As I said over bug 1549593, the eventual goal is to use ArcSlice in all
inherited properties. But this seemed like a good first candidate that doesn't
require me to move around a lot more code, since we were already using cbindgen
for the path commands.
Differential Revision: https://phabricator.services.mozilla.com/D30134
--HG--
extra : moz-landing-system : lando
We could make the header PhantomData or something, but then we wouldn't be able
to bind to C++, since C++ doesn't have ZSTs. So add a canary instead to add a
runtime check of stuff being sane.
Differential Revision: https://phabricator.services.mozilla.com/D30133
--HG--
extra : moz-landing-system : lando
This enables destructors for tagged unions in cbindgen, implemented in:
* https://github.com/eqrion/cbindgen/pull/333
Which allow us to properly generate a destructor for the cbindgen-generated
StyleBasicShape (which now contains an OwnedSlice).
For now, we still use the glue code to go from Box<BasicShape> to
UniquePtr<BasicShape>. But that will change in the future when we generate even
more stuff and remove all the glue.
I could add support for copy-constructor generation to cbindgen for tagged
enums, but I'm not sure if it'll end up being needed, and copy-constructing
unions in C++ is always very tricky.
Differential Revision: https://phabricator.services.mozilla.com/D29769
--HG--
extra : moz-landing-system : lando
Passing these by value won't be ok of course, but that's fine.
I plan to combine this with https://github.com/eqrion/cbindgen/pull/333 to
actually be able to share representation for ~all the things, this is just the
first bit.
Box<T>, Atom and Arc<T> will be much easier since cbindgen can understand them
without issues.
It's boxed slices the only ones I should need something like this. I could avoid
it if I rely on Rust's internal representation, which we can per [1], but then I
need to teach cbindgen all about slices, which is generally hard, I think.
[1]: https://github.com/rust-lang/unsafe-code-guidelines/blob/master/reference/src/layout/pointers.md
Differential Revision: https://phabricator.services.mozilla.com/D29768
--HG--
extra : moz-landing-system : lando
I'm hitting this when trying to add bindings for box shadows, which are 32-bit
aligned.
Depends on D30352
Differential Revision: https://phabricator.services.mozilla.com/D30353
--HG--
extra : moz-landing-system : lando
The change from [T; 1] to [T; 0] shouldn't change behavior since they have the
right alignment and we never poke at that particular array, but feels more
correct to avoid creating types that point to uninitialized data or outside of
their allocation, now that we allow empty slices.
Differential Revision: https://phabricator.services.mozilla.com/D30352
--HG--
extra : moz-landing-system : lando
I thought a bit about how to test it and it's not particularly great.
test_css_parse_error_smoketest.html is great to assert that something _gets_
reported, but not that it doesn't :)
Differential Revision: https://phabricator.services.mozilla.com/D30201
I thought a bit about how to test it and it's not particularly great.
test_css_parse_error_smoketest.html is great to assert that something _gets_
reported, but not that it doesn't :)
Differential Revision: https://phabricator.services.mozilla.com/D30201
--HG--
extra : moz-landing-system : lando
Also, buffer the errors, since we're going to want to look at the whole
declaration block to skip reporting them.
This shouldn't change behavior, just moves some work to the caller, and defers a
bit the work so that it happens only when error reporting is enabled.
Differential Revision: https://phabricator.services.mozilla.com/D30200
--HG--
extra : moz-landing-system : lando
The initial value for this is indeed `none` (and thus empty). The Rust code was
confused.
This property is disabled by default these days, and I think the
get_initial_value() function, which is what could get confused, is not called
for this property, so I think this shouldn't be observable.
Differential Revision: https://phabricator.services.mozilla.com/D30124
--HG--
extra : moz-landing-system : lando
No longer needed now that the case-sensitive flag for attributes selectors is
supported.
Update user-agent CSS sheet to use the standard flag.
Fixes#23227
This cherry-picks https://github.com/servo/servo/pull/23228.
Converting the NS_STYLE_BORDER definitions in to enumerated classes as
per bug 1277133.
The original constants broke the convention used by the rest of the
definitions as the CSS property being described is `border-collapse`,
so corrections were made with the migration to the enumerated class.
Differential Revision: https://phabricator.services.mozilla.com/D29951
--HG--
extra : moz-landing-system : lando
Additionally, this sorts out the order of member variables for minimizing the
instance size.
And also this changes `enum RenderFlags` to `enum class RenderingStateFlags`.
Differential Revision: https://phabricator.services.mozilla.com/D29312
--HG--
extra : moz-landing-system : lando
Disabled for now of course. This should be pretty uncontroversial I'd think.
Differential Revision: https://phabricator.services.mozilla.com/D28060
--HG--
extra : moz-landing-system : lando
We always include the combinator for pseudo-elements now (not including it was
just an optimization) in order to not match when nested pseudo-elements are
involved.
We could add a more generic check in `matches_simple_selector` like:
```
if element.is_pseudo_element() {
match *selector {
Component::PseudoElement(..) |
Component::NonTSPseudoClass(..) => {},
_ => return false,
}
}
```
But even that wouldn't be enough to make selectors like `:hover::marker` not
match on the `::before::marker` pseudo-element, plus that code is really hot.
So for now do the check on the `next_element_for_combinator` function. It's a
bit hacky but it's the best I could came up with...
While at it, simplify some checks to use is_pseudo_element() instead of
implemented_pseudo_element() directly.
Only the Rust patch as-is would make markers for ::before and ::after on list
items not show up, so we also need to switch ::marker to use ProbeMarkerStyle()
rather than ProbePseudoElementStyle(), since the marker should exist even if it
matches no rules.
Differential Revision: https://phabricator.services.mozilla.com/D27529
--HG--
extra : moz-landing-system : lando
I'm going to unconditionally generate the PseudoElement combinator, and this
causes issues since we'll put the raw `::pseudo` selectors in the host bucket,
which is obviously wrong.
Differential Revision: https://phabricator.services.mozilla.com/D27528
--HG--
extra : moz-landing-system : lando
The previous commit removed the dependence on the discriminant value, so we
don't need to keep discriminants different from text-align anymore.
Differential Revision: https://phabricator.services.mozilla.com/D29361
--HG--
extra : moz-landing-system : lando
This doesn't clean up all that much, yet, but it's a step in the right
direction.
Differential Revision: https://phabricator.services.mozilla.com/D29168
--HG--
extra : moz-landing-system : lando
This won't reintroduce any of the regressions that were triggered by our
previous attempts to turn off -moz prefixed gradients, and lets us massively
simplify the gradient code, if it sticks.
Differential Revision: https://phabricator.services.mozilla.com/D29346
--HG--
extra : moz-landing-system : lando
Turns out removing the pseudo-class and such ends up not being quite as trivial
as I initially thought, or possible at all, since the fact that the <symbol> is
a <symbol> is observable via selectors, added a test for that.
Differential Revision: https://phabricator.services.mozilla.com/D29131
--HG--
extra : moz-landing-system : lando
We want to drop the cascade data memory as soon as possible, so bug 1544546
introduced an UpdateStylistIfNeeded call from ShellDetachedFromDocument.
Unfortunately, this call can reenter into the global user-agent cascade data if
some of the CSS values kept alive by the cascade data keep alive an SVG
document, see the stack on this bug for an example. Make sure to drop the
user-agent cascade datas when not holding the cache lock to avoid this
situation.
Before bug 1535788, we just destroyed the stylist, so we kept holding a
reference from the cache, and that reference will be dropped sometime later when
other document updated their user-agent stylesheets (if they happened not to
match the cache of course).
Seems to me this doesn't ended up happening in our automation, but it could
happen in the wild, in theory at least.
It's nice that Rust made this a safe deadlock caught by our tests rather than a
very subtle and infrequent memory corruption.
The relevant SVG documents are probably the <input type=number> rules:
https://searchfox.org/mozilla-central/rev/d80f0a570736dce76a2eb184fb65517462089e8a/layout/style/res/forms.css#1050
Differential Revision: https://phabricator.services.mozilla.com/D28299
--HG--
extra : moz-landing-system : lando
Just one set of stylesheets is enough. While at it, unify SheetType and Origin.
Differential Revision: https://phabricator.services.mozilla.com/D27564
--HG--
extra : moz-landing-system : lando