Commit 5c82f8a0 authored by Alyssa Rosenzweig's avatar Alyssa Rosenzweig 💜 Committed by Marge Bot
Browse files

panfrost: Add theory for new framebuffer lowering



We take a somewhat different strategy that should be more flexible.
Signed-off-by: Alyssa Rosenzweig's avatarAlyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <mesa/mesa!5265>
parent 5a175e4a
......@@ -87,4 +87,5 @@ util_FILES := \
util/pan_ir.c \
util/pan_ir.h \
util/pan_liveness.c \
util/pan_lower_framebuffer.c \
util/pan_sysval.c \
......@@ -26,6 +26,7 @@ libpanfrost_util_files = files(
'pan_ir.c',
'pan_ir.h',
'pan_liveness.c',
'pan_lower_framebuffer.c',
'pan_sysval.c',
)
......
/*
* Copyright (C) 2020 Collabora, Ltd.
*
* 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.
*
* Authors (Collabora):
* Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
*/
/**
* Implements framebuffer format conversions in software for Midgard/Bifrost
* blend shaders. This pass is designed for a single render target; Midgard
* duplicates blend shaders for MRT to simplify everything. A particular
* framebuffer format may be categorized as 1) typed load available, 2) typed
* unpack available, or 3) software unpack only, and likewise for stores. The
* first two types are handled in the compiler backend directly, so this module
* is responsible for identifying type 3 formats (hardware dependent) and
* inserting appropriate ALU code to perform the conversion from the packed
* type to a designated unpacked type, and vice versa.
*
* The unpacked type depends on the format:
*
* - For 32-bit float formats, 32-bit floats.
* - For other floats, 16-bit floats.
* - For 32-bit ints, 32-bit ints.
* - For 8-bit ints, 8-bit ints.
* - For other ints, 16-bit ints.
*
* The rationale is to optimize blending and logic op instructions by using the
* smallest precision necessary to store the pixel losslessly.
*/
#include "compiler/nir/nir.h"
#include "compiler/nir/nir_builder.h"
#include "compiler/nir/nir_format_convert.h"
#include "util/format/u_format.h"
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment