We introduce a new type of callback, LazyGenericCallback and use it in
websocket initialization.
The basic requirement is that for websocket the place where the callback
is called and where it is constructed are over several functions,
including async functions. To make this generic, we implement
LazyGenericCallback and a GeneralCallbackSetter. The CallbackSetter can
be used to set the callback which then will be transfered to the
LazyGenericCallback which calls it on messages send to it.
Finally, we use this to "Generify" the initialization of WebSockets.
Signed-off-by: Narfinger <Narfinger@users.noreply.github.com>
Testing: We have new tests for LazyGenericCallback and wpt test run
here: https://github.com/Narfinger/servo/actions/runs/21246091767
---------
Signed-off-by: Narfinger <Narfinger@users.noreply.github.com>
Co-authored-by: Jonathan Schwender <55576758+jschwe@users.noreply.github.com>
The frontend connection metadata was not consistently synchronized with
backend object store state during open/upgrade, and upgrade abort did
not fully restore the connection object store set in all cases.
Testing: More indexeddb tests should pass.
part of https://github.com/servo/servo/issues/40983
---------
Signed-off-by: Taym Haddadi <haddadi.taym@gmail.com>
DomRefCell::borrow inside of layout code is risky because it leads to
memory races if the code is ever called from layout worker threads. This
can be caught via TSAN, but we can also catch it deterministically in CI
by using our existing thread state debug assertions correctly.
Testing: Existing WPT test coverage is sufficient.
Fixes: #42962
---------
Signed-off-by: Josh Matthews <josh@joshmatthews.net>
In the WebCrypto modern algorithm specification, the issue
(https://github.com/WICG/webcrypto-modern-algos/issues/47) on algorithm
name referencing in the export key operation of ML-KEM had been resolved
by the following commit in the specification repository:
705f8ec6ce
Our implementation actually matches the new specification. We simply
update the specification text, with some minor refactoring accordingly.
This is equivalent the changes on ML-KEM in #42918 in which I forgot to
include the counterpart for ML-DSA :(
Testing: Refactoring. Existing tests suffice.
Signed-off-by: Kingsley Yung <kingsley@kkoyung.dev>
Testing: New tests start to pass. Most related tests still fail because
we dispatch mouse and pointer events in the wrong order.
---------
Signed-off-by: Simon Wülker <simon.wuelker@arcor.de>
Moves the `Drop` logic for `GPURenderBundle` into a dedicated helper
struct, `DroppableGPURenderBundle`. This change ensures that
`GPURenderBundle` itself does not directly implement `Drop`, aligning
with the policy to prevent direct `Drop` implementations on DOM types.
Removes the `allowDropImpl` configuration for `GPURenderBundle` from
script bindings.
Testing: WebGPU tests cover all related cases
Fixes: Partially #26488
Signed-off-by: Domenico Rizzo <domenico.rizzo@gmail.com>
This changes `Element::selector_flags` from `Cell<ElementSelectorFlags>`
to `AtomicUsize`, thus avoiding the risk of data races.
Testing: Tested manually. We don't run tests using TSAN builds.
Fixes: #37495
Signed-off-by: Oriol Brufau <obrufau@igalia.com>
Adds support for more pointer events: pointerenter, pointerout,
pointerleave, pointerover
Also add global event handlers that were missing.
Testing: WPT expectations are updated.
cc @yezhizhen
Signed-off-by: webbeef <me@webbeef.org>
This change adds very basic support for tab navigation, but without
support for focus scopes. Followup changes will refine the behavior of
this implementation to follow the specification around focus scopes,
shadow DOM, and slots. In particular `delegatesFocus` is not supported
here yet.
Testing: This causes quite a few WPT tests to start passing.
Signed-off-by: Martin Robinson <mrobinson@igalia.com>
Co-authored-by: Oriol Brufau <obrufau@igalia.com>
Moves the `Drop` implementation for `GPUPipelineLayout` to a dedicated
private helper struct, `DroppableGPUPipelineLayout`. This ensures that
DOM types do not directly implement `Drop`, aligning with best practices
for resource management in the DOM. The associated `Bindings.conf` entry
is removed as direct `Drop` is no longer needed for this type.
Testing: WebGpu just coverages cases with proper tests
Fixes: Partially #26488
Signed-off-by: Domenico Rizzo <domenico.rizzo@gmail.com>
When the async parser thread finds a `<meta charset>` tag then it asks
the main thread whether parsing should restart. The main thread wasn't
responding, so the parser got stuck.
Part of https://github.com/servo/servo/issues/37418
Testing: This change adds a test
---------
Signed-off-by: Simon Wülker <simon.wuelker@arcor.de>
This is a preparation for publishing to crates.io. Changes include:
- Add `servo-` prefixes to avoid name collisions on crates.io
- Use `-` instead of `_` in package names.
- Rename the crates to their original names in Cargo.toml,
to keep the diff minimal
- Rename `media` to `servo-media-thread` to avoid name collision with
`servo-media` (originally from the media repository).
This is an outcome of the previous discussion at [#general > Switch
remaining git dependencies to
crates.io](https://servo.zulipchat.com/#narrow/channel/263398-general/topic/Switch.20remaining.20git.20dependencies.20to.20crates.2Eio/with/576336288)
Testing: This should be mostly covered by our CI, but some amount of
breakage is to be expected, since some package names could still be
referenced from scripts which are not tested or run in CI. [mach try
run](https://github.com/jschwe/servo/actions/runs/22502945949)
---------
Signed-off-by: Jonathan Schwender <schwenderjonathan@gmail.com>
The specification of WebCrypto and Modern Algorithms in WebCrypto has
updated to reorder reading algorithm and data arguments in "encrypt",
"decrypt", "sign", "verify", "digest", "importKey", "unwrapKey",
"decapsulateKey" and "decapsulateBits" methods. This patch updates our
implementation accordingly.
The relevant commits in specification repositories:
- WebCrypto:
5b57233c0a
- Modern Algorithms in WebCrypto:
ae72ee6cf4
Testing: Pass new WPT tests added in
https://github.com/web-platform-tests/wpt/pull/57614 and
https://github.com/servo/servo/pull/42925
Signed-off-by: Kingsley Yung <kingsley@kkoyung.dev>
In the WebCrypto modern algorithm specification, the issue
(https://github.com/WICG/webcrypto-modern-algos/issues/47) on algorithm
name referencing in the export key operation of ML-KEM had been resolved
by the following commit in the specification repository:
705f8ec6ce
Our implementation actually matches the new specification. We simply
update the specification text, with some minor refactoring accordingly.
Testing: Refactoring. Existing tests suffice.
Signed-off-by: Kingsley Yung <kingsley@kkoyung.dev>
This is exposes the `tabIndex` property for `HTMLOrSVGElement` according
to the HTML specification. This is the first step toward implementation
of tab navigation in Servo.
Testing: This causes many WPT subtests to start passing, but causes a
few to
start failing. This is likely because we do not have a full
implementation of
the `delegatesFocus` parameter of `attachShadow` yet.
Fixes: This is part of #25001 and #32169.
Signed-off-by: Martin Robinson <mrobinson@igalia.com>
Co-authored-by: Oriol Brufau <obrufau@igalia.com>
`NormalizedAlgorithm::encrypt`, `NormalizedAlgorithm::decrypt` and other
similar functions have some catch-all `match` arms that are unreachable.
They are unreachable because we rely on the name attribute in `String`
of the subtle dictionaries and the `NormalizedAlgorithm` enum to
determine which cryptographic algorithm to use, while the algorithm
normalization mechanism guarantees that some combinations of name and
enum variants won't exist.
This patch tries to get rid of those unreachable `match` arms to make
our WebCrypto code more idiomatic so that the Rust compiler can help us
ensure the correctness in the future.
To achieve this, we break the enum `NormalizedAlgorithm` into multiple
enums: `EncryptAlgorithm`, `DecryptAlgorithm`, `SignAlgorithm`, etc.
Each one is associated to a cryptographic operation, and its variants
are the cryptographic algorithms that support the associated operation.
The inner type of each variant is the desired parameter dictionary.
Therefore, when the call `EncryptAlgorithm::encrypt`,
`DecryptAlgorithm::decrypt` and other similar functions, we can have
`match` statements that cover all patterns since those enums only
contains necessary variants.
To make this change, we also need to change the algorithm registration
mechanism. Instead of using the `SupportedAlgorithm` enum and its method
`SupportedAlgorithm::support` to register the operations of the
algorithms, the algorithm registration is now done in the function
`from_object_value` of a new trait named `NormalizedAlgorithm`, which
the new enums `EncryptAlgorithm`, `DecryptAlgorithm` and so implement.
(Note that the existing enum named `NormalizedAlgorithm` is removed.)
Some refactoring in also done in the `normalize_algorithm` function to
adapt the above changes.
This new design of algorithm registration is also closer to the
WebCrypto specification, as explained in the comment block below the
`normalize_algorithm` function.
The crate `strum` is also used to reduce some boilerplate code.
Testing: Refactoring. Existing tests suffice.
Fixes: Part of #42579
---------
Signed-off-by: Kingsley Yung <kingsley@kkoyung.dev>
Before we were creating a new frame actor each time we paused, even if
the frame object in debugger.js was the same. Now we avoid this by
reusing the same frame actor id.
This helps our upcoming work on onStep, onPop, and onEnterFrame hooks.
Testing: Ran `mach test-devtools` and manually check that it works
Part of: #36027
Signed-off-by: eri <eri@igalia.com>
Co-authored-by: atbrakhi <atbrakhi@igalia.com>
This fixes a crash on the attached testcase when running with the async
html parser enabled. The crash also triggers on wpt. When executing
parse ops received from the async parser, we were not considering the
case where one of these operations causes the parser to abort (in the
case above, said operation is the `<popup-info></popup-info>`).
```html
<body>
<script>
class PopupInfo extends HTMLElement {
connectedCallback() {
document.open();
}
}
customElements.define('popup-info', PopupInfo);
</script>
<popup-info></popup-info>
</body>
```
The bug is very obscure and only triggers if `document.open` is called
while there is no parser-blocking script **and** the document parser is
still active.
Testing: We don't run tests on the async parser yet
Fixes: Part of https://github.com/servo/servo/issues/37418
Signed-off-by: Simon Wülker <simon.wuelker@arcor.de>
This PR fixes two related issues with Content Security Policy (CSP)
nonce validation for external scripts:
1. Missing nonce validation for external scripts with malformed
attributes
2. Incorrect violation event reporting for blocked external resources
This makes servo closer to passing the `nonce-enforce-blocked` wpt test.
The remaining failures are blocked by required changes in the html
parser.
1. Svg script support (https://github.com/servo/html5ever/issues/118)
```html
<svg xmlns="http://www.w3.org/2000/svg">
<script attribute attribute nonce="abc">
t.unreached_func("Duplicate attribute in SVG, no execution.")();
</script>
</svg>
```
2. Duplicate attrs check
the html parser needs to provide this flag, as mentioned on the original
commit message
(4821bc0ab0)
```html
<script attribute attribute nonce="abc">
t.unreached_func("Duplicate attribute, no execution.")();
</script>
<script attribute attribute=<style nonce="abc">
t.unreached_func("2# Duplicate attribute, no execution.")();
</script>
[...]
<script src="../support/nonce-should-be-blocked.js?5" attribute attribute nonce="abc"></script>
```
I've also created a PR to implement the duplicate attrs flag on
html5ever https://github.com/servo/html5ever/pull/695
Testing: doesn't fixes the aforementioned wpt test yet.
Fixes: part of #36437
---------
Signed-off-by: Dyego Aurélio <dyegoaurelio@gmail.com>
Since `evaluate_js_on_global` called `enter_realm` and there were
already some callee that did it, I've actually passed `&mut
CurrentRealm`.
Also converted `Window::WebdriverException` to pass `&mut JSContext`
inside `javascript_error_info_from_error_info`.
Testing: A successful build is enough
Part of #40600
---------
Signed-off-by: Gae24 <96017547+Gae24@users.noreply.github.com>
devtools: Make `why` attribute use `PauseReasons`
Currently we have why attribute hardcoded as per the event. In this PR
we introduce `PAUSE_REASON` and make sure `why` attribute is using
`PAUSE_REASONS`.
This is in order to support the upcoming work on `onStep`, `onPop`, and
`onEnterFrame` hook.
Testing: all existing tests are passing
Fixes: Part of #36027
Signed-off-by: atbrakhi <atbrakhi@igalia.com>
Co-authored-by: eri <eri@igalia.com>
This patch removes `js_*` preferences that are not currently wired up to
the SM engine. The `js_mem_gc_allocation_threshold_factor` was being
used previously but got that usage was commented out in #28092, so the
dead code is removed.
The `dom_worklet_blockingsleep` preference is also renamed to
`dom_worklet_blockingsleep_enabled` to match the declared name in
`PaintWorkletGlobalScope.idl` which also matches the naming convention
of the other the boolean preferences.
Finally, the `network_mime_sniff` is also unused and remmoved in this
patch.
Testing: Should be covered by existing tests.
Fixes: #42868
Signed-off-by: Mukilan Thiyagarajan <mukilan@igalia.com>
When creating a HTML tokenizer then we check the
`dom_servoparser_async_html_tokenizer_enabled` preference and create a
sync or async tokenizer accordingly. We don't do that in `document.open`
though, this change fixes that.
Part of https://github.com/servo/servo/issues/37418
Testing: We don't run tests for the async tokenizer yet
Signed-off-by: Simon Wülker <simon.wuelker@arcor.de>
Moves the `Drop` implementation for `GPUDevice` into a private
`DroppableGPUDevice` struct.
Also updates `Bindings.conf` to prevent the generation of `Drop`
implementations for `GPUDevice`.
Testing: WebGpu tests just cover the cases
Fixes: Partially #26488
Signed-off-by: Domenico Rizzo <domenico.rizzo@gmail.com>
This reverts commit 1f42551042.
This change led to a decrease in Speedometer performance and needs to be
reworked to avoid that.
Testing: This is just a revert.
Signed-off-by: Martin Robinson <mrobinson@igalia.com>
We need a map between frame actor ids (from devtools) and frame objects
(from debugger.js) to implement stepping hooks in the future.
To achieve this, we register the frame actor with a call to the devtools
before entering the event loop when the debugger is paused. Instead of
creating the frame actor in `handle_debugger_pause`, we create it
before.
Testing: It passes existing devtools tests
Part of: #36027
Signed-off-by: eri <eri@igalia.com>
Co-authored-by: atbrakhi <atbrakhi@igalia.com>
This changes does three things:
- Updates all specification links to reflect version 3 of the
specification
- Renames the `IDBDatabase::close` member to
`IDBDatabase::close_pending` to better match the specification
- Adds a `OpenRequest::processed` member, again to better match the
specification
Testing: This should not change behavior, so is covered by existing
IndexedDB WPT tests.
Fixes: Part of #40983
---------
Signed-off-by: Taym Haddadi <haddadi.taym@gmail.com>
Co-authored-by: Martin Robinson <mrobinson@igalia.com>
Remove embedder defined `ScrollEvent` and merge the payload of it to the
`SetScrollStates` message which tell the `ScriptThread` to update the
scroll state.
In the past, `ScrollEvent` is defined as a embedder's `InputEvent` and
being used only to forward the external scroll node id for a node that
is considered scrolled. This make us sends an additional message to
constellation and additionally, `ScrollEvent` went through lengthy
pipelines as it was deemed as an embedder input event, albeit being a
synthetic input fired by `WebviewRenderer`.
Subsequently, we could introduce a flag to detect whether the scrolling
is still ongoing or not (like whether it is still flinging) for
`scrollend` event.
Testing: No WPT changes
---------
Signed-off-by: Jo Steven Novaryo <steven.novaryo@gmail.com>
Signed-off-by: Martin Robinson <mrobinson@igalia.com>
Co-authored-by: Martin Robinson <mrobinson@igalia.com>
Now `ServoTestUtils.forceLayout()` will provide the number of fragments
that have been restyled and rebuilt. This will be useful to test that
incremental layout works well.
Testing: Adds a test using this new API
---------
Signed-off-by: Oriol Brufau <obrufau@igalia.com>
Signed-off-by: Martin Robinson <mrobinson@igalia.com>
Co-authored-by: Luke Warlow <lwarlow@igalia.com>
Co-authored-by: Martin Robinson <mrobinson@igalia.com>
Align with the spec and implement part of its lazy loading
steps. Most tests use non-srcdoc lazy loading URLs, so most
tests will not pass until the other parts are implemented
as well.
Signed-off-by: Tim van der Lippe <tvanderlippe@gmail.com>
`OutsideMarker::repair_style()` was setting `list_item_style` to the
style of the marker, not the style of the list item.
This patch sets it to `node.parent_style(context)` instead, and it fixes
`ServoThreadSafeLayoutNode::parent_style()` to take the pseudo-element
chain into account. This requires modifying a bunch of logic to pass the
`SharedStyleContext` around.
Testing: Adding a new test
Signed-off-by: Oriol Brufau <obrufau@igalia.com>
Currently, `script` and `devtools` use a node's unique id to identify it
across requests. The unique ID is part of a node's rare data field and
is really only meant for debugging. Instantiating it on a node causes
it's memory usage to go up significantly. Now, when the devtools ask for
information about a specific `Node` then they send the unique ID to
`script`, and the script thread then walks the whole DOM tree searching
for that specific ID. This happens here:
6d0b651218/components/script/devtools.rs (L142-L153)
So, in the worst case, all of the nodes in the tree now have a unique
ID. That's not great!
Also, when `script` notifies `devtools` about changes to a DOM node then
`devtools` expects a `NodeActor` to exist for that Node. The actor might
not exist if the inspector doesn't know about that node yet - that
happens when the user hasn't expanded their parent yet.
That is an oversight from https://github.com/servo/servo/pull/42601 and
causes problems now(https://github.com/servo/servo/issues/42784) because
for the longest time, `devtools` was mostly the one sending requests. Of
course, we could make `devtools` simply ignore updates for nodes that it
doesn't know about, but ideally we shouldn't send these updates in the
first place.
This change implements a lookup map on the `ScriptThread` that contains
all nodes that have been sent to the devtools inspector. The map allows
us to efficiently resolve a unique ID to a `Node` in O(1), without
creating unique IDs for the whole tree. It also allows us to only send
DOM updates for nodes that the inspector cares about.
For now, entries from the cache are not evicted unless the relevant
pipeline is closed. That reflects reality, because the inspector also
keeps using them forever.
In the future we will tell the inspector when nodes are removed from the
tree - then it can't interact with them anymore, and we can remove them
from the script-side map.
This is change is not all that complicated but it involves moving a lot
of code around, so feel free to ask for clarification when something is
unclear!
Testing: This change adds a test
Fixes part of https://github.com/servo/servo/issues/42784
---------
Signed-off-by: Simon Wülker <simon.wuelker@arcor.de>
When processing `elementFromPoint` on a `Document` or a `ShadowRoot`,
better handle the situation where the result of the hit test is the root
element, fixing a crash.
Testing: This fixes a crash in the WPT tests.
Signed-off-by: Martin Robinson <mrobinson@igalia.com>
Co-authored-by: Oriol Brufau <obrufau@igalia.com>
Co-authored-by: Luke Warlow <lwarlow@igalia.com>
This is part of replacing can_gc with the new &JSContext/&mut JSContext
method.
We modify CheckValidity and ReportValidity in various HTML Elements to
use the context instead of CanGc.
Testing: Compilation is the test as this is mostly mechanical.
---------
Signed-off-by: Narfinger <Narfinger@users.noreply.github.com>
As started in https://github.com/servo/servo/issues/42168 let's report
memory pressure for webgl objects. CanvasContexts report memory twice:
once because of underlying texture object and once by themself, but
that's okay as we also need to account for swapchain textures.
Computing exact size would be to much work and code so we report rough
estimations.
Testing: None
Signed-off-by: sagudev <16504129+sagudev@users.noreply.github.com>
As mentioned in #40600 we will need this to properly handle &mut
JSContext (we cannot pass it to drop) and I think this wrappers are
preferred ways of doing this in rust as it is safe to forget (not call
drop) which is unsound in this case (due to realm stack at least).
Revieable per commits.
Testing: Just a refactor, but should be covered by WPT
---------
Signed-off-by: sagudev <16504129+sagudev@users.noreply.github.com>
I only wanted to get `&mut JSContext` in microtask chunk and checkpoint,
but this in turn needed `&mut JSContext` in servoparser, which then
caused need for even more changes in script.
I tried to limit the size by putting some `temp_cx` in:
- drops of `LoadBlocker`, `GenericAutoEntryScript`
- methods of `VirtualMethods`
- methods of `FetchResponseListener`
Testing: Just refactor, but should be covered by WPT tests.
Part of #40600
---------
Signed-off-by: sagudev <16504129+sagudev@users.noreply.github.com>
