This is a relatively small code size regression (130k on windows and
macOS, 180k on Linux), for a few high confidence improvements in
speedometer 3. See link in the bug.
If this sticks, we can actually clean up a bunch of code, and eventually
unify RefPtr and nsCOMPtr. But I want this to be on the tree for a while
before doing more aggressive refactorings / actually removing the code.
Differential Revision: https://phabricator.services.mozilla.com/D178267
See blocked bug.
For non-Locked types we can just use the same underlying type at the FFI
boundary. Port StylesheetContents and CssUrlData to this set-up.
CssUrlData is already generated by cbindgen anyways.
Differential Revision: https://phabricator.services.mozilla.com/D177559
After the previous patches we only have one trait which we should also
tweak / rework, so let's put it all on that single trait.
Differential Revision: https://phabricator.services.mozilla.com/D177515
HasBoxFFI and HasArcFFI aren't great, see bug 1831242 as for examples of
why.
HasArcFFI requires a bit more care, but HasBoxFFI doesn't give us much
benefit. Instead use the same type in the FFI boundary.
Differential Revision: https://phabricator.services.mozilla.com/D177252
Generated by running
find servo/components/style -name "*.rs" -exec perl -p -i -e "s/write_str\(\"(.)\"\)/write_char('\1')/g" {} \;
(and then added `use std::fmt::Write;` in a couple of places to fix build errors that arose).
Differential Revision: https://phabricator.services.mozilla.com/D168217
Named scroll progress timelines are declared in the coordinated value list
constructed from the longhands of the scroll-timeline shorthand property,
which form a coordinating list property group with scroll-timeline-name as
the coordinating list base property.
In the meantime, we also update its shorthand to match the current spec.
Differential Revision: https://phabricator.services.mozilla.com/D166596
view-timeline-name: `none | <custom-ident>#`
view-timeline-axis: `[ block | inline | vertical | horizontal ]#`
Note:
Both view-timeline-name and scroll-timeline-name should accept `auto`.
We will fix it in this patch series.
Differential Revision: https://phabricator.services.mozilla.com/D166242
It's unclear to me when they were disabled, but we do want to enable
these as otherwise there's no way to catch bindgen issues that can end
up in subtle bugs at best, or memory corruption at worst.
Differential Revision: https://phabricator.services.mozilla.com/D162997
This changes font-family storage to reuse the rust types, removing a
bunch of code while at it. This allows us to, for example, use a single
static font family for -moz-bullet and clone it, rather than creating a
lot of expensive copies.
Differential Revision: https://phabricator.services.mozilla.com/D118011
And move the useful bits of it somewhere else (ServoStyleConstInlines.h for the
inline function definitions, and nsFrame.cpp for the static assertions).
Differential Revision: https://phabricator.services.mozilla.com/D36120
Right now we do a lot of useless string copying. In order to avoid transcoding
to utf-16 during layout, make sure to use nsCString at a few related places.
I may revisit this since we're storing other line names as atoms in some places.
So it may be better to just use atoms everywhere.
But that'd be a different patch either way.
Differential Revision: https://phabricator.services.mozilla.com/D35117
This doesn't clean up as much as a whole, but it's a step in the right
direction. In particular, it allows us to start using simple bindings for:
* Filters
* Shapes and images, almost. Need to:
* Get rid of the complex -moz- gradient parsing (let
layout.css.simple-moz-gradient.enabled get to release).
* Counters, almost. Need to:
* Share the Attr representation with Gecko, by not using Option<>.
* Just another variant should be enough (ContentItem::{Attr,Prefixedattr},
maybe).
Which in turn allows us to remove a whole lot of bindings in followups to this.
The setup changes a bit. This also removes the double pointer I complained about
while reviewing the shared UA sheet patches. The old setup is:
```
SpecifiedUrl
* CssUrl
* Arc<CssUrlData>
* String
* UrlExtraData
* UrlValueSource
* Arc<CssUrlData>
* load id
* resolved uri
* CORS mode.
* ...
```
The new one removes the double reference to the url data via URLValue, and looks
like:
```
SpecifiedUrl
* CssUrl
* Arc<CssUrlData>
* String
* UrlExtraData
* CorsMode
* LoadData
* load id
* resolved URI
```
The LoadData is the only mutable bit that C++ can change, and is not used from
Rust. Ideally, in the future, we could just use rust-url to resolve the URL
after parsing or something, and make it all immutable. Maybe.
I've verified that this approach still works with the UA sheet patches (via the
LoadDataSource::Lazy).
The reordering of mWillChange is to avoid nsStyleDisplay from going over the
size limit. We want to split it up anyway in bug 1552587, but mBinding gains a
tag member, which means that we were having a bit of extra padding.
One thing I want to explore is to see if we can abuse rustc's non-zero
optimizations to predict the layout from C++, but that's something to explore at
some other point in time and with a lot of care and help from Michael (who sits
next to me and works on rustc ;)).
Differential Revision: https://phabricator.services.mozilla.com/D31742
This avoids the expensive conversion, and cleans up a bunch.
Further cleanup is possible, just not done yet to avoid growing the patch even
more.
Differential Revision: https://phabricator.services.mozilla.com/D30748
This depends on https://github.com/eqrion/cbindgen/pull/308. Other than that,
this should be ready to go.
There's still a bit more magic than what I'd like to eventually. I should be
able to make cbindgen not rename types if it doesn't know about them, or
something.
But this removes most of the manual binding function implementations (all but
the ones that are declared via macros, which cbindgen doesn't see across).
I need to give up on the _Drop functions taking an Owned<T> because of
instantiation order fiasco. In order to define DefaultDelete I need Owned to be
complete, but I cannot do it after including the generated file since some
declarations already instantiate the specialization. Oh well.
Differential Revision: https://phabricator.services.mozilla.com/D24798
I kept it building the most straight-forward way possible (pub use) because it
seems to me that bindings is not a bad name, and we should probably move
structs.rs to be bindings.rs rather than the other way around.
But that's a different bug in any case, need to think more about it.
Differential Revision: https://phabricator.services.mozilla.com/D24713
And make them actually sound. We're defining functions on Rust-land that get
structs as arguments, but declaring them in C++ as getting pointers.
This is another step in order to be able to autogenerate ServoBindings.h and
remove bindings.rs altogether.
We remove FooOwned in favor of Owned<Foo>, which is generated via cbindgen.
It'd be good to actually mark Owned and such as MOZ_MUST_USE_TYPE, so I sent
https://github.com/eqrion/cbindgen/pull/307 for that.
Differential Revision: https://phabricator.services.mozilla.com/D24681
This reduces a lot the boilerplate that's needed in order to add simple binding
functions.
This starts using &Foo and Option<&Foo> instead, and as a result we need to
remove the servo_function_signatures test, which is a bit unfortunate.
I think it's worth though, this causes problems on some platforms (see bug
1534844), and messing up the functions signature is not something that I've ever
seen (other than bug 1308234, which already had all the FooBorrowed mess which
I'm removing).
Also, cbindgen understands references and Option<&Foo>, so it will be the way to
go in the future.
After this patch we can also remove HasSimpleFFI, but I've kept it for now since
I still use it in a few places, and this patch is quite big on its own.
Differential Revision: https://phabricator.services.mozilla.com/D24092
I want to do this so that I can get rid of Either<>. The reasons for getting rid
of either are multiple:
* It doesn't generate as nice C++ code using cbindgen.
* It isn't that nice to use either from Rust.
* cbindgen has bugs with zero-sized types.
I started using this for ColorOrAuto and a few others, for now.
Differential Revision: https://phabricator.services.mozilla.com/D19844
Not the prettiest, but it will work, and LengthPercentage will be 12 bytes which
is pretty good (we could do better if wanted I guess):
* Au(i32) length;
* f32 percentage;
* AllowedNumericType(u8) clamping_mode;
* bool has_percentage;
* bool was_calc;
This will allow me to start moving C++ stuff to use this representation.
Differential Revision: https://phabricator.services.mozilla.com/D16929
It does not represent `<length> | <percentage>`, but `<length-percentage>`, so
`LengthOrPercentage` is not the right name.
This patch is totally autogenerated using:
rg 'LengthOrPercentage' servo | cut -d : -f 1 | sort | uniq > files
for file in $(cat files); do sed -i "s#LengthOrPercentage#LengthPercentage#g" $file; done
Differential Revision: https://phabricator.services.mozilla.com/D15812
This is a first step to share LengthOrPercentage representation between Rust and
Gecko.
We need to preserve whether the value came from a calc() expression, for now at
least, since we do different things depending on whether we're calc or not right
now. See https://github.com/w3c/csswg-drafts/issues/3482 and dependent bugs for
example.
That means that the gecko conversion code needs to handle calc() in a bit of an
awkward way until I change it to not be needed (patches for that incoming in the
next few weeks I hope).
I need to add a hack to exclude other things from the PartialEq implementation
because the new conversion code is less lossy than the old one, and we relied on
the lousiness in AnimationValue comparison (in order to start transitions and
such, in [1] for example).
I expect to remove that manual PartialEq implementation as soon as I'm done with
the conversion.
The less lossy conversion does fix a few serialization bugs for animation values
though, like not loosing 0% values in calc() when interpolating lengths and
percentages, see the two modified tests:
* property-types.js
* test_animation_properties.html
Differential Revision: https://phabricator.services.mozilla.com/D15793
