| Commit message (Collapse) | Author | Age | Files | Lines |
|---|
| |
|
|
|
|
|
|
|
|
|
|
|
|
| |
Previously only WalkerPasses had access to the `getPassRunner` and
`getPassOptions` methods. Move those methods to `Pass` so all passes can use
them. As a result, the `PassRunner` passed to `Pass::run` and
`Pass::runOnFunction` is no longer necessary, so remove it.
Also update `Pass::create` to return a unique_ptr, which is more efficient than
having it return a raw pointer only to have the `PassRunner` wrap that raw
pointer in a `unique_ptr`.
Delete the unused template `PassRunner::getLast()`, which looks like it was
intended to enable retrieving previous analyses and has been in the code base
since 2015 but is not implemented anywhere.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
An overview of this is in the README in the diff here (conveniently, it is near the
top of the diff). Basically, we fix up nn locals after each pass, by default. This keeps
things easy to reason about - what validates is what is valid wasm - but there are
some minor nuances as mentioned there, in particular, we ignore nameless blocks
(which are commonly added by various passes; ignoring them means we can keep
more locals non-nullable).
The key addition here is LocalStructuralDominance which checks which local
indexes have the "structural dominance" property of 1a, that is, that each get has
a set in its block or an outer block that precedes it. I optimized that function quite
a lot to reduce the overhead of running that logic after each pass. The overhead
is something like 2% on J2Wasm and 0% on Dart (0%, because in this mode we
shrink code size, so there is less work actually, and it balances out).
Since we run fixups after each pass, this PR removes logic to manually call the
fixup code from various places we used to call it (like eh-utils and various passes).
Various passes are now marked as requiresNonNullableLocalFixups => false.
That lets us skip running the fixups after them, which we normally do automatically.
This helps avoid overhead. Most passes still need the fixups, though - any pass
that adds a local, or a named block, or moves code around, likely does.
This removes a hack in SimplifyLocals that is no longer needed. Before we
worked to avoid moving a set into a try, as it might not validate. Now, we just do it
and let fixups happen automatically if they need to: in the common code they
probably don't, so the extra complexity seems not worth it.
Also removes a hack from StackIR. That hack tried to avoid roundtrip adding a
nondefaultable local. But we have the logic to fix that up now, and opts will
likely keep it non-nullable as well.
Various tests end up updated here because now a local can be non-nullable -
previous fixups are no longer needed.
Note that this doesn't remove the gc-nn-locals feature. That has been useful for
testing, and may still be useful in the future - it basically just allows nn locals in
all positions (that can't read the null default value at the entry). We can consider
removing it separately.
Fixes #4824
|
| |
|
|
|
|
|
| |
RTTs were removed from the GC spec and if they are added back in in the future,
they will be heap types rather than value types as in our implementation.
Updating our implementation to have RTTs be heap types would have been more work
than deleting them for questionable benefit since we don't know how long it will
be before they are specced again.
|
| |
|
|
|
|
|
|
| |
removeUnneededBlocks() can force us to use a local when we
load the emitted wasm, which can't work for something nondefaultable
like an RTT.
For now, just don't run that optimization if GC is enabled. Eventually,
perhaps we'll want to enable this optimization in a different way.
|
| |
|
|
|
|
|
| |
Some passes need setInfluences but not getInfluences, but were
computing them nonetheless.
This makes e.g. MergeLocals 12% faster. It will also help use LocalGraph
in new passes with less worries about speed.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This adds support for reading/writing of the new `delegate` instruction
in the folded wast format, the stack IR format, the poppy IR format, and
the binary format in Binaryen. We don't have a formal spec written down
yet, but please refer to WebAssembly/exception-handling#137 and
WebAssembly/exception-handling#146 for the informal semantics. In the
current version of spec `delegate` is basically a rethrow, but with
branch-like immediate argument so that it can bypass other
catches/delegates in between.
`delegate` is not represented as a new `Expression`, but it is rather
an option within a `Try` class, like `catch`/`catch_all`.
One special thing about `delegate` is, even though it is written
_within_ a `try` in the folded wat format, like
```wasm
(try
(do
...
)
(delegate $l)
)
```
In the unfolded wat format or in the binary format, `delegate` serves as
a scope end instruction so there is no separate `end`:
```wasm
try
...
delegate $l
```
`delegate` semantically targets an outer `catch` or `delegate`, but we
write `delegate` target as a `try` label because we only give labels to
block-like scoping expressions. So far we have not given `Try` a label
and used inner blocks or a wrapping block in case a branch targets the
`try`. But in case of `delegate`, it can syntactically only target `try`
and if it targets blocks or loops it is a validation failure.
So after discussions in #3497, we give `Try` a label but this label can
only be targeted by `delegate`s. Unfortunately this makes parsing and
writing of `Try` expression somewhat complicated. Also there is one
special case; if the immediate argument of `try` is the same as the
depth of control flow stack, this means the 'delegate' delegates to the
caller. To handle this case this adds a fake label
`DELEGATE_CALLER_TARGET`, and when writing it back to the wast format
writes it as an immediate value, unlike other cases in which we write
labels.
This uses `DELEGATE_FIELD_SCOPE_NAME_DEF/USE` to represent `try`'s label
and `delegate`'s target. There are many cases that `try` and
`delegate`'s labels need to be treated in the same way as block and
branch labels, such as for hashing or comparing. But there are routines
in which we automatically assume all label uses are branches. I thought
about adding a new kind of defines such as
`DELEGATE_FIELD_TRY_NAME_DEF/USE`, but I think it will also involve some
duplication of existing routines or classes. So at the moment this PR
chooses to use the existing `DELEGATE_FIELD_SCOPE_NAME_DEF/USE` for
`try` and `delegate` labels and makes only necessary amount of changes
in branch-utils. We can revisit this decision later if necessary.
Many of changes to the existing test cases are because now all `try`s
are automatically assigned a label. They will be removed in
`RemoveUnusedNames` pass in the same way as block labels if not targeted
by any delegates.
This only supports reading and writing and has not been tested against
any optimization passes yet.
---
Original unfolded wat file to generate test/try-delegate.wasm:
```wasm
(module
(event $e)
(func
try
try
delegate 0
catch $e
end)
(func
try
try
catch $e
i32.const 0
drop
try
delegate 1
end
catch $e
end
)
)
```
|
| |
|
|
|
| |
This adds missing stack IR printing support for the new form of
try-catch-catch_all. Also uses `printMedium` when printing instructions
consistently.
|
| |
|
| |
Also includes a lot of new spec tests that eventually need to go into the spec repo
|
| | |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
That was needed for super-old wasm type system, where we allowed
(block $x
(br_if $x
(unreachable)
(nop)
)
)
That is, we differentiated "taken" branches from "named" ones (just
referred to by name, but not actually taken as it's in unreachable code).
We don't need to differentiate those any more. Remove the ReFinalize
code that considered it, and also remove the named/taken distinction in
other places.
|
| |
|
|
|
|
|
|
|
|
|
|
|
| |
Adds the ability to create multivalue types from vectors of concrete value
types. All types are transparently interned, so their representation is still a
single uint32_t. Types can be extracted into vectors of their component parts,
and all the single value types expand into vectors containing themselves.
Multivalue types are not yet used in the IR, but their creation and inspection
functionality is exposed and tested in the C and JS APIs.
Also makes common type predicates methods of Type and improves the ergonomics of
type printing.
|
| | |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Previously `StackWriter` and its subclasses had routines for all three
modes (`Binaryen2Binary`, `Binaryen2Stack`, and `Stack2Binary`) within a
single class. This splits routines for each in a separate class and
also factors out binary writing into a separate class
(`BinaryInstWriter`) so other classes can make use of it.
The new classes are:
- `BinaryInstWriter`:
Binary instruction writer. Only responsible for emitting binary
contents and no other logic
- `BinaryenIRWriter`: Converts binaryen IR into something else
- `BinaryenIRToBinaryWriter`: Writes binaryen IR to binary
- `StackIRGenerator`: Converts binaryen IR to stack IR
- `StackIRToBinaryWriter`: Writes stack IR to binary
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
- Reflected new renamed instruction names in code and tests:
- `get_local` -> `local.get`
- `set_local` -> `local.set`
- `tee_local` -> `local.tee`
- `get_global` -> `global.get`
- `set_global` -> `global.set`
- `current_memory` -> `memory.size`
- `grow_memory` -> `memory.grow`
- Removed APIs related to old instruction names in Binaryen.js and added
APIs with new names if they are missing.
- Renamed `typedef SortedVector LocalSet` to `SetsOfLocals` to prevent
name clashes.
- Resolved several TODO renaming items in wasm-binary.h:
- `TableSwitch` -> `BrTable`
- `I32ConvertI64` -> `I32WrapI64`
- `I64STruncI32` -> `I64SExtendI32`
- `I64UTruncI32` -> `I64UExtendI32`
- `F32ConvertF64` -> `F32DemoteI64`
- `F64ConvertF32` -> `F64PromoteF32`
- Renamed `BinaryenGetFeatures` and `BinaryenSetFeatures` to
`BinaryenModuleGetFeatures` and `BinaryenModuleSetFeatures` for
consistency.
|
| |
|
| |
Applies the changes in #2065, and temprarily disables the hook since it's too slow to run on a change this large. We should re-enable it in a later commit.
|
| |
|
| |
Mass change to apply clang-format to everything. We are applying this in a PR by me so the (git) blame is all mine ;) but @aheejin did all the work to get clang-format set up and all the manual work to tidy up some things to make the output nicer in #2048
|
| |
|
|
|
|
| |
Automated renaming according to
https://github.com/WebAssembly/spec/issues/884#issuecomment-426433329.
|
| |
|
| |
This now makes --generate-stack-ir --print-stack-ir emit a fully valid .wat wasm file, in stacky format.
|
|
|
This adds a new IR, "Stack IR". This represents wasm at a very low level, as a simple stream of instructions, basically the same as wasm's binary format. This is unlike Binaryen IR which is structured and in a tree format.
This gives some small wins on binary sizes, less than 1% in most cases, usually 0.25-0.50% or so. That's not much by itself, but looking forward this prepares us for multi-value, which we really need an IR like this to be able to optimize well. Also, it's possible there is more we can do already - currently there are just a few stack IR optimizations implemented,
DCE
local2stack - check if a set_local/get_local pair can be removed, which keeps the set's value on the stack, which if the stars align it can be popped instead of the get.
Block removal - remove any blocks with no branches, as they are valid in wasm binary format.
Implementation-wise, the IR is defined in wasm-stack.h. A new StackInst is defined, representing a single instruction. Most are simple reflections of Binaryen IR (an add, a load, etc.), and just pointers to them. Control flow constructs are expanded into multiple instructions, like a block turns into a block begin and end, and we may also emit extra unreachables to handle the fact Binaryen IR has unreachable blocks/ifs/loops but wasm does not. Overall, all the Binaryen IR differences with wasm vanish on the way to stack IR.
Where this IR lives: Each Function now has a unique_ptr to stack IR, that is, a function may have stack IR alongside the main IR. If the stack IR is present, we write it out during binary writing; if not, we do the same binaryen IR => wasm binary process as before (this PR should not affect speed there). This design lets us use normal Passes on stack IR, in particular this PR defines 3 passes:
Generate stack IR
Optimize stack IR (might be worth splitting out into separate passes eventually)
Print stack IR for debugging purposes
Having these as normal passes is convenient as then they can run in parallel across functions and all the other conveniences of our current Pass system. However, a downside of keeping the second IR as an option on Functions, and using normal Passes to operate on it, means that we may get out of sync: if you generate stack IR, then modify binaryen IR, then the stack IR may no longer be valid (for example, maybe you removed locals or modified instructions in place etc.). To avoid that, Passes now define if they modify Binaryen IR or not; if they do, we throw away the stack IR.
Miscellaneous notes:
Just writing Stack IR, then writing to binary - no optimizations - is 20% slower than going directly to binary, which is one reason why we still support direct writing. This does lead to some "fun" C++ template code to make that convenient: there is a single StackWriter class, templated over the "mode", which is either Binaryen2Binary (direct writing), Binaryen2Stack, or Stack2Binary. This avoids a lot of boilerplate as the 3 modes share a lot of code in overlapping ways.
Stack IR does not support source maps / debug info. We just don't use that IR if debug info is present.
A tiny text format comment (if emitting non-minified text) indicates stack IR is present, if it is ((; has Stack IR ;)). This may help with debugging, just in case people forget. There is also a pass to print out the stack IR for debug purposes, as mentioned above.
The sieve binaryen.js test was actually not validating all along - these new opts broke it in a more noticeable manner. Fixed.
Added extra checks in pass-debug mode, to verify that if stack IR should have been thrown out, it was. This should help avoid any confusion with the IR being invalid.
Added a comment about the possible future of stack IR as the main IR, depending on optimization results, following some discussion earlier today.
|
