ClangBuildAnalyzer VS cppcoro

> On another note: C++ compiler should by default keep statistics about the chain of #include's / parsing during compilation and dump it to a file at the end and also summarize how badly you're re-parsing the same .h files during build.

Not exactly that, but do you know clang's -ftime-trace and tools like https://github.com/aras-p/ClangBuildAnalyzer which help analyzing where time is actually spent? (In small repeated headers I don't see much of a problem, but they of course may contain not so small things ...)

You can also use the following when you want to inspect multiple files: https://github.com/aras-p/ClangBuildAnalyzer

ClangBuildAnalyzer reports on parsing, build, and link time, whereas IncludeGuardian only reports on parsing time.

https://github.com/aras-p/ClangBuildAnalyzer is a very useful tool to quantify the cost of different headers (and other costly parts of the compile such as template instantiations). It doesn’t help with actually fixing such problems, but it’s a pretty good ruler to measure where the time is spent.

If you can compile with Clang, I suggest you to try ClangBuildAnalyzer

PS: Take a look at cppcoro; this might help as well, especially generator<>, if you're looking to generate numbers, and stuff;

This is the only viable implementation.

Kind of sounds like whatever library you were using provided leaky abstractions. Something like cppcoro provides really good abstractions for coroutines, the user really doesn't need to understand why any of it works.

Is there a usecase for copying/serializing such coroutines? If not, I would use the normal C++20 coroutines (cppcoro?).

C++ has the popular CppCoro library. Async_mutex is its equivalent of Tokio::sync::Mutex, providing exclusive access to data shared between tasks.

Because the whole application is running under a single thread there is no need for atomic operations in synchronization primitives(which most of the time requires seq_cst memory order and CMPXCHG which is an expensive instruction in CPU). for example what async_mutex would look like if it knows it's running in a single-threaded scheduler (a non-atomic state variable and waiters queue).

Maybe not old, but I wish cppcoro was still updated. It was such a nice start!

You can get generator<> from https://github.com/lewissbaker/cppcoro

It is possible to compose them more easily than described in the article; Lewis Baker's cppcoro library for example provides a recursive_generator<> type[0] that allows this without using any macros. It's up to the library part of coroutines to make things easy, end users are not expected to write low-level coroutine code themselves.

I wonder about the allocation elision. Return value optimization became mandatory, and some compilers can already elide calls to new/delete and malloc()/free() in normal code, so perhaps it will be possible to guarantee allocation elision in the future in the most used cases.

[0]: https://github.com/lewissbaker/cppcoro#recursive_generatort