By aligning Array.prototype.sort with the spec, this removes the direct
invocation of CompareArrayElements from array_merge_sort. This opens the
door for TypedArray to use SortIndexedProperties as well because now
array_merge_sort does not assume what kind of array it is invoked with.
Further, this addresses a FIXME to avoid an extra JS heap allocation.
This isn't actually much of an issue because if the LHS side value is -0
and the RHS value is +0, errantly returning 0 in the comparison function
here has the same effect as correctly returning -1. In both cases, the
-0 is left on the LHS.
Remove the static PNG bitmaps we've been using for GUI radio buttons
and replace them with on-the-fly pixel painting.
This fixes a long-standing issue where radio buttons always looked the
same, regardless of system theme settings. :^)
This is not the most useful keyboard binding anyways, plus it will be
extremely hacky to implement it with the generic processor parameter
widgets. Therefore, we'll get rid of it and add back a more generic
keyboard binding system later.
This automatically creates the correct collection of name label, value
label and "editor" (knob, checkbox, dropdown) depending on the processor
type and layouts them vertically.
If we wait until after the parent context has laid out the flex
container, abspos children are able to use the final results of the
parent sizing the flex container.
This makes `height:auto` work on abspos children of a flex container.
Layout cleared the list of bars in the flame graph, but didn't clear the
reference m_hovered_bar. This could cause a crash in mousedown_event()
when clicking twice: the first click caused layout, the second used
the old reference.
This now shows the shortcut to enable mouse pointer highlighting below
the preview. The GML has also been updated to take advantage of the new
layout features rather than using fixed heights.
Previously we would always select the left most column when selecting a
range of rows.
This patch fixes this issue by always applying a selection to the column
in which the selection ends.
AbstractView currently assumes a certain layout of rows and colums to
perform a range selection on.
This patch moves the implementation into a helper that can be overriden
by other views.
We should actually start counting from the parent directory and not from
the symbolic link as it will represent a wrong count of hops from the
actual mountpoint.
The symlinks in /sys/dev/block and /sys/dev/char worked only by luck,
because I have set it to the wrong parent directory which is the
/sys/dev directory, so with the symlink it was 3 hops to /sys, together
with the root directory, therefore, everything seemed to work.
Now that the device symlinks in /sys/dev/block and /sys/dev/char are set
to the right parent directory and we start measure hops from root
directory with the parent directory of a symlink, everything seem to
work correctly now.
All of these conditions should make du just not report the file size
individually, but it should still count them into the grand total.
In the case of the `--threshold` option, this was actually implemented
incorrectly before, as it would report size 0 for files that did not
match the threshold.
We may very well dip into files larger than 4G at some point, so 32-bit
values are not enough, and the 64-bit sized `off_t` doesn't fully make
sense either, as it features negative values.
Instead, switch to the explicit type of `u64` everywhere, which is the
same size on all platforms and is unsigned. The exception to this is
the threshold, which needs to be signed instead of unsigned.
The `--max-depth` option only controls until which depth individual file
sizes are printed, it does not stop the utility from traversing that
branch further (as the file sizes would be wrong otherwise).
Restructure the program to track the current depth, and return early if
the current depth is higher than the maximum allowed depth, which skips
all parts of the logic that are concerned with user output.
This makes them look a bit more like a progress bar, especially
on white backgrounds (for the default theme) where you otherwise
cannot see the unfilled part of the bar.
Now that the infrastructure of the Graphics subsystem is quite stable,
it is time to try to fix a long-standing problem, which is the lack of
locking on display connector devices. Reading and writing from multiple
processes to a framebuffer controlled by the display connector is not a
huge problem - it could be solved with POSIX locking.
The real problem is some program that will try to do ioctl operations on
a display connector without the WindowServer being aware of that which
can lead to very bad situations, for example - assuming a framebuffer is
encoded at a known resolution and certain display timings, but another
process changed the ModeSetting of the display connector, leading to
inconsistency on the properties of the current ModeSetting.
To solve this, there's a new "master" ioctl to take "ownership" and
another one to release that ownership of a display connector device. To
ensure we will not hold a Process object forever just because it has an
ownership over a display connector, we hold it with a weak reference,
and if the process is gone, someone else can take an ownership.
This header file represents the entire interface between the kernel and
userland, and as such, no longer should be called FB.h but something
that represents the whole graphics subsystem.
