[PATCH] D75889: AMDGPU: Use V_MAC_F32 for fmad.ftz

Matt Arsenault via Phabricator via llvm-commits llvm-commits at lists.llvm.org
Tue Mar 10 14:49:22 PDT 2020


arsenm added a comment.

In D75889#1915780 <https://reviews.llvm.org/D75889#1915780>, @rampitec wrote:

> In D75889#1915712 <https://reviews.llvm.org/D75889#1915712>, @arsenm wrote:
>
> > In D75889#1915703 <https://reviews.llvm.org/D75889#1915703>, @rampitec wrote:
> >
> > > In D75889#1915688 <https://reviews.llvm.org/D75889#1915688>, @arsenm wrote:
> > >
> > > > In D75889#1915394 <https://reviews.llvm.org/D75889#1915394>, @rampitec wrote:
> > > >
> > > > > Please run PSDB here anyway as if there are problems conformance should catch it.
> > > >
> > > >
> > > > It passes, but there's no way we have anything testing the f16 path for this
> > >
> > >
> > > There are half tests there, I suppose mad should be covered. LGTM.
> >
> >
> > We don't have tests that run with f16 denormals enabled. We also don't use the f16 intrinsic for div expansion (although I suppose we could in some cases?)
>
>
> f16 denormals are always enabled.


I

In D75889#1915780 <https://reviews.llvm.org/D75889#1915780>, @rampitec wrote:

> In D75889#1915712 <https://reviews.llvm.org/D75889#1915712>, @arsenm wrote:
>
> > In D75889#1915703 <https://reviews.llvm.org/D75889#1915703>, @rampitec wrote:
> >
> > > In D75889#1915688 <https://reviews.llvm.org/D75889#1915688>, @arsenm wrote:
> > >
> > > > In D75889#1915394 <https://reviews.llvm.org/D75889#1915394>, @rampitec wrote:
> > > >
> > > > > Please run PSDB here anyway as if there are problems conformance should catch it.
> > > >
> > > >
> > > > It passes, but there's no way we have anything testing the f16 path for this
> > >
> > >
> > > There are half tests there, I suppose mad should be covered. LGTM.
> >
> >
> > We don't have tests that run with f16 denormals enabled. We also don't use the f16 intrinsic for div expansion (although I suppose we could in some cases?)
>
>
> f16 denormals are always enabled.


I meant disabled. They're always enabled, so we have no useful tests for the case where they're forced off


CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D75889/new/

https://reviews.llvm.org/D75889





More information about the llvm-commits mailing list