aboutsummaryrefslogtreecommitdiffstats
path: root/src
Commit message (Collapse)AuthorAgeFilesLines
* Merge GH #1213 Improve ARM processor detection for systems without getauxvalJack Lloyd2017-09-188-502/+670
|\
| * Add ARM feature detection for systems without getauxvalJack Lloyd2017-09-172-15/+129
| | | | | | | | | | | | | | | | For iOS use sysctl to get the product name and use a static table of minimum versions. For everything else (not Linux/Android or iOS) try probe functions. Only for Aarch64 to keep things simple.
| * Split up cpuid.cpp into arch-specific filesJack Lloyd2017-09-178-502/+556
| | | | | | | | | | As more archs and detection methods are used, the file was getting a little hard to follow.
* | Be less forgiving in maintainer modeJack Lloyd2017-09-172-2/+2
| |
* | Avoid int vs size_t warning in OpenSSL codeJack Lloyd2017-09-171-1/+1
| |
* | Avoid old style casts in Ed25519Jack Lloyd2017-09-172-270/+262
| |
* | Fix Doxygen comment problemsJack Lloyd2017-09-171-4/+4
| |
* | Move socket classes to anon namespace.Jack Lloyd2017-09-171-172/+181
| | | | | | | | A little cleaner than in the function.
* | Fix Wshadow warningJack Lloyd2017-09-171-5/+5
|/
* Guh setting PATH in subshell...Jack Lloyd2017-09-171-2/+1
|
* Simplify code in get_processor_timestampJack Lloyd2017-09-161-21/+12
|
* Fix lintJack Lloyd2017-09-161-4/+4
|
* Fix an off by 8Jack Lloyd2017-09-161-1/+1
|
* Fix building Perl wrapper - ability to add OIDs removedJack Lloyd2017-09-162-18/+1
| | | | | | Still fails most of its tests though [ci skip]
* Correct output of `botan config libs`Jack Lloyd2017-09-161-2/+1
| | | | It output -lbotan-2.x instead of -lbotan-2
* Fix compareJack Lloyd2017-09-161-1/+1
|
* Revert SIMDing of xor_buf in f39aa7bbJack Lloyd2017-09-161-29/+27
| | | | | | | | | Forgot this causes all kinds of issues when the SIMD ISA is not part of the base instruction set (needing -maltivec, etc). It just happens to work on x86-64 because SSE2 is part of the base ISA there. Instead just unroll by 16x this seems to be enough that GCC figures it out.
* Small tweaks for CIJack Lloyd2017-09-163-4/+8
|
* Use constant_time_compare instead of same_memJack Lloyd2017-09-1622-23/+36
| | | | New name, same great operation
* De-inline xor_buf, add SIMD and unrollingJack Lloyd2017-09-163-34/+133
| | | | Improves CBC and OCB performance with AES-NI quite noticably
* Small cleanup in CBC modeJack Lloyd2017-09-162-21/+23
|
* Small cleanup in bcryptJack Lloyd2017-09-161-3/+3
|
* Add missing override annotationsJack Lloyd2017-09-161-6/+6
|
* Small simplification in CAST-128Jack Lloyd2017-09-161-41/+41
|
* Merge duplicated logic for clearing CPUID bitsJack Lloyd2017-09-164-49/+48
|
* Minor simplifications in SIMD_32 constructorJack Lloyd2017-09-161-2/+5
|
* Update Wycheproof GCM testsJack Lloyd2017-09-151-6/+38
|
* some formattingAtanas Filyanov2017-09-151-1/+5
|
* reenable 8 byte gcm tagsAtanas Filyanov2017-09-152-1/+11
|
* Allow GCM tag size in the range 96..128 bitsAtanas Filyanov2017-09-142-16/+45
|
* Merge GH #1208 Fix decoding of RFC 5915 ECC keysJack Lloyd2017-09-133-5/+31
|\
| * Remove inner/outer OID check in EC_PrivateKey ctorRené Korthaus2017-09-133-5/+31
| | | | | | | | | | | | | | | | RFC 5915 mandates that the OID of an ECDSA ECPrivateKey shall be id-ecPublicKey with the named curve as the parameters. ECPrivateKey may contain the named curve OID, too, which is compared to id-ecPublicKey. Such keys could not be loaded. We remove this check and add a test vector from strongswan.
* | allow 12 byte GCM tags and add test vectorsAtanas Filyanov2017-09-132-1/+49
|/
* Reference the right pipeJack Lloyd2017-09-111-1/+1
| | | | This didn't affect the results [ci skip]
* Merge GH #1202 Fix CFB IV carryover closes #1200Jack Lloyd2017-09-113-47/+115
|\
| * Address CFB carryover bugJack Lloyd2017-09-103-47/+115
| | | | | | | | | | | | Test data generated by 1.10 so hopefully no further issues here. GH #1200
* | Change SM2 encryption to match updated standard.Jack Lloyd2017-09-102-6/+7
| | | | | | | | | | | | | | | | Unfortunately it seems the SM2 format changed between 2010 and 2012, now the ciphertext is C1 || C3 || C2. Unfortunate no matter how you slice it, but at least it's easy to convert from one form to another.
* | Use prebuilt ccache binaryJack Lloyd2017-09-101-6/+1
| |
* | Remove test branch from appveyor yamlJack Lloyd2017-09-101-1/+0
| |
* | Fix a couple more VC warningsJack Lloyd2017-09-102-3/+3
| |
* | Fix bogus ffi testJack Lloyd2017-09-101-1/+1
| | | | | | | | Caught by a warning from VC
* | Enable clcacheJack Lloyd2017-09-102-1/+16
| |
* | Merge GH #1199 Run Windows CI build via Python scriptJack Lloyd2017-09-108-67/+126
|\ \
| * | Fix Sonar buildJack Lloyd2017-09-081-1/+1
| | |
| * | Fix var assignmentJack Lloyd2017-09-081-2/+2
| | |
| * | Fix ccache detectionJack Lloyd2017-09-081-1/+2
| | |
| * | Script the Windows CI buildJack Lloyd2017-09-088-66/+124
| |/
* | Merge GH #1203 Add Win32 filesystem API access for MinGWJack Lloyd2017-09-102-6/+55
|\ \
| * | Make it actually workJack Lloyd2017-09-101-2/+6
| | | | | | | | | | | | | | | Weirdly Wine at least does not set FILE_ATTRIBUTE_NORMAL on the files, instead sets FILE_ATTRIBUTE_ARCHIVE. ?? I have no idea what that's about.
| * | Well, it compilesJack Lloyd2017-09-102-6/+51
| |/
generated by cgit v1.2.3 (git 2.39.1) at 2024-12-28 17:54:28 +0000