From: Jakub Kicinski <jakub.kicinski@netronome.com>
To: Willem de Bruijn <willemdebruijn.kernel@gmail.com>
Cc: John Fastabend <john.fastabend@gmail.com>,
David Miller <davem@davemloft.net>,
Network Development <netdev@vger.kernel.org>,
davejwatson@fb.com, borisp@mellanox.com, aviadye@mellanox.com,
Daniel Borkmann <daniel@iogearbox.net>,
Eric Dumazet <edumazet@google.com>,
Alexei Starovoitov <alexei.starovoitov@gmail.com>,
oss-drivers@netronome.com
Subject: Re: [PATCH net v3] net/tls: prevent skb_orphan() from leaking TLS plain text with offload
Date: Thu, 8 Aug 2019 10:31:48 -0700 [thread overview]
Message-ID: <20190808103148.164bec9f@cakuba.netronome.com> (raw)
In-Reply-To: <CA+FuTSc7H6X+rRnxZ5NcFiNy+pw1YCONiUr+K6g800DXzT_0EA@mail.gmail.com>
On Thu, 8 Aug 2019 11:59:18 -0400, Willem de Bruijn wrote:
> > diff --git a/net/tls/tls_device.c b/net/tls/tls_device.c
> > index 7c0b2b778703..43922d86e510 100644
> > --- a/net/tls/tls_device.c
> > +++ b/net/tls/tls_device.c
> > @@ -373,9 +373,9 @@ static int tls_push_data(struct sock *sk,
> > struct tls_context *tls_ctx = tls_get_ctx(sk);
> > struct tls_prot_info *prot = &tls_ctx->prot_info;
> > struct tls_offload_context_tx *ctx = tls_offload_ctx_tx(tls_ctx);
> > - int tls_push_record_flags = flags | MSG_SENDPAGE_NOTLAST;
> > int more = flags & (MSG_SENDPAGE_NOTLAST | MSG_MORE);
> > struct tls_record_info *record = ctx->open_record;
> > + int tls_push_record_flags;
> > struct page_frag *pfrag;
> > size_t orig_size = size;
> > u32 max_open_record_len;
> > @@ -390,6 +390,9 @@ static int tls_push_data(struct sock *sk,
> > if (sk->sk_err)
> > return -sk->sk_err;
> >
> > + flags |= MSG_SENDPAGE_DECRYPTED;
> > + tls_push_record_flags = flags | MSG_SENDPAGE_NOTLAST;
> > +
>
> Without being too familiar with this code: can this plaintext flag be
> set once, closer to the call to do_tcp_sendpages, in tls_push_sg?
>
> Instead of two locations with multiple non-trivial codepaths between
> them and do_tcp_sendpages.
>
> Or are there paths where the flag is not set? Which I imagine would
> imply already passing s/w encrypted ciphertext.
tls_push_sg() is shared with sw path which doesn't have the device
validation.
Device TLS can read tls_push_sg() via tls_push_partial_record() and
tls_push_data(). tls_push_data() is addressed directly here,
tls_push_partial_record() is again shared with SW path, so we have to
address it by adding the flag in tls_device_write_space().
The alternative is to add a conditional to tls_push_sg() which is
a little less nice from performance and layering PoV but it is a lot
simpler..
Should I change?
next prev parent reply other threads:[~2019-08-08 17:32 UTC|newest]
Thread overview: 6+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-08-08 0:03 [PATCH net v3] net/tls: prevent skb_orphan() from leaking TLS plain text with offload Jakub Kicinski
2019-08-08 15:59 ` Willem de Bruijn
2019-08-08 17:31 ` Jakub Kicinski [this message]
2019-08-08 18:10 ` Willem de Bruijn
2019-08-08 19:18 ` Boris Pismenny
2019-08-09 5:40 ` David Miller
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20190808103148.164bec9f@cakuba.netronome.com \
--to=jakub.kicinski@netronome.com \
--cc=alexei.starovoitov@gmail.com \
--cc=aviadye@mellanox.com \
--cc=borisp@mellanox.com \
--cc=daniel@iogearbox.net \
--cc=davejwatson@fb.com \
--cc=davem@davemloft.net \
--cc=edumazet@google.com \
--cc=john.fastabend@gmail.com \
--cc=netdev@vger.kernel.org \
--cc=oss-drivers@netronome.com \
--cc=willemdebruijn.kernel@gmail.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).