Commit Graph

2898 Commits

Author SHA1 Message Date
Steven Fackler d3104955dd Minimize test duplication 2019-06-28 19:51:52 -07:00
Steven Fackler 1b3e0c8a15 Add mozilla v5 configurations
Closes #1134
2019-06-28 19:42:29 -07:00
Steven Fackler 0d9f37be5b
Merge pull request #1131 from mbelop/ec_point_dup
Expose EC_POINT_dup as EcPoint::to_owned
2019-06-19 20:57:00 -07:00
Mike Belopuhov aef0517dcf Expose EC_POINT_dup as EcPoint::to_owned 2019-06-19 21:34:48 +02:00
Steven Fackler 39e692fac5
Merge pull request #1130 from mbelop/cofactor
Expose EC_GROUP_get_cofactor as EcGroup::cofactor
2019-06-15 11:55:56 -07:00
Mike Belopuhov e8fc907da3 Expose EC_GROUP_get_cofactor as EcGroup::cofactor 2019-06-14 16:41:47 +02:00
Steven Fackler ccb2fd49ca
Merge pull request #1128 from mbelop/mike/generator
Expose EC_GROUP_get0_generator as EcGroup::generator
2019-06-13 09:01:13 -07:00
Mike Belopuhov 390d71f1e5 Expose EC_GROUP_get0_generator as EcGroup::generator 2019-06-13 03:09:45 +02:00
Steven Fackler 89bd6d6168
Merge pull request #1127 from snapview/cms-pem
Add `to_pem()` and `from_pem()` for `CmsContentInfo`
2019-06-12 08:31:51 -07:00
Daniel Abramov fab6ea4727 Conditionally compile PEM functions for CMS
Apparently libressl does not quite support all CMS functions (well, at
least the bindings for CMS are currently compile-time guarded), so CI
checks inside the systest fail during the verification on libressl.
This is an attempt to fix it.
2019-06-12 16:48:16 +02:00
Daniel Abramov ed966a09ac Extend CMS unit tests (pem/der conversions) 2019-06-12 10:36:44 +02:00
Daniel Abramov 69ee79d435 Fix formatting in cms `to_der` implementation 2019-06-12 10:24:53 +02:00
Daniel Abramov 124c05d058 Add `CmsContentInfo` <-> PEM bindings 2019-06-12 10:23:48 +02:00
Steven Fackler 3629bb6f26
Merge pull request #1124 from Leo1003/eckey_debug
Add Debug impl for EcKey
2019-06-10 08:53:33 -07:00
Leo d2e48e8d1f Add Debug impl for EcKey 2019-06-10 21:57:10 +08:00
Steven Fackler 179758010a
Merge pull request #1122 from russelltg/aes_wrap
Add AES_wrap_key and AES_unwrap_key functionality
2019-06-07 21:06:25 -07:00
Russell Greene 1c3f5b5f57 Address comments 2019-06-07 19:32:36 -06:00
Russell Greene b9341856b1 Add AES_wrap_key and AES_unwrap_key functionality 2019-06-05 21:03:34 -06:00
Steven Fackler ef86438a10 Report the vendored install location
Closes #1117
2019-05-31 19:08:58 -07:00
Steven Fackler d7a768ea91
Merge pull request #1119 from sfackler/openssl-111c
Bump CI versions
2019-05-31 08:48:19 -07:00
Steven Fackler 958c1811b0 Fix constness for 1.1.1c 2019-05-31 07:29:00 -07:00
Steven Fackler 49fc65f366 Bump CI versions 2019-05-30 22:01:54 -07:00
Steven Fackler 0202c20ca3
Merge pull request #1118 from animalsiknow/error-display-reason
Display for Error incorrectly showing func instead of reason.
2019-05-30 13:06:23 -07:00
Simon Génier e3ac3f40bf Display for Error was incorrectly showing func instead of reason. 2019-05-30 15:06:14 -04:00
Steven Fackler 3b064fdb02 Release openssl v0.10.23 2019-05-18 12:13:39 -07:00
Steven Fackler b6d968b378 Release openssl-sys v0.9.47 2019-05-18 12:10:57 -07:00
Steven Fackler efd2c53843
Merge pull request #1116 from sfackler/fix-sess-cbs
Fix handling of session callbacks
2019-05-18 12:07:21 -07:00
Steven Fackler b39a712076 Fix handling of session callbacks
The session context is used for session callbacks rather than the normal
context, which breaks state lookup when the context has been swapped out
(e.g. for SNI). Since there isn't an accessor for the session context,
we just store an extra reference in the SSL's ex data.

Closes #1115
2019-05-18 10:27:40 -07:00
Steven Fackler d861eb16dd
Merge pull request #1113 from Metaswitch/add_client_ca
Add SSL_CTX_add_client_CA on OpenSSL
2019-05-13 11:33:01 -07:00
Andy Caldwell 2e37753790
Const-correctness 2019-05-13 19:11:15 +01:00
Andy Caldwell 41fea135ad
Allow passing by non-owned reference 2019-05-13 18:49:09 +01:00
Andy Caldwell 628c3b338a
Add SSL_CTX_add_client_CA on OpenSSL 2019-05-13 15:08:02 +01:00
Steven Fackler 6686092edf Release openssl v0.10.22 2019-05-08 18:46:43 -07:00
Steven Fackler aabaf97935 Release openssl-sys v0.9.46 2019-05-08 18:44:41 -07:00
Steven Fackler a9b9f818a1
Merge pull request #1097 from vishwin/master
Support LibreSSL 2.9.1
2019-05-08 18:40:18 -07:00
Steven Fackler 801a236413 Ignore SRTP tests on libressl 2.9.1
SRTP is broken in that release!
2019-05-08 18:19:42 -07:00
Steven Fackler 2f5ecb4267 Release openssl-sys v0.9.45 2019-05-03 19:52:18 -07:00
Steven Fackler d4f85a323f
Merge pull request #1107 from sfackler/revert-1100-patch-1
Revert "fix build err for taget `*-pc-windows-gnu`"
2019-05-03 19:38:15 -07:00
Steven Fackler 62b211990f
Revert "fix build err for taget `*-pc-windows-gnu`" 2019-05-03 18:44:53 -07:00
Steven Fackler d2b169dae6 Fix doc link bugs 2019-05-03 07:32:05 -07:00
Steven Fackler 9f3cf9907f
Merge pull request #1103 from alexcrichton/autocfg
Switch from `rustc_version` to `autocfg`
2019-05-01 08:39:49 -07:00
Alex Crichton 06577cbf9c Switch from `rustc_version` to `autocfg`
This switches the `openssl-sys` crate from using `rustc_version` as a
crate to check the version of rustc to using `autocfg`. While
functionally the same this has a few advantages:

* The `autocfg` crate has fewer dependencies and compiles faster
* If the `semver` crate has the `serde` feature activated, turns out
  `openssl-sys` gets compiled quite late in the dependency graph which
  can push back further C compilations. This is due to the slower
  compilation time of `serde` itself.
* The `autocfg` crate I believe is a bit more robust in terms of being
  flexible with the output of rustc itself.
2019-05-01 08:09:40 -07:00
Steven Fackler 3331908a1d Release openssl v0.10.21 2019-04-30 21:59:02 -07:00
Steven Fackler dedbcc6570 Release v0.9.44 2019-04-30 21:54:53 -07:00
Steven Fackler d8a9a7fa11
Merge pull request #1100 from nanpuyue/patch-1
fix build err for taget `*-pc-windows-gnu`
2019-04-29 19:17:29 -07:00
Steven Fackler ee5b8b1f44
Merge pull request #1102 from npmccallum/master
Fix output size check for stream ciphers
2019-04-29 15:54:45 -07:00
Nathaniel McCallum 7ad3208937 Fix output size check for stream ciphers
The previous output size check presumed a block cipher. Therefore, it
enforced an unnecessary extra byte in the case of stream ciphers. This
patch ensures that our size checks don't force the caller to
overallocate for stream ciphers.
2019-04-29 17:45:12 -04:00
Charlie Li 995f9a9533 Whitelist future LibreSSL 2.9.x versions, as 2.9.1 is the first stable release. 2019-04-26 14:35:11 -04:00
南浦月 844c3c445a
fix build err for taget `*-pc-windows-gnu` 2019-04-26 15:27:16 +08:00
Steven Fackler 18eaa3d09e
Merge pull request #1099 from sfackler/atomic-deprecation
Fix deprecation warnings in tests
2019-04-25 10:10:34 -07:00