Re: [boost] Re: SunPro CC & less<smart_ptr<T> >

From: George A. Heintzelman (georgeh_at_[hidden])
Date: 2001-12-10 08:51:17

> From: "George A. Heintzelman" <georgeh_at_[hidden]>
> > > Turn BOOST_NO_PARTIAL_SPECIALIZATION on. It's working in majority
> > > cases.
> >
> > My experience so far is that it is not needed in many cases. Several of
> > the test files (E.g., bind_test) seem to work better with it off than
> > with it on.
> 
> Sorry, what is "on" in the above?

"on" means with BOOST_NO_PARTIAL_SPECIALIZATION defined. 

> Judging by the regression test output I decided that SunPro handles the
> partial specialization version of boost::_bi::add_value fine but has trouble
> with the MSVC-style workaround, so the condition there is now
> 
> #if !defined(BOOST_NO_TEMPLATE_PARTIAL_SPECIALIZATION) || (__SUNPRO_CC >=
> 0x530)

Right. I've just removed the #define in the config/compiler/supro_cc.hpp
 for SUNPRO_CC >= 0x530. As far as I can tell it doesn't cause any new 
failures elsewhere, but this probably should be checked more thoroughly.

> Unfortunately the full bind_test.cpp fails either way.

Right. This can be fixed by adding the std:: qualifier inside the 
definition of less<shared_ptr<T> > that I mentioned:

./regression -o temp.html --compiler suncc run libs/bind/bind_test.cpp
*** libs/bind/bind_test.cpp ***

** Sun WorkShop 6 update 2 C++ 5.3 2001/05/15
rm -rf SunWS_cache && CC -o boosttmp.exe -I.. ../libs/bind/bind_test.cpp
 -lrt
CPU time: 19.79 s user, 1.77 s system
./boosttmp.exe 
no errors detected
CPU time: 0 s user, 0.02 s system
Pass

George Heintzelman
georgeh_at_[hidden]