Initial implementation of WebGLQueries
This patch adds initial support for WeGLQueries. Most related WebGL functions and objects ([1]) are implemented; what's still missing is:
- syncing the result of `getQueryParameter` with the event loop
- `EXT_disjoint_timer_query_webgl2` support
[1]: https://www.khronos.org/registry/webgl/specs/latest/2.0/#3.7.12
<!-- Please describe your changes on the following line: -->
cc @jdm @zakorgy
---
<!-- Thank you for contributing to Servo! Please replace each `[ ]` by `[X]` when the step is complete, and replace `___` with appropriate data: -->
- [x] `./mach build -d` does not report any errors
- [ ] `./mach test-tidy` does not report any errors
- [ ] These changes fix #___ (GitHub issue number if applicable)
<!-- Either: -->
- [x] There are tests for these changes OR
- [ ] These changes do not require tests because ___
<!-- Also, please make sure that "Allow edits from maintainers" checkbox is checked, so that we can help you if you get stuck somewhere along the way.-->
<!-- Pull requests that do not address these steps are welcome, but they will require additional verification as part of the review process. -->
<!-- Reviewable:start -->
---
This change is [<img src="https://reviewable.io/review_button.svg" height="34" align="absmiddle" alt="Reviewable"/>](https://reviewable.io/reviews/servo/servo/24178)
<!-- Reviewable:end -->
Remove redundant webidl_must_inherit compiler plugin lint
At first I was considering moving it to a procedural macro (source-level information should be sufficient), and started by trying to reproduce the error case by changing `htmldivelement.rs` to use `Element` instead of `HTMLElement` as the first field.
The output was:
```rust
error[E0308]: mismatched types
--> /home/simon/servo2/target/debug/build/script-4caa244faca7d10f/out/Bindings/HTMLDivElementBinding.rs:665:31
|
665 | let _: &HTMLElement = self.as_parent();
| ^^^^^^^^^^^^^^^^ expected struct `dom::htmlelement::HTMLElement`, found struct `dom::element::Element`
|
= note: expected type `&dom::htmlelement::HTMLElement`
found type `&dom::element::Element`
```
This line number is inside a generated method called `__assert_parent_type`. As far as I can tell, any case where this lint would error is already caught by such methods. The lint is therefore redundant and can safely be removed.
<!-- Reviewable:start -->
---
This change is [<img src="https://reviewable.io/review_button.svg" height="34" align="absmiddle" alt="Reviewable"/>](https://reviewable.io/reviews/servo/servo/24319)
<!-- Reviewable:end -->
Change setBody to be less strict when getting the immediate child
Step four of the dom document body calls for throwing an exception
if no document element was found. The current setBody implementation
was looking for a document element that is also an html element. It
has been updated to use GetDocumentElement which returns any document
element.
- [X] `./mach build -d` does not report any errors
- [X] `./mach test-tidy` does not report any errors
- [X] These changes fix#24291
<!-- Either: -->
- [X] There are tests for these changes OR
- [ ] These changes do not require tests because ___
<!-- Reviewable:start -->
---
This change is [<img src="https://reviewable.io/review_button.svg" height="34" align="absmiddle" alt="Reviewable"/>](https://reviewable.io/reviews/servo/servo/24316)
<!-- Reviewable:end -->
WebIDL codegen: Replace cmake with a single Python script
When [playing around with Cargo’s new timing visualization](https://internals.rust-lang.org/t/exploring-crate-graph-build-times-with-cargo-build-ztimings/10975/21), I was surprised to see the `script` crate’s build script take 76 seconds. I did not expect WebIDL bindings generation to be *that* computationally intensive.
It turns out almost all of this time is overhead. The build script uses CMake to generate bindings for each WebIDL file in parallel, but that causes a lot of work to be repeated 366 times:
* Starting up a Python VM
* Importing (parts of) the Python standard library
* Importing ~16k lines of our Python code
* Recompiling the latter to bytecode, since we used `python -B` to disable writing `.pyc` files
* Deserializing with `cPickle` and recreating in memory the results of parsing all WebIDL files
----
This commit remove the use of CMake and cPickle for the `script` crate. Instead, all WebIDL bindings generation is done sequentially in a single Python process. This takes 2 to 3 seconds.
<!-- Reviewable:start -->
---
This change is [<img src="https://reviewable.io/review_button.svg" height="34" align="absmiddle" alt="Reviewable"/>](https://reviewable.io/reviews/servo/servo/24303)
<!-- Reviewable:end -->
At first I was considering moving it to a procedural macro
(source-level information should be sufficient),
and started by trying to reproduce the error case
by changing `htmldivelement.rs` to use `Element` instead of `HTMLElement`
as the first field.
The output was:
```rust
error[E0308]: mismatched types
--> /home/simon/servo2/target/debug/build/script-4caa244faca7d10f/out/Bindings/HTMLDivElementBinding.rs:665:31
|
665 | let _: &HTMLElement = self.as_parent();
| ^^^^^^^^^^^^^^^^ expected struct `dom::htmlelement::HTMLElement`, found struct `dom::element::Element`
|
= note: expected type `&dom::htmlelement::HTMLElement`
found type `&dom::element::Element`
```
This line number is inside a generated method called `__assert_parent_type`.
As far as I can tell, any case where this lint would error is already caught
by such methods. The lint is therefore redundant and can safely be removed.
Step four of the dom document body calls for throwing an exception
if no document element was found. The current setBody implementation
was looking for a document element that is also an html element. It
has been updated to use GetDocumentElement which returns any document
element.
Update a bunch of stuff to syn 1
This updates a lot of our crates to syn 1. Not all of them, but syn 1 was already in tree so it doesn't matter much either way.
<!-- Reviewable:start -->
---
This change is [<img src="https://reviewable.io/review_button.svg" height="34" align="absmiddle" alt="Reviewable"/>](https://reviewable.io/reviews/servo/servo/24308)
<!-- Reviewable:end -->
Remove use of packed_simd
This crate takes about 40 seconds to build of a fast desktop, and was used in exactly one place. Additionally, that use was previously accidentally disabled (a553964123) and we only noticed more than a year later when the build broke (d1efad6763).
<!-- Reviewable:start -->
---
This change is [<img src="https://reviewable.io/review_button.svg" height="34" align="absmiddle" alt="Reviewable"/>](https://reviewable.io/reviews/servo/servo/24304)
<!-- Reviewable:end -->
Improve spec compliance of discarding BCs
<!-- Please describe your changes on the following line: -->
Came-up at https://github.com/servo/servo/pull/23637#issuecomment-528254988
---
<!-- Thank you for contributing to Servo! Please replace each `[ ]` by `[X]` when the step is complete, and replace `___` with appropriate data: -->
- [ ] `./mach build -d` does not report any errors
- [ ] `./mach test-tidy` does not report any errors
- [ ] These changes fix #___ (GitHub issue number if applicable)
<!-- Either: -->
- [ ] There are tests for these changes OR
- [ ] These changes do not require tests because ___
<!-- Also, please make sure that "Allow edits from maintainers" checkbox is checked, so that we can help you if you get stuck somewhere along the way.-->
<!-- Pull requests that do not address these steps are welcome, but they will require additional verification as part of the review process. -->
<!-- Reviewable:start -->
---
This change is [<img src="https://reviewable.io/review_button.svg" height="34" align="absmiddle" alt="Reviewable"/>](https://reviewable.io/reviews/servo/servo/24143)
<!-- Reviewable:end -->
When playing around with Cargo’s new timing visualization:
https://internals.rust-lang.org/t/exploring-crate-graph-build-times-with-cargo-build-ztimings/10975/21
… I was surprised to see the `script` crate’s build script take 76 seconds.
I did not expect WebIDL bindings generation to be *that* computationally
intensive.
It turns out almost all of this time is overhead. The build script uses CMake
to generate bindings for each WebIDL file in parallel, but that causes a lot
of work to be repeated 366 times:
* Starting up a Python VM
* Importing (parts of) the Python standard library
* Importing ~16k lines of our Python code
* Recompiling the latter to bytecode, since we used `python -B` to disable
writing `.pyc` file
* Deserializing with `cPickle` and recreating in memory the results
of parsing all WebIDL files
----
This commit remove the use of CMake and cPickle for the `script` crate.
Instead, all WebIDL bindings generation is done sequentially
in a single Python process. This takes 2 to 3 seconds.
This crate takes about 40 seconds to build of a fast desktop,
and was used in exactly one place. Additionally, that use was previously
accidentally disabled (a553964123)
and we only noticed more than a year later when the build broke
(d1efad6763).
Add check for empty data on ImageData constructor.
<!-- Please describe your changes on the following line: -->
This commit implements the nonzero check for ImageData::new_with_jsobject() according to the ImageData spec.
---
<!-- Thank you for contributing to Servo! Please replace each `[ ]` by `[X]` when the step is complete, and replace `___` with appropriate data: -->
- [X] `./mach build -d` does not report any errors
- [X] `./mach test-tidy` does not report any errors
- [X] These changes fix#24274 (GitHub issue number if applicable)
<!-- Either: -->
- [X] There are tests for these changes OR
- [ ] These changes do not require tests because ___
<!-- Also, please make sure that "Allow edits from maintainers" checkbox is checked, so that we can help you if you get stuck somewhere along the way.-->
<!-- Pull requests that do not address these steps are welcome, but they will require additional verification as part of the review process. -->
<!-- Reviewable:start -->
---
This change is [<img src="https://reviewable.io/review_button.svg" height="34" align="absmiddle" alt="Reviewable"/>](https://reviewable.io/reviews/servo/servo/24275)
<!-- Reviewable:end -->
Temporary implementation of Canvas.MeasureText
@jdm: that will help with the babylonjs demos. Another approach is to implement that in JS and add it to the demos.
<!-- Reviewable:start -->
---
This change is [<img src="https://reviewable.io/review_button.svg" height="34" align="absmiddle" alt="Reviewable"/>](https://reviewable.io/reviews/servo/servo/24261)
<!-- Reviewable:end -->
Allow playing video/webm on Windows
- [X] `./mach build -d` does not report any errors
- [X] `./mach test-tidy` does not report any errors
This makes video/webm play on non-UWP Windows. I couldn't find the equivalent gstmatroska.dll on our GStreamer binaries for UWP.
<!-- Reviewable:start -->
---
This change is [<img src="https://reviewable.io/review_button.svg" height="34" align="absmiddle" alt="Reviewable"/>](https://reviewable.io/reviews/servo/servo/24280)
<!-- Reviewable:end -->
Update WebIDL parser
<!-- Please describe your changes on the following line: -->
This includes IDL mixin support.
---
<!-- Thank you for contributing to Servo! Please replace each `[ ]` by `[X]` when the step is complete, and replace `___` with appropriate data: -->
- [x] `./mach build -d` does not report any errors
- [x] `./mach test-tidy` does not report any errors
- [x] These changes does not fix but helps #22539 / fixes#16244
<!-- Either: -->
- [x] There are tests for these changes
<!-- Also, please make sure that "Allow edits from maintainers" checkbox is checked, so that we can help you if you get stuck somewhere along the way.-->
<!-- Pull requests that do not address these steps are welcome, but they will require additional verification as part of the review process. -->
<!-- Reviewable:start -->
---
This change is [<img src="https://reviewable.io/review_button.svg" height="34" align="absmiddle" alt="Reviewable"/>](https://reviewable.io/reviews/servo/servo/23987)
<!-- Reviewable:end -->
Add PerformanceResourceTiming:TimingAllowCheck
Added timing allow check to http_loader.rs in `fn http_network_fetch`.
<!-- Thank you for contributing to Servo! Please replace each `[ ]` by `[X]` when the step is complete, and replace `___` with appropriate data: -->
- [x] `./mach build -d` does not report any errors
- [x] `./mach test-tidy` does not report any errors
- [x] These changes fix#21270
### Things to Do
- [x] Map header values from Timing-Allow-Origin to URL types using [Url::Parse](https://docs.rs/url/2.0.0/url/)
- [x] Check equality of those header URL origin with the origin in question
- [x] Just use url instead of `res.origin`
- [x] Change `.set_attribute(ResourceAttribute::RedirectStart(0))` to `.set_attribute(ResourceAttribute::RedirectStart(RedirectStartValue::Zero))`
- [x] Change `.set_attribute(ResourceAttribute::RedirectEnd(0))` to `.set_attribute(ResourceAttribute::RedirectEnd(RedirectEndValue::Zero))`
- [x] Figure out how to set DomainLookupStart, RequestStart, and ResponseStart without directly passing value as part of enum
- [x] Figure out how to handle lack of DomainLookupEnd (#21260) and SecureConnectionStart (#21268)
- [x] add a flag to ResourceFetchTiming that indicates if the timing check passed, and only update timing attributes if that flag is false
- [x] add function to mark timing test as failed and set all attributes to 0
- [x] resolve compile error regarding move of header_strings variable
- [x] resolve URL parse errors that appear during test execution
- [x] ~~Fix /resource-timing/crossorigin-sandwich-no-TAO.sub.html~~ (Refer to resource_TAO_cross_origin_redirect_chain.html)
- [x] ~~Fix /resource-timing/crossorigin-sandwich-TAO.sub.html~~ (Refer to resource_TAO_cross_origin_redirect_chain.html)
- [x] ~~Fix /resource-timing/resource-reload-TAO.sub.html~~ (Get TIMEOUT, fails on Firefox too?)
- [x] ~~Fix /resource-timing/resource_TAO_cross_origin_redirect_chain.html~~ (Problem seems to lie in loading the iFrame, when the request is made, the URL attached to the PerformanceResourceTiming interface is the initial URL set on the iFrame instead of the URL that is ultimately loaded)
- [x] ~~Fix resource-timing/resource_TAO_multi_wildcard.html~~ (Doesn't work because IMG element doesn't generate HTTP request with Origin field)
- [x] Fix /resource-timing/resource_TAO_match_origin.htm
- [x] Fix /resource-timing/resource_TAO_match_wildcard.htm
- [x] Fix /resource-timing/resource_TAO_multi.htm
- [x] Fix /resource-timing/resource_TAO_wildcard.htm
- [x] Fix /resource-timing/resource_TAO_zero.htm
- [x] Fix /resource-timing/resource_TAO_null.htm
- [x] Fix /resource-timing/resource_TAO_origin.htm (tests for responseStart and domainLookupEnd fail because #21260 and #21271 haven't been resolved)
- [x] Fix /resource-timing/resource_TAO_space.htm
- [x] Fix /resource-timing/resource_TAO_origin_uppercase.htm
- [x] ~~Fix /resource-timing/resource_timing_TAO_cross_origin_redirect.html~~ (Refer to resource_TAO_cross_origin_redirect_chain.html)
- [x] ~~Fix /resource-timing/TAO-case-insensitive-null-opaque-origin.sub.html~~ (TIMEOUT, doesn't seem to parse iFrame SRC correctly? There doesn't seem to be any sign that it makes a request to TAOResponse.py)
- [x] Fix /resource-timing/TAO-crossorigin-port.sub.html
- [x] ~~Fix /resource-timing/TAO-null-opaque-origin.sub.html~~ (Refer to /resource-timing/TAO-crossorigin-port.sub.html)
- [x] /navigation-timing/nav2_test_redirect_chain_xserver_partial_opt_in.html
- [x] /navigation-timing/nav2_test_document_open.html
- [x] /navigation-timing/nav2_test_frame_removed.html
- [x] /performance-timeline/not-clonable.html
- [x] /navigation-timing/nav2_test_redirect_xserver.html
- [x] /resource-timing/resource_connection_reuse.https.html
- [x] /resource-timing/resource_reparenting.html
- [x] /resource-timing/resource_connection_reuse.html
- [x] /resource-timing/resource_script_types.html
- [x] /resource-timing/idlharness.any.html
- [x] /resource-timing/clear_resource_timing_functionality.html
- [x] /resource-timing/idlharness.any.worker.html
- [x] /resource-timing/resource_cached.htm
- [x] /resource-timing/resource_connection_reuse_mixed_content_redirect.html
- [x] /resource-timing/resource_connection_reuse_mixed_content.html
- [x] /resource-timing/resource_timing_buffer_full_when_shrink_buffer_size.html
- [x] /navigation-timing/idlharness.window.html
- [x] /navigation-timing/nav2_test_navigate_iframe.html
- [x] /navigation-timing/nav2_test_navigate_within_document.html
- [x] /resource-timing/resource_reuse.sub.html
- [x] /navigation-timing/nav2_test_instance_accessible_from_the_start.html
- [x] /resource-timing/resource_dedicated_worker.html
- [x] /navigation-timing/unload-event-same-origin-check.html
- [x] /navigation-timing/nav2_test_navigation_type_backforward.html
<!-- Either: -->
- [x] There are tests for these changes OR
- [ ] These changes do not require tests because ___
<!-- Also, please make sure that "Allow edits from maintainers" checkbox is checked, so that we can help you if you get stuck somewhere along the way.-->
<!-- Pull requests that do not address these steps are welcome, but they will require additional verification as part of the review process. -->
<!-- Reviewable:start -->
---
This change is [<img src="https://reviewable.io/review_button.svg" height="34" align="absmiddle" alt="Reviewable"/>](https://reviewable.io/reviews/servo/servo/23873)
<!-- Reviewable:end -->
ISSUE-23995: lazily generate UniqueIds for Nodes
<!-- Please describe your changes on the following line: -->
Updated `Node` so that its `unique_id` is only generated when `unique_id()` is called. This getter now also has assumed the responsibility for saving the `id` to the global hashset on `ScriptThread`.
Felt like I was fighting the borrow-checker a little bit on this one. Definitely a good learning experience for me, but the result strikes me as a little messy?
---
<!-- Thank you for contributing to Servo! Please replace each `[ ]` by `[X]` when the step is complete, and replace `___` with appropriate data: -->
- [X] `./mach build -d` does not report any errors
- [X] `./mach test-tidy` does not report any errors
- [X] These changes fix#23995 (GitHub issue number if applicable)
<!-- Either: -->
- [X] There are tests for these changes OR
- [ ] These changes do not require tests because ___
<!-- Also, please make sure that "Allow edits from maintainers" checkbox is checked, so that we can help you if you get stuck somewhere along the way.-->
<!-- Pull requests that do not address these steps are welcome, but they will require additional verification as part of the review process. -->
<!-- Reviewable:start -->
---
This change is [<img src="https://reviewable.io/review_button.svg" height="34" align="absmiddle" alt="Reviewable"/>](https://reviewable.io/reviews/servo/servo/24217)
<!-- Reviewable:end -->
The purpose of this commit is to expose stub attributes on
PerformanceResourceTiming in order to enable tests to pass as well as
set the type of performance timing entry to navigation when the
initiator is navigation. The purpose of this is to ensure that
document performance entries are correctly classified as navigation
entries rather than resource entries.
Fix immersive mode panic on three.js rollercoaster on hololens
We have some special logic about texture formats when creating drawing buffers for WebGL that needs to be shared with the code that creates a separate framebuffer for immersive mode.
---
- [x] `./mach build -d` does not report any errors
- [x] `./mach test-tidy` does not report any errors
- [x] These changes fix#24083 and fix#20595.
- [x] These changes do not require tests because no CI for hololens; tested manually in the emulator.
<!-- Reviewable:start -->
---
This change is [<img src="https://reviewable.io/review_button.svg" height="34" align="absmiddle" alt="Reviewable"/>](https://reviewable.io/reviews/servo/servo/24242)
<!-- Reviewable:end -->
