String performance regression from python 3.2 to 3.3

[Roy Smith]

In article <HuKdnTBUKc2AcdnMnZ2dnUVZ_vqdnZ2d at westnet.com.au>,
 Neil Hodgson <nhodgson at iinet.net.au> wrote:
 
>      Low-level string manipulation often deals with blocks larger than 
> an individual character for speed. Generally 32 or 64-bits at a time 
> using the CPU or 128 or 256 using the vector unit. Then there may be 
> entry/exit code to handle initial alignment to a block boundary and 
> dealing with a smaller than block-size tail.

Duff's Device!