On Wed, May 30, 2018 at 07:31:13PM +0200, Alexandre Belloni wrote:
> Hi,
>
> On 29/05/2018 17:31:02+0200, Greg Kroah-Hartman wrote:
> > When calling debugfs functions, there is no need to ever check the
> > return value. The function can work or not, but the code logic should
> > never do something different based on this.
> >
> > There is also no need to keep the file dentries around at all, so remove
> > those variables from the device structure.
> >
> > Cc: Nicolas Ferre <nicolas.fe...@microchip.com>
> > Cc: Felipe Balbi <ba...@kernel.org>
> > Cc: Alexandre Belloni <alexandre.bell...@bootlin.com>
> > Signed-off-by: Greg Kroah-Hartman <gre...@linuxfoundation.org>
> > ---
> > drivers/usb/gadget/udc/atmel_usba_udc.c | 71 ++++---------------------
> > drivers/usb/gadget/udc/atmel_usba_udc.h | 4 --
> > 2 files changed, 11 insertions(+), 64 deletions(-)
> >
> > diff --git a/drivers/usb/gadget/udc/atmel_usba_udc.c
> > b/drivers/usb/gadget/udc/atmel_usba_udc.c
> > index 2f586f2bda7e..a4d99bf50f2f 100644
> > --- a/drivers/usb/gadget/udc/atmel_usba_udc.c
> > +++ b/drivers/usb/gadget/udc/atmel_usba_udc.c
> > @@ -206,94 +206,45 @@ static void usba_ep_init_debugfs(struct usba_udc *udc,
> > struct dentry *ep_root;
> >
> > ep_root = debugfs_create_dir(ep->ep.name, udc->debugfs_root);
> > - if (!ep_root)
> > - goto err_root;
> > ep->debugfs_dir = ep_root;
> >
> > - ep->debugfs_queue = debugfs_create_file("queue", 0400, ep_root,
> > - ep, &queue_dbg_fops);
> > - if (!ep->debugfs_queue)
> > - goto err_queue;
> > -
> > - if (ep->can_dma) {
> > - ep->debugfs_dma_status
> > - = debugfs_create_u32("dma_status", 0400, ep_root,
> > - &ep->last_dma_status);
> > - if (!ep->debugfs_dma_status)
> > - goto err_dma_status;
> > - }
> > - if (ep_is_control(ep)) {
> > - ep->debugfs_state
> > - = debugfs_create_u32("state", 0400, ep_root,
> > - &ep->state);
> > - if (!ep->debugfs_state)
> > - goto err_state;
> > - }
> > -
> > - return;
> > -
> > -err_state:
> > + debugfs_create_file("queue", 0400, ep_root, ep, &queue_dbg_fops);
>
> What happens here if debugfs_create_dir returned NULL? I guess the file
> will be placed at the root of the debugfs filesystem which is not great.
It doesn't really matter :)
The only way it can return NULL is if something really bad happens to
debugfs, and if that occurs, no one cares. The function can only return
NULL if you pass a bad dentry into it, or if the directory is being
removed at that exact point in time. Either of which are bigger issues
and then who cares what happens in debugfs.
No normal code should ever change its functionality based on if debugfs
is working or not.
> Should we stop caring about that and assume that if debugfs_create_dir,
> the following debugfs_create_* calls will fail?
Yes. You should not care :)
thanks,
greg k-h
--
To unsubscribe from this list: send the line "unsubscribe linux-usb" in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
