[Bug tree-optimization/82518] [8 regression] gfortran.fortran-torture/execute/in-pack.f90 fails on armeb since r252917

[aldyh at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=82518

--- Comment #30 from Aldy Hernandez <aldyh at gcc dot gnu.org> ---
(In reply to Christophe Lyon from comment #29)
> I still haven't found a commit where the test passes with
> -fno-vect-cost-model (before -O3).
> 
> I went back to r193053 (Nov 1, 2012), where I was able to build GCC but the
> test fails.
> With a revision 1 month earlier, the GCC fails to build.
> I tried with earlier revision, but soon reached a point where
> "-mfloat-abi=hard and VFP" is not implemented.

Ok, bisecting isn't getting us anywhere.  I apologize for making you going
through all this in vain.

Could you further reduce the testcase as I suggested in comment 12, and perhaps
we can start looking at the assembly to see where things are going wrong?