Received-SPF: pass (google.com: domain of linux-kernel+bounces-172522-linux.lists.archive=gmail.com@vger.kernel.org designates 2604:1380:45d1:ec00::1 as permitted sender) client-ip=2604:1380:45d1:ec00::1;
Precedence: bulk
MIME-Version: 1.0
References: <20240507135234.1356855-1-yangcong5@huaqin.corp-partner.google.com>
 <20240507135234.1356855-3-yangcong5@huaqin.corp-partner.google.com>
In-Reply-To: <20240507135234.1356855-3-yangcong5@huaqin.corp-partner.google.com>
From: Doug Anderson <dianders@chromium.org>
Date: Tue, 7 May 2024 16:35:20 -0700
Message-ID: <CAD=FV=Wj5WKcVbNGHQ_BbZa_fsVJkpYb2C8TE8bjhvJvx+N_hw@mail.gmail.com>
Subject: Re: [PATCH v4 2/7] drm/panel: himax-hx83102: Break out as separate driver
To: Cong Yang <yangcong5@huaqin.corp-partner.google.com>
Cc: sam@ravnborg.org, neil.armstrong@linaro.org, daniel@ffwll.ch, 
	linus.walleij@linaro.org, krzysztof.kozlowski+dt@linaro.org, 
	robh+dt@kernel.org, conor+dt@kernel.org, airlied@gmail.com, 
	dri-devel@lists.freedesktop.org, devicetree@vger.kernel.org, 
	linux-kernel@vger.kernel.org, xuxinxiong@huaqin.corp-partner.google.com
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

Hi,

On Tue, May 7, 2024 at 6:53=E2=80=AFAM Cong Yang
<yangcong5@huaqin.corp-partner.google.com> wrote:
>
> +static int hx83102_enable_extended_cmds(struct hx83102 *ctx, bool enable=
)
> +{
> +       struct mipi_dsi_multi_context dsi_ctx =3D { .dsi =3D ctx->dsi };
> +
> +       if (enable)
> +               mipi_dsi_dcs_write_seq_multi(&dsi_ctx, HX83102_SETEXTC, 0=
x83, 0x10, 0x21, 0x55, 0x00);
> +       else
> +               mipi_dsi_dcs_write_seq_multi(&dsi_ctx, HX83102_SETEXTC, 0=
x00, 0x00, 0x00);
> +
> +       return 0;

You're throwing away the error codes returned by the
mipi_dsi_dcs_write_seq_multi(), which you shouldn't do. You have two
options:

Option #1: return dsi_ctx.accum_err here and then check the return
value in callers.

Option #2: instead of having this function take "struct hx83102 *ctx",
just have it take "struct mipi_dsi_multi_context *dsi_ctx". Then it
can return void and everything will be fine.

I'd prefer option #2 but either is OK w/ me.


> +static int starry_himax83102_j02_init(struct hx83102 *ctx)
> +{
> +       struct mipi_dsi_multi_context dsi_ctx =3D { .dsi =3D ctx->dsi };
> +
> +       hx83102_enable_extended_cmds(ctx, true);
> +       mipi_dsi_dcs_write_seq_multi(&dsi_ctx, HX83102_SETPOWER, 0x2c, 0x=
b5, 0xb5, 0x31, 0xf1,
> +                                        0x31, 0xd7, 0x2f, 0x36, 0x36, 0x=
36, 0x36, 0x1a, 0x8b, 0x11,
> +                                        0x65, 0x00, 0x88, 0xfa, 0xff, 0x=
ff, 0x8f, 0xff, 0x08, 0x74,
> +                                        0x33);

The indentation is still off here. You have 5 tabs followed by a
space. To make things line up with the opening brace I think it should
be 4 tabs followed by 5 spaces.


> +static int hx83102_enable(struct drm_panel *panel)
> +{
> +       struct hx83102 *ctx =3D panel_to_hx83102(panel);
> +       struct mipi_dsi_device *dsi =3D ctx->dsi;
> +       struct device *dev =3D &dsi->dev;
> +       int ret;
> +
> +       ret =3D ctx->desc->init(ctx);
> +       if (ret)
> +               return ret;

You're still changing behavior here. In the old boe-tv101wum-nl6
driver the init() function was invoked at the end of prepare(). Now
you've got it at the beginning of enable(). If this change is
important it should be in a separate commit and explained.


> +       ret =3D mipi_dsi_dcs_exit_sleep_mode(dsi);
> +       if (ret) {
> +               dev_err(dev, "Failed to exit sleep mode: %d\n", ret);
> +               return ret;
> +       }
> +
> +       msleep(120);
> +
> +       ret =3D mipi_dsi_dcs_set_display_on(dsi);
> +       if (ret) {
> +               dev_err(dev, "Failed to turn on the display: %d\n", ret);
> +       }

The old boe-tv101wum-nl6 driver didn't call
mipi_dsi_dcs_exit_sleep_mode() nor mipi_dsi_dcs_set_display_on() in
its enable routine, did it? If this change is important please put it
in a separate change and justify it.


-Doug