_Rb_tree regression

[Jonathan Wakely]

On 14/12/16 11:27 +0000, Jonathan Wakely wrote:
>On 13/12/16 22:51 +0100, François Dumont wrote:
>>Hi
>>
>>   I have been reported privately by Christophe in copy a regression 
>>resulting from my recent changes to _Rb_tree. I removed a 
>>constructor still necessary in C++03 mode or before. Tests would 
>>have shown it if I had run them in C++03.
>
>For reference:
>
> make check RUNTESTFLAGS=--target_board=unix/-std=gnu++03
>
>On trunk this will run almost the entire testsuite as C++03.
>
>>   * include/bits/stl_tree.h
>>   (_Rb_tree_impl(const _Key_compare&, const _Node_allocator&): Restore
>>   before C++11 mode.
>>
>>   I prefer to restore it only before C++11 while it used to be 
>>available in any mode. It is only call with rvalue references. Don't
>
>It's only called with *rvalues* (not rvalue references) but for C++11
>and up we have a constructor taking an rvalue reference, so I agree
>that the constructor taking a const lvalue reference is only needed
>for C++03.
>
>>hesitate to commit it yourself if you prefer to fix it quickly 
>>otherwise I'll do it after validation in about 24 hours.
>
>OK, thanks, please commit as soon as testing completes.

I've confirmed that the change fixes these bad test results for C++03
testing:

UNRESOLVED:     12
FAIL:   32