From mboxrd@z Thu Jan  1 00:00:00 1970
From: Claude Code Review Bot <claude-review@example.com>
To: dri-devel-reviews@example.com
Subject: Claude review: drm/rockchip: analogix_dp: Add support for RK3576
Date: Tue, 26 May 2026 07:12:14 +1000
Message-ID: <review-patch10-20260525125331.140059-11-damon.ding@rock-chips.com>
In-Reply-To: <20260525125331.140059-11-damon.ding@rock-chips.com>
References: <20260525125331.140059-1-damon.ding@rock-chips.com> <20260525125331.140059-11-damon.ding@rock-chips.com>
X-Mailer: Claude Code Patch Reviewer
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: 7bit
MIME-Version: 1.0

Patch Review

Adds chip data and OF match table entry:

```c
+static const struct rockchip_dp_chip_data rk3576_edp[] = {
+	{
+		.chip_type = RK3576_EDP,
+		.reg = 0x27dc0000,
+	},
+	{ /* sentinel */ }
+};
```

Worth noting: the RK3576 chip data intentionally omits `.edp_mode` and `.lcdc_sel` GRF fields (unlike RK3588 which sets `.edp_mode`). This is safe -- `rockchip_grf_field_write()` at line 101 of the driver checks `field->valid` (which is `false` when zero-initialized) and returns 0 immediately:

```c
static int rockchip_grf_field_write(struct regmap *grf,
				    const struct rockchip_grf_reg_field *field,
				    u32 val)
{
	if (!field->valid)
		return 0;
```

So the `edp_mode` GRF writes in `rockchip_dp_poweron()`/`rockchip_dp_powerdown()` become no-ops for RK3576. This is presumably correct if RK3576 doesn't need GRF-based eDP mode switching, but a brief comment in the chip data or commit message noting this intentional omission would help future readers understand why.

No blocking issues. Series looks ready to merge.

---
Generated by Claude Code Patch Reviewer