Re: [PATCH linux-phy v2 1/4] string.h: Add str_has_proper_prefix()

["Marek Behún" <kabel@kernel.org>]

On Thu, 18 Aug 2022 22:10:58 +0300
Andy Shevchenko <andy.shevchenko@gmail.com> wrote:

> On Wed, Aug 17, 2022 at 11:06 PM Marek Behún <kabel@kernel.org> wrote:
> >
> > Add str_has_proper_prefix(), similar to str_has_prefix(), but requires
> > that the prefix is proper: the string itself must be longer than the
> > prefix.
> >
> > Signed-off-by: Marek Behún <kabel@kernel.org>
> > ---
> > Andy, Kees, could you ack this if it is ok?  
> 
> Seems to me there are too many strlen():s. One is hidden in strncmp().

I thought this was ok cause gcc has optimizations for this in
https://github.com/gcc-mirror/gcc/blob/master/gcc/tree-ssa-strlen.cc

But now I see that kernel does not declare these functions as inline
that call __builtin_strlen()... so probably the optimizations are not
used.

> Besides not the good naming (what 'proper' means),

The naming comes from similar naming in math: proper subset is as
subset that is not equal to the superset. See
https://en.wikipedia.org/wiki/Substring :
  "A proper prefix of a string is not equal to the string itself"

> the entire function is not needed. You may simply call
> 
>   str_has_prefix() && p[len] != '\0';
> 
> Correct?

Do you mean that I should implement this function to simply return
  str_has_prefix() && p[len] != '\0'
or that this function should not exist at all and I should do that in
the code where I would have used the function?

Thanks.

Marek