This is a big commit that changes the way runtime type information is stored in
the binary. Instead of compressing it and storing it in a number of sidetables,
it is stored similar to how the Go compiler toolchain stores it (but still more
compactly).
This has a number of advantages:
* It is much easier to add new features to reflect support. They can simply
be added to these structs without requiring massive changes (especially in
the reflect lowering pass).
* It removes the reflect lowering pass, which was a large amount of hard to
understand and debug code.
* The reflect lowering pass also required merging all LLVM IR into one
module, which is terrible for performance especially when compiling large
amounts of code. See issue 2870 for details.
* It is (probably!) easier to reason about for the compiler.
The downside is that it increases code size a bit, especially when reflect is
involved. I hope to fix some of that in later patches.
This was actually surprising once I got TinyGo to build on Windows 11
ARM64. All the changes are exactly what you'd expect for a new
architecture, there was no special weirdness just for arm64.
Actually getting TinyGo to build was kind of involved though. The very
short summary is: install arm64 versions of some pieces of software
(like golang, cmake) instead of installing them though choco. In
particular, use the llvm-mingw[1] toolchain instead of using standard
mingw.
[1]: https://github.com/mstorsjo/llvm-mingw/releases
These are reserved registers on MacOS so don't need to be clobbered, and
result in a compiler waring:
ld.lld-14: warning: inline asm clobber list contains reserved registers: X18, FP
Reserved registers on the clobber list may not be preserved across the asm statement, and clobbering them may lead to undefined behaviour.
You can see that it works with the following command:
tinygo run -target=simavr ./testdata/recover.go
This also gets the following tests to pass again:
go test -run=Build -target=simavr -v
Adding support for AVR was a bit more compliated because it's also
necessary to save and restore the Y register.
For example, this commit moves the 'throw' branch of an assertion (nil
check, slice index check, etc) to the end of the function while
inserting the "continue" branch right after the insert location. This
makes the resulting IR easier to follow.
For some reason, this also reduces code size a bit on average. The
TinyGo smoke tests saw a reduction of 0.22%, mainly from WebAssembly.
The drivers repo saw little average change in code size (-0.01%).
This commit also adds a few compiler tests for the defer keyword.
This removes the parentHandle argument from the internal calling convention.
It was formerly used to implment coroutines.
Now that coroutines have been removed, it is no longer necessary.
This layout parameter is currently always nil and ignored, but will
eventually contain a pointer to a memory layout.
This commit also adds module verification to the transform tests, as I
found out that it didn't (and therefore didn't initially catch all
bugs).
This commit simplifies the IR a little bit: instead of calling
pseudo-functions runtime.interfaceImplements and
runtime.interfaceMethod, real declared functions are being called that
are then defined in the interface lowering pass. This should simplify
the interaction between various transformation passes. It also reduces
the number of lines of code, which is generally a good thing.
This commit switches from the previous behavior of compiling the whole
program at once, to compiling every package in parallel and linking the
LLVM bitcode files together for further whole-program optimization.
This is a small performance win, but it has several advantages in the
future:
- There are many more things that can be done per package in parallel,
avoiding the bottleneck at the end of the compiler phase. This
should speed up the compiler futher.
- This change is a necessary step towards a non-LTO build mode for
fast incremental builds that only rebuild the changed package, when
compiler speed is more important than binary size.
- This change refactors the compiler in such a way that it will be
easier to inspect the IR for one package only. Inspecting this IR
will be very helpful for compiler developers.
Moving settings to a separate config struct has two benefits:
- It decouples the compiler a bit from other packages, most
importantly the compileopts package. Decoupling is generally a good
thing.
- Perhaps more importantly, it precisely specifies which settings are
used while compiling and affect the resulting LLVM module. This will
be necessary for caching the LLVM module.
While it would have been possible to cache without this refactor, it
would have been very easy to miss a setting and thus let the
compiler work with invalid/stale data.
This package was long making the design of the compiler more complicated
than it needs to be. Previously this package implemented several
optimization passes, but those passes have since moved to work directly
with LLVM IR instead of Go SSA. The only remaining pass is the SimpleDCE
pass.
This commit removes the *ir.Function type that permeated the whole
compiler and instead switches to use *ssa.Function directly. The
SimpleDCE pass is kept but is far less tightly coupled to the rest of
the compiler so that it can easily be removed once the switch to
building and caching packages individually happens.
This change extends defer support to all supported builitin functions.
Not all of them make sense (such as len, cap, real, imag, etc) but this
change for example adds support for `defer(delete(m, key))` which is
used in the Go 1.15 encoding/json package.
Previously, the typecode was passed via a direct reference, which results in invalid IR when the defer is not reached in all return paths.
It also results in incorrect behavior if the defer is in a loop, causing all defers to use the typecode of the last iteration.
This is a fairly big commit, but it actually changes very little.
getValue should really be a property of the builder (or frame), where
the previously created instructions are kept.
Previously using defer with an exported function generated an invalid function call due to differences between TinyGo's calling convention and the C calling convention.
This code is required by transformation passes which are being moved
into a separate package, but is too complicated to simply copy.
Therefore, I decided to move them into a new package.
In Go, it is not possible to construct pointers that are out of bounds
(and not null), so let LLVM know about this fact.
This leads to a significant code size reduction, around 3% in many
cases.
Most of these errors are actually "todo" or "unimplemented" errors, so
the return type is known. This means that compilation can proceed (with
errors) even though the output will be incorrect. This is useful because
this way, all errors in a compilation unit can be shown together to the
user.
This commit adds getValue which gets a const, global, or result of a
local SSA expression and replaces (almost) all uses of parseExpr with
getValue. The only remaining use is in parseInstr, which makes sure an
instruction is only evaluated once.
This commit replaces "unknown type" errors in getLLVMType with panics.
The main reason this is done is that it simplifies the code *a lot*.
Many `if err != nil` lines were there just because of type information.
Additionally, simply panicking is probably a better approach as the only
way this error can be produced is either with big new language features
or a serious compiler bug. Panicking is probably a better way to handle
this error anyway.
The LLVM library we use does not (yet) provide a llvm.Zero (like it
provides a llvm.Undef) so we have implemented our own. However, in
theory it might return an error in some cases.
No real-world errors have been seen in a while and errors would likely
indicate a serious compiler bug anyway (not an external error), so make
it panic instead of returning an error.
Before this commit, goroutine support was spread through the compiler.
This commit changes this support, so that the compiler itself only
generates simple intrinsics and leaves the real support to a compiler
pass that runs as one of the TinyGo-specific optimization passes.
The biggest change, that was done together with the rewrite, was support
for goroutines in WebAssembly for JavaScript. The challenge in
JavaScript is that in general no blocking operations are allowed, which
means that programs that call time.Sleep() but do not start goroutines
also have to be scheduled by the scheduler.
This reduces complexity in the compiler without affecting binary sizes
too much.
Cortex-M0: no changes
Linux x64: no changes
WebAssembly: some testcases (calls, coroutines, map) are slightly bigger
Implement defer in a different way, which results in smaller binaries.
The binary produced from testdata/calls.go (the only test case with
defer) is reduced a bit in size, but the savings in bytes greatly vary
by architecture:
Cortex-M0: -96 .text / flash
WebAssembly: -215 entire file
Linux x64: -32 .text
Deferred functions in TinyGo were implemented by creating a linked list
of struct objects that contain a function pointer to a thunk, a pointer
to the next object, and a list of parameters. When it was time to run
deferred functions, a helper runtime function called each function
pointer (the thunk) with the struct pointer as a parameter. This thunk
would then in turn extract the saved function parameter from the struct
and call the real function.
What this commit changes, is that the loop to call deferred functions is
moved into the end of the function (practically inlining it) and
replacing the thunks with direct calls inside this loop. This makes it
much easier for LLVM to perform all kinds of optimizations like inlining
and dead argument elimination.
