Hi, On 11/18/22 11:48, Andy Shevchenko wrote: > On Fri, Nov 18, 2022 at 10:25:50AM +0100, Hans de Goede wrote: >> The driver has DMI-quirk tables for force-enabling 5 button array support >> and for 2 different ways of enabling SW_TABLET_MODE reporting. >> >> Add module parameters to allow user to enable the driver behavior currently >> only available through DMI quirks. >> >> This is useful for users to test this in bug-reports and for users to use >> as a workaround while new DMI quirks find their way upstream. > > Lately you have been adding tons of module parameters here and there. I'm not sure I would call it "tons of" I've added a few parameters to allow users to test behavior which before then was only available through DMI quirk tables. This is useful for users to easily check if their model needs to be added to a DMI quirk table. > Taking into account that we discourage to do that, but at the same time > understanding your point, wouldn't be better before doing that, provide > a new type of the module parameters "for debug purposes only". One way > is to provide necessary macros > > MODULE_PARAM_DEBUG() > > And always have the parameters suffixed with _debug. OR introduce an additional > option that user may put before "open" debugging module / kernel command line > parameters? > > In this case we delimit the old/existing parameters with the parameters for > debugging. Moreover, it may be excluded by introducing Kconfig option at compile > time. Adding special infrastructure for just these 5-6 module parameters which I have added recently seems a bit too much to me. Regards, Hans p.s. I notice that I have forgotten to Cc the list when submitting that patch, I've added the list to the Cc now and I'll resend the patch with the list in the Cc.
