[Mesa-dev] [PATCH v2 26/53] intel/compiler: Extended Math is limited to SIMD8 on half-float

[Iago Toral]

On Mon, 2018-12-31 at 13:26 +0200, Pohjolainen, Topi wrote:
> On Wed, Dec 19, 2018 at 12:50:54PM +0100, Iago Toral Quiroga wrote:
> > From the Skylake PRM, Extended Math Function:
> > 
> >   "The execution size must be no more than 8 when half-floats
> >    are used in source or destination operand."
> > 
> > Earlier generations do not support Extended Math with half-float.
> > 
> > v2
> >  - Rewrite the code to make it more readable (Jason).
> > 
> > Reviewed-by: Topi Pohjolainen <topi.pohjolainen at intel.com> (v1)
> > ---
> >  src/intel/compiler/brw_fs.cpp | 29 ++++++++++++++++++++---------
> >  1 file changed, 20 insertions(+), 9 deletions(-)
> > 
> > diff --git a/src/intel/compiler/brw_fs.cpp
> > b/src/intel/compiler/brw_fs.cpp
> > index 3125e5feb1d..933b0b6ffc4 100644
> > --- a/src/intel/compiler/brw_fs.cpp
> > +++ b/src/intel/compiler/brw_fs.cpp
> > @@ -5406,18 +5406,29 @@ get_lowered_simd_width(const struct
> > gen_device_info *devinfo,
> >     case SHADER_OPCODE_EXP2:
> >     case SHADER_OPCODE_LOG2:
> >     case SHADER_OPCODE_SIN:
> > -   case SHADER_OPCODE_COS:
> > +   case SHADER_OPCODE_COS: {
> >        /* Unary extended math instructions are limited to SIMD8 on
> > Gen4 and
> > -       * Gen6.
> > +       * Gen6. Extended Math Function is limited to SIMD8 with
> > half-float.
> >         */
> > -      return (devinfo->gen >= 7 ? MIN2(16, inst->exec_size) :
> > -              devinfo->gen == 5 || devinfo->is_g4x ? MIN2(16,
> > inst->exec_size) :
> > -              MIN2(8, inst->exec_size));
> > +      if (devinfo->gen == 6 || (devinfo->gen == 4 && !devinfo-
> > >is_g4x))
> > +         return MIN2(8, inst->exec_size);
> > +      if (inst->dst.type == BRW_REGISTER_TYPE_HF)
> > +         return MIN2(8, inst->exec_size);
> > +      else
> > +         return MIN2(16, inst->exec_size);
> 
> I was expecting either of these two:
> 
> 
>          if ()
>             return
>          else if ()
>             return
>          else
>             return
> 
> 
> or
> 
> 
>          if ()
>             return
> 
>          if ()
>             return
> 
>          return

Yes, makes sense. I'll go with the second option if there is no
preference.

Iago

> > +   }
> >  
> > -   case SHADER_OPCODE_POW:
> > -      /* SIMD16 is only allowed on Gen7+. */
> > -      return (devinfo->gen >= 7 ? MIN2(16, inst->exec_size) :
> > -              MIN2(8, inst->exec_size));
> > +   case SHADER_OPCODE_POW: {
> > +      /* SIMD16 is only allowed on Gen7+. Extended Math Function
> > is limited
> > +       * to SIMD8 with half-float
> > +       */
> > +      if (devinfo->gen < 7)
> > +         return MIN2(8, inst->exec_size);
> > +      if (inst->dst.type == BRW_REGISTER_TYPE_HF)
> > +         return MIN2(8, inst->exec_size);
> > +      else
> > +         return MIN2(16, inst->exec_size);
> > +   }
> >  
> >     case SHADER_OPCODE_INT_QUOTIENT:
> >     case SHADER_OPCODE_INT_REMAINDER:
> > -- 
> > 2.17.1
> > 
> > _______________________________________________
> > mesa-dev mailing list
> > mesa-dev at lists.freedesktop.org
> > https://lists.freedesktop.org/mailman/listinfo/mesa-dev
> 
>