Claude review: drm/vmwgfx: Add some checks to vmw_cursor_plane_atomic_update

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

Patch Review

**C99 mixed declarations issue.** The patch adds early-return guards for null `new_state` and `old_state`, which is correct defensively, but introduces a variable declaration after a statement, which is problematic for kernel coding style:

```c
	new_state = drm_atomic_get_new_plane_state(state, plane);
	if (!new_state)
		return;
	old_state = drm_atomic_get_old_plane_state(state, plane);
	struct drm_crtc *crtc = new_state->crtc ? new_state->crtc :
		(old_state ? old_state->crtc : NULL);
```

The `struct drm_crtc *crtc` declaration appears after the `if (!new_state) return;` statement. While C99/C11 allows this, mixing declarations and statements like this after control flow is poor style. `crtc` should be declared at the top of the function alongside the other variables, and assigned later.

**Questionable whether `new_state` can actually be NULL.** In the atomic commit path, `drm_atomic_get_new_plane_state` should always return a valid state for a plane that is part of the atomic update. If this function is being called, the plane was part of the commit. If the null dereference in the IGT test is real, it might indicate a deeper issue worth investigating rather than just papering over with null checks. That said, defensive programming here is not unreasonable as a crash fix.

**The `old_state` null handling is good.** Guarding `old_state->crtc` with `old_state ? old_state->crtc : NULL` is the right fix for the originally reported null dereference — on first enable there may be no old state, and the `?:` operator unconditionally dereferences `old_state`.

---
Generated by Claude Code Patch Reviewer