[Bug tree-optimization/61688] New: [4.10 Regression] ICE: tree check: expected record_type or union_type or qual_union_type, have array_type in ipa_binfo_from_known_type_jfunc

[Bug tree-optimization/61688] New: [4.10 Regression] ICE: tree check: expected record_type or union_type or qual_union_type, have array_type in ipa_binfo_from_known_type_jfunc

[rmansfield at qnx dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=61688

            Bug ID: 61688
           Summary: [4.10 Regression] ICE: tree check: expected
                    record_type or union_type or qual_union_type, have
                    array_type in ipa_binfo_from_known_type_jfunc
           Product: gcc
           Version: 4.10.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: tree-optimization
          Assignee: unassigned at gcc dot gnu.org
          Reporter: rmansfield at qnx dot com

Created attachment 33062
  --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=33062&action=edit
preprocessed src

$ ./xgcc -B. ~/ice.ii -c -O2
/home/ryan/ice.ii:30:27: warning: inline function 'UBool
icu::UnicodeString::isEmpty() const' used but never defined
        inline UBool       isEmpty (void) const;
                           ^
/home/ryan/ice.ii: In member function 'void
icu::DateIntervalFormat::concatSingleDate2TimeInterval(const UChar*,
std::int32_t, const icu::UnicodeString&, UCalendarDateFields, UErrorCode&)':
/home/ryan/ice.ii:58:15: internal compiler error: tree check: expected
record_type or union_type or qual_union_type, have array_type in
ipa_binfo_from_known_type_jfunc, at ipa-prop.c:555
      void     DateIntervalFormat::concatSingleDate2TimeInterval (const UChar *
format,              int32_t formatLen,              const UnicodeString &     
        datePattern,              UCalendarDateFields              field,      
       UErrorCode & status)     {
               ^
0xd98154 tree_check_failed(tree_node const*, char const*, int, char const*,
...)
    ../../gcc/tree.c:9220
0x9e1327 tree_check3
    ../../gcc/tree.h:2771
0x9e1327 ipa_binfo_from_known_type_jfunc(ipa_jump_func*)
    ../../gcc/ipa-prop.c:555
0x9e1eba ipa_intraprocedural_devirtualization(gimple_statement_base*)
    ../../gcc/ipa-prop.c:2467
0xcb7f8f eliminate_dom_walker::before_dom_children(basic_block_def*)
    ../../gcc/tree-ssa-pre.c:4365
0xfca602 dom_walker::walk(basic_block_def*)
    ../../gcc/domwalk.c:177
0xcb6412 eliminate
    ../../gcc/tree-ssa-pre.c:4493
0xcb6693 execute
    ../../gcc/tree-ssa-pre.c:4910
Please submit a full bug report,
with preprocessed source if appropriate.
Please include the complete backtrace with any bug report.
See <http://gcc.gnu.org/bugs.html> for instructions.

gcc version 4.10.0 20140702 (experimental) [trunk revision 212230] (GCC)

[Bug ipa/61688] [4.10 Regression] ICE: tree check: expected record_type or union_type or qual_union_type, have array_type in ipa_binfo_from_known_type_jfunc

[trippels at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=61688

Markus Trippelsdorf <trippels at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
   Last reconfirmed|                            |2014-07-03
                 CC|                            |trippels at gcc dot gnu.org
          Component|tree-optimization           |ipa
   Target Milestone|---                         |4.10.0
     Ever confirmed|0                           |1

--- Comment #1 from Markus Trippelsdorf <trippels at gcc dot gnu.org> ---
Further reduced:

class UObject
{
public:
  virtual ~UObject ();
};
class A : UObject
{
public:
  A ();
};
void fn1 () { A a[1]; }

[Bug ipa/61688] [4.10 Regression] ICE: tree check: expected record_type or union_type or qual_union_type, have array_type in ipa_binfo_from_known_type_jfunc

[rmansfield at qnx dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=61688

--- Comment #3 from Ryan Mansfield <rmansfield at qnx dot com> ---
FWIW, it started happening for me at rev212222

[Bug ipa/61688] [4.10 Regression] ICE: tree check: expected record_type or union_type or qual_union_type, have array_type in ipa_binfo_from_known_type_jfunc

[trippels at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=61688

--- Comment #4 from Markus Trippelsdorf <trippels at gcc dot gnu.org> ---
(In reply to Ryan Mansfield from comment #3)
> FWIW, it started happening for me at rev212222

Yes, sorry I should have written: git blame points to r198088.
I didn't bisect.

[Bug ipa/61688] [4.10 Regression] ICE: tree check: expected record_type or union_type or qual_union_type, have array_type in ipa_binfo_from_known_type_jfunc

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=61688

Jakub Jelinek <jakub at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |hubicka at gcc dot gnu.org,
                   |                            |jakub at gcc dot gnu.org

--- Comment #5 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
Indeed, this is Honza's r212222.

[Bug ipa/61688] [4.10 Regression] ICE: tree check: expected record_type or union_type or qual_union_type, have array_type in ipa_binfo_from_known_type_jfunc

[burnus at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=61688

--- Comment #6 from Tobias Burnus <burnus at gcc dot gnu.org> ---
I think the PR has been fixed in the meanwhile. My guess is that the fix was
r212304.

[Bug ipa/61688] [5 Regression] ICE: tree check: expected record_type or union_type or qual_union_type, have array_type in ipa_binfo_from_known_type_jfunc

[hubicka at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=61688

Jan Hubicka <hubicka at gcc dot gnu.org> changed:

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

--- Comment #7 from Jan Hubicka <hubicka at gcc dot gnu.org> ---
Indeed, this one was fixed.