[Mesa-dev] [PATCH 01/10] i965: Add functions to convert float <-> VF.

[Ville Syrjälä]

On Thu, Oct 23, 2014 at 11:19:04PM -0700, Matt Turner wrote:
> On Thu, Oct 23, 2014 at 11:01 PM, Ville Syrjälä <syrjala at sci.fi> wrote:
> > On Thu, Oct 23, 2014 at 04:44:03PM -0700, Matt Turner wrote:
> >> ---
> >>  src/mesa/drivers/dri/i965/Makefile.sources   |  1 +
> >>  src/mesa/drivers/dri/i965/brw_packed_float.c | 74 ++++++++++++++++++++++++++++
> >>  src/mesa/drivers/dri/i965/brw_packed_float.h | 25 ++++++++++
> >>  3 files changed, 100 insertions(+)
> >>  create mode 100644 src/mesa/drivers/dri/i965/brw_packed_float.c
> >>  create mode 100644 src/mesa/drivers/dri/i965/brw_packed_float.h
> >>
> >> diff --git a/src/mesa/drivers/dri/i965/Makefile.sources b/src/mesa/drivers/dri/i965/Makefile.sources
> >> index 9c006da..6b0f601 100644
> >> --- a/src/mesa/drivers/dri/i965/Makefile.sources
> >> +++ b/src/mesa/drivers/dri/i965/Makefile.sources
> >> @@ -78,6 +78,7 @@ i965_FILES = \
> >>       brw_meta_fast_clear.c \
> >>       brw_misc_state.c \
> >>       brw_object_purgeable.c \
> >> +     brw_packed_float.c \
> >>       brw_performance_monitor.c \
> >>       brw_program.c \
> >>       brw_primitive_restart.c \
> >> diff --git a/src/mesa/drivers/dri/i965/brw_packed_float.c b/src/mesa/drivers/dri/i965/brw_packed_float.c
> >> new file mode 100644
> >> index 0000000..d25e7dd
> >> --- /dev/null
> >> +++ b/src/mesa/drivers/dri/i965/brw_packed_float.c
> >> @@ -0,0 +1,74 @@
> >> +/*
> >> + * Copyright © 2014 Intel Corporation
> >> + *
> >> + * Permission is hereby granted, free of charge, to any person obtaining a
> >> + * copy of this software and associated documentation files (the "Software"),
> >> + * to deal in the Software without restriction, including without limitation
> >> + * the rights to use, copy, modify, merge, publish, distribute, sublicense,
> >> + * and/or sell copies of the Software, and to permit persons to whom the
> >> + * Software is furnished to do so, subject to the following conditions:
> >> + *
> >> + * The above copyright notice and this permission notice (including the next
> >> + * paragraph) shall be included in all copies or substantial portions of the
> >> + * Software.
> >> + *
> >> + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
> >> + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
> >> + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
> >> + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
> >> + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
> >> + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
> >> + * IN THE SOFTWARE.
> >> + */
> >> +#include "brw_packed_float.h"
> >> +
> >> +union fu {
> >> +   float f;
> >> +   unsigned u;
> >> +   struct {
> >> +      unsigned mantissa:23;
> >> +      unsigned exponent:8;
> >> +      unsigned sign:1;
> >> +   };
> >> +};
> >> +
> >> +int
> >> +brw_float_to_vf(float f)
> >> +{
> >> +   union fu fu = { .f = f };
> >> +
> >> +   /* ±0.0f is special cased. */
> >> +   if (f == 0.0f)
> >> +      return fu.sign << 7;
> >> +
> >> +   unsigned mantissa = fu.mantissa >> (23 - 4);
> >> +   unsigned exponent = fu.exponent - (127 - 3);
> >
> > & 0x7
> 
> No, the exponent >= 8 below handles that.

Oh I thought you were testing the original exponent, but you indeed
check the new one. Yeah that makes more sense.

-- 
Ville Syrjälä
syrjala at sci.fi
http://www.sci.fi/~syrjala/