[PATCH v2] Documentation: checkpatch: add UNNECESSARY_CASTS message.

[Vinesh Vilas Raut]

On Sat, Aug 7, 2021 at 1:36 AM Dwaipayan Ray <dwaipayanray1 at gmail.com> wrote:
>
> On Fri, Aug 6, 2021 at 8:49 PM Vinesh Raut <vineshraut2017 at gmail.com> wrote:
> >
> > Add verbose description for UNNECESSARY_CASTS message type.
> >
> > Suggested-by: Lukas Bulwahn <lukas.bulwahn at gmail.com>
> > Signed-off-by: Vinesh Raut <vineshraut2017 at gmail.com>
> > ---
> >  Documentation/dev-tools/checkpatch.rst | 9 +++++++++
> >  1 file changed, 9 insertions(+)
> >
> > diff --git a/Documentation/dev-tools/checkpatch.rst b/Documentation/dev-tools/checkpatch.rst
> > index f0956e9ea2d8..9cb6b9059fc6 100644
> > --- a/Documentation/dev-tools/checkpatch.rst
> > +++ b/Documentation/dev-tools/checkpatch.rst
> > @@ -268,6 +268,15 @@ Allocation style
> >
> >      See: https://www.kernel.org/doc/html/latest/core-api/memory-allocation.html
> >
> > +   **UNNECESSAR_CASTS**

+   **UNNECESSARY_CASTS**

> > +   Cast should not be done to any alloc functions.
> > +   Appropriate type of pointer is pick by compiler at run time

+   Appropriate type of pointer is pick by compiler at runtime

> s/run time/runtime
>
> > +   Cast may hide bugs, as compiler suppresses the warning after cast.


+   If a pointer is cast (wrongly), bugs like memory corruption,
segmentation faults can occur.
+   Also, It can affect functionality of the program.


>
> The question which arises here is what kind of bug is suppressed.
> Maybe follow http://c-faq.com/malloc/mallocnocast.html as checkpatch
> says to write up a paragraph on why it is bad.
>
> > +
> > +   See:
> > +   1] https://lore.kernel.org/lkml/20070808024321.GA6316@redhat.com/
> > +   2] https://lore.kernel.org/lkml/a5e60a2b93e10baf84af063f6c8e56402273105d.camel@perches.com/
>
> These links are just patches that introduced the test. There is not much
> information about "why" they were introduced.

Should I remove it?
And I didnt found any discussion on LKML regarding  UNNECESSARY_CASTS issue,
other than this patches.

>
> Other than that change the enumeration type to be
> uniform with the rest of the documentation maybe?
> The thing followed till now is:
>
> See:
>   1. link_1
>   2. link_2

+   See:
+   1. https://lore.kernel.org/lkml/20070808024321.GA6316@redhat.com/
+   2. https://lore.kernel.org/lkml/a5e60a2b93e10baf84af063f6c8e56402273105d.camel@perches.com/

>
> Dwaipayan.