http://d.puremagic.com/issues/show_bug.cgi?id=10482

           Summary: Regression (2.063): Compiler allows constant global
                    functions
           Product: D
           Version: D2
          Platform: All
        OS/Version: All
            Status: NEW
          Keywords: accepts-invalid
          Severity: regression
          Priority: P2
         Component: DMD
        AssignedTo: nobody@puremagic.com
        ReportedBy: andrej.mitrovich@gmail.com


--- Comment #0 from Andrej Mitrovic <andrej.mitrovich@gmail.com> 2013-06-26 17:42:58 PDT ---
-----
const int * foo();

void main() { }
-----

2.062:
$ dmd test.d
> test.d(3): Error: function test.foo without 'this' cannot be const/immutable

2.063:
$ dmd test.d
> 

I'm not sure how this happened, I remember we had a test-suite covering these.

-- 
Configure issuemail: http://d.puremagic.com/issues/userprefs.cgi?tab=email
------- You are receiving this mail because: -------

http://d.puremagic.com/issues/show_bug.cgi?id=10482


Henning Pohl <henning@still-hidden.de> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |henning@still-hidden.de


--- Comment #1 from Henning Pohl <henning@still-hidden.de> 2013-06-27 10:19:26 PDT ---
This behaviour is wanted, moving const after the function name will produce an error:

int* foo() const;

----
Error: function main.foo without 'this' cannot be const
----

See http://d.puremagic.com/issues/show_bug.cgi?id=10150

-- 
Configure issuemail: http://d.puremagic.com/issues/userprefs.cgi?tab=email
------- You are receiving this mail because: -------

http://d.puremagic.com/issues/show_bug.cgi?id=10482


bearophile_hugs@eml.cc changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |bearophile_hugs@eml.cc


--- Comment #2 from bearophile_hugs@eml.cc 2013-06-30 08:58:27 PDT ---
See also Issue 10511

-- 
Configure issuemail: http://d.puremagic.com/issues/userprefs.cgi?tab=email
------- You are receiving this mail because: -------

http://d.puremagic.com/issues/show_bug.cgi?id=10482


Dicebot <m.strashun@gmail.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |m.strashun@gmail.com


--- Comment #3 from Dicebot <m.strashun@gmail.com> 2013-06-30 09:39:08 PDT ---
This is irrelevant to Issue 10511. Global const functions are still invalid, no regressions here : http://dpaste.1azy.net/a2500829

Looks like 2.062 had bug in parser and coupled leading 'const' with function type instead of return type. IMHO this is invalid / won't fix.

-- 
Configure issuemail: http://d.puremagic.com/issues/userprefs.cgi?tab=email
------- You are receiving this mail because: -------

http://d.puremagic.com/issues/show_bug.cgi?id=10482



--- Comment #4 from Andrej Mitrovic <andrej.mitrovich@gmail.com> 2013-06-30 09:56:33 PDT ---
(In reply to comment #3)
> IMHO this is invalid / won't fix.

See my comments: http://d.puremagic.com/issues/show_bug.cgi?id=10150#c4 http://d.puremagic.com/issues/show_bug.cgi?id=10150#c5

I'm reposting them here:

(In reply to comment #4)
> I disagree with this change, this is extremely dangerous behavior when interfacing with C. Take a look at the following:
> 
> -----
> extern(C) const int *foo();
> 
> void main()
> {
>     *foo() = 1;  // compiles
> }
> -----
> 
> The proper definition should have been:
> 
> -----
> extern(C) const(int)* foo();
> 
> void main()
> {
>     *foo() = 1;  // fails
> }
> -----
> 
> It is very easy to make this mistake, it should not be silent, at least not for module-scoped functions.
> 
> Alternatively as a compromise I suggest we at least add this check for extern(C) functions, because this is where this problem can occur very frequently.


(In reply to comment #4)
> I disagree with this change, this is extremely dangerous behavior when interfacing with C. Take a look at the following:
> 
> -----
> extern(C) const int *foo();
> 
> void main()
> {
>     *foo() = 1;  // compiles
> }
> -----
> 
> The proper definition should have been:
> 
> -----
> extern(C) const(int)* foo();
> 
> void main()
> {
>     *foo() = 1;  // fails
> }
> -----
> 
> It is very easy to make this mistake, it should not be silent, at least not for module-scoped functions.
> 
> Alternatively as a compromise I suggest we at least add this check for extern(C) functions, because this is where this problem can occur very frequently.

(In reply to comment #5)
> Another alternative is to make the compiler smarter, and only disallow const where it's only applied to one function, for example:
> 
> -----
> const int * foo();  // disallowed
> 
> const
> {
>     int * foo();  // ok
>     int * bar();  // ok
> }
> 
> const:
> int * foo();  // ok
> -----
> 
> This would be for the sake of convenience, to avoid breaking existing code.

-- 
Configure issuemail: http://d.puremagic.com/issues/userprefs.cgi?tab=email
------- You are receiving this mail because: -------

http://d.puremagic.com/issues/show_bug.cgi?id=10482



--- Comment #5 from Andrej Mitrovic <andrej.mitrovich@gmail.com> 2013-06-30 09:57:38 PDT ---
Sorry for the duplicate paste of comment 4.

-- 
Configure issuemail: http://d.puremagic.com/issues/userprefs.cgi?tab=email
------- You are receiving this mail because: -------

http://d.puremagic.com/issues/show_bug.cgi?id=10482



--- Comment #6 from Dicebot <m.strashun@gmail.com> 2013-06-30 10:31:52 PDT ---
Yeah, have figure out that it is actually no-op in prefix mode now. Still not what title claims, but is.. inconvenient. Moving in that thread.

-- 
Configure issuemail: http://d.puremagic.com/issues/userprefs.cgi?tab=email
------- You are receiving this mail because: -------

http://d.puremagic.com/issues/show_bug.cgi?id=10482


Kenji Hara <k.hara.pg@gmail.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Severity|regression                  |enhancement


--- Comment #7 from Kenji Hara <k.hara.pg@gmail.com> 2013-07-02 07:25:01 PDT ---
This is an intended change comes from issue 10150, so is not a regression.

Today dmd never distinguishes prefix storage class and scopes/labeled ones. That's a design decision which sometimes talked by Walter.

But also, I can agree that it is a little bug-prone behavior. So I switch this into enhancement issue.

-- 
Configure issuemail: http://d.puremagic.com/issues/userprefs.cgi?tab=email
------- You are receiving this mail because: -------

http://d.puremagic.com/issues/show_bug.cgi?id=10482



--- Comment #8 from Dicebot <m.strashun@gmail.com> 2013-07-02 08:43:42 PDT ---
(In reply to comment #7)
> This is an intended change comes from issue 10150, so is not a regression.
> 
> Today dmd never distinguishes prefix storage class and scopes/labeled ones. That's a design decision which sometimes talked by Walter.
> 
> But also, I can agree that it is a little bug-prone behavior. So I switch this into enhancement issue.

Yep, it is not a regression but a breaking change introduces by 10150, which should not have been implemented in the first place. I wish I have noticed that pull request before it was merged. There is currently minor discussion on topic in 10150 thread.

-- 
Configure issuemail: http://d.puremagic.com/issues/userprefs.cgi?tab=email
------- You are receiving this mail because: -------