On Thu, May 25, 2023 at 11:52:58AM -0700, Dave Hansen wrote: > On 5/24/23 00:30, Luis Chamberlain wrote: > >> It doesn't actually have anything to do with moving the > >> show_unhandled_signals sysctl, right? > > Well in my case it is making sure the sysctl variable used is declared > > as well. > > But what does this have to do with _this_ patch? This: Because to create consistency for the users. > > --- a/arch/x86/kernel/umip.c > > +++ b/arch/x86/kernel/umip.c > > @@ -12,6 +12,7 @@ > > #include <asm/insn.h> > > #include <asm/insn-eval.h> > > #include <linux/ratelimit.h> > > +#include <linux/signal.h> > > For instance. You don't move things to another header or make *ANY* > change to the compilation of umip.c. So why patch it? > > It looks to me like a _fundamentally_ superfluous change. That hunk > literally *can't* be related to the rest of the patch. I suspect it is not needed as otherwise compilation would have failed. So I'll just drop it. > >> If that's the case, it would be nice to have this in its own patch. > > If its not really fixing any build bugs or functional bugs I don't see > > the need. But if you really want it, I can do it. > > > > Let me know! > > Yes, I really want it. > > Please remove all the x86 bits from _this_ patch. If x86 has a > separate, preexisting problem, please send that patch separately with a > separate changelog and justification. > > We'll take a look. Sounds good. Luis
