[PATCH] S/390: PR target/79240: Fix assertion in s390_extzv_shift_ok.

[Dominik Vogt]

On Thu, Jan 26, 2017 at 05:33:18PM +0100, Jakub Jelinek wrote:
> On Thu, Jan 26, 2017 at 05:11:27PM +0100, Dominik Vogt wrote:
> > The patch fixes the s390x crash reported in PR 79240:
> > https://gcc.gnu.org/bugzilla/show_bug.cgi?id=79240
> > Regression tested and bootstrapped on s390x biarch and s390.
> 
> > gcc/ChangeLog-pr79240
> > 
> > 	PR target/79240
> > 	* config/s390/s390.c (s390_extzv_shift_ok): Returns false, don't assert.
> > 	* config/s390/s390.md ("*r<noxa>sbg_<mode>_srl_bitmask")
> > 	("*r<noxa>sbg_<mode>_sll_bitmask")
> > 	("*extzv_<mode>_srl<clobbercc_or_nocc>")
> > 	("*extzv_<mode>_sll<clobbercc_or_nocc>"):
> > 	Use contiguous_bitmask_nowrap_operand
> > gcc/testsuite/ChangeLog-pr79240
> > 
> > 	* gcc.target/s390/pr79240.c: New test.
> 
> > diff --git a/gcc/config/s390/s390.c b/gcc/config/s390/s390.c
> > index fe65846..b1dfbfa 100644
> > --- a/gcc/config/s390/s390.c
> > +++ b/gcc/config/s390/s390.c
> > @@ -2472,7 +2472,8 @@ s390_extzv_shift_ok (int bitsize, int rotl, unsigned HOST_WIDE_INT contig)
> >    bool ok;
> >  
> >    ok = s390_contiguous_bitmask_nowrap_p (contig, bitsize, &start, &end);
> > -  gcc_assert (ok);
> > +  if (!ok)
> > +    return false;
> 
> If the predicates are supposed to ensure it, then I think the assert is
> fine.

Is it guaranteed that the predicate conditions are evaluated
before executing the conditions?

Ciao

Dominik ^_^  ^_^

-- 

Dominik Vogt
IBM Germany