Add an ExpandableHeap and switch kmalloc to use it, which allows
for the kmalloc heap to grow as needed.
In order to make heap expansion to work, we keep around a 1 MiB backup
memory region, because creating a region would require space in the
same heap. This means, the heap will grow as soon as the reported
utilization is less than 1 MiB. It will also return memory if an entire
subheap is no longer needed, although that is rarely possible.
The move constructor of a lambda just copies it anyway.
Even if the first move() left an 'empty' closure behind, then
'm_editor->on_cursor_change' would only be able to see an empty
closure, which is certainly not what was intended.
Under the hood, a lambda is just a struct full of pointers/references/copies and whatever else
the compiler deems necessary. In the case of 'update_demo', the struct lives on the stack
frame of FontEditorWidget::FontEditorWidget(). Hence it is still alive when it's called
during the constructor.
However, when 'fixed_width_checkbox.on_checked' fires, that stack frame is no longer alive,
and thus the *reference* to the (struct of) the lambda is invalid\! This meant that
'update_demo' silently read invalid data, tried to call '.update()' on some innocent arbitrary
memory address, and it crashed somewhere unrelated.
Passing 'update_demo' by value (like with all the other event handlers) fixes this issue.
Note that this solution only works because 'update_demo' itself has no state; otherwise
the various copies of 'update_demo' might notice that they are, in fact, independent copies
of the original lambda. But that doesn't matter here.
GIFLoader now uses a single frame buffer to cache the last decoded
frame. This drastically reduces memory usage at the small expense of
re-decoding frames on each loop.
RestoreBackground disposal mode is now a transparent fill to allow
background to show through.
RestorePrevious disposal mode now restores the previous frame.
This avoids silly problems like broken debug code. When debugging, I don't first
want to have to debug the debug code! :P
With all debug options enabled, `test-js` outputs 858081 lines. This is too much
for Travis, so we need to throw the debug output away on Travis. Note that
this still preserves information like *which* test failed, as well as the
precise error message.
A change introduced in 5e01234 made it the resposibility of each
filesystem to have the file types returned from
'traverse_as_directory' match up with the DT_* types.
However, this caused corruption of the Ext2FS file format because
the Ext2FS uses 'traverse_as_directory' internally when manipulating
the file system. The result was a mixture between EXT2_FT_* and DT_*
file types in the internal Ext2FS structures.
Starting with this commit, the conversion from internal filesystem file
types to the user facing DT_* types happens at a later stage,
in the 'FileDescription::get_dir_entries' function which is directly
used by sys$get_dir_entries.
Until we have better control over cell content alignment, let's make
them all right-aligned by default since that makes numbers look nice,
and numbers are the bread & butter of spreadsheets. :^)
