| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
|
| |
Uninitialized variables, missing divide by zero checks, missing
virtual destructor, etc. Only thing serious is bug in TLS maximum
fragment decoder; missing breaks in switch statement meant receiver
would treat any negotiated max frament as 4k limit.
|
| |
|
| |
|
|
|
|
|
|
| |
Fix two memory leaks (in TLS and modes) caused by calling get_foo and
then cloning the result before saving it (leaking the original object),
a holdover from the conversion between construction techniques in 1.11.14
|
|
|
|
|
|
|
|
|
|
|
|
| |
ctypes Python wrapper that uses it. The API is intentionally designed
to have a very simple ABI (extern "C", all structs are opaque, no
memory ownership passing the FFI boundary, limited set of simple types
as args) so the ctypes wrapper is quite simple.
Currently ffi provides ciphers, hashes, MACs, RNGs, PBKDF, KDF,
bcrypt, and most public key operations.
Remove the old boost.python wrapper and all the build code for it.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Convert all uses of Algorithm_Factory and the engines to using Algo_Registry
The shared pool of entropy sources remains but is moved to EntropySource.
With that and few remaining initializations (default OIDs and aliases)
moved elsewhere, the global state is empty and init and shutdown are no-ops.
Remove almost all of the headers and code for handling the global
state, except LibraryInitializer which remains as a compatability stub.
Update seeding for blinding so only one hacky almost-global RNG
instance needs to be setup instead of across all pubkey uses (it uses
either the system RNG or an AutoSeeded_RNG if the system RNG is not
available).
|
|
|
|
| |
Remove global PRNG.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Transforms and BlockCiphers. Registration for all types is done at
startup but is very cheap as just a std::function and a std::map entry
are created, no actual objects are created until needed. This is a
huge improvement over Algorithm_Factory which used T::clone() as the
function and thus kept a prototype object of each type in memory.
Replace existing lookup mechanisms for ciphers, AEADs, and compression
to use the transform lookup. The existing Engine framework remains in
place for BlockCipher, but the engines now just call to the registry
instead of having hardcoded lookups.
s/Transformation/Transform/ with typedefs for compatability.
Remove lib/selftest code (for runtime selftesting): not the right approach.
|
| |
|
| |
|
| |
|
|
|
|
|
| |
and enable them in the default build, though still not enabled in the
runtime policy.
|
| |
|
| |
|
|
|
|
|
| |
Update license header line to specify the terms and refer to the file,
neither of which it included before.
|
| |
|
|
|
|
|
| |
random instead of all-zero padding. Check on sanity of received pads
to the extent possible. Bugzilla 269.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Github issue 27.
Refactor server hello handling to make it easier to handle other extensions.
The manual specified that 224 bit NIST primes were disabled by default
for TLS but they were not. Additionaly disable the 256k1 curve and
reorder the remaining curves by size.
Rewrite the max fragment length extension code to roughly what an
ideal compiler would have turned the original code into, using a
switch instead of a lookup into a small constant std::map.
|
| |
|
| |
|
|
|
|
|
|
| |
in places other than sqlite3, though sqlite3 remains the only
implementation. The interface is currently limited to precisely the
functionality the TLS session manager needs and will likely expand.
|
| |
|
|
|
|
|
| |
with the setting set in the constructor. This prevents various surprising
things from happening to applications and simplifies record processing.
|
|
|
|
| |
of overloads in the base class with the same name.
|
| |
|
| |
|
|
|
|
| |
the nonce sizes should be down to the ciphersuite generating script.
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
|
|
|
| |
offer. Previously the client only checked a couple of special cases.
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
|
|
|
| |
Way back when, following IEEE 1363 naming seemed like a good idea. But not so much.
|
|
|
|
|
|
|
|
|
|
|
| |
Add `website` target to makefile.
Some progress towards fixing minimized builds.
TLS now hard requires ECDSA and GCM since otherwise a minimized build
has only insecure options.
Remove boost_thread dependency in command line tool
|
|
|