[Mesa-dev] [PATCH 8/9] nir: Add code to fixup bitfield_insert/extract.
Jason Ekstrand
jason at jlekstrand.net
Mon Jan 11 18:58:29 PST 2016
On Jan 11, 2016 2:47 PM, "Matt Turner" <mattst88 at gmail.com> wrote:
>
> The OpenGL specifications for bitfieldInsert() and bitfieldExtract() says:
>
> The result will be undefined if <offset> or <bits> is negative, or if
> the sum of <offset> and <bits> is greater than the number of bits
> used to store the operand.
>
> Therefore passing bits=32, offset=0 is legal and defined in GLSL.
>
> But the earlier DX11/SM5 bfi/ibfe/ubfe opcodes are specified to accept a
> bitfield width ranging from 0-31. As such, Intel and AMD instructions
> read only the low 5 bits of the width operand, making them not able to
> implement the GLSL-specified behavior directly.
>
> This commit adds a pass that inserts code to implement the trivial cases
> of <bits> = 32 as
>
> bitfieldInsert:
> bits > 31 ? insert : bitfieldInsert(base, insert, offset, bits)
>
> bitfieldExtract:
> bits > 31 ? value : bitfieldExtract(value, offset, bits)
Why not just add this to nir_opt_algebraic with a flag in
nir_compiler_options? That's the way we've done a bunch of other lowering.
> ---
> src/glsl/Makefile.am | 5 +++
> src/glsl/Makefile.sources | 1 +
> src/glsl/nir/nir.h | 2 ++
> src/glsl/nir/nir_fixup_bitfield_insert_extract.py | 40
+++++++++++++++++++++++
> 4 files changed, 48 insertions(+)
> create mode 100644 src/glsl/nir/nir_fixup_bitfield_insert_extract.py
>
> diff --git a/src/glsl/Makefile.am b/src/glsl/Makefile.am
> index ba7af7c..33d28ec 100644
> --- a/src/glsl/Makefile.am
> +++ b/src/glsl/Makefile.am
> @@ -46,6 +46,7 @@ EXTRA_DIST = tests glcpp/tests README TODO
glcpp/README \
> nir/nir_algebraic.py \
> nir/nir_builder_opcodes_h.py \
> nir/nir_constant_expressions.py \
> + nir/nir_fixup_bitfield_insert_extract.py \
> nir/nir_opcodes.py \
> nir/nir_opcodes_c.py \
> nir/nir_opcodes_h.py \
> @@ -265,6 +266,10 @@ nir/nir_opt_algebraic.c: nir/nir_opt_algebraic.py
nir/nir_algebraic.py
> $(MKDIR_GEN)
> $(PYTHON_GEN) $(srcdir)/nir/nir_opt_algebraic.py > $@ || ($(RM)
$@; false)
>
> +nir/nir_fixup_bitfield_insert_extract.c:
nir/nir_fixup_bitfield_insert_extract.py nir/nir_algebraic.py
> + $(MKDIR_GEN)
> + $(PYTHON_GEN) $(srcdir)/nir/nir_fixup_bitfield_insert_extract.py
> $@ || ($(RM) $@; false)
> +
> nir_tests_control_flow_tests_SOURCES = \
> nir/tests/control_flow_tests.cpp
> nir_tests_control_flow_tests_CFLAGS = \
> diff --git a/src/glsl/Makefile.sources b/src/glsl/Makefile.sources
> index fd28f5c..bb79d84 100644
> --- a/src/glsl/Makefile.sources
> +++ b/src/glsl/Makefile.sources
> @@ -13,6 +13,7 @@ LIBGLCPP_GENERATED_FILES = \
> NIR_GENERATED_FILES = \
> nir/nir_builder_opcodes.h \
> nir/nir_constant_expressions.c \
> + nir/nir_fixup_bitfield_insert_extract.c \
> nir/nir_opcodes.c \
> nir/nir_opcodes.h \
> nir/nir_opt_algebraic.c
> diff --git a/src/glsl/nir/nir.h b/src/glsl/nir/nir.h
> index 23aec69..4b97b5e 100644
> --- a/src/glsl/nir/nir.h
> +++ b/src/glsl/nir/nir.h
> @@ -2064,6 +2064,8 @@ void nir_lower_to_source_mods(nir_shader *shader);
>
> bool nir_lower_gs_intrinsics(nir_shader *shader);
>
> +bool nir_fixup_bitfield_insert_extract(nir_shader *shader);
> +
> bool nir_normalize_cubemap_coords(nir_shader *shader);
>
> void nir_live_ssa_defs_impl(nir_function_impl *impl);
> diff --git a/src/glsl/nir/nir_fixup_bitfield_insert_extract.py
b/src/glsl/nir/nir_fixup_bitfield_insert_extract.py
> new file mode 100644
> index 0000000..21ff47c
> --- /dev/null
> +++ b/src/glsl/nir/nir_fixup_bitfield_insert_extract.py
> @@ -0,0 +1,40 @@
> +#! /usr/bin/env python
> +#
> +# Copyright (C) 2015 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.
> +
> +import nir_algebraic
> +
> +fixup = [
> + (('bitfield_insert', 'base', 'insert', 'offset', 'bits'),
> + ('bcsel', ('ilt', 31, 'bits'), 'insert',
> + ('bitfield_insert', 'base', 'insert', 'offset', 'bits'))),
> +
> + (('ibitfield_extract', 'value', 'offset', 'bits'),
> + ('bcsel', ('ilt', 31, 'bits'), 'value',
> + ('ibitfield_extract', 'value', 'offset', 'bits'))),
> +
> + (('ubitfield_extract', 'value', 'offset', 'bits'),
> + ('bcsel', ('ult', 31, 'bits'), 'value',
> + ('ubitfield_extract', 'value', 'offset', 'bits'))),
> +]
> +
> +print nir_algebraic.AlgebraicPass("nir_fixup_bitfield_insert_extract",
fixup).render()
> --
> 2.4.9
>
> _______________________________________________
> mesa-dev mailing list
> mesa-dev at lists.freedesktop.org
> http://lists.freedesktop.org/mailman/listinfo/mesa-dev
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.freedesktop.org/archives/mesa-dev/attachments/20160111/7268b5ee/attachment.html>
More information about the mesa-dev
mailing list