This commit replaces SinglyLinkedListIterator::universal_end() with an
empty SinglyLinkedListIterator(). Piano needs this in order to
initialize a member array of iterators without 84 lines of
universal_end().
set_interrupted_by_death was never called whenever a thread that had
a joiner died, so the joiner remained with the joinee pointer there,
resulting in an assertion fail in JoinBlocker: m_joinee pointed to
a freed task, filled with garbage.
Thread::current->m_joinee may not be valid after the unblock
Properly return the joinee exit value to the joiner thread.
On 32-bit platforms, INT32_MIN == -INT32_MIN, so we can't expect this
to always work:
if (pid < 0)
positive_pid = -pid; // may still be negative!
This happens because the -INT32_MIN expression becomes a long and is
then truncated back to an int.
Fixes#1312.
It was possible to send signals to processes that you were normally not
allowed to send signals to, by calling ioctl(tty, TIOCSPGRP, targetpid)
and then generating one of the TTY-related signals on the calling
process's TTY (e.g by pressing ^C, ^Z, etc.)
This allows a process wich has more than 1 thread to call exec, even
from a thread. This kills all the other threads, but it won't wait for
them to finish, just makes sure that they are not in a running/runable
state.
In the case where a thread does exec, the new program PID will be the
thread TID, to keep the PID == TID in the new process.
This introduces a new function inside the Process class,
kill_threads_except_self which is called on exit() too (exit with
multiple threads wasn't properly working either).
Inside the Lock class, there is the need for a new function,
clear_waiters, which removes all the waiters from the
Process::big_lock. This is needed since after a exit/exec, there should
be no other threads waiting for this lock, the threads should be simply
killed. Only queued threads should wait for this lock at this point,
since blocked threads are handled in set_should_die.
Previously, this function was using `AK::String::to_uint()`, which is
wrong considering the function returns type `int`. This also means that
configuration files would revert to the default value on negative
values.
This addresses the issue found in #1261. Previously we weren't doing
any form of sanity checking on the values in the configuration file,
meaning that a user could input a ridiculous number of mines such that
the number of mines is larger than the number of squares on the field.
This patch adds a new column to the per-process memory regions view in
SystemMonitor. It's a scaled view of the underlying pagemap of a region
that tells you which chunks of the region are resident/null/zero.
If a directory is renamed or deleted before 'make clean', git will
delete the Makefile but leave all of the object and dependency files
around. When make would try to recurse into that directory from the
wildcard, it would error out since there is no Makefile.
AbstractView does not know which column it's displaying which makes it
impossible to implement the select_all functionality up there. Now
descendants override the pure virtual select_all method and implement
it themselves.
This should make stuff like placement new work correctly when building
outside of Serenity. This stuff is a bit delicate due to the weirdly
staged toolchain build at the moment. Hopefully we can unify this stuff
in the future.
SystemTheme.h:81:1: error: 'SystemTheme' defined as a struct here but previously declared as a class; this is valid, but may result in linker errors under the Microsoft C++ ABI
We were only forwarding the value to the subviews, but not storing it
in m_model_column. This would cause MultiView::model_column() to return
the wrong value.
Thanks to Daniel Bos for spotting this! :^)
In the old model, before bc319d9e88, the parent
(the background thread) would delete us when it exits (i.e. never), so we had to
keep track of our own refcount in order to destroy ourselves when we're done.
With bc319d9e88, the parent keeps additional
reference to us, so:
* There should be no need to explicitly ref() ourselves
* The unref() would not get rid of the last reference to us anymore
The latter is why all the BackgroundAction's were getting leaked. Fix this by
simply unparenting ourselves from the background thread when we're done.
FilePicker was not showing thumbnails correctly because once each
thumbnail rendering BackgroundAction completed, it posted a deferred
invocation event to the *main* event loop.
Since FilePicker runs in a nested event loop, those completion
callbacks never ran until it was too late and the FilePicker was gone.
There is some sort of issue with using a SortingProxyModel together
with ColumnsView. This is a workaround to allow FilePicker to use a
MultiView for now, but this needs to be fixed separately somehow.
A MultiView is a combination of ItemView, TableView and ColumnsView
smashed into a single widget. You can switch between the view modes
by calling MultiView::set_view_mode().
Note that MultiView inherits from StackWidget, not AbstractView.
That's purely for practical reasons, although I'm not entirely sure
if there would be some benefit to having it inherit from AbstractView.
This is a little bit awkward since it's only used for generating
thumbnails on a background thread and it's not like I care about
thumbnails very much in a text editor, but for now let's just pledge
"thread" so I can get on with the thing I wanted to get on with.
Previously it was only possible to change these window attributes when
creating a new window. This patch adds an IPC message that allows you
to change them at runtime.
