Re: rcu_init allocation failure trace on bootup

[Dave Jones <davej@redhat.com>]

On Mon, May 06, 2013 at 02:02:49PM -0700, Paul E. McKenney wrote:

 > >  > You are quite right, I am getting ahead of myself.  Please see below
 > >  > for the patch.
 > > 
 > > Doesn't fix it for me.
 > > 
 > > Is it this one instead perhaps ?
 > > 
 > >   89 #ifndef CONFIG_RCU_NOCB_CPU_NONE
 > >   90         if (!have_rcu_nocb_mask) {
 > >   91                 alloc_bootmem_cpumask_var(&rcu_nocb_mask);
 > >   92                 have_rcu_nocb_mask = true;
 > >   93         }
 > 
 > Hmmm...  That was in fact the one that the patch was intended to change.
 > Just out of curiosity, where did the patch apply instead?

I think this might be my screwup. I did it by hand, and misapplied it
to the top of rcu_nocb_setup() instead.

	Dave