[PATCH V5 1/9] drivers core: Add support for Wifi band RF mitigations

Andrew Lunn andrew at lunn.ch
Sat Jul 1 00:25:17 UTC 2023


> Right now there are stubs for non CONFIG_WBRF as well as other patches are
> using #ifdef CONFIG_WBRF or having their own stubs.  Like mac80211 patch
> looks for #ifdef CONFIG_WBRF.
> 
> I think we should pick one or the other.
> 
> Having other subsystems #ifdef CONFIG_WBRF will make the series easier to
> land through multiple trees; so I have a slight leaning in that direction.

#ifdef in C files is generally not liked because it makes build
testing harder. There are more permutations to build. It is better to use

if (IS_ENABLED(CONFIG_WBTR)) {
}

so that the code is compiled, and them throw away because
IS_ENABLED(CONFIG_WBTR) evaluates to false.

However, if the stubs are done correctly, the driver should not
care. I doubt this is used in any sort of hot path where every
instruction counts.

	Andrew


More information about the dri-devel mailing list