On Wed, Jun 5, 2019 at 5:32 PM Eduardo Valentin <eduval@xxxxxxxxxx> wrote: > On Wed, Jun 05, 2019 at 11:25:39AM +0300, Andy Shevchenko wrote: > > On Wed, Jun 5, 2019 at 6:30 AM Eduardo Valentin <eduval@xxxxxxxxxx> wrote: > Well, yes, but the point is you would be switching from a simple AND (&) operation > to a division... > > I am keeping the power of 2 dep so that we can keep this with a simple &. Works for me. > > > > > + .of_match_table = of_match_ptr(i2c_slave_mqueue_of_match), > > > > > > > > Wouldn't compiler warn you due to unused data? > > > > Perhaps drop of_match_ptr() for good... > > > > > > Not sure what you meant here. I dont see any compiler warning. > > > Also, of_match_ptr seams to be well spread in the kernel. > > > > If this will be compiled with CONFIG_OF=n... > > I see.. I obviously did not test with that config.. > > > Though I didn't check all dependencies to see if it even possible. In > > any case of_match_ptr() is redundant in both cases here. > > Either you need to protect i2c_slave_mqueue_of_match with #ifdef > > CONFIG_OF, or drop the macro use. > > I will wrap it into CONFIG_OF.. Would be this expected to work in the case of CONFIG_OF=n? If no, why to introduce ugly #ifdef:s and additional macros? Wouldn't be better to have depends on OF || COMPILE_TEST ? -- With Best Regards, Andy Shevchenko
