1 msgMaking signaling nans signal?
2 msgUsing --with-sysroot and the location of system...
2 msgPossible accepts invalid in C++: Thing::Thing::...
12 msgProblems with building gcc 4.2.1
4 msgforce symbols for inline functions to be addres...
3 msgundefined symbol

error: forming reference to void
\ Arturs Zoldners (24 Sep 2007)
. \ John Love-Jensen (24 Sep 2007)

3 msgGCC 4.2.1 build issue on Solaris 5.9 ?
11 msgexecution order
2 msggcc crosscompile for PPC, Please help!
3 msggcc compilation errors(c++)
3 msgWhen to use if-if, else-if, switch, goto and lo...
5 msgRe: Build on powerpc-ibm-aix4.3.1.0
3 msg-ffast-math individual optimizations
1 msgCross toolchain problem.
4 msgGCC ARM assembly questions
4 msgis it possible to size a pointer's target?
2 msgPassing temporary class instantiation as an arg...
1 msgvisibility and nested classes - im a bit confused
6 msglocal variable alignment on Solaris/SPARC, with...
Subject:Re: error: forming reference to void
Group:Gcc-help
From:John Love-Jensen
Date:24 Sep 2007


Hi Arturs,

> The question is, why gcc tries to resolve "operator *()" at variable
> declaration time, but not at operator invocation time? Yeah, *(void*) is
> meaningless, but so is (int)->.

I may be mistaken, but I think the methods of the instantiated template are
created when the template is instantiated, not when those methods are used.

In the case you've run into, a template specialization would be the
appropriate way to handle this situation.

HTH,
--Eljay



© 2004-2008 readlist.com