Re: [PATCH] Speed up "copy4", "copy64" on little-endian systems.

On 21/11/12 17:46, Andrew Stubbs wrote:
> The copy64 function implements an endian-safe copy routine for
> an array of 16 32-bit integers, but this is sub-optimal on machines
> where the byte-order is already correct. Likewise for copy4.

I should add, I also had a look at "mdfour64", the other function that 
shows up at the top of a profile's hot-functions list, but I found 
nothing obviously inefficient about that implementation.

Given that the algorithm is completely linear (with every line depending 
on the result of the previous), and that everything is completely 
inlined already, I don't see that fiddling with the C code could do 
anything that the compiler's transformations don't already do.

Andrew