[mirror_ubuntu-artful-kernel.git] / Documentation / hwmon / submitting-patches

	How to Get Your Patch Accepted Into the Hwmon Subsystem
	-------------------------------------------------------

This text is a collection of suggestions for people writing patches or
drivers for the hwmon subsystem. Following these suggestions will greatly
increase the chances of your change being accepted.


1. General
----------

* It should be unnecessary to mention, but please read and follow
    Documentation/process/submit-checklist.rst
    Documentation/process/submitting-drivers.rst
    Documentation/process/submitting-patches.rst
    Documentation/process/coding-style.rst

* Please run your patch through 'checkpatch --strict'. There should be no
  errors, no warnings, and few if any check messages. If there are any
  messages, please be prepared to explain.

* If your patch generates checkpatch errors, warnings, or check messages,
  please refrain from explanations such as "I prefer that coding style".
  Keep in mind that each unnecessary message helps hiding a real problem,
  and a consistent coding style makes it easier for others to understand
  and review the code.

* Please test your patch thoroughly. We are not your test group.
  Sometimes a patch can not or not completely be tested because of missing
  hardware. In such cases, you should test-build the code on at least one
  architecture. If run-time testing was not achieved, it should be written
  explicitly below the patch header.

* If your patch (or the driver) is affected by configuration options such as
  CONFIG_SMP, make sure it compiles for all configuration variants.


2. Adding functionality to existing drivers
-------------------------------------------

* Make sure the documentation in Documentation/hwmon/<driver_name> is up to
  date.

* Make sure the information in Kconfig is up to date.

* If the added functionality requires some cleanup or structural changes, split
  your patch into a cleanup part and the actual addition. This makes it easier
  to review your changes, and to bisect any resulting problems.

* Never mix bug fixes, cleanup, and functional enhancements in a single patch.


3. New drivers
--------------

* Running your patch or driver file(s) through checkpatch does not mean its
  formatting is clean. If unsure about formatting in your new driver, run it
  through Lindent. Lindent is not perfect, and you may have to do some minor
  cleanup, but it is a good start.

* Consider adding yourself to MAINTAINERS.

* Document the driver in Documentation/hwmon/<driver_name>.

* Add the driver to Kconfig and Makefile in alphabetical order.

* Make sure that all dependencies are listed in Kconfig.

* Please list include files in alphabetic order.

* Please align continuation lines with '(' on the previous line.

* Avoid forward declarations if you can. Rearrange the code if necessary.

* Avoid macros to generate groups of sensor attributes. It not only confuses
  checkpatch, but also makes it more difficult to review the code.

* Avoid calculations in macros and macro-generated functions. While such macros
  may save a line or so in the source, it obfuscates the code and makes code
  review more difficult. It may also result in code which is more complicated
  than necessary. Use inline functions or just regular functions instead.

* Limit the number of kernel log messages. In general, your driver should not
  generate an error message just because a runtime operation failed. Report
  errors to user space instead, using an appropriate error code. Keep in mind
  that kernel error log messages not only fill up the kernel log, but also are
  printed synchronously, most likely with interrupt disabled, often to a serial
  console. Excessive logging can seriously affect system performance.

* Use devres functions whenever possible to allocate resources. For rationale
  and supported functions, please see Documentation/driver-model/devres.txt.
  If a function is not supported by devres, consider using devm_add_action().

* If the driver has a detect function, make sure it is silent. Debug messages
  and messages printed after a successful detection are acceptable, but it
  must not print messages such as "Chip XXX not found/supported".

  Keep in mind that the detect function will run for all drivers supporting an
  address if a chip is detected on that address. Unnecessary messages will just
  pollute the kernel log and not provide any value.

* Provide a detect function if and only if a chip can be detected reliably.

* Only the following I2C addresses shall be probed: 0x18-0x1f, 0x28-0x2f,
  0x48-0x4f, 0x58, 0x5c, 0x73 and 0x77. Probing other addresses is strongly
  discouraged as it is known to cause trouble with other (non-hwmon) I2C
  chips. If your chip lives at an address which can't be probed then the
  device will have to be instantiated explicitly (which is always better
  anyway.)

* Avoid writing to chip registers in the detect function. If you have to write,
  only do it after you have already gathered enough data to be certain that the
  detection is going to be successful.

  Keep in mind that the chip might not be what your driver believes it is, and
  writing to it might cause a bad misconfiguration.

* Make sure there are no race conditions in the probe function. Specifically,
  completely initialize your chip and your driver first, then register with
  the hwmon subsystem.

* Use devm_hwmon_device_register_with_groups() or, if your driver needs a remove
  function, hwmon_device_register_with_groups() to register your driver with the
  hwmon subsystem. Try using devm_add_action() instead of a remove function if
  possible. Do not use hwmon_device_register().

* Your driver should be buildable as module. If not, please be prepared to
  explain why it has to be built into the kernel.

* Do not provide support for deprecated sysfs attributes.

* Do not create non-standard attributes unless really needed. If you have to use
  non-standard attributes, or you believe you do, discuss it on the mailing list
  first. Either case, provide a detailed explanation why you need the
  non-standard attribute(s).
  Standard attributes are specified in Documentation/hwmon/sysfs-interface.

* When deciding which sysfs attributes to support, look at the chip's
  capabilities. While we do not expect your driver to support everything the
  chip may offer, it should at least support all limits and alarms.

* Last but not least, please check if a driver for your chip already exists
  before starting to write a new driver. Especially for temperature sensors,
  new chips are often variants of previously released chips. In some cases,
  a presumably new chip may simply have been relabeled.
Commit	Line	Data
c3a2f0ad GR	1	How to Get Your Patch Accepted Into the Hwmon Subsystem
	2	-------------------------------------------------------
	3
ac20ad14	4	This text is a collection of suggestions for people writing patches or
c3a2f0ad GR	5	drivers for the hwmon subsystem. Following these suggestions will greatly
	6	increase the chances of your change being accepted.
	7
	8
	9	1. General
	10	----------
	11
	12	* It should be unnecessary to mention, but please read and follow
8c27ceff MCC	13	Documentation/process/submit-checklist.rst
	14	Documentation/process/submitting-drivers.rst
	15	Documentation/process/submitting-patches.rst
	16	Documentation/process/coding-style.rst
c3a2f0ad	17
165720d9 GR	18	* Please run your patch through 'checkpatch --strict'. There should be no
	19	errors, no warnings, and few if any check messages. If there are any
	20	messages, please be prepared to explain.
	21
	22	* If your patch generates checkpatch errors, warnings, or check messages,
	23	please refrain from explanations such as "I prefer that coding style".
	24	Keep in mind that each unnecessary message helps hiding a real problem,
	25	and a consistent coding style makes it easier for others to understand
	26	and review the code.
c3a2f0ad GR	27
	28	* Please test your patch thoroughly. We are not your test group.
	29	Sometimes a patch can not or not completely be tested because of missing
	30	hardware. In such cases, you should test-build the code on at least one
	31	architecture. If run-time testing was not achieved, it should be written
	32	explicitly below the patch header.
	33
	34	* If your patch (or the driver) is affected by configuration options such as
40b31360	35	CONFIG_SMP, make sure it compiles for all configuration variants.
c3a2f0ad GR	36
	37
	38	2. Adding functionality to existing drivers
	39	-------------------------------------------
	40
	41	* Make sure the documentation in Documentation/hwmon/<driver_name> is up to
	42	date.
	43
	44	* Make sure the information in Kconfig is up to date.
	45
	46	* If the added functionality requires some cleanup or structural changes, split
	47	your patch into a cleanup part and the actual addition. This makes it easier
	48	to review your changes, and to bisect any resulting problems.
	49
	50	* Never mix bug fixes, cleanup, and functional enhancements in a single patch.
	51
	52
	53	3. New drivers
	54	--------------
	55
	56	* Running your patch or driver file(s) through checkpatch does not mean its
	57	formatting is clean. If unsure about formatting in your new driver, run it
	58	through Lindent. Lindent is not perfect, and you may have to do some minor
	59	cleanup, but it is a good start.
	60
	61	* Consider adding yourself to MAINTAINERS.
	62
	63	* Document the driver in Documentation/hwmon/<driver_name>.
	64
	65	* Add the driver to Kconfig and Makefile in alphabetical order.
	66
07d33600	67	* Make sure that all dependencies are listed in Kconfig.
c3a2f0ad	68
165720d9 GR	69	* Please list include files in alphabetic order.
	70
	71	* Please align continuation lines with '(' on the previous line.
	72
c3a2f0ad GR	73	* Avoid forward declarations if you can. Rearrange the code if necessary.
c3a2f0ad GR	74
165720d9 GR	75	* Avoid macros to generate groups of sensor attributes. It not only confuses
	76	checkpatch, but also makes it more difficult to review the code.
	77
c3a2f0ad GR	78	* Avoid calculations in macros and macro-generated functions. While such macros
	79	may save a line or so in the source, it obfuscates the code and makes code
	80	review more difficult. It may also result in code which is more complicated
	81	than necessary. Use inline functions or just regular functions instead.
	82
165720d9 GR	83	* Limit the number of kernel log messages. In general, your driver should not
	84	generate an error message just because a runtime operation failed. Report
	85	errors to user space instead, using an appropriate error code. Keep in mind
	86	that kernel error log messages not only fill up the kernel log, but also are
	87	printed synchronously, most likely with interrupt disabled, often to a serial
	88	console. Excessive logging can seriously affect system performance.
	89
3fafb0ce GR	90	* Use devres functions whenever possible to allocate resources. For rationale
3fafb0ce GR	91	and supported functions, please see Documentation/driver-model/devres.txt.
165720d9	92	If a function is not supported by devres, consider using devm_add_action().
3fafb0ce	93
c3a2f0ad GR	94	* If the driver has a detect function, make sure it is silent. Debug messages
	95	and messages printed after a successful detection are acceptable, but it
	96	must not print messages such as "Chip XXX not found/supported".
	97
	98	Keep in mind that the detect function will run for all drivers supporting an
	99	address if a chip is detected on that address. Unnecessary messages will just
	100	pollute the kernel log and not provide any value.
	101
	102	* Provide a detect function if and only if a chip can be detected reliably.
	103
97607f98 JD	104	* Only the following I2C addresses shall be probed: 0x18-0x1f, 0x28-0x2f,
	105	0x48-0x4f, 0x58, 0x5c, 0x73 and 0x77. Probing other addresses is strongly
	106	discouraged as it is known to cause trouble with other (non-hwmon) I2C
	107	chips. If your chip lives at an address which can't be probed then the
	108	device will have to be instantiated explicitly (which is always better
	109	anyway.)
	110
c3a2f0ad GR	111	* Avoid writing to chip registers in the detect function. If you have to write,
	112	only do it after you have already gathered enough data to be certain that the
	113	detection is going to be successful.
	114
	115	Keep in mind that the chip might not be what your driver believes it is, and
	116	writing to it might cause a bad misconfiguration.
	117
	118	* Make sure there are no race conditions in the probe function. Specifically,
165720d9 GR	119	completely initialize your chip and your driver first, then register with
	120	the hwmon subsystem.
	121
	122	* Use devm_hwmon_device_register_with_groups() or, if your driver needs a remove
	123	function, hwmon_device_register_with_groups() to register your driver with the
	124	hwmon subsystem. Try using devm_add_action() instead of a remove function if
	125	possible. Do not use hwmon_device_register().
	126
	127	* Your driver should be buildable as module. If not, please be prepared to
	128	explain why it has to be built into the kernel.
c3a2f0ad GR	129
	130	* Do not provide support for deprecated sysfs attributes.
	131
	132	* Do not create non-standard attributes unless really needed. If you have to use
	133	non-standard attributes, or you believe you do, discuss it on the mailing list
	134	first. Either case, provide a detailed explanation why you need the
	135	non-standard attribute(s).
	136	Standard attributes are specified in Documentation/hwmon/sysfs-interface.
	137
	138	* When deciding which sysfs attributes to support, look at the chip's
	139	capabilities. While we do not expect your driver to support everything the
	140	chip may offer, it should at least support all limits and alarms.
	141
	142	* Last but not least, please check if a driver for your chip already exists
	143	before starting to write a new driver. Especially for temperature sensors,
	144	new chips are often variants of previously released chips. In some cases,
	145	a presumably new chip may simply have been relabeled.