These are typically seen in floating point processing that depend on possible bit patterns.
./src/sun_math/Double_precision/e_exp.c ./src/sun_math/Double_precision/k_rem_pio2.c