linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Jonathan Cameron <Jonathan.Cameron@Huawei.com>
To: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
Cc: Sakari Ailus <sakari.ailus@linux.intel.com>,
	<linuxarm@huawei.com>, <mauro.chehab@huawei.com>,
	Mauro Carvalho Chehab <mchehab@kernel.org>,
	<linux-kernel@vger.kernel.org>, <linux-media@vger.kernel.org>
Subject: Re: [PATCH 25/25] media: i2c: ccs-core: fix pm_runtime_get_sync() usage count
Date: Wed, 5 May 2021 13:35:48 +0100	[thread overview]
Message-ID: <20210505133548.00005c1a@Huawei.com> (raw)
In-Reply-To: <20210505125857.7f30d8fa@coco.lan>

On Wed, 5 May 2021 12:58:57 +0200
Mauro Carvalho Chehab <mchehab+huawei@kernel.org> wrote:

> Em Wed, 5 May 2021 12:57:00 +0200
> Mauro Carvalho Chehab <mchehab+huawei@kernel.org> escreveu:
> 
> > Em Wed, 5 May 2021 13:34:09 +0300
> > Sakari Ailus <sakari.ailus@linux.intel.com> escreveu:
> >   
> > > Hi Mauro,
> > > 
> > > Thanks for the patch.
> > > 
> > > On Wed, May 05, 2021 at 11:42:15AM +0200, Mauro Carvalho Chehab wrote:  
> > > > The pm_runtime_get_sync() internally increments the
> > > > dev->power.usage_count without decrementing it, even on errors.
> > > > 
> > > > There is a bug at ccs_pm_get_init(): when this function returns
> > > > an error, the stream is not started, and RPM usage_count
> > > > should not be incremented. However, if the calls to
> > > > v4l2_ctrl_handler_setup() return errors, it will be kept
> > > > incremented.
> > > > 
> > > > At ccs_suspend() the best is to replace it by the new
> > > > pm_runtime_resume_and_get(), introduced by:
> > > > commit dd8088d5a896 ("PM: runtime: Add pm_runtime_resume_and_get to deal with usage counter")
> > > > in order to properly decrement the usage counter automatically,
> > > > in the case of errors.
> > > > 
> > > > Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>    
> > > 
> > > Could you add Fixes: line and Cc: stable?  
> > 
> > Sure. See the fixes one enclosed.
> >   
> > > The patch that breaks this is 96e3a6b92f23a .
> > > 
> > > It would be better to fix the bug first so the patch to the stable trees
> > > doesn't need special handling.
> > >   
> > > > ---
> > > >  drivers/media/i2c/ccs/ccs-core.c | 39 ++++++++++++++++++++------------
> > > >  1 file changed, 24 insertions(+), 15 deletions(-)
> > > > 
> > > > diff --git a/drivers/media/i2c/ccs/ccs-core.c b/drivers/media/i2c/ccs/ccs-core.c
> > > > index b05f409014b2..04c3ab9e37b4 100644
> > > > --- a/drivers/media/i2c/ccs/ccs-core.c
> > > > +++ b/drivers/media/i2c/ccs/ccs-core.c
> > > > @@ -1880,21 +1880,33 @@ static int ccs_pm_get_init(struct ccs_sensor *sensor)
> > > >  	struct i2c_client *client = v4l2_get_subdevdata(&sensor->src->sd);
> > > >  	int rval;
> > > >  
> > > > +	/*
> > > > +	 * It can't use pm_runtime_resume_and_get() here, as the driver
> > > > +	 * relies at the returned value to detect if the device was already
> > > > +	 * active or not.
> > > > +	 */
> > > >  	rval = pm_runtime_get_sync(&client->dev);
> > > > -	if (rval < 0) {
> > > > -		pm_runtime_put_noidle(&client->dev);
> > > > +	if (rval < 0)
> > > > +		goto error;
> > > >  
> > > > -		return rval;
> > > > -	} else if (!rval) {
> > > > -		rval = v4l2_ctrl_handler_setup(&sensor->pixel_array->
> > > > -					       ctrl_handler);
> > > > -		if (rval)
> > > > -			return rval;
> > > > +	/* Device was already active, so don't set controls */
> > > > +	if (rval == 1)
> > > > +		return 0;
> > > >  
> > > > -		return v4l2_ctrl_handler_setup(&sensor->src->ctrl_handler);
> > > > -	}
> > > > +	/* Restore V4L2 controls to the suspended device */
> > > > +	rval = v4l2_ctrl_handler_setup(&sensor->pixel_array->ctrl_handler);
> > > > +	if (rval)
> > > > +		goto error;
> > > >  
> > > > +	rval = v4l2_ctrl_handler_setup(&sensor->src->ctrl_handler);
> > > > +	if (rval)
> > > > +		goto error;
> > > > +
> > > > +	/* Keep PM runtime usage_count incremented on success */
> > > >  	return 0;
> > > > +error:
> > > > +	pm_runtime_put_noidle(&client->dev);    
> > > 
> > > This needs to be pm_runtime_put() as the device has been successfully.  
> > 
> > Ok.
> >   
> > >   
> > > > +	return rval;
> > > >  }
> > > >  
> > > >  static int ccs_set_stream(struct v4l2_subdev *subdev, int enable)
> > > > @@ -3089,12 +3101,9 @@ static int __maybe_unused ccs_suspend(struct device *dev)
> > > >  	bool streaming = sensor->streaming;
> > > >  	int rval;
> > > >  
> > > > -	rval = pm_runtime_get_sync(dev);
> > > > -	if (rval < 0) {
> > > > -		pm_runtime_put_noidle(dev);
> > > > -
> > > > +	rval = pm_runtime_resume_and_get(dev);
> > > > +	if (rval < 0)
> > > >  		return rval;
> > > > -	}
> > > >  
> > > >  	if (sensor->streaming)
> > > >  		ccs_stop_streaming(sensor);    
> > >   
> > 
> > Thanks,
> > Mauro
> > 
> > ---
> > 
> > [PATCH] media: i2c: ccs-core: fix pm_runtime_get_sync() usage count
> > 
> > The pm_runtime_get_sync() internally increments the
> > dev->power.usage_count without decrementing it, even on errors.
> > 
> > There is a bug at ccs_pm_get_init(): when this function returns
> > an error, the stream is not started, and RPM usage_count
> > should not be incremented. However, if the calls to
> > v4l2_ctrl_handler_setup() return errors, it will be kept
> > incremented.
> > 
> > At ccs_suspend() the best is to replace it by the new
> > pm_runtime_resume_and_get(), introduced by:
> > commit dd8088d5a896 ("PM: runtime: Add pm_runtime_resume_and_get to deal with usage counter")
> > in order to properly decrement the usage counter automatically,
> > in the case of errors.
> > 
> > Fixes: 96e3a6b92f23 ("media: smiapp: Avoid maintaining power state information")
> > Cc: stable@vger.kernel.org
> > Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>

> > 
> > diff --git a/drivers/media/i2c/ccs/ccs-core.c b/drivers/media/i2c/ccs/ccs-core.c
> > index b05f409014b2..5ea471fefa3a 100644
> > --- a/drivers/media/i2c/ccs/ccs-core.c
> > +++ b/drivers/media/i2c/ccs/ccs-core.c
> > @@ -1880,21 +1880,33 @@ static int ccs_pm_get_init(struct ccs_sensor *sensor)
> >  	struct i2c_client *client = v4l2_get_subdevdata(&sensor->src->sd);
> >  	int rval;
> >  
> > +	/*
> > +	 * It can't use pm_runtime_resume_and_get() here, as the driver
> > +	 * relies at the returned value to detect if the device was already
> > +	 * active or not.
> > +	 */
> >  	rval = pm_runtime_get_sync(&client->dev);
> > -	if (rval < 0) {
> > -		pm_runtime_put_noidle(&client->dev);
> > +	if (rval < 0)
> > +		goto error;
> >  
> > -		return rval;
> > -	} else if (!rval) {
> > -		rval = v4l2_ctrl_handler_setup(&sensor->pixel_array->
> > -					       ctrl_handler);
> > -		if (rval)
> > -			return rval;
> > +	/* Device was already active, so don't set controls */
> > +	if (rval == 1)
> > +		return 0;
> >  
> > -		return v4l2_ctrl_handler_setup(&sensor->src->ctrl_handler);
> > -	}
> > +	/* Restore V4L2 controls to the suspended device */  
> 
> In time: I'll fold this at the patch:
> 
> diff --git a/drivers/media/i2c/ccs/ccs-core.c b/drivers/media/i2c/ccs/ccs-core.c
> index 5ea471fefa3a..4a848ac2d2cd 100644
> --- a/drivers/media/i2c/ccs/ccs-core.c
> +++ b/drivers/media/i2c/ccs/ccs-core.c
> @@ -1896 +1896 @@ static int ccs_pm_get_init(struct ccs_sensor *sensor)
> -       /* Restore V4L2 controls to the suspended device */
> +       /* Restore V4L2 controls to the previously suspended device */
> 
> Regards,
> Mauro


  reply	other threads:[~2021-05-05 12:37 UTC|newest]

Thread overview: 73+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-05-05  9:41 [PATCH 00/25] Fix some PM runtime issues at the media subsystem Mauro Carvalho Chehab
2021-05-05  9:41 ` [PATCH 01/25] staging: media: rkvdec: fix pm_runtime_get_sync() usage count Mauro Carvalho Chehab
2021-05-05 12:23   ` Jonathan Cameron
2021-05-05  9:41 ` [PATCH 02/25] staging: media: imx7-mipi-csis: " Mauro Carvalho Chehab
2021-05-05 11:06   ` Jonathan Cameron
2021-05-05 11:17     ` Mauro Carvalho Chehab
2021-05-05 13:56     ` Rui Miguel Silva
2021-05-05 14:23       ` Mauro Carvalho Chehab
2021-05-05  9:41 ` [PATCH 03/25] media: venus: Rework error fail recover logic Mauro Carvalho Chehab
2021-05-05 11:05   ` Jonathan Cameron
2021-05-05  9:41 ` [PATCH 04/25] media: s5p_cec: decrement usage count if disabled Mauro Carvalho Chehab
2021-05-05 12:24   ` Jonathan Cameron
2021-05-05  9:41 ` [PATCH 05/25] media: i2c: ccs-core: return the right error code at suspend Mauro Carvalho Chehab
2021-05-05 12:24   ` Jonathan Cameron
2021-05-05 12:51   ` Sakari Ailus
2021-05-05  9:41 ` [PATCH 06/25] media: i2c: imx334: fix the pm runtime get logic Mauro Carvalho Chehab
2021-05-05 11:10   ` Jonathan Cameron
2021-05-05 11:24     ` Mauro Carvalho Chehab
2021-05-05 12:26       ` Jonathan Cameron
2021-05-05  9:41 ` [PATCH 07/25] media: exynos-gsc: don't resume at remove time Mauro Carvalho Chehab
2021-05-05 12:27   ` Jonathan Cameron
2021-05-05  9:41 ` [PATCH 08/25] media: atmel: properly get pm_runtime Mauro Carvalho Chehab
2021-05-05 12:08   ` Jonathan Cameron
2021-06-10  9:04     ` Eugen.Hristev
2021-06-10  9:38       ` Mauro Carvalho Chehab
2021-06-10 12:00         ` Eugen.Hristev
2021-06-16  8:03           ` Mauro Carvalho Chehab
2021-09-06  8:03           ` Mauro Carvalho Chehab
2021-09-06  8:13             ` Eugen.Hristev
2021-09-13 10:26               ` Eugen.Hristev
2021-05-05  9:41 ` [PATCH 09/25] media: hantro: do a PM resume earlier Mauro Carvalho Chehab
2021-05-05 11:34   ` Jonathan Cameron
2021-05-05 13:22   ` Ezequiel Garcia
2021-05-05 13:46     ` Mauro Carvalho Chehab
2021-05-05 14:01       ` Ezequiel Garcia
2021-05-05 14:15         ` Mauro Carvalho Chehab
2021-05-05  9:42 ` [PATCH 10/25] media: marvel-ccic: fix some issues when getting pm_runtime Mauro Carvalho Chehab
2021-05-05  9:42 ` [PATCH 11/25] media: mdk-mdp: fix pm_runtime_get_sync() usage count Mauro Carvalho Chehab
2021-05-05  9:42 ` [PATCH 12/25] media: rcar_fdp1: simplify error check logic at fdp_open() Mauro Carvalho Chehab
2021-05-05  9:48   ` Sergei Shtylyov
2021-05-05  9:42 ` [PATCH 13/25] media: rcar_fdp1: fix pm_runtime_get_sync() usage count Mauro Carvalho Chehab
2021-05-05 12:31   ` Jonathan Cameron
2021-05-05  9:42 ` [PATCH 14/25] media: renesas-ceu: Properly check for PM errors Mauro Carvalho Chehab
2021-05-05  9:56   ` Jacopo Mondi
2021-05-05  9:42 ` [PATCH 15/25] media: s5p: fix pm_runtime_get_sync() usage count Mauro Carvalho Chehab
2021-05-05 12:31   ` Jonathan Cameron
2021-05-05  9:42 ` [PATCH 16/25] media: am437x: " Mauro Carvalho Chehab
2021-05-05 12:32   ` Jonathan Cameron
2021-05-08 13:01   ` Lad, Prabhakar
2021-05-05  9:42 ` [PATCH 17/25] media: sh_vou: " Mauro Carvalho Chehab
2021-05-05  9:42 ` [PATCH 18/25] media: mtk-vcodec: fix PM runtime get logic Mauro Carvalho Chehab
2021-05-05 12:32   ` Jonathan Cameron
2021-05-05  9:42 ` [PATCH 19/25] media: s5p-jpeg: fix pm_runtime_get_sync() usage count Mauro Carvalho Chehab
2021-05-05 12:33   ` Jonathan Cameron
2021-05-05  9:42 ` [PATCH 20/25] media: sti/delta: use pm_runtime_resume_and_get() Mauro Carvalho Chehab
2021-05-05 12:01   ` Jonathan Cameron
2021-05-05 12:33     ` Jonathan Cameron
2021-05-05  9:42 ` [PATCH 21/25] media: sunxi: fix pm_runtime_get_sync() usage count Mauro Carvalho Chehab
2021-05-05  9:42 ` [PATCH 22/25] media: sti/bdisp: " Mauro Carvalho Chehab
2021-05-05 12:34   ` Jonathan Cameron
2021-05-05  9:42 ` [PATCH 23/25] media: exynos4-is: " Mauro Carvalho Chehab
2021-05-05 12:20   ` Jonathan Cameron
2021-05-05  9:42 ` [PATCH 24/25] media: exynos-gsc: " Mauro Carvalho Chehab
2021-05-05 12:34   ` Jonathan Cameron
2021-05-05  9:42 ` [PATCH 25/25] media: i2c: ccs-core: " Mauro Carvalho Chehab
2021-05-05 10:34   ` Sakari Ailus
2021-05-05 10:57     ` Mauro Carvalho Chehab
2021-05-05 10:58       ` Mauro Carvalho Chehab
2021-05-05 12:35         ` Jonathan Cameron [this message]
2021-05-05 14:06           ` Mauro Carvalho Chehab
2021-05-05 16:36             ` Jonathan Cameron
2021-05-05 11:02       ` Sakari Ailus
2021-05-06 15:11 ` [PATCH 00/25] Fix some PM runtime issues at the media subsystem Mauro Carvalho Chehab

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20210505133548.00005c1a@Huawei.com \
    --to=jonathan.cameron@huawei.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-media@vger.kernel.org \
    --cc=linuxarm@huawei.com \
    --cc=mauro.chehab@huawei.com \
    --cc=mchehab+huawei@kernel.org \
    --cc=mchehab@kernel.org \
    --cc=sakari.ailus@linux.intel.com \
    --subject='Re: [PATCH 25/25] media: i2c: ccs-core: fix pm_runtime_get_sync() usage count' \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).