[PR70920] transform (intptr_t) x eq/ne CST to x eq/ne (typeof x) cst

[Richard Biener]

On Thu, 4 Aug 2016, Prathamesh Kulkarni wrote:

> On 3 August 2016 at 17:27, Matthew Wahab <matthew.wahab@foss.arm.com> wrote:
> > On 29/07/16 15:32, Prathamesh Kulkarni wrote:
> >>
> >> On 29 July 2016 at 12:42, Richard Biener <rguenther@suse.de> wrote:
> >>>
> >>> On Fri, 29 Jul 2016, Prathamesh Kulkarni wrote:
> >>>
> >>>> On 28 July 2016 at 19:18, Richard Biener <rguenther@suse.de> wrote:
> >>>>>
> >>>>> On Thu, 28 Jul 2016, Prathamesh Kulkarni wrote:
> >>>>>
> >>>>>> On 28 July 2016 at 15:58, Andreas Schwab <schwab@suse.de> wrote:
> >>>>>>>
> >>>>>>> On Mo, Jul 25 2016, Prathamesh Kulkarni
> >>>>>>> <prathamesh.kulkarni@linaro.org> wrote:
> >>>>>>>
> >>>>>>>> diff --git a/gcc/testsuite/gcc.dg/pr70920-4.c
> >>>>>>>> b/gcc/testsuite/gcc.dg/pr70920-4.c
> >>>>>>>> new file mode 100644
> >>>>>>>> index 0000000..dedb895
> >>>>>>>> --- /dev/null
> >>>>>>>> +++ b/gcc/testsuite/gcc.dg/pr70920-4.c
> >>>>>>>> @@ -0,0 +1,21 @@
> >>>>>>>> +/* { dg-do compile } */
> >>>>>>>> +/* { dg-options "-O2 -fdump-tree-ccp-details
> >>>>>>>> -Wno-int-to-pointer-cast" } */
> >>>>>>>> +
> >>>>>>>> +#include <stdint.h>
> >>>>>>>> +
> >>>>>>>> +void f1();
> >>>>>>>> +void f2();
> >>>>>>>> +
> >>>>>>>> +void
> >>>>>>>> +foo (int a)
> >>>>>>>> +{
> >>>>>>>> +  void *cst = 0;
> >>>>>>>> +  if ((int *) a == cst)
> >>>>>>>> +    {
> >>>>>>>> +      f1 ();
> >>>>>>>> +      if (a)
> >>>>>>>> +     f2 ();
> >>>>>>>> +    }
> >>>>>>>> +}
> >>>>>>>> +
> >>>>>>>> +/* { dg-final { scan-tree-dump "gimple_simplified to if
> >>>>>>>> \\(_\[0-9\]* == 0\\)" "ccp1" } } */
> >>>>>>>
> >>>>>>>
> >>>>>>> This fails on all ilp32 platforms.
> >
> > [..]
> >>>
> >>>
> >>> I don't think just matching == 0 is a good idea.  I suggest to
> >>> restrict the testcase to lp64 targets and maybe add a ilp32 variant.
> >>
> >> Hi,
> >> I restricted the test-case to lp64 targets.
> >> Is this OK to commit ?
> >
> >
> > Hello,
> >
> > The test case is failing for arm-none-linux-gnueabihf.
> Oops, sorry about that.
> >
> > It is correctly skipped if the 'dg-require-effective-target lp64' you added
> > is moved to the end of the directives (after the dg-options).
> Indeed, it is skipped after moving to end.
> Is it OK to commit the attached patch ?

I believe the canonical place is after do-do but before dg-options.
Can you check if that works, too?

Richard.

> Thanks,
> Prathamesh
> >
> > Matthew
> >
> 

-- 
Richard Biener <rguenther@suse.de>
SUSE LINUX GmbH, GF: Felix Imendoerffer, Jane Smithard, Graham Norton, HRB 21284 (AG Nuernberg)