Re: [PATCH 3/4] usb: typec: fusb302: Fix debugfs issue

[Heikki Krogerus]

On Mon, Apr 30, 2018 at 06:34:03AM -0700, Guenter Roeck wrote:
> On 04/30/2018 05:41 AM, Heikki Krogerus wrote:
> > Removing the "fusb302" debugfs directory when unloading
> > the driver. That allows the driver to be loaded more then
> > ones.
> > 
> > This fixes an issue where the driver, if unloaded, can't be
> > re-loaded, as the "fusb302" debugfs directory already
> > exists. While here, removing useless condition when creating
> > the debugfs directory.
> > 
> > Fixes: 76f0c53d08b9 ("usb: typec: fusb302: Move out of staging")
> > Signed-off-by: Heikki Krogerus <heikki.krogerus@xxxxxxxxxxxxxxx>
> > ---
> >   drivers/usb/typec/fusb302/fusb302.c | 9 ++++-----
> >   1 file changed, 4 insertions(+), 5 deletions(-)
> > 
> > diff --git a/drivers/usb/typec/fusb302/fusb302.c b/drivers/usb/typec/fusb302/fusb302.c
> > index eba6bb890b17..0e5d0aa052f5 100644
> > --- a/drivers/usb/typec/fusb302/fusb302.c
> > +++ b/drivers/usb/typec/fusb302/fusb302.c
> > @@ -218,11 +218,9 @@ static struct dentry *rootdir;
> >   static int fusb302_debugfs_init(struct fusb302_chip *chip)
> >   {
> >   	mutex_init(&chip->logbuffer_lock);
> > -	if (!rootdir) {
> > -		rootdir = debugfs_create_dir("fusb302", NULL);
> > -		if (!rootdir)
> > -			return -ENOMEM;
> > -	}
> 
> I think the idea here was to permit more than one instance of the driver,
> and have all debugfs file entries under the fusb302 directory. Loading the
> second instance will fail after this patch is applied.

OK, need to come up with something else for this issue then.

> > +	rootdir = debugfs_create_dir("fusb302", NULL);
> > +	if (!rootdir)
> > +		return -ENOMEM;
> >   	chip->dentry = debugfs_create_file(dev_name(chip->dev),
> >   					   S_IFREG | 0444, rootdir,
> > @@ -234,6 +232,7 @@ static int fusb302_debugfs_init(struct fusb302_chip *chip)
> >   static void fusb302_debugfs_exit(struct fusb302_chip *chip)
> >   {
> >   	debugfs_remove(chip->dentry);
> > +	debugfs_remove(rootdir); >   }
> >   #else
> > 

Thanks,

-- 
heikki