* cipher/random.c (mix_pool): Store the first hash at the end of the
pool.
--
This fixes a long standing bug (since 1998) in Libgcrypt and GnuPG.
An attacker who obtains 580 bytes of the random number from the
standard RNG can trivially predict the next 20 bytes of output.
This bug does not affect the default generation of
keys because running gpg for key creation creates at most 2 keys from
the pool: For a single 4096 bit RSA key 512 byte of random are
required and thus for the second key (encryption subkey), 20 bytes
could be predicted from the the first key. However, the security of
an OpenPGP key depends on the primary key (which was generated first)
and thus the 20 predictable bytes should not be a problem. For the
default key length of 2048 bit nothing will be predictable.
For the former default of DSA+Elgamal key it is complicate to give an
answer: For 2048 bit keys a pool of 30 non-secret candidate primes of
about 300 bits each are first created. This reads at least 1140 bytes
from the pool and thus parts could be predicted. At some point a 256
bit secret is read from the pool; which in the worst case might be
partly predictable.
The bug was found and reported by Felix Dörre and Vladimir Klebanov,
Karlsruhe Institute of Technology. A paper describing the problem in
detail will shortly be published.
CVE-id: CVE-2016-6313
Signed-off-by: Werner Koch <wk@gnupg.org>
* g10/gpg.c (main): initialize opt.emit_version to 0
* doc/gpg.texi: document different default for --emit-version
--
The version of GnuPG in use is not particularly helpful. It is not
cryptographically verifiable, and it doesn't distinguish between
significant version differences like 2.0.x and 2.1.x.
Additionally, it leaks metadata that can be used to distinguish users
from one another, and can potentially be used to target specific
attacks if there are known behaviors that differ between major
versions.
It's probably better to take the more parsimonious approach to
metadata production by default.
(backport of master commit c9387e41db7520d176edd3d6613b85875bdeb32c)
Signed-off-by: Daniel Kahn Gillmor <dkg@fifthhorseman.net>
* README, cipher/cipher.c, cipher/pubkey.c, doc/gpg.texi: replace
"allow to" with clearer text
In standard English, the normal construction is "${XXX} allows ${YYY}
to" -- that is, the subject (${XXX}) of the sentence is allowing the
object (${YYY}) to do something. When the object is missing, the
phrasing sounds awkward, even if the object is implied by context.
There's almost always a better construction that isn't as awkward.
These changes should make the language a bit clearer.
Signed-off-by: Daniel Kahn Gillmor <dkg@fifthhorseman.net>
* checks/armor.test, cipher/des.c, g10/ccid-driver.c, g10/pkclist.c,
util/regcomp.c, util/regex_internal.c: correct the spelling of
"occured" to "occurred"
Signed-off-by: Daniel Kahn Gillmor <dkg@fifthhorseman.net>
* g10/sig-check.c (signature_check2): Not only subkey, but also primary
key should have flags.valid=1.
--
(backport of master
commit 6f284e6ed63f514b15fe610f490ffcefc87a2164)
Signed-off-by: NIIBE Yutaka <gniibe@fsij.org>
* g10/gpgv.c (main): Set opt.no_sig _cache, so that it doesn't depend on
cached status. Similarly, set opt.flags.require_cross_cert for backsig
validation for subkey signature.
--
(backport of master
commit e32c575e0f3704e7563048eea6d26844bdfc494b)
It is common that an organization distributes binary keyrings with
signature cache (Tag 12, Trust Packet) and people use gpgv to validate
signature with such keyrings. In such a use case, it is possible that
the key validation itself is skipped.
For the purpose of gpgv validation of signatures, we should not depend
on signature cache in keyrings (if any), but we should validate the key
by its self signature for primary key, and back signature for subkey.
Signed-off-by: NIIBE Yutaka <gniibe@fsij.org>
* g10/gpg.c (main): Call set_packet_list_mode after assignment of
opt.list_packets.
* g10/mainproc.c (do_proc_packets): Don't stop processing with
--list-packets as the comment says.
* g10/options.h (list_packets): Fix the comment.
* g10/parse-packet.c: Fix the condition for opt.list_packets.
--
(backport from 2.0 commit 4f336ed780cc2783395f3ff2b12b3ebb8e097f7b
which is backport of master
commit 52f65281f9743c42a48bf5a3354c9ab0ecdb681a)
Debian-bug-id: 828109
Signed-off-by: NIIBE Yutaka <gniibe@fsij.org>
* g10/tdbio.c (create_version_record): Call create_hashtable to always
make hashtable, together with the version record.
(get_trusthashrec): Remove call to create_hashtable.
--
GnuPG-bug-id: 1675
Thanks to Scott Moser to reproducible script and patience.
Signed-off-by: NIIBE Yutaka <gniibe@fsij.org>
(backport from master
commit 35a3ce2acf78a95fecbccfd8db0560cca24232df)
--
GnuPG-bug-id: 1394
Note that --try-secret-key was already removed with commit
2889a70c102271a1b6ff529bafb6748c4e773014
Signed-off-by: Werner Koch <wk@gnupg.org>
* g10/tdbio.c (tdbio_set_dbname): Return earlier if !CREATE. Check
the directory and create it if none before calling take_write_lock.
--
Thanks to Marc Deslauriers for the bug report and his patch.
GnuPG-bug-id: 2246
Signed-off-by: NIIBE Yutaka <gniibe@fsij.org>
(backport from master
commit 2f3e42047d17313eeb38d354048f343158402a8d)
* cipher/rndunix.c [HAVE_STDINT_H]: Include stdint.h.
(start_gatherer): Detect misbehaving sysconf.
--
See
GnuPG-bug-id: 1778
for the reason of this patch. There is no concrete bug report but this
chnage should not harm.
Signed-off-by: Werner Koch <wk@gnupg.org>
* 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:
76afaed65e3b0ddfa4923cb577ada43217dd4b18
b98939812abf6c643c752ce7c325f98039a1a9e2
91015d021b3dcbe21ad0e580a4f34c523abf9e72
)
* g10/main.h (DEFAULT_CIPHER_ALGO): Change to AES or CAST5 or 3DES
depending on configure options.
* g10/gpg.c (main): Set opt.s2k_cipher_algo to DEFAULT_CIPHER_ALGO.
--
(cherry picked from commit 57df1121c18b004dd763b35eabf7b51fc9e8ec38)
Signed-off-by: Werner Koch <wk@gnupg.org>
* g10/passphrase.c (stdenvnames): Add DBUS_SESSION_BUS_ADDRESS.
--
pinentry-gnome3 talks to the gcr prompter via dbus. Without this
environment variable, it can't find the correct session to talk to.
* g10/trustdb.c (validate_keys): Call dump_key_array only in debug
mode.
--
I guess that is a left-over from an early attempt to output
information on the trustdb for use by other tools. Maybe related to
the former --list-trust-path command. Sending it to stdout is
probably useful so we do this now only in debug mode.
Signed-off-by: Werner Koch <wk@gnupg.org>
Backported to STABLE-BRANCH-1-4 from
b03a2647299a6c8764a2574590cbaccdff9e497d by dkg
* cipher/rsa.c (rsa_sign): Verify after sign.
* g10/gpg.c (opts): Make --no-sig-create-check a NOP.
* g10/options.h (opt): Remove field "no_sig_create_check".
* g10/sign.c (do_sign): Do check only for DSA.
Signed-off-by: Werner Koch <wk@gnupg.org>
g10/mainproc.c (proc_pubkey_enc): Only allow relevant algorithms for
encryption.
--
(backported from 2.1 commit c771963140cad7c1c25349bcde27e427effc0058)
* g10/getkey.c (get_seckey): Return G10ERR_NO_PUBKEY when it's not
exact match.
--
In the situation of corrupted .gnupg/ where only private subkey is
available but no corresponding public key of the subkey, the code
returned public primary key which caused mysterious error (for a
user). This fix detects an error earlier.
GnuPG-bug-id: 1422
Debian-Bug-Id: #638619
* g10/free-packet.c (cmp_public_keys, cmp_secret_keys): Compare opaque
data at the first entry of the array when it's unknown algo.
* mpi/mpi-cmp.c (mpi_cmp): Backport libgcrypt 1.5.0's semantics.
--
(backported from 2.0 commit 43429c7869152f301157e4b24790b3801dce0f0a)
GnuPG-bug-id: 1962
* g10/parse-packet.c (parse_key): Check PKTLEN before calling mpi_read
et al.
--
Due to the missing length checks PKTLEN may turn negative. Because
PKTLEN is an unsigned int the malloc in read_rest would try to malloc
a too large number and terminate the process with "error reading rest
of packet: Cannot allocate memory".
Reported-by: Hanno Böck.
Signed-off-by: Werner Koch <wk@gnupg.org>
(backported from 2.0 commit 0aac920f23fd07e152fdb7385299c92bb9a4ade3)
