Re: [google/gcc-4_7] Allow static const floats unless -pedantic is passed (issue6212051)

[Jeffrey Yasskin]

Looks good to me. Yay for kicking the can down the road. ;)

On Tue, May 15, 2012 at 12:57 PM, Ollie Wild a...@google.com wrote:
 To be applied to google/gcc-4_7.

 Allow static const floats unless -pedantic is passed.

 This patch allows us to migrate to C++11 more incrementally, since we can 
 leave
 the static const float initializations in place, flip the switch, and then
 change it to use constexpr.

 This is a forward port of r180638 from google/gcc-4_6 (despite the fact that
 that revision says to NOT forward-port this).  Minor additional fixups have
 been applied.

 2012-05-15   Ollie Wild  a...@google.com

        * gcc/cp/decl.c (check_static_variable_definition): Only generate a
        constexpr warning when -pedantic is enabled.
        * gcc/testsuite/g++.dg/cpp0x/constexpr-static8.C: Replace -fpermissive
        with -pedantic.
        * gcc/testsuite/g++.dg/cpp0x/constexpr-static8_nonpedantic.C: New test.
        * gcc/testsuite/g++.old-deja/g++.ext/memconst.C: Compile with -pedantic
        -pedantic-errors to work around test failures with -std=gnu++11.

 diff --git a/gcc/cp/decl.c b/gcc/cp/decl.c
 index bb26d15..d2fe731 100644
 --- a/gcc/cp/decl.c
 +++ b/gcc/cp/decl.c
 @@ -7861,9 +7861,18 @@ check_static_variable_definition (tree decl, tree type)
        error (in-class initialization of static data member %q#D of 
               incomplete type, decl);
       else if (literal_type_p (type))
 -       permerror (input_location,
 -                  %constexpr% needed for in-class initialization of 
 -                  static data member %q#D of non-integral type, decl);
 +       {
 +          /* FIXME google: This local modification allows us to
 +             transition from C++98 to C++11 without moving static
 +             const floats out of the class during the transition.  It
 +             should not be forward-ported to a 4.8 branch, since by
 +             then we should be able to just fix the code to use
 +             constexpr.  */
 +          pedwarn (input_location, OPT_pedantic,
 +                   %constexpr% needed for in-class initialization of 
 +                   static data member %q#D of non-integral type, decl);
 +          return 0;
 +       }
       else
        error (in-class initialization of static data member %q#D of 
               non-literal type, decl);
 diff --git a/gcc/testsuite/g++.dg/cpp0x/constexpr-static8.C 
 b/gcc/testsuite/g++.dg/cpp0x/constexpr-static8.C
 index 7c84cf8..658a458 100644
 --- a/gcc/testsuite/g++.dg/cpp0x/constexpr-static8.C
 +++ b/gcc/testsuite/g++.dg/cpp0x/constexpr-static8.C
 @@ -1,5 +1,5 @@
  // PR c++/50258
 -// { dg-options -std=c++0x -fpermissive }
 +// { dg-options -std=c++0x -pedantic }

  struct Foo {
   static const double d = 3.14; // { dg-warning constexpr }
 diff --git a/gcc/testsuite/g++.dg/cpp0x/constexpr-static8_nonpedantic.C 
 b/gcc/testsuite/g++.dg/cpp0x/constexpr-static8_nonpedantic.C
 new file mode 100644
 index 000..28d34a1
 --- /dev/null
 +++ b/gcc/testsuite/g++.dg/cpp0x/constexpr-static8_nonpedantic.C
 @@ -0,0 +1,7 @@
 +// PR c++/50258
 +// { dg-options -std=c++0x }
 +
 +struct Foo {
 +  static const double d = 3.14; // no warning
 +};
 +const double Foo::d;
 diff --git a/gcc/testsuite/g++.old-deja/g++.ext/memconst.C 
 b/gcc/testsuite/g++.old-deja/g++.ext/memconst.C
 index d934763..7e86156 100644
 --- a/gcc/testsuite/g++.old-deja/g++.ext/memconst.C
 +++ b/gcc/testsuite/g++.old-deja/g++.ext/memconst.C
 @@ -1,5 +1,5 @@
  // { dg-do assemble  }
 -// { dg-options  }
 +// { dg-options -pedantic -pedantic-errors }
  // From: ove.ewer...@syscon.uu.se (Ove Ewerlid)
  // Subject: ss-940630:cc1plus: internal error
  // Date: Sat, 2 Jul 1994 05:07:20 +0200

 --
 This patch is available for review at http://codereview.appspot.com/6212051

Re: [google/integration] Extend C++11 UDLs to be compatible with inttypes.h macros (issue6104051)

[Jeffrey Yasskin]

Could you try to get this into mainline instead of just the google
branches? In http://gcc.gnu.org/PR52538, Jonathan sounded like he'd
consider accepting it.

On Sun, Apr 22, 2012 at 7:54 PM, Ollie Wild a...@google.com wrote:
 Add new option, -Wreserved-user-defined-literal.

 This option, which is enabled by default, causes the preprocessor to warn
 when a string or character literal is followed by a ud-suffix which does
 not begin with an underscore.  According to [lex.ext]p10, this is
 ill-formed.

 Also modifies the preprocessor to treat such ill-formed suffixes as separate
 preprocessing tokens.  This is consistent with the Clang front end (see
 http://llvm.org/viewvc/llvm-project?view=revrevision=152287), and enables
 backwards compatibility with code that uses formatting macros from
 inttypes.h, as in the following code block:

  int main() {
    int64_t i64 = 123;
    printf(My int64: %PRId64\n, i64);
  }

 Google ref b/6377711.

 2012-04-22   Ollie Wild  a...@google.com

        * gcc/c-family/c-common.c:
        * gcc/c-family/c-opts.c (c_common_handle_option):
        * gcc/c-family/c.opt:
        * gcc/doc/invoke.texi (struct A):
        * gcc/testsuite/g++.dg/cpp0x/Wreserved-user-defined-literal.C (test):
        (main):
        * libcpp/include/cpplib.h (struct cpp_options):
        * libcpp/init.c (cpp_create_reader):
        * libcpp/lex.c (lex_raw_string):
        (lex_string):

 diff --git a/gcc/c-family/c-common.c b/gcc/c-family/c-common.c
 index 1d19251..915dc25 100644
 --- a/gcc/c-family/c-common.c
 +++ b/gcc/c-family/c-common.c
 @@ -8724,6 +8724,7 @@ static const struct reason_option_codes_t 
 option_codes[] = {
   {CPP_W_NORMALIZE,                    OPT_Wnormalized_},
   {CPP_W_INVALID_PCH,                  OPT_Winvalid_pch},
   {CPP_W_WARNING_DIRECTIVE,            OPT_Wcpp},
 +  {CPP_W_RESERVED_USER_LITERALS,       OPT_Wreserved_user_defined_literal},
   {CPP_W_NONE,                         0}
  };

 diff --git a/gcc/c-family/c-opts.c b/gcc/c-family/c-opts.c
 index 5118928..dab6ce5 100644
 --- a/gcc/c-family/c-opts.c
 +++ b/gcc/c-family/c-opts.c
 @@ -509,6 +509,10 @@ c_common_handle_option (size_t scode, const char *arg, 
 int value,
          break;
        }

 +    case OPT_Wreserved_user_defined_literal:
 +      cpp_opts-warn_reserved_user_literals = value;
 +      break;
 +
     case OPT_Wreturn_type:
       warn_return_type = value;
       break;
 diff --git a/gcc/c-family/c.opt b/gcc/c-family/c.opt
 index 40ff96c..f610513 100644
 --- a/gcc/c-family/c.opt
 +++ b/gcc/c-family/c.opt
 @@ -589,6 +589,10 @@ Wreorder
  C++ ObjC++ Var(warn_reorder) Warning
  Warn when the compiler reorders code

 +Wreserved-user-defined-literal
 +C++ ObjC++ Warning
 +Warn when a string or character literal is followed by a ud-suffix which 
 does not begin with an underscore.
 +
  Wreturn-type
  C ObjC C++ ObjC++ Var(warn_return_type) Warning
  Warn whenever a function's return type defaults to \int\ (C), or about 
 inconsistent return types (C++)
 diff --git a/gcc/doc/invoke.texi b/gcc/doc/invoke.texi
 index 1b61e76..d425079 100644
 --- a/gcc/doc/invoke.texi
 +++ b/gcc/doc/invoke.texi
 @@ -198,7 +198,7 @@ in the following sections.
  -fvisibility-ms-compat @gol
  -Wabi  -Wconversion-null  -Wctor-dtor-privacy @gol
  -Wdelete-non-virtual-dtor -Wnarrowing -Wnoexcept @gol
 --Wnon-virtual-dtor  -Wreorder @gol
 +-Wnon-virtual-dtor  -Wreorder -Wreserved-user-defined-literal @gol
  -Weffc++  -Wstrict-null-sentinel @gol
  -Wno-non-template-friend  -Wold-style-cast @gol
  -Woverloaded-virtual  -Wno-pmf-conversions @gol
 @@ -2474,6 +2474,30 @@ struct A @{
  The compiler will rearrange the member initializers for @samp{i}
  and @samp{j} to match the declaration order of the members, emitting
  a warning to that effect.  This warning is enabled by @option{-Wall}.
 +
 +@item -Wreserved-user-defined-literal @r{(C++ and Objective-C++ only)}
 +@opindex Wreserved-user-defined-literal
 +@opindex Wno-reserved-user-defined-literal
 +Warn when a string or character literal is followed by a ud-suffix which does
 +not begin with an underscore.  As a conforming extension, GCC treats such
 +suffixes as separate preprocessing tokens in order to maintain backwards
 +compatibility with code that uses formatting macros from @code{inttypes.h}.
 +For example:
 +
 +@smallexample
 +#define __STDC_FORMAT_MACROS
 +#include inttypes.h
 +#include stdio.h
 +
 +int main() @{
 +  int64_t i64 = 123;
 +  printf(My int64: %PRId64\n, i64);
 +@}
 +@end smallexample
 +
 +In this case, @code{PRId64} is treated as a separate preprocessing token.
 +
 +This warning is enabled by default.
  @end table

  The following @option{-W@dots{}} options are not affected by @option{-Wall}.
 diff --git a/gcc/testsuite/g++.dg/cpp0x/Wreserved-user-defined-literal.C 
 b/gcc/testsuite/g++.dg/cpp0x/Wreserved-user-defined-literal.C
 new file mode 100644
 index 000..66de5c0
 --- /dev/null
 +++ b/gcc/testsuite/g++.dg/cpp0x/Wreserved-user-defined-literal.C
 @@ -0,0 

Re: [google/integration] Extend C++11 UDLs to be compatible with inttypes.h macros (issue6104051)

[Jeffrey Yasskin]

Let's let the discussion _start_ before assuming it'll be protracted.
;) I don't think it'll kill us to run the next round of testing in
C++98 mode. If the patch doesn't look close to acceptance in a couple
days, I think it'll make sense to put the then-current version into
the google branches while people are agreeing about what to do in the
long run.

On Sun, Apr 22, 2012 at 8:14 PM, Ollie Wild a...@google.com wrote:
 I'd like to get this into the google branches first because this is blocking
 testing.

 I fully expect this to result in some protracted discussion before it can be
 accepted into trunk.

 Ollie


 On Sun, Apr 22, 2012 at 10:10 PM, Jeffrey Yasskin jyass...@google.com
 wrote:

 Could you try to get this into mainline instead of just the google
 branches? In http://gcc.gnu.org/PR52538, Jonathan sounded like he'd
 consider accepting it.

 On Sun, Apr 22, 2012 at 7:54 PM, Ollie Wild a...@google.com wrote:
  Add new option, -Wreserved-user-defined-literal.
 
  This option, which is enabled by default, causes the preprocessor to
  warn
  when a string or character literal is followed by a ud-suffix which does
  not begin with an underscore.  According to [lex.ext]p10, this is
  ill-formed.
 
  Also modifies the preprocessor to treat such ill-formed suffixes as
  separate
  preprocessing tokens.  This is consistent with the Clang front end (see
  http://llvm.org/viewvc/llvm-project?view=revrevision=152287), and
  enables
  backwards compatibility with code that uses formatting macros from
  inttypes.h, as in the following code block:
 
   int main() {
     int64_t i64 = 123;
     printf(My int64: %PRId64\n, i64);
   }
 
  Google ref b/6377711.
 
  2012-04-22   Ollie Wild  a...@google.com
 
         * gcc/c-family/c-common.c:
         * gcc/c-family/c-opts.c (c_common_handle_option):
         * gcc/c-family/c.opt:
         * gcc/doc/invoke.texi (struct A):
         * gcc/testsuite/g++.dg/cpp0x/Wreserved-user-defined-literal.C
  (test):
         (main):
         * libcpp/include/cpplib.h (struct cpp_options):
         * libcpp/init.c (cpp_create_reader):
         * libcpp/lex.c (lex_raw_string):
         (lex_string):
 
  diff --git a/gcc/c-family/c-common.c b/gcc/c-family/c-common.c
  index 1d19251..915dc25 100644
  --- a/gcc/c-family/c-common.c
  +++ b/gcc/c-family/c-common.c
  @@ -8724,6 +8724,7 @@ static const struct reason_option_codes_t
  option_codes[] = {
    {CPP_W_NORMALIZE,                    OPT_Wnormalized_},
    {CPP_W_INVALID_PCH,                  OPT_Winvalid_pch},
    {CPP_W_WARNING_DIRECTIVE,            OPT_Wcpp},
  +  {CPP_W_RESERVED_USER_LITERALS,
  OPT_Wreserved_user_defined_literal},
    {CPP_W_NONE,                         0}
   };
 
  diff --git a/gcc/c-family/c-opts.c b/gcc/c-family/c-opts.c
  index 5118928..dab6ce5 100644
  --- a/gcc/c-family/c-opts.c
  +++ b/gcc/c-family/c-opts.c
  @@ -509,6 +509,10 @@ c_common_handle_option (size_t scode, const char
  *arg, int value,
           break;
         }
 
  +    case OPT_Wreserved_user_defined_literal:
  +      cpp_opts-warn_reserved_user_literals = value;
  +      break;
  +
      case OPT_Wreturn_type:
        warn_return_type = value;
        break;
  diff --git a/gcc/c-family/c.opt b/gcc/c-family/c.opt
  index 40ff96c..f610513 100644
  --- a/gcc/c-family/c.opt
  +++ b/gcc/c-family/c.opt
  @@ -589,6 +589,10 @@ Wreorder
   C++ ObjC++ Var(warn_reorder) Warning
   Warn when the compiler reorders code
 
  +Wreserved-user-defined-literal
  +C++ ObjC++ Warning
  +Warn when a string or character literal is followed by a ud-suffix
  which does not begin with an underscore.
  +
   Wreturn-type
   C ObjC C++ ObjC++ Var(warn_return_type) Warning
   Warn whenever a function's return type defaults to \int\ (C), or
  about inconsistent return types (C++)
  diff --git a/gcc/doc/invoke.texi b/gcc/doc/invoke.texi
  index 1b61e76..d425079 100644
  --- a/gcc/doc/invoke.texi
  +++ b/gcc/doc/invoke.texi
  @@ -198,7 +198,7 @@ in the following sections.
   -fvisibility-ms-compat @gol
   -Wabi  -Wconversion-null  -Wctor-dtor-privacy @gol
   -Wdelete-non-virtual-dtor -Wnarrowing -Wnoexcept @gol
  --Wnon-virtual-dtor  -Wreorder @gol
  +-Wnon-virtual-dtor  -Wreorder -Wreserved-user-defined-literal @gol
   -Weffc++  -Wstrict-null-sentinel @gol
   -Wno-non-template-friend  -Wold-style-cast @gol
   -Woverloaded-virtual  -Wno-pmf-conversions @gol
  @@ -2474,6 +2474,30 @@ struct A @{
   The compiler will rearrange the member initializers for @samp{i}
   and @samp{j} to match the declaration order of the members, emitting
   a warning to that effect.  This warning is enabled by @option{-Wall}.
  +
  +@item -Wreserved-user-defined-literal @r{(C++ and Objective-C++ only)}
  +@opindex Wreserved-user-defined-literal
  +@opindex Wno-reserved-user-defined-literal
  +Warn when a string or character literal is followed by a ud-suffix
  which does
  +not begin with an underscore.  As a conforming extension, GCC treats
  such
  +suffixes as separate preprocessing

Re: [google/integration] Extend C++11 UDLs to be compatible with inttypes.h macros (issue6104051)

[Jeffrey Yasskin]

Thanks!

On Sun, Apr 22, 2012 at 8:32 PM, Ollie Wild a...@google.com wrote:
 Okay, I'll send out a trunk patch for review now, too.

 Ollie

 On Sun, Apr 22, 2012 at 10:29 PM, Jeffrey Yasskin jyass...@google.com wrote:

 Let's let the discussion _start_ before assuming it'll be protracted.
 ;) I don't think it'll kill us to run the next round of testing in
 C++98 mode. If the patch doesn't look close to acceptance in a couple
 days, I think it'll make sense to put the then-current version into
 the google branches while people are agreeing about what to do in the
 long run.

 On Sun, Apr 22, 2012 at 8:14 PM, Ollie Wild a...@google.com wrote:
  I'd like to get this into the google branches first because this is
  blocking
  testing.
 
  I fully expect this to result in some protracted discussion before it
  can be
  accepted into trunk.
 
  Ollie
 
 
  On Sun, Apr 22, 2012 at 10:10 PM, Jeffrey Yasskin jyass...@google.com
  wrote:
 
  Could you try to get this into mainline instead of just the google
  branches? In http://gcc.gnu.org/PR52538, Jonathan sounded like he'd
  consider accepting it.
 
  On Sun, Apr 22, 2012 at 7:54 PM, Ollie Wild a...@google.com wrote:
   Add new option, -Wreserved-user-defined-literal.
  
   This option, which is enabled by default, causes the preprocessor to
   warn
   when a string or character literal is followed by a ud-suffix which
   does
   not begin with an underscore.  According to [lex.ext]p10, this is
   ill-formed.
  
   Also modifies the preprocessor to treat such ill-formed suffixes as
   separate
   preprocessing tokens.  This is consistent with the Clang front end
   (see
   http://llvm.org/viewvc/llvm-project?view=revrevision=152287), and
   enables
   backwards compatibility with code that uses formatting macros from
   inttypes.h, as in the following code block:
  
    int main() {
      int64_t i64 = 123;
      printf(My int64: %PRId64\n, i64);
    }
  
   Google ref b/6377711.
  
   2012-04-22   Ollie Wild  a...@google.com
  
          * gcc/c-family/c-common.c:
          * gcc/c-family/c-opts.c (c_common_handle_option):
          * gcc/c-family/c.opt:
          * gcc/doc/invoke.texi (struct A):
          * gcc/testsuite/g++.dg/cpp0x/Wreserved-user-defined-literal.C
   (test):
          (main):
          * libcpp/include/cpplib.h (struct cpp_options):
          * libcpp/init.c (cpp_create_reader):
          * libcpp/lex.c (lex_raw_string):
          (lex_string):
  
   diff --git a/gcc/c-family/c-common.c b/gcc/c-family/c-common.c
   index 1d19251..915dc25 100644
   --- a/gcc/c-family/c-common.c
   +++ b/gcc/c-family/c-common.c
   @@ -8724,6 +8724,7 @@ static const struct reason_option_codes_t
   option_codes[] = {
     {CPP_W_NORMALIZE,                    OPT_Wnormalized_},
     {CPP_W_INVALID_PCH,                  OPT_Winvalid_pch},
     {CPP_W_WARNING_DIRECTIVE,            OPT_Wcpp},
   +  {CPP_W_RESERVED_USER_LITERALS,
   OPT_Wreserved_user_defined_literal},
     {CPP_W_NONE,                         0}
    };
  
   diff --git a/gcc/c-family/c-opts.c b/gcc/c-family/c-opts.c
   index 5118928..dab6ce5 100644
   --- a/gcc/c-family/c-opts.c
   +++ b/gcc/c-family/c-opts.c
   @@ -509,6 +509,10 @@ c_common_handle_option (size_t scode, const char
   *arg, int value,
            break;
          }
  
   +    case OPT_Wreserved_user_defined_literal:
   +      cpp_opts-warn_reserved_user_literals = value;
   +      break;
   +
       case OPT_Wreturn_type:
         warn_return_type = value;
         break;
   diff --git a/gcc/c-family/c.opt b/gcc/c-family/c.opt
   index 40ff96c..f610513 100644
   --- a/gcc/c-family/c.opt
   +++ b/gcc/c-family/c.opt
   @@ -589,6 +589,10 @@ Wreorder
    C++ ObjC++ Var(warn_reorder) Warning
    Warn when the compiler reorders code
  
   +Wreserved-user-defined-literal
   +C++ ObjC++ Warning
   +Warn when a string or character literal is followed by a ud-suffix
   which does not begin with an underscore.
   +
    Wreturn-type
    C ObjC C++ ObjC++ Var(warn_return_type) Warning
    Warn whenever a function's return type defaults to \int\ (C), or
   about inconsistent return types (C++)
   diff --git a/gcc/doc/invoke.texi b/gcc/doc/invoke.texi
   index 1b61e76..d425079 100644
   --- a/gcc/doc/invoke.texi
   +++ b/gcc/doc/invoke.texi
   @@ -198,7 +198,7 @@ in the following sections.
    -fvisibility-ms-compat @gol
    -Wabi  -Wconversion-null  -Wctor-dtor-privacy @gol
    -Wdelete-non-virtual-dtor -Wnarrowing -Wnoexcept @gol
   --Wnon-virtual-dtor  -Wreorder @gol
   +-Wnon-virtual-dtor  -Wreorder -Wreserved-user-defined-literal @gol
    -Weffc++  -Wstrict-null-sentinel @gol
    -Wno-non-template-friend  -Wold-style-cast @gol
    -Woverloaded-virtual  -Wno-pmf-conversions @gol
   @@ -2474,6 +2474,30 @@ struct A @{
    The compiler will rearrange the member initializers for @samp{i}
    and @samp{j} to match the declaration order of the members, emitting
    a warning to that effect.  This warning is enabled by
   @option{-Wall

Re: [patch] Fix PR52822 (stable_partition move-assigns object to itself) in trunk, 4.7, and 4.6

[Jeffrey Yasskin]

On Wed, Apr 4, 2012 at 11:09 PM, Paolo Carlini paolo.carl...@oracle.com wrote:
 Hi,

 The attached patches fix http://gcc.gnu.org/PR52822, and have been
 tested with `make check-c++` on linux-x86_64. The trunk patch applies
 and tests cleanly on gcc-4_7-branch. The gcc-4_6-branch patch is
 significantly simpler, as Paolo suggested on the bug.

 A few small issues.

 For the 4.6 version of the patch, you want to use std::__addressof, instead
 of simply , which may be overloaded.
 And by the way what's wrong with just comparing the iterators?

Nothing's wrong with comparing the iterators. Switched.

 For the mainline/4.7 version, you want to cast __pred to bool:
 !bool(__pred(*__first)). Also, isn't clear to me why you have __local_len in
 __find_if_not_n instead of just working on __len (and in any case please
 prefer just __local_len as condition instead of the redundant __local_len !=
 0; likewise in many other similar situations).

I was micro-optimizing, since the compiler might think __len is
modified in __pred if it doesn't inline enough. I've removed
__local_len, casted pred results to bool, and avoided !=0.

 Also, very minor nit, you will be touching the libstdc++-v3 Changelog thus
 no libstdc++-v3 prefixes in the ChangeLog entry.

Yep, done.


pr52822_gcc46.patch
Description: Binary data


pr52822_trunk.patch
Description: Binary data

[patch] Fix PR52822 (stable_partition move-assigns object to itself) in trunk, 4.7, and 4.6

[Jeffrey Yasskin]

The attached patches fix http://gcc.gnu.org/PR52822, and have been
tested with `make check-c++` on linux-x86_64. The trunk patch applies
and tests cleanly on gcc-4_7-branch. The gcc-4_6-branch patch is
significantly simpler, as Paolo suggested on the bug.

The test is still inadequate, but given that there's no way to force
get_temporary_buffer to return a short buffer, I'm not sure how to
improve it. It might be possible to use set_memory_limits(), but that
would require a very complex test program to make sure exactly the
right operator new call fails.

Jeffrey Yasskin


pr52822_trunk.patch
Description: Binary data


pr52822_gcc46.patch
Description: Binary data

[google] Fix PR52822 in google's gcc-4.6 branch (issue5975063)

[Jeffrey Yasskin]

I'll send the patches for trunk, gcc-4_7-branch, and gcc-4_6-branch in
a separate thread later today or tomorrow, but I need to get the
google branch fixed before that. Paolo suggested this simple fix was
the right approach for the gcc-4_6-branch, so I'm using it for
google/gcc-4_6 too. I'll be happy to fix up the google branch to match
whatever gets accepted for gcc-4_6-branch once that's decided.

Tested with `make check` on x86-64-linux, and I'll check this against
the build I originally noticed the problem in before committing it.

2012-04-02   Jeffrey Yasskin  jyass...@google.com

* libstdc++-v3/include/bits/stl_algo.h (__stable_partition_adaptive):
Avoid move-assigning an object to itself by explicitly testing
for identity.
* libstdc++-v3/testsuite/25_algorithms/stable_partition/pr52822.cc:
Test vectors, which have a destructive self-move-assignment.
* libstdc++-v3/testsuite/25_algorithms/stable_partition/moveable.cc 
(test02):
Test with rvalstruct, which explicitly checks self-move-assignment.

Index: libstdc++-v3/include/bits/stl_algo.h
===
--- libstdc++-v3/include/bits/stl_algo.h(revision 186093)
+++ libstdc++-v3/include/bits/stl_algo.h(working copy)
@@ -1844,7 +1844,8 @@ _GLIBCXX_BEGIN_NAMESPACE_VERSION
  for (; __first != __last; ++__first)
if (__pred(*__first))
  {
-   *__result1 = _GLIBCXX_MOVE(*__first);
+if (*__result1 != *__first)
+  *__result1 = _GLIBCXX_MOVE(*__first);
++__result1;
  }
else
Index: libstdc++-v3/testsuite/25_algorithms/stable_partition/pr52822.cc
===
--- libstdc++-v3/testsuite/25_algorithms/stable_partition/pr52822.cc
(revision 0)
+++ libstdc++-v3/testsuite/25_algorithms/stable_partition/pr52822.cc
(revision 0)
@@ -0,0 +1,43 @@
+// { dg-options -std=gnu++0x }
+
+// Copyright (C) 2012 Free Software Foundation, Inc.
+//
+// This file is part of the GNU ISO C++ Library.  This library is free
+// software; you can redistribute it and/or modify it under the
+// terms of the GNU General Public License as published by the
+// Free Software Foundation; either version 3, or (at your option)
+// any later version.
+
+// This library is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without Pred the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU General Public License for more details.
+
+// You should have received a copy of the GNU General Public License along
+// with this library; see the file COPYING3.  If not see
+// http://www.gnu.org/licenses/.
+
+// 25.2.12 [lib.alg.partitions] Partitions.
+
+#include algorithm
+#include vector
+#include testsuite_hooks.h
+
+bool true_vector_pred(const std::vectorint) { return true; }
+
+void
+test01()
+{
+  std::vectorstd::vectorint  v(1);
+  v[0].push_back(7);
+  VERIFY( v[0].size() == 1 );
+  std::stable_partition(v.begin(), v.end(), true_vector_pred);
+  VERIFY( v[0].size() == 1 );
+}
+
+int
+main()
+{
+  test01();
+  return 0;
+}
Index: libstdc++-v3/testsuite/25_algorithms/stable_partition/moveable.cc
===
--- libstdc++-v3/testsuite/25_algorithms/stable_partition/moveable.cc   
(revision 186093)
+++ libstdc++-v3/testsuite/25_algorithms/stable_partition/moveable.cc   
(working copy)
@@ -1,6 +1,6 @@
 // { dg-options -std=gnu++0x }
 
-// Copyright (C) 2009, 2010 Free Software Foundation, Inc.
+// Copyright (C) 2009, 2010, 2012 Free Software Foundation, Inc.
 //
 // This file is part of the GNU ISO C++ Library.  This library is free
 // software; you can redistribute it and/or modify it under the
@@ -39,6 +39,12 @@ const int A[] = {1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11
 const int B[] = {2, 4, 6, 8, 10, 12, 14, 16, 1, 3, 5, 7, 9, 11, 13, 15, 17};
 const int N = sizeof(A) / sizeof(int);
 
+// Check that starting with a value the predicate returns true for
+// works too. (PR52822)
+const int A2[] = {2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17};
+const int B2[] = {2, 4, 6, 8, 10, 12, 14, 16, 3, 5, 7, 9, 11, 13, 15, 17};
+const int N2 = sizeof(A2) / sizeof(int);
+
 struct Pred
 {
   bool
@@ -46,7 +52,8 @@ struct Pred
   { return (x.val % 2) == 0; }
 };
 
-// 25.2.12 stable_partition()
+// 25.2.12 stable_partition(), starting with a value for which the
+// predicate returns false.
 void
 test01()
 {
@@ -60,9 +67,25 @@ test01()
   VERIFY( std::equal(s1, s1 + N, B) );
 }
 
+// 25.2.12 stable_partition(), starting with a value for which the
+// predicate returns true.
+void
+test02()
+{
+  bool test __attribute__((unused)) = true;
+
+  rvalstruct s1[N2];
+  std::copy(A2, A2 + N2, s1);
+  Container con(s1, s1 + N2);
+
+  std::stable_partition(con.begin(), con.end(), Pred());
+  VERIFY

[google] Work around PR52796 by replacing empty packs with explicit overloads. (issue5971053)

[Jeffrey Yasskin]

Work around http://gcc.gnu.org/PR52796 in gcc-4.6 by adding an
overload of each function that passes a parameter pack directly as the
only arguments of an object's constructor, which explicitly takes no
arguments in place of the pack.

Tested with check-c++ and by trying to provoke the bug in valgrind for
each changed location.  Some of the insert_aux locations appear
inaccessible because of missing emplace() functions.

I plan to only apply this to the google/gcc-4_6 branch, since gcc-4.7
already makes these cases work properly.

2012-03-30   Jeffrey Yasskin  jyass...@google.com

* libstdc++-v3/include/ext/pool_allocator.h: Add 1-argument construct() 
method.
* libstdc++-v3/include/ext/bitmap_allocator.h: Likewise.
* libstdc++-v3/include/ext/new_allocator.h: Likewise.
* libstdc++-v3/include/ext/malloc_allocator.h: Likewise.
* libstdc++-v3/include/ext/array_allocator.h: Likewise.
* libstdc++-v3/include/ext/mt_allocator.h: Likewise.
* libstdc++-v3/include/ext/extptr_allocator.h: Likewise.
* libstdc++-v3/include/bits/stl_construct.h:Add 1-argument _Construct 
function.
* libstdc++-v3/include/bits/stl_list.h: Add default _List_node 
constructor.
* libstdc++-v3/include/bits/hashtable_policy.h: Add default _Hash_node 
constructor.
* libstdc++-v3/include/bits/forward_list.h:Add default _Fwd_list_node 
constructor.
* libstdc++-v3/include/bits/stl_tree.h:Add default _Rb_tree_node 
constructor.
* 
libstdc++-v3/testsuite/23_containers/forward_list/requirements/dr438/assign_neg.cc:
 Update line numbers.
* 
libstdc++-v3/testsuite/23_containers/forward_list/requirements/dr438/insert_neg.cc:
 Likewise.
* 
libstdc++-v3/testsuite/23_containers/forward_list/requirements/dr438/constructor_1_neg.cc:
 Likewise.
* 
libstdc++-v3/testsuite/23_containers/forward_list/requirements/dr438/constructor_2_neg.cc:
 Likewise.
* 
libstdc++-v3/testsuite/23_containers/list/requirements/dr438/assign_neg.cc: 
Likewise.
* 
libstdc++-v3/testsuite/23_containers/list/requirements/dr438/insert_neg.cc: 
Likewise.
* 
libstdc++-v3/testsuite/23_containers/list/requirements/dr438/constructor_1_neg.cc:
 Likewise.
* 
libstdc++-v3/testsuite/23_containers/list/requirements/dr438/constructor_2_neg.cc:
 Likewise.

Index: libstdc++-v3/include/ext/pool_allocator.h
===
--- libstdc++-v3/include/ext/pool_allocator.h   (revision 186024)
+++ libstdc++-v3/include/ext/pool_allocator.h   (working copy)
@@ -165,6 +165,12 @@ _GLIBCXX_BEGIN_NAMESPACE_VERSION
   { ::new((void *)__p) _Tp(__val); }
 
 #ifdef __GXX_EXPERIMENTAL_CXX0X__
+  // Work around PR52796 by avoiding 0-length parameter packs
+  // passed to constructors.
+  void
+  construct(pointer __p)
+  { ::new((void *)__p) _Tp(); }
+
   templatetypename... _Args
 void
 construct(pointer __p, _Args... __args)
Index: libstdc++-v3/include/ext/bitmap_allocator.h
===
--- libstdc++-v3/include/ext/bitmap_allocator.h (revision 186024)
+++ libstdc++-v3/include/ext/bitmap_allocator.h (working copy)
@@ -1058,6 +1058,12 @@ _GLIBCXX_BEGIN_NAMESPACE_VERSION
   { ::new((void *)__p) value_type(__data); }
 
 #ifdef __GXX_EXPERIMENTAL_CXX0X__
+  // Work around PR52796 by avoiding 0-length parameter packs
+  // passed to constructors.
+  void
+  construct(pointer __p)
+  { ::new((void *)__p) _Tp(); }
+
   templatetypename... _Args
 void
 construct(pointer __p, _Args... __args)
@@ -1109,4 +1115,3 @@ _GLIBCXX_END_NAMESPACE_VERSION
 } // namespace __gnu_cxx
 
 #endif 
-
Index: libstdc++-v3/include/ext/new_allocator.h
===
--- libstdc++-v3/include/ext/new_allocator.h(revision 186024)
+++ libstdc++-v3/include/ext/new_allocator.h(working copy)
@@ -115,6 +115,12 @@ _GLIBCXX_BEGIN_NAMESPACE_VERSION
   { ::new((void *)__p) _Tp(__val); }
 
 #ifdef __GXX_EXPERIMENTAL_CXX0X__
+  // Work around PR52796 by avoiding 0-length parameter packs
+  // passed to constructors.
+  void
+  construct(pointer __p)
+  { ::new((void *)__p) _Tp(); }
+
   templatetypename... _Args
 void
 construct(pointer __p, _Args... __args)
Index: libstdc++-v3/include/ext/malloc_allocator.h
===
--- libstdc++-v3/include/ext/malloc_allocator.h (revision 186024)
+++ libstdc++-v3/include/ext/malloc_allocator.h (working copy)
@@ -111,6 +111,12 @@ _GLIBCXX_BEGIN_NAMESPACE_VERSION
   { ::new((void *)__p) value_type(__val); }
 
 #ifdef __GXX_EXPERIMENTAL_CXX0X__
+  // Work around PR52796 by avoiding 0-length parameter packs
+  // passed to constructors.
+  void
+  construct(pointer __p)
+  { ::new((void *)__p) _Tp

[patch][google] Backport r176665 to google/gcc-4_6 making narrowing conversions a pedwarn.

[Jeffrey Yasskin]

I'm backporting this so that, if we run into a particularly obnoxious
set of narrowing conversions in the C++11 transition, we can just
-Wno-narrowing instead of rolling a new gcc release.

Tested with check-c++ on ubuntu x86_64. Since this is going into a
google branch, I'll just commit it in a few hours unless I hear
complaints.

c-family/ChangeLog.google-4_6:
2011-11-02  Jeffrey Yasskin  jyass...@google.com

Backport from rev 176665

2011-07-22  Jason Merrill  ja...@redhat.com

* c.opt (Wnarrowing): New.

cp/ChangeLog.google-4_6:
2011-11-02  Jeffrey Yasskin  jyass...@google.com

Backport from rev 176665

2011-07-22  Jason Merrill  ja...@redhat.com

* typeck2.c (check_narrowing): Downgrade permerror to pedwarn.
Make conditional on -Wnarrowing.

testsuite/ChangeLog.google-4_6:
2011-11-02  Jeffrey Yasskin  jyass...@google.com

Backport from rev 176665

2011-07-22  Jason Merrill  ja...@redhat.com

* g++.dg/cpp0x/initlist55.C: New.

You can also review this at http://codereview.appspot.com/5330065/.

Property changes on: .
___
Modified: svn:mergeinfo
   Merged /trunk:r176665


Property changes on: libjava/classpath
___
Modified: svn:mergeinfo
   Merged /trunk/libjava/classpath:r176665

Index: gcc/c-family/c.opt
===
--- gcc/c-family/c.opt	(revision 180783)
+++ gcc/c-family/c.opt	(working copy)
@@ -493,6 +493,10 @@
 C ObjC C++ ObjC++ Warning
 Warn about use of multi-character character constants
 
+Wnarrowing
+C ObjC C++ ObjC++ Warning Var(warn_narrowing) Init(1)
+-Wno-narrowing	  In C++0x mode, ignore ill-formed narrowing conversions within { }
+
 Wnested-externs
 C ObjC Var(warn_nested_externs) Warning
 Warn about \extern\ declarations not at file scope
Property changes on: gcc/testsuite/gcc.target/powerpc/ppc-round.c
___
Modified: svn:mergeinfo
   Merged /trunk/gcc/testsuite/gcc.target/powerpc/ppc-round.c:r176665

Index: gcc/testsuite/g++.dg/cpp0x/initlist5.C
===
--- gcc/testsuite/g++.dg/cpp0x/initlist5.C	(revision 180783)
+++ gcc/testsuite/g++.dg/cpp0x/initlist5.C	(working copy)
@@ -1,5 +1,5 @@
 // Test for narrowing diagnostics
-// { dg-options -std=c++0x }
+// { dg-options -std=c++0x -pedantic-errors }
 
 #include initializer_list
 
Index: gcc/testsuite/g++.dg/cpp0x/initlist7.C
===
--- gcc/testsuite/g++.dg/cpp0x/initlist7.C	(revision 180783)
+++ gcc/testsuite/g++.dg/cpp0x/initlist7.C	(working copy)
@@ -1,5 +1,5 @@
 // PR c++/37932
-// { dg-options -std=c++0x }
+// { dg-options -std=c++0x -pedantic-errors }
 
 typedef enum { AA=1, BB=2 } my_enum;
 
Index: gcc/testsuite/g++.dg/cpp0x/initlist17.C
===
--- gcc/testsuite/g++.dg/cpp0x/initlist17.C	(revision 180783)
+++ gcc/testsuite/g++.dg/cpp0x/initlist17.C	(working copy)
@@ -1,4 +1,4 @@
-// { dg-options -std=c++0x }
+// { dg-options -std=c++0x -pedantic-errors }
 
 void f(int i);
 
Index: gcc/testsuite/g++.dg/cpp0x/initlist36.C
===
--- gcc/testsuite/g++.dg/cpp0x/initlist36.C	(revision 180783)
+++ gcc/testsuite/g++.dg/cpp0x/initlist36.C	(working copy)
@@ -1,5 +1,5 @@
 // PR c++/44358
-// { dg-options -std=c++0x }
+// { dg-options -std=c++0x -pedantic-errors }
 
 #include initializer_list
 
Index: gcc/cp/typeck2.c
===
--- gcc/cp/typeck2.c	(revision 180783)
+++ gcc/cp/typeck2.c	(working copy)
@@ -733,7 +733,7 @@
   bool ok = true;
   REAL_VALUE_TYPE d;
 
-  if (!ARITHMETIC_TYPE_P (type))
+  if (!warn_narrowing || !ARITHMETIC_TYPE_P (type))
 return;
 
   init = maybe_constant_value (init);
@@ -781,8 +781,8 @@
 }
 
   if (!ok)
-permerror (input_location, narrowing conversion of %qE from %qT to %qT inside { },
-	   init, ftype, type);
+pedwarn (input_location, OPT_Wnarrowing, narrowing conversion of %qE 
+	 from %qT to %qT inside { }, init, ftype, type);
 }
 
 /* Process the initializer INIT for a variable of type TYPE, emitting
Property changes on: gcc/config/rs6000/rs6000.c
___
Modified: svn:mergeinfo
   Merged /trunk/gcc/config/rs6000/rs6000.c:r176665


Property changes on: gcc/config/rs6000/rs6000.h
___
Modified: svn:mergeinfo
   Merged /trunk/gcc/config/rs6000/rs6000.h:r176665

Re: [patch][google] Allow static const floats unless -pedantic is passed. (issue 5306071)

[Jeffrey Yasskin]

Thanks Diego,

Here's a new version of the patch with fixes for your comments. I'll
submit it in a couple hours unless I hear objections.

On Fri, Oct 28, 2011 at 7:57 AM, Diego Novillo dnovi...@google.com wrote:
 On Thu, Oct 27, 2011 at 09:27,  jyass...@google.com wrote:
 Reviewers: Diego Novillo,

 Message:
 This patch is intended for the google/gcc-4_6 branch. Tested with make
 check-c++ on ubuntu x86-64.

 Should this go to gcc-patches@gcc.gnu.org too, or just the internal
 list?

 As you prefer.  Strictly speaking, yes, in case other C++ maintainers
 have feedback on your patch.  But given that it is a patch that you
 intend to keep in a google release only, then it does not really
 matter all that much.


 Description:
This patch allows us to migrate to C++11 more incrementally, since we can leave
the static const float initializations in place, flip the switch, and then
change them to use constexpr.

We should NOT forward-port this to any gcc-4.7 branches.


gcc/cp/ChangeLog.google-4_6
2011-10-28  Jeffrey Yasskin  jyass...@google.com

google ref 5514746; backport of r179121

Modified locally to only block static const literals in -pedantic
mode.

2011-09-23  Paolo Carlini  paolo.carl...@oracle.com

* decl.c (check_static_variable_definition): Allow in-class
initialization of static data member of non-integral type in
permissive mode.


gcc/testsuite/ChangeLog.google-4_6
2011-10-28  Jeffrey Yasskin  jyass...@google.com

google ref 5514746; backport of r179121

Modified locally to only block static const literals in -pedantic
mode.

* g++.dg/cpp0x/constexpr-static8_nonpedantic.C: New.

2011-09-23  Paolo Carlini  paolo.carl...@oracle.com

* g++.dg/cpp0x/constexpr-static8.C: New.

 You can review this at http://codereview.appspot.com/5306071/

 Affected files:
  M     gcc/cp/ChangeLog.google-4_6
  M     gcc/cp/decl.c
  M     gcc/testsuite/ChangeLog.google-4_6
  A     gcc/testsuite/g++.dg/cpp0x/constexpr-static8.C
  A     gcc/testsuite/g++.dg/cpp0x/constexpr-static8_nonpedantic.C

...
 Index: gcc/cp/decl.c
 ===
 --- gcc/cp/decl.c       (revision 180546)
 +++ gcc/cp/decl.c       (working copy)
 @@ -7508,8 +7508,12 @@ check_static_variable_definition (tree decl, tree
   else if (cxx_dialect = cxx0x  !INTEGRAL_OR_ENUMERATION_TYPE_P (type))
     {
       if (literal_type_p (type))
 -       error (%constexpr% needed for in-class initialization of static 
 -              data member %q#D of non-integral type, decl);
 +        {
 +          pedwarn (input_location, OPT_pedantic,
 +                   %constexpr% needed for in-class initialization of 
 +                   static data member %q#D of non-integral type, decl);
 +          return 0;
 +        }

 Add a 'FIXME google' here?  Describe why this is different than
 upstream.  Helps with merge conflicts.

Done.

 OK with those changes.


 Diego.

Index: gcc/testsuite/g++.dg/cpp0x/constexpr-static8.C
===
--- gcc/testsuite/g++.dg/cpp0x/constexpr-static8.C	(revision 0)
+++ gcc/testsuite/g++.dg/cpp0x/constexpr-static8.C	(revision 0)
@@ -0,0 +1,7 @@
+// PR c++/50258
+// { dg-options -std=c++0x -pedantic }
+
+struct Foo {
+  static const double d = 3.14; // { dg-warning constexpr }
+};
+const double Foo::d;
Index: gcc/testsuite/g++.dg/cpp0x/constexpr-static8_nonpedantic.C
===
--- gcc/testsuite/g++.dg/cpp0x/constexpr-static8_nonpedantic.C	(revision 0)
+++ gcc/testsuite/g++.dg/cpp0x/constexpr-static8_nonpedantic.C	(revision 0)
@@ -0,0 +1,7 @@
+// PR c++/50258
+// { dg-options -std=c++0x }
+
+struct Foo {
+  static const double d = 3.14; // no warning
+};
+const double Foo::d;
Index: gcc/cp/decl.c
===
--- gcc/cp/decl.c	(revision 180546)
+++ gcc/cp/decl.c	(working copy)
@@ -7508,8 +7508,18 @@
   else if (cxx_dialect = cxx0x  !INTEGRAL_OR_ENUMERATION_TYPE_P (type))
 {
   if (literal_type_p (type))
-	error (%constexpr% needed for in-class initialization of static 
-	   data member %q#D of non-integral type, decl);
+{
+  /* FIXME google: This local modification allows us to
+ transition from C++98 to C++11 without moving static
+ const floats out of the class during the transition.  It
+ should not be forward-ported to a 4.7 branch, since by
+ then we should be able to just fix the code to use
+ constexpr.  */
+  pedwarn (input_location, OPT_pedantic,
+   %constexpr% needed for in-class initialization of 
+   static data member %q#D of non-integral type, decl);
+  return 0;
+}
   else
 	error (in-class initialization of static data member

Re: [PATCH] Propagate source locations from function_decls to their template_decls

[Jeffrey Yasskin]

Thanks. Committed to gcc-4_6-branch in r176851 and the google/gcc-4_6
branch in r177072.

On Wed, Jul 27, 2011 at 11:46 AM, Jason Merrill ja...@redhat.com wrote:
 Yes.

 Jeffrey Yasskin jyass...@google.com wrote:

 Thanks. I'll commit to trunk in the morning when I can be around to
 watch for breakage.

 Is this also ok for gcc-4_6-branch?

 On Tue, Jul 26, 2011 at 7:16 PM, Jason Merrill ja...@redhat.com wrote:
 Ok.

 Jeffrey Yasskin jyass...@google.com wrote:

 Hi Jason. Paolo suggested I ping you directly about this patch for the
 C++ parser. Thanks in advance for taking a look.

 On Tue, Jul 26, 2011 at 2:20 PM, Jeffrey Yasskin jyass...@google.com wrote:
 This patch copies the source location of a FUNCTION_DECL to the
 TEMPLATE_DECL that build_template_decl() builds out of it. Otherwise,
 the TEMPLATE_DECL's location becomes input_location, which is the end
 of the parameter list, while the FUNCTION_DECL's location is the
 location of the name of the function. Depending on what order
 templates are defined and used, gcc may emit either the
 FUNCTION_DECL's or TEMPLATE_DECL's location into the debug location,
 which causes gold's ODR checker to emit false positives.

 Tested with a bootstrap+`make -k check-c++` on
 x86_64-unknown-linux-gnu. I'm looking to check it in to trunk, and
 will propagate it to the gcc-4_6-branch if you think that's the right
 thing to do.

 No more tests fail than in
 http://gcc.gnu.org/ml/gcc-testresults/2011-07/msg02995.html.

 gcc/cp/ChangeLog:
 2011-07-26 ? Jeffrey Yasskin ?jyass...@google.com

 ? ? ? ?* pt.c (build_template_decl): Copy the function_decl's source
 ? ? ? ?location to the new template_decl.

 gcc/testsuite/ChangeLog:
 2011-07-26 ? Jeffrey Yasskin ?jyass...@google.com

 ? ? ? ?* g++.old-deja/g++.pt/crash60.C: Updated.

 libstdc++-v3/ChangeLog:
 2011-07-26 ? Jeffrey Yasskin ?jyass...@google.com

 ? ? ? ?* testsuite/20_util/weak_ptr/comparison/cmp_neg.cc: Updated.

Re: [google][patch] Put make_heap's declaration on a single line to work around inconsistent debug locations

[Jeffrey Yasskin]

And reverted in r177083 because I fixed the underlying problem.

On Fri, Jul 22, 2011 at 11:06 PM, Ollie Wild a...@google.com wrote:
 Ok for inclusion in google/gcc-4_6.

 Ollie

 On Fri, Jul 22, 2011 at 5:46 PM, Jeffrey Yasskin jyass...@google.com wrote:

 For the google/gcc-4_6 branch _only_. I'll fix the inconsistency in
 debug locations later for trunk and google/main.

 In some translation units the debug location of make_heap is the
 location of its name; in other TUs it's the location of the closing
 ')'. This causes false positives in gold's ODR checker. Until I can
 find why the locations are inconsistent, we can work around the
 problem by putting the name and closing ')' on the same line.

 libstdc++-v3/ChangeLog
 2011-07-22   Jeffrey Yasskin  jyass...@google.com

        * include/bits/stl_heap.h(make_heap): Remove a newline.

Re: [PATCH] Propagate source locations from function_decls to their template_decls

[Jeffrey Yasskin]

Thanks. I'll commit to trunk in the morning when I can be around to
watch for breakage.

Is this also ok for gcc-4_6-branch?

On Tue, Jul 26, 2011 at 7:16 PM, Jason Merrill ja...@redhat.com wrote:
 Ok.

 Jeffrey Yasskin jyass...@google.com wrote:

 Hi Jason. Paolo suggested I ping you directly about this patch for the
 C++ parser. Thanks in advance for taking a look.

 On Tue, Jul 26, 2011 at 2:20 PM, Jeffrey Yasskin jyass...@google.com wrote:
 This patch copies the source location of a FUNCTION_DECL to the
 TEMPLATE_DECL that build_template_decl() builds out of it. Otherwise,
 the TEMPLATE_DECL's location becomes input_location, which is the end
 of the parameter list, while the FUNCTION_DECL's location is the
 location of the name of the function. Depending on what order
 templates are defined and used, gcc may emit either the
 FUNCTION_DECL's or TEMPLATE_DECL's location into the debug location,
 which causes gold's ODR checker to emit false positives.

 Tested with a bootstrap+`make -k check-c++` on
 x86_64-unknown-linux-gnu. I'm looking to check it in to trunk, and
 will propagate it to the gcc-4_6-branch if you think that's the right
 thing to do.

 No more tests fail than in
 http://gcc.gnu.org/ml/gcc-testresults/2011-07/msg02995.html.

 gcc/cp/ChangeLog:
 2011-07-26 ? Jeffrey Yasskin ?jyass...@google.com

 ? ? ? ?* pt.c (build_template_decl): Copy the function_decl's source
 ? ? ? ?location to the new template_decl.

 gcc/testsuite/ChangeLog:
 2011-07-26 ? Jeffrey Yasskin ?jyass...@google.com

 ? ? ? ?* g++.old-deja/g++.pt/crash60.C: Updated.

 libstdc++-v3/ChangeLog:
 2011-07-26 ? Jeffrey Yasskin ?jyass...@google.com

 ? ? ? ?* testsuite/20_util/weak_ptr/comparison/cmp_neg.cc: Updated.

[PATCH] Propagate source locations from function_decls to their template_decls

[Jeffrey Yasskin]

This patch copies the source location of a FUNCTION_DECL to the
TEMPLATE_DECL that build_template_decl() builds out of it. Otherwise,
the TEMPLATE_DECL's location becomes input_location, which is the end
of the parameter list, while the FUNCTION_DECL's location is the
location of the name of the function. Depending on what order
templates are defined and used, gcc may emit either the
FUNCTION_DECL's or TEMPLATE_DECL's location into the debug location,
which causes gold's ODR checker to emit false positives.

Tested with a bootstrap+`make -k check-c++` on
x86_64-unknown-linux-gnu. I'm looking to check it in to trunk, and
will propagate it to the gcc-4_6-branch if you think that's the right
thing to do.

No more tests fail than in
http://gcc.gnu.org/ml/gcc-testresults/2011-07/msg02995.html.

gcc/cp/ChangeLog:
2011-07-26   Jeffrey Yasskin  jyass...@google.com

* pt.c (build_template_decl): Copy the function_decl's source
location to the new template_decl.

gcc/testsuite/ChangeLog:
2011-07-26   Jeffrey Yasskin  jyass...@google.com

* g++.old-deja/g++.pt/crash60.C: Updated.

libstdc++-v3/ChangeLog:
2011-07-26   Jeffrey Yasskin  jyass...@google.com

* testsuite/20_util/weak_ptr/comparison/cmp_neg.cc: Updated.
diff --git a/gcc/cp/pt.c b/gcc/cp/pt.c
index 178685c..b9e09af 100644
--- a/gcc/cp/pt.c
+++ b/gcc/cp/pt.c
@@ -4121,6 +4121,7 @@ build_template_decl (tree decl, tree parms, bool member_template_p)
   tree tmpl = build_lang_decl (TEMPLATE_DECL, DECL_NAME (decl), NULL_TREE);
   DECL_TEMPLATE_PARMS (tmpl) = parms;
   DECL_CONTEXT (tmpl) = DECL_CONTEXT (decl);
+  DECL_SOURCE_LOCATION (tmpl) = DECL_SOURCE_LOCATION (decl);
   DECL_MEMBER_TEMPLATE_P (tmpl) = member_template_p;
 
   return tmpl;
diff --git a/gcc/testsuite/g++.old-deja/g++.pt/crash60.C b/gcc/testsuite/g++.old-deja/g++.pt/crash60.C
index 747af9b..1be4678 100644
--- a/gcc/testsuite/g++.old-deja/g++.pt/crash60.C
+++ b/gcc/testsuite/g++.old-deja/g++.pt/crash60.C
@@ -5,9 +5,9 @@
 // We ICE'd rather than fail to instantiate.
 
 template typename SID, class SDR 
-void k( SID sid, SDR* p,
+void k( SID sid, SDR* p,	// { dg-error no type named 'T' }
  void (SDR::*)
- ( typename SID::T ) );		// { dg-error no type named 'T' }
+ ( typename SID::T ) );
 
 struct E { };
 struct S { void f( int ); };
diff --git a/libstdc++-v3/testsuite/20_util/weak_ptr/comparison/cmp_neg.cc b/libstdc++-v3/testsuite/20_util/weak_ptr/comparison/cmp_neg.cc
index df18712..6eecc2d 100644
--- a/libstdc++-v3/testsuite/20_util/weak_ptr/comparison/cmp_neg.cc
+++ b/libstdc++-v3/testsuite/20_util/weak_ptr/comparison/cmp_neg.cc
@@ -44,16 +44,16 @@ main()
 
 // { dg-warning note  { target *-*-* } 370 }
 // { dg-warning note  { target *-*-* } 365 }
-// { dg-warning note  { target *-*-* } 357 }
+// { dg-warning note  { target *-*-* } 356 }
 // { dg-warning note  { target *-*-* } 1103 }
 // { dg-warning note  { target *-*-* } 1098 }
-// { dg-warning note  { target *-*-* } 1090 }
+// { dg-warning note  { target *-*-* } 1089 }
 // { dg-warning note  { target *-*-* } 485 }
 // { dg-warning note  { target *-*-* } 479 }
-// { dg-warning note  { target *-*-* } 469 }
-// { dg-warning note  { target *-*-* } 814 }
-// { dg-warning note  { target *-*-* } 1056 }
-// { dg-warning note  { target *-*-* } 1050 }
-// { dg-warning note  { target *-*-* } 342 }
-// { dg-warning note  { target *-*-* } 292 }
+// { dg-warning note  { target *-*-* } 468 }
+// { dg-warning note  { target *-*-* } 813 }
+// { dg-warning note  { target *-*-* } 1055 }
+// { dg-warning note  { target *-*-* } 1049 }
+// { dg-warning note  { target *-*-* } 341 }
+// { dg-warning note  { target *-*-* } 291 }
 // { dg-warning note  { target *-*-* } 224 }

[google][patch] Put make_heap's declaration on a single line to work around inconsistent debug locations

[Jeffrey Yasskin]

For the google/gcc-4_6 branch _only_. I'll fix the inconsistency in
debug locations later for trunk and google/main.

In some translation units the debug location of make_heap is the
location of its name; in other TUs it's the location of the closing
')'. This causes false positives in gold's ODR checker. Until I can
find why the locations are inconsistent, we can work around the
problem by putting the name and closing ')' on the same line.

libstdc++-v3/ChangeLog
2011-07-22   Jeffrey Yasskin  jyass...@google.com

* include/bits/stl_heap.h(make_heap): Remove a newline.
Index: libstdc++-v3/include/bits/stl_heap.h
===
--- libstdc++-v3/include/bits/stl_heap.h	(revision 176544)
+++ libstdc++-v3/include/bits/stl_heap.h	(working copy)
@@ -412,8 +412,7 @@
   */
   templatetypename _RandomAccessIterator, typename _Compare
 void
-make_heap(_RandomAccessIterator __first, _RandomAccessIterator __last,
-	  _Compare __comp)
+make_heap(_RandomAccessIterator __first, _RandomAccessIterator __last, _Compare __comp)
 {
   typedef typename iterator_traits_RandomAccessIterator::value_type
 	  _ValueType;

[google] Merge r173574 to google/gcc-4_6 to fix an incompatibility between C++98 and C++0x (issue4592057)

[Jeffrey Yasskin]

In C++0x mode, without this patch, calls to a user-defined trunc() function 
with an argument in namespace std and a parameter type that has an implicit 
conversion from the argument's type, cause infinite recursion in std::trunc().

This patch also includes 
http://gcc.gnu.org/viewcvs/trunk/libstdc%2B%2B-v3/testsuite/26_numerics/headers/cmath/overloads_c%2B%2B0x_neg.cc?view=markuppathrev=173574
 and 
http://gcc.gnu.org/viewcvs/trunk/libstdc%2B%2B-v3/testsuite/tr1/8_c_compatibility/cmath/overloads_neg.cc?view=markuppathrev=173574,
 but `svn diff` didn't capture them.

Tested with `make check-c++` on x86_64-unknown-linux-gnu.

2011-06-14  Jeffrey Yasskin  jyass...@google.com

Merge r173574 to google/gcc-4_6.
* include/c_global/cmath (acosh, asinh, atanh, cbrt, copysign,
erf, erfc, exp2, expm1, fdim, fma, fmax, hypot, ilogb, lgamma,
llrint, llround, log1p, log2, logb, lrint, lround, nearbyint,
nextafter, nexttoward, remainder, remquo, rint, round, scalbln,
scalbn, tgamma, trunc): Use __enable_if on the return type.
* include/tr1/cmath: Likewise.
* testsuite/26_numerics/headers/cmath/overloads_c++0x_neg.cc: New.
* testsuite/tr1/8_c_compatibility/cmath/overloads_neg.cc: Likewise.

Property changes on: .
___
Modified: svn:mergeinfo
   Merged /trunk:r173574

Index: libstdc++-v3/include/c_global/cmath
===
--- libstdc++-v3/include/c_global/cmath (revision 175001)
+++ libstdc++-v3/include/c_global/cmath (working copy)
@@ -1,7 +1,7 @@
 // -*- C++ -*- C forwarding header.
 
 // Copyright (C) 1997, 1998, 1999, 2000, 2001, 2002, 2003, 2004, 2005,
-// 2006, 2007, 2008, 2009, 2010
+// 2006, 2007, 2008, 2009, 2010, 2011
 // Free Software Foundation, Inc.
 //
 // This file is part of the GNU ISO C++ Library.  This library is free
@@ -1120,12 +1120,10 @@
   { return __builtin_acoshl(__x); }
 
   templatetypename _Tp
-inline typename __gnu_cxx::__promote_Tp::__type 
+inline typename __gnu_cxx::__enable_if__is_integer_Tp::__value, 
+  double::__type
 acosh(_Tp __x)
-{
-  typedef typename __gnu_cxx::__promote_Tp::__type __type;
-  return acosh(__type(__x));
-}
+{ return __builtin_acosh(__x); }
 
   inline float
   asinh(float __x)
@@ -1136,12 +1134,10 @@
   { return __builtin_asinhl(__x); }
 
   templatetypename _Tp
-inline typename __gnu_cxx::__promote_Tp::__type 
+inline typename __gnu_cxx::__enable_if__is_integer_Tp::__value, 
+  double::__type
 asinh(_Tp __x)
-{
-  typedef typename __gnu_cxx::__promote_Tp::__type __type;
-  return asinh(__type(__x));
-}
+{ return __builtin_asinh(__x); }
 
   inline float
   atanh(float __x)
@@ -1152,12 +1148,10 @@
   { return __builtin_atanhl(__x); }
 
   templatetypename _Tp
-inline typename __gnu_cxx::__promote_Tp::__type 
+inline typename __gnu_cxx::__enable_if__is_integer_Tp::__value, 
+  double::__type
 atanh(_Tp __x)
-{
-  typedef typename __gnu_cxx::__promote_Tp::__type __type;
-  return atanh(__type(__x));
-}
+{ return __builtin_atanh(__x); }
 
   inline float
   cbrt(float __x)
@@ -1168,12 +1162,10 @@
   { return __builtin_cbrtl(__x); }
 
   templatetypename _Tp
-inline typename __gnu_cxx::__promote_Tp::__type 
+inline typename __gnu_cxx::__enable_if__is_integer_Tp::__value, 
+  double::__type
 cbrt(_Tp __x)
-{
-  typedef typename __gnu_cxx::__promote_Tp::__type __type;
-  return cbrt(__type(__x));
-}
+{ return __builtin_cbrt(__x); }
 
   inline float
   copysign(float __x, float __y)
@@ -1184,7 +1176,11 @@
   { return __builtin_copysignl(__x, __y); }
 
   templatetypename _Tp, typename _Up
-inline typename __gnu_cxx::__promote_2_Tp, _Up::__type
+inline
+typename __gnu_cxx::__promote_2
+typename __gnu_cxx::__enable_if__is_arithmetic_Tp::__value
+__is_arithmetic_Up::__value,
+   _Tp::__type, _Up::__type
 copysign(_Tp __x, _Up __y)
 {
   typedef typename __gnu_cxx::__promote_2_Tp, _Up::__type __type;
@@ -1200,12 +1196,10 @@
   { return __builtin_erfl(__x); }
 
   templatetypename _Tp
-inline typename __gnu_cxx::__promote_Tp::__type 
+inline typename __gnu_cxx::__enable_if__is_integer_Tp::__value, 
+  double::__type
 erf(_Tp __x)
-{
-  typedef typename __gnu_cxx::__promote_Tp::__type __type;
-  return erf(__type(__x));
-}
+{ return __builtin_erf(__x); }
 
   inline float
   erfc(float __x)
@@ -1216,12 +1210,10 @@
   { return __builtin_erfcl(__x); }
 
   templatetypename _Tp
-inline typename __gnu_cxx::__promote_Tp::__type 
+inline

Re: [google] Merge r173574 to google/gcc-4_6 to fix an incompatibility between C++98 and C++0x (issue4592057)

[Jeffrey Yasskin]

On Tue, Jun 14, 2011 at 12:38 PM, Diego Novillo dnovi...@google.com wrote:
 On Tue, Jun 14, 2011 at 14:45, Jeffrey Yasskin jyass...@google.com wrote:
 In C++0x mode, without this patch, calls to a user-defined trunc() function 
 with an argument in namespace std and
 a parameter type that has an implicit conversion from the argument's type, 
 cause infinite recursion in std::trunc().

 This patch also includes 
 http://gcc.gnu.org/viewcvs/trunk/libstdc%2B%2B-v3/testsuite/26_numerics/headers/cmath/overloads_c%2B%2B0x_neg.cc?view=markuppathrev=173574
  and 
 http://gcc.gnu.org/viewcvs/trunk/libstdc%2B%2B-v3/testsuite/tr1/8_c_compatibility/cmath/overloads_neg.cc?view=markuppathrev=173574,
  but `svn diff` didn't capture them.

 Yeah, svn diff never picks up added files.  Silly thing.


 Tested with `make check-c++` on x86_64-unknown-linux-gnu.

 2011-06-14  Jeffrey Yasskin  jyass...@google.com

        Merge r173574 to google/gcc-4_6.
        * include/c_global/cmath (acosh, asinh, atanh, cbrt, copysign,
        erf, erfc, exp2, expm1, fdim, fma, fmax, hypot, ilogb, lgamma,
        llrint, llround, log1p, log2, logb, lrint, lround, nearbyint,
        nextafter, nexttoward, remainder, remquo, rint, round, scalbln,
        scalbn, tgamma, trunc): Use __enable_if on the return type.
        * include/tr1/cmath: Likewise.
        * testsuite/26_numerics/headers/cmath/overloads_c++0x_neg.cc: New.
        * testsuite/tr1/8_c_compatibility/cmath/overloads_neg.cc: Likewise.


 Any reason not to put this in google/main?

It's already in trunk, so my impression was that it was going to be
automatically merged to google/main. I only need a manual merge to get
it into our release branches.

Jeffrey

[patch] Import iota and is_sorted from std to __gnu_cxx in C++0x mode. (issue4384043)

[Jeffrey Yasskin]

In C++0x mode, import iota and is_sorted from namespace std into
namespace __gnu_cxx to avoid ambiguous call errors.  copy_n would be a
good candidate for this too, except it was standardized with a
different return type.

I think this is a candidate to backport to libstdc++-4.6.1.

Tested:
  `make -k check-c++` on x86_64-unknown-linux-gnu.  (still running)

libstdc++-v3/ChangeLog:

2011-04-06  Jeffrey Yasskin  jyass...@google.com

* include/ext/algorithm (is_sorted): In C++0x mode import from
namespace std.
* include/ext/numeric (iota): In C++0x mode import from
namespace std.
* testsuite/ext/is_sorted/cxx0x.cc: New.
* testsuite/ext/iota/cxx0x.cc: New.

diff --git a/libstdc++-v3/include/ext/algorithm 
b/libstdc++-v3/include/ext/algorithm
index 368b591..417a03a 100644
--- a/libstdc++-v3/include/ext/algorithm
+++ b/libstdc++-v3/include/ext/algorithm
@@ -471,6 +471,9 @@ _GLIBCXX_BEGIN_NAMESPACE_VERSION
 }
 #endif
 
+#ifdef __GXX_EXPERIMENTAL_CXX0X__
+  using std::is_sorted;
+#else
   // is_sorted, a predicated testing whether a range is sorted in
   // nondescending order.  This is an extension, not part of the C++
   // standard.
@@ -526,6 +529,7 @@ _GLIBCXX_BEGIN_NAMESPACE_VERSION
  return false;
   return true;
 }
+#endif  // __GXX_EXPERIMENTAL_CXX0X__
 
   /**
*  @brief Find the median of three values.
diff --git a/libstdc++-v3/include/ext/numeric b/libstdc++-v3/include/ext/numeric
index d4a367f..b389177 100644
--- a/libstdc++-v3/include/ext/numeric
+++ b/libstdc++-v3/include/ext/numeric
@@ -123,6 +123,9 @@ _GLIBCXX_BEGIN_NAMESPACE_VERSION
 power(_Tp __x, _Integer __n)
 { return __power(__x, __n); }
 
+#ifdef __GXX_EXPERIMENTAL_CXX0X__
+  using std::iota;
+#else
   /**
*  This is an SGI extension.
*  @ingroup SGIextensions
@@ -141,9 +144,9 @@ _GLIBCXX_BEGIN_NAMESPACE_VERSION
   while (__first != __last)
*__first++ = __value++;
 }
+#endif  // __GXX_EXPERIMENTAL_CXX0X__
 
 _GLIBCXX_END_NAMESPACE_VERSION
 } // namespace
 
 #endif
-
diff --git a/libstdc++-v3/testsuite/ext/iota/cxx0x.cc 
b/libstdc++-v3/testsuite/ext/iota/cxx0x.cc
new file mode 100644
index 000..84917ad
--- /dev/null
+++ b/libstdc++-v3/testsuite/ext/iota/cxx0x.cc
@@ -0,0 +1,28 @@
+// { dg-do compile }
+// { dg-options -std=gnu++0x }
+
+// Copyright (C) 2011 Free Software Foundation, Inc.
+//
+// This file is part of the GNU ISO C++ Library.  This library is free
+// software; you can redistribute it and/or modify it under the
+// terms of the GNU General Public License as published by the
+// Free Software Foundation; either version 3, or (at your option)
+// any later version.
+
+// This library is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU General Public License for more details.
+
+// You should have received a copy of the GNU General Public License along
+// with this library; see the file COPYING3.  If not see
+// http://www.gnu.org/licenses/.
+
+#include ext/numeric
+#include vector
+
+void foo()
+{
+  std::vectorint v;
+  iota(v.begin(), v.end(), 0);
+}
diff --git a/libstdc++-v3/testsuite/ext/is_sorted/cxx0x.cc 
b/libstdc++-v3/testsuite/ext/is_sorted/cxx0x.cc
new file mode 100644
index 000..518716c
--- /dev/null
+++ b/libstdc++-v3/testsuite/ext/is_sorted/cxx0x.cc
@@ -0,0 +1,28 @@
+// { dg-do compile }
+// { dg-options -std=gnu++0x }
+
+// Copyright (C) 2011 Free Software Foundation, Inc.
+//
+// This file is part of the GNU ISO C++ Library.  This library is free
+// software; you can redistribute it and/or modify it under the
+// terms of the GNU General Public License as published by the
+// Free Software Foundation; either version 3, or (at your option)
+// any later version.
+
+// This library is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU General Public License for more details.
+
+// You should have received a copy of the GNU General Public License along
+// with this library; see the file COPYING3.  If not see
+// http://www.gnu.org/licenses/.
+
+#include ext/algorithm
+#include vector
+
+void foo()
+{
+  std::vectorint v;
+  is_sorted(v.begin(), v.end());
+}

--
This patch is available for review at http://codereview.appspot.com/4384043

[patch] Fix two C++ errors in libstdc++. (issue4341041)

[Jeffrey Yasskin]

I think this is a candidate to backport to libstdc++-4.6.1.

exception_ptr.h needs the forward declaration because it's
included from typeinfo before typeinfo defines std::type_info.

Tested:
  `make check` in x86_64-unknown-linux-gnu/libstdc++-v3.  The
  abi_check test fails, but also did before this change.

libstdc++-v3/ChangeLog:

2011-03-31  Jeffrey Yasskin  jyass...@google.com

* libsupc++/exception_ptr.h: Forward-declare std::type_info.
* libsupc++/nested_exception.h (__throw_with_nested): Remove a
redundant default argument from std::__throw_with_nested.

diff --git a/libstdc++-v3/libsupc++/exception_ptr.h 
b/libstdc++-v3/libsupc++/exception_ptr.h
index ef826f6..26117cd 100644
--- a/libstdc++-v3/libsupc++/exception_ptr.h
+++ b/libstdc++-v3/libsupc++/exception_ptr.h
@@ -137,7 +137,7 @@ namespace std
   operator==(const exception_ptr, const exception_ptr) throw() 
   __attribute__ ((__pure__));
 
-  const type_info*
+  const class type_info*
   __cxa_exception_type() const throw() __attribute__ ((__pure__));
 };
 
diff --git a/libstdc++-v3/libsupc++/nested_exception.h 
b/libstdc++-v3/libsupc++/nested_exception.h
index 6a4f04e..d4804bb 100644
--- a/libstdc++-v3/libsupc++/nested_exception.h
+++ b/libstdc++-v3/libsupc++/nested_exception.h
@@ -117,7 +117,7 @@ namespace std
   // with a type that has an accessible nested_exception base.
   templatetypename _Ex
 inline void
-__throw_with_nested(_Ex __ex, const nested_exception* = 0)
+__throw_with_nested(_Ex __ex, const nested_exception*)
 { throw __ex; }
 
   templatetypename _Ex

--
This patch is available for review at http://codereview.appspot.com/4341041

Re: [patch] Fix two C++ errors in libstdc++. (issue4341041)

[Jeffrey Yasskin]

On Thu, Mar 31, 2011 at 12:49 PM, Paolo Carlini
paolo.carl...@oracle.com wrote:
 On 03/31/2011 09:41 PM, Paolo Carlini wrote:

 I think this is a candidate to backport to libstdc++-4.6.1.

 exception_ptr.h needs the forward declaration because it's
 included fromtypeinfo  beforetypeinfo  defines std::type_info.

 before going ahead with the patch - I'm not at all sure it's enough, for
 exception itself for example

 I had a closer look to what we have and indeed the patch seems enough to
 solve the existing issues. Thanks. I think we want indeed to apply the
 patchlet to mainline and 4_6-branch, but my basic puzzlement stays, indeed
 it's the real reason why the problem remained unnoticed for so much time!

Thanks. Is that an OK that I can commit this to mainline and 4_6-branch?

[For libstdc++@ people, the patch is at
http://codereview.appspot.com/4341041 and
http://gcc.gnu.org/ml/gcc-patches/2011-03/msg02261.html]