Netdev Archive on lore.kernel.org
help / color / mirror / Atom feed
From: Lorenz Bauer <lmb@cloudflare.com>
To: Yonghong Song <yhs@fb.com>
Cc: Jakub Sitnicki <jakub@cloudflare.com>,
	John Fastabend <john.fastabend@gmail.com>,
	Alexei Starovoitov <ast@kernel.org>,
	Daniel Borkmann <daniel@iogearbox.net>,
	kernel-team <kernel-team@cloudflare.com>,
	Networking <netdev@vger.kernel.org>, bpf <bpf@vger.kernel.org>,
	LKML <linux-kernel@vger.kernel.org>
Subject: Re: [PATCH bpf-next 4/6] bpf: override the meaning of ARG_PTR_TO_MAP_VALUE for sockmap and sockhash
Date: Thu, 20 Aug 2020 13:33:57 +0100	[thread overview]
Message-ID: <CACAyw9_4_PZ1bHd0W-mAN5b0i-ZriTZSxWtoS59baXdRg6wk0g@mail.gmail.com> (raw)
In-Reply-To: <5d64158b-35ed-d28d-9857-6ee725d287f2@fb.com>

On Wed, 19 Aug 2020 at 21:13, Yonghong Song <yhs@fb.com> wrote:
>
>
>
> On 8/19/20 2:24 AM, Lorenz Bauer wrote:
> > The verifier assumes that map values are simple blobs of memory, and
> > therefore treats ARG_PTR_TO_MAP_VALUE, etc. as such. However, there are
> > map types where this isn't true. For example, sockmap and sockhash store
> > sockets. In general this isn't a big problem: we can just
> > write helpers that explicitly requests PTR_TO_SOCKET instead of
> > ARG_PTR_TO_MAP_VALUE.
> >
> > The one exception are the standard map helpers like map_update_elem,
> > map_lookup_elem, etc. Here it would be nice we could overload the
> > function prototype for different kinds of maps. Unfortunately, this
> > isn't entirely straight forward:
> > We only know the type of the map once we have resolved meta->map_ptr
> > in check_func_arg. This means we can't swap out the prototype
> > in check_helper_call until we're half way through the function.
> >
> > Instead, modify check_func_arg to treat ARG_PTR_TO_MAP_VALUE* to
> > mean "the native type for the map" instead of "pointer to memory"
> > for sockmap and sockhash. This means we don't have to modify the
> > function prototype at all
> >
> > Signed-off-by: Lorenz Bauer <lmb@cloudflare.com>
> > ---
> >   kernel/bpf/verifier.c | 40 ++++++++++++++++++++++++++++++++++++++++
> >   1 file changed, 40 insertions(+)
> >
> > diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c
> > index b6ccfce3bf4c..47f9b94bb9d4 100644
> > --- a/kernel/bpf/verifier.c
> > +++ b/kernel/bpf/verifier.c
> > @@ -3872,6 +3872,38 @@ static int int_ptr_type_to_size(enum bpf_arg_type type)
> >       return -EINVAL;
> >   }
> >
> > +static int override_map_arg_type(struct bpf_verifier_env *env,
> > +                              const struct bpf_call_arg_meta *meta,
> > +                              enum bpf_arg_type *arg_type)
> > +{
> > +     if (!meta->map_ptr) {
> > +             /* kernel subsystem misconfigured verifier */
> > +             verbose(env, "invalid map_ptr to access map->type\n");
> > +             return -EACCES;
> > +     }
> > +
> > +     switch (meta->map_ptr->map_type) {
> > +     case BPF_MAP_TYPE_SOCKMAP:
> > +     case BPF_MAP_TYPE_SOCKHASH:
> > +             switch (*arg_type) {
> > +             case ARG_PTR_TO_MAP_VALUE:
> > +                     *arg_type = ARG_PTR_TO_SOCKET;
> > +                     break;
> > +             case ARG_PTR_TO_MAP_VALUE_OR_NULL:
> > +                     *arg_type = ARG_PTR_TO_SOCKET_OR_NULL;
> > +                     break;
> > +             default:
> > +                     verbose(env, "invalid arg_type for sockmap/sockhash\n");
> > +                     return -EINVAL;
> > +             }
> > +             break;
> > +
> > +     default:
> > +             break;
> > +     }
> > +     return 0;
> > +}
> > +
> >   static int check_func_arg(struct bpf_verifier_env *env, u32 arg,
> >                         struct bpf_call_arg_meta *meta,
> >                         const struct bpf_func_proto *fn)
> > @@ -3904,6 +3936,14 @@ static int check_func_arg(struct bpf_verifier_env *env, u32 arg,
> >               return -EACCES;
> >       }
> >
> > +     if (arg_type == ARG_PTR_TO_MAP_VALUE ||
> > +         arg_type == ARG_PTR_TO_UNINIT_MAP_VALUE ||
> > +         arg_type == ARG_PTR_TO_MAP_VALUE_OR_NULL) {
>
> We probably do not need ARG_PTR_TO_UNINIT_MAP_VALUE here.
>
> Do we need ARG_PTR_TO_MAP_VALUE_OR_NULL? bpf_map_update_elem arg type
> is ARG_PTR_TO_MAP_VALUE.

I did this to be consistent: in a single function definition you
either get the map specific
types or the regular semantics. You don't get to mix and match them.
For the same
reason I included ARG_PTR_TO_UNINIT_MAP_VALUE: the semantics don't make
sense for sockmap, so a function using this doesn't make sense either.

>
> > +             err = override_map_arg_type(env, meta, &arg_type);
> > +             if (err)
> > +                     return err;
> > +     }
> > +
> >       if (arg_type == ARG_PTR_TO_MAP_KEY ||
> >           arg_type == ARG_PTR_TO_MAP_VALUE ||
> >           arg_type == ARG_PTR_TO_UNINIT_MAP_VALUE ||
> >



-- 
Lorenz Bauer  |  Systems Engineer
6th Floor, County Hall/The Riverside Building, SE1 7PB, UK

www.cloudflare.com

  reply	other threads:[~2020-08-20 12:34 UTC|newest]

Thread overview: 25+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-08-19  9:24 [PATCH bpf-next 0/6] Allow updating sockmap / sockhash from BPF Lorenz Bauer
2020-08-19  9:24 ` [PATCH bpf-next 1/6] net: sk_msg: simplify sk_psock initialization Lorenz Bauer
2020-08-19 20:05   ` John Fastabend
2020-08-19  9:24 ` [PATCH bpf-next 2/6] bpf: sockmap: merge sockmap and sockhash update functions Lorenz Bauer
2020-08-19 15:48   ` Jakub Kicinski
2020-08-19 18:50   ` Yonghong Song
2020-08-19 20:11   ` John Fastabend
2020-08-19  9:24 ` [PATCH bpf-next 3/6] bpf: sockmap: call sock_map_update_elem directly Lorenz Bauer
2020-08-19 19:15   ` Yonghong Song
2020-08-19 20:29   ` John Fastabend
2020-08-19  9:24 ` [PATCH bpf-next 4/6] bpf: override the meaning of ARG_PTR_TO_MAP_VALUE for sockmap and sockhash Lorenz Bauer
2020-08-19 20:13   ` Yonghong Song
2020-08-20 12:33     ` Lorenz Bauer [this message]
2020-08-19 20:51   ` John Fastabend
2020-08-19  9:24 ` [PATCH bpf-next 5/6] bpf: sockmap: allow update from BPF Lorenz Bauer
2020-08-19 20:35   ` Yonghong Song
2020-08-19 21:22   ` John Fastabend
2020-08-19 22:41     ` John Fastabend
2020-08-20 11:33       ` Lorenz Bauer
2020-08-20 14:45         ` Yonghong Song
2020-08-19  9:24 ` [PATCH bpf-next 6/6] selftests: bpf: test sockmap " Lorenz Bauer
2020-08-19 20:45   ` Yonghong Song
2020-08-20 11:58     ` Lorenz Bauer
2020-08-20 14:49       ` Yonghong Song
2020-08-20 16:12         ` Lorenz Bauer

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=CACAyw9_4_PZ1bHd0W-mAN5b0i-ZriTZSxWtoS59baXdRg6wk0g@mail.gmail.com \
    --to=lmb@cloudflare.com \
    --cc=ast@kernel.org \
    --cc=bpf@vger.kernel.org \
    --cc=daniel@iogearbox.net \
    --cc=jakub@cloudflare.com \
    --cc=john.fastabend@gmail.com \
    --cc=kernel-team@cloudflare.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=netdev@vger.kernel.org \
    --cc=yhs@fb.com \
    --subject='Re: [PATCH bpf-next 4/6] bpf: override the meaning of ARG_PTR_TO_MAP_VALUE for sockmap and sockhash' \
    /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

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).