[SMT-LIB] wrong answers in QF_NRA benchmarks

[Leonardo de Moura]

Hi,

It seems the benchmarks QF_NRA\zankl\gen-19.smt2  and QF_NRA\zankl\gen-16.smt2 are incorrectly tagged as unsat.
They are very small, and it is easy to manually check that they are actually satisfiable.

For gen-16.smt2, the model is
a = sqrt(2)
b = -sqrt(3)
where sqrt(n) is the square root of n.

For gen-19.smt2, the model is
a = 1/2
b = -sqrt(2)

Cheers,
Leo