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/msm: Fix GMEM_BASE for A650 Date: Mon, 16 Mar 2026 12:09:20 +1000 Message-ID: In-Reply-To: <20260314-fix-gmem-base-a650-v1-1-3308f60cf74c@pm.me> References: <20260314-fix-gmem-base-a650-v1-1-3308f60cf74c@pm.me> <20260314-fix-gmem-base-a650-v1-1-3308f60cf74c@pm.me> X-Mailer: Claude Code Patch Reviewer Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: 7bit MIME-Version: 1.0 Patch Review **Correctness: Good.** The enum in `adreno_gpu.h:47-50` shows: ``` ADRENO_6XX_GEN1, /* a630 family */ ADRENO_6XX_GEN2, /* a640 family */ ADRENO_6XX_GEN3, /* a650 family */ ADRENO_6XX_GEN4, /* a660 family */ ``` The `adreno_is_a650_family()` helper at `adreno_gpu.h:540-541` explicitly includes both GEN3 and GEN4: ```c return gpu->info->family == ADRENO_6XX_GEN3 || gpu->info->family == ADRENO_6XX_GEN4; ``` So the original intent was to return `GMEM_BASE = 0` for a650 family and above. The regressing commit changed the `>=` threshold to `ADRENO_6XX_GEN4`, excluding GEN3 (a650). Changing it to `>= ADRENO_6XX_GEN3` restores the correct behavior and is also a cleaner expression than the original pair of `adreno_is_a650_family() || adreno_is_a740_family()` checks, since all families from GEN3 onward should use GMEM_BASE of 0. **Commit message:** Clear and well-written. Correctly identifies the regressing commit, the root cause, and the symptom (rendering corruption). Has the appropriate `Fixes:` tag. **No issues found.** This is a clean, minimal one-line regression fix. Reviewed-by worthy. No changes requested. --- Generated by Claude Code Patch Reviewer