[RFC] Clean up SSE variable shifts

[Richard Henderson]

On 08/23/2010 10:07 AM, Ulrich Drepper wrote:
> As a forth variant try this patch with the following change.
> Considering the reduced memory footprint (one cache line instead of
> four) and the improved capabilities for unaligned SSE memory
> operations in modern CPUs it might be even faster.

Nice.  I wish I'd thought of that in the first place.


r~