Boost logo

Boost Testing :

Subject: Re: [Boost-testing] [boost]filesystem/v2 failure breaking Darwin testing
From: Belcourt, Kenneth (kbelco_at_[hidden])
Date: 2010-06-16 11:10:29


Hi Beman,

On Jun 15, 2010, at 6:33 PM, Beman Dawes wrote:

> On Tue, Jun 15, 2010 at 1:47 PM, Belcourt, Kenneth
> <kbelco_at_[hidden]> wrote:
>>
>> On Jun 14, 2010, at 11:43 AM, Beman Dawes wrote:
>>
>> ...
>>> I've committed a fix, but it's untested. Please let me know if the
>>> failure persists.
>>
>> Darwin looks okay though it looks like gcc-3.4.6 and Sun are still
>> broken.
>> Here's the gcc-3.4.6 error.
>> ...
>
> The gcc-3.4.6 problem seems to be an overload resolution bug that must
> have been fixed in later gcc versions.

Still broken with revision 63011.

gcc.compile.c++ /tmp/kbelco/boost/boost/bin.v2/libs/filesystem/build/
gcc-3.4.6/release/link-static/v3/src/path.o

"/sierra/Sntools/extras/compilers/gcc-3.4.6/bin/g++" -ftemplate-
depth-128 -O3 -finline-functions -Wno-inline -Wall -
DBOOST_ALL_NO_LIB=1 -DBOOST_SYSTEM_STATIC_LINK=1 -DNDEBUG -I"/tmp/
kbelco/boost/boost" -c -o "/tmp/kbelco/boost/boost/bin.v2/libs/
filesystem/build/gcc-3.4.6/release/link-static/v3/src/path.o" "/tmp/
kbelco/boost/boost/libs/filesystem/v3/src/path.cpp"

/tmp/kbelco/boost/boost/boost/filesystem/v3/path.hpp: In constructor
`boost::filesystem3::path::path(const Source&) [with Source = char[2]]':
/tmp/kbelco/boost/boost/libs/filesystem/v3/src/path.cpp:87:
instantiated from here/tmp/kbelco/boost/boost/boost/filesystem/v3/
path.hpp:161: error: call of overloaded `dispatch(const char[2],
std::basic_string<char, std::char_traits<char>, std::allocator<char>
>&, const std::codecvt<wchar_t, char, mbstate_t>&)' is ambiguous
/tmp/kbelco/boost/boost/boost/filesystem/v3/path_traits.hpp:140: note:
candidates are: void boost::filesystem3::path_traits::dispatch(const
std::string&, U&, const
boost::filesystem3::path_traits::codecvt_type&) [with U =
std::basic_string<char, std::char_traits<char>, std::allocator<char> >]
/tmp/kbelco/boost/boost/boost/filesystem/v3/path_traits.hpp:166:
note: void
boost::filesystem3::path_traits::dispatch(const Container&,
U&, const boost::filesystem3::path_traits::codecvt_type&) [with
Container = char[2], U = std::basic_string<char,
std::char_traits<char>, std::allocator<char> >]/tmp/kbelco/boost/boost/
boost/filesystem/v3/path_traits.hpp:186: note: void
boost::filesystem3::path_traits::dispatch(T (&)[N], U&, cons
t boost::filesystem3::path_traits::codecvt_type&) [with T = const
char, long unsigned int N = 2ul, U = std::basic_string<char,
std::char_traits<char>, s
td::allocator<char> >]
/tmp/kbelco/boost/boost/boost/filesystem/v3/path.hpp: In constructor
`boost::filesystem3::path::path(const Source&) [with Source = char[3]]':
/tmp/kbelco/boost/boost/libs/filesystem/v3/src/path.cpp:88:
instantiated from here
/tmp/kbelco/boost/boost/boost/filesystem/v3/path.hpp:161: error: call
of overloaded `dispatch(const char[3], std::basic_string<char,
std::char_traits<char>, std::allocator<char> >&, const
std::codecvt<wchar_t, char, mbstate_t>&)' is ambiguous
/tmp/kbelco/boost/boost/boost/filesystem/v3/path_traits.hpp:140: note:
candidates are: void boost::filesystem3::path_traits::dispatch(const
std::string&, U&, const
boost::filesystem3::path_traits::codecvt_type&) [with U =
std::basic_string<char, std::char_traits<char>, std::allocator<char> >]
/tmp/kbelco/boost/boost/boost/filesystem/v3/path_traits.hpp:166:
note: void
boost::filesystem3::path_traits::dispatch(const Container&, U&, const
boost::filesystem3::path_traits::codecvt_type&) [with Container =
char[3], U = std::basic_string<char, std::char_traits<char>,
std::allocator<char> >]
/tmp/kbelco/boost/boost/boost/filesystem/v3/path_traits.hpp:186:
note: void boost::filesystem3::path_traits::dispatch(T
(&)[N], U&, const boost::filesystem3::path_traits::codecvt_type&)
[with T = const char, long unsigned int N = 3ul, U =
std::basic_string<char, std::char_traits<char>, std::allocator<char> >]

...failed gcc.compile.c++ /tmp/kbelco/boost/boost/bin.v2/libs/
filesystem/build/gcc-3.4.6/release/link-static/v3/src/path.o...

> I'm guessing from the line numbers that the gcc-3.4.6 tests were run
> before revision 62976. That fixed a clang problem in the same code, so
> I'm wondering if 62976 happened to fix the gcc-3.4.6 problem too.

No, doesn't look like it.

> I have an unrelated problem with the Sun compiler, so I'll try to get
> Solaris running in a virtual machine to allow testing here. Otherwise
> I'll have to ask you to try some different things.

-- Noel


Boost-testing list run by mbergal at meta-comm.com