diff --git a/.ci/test.sh b/.ci/test.sh
index 516581ff0..b0e65abc4 100755
--- a/.ci/test.sh
+++ b/.ci/test.sh
@@ -2,7 +2,7 @@
 
 set -e
 
-python -m pytest -v -x -W always --cov PIL --cov Tests --cov-report term Tests
+python -m pytest -v -x -ra -W always --cov PIL --cov Tests --cov-report term Tests
 
 # Docs
 if [ "$TRAVIS_PYTHON_VERSION" == "3.8" ] && [ "$TRAVIS_CPU_ARCH" == "amd64" ]; then
diff --git a/.github/workflows/test-windows.yml b/.github/workflows/test-windows.yml
index 7ae26b883..1989bc00c 100644
--- a/.github/workflows/test-windows.yml
+++ b/.github/workflows/test-windows.yml
@@ -113,7 +113,7 @@ jobs:
     - name: Test Pillow
       run: |
         path %GITHUB_WORKSPACE%\\winbuild\\build\\bin;%PATH%
-        python.exe -m pytest -vx -W always --cov PIL --cov Tests --cov-report term --cov-report xml Tests
+        python.exe -m pytest -vxra -W always --cov PIL --cov Tests --cov-report term --cov-report xml Tests
       shell: cmd
 
     - name: Prepare to upload errors
diff --git a/Tests/test_image_resample.py b/Tests/test_image_resample.py
index 764a3ca49..35eae128b 100644
--- a/Tests/test_image_resample.py
+++ b/Tests/test_image_resample.py
@@ -218,7 +218,7 @@ class TestImagingCoreResampleAccuracy:
         assert_image_equal(im, ref)
 
 
-class CoreResampleConsistencyTest:
+class TestCoreResampleConsistency:
     def make_case(self, mode, fill):
         im = Image.new(mode, (512, 9), fill)
         return im.resize((9, 512), Image.LANCZOS), im.load()[0, 0]
@@ -253,7 +253,7 @@ class CoreResampleConsistencyTest:
         self.run_case(self.make_case("F", 1.192093e-07))
 
 
-class CoreResampleAlphaCorrectTest:
+class TestCoreResampleAlphaCorrect:
     def make_levels_case(self, mode):
         i = Image.new(mode, (256, 16))
         px = i.load()
@@ -274,7 +274,7 @@ class CoreResampleAlphaCorrectTest:
                 len(used_colors), y
             )
 
-    @pytest.mark.skip("Current implementation isn't precise enough")
+    @pytest.mark.xfail(reason="Current implementation isn't precise enough")
     def test_levels_rgba(self):
         case = self.make_levels_case("RGBA")
         self.run_levels_case(case.resize((512, 32), Image.BOX))
@@ -283,7 +283,7 @@ class CoreResampleAlphaCorrectTest:
         self.run_levels_case(case.resize((512, 32), Image.BICUBIC))
         self.run_levels_case(case.resize((512, 32), Image.LANCZOS))
 
-    @pytest.mark.skip("Current implementation isn't precise enough")
+    @pytest.mark.xfail(reason="Current implementation isn't precise enough")
     def test_levels_la(self):
         case = self.make_levels_case("LA")
         self.run_levels_case(case.resize((512, 32), Image.BOX))
@@ -329,7 +329,7 @@ class CoreResampleAlphaCorrectTest:
         self.run_dirty_case(case.resize((20, 20), Image.LANCZOS), (255,))
 
 
-class CoreResamplePassesTest:
+class TestCoreResamplePasses:
     @contextmanager
     def count(self, diff):
         count = Image.core.get_stats()["new_count"]
@@ -372,7 +372,7 @@ class CoreResamplePassesTest:
         assert_image_similar(with_box, cropped, 0.1)
 
 
-class CoreResampleCoefficientsTest:
+class TestCoreResampleCoefficients:
     def test_reduce(self):
         test_color = 254
 
@@ -401,7 +401,7 @@ class CoreResampleCoefficientsTest:
         assert histogram[0x100 * 3 + 0xFF] == 0x10000
 
 
-class CoreResampleBoxTest:
+class TestCoreResampleBox:
     def test_wrong_arguments(self):
         im = hopper()
         for resample in (
diff --git a/Tests/test_imagefont.py b/Tests/test_imagefont.py
index bd79f08e3..65d749b14 100644
--- a/Tests/test_imagefont.py
+++ b/Tests/test_imagefont.py
@@ -443,7 +443,7 @@ class TestImageFont:
         with pytest.raises(UnicodeEncodeError):
             font.getsize("’")
 
-    @pytest.mark.skipif(is_pypy(), reason="failing on PyPy")
+    @pytest.mark.xfail(is_pypy(), reason="failing on PyPy with Raqm")
     def test_unicode_extended(self):
         # issue #3777
         text = "A\u278A\U0001F12B"