New propagated errors comes from an endgrent call, and all chown+chmod
calls all over the file. For the two last functions, custom C++ wrappers
were used.
Those wrappers were discarding `ENOENT` errors and naively crashing on
other errors, this change introduced in 76e12a48 was used to prevent a
crash if a mouse wasn't detected. There are no reasons to discard and to
not propagate other errors.
All DisplayConnectors should support the mmap interface and it should
provide better performance now, so let's just use it and drop support
for the DisplayConnector's write interface from the WindowServer side.
This function is added to separate the desktop toggling functionality
from the "Show Desktop" button in the taskbar and to enable calling this
behaviour via a keyboard shortcut in a later commit.
Previously when setting an action's icon we would only change the bitmap
stored by the action. This patch adds logic to propagate that change to
toolbar buttons as well as window menus.
This fixes an issue in SoundPlayer that would cause the play button not
to reflect the play state.
Resetting the interface just based on the fact that it already has an IP
assigned doesn't make much sense, considering that we frequently end up
here after having configured an interface via DHCP already.
Instead, just keep the part that prevents us from sending DHCP
discoveries to interfaces that shouldn't be using DHCP.
While we are at it, place some of the logging behind a debug flag, as
this function is apparently meant to be run frequently.
This partially reverts commit e14d4482a1.
When the user executes chres to change to a new resolution, the
WindowManager removes for each window its intersections with the
screens (window.screens()) and recalculates its rect. Finally, a
Window::set_rect call sets the window's new rectangle. The set_rect
call also triggers a call to Compositor::invalidate_occlusions which
fills for each window the intersections with the screens again in
window.screens().
In case chres switches to an already present resolution the set_rect
call exits prematurely as it checks if the window's rect really
changed. This means that nobody calls invalidate_occlusions
resulting in a rendering issue for each window.
Moving the call to Compositor::screen_resolution_changed after the
clearing of window.screens() and recalc of the window rect for each
window resolves the rendering issue as screen_resolution_changed
calls invalidate_occlusions.
This commit adds an IPv4Gateway to Network.ini. If that option is set to
value other than 0.0.0.0, the NetworkServer adds a default route (e.g.
with address 0.0.0.0/0) with the specified destination.
This service is responsible for loading network configuration from a
/etc/Network.ini config file. It sets up static IP address + mask or
starts DHCPClient depending on configuration.
Now, the caller needs to give interface names in command-line arguments.
The DHCPClient will perform DHCP discovery only on these adapters. The
service now immediately closes when no interfaces were given.
We don't check if interface has already IP address assigned; we just
reset it to zero so that DHCP resolution will not fail.
This became apparent when using the VirtIO graphics device, because the
HardwareScreenBackend object needs to allow flushing of the framebuffer
constantly, and due to incorrect if-else flow, even a non-error response
from the ioctl was leading to a debug spam.
This makes the wrapper more like the rest in LibCore, and also
removes the annoying limitation of not supporting arguments.
There are three overloads one for String, char const *, and StringView
argument lists. As long as there are <= 10 arguments the argv list
will be allocated inline, otherwise on the heap.
Previously, the device wasn't being created.
The new numbers correspond to those in the FullDevice (which is also
a Linux's full device number).
Also, I've tweaked the permissions to give everyone read-write access
to it.
This fixes a crash where if you switched to a theme that has hover
icons for title buttons, then back to a theme that does not. Then
when you next hover over the title buttons the window server would
crash.
This was due to the hover_bitmap multi-scale bitmap pointer being
non-null, but not containing any bitmaps, so hitting an assertion
when painting.
This is the final component that required LibProtocol as a dependency
of LibWeb. With this, we can now remove the dependency, and LibWeb no
longer requires IPC to work :^)
Much like the ImageDecoder change, this moves the underlying connection
of the Web::WebSockets class from LibWeb to LibWebView, removing the
need for LibProtocol in LibWeb for this specific use-case.
After this change, LibWeb now expects Web::ImageDecoding::Decoder to be
pre-initialized with a concrete implementation before using the webpage.
The previous implementation, based on the ImageDecoder service, has been
provided directly through an adapter in LibWebClient, and is now used as
the default value by WebContent.
Also moves WebContentClient and the references to the generated IPC
descriptions, since they are all components of OutOfProcessWebView.
This patch has no functional changes.
The compiler would complain about `__builtin_memcpy` in ByteBuffer::copy
writing out of bounds, as it isn't able to deduce the invariant that the
inline buffer is only used when the requested size is smaller than the
inline capacity.
The other change is more bizarre. If the destructor's declaration
exists, gcc complains about a `delete` operation causing an
out-of-bounds array access.
error: array subscript 'DHCPv4Client::__as_base [0]' is partly outside
array bounds of 'unsigned char [8]' [-Werror=array-bounds]
14 | ~DHCPv4Client() = default;
| ^
This looks like a compiler bug, and I'll report it if I find a suitable
reduced reproducer.
When user code requests the current cursor theme name with
`GUI::ConnectionToWindowServer::the().get_cursor_theme()`, that reads
the name from the config file. If we don't write that out when it's
changed, then users get given an outdated cursor theme instead of the
current one.
In particular, changing the theme in MouseSettings, saving and closing
it, then reopening MouseSettings, would show the wrong theme selected.
So, that's fixed now. :^)
Now, when windows flash, the "active" color of the window title frame's
flash is the highlight color instead of the standard window color. The
"inactive" color of the flash is still the disabled color as before.
Reasoning behind this change in aesthetics: There are four [1] window
title frame colors with specific UI purposes:
1. "Active" for the normal active window, obvious purpose.
2. "Moving" for the window that is being dragged or resized by the user.
Responsible for acting as a visual click feedback as a kind of
subdued button.
3. "Inactive" for any inactive windows, obvious purpose.
4. "Highlight".
The purpose of "Highlight" is in the name, though it's non-obvious what
that exactly entails. Before, only alt-tab selecting windows would use
the highlight color for showing the current target window. In my opinion
this points to the purpose of "highlight" being to move the user's focus
to another window when the simple active state is not enough. Then it
makes sense to have the window flashing use the highlight color. The
entire purpose of window flashing is to shift the user's focus to a
dialog window that might not be close to the window they just clicked.
Using the highlight color enables an even stronger emphasis than before.
It enables a cleaner separation between the purpose of the two frame
colors, as well as making the "Highlight" frame more common.
[1] Technically there are eight, as the title frame has a gradient by
default. We can count the gradient as one color for this purpose.
Multiply the cursor position by the current scaling mode multiplier when
getting the color under the cursor. Also multiply the screen rectangle
before checking if the cursor is within bounds.
Color picker would not account for scaling when getting the color under
the cursor.
Fixes#13906.
Using a Vector<Value> is unsafe as GC cannot see the stored values.
This is then vended to outside users of ConsoleClient, e.g. LibWeb and
WebContent, which is then outside of LibJS's control.
An example issue is if the client stores it for later use and forgets
to visit the stored values, meaning they can be destroyed at any time.
We can save the client from this by vending a MarkedVector<Value> to
them.
Physical hardware doesn't care about scale factors as this is a concept
being related to WindowServer and userland applications. To ensure we
provide the correct display resolution details to HardwareScreenBackend
objects, we must keep a separate Gfx::IntRect object that reserve the
correct details.
This hack is not necessary anymore, because WindowServer will try
constantly to write the framebuffer contents to the display connector
devices. After a switch from console mode to graphical mode, the write
syscall on these devices will not be silently ignored but will actually
write to the framebuffer screen.
As the framebuffer devices are gone by now, searching for them makes no
sense, so instead, let's remove this check now. If the user requested to
boot into text mode, he probably knows what he is doing and we should
not try to warn him about this.
Unconditionally adding the item height offset to the adjusted vertical
position of a menu leads to an incorrect position if we previously
clamped the translation offset so the menu doesn't go off-screen.
Also add some comments to this coordinate math. Always add comments to
coordinate math.
This ioctl is more appropriate when the hardware supports flushing of
the entire framebuffer, so we use that instead of the previous default
FB_IOCTL_FLUSH_HEAD_BUFFERS ioctl.
Such mechanism will be used by the Intel Graphics driver, because we
lack support of changing the resolution on this driver currently, so,
when WindowServer will try to mode-set the display then it will fail,
and will use the safe mode-setting call instead to be able to show
something on screen.
The file is now renamed to Queue.h, and the Resampler APIs with
LegacyBuffer are also removed. These changes look large because nobody
actually needs Buffer.h (or Queue.h). It was mostly transitive
dependencies on the massive list of includes in that header, which are
now almost all gone. Instead, we include common things like Sample.h
directly, which should give faster compile times as very few files
actually need Queue.h.
This allows adding "-hover.png" variants of the title button icons.
This can be useful for themes which use the TitleButtonsIconOnly
flag, which otherwise don't have a way of showing the hover state.
This will allow using the console tty and WindowServer regardless of
your kernel command line. Also this fixes a bug where, when booting in
text mode, the console was in graphical mode, and would not accept
input.
The current message distinguishes between a user that doesn't exist, and
an invalid password. This is considered to be bad practice, because an
attack can first check if a user exists before guessing that users
password.
Also it's just tradition or something.
This patch fixes the issue of pressing the ok button of a settings menu
without saving the changes, or not reverting the changes when pressing
the cancel button because the app has died before the new values make
it to the other end.
With this flag set to true only the icon of the title button is painted.
This is useful for themes with a more non-serenity look such as
Coffee and Cupertino (that currently try to hide the button).
Previously, we were sending Buffers to the server whenever we had new
audio data for it. This meant that for every audio enqueue action, we
needed to create a new shared memory anonymous buffer, send that
buffer's file descriptor over IPC (+recfd on the other side) and then
map the buffer into the audio server's memory to be able to play it.
This was fine for sending large chunks of audio data, like when playing
existing audio files. However, in the future we want to move to
real-time audio in some applications like Piano. This means that the
size of buffers that are sent need to be very small, as just the size of
a buffer itself is part of the audio latency. If we were to try
real-time audio with the existing system, we would run into problems
really quickly. Dealing with a continuous stream of new anonymous files
like the current audio system is rather expensive, as we need Kernel
help in multiple places. Additionally, every enqueue incurs an IPC call,
which are not optimized for >1000 calls/second (which would be needed
for real-time audio with buffer sizes of ~40 samples). So a fundamental
change in how we handle audio sending in userspace is necessary.
This commit moves the audio sending system onto a shared single producer
circular queue (SSPCQ) (introduced with one of the previous commits).
This queue is intended to live in shared memory and be accessed by
multiple processes at the same time. It was specifically written to
support the audio sending case, so e.g. it only supports a single
producer (the audio client). Now, audio sending follows these general
steps:
- The audio client connects to the audio server.
- The audio client creates a SSPCQ in shared memory.
- The audio client sends the SSPCQ's file descriptor to the audio server
with the set_buffer() IPC call.
- The audio server receives the SSPCQ and maps it.
- The audio client signals start of playback with start_playback().
- At the same time:
- The audio client writes its audio data into the shared-memory queue.
- The audio server reads audio data from the shared-memory queue(s).
Both sides have additional before-queue/after-queue buffers, depending
on the exact application.
- Pausing playback is just an IPC call, nothing happens to the buffer
except that the server stops reading from it until playback is
resumed.
- Muting has nothing to do with whether audio data is read or not.
- When the connection closes, the queues are unmapped on both sides.
This should already improve audio playback performance in a bunch of
places.
Implementation & commit notes:
- Audio loaders don't create LegacyBuffers anymore. LegacyBuffer is kept
for WavLoader, see previous commit message.
- Most intra-process audio data passing is done with FixedArray<Sample>
or Vector<Sample>.
- Improvements to most audio-enqueuing applications. (If necessary I can
try to extract some of the aplay improvements.)
- New APIs on LibAudio/ClientConnection which allows non-realtime
applications to enqueue audio in big chunks like before.
- Removal of status APIs from the audio server connection for
information that can be directly obtained from the shared queue.
- Split the pause playback API into two APIs with more intuitive names.
I know this is a large commit, and you can kinda tell from the commit
message. It's basically impossible to break this up without hacks, so
please forgive me. These are some of the best changes to the audio
subsystem and I hope that that makes up for this :yaktangle: commit.
:yakring:
With the following change in how we send audio, the old Buffer type is
not really needed anymore. However, moving WavLoader to the new system
is a bit more involved and out of the scope of this PR. Therefore, we
need to keep Buffer around, but to make it clear that it's the old
buffer type which will be removed soon, we rename it to LegacyBuffer.
Most of the users will be gone after the next commit anyways.
SystemServer had safety fallbacks to boot into text mode if the user
errorneously specified graphical mode but no video hardware was present.
As it's now possible to do exactly this intentionally, we should allow
it. This would of course make WindowServer fall over and die if
configured improperly, but if you're messing with the kernel command
line in strange ways, you should be able to fix that.
This screen backend is just memory-backed and doesn't connect to any
screen hardware. That way, we can boot Serenity without video hardware
but in full graphical mode :^)
To create a virtual screen, put something like this in your
WindowServer.ini. There's no way yet to do this through Display
Settings, though an existing virtual screen's settings can be changed
there.
```ini
[Screen0]
Mode=Virtual
Left=1024
Top=0
Width=1920
Height=1080
ScaleFactor=1
```
This will allow us to change between a couple of properties, for now
it's only Device and Virtual. (How about Remote :^) ) These get handled
by a different screen backend in the Screen.
The ScreenBackend is a thin wrapper around the actual screen hardware
connection. It contains all the variables specific to that hardware and
abstracts away operations that deal with controlling the hardware. The
standard ScreenBackend implementor is HardwareScreenBackend, which
contains all the existing frame buffer & ioctl handling code of Screen.
I took this opportunity to introduce ErrorOr wherever sensible.
This was very badly named. All that the "FBData" struct contains is the
currently to-be-flushed rectangles plus a fullness flag, so it should
better be called FlushRectData. This rename is similarly applied to all
variable names.
A mistake I've repeatedly made is along these lines:
```c++
auto nread = TRY(source_file->read(buffer));
TRY(destination_file->write(buffer));
```
It's a little clunky to have to create a Bytes or StringView from the
buffer's data pointer and the nread, and easy to forget and just use
the buffer. So, this patch changes the read() function to return a
Bytes of the data that were just read.
The other read_foo() methods will be modified in the same way in
subsequent commits.
Fixes#13687
This fixes issues with carriage return sequences.
Before, using <CR><NUL> as the return sequence wouldn't work at all,
and when using <CR><LF> there was an extra newline after every newline.
After this patch, the behaviour should be closer to the Telnet RFC.
If the .config directory (or its children, like lib) was deleted,
ConfigFile would crash because it would try to open or create a file in
a directory that didn't exist. Therefore, for user and library configs
(but not system configs), ensure that the parent directories exist. This
allows the user to delete the entire .config folder and all apps still
work. (Except those which can't handle missing config. That's a separate
issue though.)
Fixes#13555
Note: Some changes to pledges and unveils are necessary for this to
work. The only one who can recreate .config at the moment is
ConfigServer, as others probably don't pledge the user home directory.
This prevents a crash when the inspected process closes the socket (i.e.
when the app is exited). This crash was caused by the InspectableProcess
removing itself from the global process table within a callback Function
that is stored as part of the InspectableProcess.
Previously, GUI::Window::is_maximized() had to make a synchronous IPC
request to WindowServer in order to find out if the window was indeed
maximized.
This patch removes the need for synchronous IPC by instead pushing the
maximization state to clients when it changes.
The motivation for this change was that GUI::Statusbar was checking
if the containing window was maximized in its resize_event(), causing
all windows with a statusbar to block on sync IPC *during* resize.
Browser would typically block for ~15 milliseconds here every time
on my machine, continuously during live resize.
If there are pending paint requests waiting to be processed when the
client asks us to remove a backing store, we now prune them from the
request queue.
This avoids doing completely wasted painting work while resizing the
browser window. :^)
The storage inspector now has a new tab for local storage. The next step
would be to persist local storage and receive real-time notifications
for changes to update the table view.
With this change you can now set the theme and background color at the
same time in the Display Settings. Before if both were changed
before hitting 'apply' the theme background color would overwrite
the custom background.
Some ISPs may MITM DNS requests coming from clients, changing the case
of domain name in response. LookupServer will refuse responses from
any DNS server in that case. This commit changes the behaviour to
perform a case-insensitive equality check.
This is useful, for instance, in games in which you can switch held
items using the scroll wheel. In order to implement this, they
previously would have to either add a hard-coded division by 4, or look
up your mouse settings to adjust correctly.
This commit adds an MouseEvent.wheel_raw_delta_x() and
MouseEvent.wheel_raw_delta_y().
We now distribute the line-height evenly between the space above and
below inline-level boxes. This noticeably improves our baseline
alignment in many cases.
Note that the "vertical-align: <length>" case is quite awkward, as the
extra height added by the offset baseline must count towards the line
box height.
There's a lot of room for improvement here, but this makes the buckets
container on Acid3 show up in the right place, with the right size.
This converts the return value of File::read_link() from String to
ErrorOr<String>.
The rest of the change is to support the potential of an Error being
returned and subsequent release of the value when no Error is returned.
Unfortunately at this stage none of the places affected can utililize
our TRY() macro.
If an Animation's on_stop handler cleared itself inside the on_stop
event handler, it would remove itself from the animation map that was
still being iterated, leading to a crash.
To solve this, we'll iterate over the animations using the
remove_all_matching function, which enables us to delete it by simply
returning true when the animation finished. In the event that the
Animation is kept alive elsewhere and the on_stop event clears its own
reference, we need to temporarily bump the reference count. Another
advantage is that we only need to bump the reference count when an
animation is finished, whereas before this we bumped it
unconditionally.
`static const` variables can be computed and initialized at run-time
during initialization or the first time a function is called. Change
them to `static constexpr` to ensure they are computed at
compile-time.
This allows some removal of `strlen` because the length of the
`StringView` can be used which is pre-computed at compile-time.
It makes no sense to require passing a global object and doing a stack
space check in some cases where running out of stack is highly unlikely,
we can't recover from errors, and currently ignore the result anyway.
This is most commonly in constructors and when setting things up, rather
than regular function calls.
Let's make it very clear that these are *computed* values, and not at
all the specified values. The specified values are currently discarded
by the CSS cascade algorithm.
The "paintable" state in Layout::Box was actually not safe to access
until after layout had been performed.
As a first step towards making this harder to mess up accidentally,
this patch moves painting information from Layout::Box to a new class:
Painting::Box. Every layout can have a corresponding paint box, and
it holds the final used metrics determined by layout.
The paint box is created and populated by FormattingState::commit().
I've also added DOM::Node::paint_box() as a convenient way to access
the paint box (if available) of a given DOM node.
Going forward, I believe this will allow us to better separate data
that belongs to layout vs painting, and also open up opportunities
for naturally invalidating caches in the paint box (since it's
reconstituted by every layout.)
The blank string "" does not parse as JSON, and so the InspectorWidget
would fail to update the box-model information when inspecting elements
with no box, (for example, `<head>`) showing stale values instead. Now,
they show all 0s.
You could argue that InspectorWidget should be more resilient when given
invalid JSON strings, but making sure we only pass valid ones works
too.
This expands the InspectorWidget::Selection to include an optional
PseudoElement, which is then passed over IPC to request style
information for it.
As noted, this has some pretty big limitations because pseudo-elements
don't have DOM nodes:
- Declared style has to be recalculated when it's requested.
- We don't display the computed style.
- We don't display custom properties.
This Adds an element size preview widget to the inspector widget
in a new tab. This functions similar to chrome and firefox and
shows the margin, border, padding, and content size of the selected
element in the inspector.
The colors for the size preview widget are taken from the chrome
browser.
The ARGB32 typedef is used for 32-bit #AARRGGBB quadruplets. As such,
the name RGBA32 was misleading, so let's call it ARGB32 instead.
Since endianness is a thing, let's not encode any assumptions about byte
order in the name of this type. ARGB32 is basically a "machine word"
of color.
`CharacterBitmap` instances are generated at run-time and put on the
heap, but they can be created in a `constexpr` context and stored in
static memory.
Also, remove additional `width` and `height` `static` values in favor
of using the `constexpr` member functions of `CharacterBitmap`.
These changes also include the removal of some initialization code
which tests if the `CharacterBitmap` is created since it is always
created and removes function-local `static` values which cause
run-time branches to ensure it is initialized each time the function
is called.
This adds a keyboard event for Super+0 to Super+9. Later to be consumed
in the taskbar.
Currently only this keyboard sequence is supported:
- Super key down
- Digit key down
But not this:
- Super key down
- Digit key down
- Digit key up
- Digit key down
We need to set Window::m_invalidated_frame to true when invalidating
the title, otherwise we may miss re-rendering the frame if nothing
else triggers it.
This fixes a crash of the browser when loading any page. LibWeb
immediately pops the 'running execution context' after creating an
interpreter, but it's needed to have a 'current realm' during
initialization of the ConsoleGlobalObject for NativeFunction::create()
to work.
Once this is done, we can immediately pop the execution context again.
Currently this method always succeeds, but that won't be true once we
switch to the Core::Stream API. :^)
Some of these places would ideally show an error message to the user,
since failure to save a file is significant, but let's not get
distracted right now.
I've attempted to handle the errors gracefully where it was clear how to
do so, and simple, but a lot of this was just adding
`release_value_but_fixme_should_propagate_errors()` in places.
If we're waiting for the client (typically Browser) to respond to a
synchronous IPC message from our side (e.g window.alert()) and the
client disconnects instead, just exit peacefully.
Ultimately a WebContent process lives to serve its client. When the
client dies, there is no need for WebContent anymore.
The WindowServer _really_ does not need to know the filesystem path to
it's wallpaper, and allows setting arbitrary wallpapers (those outside
of `/res/wallpapers`).
The GUI::Desktop will keep track of the path to the wallpaper (if any),
and save it to config if desired (to be persisted).
This avoids the need to `unveil` paths to the wallpaper, fixing #11158
Calculating tiled and miximized window frame have a lot in common. In
fact, we can look at maximized window state as a special case of the
tile type. It simplifies the code since there is a lot of cases when
we take an action only if the window is maximized or tiled.
For now, just use the first audio channel in the /dev/audio directory.
In the future we can add support for watching and loading other channels
so we can route audio to multiple sound cards on the system.
I saw what looked like a UAF of this URL in a RequestServer crash,
and it seems reasonable to make a copy here since we end up passing
them to Core::deferred_invoke().
pledge_domains() that takes only one String argument was specifically
added as a shortcut for pledging a single domain. So, it makes sense to
use singular here.
The result of a SQL statement execution is either:
1. An error.
2. The list of rows inserted, deleted, selected, etc.
(2) is currently represented by a combination of the Result class and
the ResultSet list it holds. This worked okay, but issues start to
arise when trying to use Result in non-statement contexts (for example,
when introducing Result to SQL expression execution).
What we really need is for Result to be a thin wrapper that represents
both (1) and (2), and to not have any explicit members like a ResultSet.
So this commit removes ResultSet from Result, and introduces ResultOr,
which is just an alias for AK::ErrorOrr. Statement execution now returns
ResultOr<ResultSet> instead of Result. This further opens the door for
expression execution to return ResultOr<Value> in the future.
Lastly, this moves some other context held by Result over to ResultSet.
This includes the row count (which is really just the size of ResultSet)
and the command for which the result is for.
Previously we would re-run the entire CSS selector machinery for each
property resolved. Instead of doing that, we now resolve a final set of
custom property key/value pairs at the start of the cascade.
`it.is_end()` could be updated to return false for a previously-invalid
iterator after we append a new socket, copy its value out to a local
variable to not hit this behaviour.
The environment settings object is effectively the context a piece of
script is running under, for example, it contains the origin,
responsible document, realm, global object and event loop for the
current context. This effectively replaces ScriptExecutionContext, but
it cannot be removed in this commit as EventTarget still depends on it.
https://html.spec.whatwg.org/multipage/webappapis.html#environment-settings-object
Previously windows would end up in awkward positions relative to
the move cursor when dragging between tile types or unmaximizing.
This feels a bit more ergonomic.
VerticallyMaximized tiling replaces set_vertically_maximized() to
take advantage of tiling ergonomics.
Middle-clicking a window's maximize button now tiles vertically;
secondary-clicking tiles horizontally.
Adds Super+Alt+Arrow shortcuts for both. Super+Left/Right tiling
shortcuts now let windows shift between tile types directly.
Previously, different rects were used to restore tiled and maximized
windows, creating edge cases for inconsistent restoration. All states
now restore m_floating_rect, which saves the last valid size and
location of a window while free-floating.
Since VM::exception() no longer exists this is now useless. All of these
calls to clear_exception were just to clear the VM state after some
(potentially) failed evaluation and did not use the exception itself.
There's a possible window where the notifications are disabled, and any
request coming at that time will never get any data if it relies on
socket notifications.
Now that the GML formatter is both perserving comments and also mostly
agrees to the existing GML style, it can be used to auto-format all the
GML files in the system. This commit does not only contain the scripts
for running the formatting on CI and the pre-commit hook, but also
initially formats all the existing GML files so that the hook is
successfull.
This commit converts TLS::TLSv12 to a Core::Stream object, and in the
process allows TLS to now wrap other Core::Stream::Socket objects.
As a large part of LibHTTP and LibGemini depend on LibTLS's interface,
this also converts those to support Core::Stream, which leads to a
simplification of LibHTTP (as there's no need to care about the
underlying socket type anymore).
Note that RequestServer now controls the TLS socket options, which is a
better place anyway, as RS is the first receiver of the user-requested
options (though this is currently not particularly useful).
In order to avoid having multiple instances, we were keeping a pointer
to these singleton objects and only allocating them when it was null.
We have `__cxa_guard_{acquire,release}` in the userland, so there's no
need to do this dance, as the compiler will ensure that the constructors
are only called once.
We previously only replaced disconnected sockets on the queued-request
path, leading to attempts to send requests on a disconnected socket if
the disconnection happened in the deletion grace period.
This reverts most of commit ede5c9548e.
The one change not reverted is ClockWidget.h, so that the taskbar clock
can continue to notice time zone changes.
