On Thu, Feb 9, 2023 at 11:17 AM Stefan Metzmacher <metze@xxxxxxxxx> wrote:
>
> Any comment about the idea of having a preadv2() flag that
> asks for a dma copy with something like async_memcpy() instead
> of the default that ends up in copy_user_enhanced_fast_string()?
I guarantee that you will only slow things down with some odd async_memcpy.
There are zero DMA engines that do memory copying better than memcpy -
the only way you can do better if is the destination isn't memory in
the first place but the final device, and then we just call it "dma".
Linus
