aboutsummaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* Added a EVP_CIPHER_CTX_init() callPaul Kehrer2013-09-302-0/+3
| | | | | | | * In OpenSSL when you alloc an EVP_CIPHER_CTX you must then init it (which just zeroes the allocated memory). If you do not then it is possible for things to kerplode when passing the uninitialized context to EVP_EncryptInit_ex(). This patch fixes that.
* Merge pull request #75 from alex/expose-apiDonald Stufft2013-09-304-25/+63
|\ | | | | Make the lib and ffi public for the OpenSSL binding and document them
| * Put a warningAlex Gaynor2013-09-301-0/+7
| |
| * Make the lib and ffi public for the OpenSSL binding and document themAlex Gaynor2013-09-304-25/+56
|/
* Merge pull request #74 from alex/copyrightDonald Stufft2013-09-291-1/+1
|\ | | | | Corrected the copyright
| * Corrected the copyrightAlex Gaynor2013-09-291-1/+1
|/
* Merge pull request #73 from alex/dedentDonald Stufft2013-09-282-19/+21
|\ | | | | Dedent the C declarations to give us more space to work with
| * Dedent the C declarations to give us more space to work withAlex Gaynor2013-09-282-19/+21
|/
* Merge pull request #71 from reaperhulk/cryptrec-openssl-loaderAlex Gaynor2013-09-272-2/+203
|\ | | | | add cryptrec and openssl test vector loaders + tests
| * simplify cryptrec loader, improve commentsPaul Kehrer2013-09-271-27/+17
| |
| * remove distinction between encrypt/decrypt in the openssl loaderPaul Kehrer2013-09-242-124/+13
| | | | | | | | | | | | | | * This was done because all the OpenSSL test vectors we're currently using are identical between encrypt/decrypt. * Removed a bunch of unneeded unit tests that checked the encrypt/decrypt vectors
| * simplify and annotate loaders based on review feedbackPaul Kehrer2013-09-231-14/+17
| |
| * crazy pep8 contortionsPaul Kehrer2013-09-151-25/+42
| |
| * sort the cryptrec loader dict's items to test reliablyPaul Kehrer2013-09-151-1/+1
| |
| * add cryptrec and openssl test vector loaders + testsPaul Kehrer2013-09-152-2/+304
|/
* Merge pull request #69 from reaperhulk/camellia-vectorsAlex Gaynor2013-09-156-0/+11687
|\ | | | | Camellia test vectors
| * Camellia test vectorsPaul Kehrer2013-09-146-0/+11687
|/ | | | | | * ECB tests come from the NTT Camellia pages (as linked to by the latest [CRYPTREC](http://www.cryptrec.go.jp/english/method.html)). * CFB, OFB, CBC tests from OpenSSL evptests
* Merge pull request #68 from alex/cleanupDonald Stufft2013-09-131-28/+14
|\ | | | | Slightly clean up the OpenSSL binding
| * Inline this function, it didn't do muchAlex Gaynor2013-09-131-9/+3
| |
| * Slightly clean up the OpenSSL bindingAlex Gaynor2013-09-131-23/+15
|/ | | | | - reduce the amount of state needed on the API instance - fixed import ordering
* Merge pull request #67 from reaperhulk/api-refactorAlex Gaynor2013-09-134-30/+93
|\ | | | | API refactor to separate cffi declarations
| * remove import function, rename attrs to upper in modulesPaul Kehrer2013-09-133-16/+9
| |
| * Various PR review fixesPaul Kehrer2013-09-135-67/+63
| | | | | | | | | | | | | | | | | | | | | | | | | | | | * Changed some single quotes to double quotes * Moved the files in the cffi package back into the openssl package * evp, opensslv * Changed attrs from ALL CAPS * Fixed up docstrings * Stopped using fromlist=['*'] * No fallback on definition imports. You must supply includes, types, functions * Change includes, types, functions to be strings rather than lists for now * Removed teardown since we're not using it right now
| * API refactor to separate cffi declarationsPaul Kehrer2013-09-135-30/+104
|/ | | | | | | * Moved to structure similar to OpenTLS * api modified to load the cffi modules * cffi verify methods broken out * Jarret Raim and I pair programmed this
* Merge pull request #60 from dreid/ofb-iv-not-nonceDonald Stufft2013-09-124-15/+11
|\ | | | | OFB uses an IV instead of a nonce.
| * OFB uses an initialization vector instead a nonce.David Reid2013-09-111-6/+8
| |
| * ModeWithNonce no longer exists.David Reid2013-09-111-2/+0
| |
| * ModeWithNonce is no longer used.David Reid2013-09-111-4/+0
| |
| * OFB is specified as having an IV instead of a Nonce.David Reid2013-09-111-3/+3
| | | | | | https://en.wikipedia.org/wiki/Block_cipher_mode_of_operation#Output_feedback_.28OFB.29
* | Merge pull request #62 from dstufft/test-pypyAlex Gaynor2013-09-121-0/+6
|\ \ | |/ |/| Enable testing on PyPy
| * Use the version of PyPy from PPAs to get PyPy 2.1Donald Stufft2013-09-121-0/+6
|/
* Merge pull request #59 from reaperhulk/cfb-supportDavid Reid2013-09-113-0/+69
|\ | | | | CFB support
| * add CFB to documentationPaul Kehrer2013-09-111-0/+13
| |
| * CFB supportPaul Kehrer2013-09-112-0/+56
|/ | | | | | | | | | This requires a bit of explanation. OpenSSL has methods that implement standard CFB, 1-bit CFB (cfb1), and 8-bit CFB (cfb8). Unfortunately, while old (read: 0.9.7) versions of OpenSSL appear to test these variants, newer versions have a comment stating that cfb{1,8} are unsupported. Accordingly, I've backed out any support for the variants for now. We can add it back into the CFB class if and when we gain a backend that supports arbitrary s for 1 <= s <= block_size
* Merge pull request #57 from reaperhulk/ofb-supportDonald Stufft2013-09-105-0/+73
|\ | | | | Output feedback mode support + test vectors (aes)
| * change OFB iv to nonce to reflect dstufft nomenclature pitchPaul Kehrer2013-09-104-11/+15
| | | | | | | | | | | | * Namely, we should try to call things IV if reuse leaks a small amount of data and nonce if reuse can result in a complete break. This can be somewhat ambiguous, but we'll track in #58
| * add output feedback mode support + test vectors (aes)Paul Kehrer2013-09-103-0/+69
| |
* | Merge pull request #56 from alex/docs-cleanupDonald Stufft2013-09-101-5/+9
|\ \ | |/ |/| Cleaned up the docs for ECB
| * Cleaned up the docs for ECBAlex Gaynor2013-09-101-5/+9
|/
* Merge pull request #53 from reaperhulk/ecb-support-im-sorryAlex Gaynor2013-09-107-5/+103
|\ | | | | ECB Support
| * address style on import + remove a tox dependency (in setup.py already)Paul Kehrer2013-09-102-1/+1
| |
| * move abc, inline introspect method, use six for abcsPaul Kehrer2013-09-107-17/+14
| | | | | | | | | | | | | | * abc moved to cryptography.primitive.interfaces * six added to dependencies * six used to have py2x/py3x compatible abc * nonce abc removed for now
| * modify modes to use abc so api can determine what attribute to callPaul Kehrer2013-09-106-12/+36
| | | | | | | | | | | | | | * Due to a circular dependency issue I had to put the abcs in cryptography.primitives.abc.block.modes * The ABCs look like they do because that is the form that is compatible with 2.x and 3.x
| * rename get_null_for_ecb to get_iv_for_ecb per alex's commentsPaul Kehrer2013-09-103-4/+4
| |
| * add get_iv_or_nonce() methods to replace _get_iv() on apiPaul Kehrer2013-09-103-18/+13
| |
| * remove unneeded init in ECB class, add warning to docs for ECB modePaul Kehrer2013-09-102-3/+1
| |
| * add ECB support to create_block_cipher_contextPaul Kehrer2013-09-102-1/+19
| | | | | | | | | | * This is a basic refactor to support ECB and CBC mode in this method. We can use this as a starting point to discuss a better solution.
| * Add ECB class + docs + testsPaul Kehrer2013-09-103-0/+66
|/ | | | | * Slightly refactors test_nist to allow fetching of data that has no IV * Does not modify create_block_cipher_context (next commit)
* Merge pull request #55 from reaperhulk/parameterize-encrypt-changesAlex Gaynor2013-09-101-22/+30
|\ | | | | Changes to parameterize_encrypt_test to support other NIST vectors
| * changes to parameterize_encrypt_test to support other NIST vectorsPaul Kehrer2013-09-101-22/+30
| | | | | | | | | | * ECB has no IV so we need to be able to pass params rather than have them hardcoded in the method
generated by cgit v1.2.3 (git 2.25.1) at 2025-08-19 02:45:34 +0000