BUG: isna() does not catch np.NaN when datatype is Float64

[mortnstak]

Pandas version checks

• I have checked that this issue has not already been reported.

• I have confirmed this bug exists on the latest version of pandas.

• I have confirmed this bug exists on the main branch of pandas.

Reproducible Example

import pandas as pd
import numpy as np

data={"x":[1,0],"y":[1,0]}
df=pd.DataFrame(data,dtype="Float64")
df['z']=df['y']/df['x']
df['z'].isna()

Issue Description

The pandas isna() function does not catch NaN values that are of type np.NaN when using the Float64 datatype. The call df['z'].isna() returns a series with following rows.

0    False
1    False
Name: z, dtype: bool

Using the code above, both rows return a false value. Using df['z'].apply(np.isnan) correctly returns false for the first row, and true for the second row.

0    False
1     True
Name: z, dtype: boolean

Expected Behavior

I would expect the pandas isna() function to also classify the np.NaN type as a null or nan value when using the Float64 datatype.
The returned value of df['z'].isna() should be a series with following rows.

0    False
1     True
Name: z, dtype: bool

Installed Versions

INSTALLED VERSIONS

commit : 0691c5c
python : 3.11.9
python-bits : 64
OS : Windows
OS-release : 10
Version : 10.0.22631
machine : AMD64
processor : Intel64 Family 6 Model 141 Stepping 1, GenuineIntel
byteorder : little
LC_ALL : None
LANG : None
LOCALE : Norwegian Bokmål_Norway.1252

pandas : 2.2.3
numpy : 1.26.2
pytz : 2023.3.post1
dateutil : 2.8.2
pip : 24.2
Cython : None
sphinx : None
IPython : 8.18.1
adbc-driver-postgresql: None
adbc-driver-sqlite : None
bs4 : None
blosc : None
bottleneck : None
dataframe-api-compat : None
fastparquet : None
fsspec : None
html5lib : None
hypothesis : None
gcsfs : None
jinja2 : None
lxml.etree : None
matplotlib : None
numba : None
numexpr : None
odfpy : None
openpyxl : None
pandas_gbq : None
psycopg2 : None
pymysql : None
pyarrow : 14.0.1
pyreadstat : None
pytest : None
python-calamine : None
pyxlsb : None
s3fs : None
scipy : None
sqlalchemy : None
tables : None
tabulate : None
xarray : None
xlrd : None
xlsxwriter : None
zstandard : None
tzdata : 2023.3
qtpy : None
pyqt5 : None

[mortnstak]

Also - this is correctly handled if datatype the non-nullable "float64"

[rhshadrach]

cc @jbrockmendel @jorisvandenbossche @mroeschke - there is one test that hits this where we check that NaN in the float values gives False with pd.isna:

pandas/pandas/tests/arrays/floating/test_contains.py

Line 6 in 68d9dca

def test_contains_nan():

But it seems unexpected to me to have float64 and Float64 act differently here. Can't find a discussion on the current behavior, has this been considered before?