Can't determine which overloaded function to call

classic Classic list List threaded Threaded
6 messages Options
Reply | Threaded
Open this post in threaded view
|

Can't determine which overloaded function to call

LacaK
Hi *,

this code compiles for target Win32 but does not compile for
Win64/x86-64. Why? Is there workaround?
(Error: Can't determine which overloaded function to call)
Thank you
-Laco.

=== code sample ===

  TRec1 = record
    x,y: integer;
    function Offset(const Ax,Ay: integer): TRec1; overload;
    function Offset(const Ax,Ay: single): TRec1; overload;
  end;

function TRec1.Offset(const Ax,Ay: integer): TRec1;
begin
end;

function TRec1.Offset(const Ax,Ay: single): TRec1;
begin
end;

var
  r1: TRec1;
  a,b: integer;

begin
  r1 := r1.Offset(a-1,b-1); // HERE Error:
end.  

============================

_______________________________________________
fpc-pascal maillist  -  [hidden email]
http://lists.freepascal.org/cgi-bin/mailman/listinfo/fpc-pascal
Reply | Threaded
Open this post in threaded view
|

Re: Can't determine which overloaded function to call

Santiago A.
El 04/12/18 a las 13:21, LacaK escribió:
> Hi *,
>
> this code compiles for target Win32 but does not compile for
> Win64/x86-64. Why? Is there workaround?
> (Error: Can't determine which overloaded function to call)
> Thank you
For me, in win32 works fine

try this

r1.Offset(a-Integer(1),b-Integer(1));

or

r1.Offset(Integer(a-1),Integer(b-1));


--
Saludos

Santiago A.

_______________________________________________
fpc-pascal maillist  -  [hidden email]
http://lists.freepascal.org/cgi-bin/mailman/listinfo/fpc-pascal
Reply | Threaded
Open this post in threaded view
|

Re: Can't determine which overloaded function to call

LacaK

>
>> Hi *,
>>
>> this code compiles for target Win32 but does not compile for
>> Win64/x86-64. Why? Is there workaround?
>> (Error: Can't determine which overloaded function to call)
>> Thank you
> For me, in win32 works fine

Yes in Win32 works also for me. And this is my question why it does work
in Win32 and does not work in Win64?


>
> try this
>
> r1.Offset(a-Integer(1),b-Integer(1));

does not helps


>
> or
>
> r1.Offset(Integer(a-1),Integer(b-1));

helps, but why is it not needed in Win32? Why for Win32
"integer"-"integer" is considered as "integer" so compiler can determine
which overloaded function to call and for Win64 compiler compiler can
NOT determine which overloaded function to call?

It seems to me as inconsistent?

L.

=== code sample ===

  TRec1 = record
    x,y: integer;
    function Offset(const Ax,Ay: integer): TRec1; overload;
    function Offset(const Ax,Ay: single): TRec1; overload;
  end;

function TRec1.Offset(const Ax,Ay: integer): TRec1;
begin
end;

function TRec1.Offset(const Ax,Ay: single): TRec1;
begin
end;

var
  r1: TRec1;
  a,b: integer;

begin
  r1 := r1.Offset(a-1,b-1); // HERE Error:
end.  

============================


_______________________________________________
fpc-pascal maillist  -  [hidden email]
http://lists.freepascal.org/cgi-bin/mailman/listinfo/fpc-pascal
Reply | Threaded
Open this post in threaded view
|

Re: Can't determine which overloaded function to call

Jonas Maebe-3
On 05/12/18 07:51, LacaK wrote:
> helps, but why is it not needed in Win32? Why for Win32
> "integer"-"integer" is considered as "integer" so compiler can determine
> which overloaded function to call and for Win64 compiler compiler can
> NOT determine which overloaded function to call?

It is because as documented at
https://www.freepascal.org/docs-html/ref/refsu4.html (remarks under
table 3.2), FPC evaluates integer expressions using the native integer
type of the platform. On Win32 this is 32 bit, while on Win64 this is 64
bit. On the other hand, "integer" is always 32 bit in Delphi mode.

This means that on Win32, there is an exact match for overload selection
in your test program, while on Win64 there is not and the int64 ->
integer and int64 -> single type conversions have the same priority.


Jonas
_______________________________________________
fpc-pascal maillist  -  [hidden email]
http://lists.freepascal.org/cgi-bin/mailman/listinfo/fpc-pascal
Reply | Threaded
Open this post in threaded view
|

Re: Can't determine which overloaded function to call

LacaK

>> helps, but why is it not needed in Win32? Why for Win32
>> "integer"-"integer" is considered as "integer" so compiler can determine
>> which overloaded function to call and for Win64 compiler compiler can
>> NOT determine which overloaded function to call?
>
> It is because as documented at
> https://www.freepascal.org/docs-html/ref/refsu4.html (remarks under
> table 3.2), FPC evaluates integer expressions using the native integer
> type of the platform.

Do you mean: "2. Every integer smaller than the ”native” size is
promoted to a signed version of the ”native” size. Integers equal to the
”native” size keep their signedness. "?


> On Win32 this is 32 bit, while on Win64 this is 64 bit. On the other
> hand, "integer" is always 32 bit in Delphi mode.
>
> This means that on Win32, there is an exact match for overload
> selection in your test program, while on Win64 there is not and the
> int64 -> integer and int64 -> single type conversions have the same
> priority.

I understand now. Although I must say that it is bit unintuitive for me ;-)

So I can workaround using:
    function Offset(const Ax,Ay: PtrInt): TRec1; overload;

Thank you for explanation

-Laco.

_______________________________________________
fpc-pascal maillist  -  [hidden email]
http://lists.freepascal.org/cgi-bin/mailman/listinfo/fpc-pascal
Reply | Threaded
Open this post in threaded view
|

Re: Can't determine which overloaded function to call

Santiago A.
In reply to this post by Jonas Maebe-3
El 05/12/18 a las 08:46, Jonas Maebe escribió:

> On 05/12/18 07:51, LacaK wrote:
>> helps, but why is it not needed in Win32? Why for Win32
>> "integer"-"integer" is considered as "integer" so compiler can determine
>> which overloaded function to call and for Win64 compiler compiler can
>> NOT determine which overloaded function to call?
>
> It is because as documented at
> https://www.freepascal.org/docs-html/ref/refsu4.html (remarks under
> table 3.2), FPC evaluates integer expressions using the native integer
> type of the platform. On Win32 this is 32 bit, while on Win64 this is
> 64 bit. On the other hand, "integer" is always 32 bit in Delphi mode.
>
> This means that on Win32, there is an exact match for overload
> selection in your test program, while on Win64 there is not and the
> int64 -> integer and int64 -> single type conversions have the same
> priority.
do  (LongInt -> Integer) and (Longint -> single) have the same priority?


--
Saludos

Santiago A.

_______________________________________________
fpc-pascal maillist  -  [hidden email]
http://lists.freepascal.org/cgi-bin/mailman/listinfo/fpc-pascal