linux-kernel - Re: [PATCH] media: ov5640: Update last busy timestamp to reset autosuspend timer

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:   Sat, 14 Jan 2023 23:49:02 +0200
From:   Sakari Ailus <sakari.ailus@....fi>
To:     Jacopo Mondi <jacopo.mondi@...asonboard.com>
Cc:     Andrey Skvortsov <andrej.skvortzov@...il.com>,
        Steve Longerbeam <slongerbeam@...il.com>,
        Mauro Carvalho Chehab <mchehab@...nel.org>,
        linux-media@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH] media: ov5640: Update last busy timestamp to reset
 autosuspend timer

Hi Jacopo,

On Sat, Jan 14, 2023 at 04:19:06PM +0100, Jacopo Mondi wrote:
> Hi Sakari
> 
> On Sat, Jan 14, 2023 at 02:29:14PM +0200, Sakari Ailus wrote:
> > Hi Andrey,
> >
> > On Sat, Jan 14, 2023 at 02:21:09PM +0300, Andrey Skvortsov wrote:
> > > Otherwise autosuspend delay doesn't work and power is cut off
> > > immediately as device is freed.
> > >
> > > Signed-off-by: Andrey Skvortsov <andrej.skvortzov@...il.com>
> > > ---
> > >  drivers/media/i2c/ov5640.c | 9 ++++++++-
> > >  1 file changed, 8 insertions(+), 1 deletion(-)
> > >
> > > diff --git a/drivers/media/i2c/ov5640.c b/drivers/media/i2c/ov5640.c
> > > index ac35350..f71c0f7 100644
> > > --- a/drivers/media/i2c/ov5640.c
> > > +++ b/drivers/media/i2c/ov5640.c
> > > @@ -1238,6 +1238,7 @@ static int ov5640_write_reg(struct ov5640_dev *sensor, u16 reg, u8 val)
> > >  		return ret;
> > >  	}
> > >
> > > +	pm_runtime_mark_last_busy(&sensor->i2c_client->dev);
> > >  	return 0;
> > >  }
> > >
> > > @@ -1305,6 +1306,7 @@ static int ov5640_read_reg(struct ov5640_dev *sensor, u16 reg, u8 *val)
> > >  	}
> > >
> > >  	*val = buf[0];
> > > +	pm_runtime_mark_last_busy(&sensor->i2c_client->dev);
> >
> > I wouldn't add these calls to register accesses. It's generally relevant
> > with autosuspend_put().
> >
> > The rest seems fine to me.
> >
> 
> Does it mean the same should be done for the all the sensor drivers that
> use autosuspend ? I count 8 of them, not a huge number.

pm_runtime_mark_last_busy() should be called before
pm_runtime_put_autosuspend(). I was thinking of adding a helper that would
simply call both of them. We don't have that yet though.

> 
> > >  	return 0;
> > >  }
> > >
> > > @@ -3615,6 +3617,7 @@ static int ov5640_g_volatile_ctrl(struct v4l2_ctrl *ctrl)
> > >  		break;
> > >  	}
> > >
> > > +	pm_runtime_mark_last_busy(&sensor->i2c_client->dev);
> > >  	pm_runtime_put_autosuspend(&sensor->i2c_client->dev);
> > >
> > >  	return 0;
> > > @@ -3702,6 +3705,7 @@ static int ov5640_s_ctrl(struct v4l2_ctrl *ctrl)
> > >  		break;
> > >  	}
> > >
> > > +	pm_runtime_mark_last_busy(&sensor->i2c_client->dev);
> > >  	pm_runtime_put_autosuspend(&sensor->i2c_client->dev);
> > >
> > >  	return ret;
> > > @@ -4034,8 +4038,10 @@ static int ov5640_s_stream(struct v4l2_subdev *sd, int enable)
> > >  out:
> > >  	mutex_unlock(&sensor->lock);
> > >
> > > -	if (!enable || ret)
> > > +	if (!enable || ret) {
> > > +		pm_runtime_mark_last_busy(&sensor->i2c_client->dev);
> > >  		pm_runtime_put_autosuspend(&sensor->i2c_client->dev);
> > > +	}
> > >
> > >  	return ret;
> > >  }
> > > @@ -4203,6 +4209,7 @@ static int ov5640_probe(struct i2c_client *client)
> > >
> > >  	pm_runtime_set_autosuspend_delay(dev, 1000);
> > >  	pm_runtime_use_autosuspend(dev);
> > > +	pm_runtime_mark_last_busy(dev);
> > >  	pm_runtime_put_autosuspend(dev);
> > >
> > >  	return 0;

-- 
Kind regards,

Sakari Ailus