Messages in this thread | | | Date | Tue, 3 Sep 2013 21:56:25 -0400 | From | Steven Rostedt <> | Subject | Re: [RFC][PATCH 01/18 v2] ftrace: Add hash list to save RCU unsafe functions |
| |
On Tue, 3 Sep 2013 21:51:52 -0400 Steven Rostedt <rostedt@goodmis.org> wrote:
> > > void ftrace_unsafe_rcu_checker_disable(void) > > > { > > > atomic_inc(&ftrace_unsafe_rcu_disabled); > > > + /* Make sure the update is seen immediately */ > > > + smp_wmb(); > > > > smp_mb__after_atomic_inc()? > > Yep. > > > > > > } > > > > > > void ftrace_unsafe_rcu_checker_enable(void) > > > { > > > atomic_dec(&ftrace_unsafe_rcu_disabled); > > > + /* Make sure the update is seen immediately */ > > > + smp_wmb(); > > > > smp_mb__after_atomic_dec()? > > Yep. > > > > > > } > > > > > > static DEFINE_PER_CPU(unsigned long, ftrace_rcu_func); > > > @@ -588,15 +593,14 @@ static void > > > ftrace_unsafe_callback(unsigned long ip, unsigned long parent_ip, > > > struct ftrace_ops *op, struct pt_regs *pt_regs) > > > { > > > - int bit; > > > - > > > + /* Make sure we see disabled or not first */ > > > + smp_rmb(); > > > > smp_mb__before_atomic_inc()? > > Yep. > > I'll update it and restart my tests. >
Although, this will punish archs that need the mb(), as IIUC, smp_mb() is heavier weight than smp_wmb() or smp_rmb().
-- Steve
| |