Re: [boost] [thread] Failed assertions under Borland?

From: Anthony Williams (anthony_w.geo_at_[hidden])
Date: 2007-10-24 06:04:32

"John Maddock" <john_at_[hidden]> writes:

> Anthony Williams wrote:
>> "John Maddock" <john_at_[hidden]> writes:
>>
>>> I'm seeing new regex regressions caused by assertions being fired in
>>> Boost.Thread: http://tinyurl.com/2okqzn  Any ideas?
>>
>> It looks like list_entry is not being zeroed out with list_entry().
>> I've committed a change that should fix that (by adding a proper
>> default constructor), but I'm having lots of problems with Borland
>> crashing in the test framework, so I don't know if it fixes the
>> problem or not.
>
> Hmm, now I'm seeing a different problem:
>
> borland.compile.c++ 
> ..\..\..\bin.v2\libs\thread\build\borland-5.6.4\debug\thread
> ing-multi\thread.obj
> ..\..\..\libs\thread\src\win32\thread.cpp:
> Error E2015 ..\..\..\libs\thread\src\win32\thread.cpp 187: Ambiguity between 
> 'thread::operator move_t<thread>()' and 'move_t<thread>::move_t(thread &)' 
> in function thread::operator move_t<thread>()
> Error E2015 ..\..\..\libs\thread\src\win32\thread.cpp 192: Ambiguity between 
> 'thread::operator move_t<thread>()' and 'move_t<thread>::move_t(thread &)' 
> in function thread::move()
> *** 2 errors in Compile ***

That code is there to support move semantics for thread. It compiles OK on
Borland 5.8.2.

This should be a direct constructor call. Any ideas for a Borland 5.6.4
workaround?

Anthony
-- 
Anthony Williams
Just Software Solutions Ltd - http://www.justsoftwaresolutions.co.uk
Registered in England, Company Number 5478976.
Registered Office: 15 Carrallack Mews, St Just, Cornwall, TR19 7UL