==2525== Cachegrind, a cache and branch-prediction profiler
==2525== Copyright (C) 2002-2017, and GNU GPL'd, by Nicholas Nethercote et al.
==2525== Using Valgrind-3.14.0 and LibVEX; rerun with -h for copyright info
==2525== Command: _bin/cxx-opt/osh -n --ast-format none benchmarks/testdata/t4014-format-patch.sh
==2525== Parent PID: 2518
==2525== 
--2525-- warning: L3 cache found, using its data for the LL simulation.
==2525== 
==2525== I   refs:      11,915,705
==2525== I1  misses:        43,190
==2525== LLi misses:         8,520
==2525== I1  miss rate:       0.36%
==2525== LLi miss rate:       0.07%
==2525== 
==2525== D   refs:       5,263,569  (2,965,258 rd   + 2,298,311 wr)
==2525== D1  misses:        57,897  (   30,217 rd   +    27,680 wr)
==2525== LLd misses:        38,275  (   13,356 rd   +    24,919 wr)
==2525== D1  miss rate:        1.1% (      1.0%     +       1.2%  )
==2525== LLd miss rate:        0.7% (      0.5%     +       1.1%  )
==2525== 
==2525== LL refs:          101,087  (   73,407 rd   +    27,680 wr)
==2525== LL misses:         46,795  (   21,876 rd   +    24,919 wr)
==2525== LL miss rate:         0.3% (      0.1%     +       1.1%  )