[Bug tree-optimization/24227] New: ICE in compare_values, at tree-vrp.c:415

rguenth at gcc dot gnu dot org gcc-bugzilla@gcc.gnu.org
Thu Oct 6 08:49:00 GMT 2005


/usr/lib/gcc/i586-suse-linux/4.1.0/cc1 -fpreprocessed window.i -quiet -dumpbase
window.c -march=i586 -mtune=i686 -auxbase window -O2 -Wall -Wno-pointer-sign
-version -fmessage-length=0 -fno-strict-aliasing -o window.s
window.c: In function ‘Fdisplay_buffer’:
window.c:2921: internal compiler error: in compare_values, at tree-vrp.c:415
Please submit a full bug report,
with preprocessed source if appropriate.
See <URL:http://www.suse.de/feedback> for instructions.

#0  fancy_abort (file=0x87533f0 "../../../src/gcc/gcc/tree-vrp.c", line=415,
    function=0x8753492 "compare_values") at diagnostic.c:590
#1  0x085376cb in compare_values (val1=0x406248f0, val2=0x4062498c)
    at tree-vrp.c:414
#2  0x0853b03b in compare_ranges (comp=EQ_EXPR, vr0=0xbfae1b00, vr1=0xbfae1af0)
    at tree-vrp.c:1790
#3  0x08540424 in compare_names (comp=EQ_EXPR, n1=0x406248f0, n2=0x4062498c)
    at tree-vrp.c:3143
#4  0x085408c4 in vrp_evaluate_conditional (cond=0x4052f4a4, 
    use_equiv_p=1 '\001') at tree-vrp.c:3222
#5  0x0819dd68 in fold_predicate_in (stmt=0x4051ae88)
    at tree-ssa-propagate.c:1037
#6  0x0819e05e in substitute_and_fold (prop_value=0x897a9b0,
    use_ranges_p=1 '\001') at tree-ssa-propagate.c:1121
#7  0x0854374f in vrp_finalize () at tree-vrp.c:3958
#8  0x08543873 in execute_vrp () at tree-vrp.c:4028
#9  0x0852f0cd in execute_one_pass (pass=0x8777180) at passes.c:827

414       gcc_assert (POINTER_TYPE_P (TREE_TYPE (val1))
(gdb) call debug_tree(val1)
 <ssa_name 0x406248f0
    type <integer_type 0x4019d2e0 unsigned int sizes-gimplified public unsigned
SI
        size <integer_cst 0x4018d408 constant invariant 32>
        unit size <integer_cst 0x4018d198 constant invariant 4>
        align 32 symtab 0 alias set -1 precision 32 min <integer_cst 0x4018d480
0> max <integer_cst 0x4018d468 4294967295>
        pointer_to_this <pointer_type 0x403272e0>>
    visited var <var_decl 0x40527cb8 D.10670> def_stmt <modify_expr 0x4052f60c>
    version 328>


(gdb) up 3
#4  0x085408c4 in vrp_evaluate_conditional (cond=0x4052f4a4,
    use_equiv_p=1 '\001') at tree-vrp.c:3222
3222                return compare_names (TREE_CODE (cond), op0, op1);
(gdb) call debug_tree(cond)
 <eq_expr 0x4052f4a4
    type <boolean_type 0x4019d4ac _Bool public unsigned QI
        size <integer_cst 0x4018d210 constant invariant 8>
        unit size <integer_cst 0x4018d228 constant invariant 1>
        align 8 symtab 0 alias set -1 precision 1 min <integer_cst 0x4018d600
0> max <integer_cst 0x4018d630 1>>

    arg 0 <ssa_name 0x406248f0
        type <integer_type 0x4019d2e0 unsigned int sizes-gimplified public
unsigned SI
            size <integer_cst 0x4018d408 constant invariant 32>
            unit size <integer_cst 0x4018d198 constant invariant 4>
            align 32 symtab 0 alias set -1 precision 32 min <integer_cst
0x4018d480 0> max <integer_cst 0x4018d468 4294967295>
            pointer_to_this <pointer_type 0x403272e0>>
        visited var <var_decl 0x40527cb8 D.10670> def_stmt <modify_expr
0x4052f60c>
        version 328>
    arg 1 <ssa_name 0x4062498c
        type <pointer_type 0x40347cf0 type <record_type 0x40347c94 buffer>
            sizes-gimplified public unsigned SI size <integer_cst 0x4018d408
32> unit size <integer_cst 0x4018d198 4>
            align 32 symtab 0 alias set -1>
        visited var <var_decl 0x40527d10 D.10671> def_stmt <modify_expr
0x4052f678>
        version 331
        ptr-info 0x40780240>>

Looks like we start with mismatched types in the EQ_EXPR.


-- 
           Summary: ICE in compare_values, at tree-vrp.c:415
           Product: gcc
           Version: 4.1.0
            Status: UNCONFIRMED
          Keywords: ice-on-invalid-code
          Severity: normal
          Priority: P2
         Component: tree-optimization
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: rguenth at gcc dot gnu dot org


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=24227



More information about the Gcc-bugs mailing list