[CI] Fix ROCm CI (#1043)

* [CI] fix ROCm CI

* feat: add a hook to error out on no test runs

.github/workflows/ci.yml

@@ -90,7 +90,7 @@ jobs:
             name: self-hosted-amd
             # Format: [Nightly-]ROCm-<major>.<minor>[.<patch>]. E.g., "ROCm-6.4" or "Nightly-ROCm-7.0".
             # Use "Nightly-" prefix to use torch nightly builds.
-            toolkit: Nightly-ROCm-7.0
+            toolkit: ROCm-6.3
           - tags: [macos-latest]
             name: macos-latest
             toolkit: Metal # or Nightly-Metal
@@ -352,8 +352,6 @@ jobs:
       - name: Run ROCm tests with Python ${{ matrix.python-version }} (${{ matrix.runner.toolkit }})
         id: rocm-tests
         if: contains(matrix.runner.toolkit, 'ROCm')
-        # FIXME: ROCm test incorrectly skips tests
-        continue-on-error: true
         run: |
           cd testing
           PYTEST=(
@@ -362,7 +360,6 @@ jobs:
           )
           "${PYTEST[@]}" --maxfail=3 --numprocesses=4 \
             ./python/amd/test_tilelang_test_amd.py
-          echo "::error::ROCm tests are known to be skipped incorrectly due to ROCm TVM build issues." >&2
 
       # Apple Metal tests
       - name: Run Metal tests with Python ${{ matrix.python-version }} (${{ matrix.runner.toolkit }})

examples/conftest.py

 import os
 import random
+import pytest
 
 os.environ["PYTHONHASHSEED"] = "0"
 
@@ -18,3 +19,26 @@ except ImportError:
     pass
 else:
     np.random.seed(0)
+
+
+def pytest_terminal_summary(terminalreporter, exitstatus, config):
+    """Ensure that at least one test is collected. Error out if all tests are skipped."""
+    known_types = {
+        "failed",
+        "passed",
+        "skipped",
+        "deselected",
+        "xfailed",
+        "xpassed",
+        "warnings",
+        "error",
+    }
+    if (sum(
+            len(terminalreporter.stats.get(k, []))
+            for k in known_types.difference({"skipped", "deselected"})) == 0):
+        terminalreporter.write_sep(
+            "!",
+            (f"Error: No tests were collected. "
+             f"{dict(sorted((k, len(v)) for k, v in terminalreporter.stats.items()))}"),
+        )
+        pytest.exit("No tests were collected.", returncode=5)

testing/conftest.py

 import os
 import random
+import pytest
 
 os.environ["PYTHONHASHSEED"] = "0"
 
@@ -18,3 +19,26 @@ except ImportError:
     pass
 else:
     np.random.seed(0)
+
+
+def pytest_terminal_summary(terminalreporter, exitstatus, config):
+    """Ensure that at least one test is collected. Error out if all tests are skipped."""
+    known_types = {
+        "failed",
+        "passed",
+        "skipped",
+        "deselected",
+        "xfailed",
+        "xpassed",
+        "warnings",
+        "error",
+    }
+    if (sum(
+            len(terminalreporter.stats.get(k, []))
+            for k in known_types.difference({"skipped", "deselected"})) == 0):
+        terminalreporter.write_sep(
+            "!",
+            (f"Error: No tests were collected. "
+             f"{dict(sorted((k, len(v)) for k, v in terminalreporter.stats.items()))}"),
+        )
+        pytest.exit("No tests were collected.", returncode=5)