Claude review: drm: bridge: dw_hdmi: Misc enable/disable, CEC and EDID cleanup

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

Overall Series Review

Subject: drm: bridge: dw_hdmi: Misc enable/disable, CEC and EDID cleanup
Author: Jonas Karlman <jonas@kwiboo.se>
Patches: 19
Reviewed: 2026-05-05T07:58:54.413382

---

This is a well-structured 17-patch cleanup series for the Synopsys DesignWare HDMI bridge driver (`dw-hdmi.c`). The series systematically simplifies the HPD/RXSENSE interrupt handling, moves power on/off into the atomic bridge callbacks, eliminates stale state tracking (`bridge_is_on`, `previous_mode`, `rxsense`, `sink_is_hdmi`, `sink_has_audio`), switches CEC notifier handling to use the generic CEC helpers, and introduces debounced HPD processing via `delayed_work`.

The overall direction is excellent — the driver becomes significantly simpler (net -183 lines), and the changes align with how the DRM bridge infrastructure expects things to work. Each patch is well-scoped and the commit messages are clear and well-written with good rationale.

Most patches carry Reviewed-by tags already. The new v4 patches (12, 13, 14, 15, 16, 17) do not yet have reviews, and these are the ones that deserve the most scrutiny.

I have a few observations and one concern:

**Main concern**: Patch 14 (RXSENSE rework) removes the `force` field from `struct dw_hdmi` and the `disabled` field, but patch 10 (which applies earlier) still writes to `hdmi->force` in `dw_hdmi_connector_force()`. This looks like a build-breaking ordering dependency — patch 10 assigns `hdmi->force` but patch 14 removes the field. The intermediary patches (11, 12, 13) don't appear to remove the `force` usage. This needs to be verified carefully — the mbox ordering doesn't exactly match patch numbering so I may be misreading the incremental diffs, but if the `force` field is accessed after it's removed, that's a compile error.

**Minor observations** are noted per-patch below.

---
Generated by Claude Code Patch Reviewer