Subject: Re: [Boost-build] Compile-time probing (Ticket #2597)
From: Vladimir Prus (ghost_at_[hidden])
Date: 2008-12-19 15:12:40
On Friday 19 December 2008 04:19:02 Steve M. Robbins wrote:
> Boost.Math builds a test file to check whether long double math is
> supported. If it is not, the build fails with an error which is
> misleading because it is not an error but, rather, a successful build
> of all parts except the long double version of boost_math_c99 and
> This issue causes the builds to fail for Debian;
> c.f. http://tinyurl.com/45feh8
> I have opened a Trac ticket about this
> (https://svn.boost.org/trac/boost/ticket/2597) and have been told by
> John Maddock that this is just the way Boost.Build works.
> Maybe the Boost.Build cogniscenti can help out. How can I
> express the following in Boost.Build:
> try to compile foo.cpp
> if successful:
> build library libfoo
> emit warning
> The key point is that the build should *succeed* in both cases;
> here, "succeed" means "exit with status 0".
> In short, we need a compile-time test similar to what a system like
> autotools has.
I have bad news, and good news. The bad news is that there's no way
to do this currently.
The good news is that I hope to implement this mechanism soonish. I'll
keep you posted.
Boost-Build list run by bdawes at acm.org, david.abrahams at rcn.com, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk