[Bug c/114181] New: issubnormal is a macro

[Bug c/114181] New: issubnormal is a macro

[g.peterhoff@t-online.de]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=114181

            Bug ID: 114181
           Summary: issubnormal is a macro
           Product: gcc
           Version: unknown
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: c
          Assignee: unassigned at gcc dot gnu.org
          Reporter: g.peterhoff@t-online.de
  Target Milestone: ---

issubnormal is a macro and therefore not a (builtin)function.
This is incorrect, as no further issubnormal functions can be implemented, e.g.
for C++
namespace std
{
bool issubnormal(...);
}

thx
Gero

[Bug c/114181] issubnormal is a macro

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=114181

Andrew Pinski <pinskia at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
     Ever confirmed|0                           |1
   Last reconfirmed|                            |2024-03-01
             Status|UNCONFIRMED                 |WAITING

--- Comment #1 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
Gcc does not provide math.h. nor defines the macro issubnormal.


In libstdc++, there are #undef for some macros to get around libc that define
some functions as macros.

If you are implementing something similar then you should be able to do that
too.

[Bug c/114181] issubnormal is a macro

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=114181

--- Comment #2 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
Also please provide a full testcase to show what you are doing because this
code snippet is not even valid c++.

[Bug c/114181] issubnormal is a macro

[g.peterhoff@t-online.de]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=114181

--- Comment #3 from g.peterhoff@t-online.de ---
Of course issubnormal is defined in math.h (in my case line 1088, gcc 13.2).

[Bug c/114181] issubnormal is a macro

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=114181

Andrew Pinski <pinskia at gcc dot gnu.org> changed:

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

--- Comment #4 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
(In reply to g.peterhoff from comment #3)
> Of course issubnormal is defined in math.h (in my case line 1088, gcc 13.2).

libstdc++'s cmath does:
```
#define _GLIBCXX_INCLUDE_NEXT_C_HEADERS
#include_next <math.h>
#undef _GLIBCXX_INCLUDE_NEXT_C_HEADERS

...

#undef fpclassify
#undef isfinite
#undef isinf
#undef isnan
#undef isnormal
#undef signbit
#undef isgreater
#undef isgreaterequal
#undef isless
#undef islessequal
#undef islessgreater
#undef isunordered
```


And then libstdc++'s math.h does:
```
#if !defined __cplusplus || defined _GLIBCXX_INCLUDE_NEXT_C_HEADERS
# include_next <math.h>
#else

#ifndef _GLIBCXX_MATH_H
#define _GLIBCXX_MATH_H 1

# include <cmath>
```

If you are implementing a cmath for a C++ implementation, you need to a similar
thing and `#undef` it.

The math.h that defines issubnormal comes from glibc.

[Bug c/114181] issubnormal is a macro

[g.peterhoff@t-online.de]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=114181

g.peterhoff@t-online.de changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
         Resolution|MOVED                       |FIXED

--- Comment #5 from g.peterhoff@t-online.de ---
> If you are implementing a cmath for a C++ implementation, you need to a similar thing and `#undef` it.
> The math.h that defines issubnormal comes from glibc.

That's what I mean. See also e.g.
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=77925
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=77926

[Bug c/114181] issubnormal is a macro

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=114181

Andrew Pinski <pinskia at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
         Resolution|FIXED                       |INVALID

--- Comment #6 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
.

[Bug c/114181] issubnormal is a macro

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=114181

--- Comment #7 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
C++ standard does not define a std::issubnormal yet.

[Bug c/114181] issubnormal is a macro

[g.peterhoff@t-online.de]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=114181

g.peterhoff@t-online.de changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
         Resolution|INVALID                     |FIXED

--- Comment #8 from g.peterhoff@t-online.de ---
Of course, std::issubnormal is not yet available at the moment. To be able to
implement this at all, issubnormal from math.h must not be a macro!

[Bug c/114181] issubnormal is a macro

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=114181

Andrew Pinski <pinskia at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
         Resolution|FIXED                       |INVALID

--- Comment #9 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
To implement it in libstdc++, you implement it the same way isnormal, etc. Is
done.

[Bug c/114181] issubnormal is a macro

[g.peterhoff@t-online.de]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=114181

g.peterhoff@t-online.de changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
         Resolution|INVALID                     |FIXED

--- Comment #10 from g.peterhoff@t-online.de ---
Exactly that does not work, because issubnormal is a simple macro.
Only if before the implementation
#undef issubnormal
is made before implementation: https://godbolt.org/z/z3PG3hYev

That is incorrect. 
* I have no idea what happens if math.h is already included somewhere and I
subsequently undefine issubnormal.
* It is not my job to program around any (compiler-specific) problems. The
compiler has to do it right or it doesn't support it at all.

Therefore issubnormal must be provided as a "real" function or via builtin.

[Bug c/114181] issubnormal is a macro

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=114181

Andrew Pinski <pinskia at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
         Resolution|FIXED                       |INVALID

--- Comment #11 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
In this case gcc has nothing to do with this. Since defining issubnormal inside
std is undefined behavior there is nothing to be done.

Also you don't need a builtin since fpclassify is there already. 

When issubnormal is added to the c++ standard, libstdc++'s will add it too.

[Bug c/114181] issubnormal is a macro

[g.peterhoff@t-online.de]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=114181

g.peterhoff@t-online.de changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
         Resolution|INVALID                     |FIXED

--- Comment #12 from g.peterhoff@t-online.de ---
If this comes into the C++ standard I would have to rewrite it anyway. Why not
now that I have reported this error?

Are there already plans how to deal with the
https://en.cppreference.com/w/c/experimental/fpext1
https://en.cppreference.com/w/c/experimental/fpext4
regarding C++?

thx
Gero

[Bug c/114181] issubnormal is a macro

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=114181

Andrew Pinski <pinskia at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
         Resolution|FIXED                       |INVALID

--- Comment #13 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
(In reply to g.peterhoff from comment #12) 
> Are there already plans how to deal with the
> https://en.cppreference.com/w/c/experimental/fpext1
> https://en.cppreference.com/w/c/experimental/fpext4
> regarding C++?

Well dfp is supported, or kinda of supported already for GCC, see PR 51364
also.

as far as fpext4 is concern, I suspect it might be better to take it up with
the C++ standards committee ...

[Bug c/114181] issubnormal is a macro

[redi at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=114181

--- Comment #14 from Jonathan Wakely <redi at gcc dot gnu.org> ---
Libc headers define lots of macros, we don't #undef them in C++ headers unless
they use the name of a function defined by the C++ library (because C++ says
its library functions must be real functions and must not be hidden by macros).

If you want to provide an implementation of issubnormal (e.g. as
boost::issubnormal) then you need to do:

#include <math.h>
#undef issubnormal

I don't think it's a libstdc++ bug that we don't #undef non-standard macros.

It's certainly not a GCC component=c bug that Glibc defines a macro in its
<math.h>.

So there's no GCC bug here.

Like Andrew said, iff we need to define std::issubnormal then we'll #undef it,
but we're not going to do that until we need to.

[Bug c/114181] issubnormal is a macro

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=114181

Jakub Jelinek <jakub at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |jakub at gcc dot gnu.org

--- Comment #15 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
You could as well use namespace std { bool (issubnormal) (...); } if you don't
want macro expansion for it.