<h1>Linux stable releases</h1>
<ul>
<li>
- Linus releases a 'stable' 3.<var>x</var> every 2-3 months after
+ Linus releases a stable 3.<var>x</var> every 2-3 months after
a series of release candidates
</li>
<li>
- Greg K-H maintains a 'stable branch' for each Linus release, with
- releases numbered 3.<var>x</var>.<var>y</var>
+ Greg K-H maintains a <dfn>stable branch</dfn> for each Linus
+ release, with <dfn>stable update</dfn> releases numbered
+ 3.<var>x</var>.<var>y</var>
</li>
<li>
- Stable branch includes cherry-picked or backported changes to
- fix serious bugs and add support for new devices
+ Branch includes cherry-picked or backported changes to fix
+ serious bugs and add support for new devices
</li>
<li>
- Stable update releases roughly every 1-2 weeks, at least until
+ Updates released roughly every 1-2 weeks, at least until
3.<var>x</var>+1 is available
</li>
<li>
- A 'long-term' stable branch may be maintained by Greg or another
- developer beyond this period
+ A <dfn>longterm</dfn> stable branch may be maintained by Greg or
+ another developer beyond this period (I look after 3.2)
</li>
</ul>
</div>
notified and may provide a backported version
</li>
<li>
- Any commit in Linus's tree, or a fix for a stable-only
- regression, can be nominated by mail to this list, if it
- meets the criteria
+ Any commit in Linus's tree can be nominated by mail to this list, if it
+ meets the criteria - see
+ <a href="http://www.kernel.org/doc/Documentation/stable_kernel_rules.txt"><tt>Documentation/stable_kernel_rules.txt</tt></a>
+ </li>
+ <li>
+ Occasional regression fix in stable without a corresponding
+ commit in Linus's tree, because the change that regressed was
+ OK in mainline
+ </li>
+ </ul>
+</div>
+
+<div class="slide">
+ <h1>The stable update process (2)</h1>
+ <ul>
+ <li>
+ Maintainer sends pending changes to mailing list, original
+ authors, etc., for time-limited review and testing
+ <ul>
+ <li>
+ Fix might not be needed in a stable branch, or might not
+ work due to missing dependencies
+ </li>
+ <li>
+ Fix might have been found to introduce a regression in
+ mainline, so must wait until second fix available
+ </li>
+ </ul>
+ </li>
+ <li>
+ Maintainer drops/adds changes based on review, then applies
+ to stable branch and tags release
+ </li>
+ <li>
+ Greg pushes tag to kernel.org (possibly after pulling from
+ another maintainer) which generates tarballs, updates the front
+ page, etc.
+ </li>
+ <li>
+ Maintainer announces the release, shortly followed by LWN
+ and other news media
+ </li>
+ </ul>
+</div>
+
+<div class="slide">
+ <h1>Distributions and stable updates</h1>
+ <ul>
+ <li>
+ Package maintainers want to get bug fixes without waiting for
+ the next release - particularly for security
+ </li>
+ <li>
+ Maintainers report bugs upstream, find and sometimes write
+ fixes, and stable updates are a way to share these fixes across
+ distributions
<ul>
<li>
- See
- <a href="http://www.kernel.org/doc/Documentation/stable_kernel_rules.txt"><tt>Documentation/stable_kernel_rules.txt</tt></a>
+ Are you doing this? If not, what's stopping you?
</li>
</ul>
</li>
+ <li>
+ Distribution stable releases have a much longer support period
+ than kernel release cycle, but updating to every new Linus
+ release is too risky
+ </li>
+ <li>
+ So the longterm stable branches are very important for most
+ distributions with stable releases
+ </li>
</ul>
</div>
+<div class="slide">
+ <h1>Coordinating longterm branches</h1>
+ <ul>
+ <li>
+ Linux 2.6.32 was chosen as the basis for RHEL 6, and other
+ distributions preparing a stable release in 2010 opted to do
+ the same
+ </li>
+ <li>
+ The 2.6.32.<var>y</var> longterm branch is the basis for kernel
+ packages in Debian 6.0, SLE11 SP1 and Ubuntu 10.04 LTS and
+ has over 3,500 changes
+ </li>
+ <li>
+ Other longterm branches have not been quite as widely used or as
+ active - maybe because release schedules didn't align as well
+ </li>
+ <li>
+ Do package maintainers expect/want there to be a longterm stable
+ branch for the kernel version in a stable release?
+ </li>
+ <li>
+ Should we be coordinating more explicitly to ensure that this
+ happens?
+ </li>
+ <li>
+ Would you be prepared to maintain such a branch at kernel.org?
+ </li>
+ </ul>
+</div>
+
+<div class="slide">
+ <h1>Questions?</h1>
+ <p>
+ The FAQ is
+ <a href="http://www.kernel.org/doc/Documentation/stable_kernel_rules.txt"><tt>Documentation/stable_kernel_rules.txt</tt></a>
+ </p>
+</div>
+
<div class="slide">
<h1>Credits</h1>
<ul>