2 ms·
Aside from this contrived nuclear meltdown example (which could be more efficiently handled using interrupts rather than polling), when would you ever encounter
by kstenerud 14y ago
Aside from this contrived nuclear meltdown example (which could be more efficiently handled using interrupts rather than polling), when would you ever encounter a situation where a few nanoseconds/picoseconds would make enough of a difference that you want to force a pipeline hit on the unlikely code path and a pipeline miss on the likely path?
And really, considering how rare (and implausible) such situations are, why not just put a comment in the code to communicate your intent to maintenance engineers?
Rejiggering the 99.999% case to accomodate the 0.001% case is a major cause of unnecessary complexity.