According to Dr. POSIX, we should allow to call mmap on inodes even on
ranges that currently don't map to any actual data. Trying to read or
write to those ranges should result in SIGBUS being sent to the thread
that did violating memory access.
We make these methods non-virtual because we want to ensure we properly
enforce locking of the m_inode_lock mutex. Also, for write operations,
we want to call prepare_to_write_data before the actual write. The
previous design required us to ensure the callers do that at various
places which lead to hard-to-find bugs. By moving everything to a place
where we call prepare_to_write_data only once, we eliminate a possibilty
of forgeting to call it on some code path in the kernel.
The block list required a bit of work, and now the only method being
declared const to bypass its const-iness is the read_bytes method that
calls a new method called compute_block_list_with_exclusive_locking that
takes care of proper locking before trying to update the block list data
of the ext2 inode.
This script is useful when wanting to install lagom libraries for
projects using Lagom via FetchContent, but trips over itself if the
project links other non-Lagom imported targets to itself. So, let's just
skip them.
This implements all the filters other than `saturate()`,
`hue-rotate()`, and `drop-shadow()`.
There are still a lot of FIXMEs to handle in the actual implementation
though, particularly around supporting transforms, but this handles
the most common use cases :^)
This will be needed so we can apply filter effects to the backdrop
of an element in LibWeb.
This now also allows getting a crop of a bitmap in a different format
than the source bitmap. This is for if the painter's bitmap does not
have an alpha channel, but you want to ensure the cropped bitmap does.
This amount can be handled in the filter's implementation or if
not it will default to mixing between the new and previous pixel.
This behaviour is used for implementing CSS filters that allow stuff
like grayscale(70%).
This style value holds a list of CSS filter function calls e.g.
blur(10px) invert() grayscale()
It will be used to implement backdrop-filter, but the same style value
can be used for the image filter property.
(The name is a little awkward but it's referenced to as
filter-value-list in the spec too).
We now have a proper aligned allocation implementation, and the
toolchain patch to make Clang use the intermediary implementation
has already been removed in an earlier iteration.
Those have been added to the Clang build script in
d4d6f2d945, and now that the Clang version
has been updated, everyone should have those symlinks already anyways.
LLVM 15 switched around what it's basing its `nullptr_t` definitions on,
it's now defining `std::nullptr_t` using `::nullptr_t` instead of the
other way around.
Work around any errors that result from that by just defining it both in
the global namespace as well as in `std` ourselves.
This caused the m_allocation_enabled_previously member to be technically
uninitialized when the compiler emits the implicit destructor call for
stack allocated classes.
This was pointed out by gcc on lagom builds, no clue how this was flying
under the radar for so long and is not triggering CI.
When matching selectors in HTML documents, we know that all the elements
have lowercase local names already (the parser makes sure of this.)
Style sheets still need to remember the original name strings, in case
we want to match against non-HTML content like XML/SVG. To make the
common HTML case faster, we now cache a lowercase version of the name
with each type/class/id SimpleSelector.
This makes tag type checks O(1) instead of O(n).
This remained undetected for a long time as HeaderCheck is disabled by
default. This commit makes the following file compile again:
// file: compile_me.cpp
#include <LibWeb/CSS/GridTrackSize.h>
// That's it, this was enough to cause a compilation error.
The purpose of this patch is to support addition, subtraction,
multiplication and division without using conversion to double. To this
end, we use the BigFraction class of LibCrypto. With this solution, we
can store values without any losses and forward rounding as the last
step before displaying.
SignedBigInteger::operator==(const UnsignedBigInteger&) was rejecting
all negative value before testing for equality. It now accepts negative
zero and test for a value equality with the UnsignedBigInteger.
This mainly changes two aspects:
- The category can now be a single letter, such as 'w' to indicate the
file Utilities/w.cpp
- Spaces in the category (or list) are no longer allowed. This follows
the lived practice of writing category lists as "Foo+Bar: Quux"
Closes#15243.
Instead of just keeping them in an unsorted Vector, which led to
increasingly noticeable O(n) lookups, we now cache a list of Typefaces
per family name.
The previous version relied on manually setting the amount of data to
read for the next chunk and was overall unclear. The new version uses
the Bytes API to vastly improve readability, and fixes a bug where
reading from files where a single read that wasn't of equal size to the
block size would cause the byte buffer to be incorrectly resized causing
corrupted output.
The document.domain setter is currently stubbed as that is a doozy to
implement, given how much restrictions there are in place to try and
prevent use of it and potential multi-process implications.
This was the only thing preventing us from being able to start
displaying ads delivered via Google Syndication.
In a subclass of Cell, we cannot use Cell::vm() before the base Cell
object itself is constructed. Use the Realm's VM instead.
This was caught by UBSAN with vptr sanitation enabled.
IsArray returns true if the object is an Array *or* if it is a
ProxyObject whose target is an Array. Therefore, we cannot downcast to
an Array based on IsArray.
Luckily, we don't actually need an Array here; SerializeJSONArray only
needs an Object.
This was caught by UBSAN with vptr sanitation enabled.
