On Tue, 2023-02-14 at 14:54 +0100, Ard Biesheuvel wrote:
[...]
> and the requirement to use
> scatterlists for skciphers is especially horrid.
Just by way of irony, we do have one place that does use a two element
scatterlist. It's how I found this bug: 95ec01ba1ef0 ("crypto: ecdh -
fix to allow multi segment scatterlists"). And that does mean I could
do with a library version of ecdh with the nist P-256 curve ... pretty
please ...
Regards,
James
