HELP: function pointer

[emanuela]

Hi,
I've a little question about C language:

according to the pthread_create signature:
int pthread_create( pthread_t *thread,
const pthread_attr_t *attr,
void *(*start_routin e)(void*),
void *arg);
which syntax is more correct?

1) void* function (void* args) {...}

main()
{ ...
pthread_create( &mythread, NULL, function, NULL)
...
}

2) void* function (void* args) {...}

main()
{ ...
pthread_create( &mythread, NULL, &function, NULL)
...
}

According to me, only the second one is correct
but I succeed in compiling both of them (gcc - Linux Fedora Core 1 (2.4.))
Why?

Thanks

Emanuela

[Simon Massey]

Re: function pointer

"emanuela" <em614@virgilio .it> wrote in message
news:833383f9.0 406140254.7865e 63d@posting.goo gle.com...[color=blue]
> Hi,
> I've a little question about C language:
> which syntax is more correct?
>
> 1) void* function (void* args) {...}
>
> main()
> { ...
> pthread_create( &mythread, NULL, function, NULL)
> pthread_create( &mythread, NULL, &function, NULL)
> ...
> }
>
> According to me, only the second one is correct
> but I succeed in compiling both of them (gcc - Linux Fedora Core 1 (2.4.))
> Why?[/color]

Both are correct (perhaps &function, slightly more so) as the name of a
function is it's address. The leading & is ignored with function names as
the only thing you can do with a function is call it (using ()) or take it's
address.

[red floyd]

Re: HELP: function pointer

em614@virgilio. it (emanuela) wrote in message news:<833383f9. 0406140254.7865 e63d@posting.go ogle.com>...[color=blue]
> Hi,
> I've a little question about C language:
>
> according to the pthread_create signature:
> int pthread_create( pthread_t *thread,
> const pthread_attr_t *attr,
> void *(*start_routin e)(void*),
> void *arg);
> which syntax is more correct?
>
> 1) void* function (void* args) {...}
>
> main()
> { ...
> pthread_create( &mythread, NULL, function, NULL)
> ...
> }
>
> 2) void* function (void* args) {...}
>
> main()
> { ...
> pthread_create( &mythread, NULL, &function, NULL)
> ...
> }[/color]

Either one is correct. IIRC, in the original K&R, you got a warning
with the second form. The naked name of a function (without parens)
refers to its address.

[Mark A. Odell]

Re: HELP: function pointer

em614@virgilio. it (emanuela) wrote in
news:833383f9.0 406140254.7865e 63d@posting.goo gle.com:
[color=blue]
> Hi,
> I've a little question about C language:
>
> according to the pthread_create signature:
> int pthread_create( pthread_t *thread,
> const pthread_attr_t *attr,
> void *(*start_routin e)(void*),
> void *arg);
> which syntax is more correct?
>
> 1) void* function (void* args) {...}
>
> main()
> { ...
> pthread_create( &mythread, NULL, function, NULL)
> ...
> }
>
> 2) void* function (void* args) {...}
>
> main()
> { ...
> pthread_create( &mythread, NULL, &function, NULL)
> ...
> }
>
> According to me, only the second one is correct
> but I succeed in compiling both of them (gcc - Linux Fedora Core 1
> (2.4.)) Why?[/color]

Because you are wrong. Similar to how you can use char array[10]; 'array'
as the address of the first element or '&array[0]' so too can you use
'function' or '&function'. Also, your main() is wrong, it must return
'int'. Do not rely on implicit 'returns int' of non-prototyped functions.

--
- Mark ->
--

[Jack Klein]

Re: function pointer

On Mon, 14 Jun 2004 12:52:23 +0100, "Simon Massey"
<sma@prismtechn ologies.com> wrote in comp.lang.c:
[color=blue]
> "emanuela" <em614@virgilio .it> wrote in message
> news:833383f9.0 406140254.7865e 63d@posting.goo gle.com...[color=green]
> > Hi,
> > I've a little question about C language:
> > which syntax is more correct?
> >
> > 1) void* function (void* args) {...}
> >
> > main()
> > { ...
> > pthread_create( &mythread, NULL, function, NULL)
> > pthread_create( &mythread, NULL, &function, NULL)
> > ...
> > }
> >
> > According to me, only the second one is correct
> > but I succeed in compiling both of them (gcc - Linux Fedora Core 1 (2.4.))
> > Why?[/color]
>
> Both are correct (perhaps &function, slightly more so) as the name of a
> function is it's address. The leading & is ignored with function names as
> the only thing you can do with a function is call it (using ()) or take it's
> address.[/color]

No, the name of a function is not an address, anymore than the name of
an array is an address. It is a "function designator", defined as
such by paragraph 4 of 6.3.2.1 of the current C standard:

"A function designator is an expression that has function type. Except
when it is the operand of the sizeof operator(54) or the unary &
operator, a function designator with type ‘‘function returning type’’
is converted to an expression that has type ‘‘pointer to
function returning type’’."

So a function designator is automatically and silently converted to a
function pointer in all contexts except when it is used as the operand
of the sizeof or & operators. The use of the & operator, as when this
operator is used on an lvalue of array type, explicitly converts the
name to a pointer.

The sizeof operator applied to a function designator, unlike applying
it to an lvalue of array type, is invalid, as stated explicitly in the
text of footnote 54, referenced above:

"54) Because this conversion does not occur, the operand of the sizeof
operator remains a function designator and violates the constraint in
6.5.3.4."

So given:

#include <stdio.h>
int func(int x)
{
return x + 1;
}

int (*fptr)(int) = func;

int main(void)
{
printf("size of function pointer %lu\n",
(unsigned long)sizeof(fpt r));
printf("size of function %lu\n",
(unsigned long)sizeof(fun c));
return 0;
}

....a conforming compiler must issue a diagnostic for the sizeof
expression in the second printf() call.

As shown by:

Error c:\prog\lcc\pro jects\sample\sa mple.c: 14 invalid type argument
'int function(int)' to 'sizeof'

....from one compiler and this:

Compiling...
sample.c
c:\program files\microsoft visual
studio\myprojec ts\sample\sampl e.c(14) : error C2070: illegal sizeof
operand
Error executing cl.exe.

sample.obj - 1 error(s), 0 warning(s)

....from another.

--
Jack Klein
Home: http://JK-Technology.Com
FAQs for
comp.lang.c http://www.eskimo.com/~scs/C-faq/top.html
comp.lang.c++ http://www.parashift.com/c++-faq-lite/
alt.comp.lang.l earn.c-c++

http://www.contrib.andrew.cmu.edu/~ajo/docs/FAQ-acllc.html

[emanuela]

Re: HELP: function pointer

thanks

Emanuela

"Mark A. Odell" <odellmark@hotm ail.com> wrote in message news:<Xns95086F 34AE6E5Copyrigh tMarkOdell@130. 133.1.4>...[color=blue]
> em614@virgilio. it (emanuela) wrote in
> news:833383f9.0 406140254.7865e 63d@posting.goo gle.com:
>[color=green]
> > Hi,
> > I've a little question about C language:
> >
> > according to the pthread_create signature:
> > int pthread_create( pthread_t *thread,
> > const pthread_attr_t *attr,
> > void *(*start_routin e)(void*),
> > void *arg);
> > which syntax is more correct?
> >
> > 1) void* function (void* args) {...}
> >
> > main()
> > { ...
> > pthread_create( &mythread, NULL, function, NULL)
> > ...
> > }
> >
> > 2) void* function (void* args) {...}
> >
> > main()
> > { ...
> > pthread_create( &mythread, NULL, &function, NULL)
> > ...
> > }
> >
> > According to me, only the second one is correct
> > but I succeed in compiling both of them (gcc - Linux Fedora Core 1
> > (2.4.)) Why?[/color]
>
> Because you are wrong. Similar to how you can use char array[10]; 'array'
> as the address of the first element or '&array[0]' so too can you use
> 'function' or '&function'. Also, your main() is wrong, it must return
> 'int'. Do not rely on implicit 'returns int' of non-prototyped functions.[/color]