aboutsummaryrefslogtreecommitdiffstats
path: root/src/lib/hash
Commit message (Collapse)AuthorAgeFilesLines
* Make ssse3/sse2 dependencies explicit rather than implicitJack Lloyd2019-09-042-0/+2
| | | | Previously --disable-sse2/--disable-ssse3 would not work as expected
* Add BOTAN_FORCE_INLINE and use it in BLAKE2bJack Lloyd2019-08-311-3/+3
| | | | Closes #2089
* Fix a carry handling bug in StreebogJack Lloyd2019-08-261-3/+4
| | | | Closes #2082
* Remove BearSSL providerJack Lloyd2019-07-051-16/+1
| | | | | | BearSSL is much slower than Botan's builtins, and it is not commonly included in distributions so doesn't even have the advantage of ubiquity.
* Fix conflict when BMI2 is enabled and --single-amalgamation-fileJack Lloyd2019-06-191-3/+3
| | | | Not sure why this wasn't causing an error in the MSVC CI builds.
* Make the ISA list a listJack Lloyd2019-04-178-8/+28
|
* Fix some extra semicolonsJack Lloyd2019-03-271-1/+1
|
* Work around problem with GCC 4.8Jack Lloyd2019-03-262-2/+4
|
* Fix some warnings from GCC 9Jack Lloyd2019-03-252-2/+2
| | | | New redundant-move and pessimizing-move warnings found some
* s/Blake2b/BLAKE2b/Jack Lloyd2019-02-222-18/+20
| | | | As that is the proper name of the hash. Add a typedef for compat.
* Unroll SHA-3Jack Lloyd2019-02-212-132/+158
| | | | Improves performance by about 10-12%
* Recognize BLAKE2b also [ci skip]Jack Lloyd2019-01-311-1/+1
| | | | See #1822
* Add BMI2 variants for SHA-512 and SHA-3Jack Lloyd2019-01-1810-10/+413
| | | | Both about 33% faster on Skylake
* Avoid including rotate.h in bswap.hJack Lloyd2018-12-219-0/+9
| | | | | | | It was only needed for one case which is easily hardcoded. Include rotate.h in all the source files that actually use rotr/rotl but implicitly picked it up via loadstor.h -> bswap.h -> rotate.h include chain.
* Fix more MSVC warningsJack Lloyd2018-12-101-1/+1
|
* Fix some MSVC warningsJack Lloyd2018-12-101-1/+1
|
* Merge GH #1776 Clean ups in MDx_HashFunctionJack Lloyd2018-12-082-33/+49
|\
| * Require MDx_HashFunction block size to be a power of 2Jack Lloyd2018-12-082-14/+25
| | | | | | | | | | | | Allows replacing div/mod by a variable with a shift/mask. Allows storing just the bit count, which saves a few bytes.
| * Cleanups in MDx_HashFunctionJack Lloyd2018-12-082-32/+37
| |
* | Clean up Streebog and fix use of unaligned loadsJack Lloyd2018-12-082-94/+114
|/
* Avoid calling memset, memcpy within library codeJack Lloyd2018-11-171-1/+3
| | | | | | | | | Prefer using wrappers in mem_utils for this. Current exception is where memcpy is being used to convert between two different types, since copy_mem requires input and output pointers have the same type. There should be a new function to handle conversion-via-memcpy operation.
* Fix some warnings in ARM specific codeJack Lloyd2018-10-011-12/+12
|
* Add support for hashing with CommonCryptoJose Pereira2018-09-031-3/+17
|
* Tiny optimization in MDx_HashFunction::final_resultJack Lloyd2018-05-281-2/+1
| | | | Typically not a bottleneck but this shows up in XMSS profiling
* Add BMI2-specific SHA-256Jack Lloyd2018-05-274-2/+167
| | | | | Currently just a copy of the baseline compression function, but compiled with BMI2 flags. On Skylake improves performance by about 40%.
* Add message to BOTAN_ARG_CHECK and use it more widelyJack Lloyd2018-05-131-2/+2
|
* Rename threefish module to threefish_512Jack Lloyd2018-03-062-2/+2
| | | | GH #1477
* Optimize SHA_3::expandJack Lloyd2018-01-311-15/+14
| | | | Noticable speedup for SHAKE esp with longer output lengths
* WhitespaceJack Lloyd2018-01-301-5/+5
|
* Use copy_out_vec_le instead of explicit loop in SHA-3 and KeccakJack Lloyd2018-01-302-4/+2
|
* Reorganize SHA-3 source file [ci skip]Jack Lloyd2018-01-301-35/+35
| | | | Put all the statics at beginning followed by member functions.
* Avoid allocating zero bytes for SHA-3 paddingJack Lloyd2018-01-304-27/+28
| | | | Inspired by #1433
* ABI for Aarch64 cryptoJack Lloyd2018-01-122-8/+2
|
* Make stream, block, hash and cipher mode base classes optionalJack Lloyd2018-01-122-0/+8
|
* Add missing ISA annotationsJack Lloyd2018-01-043-2/+3
| | | | Lack of these broke single file amalgamation (GH #1386)
* Avoid macros in Blake2b to workaround Visual C++ 2017 infinite loopJack Lloyd2017-12-301-39/+44
| | | | | Was fixed in 2017 SP1. Same bug hit Crypto++ - https://gihub.com/weidai11/cryptopp/issues/527
* Rename SSE4.x names to avoid underscoresJack Lloyd2017-12-112-2/+2
| | | | | | This breaks how we determine the ISA flags for amalgamation files. The code for doing that is kind of a hack but I don't want to mess with it right now, easier to just rename the ISA internally.
* Rename the SSE4 ISA extensionsJack Lloyd2017-12-112-2/+2
| | | | Simplifies macro generation
* Rename CRC24 tablesJack Lloyd2017-12-101-9/+9
| | | | These conflict with name of temp variables and MSVC gets noisy.
* Fix a typo in sha1_sse2.cppFelix Yan2017-12-031-1/+1
|
* Minor documentation fix in HashFunction::create_or_throw.Marcus Brinkmann2017-11-261-1/+1
|
* Merge GH #1281 CRC24 optimizationsJack Lloyd2017-10-292-67/+213
|\
| * CRC24 performance improvement (32 bits in parallel)Krzysztof Kwiatkowski2017-10-292-67/+213
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Algorithm uses 4 tables of precalculated CRC24 values, thanks to which it can process in parallel 32 bits of data. This tric doubles performance Further improvements are possible. Results - (tested with RNP) processing 1GB armor data ``` OLD: rnp --enarmor=msg /tmp/1gb.rnd --output 4.48s user 0.89s system 98% cpu 5.429 total NEW: rnp --enarmor=msg /tmp/1gb.rnd --output 2.38s user 0.86s system 79% cpu 4.089 total OLD: rnp --dearmor out.xxx --output out.d 5.58s user 0.65s system 98% cpu 6.338 total NEW: rnp --dearmor out.xxx --output out.d 3.28s user 0.84s system 96% cpu 4.275 total ```
* | Include <memory> in base type headersJack Lloyd2017-10-291-0/+1
|/ | | | Needed for the create calls
* Fix build on 32-bitJack Lloyd2017-10-261-1/+1
|
* Avoid invalid iterator woesJack Lloyd2017-10-261-8/+18
|
* Blake2b optimizationsJack Lloyd2017-10-263-133/+94
| | | | | Nothing major but does improve perf for large buffers from 910 MB/s to 970 MB/s on Skylake.
* Convert http:// links to https:// where possibleJack Lloyd2017-10-242-2/+2
|
* Interleave SM3 message expansionJack Lloyd2017-10-121-141/+142
| | | | Reduces stack usage and a bit faster
* Add compile-time rotation functionsJack Lloyd2017-10-1211-370/+381
| | | | | | | | | | | | | | | | | The problem with asm rol/ror is the compiler can't schedule effectively. But we only need asm in the case when the rotation is variable, so distinguish the two cases. If a compile time constant, then static_assert that the rotation is in the correct range and do the straightforward expression knowing the compiler will probably do the right thing. Otherwise do a tricky expression that both GCC and Clang happen to have recognize. Avoid the reduction case; instead require that the rotation be in range (this reverts 2b37c13dcf). Remove the asm rotations (making this branch illnamed), because now both Clang and GCC will create a roll without any extra help. Remove the reduction/mask by the word size for the variable case. The compiler can't optimize that it out well, but it's easy to ensure it is valid in the callers, especially now that the variable input cases are easy to grep for.
generated by cgit v1.2.3 (git 2.39.1) at 2024-12-27 04:09:49 +0000