netdev.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Roberto Sassu <roberto.sassu@huawei.com>
To: KP Singh <kpsingh@kernel.org>, Daniel Borkmann <daniel@iogearbox.net>
Cc: "ast@kernel.org" <ast@kernel.org>,
	"andrii@kernel.org" <andrii@kernel.org>,
	"bpf@vger.kernel.org" <bpf@vger.kernel.org>,
	"netdev@vger.kernel.org" <netdev@vger.kernel.org>,
	"linux-kselftest@vger.kernel.org"
	<linux-kselftest@vger.kernel.org>,
	"linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
	"john.fastabend@gmail.com" <john.fastabend@gmail.com>
Subject: RE: [PATCH v2 1/3] bpf: Add bpf_verify_pkcs7_signature() helper
Date: Wed, 8 Jun 2022 15:13:40 +0000	[thread overview]
Message-ID: <48b67de1c99f4b9f97a12016e6e99081@huawei.com> (raw)
In-Reply-To: <CACYkzJ4VU+JQzsCZHgeAY9Aej5W_k7bJFSeDP93Nq=uM_v7c8Q@mail.gmail.com>

> From: KP Singh [mailto:kpsingh@kernel.org]
> Sent: Wednesday, June 8, 2022 4:45 PM
> On Wed, Jun 8, 2022 at 4:43 PM Daniel Borkmann <daniel@iogearbox.net>
> wrote:
> >
> > On 6/8/22 1:12 PM, Roberto Sassu wrote:
> > > Add the bpf_verify_pkcs7_signature() helper, to give the ability to eBPF
> > > security modules to check the validity of a PKCS#7 signature against
> > > supplied data.
> 
> Can we keep the helper generic so that it can be extended to more types of
> signatures and pass the signature type as an enum?
> 
> bpf_verify_signature and a type SIG_PKCS7 or something.

Hi KP

makes sense. Otherwise, we have to add a new helper every time
a new signature verification function is introduced (for example
one would be needed for PGP).

I will reuse enum pkey_id_type in module_signature.h

Thanks

Roberto

HUAWEI TECHNOLOGIES Duesseldorf GmbH, HRB 56063
Managing Director: Li Peng, Yang Xi, Li He

> > > Use the 'keyring' parameter to select the keyring containing the
> > > verification key: 0 for the primary keyring, 1 for the primary and
> > > secondary keyrings, 2 for the platform keyring.
> > >
> > > Signed-off-by: Roberto Sassu <roberto.sassu@huawei.com>
> > > ---
> > >   include/uapi/linux/bpf.h       |  8 ++++++++
> > >   kernel/bpf/bpf_lsm.c           | 32 ++++++++++++++++++++++++++++++++
> > >   tools/include/uapi/linux/bpf.h |  8 ++++++++
> > >   3 files changed, 48 insertions(+)
> > >
> > > diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h
> > > index f4009dbdf62d..40d0fc0d9493 100644
> > > --- a/include/uapi/linux/bpf.h
> > > +++ b/include/uapi/linux/bpf.h
> > > @@ -5249,6 +5249,13 @@ union bpf_attr {
> > >    *          Pointer to the underlying dynptr data, NULL if the dynptr is
> > >    *          read-only, if the dynptr is invalid, or if the offset and length
> > >    *          is out of bounds.
> > > + *
> > > + * long bpf_verify_pkcs7_signature(u8 *data, u32 datalen, u8 *sig, u32
> siglen, u64 keyring)
> > > + *   Description
> > > + *           Verify the PKCS#7 *sig* with length *siglen*, on *data* with
> > > + *           length *datalen*, with key in *keyring*.
> >
> > Could you also add a description for users about the keyring argument and
> guidance on when
> > they should use which in their programs? Above is a bit too terse, imho.
> >
> > > + *   Return
> > > + *           0 on success, a negative value on error.
> > >    */
> > >   #define __BPF_FUNC_MAPPER(FN)               \
> > >       FN(unspec),                     \
> > > @@ -5455,6 +5462,7 @@ union bpf_attr {
> > >       FN(dynptr_read),                \
> > >       FN(dynptr_write),               \
> > >       FN(dynptr_data),                \
> > > +     FN(verify_pkcs7_signature),     \
> > >       /* */
> > >
> > >   /* integer value in 'imm' field of BPF_CALL instruction selects which helper
> > > diff --git a/kernel/bpf/bpf_lsm.c b/kernel/bpf/bpf_lsm.c
> > > index c1351df9f7ee..1cda43cb541a 100644
> > > --- a/kernel/bpf/bpf_lsm.c
> > > +++ b/kernel/bpf/bpf_lsm.c
> > > @@ -16,6 +16,7 @@
> > >   #include <linux/bpf_local_storage.h>
> > >   #include <linux/btf_ids.h>
> > >   #include <linux/ima.h>
> > > +#include <linux/verification.h>
> > >
> > >   /* For every LSM hook that allows attachment of BPF programs, declare a
> nop
> > >    * function where a BPF program can be attached.
> > > @@ -132,6 +133,35 @@ static const struct bpf_func_proto
> bpf_get_attach_cookie_proto = {
> > >       .arg1_type      = ARG_PTR_TO_CTX,
> > >   };
> > >
> > > +BPF_CALL_5(bpf_verify_pkcs7_signature, u8 *, data, u32, datalen, u8 *, sig,
> > > +        u32, siglen, u64, keyring)
> > > +{
> > > +     int ret = -EOPNOTSUPP;
> > > +
> > > +#ifdef CONFIG_SYSTEM_DATA_VERIFICATION
> > > +     if (keyring > (unsigned long)VERIFY_USE_PLATFORM_KEYRING)
> > > +             return -EINVAL;
> > > +
> > > +     ret = verify_pkcs7_signature(data, datalen, sig, siglen,
> > > +                                  (struct key *)keyring,
> > > +                                  VERIFYING_UNSPECIFIED_SIGNATURE, NULL,
> > > +                                  NULL);
> > > +#endif
> > > +     return ret;
> > > +}
> >
> > Looks great! One small nit, I would move all of the BPF_CALL and _proto under
> the
> > #ifdef CONFIG_SYSTEM_DATA_VERIFICATION ...
> >
> > > +static const struct bpf_func_proto bpf_verify_pkcs7_signature_proto = {
> > > +     .func           = bpf_verify_pkcs7_signature,
> > > +     .gpl_only       = false,
> > > +     .ret_type       = RET_INTEGER,
> > > +     .arg1_type      = ARG_PTR_TO_MEM,
> > > +     .arg2_type      = ARG_CONST_SIZE_OR_ZERO,
> > > +     .arg3_type      = ARG_PTR_TO_MEM,
> > > +     .arg4_type      = ARG_CONST_SIZE_OR_ZERO,
> > > +     .arg5_type      = ARG_ANYTHING,
> > > +     .allowed        = bpf_ima_inode_hash_allowed,
> > > +};
> > > +
> > >   static const struct bpf_func_proto *
> > >   bpf_lsm_func_proto(enum bpf_func_id func_id, const struct bpf_prog
> *prog)
> > >   {
> > > @@ -158,6 +188,8 @@ bpf_lsm_func_proto(enum bpf_func_id func_id,
> const struct bpf_prog *prog)
> > >               return prog->aux->sleepable ? &bpf_ima_file_hash_proto : NULL;
> > >       case BPF_FUNC_get_attach_cookie:
> > >               return bpf_prog_has_trampoline(prog) ?
> &bpf_get_attach_cookie_proto : NULL;
> > > +     case BPF_FUNC_verify_pkcs7_signature:
> > > +             return prog->aux->sleepable ? &bpf_verify_pkcs7_signature_proto :
> NULL;
> >
> > ... same here:
> >
> > #ifdef CONFIG_SYSTEM_DATA_VERIFICATION
> >         case BPF_FUNC_verify_pkcs7_signature:
> >                 return prog->aux->sleepable ? &bpf_verify_pkcs7_signature_proto :
> NULL;
> > #endif
> >
> > So that bpftool or other feature probes can check for its availability.
> Otherwise, apps have
> > a hard time checking whether bpf_verify_pkcs7_signature() helper is available
> for use or not.
> >
> > >       default:
> > >               return tracing_prog_func_proto(func_id, prog);
> > >       }
> > > diff --git a/tools/include/uapi/linux/bpf.h b/tools/include/uapi/linux/bpf.h
> > > index f4009dbdf62d..40d0fc0d9493 100644
> > > --- a/tools/include/uapi/linux/bpf.h
> > > +++ b/tools/include/uapi/linux/bpf.h
> > > @@ -5249,6 +5249,13 @@ union bpf_attr {
> > >    *          Pointer to the underlying dynptr data, NULL if the dynptr is
> > >    *          read-only, if the dynptr is invalid, or if the offset and length
> > >    *          is out of bounds.
> > > + *
> > > + * long bpf_verify_pkcs7_signature(u8 *data, u32 datalen, u8 *sig, u32
> siglen, u64 keyring)
> > > + *   Description
> > > + *           Verify the PKCS#7 *sig* with length *siglen*, on *data* with
> > > + *           length *datalen*, with key in *keyring*.
> > > + *   Return
> > > + *           0 on success, a negative value on error.
> > >    */
> > >   #define __BPF_FUNC_MAPPER(FN)               \
> > >       FN(unspec),                     \
> > > @@ -5455,6 +5462,7 @@ union bpf_attr {
> > >       FN(dynptr_read),                \
> > >       FN(dynptr_write),               \
> > >       FN(dynptr_data),                \
> > > +     FN(verify_pkcs7_signature),     \
> > >       /* */
> > >
> > >   /* integer value in 'imm' field of BPF_CALL instruction selects which helper
> > >
> >

  reply	other threads:[~2022-06-08 15:21 UTC|newest]

Thread overview: 13+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-06-08 11:12 [PATCH v2 0/3] bpf: Add bpf_verify_pkcs7_signature() helper Roberto Sassu
2022-06-08 11:12 ` [PATCH v2 1/3] " Roberto Sassu
2022-06-08 14:43   ` Daniel Borkmann
2022-06-08 14:44     ` KP Singh
2022-06-08 15:13       ` Roberto Sassu [this message]
2022-06-08 15:09     ` Roberto Sassu
2022-06-08 14:48   ` kernel test robot
2022-06-08 11:12 ` [PATCH v2 2/3] selftests/bpf: Add test_progs opts for sign-file and kernel priv key + cert Roberto Sassu
2022-06-09  0:12   ` Alexei Starovoitov
2022-06-09  9:00     ` Roberto Sassu
2022-06-09 15:38       ` Alexei Starovoitov
2022-06-10 12:10         ` Roberto Sassu
2022-06-08 11:12 ` [PATCH v2 3/3] selftests/bpf: Add test for bpf_verify_pkcs7_signature() helper Roberto Sassu

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=48b67de1c99f4b9f97a12016e6e99081@huawei.com \
    --to=roberto.sassu@huawei.com \
    --cc=andrii@kernel.org \
    --cc=ast@kernel.org \
    --cc=bpf@vger.kernel.org \
    --cc=daniel@iogearbox.net \
    --cc=john.fastabend@gmail.com \
    --cc=kpsingh@kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-kselftest@vger.kernel.org \
    --cc=netdev@vger.kernel.org \
    /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).