Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
MIME-Version: 1.0
References: <1597380295-6297-1-git-send-email-wuxy@bitland.com.cn>
 <1597380295-6297-5-git-send-email-wuxy@bitland.com.cn> <20200831174057.GO31019@paasikivi.fi.intel.com>
In-Reply-To: <20200831174057.GO31019@paasikivi.fi.intel.com>
From:   Tomasz Figa <tfiga@chromium.org>
Date:   Thu, 3 Sep 2020 00:59:20 +0200
Message-ID: <CAAFQd5ARYNVMjScuk5-w_z5Pt6jD=CPkCYG+rM2Znvt9j1Od6g@mail.gmail.com>
Subject: Re: [PATCH V3 3/3] media: i2c: gc5035: Add GC5035 image sensor driver
To:     Sakari Ailus <sakari.ailus@linux.intel.com>
Cc:     Xingyu Wu <wuxy@bitland.com.cn>,
        Mauro Carvalho Chehab <mchehab@kernel.org>,
        Rob Herring <robh+dt@kernel.org>,
        "David S. Miller" <davem@davemloft.net>,
        "shawnx.tu" <shawnx.tu@intel.com>,
        Hans Verkuil <hverkuil-cisco@xs4all.nl>,
        "dave.stevenson" <dave.stevenson@raspberrypi.com>,
        Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>,
        Cao Bing Bu <bingbu.cao@intel.com>,
        Nicolas Boichat <drinkcat@chromium.org>,
        Linux Media Mailing List <linux-media@vger.kernel.org>,
        Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
        linux-devicetree <devicetree@vger.kernel.org>,
        Dongchun Zhu <dongchun.zhu@mediatek.com>,
        Sj Huang <sj.huang@mediatek.com>,
        darfur_liu <darfur_liu@gcoreinc.com>,
        "hao.he7" <hao.he7@gmail.com>,
        =?UTF-8?B?5L2V5rWpQjAzMjA1?= <hao.he@bitland.com.cn>
Content-Type: text/plain; charset="UTF-8"
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

Hi Sakari,

On Mon, Aug 31, 2020 at 7:41 PM Sakari Ailus
<sakari.ailus@linux.intel.com> wrote:
>
> Hi Xingyu,
>
> Thanks for the update. I've got a few more comments below.
>
> Do you happen to have some insight on what the OTP data contains and what
> does the driver do with it?
>
> At least in principle the OTP data may be programmed for the customer so
> the same sensor could contain something else what the driver expects to
> find there.
>

Thanks for the review. For anything without my reply, assume fixed. :)

As far as I can see, the data is being read from an area that is
supposed to be reserved for Galaxycore, so I'd assume it doesn't
depend on the customer.

[snip]
> > diff --git a/drivers/media/i2c/Kconfig b/drivers/media/i2c/Kconfig
> > index da11036..aeaf594 100644
> > --- a/drivers/media/i2c/Kconfig
> > +++ b/drivers/media/i2c/Kconfig
> > @@ -712,6 +712,18 @@ config VIDEO_APTINA_PLL
> >  config VIDEO_SMIAPP_PLL
> >       tristate
> >
> > +config VIDEO_GC5035
> > +     tristate "Galaxycore GC5035 sensor support"
> > +     depends on I2C && VIDEO_V4L2
> > +     select MEDIA_CONTROLLER
> > +     select VIDEO_V4L2_SUBDEV_API
>
> Add:
>
>         V4L2_FWNODE
>         OF

This driver doesn't depend on OF. It uses the firmware-independent
property access API. (v4 I sent actually uses device_property_*()).

[snip]
> > +static int __gc5035_power_on(struct gc5035 *gc5035)
> > +{
> > +     struct device *dev = &gc5035->client->dev;
> > +     int i, ret;
> > +
> > +     ret = clk_prepare_enable(gc5035->xvclk);
> > +     if (ret < 0) {
> > +             dev_err(dev, "Failed to enable xvclk\n");
> > +             return ret;
> > +     }
> > +
> > +     gpiod_set_value_cansleep(gc5035->reset_gpio, 1);
> > +
> > +     for (i = 0; i < GC5035_NUM_SUPPLIES; i++) {
> > +             ret = regulator_enable(gc5035->supplies[i].consumer);
> > +             if (ret) {
> > +                     dev_err(dev, "Failed to enable %s: %d\n",
> > +                             gc5035->supplies[i].supply, ret);
> > +                     goto disable_reg_clk;
>
> Please use regulator_bulk_enable() here, and regulator_bulk_disable()
> below.
>

This actually needs to have one of the regulators (iovdd) enabled
before the other ones, but regulator_bulk_enable() is async. In v4 I
used regulator_enable() for iovdd and regulator_bulk_enable() for the
other two for optimal sequencing.

Best regards,
Tomasz