I looked at this because in bug 2014331 I need to mark more pseudos as
element-backed, and it's kinda silly the state our setup is in right
now.
This is all a somewhat historical mess, because in the old style system
pseudo-elements were just atoms, and we had that different types of atom
setup. The source of truth for these right now lives in a header macro
file, which isn't very flexible.
But now we have PseudoStyleType which represents both pseudos and anon
boxes, and is similar to the rust side PseudoElement enum.
Plus, the ::-moz-tree pseudo-elements are defined as anon boxes which
doesn't make any sense, so move them to pseudo-elements while at it.
Unify this in a similar way we treat CSS properties: Source of truth is
in pseudo_elements.toml / anon_boxes.toml, and we generate Rust and C++
as needed. This allows removing the nsStaticAtom subclass thing from
XPCOM, too.
This avoids some nasty duplication around which pseudos are pref-gated
and what not as well.
Differential Revision: https://phabricator.services.mozilla.com/D282076
I looked at this because in bug 2014331 I need to mark more pseudos as
element-backed, and it's kinda silly the state our setup is in right
now.
This is all a somewhat historical mess, because in the old style system
pseudo-elements were just atoms, and we had that different types of atom
setup. The source of truth for these right now lives in a header macro
file, which isn't very flexible.
But now we have PseudoStyleType which represents both pseudos and anon
boxes, and is similar to the rust side PseudoElement enum.
Plus, the ::-moz-tree pseudo-elements are defined as anon boxes which
doesn't make any sense, so move them to pseudo-elements while at it.
Unify this in a similar way we treat CSS properties: Source of truth is
in pseudo_elements.toml / anon_boxes.toml, and we generate Rust and C++
as needed. This allows removing the nsStaticAtom subclass thing from
XPCOM, too.
This avoids some nasty duplication around which pseudos are pref-gated
and what not as well.
Differential Revision: https://phabricator.services.mozilla.com/D282076
Extend the per-frame-class bit we have to devirtualize IsLeaf to also
devirtualize IsFrameOfType. That is, move this data to FrameClasses.py.
This was done by going through all the frame classes, trying to preserve
behavior.
The only quirky thing is that I had to add two more trivial frame
classes, `nsAudioFrame` for audio elements, and
`nsFloatingFirstLetterFrame`. That's because these frame classes were
returning different answers at runtime, but they do this only on
conditions that trigger frame reconstruction (floating, and being an
audio element, respectively).
Differential Revision: https://phabricator.services.mozilla.com/D194703
While looking at moving the flag around I realized that the only reason
we have FontMetricsProvider and co is because we didn't have access to
the per-document font-prefs cache. That's trivial to fix tho, so do
that and simplify the setup for font queries even more.
Differential Revision: https://phabricator.services.mozilla.com/D157589
gfxFontCache acquires and releases its mutex during various operations.
In order to keep the state internally consistent, we should only release
the lock after the full operation is complete. This involves moving the
deletion of gfxFont to outside the lock via a temporary discard array.
The expiration state should not be protected by the gfxFont's mutex
since we don't hold it during most operations. Instead we should hold
gfxFontCache's mutex because then we can guarantee the operation is
atomic, particularly when a worker wants a font, and the main thread is
aging the generations.
When a font is returned from gfxFontCache, we now return it already
removed from the tracker, and with its refcount updated. This avoids any
potential races between the expiration timer and a worker accessing the
font, as well as simplying the callers so they don't need to be aware of
addref-ing manually in case the result is to be discarded (so that it
gets readded to the tracker).
Differential Revision: https://phabricator.services.mozilla.com/D151821
gfxFontCache acquires and releases its mutex during various operations.
In order to keep the state internally consistent, we should only release
the lock after the full operation is complete. This involves moving the
deletion of gfxFont to outside the lock via a temporary discard array.
The expiration state should not be protected by the gfxFont's mutex
since we don't hold it during most operations. Instead we should hold
gfxFontCache's mutex because then we can guarantee the operation is
atomic, particularly when a worker wants a font, and the main thread is
aging the generations.
When a font is returned from gfxFontCache, we now return it already
removed from the tracker, and with its refcount updated. This avoids any
potential races between the expiration timer and a worker accessing the
font, as well as simplying the callers so they don't need to be aware of
addref-ing manually in case the result is to be discarded (so that it
gets readded to the tracker).
Differential Revision: https://phabricator.services.mozilla.com/D151821
The math-depth implementation is refined to take into account the
ScriptPercentScaleDown and ScriptScriptPercentScaleDown constants (if the
parent's first valid font has a MATH table) in order to calculate the
scale factor between math-deth 0 and 1, and between 0 and 2 respectively.
Behavior is unchanged if the legacy scriptsizemultiplier attribute is
specified or if no MATH table is available.
The preference layout.css.math-depth.enabled remains disabled in nightly
until the remaining bit (support for font-size: math) is implemented in
bug 1667090.
Differential Revision: https://phabricator.services.mozilla.com/D91604
Emilio Cobos Álvarez