All of lore.kernel.org
 help / color / mirror / Atom feed
From: Guenter Roeck <linux-0h96xk9xTtrk1uMJSBkQmQ@public.gmane.org>
To: Wolfram Sang <wsa-z923LK4zBo2bacvFa/9K2g@public.gmane.org>
Cc: linux-i2c-u79uwXL29TY76Z2rM5mHXA@public.gmane.org,
	linux-kernel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
Subject: Re: [PATCH] i2c: i2c-dev: Create 'name' attribute automatically
Date: Wed, 25 Sep 2013 14:11:05 -0700	[thread overview]
Message-ID: <20130925211104.GA2784@roeck-us.net> (raw)
In-Reply-To: <20130925202035.GB4290@katana>

On Wed, Sep 25, 2013 at 10:20:35PM +0200, Wolfram Sang wrote:
> On Sat, Aug 31, 2013 at 09:30:19AM -0700, Guenter Roeck wrote:
> > The 'name' attribute is needed for all i2c-dev class devices, meaning
> > it can be created automatically by pointing to it in the class data
> > structure. This simplifies the code and reduces the probability for race
> > conditions.
> 
> What race condition?
> 

Userspace trying to access the name attribute before it is created.
See http://kroah.com/log/blog/2013/06/26/how-to-create-a-sysfs-file-correctly/
for details; Greg describes the problem and solution much better than I could
ever do.

> > 
> > Signed-off-by: Guenter Roeck <linux-0h96xk9xTtrk1uMJSBkQmQ@public.gmane.org>
> > ---
> >  drivers/i2c/i2c-dev.c |   32 ++++++++++++++++----------------
> >  1 file changed, 16 insertions(+), 16 deletions(-)
> > 
> > diff --git a/drivers/i2c/i2c-dev.c b/drivers/i2c/i2c-dev.c
> > index c3ccdea..46eea02 100644
> > --- a/drivers/i2c/i2c-dev.c
> > +++ b/drivers/i2c/i2c-dev.c
> > @@ -111,7 +111,11 @@ static ssize_t show_adapter_name(struct device *dev,
> >  		return -ENODEV;
> >  	return sprintf(buf, "%s\n", i2c_dev->adap->name);
> >  }
> > -static DEVICE_ATTR(name, S_IRUGO, show_adapter_name, NULL);
> > +
> > +static struct device_attribute i2c_dev_attributes[] = {
> > +	__ATTR(name, S_IRUGO, show_adapter_name, NULL),
> > +	{ }
> > +};
> >  
> >  /* ------------------------------------------------------------------------- */
> >  
> > @@ -538,7 +542,11 @@ static const struct file_operations i2cdev_fops = {
> >  
> >  /* ------------------------------------------------------------------------- */
> >  
> > -static struct class *i2c_dev_class;
> > +static struct class i2c_dev_class = {
> > +	.owner = THIS_MODULE,
> > +	.name = "i2c-dev",
> > +	.dev_attrs = i2c_dev_attributes,
> > +};
> 
> Have you tried this with two instances? I don't think it will work since
> class_register modifies the class struct.
> 
Half a dozen instances, actually. Unless I am missing something, seems to be
unlikely that there is a problem, since the kernel would fail all over the place
if there was a problem with class->dev_attrs (or its replacement dev_groups);

Anyway, the patch is obsolete; one is supposed to use class->dev_groups now.

Thanks,
Guenter

WARNING: multiple messages have this Message-ID (diff)
From: Guenter Roeck <linux@roeck-us.net>
To: Wolfram Sang <wsa@the-dreams.de>
Cc: linux-i2c@vger.kernel.org, linux-kernel@vger.kernel.org
Subject: Re: [PATCH] i2c: i2c-dev: Create 'name' attribute automatically
Date: Wed, 25 Sep 2013 14:11:05 -0700	[thread overview]
Message-ID: <20130925211104.GA2784@roeck-us.net> (raw)
In-Reply-To: <20130925202035.GB4290@katana>

On Wed, Sep 25, 2013 at 10:20:35PM +0200, Wolfram Sang wrote:
> On Sat, Aug 31, 2013 at 09:30:19AM -0700, Guenter Roeck wrote:
> > The 'name' attribute is needed for all i2c-dev class devices, meaning
> > it can be created automatically by pointing to it in the class data
> > structure. This simplifies the code and reduces the probability for race
> > conditions.
> 
> What race condition?
> 

Userspace trying to access the name attribute before it is created.
See http://kroah.com/log/blog/2013/06/26/how-to-create-a-sysfs-file-correctly/
for details; Greg describes the problem and solution much better than I could
ever do.

> > 
> > Signed-off-by: Guenter Roeck <linux@roeck-us.net>
> > ---
> >  drivers/i2c/i2c-dev.c |   32 ++++++++++++++++----------------
> >  1 file changed, 16 insertions(+), 16 deletions(-)
> > 
> > diff --git a/drivers/i2c/i2c-dev.c b/drivers/i2c/i2c-dev.c
> > index c3ccdea..46eea02 100644
> > --- a/drivers/i2c/i2c-dev.c
> > +++ b/drivers/i2c/i2c-dev.c
> > @@ -111,7 +111,11 @@ static ssize_t show_adapter_name(struct device *dev,
> >  		return -ENODEV;
> >  	return sprintf(buf, "%s\n", i2c_dev->adap->name);
> >  }
> > -static DEVICE_ATTR(name, S_IRUGO, show_adapter_name, NULL);
> > +
> > +static struct device_attribute i2c_dev_attributes[] = {
> > +	__ATTR(name, S_IRUGO, show_adapter_name, NULL),
> > +	{ }
> > +};
> >  
> >  /* ------------------------------------------------------------------------- */
> >  
> > @@ -538,7 +542,11 @@ static const struct file_operations i2cdev_fops = {
> >  
> >  /* ------------------------------------------------------------------------- */
> >  
> > -static struct class *i2c_dev_class;
> > +static struct class i2c_dev_class = {
> > +	.owner = THIS_MODULE,
> > +	.name = "i2c-dev",
> > +	.dev_attrs = i2c_dev_attributes,
> > +};
> 
> Have you tried this with two instances? I don't think it will work since
> class_register modifies the class struct.
> 
Half a dozen instances, actually. Unless I am missing something, seems to be
unlikely that there is a problem, since the kernel would fail all over the place
if there was a problem with class->dev_attrs (or its replacement dev_groups);

Anyway, the patch is obsolete; one is supposed to use class->dev_groups now.

Thanks,
Guenter

  reply	other threads:[~2013-09-25 21:11 UTC|newest]

Thread overview: 9+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2013-08-31 16:30 [PATCH] i2c: i2c-dev: Create 'name' attribute automatically Guenter Roeck
2013-08-31 16:30 ` Guenter Roeck
     [not found] ` <1377966619-20197-1-git-send-email-linux-0h96xk9xTtrk1uMJSBkQmQ@public.gmane.org>
2013-09-25 20:20   ` Wolfram Sang
2013-09-25 20:20     ` Wolfram Sang
2013-09-25 21:11     ` Guenter Roeck [this message]
2013-09-25 21:11       ` Guenter Roeck
2013-09-26  8:28       ` Wolfram Sang
2013-09-26 10:37         ` Guenter Roeck
2013-09-26 10:37           ` Guenter Roeck

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=20130925211104.GA2784@roeck-us.net \
    --to=linux-0h96xk9xttrk1umjsbkqmq@public.gmane.org \
    --cc=linux-i2c-u79uwXL29TY76Z2rM5mHXA@public.gmane.org \
    --cc=linux-kernel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org \
    --cc=wsa-z923LK4zBo2bacvFa/9K2g@public.gmane.org \
    /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
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.