| Message ID | BANLkTim5nMSRJddwRzU6y69-LZ9pdCbQxw@mail.gmail.com |
|---|---|
| State | New |
| Headers | show |
OK for trunk and 4.6. Jason
On 15 May 2011 23:19, Jason Merrill wrote:
> OK for trunk and 4.6.
The bug isn't present on the 4.6 branch and I'm not sure where the
change should go, if it's needed at all, so I've only committed it to
trunk.
On 05/15/2011 07:07 PM, Jonathan Wakely wrote: > On 15 May 2011 23:19, Jason Merrill wrote: >> OK for trunk and 4.6. > > The bug isn't present on the 4.6 branch and I'm not sure where the > change should go, if it's needed at all, so I've only committed it to > trunk. Ah, my mistake. Sounds good. Jason
Index: cp/parser.c =================================================================== --- cp/parser.c (revision 173683) +++ cp/parser.c (working copy) @@ -8682,7 +8682,7 @@ static tree cp_parser_perform_range_for_lookup (tree range, tree *begin, tree *end) { - if (!COMPLETE_TYPE_P (TREE_TYPE (range))) + if (!COMPLETE_TYPE_P (complete_type (TREE_TYPE (range)))) { error ("range-based %<for%> expression of type %qT " "has incomplete type", TREE_TYPE (range)); Index: testsuite/g++.dg/cpp0x/range-for18.C =================================================================== --- testsuite/g++.dg/cpp0x/range-for18.C (revision 0) +++ testsuite/g++.dg/cpp0x/range-for18.C (revision 0) @@ -0,0 +1,17 @@ +// PR c++/48994 + +// { dg-do compile } +// { dg-options "-std=c++0x" } + +template <typename T> +struct myvec +{ + T* begin() const; + T* end() const; +}; + +void f(const myvec<int>& v) +{ + for (int i : v) + ; +}
cp/ChangeLog PR c++/48994 * parser.c (cp_parser_perform_range_for_lookup): Call complete_type. testsuite/ChangeLog PR c++/48994 * g++.dg/cpp0x/range-for18.C: New. Tested x86_64-linux, ok for trunk?