Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::3:1 as permitted sender) client-ip=2620:137:e000::3:1;
MIME-Version: 1.0
References: <4a096ba9cf981a588aa87235bb91e933ee162b3d.1695542544.git.christophe.jaillet@wanadoo.fr>
 <65105d3a8e70e_12c73e29410@willemb.c.googlers.com.notmuch> <f705117e-41dd-cb2f-ed06-6c47876fd6a2@wanadoo.fr>
In-Reply-To: <f705117e-41dd-cb2f-ed06-6c47876fd6a2@wanadoo.fr>
From:   Willem de Bruijn <willemdebruijn.kernel@gmail.com>
Date:   Mon, 25 Sep 2023 22:50:24 +0200
Message-ID: <CAF=yD-+8Q19sO4XC0dNSiYWSCj03rS9tfV+mVjvZOBnnwpnS5g@mail.gmail.com>
Subject: Re: [PATCH net-next] udp_tunnel: Use flex array to simplify code
To:     Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Cc:     "David S. Miller" <davem@davemloft.net>,
        David Ahern <dsahern@kernel.org>,
        Eric Dumazet <edumazet@google.com>,
        Jakub Kicinski <kuba@kernel.org>,
        Paolo Abeni <pabeni@redhat.com>, linux-kernel@vger.kernel.org,
        kernel-janitors@vger.kernel.org, netdev@vger.kernel.org
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Precedence: bulk

On Mon, Sep 25, 2023 at 6:26=E2=80=AFPM Christophe JAILLET
<christophe.jaillet@wanadoo.fr> wrote:
>
> Le 24/09/2023 =C3=A0 18:00, Willem de Bruijn a =C3=A9crit :
> > Christophe JAILLET wrote:
> >> 'n_tables' is small, UDP_TUNNEL_NIC_MAX_TABLES       =3D 4 as a maximu=
m. So there
> >> is no real point to allocate the 'entries' pointers array with a dedic=
ate
> >> memory allocation.
> >>
> >> Using a flexible array for struct udp_tunnel_nic->entries avoids the
> >> overhead of an additional memory allocation.
> >>
> >> This also saves an indirection when the array is accessed.
> >>
> >> Finally, __counted_by() can be used for run-time bounds checking if
> >> configured and supported by the compiler.
> >>
> >> Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>

Reviewed-by: Willem de Bruijn <willemb@google.com>

> >> ---
> >>   net/ipv4/udp_tunnel_nic.c | 11 ++---------
> >>   1 file changed, 2 insertions(+), 9 deletions(-)
> >>
> >> diff --git a/net/ipv4/udp_tunnel_nic.c b/net/ipv4/udp_tunnel_nic.c
> >> index 029219749785..b6d2d16189c0 100644
> >> --- a/net/ipv4/udp_tunnel_nic.c
> >> +++ b/net/ipv4/udp_tunnel_nic.c
> >> @@ -47,7 +47,7 @@ struct udp_tunnel_nic {
> >>
> >>      unsigned int n_tables;
> >>      unsigned long missed;
> >> -    struct udp_tunnel_nic_table_entry **entries;
> >> +    struct udp_tunnel_nic_table_entry *entries[] __counted_by(n_table=
s);
> >>   };
> >>
> >>   /* We ensure all work structs are done using driver state, but not t=
he code.
> >> @@ -725,16 +725,12 @@ udp_tunnel_nic_alloc(const struct udp_tunnel_nic=
_info *info,
> >>      struct udp_tunnel_nic *utn;
> >>      unsigned int i;
> >>
> >> -    utn =3D kzalloc(sizeof(*utn), GFP_KERNEL);
> >> +    utn =3D kzalloc(struct_size(utn, entries, n_tables), GFP_KERNEL);
> >>      if (!utn)
> >>              return NULL;
> >>      utn->n_tables =3D n_tables;
> >
> > Should utn->n_tables be initialized before first use of
> > struct_size(utn, entries, n_tables)?
> >
>
> It can't be.
> struct_size() is used to compute the memory size to allocate.
>
> Before the kzalloc() call, utn does not exist, so we can't write
> anything to utn->n_tables. It is undefined at this point.
>
> It is initialized the line just after, after the allocation, but before
> any use.

Of course, sorry. I confused __counted_by's reference to a field
member in the struct, with the normal argument passing of struct_size
and flex_array_size.