Re: [cocci] [1/36] coccinelle: misc: add cond_return_no_effect.cocci

>> I would appreciate to clarify such an approach for a bit of dead-code e=
limination
>> in more detail.
>=20
> I have no idea what you are talking about.  What dead code are you
> referring to?

Variable definitions which became unused according to transformation rules
by the semantic patch language.
https://en.wikipedia.org/wiki/Dead-code_elimination


>> I got the impression that some versions of the C programming language d=
o
>> not handle variable declarations also as =E2=80=9Cstatements=E2=80=9D.
>> https://en.cppreference.com/c/language/statements
>>
>> See also:
>> https://en.cppreference.com/cpp/language/statements#Declaration_stateme=
nts
>=20
> And the conclusion would be what?  First, Coccinelle has it's own parser=
.

This is generally fine.


> It decides what is a statement, not some wikipedia page.

Does your supported data model deviate from known programming language sta=
ndard specifications?

Can users determine relevant differences from published software documenta=
tion?


> Second, Coccinelle does not consider declarations to be statements.
> That's what the D is there for.

You suggested to apply two metavariable types in an SmPL disjunction
while trying to delete special variable definitions, didn't you?


>>> The first part ensures that it is not a top-level declaration and the
>>> second part does the check you had previously.
>>
>> I do not see so far that the intended source code restriction would be =
explicitly
>> expressed here.
>=20
> No idea what this means.  What intended code restriction are you referri=
ng to?
Do you try to distinguish global variable declarations from local definiti=
ons anyhow?

How are variable scopes be taken into account here?
https://github.com/coccinelle/coccinelle/issues/322

Regards,
Markus