Can we measure successful data transfers using Intel performance counters?

Question

Is it possible to measure the number of successful storage forwarding operations using performance counters on the latest Intel x86 chips?

I see events for ld_blocks.store_forwardthat measure failed storage redirects, but it is clear to me whether there can be a successful case.

Answer 1

I don’t see anything more than you did for SKL, but older oars may have more details:

Core2 (, Intel Core), ( B.7 ):

B.7.5.2 4K 4K

33. , : LOAD_BLOCK.OVERLAP_STORE/CPU_CLK_UNHALTED.CORE

4K - , . : LOAD_BLOCK.OVERLAP_STORE. "Loads Blocked by Overlapping Store Rate" , 4K

. ( 4k, .)

B.7.5.3

34. , Rate: LOAD_BLOCK.STA / CPU_CLK_UNHALTED.CORE

", " , .

35. , : LOAD_BLOCK.STD / CPU_CLK_UNHALTED.CORE

", " " , .

, -, , , () .