Commit graph

342 commits

Author SHA1 Message Date
Manuel Pégourié-Gonnard 6df09578bb Revert "Mutex call in x509_crt.c depended on PTHREAD specific instead of generic"
This reverts commit 9eae7aae80.
2014-02-12 09:29:05 +01:00
Paul Bakker 2ceda57989 Ability to force the Entropy module to use SHA-256 as its basis
By default the SHA-512 module is used if both are available. On some
systems, SHA-256 is the better choice.

Contributed by: Gergely Budai
2014-02-06 15:55:25 +01:00
Paul Bakker cd6d69a414 Fixed testing with out-of-source builds using cmake 2014-02-06 15:44:11 +01:00
Paul Bakker f2561b3f69 Ability to provide alternate timing implementation 2014-02-06 15:32:26 +01:00
Paul Bakker 47703a0a80 More entropy functions made thread-safe (add_source, update_manual, gather) 2014-02-06 15:01:20 +01:00
Paul Bakker 9eae7aae80 Mutex call in x509_crt.c depended on PTHREAD specific instead of generic
threading
2014-02-06 14:51:53 +01:00
Paul Bakker 6a28e722c9 Merged platform compatibility layer 2014-02-06 13:44:19 +01:00
Paul Bakker ab50d8d30c Merged RSA-PSS support in Certificate, CSR and CRL 2014-02-06 13:14:56 +01:00
Paul Bakker e31b1d992a Added Curve preference order to ChangeLog 2014-02-06 13:08:02 +01:00
Manuel Pégourié-Gonnard fbf0915404 Fix bug in RSA PKCS#1 v1.5 "reversed" operations 2014-02-05 17:01:24 +01:00
Paul Bakker 5fb8efe71e Merged HMAC-DRBG code 2014-02-05 15:55:18 +01:00
Manuel Pégourié-Gonnard 6e8e34d61e Fix ecp_gen_keypair()
Too few tries caused failures for some curves (esp. secp224k1)
2014-02-05 15:53:45 +01:00
Paul Bakker 2cb1a0c400 Fixed indication for TrustInSoft fix 2014-01-27 13:36:23 +01:00
Paul Bakker b84582b7a6 Fixed up ChangeLog with missing bug fixes 2014-01-27 12:23:43 +01:00
Paul Bakker 2aca241425 Ready for release 1.3.4 2014-01-27 11:59:30 +01:00
Paul Bakker e6c2ddb0b8 Updated ChangeLog with deterministic ECDSA 2014-01-27 11:59:29 +01:00
Paul Bakker 42099c3155 Revert "Add pk_rsa_set_padding() and rsa_set_padding()"
This reverts commit b4fae579e8.

Conflicts:
	library/pk.c
	tests/suites/test_suite_pk.data
	tests/suites/test_suite_pk.function
2014-01-27 11:59:29 +01:00
Manuel Pégourié-Gonnard 5cac583482 Factor out some common code 2014-01-25 12:48:58 +01:00
Paul Bakker d75ba40cc3 SMTP lines are officially terminated with CRLF, ssl_mail_client fixed 2014-01-24 16:12:18 +01:00
Paul Bakker 556efba51c Added AES CFB8 mode 2014-01-24 15:38:12 +01:00
Paul Bakker 5862eee4ca Merged RIPEMD-160 support 2014-01-22 14:18:34 +01:00
Paul Bakker 0ac99ca7bc Merged support for secp224k1, secp192k1 and secp25k1 2014-01-22 13:10:48 +01:00
Paul Bakker 3eb9673e19 Updated ChangeLog with recent changes 2014-01-22 13:08:19 +01:00
Paul Bakker a8fd3e31ed Removed POLARSSL_THREADING_DUMMY option 2013-12-31 11:54:08 +01:00
Paul Bakker 6992eb762c Fixed potential overflow in certificate size in ssl_write_certificate() 2013-12-31 11:38:33 +01:00
Paul Bakker 6ea1a95ce8 Added missing MPI_CHK() around some statements 2013-12-31 11:17:14 +01:00
Paul Bakker 5bc07a3d30 Prepped for 1.3.3 2013-12-31 10:57:44 +01:00
Paul Bakker c73879139e Merged ECP memory usage optimizations 2013-12-31 10:33:47 +01:00
Paul Bakker 956c9e063d Reduced the input / output overhead with 200+ bytes and covered corner
case

The actual input / output buffer overhead is only 301 instead of 512.
This requires a proper check on the padding_idx to prevent out of bounds
reads.

Previously a remote party could potentially trigger an access error and
thus stop the application when sending a malicious packet having
MAX_CONTENT_LEN of data, 32 bytes of MAC and a decrypted padlen of .
This would result in reading from in_ctr + 13 + 32 + MAX_CONTENT_LEN - 1 - 1
for 256 bytes (including fake padding check). Or 13 + 32 bytes over the
buffer length.

We now reset padding_idx to 0, if it's clear that it will never be a
valid padding (padlen > msg_len || msg_len + padlen + 256 > buffer_len)
2013-12-30 15:00:51 +01:00
Paul Bakker f9c4953e39 Added version of the SSL pthread server example 2013-12-30 14:55:54 +01:00
Paul Bakker 23116fdb53 Merged AES-NI support for AES, AES-GCM and AES key scheduling 2013-12-30 14:10:35 +01:00
Paul Bakker 1a56fc96a3 Fixed x509_crt_parse_path() bug on Windows platforms 2013-12-19 13:52:33 +01:00
Paul Bakker 5a607d26b7 Merged IPv6 support in the NET module 2013-12-17 14:34:19 +01:00
Paul Bakker 5ab68ba679 Merged storing curves fully in ROM 2013-12-17 13:11:18 +01:00
Paul Bakker fdf946928d Merged support for ECDH-RSA / ECDH-ECDSA key exchanges and ciphersuites 2013-12-17 13:10:27 +01:00
Paul Bakker f70fe81a6e Fixed memory leak in benchmark application 2013-12-17 13:09:12 +01:00
Paul Bakker 6f0636a09f Potential memory leak in ssl_ticket_keys_init() 2013-12-17 13:09:12 +01:00
Paul Bakker 48d78a5e60 Merged support for Curve25519 2013-12-05 16:12:40 +01:00
Manuel Pégourié-Gonnard 9a4a5ac4de Fix bug in mpi_set_bit 2013-12-05 15:58:38 +01:00
Paul Bakker b14817d10a Updated ChangeLog for splitting off curves from ecp.c 2013-12-02 22:03:23 +01:00
Paul Bakker 9dc53a9967 Merged client ciphersuite order preference option 2013-12-02 14:56:27 +01:00
Paul Bakker 014f143c2a Merged EC key generation support 2013-12-02 14:55:09 +01:00
Paul Bakker 4040d7e95c Merged more constant-time checking in RSA 2013-12-02 14:53:23 +01:00
Paul Bakker c3d0d07a7a Merged change from readdir_r() to readdir() + threading 2013-12-02 14:52:50 +01:00
Paul Bakker 7aa0375b78 Updated ChangeLog to reflect recent changes 2013-11-26 17:37:31 +01:00
Paul Bakker a9a028ebd0 SSL now gracefully handles missing RNG 2013-11-21 17:31:06 +01:00
Paul Bakker f2b4d86452 Fixed X.509 hostname comparison (with non-regular characters)
In situations with 'weird' certificate names or hostnames (containing
non-western allowed names) the check would falsely report a name or
wildcard match.
2013-11-21 17:30:23 +01:00
Paul Bakker f4dc186818 Prep for PolarSSL 1.3.2 2013-11-04 17:29:42 +01:00
Paul Bakker e1121b6217 Update ChangeLog for renegotiation changes 2013-10-31 15:57:22 +01:00
Paul Bakker 7b0be68977 Support for serialNumber, postalAddress and postalCode in X509 names 2013-10-29 14:24:37 +01:00