explained in detail in this blog post: https://quuxplusone.github.io/blog/2020/02/05/vector-is-copyable-except-when-its-not/.

In short, if we want to be able to use the container with an incomplete type, e.g. recursively like

Looking at the header here: https://github.com/hunspell/hunspell/blob/master/src/hunspell/hunspell.h#L80

I tested the same code in Visual Studio 2022 and got the same error. After my exploration, I found the problem.

Firstly, I copied the contents of A.h and B.h into main.cpp and removed the #include directive. After compiling, I still got the same error.

Then I tested and found that as soon as I moved namespace std {...} after the definition of class B, the error went away.

I read the assembly code generated by the compiler and found that the names generated for GetMap in main.cpp and b.cpp are different:

main.asm:

GetMap@B@@QEBAAEBV?$unordered_map@UA@@HV?$hash@UA@@@std@@U?$equal_to@UA@@@3@V?$allocator@U?$pair@$$CBUA@@H@std@@@3@@std@@XZ

b.asm:

GetMap@B@@QEBAAEBV?$unordered_map@UA@@HU?$hash@UA@@@std@@U?$equal_to@UA@@@3@V?$allocator@U?$pair@$$CBUA@@H@std@@@3@@std@@XZ

I looked up MSVC's name mangling rules and found U for struct and V for class. So I change the definition of template<> class hash to template<> struct hash. Then the error disappeared.

Your type is an aggregate

Maybe I'm wrong, but it seems that last part:

The standard containers will use the allocator provided to them to allocate dynamic memory. By default, that is std::allocator.

For most other dynamic memory uses in the standard library, the standard doesn't specify how the implementation should acquire memory, and the implementation has the freedom to do what they want.

As for tracking memory allocations, I recommend wrapping malloc.

The bug has been fixed and the example now compiles.

It can be a bug of that Flutter package, indeed. Have you tried to create an issue in GitHub of that package?

Secondly, during my development, I see several times when emulators just fail and real devices always work. The solution I used is - simply to do not test them on simulators. Real users never use simulators, aren't they?

It can be a bug of the library when running on x86 arch (the arch simulators use). Then, nobody with a real device (arm arch) will ever see the bug.

Thirdly, what about trying to use "cloud real devices" to test whether they work on real Pixel devices that you are worried about. There are many platforms that host some real devices and you can connect to them via a webpage and test your app.

Assuming you are using Microsoft's C/C++ extension, you must configure the extension to use C++ 20 standard for intellisense.

The easiest way to do this is to add the line "C_Cpp.default.cppStandard": "c++20" to your settings.json file. You can also find the setting in the GUI under the name "Cpp Standard". Selecting c++20 from its dropdown will achieve the same result.

Note that this setting is, by default, set as a global user defaults. You can configure it per-workspace by selecting the Workspace tab in the settings GUI and changing that Cpp Standard dropdown to c++20.

As for why adding the -std=c++20 flag didn't work: -std=c++20 just tells your compiler which standard to use to build your code. 'Intellisense' does not receive this flag because it is a separate tool from the compiler and is therefore not required to support all the standards the compiler supports. It may support less even, although Intellisense tools usually support as current a standard as possible. Therefore the language standard for Intellisense must be configured separately from the compiler (in this case).

Final Note: After changing the setting, try closing and re-opening VS Code. In my experience changing the language standard setting can cause some weirdness to happen. Closing and re-opening VS Code seems to ensure the setting changes take full effect.