lockfree VS slice_deque

If you're interested about lock-free data structures, I wrote [lockfree](https://github.com/DNedic/lockfree) a collection of lock-free data structures meant to be readable and both hosted system and embedded friendly.

If you want more than a spsc queue, I've written `lockfree`, a collection of SPSC and MPMC data structures along the same principles the author here used:https://github.com/DNedic/lockfree.

The library is written in standard C++11 (but additional API's for higher C++ versions have been added), uses no dynamic allocation and is configurable so it is both big metal and deeply embedded friendly.

- A lot of code won't work for types with no default constructors, but that is at least compile error

- Using memcpy[0] for arbitrary types is just wrong, see [1]

[0] https://github.com/DNedic/lockfree/blob/main/lockfree/inc/bi...

[1] https://www.open-std.org/jtc1/sc22/wg21/docs/papers/2023/p11...

My friend wrote a few lock-free data structures and he is now looking for some feedback. Here is the link: https://github.com/DNedic/lockfree

> It's not an attack on the wording, but the correctness of your first bullet point. `unsafe` is appropriate for the initialization of a ring buffer in Rust. That's true for using `mmap` or anything in "pure" Rust using the allocator API to get the most idiomatic representation (which can't be done in safe or stable Rust). It's not one line. It's also not platform dependent, the code is the same on MacOS, Linux, and Windows the last I tried it.

We're not talking about the same thing then.

I'm talking about this code here: <https://github.com/gnzlbg/slice_deque/tree/master/src/mirror...> It is absolutely platform specific.

Yes, most ring buffer implementations feature a little bit of `unsafe` code. No, it doesn't make sense to say "I have a tiny amount of `unsafe` already, so adding more has no cost."

> But if your bottleneck is determined by the frequency at which channels get created or how many exist then I would call architecture into the question. ... This last month I've written a lock-free ring buffer to solve a problem and there's exactly one in an application that spawns millions of concurrent tasks.

Okay, but a lot of applications or libraries are written to support many connections, and you don't necessarily know when writing the code (or even when your server receives them) if those connections will be just cycled very quickly or will be high-throughput long-lived affairs. Each of those probably has a send buffer and a receive buffer. So while it might make sense for your application to have a single ring buffer for its life, applications which churn through them heavily are completely valid.

There is a single contiguous memory allocation, which mirrors itself.

One thread produces elements and pushes them at the tail (e.g. I/O bytes, in batch), and one thread consumes as many elements as possible in batch from the other end (e.g. all bytes available, in batch).

The mirror is required to allow processing all elements in the deque as if they were adjacent in memory.

This is the library i am using, the array contains an explanation : https://github.com/gnzlbg/slice_deque