Commit Graph

4532 Commits

Author SHA1 Message Date
Simon Butcher
6ee1af5aab Update mbed-drivers dependency to v1.0.0 2016-03-09 19:32:10 +00:00
Simon Butcher
f59e66ba24 Remove redundant test certificates and clarify ChangeLog 2016-03-09 19:32:10 +00:00
Janos Follath
b437b4b125 X509: Fix bug triggered by future CA among trusted
Fix an issue that caused valid certificates being rejected whenever an
expired or not yet valid version of the trusted certificate was before the
valid version in the trusted certificate list.
2016-03-09 19:32:10 +00:00
Janos Follath
df4bca2029 X509: Future CA among trusted: add more tests 2016-03-09 19:32:10 +00:00
Janos Follath
12c868c5d6 X509: Future CA among trusted: add unit tests 2016-03-09 19:32:10 +00:00
Simon Butcher
64d60da4f6 Fix typos and add copyright statement to generate_code.pl 2016-03-09 19:32:10 +00:00
SimonB
0269dad5e5 Refactored test suite template code
Restructed test suite helper and main code to support tests suite helper
functions, changed C++ comments to C-style, and made the generated
source code more navigable.
2016-03-09 19:32:10 +00:00
SimonB
152ea18037 Added support for per test suite helper functions
Added to generate_code.pl:
    - support for per test suite helper functions
    - description of the structure of the files the script uses to construct
      the test suite file
    - delimiters through the source code to make the machine generated code
      easier to understand
2016-03-09 19:32:10 +00:00
SimonB
3ddf35526a Clarified purpose and usage of generate_code.pl
Added comments to explain purpose and usage of generate_code.pl
2016-03-09 19:32:09 +00:00
Simon Butcher
ef50c0da6e Added script to split the test case data files
Script generate-afl-tests.sh will split the test suite data files into
individual test case files, suitable for fuzzing.
2016-03-09 19:32:09 +00:00
Simon Butcher
aad787f1c7 Parameterised the test suite applications
All test suites can now take an arbitrary test file.
2016-03-09 19:32:09 +00:00
Simon Butcher
e846b5128f Use the SSL IO and time callback typedefs consistently
The callback typedefs defined for mbedtls_ssl_set_bio() and
mbedtls_ssl_set_timer_cb() were not used consistently where the callbacks were
referenced in structures or in code.
2016-03-09 19:32:09 +00:00
Simon Butcher
c0957bdc13 Fix some minor typos in comments
Fix spelling mistakes and typos.
2016-03-09 19:32:09 +00:00
Manuel Pégourié-Gonnard
0c6aad90f2 x509: remove obsolete TODO comment
- basicContraints checks are done during verification
- there is no need to set extensions that are not present to default values,
  as the code using the extension will check if it was present using
ext_types. (And default values would not make sense anyway.)
2016-03-09 19:32:09 +00:00
Manuel Pégourié-Gonnard
986bbf24ce x509:
-
2016-03-09 19:32:09 +00:00
Manuel Pégourié-Gonnard
d1b7f2b8cf ssl: ignore CertificateRequest's content for real
- document why we made that choice
- remove the two TODOs about checking hash and CA
- remove the code that parsed certificate_type: it did nothing except store
  the selected type in handshake->cert_type, but that field was never accessed
afterwards. Since handshake_params is now an internal type, we can remove that
field without breaking the ABI.
2016-03-09 19:32:09 +00:00
Manuel Pégourié-Gonnard
56e9ae2bf2 Remove unnecessary TODO comment
We don't implement anonymous key exchanges, and we don't intend to, so it can
never happen that an unauthenticated server requests a certificate from us.
2016-03-09 19:32:09 +00:00
Manuel Pégourié-Gonnard
eeef947040 Clarify documentation about missing CRLs
Also tune up some working while at it.
2016-03-09 19:32:08 +00:00
Manuel Pégourié-Gonnard
214a84889c Update note about hardcoded verify_data_length 2016-03-09 19:32:08 +00:00
Manuel Pégourié-Gonnard
967994a05e Remove unused code.
After the record contents are decompressed, in_len is no longer
accessed directly, only in_msglen is accessed. in_len is only read by
ssl_parse_record_header() which happens before ssl_prepare_record_contents().

This is also made clear by the fact that in_len is not touched after
decrypting anyway, so if it was accessed after that it would be wrong unless
decryption is used - as this is not the case, it show in_len is not accessed.
2016-03-09 19:32:08 +00:00
Manuel Pégourié-Gonnard
9d6241269a Add note about not implementing PSK id_hint 2016-03-09 19:32:08 +00:00
Manuel Pégourié-Gonnard
90ab4a45b5 Fix Unix detection in mini_client
fixes #398
2016-03-09 19:32:08 +00:00
Simon Butcher
157cb656a9 Clarified mbedtls_ssl_conf_alpn_protocols() doc
Clarified the lifetime of the protos parameter passed in the
function mbedtls_ssl_conf_alpn_protocols().
2016-03-09 19:32:08 +00:00
Simon Butcher
8b4a1bdbb0 Update the ChangeLog 2016-03-07 23:30:50 +00:00
Simon Butcher
342671f982 Update interop tests to default configuration
Removed SSLv3 from the default tests in compat.sh, and adapted the test
cases in all.sh to include an additional SSLv3 regression test suite.
2016-03-07 23:22:10 +00:00
Simon Butcher
14ecd0439f Fix yotta builds for change in default configs
The change to defaults configurations in a720ced broke the yotta build.
This fix addresses that.
2016-03-07 23:07:04 +00:00
Simon Butcher
29b2150016 Fix the 'all tests' script for baremetal builds
Fixes the test script test/scripts/all.sh which was failing at the
baremetal ARM builds due to the entropy platform check introduced in
7ff4b77.
2016-03-07 23:06:27 +00:00
Janos Follath
a720ced403 Update default configuration
Change the default settings for SSL and modify the tests accordingly.
2016-03-07 15:57:05 +00:00
Simon Butcher
b3c6978c7e Add copright, and better documentation to curves.pl
The purpose and use of the test script, curves.pl was not obvious without
reading the source code, plus the file was missing a copyright statement.
2016-03-04 23:26:57 +00:00
Simon Butcher
5d23716e20 Add missing dependencies to X509 Parse test suite for P-384 curve
The test script curves.pl was failing on testing dependencies for the P-384
curve on the new test cases introduced by ede75f0 and 884b4fc.
2016-03-04 22:21:52 +00:00
Simon Butcher
25f2c4c028 Update mbed-drivers dependency to v1.0.0 2016-03-02 17:00:16 +00:00
Simon Butcher
a418ff8eb5 Remove redundant test certificates and clarify ChangeLog 2016-03-01 20:26:16 +00:00
Janos Follath
c72d642595 X509: Fix bug triggered by future CA among trusted
Fix an issue that caused valid certificates being rejected whenever an
expired or not yet valid version of the trusted certificate was before the
valid version in the trusted certificate list.
2016-03-01 19:13:48 +00:00
Janos Follath
884b4fc2e9 X509: Future CA among trusted: add more tests 2016-03-01 19:13:48 +00:00
Janos Follath
ede75f06c5 X509: Future CA among trusted: add unit tests 2016-03-01 19:13:48 +00:00
Simon Butcher
f18e02c22e Fix typos and add copyright statement to generate_code.pl 2016-03-01 18:37:49 +00:00
SimonB
beff85aaee Refactored test suite template code
Restructed test suite helper and main code to support tests suite helper
functions, changed C++ comments to C-style, and made the generated
source code more navigable.
2016-03-01 18:37:49 +00:00
SimonB
6fb9db3afd Added support for per test suite helper functions
Added to generate_code.pl:
    - support for per test suite helper functions
    - description of the structure of the files the script uses to construct
      the test suite file
    - delimiters through the source code to make the machine generated code
      easier to understand
2016-03-01 18:37:49 +00:00
SimonB
718548d5c9 Clarified purpose and usage of generate_code.pl
Added comments to explain purpose and usage of generate_code.pl
2016-03-01 18:37:49 +00:00
Simon Butcher
2bed20d670 Added script to split the test case data files
Script generate-afl-tests.sh will split the test suite data files into
individual test case files, suitable for fuzzing.
2016-03-01 18:37:49 +00:00
Simon Butcher
6c545a87c2 Parameterised the test suite applications
All test suites can now take an arbitrary test file.
2016-03-01 18:37:49 +00:00
Simon Butcher
1b6044ded2 Use the SSL IO and time callback typedefs consistently
The callback typedefs defined for mbedtls_ssl_set_bio() and
mbedtls_ssl_set_timer_cb() were not used consistently where the callbacks were
referenced in structures or in code.
2016-03-01 17:37:12 +00:00
Simon Butcher
a766576a74 Fix some minor typos in comments
Fix spelling mistakes and typos.
2016-03-01 17:37:12 +00:00
Manuel Pégourié-Gonnard
347700ebe2 x509: remove obsolete TODO comment
- basicContraints checks are done during verification
- there is no need to set extensions that are not present to default values,
  as the code using the extension will check if it was present using
ext_types. (And default values would not make sense anyway.)
2016-03-01 17:37:11 +00:00
Manuel Pégourié-Gonnard
a0e924fa7b x509:
-
2016-03-01 17:37:11 +00:00
Manuel Pégourié-Gonnard
04d39d2825 ssl: ignore CertificateRequest's content for real
- document why we made that choice
- remove the two TODOs about checking hash and CA
- remove the code that parsed certificate_type: it did nothing except store
  the selected type in handshake->cert_type, but that field was never accessed
afterwards. Since handshake_params is now an internal type, we can remove that
field without breaking the ABI.
2016-03-01 17:37:11 +00:00
Manuel Pégourié-Gonnard
b222cd92c1 Remove unnecessary TODO comment
We don't implement anonymous key exchanges, and we don't intend to, so it can
never happen that an unauthenticated server requests a certificate from us.
2016-03-01 17:37:11 +00:00
Manuel Pégourié-Gonnard
0fa5b055c9 Clarify documentation about missing CRLs
Also tune up some working while at it.
2016-03-01 17:37:11 +00:00
Manuel Pégourié-Gonnard
982b9adc96 Update note about hardcoded verify_data_length 2016-03-01 17:37:11 +00:00
Manuel Pégourié-Gonnard
fc0e286c0e Remove unused code.
After the record contents are decompressed, in_len is no longer
accessed directly, only in_msglen is accessed. in_len is only read by
ssl_parse_record_header() which happens before ssl_prepare_record_contents().

This is also made clear by the fact that in_len is not touched after
decrypting anyway, so if it was accessed after that it would be wrong unless
decryption is used - as this is not the case, it show in_len is not accessed.
2016-03-01 17:37:11 +00:00