blob: 3049a612291b1ad8651da72c6081539bb4e83a74 [file] [log] [blame]
Paul Gortmaker2584f522012-06-05 11:15:50 -04001Everything you ever wanted to know about Linux -stable releases.
Greg KHfc185d92005-07-29 12:14:34 -07002
Jesper Juhle48e9902006-01-09 20:53:59 -08003Rules on what kind of patches are accepted, and which ones are not, into the
4"-stable" tree:
Greg KHfc185d92005-07-29 12:14:34 -07005
6 - It must be obviously correct and tested.
Matt LaPlante84eb8d02006-10-03 22:53:09 +02007 - It cannot be bigger than 100 lines, with context.
Greg KHfc185d92005-07-29 12:14:34 -07008 - It must fix only one thing.
9 - It must fix a real bug that bothers people (not a, "This could be a
Jesper Juhle48e9902006-01-09 20:53:59 -080010 problem..." type thing).
Greg KHfc185d92005-07-29 12:14:34 -070011 - It must fix a problem that causes a build error (but not for things
12 marked CONFIG_BROKEN), an oops, a hang, data corruption, a real
Jesper Juhle48e9902006-01-09 20:53:59 -080013 security issue, or some "oh, that's not good" issue. In short, something
14 critical.
Mel Gorman8191e0d2012-06-21 11:36:50 +010015 - Serious issues as reported by a user of a distribution kernel may also
16 be considered if they fix a notable performance or interactivity issue.
17 As these fixes are not as obvious and have a higher risk of a subtle
18 regression they should only be submitted by a distribution kernel
19 maintainer and include an addendum linking to a bugzilla entry if it
20 exists and additional information on the user-visible impact.
Josh Boyer79cd80a2008-10-24 10:10:42 -040021 - New device IDs and quirks are also accepted.
Jesper Juhle48e9902006-01-09 20:53:59 -080022 - No "theoretical race condition" issues, unless an explanation of how the
23 race can be exploited is also provided.
Matt LaPlante84eb8d02006-10-03 22:53:09 +020024 - It cannot contain any "trivial" fixes in it (spelling changes,
Jesper Juhle48e9902006-01-09 20:53:59 -080025 whitespace cleanups, etc).
Jesper Juhle48e9902006-01-09 20:53:59 -080026 - It must follow the Documentation/SubmittingPatches rules.
Stefan Richter959fb1c2010-04-01 22:01:52 +020027 - It or an equivalent fix must already exist in Linus' tree (upstream).
Greg KHfc185d92005-07-29 12:14:34 -070028
29
30Procedure for submitting patches to the -stable tree:
31
Dave Chilukb76fc282014-06-24 10:11:26 -050032 - If the patch covers files in net/ or drivers/net please follow netdev stable
33 submission guidelines as described in
34 Documentation/networking/netdev-FAQ.txt
Brian Norris5de61e72014-12-18 14:55:53 -080035 - Security patches should not be handled (solely) by the -stable review
36 process but should follow the procedures in Documentation/SecurityBugs.
37
38For all other submissions, choose one of the following procedures:
39
40 --- Option 1 ---
41
42 To have the patch automatically included in the stable tree, add the tag
Josh Boyer5fa22422011-10-17 21:16:39 -040043 Cc: stable@vger.kernel.org
Sebastian Andrzej Siewior8e9b9362009-12-06 12:24:31 +010044 in the sign-off area. Once the patch is merged it will be applied to
45 the stable tree without anything else needing to be done by the author
46 or subsystem maintainer.
Brian Norris5de61e72014-12-18 14:55:53 -080047
48 --- Option 2 ---
49
50 After the patch has been merged to Linus' tree, send an email to
51 stable@vger.kernel.org containing the subject of the patch, the commit ID,
52 why you think it should be applied, and what kernel version you wish it to
53 be applied to.
54
55 --- Option 3 ---
56
57 Send the patch, after verifying that it follows the above rules, to
58 stable@vger.kernel.org. You must note the upstream commit ID in the
59 changelog of your submission, as well as the kernel version you wish
60 it to be applied to.
61
Christoffer Dallbde1b292015-05-13 22:01:07 +020062Option 1 is *strongly* preferred, is the easiest and most common. Options 2 and
633 are more useful if the patch isn't deemed worthy at the time it is applied to
64a public git tree (for instance, because it deserves more regression testing
65first). Option 3 is especially useful if the patch needs some special handling
66to apply to an older kernel (e.g., if API's have changed in the meantime).
67
68Note that for Option 3, if the patch deviates from the original upstream patch
69(for example because it had to be backported) this must be very clearly
70documented and justified in the patch description.
71
72The upstream commit ID must be specified with a separate line above the commit
73text, like this:
74
75 commit <sha1> upstream.
Brian Norris5de61e72014-12-18 14:55:53 -080076
77Additionally, some patches submitted via Option 1 may have additional patch
78prerequisites which can be cherry-picked. This can be specified in the following
79format in the sign-off area:
Sebastian Andrzej Siewior8e9b9362009-12-06 12:24:31 +010080
Paul Gortmaker2584f522012-06-05 11:15:50 -040081 Cc: <stable@vger.kernel.org> # 3.3.x: a1f84a3: sched: Check for idle
82 Cc: <stable@vger.kernel.org> # 3.3.x: 1b9508f: sched: Rate-limit newidle
83 Cc: <stable@vger.kernel.org> # 3.3.x: fd21073: sched: Fix affinity logic
84 Cc: <stable@vger.kernel.org> # 3.3.x
Sebastian Andrzej Siewior8e9b9362009-12-06 12:24:31 +010085 Signed-off-by: Ingo Molnar <mingo@elte.hu>
86
87 The tag sequence has the meaning of:
88 git cherry-pick a1f84a3
89 git cherry-pick 1b9508f
90 git cherry-pick fd21073
91 git cherry-pick <this commit>
92
David Daneyfdc81b72015-03-26 10:09:08 -070093Also, some patches may have kernel version prerequisites. This can be
94specified in the following format in the sign-off area:
95
96 Cc: <stable@vger.kernel.org> # 3.3.x-
97
98 The tag has the meaning of:
99 git cherry-pick <this commit>
100
101 For each "-stable" tree starting with the specified version.
102
Brian Norris5de61e72014-12-18 14:55:53 -0800103Following the submission:
104
Jesper Juhle48e9902006-01-09 20:53:59 -0800105 - The sender will receive an ACK when the patch has been accepted into the
106 queue, or a NAK if the patch is rejected. This response might take a few
107 days, according to the developer's schedules.
108 - If accepted, the patch will be added to the -stable queue, for review by
Stefan Richter46cdf8712008-02-08 23:26:02 +0100109 other developers and by the relevant subsystem maintainer.
Greg KHfc185d92005-07-29 12:14:34 -0700110
111
112Review cycle:
113
Jesper Juhle48e9902006-01-09 20:53:59 -0800114 - When the -stable maintainers decide for a review cycle, the patches will be
115 sent to the review committee, and the maintainer of the affected area of
116 the patch (unless the submitter is the maintainer of the area) and CC: to
117 the linux-kernel mailing list.
118 - The review committee has 48 hours in which to ACK or NAK the patch.
Greg KHfc185d92005-07-29 12:14:34 -0700119 - If the patch is rejected by a member of the committee, or linux-kernel
Jesper Juhle48e9902006-01-09 20:53:59 -0800120 members object to the patch, bringing up issues that the maintainers and
121 members did not realize, the patch will be dropped from the queue.
122 - At the end of the review cycle, the ACKed patches will be added to the
123 latest -stable release, and a new -stable release will happen.
124 - Security patches will be accepted into the -stable tree directly from the
125 security kernel team, and not go through the normal review cycle.
Greg KHfc185d92005-07-29 12:14:34 -0700126 Contact the kernel security team for more details on this procedure.
127
Paul Gortmaker2584f522012-06-05 11:15:50 -0400128Trees:
129
130 - The queues of patches, for both completed versions and in progress
131 versions can be found at:
132 http://git.kernel.org/?p=linux/kernel/git/stable/stable-queue.git
133 - The finalized and tagged releases of all stable kernels can be found
134 in separate branches per version at:
135 http://git.kernel.org/?p=linux/kernel/git/stable/linux-stable.git
136
Greg KHfc185d92005-07-29 12:14:34 -0700137
Matt LaPlante5d3f0832006-11-30 05:21:10 +0100138Review committee:
Greg KHfc185d92005-07-29 12:14:34 -0700139
Jesper Juhle48e9902006-01-09 20:53:59 -0800140 - This is made up of a number of kernel developers who have volunteered for
141 this task, and a few that haven't.