https://gcc.gnu.org/bugzilla/show_bug.cgi?id=127268
Hongtao Liu <liuhongt at gcc dot gnu.org> changed:
What |Removed |Added
----------------------------------------------------------------------------
CC| |liuhongt at gcc dot gnu.org
--- Comment #1 from Hongtao Liu <liuhongt at gcc dot gnu.org> ---
Note, vec_to_scalar/scalar_to_vect also include floating point.
At the stage of vectorization, we don't know if the integer is allocated to gpr
and moved to sse register or it's already allocated to sse register, it may
directly load to sse register.
That's why we only conservatively add integer_to_sse cost in
x86_vector_costs::add_slp_cost
