[Bug c++/101795] New: (x > QNaNf) is not a constant expression

[Bug c++/101795] New: (x > QNaNf) is not a constant expression

[jonas.rahlf.basf at gmail dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=101795

            Bug ID: 101795
           Summary: (x > QNaNf) is not a constant expression
           Product: gcc
           Version: 11.2.1
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: c++
          Assignee: unassigned at gcc dot gnu.org
          Reporter: jonas.rahlf.basf at gmail dot com
  Target Milestone: ---

Somehow (x > QNaNf) is not a constant expression, according to g++. I could not
find anything in the c++ standard that would support this behaviour.

Both clang and MSVC compile (x > QNaNf) as a constant expression.
Interestingly, (x == QNaNf) seems to work fine with g++.


Example:
#include <limits>
constexpr bool canCompile = std::numeric_limits<float>::quiet_NaN() > 1;

Compiler output: error: '(+QNaNf > 1.0e+0f)' is not a constant expression
See also: https://godbolt.org/z/vW651jPYh

No extra compile options needed. 
I noticed this when attempting to compile std::lerp(a,b,NaN) (which is supposed
to be constexpr) as a constant expression.

[Bug c++/101795] (x > QNaNf) is not a constant expression

[glisse at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=101795

--- Comment #1 from Marc Glisse <glisse at gcc dot gnu.org> ---
Hint: -fno-trapping-math lets it compile.
It should probably be accepted in a manifestly_const_eval context, although
some in the committee wanted to prevent the use of NaN (and sometimes even
infinity!) in constant expressions...

[Bug c++/101795] (x > QNaNf) is not a constant expression

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=101795

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Keywords|                            |needs-bisection

--- Comment #2 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
Seems to be fixed in GCC 12+.

[Bug c++/101795] (x > QNaNf) is not a constant expression

[marxin at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=101795

Martin Liška <marxin at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |sayle at gcc dot gnu.org
           Keywords|needs-bisection             |

--- Comment #3 from Martin Liška <marxin at gcc dot gnu.org> ---
Fixed since r12-3660-ge9e46864cd0695d2.

[Bug c++/101795] (x > QNaNf) is not a constant expression

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=101795

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

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

--- Comment #4 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
Dup of bug 88173.

*** This bug has been marked as a duplicate of bug 88173 ***