==2984== Cachegrind, a cache and branch-prediction profiler
==2984== Copyright (C) 2002-2017, and GNU GPL'd, by Nicholas Nethercote et al.
==2984== Using Valgrind-3.19.0 and LibVEX; rerun with -h for copyright info
==2984== Command: _bin/cxx-opt-sh/osh -n --ast-format none benchmarks/testdata/t9300-fast-import.sh
==2984== Parent PID: 2977
==2984== 
--2984-- warning: L3 cache found, using its data for the LL simulation.
==2984== 
==2984== I   refs:      11,870,620
==2984== I1  misses:        38,018
==2984== LLi misses:         7,801
==2984== I1  miss rate:       0.32%
==2984== LLi miss rate:       0.07%
==2984== 
==2984== D   refs:       5,173,141  (3,106,391 rd   + 2,066,750 wr)
==2984== D1  misses:        63,704  (   29,146 rd   +    34,558 wr)
==2984== LLd misses:        45,266  (   12,769 rd   +    32,497 wr)
==2984== D1  miss rate:        1.2% (      0.9%     +       1.7%  )
==2984== LLd miss rate:        0.9% (      0.4%     +       1.6%  )
==2984== 
==2984== LL refs:          101,722  (   67,164 rd   +    34,558 wr)
==2984== LL misses:         53,067  (   20,570 rd   +    32,497 wr)
==2984== LL miss rate:         0.3% (      0.1%     +       1.6%  )