PERF: avoid casting to float in IntegerArray reducing ops

[jorisvandenbossche]

Currently, we cast to float in the IntegerArray reducing ops:

pandas/pandas/core/arrays/integer.py

Lines 673 to 678 in db62039

	# coerce to a nan-aware float if needed
	if mask.any():
	data = self._data.astype("float64")
	data[mask] = self._na_value
	op = getattr(nanops, "nan" + name)

However, the nanops functions can already handle a mask. So with the appropriate fill_value, there should be no need to cast the integers to float (and check for missing values). At least this is the case for skipna=True, the case of skipna=False might need to be handled separately.