stochastic gecode-solver presolving phase #25

I solved gcc.attribs.handle_noinline_attribute twice. The first time it printed:

[pre] cost: {[19..20]}, optimality gap: {5.26%}

The second time it printed:

[pre] cost: {[18..20]}, optimality gap: {11.11%}

Settings:

export UNISON_BENCHMARK=toplas
export UNISON_BENCHMARK_LIST=functions
export UNISON_TARGET=Hexagon
export UNISON_GOAL=speed