From: Jonathan Marek <jonathan@marek.ca>
Date: Sun, 15 Dec 2019 18:43:07 +0000 (-0500)
Subject: nir: add option to lower half packing opcodes
X-Git-Url: https://git.libre-soc.org/?a=commitdiff_plain;h=004797002fb87db0c2323955eed2f8f6871c3bfb;p=mesa.git

nir: add option to lower half packing opcodes

Signed-off-by: Jonathan Marek <jonathan@marek.ca>
Reviewed-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Reviewed-by: Eric Anholt <eric@anholt.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/merge_requests/3106>
---

diff --git a/src/compiler/nir/nir.h b/src/compiler/nir/nir.h
index fc46a3e6604..fbef503f0fa 100644
--- a/src/compiler/nir/nir.h
+++ b/src/compiler/nir/nir.h
@@ -2761,11 +2761,13 @@ typedef struct nir_shader_compiler_options {
    bool lower_ldexp;
 
    bool lower_pack_half_2x16;
+   bool lower_pack_half_2x16_split;
    bool lower_pack_unorm_2x16;
    bool lower_pack_snorm_2x16;
    bool lower_pack_unorm_4x8;
    bool lower_pack_snorm_4x8;
    bool lower_unpack_half_2x16;
+   bool lower_unpack_half_2x16_split;
    bool lower_unpack_unorm_2x16;
    bool lower_unpack_snorm_2x16;
    bool lower_unpack_unorm_4x8;
diff --git a/src/compiler/nir/nir_opt_algebraic.py b/src/compiler/nir/nir_opt_algebraic.py
index f101ad04567..633ee8792d9 100644
--- a/src/compiler/nir/nir_opt_algebraic.py
+++ b/src/compiler/nir/nir_opt_algebraic.py
@@ -1117,6 +1117,18 @@ optimizations.extend([
                                            127.0))),
      'options->lower_unpack_snorm_4x8'),
 
+   (('pack_half_2x16_split', 'a@32', 'b@32'),
+    ('ior', ('ishl', ('u2u32', ('f2f16', b)), 16), ('u2u32', ('f2f16', a))),
+    'options->lower_pack_half_2x16_split'),
+
+   (('unpack_half_2x16_split_x', 'a@32'),
+    ('f2f32', ('u2u16', a)),
+    'options->lower_unpack_half_2x16_split'),
+
+   (('unpack_half_2x16_split_y', 'a@32'),
+    ('f2f32', ('u2u16', ('ushr', a, 16))),
+    'options->lower_unpack_half_2x16_split'),
+
    (('isign', a), ('imin', ('imax', a, -1), 1), 'options->lower_isign'),
    (('fsign', a), ('fsub', ('b2f', ('flt', 0.0, a)), ('b2f', ('flt', a, 0.0))), 'options->lower_fsign'),