Re: [PATCH 16/22] fwd: Helpers to clarify what host addresses aren't guest accessible

[Stefano Brivio <sbrivio@redhat.com>]

On Fri, 16 Aug 2024 15:39:57 +1000
David Gibson <david@gibson.dropbear.id.au> wrote:

> We usually avoid NAT, but in a few cases we need to apply address
> translations.  For inbound connections that happens for addresses which
> make sense to the host but are either inaccessible, or mean a different
> location from the guest's point of view.
> 
> Add some helper functions to determine such addresses, and use them in
> fwd_nat_from_host().  In doing so clarify some of the reasons for the
> logic.  We'll also have further use for these helpers in future.
> 
> While we're there fix one unneccessary inconsistency between IPv4 and IPv6.
> We always translated the guest's observed address, but for IPv4 we didn't
> translate the guest's assigned address, whereas for IPv6 we did.  Change
> this to translate both in all cases for consistency.
> 
> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
> ---
>  fwd.c | 98 ++++++++++++++++++++++++++++++++++++++++++++++++++++-------
>  1 file changed, 87 insertions(+), 11 deletions(-)
> 
> diff --git a/fwd.c b/fwd.c
> index 75dc0151..1baae338 100644
> --- a/fwd.c
> +++ b/fwd.c
> @@ -170,6 +170,85 @@ static bool is_dns_flow(uint8_t proto, const struct flowside *ini)
>  		((ini->oport == 53) || (ini->oport == 853));
>  }
>  
> +/**
> + * fwd_guest_accessible4() - Is IPv4 address guest accessible

Nit: I wonder if we should say "guest-accessible" in all these cases,
it's a bit easier for me to decode, but not necessarily more correct.
It's fine by me either way.

-- 
Stefano