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/msm/dsi: Add support for RGB101010 pixel format
Date: Sun, 22 Mar 2026 04:21:56 +1000
Message-ID: <review-patch3-20260319-dsi-rgb101010-support-v3-3-85b99df2d090@pm.me>
In-Reply-To: <20260319-dsi-rgb101010-support-v3-3-85b99df2d090@pm.me>
References: <20260319-dsi-rgb101010-support-v3-0-85b99df2d090@pm.me> <20260319-dsi-rgb101010-support-v3-3-85b99df2d090@pm.me>
X-Mailer: Claude Code Patch Reviewer
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: quoted-printable
MIME-Version: 1.0

Patch Review

**Critical dependency issue:** This patch references `MIPI_DSI_FMT_RGB10101=
0`, `VID_DST_FORMAT_RGB101010`, and `CMD_DST_FORMAT_RGB101010`. The latter =
two are added by the dsi.xml changes in this patch, but `MIPI_DSI_FMT_RGB10=
1010` is not defined anywhere in the current kernel's `include/drm/drm_mipi=
_dsi.h`. This series will not compile without a prerequisite patch adding t=
hat enum value. The cover letter should state this dependency.

**Register definition change:**

```xml
-		<bitfield name=3D"DST_FORMAT" low=3D"4" high=3D"5" type=3D"dsi_vid_dst_f=
ormat"/>
+		<!-- high was 5 before DSI 6G 2.1.0 -->
+		<bitfield name=3D"DST_FORMAT" low=3D"4" high=3D"6" type=3D"dsi_vid_dst_f=
ormat"/>
```

Extending VID_CFG0 DST_FORMAT from 2 bits (high=3D5) to 3 bits (high=3D6) i=
s necessary to encode value 4 (`RGB101010`). The comment noting this change=
d in DSI 6G 2.1.0 is helpful. However, this means on older hardware (< V2.1=
.0), writing value 4 would spill into bit 6 which could be problematic =E2=
=80=94 but the version guard in `dsi_host_attach` prevents `RGB101010` from=
 being used on older hardware, so this is safe.

**Guard logic is correct:**

```c
+	if (dsi->format =3D=3D MIPI_DSI_FMT_RGB101010 &&
+	    !msm_dsi_host_version_ge(msm_host, MSM_DSI_VER_MAJOR_6G,
+				     MSM_DSI_6G_VER_MINOR_V2_1_0)) {
```

This correctly rejects RGB101010 on controllers older than V2.1.0. Note thi=
s also rejects non-6G controllers (major !=3D 0x03) which is appropriate.

**Minor observation:** The `default:` case in both `dsi_get_vid_fmt` and `d=
si_get_cmd_fmt` silently returns `RGB888`/`RGB888` for any unknown format. =
This was already the behavior and is unchanged, so not a concern introduced=
 by this patch.

---
Generated by Claude Code Patch Reviewer