All of lore.kernel.org
 help / color / mirror / Atom feed
From: Daniel Axtens <dja@axtens.net>
To: Charles Duffy <charles@dyfis.net>, grub-devel@gnu.org
Subject: Re: [PATCH] pgp: Recognize issuer subpackets in either hashed or unhashed sections
Date: Wed, 10 Jun 2020 14:29:30 +1000	[thread overview]
Message-ID: <87d067tufp.fsf@dja-thinkpad.axtens.net> (raw)
In-Reply-To: <CAEHm_shGEqkE7cRvZV9u5+9fwHA=DJskkxPgje-9RaAkM1o7aA@mail.gmail.com>

Charles Duffy <charles@dyfis.net> writes:

> Looking at the mailing list archive, it appears that the patch may have
> been munged by line wrap (unclear as to why; used `mutt -H` to send `git
> format-patch`'s output).

I use git-send-email, that might give better results?

Anyway I think it might be worth adding a test to signature_test. I know
the test suite is a bit crufty so it may end up being too big an
undertaking, but I think it would be good to at least try.

Regards,
Daniel

>
> Regardless, if it doesn't apply for anyone, a byte-for-byte unmodified copy
> can also be found at
> https://raw.githubusercontent.com/charles-dyfis-net/grub-openpgp-compat-test/fa5029ee0c2a8be073b05245c247c2610b5f864d/0001-pgp-Recognize-issuer-subpackets-in-either-hashed-or-.patch
> <https://raw.githubusercontent.com/charles-dyfis-net/grub-openpgp-compat-test/master/0001-pgp-Recognize-issuer-subpackets-in-either-hashed-or-.patch>
>
> Thanks/apologies/&c.,
> -- Charles
>
> On Sat, May 30, 2020 at 4:20 PM Charles Duffy <charles@dyfis.net> wrote:
>
>> Currently, GRUB's OpenPGP signature parsing searches for the issuer
>> field (specifying the key to use) only in the unhashed portion of the
>> signature.
>>
>> RFC 4880 permits almost all fields (with the sole exception of signature
>> creation time, which MUST be recognized only in the hashed area) to be
>> present either in hashed or unhashed areas; and specifies that
>> implementations should use the unhashed area only for "advisory
>> information".
>>
>> While GnuPG's decision to consider issuer ID advisory is defensible
>> (after all, one could simply do an exhaustive search of known public
>> keys in its absence), it is not the only valid decision; in particular,
>> the Go x/crypto/openpgp library chooses to store issuer ID in the hashed
>> area.
>>
>> Without this patch, trying to verify a valid signature made by
>> x/crypto/openpgp results in `error: public key 00000000 not found.`,
>> because the `keyid` variable is unpopulated.
>>
>> This patch, originally written by Ignat Korchagin and ported to GRUB
>> 2.04 by Daniel Axtens, remedies this. I (Charles Duffy) have tried to
>> address review comments on the original requesting that named constants
>> be used to enhance readability.
>>
>> There are still outstanding compatibility issues parsing public keys
>> not serialized by GnuPG; these may be addressed in a later patch.
>>
>> Signed-off-by: Ignat Korchagin <ignat@cloudflare.com>
>> Signed-off-by: Daniel Axtens <dja@axtens.net>
>> Signed-off-by: Charles Duffy <charles@dyfis.net>
>> ---
>>  grub-core/commands/pgp.c | 137 +++++++++++++++++++++++++++------------
>>  1 file changed, 97 insertions(+), 40 deletions(-)
>>
>> diff --git a/grub-core/commands/pgp.c b/grub-core/commands/pgp.c
>> index bbf6871fe..b49b997c4 100644
>> --- a/grub-core/commands/pgp.c
>> +++ b/grub-core/commands/pgp.c
>> @@ -39,6 +39,17 @@ enum
>>      OPTION_SKIP_SIG = 0
>>    };
>>
>> +enum
>> +  {
>> +    OPENPGP_SIG_SUBPACKET_TYPE_ISSUER = 16 /* subpacket contains 8-octet
>> key id, see RFC4880 5.2.3.5 */
>> +  };
>> +
>> +enum
>> +  {
>> +    OPENPGP_SIG_SUBPACKET_LEN_2BYTE_LIMIT = 192,
>> +    OPENPGP_SIG_SUBPACKET_LEN_5BYTE_CONST = 255
>> +  };
>> +
>>  static const struct grub_arg_option options[] =
>>    {
>>      {"skip-sig", 's', 0,
>> @@ -448,6 +459,47 @@ struct grub_pubkey_context
>>    void *hash_context;
>>  };
>>
>> +static grub_uint64_t
>> +grub_subpacket_keyid_search (const grub_uint8_t * sub, grub_ssize_t
>> sub_len)
>> +{
>> +  const grub_uint8_t *ptr;
>> +  grub_uint32_t l;
>> +  grub_uint64_t keyid = 0;
>> +
>> +  for (ptr = sub; ptr < sub + sub_len; ptr += l)
>> +    {
>> +      /* this algorithm is expressely given in RFC 4880 5.2.3.1 to parse
>> length
>> +       * specifications, which may be 1-byte, 2-byte or 5-bytes long */
>> +      if (*ptr < OPENPGP_SIG_SUBPACKET_LEN_2BYTE_LIMIT)
>> +       l = *ptr++;
>> +      /* 2-octet length field; the two high bits used to specify this
>> format
>> +       * are not part of the data, and the value as a whole is offset to
>> avoid
>> +       * having multiple ways to specify values that would fit in the
>> 1-byte
>> +       * form */
>> +      else if (*ptr < OPENPGP_SIG_SUBPACKET_LEN_5BYTE_CONST)
>> +       {
>> +         if (ptr + 1 >= sub + sub_len)
>> +           break;
>> +         l = (((ptr[0] - OPENPGP_SIG_SUBPACKET_LEN_2BYTE_LIMIT) <<
>> GRUB_CHAR_BIT) | ptr[1])
>> +                + OPENPGP_SIG_SUBPACKET_LEN_2BYTE_LIMIT;
>> +         ptr += 2;
>> +       }
>> +      /* 5-octet length field, 0xff constant followed by 4-byte value */
>> +      else
>> +       {
>> +         if (ptr + 5 >= sub + sub_len)
>> +           break;
>> +         l = grub_be_to_cpu32 (grub_get_unaligned32 (ptr + 1));
>> +         ptr += 5;
>> +       }
>> +      /* determine whether we found the packet we're looking for */
>> +      if (*ptr == OPENPGP_SIG_SUBPACKET_TYPE_ISSUER && l >= 8)
>> +       keyid = grub_get_unaligned64 (ptr + 1);
>> +    }
>> +
>> +  return keyid;
>> +}
>> +
>>  static grub_err_t
>>  grub_verify_signature_init (struct grub_pubkey_context *ctxt, grub_file_t
>> sig)
>>  {
>> @@ -520,7 +572,7 @@ grub_verify_signature_real (struct grub_pubkey_context
>> *ctxt,
>>    gcry_mpi_t mpis[10];
>>    grub_uint8_t pk = ctxt->v4.pkeyalgo;
>>    grub_size_t i;
>> -  grub_uint8_t *readbuf = NULL;
>> +  grub_uint8_t *readbuf = NULL, *subpacket_buf = NULL;
>>    unsigned char *hval;
>>    grub_ssize_t rem = grub_be_to_cpu16 (ctxt->v4.hashed_sub);
>>    grub_uint32_t headlen = grub_cpu_to_be32 (rem + 6);
>> @@ -538,17 +590,29 @@ grub_verify_signature_real (struct
>> grub_pubkey_context *ctxt,
>>
>>    ctxt->hash->write (ctxt->hash_context, &ctxt->v, sizeof (ctxt->v));
>>    ctxt->hash->write (ctxt->hash_context, &ctxt->v4, sizeof (ctxt->v4));
>> -  while (rem)
>> +
>> +  subpacket_buf = grub_malloc (rem);
>> +  if (!subpacket_buf)
>> +    goto fail;
>> +
>> +  r = 0;
>> +  while (r < rem)
>>      {
>> -      r = grub_file_read (ctxt->sig, readbuf,
>> -                         rem < READBUF_SIZE ? rem : READBUF_SIZE);
>> -      if (r < 0)
>> -       goto fail;
>> -      if (r == 0)
>> +      grub_ssize_t rr = grub_file_read (ctxt->sig, subpacket_buf + r, rem
>> - r);
>> +      if (rr < 0)
>> +        goto fail;
>> +      if (rr == 0)
>>         break;
>> -      ctxt->hash->write (ctxt->hash_context, readbuf, r);
>> -      rem -= r;
>> +      r += rr;
>>      }
>> +  if (r != rem)
>> +    goto fail;
>> +  ctxt->hash->write (ctxt->hash_context, subpacket_buf, rem);
>> +
>> +  keyid = grub_subpacket_keyid_search (subpacket_buf, rem);
>> +  grub_free (subpacket_buf);
>> +  subpacket_buf = NULL;
>> +
>>    ctxt->hash->write (ctxt->hash_context, &ctxt->v, sizeof (ctxt->v));
>>    s = 0xff;
>>    ctxt->hash->write (ctxt->hash_context, &s, sizeof (s));
>> @@ -556,37 +620,27 @@ grub_verify_signature_real (struct
>> grub_pubkey_context *ctxt,
>>    r = grub_file_read (ctxt->sig, &unhashed_sub, sizeof (unhashed_sub));
>>    if (r != sizeof (unhashed_sub))
>>      goto fail;
>> -  {
>> -    grub_uint8_t *ptr;
>> -    grub_uint32_t l;
>> -    rem = grub_be_to_cpu16 (unhashed_sub);
>> -    if (rem > READBUF_SIZE)
>> -      goto fail;
>> -    r = grub_file_read (ctxt->sig, readbuf, rem);
>> -    if (r != rem)
>> -      goto fail;
>> -    for (ptr = readbuf; ptr < readbuf + rem; ptr += l)
>> -      {
>> -       if (*ptr < 192)
>> -         l = *ptr++;
>> -       else if (*ptr < 255)
>> -         {
>> -           if (ptr + 1 >= readbuf + rem)
>> -             break;
>> -           l = (((ptr[0] & ~192) << GRUB_CHAR_BIT) | ptr[1]) + 192;
>> -           ptr += 2;
>> -         }
>> -       else
>> -         {
>> -           if (ptr + 5 >= readbuf + rem)
>> -             break;
>> -           l = grub_be_to_cpu32 (grub_get_unaligned32 (ptr + 1));
>> -           ptr += 5;
>> -         }
>> -       if (*ptr == 0x10 && l >= 8)
>> -         keyid = grub_get_unaligned64 (ptr + 1);
>> -      }
>> -  }
>> +
>> +  rem = grub_be_to_cpu16 (unhashed_sub);
>> +  subpacket_buf = grub_malloc (rem);
>> +  if (!subpacket_buf)
>> +    goto fail;
>> +
>> +  r = 0;
>> +  while (r < rem)
>> +    {
>> +     grub_ssize_t rr = grub_file_read (ctxt->sig, subpacket_buf + r, rem
>> - r);
>> +     if (rr < 0)
>> +       goto fail;
>> +     if (rr == 0)
>> +       break;
>> +     r += rr;
>> +    }
>> +  if (r != rem)
>> +    goto fail;
>> +
>> +  if (keyid == 0)
>> +    keyid = grub_subpacket_keyid_search (subpacket_buf, rem);
>>
>>    ctxt->hash->final (ctxt->hash_context);
>>
>> @@ -656,10 +710,13 @@ grub_verify_signature_real (struct
>> grub_pubkey_context *ctxt,
>>      goto fail;
>>
>>    grub_free (readbuf);
>> +  grub_free (subpacket_buf);
>>
>>    return GRUB_ERR_NONE;
>>
>>   fail:
>> +  if (subpacket_buf)
>> +    grub_free (subpacket_buf);
>>    grub_free (readbuf);
>>    if (!grub_errno)
>>      return grub_error (GRUB_ERR_BAD_SIGNATURE, N_("bad signature"));
>> --
>> 2.25.4
>>
>>
> _______________________________________________
> Grub-devel mailing list
> Grub-devel@gnu.org
> https://lists.gnu.org/mailman/listinfo/grub-devel


      reply	other threads:[~2020-06-10  4:29 UTC|newest]

Thread overview: 3+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-05-30 21:20 [PATCH] pgp: Recognize issuer subpackets in either hashed or unhashed sections Charles Duffy
2020-05-31 17:51 ` Charles Duffy
2020-06-10  4:29   ` Daniel Axtens [this message]

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=87d067tufp.fsf@dja-thinkpad.axtens.net \
    --to=dja@axtens.net \
    --cc=charles@dyfis.net \
    --cc=grub-devel@gnu.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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.