[PR 17431] Fix copy ctor access check

[Nathan Sidwell]

I've installed this patch for 17431.  It was yet another case of us not
implementing overloading exactly how the std describes it.  Specifically
there is no derived->base object  standard conversion.

I discovered that we do too many temporary copies when applying
a user conversion to a class type, but that is not a regression.  I'll
add a PR for that once I've constructed the correct test cases.

I'll backport this to 3.4 shortly

booted & tested on i686-pc-linux-gnu.

nathan
-- 
Nathan Sidwell    ::   http://www.codesourcery.com   ::     CodeSourcery LLC
nathan@codesourcery.com    ::     http://www.planetfall.pwp.blueyonder.co.uk

-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: 17431.patch
URL: <http://gcc.gnu.org/pipermail/gcc-patches/attachments/20041201/52bf4356/attachment.ksh>
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: 17431-1.C
URL: <http://gcc.gnu.org/pipermail/gcc-patches/attachments/20041201/52bf4356/attachment-0001.ksh>
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: 17431-2.C
URL: <http://gcc.gnu.org/pipermail/gcc-patches/attachments/20041201/52bf4356/attachment-0002.ksh>
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: 17431-3.C
URL: <http://gcc.gnu.org/pipermail/gcc-patches/attachments/20041201/52bf4356/attachment-0003.ksh>
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: 17431-4.C
URL: <http://gcc.gnu.org/pipermail/gcc-patches/attachments/20041201/52bf4356/attachment-0004.ksh>