==2630== Cachegrind, a cache and branch-prediction profiler
==2630== Copyright (C) 2002-2017, and GNU GPL'd, by Nicholas Nethercote et al.
==2630== Using Valgrind-3.19.0 and LibVEX; rerun with -h for copyright info
==2630== Command: bash -n benchmarks/testdata/t9300-fast-import.sh
==2630== Parent PID: 2623
==2630== 
--2630-- warning: L3 cache found, using its data for the LL simulation.
==2630== 
==2630== I   refs:      16,659,503
==2630== I1  misses:         5,755
==2630== LLi misses:         2,979
==2630== I1  miss rate:       0.03%
==2630== LLi miss rate:       0.02%
==2630== 
==2630== D   refs:       5,500,234  (3,914,613 rd   + 1,585,621 wr)
==2630== D1  misses:        10,728  (    7,247 rd   +     3,481 wr)
==2630== LLd misses:         7,301  (    4,210 rd   +     3,091 wr)
==2630== D1  miss rate:        0.2% (      0.2%     +       0.2%  )
==2630== LLd miss rate:        0.1% (      0.1%     +       0.2%  )
==2630== 
==2630== LL refs:           16,483  (   13,002 rd   +     3,481 wr)
==2630== LL misses:         10,280  (    7,189 rd   +     3,091 wr)
==2630== LL miss rate:         0.0% (      0.0%     +       0.2%  )