Re: [PATCH v2 0/2] MR5020: Draft: kernelbase/file: implement CopyFile2

[Alfred Agrell (＠Alcaro)]

On Mon Feb  5 19:23:01 2024 +0000, Zebediah Figura wrote:
> > Maybe having CopyFile2 functionality but with `SetLastError`-style
>
> error handling, and then write a wrapper over it for `CopyFile2` to
> convert to HRESULT?
> This sounds like what I'd do, at least.

Yes, HRESULT_CODE would work for now.

But if that changes in the future, it'll return random errors like ERROR_BAD_EXE_FORMAT. If you choose that approach, I vote check HRESULT_FACILITY(), and map everything that isn't FACILITY_WIN32 to some random hardcoded error (ERROR_CANNOT_COPY, for example).

Alternatively, do your proposal from yesterday. It sounds cleaner in the long run.