[Bug gold/14344] New: FAIL: gdb_index_test_[12].sh

http://sourceware.org/bugzilla/show_bug.cgi?id=14344

             Bug #: 14344
           Summary: FAIL: gdb_index_test_[12].sh
           Product: binutils
           Version: 2.23 (HEAD)
            Status: NEW
          Severity: normal
          Priority: P2
         Component: gold
        AssignedTo: ian@...
        ReportedBy: hjl.tools@...
                CC: ccoutant@...
    Classification: Unclassified


On Fedora 17/x86-64, I got

FAIL: gdb_index_test_1.sh
FAIL: gdb_index_test_2.sh

[hjl@gnu-6 testsuite]$ grep "::c2" gdb_index_test_*.stdout | grep const
gdb_index_test_1.stdout:[139] two::c2<const int*>::c2: 0 [no symbol
information]
gdb_index_test_1.stdout:[326] check<two::c2<const int*> >: 0 [no symbol
information]
gdb_index_test_1.stdout:[468] two::c2<const int*>::val: 0 [no symbol
information]
gdb_index_test_1.stdout:[616] two::c2<const int*>::~c2: 0 [no symbol
information]
gdb_index_test_1.stdout:[752] two::c2<const int*>: 0 [no symbol information]
gdb_index_test_2.stdout:[139] two::c2<const int*>::c2: 0 [no symbol
information]
gdb_index_test_2.stdout:[326] check<two::c2<const int*> >: 0 [no symbol
information]
gdb_index_test_2.stdout:[468] two::c2<const int*>::val: 0 [no symbol
information]
gdb_index_test_2.stdout:[616] two::c2<const int*>::~c2: 0 [no symbol
information]
gdb_index_test_2.stdout:[752] two::c2<const int*>: 0 [no symbol information]
[hjl@gnu-6 testsuite]$

But

FAIL: gdb_index_test_1.sh (exit: 1)
===================================

Did not find expected output:
   ^\[ *[0-9]*\] two::c2<int const\*>:

The difference is "const int" vs "int const".

--
Configure bugmail: http://sourceware.org/bugzilla/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

_______________________________________________
bug-binutils mailing list
bug-binutils@...
https://lists.gnu.org/mailman/listinfo/bug-binutils

http://sourceware.org/bugzilla/show_bug.cgi?id=14344

--- Comment #1 from H.J. Lu <hjl.tools at gmail dot com> 2012-07-10 06:54:24 UTC ---
The difference is GCC 4.6 vs GCC 4.7. GCC 4.7 mangles "const int"
as "const int" while GCC 4.6 uses "int const".

--
Configure bugmail: http://sourceware.org/bugzilla/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

_______________________________________________
bug-binutils mailing list
bug-binutils@...
https://lists.gnu.org/mailman/listinfo/bug-binutils

http://sourceware.org/bugzilla/show_bug.cgi?id=14344

--- Comment #2 from H.J. Lu <hjl.tools at gmail dot com> 2012-07-10 06:59:10 UTC ---
I think tests should accept both "const int" and "int const".

--
Configure bugmail: http://sourceware.org/bugzilla/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

_______________________________________________
bug-binutils mailing list
bug-binutils@...
https://lists.gnu.org/mailman/listinfo/bug-binutils

http://sourceware.org/bugzilla/show_bug.cgi?id=14344

--- Comment #3 from Cary Coutant <ccoutant at google dot com> 2012-07-10 08:15:28 UTC ---
> I think tests should accept both "const int" and "int const".

The point of the test is to make sure that it's in the canonical form
expected by GDB. For GCC 4.6, perhaps we should just treat this as an
XFAIL, or maybe have a configure check to turn off the test.

-cary

--
Configure bugmail: http://sourceware.org/bugzilla/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

_______________________________________________
bug-binutils mailing list
bug-binutils@...
https://lists.gnu.org/mailman/listinfo/bug-binutils

http://sourceware.org/bugzilla/show_bug.cgi?id=14344

--- Comment #4 from H.J. Lu <hjl.tools at gmail dot com> 2012-07-13 17:09:24 UTC ---
I got different failures in GCC 4.1.2 from Red Hat:

FAIL: gdb_index_test_1.sh (exit: 1)
===================================

Did not find expected output:
   ^\[ *[0-9]*\] anonymous_union_var:

FAIL: gdb_index_test_2.sh (exit: 1)
===================================

Did not find expected output:
   ^\[ *[0-9]*\] anonymous_union_var:

--
Configure bugmail: http://sourceware.org/bugzilla/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

_______________________________________________
bug-binutils mailing list
bug-binutils@...
https://lists.gnu.org/mailman/listinfo/bug-binutils

http://sourceware.org/bugzilla/show_bug.cgi?id=14344

--- Comment #5 from H.J. Lu <hjl.tools at gmail dot com> 2012-07-13 17:36:43 UTC ---
Created attachment 6527
  --> http://sourceware.org/bugzilla/attachment.cgi?id=6527
A testcase

[hjl@gnu-6 pr14344]$ nm  --demangle gcc-4.7  | grep const
000000000040081a W bool check<two::c2<int const*> >(two::c2<int const*>*)
0000000000400864 W two::c2<int const*>::val()
000000000040077c W two::c2<int const*>::c2(int const*)
000000000040077c W two::c2<int const*>::c2(int const*)
00000000004008aa W two::c2<int const*>::~c2()
00000000004008aa W two::c2<int const*>::~c2()
[hjl@gnu-6 pr14344]$ readelf --debug-dump=gdb_index gcc-4.7  | grep const
readelf: Warning: Version 5 does not include inlined functions.
readelf: Warning: Version 6 does not include symbol attributes.
[139] two::c2<const int*>::c2: 0 [no symbol information]
[326] check<two::c2<const int*> >: 0 [no symbol information]
[468] two::c2<const int*>::val: 0 [no symbol information]
[616] two::c2<const int*>::~c2: 0 [no symbol information]
[752] two::c2<const int*>: 0 [no symbol information]
[hjl@gnu-6 pr14344]$

For GCC 4.7, --gdb-index generates different demangled symbol names.

--
Configure bugmail: http://sourceware.org/bugzilla/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

_______________________________________________
bug-binutils mailing list
bug-binutils@...
https://lists.gnu.org/mailman/listinfo/bug-binutils

http://sourceware.org/bugzilla/show_bug.cgi?id=14344

--- Comment #6 from cvs-commit at gcc dot gnu.org <cvs-commit at gcc dot gnu.org> 2012-07-19 00:19:38 UTC ---
CVSROOT:    /cvs/src
Module name:    src
Changes by:    ccoutant@...    2012-07-19 00:19:35

Modified files:
    gold           : ChangeLog configure configure.ac
    gold/testsuite : Makefile.am Makefile.in gdb_index_test_1.sh
                     gdb_index_test_2.sh gdb_index_test_3.sh
Added files:
    gold/testsuite : gdb_index_test_4.sh gdb_index_test_comm.sh

Log message:
    2012-07-18  Cary Coutant  <ccoutant@...>

    PR gold/14344
    * configure.ac: Add check for -gpubnames support.
    * configure: Regenerate.
    * testsuite/Makefile.am (gdb_index_test_1): Add check for -gpubnames
    support; force -gno-pubnames.
    (gdb_index_test_2, gdb_index_test_3): Add check for -gpubnames
    support.
    (gdb_index_test_4): New test.
    * testsuite/Makefile.in: Regenerate.
    * testsuite/gdb_index_test_1.sh: Refactor code into common file.
    * testsuite/gdb_index_test_2.sh: Likewise.
    * testsuite/gdb_index_test_3.sh: Don't look for space after colon.
    * testsuite/gdb_index_test_4.sh: New script.
    * testsuite/gdb_index_test_comm.sh: New script with common code;
    don't look for space after colon.

Patches:
http://sourceware.org/cgi-bin/cvsweb.cgi/src/gold/ChangeLog.diff?cvsroot=src&r1=1.927&r2=1.928
http://sourceware.org/cgi-bin/cvsweb.cgi/src/gold/configure.diff?cvsroot=src&r1=1.78&r2=1.79
http://sourceware.org/cgi-bin/cvsweb.cgi/src/gold/configure.ac.diff?cvsroot=src&r1=1.75&r2=1.76
http://sourceware.org/cgi-bin/cvsweb.cgi/src/gold/testsuite/gdb_index_test_4.sh.diff?cvsroot=src&r1=NONE&r2=1.1
http://sourceware.org/cgi-bin/cvsweb.cgi/src/gold/testsuite/gdb_index_test_comm.sh.diff?cvsroot=src&r1=NONE&r2=1.1
http://sourceware.org/cgi-bin/cvsweb.cgi/src/gold/testsuite/Makefile.am.diff?cvsroot=src&r1=1.193&r2=1.194
http://sourceware.org/cgi-bin/cvsweb.cgi/src/gold/testsuite/Makefile.in.diff?cvsroot=src&r1=1.203&r2=1.204
http://sourceware.org/cgi-bin/cvsweb.cgi/src/gold/testsuite/gdb_index_test_1.sh.diff?cvsroot=src&r1=1.1&r2=1.2
http://sourceware.org/cgi-bin/cvsweb.cgi/src/gold/testsuite/gdb_index_test_2.sh.diff?cvsroot=src&r1=1.1&r2=1.2
http://sourceware.org/cgi-bin/cvsweb.cgi/src/gold/testsuite/gdb_index_test_3.sh.diff?cvsroot=src&r1=1.1&r2=1.2

--
Configure bugmail: http://sourceware.org/bugzilla/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

_______________________________________________
bug-binutils mailing list
bug-binutils@...
https://lists.gnu.org/mailman/listinfo/bug-binutils

http://sourceware.org/bugzilla/show_bug.cgi?id=14344

Cary Coutant <ccoutant at google dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED
         AssignedTo|ian at airs dot com         |ccoutant at google dot com

--- Comment #7 from Cary Coutant <ccoutant at google dot com> 2012-07-19 00:30:59 UTC ---
Added configure check to disable the gdb_index tests when not using a compiler
known to produce reliable pubnames information.

--
Configure bugmail: http://sourceware.org/bugzilla/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

_______________________________________________
bug-binutils mailing list
bug-binutils@...
https://lists.gnu.org/mailman/listinfo/bug-binutils