On Thu, Apr 1, 2021 at 9:27 AM Mugilraj Dhavachelvan
<[email protected]> wrote:
>
> Change function's name to %s and __func__ to fix checkpatch.pl errors.
No, just drop the __func__ completely. First of all, we have a device
name, and uniqueness of the messages in the driver should guarantee
easy findings. Second, specific to _dbg() variants, with enabled
Dynamic Debug it can be chosen at run time!
I recommend going through all drivers and drop __func__ near to
dev_dbg() and pr_debug().
--
With Best Regards,
Andy Shevchenko
On Thu, 1 Apr 2021 12:24:50 +0300
Andy Shevchenko <[email protected]> wrote:
> On Thu, Apr 1, 2021 at 9:27 AM Mugilraj Dhavachelvan
> <[email protected]> wrote:
> >
> > Change function's name to %s and __func__ to fix checkpatch.pl errors.
>
> No, just drop the __func__ completely. First of all, we have a device
> name, and uniqueness of the messages in the driver should guarantee
> easy findings. Second, specific to _dbg() variants, with enabled
> Dynamic Debug it can be chosen at run time!
>
> I recommend going through all drivers and drop __func__ near to
> dev_dbg() and pr_debug().
>
Agreed. Though beware that some maintainers will count this
as noise and get grumpy.
I'm fine with such patches for IIO.
Jonathan
On 01/04/21 4:00 pm, Jonathan Cameron wrote:
> On Thu, 1 Apr 2021 12:24:50 +0300
> Andy Shevchenko <[email protected]> wrote:
>
>> On Thu, Apr 1, 2021 at 9:27 AM Mugilraj Dhavachelvan
>> <[email protected]> wrote:
>>>
>>> Change function's name to %s and __func__ to fix checkpatch.pl errors.
>>
>> No, just drop the __func__ completely. First of all, we have a device
>> name, and uniqueness of the messages in the driver should guarantee
>> easy findings. Second, specific to _dbg() variants, with enabled
>> Dynamic Debug it can be chosen at run time!
>>
>> I recommend going through all drivers and drop __func__ near to
>> dev_dbg() and pr_debug().
>>
>
> Agreed. Though beware that some maintainers will count this
> as noise and get grumpy.
>
> I'm fine with such patches for IIO.
Sorry for the noise. I just seen the docs about dynamic debug.
So, if we use dev_dbg("log_msg") it will print statement like
filename:lineno [module]func flag log_msg, If I get it correctly.
And no need of specifying __func__ in dev_dbg() and dp_dbg() right!!
Jonathan do you have any TODO's?
>
> Jonathan
>
On Thu, 1 Apr 2021 20:22:47 +0530
Mugilraj D <[email protected]> wrote:
> On 01/04/21 4:00 pm, Jonathan Cameron wrote:
> > On Thu, 1 Apr 2021 12:24:50 +0300
> > Andy Shevchenko <[email protected]> wrote:
> >
> >> On Thu, Apr 1, 2021 at 9:27 AM Mugilraj Dhavachelvan
> >> <[email protected]> wrote:
> >>>
> >>> Change function's name to %s and __func__ to fix checkpatch.pl errors.
> >>
> >> No, just drop the __func__ completely. First of all, we have a device
> >> name, and uniqueness of the messages in the driver should guarantee
> >> easy findings. Second, specific to _dbg() variants, with enabled
> >> Dynamic Debug it can be chosen at run time!
> >>
> >> I recommend going through all drivers and drop __func__ near to
> >> dev_dbg() and pr_debug().
> >>
> >
> > Agreed. Though beware that some maintainers will count this
> > as noise and get grumpy.
> >
> > I'm fine with such patches for IIO.
>
> Sorry for the noise. I just seen the docs about dynamic debug.
> So, if we use dev_dbg("log_msg") it will print statement like
> filename:lineno [module]func flag log_msg, If I get it correctly.
> And no need of specifying __func__ in dev_dbg() and dp_dbg() right!!
>
> Jonathan do you have any TODO's?
I tend not to mind cleanup patches (within reason)in IIO so I'm absolutely
fine with a series removing any __func__ items like this. One patch per
driver preferred because it avoids issues with this interfering with backports
etc. There will end up being about 18 patches from a quick grep. Perhaps send
a small number first though to avoid having to put in too much effort as
any issues likely to affect the whole set.
Just be careful with other maintainers, they sometimes strike a different
balance for what they consider noise vs useful.
Jonathan
>
> >
> > Jonathan
> >