--- src/compiler/nir/nir_lower_alu_to_scalar.c | 4 +++ src/compiler/nir/nir_opcodes.py | 29 ++++++++++++++++++++++ 2 files changed, 33 insertions(+)
diff --git a/src/compiler/nir/nir_lower_alu_to_scalar.c b/src/compiler/nir/nir_lower_alu_to_scalar.c index e424dff25c4..4f97472a87d 100644 --- a/src/compiler/nir/nir_lower_alu_to_scalar.c +++ b/src/compiler/nir/nir_lower_alu_to_scalar.c @@ -197,6 +197,10 @@ lower_alu_instr_scalar(nir_alu_instr *instr, nir_builder *b) return false; LOWER_REDUCTION(nir_op_fdot, nir_op_fmul, nir_op_fadd); + LOWER_REDUCTION(nir_op_ball_fequal, nir_op_feq, nir_op_iand); + LOWER_REDUCTION(nir_op_ball_iequal, nir_op_ieq, nir_op_iand); + LOWER_REDUCTION(nir_op_bany_fnequal, nir_op_fne, nir_op_ior); + LOWER_REDUCTION(nir_op_bany_inequal, nir_op_ine, nir_op_ior); LOWER_REDUCTION(nir_op_b32all_fequal, nir_op_feq32, nir_op_iand); LOWER_REDUCTION(nir_op_b32all_iequal, nir_op_ieq32, nir_op_iand); LOWER_REDUCTION(nir_op_b32any_fnequal, nir_op_fne32, nir_op_ior); diff --git a/src/compiler/nir/nir_opcodes.py b/src/compiler/nir/nir_opcodes.py index d349f74ed2a..aee505667b8 100644 --- a/src/compiler/nir/nir_opcodes.py +++ b/src/compiler/nir/nir_opcodes.py @@ -89,6 +89,7 @@ class Opcode(object): # helper variables for strings tfloat = "float" tint = "int" +tbool = "bool1" tbool32 = "bool32" tuint = "uint" tuint16 = "uint16" @@ -192,6 +193,10 @@ for src_t in [tint, tuint, tfloat]: # We'll hand-code the to/from bool conversion opcodes. Because bool doesn't # have multiple bit-sizes, we can always infer the size from the other type. +unop_convert("f2b", tbool, tfloat, "src0 != 0.0") +unop_convert("i2b", tbool, tint, "src0 != 0") +unop_convert("b2f", tfloat, tbool, "src0 ? 1.0 : 0.0") +unop_convert("b2i", tint, tbool, "src0 ? 1 : 0") unop_convert("f2b32", tbool32, tfloat, "src0 != 0.0") unop_convert("i2b32", tbool32, tint, "src0 != 0") unop_convert("b322f", tfloat, tbool32, "src0 ? 1.0 : 0.0") @@ -405,6 +410,9 @@ def binop_convert(name, out_type, in_type, alg_props, const_expr): def binop(name, ty, alg_props, const_expr): binop_convert(name, ty, ty, alg_props, const_expr) +def binop_compare(name, ty, alg_props, const_expr): + binop_convert(name, tbool, ty, alg_props, const_expr) + def binop_compare32(name, ty, alg_props, const_expr): binop_convert(name, tbool32, ty, alg_props, const_expr) @@ -488,6 +496,16 @@ binop("frem", tfloat, "", "src0 - src1 * truncf(src0 / src1)") # these integer-aware comparisons return a boolean (0 or ~0) +binop_compare("flt", tfloat, "", "src0 < src1") +binop_compare("fge", tfloat, "", "src0 >= src1") +binop_compare("feq", tfloat, commutative, "src0 == src1") +binop_compare("fne", tfloat, commutative, "src0 != src1") +binop_compare("ilt", tint, "", "src0 < src1") +binop_compare("ige", tint, "", "src0 >= src1") +binop_compare("ieq", tint, commutative, "src0 == src1") +binop_compare("ine", tint, commutative, "src0 != src1") +binop_compare("ult", tuint, "", "src0 < src1") +binop_compare("uge", tuint, "", "src0 >= src1") binop_compare32("flt32", tfloat, "", "src0 < src1") binop_compare32("fge32", tfloat, "", "src0 >= src1") binop_compare32("feq32", tfloat, commutative, "src0 == src1") @@ -501,6 +519,15 @@ binop_compare32("uge32", tuint, "", "src0 >= src1") # integer-aware GLSL-style comparisons that compare floats and ints +binop_reduce("ball_fequal", 1, tbool, tfloat, "{src0} == {src1}", + "{src0} && {src1}", "{src}") +binop_reduce("bany_fnequal", 1, tbool, tfloat, "{src0} != {src1}", + "{src0} || {src1}", "{src}") +binop_reduce("ball_iequal", 1, tbool, tint, "{src0} == {src1}", + "{src0} && {src1}", "{src}") +binop_reduce("bany_inequal", 1, tbool, tint, "{src0} != {src1}", + "{src0} || {src1}", "{src}") + binop_reduce("b32all_fequal", 1, tbool32, tfloat, "{src0} == {src1}", "{src0} && {src1}", "{src}") binop_reduce("b32any_fnequal", 1, tbool32, tfloat, "{src0} != {src1}", @@ -694,6 +721,8 @@ triop("fmed3", tfloat, "fmaxf(fminf(fmaxf(src0, src1), src2), fminf(src0, src1)) triop("imed3", tint, "MAX2(MIN2(MAX2(src0, src1), src2), MIN2(src0, src1))") triop("umed3", tuint, "MAX2(MIN2(MAX2(src0, src1), src2), MIN2(src0, src1))") +opcode("bcsel", 0, tuint, [0, 0, 0], + [tbool, tuint, tuint], "", "src0 ? src1 : src2") opcode("b32csel", 0, tuint, [0, 0, 0], [tbool32, tuint, tuint], "", "src0 ? src1 : src2") -- 2.19.1 _______________________________________________ mesa-dev mailing list mesa-dev@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/mesa-dev