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/radeon: fix iMac11,1 dark display Date: Thu, 04 Jun 2026 13:37:51 +1000 Message-ID: In-Reply-To: <20260518211409.4868-3-gilles.risch@gmail.com> References: <20260518211409.4868-3-gilles.risch@gmail.com> X-Mailer: Claude Code Patch Reviewer Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: 7bit MIME-Version: 1.0 Overall Series Review Subject: drm/radeon: fix iMac11,1 dark display Author: Gilles Risch Patches: 21 Reviewed: 2026-06-04T13:37:51.633170 --- This mbox contains the evolution of a fix for the Apple iMac11,1 (Late 2009, Mobility Radeon HD 4850/RV770) dark display under KMS, spanning v1 through v5. The root cause is well-understood: the RV770 is DCE3.1 but the existing ASIC_IS_DCE32() macro starts at CHIP_RV730, which comes *after* CHIP_RV770 in the enum, so the RV770 never matched the DCE3.2 PLL and encoder logic paths. Additionally, Apple routes the internal eDP panel through Link B instead of Link A. The author iterated productively in response to maintainer feedback: v1-v3 used a broad `ASIC_IS_DCE31()` macro that could affect other DCE3.1 boards; v4-v5 narrowed the fix to `dmi_match()` for iMac11,1 specifically. The v5 further improves the encoder picker restructuring, separating the Apple DMI quirk from the DCE32 version logic entirely. **The latest patch (v5) is the one that should be reviewed for merge.** It is a well-scoped, low-risk quirk for a specific Apple machine. A few minor issues remain. The earlier patches (v1 through v4) are superseded and should be disregarded. --- --- Generated by Claude Code Patch Reviewer