[Bug target/125638] openmp performance problem #pragma omp simd is necessary to get fast vectorized loops on gpu, whereas on cpu, automatic optimization is much faster than the simd statement for some loops

[rguenth at gcc dot gnu.org via Gcc-bugs]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=125638

Richard Biener <rguenth at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Blocks|                            |53947
           Keywords|                            |missed-optimization, openmp

--- Comment #1 from Richard Biener <rguenth at gcc dot gnu.org> ---
I'm not sure if it is possible to elide OMP SIMD lowering for the CPU target
and still get correct semantics.  But I also think that the lowering happens
before the split to CPU/offload target.


Referenced Bugs:

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=53947
[Bug 53947] [meta-bug] vectorizer missed-optimizations