• Open Menu Close Menu
  • Apple
  • Shopping Bag
  • Apple
  • Mac
  • iPad
  • iPhone
  • Watch
  • TV
  • Music
  • Support
  • Search apple.com
  • Shopping Bag

Lists

Open Menu Close Menu
  • Terms and Conditions
  • Lists hosted on this site
  • Email the Postmaster
  • Tips for posting to public mailing lists
Re: building gcc4... HELP!
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: building gcc4... HELP!


  • Subject: Re: building gcc4... HELP!
  • From: Hugh Hoover <email@hidden>
  • Date: Wed, 11 May 2005 09:53:12 -0700


On May 11, 2005, at 07:51, David Leimbach wrote:
So
class A
{
virtual void foo ();
};

template <typename T>
class B : public A
{

};

template <typename T>
void blah (A * obj)
{
B<T> * bobj = std::dynamic_cast <B<T> * >(obj);
if (!bobj)
{
//fails even if B<T>'s instance is loaded...
}
}

That's it precisely.

The explicit instantiations of  B<T>  are implemented as "dylibs" or
some other dynamically loadable  module of sorts.

Not necessarily explicit instantiations...

I'm not sure this is a real bug.  Do you at least declare the explicit
instantiation to exist in a common header file?

No - but when I attempt that, I get multiply-defined symbol errors. (with -fno-implicit-templates)


Templates are a compile time construct of the C++ language.  There's
no easy way I know around this.  The compilation unit pretty much has
to have the definition for any of the instantiations you might use
[due to the lack of "export" keyword support in gcc, but even then...
you still need to access the source]

That's it in a nutshell :)

From further poking around, I find that the Itanium ABI spec, which is referenced from the Linux Base Spec for C++, which is a current attempt to standardize these things, explicitly says type_info's may be compared by pointer (canonical comparison not apparently required).

In addition, in the gcc4 release notes from Apple, I find:
"GCC 4.0 supports full coalescing both within and across dylibs. If a template instantiation has external visibility then all copies of it, in all dylibs, will be merged into a single copy.


If you want to ensure that a template instantiation in a dylib is not merged with template instantiations in any other dylib, use the visibility attribute to give the template hidden visibility."

That says to ME that I should be getting an automatic coalescing of the type_info across the dylibs, but it's certainly NOT happening for me. Of course, they were talking about static members rather than type_info, but they're really the same thing. I do NOT have the new "hide everything" flags turned on.

So - I'm more confused than ever - at least with respect to a workaround / solution.

Thanks again for your help - it's proving very useful to me!

Hugh Hoover
Enumclaw Software
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Darwin-dev mailing list      (email@hidden)
Help/Unsubscribe/Update your Subscription:
This email sent to email@hidden


References: 
 >building gcc4... HELP! (From: Hugh Hoover <email@hidden>)
 >Re: building gcc4... HELP! (From: David Leimbach <email@hidden>)
 >Re: building gcc4... HELP! (From: Hugh Hoover <email@hidden>)
 >Re: building gcc4... HELP! (From: David Leimbach <email@hidden>)

  • Prev by Date: Re: building gcc4... HELP!
  • Next by Date: Unable to register IPC port (problem restarting a daemon)
  • Previous by thread: Re: building gcc4... HELP!
  • Next by thread: Re: building gcc4... HELP!
  • Index(es):
    • Date
    • Thread