Claude review: kunit: Add documentation for warning backtrace suppression API

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

Patch Review

The documentation is clear and shows both API forms with examples. 

**Minor nit:** The documentation shows the direct API example checking the warning count *after* `kunit_end_suppress_warning()`:

```rst
+		w = kunit_start_suppress_warning(test);
+		trigger_backtrace();
+		kunit_end_suppress_warning(test, w);
+
+		KUNIT_EXPECT_EQ(test, kunit_suppressed_warning_count(w), 1);
```

This works because the handle memory is managed by `kunit_kzalloc()` and remains valid until test exit — the `end` call only removes the node from the RCU list, it doesn't free the struct. This is an important property that makes the direct API more ergonomic than the scoped form for post-scope count checks. The code comment in patch 1 explains this, and the doc example demonstrates it. Good.

**Also good:** The patch fixes a missing newline at end of file in usage.rst:
```
-	}
\ No newline at end of file
+	}
```

**No issues found.**

---

**Summary:** This is a well-crafted v13 series with solid design, comprehensive tests, and clean code. The RCU + spinlock synchronization is standard kernel practice, the static branch gives zero overhead outside tests, and the `in_task()` guard prevents false suppression from interrupt context. The series has appropriate reviews from KUnit and security maintainers. No blocking issues found from a DRM subsystem perspective; the drm_rect_test changes in patch 3 are clean and correct.

---
Generated by Claude Code Patch Reviewer