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/rockchip: dw_hdmi_qp: Do not send HPD events for all connectors Date: Tue, 03 Mar 2026 12:42:55 +1000 Message-ID: In-Reply-To: <20260303-dw-hdmi-qp-scramb-v4-4-317d3b8bd219@collabora.com> References: <20260303-dw-hdmi-qp-scramb-v4-0-317d3b8bd219@collabora.com> <20260303-dw-hdmi-qp-scramb-v4-4-317d3b8bd219@collabora.com> X-Mailer: Claude Code Patch Reviewer Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: quoted-printable MIME-Version: 1.0 Patch Review This patch switches from `drm_helper_hpd_irq_event()` (which probes all con= nectors) to `drm_connector_helper_hpd_irq_event()` (which probes only the s= pecific connector). The IRQ registration is also moved after bridge connect= or initialization so `hdmi->connector` is available. **Looks good.** The reordering is correct: 1. `drm_bridge_connector_init` =E2=86=92 sets `hdmi->connector` 2. `drm_connector_attach_encoder` =E2=86=92 attaches it 3. `devm_request_threaded_irq` =E2=86=92 now HPD IRQ can safely access `hdm= i->connector` The rename of `irq` to `hpd_irq` is a nice clarity improvement and avoids s= hadowing. **One observation:** The old code had a `if (drm)` NULL check before callin= g `drm_helper_hpd_irq_event`. The new code unconditionally calls `drm_conne= ctor_helper_hpd_irq_event(hdmi->connector)`. Since the IRQ is only register= ed after the connector is initialized, `hdmi->connector` should never be NU= LL when the work runs, so this is fine. --- Generated by Claude Code Patch Reviewer