MikeDvorskiy

> For host_iterator inputs, wouldn't that add an unnecessary copy-back for data1 and data2 which were previously read (copy-in) only? right.... In that case - "transform" + read_write mode for...

BTW, my general thoughts. Looking at a scenario like this wider - actually we need a way(approach) to "transfer data by mask" between device and host. If we consider an...

> * move backend public API functions into these structures; Unfortunately, it breaks one of tag dispatching advantage - to have possibility to overload (via C++ function overloading mechanism) backend...

There is a prototype to fix the issue: https://github.com/oneapi-src/oneDPL/pull/1513 It is we can do the best on the current moment.

> For other tested APIs we preferred to adjust libc++ tests, Have you looked at https://github.com/llvm/llvm-project/tree/main/libcxx/test/std/ranges? Does it make sense to use these for testing, and if not - why?...

> RT asserts Update: it seems that works, it leads to std::abort call... (https://intel.github.io/llvm-docs/design/Assert.html)

> The second concern - there are many files on each std range. Each file checks just one part of functionality of the range. as an example, I have adapted...