nir/algebraic: mark some optimizations with fsat(NaN) as inexact

[mesa.git] / src / compiler / nir / nir_opt_algebraic.py

diff --git a/src/compiler/nir/nir_opt_algebraic.py b/src/compiler/nir/nir_opt_algebraic.py
index 4a2efa8252f62963e1a8b294d22ad8dcfa54b43a..d775a23fceec574d8c56ac77095aca36894823e5 100644 (file)
--- a/src/compiler/nir/nir_opt_algebraic.py
+++ b/src/compiler/nir/nir_opt_algebraic.py
@@ -329,10 +329,14 @@ optimizations.extend([
    (('fneu', ('fneg', a), -1.0), ('fneu', 1.0, a)),
    (('feq', -1.0, ('fneg', a)), ('feq', a, 1.0)),
 
-   (('flt', ('fsat(is_used_once)', a), '#b(is_gt_0_and_lt_1)'), ('flt', a, b)),
+   # flt(fsat(a), b > 0 && b < 1) is inexact if a is NaN (fsat(NaN) is 0)
+   # because it returns True while flt(a, b) always returns False.
+   (('~flt', ('fsat(is_used_once)', a), '#b(is_gt_0_and_lt_1)'), ('flt', a, b)),
    (('flt', '#b(is_gt_0_and_lt_1)', ('fsat(is_used_once)', a)), ('flt', b, a)),
    (('fge', ('fsat(is_used_once)', a), '#b(is_gt_0_and_lt_1)'), ('fge', a, b)),
-   (('fge', '#b(is_gt_0_and_lt_1)', ('fsat(is_used_once)', a)), ('fge', b, a)),
+   # fge(b > 0 && b < 1, fsat(a)) is inexact if a is NaN (fsat(NaN) is 0)
+   # because it returns True while fge(b, a) always returns False.
+   (('~fge', '#b(is_gt_0_and_lt_1)', ('fsat(is_used_once)', a)), ('fge', b, a)),
    (('feq', ('fsat(is_used_once)', a), '#b(is_gt_0_and_lt_1)'), ('feq', a, b)),
    (('fneu', ('fsat(is_used_once)', a), '#b(is_gt_0_and_lt_1)'), ('fneu', a, b)),
 
@@ -1153,14 +1157,10 @@ optimizations.extend([
    (('bcsel', a, ('bcsel', b, c, d), d), ('bcsel', ('iand', a, b), c, d)),
    (('bcsel', a, b, ('bcsel', c, b, d)), ('bcsel', ('ior', a, c), b, d)),
 
-   (('fmin3@64', a, b, c), ('fmin@64', a, ('fmin@64', b, c))),
-   (('fmax3@64', a, b, c), ('fmax@64', a, ('fmax@64', b, c))),
-   (('fmed3@64', a, b, c), ('fmax@64', ('fmin@64', ('fmax@64', a, b), c), ('fmin@64', a, b))),
-
    # Misc. lowering
    (('fmod', a, b), ('fsub', a, ('fmul', b, ('ffloor', ('fdiv', a, b)))), 'options->lower_fmod'),
    (('frem', a, b), ('fsub', a, ('fmul', b, ('ftrunc', ('fdiv', a, b)))), 'options->lower_fmod'),
-   (('uadd_carry@32', a, b), ('b2i', ('ult', ('iadd', a, b), a)), 'options->lower_uadd_carry'),
+   (('uadd_carry', a, b), ('b2i', ('ult', ('iadd', a, b), a)), 'options->lower_uadd_carry'),
    (('usub_borrow@32', a, b), ('b2i', ('ult', a, b)), 'options->lower_usub_borrow'),
 
    (('bitfield_insert', 'base', 'insert', 'offset', 'bits'),
@@ -1917,7 +1917,9 @@ late_optimizations = [
    (('fneu', ('fsat(is_used_once)', a), '#b(is_gt_0_and_lt_1)'), ('fneu', a, b)),
 
    (('fge', ('fsat(is_used_once)', a), 1.0), ('fge', a, 1.0)),
-   (('flt', ('fsat(is_used_once)', a), 1.0), ('flt', a, 1.0)),
+   # flt(fsat(a), 1.0) is inexact because it returns True if a is NaN
+   # (fsat(NaN) is 0), while flt(a, 1.0) always returns FALSE.
+   (('~flt', ('fsat(is_used_once)', a), 1.0), ('flt', a, 1.0)),
 
    (('~fge', ('fmin(is_used_once)', ('fadd(is_used_once)', a, b), ('fadd', c, d)), 0.0), ('iand', ('fge', a, ('fneg', b)), ('fge', c, ('fneg', d)))),