==3262== Cachegrind, a cache and branch-prediction profiler
==3262== Copyright (C) 2002-2017, and GNU GPL'd, by Nicholas Nethercote et al.
==3262== Using Valgrind-3.14.0 and LibVEX; rerun with -h for copyright info
==3262== Command: mksh -n benchmarks/testdata/Build.sh
==3262== Parent PID: 3257
==3262== 
--3262-- warning: L3 cache found, using its data for the LL simulation.
==3262== 
==3262== I   refs:      16,088,814
==3262== I1  misses:        28,067
==3262== LLi misses:         2,028
==3262== I1  miss rate:       0.17%
==3262== LLi miss rate:       0.01%
==3262== 
==3262== D   refs:       5,923,060  (3,738,817 rd   + 2,184,243 wr)
==3262== D1  misses:        28,263  (   19,842 rd   +     8,421 wr)
==3262== LLd misses:         6,919  (    2,921 rd   +     3,998 wr)
==3262== D1  miss rate:        0.5% (      0.5%     +       0.4%  )
==3262== LLd miss rate:        0.1% (      0.1%     +       0.2%  )
==3262== 
==3262== LL refs:           56,330  (   47,909 rd   +     8,421 wr)
==3262== LL misses:          8,947  (    4,949 rd   +     3,998 wr)
==3262== LL miss rate:         0.0% (      0.0%     +       0.2%  )