==2364== Cachegrind, a cache and branch-prediction profiler
==2364== Copyright (C) 2002-2017, and GNU GPL'd, by Nicholas Nethercote et al.
==2364== Using Valgrind-3.14.0 and LibVEX; rerun with -h for copyright info
==2364== Command: _bin/cxx-opt/osh -n --ast-format none benchmarks/testdata/t9300-fast-import.sh
==2364== Parent PID: 2357
==2364== 
--2364-- warning: L3 cache found, using its data for the LL simulation.
==2364== 
==2364== I   refs:      12,806,974
==2364== I1  misses:        36,343
==2364== LLi misses:         7,418
==2364== I1  miss rate:       0.28%
==2364== LLi miss rate:       0.06%
==2364== 
==2364== D   refs:       5,607,219  (3,310,365 rd   + 2,296,854 wr)
==2364== D1  misses:        71,632  (   38,267 rd   +    33,365 wr)
==2364== LLd misses:        43,695  (   13,420 rd   +    30,275 wr)
==2364== D1  miss rate:        1.3% (      1.2%     +       1.5%  )
==2364== LLd miss rate:        0.8% (      0.4%     +       1.3%  )
==2364== 
==2364== LL refs:          107,975  (   74,610 rd   +    33,365 wr)
==2364== LL misses:         51,113  (   20,838 rd   +    30,275 wr)
==2364== LL miss rate:         0.3% (      0.1%     +       1.3%  )