Re: [PATCH] ACPI: Add GPIO-signaled event emulator.

[Tomasz Nowicki]

On 21.08.2014 16:54, Mika Westerberg wrote:
> On Thu, Aug 21, 2014 at 04:39:46PM +0200, Tomasz Nowicki wrote:
> > > > +{
> > > > +	struct gpio_pin_data *pin_data, *temp;
> > > > +
> > > > +	list_for_each_entry_safe(pin_data, temp, &pin_data_list, list)
> > > > +		kfree(pin_data);
> > >
> > > I suppose you want to first remove the directory entries and then the
> > > pin data. Otherwise if you get pre-empted at this point and someone
> > > tries to use your debugfs files, bad things might happen.
> > Good catch!
> >
> > > > +
> > > > +	debugfs_remove_recursive(acpi_evt_debugfs_dir);
> > >
> > > Since this already removes everything below this dentry, why do you need
> > > to store the pointer in gpio_pin_parent_data?
> > Not sure I got the question.
> >
> > GPIO device instance (debugfs dir) is parent for all its pins (debugfs
> > nodes). I am using gpio_pin_parent_data as container to pass info for all
> > children so I can create debugfs node inside of parent related debugfs dir.
> >
> > pin_data_list, however, is used to keep pointers to allocated memory (one
> > for each pins). debugfs_remove_recursive won't free it.
>
> Yes but it removes the debugfs entries below so you don't need to store
> the children debugfs dentries.
I don't, gpio_pin_parent_data (local variable) keeps pointer to *parent 
debugfs dentries* only. gpio_pin_parent_data name might be misleading, I 
will rename it to gpio_parent_data in next version.

Regards,
Tomasz
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/