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/adreno: use qcom_ubwc_version_tag() helper Date: Mon, 09 Mar 2026 08:32:15 +1000 Message-ID: In-Reply-To: <20260306-ubwc-rework-v1-5-9cfdff12f2bb@oss.qualcomm.com> References: <20260306-ubwc-rework-v1-0-9cfdff12f2bb@oss.qualcomm.com> <20260306-ubwc-rework-v1-5-9cfdff12f2bb@oss.qualcomm.com> X-Mailer: Claude Code Patch Reviewer Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: 7bit MIME-Version: 1.0 Patch Review Replaces hardcoded `mode` values in `a8xx_set_ubwc_config` with `qcom_ubwc_version_tag(cfg)` and removes the inline version ladder from `msm_mdss_setup_ubwc_v6`. **Potential bug in MDSS**: The MDSS change removes the `ver` variable and the `writel_relaxed(ver, ...)` to `REG_MDSS_UBWC_CTRL_2`: ```c - writel_relaxed(ver, msm_mdss->mmio + REG_MDSS_UBWC_CTRL_2); writel_relaxed(prediction_mode, msm_mdss->mmio + REG_MDSS_UBWC_PREDICTION_MODE); ``` This looks like it drops the UBWC_CTRL_2 register write entirely. This register was being written in both the old `msm_mdss_setup_ubwc_dec_40` and `msm_mdss_setup_ubwc_dec_50` paths. If this is intentional, the commit message should explain why. If not, this is a regression -- the line should be: ```c writel_relaxed(qcom_ubwc_version_tag(data), msm_mdss->mmio + REG_MDSS_UBWC_CTRL_2); ``` --- Generated by Claude Code Patch Reviewer