Difference between revisions of "Performance Counters"

From dankwiki
 
Line 5: Line 5:
 
* This is the preferred way we get at [[x86 timing|TSC]] (TSC has fallen out of favor, though)
 
* This is the preferred way we get at [[x86 timing|TSC]] (TSC has fallen out of favor, though)
 
* Collectors include [[oprofile]] and [[perf]].
 
* Collectors include [[oprofile]] and [[perf]].
{| border="1"
+
 
 +
{| border="1" class="wikitable"
 
! Counter
 
! Counter
 
! Introduced
 
! Introduced

Latest revision as of 22:14, 22 September 2019

Counter Introduced Description
MACHINE_CLEARS.ASSIST_FUSION Core i7 Number of macro-fusions performed
LSD.UOPS Core i7 Number of μops delivered by the Loop Stream Decoder