[PATCH] D72906: [X86] Improve X86 cmpps/cmppd/cmpss/cmpsd intrinsics with strictfp

[Ulrich Weigand via Phabricator via cfe-commits]

uweigand added a comment.

In D72906#1837905 <https://reviews.llvm.org/D72906#1837905>, @craig.topper wrote:

> In D72906#1826849 <https://reviews.llvm.org/D72906#1826849>, @uweigand wrote:
>
> > In D72906#1826122 <https://reviews.llvm.org/D72906#1826122>, @craig.topper wrote:
> >
> > > In D72906#1826061 <https://reviews.llvm.org/D72906#1826061>, @uweigand wrote:
> > >
> > > > > The constrained fcmp intrinsics don't allow the TRUE/FALSE predicates.
> > > >
> > > > Hmm, maybe they should then?   The only reason I didn't add them initially was that I wasn't sure they were useful for anything; if they are, it should be straightforward to add them back.
> > >
> > >
> > > What would we lower it to on a target that doesn’t support it natively?
> >
> >
> > Any supported compare (quiet or signaling as appropriate, just so we get the correct exceptions), and then ignore the result (and use true/false constant result instead)?
>
>
> Sure. Is that something we want to force all targets to have to implement just to handle this case for X86? Unless we can come up with a generic DAG combine to pick a valid condition alternate so that the lowering code for each target doesn't have to deal with it.


Hmm, OK.  Given that this X86-specific builtin code seems to be the only place in clang where FCMP_TRUE/FCMP_FALSE can ever get emitted anyway, I guess you might as well implement the workaround (use some other compare to get the exception) right there, just for X86.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D72906/new/

https://reviews.llvm.org/D72906