| Age | Commit message (Collapse) | Author |
|---|
|
"Improve" is an understatement, since this commit makes all FLAC files
seek without errors, mostly with high accuracy, and partially even fast:
- A new generic seek table type is introduced, which keeps an
always-sorted list of seek points, which allows it to use binary
search and fast insertion.
- Automatic seek points are inserted according to two heuristics
(distance between seek points and minimum seek precision), which not
only builds a seek table for already-played sections of the file, but
improves seek precision even for files with an existing seek table.
- Manual seeking by skipping frames works properly now and is still used
as a last resort.
|
|
It appears that QEMU on macOS doesn't have the VirtIO GPU variants that
support VGA functionality. Those variants are not especially important
to us, because we don't use any kind of VGA functionality in our kernel
anyway.
Therefore, for macOS, we could decide to use virtio-gpu-gl-pci and
virtio-gpu-pci devices instead.
|
|
It appeared that we sometimes failed to invoke synchronous commands on
the GPU. To temporarily fix this, wait 10 milliseconds for commands to
complete before failing.
|
|
According to the specification, modesetting can be invoked with no need
for flushing the framebuffer nor with DMA to transfer the framebuffer
rendering.
|
|
This configuration exposes a suboptimal mechanism to access other
VirtIO device configurations. It is also the only configuration to use a
zero length for a configuration structure, and specify a valid BAR which
triggered a kernel panic when attaching a virtio-gpu-pci device before
95b15e49010299902abd2aa65bcc71e4158b7326 was applied.
The real solution for that problem is to ignore this configuration type
because we never actually use it. It means that we can VERIFY that all
other configuration types have a valid length, as being expected.
|
|
Turns out LLVMFuzzerTestOneInput may be called more than once per
process.
|
|
Prior to this commit, when you double-click a .zip file to open it, it
gets opened in Text-Editor as there is no other file association.
Now, when FileManager is invoked with a .zip file as the first argument,
a temporary directory will be created and the .zip will be extracted
into it. Once the FileManager window is closed, Core::TempFile will
delete the temporary directory.
This adds something like what we see in other operating systems' file
explorers, except for the fact that most other operating systems will
treat the .zip file as its own independent read-only filesystem. It
would be nice to do that in the future, but I feel like this is
sufficient for now.
|
|
- We were using primitive versions of mkstemp and mkdtemp, they have
been converted to use LibCore::System.
- If an error occurred whilst creating a temporary directory or file, it
was thrown and the program would crash. Now, we use ErrorOr<T> so that
the caller can handle the error accordingly
- The `Type` enumeration has been made private, and `create_temp` has
been "split" (although rewritten) into create_temp_directory and
create_temp_file. The old pattern of TempFile::create_temp(Type::File)
felt a bit awkward, and TempFile::create_temp_file() feels a bit nicer
to use! :^)
Once the Core::Filesystem PR is merged (#17789), it would be better for
this helper to be merged in with that. But until then, this is a nice
improvement.
|
|
|
|
At the moment, we can't resolve CSS calc() values without having a
LayoutNode. The new StyleProperties::line_height() overload was trying
to do exactly that, which led to an assertion.
This patch makes `line-height: calc(...)` behave the same as
`line-height: normal` for now and adds a FIXME.
|
|
Don't crash in case the elements parent is null.
|
|
If the PaintableBox had children, but we didn't hit any of them, we
default to saying that you hit the PaintableBox itself.
However, if said PaintableBox has `pointer-events: none`, we should
say nothing was hit, so that the hit testing can continue.
This fixes an issue where Discord server icons were not clickable.
|
|
This makes focusing input fields by clicking on them work.
Fixes #17801
|
|
|
|
Previously, empty text boxes would fall back to the min-height: 16px
set on the <input> element. As soon as there is any content they
would usually gain height because the line height of that text is
more than 16px (depending on the font/font-size used).
Now they use height: 1lh for the inner div (which contains the
actual text), which matches the exact height of 1 line of content.
|
|
|
|
|
|
Resolving these units is somewhat tricky because of their interaction
with both font-size and line-height, but this implementation seems to
work as tested in http://wpt.live/css/css-values/lh-unit-001.html and
http://wpt.live/css/css-values/lh-unit-002.html
|
|
Co-authored-by: Luke Wilde <lukew@serenityos.org>
|
|
These are currently initialized in a [[gnu::constructor]], which has a
weird initialization order. These constructors are invoked before main()
and, incidentally, before any user-defined default constructors of the
static strings they are initializing.
This will become an issue when these strings are ported to FlyString,
which has a user-defined default constructor. In that scenario, when the
FlyString constructor is executed after the [[gnu::constructor]], the
strings will be "reset" to the empty string.
Instead of relying on a non-standard compiler extension here, let's just
initialize these strings explicitly during main-thread VM creation, as
this now happens in WebContent's main().
|
|
This is to disambiguate this type from the StringView overload.
|
|
|
|
When CallbackType::callback was converted from Object& to NNGCP<Object>,
we started comparing the addresses of NNGCPs instead of the addresses of
Objects.
That broke the Discord login form, and this patch fixes it.
Regression from 7c0c1c8f4969abeec1436346f29081b3afbcdeab.
|
|
We use recursive algorithms all over the place, and this check makes
it hard to read those algorithms in CLion, since it draws squiggles
under everything.
|
|
Explicitly check is_viewport() instead of looking at the corresponding
DOM node. (The viewport has the DOM document as its DOM node, but that's
not obvious from context here.)
|
|
Checkedness of an input element can influence sibling style, as well as
style of their children, when they use the `:checked` pseudo-class in
combination with a kind of sibling selector. That means its not
sufficient to just invalidate the input elements on style.
This is actually more commonly observable than one might expect, because
this pattern is often used as a JS-free toggle solution for things like
menus.
|
|
Invalid UTF-8 crashes JS::Script::Parse.
|
|
Though deprecated by IANA, `application/font-woff` is still in active
use as a MIME type for WOFF fonts by web servers throughout the wild
web.
|
|
We don't support all parts of the font formats we assume as "supported"
in the CSS parser. For example, if an open type font has a CFF table, we
reject loading it. This meant that until now, when such an
unsupported-supported font url was first in the list of urls, we
couldn't load it at all, even when we would support a later url.
To resolve that, try loading all font urls one after each other, in case
we are not able to load the higher priority one.
This also resolves a FIXME related to spec compliant url prioritization.
Our CSS parser already filters and prioritizes font src urls in
compliance with the spec. However, we still had to resort to brittle
file extension matching, because some websites don't set the `format`
and if the first url in a src list happened to be one we don't support,
the font could not be loaded at all. This now is unnecessary because we
can try and discard the urls instead.
|
|
This fixes an issue where e.g `height: 100%` on a flex item whose
container has indefinite height was being resolved to 0. It now
correctly behaves the same as auto.
|
|
|
|
There is no official IANA MIME type for WAV (see
https://www.iana.org/assignments/media-types/media-types.xhtml#audio),
so this will always be subjective. While
https://www.rfc-editor.org/rfc/rfc2361 suggests audio/vnd.wave, we use
audio/wav since that seems to be most common across the internet.
|
|
We were storing indices into OSC escape sequences as `u8`s, which
overflow at a length of just 256 characters. This caused a crash when
parsing OSC 8 hyperlinks pointing to long filenames.
|
|
|
|
It is a fallible operation, so this lets us abort early if it fails.
|
|
|
|
No change of behavior in this patch, but this will allow this factory to
propagate any errors from the creation of these objects.
|
|
Required by code that brand checks native constructors.
For example, Wistia brand checks XMLHttpRequest by doing:
```
XMLHttpRequest.prototype.constructor.toString()
```
It then checks if it matches either one of:
```
function XMLHttpRequest() { [native code] }
```
```
[object XMLHttpRequestConstructor]
```
If neither matches, it disables HLS playback and prints:
"The XMLHttpRequest constructor has been tampered with. Because this
affects CORS/Range XHR requests, HLS playback has been disabled.
To enable HLS playback and other important features, please remove code
that changes the definition of window.XMLHttpRequest."
We hit this path due to not giving generated constructors a name, as
we would provide `function () { [native code] }`.
|
|
|
|
Now when the bookmark button that has not yet bookmarked the current
URL is pressed, it will add the bookmark but also prompt the user
with the BookmarkEditor dialog in case they wish to make final
touches to their bookmark title or URL. If they cancel or escape
the dialog, the bookmark will be removed.
|
|
|
|
|
|
We were missing some unwind related control flow paths, and followed
some in improper ways, leading us to access a dead unwind frame in some
cases, as well as generating a technically wrong CFG.
This reorders the ways EnterUnwindContext works and alleviates those
errors.
|
|
This patch parses enough of GPOS tables to be able to support the
kerning information embedded in Inter.
Since that specific font only applies positioning offsets to the first
glyph in each pair, I was able to get away with not changing our API.
Once we start adding support for more sophisticated positioning, we'll
need to be able to communicate more than a simple "kerning offset" to
the clients of this code.
|
|
Previously HTMLProgressElement implemented is_html_input_element() not
is_html_progress_element(), so is_html_progress_element() defaulted to
always returning false. This broke is<HTML::HTMLProgressElement>().
|
|
|
|
|
|
Currently, the WebAssemblyObject implements a visitor to keep its static
objects alive. This custom attribute will be used to hook the generated
namespace object's visitor to one that we define in non-generated code.
|
|
Interfaces with a LegacyNamespace extended attribute should have their
constructors defined on the namespace identified by the LegacyNamespace
attribute value.
|
|
These will be needed by any namespace having a BufferSource parameter,
such as WebAssembly.
Similar to 49e6cb7c3d62eeeddf63c5db931038f1aeb58d53.
|
