src/compiler/nir/nir_opcodes.py

   1 #
   2 # Copyright (C) 2014 Connor Abbott
   3 #
   4 # Permission is hereby granted, free of charge, to any person obtaining a
   5 # copy of this software and associated documentation files (the "Software"),
   6 # to deal in the Software without restriction, including without limitation
   7 # the rights to use, copy, modify, merge, publish, distribute, sublicense,
   8 # and/or sell copies of the Software, and to permit persons to whom the
   9 # Software is furnished to do so, subject to the following conditions:
  10 #
  11 # The above copyright notice and this permission notice (including the next
  12 # paragraph) shall be included in all copies or substantial portions of the
  13 # Software.
  14 #
  15 # THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
  16 # IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
  17 # FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
  18 # THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
  19 # LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
  20 # FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
  21 # IN THE SOFTWARE.
  22 #
  23 # Authors:
  24 #    Connor Abbott (cwabbott0@gmail.com)
  25
  26 import re
  27
  28 # Class that represents all the information we have about the opcode
  29 # NOTE: this must be kept in sync with nir_op_info
  30
  31 class Opcode(object):
  32    """Class that represents all the information we have about the opcode
  33    NOTE: this must be kept in sync with nir_op_info
  34    """
  35    def __init__(self, name, output_size, output_type, input_sizes,
  36                 input_types, is_conversion, algebraic_properties, const_expr):
  37       """Parameters:
  38
  39       - name is the name of the opcode (prepend nir_op_ for the enum name)
  40       - all types are strings that get nir_type_ prepended to them
  41       - input_types is a list of types
  42       - is_conversion is true if this opcode represents a type conversion
  43       - algebraic_properties is a space-seperated string, where nir_op_is_ is
  44         prepended before each entry
  45       - const_expr is an expression or series of statements that computes the
  46         constant value of the opcode given the constant values of its inputs.
  47
  48       Constant expressions are formed from the variables src0, src1, ...,
  49       src(N-1), where N is the number of arguments.  The output of the
  50       expression should be stored in the dst variable.  Per-component input
  51       and output variables will be scalars and non-per-component input and
  52       output variables will be a struct with fields named x, y, z, and w
  53       all of the correct type.  Input and output variables can be assumed
  54       to already be of the correct type and need no conversion.  In
  55       particular, the conversion from the C bool type to/from  NIR_TRUE and
  56       NIR_FALSE happens automatically.
  57
  58       For per-component instructions, the entire expression will be
  59       executed once for each component.  For non-per-component
  60       instructions, the expression is expected to store the correct values
  61       in dst.x, dst.y, etc.  If "dst" does not exist anywhere in the
  62       constant expression, an assignment to dst will happen automatically
  63       and the result will be equivalent to "dst = <expression>" for
  64       per-component instructions and "dst.x = dst.y = ... = <expression>"
  65       for non-per-component instructions.
  66       """
  67       assert isinstance(name, str)
  68       assert isinstance(output_size, int)
  69       assert isinstance(output_type, str)
  70       assert isinstance(input_sizes, list)
  71       assert isinstance(input_sizes[0], int)
  72       assert isinstance(input_types, list)
  73       assert isinstance(input_types[0], str)
  74       assert isinstance(is_conversion, bool)
  75       assert isinstance(algebraic_properties, str)
  76       assert isinstance(const_expr, str)
  77       assert len(input_sizes) == len(input_types)
  78       assert 0 <= output_size <= 4
  79       for size in input_sizes:
  80          assert 0 <= size <= 4
  81          if output_size != 0:
  82             assert size != 0
  83       self.name = name
  84       self.num_inputs = len(input_sizes)
  85       self.output_size = output_size
  86       self.output_type = output_type
  87       self.input_sizes = input_sizes
  88       self.input_types = input_types
  89       self.is_conversion = is_conversion
  90       self.algebraic_properties = algebraic_properties
  91       self.const_expr = const_expr
  92
  93 # helper variables for strings
  94 tfloat = "float"
  95 tint = "int"
  96 tbool = "bool"
  97 tbool1 = "bool1"
  98 tbool32 = "bool32"
  99 tuint = "uint"
 100 tuint16 = "uint16"
 101 tfloat32 = "float32"
 102 tint32 = "int32"
 103 tuint32 = "uint32"
 104 tint64 = "int64"
 105 tuint64 = "uint64"
 106 tfloat64 = "float64"
 107
 108 _TYPE_SPLIT_RE = re.compile(r'(?P<type>int|uint|float|bool)(?P<bits>\d+)?')
 109
 110 def type_has_size(type_):
 111     m = _TYPE_SPLIT_RE.match(type_)
 112     assert m is not None, 'Invalid NIR type string: "{}"'.format(type_)
 113     return m.group('bits') is not None
 114
 115 def type_size(type_):
 116     m = _TYPE_SPLIT_RE.match(type_)
 117     assert m is not None, 'Invalid NIR type string: "{}"'.format(type_)
 118     assert m.group('bits') is not None, \
 119            'NIR type string has no bit size: "{}"'.format(type_)
 120     return int(m.group('bits'))
 121
 122 def type_sizes(type_):
 123     if type_has_size(type_):
 124         return [type_size(type_)]
 125     elif type_ == 'bool':
 126         return [1, 32]
 127     elif type_ == 'float':
 128         return [16, 32, 64]
 129     else:
 130         return [1, 8, 16, 32, 64]
 131
 132 def type_base_type(type_):
 133     m = _TYPE_SPLIT_RE.match(type_)
 134     assert m is not None, 'Invalid NIR type string: "{}"'.format(type_)
 135     return m.group('type')
 136
 137 # Operation where the first two sources are commutative.
 138 #
 139 # For 2-source operations, this just mathematical commutativity.  Some
 140 # 3-source operations, like ffma, are only commutative in the first two
 141 # sources.
 142 _2src_commutative = "2src_commutative "
 143 associative = "associative "
 144
 145 # global dictionary of opcodes
 146 opcodes = {}
 147
 148 def opcode(name, output_size, output_type, input_sizes, input_types,
 149            is_conversion, algebraic_properties, const_expr):
 150    assert name not in opcodes
 151    opcodes[name] = Opcode(name, output_size, output_type, input_sizes,
 152                           input_types, is_conversion, algebraic_properties,
 153                           const_expr)
 154
 155 def unop_convert(name, out_type, in_type, const_expr):
 156    opcode(name, 0, out_type, [0], [in_type], False, "", const_expr)
 157
 158 def unop(name, ty, const_expr):
 159    opcode(name, 0, ty, [0], [ty], False, "", const_expr)
 160
 161 def unop_horiz(name, output_size, output_type, input_size, input_type,
 162                const_expr):
 163    opcode(name, output_size, output_type, [input_size], [input_type],
 164           False, "", const_expr)
 165
 166 def unop_reduce(name, output_size, output_type, input_type, prereduce_expr,
 167                 reduce_expr, final_expr):
 168    def prereduce(src):
 169       return "(" + prereduce_expr.format(src=src) + ")"
 170    def final(src):
 171       return final_expr.format(src="(" + src + ")")
 172    def reduce_(src0, src1):
 173       return reduce_expr.format(src0=src0, src1=src1)
 174    src0 = prereduce("src0.x")
 175    src1 = prereduce("src0.y")
 176    src2 = prereduce("src0.z")
 177    src3 = prereduce("src0.w")
 178    unop_horiz(name + "2", output_size, output_type, 2, input_type,
 179               final(reduce_(src0, src1)))
 180    unop_horiz(name + "3", output_size, output_type, 3, input_type,
 181               final(reduce_(reduce_(src0, src1), src2)))
 182    unop_horiz(name + "4", output_size, output_type, 4, input_type,
 183               final(reduce_(reduce_(src0, src1), reduce_(src2, src3))))
 184
 185 def unop_numeric_convert(name, out_type, in_type, const_expr):
 186    opcode(name, 0, out_type, [0], [in_type], True, "", const_expr)
 187
 188 unop("mov", tuint, "src0")
 189
 190 unop("ineg", tint, "-src0")
 191 unop("fneg", tfloat, "-src0")
 192 unop("inot", tint, "~src0") # invert every bit of the integer
 193 unop("fsign", tfloat, ("bit_size == 64 ? " +
 194                        "((src0 == 0.0) ? 0.0 : ((src0 > 0.0) ? 1.0 : -1.0)) : " +
 195                        "((src0 == 0.0f) ? 0.0f : ((src0 > 0.0f) ? 1.0f : -1.0f))"))
 196 unop("isign", tint, "(src0 == 0) ? 0 : ((src0 > 0) ? 1 : -1)")
 197 unop("iabs", tint, "(src0 < 0) ? -src0 : src0")
 198 unop("fabs", tfloat, "fabs(src0)")
 199 unop("fsat", tfloat, ("bit_size == 64 ? " +
 200                       "((src0 > 1.0) ? 1.0 : ((src0 <= 0.0) ? 0.0 : src0)) : " +
 201                       "((src0 > 1.0f) ? 1.0f : ((src0 <= 0.0f) ? 0.0f : src0))"))
 202 unop("frcp", tfloat, "bit_size == 64 ? 1.0 / src0 : 1.0f / src0")
 203 unop("frsq", tfloat, "bit_size == 64 ? 1.0 / sqrt(src0) : 1.0f / sqrtf(src0)")
 204 unop("fsqrt", tfloat, "bit_size == 64 ? sqrt(src0) : sqrtf(src0)")
 205 unop("fexp2", tfloat, "exp2f(src0)")
 206 unop("flog2", tfloat, "log2f(src0)")
 207
 208 # Generate all of the numeric conversion opcodes
 209 for src_t in [tint, tuint, tfloat, tbool]:
 210    if src_t == tbool:
 211       dst_types = [tfloat, tint]
 212    elif src_t == tint:
 213       dst_types = [tfloat, tint, tbool]
 214    elif src_t == tuint:
 215       dst_types = [tfloat, tuint]
 216    elif src_t == tfloat:
 217       dst_types = [tint, tuint, tfloat, tbool]
 218
 219    for dst_t in dst_types:
 220       for bit_size in type_sizes(dst_t):
 221           if bit_size == 16 and dst_t == tfloat and src_t == tfloat:
 222               rnd_modes = ['_rtne', '_rtz', '']
 223               for rnd_mode in rnd_modes:
 224                   unop_numeric_convert("{0}2{1}{2}{3}".format(src_t[0], dst_t[0],
 225                                                               bit_size, rnd_mode),
 226                                        dst_t + str(bit_size), src_t, "src0")
 227           else:
 228               conv_expr = "src0 != 0" if dst_t == tbool else "src0"
 229               unop_numeric_convert("{0}2{1}{2}".format(src_t[0], dst_t[0], bit_size),
 230                                    dst_t + str(bit_size), src_t, conv_expr)
 231
 232
 233 # Unary floating-point rounding operations.
 234
 235
 236 unop("ftrunc", tfloat, "bit_size == 64 ? trunc(src0) : truncf(src0)")
 237 unop("fceil", tfloat, "bit_size == 64 ? ceil(src0) : ceilf(src0)")
 238 unop("ffloor", tfloat, "bit_size == 64 ? floor(src0) : floorf(src0)")
 239 unop("ffract", tfloat, "src0 - (bit_size == 64 ? floor(src0) : floorf(src0))")
 240 unop("fround_even", tfloat, "bit_size == 64 ? _mesa_roundeven(src0) : _mesa_roundevenf(src0)")
 241
 242 unop("fquantize2f16", tfloat, "(fabs(src0) < ldexpf(1.0, -14)) ? copysignf(0.0f, src0) : _mesa_half_to_float(_mesa_float_to_half(src0))")
 243
 244 # Trigonometric operations.
 245
 246
 247 unop("fsin", tfloat, "bit_size == 64 ? sin(src0) : sinf(src0)")
 248 unop("fcos", tfloat, "bit_size == 64 ? cos(src0) : cosf(src0)")
 249
 250 # dfrexp
 251 unop_convert("frexp_exp", tint32, tfloat, "frexp(src0, &dst);")
 252 unop_convert("frexp_sig", tfloat, tfloat, "int n; dst = frexp(src0, &n);")
 253
 254 # Partial derivatives.
 255
 256
 257 unop("fddx", tfloat, "0.0") # the derivative of a constant is 0.
 258 unop("fddy", tfloat, "0.0")
 259 unop("fddx_fine", tfloat, "0.0")
 260 unop("fddy_fine", tfloat, "0.0")
 261 unop("fddx_coarse", tfloat, "0.0")
 262 unop("fddy_coarse", tfloat, "0.0")
 263
 264
 265 # Floating point pack and unpack operations.
 266
 267 def pack_2x16(fmt):
 268    unop_horiz("pack_" + fmt + "_2x16", 1, tuint32, 2, tfloat32, """
 269 dst.x = (uint32_t) pack_fmt_1x16(src0.x);
 270 dst.x |= ((uint32_t) pack_fmt_1x16(src0.y)) << 16;
 271 """.replace("fmt", fmt))
 272
 273 def pack_4x8(fmt):
 274    unop_horiz("pack_" + fmt + "_4x8", 1, tuint32, 4, tfloat32, """
 275 dst.x = (uint32_t) pack_fmt_1x8(src0.x);
 276 dst.x |= ((uint32_t) pack_fmt_1x8(src0.y)) << 8;
 277 dst.x |= ((uint32_t) pack_fmt_1x8(src0.z)) << 16;
 278 dst.x |= ((uint32_t) pack_fmt_1x8(src0.w)) << 24;
 279 """.replace("fmt", fmt))
 280
 281 def unpack_2x16(fmt):
 282    unop_horiz("unpack_" + fmt + "_2x16", 2, tfloat32, 1, tuint32, """
 283 dst.x = unpack_fmt_1x16((uint16_t)(src0.x & 0xffff));
 284 dst.y = unpack_fmt_1x16((uint16_t)(src0.x << 16));
 285 """.replace("fmt", fmt))
 286
 287 def unpack_4x8(fmt):
 288    unop_horiz("unpack_" + fmt + "_4x8", 4, tfloat32, 1, tuint32, """
 289 dst.x = unpack_fmt_1x8((uint8_t)(src0.x & 0xff));
 290 dst.y = unpack_fmt_1x8((uint8_t)((src0.x >> 8) & 0xff));
 291 dst.z = unpack_fmt_1x8((uint8_t)((src0.x >> 16) & 0xff));
 292 dst.w = unpack_fmt_1x8((uint8_t)(src0.x >> 24));
 293 """.replace("fmt", fmt))
 294
 295
 296 pack_2x16("snorm")
 297 pack_4x8("snorm")
 298 pack_2x16("unorm")
 299 pack_4x8("unorm")
 300 pack_2x16("half")
 301 unpack_2x16("snorm")
 302 unpack_4x8("snorm")
 303 unpack_2x16("unorm")
 304 unpack_4x8("unorm")
 305 unpack_2x16("half")
 306
 307 unop_horiz("pack_uvec2_to_uint", 1, tuint32, 2, tuint32, """
 308 dst.x = (src0.x & 0xffff) | (src0.y << 16);
 309 """)
 310
 311 unop_horiz("pack_uvec4_to_uint", 1, tuint32, 4, tuint32, """
 312 dst.x = (src0.x <<  0) |
 313         (src0.y <<  8) |
 314         (src0.z << 16) |
 315         (src0.w << 24);
 316 """)
 317
 318 unop_horiz("pack_32_2x16", 1, tuint32, 2, tuint16,
 319            "dst.x = src0.x | ((uint32_t)src0.y << 16);")
 320
 321 unop_horiz("pack_64_2x32", 1, tuint64, 2, tuint32,
 322            "dst.x = src0.x | ((uint64_t)src0.y << 32);")
 323
 324 unop_horiz("pack_64_4x16", 1, tuint64, 4, tuint16,
 325            "dst.x = src0.x | ((uint64_t)src0.y << 16) | ((uint64_t)src0.z << 32) | ((uint64_t)src0.w << 48);")
 326
 327 unop_horiz("unpack_64_2x32", 2, tuint32, 1, tuint64,
 328            "dst.x = src0.x; dst.y = src0.x >> 32;")
 329
 330 unop_horiz("unpack_64_4x16", 4, tuint16, 1, tuint64,
 331            "dst.x = src0.x; dst.y = src0.x >> 16; dst.z = src0.x >> 32; dst.w = src0.w >> 48;")
 332
 333 unop_horiz("unpack_32_2x16", 2, tuint16, 1, tuint32,
 334            "dst.x = src0.x; dst.y = src0.x >> 16;")
 335
 336 # Lowered floating point unpacking operations.
 337
 338
 339 unop_convert("unpack_half_2x16_split_x", tfloat32, tuint32,
 340              "unpack_half_1x16((uint16_t)(src0 & 0xffff))")
 341 unop_convert("unpack_half_2x16_split_y", tfloat32, tuint32,
 342              "unpack_half_1x16((uint16_t)(src0 >> 16))")
 343
 344 unop_convert("unpack_32_2x16_split_x", tuint16, tuint32, "src0")
 345 unop_convert("unpack_32_2x16_split_y", tuint16, tuint32, "src0 >> 16")
 346
 347 unop_convert("unpack_64_2x32_split_x", tuint32, tuint64, "src0")
 348 unop_convert("unpack_64_2x32_split_y", tuint32, tuint64, "src0 >> 32")
 349
 350 # Bit operations, part of ARB_gpu_shader5.
 351
 352
 353 unop("bitfield_reverse", tuint32, """
 354 /* we're not winning any awards for speed here, but that's ok */
 355 dst = 0;
 356 for (unsigned bit = 0; bit < 32; bit++)
 357    dst |= ((src0 >> bit) & 1) << (31 - bit);
 358 """)
 359 unop_convert("bit_count", tuint32, tuint, """
 360 dst = 0;
 361 for (unsigned bit = 0; bit < bit_size; bit++) {
 362    if ((src0 >> bit) & 1)
 363       dst++;
 364 }
 365 """)
 366
 367 unop_convert("ufind_msb", tint32, tuint, """
 368 dst = -1;
 369 for (int bit = bit_size - 1; bit >= 0; bit--) {
 370    if ((src0 >> bit) & 1) {
 371       dst = bit;
 372       break;
 373    }
 374 }
 375 """)
 376
 377 unop("ifind_msb", tint32, """
 378 dst = -1;
 379 for (int bit = 31; bit >= 0; bit--) {
 380    /* If src0 < 0, we're looking for the first 0 bit.
 381     * if src0 >= 0, we're looking for the first 1 bit.
 382     */
 383    if ((((src0 >> bit) & 1) && (src0 >= 0)) ||
 384       (!((src0 >> bit) & 1) && (src0 < 0))) {
 385       dst = bit;
 386       break;
 387    }
 388 }
 389 """)
 390
 391 unop_convert("find_lsb", tint32, tint, """
 392 dst = -1;
 393 for (unsigned bit = 0; bit < bit_size; bit++) {
 394    if ((src0 >> bit) & 1) {
 395       dst = bit;
 396       break;
 397    }
 398 }
 399 """)
 400
 401
 402 for i in range(1, 5):
 403    for j in range(1, 5):
 404       unop_horiz("fnoise{0}_{1}".format(i, j), i, tfloat, j, tfloat, "0.0f")
 405
 406
 407 # AMD_gcn_shader extended instructions
 408 unop_horiz("cube_face_coord", 2, tfloat32, 3, tfloat32, """
 409 dst.x = dst.y = 0.0;
 410 float absX = fabs(src0.x);
 411 float absY = fabs(src0.y);
 412 float absZ = fabs(src0.z);
 413
 414 float ma = 0.0;
 415 if (absX >= absY && absX >= absZ) { ma = 2 * src0.x; }
 416 if (absY >= absX && absY >= absZ) { ma = 2 * src0.y; }
 417 if (absZ >= absX && absZ >= absY) { ma = 2 * src0.z; }
 418
 419 if (src0.x >= 0 && absX >= absY && absX >= absZ) { dst.x = -src0.z; dst.y = -src0.y; }
 420 if (src0.x < 0 && absX >= absY && absX >= absZ) { dst.x = src0.z; dst.y = -src0.y; }
 421 if (src0.y >= 0 && absY >= absX && absY >= absZ) { dst.x = src0.x; dst.y = src0.z; }
 422 if (src0.y < 0 && absY >= absX && absY >= absZ) { dst.x = src0.x; dst.y = -src0.z; }
 423 if (src0.z >= 0 && absZ >= absX && absZ >= absY) { dst.x = src0.x; dst.y = -src0.y; }
 424 if (src0.z < 0 && absZ >= absX && absZ >= absY) { dst.x = -src0.x; dst.y = -src0.y; }
 425
 426 dst.x = dst.x / ma + 0.5;
 427 dst.y = dst.y / ma + 0.5;
 428 """)
 429
 430 unop_horiz("cube_face_index", 1, tfloat32, 3, tfloat32, """
 431 float absX = fabs(src0.x);
 432 float absY = fabs(src0.y);
 433 float absZ = fabs(src0.z);
 434 if (src0.x >= 0 && absX >= absY && absX >= absZ) dst.x = 0;
 435 if (src0.x < 0 && absX >= absY && absX >= absZ) dst.x = 1;
 436 if (src0.y >= 0 && absY >= absX && absY >= absZ) dst.x = 2;
 437 if (src0.y < 0 && absY >= absX && absY >= absZ) dst.x = 3;
 438 if (src0.z >= 0 && absZ >= absX && absZ >= absY) dst.x = 4;
 439 if (src0.z < 0 && absZ >= absX && absZ >= absY) dst.x = 5;
 440 """)
 441
 442 # Sum of vector components
 443 unop_reduce("fsum", 1, tfloat, tfloat, "{src}", "{src0} + {src1}", "{src}")
 444
 445 def binop_convert(name, out_type, in_type, alg_props, const_expr):
 446    opcode(name, 0, out_type, [0, 0], [in_type, in_type],
 447           False, alg_props, const_expr)
 448
 449 def binop(name, ty, alg_props, const_expr):
 450    binop_convert(name, ty, ty, alg_props, const_expr)
 451
 452 def binop_compare(name, ty, alg_props, const_expr):
 453    binop_convert(name, tbool1, ty, alg_props, const_expr)
 454
 455 def binop_compare32(name, ty, alg_props, const_expr):
 456    binop_convert(name, tbool32, ty, alg_props, const_expr)
 457
 458 def binop_horiz(name, out_size, out_type, src1_size, src1_type, src2_size,
 459                 src2_type, const_expr):
 460    opcode(name, out_size, out_type, [src1_size, src2_size], [src1_type, src2_type],
 461           False, "", const_expr)
 462
 463 def binop_reduce(name, output_size, output_type, src_type, prereduce_expr,
 464                  reduce_expr, final_expr):
 465    def final(src):
 466       return final_expr.format(src= "(" + src + ")")
 467    def reduce_(src0, src1):
 468       return reduce_expr.format(src0=src0, src1=src1)
 469    def prereduce(src0, src1):
 470       return "(" + prereduce_expr.format(src0=src0, src1=src1) + ")"
 471    src0 = prereduce("src0.x", "src1.x")
 472    src1 = prereduce("src0.y", "src1.y")
 473    src2 = prereduce("src0.z", "src1.z")
 474    src3 = prereduce("src0.w", "src1.w")
 475    opcode(name + "2", output_size, output_type,
 476           [2, 2], [src_type, src_type], False, _2src_commutative,
 477           final(reduce_(src0, src1)))
 478    opcode(name + "3", output_size, output_type,
 479           [3, 3], [src_type, src_type], False, _2src_commutative,
 480           final(reduce_(reduce_(src0, src1), src2)))
 481    opcode(name + "4", output_size, output_type,
 482           [4, 4], [src_type, src_type], False, _2src_commutative,
 483           final(reduce_(reduce_(src0, src1), reduce_(src2, src3))))
 484
 485 binop("fadd", tfloat, _2src_commutative + associative, "src0 + src1")
 486 binop("iadd", tint, _2src_commutative + associative, "src0 + src1")
 487 binop("iadd_sat", tint, _2src_commutative, """
 488       src1 > 0 ?
 489          (src0 + src1 < src0 ? (1ull << (bit_size - 1)) - 1 : src0 + src1) :
 490          (src0 < src0 + src1 ? (1ull << (bit_size - 1))     : src0 + src1)
 491 """)
 492 binop("uadd_sat", tuint, _2src_commutative,
 493       "(src0 + src1) < src0 ? MAX_UINT_FOR_SIZE(sizeof(src0) * 8) : (src0 + src1)")
 494 binop("isub_sat", tint, "", """
 495       src1 < 0 ?
 496          (src0 - src1 < src0 ? (1ull << (bit_size - 1)) - 1 : src0 - src1) :
 497          (src0 < src0 - src1 ? (1ull << (bit_size - 1))     : src0 - src1)
 498 """)
 499 binop("usub_sat", tuint, "", "src0 < src1 ? 0 : src0 - src1")
 500
 501 binop("fsub", tfloat, "", "src0 - src1")
 502 binop("isub", tint, "", "src0 - src1")
 503
 504 binop("fmul", tfloat, _2src_commutative + associative, "src0 * src1")
 505 # low 32-bits of signed/unsigned integer multiply
 506 binop("imul", tint, _2src_commutative + associative, "src0 * src1")
 507
 508 # Generate 64 bit result from 2 32 bits quantity
 509 binop_convert("imul_2x32_64", tint64, tint32, _2src_commutative,
 510               "(int64_t)src0 * (int64_t)src1")
 511 binop_convert("umul_2x32_64", tuint64, tuint32, _2src_commutative,
 512               "(uint64_t)src0 * (uint64_t)src1")
 513
 514 # high 32-bits of signed integer multiply
 515 binop("imul_high", tint, _2src_commutative, """
 516 if (bit_size == 64) {
 517    /* We need to do a full 128-bit x 128-bit multiply in order for the sign
 518     * extension to work properly.  The casts are kind-of annoying but needed
 519     * to prevent compiler warnings.
 520     */
 521    uint32_t src0_u32[4] = {
 522       src0,
 523       (int64_t)src0 >> 32,
 524       (int64_t)src0 >> 63,
 525       (int64_t)src0 >> 63,
 526    };
 527    uint32_t src1_u32[4] = {
 528       src1,
 529       (int64_t)src1 >> 32,
 530       (int64_t)src1 >> 63,
 531       (int64_t)src1 >> 63,
 532    };
 533    uint32_t prod_u32[4];
 534    ubm_mul_u32arr(prod_u32, src0_u32, src1_u32);
 535    dst = (uint64_t)prod_u32[2] | ((uint64_t)prod_u32[3] << 32);
 536 } else {
 537    dst = ((int64_t)src0 * (int64_t)src1) >> bit_size;
 538 }
 539 """)
 540
 541 # high 32-bits of unsigned integer multiply
 542 binop("umul_high", tuint, _2src_commutative, """
 543 if (bit_size == 64) {
 544    /* The casts are kind-of annoying but needed to prevent compiler warnings. */
 545    uint32_t src0_u32[2] = { src0, (uint64_t)src0 >> 32 };
 546    uint32_t src1_u32[2] = { src1, (uint64_t)src1 >> 32 };
 547    uint32_t prod_u32[4];
 548    ubm_mul_u32arr(prod_u32, src0_u32, src1_u32);
 549    dst = (uint64_t)prod_u32[2] | ((uint64_t)prod_u32[3] << 32);
 550 } else {
 551    dst = ((uint64_t)src0 * (uint64_t)src1) >> bit_size;
 552 }
 553 """)
 554
 555 # low 32-bits of unsigned integer multiply
 556 binop("umul_low", tuint32, _2src_commutative, """
 557 uint64_t mask = (1 << (bit_size / 2)) - 1;
 558 dst = ((uint64_t)src0 & mask) * ((uint64_t)src1 & mask);
 559 """)
 560
 561
 562 binop("fdiv", tfloat, "", "src0 / src1")
 563 binop("idiv", tint, "", "src1 == 0 ? 0 : (src0 / src1)")
 564 binop("udiv", tuint, "", "src1 == 0 ? 0 : (src0 / src1)")
 565
 566 # returns a boolean representing the carry resulting from the addition of
 567 # the two unsigned arguments.
 568
 569 binop_convert("uadd_carry", tuint, tuint, _2src_commutative, "src0 + src1 < src0")
 570
 571 # returns a boolean representing the borrow resulting from the subtraction
 572 # of the two unsigned arguments.
 573
 574 binop_convert("usub_borrow", tuint, tuint, "", "src0 < src1")
 575
 576 # hadd: (a + b) >> 1 (without overflow)
 577 # x + y = x - (x & ~y) + (x & ~y) + y - (~x & y) + (~x & y)
 578 #       =      (x & y) + (x & ~y) +      (x & y) + (~x & y)
 579 #       = 2 *  (x & y) + (x & ~y) +                (~x & y)
 580 #       =     ((x & y) << 1) + (x ^ y)
 581 #
 582 # Since we know that the bottom bit of (x & y) << 1 is zero,
 583 #
 584 # (x + y) >> 1 = (((x & y) << 1) + (x ^ y)) >> 1
 585 #              =   (x & y) +      ((x ^ y)  >> 1)
 586 binop("ihadd", tint, _2src_commutative, "(src0 & src1) + ((src0 ^ src1) >> 1)")
 587 binop("uhadd", tuint, _2src_commutative, "(src0 & src1) + ((src0 ^ src1) >> 1)")
 588
 589 # rhadd: (a + b + 1) >> 1 (without overflow)
 590 # x + y + 1 = x + (~x & y) - (~x & y) + y + (x & ~y) - (x & ~y) + 1
 591 #           =      (x | y) - (~x & y) +      (x | y) - (x & ~y) + 1
 592 #           = 2 *  (x | y) - ((~x & y) +               (x & ~y)) + 1
 593 #           =     ((x | y) << 1) - (x ^ y) + 1
 594 #
 595 # Since we know that the bottom bit of (x & y) << 1 is zero,
 596 #
 597 # (x + y + 1) >> 1 = (x | y) + (-(x ^ y) + 1) >> 1)
 598 #                  = (x | y) -  ((x ^ y)      >> 1)
 599 binop("irhadd", tint, _2src_commutative, "(src0 | src1) + ((src0 ^ src1) >> 1)")
 600 binop("urhadd", tuint, _2src_commutative, "(src0 | src1) + ((src0 ^ src1) >> 1)")
 601
 602 binop("umod", tuint, "", "src1 == 0 ? 0 : src0 % src1")
 603
 604 # For signed integers, there are several different possible definitions of
 605 # "modulus" or "remainder".  We follow the conventions used by LLVM and
 606 # SPIR-V.  The irem opcode implements the standard C/C++ signed "%"
 607 # operation while the imod opcode implements the more mathematical
 608 # "modulus" operation.  For details on the difference, see
 609 #
 610 # http://mathforum.org/library/drmath/view/52343.html
 611
 612 binop("irem", tint, "", "src1 == 0 ? 0 : src0 % src1")
 613 binop("imod", tint, "",
 614       "src1 == 0 ? 0 : ((src0 % src1 == 0 || (src0 >= 0) == (src1 >= 0)) ?"
 615       "                 src0 % src1 : src0 % src1 + src1)")
 616 binop("fmod", tfloat, "", "src0 - src1 * floorf(src0 / src1)")
 617 binop("frem", tfloat, "", "src0 - src1 * truncf(src0 / src1)")
 618
 619 #
 620 # Comparisons
 621 #
 622
 623
 624 # these integer-aware comparisons return a boolean (0 or ~0)
 625
 626 binop_compare("flt", tfloat, "", "src0 < src1")
 627 binop_compare("fge", tfloat, "", "src0 >= src1")
 628 binop_compare("feq", tfloat, _2src_commutative, "src0 == src1")
 629 binop_compare("fne", tfloat, _2src_commutative, "src0 != src1")
 630 binop_compare("ilt", tint, "", "src0 < src1")
 631 binop_compare("ige", tint, "", "src0 >= src1")
 632 binop_compare("ieq", tint, _2src_commutative, "src0 == src1")
 633 binop_compare("ine", tint, _2src_commutative, "src0 != src1")
 634 binop_compare("ult", tuint, "", "src0 < src1")
 635 binop_compare("uge", tuint, "", "src0 >= src1")
 636 binop_compare32("flt32", tfloat, "", "src0 < src1")
 637 binop_compare32("fge32", tfloat, "", "src0 >= src1")
 638 binop_compare32("feq32", tfloat, _2src_commutative, "src0 == src1")
 639 binop_compare32("fne32", tfloat, _2src_commutative, "src0 != src1")
 640 binop_compare32("ilt32", tint, "", "src0 < src1")
 641 binop_compare32("ige32", tint, "", "src0 >= src1")
 642 binop_compare32("ieq32", tint, _2src_commutative, "src0 == src1")
 643 binop_compare32("ine32", tint, _2src_commutative, "src0 != src1")
 644 binop_compare32("ult32", tuint, "", "src0 < src1")
 645 binop_compare32("uge32", tuint, "", "src0 >= src1")
 646
 647 # integer-aware GLSL-style comparisons that compare floats and ints
 648
 649 binop_reduce("ball_fequal",  1, tbool1, tfloat, "{src0} == {src1}",
 650              "{src0} && {src1}", "{src}")
 651 binop_reduce("bany_fnequal", 1, tbool1, tfloat, "{src0} != {src1}",
 652              "{src0} || {src1}", "{src}")
 653 binop_reduce("ball_iequal",  1, tbool1, tint, "{src0} == {src1}",
 654              "{src0} && {src1}", "{src}")
 655 binop_reduce("bany_inequal", 1, tbool1, tint, "{src0} != {src1}",
 656              "{src0} || {src1}", "{src}")
 657
 658 binop_reduce("b32all_fequal",  1, tbool32, tfloat, "{src0} == {src1}",
 659              "{src0} && {src1}", "{src}")
 660 binop_reduce("b32any_fnequal", 1, tbool32, tfloat, "{src0} != {src1}",
 661              "{src0} || {src1}", "{src}")
 662 binop_reduce("b32all_iequal",  1, tbool32, tint, "{src0} == {src1}",
 663              "{src0} && {src1}", "{src}")
 664 binop_reduce("b32any_inequal", 1, tbool32, tint, "{src0} != {src1}",
 665              "{src0} || {src1}", "{src}")
 666
 667 # non-integer-aware GLSL-style comparisons that return 0.0 or 1.0
 668
 669 binop_reduce("fall_equal",  1, tfloat32, tfloat32, "{src0} == {src1}",
 670              "{src0} && {src1}", "{src} ? 1.0f : 0.0f")
 671 binop_reduce("fany_nequal", 1, tfloat32, tfloat32, "{src0} != {src1}",
 672              "{src0} || {src1}", "{src} ? 1.0f : 0.0f")
 673
 674 # These comparisons for integer-less hardware return 1.0 and 0.0 for true
 675 # and false respectively
 676
 677 binop("slt", tfloat32, "", "(src0 < src1) ? 1.0f : 0.0f") # Set on Less Than
 678 binop("sge", tfloat, "", "(src0 >= src1) ? 1.0f : 0.0f") # Set on Greater or Equal
 679 binop("seq", tfloat32, _2src_commutative, "(src0 == src1) ? 1.0f : 0.0f") # Set on Equal
 680 binop("sne", tfloat32, _2src_commutative, "(src0 != src1) ? 1.0f : 0.0f") # Set on Not Equal
 681
 682 # SPIRV shifts are undefined for shift-operands >= bitsize,
 683 # but SM5 shifts are defined to use the least significant bits, only
 684 # The NIR definition is according to the SM5 specification.
 685 opcode("ishl", 0, tint, [0, 0], [tint, tuint32], False, "",
 686        "src0 << (src1 & (sizeof(src0) * 8 - 1))")
 687 opcode("ishr", 0, tint, [0, 0], [tint, tuint32], False, "",
 688        "src0 >> (src1 & (sizeof(src0) * 8 - 1))")
 689 opcode("ushr", 0, tuint, [0, 0], [tuint, tuint32], False, "",
 690        "src0 >> (src1 & (sizeof(src0) * 8 - 1))")
 691
 692 opcode("urol", 0, tuint, [0, 0], [tuint, tuint32], False, "", """
 693    uint32_t rotate_mask = sizeof(src0) * 8 - 1;
 694    dst = (src0 << (src1 & rotate_mask)) |
 695          (src0 >> (-src1 & rotate_mask));
 696 """)
 697 opcode("uror", 0, tuint, [0, 0], [tuint, tuint32], False, "", """
 698    uint32_t rotate_mask = sizeof(src0) * 8 - 1;
 699    dst = (src0 >> (src1 & rotate_mask)) |
 700          (src0 << (-src1 & rotate_mask));
 701 """)
 702
 703 # bitwise logic operators
 704 #
 705 # These are also used as boolean and, or, xor for hardware supporting
 706 # integers.
 707
 708
 709 binop("iand", tuint, _2src_commutative + associative, "src0 & src1")
 710 binop("ior", tuint, _2src_commutative + associative, "src0 | src1")
 711 binop("ixor", tuint, _2src_commutative + associative, "src0 ^ src1")
 712
 713
 714 binop_reduce("fdot", 1, tfloat, tfloat, "{src0} * {src1}", "{src0} + {src1}",
 715              "{src}")
 716
 717 binop_reduce("fdot_replicated", 4, tfloat, tfloat,
 718              "{src0} * {src1}", "{src0} + {src1}", "{src}")
 719
 720 opcode("fdph", 1, tfloat, [3, 4], [tfloat, tfloat], False, "",
 721        "src0.x * src1.x + src0.y * src1.y + src0.z * src1.z + src1.w")
 722 opcode("fdph_replicated", 4, tfloat, [3, 4], [tfloat, tfloat], False, "",
 723        "src0.x * src1.x + src0.y * src1.y + src0.z * src1.z + src1.w")
 724
 725 binop("fmin", tfloat, "", "fminf(src0, src1)")
 726 binop("imin", tint, _2src_commutative + associative, "src1 > src0 ? src0 : src1")
 727 binop("umin", tuint, _2src_commutative + associative, "src1 > src0 ? src0 : src1")
 728 binop("fmax", tfloat, "", "fmaxf(src0, src1)")
 729 binop("imax", tint, _2src_commutative + associative, "src1 > src0 ? src1 : src0")
 730 binop("umax", tuint, _2src_commutative + associative, "src1 > src0 ? src1 : src0")
 731
 732 # Saturated vector add for 4 8bit ints.
 733 binop("usadd_4x8", tint32, _2src_commutative + associative, """
 734 dst = 0;
 735 for (int i = 0; i < 32; i += 8) {
 736    dst |= MIN2(((src0 >> i) & 0xff) + ((src1 >> i) & 0xff), 0xff) << i;
 737 }
 738 """)
 739
 740 # Saturated vector subtract for 4 8bit ints.
 741 binop("ussub_4x8", tint32, "", """
 742 dst = 0;
 743 for (int i = 0; i < 32; i += 8) {
 744    int src0_chan = (src0 >> i) & 0xff;
 745    int src1_chan = (src1 >> i) & 0xff;
 746    if (src0_chan > src1_chan)
 747       dst |= (src0_chan - src1_chan) << i;
 748 }
 749 """)
 750
 751 # vector min for 4 8bit ints.
 752 binop("umin_4x8", tint32, _2src_commutative + associative, """
 753 dst = 0;
 754 for (int i = 0; i < 32; i += 8) {
 755    dst |= MIN2((src0 >> i) & 0xff, (src1 >> i) & 0xff) << i;
 756 }
 757 """)
 758
 759 # vector max for 4 8bit ints.
 760 binop("umax_4x8", tint32, _2src_commutative + associative, """
 761 dst = 0;
 762 for (int i = 0; i < 32; i += 8) {
 763    dst |= MAX2((src0 >> i) & 0xff, (src1 >> i) & 0xff) << i;
 764 }
 765 """)
 766
 767 # unorm multiply: (a * b) / 255.
 768 binop("umul_unorm_4x8", tint32, _2src_commutative + associative, """
 769 dst = 0;
 770 for (int i = 0; i < 32; i += 8) {
 771    int src0_chan = (src0 >> i) & 0xff;
 772    int src1_chan = (src1 >> i) & 0xff;
 773    dst |= ((src0_chan * src1_chan) / 255) << i;
 774 }
 775 """)
 776
 777 binop("fpow", tfloat, "", "bit_size == 64 ? powf(src0, src1) : pow(src0, src1)")
 778
 779 binop_horiz("pack_half_2x16_split", 1, tuint32, 1, tfloat32, 1, tfloat32,
 780             "pack_half_1x16(src0.x) | (pack_half_1x16(src1.x) << 16)")
 781
 782 binop_convert("pack_64_2x32_split", tuint64, tuint32, "",
 783               "src0 | ((uint64_t)src1 << 32)")
 784
 785 binop_convert("pack_32_2x16_split", tuint32, tuint16, "",
 786               "src0 | ((uint32_t)src1 << 16)")
 787
 788 # bfm implements the behavior of the first operation of the SM5 "bfi" assembly
 789 # and that of the "bfi1" i965 instruction. That is, the bits and offset values
 790 # are from the low five bits of src0 and src1, respectively.
 791 binop_convert("bfm", tuint32, tint32, "", """
 792 int bits = src0 & 0x1F;
 793 int offset = src1 & 0x1F;
 794 dst = ((1u << bits) - 1) << offset;
 795 """)
 796
 797 opcode("ldexp", 0, tfloat, [0, 0], [tfloat, tint32], False, "", """
 798 dst = (bit_size == 64) ? ldexp(src0, src1) : ldexpf(src0, src1);
 799 /* flush denormals to zero. */
 800 if (!isnormal(dst))
 801    dst = copysignf(0.0f, src0);
 802 """)
 803
 804 # Combines the first component of each input to make a 2-component vector.
 805
 806 binop_horiz("vec2", 2, tuint, 1, tuint, 1, tuint, """
 807 dst.x = src0.x;
 808 dst.y = src1.x;
 809 """)
 810
 811 # Byte extraction
 812 binop("extract_u8", tuint, "", "(uint8_t)(src0 >> (src1 * 8))")
 813 binop("extract_i8", tint, "", "(int8_t)(src0 >> (src1 * 8))")
 814
 815 # Word extraction
 816 binop("extract_u16", tuint, "", "(uint16_t)(src0 >> (src1 * 16))")
 817 binop("extract_i16", tint, "", "(int16_t)(src0 >> (src1 * 16))")
 818
 819
 820 def triop(name, ty, alg_props, const_expr):
 821    opcode(name, 0, ty, [0, 0, 0], [ty, ty, ty], False, alg_props, const_expr)
 822 def triop_horiz(name, output_size, src1_size, src2_size, src3_size, const_expr):
 823    opcode(name, output_size, tuint,
 824    [src1_size, src2_size, src3_size],
 825    [tuint, tuint, tuint], False, "", const_expr)
 826
 827 triop("ffma", tfloat, _2src_commutative, "src0 * src1 + src2")
 828
 829 triop("flrp", tfloat, "", "src0 * (1 - src2) + src1 * src2")
 830
 831 # Conditional Select
 832 #
 833 # A vector conditional select instruction (like ?:, but operating per-
 834 # component on vectors). There are two versions, one for floating point
 835 # bools (0.0 vs 1.0) and one for integer bools (0 vs ~0).
 836
 837
 838 triop("fcsel", tfloat32, "", "(src0 != 0.0f) ? src1 : src2")
 839
 840 # 3 way min/max/med
 841 triop("fmin3", tfloat, "", "fminf(src0, fminf(src1, src2))")
 842 triop("imin3", tint, "", "MIN2(src0, MIN2(src1, src2))")
 843 triop("umin3", tuint, "", "MIN2(src0, MIN2(src1, src2))")
 844
 845 triop("fmax3", tfloat, "", "fmaxf(src0, fmaxf(src1, src2))")
 846 triop("imax3", tint, "", "MAX2(src0, MAX2(src1, src2))")
 847 triop("umax3", tuint, "", "MAX2(src0, MAX2(src1, src2))")
 848
 849 triop("fmed3", tfloat, "", "fmaxf(fminf(fmaxf(src0, src1), src2), fminf(src0, src1))")
 850 triop("imed3", tint, "", "MAX2(MIN2(MAX2(src0, src1), src2), MIN2(src0, src1))")
 851 triop("umed3", tuint, "", "MAX2(MIN2(MAX2(src0, src1), src2), MIN2(src0, src1))")
 852
 853 opcode("bcsel", 0, tuint, [0, 0, 0],
 854       [tbool1, tuint, tuint], False, "", "src0 ? src1 : src2")
 855 opcode("b32csel", 0, tuint, [0, 0, 0],
 856        [tbool32, tuint, tuint], False, "", "src0 ? src1 : src2")
 857
 858 # SM5 bfi assembly
 859 triop("bfi", tuint32, "", """
 860 unsigned mask = src0, insert = src1, base = src2;
 861 if (mask == 0) {
 862    dst = base;
 863 } else {
 864    unsigned tmp = mask;
 865    while (!(tmp & 1)) {
 866       tmp >>= 1;
 867       insert <<= 1;
 868    }
 869    dst = (base & ~mask) | (insert & mask);
 870 }
 871 """)
 872
 873
 874 triop("bitfield_select", tuint, "", "(src0 & src1) | (~src0 & src2)")
 875
 876 # SM5 ubfe/ibfe assembly: only the 5 least significant bits of offset and bits are used.
 877 opcode("ubfe", 0, tuint32,
 878        [0, 0, 0], [tuint32, tuint32, tuint32], False, "", """
 879 unsigned base = src0;
 880 unsigned offset = src1 & 0x1F;
 881 unsigned bits = src2 & 0x1F;
 882 if (bits == 0) {
 883    dst = 0;
 884 } else if (offset + bits < 32) {
 885    dst = (base << (32 - bits - offset)) >> (32 - bits);
 886 } else {
 887    dst = base >> offset;
 888 }
 889 """)
 890 opcode("ibfe", 0, tint32,
 891        [0, 0, 0], [tint32, tuint32, tuint32], False, "", """
 892 int base = src0;
 893 unsigned offset = src1 & 0x1F;
 894 unsigned bits = src2 & 0x1F;
 895 if (bits == 0) {
 896    dst = 0;
 897 } else if (offset + bits < 32) {
 898    dst = (base << (32 - bits - offset)) >> (32 - bits);
 899 } else {
 900    dst = base >> offset;
 901 }
 902 """)
 903
 904 # GLSL bitfieldExtract()
 905 opcode("ubitfield_extract", 0, tuint32,
 906        [0, 0, 0], [tuint32, tint32, tint32], False, "", """
 907 unsigned base = src0;
 908 int offset = src1, bits = src2;
 909 if (bits == 0) {
 910    dst = 0;
 911 } else if (bits < 0 || offset < 0 || offset + bits > 32) {
 912    dst = 0; /* undefined per the spec */
 913 } else {
 914    dst = (base >> offset) & ((1ull << bits) - 1);
 915 }
 916 """)
 917 opcode("ibitfield_extract", 0, tint32,
 918        [0, 0, 0], [tint32, tint32, tint32], False, "", """
 919 int base = src0;
 920 int offset = src1, bits = src2;
 921 if (bits == 0) {
 922    dst = 0;
 923 } else if (offset < 0 || bits < 0 || offset + bits > 32) {
 924    dst = 0;
 925 } else {
 926    dst = (base << (32 - offset - bits)) >> offset; /* use sign-extending shift */
 927 }
 928 """)
 929
 930 # Combines the first component of each input to make a 3-component vector.
 931
 932 triop_horiz("vec3", 3, 1, 1, 1, """
 933 dst.x = src0.x;
 934 dst.y = src1.x;
 935 dst.z = src2.x;
 936 """)
 937
 938 def quadop_horiz(name, output_size, src1_size, src2_size, src3_size,
 939                  src4_size, const_expr):
 940    opcode(name, output_size, tuint,
 941           [src1_size, src2_size, src3_size, src4_size],
 942           [tuint, tuint, tuint, tuint],
 943           False, "", const_expr)
 944
 945 opcode("bitfield_insert", 0, tuint32, [0, 0, 0, 0],
 946        [tuint32, tuint32, tint32, tint32], False, "", """
 947 unsigned base = src0, insert = src1;
 948 int offset = src2, bits = src3;
 949 if (bits == 0) {
 950    dst = base;
 951 } else if (offset < 0 || bits < 0 || bits + offset > 32) {
 952    dst = 0;
 953 } else {
 954    unsigned mask = ((1ull << bits) - 1) << offset;
 955    dst = (base & ~mask) | ((insert << offset) & mask);
 956 }
 957 """)
 958
 959 quadop_horiz("vec4", 4, 1, 1, 1, 1, """
 960 dst.x = src0.x;
 961 dst.y = src1.x;
 962 dst.z = src2.x;
 963 dst.w = src3.x;
 964 """)
 965
 966 # ir3-specific instruction that maps directly to mul-add shift high mix,
 967 # (IMADSH_MIX16 i.e. ah * bl << 16 + c). It is used for lowering integer
 968 # multiplication (imul) on Freedreno backend..
 969 opcode("imadsh_mix16", 1, tint32,
 970        [1, 1, 1], [tint32, tint32, tint32], False, "", """
 971 dst.x = ((((src0.x & 0xffff0000) >> 16) * (src1.x & 0x0000ffff)) << 16) + src2.x;
 972 """)