[Bug tree-optimization/109735] New: [14 Regression] ICE in vectorizable_store, at tree-vect-stmts.cc:8529 since r14-322-g821ef93976e750

[Bug tree-optimization/109735] New: [14 Regression] ICE in vectorizable_store, at tree-vect-stmts.cc:8529 since r14-322-g821ef93976e750

[marxin at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=109735

            Bug ID: 109735
           Summary: [14 Regression] ICE in vectorizable_store, at
                    tree-vect-stmts.cc:8529 since r14-322-g821ef93976e750
           Product: gcc
           Version: 14.0
            Status: UNCONFIRMED
          Keywords: ice-on-valid-code
          Severity: normal
          Priority: P3
         Component: tree-optimization
          Assignee: unassigned at gcc dot gnu.org
          Reporter: marxin at gcc dot gnu.org
                CC: rguenth at gcc dot gnu.org
  Target Milestone: ---
              Host: x86_64-linux-gnu
            Target: riscv64-unknown-linux-gnu

The following is crashing:

$ ./xgcc -B.
/home/marxin/Programming/gcc/gcc/testsuite/gcc.dg/tree-ssa/pr83329.c
-fvect-cost-model=unlimited -Ofast -c
during GIMPLE pass: slp
/home/marxin/Programming/gcc/gcc/testsuite/gcc.dg/tree-ssa/pr83329.c: In
function ‘h’:
/home/marxin/Programming/gcc/gcc/testsuite/gcc.dg/tree-ssa/pr83329.c:11:6:
internal compiler error: in vectorizable_store, at tree-vect-stmts.cc:8513
   11 | void h() {
      |      ^
0x2b97b57 vectorizable_store
        /home/marxin/Programming/gcc2/gcc/tree-vect-stmts.cc:8513
0x2ba28ee vect_transform_stmt(vec_info*, _stmt_vec_info*,
gimple_stmt_iterator*, _slp_tree*, _slp_instance*)
        /home/marxin/Programming/gcc2/gcc/tree-vect-stmts.cc:11595
0x2c0167f vect_schedule_slp_node
        /home/marxin/Programming/gcc2/gcc/tree-vect-slp.cc:8952
0x2c0217d vect_schedule_scc
        /home/marxin/Programming/gcc2/gcc/tree-vect-slp.cc:9147
0x2c0297b vect_schedule_slp(vec_info*, vec<_slp_instance*, va_heap, vl_ptr>
const&)
        /home/marxin/Programming/gcc2/gcc/tree-vect-slp.cc:9287
0x2bfc2fa vect_slp_region
        /home/marxin/Programming/gcc2/gcc/tree-vect-slp.cc:7496
0x2bfcb70 vect_slp_bbs
        /home/marxin/Programming/gcc2/gcc/tree-vect-slp.cc:7608
0x2bfd100 vect_slp_function(function*)
        /home/marxin/Programming/gcc2/gcc/tree-vect-slp.cc:7709
0x2c2055c execute
        /home/marxin/Programming/gcc2/gcc/tree-vectorizer.cc:1529
Please submit a full bug report, with preprocessed source (by using
-freport-bug).
Please include the complete backtrace with any bug report.
See <https://gcc.gnu.org/bugs/> for instructions.

[Bug tree-optimization/109735] [14 Regression] ICE in vectorizable_store, at tree-vect-stmts.cc:8529 since r14-322-g821ef93976e750

[marxin at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=109735

Martin Liška <marxin at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Assignee|unassigned at gcc dot gnu.org      |rguenth at gcc dot gnu.org
   Last reconfirmed|                            |2023-05-04
     Ever confirmed|0                           |1
             Status|UNCONFIRMED                 |ASSIGNED
   Target Milestone|---                         |14.0

[Bug tree-optimization/109735] [14 Regression] ICE in vectorizable_store, at tree-vect-stmts.cc:8529 since r14-322-g821ef93976e750

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=109735

--- Comment #1 from Richard Biener <rguenth at gcc dot gnu.org> ---
I will have a look.

[Bug tree-optimization/109735] [14 Regression] ICE in vectorizable_store, at tree-vect-stmts.cc:8529 since r14-322-g821ef93976e750

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=109735

--- Comment #2 from Richard Biener <rguenth at gcc dot gnu.org> ---
We have a sign mismatch here from vectorizing a POINTER_MINUS_EXPR.

(gdb) p debug ((slp_tree)0x39c2970)
pr83329.c:11:6: note: node 0x39c2970 (max_nunits=1, refcnt=1) vector(1) int
pr83329.c:11:6: note: op template: _2 = &g - &f;
pr83329.c:11:6: note:   stmt 0 _2 = &g - &f;
pr83329.c:11:6: note:   stmt 1 _3 = &e - &f;
pr83329.c:11:6: note:   children 0x39c2a00 0x39c2a90

but

(gdb) p debug_gimple_stmt (0x7ffff6fc3dc0)
_24 = VIEW_CONVERT_EXPR<vector(1) unsigned int>(_23);

there's code to handle this but it's conditional on a special case.  The fix
is to invoke it generally I think.

[Bug tree-optimization/109735] [14 Regression] ICE in vectorizable_store, at tree-vect-stmts.cc:8529 since r14-322-g821ef93976e750

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=109735

--- Comment #3 from CVS Commits <cvs-commit at gcc dot gnu.org> ---
The master branch has been updated by Richard Biener <rguenth@gcc.gnu.org>:

https://gcc.gnu.org/g:37b5f031292fbdb854ee791de3883362cf2afcff

commit r14-494-g37b5f031292fbdb854ee791de3883362cf2afcff
Author: Richard Biener <rguenther@suse.de>
Date:   Fri May 5 08:54:28 2023 +0200

    tree-optimization/109735 - conversion for vectorized pointer-diff

    There's handling in vectorizable_operation for POINTER_DIFF_EXPR
    requiring conversion of the result of the unsigned operation to
    a signed type.  But that's conditional on the "default" kind of
    vectorization.  In this PR it's shown the emulated vector path
    needs it and I think the masked operation case will, too (though
    we might eventually never mask an integral MINUS_EXPR).  So the
    following makes that handling unconditional.

            PR tree-optimization/109735
            * tree-vect-stmts.cc (vectorizable_operation): Perform
            conversion for POINTER_DIFF_EXPR unconditionally.

[Bug tree-optimization/109735] [14 Regression] ICE in vectorizable_store, at tree-vect-stmts.cc:8529 since r14-322-g821ef93976e750

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=109735

Richard Biener <rguenth at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
         Resolution|---                         |FIXED
             Status|ASSIGNED                    |RESOLVED

--- Comment #4 from Richard Biener <rguenth at gcc dot gnu.org> ---
Fixed.