==2584== Cachegrind, a cache and branch-prediction profiler
==2584== Copyright (C) 2002-2017, and GNU GPL'd, by Nicholas Nethercote et al.
==2584== Using Valgrind-3.19.0 and LibVEX; rerun with -h for copyright info
==2584== Command: bash -n benchmarks/testdata/ltmain.sh
==2584== Parent PID: 2577
==2584== 
--2584-- warning: L3 cache found, using its data for the LL simulation.
==2584== 
==2584== I   refs:      80,310,064
==2584== I1  misses:        27,427
==2584== LLi misses:         3,153
==2584== I1  miss rate:       0.03%
==2584== LLi miss rate:       0.00%
==2584== 
==2584== D   refs:      27,266,511  (19,288,674 rd   + 7,977,837 wr)
==2584== D1  misses:        76,614  (    51,629 rd   +    24,985 wr)
==2584== LLd misses:        22,405  (     4,205 rd   +    18,200 wr)
==2584== D1  miss rate:        0.3% (       0.3%     +       0.3%  )
==2584== LLd miss rate:        0.1% (       0.0%     +       0.2%  )
==2584== 
==2584== LL refs:          104,041  (    79,056 rd   +    24,985 wr)
==2584== LL misses:         25,558  (     7,358 rd   +    18,200 wr)
==2584== LL miss rate:         0.0% (       0.0%     +       0.2%  )