Instead of telling the Constellation to tell the embedder that new
frames are ready, have the compositor tell the embedder directly. This
should reduce frame latency. Now, after processing compositor
updates, run any pending `WebView::new_frame_ready` delegate methods.
This change also removes the `refresh` call from the Java interface as
that was the only other place that the compositor was rendering the
WebRender scene outside of event looping spinning. This `refresh` call
was completely unused.
Signed-off-by: Martin Robinson <mrobinson@igalia.com>
In addition to some small API changes, this downstream version of
WebRender no longer depends on a very old version of time. This is the
last step toward removing the dependency on `time0.1`.
The review for this commit should also include: 9f552bebab
Signed-off-by: Martin Robinson <mrobinson@igalia.com>
There are a few methods are still difficult to implement without
the help of surfman. To simplify the trait methods, all methods that
return surfman types are removed. They are either handled by
embedders themselves or abstract to simpler types that servo
components need. The most noticeable changes are:
- Methods related to native surface are moved to servo_glue. The
embedder should decide when to remove/replace the surface and it's
outside of servo's scope.
- Methods required by servo media now return exact media types for it.
The other major change is sevevral difficult trait methods that are
reuiqred by WebGL and Servo media have default implementation. So they
can be optional for users to implement.
Signed-off-by: Wu Wayne <yuweiwu@pm.me>
At some point in the past this message was only sent from the
`Constellation` to `script`, but nowadays this is sent from various
parts of servo to the `ScriptThread`, so this is a better name. In
particular, the current name makes it seeem like this message controls
the `Constellation`, which it does not.
Signed-off-by: Martin Robinson <mrobinson@igalia.com>
`EmbedderMsg` was previously paired with an implicit
`Option<WebViewId>`, even though almost all variants were either always
`Some` or always `None`, depending on whether there was a `WebView
involved.
This patch adds the `WebViewId` to as many `EmbedderMsg` variants as
possible, so we can call their associated `WebView` delegate methods
without needing to check and unwrap the `Option`. In many cases, this
required more changes to plumb through the `WebViewId`.
Notably, all `Request`s now explicitly need a `WebView` or not, in order
to ensure that it is passed when appropriate.
Signed-off-by: Delan Azabani <dazabani@igalia.com>
Co-authored-by: Martin Robinson <mrobinson@igalia.com>
This removes all uses of `EmbedderEvent` in the desktop servoshell to
use the new `WebView` API -- filling it out when necessary.
Signed-off-by: Martin Robinson <mrobinson@igalia.com>
Co-authored-by: Delan Azabani <dazabani@igalia.com>
Many types used directly in the `libservo` API are in the
`script_traits` crate, which was created to break circular dependencies.
Move all API exposed types to `embedder_traits` which now contains types
exposed via the `libservo` embedding API. Also expose these at the root
of the `libservo` `servo` crate so that the API won't break when they
move around in the future.
The idea with `embedder_traits` in the future is that it contains types
that are available throughout servo because they are used in the
embedding API and thus should have minimal dependencies on other Servo
crates (a bit like `base`).
Signed-off-by: Martin Robinson <mrobinson@igalia.com>
This patch introduces a new handle-based webview API to libservo, with
two main design goals:
1. The lifetime of the handles controls the lifetime of the webview,
giving the embedder full control over exactly when webviews are
created and destroyed. This is consistent with how WebKitGTK’s
WebView works; the engine can only create webviews via a create
request, and can only destroy them via a close request.
2. All methods are infallible; if the constellation dies, the embedder
finds out when calling Servo::handle_events.
For the moment, the embedder is only responsible for creating the
WebView id, and not the internal TopLevelBrowsingContext data
structures. This is so that the ScriptThread is able to get a handle on
the new WebView's WindowProxy in the case that it's an auxiliary
browsing context. In the future, the embedder should also be responsible
for creating the TopLevelBrowsingContext and the ScriptThread should
have mechanism to associate the two views so that WebView creation is
always executed through the same code path in the embedding layer. For
now, it's enough that the embedder can get a handle to the new WebView
when it's creation is requested.
Once we replace EmbedderMsg with a webview delegate trait, we will pass
WebView handles to the embedder, rather than webview ids. We’ll also add
detailed docs, once the design settles.
Signed-off-by: Delan Azabani <dazabani@igalia.com>
Co-authored-by: Martin Robinson <mrobinson@igalia.com>
Flatten and simplify Servo's preferences code. In addition, have both
preferences and options passed in as arguments to `Servo::new()` and
make sure not to use the globally set preferences in `servoshell` (as
much as possible now).
Instead of a complex procedural macro to generate preferences, just
expose a very simple derive macro that adds string based getters and
setters.
- All command-line parsing is moved to servoshell.
- There is no longer the concept of a missing preference.
- Preferences no longer have to be part of the resources bundle because
they now have reasonable default values.
- servoshell specific preferences are no longer part of the preferences
exposed by the Servo API.
Signed-off-by: Martin Robinson <mrobinson@igalia.com>
Not all pending display list epochs are painted immediately or at all.
Soemtimes WebRender is one or more display lists behind the latest and
sometimes an epoch isn't painted at all if a new display list arrives
before the old one is painted. These situations aren't errors, but they
might indicate performance issues.
Instead of forever printing warnings when this happens, properly clean
up pending paint metrics after compositing. If earlier pending paint
metrics for a pipeline exist, remove them. If a pipeline is no longer
known by the compositor, remove its pending paint metrics.
This change fixes situations where warnings flood the output and a small
memory leak.
Signed-off-by: Martin Robinson <mrobinson@igalia.com>
Co-authored-by: Mukilan Thiyagarajan <mukilan@igalia.com>
* Clean up tracing instrumentation
Signed-off-by: Delan Azabani <dazabani@igalia.com>
* Set all tracing spans to trace level for now
Signed-off-by: Delan Azabani <dazabani@igalia.com>
---------
Signed-off-by: Delan Azabani <dazabani@igalia.com>
* Add simple fling implementation
Add a simple fling implementation, which depends on a refresh tick from the
embedder.
Currently this refresh tick is only implemented for OpenHarmony
(using the vsync signal).
The fling implementation is very simple, without any fancy things like acceleration.
This can be improved in the future.
Signed-off-by: Jonathan Schwender <schwenderjonathan@gmail.com>
Signed-off-by: Jonathan Schwender <jonathan.schwender@huawei.com>
* Multiply initial velocity with 2
This makes the experience much more snappy.
Signed-off-by: Jonathan Schwender <jonathan.schwender@huawei.com>
* address review comments
Signed-off-by: Jonathan Schwender <schwenderjonathan@gmail.com>
* Rename constants and add todo
Signed-off-by: Jonathan Schwender <schwenderjonathan@gmail.com>
* fmt
Signed-off-by: Jonathan Schwender <schwenderjonathan@gmail.com>
* Add a few periods to make comments consistent
Signed-off-by: Martin Robinson <mrobinson@igalia.com>
---------
Signed-off-by: Jonathan Schwender <schwenderjonathan@gmail.com>
Signed-off-by: Jonathan Schwender <jonathan.schwender@huawei.com>
Signed-off-by: Martin Robinson <mrobinson@igalia.com>
Co-authored-by: Martin Robinson <mrobinson@igalia.com>
Instead of exposing many different kinds of messages to the compositor
that are routed through the constellation, expose a single message type
which can be sent across IPC channels. In addition, this IPC channel and
the route to the crossbeam channel with the compositor is created along
with the `CompositorProxy`, simplifying what needs to be passed around
during pipeline initialization.
Previously, some image updates (from video) were sent over IPC with a
special serialization routine and some were sent via crossbeam channels
(canvas). Now all updates go over the IPC channel `IpcSharedMemory` is
used to avoid serialization penalties. This should improve performance
and reduce copies for video, but add a memory copy overhead for canvas.
This will improve in the future when canvas renders directly into a
texture.
All-in-all this is a simplification which opens the path toward having a
standard compositor API and reduces the number of duplicate messages and
proxying that had to happen in libservo.
Signed-off-by: Martin Robinson <mrobinson@igalia.com>
Improves the instrumentation to skip all function arguments and also add
spans for some layout modes. This is preparation for improving the
performance of flexbox.
Signed-off-by: Martin Robinson <mrobinson@igalia.com>
Instead of exposing many different kinds of messages to the compositor
that are routed through the constellation, expose a single message type
which can be sent across IPC channels. In addition, this IPC channel and
the route to the crossbeam channel with the compositor is created along
with the `CompositorProxy`, simplifying what needs to be passed around
during pipeline initialization.
Previously, some image updates (from video) were sent over IPC with a
special serialization routine and some were sent via crossbeam channels
(canvas). Now all updates go over the IPC channel `IpcSharedMemory` is
used to avoid serialization penalties. This should improve performance
and reduce copies for video, but add a memory copy overhead for canvas.
This will improve in the future when canvas renders directly into a
texture.
All-in-all this is a simplification which opens the path toward having a
standard compositor API and reduces the number of duplicate messages and
proxying that had to happen in libservo.
Signed-off-by: Martin Robinson <mrobinson@igalia.com>
* Add document id to NewWebRenderFrame variant
Signed-off-by: Wu Wayne <yuweiwu@pm.me>
* Match the arguments order
Signed-off-by: Wu Wayne <yuweiwu@pm.me>
---------
Signed-off-by: Wu Wayne <yuweiwu@pm.me>
Instead of a blocking a layout thread on the generation of WebRender
`FontKey`s and `FontInstanceKey`s, generate the keys ahead of time and
send the font data to WebRender asynchronously. This has the benefit of
allowing use of the font much more quickly in layout, though blocking
display list sending itself on the font data upload.
In order to make this work for web fonts, `FontContext` now asks the
`SystemFontService` for a `FontKey`s and `FontInstanceKey`s for new web
fonts. This should happen much more quickly as the `SystemFontService`
is only blocking in order to load system fonts into memory now. In
practice this still drops layout thread blocking to fractions of a
millisecond instead of multiple milliseconds as before.
In addition, ensure that we don't send font data or generate keys for
fonts that are used in layout but never added to display lists. This
should help to reduce memory usage and increase performance.
Performance of this change was verified by putting a microbenchmark
around `FontContext::create_font` which is what triggered font key
generation.
Signed-off-by: Martin Robinson <mrobinson@igalia.com>
This is done by no longer forwarding compositor-bound messages through
SystemFontService and making `FontContext` non-generic:
- Messages from the `FontContext` to the `Compositor` no longer need to be
forwarded through the `SystemFontService`. Instead send these messages
directly through the script IPC channel to the `Compositor`.
- Instead of adding a mock `SystemFontServiceProxy`, simply implement a
mock `SystemFontService` on the other side of an IPC channel in the
`font_context` unit test. This allows making `FontContext`
non-generic, greatly simplifying the code. The extra complexity moves
into the unit test.
These changes necessitate adding a new kind of `FontIdentifier`,
`FontIdentifier::Mock` due to the fact that local fonts have
platform-specific identifiers. This avoids having to pretend like the
system font service can have web fonts -- which was always a bit of a
hack.
These two changes are combined into one PR because they both require
extensive and similar chages in the font_context unit test which
dependended on the details of both of them.
Signed-off-by: Martin Robinson <mrobinson@igalia.com>
This changes modifes the way that font data is sent over IPC channels.
Instead of serializing the data or sending it via IPC byte senders, font
data is copied into shared memory and a copy of the handle is sent over
the channel.
There is also the idea of sending the file handle of the on disk data of
system fonts. This could be implemented as a further followup once there
is an abstraction in `ipc-channel` over file handles.
To accomplish this, a `FontData` abstraction is added, which also allows
caching an in-memory shared `Arc<Vec<u8>>` version of the data (neeeded
by some APIs). This could also be a place for caching font tables in the
future.
Finally, the `FontCacheThread` is renamed to the `SystemFontService`
while the proxy for this is now named `SystemFontServiceProxy`.
Signed-off-by: Martin Robinson <mrobinson@igalia.com>
Co-authored-by: Mukilan Thiyagarajan <mukilan@igalia.com>
* Plumb selected tracing spans into hitrace
Signed-off-by: Delan Azabani <dazabani@igalia.com>
* Tag the spans in #33189 with the `servo_profiling` field
Signed-off-by: Delan Azabani <dazabani@igalia.com>
* Add comment about fields vs extensions
Signed-off-by: Delan Azabani <dazabani@igalia.com>
---------
Signed-off-by: Delan Azabani <dazabani@igalia.com>
Up until now, Servo was using a very old version of time to get a
cross-process monotonic timestamp (using `time::precise_time_ns()`).
This change replaces the usage of old time with a new serializable
monotonic time called `CrossProcessInstant` and uses it where `u64`
timestamps were stored before. The standard library doesn't provide this
functionality because it isn't something you can do reliably on all
platforms. The idea is that we do our best and then fall back
gracefully.
This is a big change, because Servo was using `u64` timestamps all over
the place some as raw values taken from `time::precise_time_ns()` and
some as relative offsets from the "navigation start," which is a concept
similar to DOM's `timeOrigin` (but not exactly the same). It's very
difficult to fix this situation without fixing it everywhere as the
`Instant` concept is supposed to be opaque. The good thing is that this
change clears up all ambiguity when passing times as a `time::Duration`
is unit agnostic and a `CrossProcessInstant` represents an absolute
moment in time.
The `time` version of `Duration` is used because it can both be negative
and is also serializable.
Good things:
- No need too pass around `time` and `time_precise` any longer.
`CrossProcessInstant` is also precise and monotonic.
- The distinction between a time that is unset or at `0` (at some kind
of timer epoch) is now gone.
There still a lot of work to do to clean up timing, but this is the
first step. In general, I've tried to preserve existing behavior, even
when not spec compliant, as much as possible. I plan to submit followup
PRs fixing some of the issues I've noticed.
Signed-off-by: Martin Robinson <mrobinson@igalia.com>
The compositor's `build.rs` script was parsing the `Cargo.lock` file in
order to tag WebRender captures with the WebRender version. The embedder
already knows what version of Servo we are using, which should be enough
to infer the WebRender revision. This changes does that and generally
does a bit of cleaning up of how captures are done.
- The name of the capture directory is now `webrender-captures`
- There is console output now when captures are done. Before it was hard
to know if it succeeded.
- Simplify the Compositor constructor a little to avoid passing
arguments so much.
Signed-off-by: Martin Robinson <mrobinson@igalia.com>
* Read pixels in same format (gl::RGBA) as the texture
Signed-off-by: crbrz <cristianb@gmail.com>
* Add read pixels test
Signed-off-by: crbrz <cristianb@gmail.com>
* Use patched surfman
Signed-off-by: crbrz <cristianb@gmail.com>
* Update surfman to 0.9.5
Signed-off-by: crbrz <cristianb@gmail.com>
---------
Signed-off-by: crbrz <cristianb@gmail.com>
Currently we don't propagate the scroll events across pipeline and
scroll only nodes contained within the nearest pipeline that passes hit
test at a gived point. This causes iframes to "capture" the scroll and
halt the scrolling process once the iframe has been scrolled all the
way. The expected behaviour is that the parent page begins to scroll
once the iframe has been scrolled fully.
The issue is present in both desktop and mobile, but was more noticeable
on mobile, especially on the default servo.org page, because of the relative
sizes of the YouTube video's iframe wrt to the parent.
Signed-off-by: Mukilan Thiyagarajan <mukilan@igalia.com>
Co-authored-by: Martin Robinson <mrobinson@igalia.com>
This is the first part of cleaning up unused WebRender resources.
Currently this only cleans up web font resources, but a more
full-featured implementation in the future could also clean up unused
system fonts.
Fixes#32345.
Signed-off-by: Martin Robinson <mrobinson@igalia.com>
Co-authored-by: Mukilan Thiyagarajan <mukilan@igalia.com>
The `gfx_traits` crate is gradually become just about text and fonts and
this is one of the few things there that is standing in the way of this.
Eventually `webrender_traits` will be the general cross-process
compositor API, so this sort of makes sense as well.
Instead of bouncing messages from the compositor to script and then to
layout, just have script call methods on Layout. Additionally, we do not
need to send any followup messages to script for these messages. Instead
just execute code after calling the method on Layout.
* Move WebRender related types to `webrender_traits`
This refactor moves several WebRender related types
from `compositing_traits`, `script_traits` and `net_traits`
crates to the `webrender_traits` crate.
This change also moves the `Image` type and associated
function out of `net_traits` and into the `pixels` crate.
Co-authored-by: Martin Robinson <mrobinson@igalia.com>
Signed-off-by: Mukilan Thiyagarajan <mukilan@igalia.com>
* Move `script_traits::WebrenderIpcSender` to `webrender_traits::WebRenderScriptApi`
---------
Signed-off-by: Mukilan Thiyagarajan <mukilan@igalia.com>
Co-authored-by: Martin Robinson <mrobinson@igalia.com>
For a long time, `gfx_traits` has held a lot of things unrelated to graphics
and also unrelated to the `gfx` crate (which is mostly about fonts).
This is a cleanup which does a few things:
1. Move non `gfx` crate things out of `gfx_traits`. This is important in
order to prevent dependency cycles with a different integration between
layout, script, and fonts.
2. Rename the `msg` crate to `base`. It didn't really contain anything
to do with messages and instead mostly holds ids, which are used
across many different crates in Servo. This new crate will hold the
*rare* data types that are widely used.
Details:
- All BackgroundHangMonitor-related things from base to a new
`background_hang_monitor_api` crate.
- Moved `TraversalDirection` to `script_traits`
- Moved `Epoch`-related things from `gfx_traits` to `base`.
- Moved `PrintTree` to base. This should be widely useful in Servo.
- Moved `WebrenderApi` from `base` to `webrender_traits` and renamed it
to `WebRenderFontApi`.
* clippy: Squish warnings and errors in gfx
warning: redundant closure (gfx/font.rs:415:18)
warning: useless conversion to the same type (gfx/font.rs:534:9)
warning: the following explicit lifetimes could be elided: 'a (gfx/font.rs:619:16)
error: this loop never actually loops (gfx/font_cache_thread.rs:112:9)
warning: this expression creates a reference which is immediately dereferenced by the compiler (gfx/font_cache_thread.rs:229:51)
warning: redundant closure (gfx/font_cache_thread.rs:551:18)
3 instances of:
warning: casting integer literal to `f64` is unnecessary (gfx/platform/freetype/font_list.rs:271-273)
* clippy: methods called `from_*` usually take no `self`
It reports that by standard convention, from_* methods should not take any `&self` parameter
* clippy: you should consider adding a `Default` implementation
It reports that public types with a pub fn new() -> Self should have a Default implementation since they can be constructed without arguments
* clippy: casting to the same type is unnecessary (`f32` -> `f32`)
* clippy: use of `unwrap_or_else` to construct default value
* clippy: methods called `is_*` usually take `self` by mutable reference or `self` by reference or no `self`
* clippy: manual `!RangeInclusive::contains` implementation
contains expresses the intent better and has less failure modes (such as fencepost errors or using || instead of &&)
* clippy: this function has an empty `#[must_use]` attribute, but returns a type already marked as `#[must_use]`
* clippy: Fix some new warnings
warning: this `if` statement can be collapsed (gfx/font.rs:468:130)
warning: this lifetime isn't used in the impl (gfx/platform/freetype/font.rs:341:6)
warning: field assignment outside of initializer for an instance created with Default::default() (compositor.rs:881:17)
