Re: [boost] BOOST_ASSERT and __assume on MSVC

From: Felipe Magno de Almeida (felipe.m.almeida_at_[hidden])
Date: 2005-09-19 14:50:38

On 9/19/05, Rob Stewart <stewart_at_[hidden]> wrote:
> From: "Marcin Kalicinski" <kalita_at_[hidden]>
> >
> > I think we should rather base on NDEBUG macro, the same that determines the
> > behavior of assert. The whole change then boils down to adding these 6 lines
> > at the end of assert.hpp:
> >
> > #include <boost/config.hpp>
> > #if defined(BOOST_MSVC) && defined(NDEBUG)
> > # define BOOST_ASSUME(e) { BOOST_ASSERT(e); if (e); else __assume(0); }
> 
> Interesting.  I'd have thought that calling __assume(0) was
> useless, but it tells the optimizer that the else clause won't
> occur.  I wonder if that will work at all or as well as
> BOOST_ASSERT(e); __assume(e);.

I think it works exactly(seeing documentation and some tests)
What a __assume(0) means that the assertion is always false, which
means that that place cant be executed... I think __assume should
replace assert exactly as is.

> 
> There's a real problem with your solution, however: you evaluate
> the expression twice.  Won't the simpler, more direct form work?
> 
> --
> Rob Stewart                           stewart_at_[hidden]
> Software Engineer                     http://www.sig.com
> Susquehanna International Group, LLP  using std::disclaimer;
> _______________________________________________
> Unsubscribe & other changes: http://lists.boost.org/mailman/listinfo.cgi/boost
> 

-- 
   Felipe Magno de Almeida
Developer from synergy and Computer Science student from State
University of Campinas(UNICAMP).
Unicamp: http://www.ic.unicamp.br
Synergy: http://www.synergy.com.br
"There is no dark side of the moon really. Matter of fact it's all dark."