RE: library search test fails, please help

[Bob Friesenhahn]

On Mon, 23 Feb 2009, aaragon wrote:

> Yes, I figured that the weird looking symbol is compiler specific. Then it
> also means that I won't be able to use the c++filt command with anything
> other than symbols created by the GNU C++ compiler right? I mean, if I
> compile the code with the Intel compiler, most likely this function won't
> work.

Right.  In fact, it is likely that even the GNU C++ compiler has 
altered its obfustication strategy several times due to ABI changes.

> The fact that there is no macro for C++ functions indicate that either a)
> people do not use autotools when working with with C++ programs; b) they do,
> but they are very good writing m4 macros for testing C++ code, macros that
> are not published anywhere in the web; c) the developers of autotools do not
> write C++ code and they don't care about implementing a macro to test for
> them. As for me, I'm just a user, and it would be very difficult to
> implement a macro to do these tests. Even if I come up with a test, it might
> be very nasty.

Most C++ is based on C++ classes, class methods, and templates, rather 
than "functions".  C style functions in C++ is a bit unusual.  Also it 
is quite common for C++ code to not compile or link without including 
library-specific headers, and multiple headers may be required, 
ordered in a particular way.  This makes C++ features much more 
difficult to test in a generic way.

Bob
--
Bob Friesenhahn
address@hidden, http://www.simplesystems.org/users/bfriesen/
GraphicsMagick Maintainer,    http://www.GraphicsMagick.org/