* g10/options.h: Add weak_digests linked list to opts.
* g10/main.h: Declare weakhash linked list struct and
additional_weak_digest() function to insert newly-declared weak
digests into opts.
* g10/misc.c: (additional_weak_digest): New function.
(print_digest_algo_note): Check for deprecated digests.
* g10/sig-check.c: (do_check): Reject all weak digests.
* g10/gpg.c: Add --weak-digest option to gpg.
* doc/gpg.texi: Document gpg --weak-digest option.
* g10/gpgv.c: Add --weak-digest option to gpgv.
* doc/gpgv.texi: Document gpgv --weak-digest option.
--
gpg and gpgv treat signatures made over MD5 as unreliable, unless the
user supplies --allow-weak-digests to gpg. Signatures over any other
digest are considered acceptable.
Despite SHA-1 being a mandatory-to-implement digest algorithm in RFC
4880, the collision-resistance of SHA-1 is weaker than anyone would
like it to be.
Some operators of high-value targets that depend on OpenPGP signatures
may wish to require their signers to use a stronger digest algorithm
than SHA1, even if the OpenPGP ecosystem at large cannot deprecate
SHA1 entirely today.
This changeset adds a new "--weak-digest DIGEST" option for both gpg
and gpgv, which makes it straightforward for anyone to treat any
signature or certification made over the specified digest as
unreliable.
This option can be supplied multiple times if the operator wishes to
deprecate multiple digest algorithms, and will be ignored completely
if the operator supplies --allow-weak-digests (as before).
MD5 is always considered weak, regardless of any further
--weak-digest options supplied.
Signed-off-by: Daniel Kahn Gillmor <dkg@fifthhorseman.net>
(this is a rough cherry-pick of applying the following commits to
STABLE-BRANCH-1-4:
76afaed65eb98939812a91015d021b
)
SHA-224 DER prefixes when verifying a signature. See the change
itself for more discussion.
* main.h, seskey.c (do_encode_md): Rename to pkcs1_encode_md and make
non-static.
than 160 bits as per DSA2. This will allow us to verify and issue DSA2
signatures for some backwards compatibility once we start generating DSA2
keys.
* sign.c (do_sign), sig-check.c (do_check): Change all callers.
* sign.c (do_sign): Enforce the 160-bit check for new signatures here
since encode_md_value can handle non-160-bit digests now. This will need
to come out once the standard for DSA2 is firmed up.
is no backsig present. Give a URL for more information.
* keyedit.c (menu_backsign): Small tweak to work properly with keys
originally generated with older GnuPGs that included comments in the
secret keys.
pka-lookups, not pka-lookup.
* options.h, gpg.c (main), keyedit.c [cmds], sig-check.c
(signature_check2): Rename "backsign" to "cross-certify" as a more
accurate name.
keygen.c (make_backsig): Did some backsig interop testing with the PGP
folks. All is well, so I'm turning generation of backsigs on for new
keys. Checking for backsigs on verification is still off.
keyserver to happen even if there is no global keyserver set.
* sig-check.c (do_check_messages): No need to check for Elgamal signatures
any longer. (do_check_messages, do_check, check_key_signature2):
--keyid-format conversion.
* pkclist.c (show_paths, edit_ownertrust): Remove some unused code.
backsigs.
* getkey.c (merge_selfsigs_subkey): Find 0x19 backsigs on subkey selfsigs
and verify they are valid. If DO_BACKSIGS is not defined, fake this as
always valid.
* packet.h, parse-packet.c (parse_signature): Make parse_signature
non-static so we can parse 0x19s in self-sigs.
* main.h, sig-check.c (check_backsig): Check a 0x19 signature.
(signature_check2): Give a backsig warning if there is no or a bad 0x19
with signatures from a subkey.
present if --expert is set.
* plaintext.c (handle_plaintext): Make bytecount static so it works with
multiple literal packets inside a message.
* encode.c, helptext.c (keygen.algo, keygen.algo.elg_se), keygen.c
(ask_algo), sig-check.c (do_check_messages), skclist.c (build_sk_list):
Rename "ElGamal" to "Elgamal" as that is the proper spelling nowadays.
Suggested by Jon Callas.
* keyring.c (keyring_rebuild_cache): Clear sig cache for any signatures
that we can no longer process (say, if the user removed support for a
necessary pubkey or digest algorithm).
* sig-check.c (check_revocation_keys): Comments.
* getkey.c (merge_selfsigs_main): Don't bother to check designated revoker
sigs if the key is already revoked.
* packet.h, getkey.c (merge_selfsigs_main): New "maybe_revoked" flag on
PKs. It is set when there is a revocation signature from a valid
revocation key, but the revocation key is not present to verify the
signature.
* pkclist.c (check_signatures_trust): Use it here to give a warning when
showing key trust.
* compress-bz2.c: Include stdio.h. Solaris 9 has a very old bzip2 library
and we can at least guarantee that it won't fail because of the lack of
stdio.h.
* tdbio.c: Fixed format string bugs related to the use of DB_NAME.
Reported by Florian Weimer.
here... (signature_check2): ... to here. (check_key_signature2): ... and
here. This is a minor optimization to avoid fetching a key (which can be
expensive, especially if it is not self-signed, and there are many key
signatures on it which need to be checked for ultimate trust) if the
signature would have failed anyway because of algorithm or hash problems.
encode_md_value), sig-check.c (do_check), sign.c (do_sign): Remove
--emulate-md-encode-bug as it only applied to Elgamal signatures, which
are going away.
* options.skel: Note that keyserver.pgp.com isn't synchronized, and
explain the roundrobin a bit better.
* sig-check.c (check_key_signature2), import.c (import_one,
import_revoke_cert, chk_self_sigs, delete_inv_parts, collapse_uids,
merge_blocks): Make much quieter during import of slightly munged, but
recoverable, keys. Use log_error for unrecoverable import failures.
* keyring.c (keyring_rebuild_cache): Comment.
* sign.c (mk_notation_and_policy): Making a v3 signature with notations or
policy urls is an error, not an info (i.e. increment the errorcount).
Don't print the notation or policy url to stdout since it can be mixed
into the output stream when piping and munge the stream.
Provide a signing-key-is-revoked flag. Change all callers.
* status.h, status.c (get_status_string): New REVKEYSIG status tag for a
good signature from a revoked key.
* mainproc.c (do_check_sig, check_sig_and_print): Use it here.
* import.c (import_revoke_cert, merge_blocks, merge_sigs): Compare actual
signatures on import rather than using keyid or class matching. This does
not change actual behavior with a key, but does mean that all sigs are
imported whether they will be used or not.
* parse-packet.c (parse_signature): Don't give "signature packet without
xxxx" warnings for experimental pk algorithms. An experimental algorithm
may not have a notion of (for example) a keyid (i.e. PGP's x.509 stuff).
(ask_algo): Provide key flags for DSA, Elgamal_e, and Elgamal subkeys.
(generate_keypair): Provide key flags for the default DSA/Elgamal keys.
* sig-check.c (signature_check, signature_check2, check_key_signature,
check_key_signature2): Allow passing NULLs for unused parameters in the x2
form of each function to avoid the need for dummy variables. getkey.c,
mainproc.c: Change all callers.
* trustdb.h, trustdb.c (read_trust_options): New. Returns items from the
trustdb version record.
* keylist.c (public_key_list): Use it here for the new "tru" record.
* gpgv.c (read_trust_options): Stub.
do_check): If ret_pk is set, fill in the pk used to verify the signature.
Change all callers in getkey.c, mainproc.c, and sig-check.c.
* keylist.c (list_keyblock_colon): Use the ret_pk from above to put the
fingerprint of the signing key in "sig" records during a --with-colons
--check-sigs. This requires --no-sig-cache as well since we don't cache
fingerprints.
--with-colons mode. Actually translate "Keyring" string.
* mainproc.c (proc_tree): We can't currently handle multiple signatures of
different classes or digests (we'd pretty much have to run a different
hash context for each), but if they are all the same, make an exception.
This is Debian bug #194292.
* sig-check.c (check_key_signature2): Make string translatable.
* packet.h, getkey.c (fixup_uidnode): Mark real primary uids differently
than assumed primaries.
* keyedit.c (no_primary_warning): Use the differently marked primaries
here in a new function to warn when an --edit-key command might rearrange
the self-sig dates enough to change which uid is primary. (menu_expire,
menu_set_preferences): Use no_primary_warning() here.
* Makefile.am: Use @DLLIBS@ for -ldl.
warning a verbose item and translatable. (From wk on stable branch)
* sig-check.c (check_key_signature2): Made "no subkey for subkey binding
packet" a verbose item instead of a !quiet one. There are too many
garbled keys out in the wild. (From wk on stable branch)
* filter.h: Remove const from WHAT. (From wk on stable branch)
* progress.c (handle_progress): Store a copy of NAME. (progress_filter):
Release WHAT, make sure not to print a NULL WHAT. (From wk on stable
branch)
* openfile.c (open_sigfile): Adjust free for new progress semantics. (From
wk on stable branch)
* plaintext.c (ask_for_detached_datafile): Don't dealloc pfx->WHAT. (From
wk on stable branch)
* seckey-cert.c (do_check): Issue the RSA_OR_IDEA status when the cipher
algo is IDEA to make it easier to track down the problem. (From twoaday on
stable branch)
getkey.c, keydb.c, openfile.c, plaintext.c, status.c, gpgv.c, keygen.c,
options.h, sig-check.c, tdbio.h, encode.c, mainproc.c, parse-packet.c,
signal.c, textfilter.c: Edit all preprocessor instructions to remove
whitespace before the '#'. This is not required by C89, but there are some
compilers out there that don't like it.
matter where the user puts the = sign.
* options.skel: Comment out the first three lines in case someone manually
copies the skel file to their homedir.
* sign.c (clearsign_file): Only use pgp2mode with v3 keys and MD5. This
matches what we do when decoding such messages and prevents creating a
message (v3+RIPEMD/160) that we can't verify.
* sig-check.c (signature_check2): Use G10ERR_GENERAL as the error for
signature digest conflict. BAD_SIGN implies that a signature was checked
and we may try and print out a user ID for a key that doesn't exist.
(check_key_signature2): Pass the ultimately trusted pk directly to
check_key_signature2 to avoid going through the key selection mechanism.
This prevents a deadly embrace when two keys without selfsigs each sign
the other.
function as they may not have all their fields filled in.
* sig-check.c (signature_check2): Use new is_primary flag to check rather
than comparing main_keyid with keyid as this still works in the case of a
not fully filled in pk.
* keyedit.c (menu_addrevoker): The direct key signature for revocation
keys must be at least v4 to carry the revocation key subpacket. Add a PGP
2.x warning for revocation keys.
* g10.c (check_permissions): Rearrange strings to make translating easier
(don't incorporate string parts).
* keyedit.c (sign_uids): Make strings translatable.
* sig-check.c (check_key_signature2): Make string translatable.
2002-09-13 David Shaw <dshaw@jabberwocky.com>
* getkey.c (check_revocation_keys): Move....
* main.h, sig-check.c (check_revocation_keys): to here. Also
return the signature_check error code rather than 0/1 and cache
the sig result.
* sig-check.c (check_key_signature2): Divert to
check_revocation_keys if a revocation sig is made by someone other
than the pk owner.
* getkey.c (merge_selfsigs_main): Tidy.
2002-09-13 Werner Koch <wk@gnupg.org>
* g10.c (main) [__MINGW32__]: Activate oLoadExtension.
even for cached sigs. This also serves to protect against missing a sig
expiring while cached.
* getkey.c (merge_selfsigs_main): Don't check UID self-sigs twice.
for the hash that the sig is expecting. This can happen if a onepass sig
header does not match the actual sig, and also if the clearsign "Hash:"
header is missing or does not match the actual sig.
