Write functions to handle loading and saving words a block at a time, taking into

account endian differences.

The current code does not take advantage of the knowledge of which endianness
we are running on; an optimization suggested by Yves Jerschow is to use (unsafe)
casts to speed up the load/store operations. This turns out to provide large
performance increases (30% or more) in some cases.

Even without the unsafe casts, this version seems to average a few percent
faster, probably because the longer loading loops have been partially or
fully unrolled.

This also makes the code implementing low-level algorithms like ciphers and
hashes a bit more succint.

1 files changed, 9 insertions, 13 deletions

diff --git a/src/tea.cpp b/src/tea.cpp
index 9b04aba11..aa04b1df8 100644
--- a/src/tea.cpp
+++ b/src/tea.cpp
@@ -13,8 +13,8 @@ namespace Botan {
 *************************************************/
 void TEA::enc(const byte in[], byte out[]) const
    {
-   u32bit left  = make_u32bit(in[0], in[1], in[2], in[3]),
-          right = make_u32bit(in[4], in[5], in[6], in[7]);
+   u32bit left = load_be<u32bit>(in, 0), right = load_be<u32bit>(in, 1);
+
    u32bit sum = 0;
    for(u32bit j = 0; j != 32; ++j)
       {
@@ -22,10 +22,8 @@ void TEA::enc(const byte in[], byte out[]) const
       left  += ((right << 4) + K[0]) ^ (right + sum) ^ ((right >> 5) + K[1]);
       right += ((left  << 4) + K[2]) ^ (left  + sum) ^ ((left  >> 5) + K[3]);
       }
-   out[0] = get_byte(0, left);  out[1] = get_byte(1, left);
-   out[2] = get_byte(2, left);  out[3] = get_byte(3, left);
-   out[4] = get_byte(0, right); out[5] = get_byte(1, right);
-   out[6] = get_byte(2, right); out[7] = get_byte(3, right);
+
+   store_be(out, left, right);
    }
 
 /*************************************************
@@ -33,8 +31,8 @@ void TEA::enc(const byte in[], byte out[]) const
 *************************************************/
 void TEA::dec(const byte in[], byte out[]) const
    {
-   u32bit left  = make_u32bit(in[0], in[1], in[2], in[3]),
-          right = make_u32bit(in[4], in[5], in[6], in[7]);
+   u32bit left = load_be<u32bit>(in, 0), right = load_be<u32bit>(in, 1);
+
    u32bit sum = 0xC6EF3720;
    for(u32bit j = 0; j != 32; ++j)
       {
@@ -42,10 +40,8 @@ void TEA::dec(const byte in[], byte out[]) const
       left  -= ((right << 4) + K[0]) ^ (right + sum) ^ ((right >> 5) + K[1]);
       sum   -= 0x9E3779B9;
       }
-   out[0] = get_byte(0, left);  out[1] = get_byte(1, left);
-   out[2] = get_byte(2, left);  out[3] = get_byte(3, left);
-   out[4] = get_byte(0, right); out[5] = get_byte(1, right);
-   out[6] = get_byte(2, right); out[7] = get_byte(3, right);
+
+   store_be(out, left, right);
    }
 
 /*************************************************
@@ -54,7 +50,7 @@ void TEA::dec(const byte in[], byte out[]) const
 void TEA::key(const byte key[], u32bit)
    {
    for(u32bit j = 0; j != 4; ++j)
-      K[j] = make_u32bit(key[4*j], key[4*j+1], key[4*j+2], key[4*j+3]);
+      K[j] = load_be<u32bit>(key, j);
    }
 
 }