On Mon, 15 Feb 2016, Richard Biener wrote:
> On Sun, 14 Feb 2016, Marc Glisse wrote:
>
> > On Tue, 2 Feb 2016, Richard Biener wrote:
> >
> > > *** gcc/match.pd (revision 233067)
> > > --- gcc/match.pd (working copy)
> > > *** DEFINE_INT_AND_FLOAT_ROUND_FN (RINT)
> > > *** 2094,2099
On Sun, 14 Feb 2016, Marc Glisse wrote:
> On Tue, 2 Feb 2016, Richard Biener wrote:
>
> > *** gcc/match.pd(revision 233067)
> > --- gcc/match.pd(working copy)
> > *** DEFINE_INT_AND_FLOAT_ROUND_FN (RINT)
> > *** 2094,2099
> > --- 2094,2117
> > (bit_and:c (ordered @
On Tue, 2 Feb 2016, Richard Biener wrote:
*** gcc/match.pd(revision 233067)
--- gcc/match.pd(working copy)
*** DEFINE_INT_AND_FLOAT_ROUND_FN (RINT)
*** 2094,2099
--- 2094,2117
(bit_and:c (ordered @0 @0) (ordered:c@2 @0 @1))
@2)
+ /* Simple range test si
The following is the minimal approach to fix this -Warray-bound at
this stage. We're pretty bad at optimizing range tests that
get optimizable during GIMPLE only (fold can handle quite some cases).
Too bad to optimize this before VRP gets to warn about out-of-bound
array accesses. Later reassoc
