Re: ct hardware offload ignores RST packet

[Florian Westphal <fw@strlen.de>]

Chris Mi <cmi@nvidia.com> wrote:
> > nf_tcp_handle_invalid() here resolves the problem as well?
> > Intent would be to reduce timeout but keep connecton state
> > as-is.
> > 
> > I don't think we should force customers to tweak sysctls to
> > make expiry work as intended.
> 
> It doesn't work. The if statement is not executed because the condition
> is not met.
> 
> [Mon Sep 23 18:41:59 2024] nf_tcp_handle_invalid: 756, last_dir: 0, dir: 0,
> last_index: 3

How about relaxing nf_tcp_handle_invalid() to no longer check dir and
last_index?

It already makes sure that timeout can only be reduced by such invalid
fin/rst.

I.e. also get rid of else clause and extra indent level.

> Even if the if statement is executed, the timeout is still not changed.

Hmm, why not? Can you elaborate? Is the timeout already below 2 minutes?
If so, what is the exact expectation?

Could you propose a patch? As I said, I dislike tying this to sysctls.