linux-kernel - Re: [PATCH bpf-next] bpf: Add document for 'dst_port' of 'struct bpf

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <CADxym3b-Q6LyjKqTFcrssK9dVJ8hL6QkMb0MzLyn64r4LS=xtw@mail.gmail.com>
Date:   Thu, 20 Jan 2022 22:14:14 +0800
From:   Menglong Dong <menglong8.dong@...il.com>
To:     Alexei Starovoitov <alexei.starovoitov@...il.com>
Cc:     Alexei Starovoitov <ast@...nel.org>,
        Daniel Borkmann <daniel@...earbox.net>,
        Andrii Nakryiko <andrii@...nel.org>,
        Martin KaFai Lau <kafai@...com>,
        Song Liu <songliubraving@...com>, Yonghong Song <yhs@...com>,
        John Fastabend <john.fastabend@...il.com>,
        KP Singh <kpsingh@...nel.org>,
        Network Development <netdev@...r.kernel.org>,
        bpf <bpf@...r.kernel.org>, LKML <linux-kernel@...r.kernel.org>,
        Mengen Sun <mengensun@...cent.com>, flyingpeng@...cent.com,
        mungerjiang@...cent.com, Menglong Dong <imagedong@...cent.com>
Subject: Re: [PATCH bpf-next] bpf: Add document for 'dst_port' of 'struct bpf_sock'

On Thu, Jan 20, 2022 at 12:17 PM Alexei Starovoitov
<alexei.starovoitov@...il.com> wrote:
>
> On Thu, Jan 20, 2022 at 11:02:27AM +0800, Menglong Dong wrote:
> > Hello!
> >
> > On Thu, Jan 20, 2022 at 6:03 AM Alexei Starovoitov
> > <alexei.starovoitov@...il.com> wrote:
> > >
> > [...]
> > >
> > > Looks like
> > >  __sk_buff->remote_port
> > >  bpf_sock_ops->remote_port
> > >  sk_msg_md->remote_port
> > > are doing the right thing,
> > > but bpf_sock->dst_port is not correct?
> > >
> > > I think it's better to fix it,
> > > but probably need to consolidate it with
> > > convert_ctx_accesses() that deals with narrow access.
> > > I suspect reading u8 from three flavors of 'remote_port'
> > > won't be correct.
> >
> > What's the meaning of 'narrow access'? Do you mean to
> > make 'remote_port' u16? Or 'remote_port' should be made
> > accessible with u8? In fact, '*((u16 *)&skops->remote_port + 1)'
> > won't work, as it only is accessible with u32.
>
> u8 access to remote_port won't pass the verifier,
> but u8 access to dst_port will.
> Though it will return incorrect data.
> See how convert_ctx_accesses() handles narrow loads.
> I think we need to generalize it for different endian fields.

Yeah, I understand narrower load in convert_ctx_accesses()
now. Seems u8 access to dst_port can't pass the verifier too,
which can be seen form bpf_sock_is_valid_access():

$    switch (off) {
$    case offsetof(struct bpf_sock, state):
$    case offsetof(struct bpf_sock, family):
$    case offsetof(struct bpf_sock, type):
$    case offsetof(struct bpf_sock, protocol):
$    case offsetof(struct bpf_sock, dst_port):  // u8 access is not allowed
$    case offsetof(struct bpf_sock, src_port):
$    case offsetof(struct bpf_sock, rx_queue_mapping):
$    case bpf_ctx_range(struct bpf_sock, src_ip4):
$    case bpf_ctx_range_till(struct bpf_sock, src_ip6[0], src_ip6[3]):
$    case bpf_ctx_range(struct bpf_sock, dst_ip4):
$    case bpf_ctx_range_till(struct bpf_sock, dst_ip6[0], dst_ip6[3]):
$        bpf_ctx_record_field_size(info, size_default);
$        return bpf_ctx_narrow_access_ok(off, size, size_default);
$    }

I'm still not sure what should we do now. Should we make all
remote_port and dst_port narrower accessable and endianness
right? For example the remote_port in struct bpf_sock_ops:

--- a/net/core/filter.c
+++ b/net/core/filter.c
@@ -8414,6 +8414,7 @@ static bool sock_ops_is_valid_access(int off, int size,
                                return false;
                        info->reg_type = PTR_TO_PACKET_END;
                        break;
+               case bpf_ctx_range(struct bpf_sock_ops, remote_port):
                case offsetof(struct bpf_sock_ops, skb_tcp_flags):
                        bpf_ctx_record_field_size(info, size_default);
                        return bpf_ctx_narrow_access_ok(off, size,

If remote_port/dst_port are made narrower accessable, the
result will be right. Therefore, *((u16*)&sk->remote_port) will
be the port with network byte order. And the port in host byte
order can be get with:
bpf_ntohs(*((u16*)&sk->remote_port))
or
bpf_htonl(sk->remote_port)

Thanks!
Menglong Dong