suyu/src/video_core/shader/decode/conversion.cpp
ReinUsesLisp da706cad25 shader/conversion: Implement I2I sign extension, saturation and selection
Reimplements I2I adding sign extension, saturation (clamp source value
to the destination), selection and destination sizes that are not 32
bits wide.

It doesn't implement CC yet.
2020-04-07 02:19:44 -03:00

321 lines
13 KiB
C++

// Copyright 2018 yuzu Emulator Project
// Licensed under GPLv2 or any later version
// Refer to the license.txt file included.
#include <limits>
#include <optional>
#include <utility>
#include "common/assert.h"
#include "common/common_types.h"
#include "video_core/engines/shader_bytecode.h"
#include "video_core/shader/node_helper.h"
#include "video_core/shader/shader_ir.h"
namespace VideoCommon::Shader {
using Tegra::Shader::Instruction;
using Tegra::Shader::OpCode;
using Tegra::Shader::Register;
namespace {
constexpr OperationCode GetFloatSelector(u64 selector) {
return selector == 0 ? OperationCode::FCastHalf0 : OperationCode::FCastHalf1;
}
constexpr u32 SizeInBits(Register::Size size) {
switch (size) {
case Register::Size::Byte:
return 8;
case Register::Size::Short:
return 16;
case Register::Size::Word:
return 32;
case Register::Size::Long:
return 64;
}
return 0;
}
constexpr std::optional<std::pair<s32, s32>> IntegerSaturateBounds(Register::Size src_size,
Register::Size dst_size,
bool src_signed,
bool dst_signed) {
const u32 dst_bits = SizeInBits(dst_size);
if (src_size == Register::Size::Word && dst_size == Register::Size::Word) {
if (src_signed == dst_signed) {
return std::nullopt;
}
return std::make_pair(0, std::numeric_limits<s32>::max());
}
if (dst_signed) {
// Signed destination, clamp to [-128, 127] for instance
return std::make_pair(-(1 << (dst_bits - 1)), (1 << (dst_bits - 1)) - 1);
} else {
// Unsigned destination
if (dst_bits == 32) {
// Avoid shifting by 32, that is undefined behavior
return std::make_pair(0, s32(std::numeric_limits<u32>::max()));
}
return std::make_pair(0, (1 << dst_bits) - 1);
}
}
} // Anonymous namespace
u32 ShaderIR::DecodeConversion(NodeBlock& bb, u32 pc) {
const Instruction instr = {program_code[pc]};
const auto opcode = OpCode::Decode(instr);
switch (opcode->get().GetId()) {
case OpCode::Id::I2I_R:
case OpCode::Id::I2I_C:
case OpCode::Id::I2I_IMM: {
const bool src_signed = instr.conversion.is_input_signed;
const bool dst_signed = instr.conversion.is_output_signed;
const Register::Size src_size = instr.conversion.src_size;
const Register::Size dst_size = instr.conversion.dst_size;
const u32 selector = static_cast<u32>(instr.conversion.int_src.selector);
Node value = [this, instr, opcode] {
switch (opcode->get().GetId()) {
case OpCode::Id::I2I_R:
return GetRegister(instr.gpr20);
case OpCode::Id::I2I_C:
return GetConstBuffer(instr.cbuf34.index, instr.cbuf34.GetOffset());
case OpCode::Id::I2I_IMM:
return Immediate(instr.alu.GetSignedImm20_20());
default:
UNREACHABLE();
return Immediate(0);
}
}();
// Ensure the source selector is valid
switch (instr.conversion.src_size) {
case Register::Size::Byte:
break;
case Register::Size::Short:
ASSERT(selector == 0 || selector == 2);
break;
default:
ASSERT(selector == 0);
break;
}
if (src_size != Register::Size::Word || selector != 0) {
value = SignedOperation(OperationCode::IBitfieldExtract, src_signed, std::move(value),
Immediate(selector * 8), Immediate(SizeInBits(src_size)));
}
value = GetOperandAbsNegInteger(std::move(value), instr.conversion.abs_a,
instr.conversion.negate_a, src_signed);
if (instr.alu.saturate_d) {
if (src_signed && !dst_signed) {
Node is_negative = Operation(OperationCode::LogicalUGreaterEqual, value,
Immediate(1 << (SizeInBits(src_size) - 1)));
value = Operation(OperationCode::Select, std::move(is_negative), Immediate(0),
std::move(value));
// Simplify generated expressions, this can be removed without semantic impact
SetTemporary(bb, 0, std::move(value));
value = GetTemporary(0);
if (dst_size != Register::Size::Word) {
const Node limit = Immediate((1 << SizeInBits(dst_size)) - 1);
Node is_large =
Operation(OperationCode::LogicalUGreaterThan, std::move(value), limit);
value = Operation(OperationCode::Select, std::move(is_large), limit,
std::move(value));
}
} else if (const std::optional bounds =
IntegerSaturateBounds(src_size, dst_size, src_signed, dst_signed)) {
value = SignedOperation(OperationCode::IMax, src_signed, std::move(value),
Immediate(bounds->first));
value = SignedOperation(OperationCode::IMin, src_signed, std::move(value),
Immediate(bounds->second));
}
} else if (dst_size != Register::Size::Word) {
// No saturation, we only have to mask the result
Node mask = Immediate((1 << SizeInBits(dst_size)) - 1);
value = Operation(OperationCode::UBitwiseAnd, std::move(value), std::move(mask));
}
SetInternalFlagsFromInteger(bb, value, instr.generates_cc);
SetRegister(bb, instr.gpr0, std::move(value));
break;
}
case OpCode::Id::I2F_R:
case OpCode::Id::I2F_C:
case OpCode::Id::I2F_IMM: {
UNIMPLEMENTED_IF(instr.conversion.dst_size == Register::Size::Long);
UNIMPLEMENTED_IF_MSG(instr.generates_cc,
"Condition codes generation in I2F is not implemented");
Node value = [&] {
switch (opcode->get().GetId()) {
case OpCode::Id::I2F_R:
return GetRegister(instr.gpr20);
case OpCode::Id::I2F_C:
return GetConstBuffer(instr.cbuf34.index, instr.cbuf34.GetOffset());
case OpCode::Id::I2F_IMM:
return Immediate(instr.alu.GetSignedImm20_20());
default:
UNREACHABLE();
return Immediate(0);
}
}();
const bool input_signed = instr.conversion.is_input_signed;
if (const u32 offset = static_cast<u32>(instr.conversion.int_src.selector); offset > 0) {
ASSERT(instr.conversion.src_size == Register::Size::Byte ||
instr.conversion.src_size == Register::Size::Short);
if (instr.conversion.src_size == Register::Size::Short) {
ASSERT(offset == 0 || offset == 2);
}
value = SignedOperation(OperationCode::ILogicalShiftRight, input_signed,
std::move(value), Immediate(offset * 8));
}
value = ConvertIntegerSize(value, instr.conversion.src_size, input_signed);
value = GetOperandAbsNegInteger(value, instr.conversion.abs_a, false, input_signed);
value = SignedOperation(OperationCode::FCastInteger, input_signed, PRECISE, value);
value = GetOperandAbsNegFloat(value, false, instr.conversion.negate_a);
SetInternalFlagsFromFloat(bb, value, instr.generates_cc);
if (instr.conversion.dst_size == Register::Size::Short) {
value = Operation(OperationCode::HCastFloat, PRECISE, value);
}
SetRegister(bb, instr.gpr0, value);
break;
}
case OpCode::Id::F2F_R:
case OpCode::Id::F2F_C:
case OpCode::Id::F2F_IMM: {
UNIMPLEMENTED_IF(instr.conversion.dst_size == Register::Size::Long);
UNIMPLEMENTED_IF(instr.conversion.src_size == Register::Size::Long);
UNIMPLEMENTED_IF_MSG(instr.generates_cc,
"Condition codes generation in F2F is not implemented");
Node value = [&]() {
switch (opcode->get().GetId()) {
case OpCode::Id::F2F_R:
return GetRegister(instr.gpr20);
case OpCode::Id::F2F_C:
return GetConstBuffer(instr.cbuf34.index, instr.cbuf34.GetOffset());
case OpCode::Id::F2F_IMM:
return GetImmediate19(instr);
default:
UNREACHABLE();
return Immediate(0);
}
}();
if (instr.conversion.src_size == Register::Size::Short) {
value = Operation(GetFloatSelector(instr.conversion.float_src.selector), NO_PRECISE,
std::move(value));
} else {
ASSERT(instr.conversion.float_src.selector == 0);
}
value = GetOperandAbsNegFloat(value, instr.conversion.abs_a, instr.conversion.negate_a);
value = [&] {
if (instr.conversion.src_size != instr.conversion.dst_size) {
// Rounding operations only matter when the source and destination conversion size
// is the same.
return value;
}
switch (instr.conversion.f2f.GetRoundingMode()) {
case Tegra::Shader::F2fRoundingOp::None:
return value;
case Tegra::Shader::F2fRoundingOp::Round:
return Operation(OperationCode::FRoundEven, value);
case Tegra::Shader::F2fRoundingOp::Floor:
return Operation(OperationCode::FFloor, value);
case Tegra::Shader::F2fRoundingOp::Ceil:
return Operation(OperationCode::FCeil, value);
case Tegra::Shader::F2fRoundingOp::Trunc:
return Operation(OperationCode::FTrunc, value);
default:
UNIMPLEMENTED_MSG("Unimplemented F2F rounding mode {}",
static_cast<u32>(instr.conversion.f2f.rounding.Value()));
return value;
}
}();
value = GetSaturatedFloat(value, instr.alu.saturate_d);
SetInternalFlagsFromFloat(bb, value, instr.generates_cc);
if (instr.conversion.dst_size == Register::Size::Short) {
value = Operation(OperationCode::HCastFloat, PRECISE, value);
}
SetRegister(bb, instr.gpr0, value);
break;
}
case OpCode::Id::F2I_R:
case OpCode::Id::F2I_C:
case OpCode::Id::F2I_IMM: {
UNIMPLEMENTED_IF(instr.conversion.src_size == Register::Size::Long);
UNIMPLEMENTED_IF_MSG(instr.generates_cc,
"Condition codes generation in F2I is not implemented");
Node value = [&]() {
switch (opcode->get().GetId()) {
case OpCode::Id::F2I_R:
return GetRegister(instr.gpr20);
case OpCode::Id::F2I_C:
return GetConstBuffer(instr.cbuf34.index, instr.cbuf34.GetOffset());
case OpCode::Id::F2I_IMM:
return GetImmediate19(instr);
default:
UNREACHABLE();
return Immediate(0);
}
}();
if (instr.conversion.src_size == Register::Size::Short) {
value = Operation(GetFloatSelector(instr.conversion.float_src.selector), NO_PRECISE,
std::move(value));
} else {
ASSERT(instr.conversion.float_src.selector == 0);
}
value = GetOperandAbsNegFloat(value, instr.conversion.abs_a, instr.conversion.negate_a);
value = [&]() {
switch (instr.conversion.f2i.rounding) {
case Tegra::Shader::F2iRoundingOp::RoundEven:
return Operation(OperationCode::FRoundEven, PRECISE, value);
case Tegra::Shader::F2iRoundingOp::Floor:
return Operation(OperationCode::FFloor, PRECISE, value);
case Tegra::Shader::F2iRoundingOp::Ceil:
return Operation(OperationCode::FCeil, PRECISE, value);
case Tegra::Shader::F2iRoundingOp::Trunc:
return Operation(OperationCode::FTrunc, PRECISE, value);
default:
UNIMPLEMENTED_MSG("Unimplemented F2I rounding mode {}",
static_cast<u32>(instr.conversion.f2i.rounding.Value()));
return Immediate(0);
}
}();
const bool is_signed = instr.conversion.is_output_signed;
value = SignedOperation(OperationCode::ICastFloat, is_signed, PRECISE, value);
value = ConvertIntegerSize(value, instr.conversion.dst_size, is_signed);
SetRegister(bb, instr.gpr0, value);
break;
}
default:
UNIMPLEMENTED_MSG("Unhandled conversion instruction: {}", opcode->get().GetName());
}
return pc;
}
} // namespace VideoCommon::Shader