This is the mail archive of the
gcc-help@gcc.gnu.org
mailing list for the GCC project.
Re: using __float128 and -std=c++11
- From: Jonathan Wakely <jwakely dot gcc at gmail dot com>
- To: gcc-help <gcc-help at gcc dot gnu dot org>
- Cc: Peter van Hoof <p dot vanhoof at oma dot be>
- Date: Tue, 13 Oct 2015 10:33:06 +0100
- Subject: Re: using __float128 and -std=c++11
- Authentication-results: sourceware.org; auth=none
- References: <561CCE93 dot 40700 at oma dot be> <alpine dot DEB dot 2 dot 20 dot 1510131130210 dot 2213 at laptop-mg dot saclay dot inria dot fr>
On 13 October 2015 at 10:31, Marc Glisse wrote:
> On Tue, 13 Oct 2015, Peter van Hoof wrote:
>
>> Consider the following program
>>
>> #include <quadmath.h>
>> __float128 a()
>> {
>> return FLT128_MAX;
>> }
>>
>> When I compile this with any of these commands
>>
>> g++ -c prog.cc
>> g++ -std=c++98 -c prog.cc
>> g++ -std=c++03 -c prog.cc
>>
>> it compiles just fine. However, when I use this
>>
>> g++ -std=c++11 -c prog.cc
>>
>> I get the error message
>>
>> In file included from prog.cc:1:0:
>> prog.cc: In function â__float128 a()â:
>> prog.cc:5:9: error: unable to find numeric literal operator âoperator"" Qâ
>> return FLT128_MAX;
>> ^
>> I suspect this is a compiler bug, but would like to double-check first
>> before submitting a report. The compiler doesn't complain about the use of
>> __float128, so I guess that type is deemed compliant with the C++11
>> standard. So I would assume that I should also be allowed to use __float128
>> literals...
>
>
> A more recent version of the compiler should give you a hint about using
> gnu++11 instead (or some other flag I have forgotten).
q.cc:4:16: note: use -std=gnu++11 or -fext-numeric-literals to enable
more built-in suffixes