[cfe-dev] Clang/LLVM OpenMP program not spawning threads

[andreybokhanko at gmail.com]

> 28 июля 2015 г., в 14:13, Anton Korobeynikov <anton at korobeynikov.info> написал(а):
> 
> Probably for 3.7 we'd need to issue a warning / error for -fopenmp w/o libomp.
> 

Nope! This will be a "regression" for the users Chandler referred to when he switched the default back to libgomp.

Yours,
Andrey

> On Tue, Jul 28, 2015 at 1:28 PM, David Chisnall
> <David.Chisnall at cl.cam.ac.uk> wrote:
>> On 28 Jul 2015, at 11:12, Andrey Bokhanko <andreybokhanko at gmail.com> wrote:
>>> 
>>> On Tue, Jul 28, 2015 at 12:37 PM, David Chisnall
>>> <David.Chisnall at cl.cam.ac.uk> wrote:
>>>> On 27 Jul 2015, at 23:19, Jaime Arteaga <jandres742 at gmail.com> wrote:
>>>>> 
>>>>> clang -lomp -I/.../openmp/runtime/exports/common/include -L/.../openmp/runtime/exports/lin_32e/lib ./test-openmp.c -o ./test-openmp
>>>> 
>>>> Doesn’t this need -fopenmp?
>>> 
>>> -fopenmp alone is not enough, since OpenMP support (that relies on
>>> Intel OMP API) is enabled *only* if compiler is specifically directed
>>> to target libomp. Thus, -fopenmp=libomp is required.
>> 
>> Ugh.  I thought we were going to fix that before the 3.7 release.  Having -fopenmp default to the null implementation is quite user hostile.
>> 
>> David
>> 
>> 
>> _______________________________________________
>> cfe-dev mailing list
>> cfe-dev at cs.uiuc.edu
>> http://lists.cs.uiuc.edu/mailman/listinfo/cfe-dev
> 
> 
> 
> -- 
> With best regards, Anton Korobeynikov
> Faculty of Mathematics and Mechanics, Saint Petersburg State University