LLVM目标无关代码生成器The LLVM Target-Independent Code Generator

LLVM目标无关代码生成器The LLVM Target-Independent Code Generator

Warning

This is a work in progress.

LLVM目标无关代码生成器是一个框架，它提供了一套可重用组件，用于将LLVM内部表示转换为指定目标的机器代码，无论是汇编形式（适用于静态编译器）还是二进制机器代码格式（适用于JIT编译器）。LLVM目标独立代码生成器由六个主要组件组成：

1.抽象的目标描述接口，其捕获关于机器的各个方面的重要属性，而与它们将如何使用无关。这些接口在include/llvm/Target/中定义。

2.用于表示为目标生成的代码的类。这些类旨在足够抽象，以表示任何目标机器的机器代码。这些类在include/llvm/CodeGen/中定义。在这个级别上，“常量池条目”和“跳转表”等概念被显式公开。

3.用于表示对象文件级别（MC层）代码的类和算法。这些类表示组件级构造，如标签、节和指令。在这个级别上，“常量池条目”和“跳转表”等概念并不存在。

4.用于实现本地代码生成的各个阶段（寄存器分配、调度、堆栈帧表示等）的目标无关算法。此代码位于lib/CodeGen/中。

5.特定目标的抽象目标描述接口的实现。这些机器描述使用LLVM提供的组件，并可以选择提供定制的特定于目标的传递，为特定目标构建完整的代码生成器。目标描述位于lib/Target/中。

6.目标独立的JIT组件。LLVM JIT完全独立于目标（它使用TargetJITInfo结构来处理特定于目标的问题。独立于目标的JIT的代码位于lib/ExecutionEngine/JIT中。

根据您对代码生成器的哪一部分感兴趣，其中的不同部分将对您有用。在任何情况下，您都应该熟悉目标描述和机器代码表示类。如果要为新目标添加后端，则需要为新目标实现目标描述类，并了解LLVM代码表示。如果您对实现新的代码生成算法感兴趣，那么它应该只依赖于目标描述和机器代码表示类，确保它是可移植的。

The LLVM target-independent code generator is a framework that provides a suite of reusable components for translating the LLVM internal representation to the machine code for a specified target—either in assembly form (suitable for a static compiler) or in binary machine code format (usable for a JIT compiler). The LLVM target-independent code generator consists of six main components:

  1. Abstract target description interfaces which capture important properties about various aspects of the machine, independently of how they will be used. These interfaces are defined in include/llvm/Target/.
  2. Classes used to represent the code being generated for a target. These classes are intended to be abstract enough to represent the machine code for any target machine. These classes are defined in include/llvm/CodeGen/. At this level, concepts like “constant pool entries” and “jump tables” are explicitly exposed.
  3. Classes and algorithms used to represent code at the object file level, the MC Layer. These classes represent assembly level constructs like labels, sections, and instructions. At this level, concepts like “constant pool entries” and “jump tables” don’t exist.
  4. Target-independent algorithms used to implement various phases of native code generation (register allocation, scheduling, stack frame representation, etc). This code lives in lib/CodeGen/.
  5. Implementations of the abstract target description interfaces for particular targets. These machine descriptions make use of the components provided by LLVM, and can optionally provide custom target-specific passes, to build complete code generators for a specific target. Target descriptions live in lib/Target/.
  6. The target-independent JIT components. The LLVM JIT is completely target independent (it uses the TargetJITInfo structure to interface for target-specific issues. The code for the target-independent JIT lives in lib/ExecutionEngine/JIT.

Depending on which part of the code generator you are interested in working on, different pieces of this will be useful to you. In any case, you should be familiar with the target description and machine code representation classes. If you want to add a backend for a new target, you will need to implement the target description classes for your new target and understand the LLVM code representation. If you are interested in implementing a new code generation algorithm, it should only depend on the target-description and machine code representation classes, ensuring that it is portable.

LLVM代码生成器的两部分是代码生成器的高级接口和可用于构建目标特定后端的一组可重用组件。两个最重要的接口（TargetMachine和DataLayout）是唯一需要为后端定义以适应LLVM系统的接口，但如果要使用可重用代码生成器组件，则必须定义其他接口。

这种设计有两个重要的含义。首先，LLVM可以支持完全非传统的代码生成目标。例如，C后端不需要寄存器分配、指令选择或系统提供的任何其他标准组件。因此，它只实现这两个接口，并做自己的事情。注意，自LLVM 3.1发布以来，C后端已从主干中删除。类似这样的代码生成器的另一个示例是一个（纯假设的）后端，它将LLVM转换为GCC RTL形式，并使用GCC为目标发出机器代码。

这种设计还意味着可以在LLVM系统中设计和实现完全不同的代码生成器，而不使用任何内置组件。根本不建议这样做，但对于根本不符合LLVM机器描述模型（例如FPGA）的不同目标，可能需要这样做。

The two pieces of the LLVM code generator are the high-level interface to the code generator and the set of reusable components that can be used to build target-specific backends. The two most important interfaces ( TargetMachine and DataLayout ) are the only ones that are required to be defined for a backend to fit into the LLVM system, but the others must be defined if the reusable code generator components are going to be used.

This design has two important implications. The first is that LLVM can support completely non-traditional code generation targets. For example, the C backend does not require register allocation, instruction selection, or any of the other standard components provided by the system. As such, it only implements these two interfaces, and does its own thing. Note that C backend was removed from the trunk since LLVM 3.1 release. Another example of a code generator like this is a (purely hypothetical) backend that converts LLVM to the GCC RTL form and uses GCC to emit machine code for a target.

This design also implies that it is possible to design and implement radically different code generators in the LLVM system that do not make use of any of the built-in components. Doing so is not recommended at all, but could be required for radically different targets that do not fit into the LLVM machine description model: FPGAs for example.

LLVM目标独立代码生成器旨在支持基于寄存器的标准微处理器的高效和高质量代码生成。此模型中的代码生成分为以下阶段：

1.指令选择—此阶段确定在目标指令集中表达输入LLVM代码的有效方式。此阶段为目标指令集中的程序生成初始代码，然后使用SSA形式的虚拟寄存器和物理寄存器，这些寄存器表示由于目标约束或调用约定而需要的任何寄存器分配。此步骤将LLVM代码转换为目标指令的DAG。

2.调度和形成-此阶段获取指令选择阶段生成的目标指令的DAG，确定指令的顺序，然后将指令作为具有该顺序的MachineInstrs发出。注意，我们在指令选择部分对此进行了描述，因为它在SelectionDAG上运行。

3.基于SSA的机器代码优化-此可选阶段由一系列机器代码优化组成，这些优化对指令选择器生成的SSA形式进行操作。像模调度或窥视孔优化这样的优化在这里起作用。

4.寄存器分配-目标代码从SSA形式的无限虚拟寄存器文件转换为目标使用的具体寄存器文件。此阶段引入溢出代码并从程序中删除所有虚拟寄存器引用。

5.Prolog/Epilog代码插入-一旦为函数生成了机器代码，并且所需的堆栈空间量已知（用于LLVM分配和溢出槽），就可以插入函数的Prolog和Epilog代码，并消除“抽象堆栈位置引用”。此阶段负责实现诸如帧指针消除和堆栈打包之类的优化。

6.后期机器代码优化-对“最终”机器代码进行的优化可以在这里进行，例如溢出代码调度和窥视孔优化。

7.代码发射-最后阶段实际上以目标汇编程序格式或机器代码为当前函数输出代码。

代码生成器基于以下假设：指令选择器将使用最佳模式匹配选择器来创建高质量的本地指令序列。基于模式扩展和积极迭代窥视孔优化的替代代码生成器设计要慢得多。这种设计允许在编译的任何步骤中使用不同复杂度的组件，从而允许高效编译（对于JIT环境很重要）和积极优化（在离线生成代码时使用）。

除了这些阶段之外，目标实现还可以向流中插入任意特定于目标的传递。例如，X86目标使用特殊通道来处理80x87浮点堆栈架构。其他具有特殊要求的目标可以根据需要通过自定义传递来支持。

The LLVM target-independent code generator is designed to support efficient and quality code generation for standard register-based microprocessors. Code generation in this model is divided into the following stages:

  1. Instruction Selection — This phase determines an efficient way to express the input LLVM code in the target instruction set. This stage produces the initial code for the program in the target instruction set, then makes use of virtual registers in SSA form and physical registers that represent any required register assignments due to target constraints or calling conventions. This step turns the LLVM code into a DAG of target instructions.
  2. Scheduling and Formation — This phase takes the DAG of target instructions produced by the instruction selection phase, determines an ordering of the instructions, then emits the instructions as MachineInstrs with that ordering. Note that we describe this in the instruction selection section because it operates on a SelectionDAG.
  3. SSA-based Machine Code Optimizations — This optional stage consists of a series of machine-code optimizations that operate on the SSA-form produced by the instruction selector. Optimizations like modulo-scheduling or peephole optimization work here.
  4. Register Allocation — The target code is transformed from an infinite virtual register file in SSA form to the concrete register file used by the target. This phase introduces spill code and eliminates all virtual register references from the program.
  5. Prolog/Epilog Code Insertion — Once the machine code has been generated for the function and the amount of stack space required is known (used for LLVM alloca’s and spill slots), the prolog and epilog code for the function can be inserted and “abstract stack location references” can be eliminated. This stage is responsible for implementing optimizations like frame-pointer elimination and stack packing.
  6. Late Machine Code Optimizations — Optimizations that operate on “final” machine code can go here, such as spill code scheduling and peephole optimizations.
  7. Code Emission — The final stage actually puts out the code for the current function, either in the target assembler format or in machine code.

The code generator is based on the assumption that the instruction selector will use an optimal pattern matching selector to create high-quality sequences of native instructions. Alternative code generator designs based on pattern expansion and aggressive iterative peephole optimization are much slower. This design permits efficient compilation (important for JIT environments) and aggressive optimization (used when generating code offline) by allowing components of varying levels of sophistication to be used for any step of compilation.

In addition to these stages, target implementations can insert arbitrary target-specific passes into the flow. For example, the X86 target uses a special pass to handle the 80x87 floating point stack architecture. Other targets with unusual requirements can be supported with custom passes as needed.

目标描述类需要目标体系结构的详细描述。这些目标描述通常具有大量公共信息（例如，加法指令与子指令几乎相同）。为了最大限度地减少通用性，LLVM代码生成器使用TableGen Overview工具来描述目标机器的大块，这允许使用特定于域和特定于目标的抽象来减少重复量。

随着LLVM的不断开发和完善，我们计划将越来越多的目标描述转移到.td格式。这样做会给我们带来许多好处。最重要的是，它使LLVM更容易移植，因为它减少了必须编写的C++代码的数量，并且减少了代码生成器的表面区域，在有人能够工作之前需要了解这些区域。第二，它使改变事情变得更容易。特别是，如果表和其他东西都是由tblgen发出的，我们只需要在一个地方（tblgen）进行更改，就可以将所有目标更新为新接口。

The target description classes require a detailed description of the target architecture. These target descriptions often have a large amount of common information (e.g., an add instruction is almost identical to a sub instruction). In order to allow the maximum amount of commonality to be factored out, the LLVM code generator uses the TableGen Overview tool to describe big chunks of the target machine, which allows the use of domain-specific and target-specific abstractions to reduce the amount of repetition.

As LLVM continues to be developed and refined, we plan to move more and more of the target description to the .td form. Doing so gives us a number of advantages. The most important is that it makes it easier to port LLVM because it reduces the amount of C++ code that has to be written, and the surface area of the code generator that needs to be understood before someone can get something working. Second, it makes it easier to change things. In particular, if tables and other things are all emitted by tblgen, we only need a change in one place (tblgen) to update all of the targets to a new interface.

LLVM目标描述类（位于include/LLVM/target目录中）提供独立于任何特定客户端的目标机器的抽象描述。这些类被设计为捕获目标的抽象属性（例如它所具有的指令和寄存器），并且不包含任何特定的代码生成算法。

所有的目标描述类（DataLayout类除外）都被设计为由具体的目标实现进行子类化，并实现了虚拟方法。为了实现这些实现，TargetMachine类提供了应该由目标实现的访问器。

The LLVM target description classes (located in the include/llvm/Target directory) provide an abstract description of the target machine independent of any particular client. These classes are designed to capture the abstract properties of the target (such as the instructions and registers it has), and do not incorporate any particular pieces of code generation algorithms.

All of the target description classes (except the DataLayout class) are designed to be subclassed by the concrete target implementation, and have virtual methods implemented. To get to these implementations, the TargetMachine class provides accessors that should be implemented by the target.

LLVM目标描述类（位于include/LLVM/target目录中）提供独立于任何特定客户端的目标机器的抽象描述。这些类被设计为捕获目标的抽象属性（例如它所具有的指令和寄存器），并且不包含任何特定的代码生成算法。

所有的目标描述类（DataLayout类除外）都被设计为由具体的目标实现进行子类化，并实现了虚拟方法。为了实现这些实现，TargetMachine类提供了应该由目标实现的访问器。

The TargetMachine class provides virtual methods that are used to access the target-specific implementations of the various target description classes via the get*Info methods (getInstrInfo, getRegisterInfo, getFrameInfo, etc.). This class is designed to be specialized by a concrete target implementation (e.g., X86TargetMachine) which implements the various virtual methods. The only required target description class is the DataLayout class, but if the code generator components are to be used, the other interfaces should be implemented as well.

DataLayout类是唯一必需的目标描述类，也是唯一不可扩展的类（不能从中派生新类）。DataLayout指定了有关目标如何为结构布局内存、各种数据类型的对齐要求、目标中指针的大小以及目标是小端还是大端的信息。

The DataLayout class is the only required target description class, and it is the only class that is not extensible (you cannot derive a new class from it). DataLayout specifies information about how the target lays out memory for structures, the alignment requirements for various data types, the size of pointers in the target, and whether the target is little-endian or big-endian.

TargetLowering类由基于SelectionDAG的指令选择器使用，主要用于描述LLVM代码应如何降低到SelectionDAG操作。除其他外，该类表示：

•用于各种ValueType的初始寄存器类，

•目标机器本机支持哪些操作，

•setcc操作的返回类型，

•用于班次数量的类型，以及

•各种高级特征，如将常数除法转换为乘法序列是否有利可图。

The TargetLowering class is used by SelectionDAG based instruction selectors primarily to describe how LLVM code should be lowered to SelectionDAG operations. Among other things, this class indicates:

  • an initial register class to use for various ValueTypes,
  • which operations are natively supported by the target machine,
  • the return type of setcc operations,
  • the type to use for shift amounts, and
  • various high-level characteristics, like whether it is profitable to turn division by a constant into a multiplication sequence.

TargetRegisterInfo类用于描述目标的寄存器文件以及寄存器之间的任何交互。

寄存器在代码生成器中由无符号整数表示。物理寄存器（目标描述中实际存在的寄存器）是唯一的小数字，而虚拟寄存器通常很大。注意，寄存器#0被保留为标志值。

处理器描述中的每个寄存器都有一个关联的TargetRegisterDesc条目，该条目提供了寄存器的文本名称（用于程序集输出和调试转储）和一组别名（用于指示一个寄存器是否与另一个寄存器重叠）。

除了每个寄存器的描述之外，TargetRegisterInfo类还公开一组特定于处理器的寄存器类（TargetRegisterClass类的实例）。每个寄存器类包含具有相同属性的寄存器集（例如，它们都是32位整数寄存器）。指令选择器创建的每个SSA虚拟寄存器都有一个相关的寄存器类。当寄存器分配器运行时，它用集合中的物理寄存器替换虚拟寄存器。

这些类的目标特定实现是根据寄存器文件的TableGen概述描述自动生成的。The TargetRegisterInfo class is used to describe the register file of the target and any interactions between the registers.

Registers are represented in the code generator by unsigned integers. Physical registers (those that actually exist in the target description) are unique small numbers, and virtual registers are generally large. Note that register #0 is reserved as a flag value.

Each register in the processor description has an associated TargetRegisterDesc entry, which provides a textual name for the register (used for assembly output and debugging dumps) and a set of aliases (used to indicate whether one register overlaps with another).

In addition to the per-register description, the TargetRegisterInfo class exposes a set of processor specific register classes (instances of the TargetRegisterClass class). Each register class contains sets of registers that have the same properties (for example, they are all 32-bit integer registers). Each SSA virtual register created by the instruction selector has an associated register class. When the register allocator runs, it replaces virtual registers with a physical register in the set.

The target-specific implementations of these classes is auto-generated from a TableGen Overview description of the register file.

TargetInstrInfo类用于描述目标支持的机器指令。描述定义了操作码的助记符、操作数、隐式寄存器使用和def的列表、指令是否具有特定的目标无关属性（访问内存、可交换等），以及保存任何特定于目标的标志。

The TargetInstrInfo class is used to describe the machine instructions supported by the target. Descriptions define things like the mnemonic for the opcode, the number of operands, the list of implicit register uses and defs, whether the instruction has certain target-independent properties (accesses memory, is commutable, etc), and holds any target-specific flags.

TargetFrameLowing类用于提供有关目标的堆栈帧布局的信息。它保存堆栈增长的方向、每个函数进入时的已知堆栈对齐以及到局部区域的偏移。到局部区域的偏移量是从函数入口上的堆栈指针到可以存储函数数据（局部变量、溢出位置）的第一个位置的偏移量

The TargetFrameLowering class is used to provide information about the stack frame layout of the target. It holds the direction of stack growth, the known stack alignment on entry to each function, and the offset to the local area. The offset to the local area is the offset from the stack pointer on function entry to the first location where function data (local variables, spill locations) can be stored.

TargetSubtarget类用于提供有关所针对的特定芯片集的信息。子目标通知代码生成支持哪些指令、指令延迟和指令执行行程；例如，使用哪些处理单元、顺序和时间。

The TargetSubtarget class is used to provide information about the specific chip set being targeted. A sub-target informs code generation of which instructions are supported, instruction latencies and instruction execution itinerary; i.e., which processing units are used, in what order, and for how long.

TargetJITInfo类公开了实时代码生成器用于执行特定于目标的活动（如发送存根）的抽象接口。如果TargetMachine支持JIT代码生成，它应该通过getJITInfo方法提供其中一个对象。

The TargetJITInfo class exposes an abstract interface used by the Just-In-Time code generator to perform target-specific activities, such as emitting stubs. If a TargetMachine supports JIT code generation, it should provide one of these objects through the getJITInfo method.

在高层，LLVM代码被转换为由MachineFunction、MachineBasicBlock和MachineInstr实例（在include/LLVM/CodeGen中定义）组成的机器特定表示。这种表示完全不区分目标，以最抽象的形式表示指令：一个操作码和一系列操作数。此表示法旨在支持机器代码的SSA表示法，以及寄存器分配的非SSA形式。

At the high-level, LLVM code is translated to a machine specific representation formed out of MachineFunction , MachineBasicBlock , and MachineInstr instances (defined in include/llvm/CodeGen). This representation is completely target agnostic, representing instructions in their most abstract form: an opcode and a series of operands. This representation is designed to support both an SSA representation for machine code, as well as a register allocated, non-SSA form.

目标机器指令表示为MachineInstr类的实例。此类是表示机器指令的一种极其抽象的方式。特别是，它只跟踪一个操作码编号和一组操作数。

操作码编号是一个简单的无符号整数，只对特定后端有意义。目标的所有指令都应在目标的*InstrInfo.td文件中定义。操作码枚举值是根据此描述自动生成的。MachineInstr类没有关于如何解释指令的任何信息（即，指令的语义是什么）；为此，必须引用TargetInstrInfo类。

机器指令的操作数可以有几种不同的类型：寄存器引用、常量整数、基本块引用等。此外，机器操作数应标记为def或值的使用（尽管只允许寄存器为def）。

按照惯例，LLVM代码生成器对指令操作数进行排序，以便所有寄存器定义都先于寄存器使用，即使在通常以其他顺序打印的架构上也是如此。例如，SPARC add指令：“add%i1，%i2，%i3”将“%i1”和“%i2”寄存器相加，并将结果存储到“%i3“寄存器中。在LLVM代码生成器中，操作数应存储为“%i3，%i1，%i2”：目标优先。

将目标（定义）操作数保留在操作数列表的开头有几个优点。特别是，调试打印机将按如下方式打印指令：

Target machine instructions are represented as instances of the MachineInstr class. This class is an extremely abstract way of representing machine instructions. In particular, it only keeps track of an opcode number and a set of operands.

The opcode number is a simple unsigned integer that only has meaning to a specific backend. All of the instructions for a target should be defined in the *InstrInfo.td file for the target. The opcode enum values are auto-generated from this description. The MachineInstr class does not have any information about how to interpret the instruction (i.e., what the semantics of the instruction are); for that you must refer to the TargetInstrInfo class.

The operands of a machine instruction can be of several different types: a register reference, a constant integer, a basic block reference, etc. In addition, a machine operand should be marked as a def or a use of the value (though only registers are allowed to be defs).

By convention, the LLVM code generator orders instruction operands so that all register definitions come before the register uses, even on architectures that are normally printed in other orders. For example, the SPARC add instruction: “add %i1, %i2, %i3” adds the “%i1”, and “%i2” registers and stores the result into the “%i3” register. In the LLVM code generator, the operands should be stored as “%i3, %i1, %i2”: with the destination first.

Keeping destination (definition) operands at the beginning of the operand list has several advantages. In particular, the debugging printer will print the instruction like this:

%r3 = add %i1, %i2

Also if the first operand is a def, it is easier to create instructions whose only def is the first operand.

机器指令是通过使用位于include/llvm/CodeGen/MachineInstrBuilder.h文件中的BuildMI函数创建的。BuildMI函数使构建任意机器指令变得容易。BuildMI函数的用法如下：

Machine instructions are created by using the BuildMI functions, located in the include/llvm/CodeGen/MachineInstrBuilder.h file. The BuildMI functions make it easy to build arbitrary machine instructions. Usage of the BuildMI functions look like this:

// Create a 'DestReg = mov 42' (rendered in X86 assembly as 'mov DestReg, 42')
// instruction and insert it at the end of the given MachineBasicBlock.
const TargetInstrInfo &TII = ...
MachineBasicBlock &MBB = ...
DebugLoc DL;
MachineInstr *MI = BuildMI(MBB, DL, TII.get(X86::MOV32ri), DestReg).addImm(42);
 
// Create the same instr, but insert it before a specified iterator point.
MachineBasicBlock::iterator MBBI = ...
BuildMI(MBB, MBBI, DL, TII.get(X86::MOV32ri), DestReg).addImm(42);
 
// Create a 'cmp Reg, 0' instruction, no destination reg.
MI = BuildMI(MBB, DL, TII.get(X86::CMP32ri8)).addReg(Reg).addImm(42);
 
// Create an 'sahf' instruction which takes no operands and stores nothing.
MI = BuildMI(MBB, DL, TII.get(X86::SAHF));
 
// Create a self looping branch instruction.
BuildMI(MBB, DL, TII.get(X86::JNE)).addMBB(&MBB);

如果需要添加定义操作数（可选目标寄存器除外），则必须将其显式标记为：

If you need to add a definition operand (other than the optional destination register), you must explicitly mark it as such:

MI.addReg(Reg, RegState::Define);

代码生成器需要注意的一个重要问题是固定寄存器的存在。特别是，在指令流中，寄存器分配器必须安排特定值在特定寄存器中的位置。这可能是由于指令集的限制（例如，X86只能与EAX/EDX寄存器进行32位除法），或外部因素（如调用约定）。在任何情况下，指令选择器都应该发出代码，在需要时将虚拟寄存器复制到物理寄存器中或从物理寄存器中复制出来。

例如，考虑这个简单的LLVM示例：

One important issue that the code generator needs to be aware of is the presence of fixed registers. In particular, there are often places in the instruction stream where the register allocator must arrange for a particular value to be in a particular register. This can occur due to limitations of the instruction set (e.g., the X86 can only do a 32-bit divide with the EAX/EDX registers), or external factors like calling conventions. In any case, the instruction selector should emit code that copies a virtual register into or out of a physical register when needed.

For example, consider this simple LLVM example:

define i32 @test(i32 %X, i32 %Y) {
  %Z = sdiv i32 %X, %Y
  ret i32 %Z
}

X86指令选择器可能会为div和ret生成以下机器代码：

The X86 instruction selector might produce this machine code for the div and ret:

;; Start of div
%EAX = mov %reg1024           ;; Copy X (in reg1024) into EAX
%reg1027 = sar %reg1024, 31
%EDX = mov %reg1027           ;; Sign extend X into EDX
idiv %reg1025                 ;; Divide by Y (in reg1025)
%reg1026 = mov %EAX           ;; Read the result (Z) out of EAX
 
;; Start of ret
%EAX = mov %reg1026           ;; 32-bit return value goes in EAX
ret

在代码生成结束时，寄存器分配器将合并寄存器并删除生成的标识移动，生成以下代码：

By the end of code generation, the register allocator would coalesce the registers and delete the resultant identity moves producing the following code:

;; X is in EAX, Y is in ECX
mov %EAX, %EDX
sar %EDX, 31
idiv %ECX
ret

这种方法非常通用（如果它可以处理X86体系结构，那么它可以处理任何事情！），并且允许在指令选择器中隔离关于指令流的所有目标特定知识。请注意，物理寄存器的生存期应该很短，以便生成好的代码，并且所有物理寄存器在进入和退出基本块时（在寄存器分配之前）都被认为是死的。因此，如果您需要一个值跨越基本块边界，它必须存在于虚拟寄存器中。

This approach is extremely general (if it can handle the X86 architecture, it can handle anything!) and allows all of the target specific knowledge about the instruction stream to be isolated in the instruction selector. Note that physical registers should have a short lifetime for good code generation, and all physical registers are assumed dead on entry to and exit from basic blocks (before register allocation). Thus, if you need a value to be live across basic block boundaries, it must live in a virtual register.

一些机器指令，如调用，会大量物理寄存器。可以使用MO_RegisterMask操作数，而不是为所有操作数添加<def，dead>操作数。寄存器掩码操作数保存一个保留寄存器的位掩码，其他所有内容都被视为被指令清除。

Some machine instructions, like calls, clobber a large number of physical registers. Rather than adding <def,dead> operands for all of them, it is possible to use an MO_RegisterMask operand instead. The register mask operand holds a bit mask of preserved registers, and everything else is considered to be clobbered by the instruction.

MachineInstr最初以SSA形式选择，并以SSA格式维护，直到发生寄存器分配。在大多数情况下，这很简单，因为LLVM已经是SSA形式；LLVM PHI节点变成机器代码PHI节点，虚拟寄存器只能有一个定义。

寄存器分配后，机器代码不再是SSA形式，因为代码中没有剩余的虚拟寄存器。

MachineInstr’s are initially selected in SSA-form, and are maintained in SSA-form until register allocation happens. For the most part, this is trivially simple since LLVM is already in SSA form; LLVM PHI nodes become machine code PHI nodes, and virtual registers are only allowed to have a single definition.

After register allocation, machine code is no longer in SSA-form because there are no virtual registers left in the code.

MachineBasicBlock类包含机器指令列表（MachineInstr实例）。它大致对应于输入到指令选择器的LLVM代码，但可以有一对多映射（即一个LLVM基本块可以映射到多个机器基本块）。MachineBasicBlock类有一个“getBasicBlock”方法，它返回LLVM基本块。

The MachineBasicBlock class contains a list of machine instructions ( MachineInstr instances). It roughly corresponds to the LLVM code input to the instruction selector, but there can be a one-to-many mapping (i.e. one LLVM basic block can map to multiple machine basic blocks). The MachineBasicBlock class has a “getBasicBlock” method, which returns the LLVM basic block that it comes from.

MachineFunction类包含机器基本块（MachineBasicBlock实例）的列表。它与输入到指令选择器的LLVM函数一一对应。除了基本块列表之外，MachineFunction还包含MachineConstantPool、MachineFrameInfo、MachineFunctionInfo和MachineRegisterInfo。有关详细信息，请参阅include/llvm/CodeGen/MachineFunction.h。

The MachineFunction class contains a list of machine basic blocks ( MachineBasicBlock instances). It corresponds one-to-one with the LLVM function input to the instruction selector. In addition to a list of basic blocks, the MachineFunction contains a MachineConstantPool, a MachineFrameInfo, a MachineFunctionInfo, and a MachineRegisterInfo. See include/llvm/CodeGen/MachineFunction.h for more information.

LLVM代码生成器可以将指令序列建模为MachineInstr包。MI包可以对包含任意数量的并行指令的VLIW组/包进行建模。它还可以用于对无法合法分离的指令（可能具有数据依赖性）的顺序列表进行建模（例如ARM Thumb2 It块）。

从概念上讲，MI束是一个MI，其中嵌套了许多其他MI：

LLVM code generator can model sequences of instructions as MachineInstr bundles. A MI bundle can model a VLIW group / pack which contains an arbitrary number of parallel instructions. It can also be used to model a sequential list of instructions (potentially with data dependencies) that cannot be legally separated (e.g. ARM Thumb2 IT blocks).

Conceptually a MI bundle is a MI with a number of other MIs nested within:

--------------
|   Bundle   | ---------
--------------          \
       |           ----------------
       |           |      MI      |
       |           ----------------
       |                   |
       |           ----------------
       |           |      MI      |
       |           ----------------
       |                   |
       |           ----------------
       |           |      MI      |
       |           ----------------
       |
--------------
|   Bundle   | --------
--------------         \
       |           ----------------
       |           |      MI      |
       |           ----------------
       |                   |
       |           ----------------
       |           |      MI      |
       |           ----------------
       |                   |
       |                  ...
       |
--------------
|   Bundle   | --------
--------------         \
       |
      ...

MI捆绑包支持不会更改MachineBasicBlock和MachineInstr的物理表示。所有MI（包括顶层和嵌套的MI）都存储为MI的顺序列表。“捆绑”的MI标有“InsideBundle”标志。具有特殊BUNDLE操作码的顶级MI用于表示包的开始。将BUNDLE MI与不在捆绑包内也不代表捆绑包的单个MI混合是合法的。

MachineInstr通道应作为单个单元在MI束上运行。成员方法已被教导如何正确处理捆绑包和捆绑包中的MI。MachineBasicBlock迭代器已被修改为跳过捆绑的MI，以强制执行捆绑为单个单元的概念。MachineBasicBlock中添加了另一个迭代器instr_iterator，以允许传递遍历MachineBasicBlock中的所有MI，包括嵌套在bundle中的MI。顶级BUNDLE指令必须具有正确的寄存器MachineOperand集，这些寄存器表示绑定MI的累积输入和输出。

VLIW体系结构的MachineInstr打包/绑定通常应作为寄存器分配超级传递的一部分完成。更具体地说，在代码生成器退出SSA形式之后（即，在两个地址传递、PHI消除和复制合并之后），应执行确定应将哪些MI捆绑在一起的传递。在将虚拟寄存器重写为物理寄存器后，应最终确定此类捆绑包（即添加捆绑包MI以及输入和输出寄存器MachineOperands）。这消除了向BUNDLE指令添加虚拟寄存器操作数的需要，这将有效地使虚拟寄存器定义和使用列表加倍。捆绑包可以使用虚拟寄存器并以SSA形式形成，但可能不适用于所有用例。

MI bundle support does not change the physical representations of MachineBasicBlock and MachineInstr. All the MIs (including top level and nested ones) are stored as sequential list of MIs. The “bundled” MIs are marked with the ‘InsideBundle’ flag. A top level MI with the special BUNDLE opcode is used to represent the start of a bundle. It’s legal to mix BUNDLE MIs with individual MIs that are not inside bundles nor represent bundles.

MachineInstr passes should operate on a MI bundle as a single unit. Member methods have been taught to correctly handle bundles and MIs inside bundles. The MachineBasicBlock iterator has been modified to skip over bundled MIs to enforce the bundle-as-a-single-unit concept. An alternative iterator instr_iterator has been added to MachineBasicBlock to allow passes to iterate over all of the MIs in a MachineBasicBlock, including those which are nested inside bundles. The top level BUNDLE instruction must have the correct set of register MachineOperand’s that represent the cumulative inputs and outputs of the bundled MIs.

Packing / bundling of MachineInstrs for VLIW architectures should generally be done as part of the register allocation super-pass. More specifically, the pass which determines what MIs should be bundled together should be done after code generator exits SSA form (i.e. after two-address pass, PHI elimination, and copy coalescing). Such bundles should be finalized (i.e. adding BUNDLE MIs and input and output register MachineOperands) after virtual registers have been rewritten into physical registers. This eliminates the need to add virtual register operands to BUNDLE instructions which would effectively double the virtual register def and use lists. Bundles may use virtual registers and be formed in SSA form, but may not be appropriate for all use cases.

MC层用于表示和处理原始机器代码级别的代码，没有“高级”信息，如“常量池”、“跳转表”、“全局变量”或类似的信息。在这个级别上，LLVM处理诸如标签名称、机器指令和对象文件中的部分之类的事情。该层中的代码用于许多重要用途：代码生成器的尾部使用它来编写.s或.o文件，llvmmc工具也使用它来实现独立的机器代码汇编程序和反汇编程序。

本节介绍一些重要的类。还有许多重要的子系统在该层进行交互，这些子系统将在本手册的后面进行描述。

The MC Layer is used to represent and process code at the raw machine code level, devoid of “high level” information like “constant pools”, “jump tables”, “global variables” or anything like that. At this level, LLVM handles things like label names, machine instructions, and sections in the object file. The code in this layer is used for a number of important purposes: the tail end of the code generator uses it to write a .s or .o file, and it is also used by the llvm-mc tool to implement standalone machine code assemblers and disassemblers.

This section describes some of the important classes. There are also a number of important subsystems that interact at this layer, they are described later in this manual.

MCStreamer最好被认为是一个汇编程序API。它是一个抽象API，以不同的方式实现（例如输出.s文件、输出ELF.o文件等），但其API直接对应于在.s文件中看到的内容。MCStreamer每个指令有一个方法，例如EmitLabel、EmitSymbolAttribute、switchSection、emitValue（用于.byte、.word）等，它们直接对应于程序集级指令。它还有一个EmitInstruction方法，用于向拖缆输出MCInst。

这个API对于两个客户机来说最重要：llvm-mc独立汇编程序实际上是一个解析器，它解析一行，然后调用MCStreamer上的方法。在代码生成器中，代码生成器的代码发射阶段将更高级别的LLVM IR和Machine*构造降低到MC层，通过MCStreamer发射指令。

在MCStreamer的实现端，有两个主要的实现：一个用于写出.s文件（MCAsmStreamer），另一个用于写入.o文件（MCAObjectStreamer）。MCAsmStreamer是一个简单的实现，它为每个方法打印出一个指令（例如EmitValue->.byte），但MCObjectStreamer实现了一个完整的汇编程序。

对于特定于目标的指令，MCStreamer有一个MCTargetStreamer实例。每个需要它的目标都定义了一个继承自它的类，与MCStreamer本身非常相似：每个指令有一个方法，两个继承自其的类，一个目标对象流和一个目标asm流。目标asm拖缆只打印它（emitStart->.fnstart），对象拖缆为它实现汇编程序逻辑。

要使llvm使用这些类，目标初始化必须调用TargetRegistry:：RegisterAsmStreamer和TargetRegistry::RegisterMCObjectStreamer，并通过回调来分配相应的目标流，并将其传递给createAsmStreaker或相应的对象流构造函数。

MCStreamer is best thought of as an assembler API. It is an abstract API which is implemented in different ways (e.g. to output a .s file, output an ELF .o file, etc) but whose API correspond directly to what you see in a .s file. MCStreamer has one method per directive, such as EmitLabel, EmitSymbolAttribute, switchSection, emitValue (for .byte, .word), etc, which directly correspond to assembly level directives. It also has an EmitInstruction method, which is used to output an MCInst to the streamer.

This API is most important for two clients: the llvm-mc stand-alone assembler is effectively a parser that parses a line, then invokes a method on MCStreamer. In the code generator, the Code Emission phase of the code generator lowers higher level LLVM IR and Machine* constructs down to the MC layer, emitting directives through MCStreamer.

On the implementation side of MCStreamer, there are two major implementations: one for writing out a .s file (MCAsmStreamer), and one for writing out a .o file (MCObjectStreamer). MCAsmStreamer is a straightforward implementation that prints out a directive for each method (e.g. EmitValue -> .byte), but MCObjectStreamer implements a full assembler.

For target specific directives, the MCStreamer has a MCTargetStreamer instance. Each target that needs it defines a class that inherits from it and is a lot like MCStreamer itself: It has one method per directive and two classes that inherit from it, a target object streamer and a target asm streamer. The target asm streamer just prints it (emitFnStart -> .fnstart), and the object streamer implement the assembler logic for it.

To make llvm use these classes, the target initialization must call TargetRegistry::RegisterAsmStreamer and TargetRegistry::RegisterMCObjectStreamer passing callbacks that allocate the corresponding target streamer and pass it to createAsmStreamer or to the appropriate object streamer constructor.

The MCContext class is the owner of a variety of uniqued data structures at the MC layer, including symbols, sections, etc. As such, this is the class that you interact with to create symbols and sections. This class can not be subclassed.

The MCSymbol class represents a symbol (aka label) in the assembly file. There are two interesting kinds of symbols: assembler temporary symbols, and normal symbols. Assembler temporary symbols are used and processed by the assembler but are discarded when the object file is produced. The distinction is usually represented by adding a prefix to the label, for example “L” labels are assembler temporary labels in MachO.

MCSymbols are created by MCContext and uniqued there. This means that MCSymbols can be compared for pointer equivalence to find out if they are the same symbol. Note that pointer inequality does not guarantee the labels will end up at different addresses though. It’s perfectly legal to output something like this to the .s file:

foo:
bar:
  .byte 4

In this case, both the foo and bar symbols will have the same address.

The MCSection class represents an object-file specific section. It is subclassed by object file specific implementations (e.g. MCSectionMachO, MCSectionCOFF, MCSectionELF) and these are created and uniqued by MCContext. The MCStreamer has a notion of the current section, which can be changed with the SwitchToSection method (which corresponds to a “.section” directive in a .s file).

The MCInst class is a target-independent representation of an instruction. It is a simple class (much more so than MachineInstr) that holds a target-specific opcode and a vector of MCOperands. MCOperand, in turn, is a simple discriminated union of three cases: 1) a simple immediate, 2) a target register ID, 3) a symbolic expression (e.g. “Lfoo-Lbar+42”) as an MCExpr.

MCInst is the common currency used to represent machine instructions at the MC layer. It is the type used by the instruction encoder, the instruction printer, and the type generated by the assembly parser and disassembler.

The MC layer’s object writers support a variety of object formats. Because of target-specific aspects of object formats each target only supports a subset of the formats supported by the MC layer. Most targets support emitting ELF objects. Other vendor-specific objects are generally supported only on targets that are supported by that vendor (i.e. MachO is only supported on targets supported by Darwin, and XCOFF is only supported on targets that support AIX). Additionally some targets have their own object formats (i.e. DirectX, SPIR-V and WebAssembly).

The table below captures a snapshot of object file support in LLVM:

Object File Formats

Format

Supported Targets

COFF

AArch64, ARM, X86

DXContainer

DirectX

ELF

AArch64, AMDGPU, ARM, AVR, BPF, CSKY, Hexagon, Lanai, LoongArch, M86k, MSP430, MIPS, PowerPC, RISCV, SPARC, SystemZ, VE, X86

GCOFF

SystemZ

MachO

AArch64, ARM, X86

SPIR-V

SPIRV

WASM

WebAssembly

XCOFF

PowerPC

This section documents the phases described in the high-level design of the code generator. It explains how they work and some of the rationale behind their design.

Instruction Selection is the process of translating LLVM code presented to the code generator into target-specific machine instructions. There are several well-known ways to do this in the literature. LLVM uses a SelectionDAG based instruction selector.

Portions of the DAG instruction selector are generated from the target description (*.td) files. Our goal is for the entire instruction selector to be generated from these .td files, though currently there are still things that require custom C++ code.

GlobalISel is another instruction selection framework.

The SelectionDAG provides an abstraction for code representation in a way that is amenable to instruction selection using automatic techniques (e.g. dynamic-programming based optimal pattern matching selectors). It is also well-suited to other phases of code generation; in particular, instruction scheduling (SelectionDAG’s are very close to scheduling DAGs post-selection). Additionally, the SelectionDAG provides a host representation where a large variety of very-low-level (but target-independent) optimizations may be performed; ones which require extensive information about the instructions efficiently supported by the target.

The SelectionDAG is a Directed-Acyclic-Graph whose nodes are instances of the SDNode class. The primary payload of the SDNode is its operation code (Opcode) that indicates what operation the node performs and the operands to the operation. The various operation node types are described at the top of the include/llvm/CodeGen/ISDOpcodes.h file.

Although most operations define a single value, each node in the graph may define multiple values. For example, a combined div/rem operation will define both the dividend and the remainder. Many other situations require multiple values as well. Each node also has some number of operands, which are edges to the node defining the used value. Because nodes may define multiple values, edges are represented by instances of the SDValue class, which is a <SDNode, unsigned> pair, indicating the node and result value being used, respectively. Each value produced by an SDNode has an associated MVT (Machine Value Type) indicating what the type of the value is.

SelectionDAGs contain two different kinds of values: those that represent data flow and those that represent control flow dependencies. Data values are simple edges with an integer or floating point value type. Control edges are represented as “chain” edges which are of type MVT::Other. These edges provide an ordering between nodes that have side effects (such as loads, stores, calls, returns, etc). All nodes that have side effects should take a token chain as input and produce a new one as output. By convention, token chain inputs are always operand #0, and chain results are always the last value produced by an operation. However, after instruction selection, the machine nodes have their chain after the instruction’s operands, and may be followed by glue nodes.

A SelectionDAG has designated “Entry” and “Root” nodes. The Entry node is always a marker node with an Opcode of ISD::EntryToken. The Root node is the final side-effecting node in the token chain. For example, in a single basic block function it would be the return node.

One important concept for SelectionDAGs is the notion of a “legal” vs. “illegal” DAG. A legal DAG for a target is one that only uses supported operations and supported types. On a 32-bit PowerPC, for example, a DAG with a value of type i1, i8, i16, or i64 would be illegal, as would a DAG that uses a SREM or UREM operation. The legalize types and legalize operations phases are responsible for turning an illegal DAG into a legal DAG.

SelectionDAG-based instruction selection consists of the following steps:

  1. Build initial DAG — This stage performs a simple translation from the input LLVM code to an illegal SelectionDAG.
  2. Optimize SelectionDAG — This stage performs simple optimizations on the SelectionDAG to simplify it, and recognize meta instructions (like rotates and div/rem pairs) for targets that support these meta operations. This makes the resultant code more efficient and the select instructions from DAG phase (below) simpler.
  3. Legalize SelectionDAG Types — This stage transforms SelectionDAG nodes to eliminate any types that are unsupported on the target.
  4. Optimize SelectionDAG — The SelectionDAG optimizer is run to clean up redundancies exposed by type legalization.
  5. Legalize SelectionDAG Ops — This stage transforms SelectionDAG nodes to eliminate any operations that are unsupported on the target.
  6. Optimize SelectionDAG — The SelectionDAG optimizer is run to eliminate inefficiencies introduced by operation legalization.
  7. Select instructions from DAG — Finally, the target instruction selector matches the DAG operations to target instructions. This process translates the target-independent input DAG into another DAG of target instructions.
  8. SelectionDAG Scheduling and Formation — The last phase assigns a linear order to the instructions in the target-instruction DAG and emits them into the MachineFunction being compiled. This step uses traditional prepass scheduling techniques.

After all of these steps are complete, the SelectionDAG is destroyed and the rest of the code generation passes are run.

One great way to visualize what is going on here is to take advantage of a few LLC command line options. The following options pop up a window displaying the SelectionDAG at specific times (if you only get errors printed to the console while using this, you probably need to configure your system to add support for it).

  • -view-dag-combine1-dags displays the DAG after being built, before the first optimization pass.
  • -view-legalize-dags displays the DAG before Legalization.
  • -view-dag-combine2-dags displays the DAG before the second optimization pass.
  • -view-isel-dags displays the DAG before the Select phase.
  • -view-sched-dags displays the DAG before Scheduling.

The -view-sunit-dags displays the Scheduler’s dependency graph. This graph is based on the final SelectionDAG, with nodes that must be scheduled together bundled into a single scheduling-unit node, and with immediate operands and other nodes that aren’t relevant for scheduling omitted.

The option -filter-view-dags allows to select the name of the basic block that you are interested to visualize and filters all the previous view-*-dags options.

The initial SelectionDAG is naïvely peephole expanded from the LLVM input by the SelectionDAGBuilder class. The intent of this pass is to expose as much low-level, target-specific details to the SelectionDAG as possible. This pass is mostly hard-coded (e.g. an LLVM add turns into an SDNode add while a getelementptr is expanded into the obvious arithmetic). This pass requires target-specific hooks to lower calls, returns, varargs, etc. For these features, the TargetLowering interface is used.

The Legalize phase is in charge of converting a DAG to only use the types that are natively supported by the target.

There are two main ways of converting values of unsupported scalar types to values of supported types: converting small types to larger types (“promoting”), and breaking up large integer types into smaller ones (“expanding”). For example, a target might require that all f32 values are promoted to f64 and that all i1/i8/i16 values are promoted to i32. The same target might require that all i64 values be expanded into pairs of i32 values. These changes can insert sign and zero extensions as needed to make sure that the final code has the same behavior as the input.

There are two main ways of converting values of unsupported vector types to value of supported types: splitting vector types, multiple times if necessary, until a legal type is found, and extending vector types by adding elements to the end to round them out to legal types (“widening”). If a vector gets split all the way down to single-element parts with no supported vector type being found, the elements are converted to scalars (“scalarizing”).

A target implementation tells the legalizer which types are supported (and which register class to use for them) by calling the addRegisterClass method in its TargetLowering constructor.

The Legalize phase is in charge of converting a DAG to only use the operations that are natively supported by the target.

Targets often have weird constraints, such as not supporting every operation on every supported datatype (e.g. X86 does not support byte conditional moves and PowerPC does not support sign-extending loads from a 16-bit memory location). Legalize takes care of this by open-coding another sequence of operations to emulate the operation (“expansion”), by promoting one type to a larger type that supports the operation (“promotion”), or by using a target-specific hook to implement the legalization (“custom”).

A target implementation tells the legalizer which operations are not supported (and which of the above three actions to take) by calling the setOperationAction method in its TargetLowering constructor.

If a target has legal vector types, it is expected to produce efficient machine code for common forms of the shufflevector IR instruction using those types. This may require custom legalization for SelectionDAG vector operations that are created from the shufflevector IR. The shufflevector forms that should be handled include:

  • Vector select — Each element of the vector is chosen from either of the corresponding elements of the 2 input vectors. This operation may also be known as a “blend” or “bitwise select” in target assembly. This type of shuffle maps directly to the shuffle_vector SelectionDAG node.
  • Insert subvector — A vector is placed into a longer vector type starting at index 0. This type of shuffle maps directly to the insert_subvector SelectionDAG node with the index operand set to 0.
  • Extract subvector — A vector is pulled from a longer vector type starting at index 0. This type of shuffle maps directly to the extract_subvector SelectionDAG node with the index operand set to 0.
  • Splat — All elements of the vector have identical scalar elements. This operation may also be known as a “broadcast” or “duplicate” in target assembly. The shufflevector IR instruction may change the vector length, so this operation may map to multiple SelectionDAG nodes including shuffle_vector, concat_vectors, insert_subvector, and extract_subvector.

Prior to the existence of the Legalize passes, we required that every target selector supported and handled every operator and type even if they are not natively supported. The introduction of the Legalize phases allows all of the canonicalization patterns to be shared across targets, and makes it very easy to optimize the canonicalized code because it is still in the form of a DAG.

The SelectionDAG optimization phase is run multiple times for code generation, immediately after the DAG is built and once after each legalization. The first run of the pass allows the initial code to be cleaned up (e.g. performing optimizations that depend on knowing that the operators have restricted type inputs). Subsequent runs of the pass clean up the messy code generated by the Legalize passes, which allows Legalize to be very simple (it can focus on making code legal instead of focusing on generating good and legal code).

One important class of optimizations performed is optimizing inserted sign and zero extension instructions. We currently use ad-hoc techniques, but could move to more rigorous techniques in the future. Here are some good papers on the subject:

“Widening integer arithmetic”
Kevin Redwine and Norman Ramsey
International Conference on Compiler Construction (CC) 2004

“Effective sign extension elimination”
Motohiro Kawahito, Hideaki Komatsu, and Toshio Nakatani
Proceedings of the ACM SIGPLAN 2002 Conference on Programming Language Design and Implementation.

The Select phase is the bulk of the target-specific code for instruction selection. This phase takes a legal SelectionDAG as input, pattern matches the instructions supported by the target to this DAG, and produces a new DAG of target code. For example, consider the following LLVM fragment:

%t1 = fadd float %W, %X
%t2 = fmul float %t1, %Y
%t3 = fadd float %t2, %Z

This LLVM code corresponds to a SelectionDAG that looks basically like this:

(fadd:f32 (fmul:f32 (fadd:f32 W, X), Y), Z)

If a target supports floating point multiply-and-add (FMA) operations, one of the adds can be merged with the multiply. On the PowerPC, for example, the output of the instruction selector might look like this DAG:

(FMADDS (FADDS W, X), Y, Z)

The FMADDS instruction is a ternary instruction that multiplies its first two operands and adds the third (as single-precision floating-point numbers). The FADDS instruction is a simple binary single-precision add instruction. To perform this pattern match, the PowerPC backend includes the following instruction definitions:

def FMADDS : AForm_1<59, 29,
                    (ops F4RC:$FRT, F4RC:$FRA, F4RC:$FRC, F4RC:$FRB),
                    "fmadds $FRT, $FRA, $FRC, $FRB",
                    [(set F4RC:$FRT, (fadd (fmul F4RC:$FRA, F4RC:$FRC),
                                           F4RC:$FRB))]>;
def FADDS : AForm_2<59, 21,
                    (ops F4RC:$FRT, F4RC:$FRA, F4RC:$FRB),
                    "fadds $FRT, $FRA, $FRB",
                    [(set F4RC:$FRT, (fadd F4RC:$FRA, F4RC:$FRB))]>;

The highlighted portion of the instruction definitions indicates the pattern used to match the instructions. The DAG operators (like fmul/fadd) are defined in the include/llvm/Target/TargetSelectionDAG.td file. “F4RC” is the register class of the input and result values.

The TableGen DAG instruction selector generator reads the instruction patterns in the .td file and automatically builds parts of the pattern matching code for your target. It has the following strengths:

  • At compiler-compile time, it analyzes your instruction patterns and tells you if your patterns make sense or not.
  • It can handle arbitrary constraints on operands for the pattern match. In particular, it is straight-forward to say things like “match any immediate that is a 13-bit sign-extended value”. For examples, see the immSExt16 and related tblgen classes in the PowerPC backend.
  • It knows several important identities for the patterns defined. For example, it knows that addition is commutative, so it allows the FMADDS pattern above to match “(fadd X, (fmul Y, Z))” as well as “(fadd (fmul X, Y), Z)”, without the target author having to specially handle this case.
  • It has a full-featured type-inferencing system. In particular, you should rarely have to explicitly tell the system what type parts of your patterns are. In the FMADDS case above, we didn’t have to tell tblgen that all of the nodes in the pattern are of type ‘f32’. It was able to infer and propagate this knowledge from the fact that F4RC has type ‘f32’.
  • Targets can define their own (and rely on built-in) “pattern fragments”. Pattern fragments are chunks of reusable patterns that get inlined into your patterns during compiler-compile time. For example, the integer “(not x)” operation is actually defined as a pattern fragment that expands as “(xor x, -1)”, since the SelectionDAG does not have a native ‘not’ operation. Targets can define their own short-hand fragments as they see fit. See the definition of ‘not’ and ‘ineg’ for examples.
  • In addition to instructions, targets can specify arbitrary patterns that map to one or more instructions using the ‘Pat’ class. For example, the PowerPC has no way to load an arbitrary integer immediate into a register in one instruction. To tell tblgen how to do this, it defines:
·       // Arbitrary immediate support.  Implement in terms of LIS/ORI.
·       def : Pat<(i32 imm:$imm),
·                 (ORI (LIS (HI16 imm:$imm)), (LO16 imm:$imm))>;

If none of the single-instruction patterns for loading an immediate into a register match, this will be used. This rule says “match an arbitrary i32 immediate, turning it into an ORI (‘or a 16-bit immediate’) and an LIS (‘load 16-bit immediate, where the immediate is shifted to the left 16 bits’) instruction”. To make this work, the LO16/HI16 node transformations are used to manipulate the input immediate (in this case, take the high or low 16-bits of the immediate).

  • When using the ‘Pat’ class to map a pattern to an instruction that has one or more complex operands (like e.g. X86 addressing mode), the pattern may either specify the operand as a whole using a ComplexPattern, or else it may specify the components of the complex operand separately. The latter is done e.g. for pre-increment instructions by the PowerPC back end:
·       def STWU  : DForm_1<37, (outs ptr_rc:$ea_res), (ins GPRC:$rS, memri:$dst),
·                       "stwu $rS, $dst", LdStStoreUpd, []>,
·                       RegConstraint<"$dst.reg = $ea_res">, NoEncode<"$ea_res">;
·        
·       def : Pat<(pre_store GPRC:$rS, ptr_rc:$ptrreg, iaddroff:$ptroff),
·                 (STWU GPRC:$rS, iaddroff:$ptroff, ptr_rc:$ptrreg)>;

Here, the pair of ptroff and ptrreg operands is matched onto the complex operand dst of class memri in the STWU instruction.

  • While the system does automate a lot, it still allows you to write custom C++ code to match special cases if there is something that is hard to express.

While it has many strengths, the system currently has some limitations, primarily because it is a work in progress and is not yet finished:

  • Overall, there is no way to define or match SelectionDAG nodes that define multiple values (e.g. SMUL_LOHI, LOAD, CALL, etc). This is the biggest reason that you currently still have to write custom C++ code for your instruction selector.
  • There is no great way to support matching complex addressing modes yet. In the future, we will extend pattern fragments to allow them to define multiple values (e.g. the four operands of the X86 addressing mode, which are currently matched with custom C++ code). In addition, we’ll extend fragments so that a fragment can match multiple different patterns.
  • We don’t automatically infer flags like isStore/isLoad yet.
  • We don’t automatically generate the set of supported registers and operations for the Legalizer yet.
  • We don’t have a way of tying in custom legalized nodes yet.

Despite these limitations, the instruction selector generator is still quite useful for most of the binary and logical operations in typical instruction sets. If you run into any problems or can’t figure out how to do something, please let Chris know!

The scheduling phase takes the DAG of target instructions from the selection phase and assigns an order. The scheduler can pick an order depending on various constraints of the machines (i.e. order for minimal register pressure or try to cover instruction latencies). Once an order is established, the DAG is converted to a list of MachineInstrs and the SelectionDAG is destroyed.

Note that this phase is logically separate from the instruction selection phase, but is tied to it closely in the code because it operates on SelectionDAGs.

  1. Optional function-at-a-time selection.
  2. Auto-generate entire selector from .td file.

To Be Written

Live Intervals are the ranges (intervals) where a variable is live. They are used by some register allocator passes to determine if two or more virtual registers which require the same physical register are live at the same point in the program (i.e., they conflict). When this situation occurs, one virtual register must be spilled.

The first step in determining the live intervals of variables is to calculate the set of registers that are immediately dead after the instruction (i.e., the instruction calculates the value, but it is never used) and the set of registers that are used by the instruction, but are never used after the instruction (i.e., they are killed). Live variable information is computed for each virtual register and register allocatable physical register in the function. This is done in a very efficient manner because it uses SSA to sparsely compute lifetime information for virtual registers (which are in SSA form) and only has to track physical registers within a block. Before register allocation, LLVM can assume that physical registers are only live within a single basic block. This allows it to do a single, local analysis to resolve physical register lifetimes within each basic block. If a physical register is not register allocatable (e.g., a stack pointer or condition codes), it is not tracked.

Physical registers may be live in to or out of a function. Live in values are typically arguments in registers. Live out values are typically return values in registers. Live in values are marked as such, and are given a dummy “defining” instruction during live intervals analysis. If the last basic block of a function is a return, then it’s marked as using all live out values in the function.

PHI nodes need to be handled specially, because the calculation of the live variable information from a depth first traversal of the CFG of the function won’t guarantee that a virtual register used by the PHI node is defined before it’s used. When a PHI node is encountered, only the definition is handled, because the uses will be handled in other basic blocks.

For each PHI node of the current basic block, we simulate an assignment at the end of the current basic block and traverse the successor basic blocks. If a successor basic block has a PHI node and one of the PHI node’s operands is coming from the current basic block, then the variable is marked as alive within the current basic block and all of its predecessor basic blocks, until the basic block with the defining instruction is encountered.

We now have the information available to perform the live intervals analysis and build the live intervals themselves. We start off by numbering the basic blocks and machine instructions. We then handle the “live-in” values. These are in physical registers, so the physical register is assumed to be killed by the end of the basic block. Live intervals for virtual registers are computed for some ordering of the machine instructions [1, N]. A live interval is an interval [i, j), where 1 >= i >= j > N, for which a variable is live.

Note

More to come…

The Register Allocation problem consists in mapping a program Pv, that can use an unbounded number of virtual registers, to a program Pp that contains a finite (possibly small) number of physical registers. Each target architecture has a different number of physical registers. If the number of physical registers is not enough to accommodate all the virtual registers, some of them will have to be mapped into memory. These virtuals are called spilled virtuals.

In LLVM, physical registers are denoted by integer numbers that normally range from 1 to 1023. To see how this numbering is defined for a particular architecture, you can read the GenRegisterNames.inc file for that architecture. For instance, by inspecting lib/Target/X86/X86GenRegisterInfo.inc we see that the 32-bit register EAX is denoted by 43, and the MMX register MM0 is mapped to 65.

Some architectures contain registers that share the same physical location. A notable example is the X86 platform. For instance, in the X86 architecture, the registers EAX, AX and AL share the first eight bits. These physical registers are marked as aliased in LLVM. Given a particular architecture, you can check which registers are aliased by inspecting its RegisterInfo.td file. Moreover, the class MCRegAliasIterator enumerates all the physical registers aliased to a register.

Physical registers, in LLVM, are grouped in Register Classes. Elements in the same register class are functionally equivalent, and can be interchangeably used. Each virtual register can only be mapped to physical registers of a particular class. For instance, in the X86 architecture, some virtuals can only be allocated to 8 bit registers. A register class is described by TargetRegisterClass objects. To discover if a virtual register is compatible with a given physical, this code can be used:

bool RegMapping_Fer::compatible_class(MachineFunction &mf,
                                      unsigned v_reg,
                                      unsigned p_reg) {
  assert(TargetRegisterInfo::isPhysicalRegister(p_reg) &&
         "Target register must be physical");
  const TargetRegisterClass *trc = mf.getRegInfo().getRegClass(v_reg);
  return trc->contains(p_reg);
}

Sometimes, mostly for debugging purposes, it is useful to change the number of physical registers available in the target architecture. This must be done statically, inside the TargetRegisterInfo.td file. Just grep for RegisterClass, the last parameter of which is a list of registers. Just commenting some out is one simple way to avoid them being used. A more polite way is to explicitly exclude some registers from the allocation order. See the definition of the GR8 register class in lib/Target/X86/X86RegisterInfo.td for an example of this.

Virtual registers are also denoted by integer numbers. Contrary to physical registers, different virtual registers never share the same number. Whereas physical registers are statically defined in a TargetRegisterInfo.td file and cannot be created by the application developer, that is not the case with virtual registers. In order to create new virtual registers, use the method MachineRegisterInfo::createVirtualRegister(). This method will return a new virtual register. Use an IndexedMap<Foo, VirtReg2IndexFunctor> to hold information per virtual register. If you need to enumerate all virtual registers, use the function TargetRegisterInfo::index2VirtReg() to find the virtual register numbers:

for (unsigned i = 0, e = MRI->getNumVirtRegs(); i != e; ++i) {
  unsigned VirtReg = TargetRegisterInfo::index2VirtReg(i);
  stuff(VirtReg);
}

Before register allocation, the operands of an instruction are mostly virtual registers, although physical registers may also be used. In order to check if a given machine operand is a register, use the boolean function MachineOperand::isRegister(). To obtain the integer code of a register, use MachineOperand::getReg(). An instruction may define or use a register. For instance, ADD reg:1026 := reg:1025 reg:1024 defines the registers 1024, and uses registers 1025 and 1026. Given a register operand, the method MachineOperand::isUse() informs if that register is being used by the instruction. The method MachineOperand::isDef() informs if that registers is being defined.

We will call physical registers present in the LLVM bitcode before register allocation pre-colored registers. Pre-colored registers are used in many different situations, for instance, to pass parameters of functions calls, and to store results of particular instructions. There are two types of pre-colored registers: the ones implicitly defined, and those explicitly defined. Explicitly defined registers are normal operands, and can be accessed with MachineInstr::getOperand(int)::getReg(). In order to check which registers are implicitly defined by an instruction, use the TargetInstrInfo::get(opcode)::ImplicitDefs, where opcode is the opcode of the target instruction. One important difference between explicit and implicit physical registers is that the latter are defined statically for each instruction, whereas the former may vary depending on the program being compiled. For example, an instruction that represents a function call will always implicitly define or use the same set of physical registers. To read the registers implicitly used by an instruction, use TargetInstrInfo::get(opcode)::ImplicitUses. Pre-colored registers impose constraints on any register allocation algorithm. The register allocator must make sure that none of them are overwritten by the values of virtual registers while still alive.

There are two ways to map virtual registers to physical registers (or to memory slots). The first way, that we will call direct mapping, is based on the use of methods of the classes TargetRegisterInfo, and MachineOperand. The second way, that we will call indirect mapping, relies on the VirtRegMap class in order to insert loads and stores sending and getting values to and from memory.

The direct mapping provides more flexibility to the developer of the register allocator; however, it is more error prone, and demands more implementation work. Basically, the programmer will have to specify where load and store instructions should be inserted in the target function being compiled in order to get and store values in memory. To assign a physical register to a virtual register present in a given operand, use MachineOperand::setReg(p_reg). To insert a store instruction, use TargetInstrInfo::storeRegToStackSlot(...), and to insert a load instruction, use TargetInstrInfo::loadRegFromStackSlot.

The indirect mapping shields the application developer from the complexities of inserting load and store instructions. In order to map a virtual register to a physical one, use VirtRegMap::assignVirt2Phys(vreg, preg). In order to map a certain virtual register to memory, use VirtRegMap::assignVirt2StackSlot(vreg). This method will return the stack slot where vreg’s value will be located. If it is necessary to map another virtual register to the same stack slot, use VirtRegMap::assignVirt2StackSlot(vreg, stack_location). One important point to consider when using the indirect mapping, is that even if a virtual register is mapped to memory, it still needs to be mapped to a physical register. This physical register is the location where the virtual register is supposed to be found before being stored or after being reloaded.

If the indirect strategy is used, after all the virtual registers have been mapped to physical registers or stack slots, it is necessary to use a spiller object to place load and store instructions in the code. Every virtual that has been mapped to a stack slot will be stored to memory after being defined and will be loaded before being used. The implementation of the spiller tries to recycle load/store instructions, avoiding unnecessary instructions. For an example of how to invoke the spiller, see RegAllocLinearScan::runOnMachineFunction in lib/CodeGen/RegAllocLinearScan.cpp.

With very rare exceptions (e.g., function calls), the LLVM machine code instructions are three address instructions. That is, each instruction is expected to define at most one register, and to use at most two registers. However, some architectures use two address instructions. In this case, the defined register is also one of the used registers. For instance, an instruction such as ADD %EAX, %EBX, in X86 is actually equivalent to %EAX = %EAX + %EBX.

In order to produce correct code, LLVM must convert three address instructions that represent two address instructions into true two address instructions. LLVM provides the pass TwoAddressInstructionPass for this specific purpose. It must be run before register allocation takes place. After its execution, the resulting code may no longer be in SSA form. This happens, for instance, in situations where an instruction such as %a = ADD %b %c is converted to two instructions such as:

%a = MOVE %b
%a = ADD %a %c

Notice that, internally, the second instruction is represented as ADD %a[def/use] %c. I.e., the register operand %a is both used and defined by the instruction.

An important transformation that happens during register allocation is called the SSA Deconstruction Phase. The SSA form simplifies many analyses that are performed on the control flow graph of programs. However, traditional instruction sets do not implement PHI instructions. Thus, in order to generate executable code, compilers must replace PHI instructions with other instructions that preserve their semantics.

There are many ways in which PHI instructions can safely be removed from the target code. The most traditional PHI deconstruction algorithm replaces PHI instructions with copy instructions. That is the strategy adopted by LLVM. The SSA deconstruction algorithm is implemented in lib/CodeGen/PHIElimination.cpp. In order to invoke this pass, the identifier PHIEliminationID must be marked as required in the code of the register allocator.

Instruction folding is an optimization performed during register allocation that removes unnecessary copy instructions. For instance, a sequence of instructions such as:

%EBX = LOAD %mem_address
%EAX = COPY %EBX

can be safely substituted by the single instruction:

%EAX = LOAD %mem_address

Instructions can be folded with the TargetRegisterInfo::foldMemoryOperand(...) method. Care must be taken when folding instructions; a folded instruction can be quite different from the original instruction. See LiveIntervals::addIntervalsForSpills in lib/CodeGen/LiveIntervalAnalysis.cpp for an example of its use.

The LLVM infrastructure provides the application developer with three different register allocators:

  • Fast — This register allocator is the default for debug builds. It allocates registers on a basic block level, attempting to keep values in registers and reusing registers as appropriate.
  • Basic — This is an incremental approach to register allocation. Live ranges are assigned to registers one at a time in an order that is driven by heuristics. Since code can be rewritten on-the-fly during allocation, this framework allows interesting allocators to be developed as extensions. It is not itself a production register allocator but is a potentially useful stand-alone mode for triaging bugs and as a performance baseline.
  • Greedy — The default allocator. This is a highly tuned implementation of the Basic allocator that incorporates global live range splitting. This allocator works hard to minimize the cost of spill code.
  • PBQP — A Partitioned Boolean Quadratic Programming (PBQP) based register allocator. This allocator works by constructing a PBQP problem representing the register allocation problem under consideration, solving this using a PBQP solver, and mapping the solution back to a register assignment.

The type of register allocator used in llc can be chosen with the command line option -regalloc=...:

$ llc -regalloc=linearscan file.bc -o ln.s
$ llc -regalloc=fast file.bc -o fa.s
$ llc -regalloc=pbqp file.bc -o pbqp.s

Note

To Be Written

Throwing an exception requires unwinding out of a function. The information on how to unwind a given function is traditionally expressed in DWARF unwind (a.k.a. frame) info. But that format was originally developed for debuggers to backtrace, and each Frame Description Entry (FDE) requires ~20-30 bytes per function. There is also the cost of mapping from an address in a function to the corresponding FDE at runtime. An alternative unwind encoding is called compact unwind and requires just 4-bytes per function.

The compact unwind encoding is a 32-bit value, which is encoded in an architecture-specific way. It specifies which registers to restore and from where, and how to unwind out of the function. When the linker creates a final linked image, it will create a __TEXT,__unwind_info section. This section is a small and fast way for the runtime to access unwind info for any given function. If we emit compact unwind info for the function, that compact unwind info will be encoded in the __TEXT,__unwind_info section. If we emit DWARF unwind info, the __TEXT,__unwind_info section will contain the offset of the FDE in the __TEXT,__eh_frame section in the final linked image.

For X86, there are three modes for the compact unwind encoding:

Function with a Frame Pointer (``EBP`` or ``RBP``)

EBP/RBP-based frame, where EBP/RBP is pushed onto the stack immediately after the return address, then ESP/RSP is moved to EBP/RBP. Thus to unwind, ESP/RSP is restored with the current EBP/RBP value, then EBP/RBP is restored by popping the stack, and the return is done by popping the stack once more into the PC. All non-volatile registers that need to be restored must have been saved in a small range on the stack that starts EBP-4 to EBP-1020 (RBP-8 to RBP-1020). The offset (divided by 4 in 32-bit mode and 8 in 64-bit mode) is encoded in bits 16-23 (mask: 0x00FF0000). The registers saved are encoded in bits 0-14 (mask: 0x00007FFF) as five 3-bit entries from the following table:

Compact Number

i386 Register

x86-64 Register

1

EBX

RBX

2

ECX

R12

3

EDX

R13

4

EDI

R14

5

ESI

R15

6

EBP

RBP

Frameless with a Small Constant Stack Size (``EBP`` or ``RBP`` is not used as a frame pointer)

To return, a constant (encoded in the compact unwind encoding) is added to the ESP/RSP. Then the return is done by popping the stack into the PC. All non-volatile registers that need to be restored must have been saved on the stack immediately after the return address. The stack size (divided by 4 in 32-bit mode and 8 in 64-bit mode) is encoded in bits 16-23 (mask: 0x00FF0000). There is a maximum stack size of 1024 bytes in 32-bit mode and 2048 in 64-bit mode. The number of registers saved is encoded in bits 9-12 (mask: 0x00001C00). Bits 0-9 (mask: 0x000003FF) contain which registers were saved and their order. (See the encodeCompactUnwindRegistersWithoutFrame() function in lib/Target/X86FrameLowering.cpp for the encoding algorithm.)

Frameless with a Large Constant Stack Size (``EBP`` or ``RBP`` is not used as a frame pointer)

This case is like the “Frameless with a Small Constant Stack Size” case, but the stack size is too large to encode in the compact unwind encoding. Instead it requires that the function contains “subl $nnnnnn, %esp” in its prolog. The compact encoding contains the offset to the $nnnnnn value in the function in bits 9-12 (mask: 0x00001C00).

Note

To Be Written

The code emission step of code generation is responsible for lowering from the code generator abstractions (like MachineFunction, MachineInstr, etc) down to the abstractions used by the MC layer (MCInst, MCStreamer, etc). This is done with a combination of several different classes: the (misnamed) target-independent AsmPrinter class, target-specific subclasses of AsmPrinter (such as SparcAsmPrinter), and the TargetLoweringObjectFile class.

Since the MC layer works at the level of abstraction of object files, it doesn’t have a notion of functions, global variables etc. Instead, it thinks about labels, directives, and instructions. A key class used at this time is the MCStreamer class. This is an abstract API that is implemented in different ways (e.g. to output a .s file, output an ELF .o file, etc) that is effectively an “assembler API”. MCStreamer has one method per directive, such as EmitLabel, EmitSymbolAttribute, switchSection, etc, which directly correspond to assembly level directives.

If you are interested in implementing a code generator for a target, there are three important things that you have to implement for your target:

  1. First, you need a subclass of AsmPrinter for your target. This class implements the general lowering process converting MachineFunction’s into MC label constructs. The AsmPrinter base class provides a number of useful methods and routines, and also allows you to override the lowering process in some important ways. You should get much of the lowering for free if you are implementing an ELF, COFF, or MachO target, because the TargetLoweringObjectFile class implements much of the common logic.
  2. Second, you need to implement an instruction printer for your target. The instruction printer takes an MCInst and renders it to a raw_ostream as text. Most of this is automatically generated from the .td file (when you specify something like “add $dst, $src1, $src2” in the instructions), but you need to implement routines to print operands.
  3. Third, you need to implement code that lowers a MachineInstr to an MCInst, usually implemented in “<target>MCInstLower.cpp”. This lowering process is often target specific, and is responsible for turning jump table entries, constant pool indices, global variable addresses, etc into MCLabels as appropriate. This translation layer is also responsible for expanding pseudo ops used by the code generator into the actual machine instructions they correspond to. The MCInsts that are generated by this are fed into the instruction printer or the encoder.

Finally, at your choosing, you can also implement a subclass of MCCodeEmitter which lowers MCInst’s into machine code bytes and relocations. This is important if you want to support direct .o file emission, or would like to implement an assembler for your target.

A section containing metadata on function stack sizes will be emitted when TargetLoweringObjectFile::StackSizesSection is not null, and TargetOptions::EmitStackSizeSection is set (-stack-size-section). The section will contain an array of pairs of function symbol values (pointer size) and stack sizes (unsigned LEB128). The stack size values only include the space allocated in the function prologue. Functions with dynamic stack allocations are not included.

In a Very Long Instruction Word (VLIW) architecture, the compiler is responsible for mapping instructions to functional-units available on the architecture. To that end, the compiler creates groups of instructions called packets or bundles. The VLIW packetizer in LLVM is a target-independent mechanism to enable the packetization of machine instructions.

Instructions in a VLIW target can typically be mapped to multiple functional units. During the process of packetizing, the compiler must be able to reason about whether an instruction can be added to a packet. This decision can be complex since the compiler has to examine all possible mappings of instructions to functional units. Therefore to alleviate compilation-time complexity, the VLIW packetizer parses the instruction classes of a target and generates tables at compiler build time. These tables can then be queried by the provided machine-independent API to determine if an instruction can be accommodated in a packet.

The packetizer reads instruction classes from a target’s itineraries and creates a deterministic finite automaton (DFA) to represent the state of a packet. A DFA consists of three major elements: inputs, states, and transitions. The set of inputs for the generated DFA represents the instruction being added to a packet. The states represent the possible consumption of functional units by instructions in a packet. In the DFA, transitions from one state to another occur on the addition of an instruction to an existing packet. If there is a legal mapping of functional units to instructions, then the DFA contains a corresponding transition. The absence of a transition indicates that a legal mapping does not exist and that the instruction cannot be added to the packet.

To generate tables for a VLIW target, add TargetGenDFAPacketizer.inc as a target to the Makefile in the target directory. The exported API provides three functions: DFAPacketizer::clearResources(), DFAPacketizer::reserveResources(MachineInstr *MI), and DFAPacketizer::canReserveResources(MachineInstr *MI). These functions allow a target packetizer to add an instruction to an existing packet and to check whether an instruction can be added to a packet. See llvm/CodeGen/DFAPacketizer.h for more information.

Though you’re probably reading this because you want to write or maintain a compiler backend, LLVM also fully supports building a native assembler. We’ve tried hard to automate the generation of the assembler from the .td files (in particular the instruction syntax and encodings), which means that a large part of the manual and repetitive data entry can be factored and shared with the compiler.

Note

To Be Written

Once the instruction is parsed, it enters the MatchInstructionImpl function. The MatchInstructionImpl function performs alias processing and then does actual matching.

Alias processing is the phase that canonicalizes different lexical forms of the same instructions down to one representation. There are several different kinds of alias that are possible to implement and they are listed below in the order that they are processed (which is in order from simplest/weakest to most complex/powerful). Generally you want to use the first alias mechanism that meets the needs of your instruction, because it will allow a more concise description.

The first phase of alias processing is simple instruction mnemonic remapping for classes of instructions which are allowed with two different mnemonics. This phase is a simple and unconditionally remapping from one input mnemonic to one output mnemonic. It isn’t possible for this form of alias to look at the operands at all, so the remapping must apply for all forms of a given mnemonic. Mnemonic aliases are defined simply, for example X86 has:

def : MnemonicAlias<"cbw",     "cbtw">;
def : MnemonicAlias<"smovq",   "movsq">;
def : MnemonicAlias<"fldcww",  "fldcw">;
def : MnemonicAlias<"fucompi", "fucomip">;
def : MnemonicAlias<"ud2a",    "ud2">;

… and many others. With a MnemonicAlias definition, the mnemonic is remapped simply and directly. Though MnemonicAlias’s can’t look at any aspect of the instruction (such as the operands) they can depend on global modes (the same ones supported by the matcher), through a Requires clause:

def : MnemonicAlias<"pushf", "pushfq">, Requires<[In64BitMode]>;
def : MnemonicAlias<"pushf", "pushfl">, Requires<[In32BitMode]>;

In this example, the mnemonic gets mapped into a different one depending on the current instruction set.

The most general phase of alias processing occurs while matching is happening: it provides new forms for the matcher to match along with a specific instruction to generate. An instruction alias has two parts: the string to match and the instruction to generate. For example:

def : InstAlias<"movsx $src, $dst", (MOVSX16rr8W GR16:$dst, GR8  :$src)>;
def : InstAlias<"movsx $src, $dst", (MOVSX16rm8W GR16:$dst, i8mem:$src)>;
def : InstAlias<"movsx $src, $dst", (MOVSX32rr8  GR32:$dst, GR8  :$src)>;
def : InstAlias<"movsx $src, $dst", (MOVSX32rr16 GR32:$dst, GR16 :$src)>;
def : InstAlias<"movsx $src, $dst", (MOVSX64rr8  GR64:$dst, GR8  :$src)>;
def : InstAlias<"movsx $src, $dst", (MOVSX64rr16 GR64:$dst, GR16 :$src)>;
def : InstAlias<"movsx $src, $dst", (MOVSX64rr32 GR64:$dst, GR32 :$src)>;

This shows a powerful example of the instruction aliases, matching the same mnemonic in multiple different ways depending on what operands are present in the assembly. The result of instruction aliases can include operands in a different order than the destination instruction, and can use an input multiple times, for example:

def : InstAlias<"clrb $reg", (XOR8rr  GR8 :$reg, GR8 :$reg)>;
def : InstAlias<"clrw $reg", (XOR16rr GR16:$reg, GR16:$reg)>;
def : InstAlias<"clrl $reg", (XOR32rr GR32:$reg, GR32:$reg)>;
def : InstAlias<"clrq $reg", (XOR64rr GR64:$reg, GR64:$reg)>;

This example also shows that tied operands are only listed once. In the X86 backend, XOR8rr has two input GR8’s and one output GR8 (where an input is tied to the output). InstAliases take a flattened operand list without duplicates for tied operands. The result of an instruction alias can also use immediates and fixed physical registers which are added as simple immediate operands in the result, for example:

// Fixed Immediate operand.
def : InstAlias<"aad", (AAD8i8 10)>;
 
// Fixed register operand.
def : InstAlias<"fcomi", (COM_FIr ST1)>;
 
// Simple alias.
def : InstAlias<"fcomi $reg", (COM_FIr RST:$reg)>;

Instruction aliases can also have a Requires clause to make them subtarget specific.

If the back-end supports it, the instruction printer can automatically emit the alias rather than what’s being aliased. It typically leads to better, more readable code. If it’s better to print out what’s being aliased, then pass a ‘0’ as the third parameter to the InstAlias definition.

Note

To Be Written

This section of the document explains features or design decisions that are specific to the code generator for a particular target.

Tail call optimization, callee reusing the stack of the caller, is currently supported on x86/x86-64, PowerPC, AArch64, and WebAssembly. It is performed on x86/x86-64, PowerPC, and AArch64 if:

  • Caller and callee have the calling convention fastcc, cc 10 (GHC calling convention), cc 11 (HiPE calling convention), tailcc, or swifttailcc.
  • The call is a tail call - in tail position (ret immediately follows call and ret uses value of call or is void).
  • Option -tailcallopt is enabled or the calling convention is tailcc.
  • Platform-specific constraints are met.

x86/x86-64 constraints:

  • No variable argument lists are used.
  • On x86-64 when generating GOT/PIC code only module-local calls (visibility = hidden or protected) are supported.

PowerPC constraints:

  • No variable argument lists are used.
  • No byval parameters are used.
  • On ppc32/64 GOT/PIC only module-local calls (visibility = hidden or protected) are supported.

WebAssembly constraints:

  • No variable argument lists are used
  • The ‘tail-call’ target attribute is enabled.
  • The caller and callee’s return types must match. The caller cannot be void unless the callee is, too.

AArch64 constraints:

  • No variable argument lists are used.

Example:

Call as llc -tailcallopt test.ll.

declare fastcc i32 @tailcallee(i32 inreg %a1, i32 inreg %a2, i32 %a3, i32 %a4)
 
define fastcc i32 @tailcaller(i32 %in1, i32 %in2) {
  %l1 = add i32 %in1, %in2
  %tmp = tail call fastcc i32 @tailcallee(i32 inreg %in1, i32 inreg %in2, i32 %in1, i32 %l1)
  ret i32 %tmp
}

Implications of -tailcallopt:

To support tail call optimization in situations where the callee has more arguments than the caller a ‘callee pops arguments’ convention is used. This currently causes each fastcc call that is not tail call optimized (because one or more of above constraints are not met) to be followed by a readjustment of the stack. So performance might be worse in such cases.

Sibling call optimization is a restricted form of tail call optimization. Unlike tail call optimization described in the previous section, it can be performed automatically on any tail calls when -tailcallopt option is not specified.

Sibling call optimization is currently performed on x86/x86-64 when the following constraints are met:

  • Caller and callee have the same calling convention. It can be either c or fastcc.
  • The call is a tail call - in tail position (ret immediately follows call and ret uses value of call or is void).
  • Caller and callee have matching return type or the callee result is not used.
  • If any of the callee arguments are being passed in stack, they must be available in caller’s own incoming argument stack and the frame offsets must be the same.

Example:

declare i32 @bar(i32, i32)
 
define i32 @foo(i32 %a, i32 %b, i32 %c) {
entry:
  %0 = tail call i32 @bar(i32 %a, i32 %b)
  ret i32 %0
}

The X86 code generator lives in the lib/Target/X86 directory. This code generator is capable of targeting a variety of x86-32 and x86-64 processors, and includes support for ISA extensions such as MMX and SSE.

The following are the known target triples that are supported by the X86 backend. This is not an exhaustive list, and it would be useful to add those that people test.

  • i686-pc-linux-gnu — Linux
  • i386-unknown-freebsd5.3 — FreeBSD 5.3
  • i686-pc-cygwin — Cygwin on Win32
  • i686-pc-mingw32 — MingW on Win32
  • i386-pc-mingw32msvc — MingW crosscompiler on Linux
  • i686-apple-darwin* — Apple Darwin on X86
  • x86_64-unknown-linux-gnu — Linux

The following target-specific calling conventions are known to backend:

  • x86_StdCall — stdcall calling convention seen on Microsoft Windows platform (CC ID = 64).
  • x86_FastCall — fastcall calling convention seen on Microsoft Windows platform (CC ID = 65).
  • x86_ThisCall — Similar to X86_StdCall. Passes first argument in ECX, others via stack. Callee is responsible for stack cleaning. This convention is used by MSVC by default for methods in its ABI (CC ID = 70).

The x86 has a very flexible way of accessing memory. It is capable of forming memory addresses of the following expression directly in integer instructions (which use ModR/M addressing):

SegmentReg: Base + [1,2,4,8] * IndexReg + Disp32

In order to represent this, LLVM tracks no less than 5 operands for each memory operand of this form. This means that the “load” form of ‘mov’ has the following MachineOperands in this order:

Index:        0     |    1        2       3           4          5
Meaning:   DestReg, | BaseReg,  Scale, IndexReg, Displacement Segment
OperandTy: VirtReg, | VirtReg, UnsImm, VirtReg,   SignExtImm  PhysReg

Stores, and all other instructions, treat the four memory operands in the same way and in the same order. If the segment register is unspecified (regno = 0), then no segment override is generated. “Lea” operations do not have a segment register specified, so they only have 4 operands for their memory reference.

x86 has a feature which provides the ability to perform loads and stores to different address spaces via the x86 segment registers. A segment override prefix byte on an instruction causes the instruction’s memory access to go to the specified segment. LLVM address space 0 is the default address space, which includes the stack, and any unqualified memory accesses in a program. Address spaces 1-255 are currently reserved for user-defined code. The GS-segment is represented by address space 256, the FS-segment is represented by address space 257, and the SS-segment is represented by address space 258. Other x86 segments have yet to be allocated address space numbers.

While these address spaces may seem similar to TLS via the thread_local keyword, and often use the same underlying hardware, there are some fundamental differences.

The thread_local keyword applies to global variables and specifies that they are to be allocated in thread-local memory. There are no type qualifiers involved, and these variables can be pointed to with normal pointers and accessed with normal loads and stores. The thread_local keyword is target-independent at the LLVM IR level (though LLVM doesn’t yet have implementations of it for some configurations)

Special address spaces, in contrast, apply to static types. Every load and store has a particular address space in its address operand type, and this is what determines which address space is accessed. LLVM ignores these special address space qualifiers on global variables, and does not provide a way to directly allocate storage in them. At the LLVM IR level, the behavior of these special address spaces depends in part on the underlying OS or runtime environment, and they are specific to x86 (and LLVM doesn’t yet handle them correctly in some cases).

Some operating systems and runtime environments use (or may in the future use) the FS/GS-segment registers for various low-level purposes, so care should be taken when considering them.

An instruction name consists of the base name, a default operand size, and a character per operand with an optional special size. For example:

ADD8rr      -> add, 8-bit register, 8-bit register
IMUL16rmi   -> imul, 16-bit register, 16-bit memory, 16-bit immediate
IMUL16rmi8  -> imul, 16-bit register, 16-bit memory, 8-bit immediate
MOVSX32rm16 -> movsx, 32-bit register, 16-bit memory

The PowerPC code generator lives in the lib/Target/PowerPC directory. The code generation is retargetable to several variations or subtargets of the PowerPC ISA; including ppc32, ppc64 and altivec.

LLVM follows the AIX PowerPC ABI, with two deviations. LLVM uses a PC relative (PIC) or static addressing for accessing global values, so no TOC (r2) is used. Second, r31 is used as a frame pointer to allow dynamic growth of a stack frame. LLVM takes advantage of having no TOC to provide space to save the frame pointer in the PowerPC linkage area of the caller frame. Other details of PowerPC ABI can be found at PowerPC ABI. Note: This link describes the 32 bit ABI. The 64 bit ABI is similar except space for GPRs are 8 bytes wide (not 4) and r13 is reserved for system use.

The size of a PowerPC frame is usually fixed for the duration of a function’s invocation. Since the frame is fixed size, all references into the frame can be accessed via fixed offsets from the stack pointer. The exception to this is when dynamic alloca or variable sized arrays are present, then a base pointer (r31) is used as a proxy for the stack pointer and stack pointer is free to grow or shrink. A base pointer is also used if llvm-gcc is not passed the -fomit-frame-pointer flag. The stack pointer is always aligned to 16 bytes, so that space allocated for altivec vectors will be properly aligned.

An invocation frame is laid out as follows (low memory at top):

Linkage

Parameter area

Dynamic area

Locals area

Saved registers area
 

Previous Frame

The linkage area is used by a callee to save special registers prior to allocating its own frame. Only three entries are relevant to LLVM. The first entry is the previous stack pointer (sp), aka link. This allows probing tools like gdb or exception handlers to quickly scan the frames in the stack. A function epilog can also use the link to pop the frame from the stack. The third entry in the linkage area is used to save the return address from the lr register. Finally, as mentioned above, the last entry is used to save the previous frame pointer (r31.) The entries in the linkage area are the size of a GPR, thus the linkage area is 24 bytes long in 32 bit mode and 48 bytes in 64 bit mode.

32 bit linkage area:

0

Saved SP (r1)

4

Saved CR

8

Saved LR

12

Reserved

16

Reserved

20

Saved FP (r31)

64 bit linkage area:

0

Saved SP (r1)

8

Saved CR

16

Saved LR

24

Reserved

32

Reserved

40

Saved FP (r31)

The parameter area is used to store arguments being passed to a callee function. Following the PowerPC ABI, the first few arguments are actually passed in registers, with the space in the parameter area unused. However, if there are not enough registers or the callee is a thunk or vararg function, these register arguments can be spilled into the parameter area. Thus, the parameter area must be large enough to store all the parameters for the largest call sequence made by the caller. The size must also be minimally large enough to spill registers r3-r10. This allows callees blind to the call signature, such as thunks and vararg functions, enough space to cache the argument registers. Therefore, the parameter area is minimally 32 bytes (64 bytes in 64 bit mode.) Also note that since the parameter area is a fixed offset from the top of the frame, that a callee can access its split arguments using fixed offsets from the stack pointer (or base pointer.)

Combining the information about the linkage, parameter areas and alignment. A stack frame is minimally 64 bytes in 32 bit mode and 128 bytes in 64 bit mode.

The dynamic area starts out as size zero. If a function uses dynamic alloca then space is added to the stack, the linkage and parameter areas are shifted to top of stack, and the new space is available immediately below the linkage and parameter areas. The cost of shifting the linkage and parameter areas is minor since only the link value needs to be copied. The link value can be easily fetched by adding the original frame size to the base pointer. Note that allocations in the dynamic space need to observe 16 byte alignment.

The locals area is where the llvm compiler reserves space for local variables.

The saved registers area is where the llvm compiler spills callee saved registers on entry to the callee.

The llvm prolog and epilog are the same as described in the PowerPC ABI, with the following exceptions. Callee saved registers are spilled after the frame is created. This allows the llvm epilog/prolog support to be common with other targets. The base pointer callee saved register r31 is saved in the TOC slot of linkage area. This simplifies allocation of space for the base pointer and makes it convenient to locate programmatically and during debugging.

Note

TODO - More to come.

The NVPTX code generator under lib/Target/NVPTX is an open-source version of the NVIDIA NVPTX code generator for LLVM. It is contributed by NVIDIA and is a port of the code generator used in the CUDA compiler (nvcc). It targets the PTX 3.0/3.1 ISA and can target any compute capability greater than or equal to 2.0 (Fermi).

This target is of production quality and should be completely compatible with the official NVIDIA toolchain.

Code Generator Options:

Option

Description

sm_20

Set shader model/compute capability to 2.0

sm_21

Set shader model/compute capability to 2.1

sm_30

Set shader model/compute capability to 3.0

sm_35

Set shader model/compute capability to 3.5

ptx30

Target PTX 3.0

ptx31

Target PTX 3.1

Extended BPF (or eBPF) is similar to the original (“classic”) BPF (cBPF) used to filter network packets. The bpf() system call performs a range of operations related to eBPF. For both cBPF and eBPF programs, the Linux kernel statically analyzes the programs before loading them, in order to ensure that they cannot harm the running system. eBPF is a 64-bit RISC instruction set designed for one to one mapping to 64-bit CPUs. Opcodes are 8-bit encoded, and 87 instructions are defined. There are 10 registers, grouped by function as outlined below.

R0        return value from in-kernel functions; exit value for eBPF program
R1 - R5   function call arguments to in-kernel functions
R6 - R9   callee-saved registers preserved by in-kernel functions
R10       stack frame pointer (read only)

eBPF is reusing most of the opcode encoding from classic to simplify conversion of classic BPF to eBPF. For arithmetic and jump instructions the 8-bit ‘code’ field is divided into three parts:

+----------------+--------+--------------------+
|   4 bits       |  1 bit |   3 bits           |
| operation code | source | instruction class  |
+----------------+--------+--------------------+
(MSB)                                      (LSB)

Three LSB bits store instruction class which is one of:

BPF_LD     0x0
BPF_LDX    0x1
BPF_ST     0x2
BPF_STX    0x3
BPF_ALU    0x4
BPF_JMP    0x5
(unused)   0x6
BPF_ALU64  0x7

When BPF_CLASS(code) == BPF_ALU or BPF_ALU64 or BPF_JMP, 4th bit encodes source operand

BPF_X     0x1  use src_reg register as source operand
BPF_K     0x0  use 32 bit immediate as source operand

and four MSB bits store operation code

BPF_ADD   0x0  add
BPF_SUB   0x1  subtract
BPF_MUL   0x2  multiply
BPF_DIV   0x3  divide
BPF_OR    0x4  bitwise logical OR
BPF_AND   0x5  bitwise logical AND
BPF_LSH   0x6  left shift
BPF_RSH   0x7  right shift (zero extended)
BPF_NEG   0x8  arithmetic negation
BPF_MOD   0x9  modulo
BPF_XOR   0xa  bitwise logical XOR
BPF_MOV   0xb  move register to register
BPF_ARSH  0xc  right shift (sign extended)
BPF_END   0xd  endianness conversion

If BPF_CLASS(code) == BPF_JMP, BPF_OP(code) is one of

BPF_JA    0x0  unconditional jump
BPF_JEQ   0x1  jump ==
BPF_JGT   0x2  jump >
BPF_JGE   0x3  jump >=
BPF_JSET  0x4  jump if (DST & SRC)
BPF_JNE   0x5  jump !=
BPF_JSGT  0x6  jump signed >
BPF_JSGE  0x7  jump signed >=
BPF_CALL  0x8  function call
BPF_EXIT  0x9  function return

For load and store instructions the 8-bit ‘code’ field is divided as:

+--------+--------+-------------------+
| 3 bits | 2 bits |   3 bits          |
|  mode  |  size  | instruction class |
+--------+--------+-------------------+
(MSB)                             (LSB)

Size modifier is one of

BPF_W       0x0  word
BPF_H       0x1  half word
BPF_B       0x2  byte
BPF_DW      0x3  double word

Mode modifier is one of

BPF_IMM     0x0  immediate
BPF_ABS     0x1  used to access packet data
BPF_IND     0x2  used to access packet data
BPF_MEM     0x3  memory
(reserved)  0x4
(reserved)  0x5
BPF_XADD    0x6  exclusive add

Two non-generic instructions: (BPF_ABS | <size> | BPF_LD) and (BPF_IND | <size> | BPF_LD) which are used to access packet data. Register R6 is an implicit input that must contain pointer to sk_buff. Register R0 is an implicit output which contains the data fetched from the packet. Registers R1-R5 are scratch registers and must not be used to store the data across BPF_ABS | BPF_LD or BPF_IND | BPF_LD instructions. These instructions have implicit program exit condition as well. When eBPF program is trying to access the data beyond the packet boundary, the interpreter will abort the execution of the program.

BPF_IND | BPF_W | BPF_LD is equivalent to:

R0 = ntohl(*(u32 *) (((struct sk_buff *) R6)->data + src_reg + imm32))

eBPF maps are provided for sharing data between kernel and user-space. Currently implemented types are hash and array, with potential extension to support bloom filters, radix trees, etc. A map is defined by its type, maximum number of elements, key size and value size in bytes. eBPF syscall supports create, update, find and delete functions on maps.

Function call arguments are passed using up to five registers (R1 - R5). The return value is passed in a dedicated register (R0). Four additional registers (R6 - R9) are callee-saved, and the values in these registers are preserved within kernel functions. R0 - R5 are scratch registers within kernel functions, and eBPF programs must therefor store/restore values in these registers if needed across function calls. The stack can be accessed using the read-only frame pointer R10. eBPF registers map 1:1 to hardware registers on x86_64 and other 64-bit architectures. For example, x86_64 in-kernel JIT maps them as

R0 - rax
R1 - rdi
R2 - rsi
R3 - rdx
R4 - rcx
R5 - r8
R6 - rbx
R7 - r13
R8 - r14
R9 - r15
R10 - rbp

since x86_64 ABI mandates rdi, rsi, rdx, rcx, r8, r9 for argument passing and rbx, r12 - r15 are callee saved.

An eBPF program receives a single argument and contains a single eBPF main routine; the program does not contain eBPF functions. Function calls are limited to a predefined set of kernel functions. The size of a program is limited to 4K instructions: this ensures fast termination and a limited number of kernel function calls. Prior to running an eBPF program, a verifier performs static analysis to prevent loops in the code and to ensure valid register usage and operand types.

The AMDGPU code generator lives in the lib/Target/AMDGPU directory. This code generator is capable of targeting a variety of AMD GPU processors. Refer to User Guide for AMDGPU Backend for more information.

 

参考文献链接

https://llvm.org/docs/CodeGenerator.html

 

posted @ 2023-01-03 05:15  吴建明wujianming  阅读(388)  评论(0)  编辑  收藏  举报