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/panel/boe-tv101wum-nl6: set MIPI_DSI_MODE_LPM after sending panel disable cmds
Date: Thu, 23 Apr 2026 08:22:30 +1000
Message-ID: <review-patch1-20260421153147.4378-2-brady.norander@mainlining.org>
In-Reply-To: <20260421153147.4378-2-brady.norander@mainlining.org>
References: <20260421153147.4378-2-brady.norander@mainlining.org> <20260421153147.4378-2-brady.norander@mainlining.org>
X-Mailer: Claude Code Patch Reviewer
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: quoted-printable
MIME-Version: 1.0

Patch Review

**Correctness: Good.** The change is straightforward and correct:

```c
boe->dsi->mode_flags &=3D ~MIPI_DSI_MODE_LPM;   // line 1322: clear LPM (sw=
itch to HS)

mipi_dsi_dcs_set_display_off_multi(&ctx);        // line 1324: send disable=
 in HS
mipi_dsi_dcs_enter_sleep_mode_multi(&ctx);        // line 1325: send sleep =
in HS

mipi_dsi_msleep(&ctx, 150);                      // line 1327: wait for pan=
el

boe->dsi->mode_flags |=3D MIPI_DSI_MODE_LPM;      // line 1329: restore LPM=
  <-- NEW
```

The flag is restored with `|=3D` rather than `=3D`, which is correct =E2=80=
=94 it only sets the `MIPI_DSI_MODE_LPM` bit without disturbing other flags=
. Since only `MIPI_DSI_MODE_LPM` was cleared earlier (line 1322), this perf=
ectly reverses the earlier modification.

**Pattern consistency:** The clear-before/restore-after pattern is used in =
other panel drivers (e.g., `panel-asus-z00t-tm5p5-n35596.c` lines 174/180).=
 Interestingly, some other panel drivers like `panel-ebbg-ft8719.c` and `pa=
nel-boe-td4320.c` do NOT restore LPM after their off sequence, so this is n=
ot universally consistent =E2=80=94 but restoring it is clearly the safer a=
pproach and matches the initial state set at probe time (`dsi->mode_flags =
=3D desc->mode_flags` at line 1758, where all descriptors include `MIPI_DSI=
_MODE_LPM`).

**Placement:** Restoring LPM after the `mipi_dsi_msleep()` is the right cho=
ice =E2=80=94 all DSI commands have been fully sent before the flag is chan=
ged back.

**Impact scope:** This driver covers multiple panel variants (BOE TV101WUM-=
NL6, TV105WUM-NW0, AUO KD101N80-45NA, etc.). The `boe_panel_disable()` func=
tion is shared across all of them, so this fix applies to all. The submitte=
r tested on krane sku176; other devices should also benefit since the asymm=
etry was a latent bug.

**Nits:**
- **Missing `Fixes:` tag.** This would be very helpful for stable tree back=
port consideration. It should reference either the Mediatek HS mode commit =
that exposed this, or the original commit that introduced the `mode_flags &=
=3D ~MIPI_DSI_MODE_LPM` without a corresponding restore.
- **Commit subject line** is a bit long (76 chars after the prefix). Minor =
=E2=80=94 well within tolerance.

---
Generated by Claude Code Patch Reviewer