Low hanging fruit for optimizing loops

[Walter Bright]

On 6/7/2013 9:15 PM, Juan Manuel Cabo wrote:
> Given the recent surge in interest for performance, I dusted
> off a small test that I made long ago and determined myself
> to find the cause of the performance difference.

It's great that you're doing this. You can track it down further by using inline 
assembler and trying different instruction sequences.

Also, obj2asm gives nicer disassembly :-)