Address CBC decrypt timing issues and reenable the AESNI+SHA1 stitch.

Kludge alert. This is arranged by passing padding length in unused
bits of SSL3_RECORD->type, so that orig_len can be reconstructed.

RISCs are picky and alignment granted by compiler for md_state can be
insufficient for SHA512.

Break dependency on uint64_t. It's possible to declare bits as
unsigned int, because TLS packets are limited in size and 32-bit
value can't overflow.

branches it needs to be in a "gap".

We have to use EVP in FIPS mode so we can only partially mitigate
timing differences.

Make an extra call to EVP_DigestSignUpdate to hash additonal blocks
to cover any timing differences caused by removal of padding.

Add additional check to catch this in ASN1_item_verify too.

This change updates the DTLS code to match the constant-time CBC
behaviour in the TLS.

The previous CBC patch was bugged in that there was a path through enc()
in s3_pkt.c/d1_pkt.c which didn't set orig_len. orig_len would be left
at the previous value which could suggest that the packet was a
sufficient length when it wasn't.

This patch makes the decoding of SSLv3 and TLS CBC records constant
time. Without this, a timing side-channel can be used to build a padding
oracle and mount Vaudenay's attack.

This patch also disables the stitched AESNI+SHA mode pending a similar
fix to that code.

In order to be easy to backport, this change is implemented in ssl/,
rather than as a generic AEAD mode. In the future this should be changed
around so that HMAC isn't in ssl/, but crypto/ as FIPS expects.

This change adds CRYPTO_memcmp, which compares two vectors of bytes in
an amount of time that's independent of their contents. It also changes
several MAC compares in the code to use this over the standard memcmp,
which may leak information about the size of a matching prefix.

PR: 2963 and a number of others

Submitted by: Dmitry Belyavsky, Seguei Leontiev
PR: 2821

traditional format uses MD5 which is prohibited in FIPS mode.

Submitted by: Pierre Delaage