You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.

200 lines
9.5 KiB

*************************
Cretonne compared to LLVM
*************************
`LLVM <http://llvm.org>`_ is a collection of compiler components implemented as
a set of C++ libraries. It can be used to build both JIT compilers and static
compilers like `Clang <http://clang.llvm.org>`_, and it is deservedly very
popular. `Chris Lattner's chapter about LLVM
<http://www.aosabook.org/en/llvm.html>`_ in the `Architecture of Open Source
Applications <http://aosabook.org/en/index.html>`_ book gives an excellent
overview of the architecture and design of LLVM.
Cretonne and LLVM are superficially similar projects, so it is worth
highlighting some of the differences and similarities. Both projects:
- Use an ISA-agnostic input language in order to mostly abstract away the
differences between target instruction set architectures.
- Depend extensively on SSA form.
- Have both textual and in-memory forms of their primary intermediate language.
(LLVM also has a binary bitcode format; Cretonne doesn't.)
- Can target multiple ISAs.
- Can cross-compile by default without rebuilding the code generator.
Cretonne's scope is much smaller than that of LLVM. The classical three main
parts of a compiler are:
1. The language-dependent front end parses and type-checks the input program.
2. Common optimizations that are independent of both the input language and the
target ISA.
3. The code generator which depends strongly on the target ISA.
LLVM provides both common optimizations *and* a code generator. Cretonne only
provides the last part, the code generator. LLVM additionally provides
infrastructure for building assemblers and disassemblers. Cretonne does not
handle assembly at all---it only generates binary machine code.
Intermediate representations
============================
LLVM uses multiple intermediate representations as it translates a program to
binary machine code:
`LLVM IR <http://llvm.org/docs/LangRef.html>`_
This is the primary intermediate language which has textual, binary, and
in-memory representations. It serves two main purposes:
- An ISA-agnostic, stable(ish) input language that front ends can generate
easily.
- Intermediate representation for common mid-level optimizations. A large
library of code analysis and transformation passes operate on LLVM IR.
`SelectionDAG <http://llvm.org/docs/CodeGenerator.html#instruction-selection-section>`_
A graph-based representation of the code in a single basic block is used by
the instruction selector. It has both ISA-agnostic and ISA-specific
opcodes. These main passes are run on the SelectionDAG representation:
- Type legalization eliminates all value types that don't have a
representation in the target ISA registers.
- Operation legalization eliminates all opcodes that can't be mapped to
target ISA instructions.
- DAG-combine cleans up redundant code after the legalization passes.
- Instruction selection translates ISA-agnostic expressions to ISA-specific
instructions.
The SelectionDAG representation automatically eliminates common
subexpressions and dead code.
`MachineInstr <http://llvm.org/docs/CodeGenerator.html#machine-code-representation>`_
A linear representation of ISA-specific instructions that initially is in
SSA form, but it can also represent non-SSA form during and after register
allocation. Many low-level optimizations run on MI code. The most important
passes are:
- Scheduling.
- Register allocation.
`MC <http://llvm.org/docs/CodeGenerator.html#the-mc-layer>`_
MC serves as the output abstraction layer and is the basis for LLVM's
integrated assembler. It is used for:
- Branch relaxation.
- Emitting assembly or binary object code.
- Assemblers.
- Disassemblers.
There is an ongoing "global instruction selection" project to replace the
SelectionDAG representation with ISA-agnostic opcodes on the MachineInstr
representation. Some target ISAs have a fast instruction selector that can
translate simple code directly to MachineInstrs, bypassing SelectionDAG when
possible.
:doc:`Cretonne <langref>` uses a single intermediate language to cover these
levels of abstraction. This is possible in part because of Cretonne's smaller
scope.
- Cretonne does not provide assemblers and disassemblers, so it is not
necessary to be able to represent every weird instruction in an ISA. Only
those instructions that the code generator emits have a representation.
- Cretonne's opcodes are ISA-agnostic, but after legalization / instruction
selection, each instruction is annotated with an ISA-specific encoding which
represents a native instruction.
- SSA form is preserved throughout. After register allocation, each SSA value
is annotated with an assigned ISA register or stack slot.
The Cretonne intermediate language is similar to LLVM IR, but at a slightly
lower level of abstraction.
Program structure
-----------------
In LLVM IR, the largest representable unit is the *module* which corresponds
more or less to a C translation unit. It is a collection of functions and
global variables that may contain references to external symbols too.
In Cretonne IL, the largest representable unit is the *function*. This is so
that functions can easily be compiled in parallel without worrying about
references to shared data structures. Cretonne does not have any
inter-procedural optimizations like inlining.
An LLVM IR function is a graph of *basic blocks*. A Cretonne IL function is a
graph of *extended basic blocks* that may contain internal branch instructions.
The main difference is that an LLVM conditional branch instruction has two
target basic blocks---a true and a false edge. A Cretonne branch instruction
only has a single target and falls through to the next instruction when its
condition is false. The Cretonne representation is closer to how machine code
works; LLVM's representation is more abstract.
LLVM uses `phi instructions
<http://llvm.org/docs/LangRef.html#phi-instruction>`_ in its SSA
representation. Cretonne passes arguments to EBBs instead. The two
representations are equivalent, but the EBB arguments are better suited to
handle EBBs that may contain multiple branches to the same destination block
with different arguments. Passing arguments to an EBB looks a lot like passing
arguments to a function call, and the register allocator treats them very
similarly. Arguments are assigned to registers or stack locations.
Value types
-----------
:ref:`Cretonne's type system <value-types>` is mostly a subset of LLVM's type
system. It is less abstract and closer to the types that common ISA registers
can hold.
- Integer types are limited to powers of two from :cton:type:`i8` to
:cton:type:`i64`. LLVM can represent integer types of arbitrary bit width.
- Floating point types are limited to :cton:type:`f32` and :cton:type:`f64`
which is what WebAssembly provides. It is possible that 16-bit and 128-bit
types will be added in the future.
- Addresses are represented as integers---There are no Cretonne pointer types.
LLVM currently has rich pointer types that include the pointee type. It may
move to a simpler 'address' type in the future. Cretonne may add a single
address type too.
- SIMD vector types are limited to a power-of-two number of vector lanes up to
256. LLVM allows an arbitrary number of SIMD lanes.
- Cretonne has no aggregate types. LLVM has named and anonymous struct types as
well as array types.
Cretonne has multiple boolean types, whereas LLVM simply uses `i1`. The sized
Cretonne boolean types are used to represent SIMD vector masks like ``b32x4``
where each lane is either all 0 or all 1 bits.
Cretonne instructions and function calls can return multiple result values. LLVM
instead models this by returning a single value of an aggregate type.
Instruction set
---------------
LLVM has a small well-defined basic instruction set and a large number of
intrinsics, some of which are ISA-specific. Cretonne has a larger instruction
set and no intrinsics. Some Cretonne instructions are ISA-specific.
Since Cretonne instructions are used all the way until the binary machine code
is emitted, there are opcodes for every native instruction that can be
generated. There is a lot of overlap between different ISAs, so for example the
:cton:inst:`iadd_imm` instruction is used by every ISA that can add an
immediate integer to a register. A simple RISC ISA like RISC-V can be defined
with only shared instructions, while an Intel ISA needs a number of specific
instructions to model addressing modes.
Undefined behavior
==================
Cretonne does not generally exploit undefined behavior in its optimizations.
LLVM's mid-level optimizations do, but it should be noted that LLVM's low-level code
generator rarely needs to make use of undefined behavior either.
LLVM provides ``nsw`` and ``nuw`` flags for its arithmetic that invoke
undefined behavior on overflow. Cretonne does not provide this functionality.
Its arithmetic instructions either produce a value or a trap.
LLVM has an ``unreachable`` instruction which is used to indicate impossible
code paths. Cretonne only has an explicit :cton:inst:`trap` instruction.
Cretonne does make assumptions about aliasing. For example, it assumes that it
has full control of the stack objects in a function, and that they can only be
modified by function calls if their address have escaped. It is quite likely
that Cretonne will admit more detailed aliasing annotations on load/store
instructions in the future. When these annotations are incorrect, undefined
behavior ensues.