Messages in this thread | | | Subject | Re: [PATCH 1/5] user_hooks: New user hooks subsystem | From | Steven Rostedt <> | Date | Mon, 30 Jul 2012 11:27:54 -0400 |
| |
On Mon, 2012-07-30 at 17:08 +0200, Peter Zijlstra wrote: > On Fri, 2012-07-27 at 17:40 +0200, Frederic Weisbecker wrote: > > +++ b/kernel/user_hooks.c > > @@ -0,0 +1,56 @@ > > +#include <linux/user_hooks.h> > > +#include <linux/rcupdate.h> > > +#include <linux/sched.h> > > +#include <linux/percpu.h> > > + > > +struct user_hooks { > > + bool hooking; > > + bool in_user; > > +}; > > I really detest using bool in structures.. but that's just me. Also this > really wants a comment as to wtf 'hooking' means. in_user I can just > about guess.
I'm curious to what you have against bool in structures? Would you prefer a:
struct user_hooks { unsigned int hooking:1; unsigned int in_user:1; };
instead? I haven't checked, but I would hope that gcc would optimize the struct into a single word.
But I could see that it can cause races as that would make modifying hooking and in_user dependent on each other. That is, if one CPU updates hooking as another CPU updates in_user, that could cause a read-modify-write race. At least in this case the modification is only done on local cpu variables.
-- Steve
| |