On Mon, Sep 23, 2013 at 8:00 PM, Dan Carpenter <dan.carpenter@xxxxxxxxxx> wrote:
> On Mon, Sep 23, 2013 at 06:47:17PM -0400, Lidza Louina wrote:
>> This patch removes this smatch warning:
>> unsigned '--un->un_open_count' is never less than zero
>>
>> The code decremented the un_open_count variable
>> and tested to see if it was less than zero. Because
>> un_open_count is unsigned and can't be below zero,
>> this test doesn't work. This patch tests
>> un_open_count against 0 without decrementing it.
>>
>> Signed-off-by: Lidza Louina <lidza.louina@xxxxxxxxx>
>> ---
>> drivers/staging/dgap/dgap_tty.c | 2 +-
>> 1 file changed, 1 insertion(+), 1 deletion(-)
>>
>> diff --git a/drivers/staging/dgap/dgap_tty.c b/drivers/staging/dgap/dgap_tty.c
>> index 8f0a824..f496710 100644
>> --- a/drivers/staging/dgap/dgap_tty.c
>> +++ b/drivers/staging/dgap/dgap_tty.c
>> @@ -1442,7 +1442,7 @@ static void dgap_tty_close(struct tty_struct *tty, struct file *file)
>> un->un_open_count = 1;
>> }
>>
>> - if (--un->un_open_count < 0) {
>> + if (un->un_open_count == 0) {
>> APR(("bad serial port open count of %d\n", un->un_open_count));
>> un->un_open_count = 0;
>> }
>
> This fix isn't right. We still need the decrement. Probably the best
> thing is to audit all the driver and make sure that un->un_open_count
> can never be == 0. But the next best, and still totally reasonable
> thing to do is this:
>
> if (un->un_open_count == 0) {
> APR(("bad serial port open count of %d\n", un->un_open_count));
> un->un_open_count = 1;
> }
> un->un_open_count--;
>
> regards,
> dan carpenter
>
Okay, I see why, I'll resend this.
_______________________________________________
devel mailing list
devel@xxxxxxxxxxxxxxxxxxxxxx
http://driverdev.linuxdriverproject.org/mailman/listinfo/driverdev-devel
