Hi Andy, On Mon, Dec 28, 2020 at 11:37:26PM +0200, Andy Shevchenko wrote: > On Fri, Oct 30, 2020 at 01:06:06PM +0200, Andy Shevchenko wrote: > > On Fri, Oct 30, 2020 at 08:33:56AM +0000, Will Deacon wrote: > > > On Wed, Oct 28, 2020 at 07:28:26PM +0200, Andy Shevchenko wrote: > > > > There is no evidence we need kernel.h inclusion in certain headers. > > > > > > Did you run some tools to determine this or is it just a hunch? > > > > I read files manually. So, closer to the latter. > > Also I rely on 0day and other CIs that didn't show anything for a long time. > > Will, any recommendation what should I run to be solid about this change?
Sorry, I don't have any good suggestions here. That's actually why I asked, in case you happened to have something to make these sort of changes less error-prone! So I'm not trying to block this change, it's just a shame that it's not easier to do. Will
