Turn on xfail_strict = true for all python packages

docs/cudf/source/developer_guide/testing.md

@@ -7,6 +7,23 @@ specifically the [`pytest-cov`](https://github.com/pytest-dev/pytest-cov) plugin
 Code coverage reports are uploaded to [Codecov](https://app.codecov.io/gh/rapidsai/cudf).
 Each PR also indicates whether it increases or decreases test coverage.
 
+### Configuring pytest
+
+Pytest will accept configuration in [multiple different
+files](https://docs.pytest.org/en/stable/reference/customize.html),
+with a specified discovery and precedence order. Note in particular
+that there is no automatic "include" mechanism, as soon as a matching
+configuration file is found, discovery stops.
+
+For preference, so that all tool configuration lives in the same
+place, we use `pyproject.toml`-based configuration. Test configuration
+for a given package should live in that package's `pyproject.toml`
+file.
+
+Where tests do not naturally belong to a project, for example the
+`cudf.pandas` integration tests and the cuDF benchmarks, use a
+`pytest.ini` file as close to the tests as possible.
+
 ## Test organization
 
 How tests are organized depends on which of the following two groups they fall into:

python/cudf/pyproject.toml

@@ -124,6 +124,26 @@ skip = [
     "__init__.py",
 ]
 
+[tool.pytest.ini_options]
+addopts = "--tb=native"
+filterwarnings = [
+    "error",
+    "ignore:::.*xdist.*",
+    "ignore:::.*pytest.*",
+    # some third-party dependencies (e.g. 'boto3') still using datetime.datetime.utcnow()
+    "ignore:.*datetime.*utcnow.*scheduled for removal.*:DeprecationWarning:botocore",
+    # Deprecation warning from Pyarrow Table.to_pandas() with pandas-2.2+
+    "ignore:Passing a BlockManager to DataFrame is deprecated:DeprecationWarning",
+    # PerformanceWarning from cupy warming up the JIT cache
+    "ignore:Jitify is performing a one-time only warm-up to populate the persistent cache:cupy._util.PerformanceWarning",
+    # Ignore numba PEP 456 warning specific to arm machines
+    "ignore:FNV hashing is not implemented in Numba.*:UserWarning"
+]
+markers = [
+    "spilling: mark benchmark a good candidate to run with `CUDF_SPILL=ON`"
+]
+xfail_strict = true
+
 [tool.rapids-build-backend]
 build-backend = "scikit_build_core.build"
 dependencies-file = "../../dependencies.yaml"

python/cudf_kafka/pyproject.toml

@@ -79,9 +79,11 @@ skip = [
 ]
 
 [tool.pytest.ini_options]
+addopts = "--tb=native"
 filterwarnings = [
   "error"
 ]
+xfail_strict = true
 
 [tool.scikit-build]
 build-dir = "build/{wheel_tag}"

python/cudf_polars/pyproject.toml

@@ -50,6 +50,10 @@ license-files = ["LICENSE"]
 version = {file = "cudf_polars/VERSION"}
 
 [tool.pytest.ini_options]
+addopts = "--tb=native"
+filterwarnings = [
+  "error"
+]
 xfail_strict = true
 
 [tool.coverage.report]

python/custreamz/custreamz/tests/test_dataframes.py

@@ -377,24 +377,16 @@ def test_setitem_overwrites(stream):
     [
         ({}, "sum"),
         ({}, "mean"),
-        pytest.param({}, "min"),
+        ({}, "min"),
         pytest.param(
             {},
             "median",
             marks=pytest.mark.xfail(reason="Unavailable for rolling objects"),
         ),
-        pytest.param({}, "max"),
-        pytest.param(
-            {},
-            "var",
-            marks=pytest.mark.xfail(reason="Unavailable for rolling objects"),
-        ),
-        pytest.param({}, "count"),
-        pytest.param(
-            {"ddof": 0},
-            "std",
-            marks=pytest.mark.xfail(reason="Unavailable for rolling objects"),
-        ),
+        ({}, "max"),
+        ({}, "var"),
+        ({}, "count"),
+        ({"ddof": 0}, "std"),
         pytest.param(
             {"quantile": 0.5},
             "quantile",

python/custreamz/pyproject.toml

@@ -111,11 +111,14 @@ skip = [
 ]
 
 [tool.pytest.ini_options]
+addopts = "--tb=native"
 filterwarnings = [
     "error",
     "ignore:unclosed <socket.socket:ResourceWarning",
     "ignore:Port .* is already in use.:UserWarning:distributed",
     # Should be fixed in the next streamz release
     # https://github.com/python-streamz/streamz/commit/2812f1f961dfcb3f17e948d8b12a12472975558e
     "ignore:pkg_resources is deprecated as an API:DeprecationWarning:streamz",
+    "ignore:Deprecated call to `pkg_resources.declare_namespace:DeprecationWarning"
 ]
+xfail_strict = true

python/dask_cudf/pyproject.toml

@@ -115,6 +115,7 @@ skip = [
 ]
 
 [tool.pytest.ini_options]
+addopts = "--tb=native"
 filterwarnings = [
     "error::FutureWarning",
     "error::DeprecationWarning",
@@ -125,3 +126,4 @@ filterwarnings = [
     "ignore:Passing a BlockManager to DataFrame is deprecated and will raise in a future version. Use public APIs instead.:DeprecationWarning",
     "ignore:String support for `aggregate_files` is experimental. Behavior may change in the future.:FutureWarning:dask",
 ]
+xfail_strict = true

python/pylibcudf/pyproject.toml

@@ -95,6 +95,15 @@ skip = [
     "__init__.py",
 ]
 
+[tool.pytest.ini_options]
+addopts = "--tb=native"
+filterwarnings = [
+  "error",
+  "ignore:::.*xdist.*",
+  "ignore:::.*pytest.*"
+]
+xfail_strict = true
+
 [tool.rapids-build-backend]
 build-backend = "scikit_build_core.build"
 dependencies-file = "../../dependencies.yaml"