Signed-off-by: Elie Tournier <elie.tourn...@collabora.com> --- src/compiler/glsl/builtin_float64.h | 446 ++++++++++++++++++++++++++++++++ src/compiler/glsl/builtin_functions.cpp | 4 + src/compiler/glsl/builtin_functions.h | 3 + src/compiler/glsl/float64.glsl | 68 +++++ src/compiler/glsl/glcpp/glcpp-parse.y | 1 + 5 files changed, 522 insertions(+)
diff --git a/src/compiler/glsl/builtin_float64.h b/src/compiler/glsl/builtin_float64.h index 5213751223..bfe387b0f5 100644 --- a/src/compiler/glsl/builtin_float64.h +++ b/src/compiler/glsl/builtin_float64.h @@ -11848,3 +11848,449 @@ fmul64(void *mem_ctx, builtin_available_predicate avail) sig->replace_parameters(&sig_parameters); return sig; } +ir_function_signature * +shift64Right(void *mem_ctx, builtin_available_predicate avail) +{ + ir_function_signature *const sig = + new(mem_ctx) ir_function_signature(glsl_type::void_type, avail); + ir_factory body(&sig->body, mem_ctx); + sig->is_defined = true; + + exec_list sig_parameters; + + ir_variable *const r0A04 = new(mem_ctx) ir_variable(glsl_type::uint_type, "a0", ir_var_function_in); + sig_parameters.push_tail(r0A04); + ir_variable *const r0A05 = new(mem_ctx) ir_variable(glsl_type::uint_type, "a1", ir_var_function_in); + sig_parameters.push_tail(r0A05); + ir_variable *const r0A06 = new(mem_ctx) ir_variable(glsl_type::int_type, "count", ir_var_function_in); + sig_parameters.push_tail(r0A06); + ir_variable *const r0A07 = new(mem_ctx) ir_variable(glsl_type::uint_type, "z0Ptr", ir_var_function_inout); + sig_parameters.push_tail(r0A07); + ir_variable *const r0A08 = new(mem_ctx) ir_variable(glsl_type::uint_type, "z1Ptr", ir_var_function_inout); + sig_parameters.push_tail(r0A08); + ir_variable *const r0A09 = new(mem_ctx) ir_variable(glsl_type::uint_type, "z1", ir_var_auto); + body.emit(r0A09); + ir_variable *const r0A0A = new(mem_ctx) ir_variable(glsl_type::uint_type, "z0", ir_var_auto); + body.emit(r0A0A); + ir_variable *const r0A0B = body.make_temp(glsl_type::int_type, "assignment_tmp"); + ir_expression *const r0A0C = neg(r0A06); + body.emit(assign(r0A0B, bit_and(r0A0C, body.constant(int(31))), 0x01)); + + /* IF CONDITION */ + ir_expression *const r0A0E = equal(r0A06, body.constant(int(0))); + ir_if *f0A0D = new(mem_ctx) ir_if(operand(r0A0E).val); + exec_list *const f0A0D_parent_instructions = body.instructions; + + /* THEN INSTRUCTIONS */ + body.instructions = &f0A0D->then_instructions; + + body.emit(assign(r0A09, r0A05, 0x01)); + + body.emit(assign(r0A0A, r0A04, 0x01)); + + + /* ELSE INSTRUCTIONS */ + body.instructions = &f0A0D->else_instructions; + + /* IF CONDITION */ + ir_expression *const r0A10 = less(r0A06, body.constant(int(32))); + ir_if *f0A0F = new(mem_ctx) ir_if(operand(r0A10).val); + exec_list *const f0A0F_parent_instructions = body.instructions; + + /* THEN INSTRUCTIONS */ + body.instructions = &f0A0F->then_instructions; + + ir_expression *const r0A11 = lshift(r0A04, r0A0B); + ir_expression *const r0A12 = rshift(r0A05, r0A06); + body.emit(assign(r0A09, bit_or(r0A11, r0A12), 0x01)); + + body.emit(assign(r0A0A, rshift(r0A04, r0A06), 0x01)); + + + /* ELSE INSTRUCTIONS */ + body.instructions = &f0A0F->else_instructions; + + ir_variable *const r0A13 = body.make_temp(glsl_type::uint_type, "conditional_tmp"); + /* IF CONDITION */ + ir_expression *const r0A15 = less(r0A06, body.constant(int(64))); + ir_if *f0A14 = new(mem_ctx) ir_if(operand(r0A15).val); + exec_list *const f0A14_parent_instructions = body.instructions; + + /* THEN INSTRUCTIONS */ + body.instructions = &f0A14->then_instructions; + + ir_expression *const r0A16 = bit_and(r0A06, body.constant(int(31))); + body.emit(assign(r0A13, rshift(r0A04, r0A16), 0x01)); + + + /* ELSE INSTRUCTIONS */ + body.instructions = &f0A14->else_instructions; + + body.emit(assign(r0A13, body.constant(0u), 0x01)); + + + body.instructions = f0A14_parent_instructions; + body.emit(f0A14); + + /* END IF */ + + body.emit(assign(r0A09, r0A13, 0x01)); + + body.emit(assign(r0A0A, body.constant(0u), 0x01)); + + + body.instructions = f0A0F_parent_instructions; + body.emit(f0A0F); + + /* END IF */ + + + body.instructions = f0A0D_parent_instructions; + body.emit(f0A0D); + + /* END IF */ + + body.emit(assign(r0A08, r0A09, 0x01)); + + body.emit(assign(r0A07, r0A0A, 0x01)); + + sig->replace_parameters(&sig_parameters); + return sig; +} +ir_function_signature * +fp64_to_uint(void *mem_ctx, builtin_available_predicate avail) +{ + ir_function_signature *const sig = + new(mem_ctx) ir_function_signature(glsl_type::uint_type, avail); + ir_factory body(&sig->body, mem_ctx); + sig->is_defined = true; + + exec_list sig_parameters; + + ir_variable *const r0A17 = new(mem_ctx) ir_variable(glsl_type::uvec2_type, "a", ir_var_function_in); + sig_parameters.push_tail(r0A17); + ir_variable *const r0A18 = body.make_temp(glsl_type::uint_type, "return_value"); + ir_variable *const r0A19 = new(mem_ctx) ir_variable(glsl_type::uint_type, "aFracHi", ir_var_auto); + body.emit(r0A19); + ir_variable *const r0A1A = new(mem_ctx) ir_variable(glsl_type::uint_type, "aFracLo", ir_var_auto); + body.emit(r0A1A); + ir_variable *const r0A1B = body.make_temp(glsl_type::uint_type, "extractFloat64FracLo_retval"); + body.emit(assign(r0A1B, swizzle_x(r0A17), 0x01)); + + body.emit(assign(r0A1A, r0A1B, 0x01)); + + ir_variable *const r0A1C = body.make_temp(glsl_type::uint_type, "extractFloat64FracHi_retval"); + body.emit(assign(r0A1C, bit_and(swizzle_y(r0A17), body.constant(1048575u)), 0x01)); + + body.emit(assign(r0A19, r0A1C, 0x01)); + + ir_variable *const r0A1D = body.make_temp(glsl_type::int_type, "extractFloat64Exp_retval"); + ir_expression *const r0A1E = rshift(swizzle_y(r0A17), body.constant(int(20))); + ir_expression *const r0A1F = bit_and(r0A1E, body.constant(2047u)); + body.emit(assign(r0A1D, expr(ir_unop_u2i, r0A1F), 0x01)); + + ir_variable *const r0A20 = body.make_temp(glsl_type::uint_type, "extractFloat64Sign_retval"); + body.emit(assign(r0A20, rshift(swizzle_y(r0A17), body.constant(int(31))), 0x01)); + + /* IF CONDITION */ + ir_expression *const r0A22 = nequal(r0A20, body.constant(0u)); + ir_if *f0A21 = new(mem_ctx) ir_if(operand(r0A22).val); + exec_list *const f0A21_parent_instructions = body.instructions; + + /* THEN INSTRUCTIONS */ + body.instructions = &f0A21->then_instructions; + + body.emit(assign(r0A18, body.constant(0u), 0x01)); + + + /* ELSE INSTRUCTIONS */ + body.instructions = &f0A21->else_instructions; + + /* IF CONDITION */ + ir_expression *const r0A24 = equal(r0A1D, body.constant(int(2047))); + ir_expression *const r0A25 = bit_or(r0A1C, swizzle_x(r0A17)); + ir_expression *const r0A26 = nequal(r0A25, body.constant(0u)); + ir_expression *const r0A27 = logic_and(r0A24, r0A26); + ir_if *f0A23 = new(mem_ctx) ir_if(operand(r0A27).val); + exec_list *const f0A23_parent_instructions = body.instructions; + + /* THEN INSTRUCTIONS */ + body.instructions = &f0A23->then_instructions; + + body.emit(assign(r0A18, body.constant(4294967295u), 0x01)); + + + /* ELSE INSTRUCTIONS */ + body.instructions = &f0A23->else_instructions; + + /* IF CONDITION */ + ir_expression *const r0A29 = nequal(r0A1D, body.constant(int(0))); + ir_if *f0A28 = new(mem_ctx) ir_if(operand(r0A29).val); + exec_list *const f0A28_parent_instructions = body.instructions; + + /* THEN INSTRUCTIONS */ + body.instructions = &f0A28->then_instructions; + + body.emit(assign(r0A19, bit_or(r0A1C, body.constant(1048576u)), 0x01)); + + + body.instructions = f0A28_parent_instructions; + body.emit(f0A28); + + /* END IF */ + + ir_variable *const r0A2A = body.make_temp(glsl_type::int_type, "assignment_tmp"); + body.emit(assign(r0A2A, sub(body.constant(int(1063)), r0A1D), 0x01)); + + /* IF CONDITION */ + ir_expression *const r0A2C = less(body.constant(int(0)), r0A2A); + ir_if *f0A2B = new(mem_ctx) ir_if(operand(r0A2C).val); + exec_list *const f0A2B_parent_instructions = body.instructions; + + /* THEN INSTRUCTIONS */ + body.instructions = &f0A2B->then_instructions; + + ir_variable *const r0A2D = new(mem_ctx) ir_variable(glsl_type::uint_type, "z1", ir_var_auto); + body.emit(r0A2D); + ir_variable *const r0A2E = new(mem_ctx) ir_variable(glsl_type::uint_type, "z0", ir_var_auto); + body.emit(r0A2E); + ir_variable *const r0A2F = body.make_temp(glsl_type::int_type, "assignment_tmp"); + ir_expression *const r0A30 = neg(r0A2A); + body.emit(assign(r0A2F, bit_and(r0A30, body.constant(int(31))), 0x01)); + + /* IF CONDITION */ + ir_expression *const r0A32 = equal(r0A2A, body.constant(int(0))); + ir_if *f0A31 = new(mem_ctx) ir_if(operand(r0A32).val); + exec_list *const f0A31_parent_instructions = body.instructions; + + /* THEN INSTRUCTIONS */ + body.instructions = &f0A31->then_instructions; + + body.emit(assign(r0A2D, r0A1B, 0x01)); + + body.emit(assign(r0A2E, r0A19, 0x01)); + + + /* ELSE INSTRUCTIONS */ + body.instructions = &f0A31->else_instructions; + + /* IF CONDITION */ + ir_expression *const r0A34 = less(r0A2A, body.constant(int(32))); + ir_if *f0A33 = new(mem_ctx) ir_if(operand(r0A34).val); + exec_list *const f0A33_parent_instructions = body.instructions; + + /* THEN INSTRUCTIONS */ + body.instructions = &f0A33->then_instructions; + + ir_expression *const r0A35 = lshift(r0A19, r0A2F); + ir_expression *const r0A36 = rshift(swizzle_x(r0A17), r0A2A); + ir_expression *const r0A37 = bit_or(r0A35, r0A36); + ir_expression *const r0A38 = lshift(swizzle_x(r0A17), r0A2F); + ir_expression *const r0A39 = nequal(r0A38, body.constant(0u)); + ir_expression *const r0A3A = expr(ir_unop_b2i, r0A39); + ir_expression *const r0A3B = expr(ir_unop_i2u, r0A3A); + body.emit(assign(r0A2D, bit_or(r0A37, r0A3B), 0x01)); + + body.emit(assign(r0A2E, rshift(r0A19, r0A2A), 0x01)); + + + /* ELSE INSTRUCTIONS */ + body.instructions = &f0A33->else_instructions; + + /* IF CONDITION */ + ir_expression *const r0A3D = equal(r0A2A, body.constant(int(32))); + ir_if *f0A3C = new(mem_ctx) ir_if(operand(r0A3D).val); + exec_list *const f0A3C_parent_instructions = body.instructions; + + /* THEN INSTRUCTIONS */ + body.instructions = &f0A3C->then_instructions; + + ir_expression *const r0A3E = nequal(swizzle_x(r0A17), body.constant(0u)); + ir_expression *const r0A3F = expr(ir_unop_b2i, r0A3E); + ir_expression *const r0A40 = expr(ir_unop_i2u, r0A3F); + body.emit(assign(r0A2D, bit_or(r0A19, r0A40), 0x01)); + + + /* ELSE INSTRUCTIONS */ + body.instructions = &f0A3C->else_instructions; + + /* IF CONDITION */ + ir_expression *const r0A42 = less(r0A2A, body.constant(int(64))); + ir_if *f0A41 = new(mem_ctx) ir_if(operand(r0A42).val); + exec_list *const f0A41_parent_instructions = body.instructions; + + /* THEN INSTRUCTIONS */ + body.instructions = &f0A41->then_instructions; + + ir_expression *const r0A43 = bit_and(r0A2A, body.constant(int(31))); + ir_expression *const r0A44 = rshift(r0A19, r0A43); + ir_expression *const r0A45 = lshift(r0A19, r0A2F); + ir_expression *const r0A46 = bit_or(r0A45, swizzle_x(r0A17)); + ir_expression *const r0A47 = nequal(r0A46, body.constant(0u)); + ir_expression *const r0A48 = expr(ir_unop_b2i, r0A47); + ir_expression *const r0A49 = expr(ir_unop_i2u, r0A48); + body.emit(assign(r0A2D, bit_or(r0A44, r0A49), 0x01)); + + + /* ELSE INSTRUCTIONS */ + body.instructions = &f0A41->else_instructions; + + ir_expression *const r0A4A = bit_or(r0A19, swizzle_x(r0A17)); + ir_expression *const r0A4B = nequal(r0A4A, body.constant(0u)); + ir_expression *const r0A4C = expr(ir_unop_b2i, r0A4B); + body.emit(assign(r0A2D, expr(ir_unop_i2u, r0A4C), 0x01)); + + + body.instructions = f0A41_parent_instructions; + body.emit(f0A41); + + /* END IF */ + + + body.instructions = f0A3C_parent_instructions; + body.emit(f0A3C); + + /* END IF */ + + body.emit(assign(r0A2E, body.constant(0u), 0x01)); + + + body.instructions = f0A33_parent_instructions; + body.emit(f0A33); + + /* END IF */ + + + body.instructions = f0A31_parent_instructions; + body.emit(f0A31); + + /* END IF */ + + body.emit(assign(r0A19, r0A2E, 0x01)); + + body.emit(assign(r0A1A, r0A2D, 0x01)); + + + body.instructions = f0A2B_parent_instructions; + body.emit(f0A2B); + + /* END IF */ + + /* IF CONDITION */ + ir_expression *const r0A4E = bit_and(r0A19, body.constant(4294963200u)); + ir_expression *const r0A4F = nequal(r0A4E, body.constant(0u)); + ir_if *f0A4D = new(mem_ctx) ir_if(operand(r0A4F).val); + exec_list *const f0A4D_parent_instructions = body.instructions; + + /* THEN INSTRUCTIONS */ + body.instructions = &f0A4D->then_instructions; + + ir_variable *const r0A50 = body.make_temp(glsl_type::uint_type, "conditional_tmp"); + /* IF CONDITION */ + ir_expression *const r0A52 = nequal(r0A20, body.constant(0u)); + ir_if *f0A51 = new(mem_ctx) ir_if(operand(r0A52).val); + exec_list *const f0A51_parent_instructions = body.instructions; + + /* THEN INSTRUCTIONS */ + body.instructions = &f0A51->then_instructions; + + body.emit(assign(r0A50, body.constant(0u), 0x01)); + + + /* ELSE INSTRUCTIONS */ + body.instructions = &f0A51->else_instructions; + + body.emit(assign(r0A50, body.constant(4294967295u), 0x01)); + + + body.instructions = f0A51_parent_instructions; + body.emit(f0A51); + + /* END IF */ + + body.emit(assign(r0A18, r0A50, 0x01)); + + + /* ELSE INSTRUCTIONS */ + body.instructions = &f0A4D->else_instructions; + + ir_variable *const r0A53 = new(mem_ctx) ir_variable(glsl_type::uint_type, "z1", ir_var_auto); + body.emit(r0A53); + ir_variable *const r0A54 = new(mem_ctx) ir_variable(glsl_type::uint_type, "z0", ir_var_auto); + body.emit(r0A54); + ir_expression *const r0A55 = lshift(r0A19, body.constant(int(20))); + ir_expression *const r0A56 = rshift(r0A1A, body.constant(int(12))); + body.emit(assign(r0A53, bit_or(r0A55, r0A56), 0x01)); + + body.emit(assign(r0A54, rshift(r0A19, body.constant(int(12))), 0x01)); + + /* IF CONDITION */ + ir_expression *const r0A58 = nequal(r0A20, body.constant(0u)); + ir_expression *const r0A59 = nequal(r0A53, body.constant(0u)); + ir_expression *const r0A5A = logic_and(r0A58, r0A59); + ir_if *f0A57 = new(mem_ctx) ir_if(operand(r0A5A).val); + exec_list *const f0A57_parent_instructions = body.instructions; + + /* THEN INSTRUCTIONS */ + body.instructions = &f0A57->then_instructions; + + ir_variable *const r0A5B = body.make_temp(glsl_type::uint_type, "conditional_tmp"); + /* IF CONDITION */ + ir_expression *const r0A5D = nequal(r0A20, body.constant(0u)); + ir_if *f0A5C = new(mem_ctx) ir_if(operand(r0A5D).val); + exec_list *const f0A5C_parent_instructions = body.instructions; + + /* THEN INSTRUCTIONS */ + body.instructions = &f0A5C->then_instructions; + + body.emit(assign(r0A5B, body.constant(0u), 0x01)); + + + /* ELSE INSTRUCTIONS */ + body.instructions = &f0A5C->else_instructions; + + body.emit(assign(r0A5B, body.constant(4294967295u), 0x01)); + + + body.instructions = f0A5C_parent_instructions; + body.emit(f0A5C); + + /* END IF */ + + body.emit(assign(r0A18, r0A5B, 0x01)); + + + /* ELSE INSTRUCTIONS */ + body.instructions = &f0A57->else_instructions; + + body.emit(assign(r0A18, r0A53, 0x01)); + + + body.instructions = f0A57_parent_instructions; + body.emit(f0A57); + + /* END IF */ + + + body.instructions = f0A4D_parent_instructions; + body.emit(f0A4D); + + /* END IF */ + + + body.instructions = f0A23_parent_instructions; + body.emit(f0A23); + + /* END IF */ + + + body.instructions = f0A21_parent_instructions; + body.emit(f0A21); + + /* END IF */ + + body.emit(ret(r0A18)); + + sig->replace_parameters(&sig_parameters); + return sig; +} diff --git a/src/compiler/glsl/builtin_functions.cpp b/src/compiler/glsl/builtin_functions.cpp index 448a24a2c7..1706aa25e8 100644 --- a/src/compiler/glsl/builtin_functions.cpp +++ b/src/compiler/glsl/builtin_functions.cpp @@ -3356,6 +3356,10 @@ builtin_builder::create_builtins() generate_ir::fmul64(mem_ctx, integer_functions_supported), NULL); + add_function("__builtin_fp64_to_uint", + generate_ir::fp64_to_uint(mem_ctx, integer_functions_supported), + NULL); + #undef F #undef FI #undef FIUD_VEC diff --git a/src/compiler/glsl/builtin_functions.h b/src/compiler/glsl/builtin_functions.h index 88512930e3..9d9a78624c 100644 --- a/src/compiler/glsl/builtin_functions.h +++ b/src/compiler/glsl/builtin_functions.h @@ -91,6 +91,9 @@ fadd64(void *mem_ctx, builtin_available_predicate avail); ir_function_signature * fmul64(void *mem_ctx, builtin_available_predicate avail); +ir_function_signature * +fp64_to_uint(void *mem_ctx, builtin_available_predicate avail); + } #endif /* BULITIN_FUNCTIONS_H */ diff --git a/src/compiler/glsl/float64.glsl b/src/compiler/glsl/float64.glsl index 6c3a2bcb88..ce0e6e52fe 100644 --- a/src/compiler/glsl/float64.glsl +++ b/src/compiler/glsl/float64.glsl @@ -837,3 +837,71 @@ fmul64(uvec2 a, uvec2 b) } return roundAndPackFloat64(zSign, zExp, zFrac0, zFrac1, zFrac2); } + +/* Shifts the 64-bit value formed by concatenating `a0' and `a1' right by the + * number of bits given in `count'. Any bits shifted off are lost. The value + * of `count' can be arbitrarily large; in particular, if `count' is greater + * than 64, the result will be 0. The result is broken into two 32-bit pieces + * which are stored at the locations pointed to by `z0Ptr' and `z1Ptr'. + */ +void +shift64Right(uint a0, uint a1, + int count, + inout uint z0Ptr, + inout uint z1Ptr) +{ + uint z0; + uint z1; + int negCount = (-count) & 31; + + if (count == 0) { + z1 = a1; + z0 = a0; + } else if (count < 32) { + z1 = (a0<<negCount) | (a1>>count); + z0 = a0>>count; + } else { + z1 = (count < 64) ? (a0>>(count & 31)) : 0u; + z0 = 0u; + } + z1Ptr = z1; + z0Ptr = z0; +} + +/* Returns the result of converting the double-precision floating-point value + * `a' to the unsigned integer format. The conversion is performed according + * to the IEEE Standard for Floating-Point Arithmetic. + */ +uint +fp64_to_uint(uvec2 a) +{ + uint aFracLo = extractFloat64FracLo(a); + uint aFracHi = extractFloat64FracHi(a); + int aExp = extractFloat64Exp(a); + uint aSign = extractFloat64Sign(a); + + if (aSign != 0u) + return 0u; + + if ((aExp == 0x7FF) && ((aFracHi | aFracLo) != 0u)) + return 0xFFFFFFFFu; + + if (aExp != 0) + aFracHi |= 0x00100000u; + + int shiftDist = 0x427 - aExp; + if (0 < shiftDist) + shift64RightJamming(aFracHi, aFracLo, shiftDist, aFracHi, aFracLo); + + if ((aFracHi & 0xFFFFF000u) != 0u) + return (aSign != 0u) ? 0u : ~0u; + + uint z = 0u; + uint zero = 0u; + shift64Right(aFracHi, aFracLo, 12, zero, z); + + if ((aSign != 0u) && (z != 0u)) + return (aSign != 0u) ? 0u : ~0u; + + return z; +} diff --git a/src/compiler/glsl/glcpp/glcpp-parse.y b/src/compiler/glsl/glcpp/glcpp-parse.y index 6496a2e505..b8dcd04637 100644 --- a/src/compiler/glsl/glcpp/glcpp-parse.y +++ b/src/compiler/glsl/glcpp/glcpp-parse.y @@ -2457,6 +2457,7 @@ _glcpp_parser_handle_version_declaration(glcpp_parser_t *parser, intmax_t versio add_builtin_define(parser, "__have_builtin_builtin_flt64", 1); add_builtin_define(parser, "__have_builtin_builtin_fadd64", 1); add_builtin_define(parser, "__have_builtin_builtin_fmul64", 1); + add_builtin_define(parser, "__have_builtin_builtin_fp64_to_uint", 1); } } -- 2.14.1 _______________________________________________ mesa-dev mailing list mesa-dev@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/mesa-dev