aboutsummaryrefslogtreecommitdiffstats
path: root/src
Commit message (Collapse)AuthorAgeFilesLines
* Clang supports -marchlloyd2010-08-091-0/+4
|
* Add an implementation of AES-128 using SSSE3 instructions. It runs inlloyd2010-08-094-0/+463
| | | | | | | | | | | | | | | constant time and on a Nehalem is significantly faster than the table based version. This implementation technique was invented by Mike Hamburg and described in a paper in CHES 2009 "Accelerating AES with Vector Permute Instructions". This code is basically a translation of his public domain x86-64 assembly code into intrinsics. Todo: Adding support for AES-192 and AES-256; this just requires implementing the key schedules. Currently only tested on an i7 with GCC (32 and 64 bit code); testing/optimization on 32-bit processors with SSSE3 like the Atom, and with Visual C++ and other compilers, are also todos.
* Also allow clang with 32-bit assembly code, everything seems to worklloyd2010-08-088-94/+20
| | | | fine with latest SVN.
* Clang understands at least some GCC inline asm syntax as well as whatlloyd2010-08-083-0/+3
| | | | an .S file is, so allow it for x86-64. Tested/works with Clang SVN.
* Identify a i7-860 as Nehalemlloyd2010-08-081-0/+1
|
* If we can't access cpuid, but we know that we are compiling forlloyd2010-08-081-0/+9
| | | | | | x86-64, then enable SSE2 anyway because we know any x86-64 processor does have SSE2, and the OS has to support it because it's part of the standard ABIs.
* Use clang++ instead of clang for the compiler driver, otherwise linklloyd2010-08-081-1/+1
| | | | errors can result due to not getting the C++ runtime library.
* Clang fixlloyd2010-08-081-0/+1
|
* Fix return value for set_global_state_unless_setlloyd2010-08-081-0/+3
|
* Move the functions that directly manipulate the global state singletonlloyd2010-08-066-67/+165
| | | | | | | | | | | into global_state.{h,cpp}. Move all of the functions into a new namespace Global_State_Management, though exposing global_state() into the Botan namespace for compatability. Also add new functions global_state_exists and set_global_state_unless_set which may be helpful in certain tricky initialization scenarios (eg when an application using botan also uses a library which may or may not itself use botan).
* merge of '28d57385c0f1a9a2665288ce728e8b3231634f59'lloyd2010-08-035-8/+42
|\ | | | | | | and 'a4d88442d5f6b8554234c7f7468856868919b614'
| * Forbid copying an Algorithm_Factory; could easily cause double-delete,lloyd2010-07-301-0/+4
| | | | | | | | | | | | | | | | | | | | especially in a multithreaded environment, and doesn't seem like a useful operation to support. (In principle, we could support this by adding a clone() call to Algorithm_Cache, which would in turn call clone on each of it's held prototype objects, plus adding a clone to Engine. Doesn't seem worth the bother, though.
| * Change the benchmark code to also take a buf_size, instead of using hardcodedlloyd2010-07-302-7/+28
| | | | | | | | | | | | | | | | 16 KiB buffer. Also reorder the parameters to make somewhat more sense, with the first arguments being totally mandatory and the later ones potentially optional. Provide inlined version matching the old interface that just forwards to the new call, marking it as deprecated.
| * If dynamic loading fails, include result of dlerror() in the exception msglloyd2010-07-301-1/+8
| |
| * Add name() function to DataSource_Stream for Filter interfacelloyd2010-07-301-0/+2
| |
* | We've already predeclared Engine at the start of the header, so nolloyd2010-07-291-4/+4
|/ | | | reason to say `class Engine*` later on.
* Restrict dyn_load to platforms where it might theoretically work:lloyd2010-07-281-0/+9
| | | | | | | | Linux, Solaris, and the BSDs. Solaris and BSD are untested, but it seems like they should work. Using libdl on Solaris is seemingly only required in Solaris 9 and earlier, but 10 has a stub library so it should work there as well.
* Remove redundant setting for adding libdl link on Linux in dyn_engine;lloyd2010-07-281-4/+0
| | | | | it relies on dyn_load which should be the sole source for this kind of stuff, since dyn_engine itself does not touch the OS level APIs.
* Add a version info function which returns a u32bit. The currentlylloyd2010-07-281-1/+12
| | | | | expected value is 20100728 (ie, today). This will allow for checking for and/or working around changes to interfaces.
* Expose Algorithm_Factory::clear_caches which clears out all of thelloyd2010-07-272-1/+10
| | | | | caches; this might be useful for applications which are, say, particularly sensitive to memory usage.
* There was an interesting bug affecting dynamically loaded engines.lloyd2010-07-272-4/+14
| | | | | | | | | | | | | | | | | | | | | | | The library initializer runs some self tests; this brings objects for a few select types (AES, SHA-1, etc) into the caches. Later on, when we add a dynamic engine, the engines aren't requeried because the cache has hits. So, for instance an dlopen'ed engine that provided AES-128 would not actually be used unless you called on the algo factory with a provider of "blah" - even using set_preferred_provider would have no effect, because that's just a request. Add a new function to Algorithm_Cache, clear_cache, which just deletes everything that is currently loaded (this is 90% of the destructor). Then call this on each cache in Algorithm_Factory when a new Engine is loaded. In normal use, this should be very fast because on init the engines are loaded one after another so clear_cache() won't do much work at all, but it ensures that if you load an engine later on in runtime it will always be found. It does have the downside that the app will then requery each Engine for each new algo after this point, but I think typically loading a dynamic engine will happen very early on so this won't be too much of a hassle. (And even if it happens in the middle of execution, everything still works, it just means some overhead the first time you ask for algo X).
* In Algorithm_Factory, delete the Engines after deleting the cacheslloyd2010-07-271-2/+2
| | | | | | | | rather than before. Otherwise, we run into a problem with dynamically loaded engines: the engine will be deleted (and thus, the external library unloaded), before calling the destructors on any objects which may have been cached, so we jump to a now invalid address instead of the destructor code.
* Add a new utility class Dynamically_Loaded_Library which wraps aroundlloyd2010-07-277-0/+308
| | | | | | | | | | the system dynamic linker (if any). Currently it only supports dlopen, and is only enabled on Linux. It will almost certainly work on BSDs and Solaris as well, though, and should be easy to extend to support Win32-style dynamic loading. Also add a new engine, Dynamically_Loaded_Engine, which loads up a new Engine object from a shared library/DLL.
* Rename Default_Engine to Core_Engine which describes its purposeslloyd2010-07-2713-38/+36
| | | | (slightly) better.
* merge of '17389a973545d2f8e25813894cdd2da1b01aa534'lloyd2010-07-275-56/+76
|\ | | | | | | and 'ada4c9893d70affd8934ab9664e390087feab3c9'
| * Oops, bad GCC version check herelloyd2010-07-271-1/+1
| |
| * Add support in CPUID for detecting PCMULUDQ and MOVBE instructions.lloyd2010-07-272-5/+19
| | | | | | | | Rename CPUID::has_aes_intel to has_aes_ni.
| * Add support for GCC's byteswap builtins. They were added in 4.3 andlloyd2010-07-271-4/+29
| | | | | | | | | | | | work for 32 and 64 bit integers. Using these instead of inline asm may offer better scheduling on x86, and additionally offers native byteswapping on other platforms (PPC, ARM, etc).
| * Modify CPUID so all the check functions are purely inline. Add a newlloyd2010-07-263-47/+28
| | | | | | | | | | initialize() call which must be called prior to use of any other functions.
* | Add support for Camellia in OpenSSL enginelloyd2010-07-221-0/+6
| |
* | Avoid unused argument warninglloyd2010-07-221-1/+3
| |
* | Use configured compiler for Pythonlloyd2010-07-221-2/+3
|/
* Consolidate the two engines that provided assembler implementationslloyd2010-07-1311-98/+30
| | | | | | (amd64_eng and ia32_eng) into a new asm_engine. This same engine could also be used in the event that asm code for other CPUs was added later on.
* Remove unused UI param to CMS_Decoderlloyd2010-07-092-6/+3
|
* Drop support for running configure with Python 2.4. This allowslloyd2010-07-092-2/+2
| | | | | | | | | removing several workarounds for limitations in optparse in that release, and also allows using the ternary operator added in 2.5. As far as I can tell, the only still active release of any Linux/BSD distro that uses 2.4 is RHEL5. The beta of RHEL6 has 2.6, and it seems likely that RHEL6 will be out before 1.10.0.
* Delete os defaults file - only used by Perllloyd2010-07-091-19/+0
|
* Rename S2K to PBKDF, because that is by far the most common name - S2Klloyd2010-07-0916-118/+108
| | | | | | | | | | | | | really is only used by OpenPGP, and largely it was named S2K here because the OpenPGP S2K was implemented years before the ones in PKCS #5. We have a typedef of PBKDF to S2K, and an inlined get_s2k that calls get_pbkdf for source compatability. There doesn't seem to be any reason to have a forward for the renamed s2k.h header - to actually use a PBKDF, you'd have to either include lookup.h and call get_s2k / get_pbkdf, or else include an algorithm-specific header and use it directly. In either case, including s2k.h is neither necessary nor sufficient.
* Fix constructorlloyd2010-07-091-1/+1
|
* Argh: SecureVector's constructor needs to behave differentlylloyd2010-07-071-2/+14
| | | | | | | | | | | | | | depending on if INITIAL_LEN is non-zero. Normal semantics are the vector will change size based on whatever it is constructed with, but that's bad in cases like SecureVector<byte, 4> val(buffer, 3); which in the past would be a 4 valued thing with 3 elements set and one zero trailing. (This construct showed up in base64 and possibly elsewhere). If INITIAL_LEN is set, use copy instead so the length does not change. C++0x cannot come soon enough.
* These functions are internal use only, and don't need to be exported withlloyd2010-06-303-13/+15
| | | | BOTAN_DLL
* Make round_up and round_down templates instead of fixed to use u32bitslloyd2010-06-294-13/+21
|
* If the Keyed_Filter's set_iv is called (ie, in the case that the1.9.9lloyd2010-06-281-2/+3
| | | | | filter doesn't support IVs at all), throw an exception unless the IV has zero length.
* Give all Filter objects a method for querying their namelloyd2010-06-2816-5/+64
|
* Simplify feature checkslloyd2010-06-281-7/+12
|
* Typo fixlloyd2010-06-281-1/+1
|
* For the SHA-2 classes, don't use inheritence to share a handful oflloyd2010-06-284-85/+106
| | | | | things, just share the compression function via an anon namespace member, and replicate the simple stuff like copy_out.
* Add a new configure option --maintainer-mode which turns on the fulllloyd2010-06-281-2/+3
| | | | | set of warning flags. Use just plain '-Wall -W' for regular GCC so the default build is happy on arbitrarily old versions.
* If we are going to lazily initialize, just create the type and calllloyd2010-06-251-2/+4
| | | | its constructor directly, instead of going through LibraryInitializer.
* Add detection support for upcoming AVXlloyd2010-06-251-1/+8
|
* Avoid name collision in amalgamationlloyd2010-06-221-5/+10
|
generated by cgit v1.2.3 (git 2.39.1) at 2024-12-04 21:38:34 +0000