Re:Intel i7 loop performance anomaly
by mason55 on Tuesday December 03, 2013 @02:23PM
(#6842466)
He mentioned in the comments that he ensured everything is aligned to 64-bit boundaries
[ Parent
| on HN ]
64 bit might not be the right alignment, and even aligned to 64 bit boundaries you can run into page and cacheline issues.
Don't you always have to pad back with nops (0x90) since the scheduler might be looking ahead for other instructions to execute (or maybe not, since it's seeing jmp back). Just wildly guessing here...