# Amdsmi Test Runner

> Run C++ and Python tests for amd-smi. Use when: running tests, verifying test results, checking test coverage, pre-review test validation.

- Skill: `rocm/amdsmi-test-runner` (Agent Skill, multi-file: 2 files)
- Install (CLI): `npx skillmds@latest add rocm/amdsmi-test-runner`
- Raw SKILL.md: https://api.skillmd.com/api/skills/rocm/amdsmi-test-runner/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- Author: rocm (https://skillmd.com/u/rocm)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/rocm/amdsmi-test-runner

---


# Test Runner — amd-smi

How to build and run all test suites for amd-smi. Assumes the `amdsmi-build-install` skill has already been executed (build exists and is installed).

## Prerequisites

- amd-smi built and installed (run `amdsmi-build-install` skill first)
- AMD GPU hardware available for integration tests
- `build/` directory exists with compiled test binaries

## C++ Tests (GTest)

### Build Test Binary

If not already built (the build skill builds with `-DBUILD_TESTS=ON`):

```bash
cd build && make -j$(nproc) amdsmitst
```

### Run Tests

```bash
cd build/tests
source ../../tests/amd_smi_test/amdsmitst.exclude
source ../../tests/amd_smi_test/detect_asic_filter.sh
./amdsmitst --gtest_filter="-${GTEST_EXCLUDE}"
```

### Interpret Results

| Output | Meaning | Severity |
|--------|---------|----------|
| `[  PASSED  ]` | Test passed | — |
| `[  FAILED  ]` | Test failed | ❌ BLOCKING |
| `[  SKIPPED ]` | Test skipped (blacklisted) | — |
| Binary not found | Build didn't produce test binary | ⚠️ IMPORTANT |

## Python Tests

### Unit / Integration / CLI

Run the suite files directly (each `sys.exit(0/1)`):

```bash
cd tests/python
python3 unit_tests.py -v
python3 integration_test.py -v
python3 cli_unit_test.py -v
```

### Import Verification

```bash
# System install
python3 -c "import amdsmi; print(amdsmi.__version__)"

# Verify CLI works
amd-smi version
amd-smi static
```

## Test Contexts

Tests must work in **both** install contexts:

| Context | How to Verify |
|---------|--------------|
| System install (RPM/DEB) | Default after `amdsmi-build-install` |

## Running All Tests to Logs (One-Shot)

`run-all-tests.sh` (this skill's directory) runs the C++ GTest, Python unit,
integration, and CLI suites, tees each to a timestamped log dir, and prints a
pass/fail summary. It derives the project root, so run it from anywhere in the tree:

```bash
.claude/skills/amdsmi-test-runner/run-all-tests.sh          # logs under $TMPDIR/amdsmi-tests-<stamp>/
.claude/skills/amdsmi-test-runner/run-all-tests.sh -o ./ci-logs
```

Each suite's full output is in `<logdir>/<suite>.log`; the exit code is non-zero
if any suite failed.

## Output

On success, report:
- **C++ tests:** X passed, Y failed, Z skipped
- **Python tests:** X passed, Y failed
- **Any new test failures** compared to baseline

On failure:
- **Which suite failed** (C++ GTest, Python unit, CLI)
- **Specific test names** that failed
- **Test output** for failed tests
- Test failures are **❌ BLOCKING** for the review