==3093== Cachegrind, a cache and branch-prediction profiler
==3093== Copyright (C) 2002-2017, and GNU GPL'd, by Nicholas Nethercote et al.
==3093== Using Valgrind-3.19.0 and LibVEX; rerun with -h for copyright info
==3093== Command: _bin/cxx-opt-sh/osh -n --ast-format none benchmarks/testdata/configure
==3093== Parent PID: 3086
==3093== 
--3093-- warning: L3 cache found, using its data for the LL simulation.
==3093== brk segment overflow in thread #1: can't grow to 0x4856000
==3093== (see section Limitations in user manual)
==3093== NOTE: further instances of this message will not be shown
==3093== 
==3093== I   refs:      231,141,148
==3093== I1  misses:      1,133,025
==3093== LLi misses:          8,380
==3093== I1  miss rate:        0.49%
==3093== LLi miss rate:        0.00%
==3093== 
==3093== D   refs:      100,267,034  (57,307,139 rd   + 42,959,895 wr)
==3093== D1  misses:      1,682,764  ( 1,143,221 rd   +    539,543 wr)
==3093== LLd misses:        257,068  (    13,000 rd   +    244,068 wr)
==3093== D1  miss rate:         1.7% (       2.0%     +        1.3%  )
==3093== LLd miss rate:         0.3% (       0.0%     +        0.6%  )
==3093== 
==3093== LL refs:         2,815,789  ( 2,276,246 rd   +    539,543 wr)
==3093== LL misses:         265,448  (    21,380 rd   +    244,068 wr)
==3093== LL miss rate:          0.1% (       0.0%     +        0.6%  )