==2980== Cachegrind, a cache and branch-prediction profiler
==2980== Copyright (C) 2002-2017, and GNU GPL'd, by Nicholas Nethercote et al.
==2980== Using Valgrind-3.19.0 and LibVEX; rerun with -h for copyright info
==2980== Command: _bin/cxx-opt-sh/osh -n --ast-format none benchmarks/testdata/t9300-fast-import.sh
==2980== Parent PID: 2973
==2980== 
--2980-- warning: L3 cache found, using its data for the LL simulation.
==2980== 
==2980== I   refs:      12,851,444
==2980== I1  misses:        69,057
==2980== LLi misses:         7,973
==2980== I1  miss rate:       0.54%
==2980== LLi miss rate:       0.06%
==2980== 
==2980== D   refs:       5,827,859  (3,289,530 rd   + 2,538,329 wr)
==2980== D1  misses:        72,373  (   37,260 rd   +    35,113 wr)
==2980== LLd misses:        44,373  (   12,575 rd   +    31,798 wr)
==2980== D1  miss rate:        1.2% (      1.1%     +       1.4%  )
==2980== LLd miss rate:        0.8% (      0.4%     +       1.3%  )
==2980== 
==2980== LL refs:          141,430  (  106,317 rd   +    35,113 wr)
==2980== LL misses:         52,346  (   20,548 rd   +    31,798 wr)
==2980== LL miss rate:         0.3% (      0.1%     +       1.3%  )