Commit graph

21 commits

Author SHA1 Message Date
Andreas Kling
3c74dc9f4d LibJS: Segregate GC-allocated objects by type
This patch adds two macros to declare per-type allocators:

- JS_DECLARE_ALLOCATOR(TypeName)
- JS_DEFINE_ALLOCATOR(TypeName)

When used, they add a type-specific CellAllocator that the Heap will
delegate allocation requests to.

The result of this is that GC objects of the same type always end up
within the same HeapBlock, drastically reducing the ability to perform
type confusion attacks.

It also improves HeapBlock utilization, since each block now has cells
sized exactly to the type used within that block. (Previously we only
had a handful of block sizes available, and most GC allocations ended
up with a large amount of slack in their tails.)

There is a small performance hit from this, but I'm sure we can make
up for it elsewhere.

Note that the old size-based allocators still exist, and we fall back
to them for any type that doesn't have its own CellAllocator.
2023-11-19 12:10:31 +01:00
Linus Groh
6d6ea1fffe LibJS: Add spec comments to FinalizationRegistryPrototype 2023-01-28 00:41:18 +00:00
Andreas Kling
4abdb68655 LibJS: Remove Object(Object& prototype) footgun
This constructor was easily confused with a copy constructor, and it was
possible to accidentally copy-construct Objects in at least one way that
we dicovered (via generic ThrowCompletionOr construction).

This patch adds a mandatory ConstructWithPrototypeTag parameter to the
constructor to disambiguate it.
2022-12-14 15:11:57 +01:00
Linus Groh
0585029c30 LibJS: Add Visitor::visit(GCPtr<T>) and Visitor::visit(NonnullGCPtr<T>)
Let's avoid reaching for ptr() as much as possible.
2022-10-02 23:02:27 +01:00
Andreas Kling
7fce5f102f LibJS: Don't use a Handle<Realm> in FinalizationRegistry
Instead just treat it like a normal GC cell edge.
2022-09-03 00:36:26 +02:00
Linus Groh
d74f8039eb LibJS: Replace GlobalObject with VM in Promise AOs [Part 8/19] 2022-08-23 13:58:30 +01:00
Idan Horowitz
a79796ea4a LibJS: Implement stage 3 proposal FinalizationRegistry changes
Specifically the 'Symbol as WeakMap Keys Proposal'.
2022-06-23 10:57:52 +03:00
Linus Groh
9f3f3b0864 LibJS: Remove implicit wrapping/unwrapping of completion records
This is an editorial change in the ECMA-262 spec, with similar changes
in some proposals.

See:
- https://github.com/tc39/ecma262/commit/7575f74
- https://github.com/tc39/proposal-array-grouping/commit/df899eb
- https://github.com/tc39/proposal-shadowrealm/commit/9eb5a12
- https://github.com/tc39/proposal-shadowrealm/commit/c81f527
2022-05-03 01:09:29 +02:00
Linus Groh
5b48912d35 LibJS: Remove a bunch of gratuitous JS namespace qualifiers 2022-04-03 15:19:33 +01:00
Lenny Maiorani
d00b79568f Libraries: Use default constructors/destructors in LibJS
https://isocpp.github.io/CppCoreGuidelines/CppCoreGuidelines#cother-other-default-operation-rules

"The compiler is more likely to get the default semantics right and
you cannot implement these functions better than the compiler."
2022-03-16 16:19:40 +00:00
Linus Groh
bc183dbbcb LibJS: Replace uses of MarkedValueList with MarkedVector<Value>
This is effectively a drop-in replacement.
2022-02-09 12:25:27 +00:00
Luke Wilde
4c1c6ef91c LibJS: Setup host hooks and have promise jobs work out the realm
This allows the host of LibJS (notably LibWeb in this case) to override
certain functions such as HostEnqueuePromiseJob, so it can do it's own
thing in certain situations. Notably, LibWeb will override
HostEnqueuePromiseJob to put promise jobs on the microtask queue.

This also makes promise jobs use AK::Function instead of
JS::NativeFunction. This removes the need to go through a JavaScript
function and it more closely matches the spec's idea of "abstract
closures"
2022-02-08 17:47:44 +00:00
davidot
f324d91106 LibJS: Convert FinalizationRegistry::cleanup to ThrowCompletionOr 2022-02-08 09:12:42 +00:00
mjz19910
1ef633472b Everywhere: Convert VM::call() to JS::call() 2022-01-23 15:24:45 +00:00
Andreas Kling
83bd675477 LibJS: Make WeakContainer pruning do less work
Instead of iterating *all* swept cells when pruning weak containers,
only iterate the cells actually *in* the container.

Also, instead of compiling a list of all swept cells, we can simply
check the Cell::state() flag to know if something should be pruned.
2021-10-05 18:52:00 +02:00
Andreas Kling
d42d655853 LibJS+LibWeb+Spreadsheet: Upcall visit_edges() via Base typedef
Let's use Base::visit_edges() when calling the base class, to prevent
accidentally skipping over anyone in the inheritance chain.
2021-09-11 14:10:11 +02:00
Andreas Kling
470c99a2a6 LibJS: Tweak the WeakContainer::remove_swept_cells() API a little bit
Make this API take a Span<Cell*> instead of a Vector<Cell*>&.
This is behavior neutral, but stops the API looking like it wants to
do mutable things to the Vector.
2021-09-11 12:09:55 +02:00
Andreas Kling
cd53d046b2 LibJS: Fix typo "sweeped" => "swept" everywhere 2021-06-27 22:40:49 +02:00
Andreas Kling
ba9d5c4d54 LibJS: Rename Function => FunctionObject 2021-06-27 22:36:04 +02:00
Linus Groh
e5753443ae LibJS: Consistently make prototype the last argument in Object ctors
This is so that we can reliably allocate them in a template function,
e.g. in ordinary_create_from_constructor():

    global_object.heap().allocate<T>(
        global_object, forward<Args>(args)..., *prototype);

The majority of objects already take the prototype as the last argument,
so I updated the ones that didn't.
2021-06-20 12:12:39 +02:00
Idan Horowitz
de9fa6622a LibJS: Add the FinalizationRegistry built-in object
As well as the needed functionality in VM to enqueue and run cleanup
jobs for the FinalizationRegistry instances.
2021-06-15 23:59:21 +01:00