How to Get Your Patch Accepted Into the Hwmon Subsystem
This text is 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.
* It should be unnecessary to mention, but please read and follow
* If your patch generates checkpatch warnings, please refrain from explanations
such as "I don't like that coding style". Keep in mind that each unnecessary
warning helps hiding a real problem. If you don't like the kernel coding
style, don't write kernel drivers.
* 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 or CONFIG_HOTPLUG, make sure it compiles for all configuration
2. Adding functionality to existing drivers
* Make sure the documentation in Documentation/hwmon/<driver_name> is up to
* 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. For new drivers, it
is most likely prudent to add a dependency on EXPERIMENTAL.
* Avoid forward declarations if you can. Rearrange the code if necessary.
* 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.
* 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.
* 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 first, then create sysfs entries and register
with the hwmon subsystem.
* 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
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.