Re: [PATCH v2 2/4] staging: unisys: dev_t initialization

From: Dan Carpenter
Date: Tue Mar 24 2015 - 04:14:41 EST


On Mon, Mar 16, 2015 at 04:36:13PM +0100, Greg Kroah-Hartman wrote:
> On Thu, Mar 12, 2015 at 11:00:10PM +0530, Sudip Mukherjee wrote:
> > dev_t is defined to be of unsigned int type, no use initializing
> > it to -1.
> >
> > Signed-off-by: Sudip Mukherjee <sudip@xxxxxxxxxxxxxxx>
> > ---
> > v2: it was not in v1
> >
> > drivers/staging/unisys/visorchipset/file.c | 2 +-
> > 1 file changed, 1 insertion(+), 1 deletion(-)
> >
> > diff --git a/drivers/staging/unisys/visorchipset/file.c b/drivers/staging/unisys/visorchipset/file.c
> > index 9ca7f1e..e9459af 100644
> > --- a/drivers/staging/unisys/visorchipset/file.c
> > +++ b/drivers/staging/unisys/visorchipset/file.c
> > @@ -30,7 +30,7 @@
> >
> > static struct cdev file_cdev;
> > static struct visorchannel **file_controlvm_channel;
> > -static dev_t majordev = -1; /**< indicates major num for device */
> > +static dev_t majordev; /**< indicates major num for device */
>
> I don't like this, please fix this up to handle the major number
> properly, no need for this -1 mess. And you just broke the logic with
> this change, which isn't allowed in any patch, sorry.

The patch description is bad and the patch is not complete, but it
doesn't introduce a bug.

The -1 is never used. By the time we call visorchipset_file_cleanup()
is called then majordev has been initialized to >= 0 so we can delete
the (MAJOR(majordev) >= 0) check. We can also delete the "registered"
variable because visorchipset_file_cleanup() is never called unless
"registered" is true.

"registered" is set on successful module_init() and
visorchipset_file_cleanup() is only called from module_exit().

There is the problem that visorchipset_init() doesn't release any
resources on error. But that's a separate issue and using globals like
"registered" is an ugly thing so it doesn't help.

regards,
dan carpenter

--
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/