From mboxrd@z Thu Jan 1 00:00:00 1970 From: Claude Code Review Bot To: dri-devel-reviews@example.com Subject: Claude review: drm/bridge: Implement generic USB Type-C DP HPD bridge Date: Mon, 25 May 2026 21:00:52 +1000 Message-ID: In-Reply-To: <20260521032854.103-1-kernel@airkyi.com> References: <20260521032854.103-1-kernel@airkyi.com> X-Mailer: Claude Code Patch Reviewer Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: quoted-printable MIME-Version: 1.0 Overall Series Review Subject: drm/bridge: Implement generic USB Type-C DP HPD bridge Author: Chaoyi Chen Patches: 7 Reviewed: 2026-05-25T21:00:52.011314 --- This series implements a generic USB Type-C DP HPD bridge infrastructure an= d refactors the Rockchip CDN-DP controller to support multiple bridge/PHY p= ort selection without depending on extcon. The design goal is sound: decoup= le DP hotplug detection from specific USB-C controller drivers, and allow t= he CDN-DP controller to work with either of the two combo PHYs on RK3399. The series has already collected Reviewed-by tags from Heikki Krogerus, Nei= l Armstrong, Luca Ceresoli, and Heiko Stuebner. The architecture is reasona= ble, but there are several correctness issues, most notably an **uninitiali= zed variable bug** in patch 5 that could cause a kernel crash, and some log= ic concerns around the multi-bridge HPD/EDID switching path. **Key concerns:** 1. Patch 5: `prev_port` used uninitialized =E2=80=94 potential NULL-pointer= dereference or use of garbage pointer 2. Patch 5: `cdn_dp_switch_port` sets `dp->active =3D true` but is called f= rom EDID read context, which could leave stale state 3. Patch 5: `cdn_dp_connected_port` changes semantics =E2=80=94 the old non= -`next_bridge_valid` path now falls through to the `next_bridge_valid` block 4. Patch 1: No error handling for `drm_dp_hpd_bridge_register` failure in b= us notifier 5. Patch 4: Condition change in `cdn_dp_probe` may inadvertently filter out= valid port configurations --- --- Generated by Claude Code Patch Reviewer