==3024== Cachegrind, a cache and branch-prediction profiler
==3024== Copyright (C) 2002-2017, and GNU GPL'd, by Nicholas Nethercote et al.
==3024== Using Valgrind-3.19.0 and LibVEX; rerun with -h for copyright info
==3024== Command: _bin/cxx-opt-sh/osh -n --ast-format none benchmarks/testdata/ltmain.sh
==3024== Parent PID: 3017
==3024== 
--3024-- warning: L3 cache found, using its data for the LL simulation.
==3024== brk segment overflow in thread #1: can't grow to 0x4840000
==3024== (see section Limitations in user manual)
==3024== NOTE: further instances of this message will not be shown
==3024== 
==3024== I   refs:      96,275,881
==3024== I1  misses:       549,549
==3024== LLi misses:         8,362
==3024== I1  miss rate:       0.57%
==3024== LLi miss rate:       0.01%
==3024== 
==3024== D   refs:      42,973,835  (24,566,600 rd   + 18,407,235 wr)
==3024== D1  misses:       649,255  (   369,508 rd   +    279,747 wr)
==3024== LLd misses:       165,984  (    12,812 rd   +    153,172 wr)
==3024== D1  miss rate:        1.5% (       1.5%     +        1.5%  )
==3024== LLd miss rate:        0.4% (       0.1%     +        0.8%  )
==3024== 
==3024== LL refs:        1,198,804  (   919,057 rd   +    279,747 wr)
==3024== LL misses:        174,346  (    21,174 rd   +    153,172 wr)
==3024== LL miss rate:         0.1% (       0.0%     +        0.8%  )