Claude review: drm/mipi-dbi: Provide callbacks for atomic interfaces

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

Patch Review

This is the core infrastructure patch. The new helper functions and macros are well-designed.

Minor issues:

1. **Documentation typos** in `drm_mipi_dbi_connector_helper_get_modes()`:
   - `"Sets the connecto rmodes"` → should be `"Sets the connector modes"`
   - `"get_mods callback"` → should be `"get_modes callback"`
   - The cross-reference to `drm_gem_destroy_shadow_plane_state()` seems irrelevant to a connector get_modes function — this looks like a copy-paste error.

2. **Redundant return path** in `drm_mipi_dbi_plane_helper_atomic_check()`:
   ```c
   if (ret)
       return ret;
   else if (!new_plane_state->visible)
       return 0;
   
   return 0;
   ```
   The `else if` branch and the final `return 0` are equivalent. This could simply be `return ret;` after the check call, but it's harmless and may be intentional to leave a hook point for future additions.

3. **Documentation typo** in `drm_mipi_dbi_crtc_helper_atomic_check()`:
   `"ensures that the primary plane as been set up"` → should be `"has been set up"`.

4. The `drm_mipi_dbi_plane_helper_atomic_update()` function drops the `WARN_ON(!fb)` that existed in the original `mipi_dbi_pipe_update()`. The new code silently returns on `!fb`, which is fine since the atomic framework guarantees `fb` is set when the plane is visible, but worth noting.

5. The `mipi_dbi_pipe_disable()` wrapper now calls `drm_mipi_dbi_crtc_helper_atomic_disable(&pipe->crtc, pipe->crtc.state->state)` — passing `pipe->crtc.state->state` is correct here since the simple-pipe disable callback is called from within the atomic commit, so state is valid. However, the new `drm_mipi_dbi_crtc_helper_atomic_disable()` doesn't actually use the `state` parameter, so this is fine regardless.

---
Generated by Claude Code Patch Reviewer