[Bug c/11051] -Wno-deprecated needed also for C

[Bug c/11051] -Wno-deprecated needed also for C

[pinskia at physics dot uc dot edu]

PLEASE REPLY TO gcc-bugzilla@gcc.gnu.org ONLY, *NOT* gcc-bugs@gcc.gnu.org.

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11051



------- Additional Comments From pinskia at physics dot uc dot edu  2003-08-03 17:50 -------
The code is no rejected in the mainline (20030802).  Can any one think of another case 
but I still agree that -Wno-deprecated needs to be supported in C/ObjC.  Before 
20030729, I know that #import was deprecated but that had a option to disable the 
warning.

Right now only one thing in the C front end is deprecated but that only is a pedwarn.

I argree that there should be still an option for this in the furture.

[Bug c/11051] -Wno-deprecated needed also for C

[neroden at gcc dot gnu dot org]

PLEASE REPLY TO gcc-bugzilla@gcc.gnu.org ONLY, *NOT* gcc-bugs@gcc.gnu.org.

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11051


neroden at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Severity|normal                      |enhancement
   Target Milestone|3.4                         |---


------- Additional Comments From neroden at gcc dot gnu dot org  2003-08-03 18:33 -------
No reason to believe that this is going to be in 3.4, although it's certainly a
good idea.  It's also an enhancement request, fundamentally.

[Bug c/11051] -Wno-deprecated needed also for C

[manu at gcc dot gnu dot org]

------- Comment #15 from manu at gcc dot gnu dot org  2007-05-21 09:54 -------
This bug will be fixed as soon as Tom's patch goes in.

*** This bug has been marked as a duplicate of 22168 ***


-- 

manu at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |DUPLICATE


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11051

[Bug c/11051] -Wno-deprecated needed also for C

[manu at gcc dot gnu dot org]

------- Comment #14 from manu at gcc dot gnu dot org  2007-03-12 17:22 -------
Tom,

your patch 
http://gcc.gnu.org/ml/gcc-patches/2006-12/msg01853.html
will also fix this by adding Wdeprecated to the C front-end.


-- 

manu at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |tromey at gcc dot gnu dot
                   |                            |org


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11051

[Bug c/11051] -Wno-deprecated needed also for C

[manu at gcc dot gnu dot org]

------- Comment #13 from manu at gcc dot gnu dot org  2007-02-16 16:04 -------
(In reply to comment #12)
> Subject: Re:  -Wno-deprecated needed also for C
> 
> manu at gcc dot gnu dot org wrote:
> > 
> > Wouldn't it be better to remove the dead code? Or is there a policy against
> > touching things that are not broken? I think that at least one comment on the
> > code would be appropriate.
> > 
> 
> The dead code should presumably be removed, too.
> 

OK. But back to this PR. Is there anything deprecated in the C front-end that
would be suitable for -Wno-deprecated ?


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11051

[Bug c/11051] -Wno-deprecated needed also for C

[hpa at zytor dot com]

------- Comment #12 from hpa at zytor dot com  2007-01-27 01:40 -------
Subject: Re:  -Wno-deprecated needed also for C

manu at gcc dot gnu dot org wrote:
> 
> Wouldn't it be better to remove the dead code? Or is there a policy against
> touching things that are not broken? I think that at least one comment on the
> code would be appropriate.
> 

The dead code should presumably be removed, too.

        -hpa


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11051

[Bug c/11051] -Wno-deprecated needed also for C

[manu at gcc dot gnu dot org]

------- Comment #11 from manu at gcc dot gnu dot org  2007-01-27 01:38 -------
(In reply to comment #10)
> Subject: Re:  -Wno-deprecated needed also for C
> 
> manu at gcc dot gnu dot org wrote:
> > 
> > But it doesn't produce that warning. Is that warning dead code or what?
> > 
> 
> Apparently so.  I think it should have stayed a warning, but that's a 
> long time gone, and is way too late to fix now.  I'm going to submit a 
> doc patch, since this change was never documented.
> 

Wouldn't it be better to remove the dead code? Or is there a policy against
touching things that are not broken? I think that at least one comment on the
code would be appropriate.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11051

[Bug c/11051] -Wno-deprecated needed also for C

[hpa at zytor dot com]

------- Comment #10 from hpa at zytor dot com  2007-01-27 01:09 -------
Subject: Re:  -Wno-deprecated needed also for C

manu at gcc dot gnu dot org wrote:
> 
> But it doesn't produce that warning. Is that warning dead code or what?
> 

Apparently so.  I think it should have stayed a warning, but that's a 
long time gone, and is way too late to fix now.  I'm going to submit a 
doc patch, since this change was never documented.

        -hpa


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11051

[Bug c/11051] -Wno-deprecated needed also for C

[manu at gcc dot gnu dot org]

------- Comment #9 from manu at gcc dot gnu dot org  2007-01-27 01:02 -------
(In reply to comment #8)
> Subject: Re:  -Wno-deprecated needed also for C
> 
> > manu at gcc dot gnu dot org wrote:
> > > ------- Comment #4 from manu at gcc dot gnu dot org  2007-01-23 00:01 -------
> > > The testcase given is not valid any more. Could you think in any other
> > > testcase?
> > > 
> > > In stmt.c (expand_asm_operands) there is:
> > > 
> > >             warning (0, "use of memory input without lvalue in "
> > >                        "asm operand %d is deprecated", i + noutputs);
> > > 
> > > 
> > 
> > Hang on, hang on...
> > 
> > WTF?!  Using an rvalue in an assembly input that may contain "m" is 
> > something that is highly useful, and it will break metric tons of code. 
> >   -Wno-deprecated or no -Wno-deprecated, deprecating this particular 
> > construct is a major mistake.
> 
> This has been true since 3.3.3 and in fact, this was made an error in 4.0.0,
> even though the warning remains.  

Huh? What is that suppose to mean? 

> So:
> float f(float a)
> {
>   asm(""::"mo"((double)a));
>   return a;
> }
> 
> Fails

But it doesn't produce that warning. Is that warning dead code or what?

Still, going back to this PR, is there anything deprecated in C front-end worth
of implementing -Wno-deprecated for C?


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11051

[Bug c/11051] -Wno-deprecated needed also for C

[pinskia at physics dot uc dot edu]

------- Comment #8 from pinskia at physics dot uc dot edu  2007-01-23 00:29 -------
Subject: Re:  -Wno-deprecated needed also for C

> manu at gcc dot gnu dot org wrote:
> > ------- Comment #4 from manu at gcc dot gnu dot org  2007-01-23 00:01 -------
> > The testcase given is not valid any more. Could you think in any other
> > testcase?
> > 
> > In stmt.c (expand_asm_operands) there is:
> > 
> >             warning (0, "use of memory input without lvalue in "
> >                        "asm operand %d is deprecated", i + noutputs);
> > 
> > 
> 
> Hang on, hang on...
> 
> WTF?!  Using an rvalue in an assembly input that may contain "m" is 
> something that is highly useful, and it will break metric tons of code. 
>   -Wno-deprecated or no -Wno-deprecated, deprecating this particular 
> construct is a major mistake.

This has been true since 3.3.3 and in fact, this was made an error in 4.0.0,
even though the warning remains.  Also what is deprecated/removed is not may
contain a memory constraint but an input which only contains memory
constraints.

So:
float f(float a)
{
  asm(""::"mo"((double)a));
  return a;
}

Fails but:
float f(float a)
{
  asm(""::"mr"((double)a));
  return a;
}

Works.


-- Pinski


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11051

Re: [Bug c/11051] -Wno-deprecated needed also for C

[Andrew Pinski]

> manu at gcc dot gnu dot org wrote:
> > ------- Comment #4 from manu at gcc dot gnu dot org  2007-01-23 00:01 -------
> > The testcase given is not valid any more. Could you think in any other
> > testcase?
> > 
> > In stmt.c (expand_asm_operands) there is:
> > 
> >             warning (0, "use of memory input without lvalue in "
> >                        "asm operand %d is deprecated", i + noutputs);
> > 
> > 
> 
> Hang on, hang on...
> 
> WTF?!  Using an rvalue in an assembly input that may contain "m" is 
> something that is highly useful, and it will break metric tons of code. 
>   -Wno-deprecated or no -Wno-deprecated, deprecating this particular 
> construct is a major mistake.

This has been true since 3.3.3 and in fact, this was made an error in 4.0.0,
even though the warning remains.  Also what is deprecated/removed is not may
contain a memory constraint but an input which only contains memory constraints.

So:
float f(float a)
{
  asm(""::"mo"((double)a));
  return a;
}

Fails but:
float f(float a)
{
  asm(""::"mr"((double)a));
  return a;
}

Works.


-- Pinski

[Bug c/11051] -Wno-deprecated needed also for C

[hpa at zytor dot com]

------- Comment #7 from hpa at zytor dot com  2007-01-23 00:24 -------
Subject: Re:  -Wno-deprecated needed also for C

manu at gcc dot gnu dot org wrote:
> ------- Comment #6 from manu at gcc dot gnu dot org  2007-01-23 00:20 -------
> Hey, don't look at me. I am not sure what that means, I was just looking for
> something deprecated in C front-end to make a testcase.

It looks like it's deprecating the following code:

void foo(int bar)
{
        asm volatile("" : : "m" (bar+1));
}

This type of constructions are actually quite common (although with a 
real instruction, and usually something like "rm" or "abdSm" in the 
constraints box.)

        -hpa


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11051

[Bug c/11051] -Wno-deprecated needed also for C

[manu at gcc dot gnu dot org]

------- Comment #6 from manu at gcc dot gnu dot org  2007-01-23 00:20 -------
Hey, don't look at me. I am not sure what that means, I was just looking for
something deprecated in C front-end to make a testcase.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11051

[Bug c/11051] -Wno-deprecated needed also for C

[hpa at zytor dot com]

------- Comment #5 from hpa at zytor dot com  2007-01-23 00:15 -------
Subject: Re:  -Wno-deprecated needed also for C

manu at gcc dot gnu dot org wrote:
> ------- Comment #4 from manu at gcc dot gnu dot org  2007-01-23 00:01 -------
> The testcase given is not valid any more. Could you think in any other
> testcase?
> 
> In stmt.c (expand_asm_operands) there is:
> 
>             warning (0, "use of memory input without lvalue in "
>                        "asm operand %d is deprecated", i + noutputs);
> 
> 

Hang on, hang on...

WTF?!  Using an rvalue in an assembly input that may contain "m" is 
something that is highly useful, and it will break metric tons of code. 
  -Wno-deprecated or no -Wno-deprecated, deprecating this particular 
construct is a major mistake.

        -hpa


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11051

[Bug c/11051] -Wno-deprecated needed also for C

[manu at gcc dot gnu dot org]

------- Comment #4 from manu at gcc dot gnu dot org  2007-01-23 00:01 -------
The testcase given is not valid any more. Could you think in any other
testcase?

In stmt.c (expand_asm_operands) there is:

            warning (0, "use of memory input without lvalue in "
                       "asm operand %d is deprecated", i + noutputs);


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11051