[Bug target/104467] [12 Regression] ICE: in store_constructor, at expr.cc:7199 with -mxop

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=104467

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|---                         |12.0
           Assignee|unassigned at gcc dot gnu.org      |rguenth at gcc dot gnu.org
     Ever confirmed|0                           |1
             Status|UNCONFIRMED                 |ASSIGNED
   Last reconfirmed|                            |2022-02-09

--- Comment #1 from Richard Biener <rguenth at gcc dot gnu.org> ---
Confirmed.  We have a V2DI CTOR with two V2DI elements created via

#0  0x0000000001df5d5d in gimple_simplify_BIT_FIELD_REF (
    res_op=0x7fffffffcdf0, seq=0x0, 
    valueize=0xfa2bcc <follow_all_ssa_edges(tree_node*)>, code=..., 
    type=<vector_type 0x7ffff66b0540>, _p0=<ssa_name 0x7ffff6522e10 3>, 
    _p1=<integer_cst 0x7ffff65370c0>, _p2=<integer_cst 0x7ffff6517dc8>)
    at gimple-match.cc:198464
#1  0x0000000001dfaa7b in gimple_simplify (res_op=0x7fffffffcdf0, seq=0x0, 
    valueize=0xfa2bcc <follow_all_ssa_edges(tree_node*)>, code=..., 
    type=<vector_type 0x7ffff66b0540>, _p0=<ssa_name 0x7ffff6522e10 3>, 
    _p1=<integer_cst 0x7ffff65370c0>, _p2=<integer_cst 0x7ffff6517dc8>)
    at gimple-match.cc:200018
#2  0x0000000001acccd8 in gimple_resimplify3 (seq=0x0, res_op=0x7fffffffd5d0, 
    valueize=0xfa2bcc <follow_all_ssa_edges(tree_node*)>)
    at /home/rguenther/src/gcc3/gcc/gimple-match-head.cc:397
#3  0x0000000001acd2af in gimple_match_op::resimplify (this=0x7fffffffd5d0, 
    seq=0x0, valueize=0xfa2bcc <follow_all_ssa_edges(tree_node*)>)
    at /home/rguenther/src/gcc3/gcc/gimple-match-head.cc:516
#4  0x0000000001df46b5 in gimple_simplify_BIT_FIELD_REF (
    res_op=0x7fffffffd5d0, seq=0x0, 
    valueize=0xfa2bcc <follow_all_ssa_edges(tree_node*)>, code=..., 
    type=<vector_type 0x7ffff66b0540>, _p0=<ssa_name 0x7ffff6522e58 4>, 
    _p1=<integer_cst 0x7ffff65370c0>, _p2=<integer_cst 0x7ffff6517dc8>)
    at gimple-match.cc:198207
#5  0x0000000001dfaa7b in gimple_simplify (res_op=0x7fffffffd5d0, seq=0x0, 
    valueize=0xfa2bcc <follow_all_ssa_edges(tree_node*)>, code=..., 
    type=<vector_type 0x7ffff66b0540>, _p0=<ssa_name 0x7ffff6522e58 4>, 
    _p1=<integer_cst 0x7ffff65370c0>, _p2=<integer_cst 0x7ffff6517dc8>)
    at gimple-match.cc:200018
#6  0x0000000001acccd8 in gimple_resimplify3 (seq=0x0, res_op=0x7fffffffd6d0, 
    valueize=0xfa2bcc <follow_all_ssa_edges(tree_node*)>)
    at /home/rguenther/src/gcc3/gcc/gimple-match-head.cc:397
#7  0x0000000001acd2af in gimple_match_op::resimplify (this=0x7fffffffd6d0, 
    seq=0x0, valueize=0xfa2bcc <follow_all_ssa_edges(tree_node*)>)
    at /home/rguenther/src/gcc3/gcc/gimple-match-head.cc:516
#8  0x00000000017b48ff in tree_vec_extract (gsi=0x7fffffffda20, 
    type=<vector_type 0x7ffff66b0540>, t=<ssa_name 0x7ffff6522e58 4>, 
    bitsize=<integer_cst 0x7ffff65370c0>, bitpos=<integer_cst 0x7ffff6517dc8>)
    at /home/rguenther/src/gcc3/gcc/tree-vect-generic.cc:183
#9  0x00000000017b4a2c in do_binop (gsi=0x7fffffffda20, 
    inner_type=<vector_type 0x7ffff66b0540>, a=<ssa_name 0x7ffff6522e58 4>, 
--Type <RET> for more, q to quit, c to continue without paging--
    b=<ssa_name 0x7ffff6522ea0 5>, bitpos=<integer_cst 0x7ffff6517dc8>, 
    bitsize=<integer_cst 0x7ffff65370c0>, code=BIT_AND_EXPR, 
    type=<vector_type 0x7ffff66b00a8>)
    at /home/rguenther/src/gcc3/gcc/tree-vect-generic.cc:206
#10 0x00000000017b51d5 in expand_vector_piecewise (gsi=0x7fffffffda20, 
    f=0x17b49ba <do_binop(gimple_stmt_iterator*, tree, tree, tree, tree, tree,
tree_code, tree)>, type=<vector_type 0x7ffff66b00a8>, 
    inner_type=<vector_type 0x7ffff66b0540>, a=<ssa_name 0x7ffff6522e58 4>, 
    b=<ssa_name 0x7ffff6522ea0 5>, code=BIT_AND_EXPR, parallel_p=false, 
    ret_type=<vector_type 0x7ffff66b00a8>)
    at /home/rguenther/src/gcc3/gcc/tree-vect-generic.cc:337

I will have a look.