| Commit message (Collapse) | Author | Age | Files | Lines |
|---|
| |
|
|
|
| |
We emit a select between two objects when only two objects exist of a
particular type. However, if the field is packed, we did not handle truncating the
written values.
|
| |
|
|
|
|
|
|
|
| |
(#6709)
Previously the replacement select got the debug info, but we should also copy it
to the values, as often optimizations lead to one of those values remaining by
itself.
Similar to #6652 in general form.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
(#6688)
If we have
(global $g (struct.new $S
(i32.const 1)
(struct.new $T ..)
(ref.func $f)
))
then before this PR if we wanted to read the middle field we'd stop, as it is non-constant.
However, we can un-nest it, making it constant:
(global $g.unnested (struct.new $T ..))
(global $g (struct.new $S
(i32.const 1)
(global.get $g.unnested)
(ref.func $f)
))
Now the field is a global.get of an immutable global, which is constant. Using this
technique we can handle anything in a struct field, constant or not. The cost of adding
a global is likely offset by the benefit of being able to refer to it directly, as that opens
up more opportunities later.
Concretely, this replaces the constant values we look for in GSI with a variant over
constants or expressions (we do still want to group constants, as multiple globals
with the same constant field can be treated as a whole). And we note cases where we
need to un-nest, and handle those at the end.
|
| |
|
|
| |
This is achieved by simply replacing the Literal with PossibleConstantValues, which
supports both Literals and Globals.
|
| |
|
|
| |
A later PR will add getSuperType which will mean "get the general super type -
either declared, or not".
|
| | |
|
| |
|
|
|
|
|
|
| |
None of that code is speed-sensitive, or at least doesn't need to be inlined to be
fast. Move it to cpp for faster compile times.
This caused a cascade of necessary header fixes (i.e. after removing unneeded
header inclusions in module-utils.h, files that improperly depended on that
stopped working and needed an added include).
|
| |
|
| |
Fixes #5406
|
| |
|
|
|
|
|
| |
After making the pass error when not in closed world, some testcases
required changes. One forward-looking testcase can just be removed - the
point of it was to see what happens if a type escapes, and now closed-world
errors on it - and another testcase needed to avoid types on the boundary,
and to use anyref+casts.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
We have a data structure there, typeGlobals, which maps types to the list of
globals for that type. Previously we used the convention of not having an
entry in the map to mean that a type is unoptimizable. However, this was not
used consistently, and in fact one place could insert to the map in a dangerous
way: a subtype's global is added to the list of globals of the super, and
typeGlobals[super].add(sub-global) would then effectively make an unoptimizable
super into an optimizable one.
To fix that, check for unoptimizability before propagating sub-globals. We do
still use the convention of not keeping data in typeGlobals for unoptimizable
things as it is a minor optimization to avoid wasted work.
Fixes #5379
|
| |
|
|
|
|
|
| |
If a global's type is not fully refined, then when --gsi replaces a reference with
a global.get, we end up with a type that might not be good enough. For example,
if the type is any then it is not a subtype of eq and we can't do ref.eq on it, which
this pass requires. We also can't just do struct.get on it if it is a too-distant parent
or such.
|
| |
|
|
| |
Equirecursive is no longer standards track and its implementation is extremely
complex. Remove it.
|
| |
|
|
|
|
|
|
|
|
|
|
| |
#5253 handled the case of just one possible global. It is also possible we have
multiple globals but just one value. This handles that case. (It slightly overlaps
with other passes, but as this pass actually identifies the creations of the objects
in globals, it has a guarantee of success that the others don't, and it is very easy
to just do given all the work done to handle the case of 2 values).
Also fix a minor bug in #5253 - we need to trap if the old reference were null.
That is, we know the reference must point to the only object ever created of
that type, but that is only if it is not null; if it's null we need to trap.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Expand GlobalStructInference to operate on cases with a single possible global, and
not just 2 or more. Even the case of a single global is useful, it turns out, as we can
alter the reference in places like this:
(struct.get $type 0
(..ref..)
)
No matter what ref is, if there is a single global it must refer to, we can switch to
this:
(struct.get $type 0
(global.get $global)
)
That can unlock further opts later. Note that we can do this even if we don't know
what the value actually is - we may not know what the struct.get returns, but we
do know what it reads from.
|
| | |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
| |
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.
|
| |
|
|
|
| |
We append to vectors of globals in a nondeterministically-ordered loop, which can lead to
different orderings of the vectors. This happens quite frequently in very large J2Wasm
files it turns out. As a solution, simply sort them after the nondeterministic stage.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
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
|
| |
|
|
|
|
|
|
|
|
|
| |
Nominal types don't make much sense without GC, and in particular trying to emit
them with typed function references but not GC enabled can result in invalid
binaries because nominal types do not respect the type ordering constraints
required by the typed function references proposal. Making this change was
mostly straightforward, but required fixing the fuzzer to use --nominal only
when GC is enabled and required exiting early from nominal-only optimizations
when GC was not enabled.
Fixes #4756.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
In GSI we look for a read of a global in a situation like this:
$global1: value1
$global2: value2
(struct.get $Type (ref))
If global inference shows this get must be of either $global1 or $global2, then we
can optimize to this:
(ref) == $global1 ? value1 : value2
We focus on the case of two values because 1 is handled by other passes, and >2
makes the tradeoffs less clear.
However, a simple extension is the case where there are more than 2 globals, but
there are only two values, and one value is unique to one global:
$global1: valueA
$global2: valueB
$global3: valueA
=>
(ref) == $global2 ? valueB : valueA
We can still use a single comparison here, on the global that has the
unique value. Then the else will handle all the other globals.
This increases the cases that GSI can optimize J2Wasm output by over 50%.
|
|
|
This optimizes constants in the megamorphic case of two: when we
know two function references are possible, we could in theory emit this:
(select
(ref.func A)
(ref.func B)
(ref.eq
(..ref value..) ;; globally, only 2 things are possible here, and one has
;; ref.func A as its value, and the other ref.func B
(ref.func A))
That is, compare to one of the values, and emit the two possible values there.
Other optimizations can then turn a call_ref on this select into an if over
two direct calls, leading to devirtualization.
We cannot compare a ref.func directly (since function references are not
comparable), and so instead we look at immutable global structs. If we
find a struct type that has only two possible values in some field, and
the structs are in immutable globals (which happens in the vtable case
in j2wasm for example), then we can compare the references of the struct
to decide between the two values in the field.
|
