[PATCH] Fix uses of non-reserved names for template parameters

[Jonathan Wakely]

On 20/05/19 16:05 +0000, Szabolcs Nagy wrote:
>On 20/05/2019 15:38, Jonathan Wakely wrote:
>> On 20/05/19 14:11 +0000, Szabolcs Nagy wrote:
>>> On 17/05/2019 15:15, Jonathan Wakely wrote:
>>>> Â Â Â Â * include/bits/random.h (seed_seq::param): Fix non-reserved name.
>>>> Â Â Â Â * include/experimental/type_traits (is_detected_exact)
>>>> Â Â Â Â (is_detected_exact_v): Likewise.
>>>> Â Â Â Â * include/pstl/execution_defs.h (is_execution_policy)
>>>> Â Â Â Â (is_execution_policy_v, __enable_if_execution_policy): Likewise.
>>>> Â Â Â Â * include/pstl/execution_impl.h (__policy_traits): Likewise.
>>>> Â Â Â Â * testsuite/17_intro/names.cc: Check for more non-reserved names.
>>>> Â Â Â Â * testsuite/experimental/names.cc: New test.
>>>
>>> on baremetal (newlib) targets i see
>>>
>>> FAIL: experimental/names.cc (test for excess errors)
>>>
>>> because
>>>
>>> .../gcc/libstdc++-v3/testsuite/experimental/names.cc:25: fatal error: experimental/filesystem: No such file or directory
>>
>> Here's what I've committed ...
>
>thanks, the fatal error is now fixed.
>
>so now i get a lot of other errors.
>is this supposed to work on baremetal?

It should work just as well as testsuite/experimental/net/headers.cc
does, but that probably fails for bare metal too.

I think for now I should probably remove the Networking TS headers
from that file, or only include them for GNU/Linux.