==3227== Cachegrind, a cache and branch-prediction profiler
==3227== Copyright (C) 2002-2017, and GNU GPL'd, by Nicholas Nethercote et al.
==3227== Using Valgrind-3.19.0 and LibVEX; rerun with -h for copyright info
==3227== Command: _bin/cxx-opt-sh/mycpp-nosouffle/osh -n --ast-format none benchmarks/testdata/t9300-fast-import.sh
==3227== Parent PID: 3220
==3227== 
--3227-- warning: L3 cache found, using its data for the LL simulation.
==3227== 
==3227== I   refs:      13,017,069
==3227== I1  misses:        73,219
==3227== LLi misses:         8,235
==3227== I1  miss rate:       0.56%
==3227== LLi miss rate:       0.06%
==3227== 
==3227== D   refs:       5,884,888  (3,318,720 rd   + 2,566,168 wr)
==3227== D1  misses:        73,572  (   37,670 rd   +    35,902 wr)
==3227== LLd misses:        45,285  (   12,742 rd   +    32,543 wr)
==3227== D1  miss rate:        1.3% (      1.1%     +       1.4%  )
==3227== LLd miss rate:        0.8% (      0.4%     +       1.3%  )
==3227== 
==3227== LL refs:          146,791  (  110,889 rd   +    35,902 wr)
==3227== LL misses:         53,520  (   20,977 rd   +    32,543 wr)
==3227== LL miss rate:         0.3% (      0.1%     +       1.3%  )