Details
-
Enhancement
-
Resolution: Unresolved
-
P4
-
9, 10
-
x86
Description
Originally found here:
http://stackoverflow.com/questions/25078285/replacing-a-32-bit-loop-count-variable-with-64-bit-introduces-crazy-performance
It seems newish Intel x86 processors treat the destination register for popcnt as the input, and wait for it to be ready in order to continue execution.
http://stackoverflow.com/questions/25078285/replacing-a-32-bit-loop-count-variable-with-64-bit-introduces-crazy-performance
It seems newish Intel x86 processors treat the destination register for popcnt as the input, and wait for it to be ready in order to continue execution.