On 28/03/2021 21:31, Srinivas Pandruvada wrote: > Hi Daniel, [ ... ] >> I don't see where is the problem. The protocol is still compatible >> with >> the previous version, so that does not break the existing AFAICT. >> That >> is done on purpose. > > The size of netlink message is changed. This is not a good argument to > just adding members at the end. The point I am trying that netlink now > is an ABI, which should go through same process as we are > adding/changing a sysfs attributes. > Ok I understand your point. I will provide a description in Documentation/ABI/testing, something I should have done before and take the opportunity to provide a bigger update of the thermal netlink messages with more commands. Thanks -- Daniel -- <http://www.linaro.org/> Linaro.org │ Open source software for ARM SoCs Follow Linaro: <http://www.facebook.com/pages/Linaro> Facebook | <http://twitter.com/#!/linaroorg> Twitter | <http://www.linaro.org/linaro-blog/> Blog
