ci: select PR unit tests by changed component

Map a PR's changed files to the minimal tests/unit_tests paths in the torch
job; fail-safe to the full suite on shared/unknown/CI changes. Other events
and the CLI/E2E/model tests are unchanged. Unit-tested.

Author: WangLingxun

.github/workflows/ci.yaml

@@ -337,7 +337,21 @@ jobs:
           # Note HSA_NO_SCRATCH_RECLAIM=1 must be set to avoid RCCL perf hit (TAS-8N Node), rocm ver:70125424
           export HSA_NO_SCRATCH_RECLAIM=1
           mkdir -p "${GITHUB_WORKSPACE}/test-reports"
-          pytest --maxfail=1 -s ./tests/unit_tests/ \
+          # Component-aware selection on PRs; full suite on push/release/dispatch.
+          # Fail-safe: any failure to compute the diff falls back to the full suite.
+          TARGETS="./tests/unit_tests/"
+          if [[ "${{ github.event_name }}" == "pull_request" ]]; then
+            base_sha="${{ github.event.pull_request.base.sha }}"
+            git fetch --no-tags --depth=200 origin "${{ github.base_ref }}" 2>/dev/null || true
+            changed="$(git diff --name-only "${base_sha}" HEAD 2>/dev/null || true)"
+            if [[ -n "${changed}" ]]; then
+              sel="$(printf '%s\n' "${changed}" | python tools/ci/select_tests.py)"
+              [[ -n "${sel}" ]] && TARGETS="${sel}"
+            fi
+          fi
+          echo "Selected unit-test targets: ${TARGETS}"
+          # shellcheck disable=SC2086 # intentional word-splitting of multiple paths
+          pytest --maxfail=1 -s ${TARGETS} \
             --cov=primus --cov-report=term-missing:skip-covered \
             --junitxml="${GITHUB_WORKSPACE}/test-reports/core-unit.xml" \
             --deselect=tests/unit_tests/megatron/cco/test_tp_overlap.py::TPOverlapTestCase::test_fp8_te_linear \

tests/unit_tests/ci/test_select_tests.py

@@ -0,0 +1,77 @@
+###############################################################################
+# Copyright (c) 2025, Advanced Micro Devices, Inc. All rights reserved.
+#
+# See LICENSE for license information.
+###############################################################################
+
+"""Unit tests for tools/ci/select_tests.py (component-aware PR test selection)."""
+
+import importlib.util
+from pathlib import Path
+
+_ROOT = Path(__file__).resolve().parents[3]
+_SPEC = importlib.util.spec_from_file_location("select_tests", _ROOT / "tools/ci/select_tests.py")
+_MOD = importlib.util.module_from_spec(_SPEC)
+_SPEC.loader.exec_module(_MOD)
+select = _MOD.select_targets
+
+FULL = ["tests/unit_tests/"]
+
+
+def test_empty_runs_full():
+    assert select([]) == FULL
+
+
+def test_ci_change_runs_full():
+    assert select([".github/workflows/ci.yaml"]) == FULL
+
+
+def test_tools_change_runs_full():
+    assert select(["tools/ci/select_tests.py"]) == FULL
+
+
+def test_root_requirements_runs_full():
+    assert select(["requirements.txt"]) == FULL
+    assert select(["requirements-jax.txt"]) == FULL
+
+
+def test_launcher_runs_full():
+    assert select(["primus/core/launcher/initialize.py"]) == FULL
+
+
+def test_unknown_primus_path_runs_full():
+    assert select(["primus/brand_new_area/x.py"]) == FULL
+
+
+def test_megatron_backend_maps_to_both_dirs():
+    out = select(["primus/backends/megatron/training/global_vars.py"])
+    assert "tests/unit_tests/backends/megatron/" in out
+    assert "tests/unit_tests/megatron/" in out
+
+
+def test_projection_takes_precedence_over_core():
+    assert select(["primus/core/projection/engine.py"]) == ["tests/unit_tests/core/projection/"]
+
+
+def test_generic_core_maps_to_core_tests():
+    assert select(["primus/core/trainer/base.py"]) == ["tests/unit_tests/core/"]
+
+
+def test_cli_and_runner_map_to_cli_tests():
+    assert select(["primus/cli/main.py"]) == ["tests/unit_tests/cli/"]
+    assert select(["runner/helpers/foo.sh"]) == ["tests/unit_tests/cli/"]
+    assert select(["primus_cli.py"]) == ["tests/unit_tests/cli/"]
+
+
+def test_changed_test_file_runs_its_own_dir():
+    assert select(["tests/unit_tests/agents/test_tools.py"]) == ["tests/unit_tests/agents/"]
+
+
+def test_multiple_components_are_unioned():
+    out = select(["primus/agents/a.py", "primus/cli/b.py"])
+    assert set(out) == {"tests/unit_tests/agents/", "tests/unit_tests/cli/"}
+
+
+def test_docs_only_change_is_ignored_then_full():
+    # No source/test matched -> fail-safe to full suite.
+    assert select(["README.md", "docs/guide.md"]) == FULL

tools/ci/select_tests.py

@@ -0,0 +1,94 @@
+###############################################################################
+# Copyright (c) 2025, Advanced Micro Devices, Inc. All rights reserved.
+#
+# See LICENSE for license information.
+###############################################################################
+
+"""Map a PR's changed files (stdin, one per line) to the unit-test paths to run.
+
+Fail-safe by design: shared / unknown / CI / packaging changes expand to the
+full suite, so it can only over-select, never under-select. Otherwise prints the
+union of the matched components' test dirs (a changed test file -> its own dir).
+
+    git diff --name-only "$BASE" HEAD | python tools/ci/select_tests.py
+"""
+
+import sys
+
+FULL = "tests/unit_tests/"
+
+# Changed paths matching any of these -> run the whole suite.
+FULL_TRIGGERS = (
+    ".github/",
+    "tools/",
+    "pyproject.toml",
+    "primus/__init__.py",
+    "primus/core/launcher/",
+    "primus/core/utils/",
+    "primus/core/config/",
+    "tests/utils.py",
+    "tests/conftest.py",
+    "tests/unit_tests/conftest.py",
+    "tests/run_unit_tests.py",
+)
+
+# Source-prefix -> test paths. Checked in order; first match wins, so more
+# specific prefixes (projection, megatron) must precede their generic parents.
+COMPONENT_MAP = (
+    ("primus/core/projection/", ("tests/unit_tests/core/projection/",)),
+    ("primus/backends/megatron/", ("tests/unit_tests/backends/megatron/", "tests/unit_tests/megatron/")),
+    ("primus/agents/", ("tests/unit_tests/agents/",)),
+    ("primus/cli/", ("tests/unit_tests/cli/",)),
+    ("primus_cli.py", ("tests/unit_tests/cli/",)),
+    ("runner/", ("tests/unit_tests/cli/",)),
+    ("primus/core/", ("tests/unit_tests/core/",)),
+    ("primus/backends/", ("tests/unit_tests/backends/",)),
+)
+
+
+def _is_full_trigger(path):
+    # Root-level requirements*.txt (dependency change) -> full suite.
+    if "/" not in path and path.startswith("requirements") and path.endswith(".txt"):
+        return True
+    return any(path == trig or path.startswith(trig) for trig in FULL_TRIGGERS)
+
+
+def select_targets(files):
+    files = [f.strip() for f in files if f.strip()]
+    if not files:
+        return [FULL]
+
+    targets = []
+
+    def add(path):
+        if path not in targets:
+            targets.append(path)
+
+    for path in files:
+        if _is_full_trigger(path):
+            return [FULL]
+        # A changed test file: run its own directory.
+        if path.startswith("tests/unit_tests/") and path.endswith(".py"):
+            add(path.rsplit("/", 1)[0] + "/")
+            continue
+        matched = False
+        for prefix, tests in COMPONENT_MAP:
+            if path.startswith(prefix):
+                for test_path in tests:
+                    add(test_path)
+                matched = True
+                break
+        if not matched and path.startswith("primus/"):
+            return [FULL]  # unknown source area -> be safe
+        # else: non-source path (docs, README, ...) -> ignore
+
+    return targets or [FULL]
+
+
+def main():
+    print(" ".join(select_targets(sys.stdin.read().splitlines())))
+    return 0
+
+
+if __name__ == "__main__":
+    raise SystemExit(main())