[Bug target/88892] [8 Regression] Double-to-float conversion uses wrong rounding mode when followed by memcpy

[segher at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=88892

--- Comment #15 from Segher Boessenkool <segher at gcc dot gnu.org> ---
Not in 8.2, no, that was released half a year ago already.  But 8.3, yes, it is
lined up for that.