Re: [PATCH v2 3/4] rcu: use simple wait queues where possible in rcutree

[Daniel Wagner <wagi@monom.org>]

Hi Paul,

On 10/20/2015 01:31 AM, Paul E. McKenney wrote:
> On Wed, Oct 14, 2015 at 09:43:21AM +0200, Daniel Wagner wrote:
>> From: Paul Gortmaker <paul.gortmaker@windriver.com>
>> @@ -4178,7 +4178,8 @@ static void __init rcu_init_one(struct rcu_state *rsp,
>>  		}
>>  	}
>>
>> -	init_waitqueue_head(&rsp->gp_wq);
>> +	rsp->rda = rda;
> 
> This is initialized at compile time in current mainline, so the above
> statement is not needed.

I accidentally forward ported this from 3.10.

> But now that you mention it, the second parameter to rcu_init_one() is
> a bit pointless these days.  I have queued a patch eliminating it,
> which should not conflict with our patch.

Glad I can help :)

I am about to send an updated version of this series containing which
swaps this patch with #4 to avoid lockdep warning while doing bissect.
Obviously, the above statement will be gone as well.

cheers,
daniel