[Bug middle-end/90796] [8/9 Regression] GCC: O2 vs O3 output differs on simple test

[matz at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=90796

--- Comment #17 from Michael Matz <matz at gcc dot gnu.org> ---
Author: matz
Date: Wed Nov 20 16:51:10 2019
New Revision: 278512

URL: https://gcc.gnu.org/viewcvs?rev=278512&root=gcc&view=rev
Log:
Fix PR90796

       PR middle-end/90796
       * gimple-loop-jam.c (any_access_function_variant_p): New function.
       (adjust_unroll_factor): Use it to constrain safety, new parameter.
       (tree_loop_unroll_and_jam): Adjust call and profitable unroll factor.

testsuite/
       Backport from mainline
       PR middle-end/90796
       * gcc.dg/unroll-and-jam.c: Disable loop-invariant motion and adjust.

       PR middle-end/90796
       * gcc.dg/unroll-and-jam.c: Add three invalid and one valid case.

Modified:
    branches/gcc-9-branch/gcc/ChangeLog
    branches/gcc-9-branch/gcc/gimple-loop-jam.c
    branches/gcc-9-branch/gcc/testsuite/ChangeLog
    branches/gcc-9-branch/gcc/testsuite/gcc.dg/unroll-and-jam.c