Re: [PATCH net-next v5 4/6] tls: update the RX key when a KeyUpdate is received

[Martin Belanger]

On 11/4/26 16:43, sashiko-bot@xxxxxxxxxx wrote:
> > +   memcpy(crypto_info_salt(info, cipher_desc), iv, cipher_desc->salt);
> > +   memcpy(crypto_info_iv(info, cipher_desc), iv + cipher_desc->salt,
> > +          cipher_desc->iv);
>
> [Severity: High]
> Won't these memcpy() calls trigger a CONFIG_FORTIFY_SOURCE field-spanning
> write warning when tls_check_pending_rekey() -> tls_keyupdate_rx() ->
> tls_keyupdate_rekey() runs on a socket with a registered traffic secret?

I could not reproduce this. I built with CONFIG_FORTIFY_SOURCE=y and
GCC 15.2 and ran the keyupdate selftest with TLS 1.3 AES-GCM. There
was no warning. In this build, __member_size() and __struct_size() of
both destinations are unknown (-1), so the field check is not compiled
in.

The result depends on the compiler. A small user space copy of this
pattern gives a member size of 0 with GCC. In the next version I will
take the info pointer from the union itself, as c43c1c24f3cd does for
the same reason.

Martin