80626e7584
We currently check for the IN_DESTRUCTION flag inside the
add_child_internal() function.
This check disallows calling methods that change the stacking order
within the destruction sequence, by triggering a critical warning first,
and leaving the actor in an undefined state, which then ends up being
caught by an assertion.
The reproducible sequence is:
- actor gets destroyed;
- another actor, linked to the first, will try to change the
stacking order of the first actor;
- changing the stacking order is a composite operation composed
by the following steps:
1. ref() the child;
2. remove_child_internal(), which removes the reference;
3. add_child_internal(), which adds a reference;
- the state of the actor is not changed between (2) and (3), as
it could be an expensive recomputation;
- if (3) bails out, then the actor is in an undefined state, but
still alive;
- the destruction sequence terminates, but the actor is unparented
while its state indicates being parented instead.
- assertion failure.
The obvious fix would be to decompose each set_child_*_sibling() method
into proper remove_child()/add_child(), with state validation; this may
cause excessive work, though, and trigger a cascade of other bugs in
code that assumes that a change in the stacking order is an atomic
operation.
Another potential fix is to just remove this check here, and let code
doing stacking order changes inside the destruction sequence of an actor
continue doing the work.
The third fix is to silently bail out early from every
set_child_*_sibling() and set_child_at_index() method, and avoid doing
work.
I have a preference for the second solution, since it involves the least
amount of work, and the least amount of code duplication.
See bug: https://bugzilla.gnome.org/show_bug.cgi?id=670647
|
||
|---|---|---|
| build | ||
| clutter | ||
| doc | ||
| po | ||
| tests | ||
| .gitignore | ||
| autogen.sh | ||
| ChangeLog.pre-git-import | ||
| clutter.doap | ||
| config.h.win32.in | ||
| configure.ac | ||
| COPYING | ||
| Makefile.am | ||
| NEWS | ||
| README.in | ||
| README.md | ||
Clutter
What is Clutter?
Clutter is an open source software library for creating fast, compelling, portable, and dynamic graphical user interfaces.
Requirements
Clutter currently requires:
On X11, Clutter depends on the following extensions:
- XComposite
- XDamage
- XExt
- XFixes
- XInput (1.x or 2.x)
- XKB
If you are building the API reference you will also need:
If you are building the additional documentation you will also need:
- xsltproc
- jw (optional, for generating PDFs)
If you are building the Introspection data you will also need:
If you want support for profiling Clutter you will also need:
Resources
The official Clutter website is:
http://www.clutter-project.org/
The API references for the latest stable release are available at:
http://docs.clutter-project.org/docs/clutter/stable/
http://docs.clutter-project.org/docs/cogl/stable/
http://docs.clutter-project.org/docs/cally/stable/
The Clutter Cookbook is available at:
http://docs.clutter-project.org/docs/clutter-cookbook/
New releases of Clutter are available at:
http://source.clutter-project.org/sources/clutter/
The Clutter blog is available at:
http://www.clutter-project.org/blog/
To subscribe to the Clutter mailing lists and read the archives, use the Mailman web interface available at:
http://lists.clutter-project.org/
New bug page on Bugzilla:
http://bugzilla.gnome.org/enter_bug.cgi?product=clutter
Clutter is licensed under the terms of the GNU Lesser General Public
License, version 2.1 or (at your option) later: see the COPYING file
for more information.
Building and Installation
To build Clutter from a release tarball, the usual autotool triad should be followed:
- ./configure
- make
- make install
To build Clutter from a Git clone, run the autogen.sh script instead
of the configure one. The autogen.sh script will run the configure script
for you, unless the NOCONFIGURE environment variable is set to a non-empty
value.
See also the BuildingClutter page on the wiki.
Versioning
Clutter uses the common "Linux kernel" versioning system, where even-numbered minor versions are stable and odd-numbered minor versions are development snapshots.
Different major versions break both API and ABI but are parallel installable. The same major version with differing minor version is expected to be ABI compatible with other minor versions; differing micro versions are meant just for bug fixing. On odd minor versions the newly added API might still change.
The micro version indicates the origin of the release: even micro numbers are only used for released archives; odd micro numbers are only used on the Git repository.
Contributing
If you want to hack on and improve Clutter check the HACKING file for
general implementation guidelines, and the HACKING.backends for
backend-specific implementation issues.
The CODING_STYLE file contains the rules for writing code conformant to
the style guidelines used throughout Clutter. Remember: the coding style
is mandatory; patches not conforming to it will be rejected by default.
The usual workflow for contributions should be:
- Fork the repository
- Create a branch (
git checkout -b my_work) - Commit your changes (
git commit -am "Added my awesome feature") - Push to the branch (
git push origin my_work) - Create an Bug with a link to your branch
- Sit back, relax and wait for feedback and eventual merge
Bugs
Bugs should be reported to the Clutter Bugzilla at:
http://bugzilla.gnome.org/enter_bug.cgi?product=clutter
You will need a Bugzilla account.
In the report you should include:
- what system you're running Clutter on;
- which version of Clutter you are using;
- which version of GLib and OpenGL (or OpenGL ES) you are using;
- which video card and which drivers you are using, including output of glxinfo and xdpyinfo (if applicable);
- how to reproduce the bug.
If you cannot reproduce the bug with one of the tests that come with Clutter source code, you should include a small test case displaying the bad behaviour.
If the bug exposes a crash, the exact text printed out and a stack trace obtained using gdb are greatly appreciated.