BUG: read_hdf() doesn't handle datetime64[ms] properly

[a-ma72]

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


df = pd.DataFrame(
    {
        "dates": [
            pd.to_datetime("2025-05-21 18:44:22"),
            pd.to_datetime("2025-05-21 19:12:42"),
        ],
        "tags": [
            12,
            45,
        ]
    },
)
df["dates"] = df["dates"].astype("datetime64[ms]")
print(df.dtypes)
print(df)

df.to_hdf("dates.h5", key="dates")
df2 = pd.read_hdf("dates.h5", key="dates")
print(df2)

df2["corrected"] = df2["dates"].astype("i8").astype("datetime64[ms]")
print(df2)

Issue Description

Dataframes containing dtype of "datetime64[ms]" seem to be correctly written in hdf format, but the readback is misinterpreted as “datetime64[ns]”.

The output of the code above is:

tags              int64
dtype: object
                dates  tags
0 2025-05-21 18:44:22    12
1 2025-05-21 19:12:42    45
                       dates  tags
0 1970-01-01 00:29:07.853062    12
1 1970-01-01 00:29:07.854762    45
                       dates  tags           corrected
0 1970-01-01 00:29:07.853062    12 2025-05-21 18:44:22
1 1970-01-01 00:29:07.854762    45 2025-05-21 19:12:42```

### Expected Behavior

Correct dates when read back.

### Installed Versions

```INSTALLED VERSIONS
------------------
commit                : d9cdd2ee5a58015ef6f4d15c7226110c9aab8140
python                : 3.11.9.final.0
python-bits           : 64
OS                    : Windows
OS-release            : 10
Version               : 10.0.22631
machine               : AMD64
processor             : Intel64 Family 6 Model 183 Stepping 1, GenuineIntel
byteorder             : little
LC_ALL                : None
LANG                  : en_US.UTF-8
LOCALE                : de_DE.cp1252

pandas                : 2.2.2
numpy                 : 1.26.4
pytz                  : 2024.1
dateutil              : 2.9.0.post0
setuptools            : 80.8.0
pip                   : 25.1.1
Cython                : 3.1.1
pytest                : 8.3.5
hypothesis            : 6.131.20
sphinx                : 8.2.3
blosc                 : None
feather               : None
xlsxwriter            : 3.2.3
lxml.etree            : 5.4.0
html5lib              : 1.1
pymysql               : None
psycopg2              : None
jinja2                : 3.1.6
IPython               : 8.36.0
pandas_datareader     : None
adbc-driver-postgresql: None
adbc-driver-sqlite    : None
bs4                   : 4.13.4
bottleneck            : 1.5.0
dataframe-api-compat  : None
fastparquet           : None
fsspec                : 2025.5.0
gcsfs                 : None
matplotlib            : 3.8.4
numba                 : 0.61.2
numexpr               : 2.10.2
odfpy                 : None
openpyxl              : 3.1.2
pandas_gbq            : None
pyarrow               : 17.0.0
pyreadstat            : None
python-calamine       : None
pyxlsb                : None
s3fs                  : None
scipy                 : 1.15.3
sqlalchemy            : 2.0.30
tables                : 3.10.2
tabulate              : None
xarray                : 2025.4.0
xlrd                  : 2.0.1
zstandard             : 0.23.0
tzdata                : 2025.2
qtpy                  : 2.4.3
pyqt5                 : None```