aboutsummaryrefslogtreecommitdiffstats
path: root/src
Commit message (Collapse)AuthorAgeFilesLines
* Throw Internal_Error instead of Algorithm_Not_Found if no usable RNGlloyd2009-09-081-1/+1
| | | | is enabled in the build.
* Add *s before comment lines in file headerslloyd2009-09-084-8/+8
|
* propagate from branch 'net.randombit.botan.1_8' (head ↵lloyd2009-09-043-2/+3
|\ | | | | | | | | | | 75371777750b63ef94693602202c5104f217a987) to branch 'net.randombit.botan' (head 3f53f01c349eeee89288b1922fbde45b283c958c)
| * Add a new option --no-autoload to configure.py. This will produce a minimallloyd2009-09-043-2/+3
| | | | | | | | | | | | | | | | | | | | | | build (only libstate, utils, plus dependencies), which can be extended with use of --enable-modules. To add new modules to the set of always-loaded, use 'load_on always' in info.txt Also fix a few small build problems that popped up when doing a minimal build. Requested by a user.
* | According to the Linux sources, S/390 and PowerPC can both do unalignedlloyd2009-09-044-0/+12
| | | | | | | | | | | | memory accesses. Since this can be a pretty big win, enable it for them. The m68k apparently also can, except in its (modern) Coldfire version, but it's always big endian so mark that as such.
* | propagate from branch 'net.randombit.botan.1_8' (head ↵lloyd2009-08-312-4/+14
|\| | | | | | | | | | | c2624292793f396cf940403e0d12073a9b2c7b17) to branch 'net.randombit.botan' (head 07a71effa1ba495b6ea57b2490ad38bf58a23bd0)
| * Remove redunant buffer create calls in XTSlloyd2009-08-281-4/+0
| |
| * In XTS, reset the buffer positions at the end of a message.lloyd2009-08-271-0/+4
| |
| * Thomas Moschny mentioned that OpenSSL 1.0 betas disable MD2 by default.lloyd2009-08-271-0/+10
| | | | | | | | Wrap the EVP_ calls in OPENSSL_NO_XXX checks to handle this.
* | Add Rivest's package transformlloyd2009-08-313-0/+190
| |
* | Combine the 4 sbox calculations in the key schedule so that all four arelloyd2009-08-312-25/+25
| | | | | | | | | | computed in parallel. Not a huge win but slightly faster (which affects things like Lion when using Turing), most likely due to more available ILP
* | Hoist creation of buffer in Lion encrypt looplloyd2009-08-311-4/+4
| |
* | Remove unneeded include in xtea.cpplloyd2009-08-271-1/+0
| |
* | Instead of each SSE2 implementation specifying which compilers + CPUs itlloyd2009-08-273-25/+8
| | | | | | | | | | works on, have sse2_eng rely on a specific compiler/arch; each sse2 impl depends on the engine anyway, so they will only be loaded if OK.
* | Include basefilt.h in filters.hlloyd2009-08-131-0/+2
| |
* | propagate from branch 'net.randombit.botan.encrypt-n' (head ↵lloyd2009-08-133-0/+210
|\| | | | | | | | | | | 378e7464abc6b3efcf9cb433f7fcec0adfbb9de0) to branch 'net.randombit.botan' (head dd9bdcc0cab8b761a1c9861f3a4fc625488c2ef5)
| * Add a new interface CryptoBox which provides basic password-based encryptionlloyd2009-08-133-0/+210
| | | | | | | | | | | | | | | | | | | | in a reasonable way. Low on features, which is rather intentional. There is a version code included in the format so further extensions are possible, if warranted. Inspired by the n-th mailing list request for such a class. Realized it was probably better that I design such code than random people who just want 'something that works'.
* | Use a much faster counter increment system, noticable speedups (~15%)lloyd2009-08-121-20/+21
| | | | | | | | for both Serpent and AES-128 in CTR mode.
* | In benchmark.cpp, use encrypt_n when benchmarking block cipherslloyd2009-08-121-2/+1
| |
* | For handling the last few blocks in Serpent_SSE2, invoke encrypt_n withlloyd2009-08-121-12/+2
| | | | | | | | | | however many blocks remain, rather than looping calling encrypt_n with a block size of 1 each time.
* | Add SSE2 Serpent decryptionlloyd2009-08-122-1/+307
| |
* | Small code cleanups in SSE2 Serpentlloyd2009-08-122-192/+195
| |
* | Use SSE2 unpack instructions instead of unions for input/output conversion.lloyd2009-08-122-300/+263
| | | | | | | | | | About 10% faster than previous. Currently 112 MiB/s in ECB mode, versus about 40 MiB/s in scalar mode, on my 2.4 GHz Core2
* | Add full 4-way SSE2 Serpent encryption. Load/store operations are vialloyd2009-08-123-47/+290
| | | | | | | | | | unions and can be made much faster using interleave operations I think. Currently ~2.5x faster in ECB or CTR mode on a Core2, which isn't too bad.
* | Make encrypt_n public for all BlockCipher implementations - unlike thelloyd2009-08-1131-58/+355
| | | | | | | | | | | | enc/dec functions it replaces, these are public interfaces. Add the first bits of a SSE2 implementation of Serpent. Currently incomplete.
* | Modify Keyed_Filter so it is a pure interfacelloyd2009-08-1116-150/+329
| | | | | | | | | | | | | | | | | | | | Modify ECB to use parallel encryption/decryption where possible Add toggles in build.h specifying how many blocks to process in parallel. Defaults to 8 blocks for all modes, which is sufficient that any likely parallelism can be extracted (via SIMD or concurrent execution) but not so much as to seem likely to cause cache problems (8*128 bits = 128 bytes, or two x86 cache lines)
* | Change the BlockCipher interface to support multi-block encryption andlloyd2009-08-1155-1321/+1740
|/ | | | | | | | | decryption. Currently only used for counter mode. Doesn't offer much advantage as-is (though might help slightly, in terms of cache effects), but allows for SIMD implementations to process multiple blocks in parallel when possible. Particularly thinking here of Serpent; TEA/XTEA also seem promising in this sense, as is Threefish once that is implemented as a standalone block cipher.
* Add 'Distributed under the terms of the Botan license' notices to the .Slloyd2009-08-118-41/+57
| | | | | files. Were missed by the automated script that added them to the cpp/h files, it appears.
* Use the default ABI instead of forcing n32 for mips32lloyd2009-08-103-33/+21
| | | | | | | | | | | | | | systems. This was something that for whatever reason that I have long since forogotten was a good idea on IRIX running MIPS circa a decade ago, but was reported to cause problems on the Debian builds. Add mipsel as an alias for the mips32 architecture for Debian. The mips32 submodel names were badly typoed and did not work correctly. Remove the leading mips32- and mips64- from MIPS submodel names.
* Add compiler settings for the Open64 (http://www.open64.net/) compiler,lloyd2009-08-031-0/+30
| | | | | | | based on the SGI Pro64 and Pathscale EKOpath compilers. Only tested on an x86-64 system running Linux (v4.2.1). Miscompiles a few of the block ciphers (segvs, didn't bother to diagnose further; recompile with -O1 to fix), other than that seems OK.
* Fix variable name in 32-bit bswap for VC++lloyd2009-08-031-1/+1
|
* Change how the pair is constructed for Sun Forte compatabilitylloyd2009-07-311-1/+1
|
* A typo in a macro check in bswap.h meant inline asm bswap was notlloyd2009-07-311-1/+1
| | | | used on Visual C++
* Add support for Dragonfly BSD (a fork of FreeBSD).lloyd2009-07-2517-1/+28
| | | | Contributed by Patrick Georgi
* Two changes to proc_walk:lloyd2009-07-251-2/+2
| | | | | | | | | | | | | | Don't read any file that is not world-readable. This avoids trouble when running as root, since on Linux various special files can cause odd interactions and/or blocking behavior when read (for instance /proc/kmsg). ssumption is that no such files are world-readable. This also avoids any issue of reading data that is potentially sensitive. Instead of reading the first 1 KB of each file, only read the first 128 bytes. This prevents large files (like /proc/config.gz or /proc/kallsyms) from swamping the input buffer; these inputs are pretty static and shouldn't count for much. Reducing to 128 bytes causes a poll to read about 400 different files, rather than ~30.
* Add an alias armv5tel for xscale, fixes autodetection for both Perl andlloyd2009-07-231-0/+1
| | | | | Python configure scripts. Previously Python version would give up, and the Perl one would guess i686 (!)
* Fix bigint_2word_mul - it tried to use MP_WORD_BITS which was not visiblelloyd2009-07-231-1/+1
| | | | | in this header, instead use the macro version BOTAN_MP_WORD_BITS. Only affected 64-bit SPARC builds.
* Replace 'XXX' and other such notes with FIXME for easy greppinglloyd2009-07-225-9/+10
|
* Disable STRIP_CODE_COMMENTS for Doxygen outputlloyd2009-07-211-1/+1
|
* Fix some unused variable nits pointed out by icc 10.1lloyd2009-07-214-9/+3
|
* One signature of PKCS8::PEM_encode was stray, and didn't have a coorespondinglloyd2009-07-211-5/+0
| | | | | | | | | implementation. This had been the case since at least 1.8.0. Remove it, since implementing this signature would require having the library create an AutoSeeded_RNG, which seems like it might be quite surprising to an unsuspecting user. Problem reported by M. Braun in ticket 44
* Move back to pk_codecs, propagate to pubkey-refactor donelloyd2009-07-214-0/+0
|
* Move from pk_codecs to pubkey to solve merge problemlloyd2009-07-214-0/+0
|
* Fix various auto generated dependencies that were wronglloyd2009-07-178-22/+6
|
* Correct some errors in the automatically generated dependencies.lloyd2009-07-168-7/+3
|
* Add missing info.txt fileslloyd2009-07-162-0/+24
|
* Correct source listings for moved fileslloyd2009-07-162-4/+0
|
* Move some files around to break up dependencies between directorieslloyd2009-07-1632-16/+48
|
* Add a script that reads the output of print_deps.py and rewriteslloyd2009-07-15102-328/+546
| | | | | | the info.txt files with the right module dependencies. Apply it across the codebase.
* Move the contents of pubkey/pubkey (which was kind of a catch-all tolloyd2009-07-1528-18/+6
| | | | | | | just toplevel pubkey). This was a convention I realized made sense sometime on when I was first doing the modularization changes. Move pkcs8.* and x509_key.* to pk_codecs
generated by cgit v1.2.3 (git 2.39.1) at 2024-12-27 07:05:23 +0000