[Bug target/89175] gcc's conversion code from double to unsigned int handles overflows incorrectly on x86-64

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=89175

Richard Biener <rguenth at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
   Last reconfirmed|                            |2019-02-04
                 CC|                            |jsm28 at gcc dot gnu.org,
                   |                            |uros at gcc dot gnu.org
     Ever confirmed|0                           |1

--- Comment #1 from Richard Biener <rguenth at gcc dot gnu.org> ---
Confirmed.  GCC defines __STDC_IEC_559__ and F.4 suggests this conversion
should raise INVALID.

I suspect most targets get this "wrong" for speed reasons in case they
have no HW support for the conversion to the target type.

Note that I would not like to have this "corrected" by default due
to speed and code-size implications...