Jesper Juhl | 8b7ecb1 | 2007-07-15 23:40:51 -0700 | [diff] [blame] | 1 | Linux Kernel patch submission checklist |
Mauro Carvalho Chehab | 7a71a80 | 2016-09-19 08:08:01 -0300 | [diff] [blame] | 2 | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ |
Randy Dunlap | 915a56d | 2006-06-23 02:05:49 -0700 | [diff] [blame] | 3 | |
Andrew Morton | e54695a | 2006-07-10 04:45:42 -0700 | [diff] [blame] | 4 | Here are some basic things that developers should do if they want to see their |
| 5 | kernel patch submissions accepted more quickly. |
Randy Dunlap | 915a56d | 2006-06-23 02:05:49 -0700 | [diff] [blame] | 6 | |
Andrew Morton | e54695a | 2006-07-10 04:45:42 -0700 | [diff] [blame] | 7 | These are all above and beyond the documentation that is provided in |
Mauro Carvalho Chehab | 7a71a80 | 2016-09-19 08:08:01 -0300 | [diff] [blame] | 8 | Documentation/SubmittingPatches |
| 9 | and elsewhere regarding submitting Linux kernel patches. |
Randy Dunlap | 915a56d | 2006-06-23 02:05:49 -0700 | [diff] [blame] | 10 | |
| 11 | |
Mauro Carvalho Chehab | 7a71a80 | 2016-09-19 08:08:01 -0300 | [diff] [blame] | 12 | 1) If you use a facility then #include the file that defines/declares |
Randy Dunlap | b5247c8 | 2010-03-10 15:21:59 -0800 | [diff] [blame] | 13 | that facility. Don't depend on other header files pulling in ones |
| 14 | that you use. |
| 15 | |
Mauro Carvalho Chehab | 7a71a80 | 2016-09-19 08:08:01 -0300 | [diff] [blame] | 16 | 2) Builds cleanly: |
Randy Dunlap | 915a56d | 2006-06-23 02:05:49 -0700 | [diff] [blame] | 17 | |
Mauro Carvalho Chehab | 7a71a80 | 2016-09-19 08:08:01 -0300 | [diff] [blame] | 18 | a) with applicable or modified ``CONFIG`` options ``=y``, ``=m``, and |
| 19 | ``=n``. No ``gcc`` warnings/errors, no linker warnings/errors. |
Randy Dunlap | 915a56d | 2006-06-23 02:05:49 -0700 | [diff] [blame] | 20 | |
Mauro Carvalho Chehab | 7a71a80 | 2016-09-19 08:08:01 -0300 | [diff] [blame] | 21 | b) Passes ``allnoconfig``, ``allmodconfig`` |
Randy Dunlap | 85f9642 | 2010-05-23 17:02:30 -0700 | [diff] [blame] | 22 | |
Mauro Carvalho Chehab | 7a71a80 | 2016-09-19 08:08:01 -0300 | [diff] [blame] | 23 | c) Builds successfully when using ``O=builddir`` |
| 24 | |
| 25 | 3) Builds on multiple CPU architectures by using local cross-compile tools |
Randy Dunlap | 7de3369 | 2009-12-15 16:46:59 -0800 | [diff] [blame] | 26 | or some other build farm. |
Randy Dunlap | 915a56d | 2006-06-23 02:05:49 -0700 | [diff] [blame] | 27 | |
Mauro Carvalho Chehab | 7a71a80 | 2016-09-19 08:08:01 -0300 | [diff] [blame] | 28 | 4) ppc64 is a good architecture for cross-compilation checking because it |
| 29 | tends to use ``unsigned long`` for 64-bit quantities. |
Randy Dunlap | 915a56d | 2006-06-23 02:05:49 -0700 | [diff] [blame] | 30 | |
J. Bruce Fields | df24d9a | 2008-02-13 15:03:20 -0800 | [diff] [blame] | 31 | 5: Check your patch for general style as detailed in |
Mauro Carvalho Chehab | 7a71a80 | 2016-09-19 08:08:01 -0300 | [diff] [blame] | 32 | Documentation/CodingStyle. |
| 33 | Check for trivial violations with the patch style checker prior to |
| 34 | submission (``scripts/checkpatch.pl``). |
J. Bruce Fields | df24d9a | 2008-02-13 15:03:20 -0800 | [diff] [blame] | 35 | You should be able to justify all violations that remain in |
| 36 | your patch. |
Randy Dunlap | 915a56d | 2006-06-23 02:05:49 -0700 | [diff] [blame] | 37 | |
Mauro Carvalho Chehab | 7a71a80 | 2016-09-19 08:08:01 -0300 | [diff] [blame] | 38 | 6) Any new or modified ``CONFIG`` options don't muck up the config menu. |
Randy Dunlap | 915a56d | 2006-06-23 02:05:49 -0700 | [diff] [blame] | 39 | |
Mauro Carvalho Chehab | 7a71a80 | 2016-09-19 08:08:01 -0300 | [diff] [blame] | 40 | 7) All new ``Kconfig`` options have help text. |
Randy Dunlap | 915a56d | 2006-06-23 02:05:49 -0700 | [diff] [blame] | 41 | |
Mauro Carvalho Chehab | 7a71a80 | 2016-09-19 08:08:01 -0300 | [diff] [blame] | 42 | 8) Has been carefully reviewed with respect to relevant ``Kconfig`` |
Andrew Morton | e54695a | 2006-07-10 04:45:42 -0700 | [diff] [blame] | 43 | combinations. This is very hard to get right with testing -- brainpower |
| 44 | pays off here. |
Randy Dunlap | 915a56d | 2006-06-23 02:05:49 -0700 | [diff] [blame] | 45 | |
Mauro Carvalho Chehab | 7a71a80 | 2016-09-19 08:08:01 -0300 | [diff] [blame] | 46 | 9) Check cleanly with sparse. |
Randy Dunlap | 915a56d | 2006-06-23 02:05:49 -0700 | [diff] [blame] | 47 | |
Mauro Carvalho Chehab | 7a71a80 | 2016-09-19 08:08:01 -0300 | [diff] [blame] | 48 | 10) Use ``make checkstack`` and ``make namespacecheck`` and fix any problems |
| 49 | that they find. |
| 50 | |
| 51 | .. note:: |
| 52 | |
| 53 | ``checkstack`` does not point out problems explicitly, |
| 54 | but any one function that uses more than 512 bytes on the stack is a |
| 55 | candidate for change. |
Randy Dunlap | 915a56d | 2006-06-23 02:05:49 -0700 | [diff] [blame] | 56 | |
Mauro Carvalho Chehab | 0cef67a | 2016-09-19 08:08:00 -0300 | [diff] [blame] | 57 | 11: Include :ref:`kernel-doc <kernel_doc>` to document global kernel APIs. |
| 58 | (Not required for static functions, but OK there also.) Use |
| 59 | ``make htmldocs`` or ``make pdfdocs`` to check the |
| 60 | :ref:`kernel-doc <kernel_doc>` and fix any issues. |
Randy Dunlap | 915a56d | 2006-06-23 02:05:49 -0700 | [diff] [blame] | 61 | |
Mauro Carvalho Chehab | 7a71a80 | 2016-09-19 08:08:01 -0300 | [diff] [blame] | 62 | 12) Has been tested with ``CONFIG_PREEMPT``, ``CONFIG_DEBUG_PREEMPT``, |
| 63 | ``CONFIG_DEBUG_SLAB``, ``CONFIG_DEBUG_PAGEALLOC``, ``CONFIG_DEBUG_MUTEXES``, |
| 64 | ``CONFIG_DEBUG_SPINLOCK``, ``CONFIG_DEBUG_ATOMIC_SLEEP``, |
| 65 | ``CONFIG_PROVE_RCU`` and ``CONFIG_DEBUG_OBJECTS_RCU_HEAD`` all |
| 66 | simultaneously enabled. |
Randy Dunlap | 915a56d | 2006-06-23 02:05:49 -0700 | [diff] [blame] | 67 | |
Mauro Carvalho Chehab | 7a71a80 | 2016-09-19 08:08:01 -0300 | [diff] [blame] | 68 | 13) Has been build- and runtime tested with and without ``CONFIG_SMP`` and |
| 69 | ``CONFIG_PREEMPT.`` |
Randy Dunlap | 915a56d | 2006-06-23 02:05:49 -0700 | [diff] [blame] | 70 | |
Mauro Carvalho Chehab | 7a71a80 | 2016-09-19 08:08:01 -0300 | [diff] [blame] | 71 | 14) If the patch affects IO/Disk, etc: has been tested with and without |
| 72 | ``CONFIG_LBDAF.`` |
Randy Dunlap | 915a56d | 2006-06-23 02:05:49 -0700 | [diff] [blame] | 73 | |
Mauro Carvalho Chehab | 7a71a80 | 2016-09-19 08:08:01 -0300 | [diff] [blame] | 74 | 15) All codepaths have been exercised with all lockdep features enabled. |
Randy Dunlap | 915a56d | 2006-06-23 02:05:49 -0700 | [diff] [blame] | 75 | |
Mauro Carvalho Chehab | 7a71a80 | 2016-09-19 08:08:01 -0300 | [diff] [blame] | 76 | 16) All new ``/proc`` entries are documented under ``Documentation/`` |
Andrew Morton | e54695a | 2006-07-10 04:45:42 -0700 | [diff] [blame] | 77 | |
Mauro Carvalho Chehab | 7a71a80 | 2016-09-19 08:08:01 -0300 | [diff] [blame] | 78 | 17) All new kernel boot parameters are documented in |
| 79 | ``Documentation/kernel-parameters.txt``. |
Andrew Morton | e54695a | 2006-07-10 04:45:42 -0700 | [diff] [blame] | 80 | |
Mauro Carvalho Chehab | 7a71a80 | 2016-09-19 08:08:01 -0300 | [diff] [blame] | 81 | 18) All new module parameters are documented with ``MODULE_PARM_DESC()`` |
Randy Dunlap | 1a036cdd | 2006-09-29 02:01:26 -0700 | [diff] [blame] | 82 | |
Mauro Carvalho Chehab | 7a71a80 | 2016-09-19 08:08:01 -0300 | [diff] [blame] | 83 | 19) All new userspace interfaces are documented in ``Documentation/ABI/``. |
| 84 | See ``Documentation/ABI/README`` for more information. |
Michael Kerrisk | 1d992ce | 2008-10-03 15:23:44 -0700 | [diff] [blame] | 85 | Patches that change userspace interfaces should be CCed to |
| 86 | linux-api@vger.kernel.org. |
Andrew Morton | ce584f9 | 2006-09-30 23:27:39 -0700 | [diff] [blame] | 87 | |
Mauro Carvalho Chehab | 7a71a80 | 2016-09-19 08:08:01 -0300 | [diff] [blame] | 88 | 20) Check that it all passes ``make headers_check``. |
Andrew Morton | 0dd4e5b | 2006-12-10 02:18:56 -0800 | [diff] [blame] | 89 | |
Mauro Carvalho Chehab | 7a71a80 | 2016-09-19 08:08:01 -0300 | [diff] [blame] | 90 | 21) Has been checked with injection of at least slab and page-allocation |
| 91 | failures. See ``Documentation/fault-injection/``. |
Andrew Morton | 0dd4e5b | 2006-12-10 02:18:56 -0800 | [diff] [blame] | 92 | |
| 93 | If the new code is substantial, addition of subsystem-specific fault |
| 94 | injection might be appropriate. |
Andrew Morton | a517b9f | 2007-01-22 20:40:36 -0800 | [diff] [blame] | 95 | |
Mauro Carvalho Chehab | 7a71a80 | 2016-09-19 08:08:01 -0300 | [diff] [blame] | 96 | 22) Newly-added code has been compiled with ``gcc -W`` (use |
| 97 | ``make EXTRA_CFLAGS=-W``). This will generate lots of noise, but is good |
| 98 | for finding bugs like "warning: comparison between signed and unsigned". |
Randy Dunlap | 244474b | 2007-02-28 20:12:35 -0800 | [diff] [blame] | 99 | |
Mauro Carvalho Chehab | 7a71a80 | 2016-09-19 08:08:01 -0300 | [diff] [blame] | 100 | 23) Tested after it has been merged into the -mm patchset to make sure |
Randy Dunlap | 244474b | 2007-02-28 20:12:35 -0800 | [diff] [blame] | 101 | that it still works with all of the other queued patches and various |
| 102 | changes in the VM, VFS, and other subsystems. |
Randy Dunlap | 8033fe6 | 2008-10-15 22:01:59 -0700 | [diff] [blame] | 103 | |
Mauro Carvalho Chehab | 7a71a80 | 2016-09-19 08:08:01 -0300 | [diff] [blame] | 104 | 24) All memory barriers {e.g., ``barrier()``, ``rmb()``, ``wmb()``} need a |
| 105 | comment in the source code that explains the logic of what they are doing |
| 106 | and why. |
Randy Dunlap | 7de3369 | 2009-12-15 16:46:59 -0800 | [diff] [blame] | 107 | |
Mauro Carvalho Chehab | 7a71a80 | 2016-09-19 08:08:01 -0300 | [diff] [blame] | 108 | 25) If any ioctl's are added by the patch, then also update |
| 109 | ``Documentation/ioctl/ioctl-number.txt``. |
Randy Dunlap | 85f9642 | 2010-05-23 17:02:30 -0700 | [diff] [blame] | 110 | |
Mauro Carvalho Chehab | 7a71a80 | 2016-09-19 08:08:01 -0300 | [diff] [blame] | 111 | 26) If your modified source code depends on or uses any of the kernel |
| 112 | APIs or features that are related to the following ``Kconfig`` symbols, |
| 113 | then test multiple builds with the related ``Kconfig`` symbols disabled |
| 114 | and/or ``=m`` (if that option is available) [not all of these at the |
Randy Dunlap | 85f9642 | 2010-05-23 17:02:30 -0700 | [diff] [blame] | 115 | same time, just various/random combinations of them]: |
| 116 | |
Mauro Carvalho Chehab | 7a71a80 | 2016-09-19 08:08:01 -0300 | [diff] [blame] | 117 | ``CONFIG_SMP``, ``CONFIG_SYSFS``, ``CONFIG_PROC_FS``, ``CONFIG_INPUT``, ``CONFIG_PCI``, ``CONFIG_BLOCK``, ``CONFIG_PM``, ``CONFIG_MAGIC_SYSRQ``, |
| 118 | ``CONFIG_NET``, ``CONFIG_INET=n`` (but latter with ``CONFIG_NET=y``). |