summaryrefslogtreecommitdiff
path: root/src/lib (follow)
Commit message (Collapse)AuthorAgeFilesLines
* Include TLSv1.3 in version handling code.jsing2018-11-061-3/+9
| | | | | | | This is effectively a no-op, since most of the code clamps to the maximum version supported by the TLS method (which are still at TLSv1.2). ok beck@ bluhm@ tb@
* Add TLS1_3_VERSION and SSL_OP_NO_TLSv1_3 defines under guards.jsing2018-11-062-2/+13
| | | | ok beck@ bluhm@ tb@
* Start working towards adding feature flags (rather than anti-feature flags)jsing2018-11-061-1/+10
| | | | | | | for LibreSSL. Add a (commented out) feature flag for TLSv1.3 and define the OPENSSL_NO_TLS1_3 anti-feature flag based on the feature flag. ok beck@ bluhm@ tb@
* Make use of bn_rand_interval() where appropriate.tb2018-11-055-33/+24
| | | | ok beck jsing
* Introduce bn_rand_interval() that allows specifying an interval [a, b)tb2018-11-052-2/+30
| | | | | | from which a a BIGNUM is chosen uniformly at random. ok beck jsing
* Eliminate a few "} else" branches, a few unneeded NULL checks beforetb2018-11-053-43/+33
| | | | | | freeing and indent nearby labels. ok beck jsing
* Remove two unnecessary BN_FLG_CONSTTIME dances: BN_mod_exp_ct() alreadytb2018-11-052-21/+7
| | | | | | takes care of this internally. ok beck jsing
* Rename the TLS Supported Elliptic Curves extension to Supported Groups.jsing2018-11-056-88/+92
| | | | | | | | | RFC 7919 renamed the Supported Elliptic Curves TLS extension to Supported Groups and redefined it to include finite field DH (FFDH) in addition to elliptic curve DH (ECDH). As such, rename the TLS extension and change the associated code to refer to groups rather than curves. ok beck@ tb@
* Rework the TLS extension handling code to improve readability/flexibility,jsing2018-11-051-89/+112
| | | | | | by moving the needs/build/parse functions into their own struct. ok beck@ tb@
* Implement coordinate blinding for EC_POINT.tb2018-11-0510-18/+110
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Based on OpenSSL commit 875ba8b21ecc65ad9a6bdc66971e50 by Billy Brumley, Sohaib ul Hassan and Nicola Tuveri. ok beck jsing commit 875ba8b21ecc65ad9a6bdc66971e50461660fcbb Author: Sohaib ul Hassan <soh.19.hassan@gmail.com> Date: Sat Jun 16 17:07:40 2018 +0300 Implement coordinate blinding for EC_POINT This commit implements coordinate blinding, i.e., it randomizes the representative of an elliptic curve point in its equivalence class, for prime curves implemented through EC_GFp_simple_method, EC_GFp_mont_method, and EC_GFp_nist_method. This commit is derived from the patch https://marc.info/?l=openssl-dev&m=131194808413635 by Billy Brumley. Coordinate blinding is a generally useful side-channel countermeasure and is (mostly) free. The function itself takes a few field multiplicationss, but is usually only necessary at the beginning of a scalar multiplication (as implemented in the patch). When used this way, it makes the values that variables take (i.e., field elements in an algorithm state) unpredictable. For instance, this mitigates chosen EC point side-channel attacks for settings such as ECDH and EC private key decryption, for the aforementioned curves. For EC_METHODs using different coordinate representations this commit does nothing, but the corresponding coordinate blinding function can be easily added in the future to extend these changes to such curves. Co-authored-by: Nicola Tuveri <nic.tuv@gmail.com> Co-authored-by: Billy Brumley <bbrumley@gmail.com> Reviewed-by: Tim Hudson <tjh@openssl.org> Reviewed-by: Nicola Tuveri <nic.tuv@gmail.com> Reviewed-by: Andy Polyakov <appro@openssl.org> (Merged from https://github.com/openssl/openssl/pull/6526)
* Implement C11's aligned_alloc(3). ok guenther@otto2018-11-052-5/+84
|
* Clean up the code that checks if we can choose an EC cipher suite.jsing2018-11-053-16/+7
| | | | | | | | | The tls1_check_ec_tmp_key() function is now rather misnamed, so just inline the code. Also, rather than running tls1_get_shared_curve() once per EC cipher suite, we can run it once at the start of the ssl3_choose_cipher() function. ok bluhm@ tb@
* Consolidate all of the SSL method structs/functions into a single file.jsing2018-11-058-871/+672
| | | | Discussed with tb@
* Mop up ecdh_tmp, since it is no longer used.jsing2018-11-053-25/+4
| | | | ok bluhm@ tb@
* Define OPENSSL_NO_ASYNC - our libcryptosink does not have built in asyncjsing2018-11-041-1/+1
| | | | features (and possibly never will).
* Update the opensslfeatures.h to include all of the OPENSSL_NO_* flags thatjsing2018-11-041-8/+84
| | | | | | | | | | currently exist in OpenSSL - comment out that ones that we do not already define. Some OPENSSL_NO_* flags that we define have been removed from OpenSSL (and code that depended on these to know when features are not available now think that the features have been enabled...). We keep these defined but in their own separate group. ok bluhm@ tb@
* Reformat and sort the OPENSSL_NO_* defines.jsing2018-11-041-25/+25
|
* move X509_get0_serialNumber where it belongstb2018-11-021-1/+1
|
* whitespacetb2018-11-021-2/+2
|
* Bump libcrypto/libssl/libtls majors due to symbol removals (libcrypto)jsing2018-10-243-6/+6
| | | | and changes to struct visibility/sizes (libssl).
* Make more of libssl's record layer state internal.jsing2018-10-248-88/+86
| | | | | | | | | | In January 2017, we changed large amounts of libssl's data structures to be non-visible/internal, however intentionally left things that the software ecosystem was needing to use. The four or so applications that reached into libssl for record layer related state now implement alternative code. As such, make these data structures internal. ok tb@
* Remove a GOST data symbol that should not be exported.jsing2018-10-241-1/+0
|
* Remove a bunch of ancient and highly crufty ASN.1 related code fromjsing2018-10-249-1633/+22
| | | | | | libcrypto (the "new" stuff replaced this back around 2000 or so...). ok tb@
* Remove stack related macros that should have been nuked whenjsing2018-10-241-333/+1
| | | | {CMS,KRB5,SRP} were removed.
* Avoid calling memcpy with a length <= 0. Reported due to a GCC 7.3.0tb2018-10-201-5/+6
| | | | | | | compiler warning by Pavel Kraynyukhov. A similar fix was made in OpenSSL commit 369e93398b68b8a328e6c1d766222b. ok inoguchi
* RFC 3394 section 2 states that we need at least two 64 bit blockstb2018-10-201-6/+6
| | | | | | | | | | | | | | for wrapping and, accordingly, three 64 bit blocks for unwrapping. That is: we need at least 16 bytes for wrapping and 24 bytes for unwrapping. This also matches the lower bounds that OpenSSL have in their CRYPTO_128_{un,}wrap() functions. In fact, if we pass an input with 'inlen < 8' to AES_unwrap_key(), this results in a segfault since then inlen -= 8 underflows. Found while playing with the Wycheproof keywrap test vectors. ok bcook
* As per POSIX, when str{,r}chr is comparing it should convert c to a char.martijn2018-10-014-8/+10
| | | | | | | | | | The C implementation of str{,r}chr are not linked to the build, because assembly implementations are used, but change to code for easier reference. At least the i386 and amd64 are checked and seem to do the correct thing. Found thanks to the csh any/strchr change. minor pointers and OK millert@
* bump for LibreSSL 2.8.2bcook2018-09-301-3/+3
|
* bump for LibreSSL 2.8.1libressl-v2.8.1bcook2018-09-231-3/+3
|
* Simplify initialization of asn1_cb; use correct spelling of NULL.tb2018-09-171-4/+2
|
* sync with mozilla-release (one removal, TURKTRUST, more details atsthen2018-09-121-48/+1
| | | | | | https://bugzilla.mozilla.org/show_bug.cgi?id=1439127) ok danj guenther millert
* tweak previous;jmc2018-09-122-4/+4
|
* crank to follow minor crank in libcrypto; ok tb@ jsing@djm2018-09-122-2/+2
|
* Add some accessor functions:djm2018-09-128-17/+138
| | | | | | RSA_meth_get_finish() RSA_meth_set1_name() EVP_CIPHER_CTX_(get|set)_iv() feedback and ok jsing@ tb@
* Remove now unused code for EVP_CIPH_FLAG_AEAD_CIPHER and EVP_CIPH_GCM_MODE.jsing2018-09-084-77/+28
| | | | ok inoguchi@ tb@
* SSL_MAX_DIGEST is no longer needed.jsing2018-09-082-17/+10
|
* ASN1_OBJECTs should be freed with ASN1_OBJECT_free(3), not with free(3).tb2018-09-081-2/+2
| | | | ok inoguchi, jsing
* indent labelstb2018-09-082-8/+8
|
* missing word & a couple of typostb2018-09-081-3/+3
|
* Drop SSL_CIPHER_ALGORITHM2_AEAD flag.jsing2018-09-064-31/+25
| | | | | | | All of our algorithm_mac == SSL_AEAD cipher suites use EVP_AEAD, so we can condition on that rather than having a separate redundant flag. ok tb@
* Use the newer/more sensible names for EVP_MD_CTX_* functions.jsing2018-09-055-16/+16
| | | | | | | | | | EVP_MD_CTX_create -> EVP_MD_CTX_new EVP_MD_CTX_destroy -> EVP_MD_CTX_free This should make the intent more obvious and reduce head scratching during code reviews. Raised by tb@
* Correctly clear the current cipher state, when changing cipher state.jsing2018-09-053-41/+37
| | | | | | | | | | | | | | | | | When a renegotiation results in a change of cipher suite, the renegotation would fail if it switched from AEAD to non-AEAD or vice versa. This is due to the fact that the previous EVP_AEAD or EVP_CIPHER state remained, resulting in incorrect logic that caused MAC failures. Rename ssl_clear_cipher_ctx() to ssl_clear_cipher_state() and split it into separate read/write components, then call these functions from the appropriate places when a ChangeCipherSpec message is being processed. Also, remove the separate ssl_clear_hash_ctx() calls and fold these into the ssl_clear_cipher_{read,write}_state() functions. Issue reported by Bernard Spil, who also tested this diff. ok tb@
* use timing-safe compares for checking results in signature verificationdjm2018-09-054-9/+10
| | | | | | (there are no known attacks, this is just inexpensive prudence) feedback and ok tb@ jsing@
* Stop using composite EVP_CIPHER AEADs.jsing2018-09-031-25/+7
| | | | | | | | | | | The composite AEADs are "stitched" mode ciphers, that are only supported on some architectures/CPUs and are designed to be faster than a separate EVP_CIPHER and EVP_MD implementation. The three AEADs are used for less than ideal cipher suites (if you have hardware support that these use there are better cipher suite options), plus continuing to support AEADs via EVP_CIPHER is creating additional code complexity. ok inoguchi@ tb@
* Stop handling AES-GCM via ssl_cipher_get_evp().jsing2018-09-031-20/+3
| | | | | | | All of the AES-GCM ciphersuites use the EVP_AEAD interface, so there is no need to support them via EVP_CIPHER. ok inoguchi@ tb@
* Clean up SSL_DES and SSL_IDEA remnants.jsing2018-09-031-41/+13
| | | | | | | All ciphersuites that used these encryption algorithms were removed some time ago. ok bcook@ inoguchi@ tb@
* Elliptic curve arithmetic only makes sense between points that belong totb2018-09-021-1/+5
| | | | | | | | | | | | the same curve. Some Wycheproof tests violate this assumption, making ECDH_compute_key() compute and return garbage. Check that pub_key lies on the curve of the private key so that the calculations make sense. Most paths that get here have this checked (in particular those from OpenSSH and libssl), but one might get here after using d2i_* or manual computation. discussed with & ok jsing; "good catch!" markus
* Tweak comment.tb2018-09-011-5/+2
|
* Remove unused argument to tls1_change_cipher_state_cipher().jsing2018-08-311-7/+4
|
* Nuke ssl_pending/ssl_shutdown function pointers.jsing2018-08-309-56/+14
| | | | | | | ssl3_pending() is used for all protocols and dtls1_shutdown() just calls ssl3_shutdown(), so just call the appropriate function directly instead. ok beck@ inoguchi@ tb@
generated by cgit v1.2.3-55-g6feb (git 2.39.1) at 2025-08-04 08:26:35 +0000