On 2/21/2022 3:55 AM, Richard Biener via Gcc-patches wrote:
On Mon, Feb 21, 2022 at 9:31 AM Roger Sayle wrote:
Hi Marc,
I'm assuming that the use (semantics) of a REDUC_PLUS expr allow the
reduction to be done in any order, for example the testcase requires
-ffast-math to allow the REDUC_PL
_1 from sNaN to qNaN and thus relies on
-fno-signalling-nans?
Otherwise OK.
Thanks,
Richard.
> Cheers,
> Roger
> --
>
> > -Original Message-
> > From: Marc Glisse
> > Sent: 21 February 2022 08:21
> > To: Roger Sayle
> > Cc: gcc-patches@gcc.gnu.org
&g
e
> Cc: gcc-patches@gcc.gnu.org
> Subject: Re: [PATCH] Implement constant-folding simplifications of
reductions.
>
> On Mon, 21 Feb 2022, Roger Sayle wrote:
>
> > +/* Fold REDUC (@0 op VECTOR_CST) as REDUC (@0) op REDUC
> (VECTOR_CST).
> > +*/ (for redu
On Mon, 21 Feb 2022, Roger Sayle wrote:
+/* Fold REDUC (@0 op VECTOR_CST) as REDUC (@0) op REDUC (VECTOR_CST). */
+(for reduc (IFN_REDUC_PLUS IFN_REDUC_MAX IFN_REDUC_MIN IFN_REDUC_FMAX
+IFN_REDUC_FMIN IFN_REDUC_AND IFN_REDUC_IOR IFN_REDUC_XOR)
+ op (plus max min IFN_FMAX IFN_FMI
This patch addresses a code quality regression in GCC 12 by implementing
some constant folding/simplification transformations for REDUC_PLUS_EXPR
in match.pd. The motivating example is gcc.dg/vect/pr89440.c which with
-O2 -ffast-math (with vectorization now enabled) gets optimized to:
float f (f
