==2824== Cachegrind, a cache and branch-prediction profiler
==2824== Copyright (C) 2002-2017, and GNU GPL'd, by Nicholas Nethercote et al.
==2824== Using Valgrind-3.19.0 and LibVEX; rerun with -h for copyright info
==2824== Command: dash -n benchmarks/testdata/ltmain.sh
==2824== Parent PID: 2817
==2824== 
--2824-- warning: L3 cache found, using its data for the LL simulation.
==2824== 
==2824== I   refs:      24,904,718
==2824== I1  misses:         1,690
==2824== LLi misses:         1,608
==2824== I1  miss rate:       0.01%
==2824== LLi miss rate:       0.01%
==2824== 
==2824== D   refs:       9,961,827  (6,830,780 rd   + 3,131,047 wr)
==2824== D1  misses:        14,084  (    5,224 rd   +     8,860 wr)
==2824== LLd misses:         9,969  (    2,751 rd   +     7,218 wr)
==2824== D1  miss rate:        0.1% (      0.1%     +       0.3%  )
==2824== LLd miss rate:        0.1% (      0.0%     +       0.2%  )
==2824== 
==2824== LL refs:           15,774  (    6,914 rd   +     8,860 wr)
==2824== LL misses:         11,577  (    4,359 rd   +     7,218 wr)
==2824== LL miss rate:         0.0% (      0.0%     +       0.2%  )