[Bug target/111814] [SH] __builtin_nan* returns signalling NaNs instead of quiet NaNs and vice versa

[olegendo at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=111814

--- Comment #7 from Oleg Endo <olegendo at gcc dot gnu.org> ---
(In reply to Jiaxun Yang from comment #6)
> (In reply to Andrew Pinski from comment #4)
> > Note it might need to be under the check of -mieee too ....
> 
> I'm preparing patch for this issue 

Thanks!

> and IMO this should not be under -mieee,
> but applied unconditionally to whole architecture.
> 
> We are actually not IEEE754-2008 compliant by having reversed signalling bit
> and slightly different NaN propagation rules...
> 
> Other architectures with such feature like hppa and MIPS are all doing it
> unconditionally as well, also note that unlike other -mieee features this
> won't come with any performance penalty.

Sounds good.