Re: [PATCH libnftables] Add support for ct set

[Kristian Evensen <kristian.evensen@gmail.com>]

Hi,

On Fri, Jan 10, 2014 at 2:52 PM, Patrick McHardy <kaber@trash.net> wrote:
> Just to be clear, I'm not only talking about the naming, a statement
> is something fundamentally different. Every expression has a value,
> which a statement doesn't have. Therefore every expression has a
> a value, destination register, byteorder, ..., all of which don't
> apply to statements. This is why we have an entire separate class
> for this in nftables, and this is what I'm also proposing for
> libnftables.

Thanks for the clarification and I agree. "ct set" is already part of
ct_stmt in nftables, but I was unsure of how to transfer this to
libnftables. I therefor decided to follow meta and store sreg in
nf_expr_ct.

-Kristian