[Bug fortran/31601] New: RFC: legacy-only allowed: State that code is allowed with -std=legacy ?

[Bug fortran/31601] New: RFC: legacy-only allowed: State that code is allowed with -std=legacy ?

[burnus at gcc dot gnu dot org]

Currently, it is far from obvious that rejected code might be compilable with
-std=legacy/-fno-range-check or similar (is there anything else?). I think one
could add a note to the error message:

Examples: 
"Integer too big for its kind at %C" -> "Integer too big for its kind at %C.
Use -fno-range-check to disable this check"

"Duplicate PROTECTED attribute specified at %L" -> "Duplicate PROTECTED
attribute specified at %L. Use -std=legacy to disable this check"

If one does so, one might also consider to promote some of the STD_GNU to
STD_LEGACY. My candidate would be, e.g., re-typing of used symbols ("use a,
only: sym; integer sym").


-- 
           Summary: RFC: legacy-only allowed: State that code is allowed
                    with -std=legacy ?
           Product: gcc
           Version: 4.3.0
            Status: UNCONFIRMED
          Keywords: diagnostic
          Severity: normal
          Priority: P3
         Component: fortran
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: burnus at gcc dot gnu dot org


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=31601

[Bug fortran/31601] RFC: legacy-only allowed: State that code is allowed with -std=legacy ?

[tobi at gcc dot gnu dot org]

-- 

tobi at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Severity|normal                      |enhancement
             Status|UNCONFIRMED                 |NEW
     Ever Confirmed|0                           |1
   Last reconfirmed|0000-00-00 00:00:00         |2007-04-17 14:05:03
               date|                            |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=31601

[Bug fortran/31601] RFC: legacy-only allowed: State that code is allowed with -std=legacy ?

[dominiq at lps dot ens.fr]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=31601

Dominique d'Humieres <dominiq at lps dot ens.fr> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |WAITING

--- Comment #1 from Dominique d'Humieres <dominiq at lps dot ens.fr> ---
> Examples: 
> "Integer too big for its kind at %C" -> "Integer too big for its kind at %C. 
> Use -fno-range-check to disable this check"

One now gets

This check can be disabled with the option -fno-range-check

since at least 4.3.1.

> "Duplicate PROTECTED attribute specified at %L" -> "Duplicate PROTECTED 
> attribute specified at %L. Use -std=legacy to disable this check"

From

        if (!gfc_notify_std (GFC_STD_LEGACY,
                             "Duplicate PROTECTED attribute specified at %L",
                             where))

this has not yet been implemented. If this deemed of any relevance, this is the
kind of fix I can handle, otherwise this PR can be closed as WONTFIX. It would
help me if a test case already exists, but I can do without it.

Last question, are there other cases one can think of?

[Bug fortran/31601] RFC: legacy-only allowed: State that code is allowed with -std=legacy ?

[dominiq at lps dot ens.fr]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=31601

Dominique d'Humieres <dominiq at lps dot ens.fr> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|WAITING                     |RESOLVED
         Resolution|---                         |WONTFIX

--- Comment #2 from Dominique d'Humieres <dominiq at lps dot ens.fr> ---
No answer to comment 1 since over a year and a half! Closing as WONTFIX.