On Sat, Oct 09, 2021 at 12:55:15PM +0800, liuhongt via Gcc-patches wrote:
> Pushed to trunk.
> 
> libgomp/ChangeLog:
> 
>       * testsuite/libgomp.c++/scan-10.C: Add option -fvect-cost-model=cheap.
>       * testsuite/libgomp.c++/scan-11.C: Ditto.
>       * testsuite/libgomp.c++/scan-12.C: Ditto.
>       * testsuite/libgomp.c++/scan-13.C: Ditto.
>       * testsuite/libgomp.c++/scan-14.C: Ditto.
>       * testsuite/libgomp.c++/scan-15.C: Ditto.
>       * testsuite/libgomp.c++/scan-16.C: Ditto.
>       * testsuite/libgomp.c++/scan-9.C: Ditto.
>       * testsuite/libgomp.c-c++-common/lastprivate-conditional-7.c: Ditto.
>       * testsuite/libgomp.c-c++-common/lastprivate-conditional-8.c: Ditto.
>       * testsuite/libgomp.c/scan-11.c: Ditto.
>       * testsuite/libgomp.c/scan-12.c: Ditto.
>       * testsuite/libgomp.c/scan-13.c: Ditto.
>       * testsuite/libgomp.c/scan-14.c: Ditto.
>       * testsuite/libgomp.c/scan-15.c: Ditto.
>       * testsuite/libgomp.c/scan-16.c: Ditto.
>       * testsuite/libgomp.c/scan-17.c: Ditto.
>       * testsuite/libgomp.c/scan-18.c: Ditto.
>       * testsuite/libgomp.c/scan-19.c: Ditto.
>       * testsuite/libgomp.c/scan-20.c: Ditto.
>       * testsuite/libgomp.c/scan-21.c: Ditto.
>       * testsuite/libgomp.c/scan-22.c: Ditto.

I don't think this is the right thing to do.
This just means that at some point between 2013 when -fsimd-cost-model has
been introduced and now -fsimd-cost-model= option at least partially stopped
working properly.
As documented, -fsimd-cost-model= overrides the -fvect-cost-model= setting
for OpenMP simd loops (loop->force_vectorize is true) if specified differently
from default.
In tree-vectorizer.h we have:
static inline bool
unlimited_cost_model (loop_p loop)
{
  if (loop != NULL && loop->force_vectorize
      && flag_simd_cost_model != VECT_COST_MODEL_DEFAULT)
    return flag_simd_cost_model == VECT_COST_MODEL_UNLIMITED;
  return (flag_vect_cost_model == VECT_COST_MODEL_UNLIMITED);
}
and use it in various places, but we also just use flag_vect_cost_model
in lots of places (and in one spot use flag_simd_cost_model, not sure if
we are sure it is a force_vectorize loop or what).

So, IMHO we should change the above inline function to
loop_cost_model and let it return the cost model and then just
reimplement unlimited_cost_model as
return loop_cost_model (loop) == VECT_COST_MODEL_UNLIMITED;
and then adjust the direct uses of the flag and revert these changes.

        Jakub

Reply via email to