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.
There was an awful lot of JS::Value <-> double conversion going on, even
through these AOs only work with number values anyway.
They don't need a global object either as they won't allocate or throw,
that was simply to pass it to infallible calls of ToIntegerOrInfinity.
In many cases we already know a certain value is a number, or don't have
JS values at all and would need to wrap doubles in a value. To optimize
these cases and avoid having to pass a global object into functions that
won't ever allocate or throw, add a standalone implementation of this
function that takes and returns doubles directly.
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.
Since the layout_box of a SVGGraphicsPaintable is already casted to the
right type, the underlying dom_node was also always of the right type.
This triggers a warning on Lagom builds.
This was changed a while ago so the generated files are placed in the
Build directory. Let's remove the gitignore file so any old version of
the files stop conflicting with the new generated versions.
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.
This implements the "close modified" icon on the terminal,
as well as several close warnings:
- A warning there is a foreground process running
- If there is a background process running
- Or if there are multiple background processes running
Fixes#13751
Instead of just printing 'ECMAScriptFunctionObject' (and leaking an
implementation detail in the process - this is not a public facing name)
let's instead print a different type string for each function kind, and
only keep the old class_name() printing for other JS::FunctionObject
subclasses.
The change in 3ec0183 wasn't actually correct, the spec tells us to set
the "prototype" property of the function (created with a prototype of
%GeneratorFunction.prototype% itself) to a newly created object:
OrdinaryObjectCreate(%GeneratorFunction.prototype.prototype%)
This patch removes the separate CookiesTab.gml file used to define the
layout of the cookies and local storage tabs and moves those defintions
into the StorageWidget.gml file.
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.
Clang didn't accept the friend declaration here, as the class has a
requires clause attached to it, and I couldn't immediately figure
out what it wants instead.
Add accessors for VectorN::m_data and use those where needed instead
for now.
ValueFormat::text_formatter is called with a u64 retrieved from
GraphWidget::m_values. However, the function pointer definition used
size_t and all the users of text_formatter used int. If bytes was over
~2 billion, we would interpret bytes to be negative. We then pass this
into `human_readable_size` which converts it to a u64, making it out to
be about 15.9 EiB.
This is fixed by making everything in the path take a u64.
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 DisplayConnector class is meant to replace the FramebufferDevice
class. The advantage of this class over the FramebufferDevice class is:
1. It removes the mmap interface entirely. This interface is unsafe, as
multiple processes could try to use it, and when switching to and from
text console mode, there's no "good" way to revoke a memory mapping from
this interface, let alone when there are multiple processes that call
this interface. Therefore, in the DisplayConnector class there's no
implementation for this method at all.
2. The class uses a new real-world structure called ModeSetting, which
takes into account the fact that real hardware requires more than width,
height and pitch settings to mode-set the display resolution.
3. The class assumes all instances should supply some sort of EDID,
so it facilitates such mechanism to do so. Even if a given driver does
not know what is the actual EDID, it will ask to create default-generic
EDID blob.
3. This class shifts the responsibilies of switching between console
mode and graphical mode from a GraphicsAdapter to the DisplayConnector
class, so when doing the switch, the GraphicsManagement code actually
asks each DisplayConnector object to do the switch and doesn't rely on
the GraphicsAdapter objects at all.
Our move to floating point precision has eradicated the pixel artifacts
in Quake 1, but introduced new and not so subtle rendering glitches in
games like Tux Racer. This commit changes three things to get the best
of both worlds:
1. Subpixel logic based on `i32` types was reintroduced, the number of
bits is set to 6. This reintroduces the artifacts in Quake 1 but
fixes rendering of Tux Racer.
2. Before triangle culling, subpixel coordinates are calculated and
stored in `Triangle`. These coordinates are rounded, which fixes the
Quake 1 artifacts. Tux Racer is unaffected.
3. The triangle area (actually parallelogram area) is also stored in
`Triangle` so we don't need to recalculate it later on. In our
previous subpixel code, there was a subtle disconnect between the
two calculations (one with and one without subpixel precision) which
resulted in triangles incorrectly being culled. This fixes some
remaining Quake 1 artifacts.
We were splitting these API wrappers up into different files without a
quantifiable benefit. Now, it's extremely clear where the direct API
implementation lives. :^)
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 completely removes WavLoader's dependency on LegacyBuffer: We
directly create the result sample container and write into it. I took
this opportunity to rewrite most of the sample reading functions as a
single templated function, which combined with the better error handling
makes this "ported" code super concise.
This makes the code much more readable and concise, reduces the size of
the WavLoader class itself, moves almost all fallible initialization out
of the constructor and should provide better error handling in general.
Also, a lot of now-unnecessary imports are removed.
* All clang-tidy warnings fixed except read_header cognitive complexity
* Use size_t in more places
* Replace #define's with constexpr constants
* Some variable renaming for readability
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 option sets -fprofile-instr-generate -fcoverage-mapping for Clang
builds only on almost all of Userland. Loader and LibTimeZone are
exempt. This can be used for generating code coverage reports, or even
PGO in the future.
This commit addresses the following shortcomings of our current, simple
and elegant memset function:
- REP STOSB/STOSQ has considerable startup overhead, it's impractical to
use for smaller sizes.
- Up until very recently, AMD CPUs didn't have support for "Enhanced REP
MOVSB/STOSB", so it performed pretty poorly on them.
With this commit applied, I could measure a ~5% decrease in `test-js`'s
runtime when I used qemu's TCG backend. The implementation is based on
the following article from Microsoft:
https://msrc-blog.microsoft.com/2021/01/11/building-faster-amd64-memset-routines
Two versions of the routine are implemented: one that uses the ERMS
extension mentioned above, and one that performs plain SSE stores. The
version appropriate for the CPU is selected at load time using an IFUNC.
.text sections of objects that contain textrels have to be writable
during the relocation procedure. Because of this, we would segfault if
we tried to execute IFUNC resolvers defined in them. Let's print a
meaningful error message instead.
Additionally, a warning is now printed when we load objects with
textrels, as in the future, additional security mitigations might
interfere with them being loaded.
IFUNC is a GNU extension to the ELF standard that allows a function to
have multiple implementations. A resolver function has to be called at
load time to choose the right one to use. The PLT will contain the entry
to the resolved function, so branching and more indirect jumps can be
avoided at run-time.
This mechanism is usually used when a routine can be made faster using
CPU features that are available in only some models, and a fallback
implementation has to exist for others.
We will use this feature to have two separate memset implementations for
CPUs with and without ERMS (Enhanced REP MOVSB/STOSB) support.
IFUNC resolvers depend on the resolved function's address having been
relocated by the time they are called. This means that relative
relocations have to be done first.
The linker is kind enough to put R_*_RELATIVE before R_*_IRELATIVE in
.rel.dyn, but .relr.dyn contains relative relocations too.
This check is here to make sure we only try to load serenity binaries.
However, with -fprofile-instr-generate -fcoverage-mapping, clang
sets the EI_OSABI field to 3, for GNU. The instrumentation uses a lot of
retained COMDAT sections for coverage instrumentation that get the
SHF_GNU_RETAINED section header flag set on them, forcing llvm to set
the ABI to GNU.
A function object's realm is not necessarily non-null (like when called
via the Reflect API), so we can't blindly dereference it. Instead use
the object's own GlobalObject.
Because `TimeHour TimeZoneNumericUTCOffsetNotAmbiguous[opt]
TimeZoneBracketedAnnotation[opt]` can be a subset of
`TimeHourNotValidMonth TimeZone` we would not exhaust the whole input
in some cases, which would result in an incorrectly thrown exception.
We only need to fire these callbacks when we want to change the palette.
So, when setting widget values to match a palette that has just been
loaded, or to show the value of the role selected in a ComboBox, we can
skip the callbacks. This saves some work, and means we can reliably use
on_palette_change to know when the palette has actually been modified.
There are two different things in ThemeEditor that want to know when a
palette changes:
1. The PreviewWidget subclass, so it can update its preview.
2. The ThemeEditor itself, so we know that the palette is modified.
Using a protected virtual function for 1 means that we can do 2 without
them fighting over the same on_palette_change callback.
The `TextEditor::on_change` callback now only fires if the user types in
the box, or `set_text()` is called with `AllowCallback::Yes`.
Previously that callback was what set `m_color`, so I've rearranged
things a little so that the color still updates regardless of what
source the color came from.
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.
...and change how the two parsing steps fit together.
The two steps were previously quite muddled. Both worked with the
TokenStream directly, and both were responsible for rewinding that
stream if there was an error. This is both confusing and also made it
impossible to replace the rewinding with StateTransactions.
This commit more clearly divides the work between the two functions: One
parses ComponentValues and produces a string, and the other parses that
string to produce the UnicodeRange. It also replaces manual rewinding
in the former with StateTransactions.
This should be a bit easier to follow.
parse_media_query() no longer rewinds if the media query is invalid,
because it then interprets all the tokens as a "not all" query.
This is modeled after the one in ISO8601Parser. It rolls back the
TokenStream state automatically at the end of scope unless told to
commit the changes. This should be less error-prone than remembering to
manually call `rewind_to_position()` at the correct time.
For convenience, a StateTransaction can have "child" transactions. When
a transaction is committed, it automatically commits its parents too.
This is useful in situations where you have several nested and don't
want to have to remember to manually `commit()` them all.
`a` and `b` had to be declared at the top of the function before since
they were used by the `make_return_value()` lambda. But now that
doesn't exist, we can move them to where they are used - or eliminate
them entirely.
parse_a_n_plus_b_pattern()'s job is to parse as much of the TokenStream
as it can as a An+B, and then stop. The caller can then deal with any
trailing tokens as it wishes.
...using a ParseErrorOr type alias.
This lets us replace a bunch of manual error-checking with TRY. :^)
I also replaced the ParsingResult::Done value with returning an
Optional. I wasn't happy with treating "Done" as an error when I first
wrote this, and this makes a clear distinction between the two.
Using ~~text~~ syntax will strike out the text between the two tildes.
Only missing portion is the terminal rendering of strike through text.
The ansi escape codes for strike through text are \e[9m and \e[29m but
it appears the terminal does not support these. Please correct me if I
am wrong.
I tested that the render_to_terminal function was being called by giving
it bold ANSI escape codes, and that did work so the function is being
called correctly.
The main event loop functionality was used in just two places where the
alternative is a bit simpler. Remove it in favor of referencing the
event loop directly, or just invoking `EventLoop::current()`.
Note that we don't need locking in the constructor since we're now only
modifying a thread-local `Vector`. We also don't need locking in the
old call sites to `::with_main_locked()` since we already lock the
event loop in the subsequent `::post_event()` invocation.
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.
Previously we would quit the event loop of a dialog without checking if
it is still alive in the Window::close overload. This patch updates the
implementation to make use of the existing done method, handling closes
more gracefully.
This fixes a CommandPalette crashing when opening an about dialog, as
since 1074c399f3 the command palette
dialog would handle a WindowBecameInactive event after closing itself
due to the action already being called.
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).
The issue was caused by the usage of the
selection_end_column_within_line variable as if it was the visual line.
This is fixed by taking the minimum between this value and the length of
a visual line.
Previously, when you selected to have a custom format, whatever was in
the custom-format box would get replaced with the format for
12-hour-without-seconds. Now, it keeps whatever was in the box before -
which is less disorientating, and lets you tweak the existing format.
This means that if you save and close ClockSettings with "24-hour"
and "Show seconds" both checked, then they will both be checked when
you re-open ClockSettings, instead of it showing as a "Custom" format.
Similarly to x86_64, Aarch64 is LP64, so its `uint64_t` type is
`unsigned long`.
Fixes a bunch of compiler warnings when compiling the LLVM runtime
libraries for the aarch64-pc-serenity target.
Following FIXMEs have been addressed:
- 1. Let settingsObject be this’s relevant settings object.
- 2. If settingsObject has a responsible document and it is not fully
active, then throw an "InvalidStateError" DOMException.
- 6. Let parsedURL be the result of parsing url with settingsObject's
API base URL and settingsObject’s API URL character encoding.
- 8. If the async argument is omitted, set async to true, and set
username and password to null.
This adds a simple histogram widget that visualizes the rgb-channels
and brightness for a given image. When hovering over the image it will
indicate what brightness level the pixel at the mouse position has.
While working on #13764 I noticed that DateTime::to_string() would just
return an empty String if the format included an invalid specifier
(eg `%Q`). This seems to be a mistake. POSIX date(1), which I believe
we are basing our implementation on, only replaces valid specifiers,
and any invalid ones get included as literals in the output.
For example, on Linux `date "+%Quiz"` returns "%Quiz", but we were
returning "".
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.
This new class with an admittedly long OOP-y name provides a circular
queue in shared memory. The queue is a lock-free synchronous queue
implemented with atomics, and its implementation is significantly
simplified by only accounting for one producer (and multiple consumers).
It is intended to be used as a producer-consumer communication
datastructure across processes. The original motivation behind this
class is efficient short-period transfer of audio data in userspace.
This class includes formal proofs of several correctness properties of
the main queue operations `enqueue` and `dequeue`. These proofs are not
100% complete in their existing form as the invariants they depend on
are "handwaved". This seems fine to me right now, as any proof is better
than no proof :^). Anyways, the proofs should build confidence that the
implemented algorithms, which are only roughly based on existing work,
operate correctly in even the worst-case concurrency scenarios.
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.
Previously netstat would print the whole line of an ip address or
resolved hostname. If the hostname was longer than the address column
length, it would push following columns into disaligned output.
This sets the default behavior to truncate any IP address or symbolic
hostname that is larger than the maximum address column size to provide
cleaner output. In the event the user wishes to see the whole address
name, they can then pass the wide option that will output as wide as
necessary to print the whole name.
- Don't add multiple numbers to nested steps, just the innermost one
(as rendered in the HTML document)
- "Otherwise" comments go before the else, not after it
- "FIXME:" goes before step number, not between it and the comment text
- Always add a period between number and comment text
The majority of these were introduced in #13756, but some unrelated ones
have been updated as well.
Currently, navigating through different years in the Year view of the
Calendar app or the taskbar is very slow.
Profiling results show that almost all the time is spent in
`Calendar::update_tiles`, and specifically, in `DateTime::create` and
`DateTime::set_time`.
Performance can improve substantially if the `TZ` environment variable
is set [0], but we can improve the current code to perform better
nevertheless :^)
This diff focuses on removing the need of the `Tile` struct to require
the instantiation of a `DateTime` object, which avoids _at least_ 365
object instantiations in the Year view, on each `update_tiles` call.
Instead, as the `date_time` isn't used often, we can instantiate it on
demand when a particular date is selected.
[0] https://blog.packagecloud.io/set-environment-variable-save-thousands-of-system-calls/
This patch changes the previously used contains method for matching the
user search term with all available commands to use the fuzzy match
algorithm, which makes it more typo tolerant.
This patch adds a header containing the fuzzy match algorithm
previously used in Assistant. The algorithm was moved to AK
since there are many places where a search may benefit from fuzzyness.
