Claude review: kunit: Add backtrace suppression self-tests

[Claude Code Review Bot <claude-review@example.com>]

Patch Review

Good test coverage. Tests exercise:
- Direct WARN with message (`backtrace_suppression_test_warn_direct`)
- Indirect WARN through a helper (`backtrace_suppression_test_warn_indirect`)
- Multiple warnings in one scope (`backtrace_suppression_test_warn_multi`)
- WARN_ON (no format string) (`backtrace_suppression_test_warn_on_direct/indirect`)
- Incremental counting (`backtrace_suppression_test_count`)
- Active state transitions (`backtrace_suppression_test_active_state`)
- Sequential independent scopes via direct API (`backtrace_suppression_test_multi_scope`)
- Cross-kthread isolation (`backtrace_suppression_test_cross_kthread`)

**Cross-kthread test:**

```c
+static void backtrace_suppression_test_cross_kthread(struct kunit *test)
+{
+	struct cross_kthread_data data;
+	struct task_struct *task;
+
+	data.was_active = false;
+	init_completion(&data.done);
+
+	kunit_warning_suppress(test) {
+		task = kthread_run(cross_kthread_fn, &data, "kunit-cross-test");
+		KUNIT_ASSERT_FALSE(test, IS_ERR(task));
+		wait_for_completion(&data.done);
+		kthread_stop(task);
+	}
+
+	KUNIT_EXPECT_FALSE(test, data.was_active);
+}
```

This is well-structured. The `kthread_stop()` is inside the suppress block so the kthread is guaranteed to be stopped before the suppression scope exits. The `while (!kthread_should_stop()) schedule();` loop in the kthread function ensures the kthread stays alive until `kthread_stop()` is called, avoiding use-after-free.

**Minor observation:** The helper functions `trigger_backtrace_warn()` and `trigger_backtrace_warn_on()` are marked `noinline`, which is important — without this, the compiler could inline them into the test function, and on some architectures the WARN path might behave differently with regard to the return address. Good.

**No issues found.**

---

---
Generated by Claude Code Patch Reviewer