Re: [RFC PATCH 06/13] of: Remove prefix "__of_" and prefix "__" from local function names

From: Frank Rowand
Date: Thu Oct 27 2016 - 14:26:18 EST


On 10/27/16 09:58, Rob Herring wrote:
> On Thu, Oct 27, 2016 at 11:35 AM, Frank Rowand <frowand.list@xxxxxxxxx> wrote:
>> On 10/27/16 05:47, Rob Herring wrote:
>>> On Tue, Oct 25, 2016 at 3:58 PM, <frowand.list@xxxxxxxxx> wrote:
>>>> From: Frank Rowand <frank.rowand@xxxxxxxxxxx>
>>>
>>> I prefer to leave the prefixes and this is getting into pointless churn.
>>>
>>>>
>>>> Signed-off-by: Frank Rowand <frank.rowand@xxxxxxxxxxx>
>>>> ---
>>>> drivers/of/resolver.c | 10 +++++-----
>>>> 1 file changed, 5 insertions(+), 5 deletions(-)
>>>
>>
>> If I was just submitting this as a single patch, I would agree.
>>
>> But since I am making so many other changes, I think it makes
>> sense to do this as part of this series. It is broken apart
>> as a separate patch to be easy to review and not pollute any
>> of the other patches in the series.
>>
>> The prefixes add no value for a local function, but they do
>> add noise when reading code.
>
> The value is when reading the calling function, you know the function
> is a DT related function. You don't know it's a static function

It is more than that. A common convention in drivers/of/ is that
function blah() acquires a lock, calls function __blah(), and
releases the lock. Any function other than blah() that wants
to call __blah() must also hold the proper lock. The functions
whose name this patch changes do not fit this pattern.


> without looking up the function name. That said, I wouldn't object to
> code originally written either way, I just don't see the value in
> changing it.
>
> Rob
>