Better diagnostic for shadowed function?

Better diagnostic for shadowed function?

[NightStrike]

I recently hit this problem:

#include <strings.h>
void f() {
    index[0] = 0;
}

#gcc is 11.2.0
gcc -c a.c
a.c:4:7: error: subscripted value is neither array nor pointer nor vector
    4 |  index[1] = 0;
      |       ^

-Wshadow (or all or extra) did not highlight that "index" was actually
a function from strings.h.  For the future, is there anything I could
have done to make gcc tell me what the real error was?

Re: Better diagnostic for shadowed function?

[Xi Ruoyao]

On Wed, 2022-01-12 at 01:33 -0500, NightStrike via Gcc-help wrote:
> I recently hit this problem:
> 
> #include <strings.h>
> void f() {
>     index[0] = 0;
> }
> 
> #gcc is 11.2.0
> gcc -c a.c
> a.c:4:7: error: subscripted value is neither array nor pointer nor
> vector
>     4 |  index[1] = 0;
>       |       ^
> 
> -Wshadow (or all or extra) did not highlight that "index" was actually
> a function from strings.h.

It has nothing to do with -Wshadow, because nothing is shadowed.

> For the future, is there anything I could
> have done to make gcc tell me what the real error was?

Open an issue in https://gcc.gnu.org/bugzilla and hope someone will pick
it up :).
-- 
Xi Ruoyao <xry111@mengyan1223.wang>
School of Aerospace Science and Technology, Xidian University

Re: Better diagnostic for shadowed function?

[Jonathan Wakely]

On Wed, 12 Jan 2022 at 06:34, NightStrike via Gcc-help
<gcc-help@gcc.gnu.org> wrote:
>
> I recently hit this problem:
>
> #include <strings.h>
> void f() {
>     index[0] = 0;
> }
>
> #gcc is 11.2.0
> gcc -c a.c
> a.c:4:7: error: subscripted value is neither array nor pointer nor vector
>     4 |  index[1] = 0;
>       |       ^
>
> -Wshadow (or all or extra) did not highlight that "index" was actually
> a function from strings.h.  For the future, is there anything I could
> have done to make gcc tell me what the real error was?

g++ is *slightly* more help:
s.c:3:10: error: invalid types ‘<unresolved overloaded function
type>[int]’ for array subscript

This at least gives you a hint it's a function, so probably declared
somewhere in a header. The C front end could do the same, saying
"subscripted value is a function", but it just has a generic
diagnostic for all un-subscriptable types. Asking for that to be
improved seems like a good use of a bugzilla PR.

Maybe would make sense for us to also add a note telling you where
'index' was declared, as long as it doesn't get suppressed without
-Wsystem-headers !

Re: Better diagnostic for shadowed function?

[NightStrike]

On Wed, Jan 12, 2022 at 5:06 AM Jonathan Wakely <jwakely.gcc@gmail.com> wrote:
>
> On Wed, 12 Jan 2022 at 06:34, NightStrike via Gcc-help
> <gcc-help@gcc.gnu.org> wrote:
> >
> > I recently hit this problem:
> >
> > #include <strings.h>
> > void f() {
> >     index[0] = 0;
> > }
> >
> > #gcc is 11.2.0
> > gcc -c a.c
> > a.c:4:7: error: subscripted value is neither array nor pointer nor vector
> >     4 |  index[1] = 0;
> >       |       ^
> >
> > -Wshadow (or all or extra) did not highlight that "index" was actually
> > a function from strings.h.  For the future, is there anything I could
> > have done to make gcc tell me what the real error was?
>
> g++ is *slightly* more help:
> s.c:3:10: error: invalid types ‘<unresolved overloaded function
> type>[int]’ for array subscript
>
> This at least gives you a hint it's a function, so probably declared
> somewhere in a header. The C front end could do the same, saying
> "subscripted value is a function", but it just has a generic
> diagnostic for all un-subscriptable types. Asking for that to be
> improved seems like a good use of a bugzilla PR.
>
> Maybe would make sense for us to also add a note telling you where
> 'index' was declared, as long as it doesn't get suppressed without
> -Wsystem-headers !

Added as https://gcc.gnu.org/bugzilla/show_bug.cgi?id=103996

Thanks!