Re: [PATCH] powercap: Adjust printing the constraint name with new line

[Lukasz Luba <lukasz.luba@arm.com>]

Hi Rafael,

On 11/10/20 7:19 PM, Rafael J. Wysocki wrote:
> On Mon, Nov 9, 2020 at 6:25 PM Lukasz Luba <lukasz.luba@arm.com> wrote:
>>
>> The constraint name has limit of size 30, which sometimes might be hit.
>> When this happens the new line might be lost. Prevent this and set the
>> new line when the name string is too long."
>>
>> Signed-off-by: Lukasz Luba <lukasz.luba@arm.com>
>> ---
>>   drivers/powercap/powercap_sys.c | 4 ++++
>>   1 file changed, 4 insertions(+)
>>
>> diff --git a/drivers/powercap/powercap_sys.c b/drivers/powercap/powercap_sys.c
>> index f808c5fa9838..575f9fdb810e 100644
>> --- a/drivers/powercap/powercap_sys.c
>> +++ b/drivers/powercap/powercap_sys.c
>> @@ -174,6 +174,10 @@ static ssize_t show_constraint_name(struct device *dev,
>>                                                                  "%s\n", name);
>>                          buf[POWERCAP_CONSTRAINT_NAME_LEN] = '\0';
>>                          len = strlen(buf);
>> +
>> +                       /* When the 'name' is too long, don't lose new line */
>> +                       if (strlen(name) >= POWERCAP_CONSTRAINT_NAME_LEN)
>> +                               buf[POWERCAP_CONSTRAINT_NAME_LEN - 1] = '\n';
> 
> Wouldn't it be better to pass POWERCAP_CONSTRAINT_NAME_LEN - 1 to
> snprintf() above?

My apologies for delay, I was on holidays.

The snprintf() overwrites the '\n' in that case also. I've experimented
with it a bit more and tried to come with a bit 'cleaner' solution.

What we are looking is probably: "%.*s\n" semantic.
Another solution would be:
------------------------------8<---------------------------
                 if (name) {
-                       snprintf(buf, POWERCAP_CONSTRAINT_NAME_LEN,
-                                                               "%s\n", 
name);
-                       buf[POWERCAP_CONSTRAINT_NAME_LEN] = '\0';
+                       sprintf(buf, "%.*s\n", 
POWERCAP_CONSTRAINT_NAME_LEN -1,
+                               name);
                         len = strlen(buf);
                 }

----------------------------->8----------------------------

I've check this and it behaves very well.

It looks cleaner and it is a used pattern in the kernel.
What do you think? Is it good for v2?

Regards,
Lukasz