Commit Graph

4264 Commits

Author SHA1 Message Date
Simon Butcher
59d2218f63 Clarification in ChangeLog 2015-12-23 18:53:21 +00:00
Simon Butcher
aa4114910a Merge 'iotssl-558-2.1-md5-tls-sigs-restricted' 2015-12-23 18:52:18 +00:00
Simon Butcher
35ea92dbc6 Merge 'iotssl-566-2.1-double-free-restricted'
Merge remote-tracking branch
'restricted/iotssl-566-2.1-double-free-restricted' into mbedtls-2.1
2015-12-23 16:49:46 +00:00
Simon Butcher
2bc3897a53 Typo in ChangeLog 2015-12-22 19:38:55 +00:00
Simon Butcher
e82ac57ef6 Merge remote-tracking branch 'origin/misc-2.1' into mbedtls-2.1 2015-12-22 19:36:17 +00:00
Simon Butcher
e103aa8a53 Added description of change to the Changelog
Also clarified some comments following review.
2015-12-16 01:51:01 +00:00
Manuel Pégourié-Gonnard
9055c1a011 Fix wrong length limit in GCM
See for example page 8 of
http://csrc.nist.gov/publications/nistpubs/800-38D/SP-800-38D.pdf

The previous constant probably came from a typo as it was 2^26 - 2^5 instead
of 2^36 - 2^5. Clearly the intention was to allow for a constant bigger than
2^32 as the ull suffix and cast to uint64_t show.

fixes #362
2015-12-10 15:08:37 +01:00
Manuel Pégourié-Gonnard
7f890c3e96 Avoid seemingly-possible overflow
By looking just at that test, it looks like 2 + dn_size could overflow. In
fact that can't happen as that would mean we've read a CA cert of size is too
big to be represented by a size_t.

However, it's best for code to be more obviously free of overflow without
having to reason about the bigger picture.
2015-12-10 15:07:46 +01:00
Manuel Pégourié-Gonnard
7da96958a6 Make documentation more explicit on TLS errors
fixes #358
2015-12-10 15:07:46 +01:00
Manuel Pégourié-Gonnard
3e60d2a458 Fix potential double free in cert writing code
In case an entry with the given OID already exists in the list passed to
mbedtls_asn1_store_named_data() and there is not enough memory to allocate
room for the new value, the existing entry will be freed but the preceding
entry in the list will sill hold a pointer to it. (And the following entries
in the list are no longer reachable.) This results in memory leak or a double
free.

The issue is we want to leave the list in a consistent state on allocation
failure. (We could add a warning that the list is left in inconsistent state
when the function returns NULL, but behaviour changes that require more care
from the user are undesirable, especially in a stable branch.)

The chosen solution is a bit inefficient in that there is a time where both
blocks are allocated, but at least it's safe and this should trump efficiency
here: this code is only used for generating certificates, which is unlikely to
be done on very constrained devices, or to be in the critical loop of
anything. Also, the sizes involved should be fairly small anyway.

fixes #367
2015-12-10 11:24:35 +01:00
Manuel Pégourié-Gonnard
2bbfee3cbc Add credits to ChangeLog 2015-12-08 16:14:30 +01:00
Manuel Pégourié-Gonnard
b6043f93b6 Fix levels of some debug messages
Messages about returning an error code that will be propagated all the way up
to the user should always be level 1.
2015-12-08 09:57:25 +01:00
Manuel Pégourié-Gonnard
b39528e2e8 Disable MD5 in handshake signatures by default 2015-12-04 15:13:36 +01:00
Manuel Pégourié-Gonnard
013198f30f DTLS: avoid dropping too many records
When the peer retransmits a flight with many record in the same datagram, and
we already saw one of the records in that datagram, we used to drop the whole
datagram, resulting in interoperability failure (spurious handshake timeouts,
due to ignoring record retransmitted by the peer) with some implementations
(issues with Chrome were reported).

So in those cases, we want to only drop the current record, and look at the
following records (if any) in the same datagram. OTOH, this is not something
we always want to do, as sometime the header of the current record is not
reliable enough.

This commit introduces a new return code for ssl_parse_header() that allows to
distinguish if we should drop only the current record or the whole datagram,
and uses it in mbedtls_ssl_read_record()

fixes #345
2015-12-03 19:22:55 +01:00
Manuel Pégourié-Gonnard
5a8396ed55 Fix two more compiler warnings
- declaration after statement
- always true comparison due to limited range of operand
2015-12-03 19:09:23 +01:00
Manuel Pégourié-Gonnard
3eab29adc8 Fix potential integer overflow in prev. commit
Found by Clang's -Wshift-count-overflow
2015-12-03 19:09:21 +01:00
Simon Butcher
ea303e3ece Added integer divide by as separate function
Added 64bit integer divided by 32bit integer, with remainder
2015-11-26 23:43:34 +00:00
Manuel Pégourié-Gonnard
d847f1f46a Fix ChangeLog 2015-11-19 12:17:17 +01:00
Manuel Pégourié-Gonnard
b030c33e57 Fix bug checking pathlen on first intermediate
Remove check on the pathLenConstraint value when looking for a parent to the
EE cert, as the constraint is on the number of intermediate certs below the
parent, and that number is always 0 at that point, so the constraint is always
satisfied.

The check was actually off-by-one, which caused valid chains to be rejected
under the following conditions:
- the parent certificate is not a trusted root, and
- it has pathLenConstraint == 0 (max_pathlen == 1 in our representation)

fixes #280
2015-11-19 11:26:52 +01:00
Manuel Pégourié-Gonnard
3cb2074a82 Add test case for root with max_pathlen=0
This was already working but not tested so far

(Test case from previous commit still failing.)

Test certificates generated with:

programs/pkey/gen_key type=ec ec_curve=secp256r1 filename=cert91.key
programs/pkey/gen_key type=ec ec_curve=secp256r1 filename=cert92.key

programs/x509/cert_write serial=91 output_file=cert91.crt is_ca=1 \
    issuer_key=cert91.key issuer_name="CN=Root 9,O=mbed TLS,C=UK" \
    selfsign=1 max_pathlen=0
programs/x509/cert_write serial=92 output_file=cert92.crt \
    issuer_key=cert91.key issuer_name="CN=Root 9,O=mbed TLS,C=UK" \
    subject_key=cert92.key subject_name="CN=EE 92,O=mbed TLS,C=UK"

mv cert9?.crt tests/data_files/dir4
rm cert9?.key
2015-11-19 11:25:30 +01:00
Manuel Pégourié-Gonnard
922cd9ba36 Add test case for first intermediate max_pathlen=0
!!! This test case is currently failing !!!
(See fix in next-next commit.)

Test certificates generated with the following script:

programs/pkey/gen_key type=ec ec_curve=secp256r1 filename=cert81.key
programs/pkey/gen_key type=ec ec_curve=secp256r1 filename=cert82.key
programs/pkey/gen_key type=ec ec_curve=secp256r1 filename=cert83.key

programs/x509/cert_write serial=81 output_file=cert81.crt is_ca=1 \
    issuer_key=cert81.key issuer_name="CN=Root 8,O=mbed TLS,C=UK" \
    selfsign=1
programs/x509/cert_write serial=82 output_file=cert82.crt is_ca=1 \
    issuer_key=cert81.key issuer_name="CN=Root 8,O=mbed TLS,C=UK" \
    subject_key=cert82.key subject_name="CN=Int 82,O=mbed TLS,C=UK" \
    max_pathlen=0
programs/x509/cert_write serial=83 output_file=cert83.crt \
    issuer_key=cert82.key issuer_name="CN=Int 82,O=mbed TLS,C=UK" \
    subject_key=cert83.key subject_name="CN=EE 83,O=mbed TLS,C=UK"

mv cert8?.crt tests/data_files/dir4
rm cert8?.key
2015-11-19 11:25:27 +01:00
Simon Butcher
ef43d41f67 Changed version number to 2.1.3
Changed for library
2015-11-04 22:08:33 +00:00
Simon Butcher
5b289208cb Remove debugging code left in test case
Removed debug code from tests/suites/test_suite_x509parse.function
2015-11-04 21:50:54 +00:00
Simon Butcher
73156357ed Disable Yotta tests from 'all tests' script
Yotta tests not supported in 2.1 branch
2015-11-04 00:36:30 +00:00
Simon Butcher
b2d2fec5a4 Corrected typo in ChangeLog 2015-11-03 23:12:36 +00:00
Manuel Pégourié-Gonnard
c28240596a Fix other int casts in bounds checking
Not a security issue as here we know the buffer is large enough (unless
something else if badly wrong in the code), and the value cast to int is less
than 2^16 (again, unless issues elsewhere).

Still changing to a more correct check as a matter of principle
2015-11-02 10:43:03 +09:00
Manuel Pégourié-Gonnard
5784dd5ac8 Fix other occurrences of same bounds check issue
Security impact is the same: not triggerrable remotely except in very specific
use cases
2015-11-02 10:43:03 +09:00
Manuel Pégourié-Gonnard
0d66bb959f Fix potential buffer overflow in asn1write 2015-11-02 10:42:44 +09:00
Manuel Pégourié-Gonnard
9dc66f4b2f Fix potential heap corruption on Windows
If len is large enough, when cast to an int it will be negative and then the
test if( len > MAX_PATH - 3 ) will not behave as expected.
2015-11-02 10:41:13 +09:00
Manuel Pégourié-Gonnard
ffb8180733 Fix potential double-free in ssl_conf_psk() 2015-11-02 10:40:14 +09:00
Manuel Pégourié-Gonnard
e34dcd7ec5 Use own implementation of strsep()
Not available on windows, and strtok() is not a good option
2015-11-02 06:48:40 +09:00
Manuel Pégourié-Gonnard
1cf8851a77 Add ChangeLog entry for ASN.1 DER boolean fix 2015-11-02 06:00:38 +09:00
Jonathan Leroy
e03fa7c16a Test certificate "Server1 SHA1, key_usage" reissued. 2015-11-02 05:58:58 +09:00
Jonathan Leroy
00c6b3c35a Fix boolean values according to DER specs
In BER encoding, any boolean with a non-zero value is considered as
TRUE. However, DER encoding require a value of 255 (0xFF) for TRUE.

This commit makes `mbedtls_asn1_write_bool` function uses `255` instead
of `1` for BOOLEAN values.

With this fix, boolean values are now reconized by OS X keychain (tested
on OS X 10.11).

Fixes #318.
2015-11-02 05:58:43 +09:00
Jonathan Leroy
3dd85ddfdf cert_write : fix "Destination buffer is too small" error
This commit fixes the `Destination buffer is too small` error returned
by `mbedtls_cert_write` command when the values of `subject_name` or
`issuer_name` parameters exceed 128 characters.

I have increased the size of these varaibles from 128 to 256 characters,
but I don't know if it's the best way to solve this issue...

Fixes #315.
2015-11-02 05:58:30 +09:00
Manuel Pégourié-Gonnard
621f83e5c5 Fix typo in an OID name
fixes #314
2015-11-02 05:58:10 +09:00
Manuel Pégourié-Gonnard
7a40dc686f Disable reportedly broken assembly of Sparc(64)
fixes #292
2015-11-02 05:57:49 +09:00
Manuel Pégourié-Gonnard
e55448a50f Add Changelog entries for max_pathlen fixes 2015-11-02 05:56:57 +09:00
Manuel Pégourié-Gonnard
1d9348a06f Fix a style issue 2015-11-02 05:56:08 +09:00
Manuel Pégourié-Gonnard
fd1f9e735e Fix whitespace at EOL issues 2015-11-02 05:55:58 +09:00
Manuel Pégourié-Gonnard
841caf1b74 Use symbolic constants in test data 2015-11-02 05:55:39 +09:00
Janos Follath
860f239eb9 Fixed pathlen contraint enforcement. 2015-11-02 05:55:28 +09:00
Janos Follath
36f1234d96 Additional corner cases for testing pathlen constrains. Just in case. 2015-11-02 05:55:15 +09:00
Janos Follath
c7bea3158a Added test case for pathlen constrains in intermediate certificates 2015-11-02 05:55:02 +09:00
Jonathan Leroy
1f8c20ac9a Fix help message for cert_req/cert_write programs
In cert_req and cert_write programs, "key_certificate_sign" is not an
allowed velue for "key_usage" parameter. The correct value is
"key_cert_sign".

See https://github.com/ARMmbed/mbedtls/blob/development/programs/x509/cert_req.c#L208
and https://github.com/ARMmbed/mbedtls/blob/development/programs/x509/cert_write.c#L323.
2015-10-30 16:56:44 +01:00
Manuel Pégourié-Gonnard
d13585f1b3 Small improvement to test script 2015-10-30 16:56:30 +01:00
Manuel Pégourié-Gonnard
9f44a80ea3 Try to prevent some misuse of RSA functions
fixes #331
2015-10-30 10:57:43 +01:00
Manuel Pégourié-Gonnard
8f115968da Pick up ChangeLog fixes from development 2015-10-28 13:55:28 +01:00
Manuel Pégourié-Gonnard
a7f0a42101 Mention new test script in Readme 2015-10-28 13:42:14 +01:00
Manuel Pégourié-Gonnard
93080dfacf Fix missing check for RSA key length on EE certs
- also adapt tests to use lesser requirement for compatibility with old
  testing material
2015-10-28 13:22:32 +01:00